0022-3565/99/2892-1134$03.

00/0
THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 289, No. 2
Copyright © 1999 by The American Society for Pharmacology and Experimental Therapeutics Printed in U.S.A.
JPET 289:1134 –1142, 1999

First-Pass Midazolam Metabolism Catalyzed by

1a,25-Dihydroxy Vitamin D3-Modified Caco-2 Cell Monolayers1

JEANNINE M. FISHER, STEVEN A. WRIGHTON, PAUL B. WATKINS, PHYLLISSA SCHMIEDLIN-REN,

JUSTINA C. CALAMIA, DANNY D. SHEN, KENT L. KUNZE, and KENNETH E. THUMMEL
Departments of Pharmaceutics (J.C.C., D.D.S., K.E.T.) and Medicinal Chemistry (K.L.K.), University of Washington, Seattle, Washington;
Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan (P.B.W., P.S-R.); and Department of Drug Disposition, Eli Lilly &
Co., Indianapolis, Indiana (J.M.F., S.A.W.)
Accepted for publication January 16, 1999 This paper is available online at http://www.jpet.org

Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014

ABSTRACT
Cytochrome P-450 (CYP) 3A4 accounts for approximately 50%
of all P-450s found in the small intestine (Paine et al., 1997) and
contributes to the extensive and variable first-pass extraction of
drugs such as cyclosporine and saquinavir. We recently dem-
onstrated that CYP3A4 expression in a differentiated Caco-2
subclone is increased when cell monolayers are treated with
1a,25-dihydroxy-vitamin-D3 (Schmiedlin-Ren et al., 1997). This
improved metabolic capacity permits the in vitro modeling of
first-pass intestinal metabolic kinetics. Midazolam (MDZ) 19-
hydroxylation was used as a specific probe for CYP3A-medi-
ated metabolism in modified Caco-2 monolayers. Caco-2 cells
were grown to confluence on laminin-coated culture inserts,
and then for two additional weeks in the presence of 1a,25-
dihydroxy vitamin-D3. Cell monolayers were subsequently ex-
posed to MDZ for varying lengths of time and concentrations.
The amount of MDZ in the monolayer increased rapidly after
apical drug administration, reaching a pseudo steady state
within 6 min. The cellular uptake rate was considerably slower
after a basolateral dose. By either route of administration, the
rate of 19-hydroxymidazolam formation was stable and linear
for 2 h. Under basolateral sink conditions and low apical MDZ
dosing concentration (1– 8 mM), the first-pass extraction ratio
was found to be ;15%. Higher dosing concentrations led to
saturation of the hydroxylation reaction and reduction in the
extraction ratio. The modified Caco-2 cell monolayer is an
excellent model for studying drug absorption and first-pass
intestinal metabolic kinetic processes. In this system, the se-
lective CYP3A probe MDZ was rapidly absorbed, yet exten-
sively metabolized, as is observed in vivo.

Cytochrome P-450s (CYP) are one of the most important
classes of drug-metabolizing enzymes found in humans and
are responsible for the elimination of a wide variety of xeno-
biotics (Guengerich, 1995). Although these enzymes are pri-
marily concentrated in the liver, several extrahepatic tissues
including the human intestine are also known to express
appreciable levels of this class of enzymes. CYP3A4 is the
most abundant of all P-450 isoforms in the liver and small
intestine (Watkins et al., 1987; de Waziers et al., 1990; Paine
et al., 1997). Immunoblot analysis indicates that the median
microsomal CYP3A protein content (per mg total protein) in
the duodenal mucosa is approximately one-half that found in
the liver (de Waziers et al., 1990; Paine et al., 1997).
The presence of this major oxidizing enzyme in the small
intestine, specifically at the tips of the mucosal microvilli
(Kolars et al., 1994), makes it a potentially important site for
Received for publication October 19, 1998.
1
This study was supported in part by Eli Lilly & Co. and National Insti-
tutes of Health Grant GM 32165.
first-pass drug metabolism after oral administration. For
example, expression of CYP3A4 within villous epithelial cells
is thought to result in significant first-pass metabolism of
many clinically important substrates (Thummel et al., 1997),
including cyclosporine, saquinavir, midazolam (MDZ), nifed-
ipine, verapamil, and tacrolimus, thereby contributing to
their low overall oral bioavailability. In addition, clinically
relevant drug-drug interactions can result from inhibition or
induction of intestinal CYP3A, as seen with cyclosporine and
ketoconazole (Gomez et al., 1995) and with verapamil and
rifampin (Fromm et al., 1996). Despite a growing recognition
of the potential effect of intestinal CYP3A on oral drug bio-
availability and efficacy, a clear understanding of the kinetic
factors governing the extent of intestinal first-pass metabo-
lism in the absence and presence of enzyme modifying agents
is lacking.
Caco-2 cells are used widely in pharmaceutical research as
an in vitro model for studying intestinal drug absorption and
transport processes (Meunier et al., 1995; Boulenc, 1997).

ABBREVIATIONS: CYP, cytochrome P-450; P-gp, P-glycoprotein; MDZ, midazolam; 19-OH MDZ, 19-hydroxymidazolam; 1a,25-(OH)2-D3,
1a,25-di-hydroxy vitamin-D3; FBS, fetal bovine serum; DMEM, Dulbelcco’s modified Eagle’s medium; NEAA, non-essential amino acids; DM,
differentiation medium; PET, polyethylene terephthalate; TEER, transepithelial electrical resistance; DMSO, dimethylsulfoxide; ER, extraction ratio.

1134
1999
This colon-carcinoma derived cell line is easily grown on a
solid, permeable membrane to achieve a confluent monolayer
of fully differentiated cells with apically directed microvilli
and tight occluding junctions. In addition to the relevant
structural features that make Caco-2 cells well suited for
drug absorption studies, the differentiated cells are also
known to produce important transport proteins such as P-
glycoprotein (P-gp; Hunter et al., 1993) and some xenobiotic-
metabolizing enzymes including CYP1A1, glutathione-S-
transferase, and phenol-sulfotransferase (Meunier et al.,
1995; Boulenc, 1997). Although CYP3A4 is abundant in the
epithelium of a normal human small intestine, standard
culturing conditions with Caco-2 cells do not lead to signifi-
cant expression of the enzyme. There is evidence of CYP3A5
expression and activity in Caco-2 cells and a subclone TC7
(Gan et al., 1996; Raeissi et al., 1997); however, CYP3A4 and
CYP3A5 can produce different product ratios from the same
substrates (Wrighton et al., 1990; Gorski et al., 1994). There-
fore, by both quantitative and qualitative criteria, standard
Caco-2 cell monolayer cultures are not a suitable model for
first-pass metabolism.
In an effort to rectify this deficiency, Crespi et al., (1996)
recently reported the transfection of Caco-2 cells with an
extrachromosomal vector containing the CYP3A4 gene. Mi-
crosomes prepared from these transfected cells contain ap-
proximately 45 pmol P-450/mg protein. The intact cells
showed significant catalytic activity, as measured by testos-
terone 6b-hydroxylation. However, multiple passages of
these Caco-2 cells lead to loss of the plasmid and CYP3A4
expression.
Recent work in our laboratories has shown that CYP3A4
expression can be up-regulated in Caco-2 cells with the use of
1a,25-dihydroxy vitamin-D3 (1a,25-(OH)2-D3), a naturally
occurring hormone. CYP3A4 expression in a Caco-2 subclone
was increased when cells were grown to confluence on an
extracellular matrix such as Matrigel (Collaborative Biomed-
ical Products, Bedford, MA) or laminin and then treated with
0.25 mM 1a,25-(OH)2-D3 for 2 to 3 weeks (Schmiedlin-Ren et
al., 1997). In addition to the approximately 3-fold increase in
CYP3A4 apoprotein, there was also a 50-fold increase in its
activity as measured by midazolam 19-hydroxylation.
Our previous experiments with the modified Caco-2 sub-
clone focused on MDZ incubation times up to 24 h in the
presence of apical and basolateral medium containing fetal
bovine serum (FBS). The objective of the present work was to
characterize MDZ first-pass metabolic extraction and metab-
olite distribution across 1a,25-(OH)2-D3-treated Caco-2
monolayers under conditions of first-order product formation
in the absence of extracellular serum proteins.
Experimental Procedures
Materials. Caco-2 cells (American Type Culture Collection
HTB37) were cloned by limiting dilution as described previously
(Schmiedlin-Ren et al., 1997). Dulbelcco’s modified Eagle’s medium
(DMEM), non-essential amino acids (NEAA), penicillin, streptomy-
cin, and Hanks’ balanced salt solution were obtained from GIBCO
(Grand Island, NY). FBS was purchased from Hy-Clone Laborato-
ries, Inc. (Logan, UT). Uncoated and commercially coated track-
etched polyethylene terephthalate (PET) inserts and mouse laminin
were obtained from Collaborative Biomedical Products (Bedford,
MA). 1a,25-(OH)2-D3 was obtained from Calbiochem Corp. (La Jolla,
First-Pass Midazolam Metabolism in CaCo-2 Cells 1135
CA). N-Methyl-N-(t-butyl-dimethylsilyl)trifluoroacetamide (MTB-
STFA) was purchased from Pierce Chemical Co. (Rockford, IL). MDZ,
15
N3-MDZ, 19-hydroxymidazolam (19-OH MDZ), 4-OH MDZ, and 19-
[2H2]19-OH MDZ were gifts from Roche Laboratories (Nutley, NJ).
Acetonitrile and ethyl acetate were purchased from Fisher Scientific
Co. (Santa Clara, CA). SDS-polyacrylamide gel electrophoresis re-
agents (SDS, acrylamide, ammonium persulfate, and N,N,N9,N9-
tetra-methyl-ethylenediamine) were purchased from Bio-Rad (Her-
cules, CA). Nitrocellulose was purchased from Schleicher & Schuell
(Keene, NH). 5-Bromo-4-chloro-3-indoyl phosphate and nitroblue tet-
razolium was purchased from Kirkegaard and Perry (Gaithersburg,
MD). Anti-rabbit IgG alkaline phosphatase conjugate, EDTA, and
dimethyl sulfoxide (DMSO) were purchased from Sigma (St. Louis,
MO). All other chemicals were of tissue culture grade. Stock solu-
tions of MDZ and 1a,25-(OH)2-D3 were prepared in DMSO and
absolute ethanol, respectively.
Cell Culture Conditions. The Caco-2 subclone, P27.7 (Schmied-
lin-Ren et al., 1997), was obtained at passage 12 and grown on
culture dishes in passage medium consisting of DMEM containing 25
Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014
mM glucose and 4 mM L-glutamine, 0.1 mM NEAA, 100 U/ml sodium
penicillin G, 100 mg/ml streptomycin, and 20% heat-inactivated FBS.
All experiments were performed with cells at passage number 19 or
20. Unless otherwise noted, cells were seeded onto laminin-coated
PET inserts at 5.2 3 106 cells/cm2 and grown in complete growth
medium (passage medium supplemented with 45 nM DL-a-tocoph-
erol) until confluent. Upon achieving confluence, cell monolayers
were fed for 2 weeks every 2 to 3 days with complete differentiation
medium (DM) containing DMEM, 0.1 mM NEAA, 100 U/ml sodium
penicillin, 100 mg/ml streptomycin, 0.1 mM sodium selenite, 3 mM
zinc sulfate, 45 nM DL-a-tocopherol, 0.25 mM 1a,25-(OH)2-D3, and 5%
heat-inactivated FBS.
On a given experiment day, transepithelial electrical resistance
(TEER) readings were recorded before initiation of the experiment.
DM was removed and cells were washed three times with 1 ml of
DMEM before addition of experimental medium. Modified DM (ab-
sent 1a,25-(OH)2-D3 and FBS) was spiked with midazolam to the
desired concentration and added to the appropriate culture insert
compartment (1.5 ml volume for apical and basolateral compart-
ments). The final concentration of DMSO in the dosing medium was
always ,1%. Apical and basolateral media were collected upon com-
pletion of the indicated incubation period and frozen at 220°C pend-
ing analysis. Cells were quickly rinsed with 1 ml of DMEM and scraped
into a fresh 1 ml of medium and immediately frozen at 220°C.
Except for the results described in the sections Concentration
Dependence of 19-OH and 4-OH MDZ Formation, Intraday Variabil-
ity in 19-OH MDZ Formation and Distribution, and Saturable First-
Fig. 1. TEER measurements for Caco-2 cell monolayers grown on hand-
applied laminin coated inserts and treated for increasing number of days
with 1a,25-(OH)2-D3. Measurements for each 4.2 cm2 culture insert were
obtained 14 days after the cells had reached confluence.
1136 Fisher et al. Vol. 289

TABLE 1 phosphate, 1 mM EDTA, pH 7.4) and centrifuged at 110,000g for 70

Intraday variability in 19-OH MDZ formation and distribution in min at 4°C. The final pellet was reconstituted in 250 ml of storage
Caco-2 cells buffer (100 mM potassium phosphate and 20% glycerol, pH 7.4) and
Amount of 19-OH MDZ Formeda immediately frozen with the other fractions at 280°C. Protein con-
Culture A/B Ratio centrations were determined by the method of Lowry et al. (1951)
Apical Basolateral Cellular Total
using bovine serum albumin as the reference standard.
pmol Western Blot Analysis of CYP3A. Total CYP3A protein in mod-
1 111.8 46.0 29.5 157.8 2.43 ified Caco-2 cell fractions was immunoquantified and compared to
2 139.1 49.3 27.9 188.4 2.82 human liver and duodenal microsomal CYP3A content. Liver and
3 128.3 43.6 24.2 171.8 2.95
small intestinal microsomes were prepared as described previously
4 154.8 58.5 23.5 213.3 2.65
5 124.5 50.9 20.4 175.4 2.44 (Paine et al., 1997). After dilution in buffer (final concentrations: 60
6 116.3 46.5 23.3 162.9 2.50 mM Tris-HCL, pH 7.4; 20% glycerol; 0.2% Emulgen 911; 2% SDS; 5%
Mean 129.1 49.1 24.8 178.3 2.63 b-mercaptoethanol), samples were boiled for 3 min and loaded onto a
SD 15.8 5.2 3.3 20.2 0.22 9% acrylamide/0.1% SDS gel. Protein components of Caco-2 homog-
CV 12.2% 10.7% 13.4 11.3% 8.2%
enate (50 mg), 1st pellet (50 mg), 21,400g supernatant (40 mg), and
a
Caco-2 monolayers were incubated with 4.5 nmol apically applied MDZ for 30 final 110,000g pellet (20 mg), and human intestinal (20 mg) and liver
min at 37°C.
(10 mg) microsomes were separated by electrophoresis (Paine et al.,
1997). Purified expressed human CYP3A4 and CYP3A5 standards
Pass Metabolic Extraction, all sets of data points (grouped by collec-

Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014

tion time) represent results from a single culture.
Monolayer Integrity. TEER was the primary determinant of cell
monolayer integrity. Resistance (ohm) was measured on each culture
immediately before an experiment with a Millipore Millicell electri-
cal resistance system (Bedford, MA). For a given set of cell cultures,
a single insert that did not contain cells was measured for back-
ground resistance. TEER was determined as the product of the
background-corrected resistance and the surface area of the insert
(4.2 cm2).
Extracellular Matrix Proteins. In specified experiments, cells
were seeded onto commercially available culture inserts precoated
with 25 mg/cm2 mouse laminin. In all other experiments, control
inserts were hand-coated with 5 mg/cm2 mouse laminin according to
Collaborative Biomedical Product’s specification immediately before
seeding cells. Briefly, mouse laminin was thawed at 4°C overnight
and diluted into DMEM to a final concentration of 21 mg/ml. Each 4.2
cm2 PET insert was coated with 1 ml of the diluted laminin and
incubated at room temperature for 1 h. The medium was then aspi-
rated and the coated inserts were rinsed three times with 1 ml of
fresh DMEM and then immediately seeded with Caco-2 cells. Cells
grown on hand-coated laminin inserts were morphologically identi-
cal with those grown on commercially coated inserts, as determined
by electron microscopy, although the laminin layer appeared thinner
(data not shown). Additionally, cells achieved confluence after ap-
proximately 4 days on hand-coated inserts compared with 7 days
with commercially coated inserts.
Preparation of Caco-2 Cell Fractions. Caco-2 cells were grown
on 28 commercially coated laminin inserts and treated with 1a,25-
(OH)2-D3 for 2 weeks postconfluence. After recording the TEER on
each culture, cell monolayers were rinsed three times with 1 ml
DMEM, scraped with a metal spatula into 500 ml of homogenizing
buffer (100 mM potassium phosphate, 0.25 M sucrose, and 1 mM
EDTA, pH 7.4), and collected in a 50-ml conical tube. Each insert was
then rinsed with an additional 500 ml of homogenizing buffer, which
was transferred to the same tube. The cell homogenate was subjected
to centrifugation at ;300g at 4°C for 10 min, at which time the
supernatant was aspirated and the pellet was reconstituted in ap-
proximately 6 ml of homogenizing buffer. Cells were then transferred
to a 10-ml glass Wheaton tube and hand-homogenized (20 strokes)
and drill-homogenized (five strokes) on ice with a Teflon-tipped pes-
tle. A 300 ml aliquot was immediately frozen for protein and immu-
nochemical analysis, and the remaining homogenate was subjected
to a low-speed centrifugation at 600g for 5 min, then at 21,400g for
25 min at 4°C. The resulting yellow pellet (1st pellet) was resus-
pended in 1 ml of homogenizing buffer and frozen before Western
blot analysis. A 500 ml aliquot of the supernatant was immediately
frozen, whereas the remainder was diluted to 60 ml and centrifuged
at 110,000g for 70 min at 4°C. The resulting microsomal pellet (2nd
pellet) was resuspended in 500 ml of wash buffer (10 mM potassium
(0.5 and 1 pmol each) were prepared in the same manner and loaded
onto the same gel into lanes parallel to cellular fractions. Transfer of
proteins to nitrocellulose sheets, and CYP3A immunoblot detection
and quantification were described previously (Paine et al., 1997).
MDZ, 1*-OH MDZ, and 4-OH MDZ Assays. Formation of 19-OH
MDZ was the primary measure of MDZ metabolism by Caco-2 cell
cultures at low (,3 mM) dosing concentrations (Schmiedlin-Ren et
al., 1997). Because 4-OH MDZ formation is increasingly favored at
higher concentrations in microsomal incubations (Gorski et al., 1994;
Paine et al., 1997), both the 19-OH and 4-OH metabolites were
measured in Caco-2 cultures incubated with initial apical MDZ con-
centrations greater than 3 mM to assess saturable metabolite kinet-
ics. 4-OH MDZ was not quantitated in all samples to conserve a
limited quantity of 15N3-MDZ-labeled 4-OH MDZ internal standard.
Measurement of MDZ levels in the apical and basolateral medium
and cell mass were used to calculate the extent of drug permeation
into and across the cell monolayer. Apical and basolateral media
from a single culture were assayed in duplicate for parent drug and
metabolites as described previously (Schmiedlin-Ren et al., 1997).
For the analysis of intracellular MDZ, 19-OH MDZ and 4-OH MDZ
0.02 to 0.8 ml of cell suspension were diluted with deionized water to
a 1-ml volume before the solvent extraction step.
Determination of MDZ Permeability Coefficient. The appar-
ent permeability coefficient (Papp) of a drug across Caco-2 monolay-
ers is often used as an in vitro predictor of in vivo drug absorption
(Artursson, 1990; Gres et al., 1998). The Papp (cm/s) for MDZ in the
modified Caco-2 monolayer was calculated according to the equation
described by Artursson et al., (1990):
dQ 1
Papp 5 z (1)
dt AC 0
where dQ/dt is the rate of MDZ transfer (pmol/s) into the receiver
compartment after an apical or basolateral MDZ dose under sink
conditions (when ,10% of the dose had crossed into the receiver
compartment), A is the area of the culture insert (4.2 cm2), and C0 is
the concentration of MDZ in the dosing compartment at time zero.
For these experiments, modified Caco-2 cells grown on laminin-
coated inserts were treated apically or basolaterally with 3 mM MDZ.
Determination of First-Pass Metabolic Extraction Ratio
(ER). Initial experimental results allowed us to define linear, first-
order conditions for metabolism during first-passage of MDZ across
the Caco-2 cell monolayer. Briefly, MDZ was incubated under sink
conditions (approximately 20 min in duration, when ,10% of the
apical dose had crossed into the basolateral compartment). The
amounts of 19-OH MDZ and 4-OH MDZ found in the apical, basolat-
eral, and cell compartments were summed and entered into eq. 2,
1999 First-Pass Midazolam Metabolism in CaCo-2 Cells 1137
TABLE 2 (OH)2-D3-treated cells when grown on hand-coated (717 ver-
Apical/basolateral 19-OH MDZ concentration ratios sus 589 ohm z cm2, p , .01) or commercially coated (1292
Caco-2 monolayers were treated with 3 mM MDZ in either the apical or basolateral versus 471 ohm z cm2, p , .01) inserts. Additionally, we
compartment. Total 19-OH MDZ was measured in the apical and basolateral medium
after incubating for the indicated period of time. Each ratio represents the result observed a decreasing TEER for confluent cells that were
obtained from a single culture.
grown on hand-applied laminin-coated inserts and treated
Time After Apical Dose Basolateral Dose for an increasing number of days with 1a,25-(OH)2-D3 (Fig.
Dose A/B Ratio A/B Ratio
1). A similar observation was recently reported by Chirayath
min
et al., (1998) who described a 50% decrease in TEER in
20 1.8 2.2 2-week postconfluent Caco-2 cells treated with 1028 M 1a,25-
30 2.8 2.3
45 3.8 2.7 (OH)2-D3. Their results suggest that this decrease could be
60 1.6 2.4 due to an increased permeability in the tight junctions be-
90 1.6 2.0 tween cells.
120 2.9 2.0
Mean 2.4 2.3 Stability of CYP3A4 Activity. To make the cell culture
S.D. 0.90 0.27 system more amenable for studies of first-pass metabolism,
FBS and 1a,25-(OH)2-D3 were removed from the apical (A)
along with the amount of MDZ in the basolateral compartment, to and basolateral (B) media during the MDZ incubation period.
calculate a first-pass ER: Because depletion of these agents would eventually alter cell

Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014

Apical Dose
ERCaco22 5
S(192OH MDZ) 1 S ~ 4 2 OH MDZ!
S ~ 19 2 OH MDZ! 1 S ~ 4 2 OH MDZ! 1 MDZbasolateral
(2)
When apical dosing concentrations of MDZ were relatively low (#3
mM), it was found that 4-OH MDZ did not contribute significantly to
total metabolite formation and it was not included in the calculated
extraction ratios shown for interday variability.
Statistics. A single enzyme, Michaelis-Menten model was fit sep-
arately to initial formation rates of 19-OH and 4-OH MDZ at varying
apical concentrations of MDZ using WinNonlin version 1.0 (Apex,
NC) with a constant coefficient of variance error model. All other
statistical analyses were performed with the statistical software
SPSS version 7.5 (Chicago, IL). For dose-dependent metabolite ki-
netics, ANOVA was used to determine if there was a significant
difference in mean parameter values for the four MDZ dose groups.
Results
TEER. TEER (ohm z cm2) was recorded for all cells at 2
weeks postconfluence. Values ranged from 375 to 953 ohm z
cm2 for 1a,25-(OH)2-D3-treated cells grown on hand-coated
inserts and 315 to 688 ohm z cm2 for cells grown on commer-
cially coated inserts. There was a statistically significant
difference in mean TEER between untreated and 1a,25-
CYP3A4 expression and metabolic activity, MDZ 19-hydroxy-
lation activity was evaluated after incubating cell monolay-
ers for varying intervals over a 24-h period in culture me-
dium lacking FBS and 1a,25-(OH)2-D3. As seen in Fig. 2, the
total amount of 19-OH MDZ formed during a 30-min incuba-
tion with 3 mM apically applied MDZ was relatively stable for
the first 4 h of serum-free incubation, after which time ac-
tivity gradually declined.
Apical, basolateral, and intracellular 19-OH MDZ all de-
creased in parallel with time. Twenty-four hours after re-
moval of FBS and 1a,25-(OH)2-D3, the activity was found to
be 27% of peak metabolite formation. Based on these initial
findings, all incubation experiments with MDZ were limited
to no more than 2 h in the absence of serum and 1a,25-(OH)2-
D3.
Reproducibility of Cell Culture Activity. A reliable
assessment of first-pass kinetics depends on a reproducible
measurement of the Caco-2 uptake and metabolism of MDZ
during relatively short (#30-min) incubation. Table 1 pro-
vides the intraday variability in 19-OH MDZ appearance in
the apical, basolateral, and cellular compartments from in-
dividual cultures after a 30-min incubation with 3 mM apical
MDZ. Total 19-OH MDZ formation was reasonably consistent
across cultures (c.v. 5 11.3%) as were apical/basolateral con-

TABLE 3
Dose-dependent metabolite kinetics and distribution in Caco-2 monolayers
Caco-2 monolayers were treated apically with a range of MDZ concentrations. 19-OH and 4-OH MDZ were measured in the apical and basolateral medium and in the cell
homogenate. Each value represents the mean (and S.D.) of three cultures.

19-OH MDZ 4-OH MDZ

Apical MDZ, 1⁄4c Total Rate of
a ERd
Conc. A/B Total A/B Total Ratio Metabolism
Ratio Formedb Ratio Formed

mM pmol pmol pmol/min

3 2.2 82.8 3.5 6.7 12.3 2.99 9.85
(0.2) (9.97) (0.49) (0.35) (0.98) (0.34) (1.2)
8 2.1 152 2.9 14.6 10.7 5.56 7.53
(0.2) (4.0) (0.82) (2.6) (2.1) (0.13) (0.22)
25 2.5 246 3.3 40.5 6.1 9.54 4.77
(0.4) (18.6) (0.24) (1.2) (0.47) (0.62) (0.46)
100 2.4 305 3.6 100 3.1 13.5 1.64
(0.2) (35.2) (0.08) (8.5) (0.39) (1.3) (0.30)
a
A/B denotes the apical-to-basolateral concentration ratio measured at the end of the incubation. Mean 19-OH MDZ and 4-OH MDZ A/B ratios for the four dose groups
were not different by ANOVA (p 5 .62 and p 5 .32, respectively).
b
Total metabolite formed is the sum of the amount found in the apical and basolateral medium and cell homogenate at the end of the 30-min incubation time.
c
Mean 19-OH MDZ/4-OH MDZ ratios for the four dose groups were significantly different by ANOVA (p , .001).
d
The first-pass ER was calculated by application of the sum of the 19-OH MDZ and 4-OH MDZ produced, and total MDZ in the basolateral medium, to eq. 2. Mean ER
values for the four dose groups were significantly different by ANOVA (p , .001).
1138 Fisher et al. Vol. 289

Consistent with a high level of permeability, MDZ trans-

cellular (donor/recipient) concentration ratios declined
sharply during the first 30 min of incubation after an apical
or basolateral dose (Fig. 3b). The ratio remained above unity
at the end of the 2-h incubation interval, although the apical
transcellular ratio for MDZ appeared to reach a plateau at a
value just above unity after 1 h, whereas the basolateral ratio
continued to decline towards that same level beyond 1 h. Our
previous results (Schmiedlin-Ren et al., 1997) indicate that,
after approximately 10 h of incubation with either an apical
or basolateral dose, the MDZ concentrations will eventually
equilibrate across the Caco-2 monolayer such that the apical
concentration is 1.2- to 1.4-fold higher than the basolateral
concentration.
Extracellular 1*-OH MDZ Partitioning. We previously
observed preferential sorting of 19-OH MDZ into the apical
Fig. 2. Time-course of 19-OH-MDZ formation activity in Caco-2 cell mono-
compartment (compared to basolateral) after apical or baso-
layers after conversion to serum-free, 1a,25-(OH)2-D3-free differentiation

Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014

medium. Total (■), apical (h), basolateral (Œ), and cellular (‚) 19-OH lateral dosing of MDZ in cultures incubated up to 24 h
MDZ content represent amounts measured after incubation for 30 min (Schmiedlin-Ren et al., 1997). As seen in Table 2, apically
with apically-applied MDZ. Each point represents data obtained from a directed sorting of 19-OH MDZ (;2.4:1) also occurred in the
single culture.

centration ratios (c.v. 5 8.2%). Cellular 19-OH MDZ ac-

counted for up to 18.7% of the total metabolite formed.
MDZ Flux after Apical and Basolateral Administra-
tion. Apical or basolateral dosing of MDZ (3 mM) allowed us
to study the absorption (A 3 B) and exsorption (B 3 A)
kinetics of this drug across the modified Caco-2 monolayer
under nonsaturating conditions. Consistent with our previ-
ous findings (Schmiedlin-Ren et al., 1997), we observed a
rapid decline in MDZ levels in both dosing compartments
over the 2-h incubation period (Fig. 3a). Although the initial
rate of loss of MDZ (0 – 60 min) from the basolateral compart-
ment appeared slower than the loss from the apical compart-
ment, the transcellular fluxes (as measured by the rate of
MDZ accumulation in the receiving compartment) did not
differ. Application of the initial rates of accumulation in the
receiving compartments to eq. 1 yielded mean transcellular
Papp values for MDZ dosed into either the apical (n 5 5) or
basolateral (n 5 3) compartment of 28.5 6 9.25 3 1026 cm/s
and 20.3 6 4.56 3 1026 cm/s, respectively. There was no
statistical difference in the Papp calculated after dosing into
either compartment.

TABLE 4
Interday variability of ER in modified Caco-2 cells
Experiment Incubation Total 19- Basolateral Basolateral MDZ ERc
No.a time OH-MDZb MDZ

min pmol pmol % of Apical Dose

1 30 66.1 374 9.1 15.1
2 20 35.0 264 9.9 11.7
3 30 90.9 375 8.3 19.5
4 10 53.2 280 5.2 16.0
5 20 44.6 347 10.0 11.4
6 20 57.1 374 8.9 13.2
Range 11.4–19.5
Average 14.5
SD 3.1
CV 21.1
a
Fig. 3. Time-course of incubation of 1a,25-(OH)2-D3-treated Caco-2 mono-
Experiment number refers to experiments where 4.5 nmol of MDZ was dosed layers with MDZ. A 3 mM dose was applied to either the apical (open
apically and 10% or less of this dose crossed into the basolateral compartment after
symbols) or basolateral (closed symbols) compartment. a, amount of MDZ
the indicated incubation time.
b
Total 19-OH-MDZ is the sum of the amount found in the apical and basolateral in the apical (squares) or basolateral (triangles) compartments. b, tran-
medium and the amount remaining in the cell at the end of the incubation time. scellular MDZ concentration ratios. c, total 19-OH MDZ (pmol) formation.
c
The first-pass ER was calculated by application of the total amount of 19-OH Each point represents data obtained from a single culture. The dotted line
MDZ formed and total MDZ in the basolateral medium to eq. 2. in Fig. 3b indicates unity.
1999
absence of FBS and at shorter incubation intervals. In this
experiment, apical, basolateral, and cellular 19-OH MDZ
were quantitated after 10 to 120 min of incubation with 3 mM
MDZ dosed either apically or basolaterally. Although there
was slightly more metabolite produced during the first 45
min after apical MDZ dosing, in comparison with basolateral
dosing, total product formation was comparable during the
60- to 120-min interval (Fig. 3c). In addition, the apical/
basolateral 19-OH MDZ ratio was above unity for both routes
of administration over the entire 2-h incubation period (Table
2).
First-Pass MDZ Uptake and Metabolism. Based on the
apparent speed with which MDZ gained access to intracellu-
lar CYP3A, we examined the kinetics of MDZ metabolism
during the first 10 min of incubation. After apical adminis-
tration, MDZ diffused rapidly into the cell reaching a con-
stant level within 4 to 6 min (Fig. 4a). Cellular uptake of
First-Pass Midazolam Metabolism in CaCo-2 Cells 1139
for the 19-OH and 4-OH MDZ formation pathways, respec-
tively. Because the concentration of MDZ at the active site of
the enzyme is unknown, Km,app is based on the initial apical
MDZ concentration. However, the total cellular MDZ level
correlated extremely well with the apical MDZ concentration
after the 30-min incubation period (r 5 0.998), indicating a
rapid equilibration between the two compartments. The
MDZ A/B concentration ratio increased slightly with an in-
crease in the initial MDZ dose concentration; 3.74 6 0.13,
4.24 6 0.20, 4.63 6 0.26, and 4.89 6 0.68 for 3, 8, 25, and 100
mM MDZ. Mean values for the 3 and 100 mM dose groups
were significantly different (p 5 .03).
Increasing apical dosing concentrations of MDZ also re-
sulted in an increase in the contribution of 4-OH MDZ to total
product formation (Fig. 5 and Table 3). Consistent with a
trend seen previously in incubations with human duodenal
microsomes (Paine et al., 1997), the ratio of 19-OH to 4-OH

Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014

MDZ after a basolateral dose was much slower. It steadily MDZ in Caco-2 monolayers decreased from 12.3 to 3.1 as
increased over the 10-min incubation period, reaching an MDZ dosing concentrations increased from 3 to 100 mM (p ,
amount that was ;1⁄3 of that achieved after apical dosing. .001). At the lowest dosing concentrations, total 4-OH MDZ
Interestingly, addition of MDZ to both apical and basolateral represented only 7.4% of total product formation. Similar to
compartments simultaneously caused only a 20% increase in 19-OH MDZ, 4-OH preferentially sorted (;3.5:1) to the apical
the cellular MDZ level after 10 min of incubation over that
observed after apical dosing alone.
The cumulative amount of 19-OH MDZ formed after each
route of administration of MDZ corresponded to the amount
of MDZ that accumulated in the cell monolayer (Fig. 4b).
19-OH MDZ appeared immediately after apical MDZ admin-
istration and, after 10 min, was approximately 4-fold higher
than the amount formed after basolateral dosing. Simulta-
neous administration of MDZ resulted in a 30% higher 19-OH
MDZ formation than after apical dosing, which was consis-
tent with the modest increase in the cellular MDZ level (Fig.
4b).
To evaluate the possible presence of an apical or basolat-
eral MDZ efflux pump, we examined MDZ transfer from the
dosing compartment to the receiving compartment under
sink conditions. Total 19-OH MDZ formation at the end of a
10-min incubation was only 53 and 15 pmol, or 1.5% and 0.3%
of the 4500 pmol (3 mM) apical and basolateral MDZ dose,
respectively. As seen in Fig. 4c, the rates of transfer of MDZ
from the apical to basolateral compartment (A 3 B) and from
the basolateral to apical (B 3 A) compartment were linear
during the 10-min incubation. Therefore, although the initial
accumulation of MDZ in the cellular compartment was much
lower after basolateral dosing (Fig. 4a), the rate of transfer of
MDZ across the cell monolayer was roughly equal after apical
and basolateral dosing (Fig. 4c).
Saturation Kinetics of Product Formation. Under
conditions of near constant intracellular MDZ concentrations
(0 –30 min after an apical dose), formation rates of 19-OH and
4-OH MDZ were measured over a range of initial apical MDZ
dosing concentrations (3.0 –100 mM). Results are presented
in Fig. 5 and Table 3. Each point represents the mean for-
mation rate of three cultures incubated for 30 min with the
indicated apical MDZ dose. The plot shows saturation of both
the 19-OH MDZ and 4-OH MDZ formation pathways in the
Caco-2 cell monolayer at apical MDZ concentrations exceed- Fig. 4. Early time-course of incubation of 1a,25-(OH)2-D3-treated Caco-2
ing ;25 and 80 mM, respectively. A single-enzyme, Michae- monolayers with MDZ. A 3 mM dose was applied to the apical (h),
lis-Menten kinetic model was fit to the data to yield apparent basolateral (Œ), or both (■) compartments. a, MDZ uptake into Caco-2
monolayers. b, total 19-OH MDZ (pmol) formation. c, transcellular flux of
Michaelis constants (Km,app) of 9.1 and 84.7 mM and maxi- MDZ (A 3 B or B 3 A). Each point represents data obtained from a single
mum formation rates (Vmax) of 11.1 and 6.1 pmol/min/culture culture incubated for the indicated time.
1140 Fisher et al.
Fig. 5. Dependence of initial 19-OH (F) and 4-OH MDZ (E) formation
rates on MDZ concentration. Total 19-OH and 4-OH MDZ formation were
measured after a 30-min incubation with MDZ (3–100 mM) applied to the
apical compartment. Each data point represents the mean rate of forma-
tion from three cultures. Error bars representing S.D. for some formation
rates were too small to be seen on the plots. The dotted lines represent the
fit of a single-enzyme, Michaelis-Menten model to the data; Km,app 5 9.1
mM, Vmax 5 11.1 pmol/min/culture (19-OH MDZ); Km,app 5 84.7 mM, Vmax
5 6.1 pmol/min/culture (4-OH MDZ).
Fig. 6. Western blot of 1a,25-(OH)2-D3-treated Caco-2 cell fractions and
microsomes from human liver and small intestine. Caco-2 cell homoge-
nate (50 mg), 1st centrifugation 21,400g pellet (50 mg), 1st 21,400g super-
natant (40 mg), and final 110,000g centrifugation pellet (20 mg) and
human intestinal (20 mg) and liver (10 mg) microsomes were electropho-
resed in parallel to expressed CYP3A4 and CYP3A5 standards (0.5 and 1
pmol each). Resolved proteins were transferred to nitrocellulose and
probed with an anti-human CYP3A4 polyclonal antibody, as described in
Experimental Procedures.
compartment after an apical dose, but neither metabolite
showed any dose-dependent sorting over the indicated con-
centration range.
Determination of First-Pass MDZ ER. Application of
eq. 2 to the MDZ and 19-OH MDZ results obtained from six
separate experiments allowed us to determine an interday
variability for the first-pass MDZ ER under subsaturating
conditions (Table 4). ER values were calculated from single
cultures incubated for 30 min or less with an apical MDZ
dose concentration of 3 mM. The duration of MDZ incubation
was selected based on the time when less than 10% of the
apical MDZ dose had appeared in the basolateral compart-
Vol. 289
ment. Using these criteria, ER values between 11.4 and
19.0% were obtained, with a mean of 14.5 6 3.1%.
A first-pass ER was also calculated for the Caco-2 mono-
layers dosed apically with increasing concentrations of MDZ
(3–100 mM) using 19-OH and 4-OH MDZ measurements (Ta-
ble 3). Although the incubation period was limited to 30 min,
the percentage of the initial MDZ dose that appeared in the
basolateral compartment exceeded our preset limit of 10%
(13.1 to 18.1%). Under this condition, increasing dosing con-
centrations resulted in a significant reduction in the appar-
ent first-pass ER from 9.8 6 1.2% (3 mM) to 1.6 6 0.26% (100
mM; p , .001).
Immunoquantitation of CYP3A in Modified Caco-2
Cells. Total CYP3A content in modified Caco-2 cells was
determined by Western blot analysis of a pooled homogenate
prepared from 28 replicate monolayer cultures. CYP3A4 ap-
peared to be the only protein in 1a,25-(OH)2-D3-treated cells
Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014
detected by the anti-CYP3A antibody (Fig. 6). Quantitation of
the homogenate CYP3A4 band density relative to standard
indicated a specific content of 3.7 pmol/4.2 cm2 cultured
insert. Analysis of the 1st centrifugation pellet (21,400g)
prepared from homogenate indicated that approximately
90% of CYP3A protein was removed with the plasma mem-
brane and mitochondrial fragments. Consequently, total
CYP3A yield in the 2nd pellet (110,000g) was only 5% of the
original amount. The specific content for 1a,25-(OH)2-D3
treated Caco-2 cell microsomes was found to be 8.3 pmol/mg
protein.
Discussion
Although Caco-2 cells are an excellent model for the pre-
diction of drug absorption across the intestinal epithelium,
until recently they have not been used as an in vitro model
for P-450-mediated drug metabolism. When treated with
1a,25-(OH)2-D3 for 2 weeks postconfluence, CYP3A4 mRNA
and protein, the dominant P-450 isoform in human intestinal
mucosa (de Waziers et al., 1990; Paine et al., 1997), were
increased substantially in the Caco-2 cell monolayer
(Schmiedlin-Ren et al., 1997). Results from the present study
show that CYP3A-induced Caco-2 monolayers can also be
used as a model for first-pass intestinal drug metabolism.
MDZ is a short-acting, water-soluble benzodiazepine that
is rapidly and nearly completely absorbed after oral admin-
istration (Tmax 15–30 min; Heizmann and Ziegler, 1981;
Smith et al., 1981). Consistent with these physicochemical
properties, we observed a rapid uptake of MDZ into the cell
monolayer and a rapid flux of MDZ from the apical to baso-
lateral compartment, as described by a high Papp of 28.5 3
1026 cm/s. Many substrates for CYP3A4 are transported
across barrier cell plasma membranes by P-gp (Wacher et al.,
1995). For example, efficient P-gp-mediated luminally-di-
rected efflux of cyclosporine from intestinal epithelial cells
appears to delay and limit the oral bioavailability of the drug
(Lown et al., 1997). For cyclosporine and other P-gp sub-
strates, transcellular flux in Caco-2 monolayers is faster in
the (B 3 A) direction than in the (A 3 B) direction. Thus, the
nearly equivalent rates of MDZ accumulation in the receiving
compartment we observed after apical or basolateral MDZ
administration (Fig. 4c) suggest that MDZ is not a significant
substrate for the P-gp efflux transporter. One caveat to this
conclusion is the modest 31% increase in the 30-min A/B
1999
MDZ concentration ratio that was observed with increasing
(3- to 100-mM) MDZ dose concentration. However, this trend
was more consistent with the saturation of a basolaterally
directed active transport process and not an apically directed
transporter such as P-gp. In addition, our previous results
from Caco-2 monolayer incubation with MDZ in the presence
of the selective P-gp inhibitor verapamil (Schmiedlin-Ren et
al., 1997) also support this conclusion.
A plausible explanation of the delayed rate of MDZ uptake
into the cell monolayer from the basolateral compartment
(Fig. 4a), in comparison with the apical compartment, is the
difference between apical and basolateral surface area and
the presence of the laminin barrier. The large surface area of
the apical membrane promoted a rapid uptake of MDZ into
the cell mass after apical administration, whereas the
smaller surface area and presence of the laminin coating
resulted in a relatively slow uptake from the basolateral
compartment over a 10-min incubation period. Identification
of the rate-limiting step for the delayed uptake of MDZ from
the basolateral compartment was beyond the scope of this
study.
In principle, the rate of cellular uptake of a CYP3A4 sub-
strate may have a significant effect on the extent of first-pass
and systemic intestinal metabolism. We examined this by
administration of MDZ to one or both extracellular compart-
ments. When the rate of diffusional efflux (apical or basolat-
eral dose) from the intracellular compartment was rendered
negligible by high MDZ concentrations in the receiver com-
partment (Fig. 4, a and b), one might have expected greater
cellular MDZ accumulation and product formation than
when an extracellular sink was present. However, the ab-
sence of such findings suggests that after an apical (oral)
MDZ dose, the initial pseudo-steady-state intracellular MDZ
concentration is controlled primarily by the apical MDZ con-
centration and the intracellular metabolic clearance, and not
MDZ back-diffusion across the basolateral membrane.
After its formation, 19-OH MDZ movement out of the mod-
ified Caco-2 cell monolayer could be governed by diffusion or
active efflux transport. In both this study and our previous
work (Schmiedlin-Ren et al., 1997), 19-OH MDZ preferen-
tially sorted to the apical compartment (A/B ratios greater
than unity at all incubation times) regardless of the admin-
istration compartment. Similar results were found in the
presence of FBS in both compartments after co-administra-
tion of 100 mM verapamil (Schmiedlin-Ren et al., 1997), sug-
gesting that the 19-OH MDZ sorting mechanism is also not
controlled by the apically directed P-gp pump that is ex-
pressed in these cells. This does not rule out the possibility of
an alternate transport mechanism. However, the apparent
preference for metabolite distribution into the apical com-
partment could also be explained by a purely diffusional
mechanism. Under conditions where the enzyme provides a
continuous generation of 19-OH MDZ in the cellular (C) com-
partment, it might be expected that the greater surface area
of the apical plasma membrane, compared with the basolat-
eral membrane, would result in a C 3 A flux of metabolite
that exceeds the C 3 B flux. A true equilibrium would not be
achieved unless MDZ metabolism were to cease. Further
studies are needed to discriminate between an active trans-
port system and simple diffusion. However, under more phys-
iologically relevant conditions, with extensive (;90%)
plasma protein binding of the MDZ metabolite (Mandema et
First-Pass Midazolam Metabolism in CaCo-2 Cells 1141
al., 1992), blood in the capillaries of the lamina propria may
act as a sink and override any “preferred” apical (luminal)
19-OH MDZ sorting.
Saturable enzyme kinetics were also evaluated in the mod-
ified cell monolayer by measuring total 19-OH and 4-OH
MDZ formation as a function of increasing concentrations of
apically applied MDZ. A single-enzyme, Michaelis-Menten
model appeared to describe both metabolite formation rate
versus initial apical MDZ concentration plots (Fig. 5). There
was no evidence of enzyme inactivation or nonhyperbolic
kinetics at high concentrations of MDZ, although an exten-
sive profiling of the concentration-velocity curve was not
performed. The estimated Km,app of 84.7 mM for 4-OH MDZ in
Caco-2 cultures was similar to (Gorski et al., 1994) or higher
than (Ghosal et al., 1996) the mean value reported with
human liver microsomes (86.5 mM) and cDNA expressed
CYP3A4 (38 mM), respectively. The estimated Km,app of 9.1
mM for 19-OH MDZ in the Caco-2 monolayer was also higher
Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014
than the median Km value reported for microsomes from
human small intestine (3.8, 3.7, and 4.5 mM for duodenum,
jejunum, and ileum, respectively; Paine et al., 1997) and the
Km for cDNA-expressed CYP3A4 (1.6 mM; Ghosal et al.,
1996). These data suggest that under pseudo-steady-state
conditions, the mean unbound intracellular concentration of
MDZ at the CYP3A4 active site is less than one-half the
apical MDZ concentration. Diffusion of MDZ out of the cell
monolayer and a high rate of metabolism could potentially
have reduced the unbound intracellular concentration at the
enzyme active site relative to the apical concentration, yield-
ing a higher calculated Km,app. Relative rates of cell uptake
from the dosing compartment, intracellular metabolism and
diffusion into the receiver compartment may be similarly
important factors to consider when evaluating the metabolic
extraction of other CYP3A4 substrates.
Based on the reproducibility of the rate of metabolite for-
mation of 19-OH MDZ and its short-term stability in the
absence of FBS and 1a,25-(OH)2-D3, the modified Caco-2
monolayer appears well suited for the study of first-pass
metabolism kinetics. Although it is possible to adapt cell
culture medium with the proper nutrients and growth factors
such that cells can be cultured in the absence of serum, it is
generally believed that serum is essential to both membrane
stability and cell function. The stability of the 19-OH MDZ
formation rate is likely to reflect the stability of the CYP3A
enzyme level in the modified Caco-2 monolayer. We previ-
ously estimated by immunoquantification (Schmiedlin-Ren
et al., 1997) that the 1a,25-(OH)2-D3-treated monolayer con-
tains 20.6 pmol CYP3A/mg homogenate protein. This trans-
lates into an estimated 25 pmol of CYP3A per 4.2 cm2 cul-
tured insert. However, our current Western blot result
suggests the CYP3A content is much lower than initially
estimated (3.7 pmol/4.2 cm2 cultured insert). One reason for
this disparity in CYP3A4 content may be the different West-
ern blot development techniques and standards used for the
two studies. Despite this apparent difference in CYP3A con-
tent, the initial (0 –2 h) 19-OH MDZ formation rate was
comparable after apical MDZ incubations in both studies,
suggesting that active CYP3A4 protein content was similar
and stable in both systems.
We can also approximate CYP3A4 content in our modified
Caco-2 cell monolayer on the basis of the expected and ob-
served turnover number for MDZ 19-hydroxylation. Turnover
1142 Fisher et al.
numbers of 1.6 pmol/pmol CYP3A4/min (Gorski et al., 1994)
and 5.6 pmol/pmol CYP3A4/min (Gibbs et al., 1999) have
been reported for purified and cDNA-expressed enzymes,
respectively. Based on these values and our measured Vmax
value with Caco-2 cells, the expected content of CYP3A in the
cell monolayer would be 6.9 and 2.0 pmol per culture, respec-
tively, which are in close agreement with the immunoblot
results observed in this study (3.7 pmol/culture).
The mean first-pass ER obtained from the cell cultures,
14.5 6 3.1%, is at the low end of a range of values (13.6 –
58.6%) obtained from an in vivo study where intestinal MDZ
extraction to 19-OH MDZ was measured directly in liver
transplant patients (Paine et al., 1996). One reason for an
apparent in vitro—in vivo discrepancy may be a lower level of
expression of active CYP3A4 enzyme in the Caco-2 system,
compared with intestinal enterocytes (8.3 versus 30.6
pmol/mg microsomal protein for Caco-2 cells from the present
study and human duodenal mucosa, respectively; Paine et
al., 1997), as discussed above. Although total CYP3A content
in homogenates prepared from the Caco-2 monolayer and
intestinal mucosal scrapings were relatively similar (7.7 ver-
sus 9.2 pmol/mg homogenate protein), total homogenate pro-
tein from the human intestine includes proteins from lumi-
nal contents and villous cells other than enterocytes whereas
Caco-2 homogenate protein is from a single cell type. This
difference in homogenate protein composition could yield a
lower CYP3A content in human mucosal homogenate prepa-
rations relative to preparations from enterocytes alone.
Thus, the somewhat low first-pass MDZ extraction efficiency
of the 1a,25-(OH)2-D3-modified Caco-2 monolayer, compared
with the in vivo ER, may actually be in line with the observed
difference in CYP3A expression. If this proves to be true,
scaling factors may be developed to increase the in vitro to in
vivo predictive power of the Caco-2 system.
In summary, our results describe for the first time the use
of a Caco-2 cell monolayer as an in vitro model to simulate
intestinal first-pass metabolic extraction kinetics for a
CYP3A-catalyzed reaction. Using MDZ 1-hydroxylation as an
enzyme-selective probe, we obtained an average first-pass
extraction ratio of ;15% for a subsaturating apically applied
MDZ dose. This level of metabolic activity is clearly sufficient
to support investigations of the effect of extracellular protein
(i.e., basolateral plasma proteins) as well as the effects of
enzyme modulators (i.e., inhibitors and activators) on the
first-pass elimination process. The model system should also
be suitable for examination of the first-pass metabolic extrac-
tion of other moderate to high intrinsic clearance CYP3A
substrates. Further, because the 1a,25-(OH)2-D3-treated
Caco-2 cells also express P-gp (Schmiedlin-Ren et al., 1997),
they should also be appropriate for studies investigating the
dynamic interplay between metabolism, active efflux, and
transcellular permeability.
References
Artursson P (1990) Epithelial transport of drugs in cell culture. I: A model for
studying the passive diffusion of drugs over intestinal absorptive (Caco-2) cells.
J Pharm Sci 79:476 – 482.
Boulenc X (1997) Intestinal cell models: Their use in evaluating the metabolism and
absorption of xenobiotics. S.T.P. Pharma Sciences 7:259 –269.
Chirayath MV, Gajdzik L, Hulla W, Graf J, Cross HS and Peterlik M (1998) Vitamin
D increases tight-junction conductance and paracellular Ca21 transport in Caco-2
cell cultures. Am J Physiol—Gastroint Liver Physiol 36:G389 –G396.
Vol. 289
Crespi CL, Penman BW and Hu M (1996) Development of Caco-2 cells expressing
high levels of cDNA-derived cytochrome P4503A4. Pharm Res 13:1635–1641.
De Waziers I, Cugnenc PH, Yang CS, Leroux JP and Beaune PH (1990) Cytochrome
P-450 isoenzymes, epoxide hydrolase and glutathione transferases in rat and
human hepatic and extrahepatic tissues. J Pharmacol Exp Ther 253:387–394.
Fromm MF, Busse D, Kroemer HK and Eichelbaum M (1996) Differential induction
of prehepatic and hepatic metabolism of verapamil by rifampin. Hepatology 24:
796 – 801.
Gan LL, Modeley MA, Khosla B, Augustijns PF, Bradshaw TP, Hendren RW and
Thakker DR (1996) CYP3A-like cytochrome P450-mediated metabolism and po-
larized efflux of cyclosporine A in Caco-2 cells. Drug Metab Dispos 24:344 –349.
Gibbs MA, Thummel KE, Shen DD and Knuze KL (1999) Inhibition of CYP3A in
human intestinal and liver microsomes: Comparison of Ki values and impact of
CYP3A5 expression. Drug Metab Dispos 27:180 –187.
Ghosal A, Satoh H, Thomas PE, Bush E and Moore D (1996) Inhibition and kinetics
of cytochrome P4503A activity in microsomes from rat, human, and cDNA-
expressed human cytochrome P450. Drug Metab Dispos 24:940 –947.
Gomez DY, Wacher VJ, Tomlanovich SJ, Hebert MF and Benet LZ (1995) The effects
of ketoconazole on the intestinal metabolism and bioavailability of cyclosporine.
Clin Pharmacol Ther 58:15–19.
Gorski JC, Hall SD, Jones DR, VandenBranden M and Wrighton SA (1994) Regios-
elective biotransformation of midazolam by members of the human cytochrome
P450 3A (CYP3A) subfamily. Biochem Pharmacol 47:1643–1653.
Gres M, Julian B, Bourrie M, Meunier V, Roques C, Berger M, Boulenc X, Berger Y
and Fabre G (1998) Correlation between oral drug absorption in humans, and
Downloaded from jpet.aspetjournals.org at ASPET Journals on December 22, 2014
apparent drug permeability in TC-7 cells, a human epithelial intestinal cell line:
Comparison with the parental Caco-2 cell line. Pharm Res 15:726 –33.
Guengerich FP (1995) Human cytochrome P450 enzymes, in Cytochrome P450:
Structure, Mechanism, and Biochemistry, 2nd ed., (Ortiz de Montellano PR ed) pp
473–535, Plenum Press, New York.
Heizmann P and Ziegler WH (1981) Excretion and metabolism of 14C-midazolam in
humans following oral dosing. Arzneim-Forsch 13:2220 –2223.
Hunter J, Jepson MA, Tsuruo T, Simmons NL and Hirst BH (1993) Functional
expression of P-glycoprotein in apical membranes of human intestinal Caco-2 cells.
Kinetics of vinblastine secretion and interaction with modulators. J Biol Chem
268:14991–14997.
Kolars JC, Lown KS, Schmiedlin-Ren P, Ghosh M, Fang C, Wrighton SA, Merion RM
and Watkins PB (1994) CYP3A gene expression in human gut epithelium. Phar-
macogenetics 4:247–259.
Lown KS, Mayo RR, Leichtman AB, Hsiao H, Turgeon K, Schmiedlin-Ren P, Brown
MB, Guo W, Rossi SJ, Benet LZ and Watkins PB (1997) Role of intestinal P-
glycoprotein (mdr1) in interpatient variation in the oral bioavailability of cyclo-
sporine. Clin Pharmacol Ther 62:248 –260.
Lowry OH, Rosebrough NJ, Farr AL and Randall RJ (1951) Protein measurement
with the Folin phenol reagent. J Biol Chem 193:265–275.
Mandema JW, Tuk B, van Steveninck AL, Breimer DD, Cohen AF and Danhof M
(1992) Pharmacokinetic-pharmacodynamic modeling of the central nervous sys-
tem effects of midazolam and its main metabolite a-hydroxymidazolam in healthy
volunteers. Clin Pharmacol Ther 51:715–728.
Meunier V, Bourrié M, Berger Y and Fabre G (1995) The human intestinal epithelial
cell line Caco-2; Pharmacological and pharmacokinetic applications. Cell Biol
Toxicol 11:187–194.
Paine MF, Khalighi M, Fisher JM, Shen DD, Kunze KL, Marsh CL, Perkins JD and
Thummel KE (1997) Characterization of interintestinal and intraintestinal vari-
ations in human CYP3A-dependent metabolism. J Pharmacol Exp Ther 283:1552–
1562.
Paine MF, Shen DD, Kunze KL, Perkins JD, Marsh CL, McVicar JP, Barr DM,
Gillies BS and Thummel KE (1996) First-pass metabolism of midazolam by the
human intestine. Clin Pharmacol Ther 60:14 –24.
Raeissi SD, Guo Z, Dobson GL, Artursson P and Hidalgo IJ (1997) Comparison of
CYP3A activities in a subclone of Caco-2 cells (TC7) and human intestine. Pharm
Res 14:1019 –1025.
Schmiedlin-Ren P, Thummel KE, Fisher JM, Paine MF, Lown KS and Watkins PB
(1997) Expression of enzymatically active CYP3A4 by Caco-2 cells grown on
extracellular matrix-coated permeable supports in the presence of 1a,25-(OH)2-D3.
Mol Pharmacol 51:741–754.
Smith MT, Eadie MJ and Brophy TO (1981) The pharmacokinetics of midazolam in
man. Eur J Clin Pharmacol 19:271–278.
Thummel KE, Kunze KL and Shen DD (1997) Enzyme-catalyzed processes of first-
pass hepatic and intestinal drug extraction. Adv Drug Deliv Rev 27:99 –127.
Wacher VJ, Wu C and Benet LZ (1995) Overlapping substrate specificities and tissue
distribution of cytochrome P450 3A and P-glycoprotein: Implications for drug
delivery and activity in cancer chemotherapy. Mol Carcinog 13:129 –134.
Watkins PB, Wrighton SA, Schuetz EG, Molowa DT and Guzelian PA (1987) Iden-
tification of glucocorticoid-inducible cytochromes P-450 in the intestinal mucosa of
rats and man. J Clin Invest 80:1029 –1036.
Wrighton SA, Brian WR, Sari MA, Iwasaki M and Guengerich FP (1990) Studies on
the expression and metabolic capabilities of human liver. Mol Pharmacol 38:207–
213.
Send reprint requests to: Dr. Kenneth E. Thummel, Ph.D., Department of
Pharmaceutics, Box 357610, University of Washington, Seattle, WA 98195-
7610. E-mail: thummel@u.washington.edu