Influence of some novel N-substituted azoles and pyridines on rat hepatic CYP3A activity

Article abstract of DOI:10.1016/S0006-2952(98)00096-3

A series of N-substituted heteroaromatic compounds structurally related to clotrimazole was synthesized, and the effects of these compounds on ethosuximide clearance in rats were determined as a measure of their abilities to induce cytochrome P4503A (CYP3A) activity. Ethosuximide clearance and in vitro erythromycin N-demethylase activity were shown to correlate. In this series, imidazole or other related heteroaromatic 'head groups' were linked to triphenylmethane or other phenylmethane derivatives. Within the series, it was found that 1-triphenylmethane-substituted imidazoles elicited the greatest increase in CYP3A activity, and that among the triphenylmethyl-substituted imidazoles, the highest activities were achieved by the substitution of F- or Cl- in either the meta or para position of one of the phenyl rings. Diphenylmethylsubstituted pyridine was effectively devoid of activity. Compounds eliciting the largest increase in CYP3A activity (viz. 1-[(3-fluorophenyl)diphenylmethyl]imidazole, 1-[(4- fluorophenyl)diphenylmethyl]imidazole, and 1-[tri-(4- fluorophenyl)methyl]imidazole) produced little or no increase in ethoxyresorufin O-dealkylase (EROD) activity (i.e. CYP1A), whereas benzylimidazole, which elicited only a small increase in CYP3A activity, produced an almost 9-fold increase in CYP1A activity. For a series of eleven compounds exhibiting a wide range of influence on CYP3A activity, a positive correlation was found between ethosuximide clearance and hepatic CYP3A mRNA levels.

Full text of DOI:10.1016/S0006-2952(98)00096-3

Biochemical Pharmacology, Vol. 55, pp. 1881–1892, 1998.
© 1998 Elsevier Science Inc. All rights reserved.
ISSN 0006-2952/98/$19.00 ϩ 0.00
PII S0006-2952(98)00096-3
Influence of Some Novel N-Substituted Azoles and
Pyridines on Rat Hepatic CYP3A Activity
James T. Slama,* Julie L. Hancock,† Taikyun Rho,*§ Lidia Sambucetti‡
࿣
and Kenneth A. Bachmann†
*DEPARTMENT OF MEDICINAL AND BIOLOGICAL CHEMISTRY AND †DEPARTMENT OF PHARMACOLOGY, COLLEGE OF
PHARMACY, THE UNIVERSITY OF TOLEDO, TOLEDO, OH 43606; AND ‡NOVARTIS PHARMACEUTICALS
CORPORATION, EAST HANOVER, NJ 07936, U.S.A.
ABSTRACT. A series of N-substituted heteroaromatic compounds structurally related to clotrimazole was
synthesized, and the effects of these compounds on ethosuximide clearance in rats were determined as a measure
of their abilities to induce cytochrome P4503A (CYP3A) activity. Ethosuximide clearance and in vitro
erythromycin N-demethylase activity were shown to correlate. In this series, imidazole or other related
heteroaromatic “head groups” were linked to triphenylmethane or other phenylmethane derivatives. Within the
series, it was found that 1-triphenylmethane-substituted imidazoles elicited the greatest increase in CYP3A
activity, and that among the triphenylmethyl-substituted imidazoles, the highest activities were achieved by the
substitution of F- or Cl- in either the meta or para position of one of the phenyl rings. Diphenylmethyl-
substituted pyridine was effectively devoid of activity. Compounds eliciting the largest increase in CYP3A
activity (viz. 1-[(3-fluorophenyl)diphenylmethyl]imidazole, 1-[(4-fluorophenyl)diphenylmethyl]imidazole, and
1-[tri-(4-fluorophenyl)methyl]imidazole) produced little or no increase in ethoxyresorufin O-dealkylase (EROD)
activity (i.e. CYP1A), whereas benzylimidazole, which elicited only a small increase in CYP3A activity,
produced an almost 9-fold increase in CYP1A activity. For a series of eleven compounds exhibiting a wide range
of influence on CYP3A activity, a positive correlation was found between ethosuximide clearance and hepatic
CYP3A mRNA levels. BIOCHEM PHARMACOL 55;11:1881–1892, 1998. © 1998 Elsevier Science Inc.
KEY WORDS. cytochrome P4503A; CYP3A; enzyme induction; ethosuximide; triphenylmethylimadazole;
structure–activity relationships
Drugs and other xenobiotics are oxidatively metabolized by
the family of CYP¶. It is now appreciated that drug
interactions often arise from specific inhibition or induc-
tion of CYP isozymes that alters the rate of metabolism of
other drug molecules. Because many of the commonly
prescribed drugs are metabolized by the CYP3A isoform,
understanding the structure–activity relationship for both
the inhibition and the induction of this cytochrome is key
to the prediction of many drug interactions. Many N-
substituted azoles have the ability to inhibit and/or induce
hepatic microsomal mixed-function oxidases [1–5]. Some
N-substituted imidazoles have been characterized as high
magnitude inducers of rat hepatic CYP such as N-benzy-
limidazole [6], which elicited a three-fold increase in CYP
content, a two-fold increase in erythromycin-O-demethyl-
ase activity, a two-fold increase in ethylmorphine O-
demethylase activity, a 56-fold increase in ethoxyresorufin
deethylase activity, and a 16-fold increase in pen-
toxyresorufin O-dealkylase activity. A series of N-substi-
tuted imidazoles and related pyridines were compared for
their ability to induce rat hepatic CYP after single doses [7].
Matsuura et al. [7] found that the imidazole structure, itself,
was important for CYP induction; that triazole could be
substituted for imidazole; and that the chlorine atom in
clotrimazole was not important for CYP induction. How-
ever, these authors and others [8] have commented on the
ability of imidazoles and other agents to induce multiple
isoforms of CYP. Hostetler et al. [9] have demonstrated that
clotrimazole is a relatively selective inducer of CYP3A in
rats. The induction of CYP3A was several orders of mag-
nitude greater than the induction of CYP2B.
In view of the marked induction of CYP3A by clotri-
mazole [9–11], and because the SARs for imidazoles regard-
ing CYP induction have thus far been based largely on the
measurement of hepatic CYP content [7], we synthesized
other selective inducers of CYP3A, using clotrimazole as a
lead compound, and characterized the effects of this series
with specific regard to the increase of CYP3A activity after
§ Current address: Department of Chemistry, Alteon Inc., 170 Williams
Drive, Ramsey, NJ 07446.
࿣
Corresponding author: Kenneth Bachmann, Ph.D., FCP, Department
of Pharmacology, The University of Toledo College of Pharmacy, 2801 W.
Bancroft St., Toledo, OH 43606. Tel. (419) 530-1912; FAX (419)
530-1909.
¶ Abbreviations: CYP, cytochrome P450; EF-1␣, elongation factor-1
alpha; EROD, ethoxyresorufin O-dealkylase; SAR, structure–activity re-
lationship; THF, tetrahydrofuran.
Received 9 May 1997; accepted 26 January 1998.

1882
J. T. Slama et al.
magnesium halides were added to benzophenone or to one
of a variety of commercially available substituted benzophe-
nones producing the mono-, di-, and trisubstituted tri-
phenylmethanols required for the study (Methods C and D).
Melting points were determined in open capillary tubes
using an Electrothermal model IA9100 apparatus and are
uncorrected. 1H NMR spectra were determined at 300 MHz
in CDCl3 unless otherwise noted. Chemical shifts are
reported in parts per million downfield from internal
tetramethylsilane. Microanalyses were performed by Atlan-
tic Microlabs Inc. Analyses indicated by the symbol of the
elements were within Ϯ0.4% of theoretical values. THF
was dried by distillation over sodium and benzophenone.
SCHEME 1. Synthesis of 1-triphenylmethyl-1H-imidazoles.
subchronic (8-day) treatment. This approach is empirical of
necessity, since the mechanism of CYP3A induction is as
yet unknown, and the biochemical nature of the macromo-
lecular receptor for CYP3A inducers has not been described
yet [12].
General Procedure for the Synthesis of
1-Triphenylmethyl-Substituted Azoles (Method A)
The heterocycle (15 mmol), a triphenylchloromethane (15
mmol), and triethylamine (15 mmol) were heated at reflux
in 100 mL of acetonitrile for 2–7 hr. Solvent was removed
in vacuo, 100 mL of water was added, and the product was
extracted into chloroform (two portions of 50 mL each).
Extracts were washed with water and dried (MgSO4), the
solvent was evaporated, and the product was purified by
crystallization, resulting in isolated yields ranging between
50 and 80% of theory.
MATERIALS AND METHODS
Materials
1-Benzylimidazole (CDD 3500), 4-benzylpyridine (CDD
3513), 3-benzylpyridine (CDD 3515), and 2-benzylpyridine
(CDD 3529) were purchased from the Aldrich Chemical
Co. Clotrimazole (1-[(2-chlorophenyl)diphenylmethy-
l]imidazole) was purchased from the Sigma Chemical Co.
Purities ranged from 97% to greater than 99%. 1-Tri-
phenylmethylimidazole (CDD 3501) was synthesized using
method A and crystallized from ethanol: m.p. 222–224°
(reported [13], 221–223°). 1-Triphenylmethylpyrazole
(CDD 3502), 1-triphenylmethyl-1,2,4-triazole (CDD
3504), and 1-diphenylmethylimidazole (CDD 3510) were
prepared by the procedure of Claramunt et al. [14]. 1-Tri-
phenylmethylpyrrole (CDD 3503) was prepared from tri-
phenylmethylamine using the procedure of Chadwick and
Hodgson [15]. Compounds CDD 3508, 3522, 3523, 3524,
3530, 3532, 3535, 3536, 3537, 3538, and 3540 were
synthesized as described by Buechel et al. [16]. The physical
General Procedure for the Conversion of
Triphenylmethanols to Triphenylchloromethanes
(Method B)
The tertiary alcohol (25 mmol), thionyl chloride (7 g, 60
mmol), and dry THF (100 mL) were heated at reflux for 1
hr. Solvent was evaporated in vacuo, 30 mL of fresh THF
was added to dissolve the residue, and the solvent was
evaporated again. The resulting triphenylchloromethanes
were somewhat unstable and were neither purified nor
characterized, but used immediately in a subsequent reac-
tion.
1
properties (m.p. and H NMR) determined for these com-
pounds were in agreement with that reported for the
compound in the literature cited. The purity of tested
compounds was established by obtaining combustion anal-
ysis (C, H, and N). In all cases, analysis obtained agreed
with calculated compositions to within Ϯ0.4%.
General Procedure for Conversion of a Substituted
Benzophenone to an (Aryl)diphenylmethanol
(Method C)
A 500-mL three-neck flask was fitted with a condenser, a
nitrogen inlet, and a magnetic stirrer and charged with the
benzophenone (35.71 mmol) and 120 mL of dry THF. The
mixture was stirred under an atmosphere of dry N2 as a 3.0
M solution of phenyl magnesium bromide in THF (12.2
mL, 36.6 mmol, Aldrich) was added dropwise from a
pressure-equalized dropping funnel. The mixture was
heated at reflux for 1 hr and, after the solution had cooled
to room temperature, solvent was removed in vacuo. The
residue was treated with 1% H2SO4 (100 mL), stirred for 30
min, and extracted twice with 100-mL portions of chloro-
form. The extract was washed with 100 mL of water, dried
over anhydrous MgSO4, and filtered. The solvent was
General Methods (Chemistry)
1-Triphenylmethyl-1H-imidazoles (see Scheme 1) were
prepared by treating a triphenylchloromethane or other
alkyl halide with an imidazole or a related heterocycle
(Method A). The triphenylchloromethanes were synthe-
sized by treating triphenylmethanols with thionyl chloride
(Method B). A variety of substituted triphenylmethanols
were available commercially, or were readily prepared from
commercially available benzophenones using the Grignard
reaction. Phenylmagnesium halides or substituted phenyl-

CYP3A Induction
1883
TABLE 1. Synthesis of substituted azoles*†
Compound
No.
Method of
synthesis
Crystallization
solvent
Compound name
m.p. (°)
3505
3506
3511
3512
3526
3527
3531
3533
3534
3539
3541
3542
3543
4-Nitro-1-triphenylmethylimidazole‡
4-Amino-1-triphenylmethylimidazole
1-(Triphenylmethyl)-2-methylimidazole
1-Triphenylmethylbenzimidazole
1-[␣-(4-Pyridyl)-␣-phenylbenzyl]imidazole
1-[(5-Pyrimidyl)diphenylmethyl]imidazole
1-Methyl-1-(triphenylmethyl)imidazole
1-[(3-Methoxyphenyl)diphenylmethyl]imidazole
1-(9-Phenyl-9-fluorenyl)imidazole
1-[(2-Naphthyl)diphenylmethyl]imidazole
1-[(1-Naphthyl)diphenylmethyl]imidazole
1(11-Phenyl-11-dibenzosuberenyl)imidazole
1-[Di(4-fluorophenyl)phenylmethyl]imidazole
A
Acetonitrile
Methanol
Acetonitrile
Cyclohexane
Hexane
210–211
235–236
215–216
180–181
211–212
213–214
220–221
115–117
197–198
155–157
171
Hydrogenation§
A
A
B and A
࿣
C , B, and A
Acetonitrile
Acetonitrile
Hexane
A
D, B, and A
B and A
D, B, and A
D, B, and A
D
Hexane
Hexane
Hexane
Hexane
222–224
135–138
D, B, and A
Hexane
*Combustion analysis for C, H, and N was performed for all new target compounds. In all cases, analyses agree with calculated compositions to within Ϯ0.4%.
†Each compound was characterized by 1H NMR, and the spectrum that was obtained supported the structure assigned.
‡Structure was confirmed by x-ray crystallographic analysis [17].
§CDD 3506 was hydrogenated using 10% palladium on carbon in ethanol solution for 2 hr under 20 psi H2.
࿣
In this case, a lithium reagent was formed from n-butyllithium and 5-bromopyrimidine at Ϫ95°, followed after 30 min by the addition of benzophenone.
evaporated, and the oily residue was crystallized. The
structure and purity of the alcohols were verified by H
NMR and TLC.
over anhydrous MgSO4. After evaporating solvent in vacuo,
a yellow oil was obtained and purified by Kugelrohr distil-
lation at 100° (5 mm Hg) to produce 2.5 g (48.5%) of a
1
1
colorless liquid: H NMR d 7.15–7.23 (4H, m), 6.98–7.02
(3H, m), 6.78 (1H, s), 4.09 (2H, t, J ϭ 7.1 Hz), 2.97 (2H,
t, J ϭ 7.1 Hz). Anal. (C11H12N2) C, H, N.
Alternate General Procedure for Conversion of a
Substituted Benzophenone to an (Aryl)diphenylmethanol
(Method D)
A 500-mL three-neck flask fitted with a reflux condenser, a
magnetic stirrer, and a nitrogen inlet was charged with
magnesium turnings (1.4 g, 58 mmol), the desired aryl
halide (59 mmol), and 100 mL of dry THF. The mixture
was maintained under an atmosphere of N2 and heated at
reflux until all magnesium disappeared (2 hr). A solution of
benzophenone or the desired substituted benzophenone (58
mmol) dissolved in 30 mL of dry THF was added dropwise
to the resulting gray mixture of Grignard reagent, and after
the addition was complete, was heated at reflux for 1 hr.
After cooling to room temperature, solvent was evaporated,
and 100 mL of 1% H2SO4 was added. The mixture was
stirred for 30 min and extracted twice with 100-mL por-
tions of chloroform. The organic extract was washed with
100 mL of water, dried over anhydrous MgSO4, and filtered.
Solvent was evaporated in vacuo producing a residue, which
was further purified. The structure and purity of the
1-(1-Phenylethyl)imidazole (CDD 3514)
1-Chloro-1-phenylethane (6.3 g, 44.8 mmol) and imidazole
(6.0 g, 88.1 mmol) were heated at 100° for 3 hr. After
cooling, 100 mL of chloroform was added, the organic
phase was washed twice with 100-mL portions of water, and
then dried (MgSO4). The solvent was removed and the
product was purified by column chromatography (silica
1
gel/chloroform), yielding 3.2 g (41%) of a colorless oil: H
NMR d 7.54 (1H, s), 7.32–7.26 (3H, m), 7.11–7.09 (2H,
m), 7.03 (1H, s), 6.89 (1H, s), 5.29 (1H, q, J ϭ 7.0 Hz),
1.79 (3H, d, J ϭ 7.0 Hz). Anal. (C11H12N2) C, H, N.
1-Triphenylmethyl-2-methyl-2-imidazoline (CDD 3520)
A mixture of 2-methyl-2-imidazoline (3.0 g, 35.66 mmol),
triphenylchloromethane (10.0 g, 35.87 mmol) and 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU, 5.55 g, 36.46 mmol)
in 30 mL of dimethylformamide was stirred overnight at
room temperature. The mixture was poured into 100 mL of
ice water, and the precipitate was collected. The solids were
dissolved in 100 mL of chloroform, dried (MgSO4), and
filtered. The product was purified by column chromatogra-
phy (silica gel/10% CH3OH in CHCl3) and crystallized
from chloroform/cyclohexane to yield 2.4 g (21%) of white
1
alcohols were verified by H NMR and TLC. Compounds
synthesized according to the General Procedures are listed
in Table 1.
1-(2-Phenylethyl)imidazole (CDD 3507)
Imidazole (4.0 g, 58.75 mmol) and 1-chloro-2-phe-
nylethane (4.2 g, 29.9 mmol, Aldrich) were heated at reflux
for 3 hr. Water (100 mL) was added, and the mixture
extracted twice with 100-mL portions of chloroform. The
organic phase was washed with water (100 mL) and dried
1
crystals; m.p. 189–190°. H NMR d 7.33–7.27 (15H, m),
3.61 (2H, t, J ϭ 9.6 Hz), 3.37 (2H, t, J ϭ 9.5 Hz), 1.31
(3H, s). Anal. (C23H22N) C, H, N.

1884
J. T. Slama et al.
(4-Pyridyl)diphenylmethane (CDD 3521)
Ritter and Franklin used a 3-day treatment period [18],
whereas we used an 8-day treatment period. We have
established that with a 24-hr interval following an 8-day
treatment with clotrimazole, one measures the identical
increase in CYP3A activity (ethosuximide clearance) com-
pared with a 3-day clotrimazole period followed by a 48-hr
interval. Compared with a mean control ethosuximide
clearance of 0.063 L/hr/kg (Ϯ0.01), various clotrimazole
treatment schedules produced the following ethosuximide
clearances: A 3-day treatment followed by a 24-hr interval
prior to the administration of ethosuximide elicited an
ethosuximide clearance of 0.073 L/hr/kg (Ϯ0.003). A 3-day
treatment followed by a 48-hr interval elicited an ethosux-
imide clearance of 0.171 L/hr/kg (Ϯ0.046). An 8-day
treatment followed by a 24-hr interval elicited an ethosux-
imide clearance of 0.171 L/hr/kg (Ϯ0.029). The latter two
values for ethosuximide clearance were significantly differ-
ent from the control value (P Ͻ 0.01).
Rats were anesthetized with CO2 for approximately 20
sec prior to ethosuximide injection. Whole blood was
collected by cardiac puncture under CO2 anesthesia exactly
8 hr after ethosuximide infusion. Animals were killed by
CO2 and cervical dislocation. Plasma was separated from
EDTA-anticoagulated whole blood at room temperature
using a bench-top centrifuge (3500 g ϫ 5 min). Plasma for
ethosuximide analysis was stored unpreserved at Ϫ20° for
up to 5 days. Ethosuximide concentrations in plasma were
analyzed by fluorescence polarization immunoassay using a
TDx analyzer (Abbott Laboratories) and commercially
available reagents. The assay exhibited a coefficient of
variation Ͻ5% and a detection limit of 0.5 mg/L. Ethosux-
imide clearance (CL) was estimated from plasma concen-
tration measurements in samples acquired from blood
drawn 8 hr after ethosuximide administration according to
the following equation:
A mixture of ␣-(4-pyridyl)benzhydrol (Aldrich) (3.0 g,
11.5 mmol) and sodium borohydride (4.32 g, 114.0 mmol)
was added in portions over 10 min to 100 mL of trifluoro-
acetic acid (TFA) at 0° under a dry nitrogen stream. After
stirring the mixture for 30 min at 0°, TFA was evaporated
in vacuo. Water (100 mL) was added, and the pH was
adjusted to 9 by adding KOH. The aqueous solution was
extracted twice with 100-mL portions of chloroform, and
the organic extracts were dried (MgSO4) and filtered. After
evaporating the solvent, the oily residue was crystallized
from toluene yielding 1.52 g (54%) of white crystals: m.p.
125–126°; 1H NMR d 8.52 (2H, d, J ϭ 4.2 Hz), 7.35–7.28
(6H, m), 7.12–7.05 (6H, m). Anal. (C18H15N) C, H, N.
Animals
Male Sprague–Dawley rats (Harlan Sprague–Dawley)
weighing between 140 and 270 g were used in groups of
3–6. All rats had free access to standard chow and water.
Animals were maintained in the vivarium at 24° on a 12-hr
light–dark cycle.
Treatments
The compounds were either administered by gavage or
mixed into the diet. All compounds given by gavage were
suspended in a 1% methylcellulose suspension to a final
concentration of 10% (w/v), and administered on a weight
basis (100 mg/kg) daily for 8 days. The sole exception was
CDD 3540, which was administered in a dose of 50
mg/kg/day. Compounds administered in the diet were
incorporated at 0.1% (w/w) into Purina rodent chow
supplemented with 0.25% cholic acid and 0.75% choles-
terol. For CDD 3540, the supplemented amount was 0.05%.
Based on the average daily consumption, these doses
approximated 100 mg/kg/day for all compounds except
CDD 3540, which was ingested in daily doses approximat-
ing 50 mg/kg/day. Supplemented diets were consumed for 8
days.
CL ϭ ͓ln͑D/V͒ Ϫ lnCt͔ ⅐ ͑V/t͒
where D is the dose of ethosuximide (35 mg/kg), t is the
sampling time (8 hr), Ct is the concentration of ethosux-
imide, and V is the volume of distribution of ethosuximide,
which is set at 0.73 L/kg. The strategy for estimating
ethosuximide clearance from a single plasma concentration
measurement has been described several times [10, 19, 20].
An 8-hr postinfusion sampling time was used instead of the
originally reported 24-hr [10] sampling time, since the total
clearance of ethosuximide is significantly slower subsequent
to urethane anesthesia versus CO2 anesthesia [21], i.e.
urethane slows ethosuximide clearance whereas CO2 does
not. Ethosuximide clearance is an index of CYP3A activity
in rats [10].
Ethosuximide Clearance
Ethosuximide (Sigma) was prepared as a 3.5% (w/v) solu-
tion dissolved in normal saline (0.9% sodium chloride
solution), and was administered in a single dose of 35 mg/kg
via tail vein injection 24 hr following the eighth compound
dose. Comment on this 24-hr interval is warranted here,
since Ritter and Franklin had established that a 48-hr
interval following the last dose of clotrimazole was required
in order to maximize its inductive effect, since residual
hepatic clotrimazole can function as an inhibitor of CYP
activity [18]. Thus, a 48-hr interval was recommended [18]
to preclude the possibility that the net effect of clotrimazole
on CYP activity might represent a mixture of both induc-
tive and inhibitory effects. It must be noted, however that
Enzyme Assays
Excised livers were rinsed in chilled 1.15% KCl, and then
prepared as 25% homogenates in 0.1 M of potassium
phosphate buffer (pH 7.4). A Brinkmann Polytron® ho-

CYP3A Induction
1885
FIG. 1. Correlation between induced increases in
hepatic erythromycin N-demethylase activity mea-
sured in vitro and the increase in ethosuximide
clearance (CL) measured in vivo. Each plotting sym-
bol depicts the effect of a different compound. After
an 8-day treatment, the in vivo clearance of ethosux-
imide was measured, the animal was killed, and
hepatic erythromycin N-demethylase activity was
measured.
mogenizer was used in the preparation of homogenates.
Homogenates were centrifuged at 10,000 g for 20 min at 4°
using a Beckman J2-21M/E centrifuge to remove nuclei and
mitochondria. The supernatant (3 mL) was centrifuged at
100,000 g for 60 min at 4° using a Beckman L-60 prepar-
ative ultracentrifuge. Microsomal pellets were resuspended
in 8 mL of 0.1 M of erythromycin dissolved in 0.1 M of
potassium phosphate buffer or were suspended in the buffer
with no erythromycin added. Protein was measured by the
method of Lowry et al. [22]. Erythromycin demethylase
activity was measured by the formation of formaldehyde
using the method of Nash [23].
EROD was measured spectrofluorometrically using a
Turner model 430 spectrofluorometer. A 300-␮L suspen-
sion of microsomes (1 mg of protein/mL in 0.1 M of
potassium phosphate buffer, pH 7.4) was diluted with 2.985
mL of phosphate buffer and 15 ␮L of ethoxyresorufin (1
mM) dissolved in DMSO. The mixture was placed in a
cuvette and allowed to equilibrate for 60 sec at 37°. The
reaction was started by the addition of 15 ␮L of 50 mM of
NADPH. The rate of resorufin formation, which was linear
for at least 10 min, was measured at excitation/emission
wavelengths of 530 and 585 nm, respectively, 4 min after
the start of the reaction.
water-saturated phenol. After addition of 2 mL of RNAzol
A, the tissue was homogenized immediately for ca. 30 sec
with a Polytron homogenizer, Kinematica AG (PT1200).
One hundred microliters of chloroform was added to 1 mL
of homogenate, transferred to a new tube, and shaken
vigorously for 15 sec. The homogenate was incubated on ice
for 15 min, and centrifuged at 15,000 g for 15 min at 4°.
The aqueous phase containing the RNA was transferred to
a new tube, mixed vigorously with 500 ␮L of phenol:
chloroform (5:1), and centrifuged at 15,000 g for 5 min at
4°. The RNA was precipitated by mixing the aqueous phase
with an equal volume of isopropyl alcohol, and incubating
at Ϫ70° for at least 15 min. The precipitate was collected
by centrifugation at 15,000 g for 20 min at 4°, and the
supernatant was removed. The resulting RNA pellet was
washed by addition of 0.5 mL of 70% ethanol followed by
centrifugation at 15,000 g for 5 min. The supernatant was
removed completely, and the RNA pellet was air dried and
resuspended in 200 ␮L of water treated with 0.1% dieth-
ylpyrocarbonate.
DNA oligonucleotides, antisense to the mRNAs of
CYP3A [25], and EF-1␣ [26] were used as probes for
Northern blot hybridization. Oligonucleotides were labeled
at the 5Ј end with [␥-33p]ATP (2000 Ci/mmol, Amersham
Life Science) in a reaction catalyzed by T4 polynucleotide
kinase [27].
RNA Materials and Methods
RNA was prepared from approximately 0.25 g of frozen
liver essentially by the method of Chomczynski and Sacchi
[24]. RNAzol A solution was prepared by mixing equal
volumes of solution D (4 M of guanidinium thiocyanate, 25
mM of sodium citrate, pH 7, 0.5% Sarcosyl, 0.1 M of
2-mercaptoethanol, 0.02 M of sodium acetate, pH 4) and
Statistics
Multiple comparison t-tests with a Bonferroni adjustment
were used to assess the significance of mean differences for
treatments versus controls and treatments versus clotri-
mazole. Differences were considered significant at P Ͻ

1886
J. T. Slama et al.
FIG. 2. Induced increase in in vivo
ethosuximide clearance (CL) as a
function of increased hepatic
CYP3A mRNA. Each plotting sym-
bol depicts the mean (؎SD) effect of
a different compound, each of which
is shown in the inset table (N ؍ 6).
The in vivo clearance of ethosuxi-
mide was measured after an 8-day
treatment with the specified com-
pound. Animals were killed, and he-
patic CYP3A mRNA levels were
measured. Data are plotted logarith-
mically for ease of visualization. with
details given in the tabular form in-
set. These data were analyzed statis-
tically in Table 2.
0.05. Sequences and coordinates of probes were as follows:
CYP3A (ϩ201 to ϩ157) 5Ј-CCCAGGAATCCCCTGTT
TCTTGAAAAGTCCATGTGTGCGGGTCCC-3Ј, EF-1␣
(ϩ420 to ϩ361) 5Ј-GATACCAGCTTCAAATTCACCA
ACACCAGCAGCAACAATCAGGACAGCACAGTC
AGCCTG-3Ј.
Northern blot hybridization assays were performed as fol-
lows: RNA samples were prepared for electrophoresis as
described [28]. For compound analysis, 10 ␮g of hepatic RNA
was analyzed per animal. RNA samples were fractionated
through 1.5% agarose gels containing 1ϫ MOPS buffer [0.62
M of formaldehyde, 20 mM of MOPS (3-[N-morpholino]pro-
panesulfonic acid), 5 mM of sodium acetate, 1 mM of EDTA,
pH 7.0]. After electrophoresis, the RNA was transferred to a
charged nylon membrane (Amersham Life Science, Hybond-
Nϩ) by capillary action resulting from an ascending flow of
20ϫ SSC buffer (3.0 M of NaCl, 0.3 M of sodium citrate, pH
7.0), and fixed by UV cross-linking at 120 mJ/cm2 in a
SpectroLinker (XL-1000 UV Crosslinker, Spectronics Corp.)
RNA blots were hybridized with 10 million cpm of the
radiolabeled probe per 100 cm2 membrane in a rate-
enhanced hybridization buffer (Rapid-hyb buffer, Amer-
sham) at 65° for ca. 18 hr. The blots were washed by
incubating in 100 mL of 0.5% SSC for 5 min at 65°,
followed by three 30-min washes at 60°. Following exposure
to the blots, storage phosphor screens were scanned, and
bands were quantitated by a Molecular Dynamics Phospho-
rImager. For reprobing, the CYP3A probe was removed by
heating the membrane to 90° in 0.1% SDS and cooling to
room temperature. Hybridization with the EF-1␣ probe was
performed as described above. Bands of interest were
quantitated using ImageQuant software provided with the
Molecular Dynamics PhosphorImager. A numerical value
was obtained for each band representing the amount of
radioactivity associated with CYP3A and EF1␣ mRNAs.
The amount of hybridization to CYP3A was divided by the
value for EF-1␣ in the same sample to obtain a ratio,
normalized for variations in sample preparation. EF-1␣ was
TABLE 2. Two-by-two contingency table comparing magnitude of treatment-induced increases in ethosuximide clearance and
CYP3A mRNA levels
Number of compounds
increasing ethosuximide
clearance > ؋2
Number of compounds
increasing ethosuximide
clearance < ؋2
Number of compounds increasing CYP3A mRNA Ն ϫ3
Number of compounds increasing CYP3A mRNA Ͻ ϫ3
7
0
0
4
2
␹
ϭ 11.0 (df ϭ 1), P Ͻ 0.001.

CYP3A Induction
1887
FIG. 3. Basic structural requirements for induction of ethosuximide clearance. Rats were treated with test compounds for 8 days, and
ethosuximide was injected via the tail vein 24 hr after the last treatment. Clearance was estimated as described under Materials and
Methods. The highest magnitude induction for the compounds shown was observed for triphenylmethylimidazole (CDD 3501). Each
value is the mean ؎ SD of 4–6 rats, except for CDD 3510. Due to chronic toxicity of this compound, only a single variate is depicted.
For those compounds that raised ethosuximide clearance above control values, P values for multiple t-test comparisons were calculated.
Appearing at the top of a bar, the symbol § denotes P < 0.01 as compared with control. P values for multiple t-test comparisons were
also calculated for the fold-increase in ethosuximide clearance compared to that induced by clotrimazole (CTZ) (see Fig. 5). An asterisk
(*) denotes P < 0.01 compared with CTZ.
chosen as the internal control, because its expression in
adult mammalian tissues is constitutive [29, 30].
clearance in vivo, a positive correlation between the two
parameters was shown (see Fig. 1). In addition, compounds
that induced ethosuximide clearance also induced high
levels of CYP3A mRNA (Fig. 2). The compounds tested
appeared to stratify into two groups such that a three-fold or
more increase in the hepatic content of CYP3A mRNA
appeared to be required before a significant (i.e. ϳtwo-fold
or more) increase in ethosuximide clearance occurred
(Table 2). Compounds failing to cause a three-fold or
greater increase in CYP3A mRNA failed to cause a signif-
icant increase (i.e. ϳtwo-fold or more) in CYP3A activity
as measured by ethosuximide clearance. Taken together,
these results plus earlier in vivo [10] and in vitro findings [31]
support the contention that ethosuximide clearance is a
measure of CYP3A activity in rats.
RESULTS
Through the use of selective CYP inducers and inhibitors
we have shown previously that the in vivo rate of ethosux-
imide clearance serves as a surrogate measure of CYP3A
activity in rats [10]. In vitro experiments with microsomes
enriched in CYP3A1/2 (clotrimazole treatment), CYP2E1
(isoniazid treatment), or CYP1A (␤-naphthoflavone treat-
ment) and immunoinhibition experiments have established
that ethosuximide is oxidized almost entirely by CYP3A1/2
[31]. The data reported herein further support that conten-
tion. When eight different compounds were tested for their
ability to induce erythromycin demethylase activity mea-
sured in vitro in hepatic microsomes and by ethosuximide
The effects of a series of heteroaromatic compounds on
CYP3A activity are depicted in Figs. 3–6. We used as a

1888
J. T. Slama et al.
FIG. 4. Structural requirements of the heteroaromatic group on the induction of ethosuximide clearance in vivo. Each compound
contained the heteroaromatic “head group” shown attached to triphenylmethane (R ؍ ؊CPh3). Experimental conditions and methods
were identical to those depicted in the legend of Fig. 3. The highest magnitude induction for the compounds shown was observed for
triphenylmethylimidazole (CDD 3501). For those compounds that raised ethosuximide clearance above control values, P values for
multiple t-test comparisons were calculated. Appearing at the top of a bar, the symbol § denotes P < 0.01 as compared with control.
P values for multiple t-test comparisons were also calculated for the fold-increase in ethosuximide clearance compared with that induced
by clotrimazole (CTZ) (Fig. 5). An asterisk (*) denotes P < 0.01 compared with CTZ. Values are means ؎ SD, N ؍ 4–6.
starting point the ability of benzylimidazole to function as a
high magnitude inducer of cytochrome P450. 1-Benzylimi-
dazole (CDD 3500) was a relatively poor inducer of CYP3A
activity as measured by ethosuximide clearance (Fig. 3).
Methylation (CDD 3514) or increasing chain length (CDD
3507) did not increase CYP3A activity significantly. Ad-
dition of a second phenyl substituent (CDD 3510) led to
increased CYP3A activity relative to CDD 3500, but also
produced nonspecific hepatotoxicity (Fig. 3). Addition of a
third benzene ring (CDD 3501) produced significant addi-
tional activity, measured as a 3.5-fold increase in ethosux-
imide clearance (Fig. 3). 3-Benzylpyridine (CDD 3515)
failed to induce CYP3A activity, and 2-benzylpyridine
(CDD 3529) was only a weak inducer. However, 4-ben-
zylpyridine (CDD 3513) was significantly active relative to
benzylimidazole. Introduction of an additional aryl substitu-
ent to 4-benzylpyridine (see CDD 3521) caused a signifi-
cant loss of activity in contrast to the result of addition of
a second aryl substituent to 1-benzylimidazole (CDD 3500;
Fig. 3).
heteroaromatic groups on the ability of the compound to
induce CYP3A activity. Each compound contained triph-
enylmethane [R ϭ ϪC(Ph)3] which, as shown in Fig. 3, was
most effective when combined with imidazole (CDD
3501). Although a variety of five-membered heteroaro-
matic compounds and substituted imidazoles were evalu-
ated, modification of the 1-imidazole invariably led to a
significant loss of activity when compared with the induc-
tion of CYP3A activity elicited by CDD 3501 (Fig. 4).
However, compound 3511 containing the 2-methyl-substi-
tuted imidazole still possessed the ability to increase etho-
suximide clearance by about two-fold.
The effects of the introduction of a single substituent
into one of the phenyl rings of triphenylmethylimidazole
(CDD 3501) are depicted in Fig. 5. The substituents
considered were: ϪOCH3, ϪCH3, ϪF, and ϪCl. Each
halogenated compound exhibited marked activity. All or-
tho-substituted compounds showed comparable activity;
however, the inductive activity was considerably less than
the unsubstituted compound CDD 3501 (see Fig. 4).
Introduction of methoxyl or methyl substituents into the
We evaluated the effects of the substitution of various

CYP3A Induction
1889
FIG. 5. Effect of the introduction of a substituent into one of the phenyl rings of triphenylmethylimidazole (CDD 3501) on the ability
of the compound to increase ethosuximide clearance in vivo. Experimental conditions and methods were identical to those depicted in
the legend of Fig. 3. The highest magnitude induction was associated with the introduction of halogenated substituents in either the
meta or para positions of the phenyl ring. For those compounds that raised ethosuximide clearance above control values, P values for
multiple t-test comparisons were calculated. Appearing at the top of a bar, the symbol § denotes P < 0.01 as compared with control.
P values for multiple t-test comparisons were also calculated for the fold increase in ethosuximide clearance compared with that induced
by clotrimazole (CTZ). An asterisk (*) denotes P < 0.01 compared with CTZ; a double dagger (‡) denotes P < 0.05 compared with
CTZ. Values are means ؎ SD, N ؍ 4–6.
meta or para positions led to loss of activity, whereas
halogenation at meta and para positions produced highly
active compounds. One of these compounds (CDD 3537),
bearing the meta-fluoro substituent, exhibited apparently
higher activity than the unsubstituted CDD 3501 (Fig. 5).
In Fig. 6 we considered the introduction of further
modification of the triphenylmethyl substituent of CDD
3501. Introduction of a nitrogen into one of the aryl rings
(CDD 3526) resulted in a loss of activity. Introduction of
two nitrogens led to total loss of activity (CDD 3527).
Although mono-halogenation produced highly active com-
pounds (see Fig. 5), dihalogenation (CDD 3543) and
trihalogenation (CDD 3540) led to a partial loss of activity
compared with the monohalogenated compounds. Replace-
ment of a phenyl ring with the naphthyl ring (CDD 3539
and CDD 3541) led to marked losses of activity. Directly
connecting two of the phenyl rings (CDD 3534) yielded a
highly active compound comparable to CDD 3501. Con-
necting the aryl rings with an ethylene bridge (CDD 3542),
however, led to a partial loss of activity.
animals ranged from 0.24 to 0.52 nmol/mg/min. Table 3
shows the mean fold increase (relative to respective control
experiment) in EROD activity (i.e. treated value/control
value) for each of eight compounds.
Those compounds that elicited the largest increases in
CYP3A activity (viz. CDD 3537, 3538, and 3540) pro-
duced little or no increase in EROD activity. In contrast,
N-benzylimidazole (CDD 3500), which produced a rela-
tively small increase in CYP3A activity, elicited almost a
nine-fold increase in EROD activity.
DISCUSSION
We synthesized a series of compounds in which we co-
valently linked either imidazole or other heteroaromatic
“head groups” to triphenylmethane or other phenylmeth-
ane derivatives, and evaluated each compound in the series
for its ability to induce hepatic CYP3A activity in rats.
Ethosuximide clearance was used as a measure of CYP3A
activity, as it has been shown to be a CYP3A substrate [10],
its in vivo clearance correlates with the in vitro activity of
hepatic erythromycin N-demethylase, a standard marker
enzyme for CYP3A activity, and substances that are capable
of substantially increasing hepatic CYP3A mRNA content
also increase ethosuximide clearance. More recent in vitro
findings within our laboratory, including immuno-inhibi-
The issue of selectivity of induction was investigated for
eight of the compounds synthesized by measuring changes
in the in vitro activity of hepatic EROD subsequent to 8-day
treatments as described in Materials and Methods. Mean
values for EROD activity in vehicle-treated (i.e. control)

1890
J. T. Slama et al.
FIG. 6. Effects of modification of
the triphenylmethyl group on the
ability of compounds to induce
ethosuximide clearance in vivo.
Each compound contained an im-
idazole “head group” linked via
the N1 to the substituted phenyl-
methane shown on the figure.
Experimental conditions and
methods were identical to those
depicted in the legend of Fig. 3.
The highest magnitude induction
was observed for CDD 3534, the
compound in which two of the
phenyl rings are covalently
linked. For those compounds that
raised ethosuximide clearance
above control values, P values for
multiple t-test comparisons were
calculated. Appearing at the top
of a bar, the symbol § denotes
P < 0.01 as compared with con-
trol; ¶ denotes P < 0.05 com-
pared with control. P values for
multiple t-test comparisons were
also calculated for the fold-in-
crease in ethosuximide clearance
compared with that induced by
clotrimazole (CTZ) (Fig. 5). An
asterisk (*) denotes P < 0.01
compared with CTZ. Values are
means ؎ SD, N ؍ 4–6.
tory studies, confirm that only CYP3A plays a significant
and predominant role in the oxidation of ethosuximide,
and that ethosuximide is hydroxylated primarily on the
ethyl side chain [31]. Several other putative probes for
CYP3A activity have been used by others, including
erythromycin [32, 33], dapsone [33], and midazolam [34,
35]. Unfortunately, there may not be a single ideal probe for
CYP3A activity. For example, reliable correspondence in
the characterization of CYP3A activity by various probes is
less than satisfactory [36–38].
Within this series we found that triphenylmethane-
(trityl-) substituted imidazoles elicited the greatest increase
in CYP3A activity, and that the highest inductive effects
were achieved by the substitution of F- or Cl- in either the
meta or para position of one of the phenyl rings. Since we
evaluated the inducibility of a specific subfamily of the
cytochromes P450, our findings are somewhat at variance
with those of Matsuura et al. [7] whose SAR analysis was
to a large extent based on measures of total hepatic
cytochrome P450 content. For example, Matsuura et al.
reported inductive equivalency between N-benzylimida-
zole, diphenylmethylimidazole, and triphenylmethylimi-
dazole, whereas we found a decreasing order of induction of
CYP3A activity as follows: triphenylmethylimidazole was
at least three times as effective as N-benzylimidazole.
Diphenylimidazole was intermediate between the two;
however, in our hands it was also extremely toxic, and its
inductive ranking is effectively precluded, because only a
single animal survived the 8-day treatment. We found that
triphenylmethyl-substituted triazole was ineffective in in-
creasing CYP3A activity, whereas Matsuura et al. reported
that triphenylmethyltriazole with an ortho-substituted Cl-
on one of the phenyl rings (i.e. the triazole analogue of
clotrimazole) increased the hepatic content of cytochrome
P450 to the same extent as clotrimazole. On the other
hand, our findings pertaining to triphenylmethyl-substi-
tuted pyridine were similar to those of Matsuura et al., who
reported that this compound failed to increase hepatic
cytochrome P450 content. It must be pointed out that our
treatments extended for 8 days rather than for a single dose
[7], and that we used Sprague–Dawley rather than Wistar
rats [7]. Though there are quantitative differences between
TABLE 3. Fold increase in EROD activity*
Compound
Fold increase in EROD activity†
Clotrimazole
CDD 3500
CDD 3501
CDD 3513
CDD 3530
CDD 3532
CDD 3537
CDD 3538
CDD 3540
1.79 Ϯ 0.41
8.94 Ϯ 3.43
1.48 Ϯ 0.48
1.72 Ϯ 0.13
0.82 Ϯ 0.09
1.34 Ϯ 0.50
0.86 Ϯ 0.06
0.73 Ϯ 2.47
1.24 Ϯ 0.20
*EROD activity was measured as the rate of formation of resorufin (nmol/mg/min).
†Fold increase represents the ratio of activity in compound-treated rats to that in
control (vehicle-treated) rats. Mean values for EROD activity in vehicle-treated
animals ranged from 0.24 to 0.52 nmol/mg/min. Values are means Ϯ SD, N ϭ 3.

CYP3A Induction
1891
our findings and those of Franklin [11], there are some areas
of broad agreement. Franklin reported that 1-benzylimida-
zole produced both substantial increases in EROD and
CYP3A activity, whereas we found that 1-benzylimidazole
produced a substantial increase only in EROD activity.
However, we found the extent of the benzylimidazole-
induced increase (or fold increase) in EROD activity was
about seven times greater than for the benzylimidazole-
induced increase in CYP3A activity. Franklin reported that
the effect of 1-benzylimidazole on EROD activity was about
13 times greater than the effect on CYP3A activity. Thus,
the relative effects of 1-benzylimidazole on EROD activity
compared with CYP3A activity were broadly similar in
both studies. The reason for the actual difference in the
extent of induced EROD activity is likely a reflection of our
higher control EROD values compared with those of
Franklin. Differences in the benzylimidazole effect on
CYP3A activity may simply reflect the differences in the
measures of CYP3A activity. We used in vivo ethosuximide
clearance, whereas Franklin used in vitro erythromycin
N-demethylase. These substrates may be processed, in part,
by different CYP3A isoforms.
References
1. Rodrigues AD, Gibson GG, Ioannides C and Parke DV,
Interactions of imidazole antifungal agents with purified
cytochrome P-450 proteins. Biochem Pharmacol 36: 4277–
4281, 1987.
2. Maurice M, Pichard L, Daujat M, Fabre I, Joyeux H, Domer-
gue J and Maurel P, Effects of imidazole derivatives on
cytochromes P450 from human hepatocytes in primary cul-
ture. FASEB J 6: 752–758, 1992.
3. Sheets JJ, Mason JI, Wise CA and Estabrook RW, Inhibition
of rat liver microsomal cytochrome P-450 steroid hydroxylase
reactions by imidazole antimycotic agents. Biochem Pharmacol
35: 487–491, 1986.
4. Houston JB, Humphrey MJ, Matthew DE and Tarbit MH,
Comparison of two azole antifungal drugs, ketoconazole and
fluconazole, as modifiers of rat hepatic monooxygenase activ-
ity. Biochem Pharmacol 37: 401–408, 1988.
5. Harmsworth WL and Franklin MR, Induction of hepatic and
extrahepatic cytochrome P-450 and monooxygenase activities
by N-substituted imidazoles. Xenobiotica 20: 1053–1063,
1990.
6. Papac DI and Franklin MR, N-Benzylimidazole, a high
magnitude inducer of rat hepatic cytochrome P-450 exhibit-
ing both polycyclic aromatic hydrocarbon- and phenobarbi-
tal-type induction of phase I and phase II drug-metabolizing
enzymes. Drug Metab Dispos 16: 259–264, 1988.
7. Matsuura Y, Kotani E, Lio T, Fukuda T, Tobinaga S, Yoshida
T and Kuroiwa Y, Structure–activity relationships in the
induction of hepatic microsomal cytochrome P450 by clo-
trimazole and its structurally related compounds in rats.
Biochem Pharmacol 41: 1949–1956, 1991.
8. Ritter JK and Franklin MR, Clotrimazole induction of cyto-
chrome P-450: Dose-differentiated isozyme induction. Mol
Pharmacol 31: 135–139, 1987.
9. Hostetler K, Wrighton S, Molowa D, Thomas P, Levin W and
Guzelian P, Coinduction of multiple cytochrome P-450 pro-
teins and their mRNAs in rats treated with imidazole anti-
mycotic agents. Mol Pharmacol 35: 279–285, 1989.
10. Bachmann K, Chu CA and Greear V, In vivo evidence that
ethosuximide is a substrate for cytochrome P450IIIA. Phar-
macology 45: 121–128, 1992.
11. Franklin MR, Induction of rat liver drug-metabolizing en-
zymes by heterocycle-containing mono-, di-, tri-, and tetra-
arylmethanes. Biochem Pharmacol 46: 683–689, 1993.
12. Wright M and Paine A, Induction of the cytochrome P4503A
subfamily in rat liver correlates with the binding of inducers
to a microsomal protein. Biochem Biophys Res Commun 201:
973–979, 1994.
13. Davis DP, Kirk KL and Cohen A, New synthesis of 2-nitro-
imidazoles. J Heterocycl Chem 19: 253–256, 1982.
14. Claramunt RM, Elguero J and Garceran R, Synthesis by phase
transfer catalysis of N-benzyl, N-diphenylmethyl, and N-
triphenylmethyl azoles and benzazoles: Proton NMR and
chromatographic data as a tool for identification. Heterocycles
23: 2895–2906, 1985.
15. Chadwick DJ and Hodgson ST, The protecting-directing role
of the trityl group in synthesis of pyrrole derivatives: Efficient
preparations of 1N-pyrrole-3-carboxylic acid and 3-acyl-,
3-amino-, and 3-bromo-tritylpyrroles. J Chem Soc Perkin Trans
1: 93–102, 1983.
Clotrimazole and 1-tritylimidazole (CDD 3501) were the
only other imidazoles that both we and Franklin evaluated
on both EROD and CYP3A [11]. Our findings were
qualitatively the same in that both of those imidazoles
produced substantial increases in CYP3A activity and only
marginal increases in EROD activity.
The inductive effects of this series of substituted het-
eroaromatic compounds were evaluated in rats receiving
daily treatments by gavage. To that extent we are not in a
position to offer comment on the structural requirements
for binding to a hepatic receptor. No hepatocyte receptor
governing the induction of CYP3A activity has, as yet,
been convincingly demonstrated to exist, though there is
preliminary evidence that CYP3A induction may be regu-
lated by the interaction of inducers and a microsomal
protein [12], though the protein is probably not CYP3A1
[12]. Moreover, the structural features that we describe as
being important for maximizing the expression of CYP3A
activity in the rat may exert their influence at one or more
of several levels including absorption across the gastroin-
testinal epithelium, the rate of oxidative biotransformation
of the compound itself, as well as “receptor” binding.
Structural contributions to each of these processes can only
be deduced with additional in vitro assays.
In conclusion, we have evaluated the SAR characteris-
tics involved in the expression of induced CYP3A activity
in rats, and found that maximum induction is elicited by
tritylimidazoles with no substitutions on the imidazole
“head group” and with either Cl- or F- substituted at either
the meta or para position of one of the phenyl rings.
16. Buechel KH, Draber W, Regel E and Plempel M, Synthesen
und eigenschaften von clotrimazol und weiteren antimykotis-
chen 1-triphenylmethylimidazolen. Arzneimittleforschung 22:
1260–1272, 1972.
17. Skrzypczak-Jankun E and Kurmumbail RG, 1-Trityl-4-nitro-
imidazole. Acta Crystallogr C 52: 189–191, 1996.
18. Ritter JK and Franklin MR, Induction and inhibition of rat
We acknowledge the collaboration of Dr. David Weinstein and Dr.
Tom Hughes of Novartis Pharmaceutical Corp. This work was funded,
in part, by a grant-in-aid from the Novartis Pharmaceutical Corp.,
East Hanover, NJ.

1892
J. T. Slama et al.
hepatic drug metabolism by N-substituted imidazole drugs.
expression in mammalian brain, heart, and muscle of S1, a
member of the elongation factor-1␣ gene family. J Biol Chem
267: 24064–24068, 1992.
Drug Metab Dispos 15: 335–343, 1987.
19. Bachmann K, Jahn D, Yang C and Schwartz J, Ethosuximide
disposition kinetics in rats. Xenobiotica 18: 373–380, 1988.
20. Bachmann K, The use of single sample clearance estimates to
probe hepatic drug metabolism in rats. IV. A model for
possible application to phenotyping xenobiotic influences on
human drug metabolism. Xenobiotica 19: 1449–1459, 1989.
21. Loch J, Potter J and Bachmann K, The influence of anesthetic
agents on rat hepatic cytochromes P450 in vivo. Pharmacology
50: 146–153, 1995.
31. Sarver JG, Bachmann KA, Zhu D and Klis WA, Ethosuxi-
mide is primarily metabolized by CYP3A when incubated
with isolated rat liver microsomes. Drug Metab Dispos 26:
78–82, 1998.
32. Zhang X-J and Thomas P, Erythromycin as a specific substrate
for cytochrome P4503A isozymes and identification of a
high-affinity erythromycin N-demethylase in adult female
rats. Drug Metab Dispos 24: 23–27, 1996.
22. Lowry OH, Rosebrough NJ, Farr AL and Randall RJ, Protein
measurement with the Folin phenol reagent. J Biol Chem 193:
265–275, 1951.
23. Nash T, The colorimetric estimation of formaldehyde by
means of the Hantzsch reaction. Biochem J 55: 416–421,
1953.
24. Chomczynski P and Sacchi N, Single-step method of RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform
extraction. Anal Biochem 162: 156–159, 1987.
25. Gonzalez FJ, Nebert DW, Hardwick J and Casper C, Com-
plete cDNA and protein sequence of a pregnenolone 16␣-
carbonitrile-induced cytochrome P450. J Biol Chem 260:
7435–7441, 1985.
26. Shirasawa T, Sakamoto K, Akashi T, Takahashi H and
Kawashima A, Nucleotide sequence of rat elongation factor
1-␣ cDNA. Nucleic Acids Res 20: 909, 1992.
27. Sambrook J, Fritsch EF and Maniatis T, Enzymes used in
molecular cloning. In: Molecular Cloning, Vol. 1, pp. 5.68–
5.69. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY, 1989.
33. Watkins P, Noninvasive tests of CYP3A enzymes. Pharmaco-
genetics 4: 171–184, 1994.
34. Thummel K, Shen D, Podoll T, Kunze K, Trager W, Hartwell
P, Raisys V, Marsh C, McVicar J, Barr D, Perkins J and
Carithers R, Jr, Use of midazolam as a human cytochrome
P450 3A probe: I. In vitro–in vivo correlations in liver
transplant patients. J Pharmacol Exp Ther 271: 549–556,
1994.
35. Thummel K, Shen D, Podoll T, Kunze K, Trager W, Bacchi
C, Marsh C, McVicar J, Barr D, Perkins J and Carithers R, Jr,
Use of midazolam as a human cytochrome P450 3A probe: II.
Characterization of inter- and intraindividual hepatic CYP3A
variability after liver transplantation. J Pharmacol Exp Ther
271: 557–566, 1994.
36. Kinirons MT, O’Shea D, Downing TE, Fitzwilliam AT,
Joellenbeck L, Groopman JD, Wilkinson GR and Wood AJ,
Absence of correlations among three putative in vivo probes of
human cytochrome P4503A activity in young healthy men.
Clin Pharmacol Ther 54: 621–629, 1993.
37. Krivoruk Y, Kinirons MT, Wood AJ and Wood M, Metabo-
lism of cytochrome P4503A substrates in vivo administered by
the same route: Lack of correlation between aflentanil clear-
ance and erythromycin breath test. Clin Pharmacol Ther 56:
608–614, 1994.
28. Sambrook J, Fritsch EF and Maniatis T, Extraction, purifica-
tion, and analysis of messenger RNA from eukaryotic Cells.
In: Molecular Cloning, Vol. 1, pp. 7.43–7.45. Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, NY, 1989.
29. Ann DK, Lin HH, Lee S, Tu ZJ and Wang E, Characteriza-
tion of the statin-like S1 and rat elongation factor-1␣ as two
distinctly expressed messages in rat. J Biol Chem 267: 699–
702, 1992.
38. Watkins PB, Turgeon DK, Saenger P, Lown KS, Kolars JC,
Hamilton T, Fishman K, Guzelian PS and Voorhees JJ,
Comparison of urinary 6-␤-cortisol and the erythromycin
breath test as measures of hepatic P450IIIA (CYP3A) activ-
ity. Clin Pharmacol Ther 52: 265–273, 1992.
30. Lee S, Francoeur A, Liu S and Wang E, Tissue-specific