Inhibition of purified and membrane-bound flavin-containing monooxygenase 1 by (N,N-dimethylamino)stilbene carboxylates

Article abstract of DOI:10.1021/tx950145x

(E)-[2-(4-(Dimethylamino)phenyl)vinyl]benzenes bearing a nitrile or carboxyl group in the 2', 3', or 4' position were synthesized and tested for substrate activity with purified pig liver flavin-containing monooxygenase (FMO1). Although the nitrile derivatives were too insoluble to saturate the catalytic site at pH 7.4, they appeared to be substrates with K(m)'s somewhat above their maximum solubility (?0.1 mM) in the assay medium. Of the three carboxylic acid analogs, (E)-4-[2-(4-(dimethylamino)phenyl)vinyl]benzoic acid had no detectable water solubility at pH 7.4, and measurements were restricted to (E)-3-[2-(4-(dimethylamino)phenyl)vinyl]benzoic acid (DS3CO) and (E)-2-[2-(4-(dimethylamino)phenyl)vinyl]benzoic acid (DS2CO). While DS3CO and DS2CO were substrates, they also inhibited FMO1 turnover. DS3CO was the more effective inhibitor, and at 2 mM it inhibited FMO1 and microsomal- catalyzed oxidation of methimazole (N-methyl-2-mercaptoimidazole) by 80-90%. Kinetic studies indicated that the aminostilbene carboxylates were noncompetitive with both the xenobiotic substrate, methimazole, and NADPH. However, inhibition constants calculated from double reciprocal plots of velocity vs NADPH were K(i)(comp) 130 and 150 μM for DS3CO and DS2CO, respectively, whereas the uncompetitive K(i)'s were 10-15 times higher, which suggests that inhibition of NADPH binding may be primarily responsible for inhibition of FMO1 by the aminostilbene carboxylates. This model is also consistent with inhibition of cyclohexanone monooxygenase, a bacterial analog of FMO. DS3CO and DS2CO were again noncompetitive with methimazole but primarily competitive with NADPH. The aminostilbene carboxylates had no detectable effects on activity of pig or rat liver NADPH-cytochrome P450 reductase, which suggests that they are not nonspecific flavoprotein antagonists.

Full text of DOI:10.1021/tx950145x

Chem. Res. Toxicol. 1996, 9, 599-604
599
In h ibition of P u r ified a n d Mem br a n e-Bou n d
F la vin -Con ta in in g Mon ooxygen a se 1 by
(N,N-Dim eth yla m in o)stilben e Ca r boxyla tes†
Bernd Clement,*,‡ Matthias Weide,‡ and Daniel M. Ziegler§
Pharmazeutisches Institut, Christian-Albrechts-Universita¨t Kiel, Gutenbergstrasse 76,
D-24118 Kiel, Germany, and Biochemical Institute, Department of Chemistry and Biochemistry,
The University of Texas at Austin, Austin, Texas 78712
Received August 21, 1995X
(E)-[2-(4-(Dimethylamino)phenyl)vinyl]benzenes bearing a nitrile or carboxyl group in the
2′, 3′, or 4′ position were synthesized and tested for substrate activity with purified pig liver
flavin-containing monooxygenase (FMO1). Although the nitrile derivatives were too insoluble
to saturate the catalytic site at pH 7.4, they appeared to be substrates with Km’s somewhat
above their maximum solubility (=0.1 mM) in the assay medium. Of the three carboxylic acid
analogs, (E)-4-[2-(4-(dimethylamino)phenyl)vinyl]benzoic acid had no detectable water solubility
at pH 7.4, and measurements were restricted to (E)-3-[2-(4-(dimethylamino)phenyl)vinyl]benzoic
acid (DS3CO) and (E)-2-[2-(4-(dimethylamino)phenyl)vinyl]benzoic acid (DS2CO). While
DS3CO and DS2CO were substrates, they also inhibited FMO1 turnover. DS3CO was the
more effective inhibitor, and at 2 mM it inhibited FMO1 and microsomal-catalyzed oxidation
of methimazole (N-methyl-2-mercaptoimidazole) by 80-90%. Kinetic studies indicated that
the aminostilbene carboxylates were noncompetitive with both the xenobiotic substrate,
methimazole, and NADPH. However, inhibition constants calculated from double reciprocal
plots of velocity vs NADPH were Ki(comp) 130 and 150 µM for DS3CO and DS2CO, respectively,
whereas the uncompetitive Ki’s were 10-15 times higher, which suggests that inhibition of
NADPH binding may be primarily responsible for inhibition of FMO1 by the aminostilbene
carboxylates. This model is also consistent with inhibition of cyclohexanone monooxygenase,
a bacterial analog of FMO. DS3CO and DS2CO were again noncompetitive with methimazole
but primarily competitive with NADPH. The aminostilbene carboxylates had no detectable
effects on activity of pig or rat liver NADPH-cytochrome P450 reductase, which suggests that
they are not nonspecific flavoprotein antagonists.
are substrates. Investigators of the latter two reports
concluded that position relative to the nucleophilic het-
eroatom rather than mere presence of a negative
chargesespecially on a rigid molecule like sulindac
sulfidesdetermines substrate activity.
In tr od u ction
Microsomal flavin-containing monooxygenases (FMO)1
catalyze the oxidation of structurally diverse xenobiotics
bearing electron rich centers, typically nitrogen, sulfur,
phosphorus, or selenium (1-3). While substrates ac-
cepted by these enzymes lack common structural fea-
tures, all are soft nucleophiles readily oxidized upon
contact with the enzyme-bound 4a-hydroperoxyflavin.
Kinetic and other studies defining the mechanism un-
derlying such lack of specificity have been reviewed (4).
The data also suggest that essential cellular nucleophiles
are effectively excluded from the catalytic site, but
structural elements that prevent their contact with the
enzyme-bound oxidant are still poorly defined. While
overall size (5) and physical dimensions (6) appear to
limit access to the active site, charge is also an important
factor (7). Most compounds bearing ionic groups are
excluded, but sulindac sulfide (8) and a few other sulfur
compounds (7) which exist as anions at physiological pH
This study was undertaken to test this interpretation
by measuring substrate activity of positional isomers of
(N,N-dimethylamino)stilbene carboxylic acids (Chart 1).
However, these compounds proved to be better inhibitors
than substrates that, unlike anionic detergent FMO
inhibitors (7), also inhibited microsomal-bound FMO at
pH 7.4. This report describes the synthesis and effects
of aminostilbenes bearing a carboxylate group on activi-
ties of purified and membrane-bound FMO1.
Exp er im en ta l P r oced u r es
Ma ter ia ls. The following reagents were obtained from the
sources indicated: NADP+, glucose 6-phosphate, acetylthiocho-
line, 5,5′-dithiobis(2-nitrobenzoic acid), methimazole (N-methyl-
2-mercaptoimidazole), L-mesenteriodes glucose-6-phosphate de-
hydrogenase, and catalase, Sigma Chemical Co. (St. Louis, MO).
4-(Dimethylamino)cinnamic acid and its nitrile were purchased
from Aldrich (Steinheim, Germany) and recrystallized from
ethanol. All reagents used for the organic syntheses were
synthetic grade purchased from either Firma Merck (Darmstadt,
Germany) or Aldrich (Steinheim, Germany).
† This paper is dedicated to Professor Dr. G. Seitz on the occasion
of his 60th birthday.
* Address correspondence to this author at the Pharmazeutisches
Institut, Christian-Albrechts-Universita¨t Kiel, Gutenbergstrasse 76,
D-24118 Kiel, Germany. Tel +431/880-1126, FAX +431/880-1352.
‡ Pharmazeutisches Institut, Christian-Albrechts-Universita¨t Kiel.
§ Biochemical Institute, The University of Texas at Austin.
X Abstract published in Advance ACS Abstracts, March 1, 1996.
1
Abbreviations: FMO, flavin-containing monooxygenase; CMO,
En zym e P r ep a r a tion s. FMO was isolated from pig liver
microsomes by minor modifications of the original procedure as
described in a recent report (9). Liver tissue from adult male
cyclohexanone monooxygenase; DMPU, N,N′-dimethyl-N,N′-propyl-
eneurea. Also see Chart 1 for abbreviations of the stilbene and cinnamic
acid derivatives used in this report.
0893-228x/96/2709-0599$12.00/0 © 1996 American Chemical Society
+
+

600 Chem. Res. Toxicol., Vol. 9, No. 3, 1996
Clement et al.
by Takahaschi et al. (12) for the base-catalyzed condensation of
4-(dimethylamino)benzaldehyde with methylbenzonitriles as
follows:
Ch a r t 1. Str u ctu r es of th e Stilben es (A) a n d
Rela ted Com p ou n d s (B) Used in Th is Stu d ya
DMPU (5 mL) containing 1.49 g (10 mmol) of 4-(dimethyl-
amino)benzaldehyde and 1.17 g (10 mmol) of the 2-, 3-, or
4-methylbenzonitrile was added dropwise to a stirred solution
of 2.47 g of t-BuOK (11 mmol) in 15 mL of DMPU at 10 °C (or
20 °C for DS3CN) under nitrogen. After stirring for 5 h, the
reaction mixture was poured into 100 mL of ice water containing
1.08 g (20 mmol) of ammonium chloride. The products were
extracted into ether or dichloromethane and crystallized. The
structure of each compound was determined by IR, 1H-NMR,
and 13C-NMR. The analytical data including mp and elemental
composition along with precise details for the extraction and
crystallization of each compound will be supplied upon request.
Only the E configurations were detected in the final crystalline
products of all positional isomers of the [2-(4-(dimethylamino)-
phenyl)vinyl]benzonitriles and the corresponding benzoic acids.
The vinyl protons of the 1H-NMR spectra are in the range of
(E)-stilbene (δ ) 7.1 ppm) and not of (Z)-stilbene (δ ) 6.55ppm)
a
3
Abbreviations used: (A) The 2, 3, and 4 positional isomers of
(13), and their coupling constants J (HR-Hâ) of about 16 Hz are
[2-(4-(dimethylamino)phenyl)vinyl]benzonitrile and [2-(4-(dimeth-
ylamino)phenyl)vinyl]benzoic acid, respectively. (B) 4-(Dimeth-
ylamino)cinnamic acid and its nitrile.
also only in agreement with the (E)-configuration (14).
(E)-4-[2-(4-(Dim eth yla m in o)p h en yl)vin yl]ben zoic Acid .
One millimole (0.248 g) of the nitrile, DS4CN, was heated to 90
°C in 6 mL of 50% sulfuric acid for 27 h. After cooling, 6 mL of
water was added and the pH adjusted to between 3 and 4 with
concentrated ammonia. The precipitate was collected, washed
with water, and then sublimed at 180 °C under vacuum (<10-2
mbar). The yield of yellow powdery product was 56% of
theoretical; mp 330 °C dec; IR (KBr): 3440, 2856, 1678, 1592,
1522, 962 cm-1; 1H-NMR (360 MHz, [D6]DMSO) δ (ppm) ) 2.95
(s, 6H, -N(CH3)2), 7.1 (mc, AA′BB′, 4H, ArH), 7.76 (mc, AA′BB′,
4H, ArH), 7.17 (mc, AB, 3J ) 16.4 Hz, 2H, olef H), 12.81 (br,
1H, -COOH). Calcd for C17H17NO2: C, 76.38; H, 6.41; N, 5.24.
Found: C, 76.40; H, 6.32; N, 5.24.
Sprague-Dawley rats was obtained locally (Austin). Microsomes
isolated by differential centrifugation from rat liver homoge-
nates prepared in 0.25 M sucrose containing 0.1 mM butylated
hydroxytoluene were resuspended in 10 volumes of 0.25 M
sucrose and resedimented. The pellet from each liver was
resuspended in 0.25 M sucrose and stored in small aliquots at
-70 °C. The preparations were thawed and stored on ice no
more than 6 h before use. Any of the sample not used in this
time was discarded. Cyclohexanone monooxygenase was iso-
lated from Acinetobacter by the procedure described by Dono-
ghue et al. (10).
(E)-3-[2-(4-(Dim eth yla m in o)p h en yl)vin yl]ben zoic Acid .
This compound was also synthesized by hydrolysis of the
corresponding nitrile and purified by sublimation. Yield ) 44%
of theoretical; mp 220-222 °C dec; IR (KBr): 3440, 2850, 2590,
Activity Mea su r em en ts. Activities of purified FMO1 were
measured polarographically by following substrate-dependent
oxygen uptake at 37 °C in 0.1 M potassium phosphate (pH 7.4)
containing 0.25 mM NADP+, 2.5 mM glucose 6-phosphate, 1
unit/mL glucose-6-phosphate dehydrogenase, and 1000 units/
mL catalase. After 3-4 min temperature equilibration, 0.2-
0.3 nmol of FMO1 in 10-30 µL was added through the capillary
stem of the 2 mL sealed reaction vessel, and the endogenous
rate of oxygen uptake was recorded for 1-1.5 min before adding
the xenobiotic substrate. The change in oxygen concentration
was recorded for 5-10 min. Methimazole was then routinely
added as a preliminary test for possible effects on FMO1 activity
of the xenobiotics added initially. Activities reported were
calculated from initial rates of oxygen uptake.
1682, 1630, 1606, 1578, 1522, 960 cm-1 1H-NMR (400 MHz,
;
[D6]DMSO) δ (ppm) ) 2.95 (s, 6H, -N(CH3)2), 6.72 (d, 2H, ArH),
7.13 (mc, AB, 3J ) 16.4 Hz, 2H, olef H), 7.45 (mc, 3H, ArH),
7.77 (mc, 1H, ArH), 8.07 (s, 2H, ArH), 13 (br, 1H, -COOH).
Calcd for C17H17NO2: C, 76.38; H, 6.41; N, 5.24. Found: C,
76.37; H, 6.33; N 5.22.
(E)-2-[2-(4-(Dim eth yla m in o)p h en yl)vin yl]ben zoic Acid .
This compound could not be prepared by hydrolysis of the nitrile
(DS2CN), and it was synthesized as follows: Five milliliters of
DMPU containing 1.49 g (10 mmol) of 4-(dimethylamino)-
benzaldehyde and 1.64 g (10 mmol) of ethyl 2-methylbenzoate
was added dropwise at 10 °C to a stirred solution of 22 mmol
(2.47 g) of t-BuOK in 15 mL of DMPU under nitrogen. After
stirring for 8 h under nitrogen, the reaction mixture was poured
into 100 mL of 0.2 mM ice-cold ammonium chloride. After
extracting three times with 100 mL of diethyl ether, the aqueous
phase was adjusted to pH 5-6 on ice. The product precipitated
as a yellow solid, which upon analysis proved to be the free acid.
The ethyl ester apparently hydrolyzed during extraction. The
precipitate was collected, washed once with cold water, pH 5-6,
dried, and recrystallized from acetone-water. Yield ) 36%
theoretical; mp 136 °C dec; IR (KBr): 3426, 2892, 2804, 1678,
The oxidation of methimazole catalyzed by microsomes and
the purified flavoenzymes was measured by following methi-
mazole-dependent thiocholine oxidation as described by Guo et
al. (6) in reaction medium identical to that used for the oxygen
uptake measurements except that it also contained 0.3 mM
thiocholine. The complete reaction medium minus the substrate
and enzyme preparation was preincubated in a metabolic shaker
at 37 °C for 3-4 min. Microsomes (or FMO1) were added, and
1 min later, the reaction was started by adding 1.0 mM
methimazole. Aliquots removed at 0, 3, 6, 9, and 12 min were
deproteinized with trichloracetic acid, and the concentration of
thiocholine was measured as described (6). The rate of H2O2
formation was measured by following the peroxidation of
methanol by the procedure described earlier (11).
1
1606, 1592, 1524, 968 cm-1; H-NMR (360 MHz, [D6]DMSO) δ
(ppm) ) 2.94 (s, 6H, -N(CH3)2), 7.06 (mc, AA′BB′, 4H, ArH),
7.3 (t, 1H, ArH), 7.38 (mc, AB, 3J ) 16.3 Hz, 2H, olef H), 7.51
(mc, 1H, ArH), 7.8 (mc, 2H, ArH), 12.96 (br, 1H, -COOH).
Calcd for C17H17NO2: C, 76.38; H, 6.41; N, 5.24. Found: C,
76.32; H, 6.36; N, 5.20.
Syn th esis. The melting points were determined with a
Bu¨chi 510 apparatus and are uncorrected. The IR spectra were
recorded with a Perkin-Elmer Fourier FTIR 16 PC spectrometer.
The 1H-NMR spectra were recorded with a Bruker AM 360
(360.15 MHz) and Bruker AM 400 (400.13 MHz) spectrometer.
The elemental analysis of compounds described in this section
were performed by the Ilse Betz Laboratory, Kronach, Germany.
The (E)-[2-(4-(dimethylamino)phenyl)vinyl]benzonitriles were
synthesized by minor modifications of the procedure described
Resu lts
(E )-[2-(4-(Dim e t h yla m in o)p h e n yl)vin yl]b e n zo-
n itr iles. All of the nitriles (Chart 1) stimulated FMO1
and NADPH-dependent oxygen uptake, which is good
+
+

Inhibition of FMO
Chem. Res. Toxicol., Vol. 9, No. 3, 1996 601
Ta ble 1. DS2CO a n d DS3CO In h ibition of
F MO1-Ca ta lyzed Oxid a tion of Am in e Dr u gsa
% inhibition by 1 mM
substrate
concn (mM)
DS2CO
DS3CO
tamoxifen
0.20
0.50
0.50
0.50
51 ( 10
47 ( 6
42 ( 2
50 ( 5
71 ( 6
70 ( 8
58 ( 6
54 ( 12
trifluoperazine
imipramine
methimazoleb
a
Activities were measured by following substrate-dependent O2
uptake at 37 °C in 0.1 M phosphate, pH 7.4, containing the
NADPH generating system and catalase at the concentrations
listed under Experimental Procedures, except that NADP+ was
reduced to 0.1 mM. After 4 min temperature equilibration, FMO1
was added, followed 1 min later with substrate. O2 uptake was
recorded for 3-5 min. The assay was then repeated in the
presence of the aminostilbene carboxylates. The values are
average ( SD % decrease in O2 reduction in the presence of the
F igu r e 1. FMO1- and NADPH-dependent O2 uptake as a
function of DS2CO (9) and DS3CO (() concentration. Oxygen
uptake was measured polarographically in a 2 mL cell at 37 °C
containing 0.1 M phosphate, pH 7.4, 0.25 mM NADP+, 2.5 mM
glucose 6-phosphate, 1.0 unit/mL glucose-6-phosphate dehydro-
genase, 1000 units/mL catalase, and 5 mg/mL bovine serum
albumin. After 4-5 min temperature equilibration, FMO (0.9
nmol in 20 µL) was added through the capillary stem, and 1.5-
2.0 min later the stilbenes were added in no more than 10 µL
of DMPU or 20 µL of acetonitrile. The rates listed are averages
( SD calculated from initial rates of substrate-dependent oxygen
uptake for 3 or more preparations of FMO1. The arrow indicates
activity with 0.15 mM methimazole under the same conditions.
b
inhibitors for no less than 3 determinations. The S-oxygenation
of methimazole, the most widely used marker activity of FMO1,
is included for comparison.
F igu r e 3. Effects of DS2CO and DS3CO on the methimazole
S-oxidase activity of rat and pig liver microsomes. Activities
were calculated from methimazole-dependent thiocholine oxida-
tion at pH 7.4 in media identical to that listed in the legend to
Figure 1 except that it also contained 0.3 mM thiocholine. The
reactions were carried out in 10 mL Erlenmeyer flasks mounted
in a shaker bath at 37 °C. After 4-5 min temperature equilibra-
tion, microsomes (2-3 mg/mL) were added, and 1 min later the
reaction was started by adding methimazole (1.5 mM). The
methimazole-dependent thiocholine oxidation was calculated
from loss of thiocholine in aliquots removed at 0, 3, 6, and 9
min by the procedure described in ref 6.
F igu r e 2. Effect of increasing concentrations of DS2CO and
DS3CO on methimazole-dependent O2 uptake catalyzed by
FMO1. The reactions were carried out as described in the legend
to Figure 1 in 0.1 M phosphate, pH 7.4, containing 0.25 mM
NADP+, 2.5 mM glucose 6-phosphate, 1.0 unit/mL glucose-6-
phosphate dehydrogenase, and 1000 units/mL catalase. After
3-4 min temperature equilibration, the stilbenes were added
at the concentrations indicated, followed 2 min later by 1.5 mM
methimazole. Oxygen uptake was recorded for another 2 min.
The percent change in rates from methimazole-dependent rate
in the absence of the stilbenes is the average ( SD for 3-4
determinations.
near the limits of detection spectrophotometrically, which
prevented substrate activity measurements with this
compound. Although considerably more soluble above 1
mM, DS2CN and DS3CN would often slowly come out of
solution. This could be largely prevented by adding
bovine serum albumin (5 mg/mL) to the assay medium.
Serum albumin had little, if any, effects on activity at
low concentrations of the aminostilbene carboxylates, and
where specified, the kinetic studies were carried out in
the presence of this protein.
Both DS2CO and DS3CO stimulated initial rates of
FMO1 and NADPH-dependent oxygen uptake, but nei-
ther affected the endogenous rate of H2O2 generation
(data not shown), which suggests that the increase in
oxygen uptake was not due to uncoupling of oxygen
reduction from substrate oxygenation. However, reaction
rates as a function of DS2CO or DS3CO concentration
did not follow typical saturation kinetics, and the rates
decreased with increasing concentrations above 1.0 mM
(Figure 1). Both DS2CO and DS3CO also inhibited the
oxidation of methimazole (Figure 2) at concentrations
presumptive evidence for substrate activity. However,
the concentrations of the (N,N-dimethylamino)stilbene-
nitriles required to half-saturate the enzyme were greater
than their solubility in the assay medium, which es-
sentially precluded collecting accurate kinetic constants.
From the limited values at low concentrations, the Km’s
for DS2CN, DS3CN, and DS4CN were near or somewhat
above their upper limits of solubility (=0.1 mM). How-
ever, CCN (Chart 1) was an excellent substrate with a
Vmax equal to that of N,N-dimethylaniline (data not
shown).
(E)-[2-(4-(Dim et h yla m in o)p h en yl)vin yl]b en zoic
Acid . The solubility of DS4CO in the assay medium was
+
+

602 Chem. Res. Toxicol., Vol. 9, No. 3, 1996
Clement et al.
Ta ble 2. DS2CO a n d DS3CO In h ibition Con sta n ts for th e
Oxid a tion of Meth im a zole Ca ta lyzed by F MO1 a n d CMOa
Ki’s (mM)
FMO1
CMO
varied
substrate
inhibitor
DS2CO
compb uncomp comp uncomp
methimazole
NADPH
1.1
0.15
0.65
2.2
0.63
0.070
0.54
1.91
DS3CO
methimazole
NADPH
0.30
0.13
1.9
1.2
0.20
0.062
0.17
0.69
a
The inhibition constants were calculated from the slope and
intercept changes (Figures 5 and 6) with the following equations:
K(comp) ) [I]V*maxKm/VmaxK*m - V*mKm; and K(comp) ) [I]V*max
Vmax - V*max; where [I] ) concentration of inhibitor; Vmax, Km and
V*max, K*m are the x and y intercepts in the absence and (*)
presence of [I], respectively. The values are averages of 2-3
determinations with different preparations of FMO1 or CMO and
concentrations of [I] that varied from 0.05 to 0.5 mM. None of
the values included in the averages varied more than (2.5 times
/
b
the Ki’s listed. Abbreviations: comp ) competitive; uncomp )
uncompetitive.
chemical synthesis of N-oxides to serve as chromato-
graphic standards could not be developed. Because of
the pronounced inhibition, the amount of product that
could be generated with FMO1 was too limited for
isolation and identification without an authentic stan-
dard. Although the products were not identified, spectra
(Figure 4A,B) indicate that FMO1 catalyzed the oxidation
of both DS2CO and DS3CO to compounds that are
spectroscopically different from the parent compounds.
Kinetic studies on the mechanism of DS2CO inhibition
of methimazole oxidation gave a noncompetitive (mixed)
pattern (Figure 5) with variable methimazole. However,
at saturating methimazole (1.5 mM) and variable NAD-
PH, the pattern was largely competitive (Figure 5B), as
indicated by the magnitude of inhibition constants (Table
2) calculated from slope and intercept changes of double
reciprocal plots of rates as a function of NADPH concen-
tration (Figure 5B).
DS2CO and DS3CO also inhibit CMO, the bacterial
analog of FMO (15), and as with the mammalian enzyme,
DS3CO was the more effective inhibitor. The inhibition
of CMO by the aminostilbene carboxylic acids gave a
classic noncompetitive pattern with methimazole as the
variable substrate, with Ki’s of 0.20 mM for DS3CO and
0.63 mM for DS2CO (Figure 6A, Table 2). However, with
NADPH as the variable substrate, the pattern was again
largely competitive (Figure 6B, Table 2), which suggests
that, as with FMO1, the aminostilbene carboxylates
compete with NADPH for the same catalytic intermediate
of CMO. They are not, however, nonspecific flavoprotein
antagonists. At 2 mM, neither DS2CO nor DS3CO had
any detectable effects on activity of rat or pig liver
NADPH-cytochrome P450 reductase measured by fol-
lowing reduction of cytochrome c (16).
F igu r e 4. Changes in the spectra of DS2CO (A) and DS3CO
(B) catalyzed by FMO1. The reactions were carried out in a
thermostated, stirred cell mounted in a Hewlett-Packard diode
array spectrophotometer, Model 8452A. The composition of the
assay medium was identical to that listed in Figure 1 except
FMO1 was decreased to about 0.1 µM, glucose 6-phosphate to
1.25 mM, and NADP+ to 50 µM. After setting the base line to
zero absorbance, 49 µM DS2CO (panel A) or DS3CO (panel B)
was added and spectra were recorded every 10 s for 10 min.
Only spectra taken at 1 min intervals are shown. In the absence
of either NADPH or FMO, there was no detectable change in
the spectra, but with both present the longer wavelength
absorbance decreased with a concomitant increase in absorbance
at shorter wavelengths with isosbestic points at the wavelengths
indicated.
above 0.2-0.3 mM. DS3CO was a more effective inhibi-
tor, and at 2 mM it decreased FMO-dependent O2 uptake
almost 80% with methimazole and with all other sub-
strates tested (Table 1). However, below 0.2 mM, both
DS2CO and DS3CO stimulated the oxidation of methi-
mazole 5-10%. Although small, the stimulation was
reproducible and observed consistently with all prepara-
tions of FMO1.
DS2CO and DS3CO also inhibited the oxidation of
methimazole catalyzed by the microsomal-bound FMO
(Figure 3). The pattern of inhibition was similar to that
observed with purified FMO1. DS3CO was again the
more effective inhibitor with both rat and pig liver
microsomes, but methimazole oxidase activity of rat liver
was less sensitive to DS3CO than that of pig liver
microsomes. While inhibition of FMO turnover by a
metabolite cannot be completely excluded, preincubation
for several minutes before adding methimazole did not
increase inhibition. Furthermore, the rates shown were
calculated from initial rates when the concentration of
product (or products) should be negligible.
Discu ssion
The pronounced inhibition of FMO1 by the (dimeth-
ylamino)stilbene carboxylic acids was unexpected. Ear-
lier studies with sulindac sulfide (8) and analogs of lipoic
acid (7) suggested that an anionic group on a rigid side
chain far enough removed from the nucleophilic hetero-
atom would not influence interactions of the substrate
with the enzyme. While it appears that increasing
distance between the amine and carboxylate groups does
facilitate interactions with the catalytic site (DS3CO vs
DS2CO, Figure 1), this also increases inhibitor activity
The (dimethylamino)stilbene carboxylates were pre-
sumably oxidized to N-oxides, but a procedure for the
+
+

Inhibition of FMO
Chem. Res. Toxicol., Vol. 9, No. 3, 1996 603
F igu r e 5. Lineweaver-Burk plots for FMO1-catalyzed oxidation of methimazole ( DS2CO or DS3CO with (A) methimazole and
(B) NADPH as the varied substrate. The reaction rates for the plots shown were calculated from methimazole-dependent thiocholine
oxidation at pH 7.4 minus (9) and plus 0.33 mM DS2CO (b) or 0.25 mM DS3CO (() with 0.25 mM NADPH and variable methimazole
(A) or 1.5 mM methimazole and variable NADPH with 0.5 mM DS2CO (b) or DS3CO (() (B) measured as described in the legend
to Figure 3 by following methimazole-dependent thiocholine oxidation in aliquots withdrawn at 0, 3, 6, and 9 min. The concentration
of FMO1 was 0.6 µM.
F igu r e 6. Lineweaver-Burk plots for CMO-catalyzed oxidation of methimazole ( DS2CO or DS3CO with (A) methimazole and (B)
NADPH as the varied substrate. Reaction rates were calculated from initial rate of methimazole-dependent oxygen uptake as described
in media described in the legend to Figure 2 minus (9) and plus (b) 0.5 mM DS2CO or (() 0.5 mM DS3CO. Because of the thermal
instability of CMO, the reactions were carried out at 15 °C with variable methimazole and at 20 °C with variable NADPH. Neither
DS2CO nor DS3CO had detectable substrate activity with CMO.
at concentrations above 1 mM. The distance between
the nucleophilic center and the anionic group is not,
however, the only factor. N,N-Dimethylcinnamic acid
(Chart 1, CCO) had no detectable substrate or inhibitor
activities even though molecular models suggest that the
distance between the amino and carboxylate groups is
not that different from DS2CO. This suggests that the
second planar ring may be essential for both substrate
and inhibitor activity of the aminostilbene carboxylic
acids.
may bind at different sites and conformational changes
induced by these stilbene derivatives prevent NADPH
binding.
This interpretation is also supported by the kinetic
studies carried out with CMOsa bacterial analog of
FMO. While CMO catalyzes NADPH- and O2-dependent
oxidation of xenobiotic sulfur compounds (15), it does
not accept amine or other FMO substrates bearing
nitrogen. Because DS3CO and DS2CO are not sub-
strates for CMO, they do not compete with methimazole
for the enzyme-bound hydroperoxyflavin. However, the
inhibition is at least partially competitive (Table 2),
which suggests that DS2CO and DS3CO bind with a form
of the enzyme that also interacts with methimazole. The
classic noncompetitive pattern (Figure 6A) of double
reciprocal plots of velocity as a function of methimazole
concentrations ( DS2CO or DS3CO suggests that the
inhibitors effectively remove a fraction of the enzyme
from the catalytic cycle, apparently by blocking NADPH
binding to the oxidized enzyme (Figure 6B), but have less
effect on the other intermediate forms of the enzyme. The
kinetic data strongly indicate that inhibition of NADPH
Kinetic studies with methimazole as the variable
substrate (Figure 5A) indicate that DS3CO and DS2CO
interact with two different catalytic intermediates of the
enzyme. The competitive Ki is probably due to the
interaction of these stilbene derivatives with the enzyme-
bound 4a-hydroperoxyflavin. While this form of the
enzyme has a bound NADP+, the pronounced competitive
inhibition with NADPH strongly suggests that DS3CO
and DS2CO also bind (or interact) with the oxidized
flavoprotein which is the only enzyme intermediate that
binds NADPH (4). While NADPH and the inhibitors both
interact with the oxidized form of the flavoprotein, they
+
+

604 Chem. Res. Toxicol., Vol. 9, No. 3, 1996
Clement et al.
(5) Nagata, T., Williams, D. E., and Ziegler, D. M. (1990) Substrate
specificities of liver and lung flavin-containing monooxygenase:
Differences due to substrate size. Chem. Res. Toxicol. 3, 372-
376.
(6) Guo, W.-X. A., Poulsen, L. L., and Ziegler, D. M. (1992) Use of
thiocarbamides as selective substrate probes for isoforms of flavin-
containing monooxygenases. Biochem. Pharmacol. 44, 2029-2037.
(7) Taylor, K. L., and Ziegler, D. M. (1987) Studies on substrate
specificity of the hog liver flavin-containing monooxygenase:
Anionic organic sulfur compounds. Biochem. Pharmacol. 36, 141-
146.
(8) Light, D. R., Waxman, D. J ., and Walsh, C. (1982) Studies on the
chirality of sulfoxidation catalyzed by flavoenzyme cyclohexane
monooxygenase and hog liver flavin adenine dinucleotide contain-
ing monooxygenase. Biochemistry 21, 2490-2498.
(9) Chen, G.-P., Poulsen, L. L., and Ziegler, D. M. (1995) Oxidation
of aldehydes catalyzed by pig liver flavin-containing monooxyge-
nase (FMO1). Drug Metab. Dispos., in press.
(10) Donoghue, N. A., Norris, D. B., and Trudgill, P. W. (1976) The
purification and properties of cyclohexanone oxygenase from
Nocardia globerula CLI and Acinetobacter NCIB 9871. Eur. J .
Biochem. 63, 175-192.
binding is primarily responsible for the inhibition of CMO
by the aminostilbene carboxylic acids. Whether only
flavoenzymes similar in mechanism to FMO will be
inhibited by the aminostilbene carboxylates remains to
be determined. However, the aminostilbene carboxylates
have no effect on activity of NADPH-cytochrome P450
reductase, which indicates that they are not nonspecific
flavoenzyme antagonists.
DS2CO and DS3CO are the first lipophilic anionic
compounds tested that inhibit the membrane-bound
flavoenzyme as effectively as purified FMO1 (Figures 2
and 3). Perhaps interference with NADPH binding
(which must bind to an exposed site on the purified and
microsomal-bound enzyme) accounts for this phenom-
enon.
Ack n ow led gm en t. The continued support of the
Foundation for Research and the Fonds der Chemischen
Industrie is deeply appreciated. In addition, the support
of one of us (D.M.Z.) by the Alexander von Humboldt
Foundation and the expert technical assistance of Timo-
thy Ward, along with the advice of Dr. L. L. Poulsen on
the interpretation of the kinetic data, are gratefully
acknowledged.
(11) Ziegler, D. M., and Kehrer, J . P. (1990) Oxygen radicals and
drugs: in vitro measurements. Methods Enzymol. 186, 621-626.
(12) Takahaschi, K., Okamoto, T., Yamada, K., and Iida, H. (1977) A
convenient synthesis of substituted stilbenes by condensation of
o- or p-tolunitrile with p-substituted benzaldehydes. Synthesis
1977, 58-59.
(13) Friebolin, H. (1988) Ein- und zweidimensionale NMR-Spektrosko-
pie (One- and Two-Dimensional NMR Spectroscopy), pp 122-123,
VCH Verlagsgesellschaft mbH, Weinheim, Basel, Cambridge, and
New York.
Refer en ces
(14) Friebolin, H. (1988) Ein- und zweidimensionale NMR-Spektrosko-
pie (One- and Two-Dimensional NMR Spectroscopy), p 78, VCH
Verlagsgesellschaft mbH, Weinheim, Basel, Cambridge, and New
York.
(1) Ziegler, D. M. (1993) Recent studies on the structure and function
of multisubstrate flavin-containing monooxygenases. Annu. Rev.
Pharmacol. Toxicol. 33, 179-199.
(2) Smyser, B. P., and Hodgson, E. (1985) Metabolism of phosphorous-
containing compounds by pig liver microsomal FAD-containing
monooxygenase. Biochem. Pharmacol. 34, 1145-1150.
(3) Chen, G.-P., and Ziegler, D. M. (1994) Liver microsome and flavin-
containing monooxygenase catalyzed oxidation of organic sele-
nium compounds. Arch. Biochem. Biophys. 312, 566-572.
(4) Poulsen, L. L., and Ziegler, D. M. (1995) Multisubstrate flavin-
containing monooxygenases: Applications of mechanism to speci-
ficity. Chem.-Biol. Interact. 96, 57-73.
(15) Ryerson, C. C., Ballou, D. P., and Walsh, C. (1982) Mechanistic
studies on cyclohexanone oxygenase. Biochemistry 21, 2644-2655.
(16) Masters, B. B. S. (1980) The role of NADPH-cytochrome c (P-
450) reductase in detoxication. In Enzymic basis of detoxication
(J acoby, W. B., Ed.) Vol. l, pp 183-200, Academic Press, New
York.
TX950145X
+
+