CYP3A4, CYP3A5, and MDR1 in Human Small and Large Intestinal Cell Llines Suitable for Drug Transport Studies

CYP3A4, CYP3A5, and MDR1 in Human Small and Large Intestinal Cell Lines Suitable for Drug Transport Studies HELENA A. ENGMAN,1 HANS LENNERNAÈS,1 JAN TAIPALENSUU,1 CHARLOTTA OTTER,2 BRITH LEIDVIK,2 PER ARTURSSON1 1

Department of Pharmacy, Uppsala University, Box 580, SE-751 23 Uppsala, Sweden

2

Department of Molecular Biology, AstraZeneca R&D, SE- 431 83 MoÈlndal, Sweden

Received 30 December 2000; revised 18 April 2001; accepted 1 June 2001

ABSTRACT: The aim of this study was to ®nd a cell culture model of the intestinal epithelium for use in studies of CYP3A4-mediated ®rst-pass metabolism of drugs and also for studies of the interplay between CYP3A4 metabolism and P-glycoprotein ef¯ux. For this purpose, the expression of CYP3A4, CYP3A5, and MDR1 mRNA was studied in three cell lines of the normal human intestinal epithelium and three transformed cell lines of colonic (Caco-2) origin. Surprisingly, only transformed cell lines were induced by 1a,25-dihydroxy vitamin D3 (D3) to express high amounts of CYP3A4. In contrast to the original ®ndings for this model, the monolayer integrity was maintained during D3 treatment. Levels of CYP3A mRNA expression in Caco-2 and TC7 cells differed dramatically. The highest levels of CYP3A4 and lowest levels of CYP3A5 mRNA expression were observed in D3 treated Caco-2 cells of high passage numbers, resulting in a CYP3A4/3A5 expression ratio greater than fourfold higher than that seen in TC7 cells. Functional studies, using the CYP3A probe testosterone, showed that CYP3A activity was completely absent only in uninduced Caco-2 cells. After D3 induction, high levels of the metabolite were produced in both cell lines (149.4  12.3 and 86.5  3.8 pmol 6b-OH testosterone/min/mg cellular protein from 75 mM testosterone in Caco-2 and TC7 cells, respectively). The maximum velocity (Vmax) and the apparent Michaelis constant (Km) for the 6b-hydroxylation of testosterone by CYP3A4 in intact Caco-2 monolayers were similar to those obtained from human intestinal microsomes. Only minor changes in P-glycoprotein activity were observed when the metabolically stable P-glycoprotein substrate celiprolol was used. In conclusion, these results show that the features of the generally available Caco-2 cell line from American Type Culture Collection make it suitable for studies of CYP3A4-mediated ®rst-pass metabolism and also for studies of the interplay between CYP3A4 and drug ef¯ux mechanisms. ß 2001 Wiley-Liss, Inc. and the American Pharmaceutical Association J Pharm Sci 90:1736±1751, 2001

Keywords: CYP3A; P-glycoprotein; Caco-2; intestinal epithelium; ®rst pass extraction; permeability; bioavailability

INTRODUCTION The cytochrome P450 (CYP) enzyme CYP3A4, which is responsible for the oxidative metabolism of > 60% of all clinically used drugs,1 is found Correspondence to: Per Artursson (Telephone: 46 18 471 44 71; Fax: 46 18 471 42 23; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 90, 1736±1751 (2001) ß 2001 Wiley-Liss, Inc. and the American Pharmaceutical Association

1736

predominantly in the small intestine and the liver.2 Recent studies suggest that drugs such as cyclosporin can be metabolized to a comparable degree in the mucosal lining of the intestinal epithelium and in the liver.3 It has been hypothesized that the gene product of the multidrug resistance gene MDR1, P-glycoprotein (Pgp), and CYP3A4 can act synergistically in the enterocytes to increase the extent of the intestinal CYP3A4

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

metabolism of these drugs.4 According to these models, Pgp-mediated ef¯ux of the drugs increases their intestinal residence time and may also prevent CYP3A4 product inhibition by the removal of primary metabolites. This hypothesis mainly originates from sources citing overlapping substrate speci®cities5 and similarities in gene regulation and tissue distribution.5,6 The lack of direct evidence supporting the proposed cooperation is partly due to the lack of potent inhibitors speci®c to Pgp or CYP3A4 and to dif®culties in separating the relative contributions of the two proteins in vivo. Therefore, there is a need for reproducible in vitro models allowing detailed study of the Pgp/CYP3A4 interactions. The Caco-2 cell line7 expresses functional Pgp and multidrug resistance associated protein-2 (MRP2)/canalicular multispeci®c organic anion transporter (cMOAT) at levels that allow reproducible ef¯ux studies in cell culture.8,9 Unfortunately, this cell line does not express suf®cient amounts of functional CYP3A4 for study under standard culture conditions.10 CYP3A5 activity is found in a Caco-2 cell clone named TC7,11 and we have previously used this cell line to study the interplay between CYP3A metabolism and MDR1-mediated drug ef¯ux.12 However, while human CYP3A comprises an average of 70% of the total intestinal CYP content,2 CYP3A5 mRNA is polymorphically expressed in only 20±25% of humans.13 Further, two recent reports suggest that CYP3A5 may play a minor role in human small intestinal metabolism.14,15 Therefore, a cell culture model that predominantly expresses CYP3A4 would be more valuable. Various approaches have been used to develop cell culture models of intestinal epithelial cells displaying CYP3A4 activity. In one approach, Caco-2 cells were transfected with cDNA encoding for CYP3A4. However, because an extrachromosomal vector was used, the enzyme activity declined rapidly with increasing passage numbers.16 Incorporation of oxidoreductase into the vector did not stabilize the CYP3A4 expression level.17 Another, probably more physiologically relevant, approach was to elevate the CYP3A4 activity in Caco-2 cells by enzyme induction, using the recently described CYP3A4 inducer 1a,25-dihydroxyvitamin D3 (D3). The expression of CYP3A4 was markedly increased, with almost no in¯uence on CYP3A5.18 At the onset of this study, it was not known whether this approach could be used to induce CYP3A4 activity also in intestinal epithelial cell lines derived from healthy human intestine.

1737

The aim of this study was to develop a cell culture model that would allow investigation of the relative contribution of and interplay between CYP3A, preferably CYP3A4, and MDR1 in studies of intestinal drug uptake and gut wall metabolism. For this purpose, six cell lines of human small intestinal and colonic origin were studied with regard to CYP3A4, CYP3A5, and MDR1 mRNA expression. Cell lines forming monolayers with constitutive expression of CYP3A and MDR1 were induced with D3. One cell line that displayed extensive CYP3A4 activity and a high CYP3A4/ CYP3A5 ratio after induction was identi®ed and characterized.

EXPERIMENTAL SECTION Materials D3 was purchased from Solvay Duphar (Weesp, The Netherlands). Testosterone, 6b-hydroxytestosterone, clotrimazole, 9-cis-retinoic acid, and dexamethasone were obtained from Sigma Chemical Company (St. Louis, MO). Androstenedione and 2b-hydroxytestosterone were kindly provided by AstraZeneca R&D (MoÈlndal, Sweden). Recombinant human growth hormone was a generous gift from Dr. Jonas Fransson (Pharmacia & Upjohn, Stockholm, Sweden). [14C]celiprolol (37.3 mCi mgÿ1) was provided by Rhone-Poulenc Rorer Pharmaceutical, Inc. (Collegeville, PA). The radiochemical and chemical purity of [14C]celiprolol was 95%. [14C]mannitol (52 mCi mmolÿ1) was purchased from New England Nuclear (Boston, MA). Stock solutions of clotrimazole, 9-cisretinoic acid, and dexamethasone were prepared at 0.01 M in methanol and stored at ÿ208C. Stock solutions of D3 were prepared at 0.1 M in ethanol and stored at ÿ808C. Chemicals used for high performance liquid chromatography (HPLC) and immunoblot analysis were of analytical or HPLC grade. Cells and Culture Conditions Three cell lines (BN, LG, and WT) originating from biopsies taken from healthy human duodenum were generous gifts from Dr Gerald Pang.19 Three cell lines originating from the human colorectal carcinoma Caco-2 were also investigated. The parental population of Caco-2 cells (Caco-2)7 and the morphologically well-differentiated Caco-2 clone, BBe-1,20 were obtained from

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1738

ENGMAN ET AL.

American Type Culture Collection (ATCC; Rockville, MD), while the CYP3A-expressing Caco-2 clone TC7 was a generous gift from Dr. Alain Zweibaum.21 In the ®rst part of the study, in which the constitutive expression of CYP3A4, CYP3A5, and MDR1 mRNA was studied, all cell lines were grown to con¯uence with or without a laminin-containing extracellular matrix (ECM; Matrigel, Collaborative Biomedical Products, Bedford, MA) in tissue culture ¯asks. The duodenal cell lines, BN, LG, and WT, were cultured according to Pang et al.19 and used at passage numbers 17±30. The Caco-2 cells were cultured as described in detail previously.22 Culture conditions for TC7 and BBe-1 were modi®ed according to CarrieÁre et al.21 and Peterson and Mooseker,20 respectively. Caco-2 cells were used at passage intervals 20±30 and 92±105, while TC7 and BBe-1 were used at passage intervals 22±30 and 48±58, respectively. In the second part of the study, the effects of D3 on CYP3A mRNA expression were investigated in the BN, Caco-2, and TC7 cell lines. The effects of D3 on induction of CYP3A protein expression and function were also studied in Caco-2 and TC7 cells. Finally, the effect of other potential inducers on functional CYP3A4 activity was investigated in Caco-2 cells. For these experiments, the cell lines were seeded onto Transwell cell culture inserts (®lter diameter 24 mm, mean pore size 0.45 mm; Costar, The Netherlands) coated with Matrigel 15 mg/cm2 (Collaborative Biomedical Products). At con¯uency, that is, 10 days after seeding, the culture medium was supplemented with D3 0.5 mM,18 or, in selected experiments, with clotrimazole (10 mM), 9-cis-retinoic acid (10 mM), dexamethasone (10 mM), recombinant human growth hormone (4.5 nM), or medium only. The cell culture medium was replaced with fresh medium  inducer every second day and always 24 h before an experiment. Depending on the experimental design, the monolayers were used after 4 days (activity assessment, see below),23 5 days (mRNA analysis),18 or 14 days (protein analysis,18 activity assessment,18 integrity- and permeability-assessment, see below) in the presence or absence of inducers. Monolayer Integrity Measurement of transepithelial electrical resistance (TEER) was used to determine the cell monolayer integrity.24 The TEER in monolayers grown on ECM-coated permeable supports was

measured at 378C using an Endohm tissue resistance measurement chamber connected to an Evohm resistance meter (World Precision Instruments, Sarasota, FL). The cell culture medium was replaced by Hank's balanced salt solution (HBSS) at pH 7.4 preheated to 378C. Each TEER value was corrected for the background resistance of an ECM-coated cell culture insert without cells. The paracellular marker [14C]mannitol was used as an additional integrity marker. Preparation for Transmission Electron Microscopy (TEM) Glutaraldehyde-®xed cell monolayers were immersed consecutively in 1% osmiumtetroxid and 1% uranyl acetate, dehydrated in a graded series of ethanol solutions, and embedded in Epon. Thin sections contrasted with uranyl acetate and lead citrate were examined in a Philips 420 electron microscope (Philips, Eindhoven, The Netherlands). Isolation of Total RNA Cells grown to con¯uence in tissue culture ¯asks were washed twice in phosphate-buffered saline (PBS) and lysed in a guanidium isothiocyanate buffer. Lysates were ¯ash frozen in liquid nitrogen and stored at ÿ808C pending RNA isolation. After homogenization, total RNA was isolated using the Qiagen RNeasy Midi kit (Primary source; KEBO Lab, SpaÊnga, Sweden). Isolated RNA was stored at ÿ808C pending reverse transcriptase polymerase chain reaction (RT-PCR) analysis. The same procedure was used to isolate total RNA from cells grown in induction medium on permeable ®lter inserts for 5 days. RT-PCR Analysis RT-PCR analysis was performed using a reverse transcription kit (Promega Corp., MA) and template-speci®c oligonucleotides obtained from T-AG-Copenhagen (Copenhagen, Denmark). Brie¯y, 1 mg total RNA and 30 pmol of the antisense template-speci®c oligonucleotide were exposed to an initial denaturation step before reverse transcription to cDNA by 15 U/mg avian myeloblastosis virus (AMV) reverse transcriptase. PCRs were performed in 50 mL ®nal volumes, consisting of 5 mL of the ®rst-strand cDNA reaction mixture, reverse transcription buffer, 2 mM MgCl2, sense and antisense oligonucleotides, and 2.5 U Taq

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

polymerase. The reaction mixtures were denatured and 30 PCR cycles were performed: 15 s at 948C, followed by 15 s at 558C (for CYP3A4 and MDR1) or at 508C (for CYP3A5), and 45 s at 728C. A ®nal 728C extension step, for 10 min, was included at the end of all the reactions. The PCR products were analyzed by separation in 1.35% agarose containing ethidium bromide for visualization, and the sizes of the PCR products were estimated from the migration of DNA size markers, which were run concurrently. Real-Time Quantitative PCR Detection In order to eliminate ampli®cation of contaminating chromosomal DNA in the real-time quanti®cation, the isolated total RNA was treated with DNase using RQ1 RNase-Free DNase from Promega, following the suppliers' instructions for use, prior to RT-PCR. RT reactions were performed using the SuperScriptTM Preampli®cation System for First Strand cDNA Synthesis (Life Technologies, Gibco BRL Products). Quantitative real-time PCR assay of transcripts was carried out with the use of genespeci®c double ¯uorescence labeled probes and the TaqMan1 Universal PCR Master Mix in a 7700 Sequence Detector (PE Applied Biosystems, Norwalk, CT). 6-Carboxy ¯uorescein (FAM) was used as the 50 ¯uorescent reporter, while tetramethylrhodamine (TAMRA) was added to the 30 end as a quencher. The following primers and probe sequences were used: CYP3A4 forward primer, 50 -CATTCCTCATCCCAATTCTTGAAGT-30 ; CYP3A4 reverse primer, 50 -CCACTCGGTGCTTTTGTGTATCT-30 ; CYP3A4 probe, 50 -FAMCGAGGCGACTTTCTTTCATCCTTTTTACAGATTTTC-TAMRA-30 ; CYP3A5 forward primer, 50 GCTCGCAGCCCAGTCAATA-30 ; CYP3A5 reverse primer, 50 -AGGTGGTGCCTTATTGGGC-30 ; CYP3A5 probe, 50 -FAM TGAAACCACCAGCAGTGTTCTTTCCTTCAC-TAMRA-30 . The primers and probes were designed to span exon junctions in the fully processed message in order to prevent reporting of ampli®cation of any contaminating genomic DNA. All primers and probes were purchased from PE Applied Biosystems. Standard curves were constructed with the use of serial (10-fold) dilutions ranging from 1.0 fg to 10 pg of an accurately determined plasmid containing the cDNA of interest as template. The use of the endogenous reference permitted normalization for variation in RT ef®ciency in the cDNA reactions. To ensure that the ef®ciency of

1739

the formation of the CYP3A4 and CYP3A5 cDNAs for the standard curves were the same as for the CYP3A4 and CYP3A5 targets within the samples, standard curves with serial dilutions of the samples were made. Because the ampli®cation of each target template exhibited an equivalent rate of ¯uorescent emission intensity change per amount of DNA target, plasmids containing template standard curves were selected to determine the arbitrary units of CYP3A4 and CYP3A5. As the ef®ciency in each PCR ampli®cation was approximately equal and close to 1, no standard curves were needed for the relative quanti®cation of induction. Each sample was run in triplicate two to four times unless otherwise stated, and the DRn (the ratio of the amount of reporter dye emission to the amount of quenching dye emission) and average threshold cycle (CT) values were calculated from each reaction in triplicate. Differences in the amount of RNA added to each reaction were normalized against human acidic ribosomal phosphoprotein (huPO). The relative fold of induction was calculated using the comparative CT method (DDCT)25 using huPO, described in TaqMan1 Human Endogenous Control Plate (P/N4309199; Applied Biosystem), as the endogenous control reference. The untreated samples were designated the calibrator and assigned the value of 1x for each of the CYP3A targets. The quantity of each CYP3A mRNA in each D3 treated sample was given relative to the calibrator sample. A detailed procedure for relative quanti®cation is described in ABI PRISMTM 7700 Sequence Detection Bulletin.26 Data were analyzed using the Sequence Detector V1.6 program (Perkin-Elmer). Immunoblotting of CYP3A4 Caco-2 and TC7 cell monolayers were cultured for 5 or 10 days to achieve different stages of differentiation. D3 was added as described above, and levels were maintained for 7 or 14 days. Control monolayers were cultured under the same conditions without D3. After a total culture time of 12, 17, 19, or 24 days, total cell lysates were generated and subjected to immunoblot analysis. The cell monolayers were washed twice in ice-cold PBS, cut out from the ®lter inserts, transferred to ice-cold lysis buffer [PBS without Ca2‡/Mg2‡, but containing 1% Triton1 X-100 and a protease inhibitor cocktail (CompleteTM; Boehringer Mannheim, Germany)], and then incubated on ice for 30 min. Lysates were generated

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1740

ENGMAN ET AL.

by sonication. The concentration of protein in the lysates was determined using the BCA Protein Assay (Pierce, Rockford, IL). Lysates were stored at ÿ848C pending analysis. Twenty to thirty micrograms of the total cell lysate was resuspended in a sodium dodecyl sulfate (SDS) reducing buffer and heated for 5 min at 958C. Samples underwent electrophoresis on a 12% SDS-polyacrylamide gel and were transferred to nitrocellulose membranes. The membranes were sequentially incubated with a monoclonal antibody to human CYP3A4 (Xenotech LLC, Cambridge, KS) and an alkaline phosphatase conjugated anti-rabbit IgG (Sigma Chemical Co.). The immunoblots were then visualized using the CDP-StarTM Western Blot Chemiluminescence Reagent (NENTM Life Science Products, Inc.). Assessment of CYP3A Activity The level of catalytic activity in ®lter-grown Caco-2 and TC7 monolayers was determined after 4 or 14 days' exposure to D3 (0.5 mM) using 6bhydroxylation of testosterone as a marker for CYP3A activity.27 Uninduced cell monolayers were used as controls. The cell monolayers from Caco-2 and TC7 were washed twice in prewarmed HBSS, pH 7.4. Thereafter, 2 mL HBSS containing 100 mM testosterone was added to both the apical and basolateral sides. Initial metabolism pro®les were performed with donor solutions containing 0±200 mM testosterone. The monolayers were incubated for 2 h at 378C. At speci®c time points, samples were collected from the apical and basolateral sides. Samples were frozen immediately and stored at ÿ208C pending HPLC analysis of 6b-hydroxytestosterone. The cell monolayers were washed twice in ice-cold PBS and the ®lters were cut and frozen pending analysis of intracellular 6b-hydroxytestosterone. The frozen monolayers were immersed in methanol, vortexed, and centrifuged at 3000 rpm for 13 min. A fraction (1 mL) of the supernatant was transferred to another tube, evaporated to dryness, and after reconstitution in mobile phase, injected onto the HPLC column. The intracellular 6b-hydroxytestosterone concentrations were calculated using a previously reported cell height of 17± 30 mm,28,29 a ®lter radius of 12.25 mm, and the fact that Caco-2 cell monolayers form tight monolayers of adjacent cells with a narrow paracellular space. The Vmax and apparent Km values for the 6bhydroxylation of testosterone were obtained by

®tting the concentration dependent activity data to the Michaelis-Menten equation using Prism1 (Graph Pad Software Inc., San Diego, CA). Effect of D3 on Pgp Activity In order to evaluate the effect of D3 on Pgp activity, the apparent permeability coef®cient (Papp) of the Pgp substrate [14C]celiprolol (75 mM) was determined in both the apical to basolateral (a-b) and basolateral to apical (b-a) directions in control and D3-treated cell monolayers. The ef¯ux activity was expressed as the relationship b-a/a-b. HPLC Analysis The metabolite 6b-OH-testosterone was analyzed by reversed phase HPLC with UV detection at 240 nm, using a method modi®ed from Kostrubsky et al.23 The chromatographic system consisted of a Perkin-Elmer Isocratic LC Pump 250, an autoinjector (Perkin-Elmer Advanced LC Sample Processor ISS-200), a Hichrom P5ODS3-IOC5 precolumn, a column (Beckman Ultrasphere ODS, 5 mm, 4.6  250 mm), an ultraviolet light detector (Spectra-Physics UV 100), and the CSW Chromatography Station for Windows. 6b-Hydroxytestosterone was eluted with methanol/acetonitrile/ water (45/15/40 v/v) at 0.8 mL/min and quanti®ed by comparing the absorbance to that of a standard curve of 6b-hydroxytestosterone prepared in HBSS. The limit of quanti®cation in this system was 0.05 mM. Using this HPLC system, two other metabolites (androstenedione and 2b-hydroxytestosterone) were identi®ed using reference substances. Statistics Results are expressed as means  one standard deviation (SD). Statistical comparison of mean values was made using one-way analysis of variance (ANOVA) or the unpaired, two-tailed student's t test. p < 0.05 was considered statistically signi®cant.

RESULTS CYP3A4, CYP3A5, and MDR1 mRNA in Human Small and Large Intestinal Epithelial Cell Lines The mRNA expression of CYP3A4, CYP3A5, and MDR1 was studied in the presence and absence of a laminin-containing ECM.30 RT-PCR analysis

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

showed that CYP3A4 was expressed weakly in two of the three human duodenal cell lines (BN and LG) (Figure 1). In contrast, CYP3A4 was not expressed in the three variants of the Caco-2 cell line. However, CYP3A5 was expressed in two of the three investigated Caco-2 cell lines: the original Caco-2 population from ATCC (Caco-2) and the clone TC7 (Figure 1). The strong expression of CYP3A5 in TC7 was observed independently of the presence of an ECM. All six cell lines of the human intestinal tract strongly expressed MDR1, with or without an ECM (Figure 1). Because BBe-1 and the duodenal cell line WT did not show any CYP3A expression, they were not investigated further. Monolayer Integrity of Cell Lines Expressing CYP3A The two duodenal cell lines expressing CYP3A4 and the two colonic Caco-2 cell lines expressing CYP3A5 were investigated with regard to mono-

Figure 1. Expression of CYP3A4, CYP3A5, and MDR1 mRNA in human epithelial cell lines derived from duodenal and colonic tissues. The cell lines were cultured in uncoated (ÿM) or ECM-coated (‡M) culture ¯asks. At con¯uency, cells were lysed and total RNA was prepared and subjected to RT-PCR, as described in the Experimental Section. The MDR1 gene was strongly expressed in all cell lines independently of the presence of an ECM. CYP3A4 mRNA was weakly expressed in two of the three duodenal cell lines (BN and LG), whereas CYP3A5 mRNA was expressed in two transformed cell lines (Caco-2 and TC7).

1741

layer formation on ECM-coated permeable supports. The two duodenal cell lines BN and LG formed tight monolayers after 14 days with TEER values of 2352  268 and 268  240 O  cm2, respectively. TEM showed that the BN cells formed tight monolayers of cuboid, partially differentiated epithelial cells with a few small patches of bilayers. In contrast, LG cells formed continuous bi- to multilayers, making this cell line unsuitable for transport studies (data not shown). Caco-2 and TC7 cells formed tight monolayers with TEER values of 181  16 and 115  2 O  cm2, respectively. Effect of D3 on CYP3A and MDR1 mRNA Expression The effects of the addition of D3 were studied in the three monolayer-forming cell lines that expressed either CYP3A4 or CYP3A5 mRNA. Addition of D3 to the duodenal cell line BN did not induce detectable expression of CYP3A4 mRNA, and this cell line was therefore excluded from further studies. However, CYP3A4 mRNA was strongly expressed in both Caco-2 and TC7 monolayers after 5 days' exposure to D3 (0.5 mM), as was MDR1 mRNA (Figure 2). Because preliminary experiments have shown that different subclones, populations, and passage numbers31 of Caco-2 cells express CYP3A5 to variable degrees, we compared the CYP3A4 and CYP3A5 mRNA expression after 5 days' treatment with D3 in the Caco-2 and TC7 cell lines, using real-time quantitative PCR (Table 1). The strongest induction of CYP3A4 mRNA was observed in the Caco-2 cell line at the high passage numbers (92±105) used throughout this study, where the expression increased from a background level of 0.14  0.07 to 23.58  5.63 units (n ˆ 3). The corresponding increase of CYP3A4 mRNA in Caco-2 cells of a lower passage number (20±30) was smaller, while in TC7 cells it ranged from a level just above the detection limit (0.01  0.00 units) to 9.71  2.62 units. The baseline expression of CYP3A5 mRNA was higher than that of CYP3A4 in both cell lines (Table 1). The increase in CYP3A5 on addition of D3 was less pronounced than that of CYP3A4 in Caco-2 (high passage number). Thus, CYP3A5 mRNA increased from 0.43  0.15 to 3.40 units (n ˆ 1), giving a ratio of the mRNA expression of CYP3A4 to that of CYP3A5 of 6.9. In Caco-2 cells of the lower passage numbers, a comparable CYP3A4/CYP3A5 expression ratio was obtained,

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1742

ENGMAN ET AL.

that is, 1.6. These results indicate that Caco-2 cells are a better model than TC7 cells for studies of CYP3A4 metabolism in the intestine. On the other hand, uninduced TC7 cells seem to be the better model for studies of CYP3A5, because virtually no CYP3A4 mRNA is expressed in these cells. CYP3A4 Protein Expression Western blot analysis indicated strong CYP3A4 protein expression in the two D3 induced Caco-2 cell lines (Figure 3, data not shown). In agreement with the original report, the strongest CYP3A4 protein expression was observed after a 2-week induction period with D3.18 Therefore, this induction protocol was used in all experiments reported below. Our attempts to determine the amount of CYP3A5 protein expression were unsuccessful, because all investigated CYP3A5 antibodies cross-reacted with CYP3A4. Figure 2. Expression of CYP3A4 (lane 1), MDR1 (lane 2), and CYP3A5 (lane 3) mRNA in (a) Caco-2 and (b) TC7 cell lines in response to D3. The Caco-2 cell lines were cultured in ECM-coated Transwell ®lter inserts. A D3 treatment period of 5 days was initiated at con¯uency. Cells were lysed and total RNA was prepared and subjected to RT-PCR as described in the Experimental Section. CYP3A4 mRNA was strongly expressed in D3 treated Caco-2 and TC7 monolayers, as were CYP3A5 and MDR1 mRNA (n ˆ 2 for each determination).

that is, 9.2 (Table 1). However, less CYP3A4 and more CYP3A5 mRNA was expressed in TC7 cells, and the corresponding CYP3A4/CYP3A5 expression ratio was therefore much lower in TC7 cells,

Effect of D3 on Monolayer Integrity Previous studies have reported a 50% decrease in monolayer integrity after D3 treatment.32,33 In our experience, such a large decrease in barrier function is not reproducible. Because the dramatic decrease in barrier function may affect the transport properties and therefore bias the interpretation of transport studies in the cell monolayers, we compared the TEER values and the paracellular permeability to mannitol for control monolayers with those for D3 treated monolayers. In these experiments, there were no differences in TEER between D3 treated (181  16 O  cm2 ) and control (179  18 O  cm2) monolayers. Similarly,

Table 1. mRNA Expression of CYP3A Isoenzymes in Caco-2 and TC7 Cell Lines CYP3A4 mRNA Cell Line Caco-2 lowc Caco-2 highd TC7

CYP3A5 mRNA

Ratio 3A4/3A5

Controla

D3a

Fold Inductionb

Controla

D3a

Fold Inductionb

Control

D3

0.04  0.01 0.14  0.07 0.01  0.00

5.95  2.05 23.58  5.63 9.71  2.62

203 177 1170

0.47  0.15 0.43  0.15 1.55  0.49

0.65  0.16 3.40e 6.22  2.57

2 7 5

0.1 0.3 0.01

9.2 6.9 1.6

a Values are expressed as arbitrary units of CYP3A4 and CYP3A5 mRNA corrected for the expression of the endogenous control huP0 and ®tted to CYP3A standard curves as described in the Experimental Section. For all samples, total-RNA was prepared from cell lysates pooled from 4±6 Transwell ®lter inserts (24 mm diameter). Data represent mean  SD for two to four separate determinations. b The relative fold of induction was calculated using the comparative CT method described in the Experimental Section. c Caco-2 cells at low passage numbers (20±30). d Caco-2 cells at high passage numbers (92±105). e Only one determination could be made, because of low levels of RNA in the pooled sample.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

1743

Effect of D3 on CYP3A and MDR1 Activity

Figure 3. CYP3A4 protein expression in D3 treated Caco-2 cell monolayers. Immunoblotted total cell lysates from Caco-2 monolayers, induced for 7 days (D3 treated, lane 1; control, lane 2) or 14 days (D3 treated, lane 3; control, lane 4) with start 10 days after seeding. Lane 5, empty; lane 6, human liver microsomes. Corresponding immunoblots of TC7 monolayers generated similar expression patterns (data not shown).

the permeability of D3 treated monolayers to mannitol (1.4  0.2  10ÿ7 cm/s) was comparable to that of control monolayers (1.7  0.01  10ÿ7 cm/s), indicating that the integrity of the Caco-2 monolayers was unaffected. These results suggest that the polyclonal Caco-2 cell line is more robust to manipulations than some of its subclones, or that the culture and/or induction protocols used in this study are superior to those used previously.18

The 6b-hydroxylation of testosterone was used to investigate the CYP3A4 activity in Caco-2 and TC7 cell lines. No (Caco-2) or very low (TC7) 6bhydroxytestosterone activity was found in cells cultivated on an ECM under standard cell culture conditions (Figure 4). After D3 induction, a large increase in CYP3A4 activity was observed in both cell lines. The rate of 6b-hydroxytestosterone formation at a substrate concentration of 75 mM increased from below the level of detection to 149.4  12.3 pmol/min/mg cellular protein in Caco-2 monolayers (Figure 4a) and from 3.2 to 86.5  3.8 pmol/min/mg cellular protein in TC7 monolayers (Figure 4b). The maximum velocity (Vmax) and the apparent Michaelis constant (Km) obtained in Caco-2 were 332 pmol/min/mg cellular protein and 90 mM, respectively. The more scattered data from the TC7 incubations did not permit a precise determination of the kinetic parameters. The apparent Km value determined in Caco-2 cell monolayers is similar to the Km values previously determined in human liver and intestinal microsomes.10 We conclude that the absence of CYP3A4 activity in uninduced Caco-2 cells as compared to TC7 cells makes the Caco-2 cell line more suitable for studies of the CYP3A4

Figure 4. Concentration-dependent CYP3A metabolic activity in (a) Caco-2 and (b) TC7 control (&) and D3 treated (&) cell monolayers measured as 6b-hydroxylation of testosterone. Cell monolayers were incubated for 120 min in HBSS containing 0±200 mM testosterone. Samples from the apical and basolateral sides were pooled and analyzed by HPLC with ultraviolet light detection. Each value represents the mean  SD of three cultures. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1744

ENGMAN ET AL.

barrier to drug absorption across the intestinal epithelium, because the CYP3A4 gene expression can be turned on and off. In addition, Caco-2 cells showed less variability in CYP3A4 activity as compared to the TC7 cells. Therefore, in the following, all studies were performed in the Caco-2 cell line. The effect of D3 on MDR1 function was investigated, using [14C]celiprolol, a Pgp substrate that is not metabolized by CYP3A.34 The permeability of the control cells to celiprolol was 1.00  0.01  10ÿ6 cm/s in the absorptive direction and 5.16  0.13  10ÿ6 cm/s, or 5.2-fold larger, in the secretory direction (Figure 5a). After D3induction, the values were 0.84  0.04  10ÿ6 and 6.03  0.18  10ÿ6cm/s, respectively (Figure 5b). Thus, a small but signi®cant (16.9%) increase in celiprolol ef¯ux was observed after D3 treatment. TaqMan PCR analysis showed only a modest increase in MDR1 mRNA expression in response to D3 treatment (data not shown).

was beyond the level of detection in control cell monolayers (data not shown), while the intracellular concentration in D3 treated cells was 0.1  0.0 mM (Figure 6a). However, the extracellular concentrations of 6b-hydroxytestosterone were markedly higher than the intracellular concentration in D3 treated Caco-2 cells, that is, 1.9  0.1 mM and 0.8  0.1 mM on the apical and basolateral sides, respectively (Figure 6b,c). The observed differences indicate that ef¯ux mechanisms in both apical and basolateral membranes are active in the 6b-hydroxytestosterone distribution process. The involvement of ef¯ux proteins may be further supported by an apical-to-basolateral ratio of 2.4. 2b-hydroxytestosterone was detected intracellularly but not extracellularly in D3 treated (Figure 6a and 6b,c respectively) as well as in control Caco-2 cells (data not shown), suggesting that the formation of 2b-hydroxytestosterone was catalyzed by an enzyme other than CYP3A4.

Differential Intra- and Extracellular Localization of Testosterone Metabolites

Effect of Other CYP3A4 Inducers on Enzyme Activity

Incubation of Caco-2 monolayers with testosterone generated three metabolites: 6b- and 2b-hydroxytestosterone and androstenedione (Figure 6). Intracellular 6b-hydroxytestosterone

We also investigated whether potential CYP3A4 inducers, which at least partially use different nuclear receptors from those used by D3, enhanced the CYP3A4 activity in Caco-2 cells,

Figure 5. Effect of D3 treatment on Pgp activity in the Caco-2 cell line. The cumulative fraction of celiprolol transported across (a) control and (b) D3 treated Caco-2 monolayers was determined in the apical-to-basolateral (&) and the basolateral-toapical (&) directions, respectively. D3 treatment resulted in a small (17%) but signi®cant increase in Pgp activity. Each value represents the mean  SD of three cultures. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

1745

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1746

ENGMAN ET AL.

using 6b-hydroxylation of testosterone as a marker (Table 2). In contrast to the studies above, a shorter (4 day) induction scheme was used in order to minimize the potential toxicity of some of the inducers. Further, after 4 days, the D3 induced CYP3A4 activity was about 25% of the maximum, which allowed identi®cation of compounds that are stronger inducers than D3. Surprisingly, none of the selected compounds were as effective as D3 in inducing CYP3A4 activity, and several did not affect the expression of CYP3A4 at all (Table 2). In order to investigate potential toxicity, as well as possible additive or synergistic effects on the CYP3A4 activity, each of the compounds were added to the culture medium in combination with D3. Of these agents, only 9-cis-retinoic acid reduced the CYP3A4 activity, indicating a possible toxic effect. Coincubation of the various potential inducers with D3 did not reveal any additive or synergistic effects in the Caco-2 cells. Therefore, we tentatively conclude that this cell model is not suitable for induction studies, as the responses in vitro do not correspond to those in vivo.

DISCUSSION Schmiedlin-Ren et al.18 recently showed that D3 induces CYP3A4 expression in the Caco-2 cell line. This ®nding expanded the applicability of Caco-2 cells to include studies of the intestinal ®rst-pass metabolism33 and of potentially important interactions between CYP3A4 and Pgp activities.35 However, the D3 induced Caco-2 cell model has not been fully characterized and can, therefore, be improved. Firstly, this cell line originates from the colon, where the expression of CYP is lower than in the small intestine and may be differently regulated.36 Secondly, while

different Caco-2 populations express CYP3A5 to different degrees,12,18 functional CYP3A5 protein was not detected in human small intestine by Zhang et al.14 This suggests that, for most individuals, CYP3A5 plays a minor role in the intestinal ®rst-pass extraction of CYP3A substrates in vivo.14,15 Thirdly, Caco-2 monolayers have been reported to partially lose their integrity after D3 treatment, which may bias the interpretation of transport studies in D3 induced cells.33 Finally, some studies have indicated coregulation of CYP3A4 and the drug ef¯ux protein MDR1 in Caco-2 cells, which may obscure studies of the interplay between CYP3A4 and MDR1 in these cells.6 We therefore addressed these issues one by one in this study. In an attempt to obtain higher CYP3A4 expression, we performed our initial studies on three cell lines isolated from the human duodenum.19 These cell lines were recently reported to be the ®rst monolayer-forming intestinal epithelial cell lines to be obtained from healthy human intestine. Although two of these cell lines displayed weak constitutive expression of CYP3A4 mRNA and one of the cell lines, BN, formed con¯uent monolayers of partly differentiated intestinal epithelial cells, no CYP3A4 induction was observed after D3 treatment. A possible reason for the low CYP3A4 expression in these cells could be their crypt origin because, in the intestine, the expression of CYP3A4 follows a gradient along the crypt-villus axis, from low expression in the midcrypts to high expression on the villus tips.37 Instead, we focused on the two Caco-2 cell lines, Caco-2 (the parental cell line from ATCC) and the Caco-2 clone TC7. Both of these expressed large amounts of functional CYP3A4 after D3 induction. In addition, several other Caco-2 clones31 were markedly induced with regard to CYP3A4 mRNA (data not shown). The 6b-hydroxytestosterone formation rate was comparable to, or

Figure 6. 6b- and 2b-Hydroxytestosterone formation during incubation of D3 treated Caco-2 monolayers with testosterone (100 mM) for 120 min. The HPLC chromatograms indicate (a) intracellular, (b) apical, and (c) basolateral accumulation of 6b- and 2bhydroxytestosterone (retention times 6.2±6.4 and 11.5 minutes, respectively). Intracellular (a) 6b-hydroxy testosterone concentration in D3 treated cells was 19 and 8 times lower than in the apical (b) and basolateral (c) chambers, respectively. The observed differences indicate that ef¯ux mechanisms in both apical and basolateral membranes are active in the 6b-hydroxy testosterone distribution process. In control cells, intracellular 6b-hydroxytestosterone was below the level of detection (data not shown). In D3 treated monolayers, 2b-hydroxytestosterone was detected intracellularly (a) but not extracellularly [(b) and (c)]. The same pattern was seen in control monolayers (data not shown). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

1747

Table 2. Effect of Putative CYP3A4 Inducers on Testosterone Metabolism in Caco-2 Monolayers Formation of 6b-OH Testosterone (mM)b Inducera ‡D3

D3 0.57  0.04 Ð

Dex

RetAc c

0.06  0.00 0.55  0.02

< 0.05 0.34  0.06

GH

Clo c

< 0.05 0.48  0.15

< 0.05c n.d.d

a 1a,25 (OH)2 vitamin D3 (D3), dexamethasone (Dex), 9-cis-retinoic acid (RetAc), recombinant human growth hormone (GH), and clotrimazole (Clo) are putatively involved in the control of CYP3A expression via different intracellular receptors: human vitamin D receptor (hVDR),50 human glucocorticoid receptor (hGR),51 retinoid X receptor (RXR),52 human growth hormone receptor (hGHR),53 and human pregnane X receptor (hPXR).47 b The catalytic activity of CYP3A in Caco-2 cell monolayers was determined as 6b-hydroxylation of testosterone. c Below detection limit. d Not determined.

slightly higher than, results reported previously for D3 induced Caco-2 cells or stable CYP3A4 expression systems.16,35 The apparent Km value of 90 mM determined in D3 induced Caco-2 cell monolayers was comparable to values reported for 6b-hydroxylation of testosterone in human intestinal and liver microsomes (both 90 mM).10 The corresponding Vmax was about ®vefold lower than reported for 6b-hydroxylation in intestinal microsomes.10 However, the Vmax value reported in this study was obtained from intact cell monolayers. If we assume that the preparation of Caco-2 microsomes would give the same ®ve- to tenfold enrichment in catalytic CYP3A4 activity as that reported for Caco-2 cells previously,16 the Vmax values in the present study and those from human intestinal microsomes are in good agreement. Based on these results, we conclude that the catalytic activity of functional CYP3A4 in D3 induced Caco-2 cells is comparable to that found in the normal human intestinal epithelium. CYP3A4 and CYP3A5 sometimes produce different product ratios from the same substrates.38,39 The isoenzymes may have different Km values for many substrates,39,40 although other reports indicate that CYP3A4 and CYP3A5 displays comparable catalytic activity towards several CYP3A4 substrates.41,42 Nevertheless, CYP3A5-expressing cell lines such as TC7 are not ideal models for studies of CYP3A4 metabolism. Preliminary studies of a series of Caco-2 clones revealed that the various clones31 displayed highly variable expression of CYP3A4 and CYP3A5 mRNA both before and after D3 treatment, and we therefore searched for the cell line that expressed the most advantageous CYP3A4/ CYP3A5 ratio. Surprisingly, our results showed that the most commonly used Caco-2 cell line, Caco-2, gave the highest CYP3A4/CYP3A5 ratio of all investigated

Caco-2 variants. Moreover, the lack of functional CYP3A activity against testosterone in uninduced cells suggests that Caco-2 provides a gene switch model with regard to CYP3A4 activity. We note that the Km value for testosterone is ninefold higher for CYP3A5 than for CYP3A4.40 It can therefore not be excluded that the metabolism of compounds with a much higher af®nity (lower Km) to CYP3A5 than testosterone can be partly mediated by this enzyme in Caco-2 cells. In this study, all cell lines originating from both small and large intestine expressed MDR1. This is in agreement with an earlier report that clearly showed the presence of MDR1 mRNA expression throughout the entire human intestinal tract.43 Several studies suggest that CYP3A4 and MDR1 are coregulated in epithelial cells.6 While this coregulation may be of interest in some studies, it will introduce a new variable in functional studies on the interplay between CYP3A4 metabolism and MDR1 ef¯ux. Likewise, in this study, preliminary results suggested a small up-regulation of MDR1 mRNA after D3 treatment (data not shown). This up-regulation was con®rmed in functional studies of the ef¯ux of the Pgp substrate celiprolol. However, the secretory ¯ux only increased marginally in the D3 treated cells compared to the untreated controls, indicating that this increase is too small to in¯uence the interpretation of the effects of CYP3A4 induction on the ef¯ux of formed metabolites. The TEER of the Caco-2 cell monolayer clone selected for CYP3A4 studies by Schmiedlin-Ren et al.18,33 was reduced by 50% after D3 treatment. In our experience, such a large reduction in integrity is neither stable nor reproducible,44 and will seriously affect the permeability of the Caco-2 cell monolayers to low permeability drugs.45 In contrast, we found that the D3 treated Caco-2 cell line retained monolayer integrity,

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1748

ENGMAN ET AL.

indicating that this widely available polyclonal cell line is superior for transport studies in CYP3A4-expressing cell monolayers. This is perhaps not surprising if we consider the aneuploid properties of the Caco-2 cell line. Clone-speci®c effects (e.g., loss of integrity after D3 induction) would be leveled out in a polyclonal cell population, while they would be readily detectable in isolated subclones. Unexpectedly, we noted that CYP3A5 mRNA was markedly induced in both Caco-2 and TC7 cells compared to the control cells. This is a new ®nding because an earlier report concluded that human liver CYP3A5 is not inducible.46 However, other workers have been able to induce CYP3A5 in the human colon cancer cell line LS 180 using clotrimazole and reserpine as inducing agents.6 This discrepancy might be due to tissue-speci®c regulation,47 a hypothesis that is supported by studies showing that neither CYP3A4 nor CYP3A5 in LS 180 were affected by dexamethasone, an inducer of CYP3A4 in the liver.6 Similarly, CYP3A4 protein in Caco-2 cells was not induced by dexamethasone,10 and, in line with those results, no quantitatively important increase in CYP3A activity was detected in our Caco-2 cells after dexamethasone treatment. Thus, while various Caco-2 cell lines are readily induced by D3, they do not respond to the common inducer dexamethasone, which may be due to a lack of a functional cross-talk between the human glucocorticoid receptor (hGR) and human pregnane X receptor (hPXR) signaling pathways.48 We therefore investigated the effects of a panel of various other potential CYP3A4 inducers in the Caco-2 cell line. Putative ligands for hPXR and the human growth hormone receptor (hGHR) were included in the induction protocol at concentrations known to induce CYP3A4 in other cell lines,6,11,49 alone or in combination with D3. 9-cis-retinoic acid was included in an attempt to study potential effects via the retinoic X receptor (RXR). The results were disappointing because none of the investigated compounds induced a signi®cant increase in CYP3A4 activity, alone or in combination with D3. Possible reasons for these results may be related to the tumor and/or colonic origin of Caco2 cells, resulting in tissue speci®c regulation or differences in host cell factors such as hPXR and RXR, rather than the CYP3A4 genes.47 We conclude that Caco-2 cells cannot be induced by several common inducers. Further studies of the expression of gene regulatory elements are

needed to elucidate the mechanisms behind this result. In conclusion, these results show that transformed Caco-2 cells, but not normal intestinal epithelial cell lines, can be induced by D3 to express large amounts of CYP3A4, but also, in a clone-dependent manner, variable amounts of CYP3A5. In our hands, the generally available Caco-2 cell line at high passage numbers from ATCC performed best because it: retains monolayer integrity after D3 induction, displays no CYP3A4 activity in the uninduced state, displays a high expression of CYP3A4 but a low expression of CYP3A5 after induction, and retains an almost constant function of Pgp after D3 induction. These features make the Caco-2 cells suitable for studies of the interplay between CYP3A4 and Pgp, and also for studies of drug-drug and drug-food interactions. Such studies are underway in our laboratory.

ACKNOWLEDGMENTS This work was supported by AstraZeneca R&D, MoÈlndal, Sweden, and The Swedish Medical Research Council, grant 9478. We thank Dr. Edward L. LeCluyse, University of Chapel Hill, and Drs. Tommy B Andersson, Ulf Bredberg, Carl-Gunnar RegaÊrdh, and Anna-Lena Ungell at AstraZeneca R&D MoÈlndal, Sweden, for valuable discussions.

REFERENCES

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1. Maurel P. 1996. The CYP3A family. In: Ioannides C, editor. Cytochromes P450: Metabolic and toxicological aspects. Boca Raton: CRC Press, Inc., p 241±270. 2. Watkins PB, Wrighton SA, Schuetz EG, Molowa DT, Guzelian PS. 1987. Identi®cation of glucocorticoid-inducible cytochromes P-450 in the intestinal mucosa of rats and man. J Clin Invest 80:1029± 1036. 3. Wu CY, Benet LZ, Hebert MF, Gupta SK, Rowland M, Gomez DY, Wacher VJ. 1995. Differentiation of absorption and ®rst-pass gut and hepatic metabolism in humans: studies with cyclosporine. Clin Pharmacol Ther 58:492±497. 4. Gan L-SL, Moseley MA, Khosla B, Augustijns PF, Bradshaw TP, Hendren RW, Thakker DR. 1996. CYP3A-like cytochrome P450-mediated metabolism and polarized ef¯ux of cyclosporin A in Caco2 cells: Interaction between the two biochemical

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

barriers to intestinal transport. Drug Metab Dispos 24:344±349. Wacher VJ, Wu CY, Benet LZ. 1995. Overlapping substrate speci®cities and tissue distribution of cytochrome P-450 3A and P-glycoprotein have implications for drug delivery and activity in cancer chemotherapy. Mol Carcinog 13:129±134. Schuetz EG, Beck WT, Schuetz JD. 1996. Modulators and substrates of P-glycoprotein and cytochrome P-450 3A coordinately upregulate these proteins in human colon carcinoma cells. Mol Pharmacol 49:311±318. Pinto M, Robine-Leon S, Appay MD, Redinger M, Triadou N, Dussaulx E, Lacroix B, Simon-Assman P, Haffen K, Fogh J, Zweibaum A. 1983. Enterocyte-like differentiation and polarization of the human colon carcinoma cell line Caco-2 in culture. Biol Cell 47:323±330. Hunter J, Hirst BH, Simmons NL. 1993. Drug absorption limited by P-glycoprotein-mediated secretory drug transport in human intestinal epithelial Caco-2 cell layers. Pharm Res 10:743± 749. Hirohashi T, Suzuki H, Chu X-Y, Tamai I, Tsuji A, Sugiyama Y. 2000. Function and expression of multidrug resistance associated protein family in human colon adenocarcinoma cells (Caco-2). J Pharmacol Exp Ther 292:265±270. Prueksaritanont T, Gorham LM, Hochman JH, Tran LO, Vyas KP. 1996. Comparative studies of drug-metabolising enzymes in dog, monkey, and human small intestine, and in Caco-2 cells. Drug Metab Dispos 24:634±642. Gervot L, CarrieÁre V, Costet P, Cugnenc P-H, Berger A, Beaune PH, de Waziers I. 1996. CYP 3A5 is the major P450 3A expressed in human colon and colonic cell lines. Environ Toxicol Pharmacol 2:381±388. Raeissi SD, Hidalgo IJ, Segura-Aguilar J, Artursson P. 1999. Interplay between CYP3A-mediated metabolism and polarized ef¯ux of terfendine metabolites in intestinal epithelial Caco-2 (TC7) cell monolayers. Pharm Res 16:625±632. Wrighton SA, Brian WR, Sari MA, Iwasaki M, Guengerich FP, Vandenbranden M. 1990. Studies on the expression and metabolic capabilities of human liver cytochrome P450IIIA5 (HLp3). Mol Pharmacol 38:207±213. Zhang Q-Y, Dunbar D, Ostrowska A, Zeisloft S, Yang J, Kaminsky LS. 1999. Characterization of human small intestinal cytochromes P-450. Drug Metab Disp 27:804±809. Paine MF, Khalighi M, Fisher JM, Shen DD, Kunze KL, Marsh CL, Perklins JD, Thummel KE. 1997. Characterisation of interintestinal and intraintestinal variations in human CYP3A-dependent metabolism. J Pharmacol Exp Ther 283:1552± 1562.

1749

16. Crespi CL, Penman BW, Hu M. 1996. Development of Caco-2 cells expressing high levels of cDNAderived cytochrome P450 3A4. Pharm Res 16: 1635±1641. 17. Hu M, Li Y, Davitt CM, Huang S-M, Thummel K, Penman BW, Crespi CL. 1999. Transport and metabolic characterization of Caco-2 cells expressing CYP3A4 and CYP3A4 plus oxidoreductase. Pharm Res 16:1352±1359. 18. Schmiedlin-Ren P, Thummel K, Fisher JM, Lown KS, Watkins PB. 1997. Expression of enzymatically active CYP3A4 by Caco-2 cells grown on matrixcoated permeable supports in the presence of 1a,25(OH)2-D3. Mol Pharmacol 51:741±754. 19. Pang G, Buret A, O'Loughlin E, Smith A, Batey R, Clancy R. 1996. Immunologic, functional and morphological characterization of three new human small intestinal epithelial cell lines. Gastroenterology 111:8±18. 20. Peterson MD, Mooseker MS. 1992. Characterization of the enterocyte-like brush-border cytoskeleton of the C2Be clones of the human intestinal cell line, Caco-2. J Cell Sci 102:581±600. 21. CarrieÁre V, Lesuf¯eur T, Barbat A, Rousset M, Dussaulx E, Costet P, deWaziers I, Beaune P, Zweibaum A. 1994. Expression of CYP 3A in HT29MTX cells and Caco-2 clone TC7. FEBS Lett 355:247±250. 22. Artursson P, Karlsson J. 1991. Correlation between oral drug absorption in humans and apparent drug permeability coef®cients in human intestinal epithelial (Caco-2) cells. Biochem Biophys Res Commun 175:880±885. 23. Kostrubsky VE, Ramachandran V, Venkataramanan R, Dorko K, Esplen JE, Zhang S, Sinclair JF, Wrighton SA, Strom SC. 1999. The use of human hepatocyte cultures to study the induction of cytochrome P-450. Drug Metab Dispos 27:887± 894. 24. Claude P, Goodenough DA. 1973. Fracture faces of zonulae occludentes from ``tight'' and ``leaky'' epithelia. J Cell Biol 58:390±400. 25. Heid CA, Steven SJ, Livak KJ, Williams PM. 1996. Real Time Quantitative PCR. Genome Res 6:986± 994. 26. Perkin Elmer User Bulletin No. 2. ABI PRISM Sequence Detection System, Dec. 11. 1997. ``Relative quanti®cation of gene expression.'' P/N 430859. 27. Waxman DJ, Ko A, Walsh C. 1983. Regioselectivity and stereoselectivity of androgen hydroxylations catalysed by cytochrome P-450 isozymes puri®ed from phenobarbital-induced rat liver. J Biol Chem 258:11937±11947. 28. Wilson G, Hassan IF, Dix CJ, Williamson I, Shah R, Mackay M, Artursson P. 1990. Transport and permeability properties of human Caco-2 cells: an in vitro model of the intestinal epithelial cell barrier. Contr Rel 11:25±40.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

1750

ENGMAN ET AL.

29. Hidalgo IJ, Raub TJ, Borchardt RT. 1989. Characterisation of the human colon carcinoma cell line (Caco-2) as a model system for intestinal epithelial permeability. Gastroenterology 96:736±749. 30. The cell line Intestine 407 (ATCC CCL-6) (Int 407) was initially included in the study due to its presumed ileal origin. However, during the course of this study, we were informed that Int 407 was initially established via HeLa cell contamination. Therefore, our results on Int 407 were excluded from this study (Int 407 expressed CYP3A4 and MDR1 mRNA, but did not form tight monolayers). 31. TC7, BBe-1, clone 12, and Caco-2 of low (20±30) and high passage numbers (92±105). 32. Chirayath MV, Gajdzik L, Hulla W, Graf J, Cross HS, Peterlik M. 1998. Vitamin D increases tight junction conductance and paracellular Ca2‡ transport in Caco-2 cell cultures. Am J Physiol-Gastroint Liver Physiol 36:G389±396. 33. Fisher JM, Wrighton SA, Watkins PB, SchmiedlinRen P, Shen DD, Kunze KL, Thummel KE. 1999. First-pass midazolam metabolism catalyzed by 1a,25-dihydroxy vitamin D3-modi®ed Caco-2 cell monolayers. J Pharmacol Exp Ther 289:1134±1142. 34. Hartmann C, FroÈlich M, Kraub D, Spahn-Langguth H, Kanuf K, Mutschler E. 1990. Comparative enantioselective pharmacokinetic studies of celiprolol in healthy volunteers and patients with impaired renal function. Eur J Clin Pharmacol 38:573±576. 35. Hochman JH, Chiba M, Nishime J, Yamasaki M, Lin JH. 2000. In¯uence of P-glycoprotein on the transport and metabolism of indinavir in Caco-2 cells expressing cytochrome P-450 3A4. J Pharmacol Exp Ther 292:310±318. 36. de Waziers I, Cugnenc PH, Yang CS, Leroux JP, Beaune PH. 1990. Cytochrome P450 isoenzymes, eposide hydrolase and glutathione transferase in rat and human hepatic and extrahepatic tissues. J Pharmacol Exp Ther 253:387±394. 37. Kolars JC, Lown KS, Schmiedlin-Ren P, Ghosh M, Fang C, Wrighton SA, Merion RM, Watkins PB. 1994. CYP3A expression in human gut epithelium. Pharmacogenetics 4:247±259. 38. Gorski JC, Hall SD, Jones DR, Vandenbranden M, Wrighton S. 1994. Regioselective biotransformation of midazolam by members of the human cytochrome P450 3A (CYP3A) subfamily. Biochem Pharmacol 47:1643±1653. 39. Raeissi SD, Guo Z, Dobson GL, Artursson P, Hidalgo IJ. 1997. Comparison of CYP3A activities in a subclone of Caco-2 cells (TC7) and human intestine. Pharm Res 14:1019±1025. 40. Waxman DJ, Lapenson DP, Toshifumi T, Gelboin HV, Gonzalez FJ, Korzekwa K. 1991. Steroid hormone hydroxylase speci®cities of eleven cDNAexpressed human cytochrome P450s. Arch Biochem Biophys 290:160±166.

41. Gibbs MA, Thummel KE, Shen DD, Kunze KL. 1999. Inhibition of cytochrome P450 3A (CYP3A) in human intestinal and liver microsomes: comparison of Ki-values and impact of CYP3A5 expression. Drug Metab Disp 27:180±187. 42. Gillam EMJ, Guo Z, Ueng Y-F, Yamazaki H, Cock I, Reilly PEB, Hooper WD, Guengerich FP. 1995. Expression of CYP3A5 in Escherichia coli: Effects of 50 modi®cation, puri®cation, spectral characterization, reconstitution conditions and catalytic activities. Arch Biochem Biophys 317: 374±384. 43. Thiebaut F, Tsuruo T, Hamada H, Gottesman MM, Pastan I, Willinham MC. 1987. Cellular localization of the multi-drug resistance gene product Pglycoprotein in normal human tissues. Proc Natl Acad Sci USA 84:7735±7738. 44. Anderberg EK, Nystrom C, Artursson P. 1992. Epithelial transport of drugs in cell culture. VII: Effects of pharmaceutical surfactant excipients and bile acids on transepithelial permeability in monolayers of human intestinal epithelial (Caco-2) cells. J Pharm Sci 81:879±887. 45. Lindmark T, Kimura Y, Artursson P. 1998. Absorption enhancement through intracellular regulation of tight junction permeability by medium fatty acids in Caco-2 cells. J Pharmacol Exp Ther 284:362±369. 46. Wrighton SA, Ring BJ, Watkins PB, Vandenbranden M. 1989. Identi®cation of a polymorphically expressed member of the cytochrome P-3450III family. Mol Pharmacol 36:97±105. 47. Lehmann JM, McKee DD, Watson MA, Willson TM, Moore JT, Kliewer SA. 1999. The humans orphan nuclear receptor PXR is activated by compounds that regulate CYP 3A4 gene expression and cause drug interactions. J Clin Invest 102:1016±1023. 48. Pascussi J-M, Drocourt L, Fabre J-M, Maurel P, Vilarem M-J. 2000. Dexamethasone induces pregnane X receptor and retinoid X receptor-a expression in human hepatocytes: synergistic increase of CYP3A4 induction by pregnane X receptor activators. Mol Pharmacol 58:361±372. 49. Canani RB, Bisceglia M, Bruzzese E, Mallordo G, Guarino A. 1999. Growth hormone stimulates, through tyrosine kinase, ion transport and proliferation in human intestinal cells. J Pediatr Gastroenterol Nutr 28:315±320. 50. Giuliano AR, Franceschi RT, Wood RJ. 1991. Characterization of the vitamin D receptor from the Caco-2 human colon carcinoma cell line: effect on cellular differentiation. Arch Biochem Biophys 285:261±269. 51. Pereira TM, Carlstedt-Duke J, Lechner MC, Gustafsson JA. 1998. Identi®cation of a functional glucocorticoid response element in the CYP3A1/ IGC2 gene. DNA Cell Biol 17:39±49.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001

CYP3A4, CYP3A5, AND MDR1 IN DRUG TRANSPORT STUDIES

52. Kliewer SA, Moore JT, Wade L, Staudinger JL, Watson MA, Jones SA, McKee DD, Oliver BB, Willson TM, Zetterstrom RH, Perlmann T, Lehmann JM. 1998. An orphan nuclear receptor activated by pregnanes de®nes a novel steroid signaling pathway. Cell 92:73±82.

1751

53. Liddle C, Goodwin BJ, George J, Tapner M, Farrell GC. 1994. Separate and interactive regulation of cytochrome P4503A4 by triiodothyronine, dexamethasone and growth hormone in cultured hepatocytes. J Clin Endocrin Metab 83: 2411±2416.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 11, NOVEMBER 2001