Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The cytochromes P450 (CYPs¹) belong to a superfamily of enzymes involved in a diverse array of physiological and xenobiotic metabolic pathways. Accordingly, the CYPs account for a substantial fraction of drug metabolism in humans. An area of concern in the pharmaceutical industry is the potential induction of various CYPs by pharmaceutical candidates, which if undetected can lead to toxicity or reduced efficacy due to drug-drug interactions (Park et al., 1996).

Although species differences in metabolism exist between rat and humans (Kocarek et al., 1994; Xu et al., 2000), primary rat hepatocytes cultured in vitro provide a reproducible and relatively accurate reflection of CYP induction in vivo. Progressively, primary hepatocyte culture technology has implemented hormone supplementation (Dickins and Peterson, 1980; Dich et al., 1988; Sidhu and Omiecinski, 1995), substratum attachment (Jauregui et al., 1986; Schuetz et al., 1988; LeCluyse et al., 1994), enriched medium formulations (Waxman et al., 1990; LeCluyse et al., 1999), and extracellular matrix (ECM) overlay approaches (Sidhu et al., 1994; Brown et al., 1995; Silva et al., 1998) in an attempt to recapitulate the whole organ responses to CYP inducers: these efforts have met with considerable success.

Modern screening methodologies for monitoring gene expression, such as quantitative real-time reverse transcription-polymerase chain reaction (real time RT-PCR), are finding utility in clinical settings (Emig et al., 1999; Eckert et al., 2000) and are poised for mainstream use in drug development. Real-time RT-PCR quantitation of nucleic acid is based on detection of amplified products at the end of each cycle, which in turn permits a quantitative calculation of target RNA in an unknown sample by comparison with the kinetics of PCR product accumulation in samples of known quantity.

Real-time RT-PCR fluorescence is typically monitored during amplification by the hybridization of additional gene-specific oligonucleotide(s) that are fluorescently labeled to allow detection during PCR (Heid et al., 1996). In the TaqMan assay (Applied Biosystems, Foster City, CA), detection occurs following Taq-driven release of a 5' reporter dye from a 3' quencher molecule on a single hybridized probe during polymerase extension (Livak et al., 1995). In the LightCycler method (Roche Diagnostics, Indianapolis, IN), detection occurs during the annealing phase via a fluorescence resonance energy transfer (FRET) mechanism when an acceptor probe and a donor probe hybridize to a region of the amplified product in proximity (Wittwer et al., 1997b; for a comparison of both technologies, see Nitsche et al., 1999). Another method of detection available with the LightCycler technology uses the nonspecific dye SYBR-Green I, which undergoes excitation-induced fluorescence following intercalation into double-stranded PCR products during amplification (Wittwer et al., 1997a). All of these real-time RT-PCR detection methods allow the calculation of a cycle threshold based upon fluorescence, and therefore a quantitative estimate of the RNA present in the unknown sample.

Another method for RNA detection is used in the RNA invasive cleavage assay (Eis et al., 2001) known as the mRNA Invader assay (Third Wave Technologies, Madison, WI). This assay uses an isothermal, linear signal amplification method and possesses excellent nucleotide discrimination capability, which can be very powerful for analysis of highly homologous mRNAs. It was first used as a single nucleotide polymorphism detection method for rare polymorphisms in genomic DNA (Lyamichev et al., 1999) and it has recently been adapted to detect and quantify individual mRNA isoforms as well (Eis et al., 2001). The mRNA version of this assay takes advantage of an initial reaction in which a genetically engineered 5' nuclease recognizes a ternary complex formed between the mRNA nucleic acid target, an Invader oligonucleotide and a probe (Fig. 1). Enzymatic activity during this primary reaction results in a cleaved 5' flap, which acts as the invading oligonucleotide in a second round of the Invader reaction. In the second reaction the 5' nuclease recognizes a ternary complex formed between the released 5' flap, a secondary reaction target, and a FRET probe. Enzymatic activity during the secondary reaction results in the liberation of the fluorophore from the quencher following cleavage of the FRET oligonucleotide. The net fluorescence output of the two-coupled-step Invader reaction is directly proportional to the original amount of target nucleic acid present in the sample (Fig. 1).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Principles of invasive cleavage and linear fluorescence signal amplification used in the mRNA Invader assay. In the primary reaction, both the signal oligonucleotide (consisting of a 5' flap and a 3' region complementary to the target mRNA) and the Invader oligonucleotide hybridize simultaneously to the mRNA of interest so that both oligonucleotides overlap at a single isoform-specific nucleotide in the target mRNA. Hybridization to the target mRNA isoform of interest, but not with closely related isoforms bearing a nonhomologous nucleotide at the overlap position, results in the formation of a triplex structure recognized by the cleavase. Catalytic cleavage occurs immediately upstream of the nucleotide of interest on the signal oligonucleotide and releases the "flap" region, which serves as a substrate in the subsequent reaction. Following addition of the secondary reagents, in the secondary reaction the cleaved flap region acts as the Invader oligonucleotide and results in the formation of a new triplex structure, in this instance with the secondary reaction FRET oligonucleotide and the "generic" secondary reaction template. A second round of cleavase activity results in cleavage of the FRET oligonucleotide and the release of the fluorophore from its associated quencher. The fluorescence signal increases linearly over time and is directly proportional to the amount of target mRNA present in the original sample.

In the present study, real-time RT-PCR using the LightCycler and the mRNA Invader assay were compared for their abilities to detect and quantify CYP induction in primary rat hepatocytes. LC/MS/MS-based enzyme assays were used to verify these findings and culture conditions that optimized CYP induction profiles in primary rat hepatocytes were assessed.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Reagents. Cell culture reagents were obtained from Invitrogen (Carlsbad, CA). Chemicals used were from Sigma (St. Louis, MO) or Aldrich (Milwaukee, WI) and were of the highest purity commercially available.

Hepatocyte Isolation and Culture. Hepatocytes were prepared according to the two-step collagenase perfusion method (Moldeus et al., 1978) with modifications. Sprague-Dawley male rats were anesthetized with sodium pentobarbital (200 mg/kg) and following hepatic portal vein cannulation, livers were preperfused with oxygenated calcium-free Krebs-Heinslett buffer containing 2000 mg/l glucose (Sigma) previously pH-adjusted to pH 7.4 with 7.5% sodium bicarbonate solution (Invitrogen). Preperfusion of the liver with calcium-free Krebs-Heinslett proceeded for 10 to 15 min at ~15 ml/min. Livers were subsequently perfused with Krebs-Heinslett buffer containing 1 mM CaCl₂ and 0.025% (w/v) collagenase (Collagenase type II, 389 units/mg; Worthington Biochemicals, Freehold, NJ) for 30 min at ~10 ml/min. Perfused livers were excised and dispersed in 50 ml of collagenase buffer to which was added 50 ml of Williams' E medium (Sigma) containing 10 µg/ml insulin (Invitrogen), 0.2% bovine albumin fraction V (Sigma), 10% heat inactivated fetal bovine serum (Invitrogen), and penicillin/streptomycin. Digested livers were oxygenated and shaken in a closed, sterile container at 37°C for 10 min. The hepatocyte preparation was filtered through 100-µm nylon filters (Sigma) and hepatocytes were allowed to settle by gravity for 10 min. Pellets were resuspended in ~20 ml of ice-cold Krebs-Heinslett containing 1 mM CaCl₂ and spun at 300 rpm for 40 s. The wash was repeated three times and following the final spin hepatocytes were pooled and resuspended in Williams' E medium. Hepatocytes were counted by hemacytometry, and 1 × 10⁶ cells were added to 60-mm collagen-coated dishes, or 1 × 10⁴ cells were added to each well of single 96-well collagen-coated plates (Biocoat Cell Environment Collagen I Cellware; Becton Dickinson, Bedford, MA) and allowed to attach for 4 to 5 h. Viability was approximately 90 to 95% and only preparations with viabilities greater than 85% were used in experiments. Unless otherwise stated, after attachment the medium was removed and replaced with fresh Williams' E medium (without serum) containing 0.3 mg/ml ECM gel (Sigma). Following an overnight incubation with the ECM overlay, media were replaced daily.

Cell Treatments and RNA Isolation. At 48 h after plating, the media were changed and hepatocytes were incubated with fresh Williams' E medium containing the various inducers for the times indicated. -Naphthoflavone (-NF), hydrocortisone (HC), and phenobarbital (PB) were dissolved in dimethyl sulfoxide, and various concentrations of these inducers were added to cultures to yield 0.1% dimethyl sulfoxide. After induction, total cellular RNA was prepared using the RNeasy method (Qiagen, Chatsworth, CA) with on-column DNase treatment (to avoid genomic DNA contamination) and stored at 80°C until further use.

PCR Primer and Amplicon Design. The mRNA sequences of each of the rat CYPs were downloaded from GenBank (http://www.ncbi.nlm.nih.gov/GenBank), and nucleotide sequences from each of the subfamilies were aligned in Vector NTI (Informax, Inc., Bethesda, MD). A subfamily-specific region for each CYP was selected as the site of hybridization for either the 5' or 3' CYP PCR primer, and then complementary PCR oligos were screened on the basis of 1) similar melting temperatures, 2) similar oligo length, and 3) the production of a PCR amplicon with greater than 50% GC content (Table 1). Primer sets producing amplicons with the highest GC content were selected for the current experiments in an attempt to maximize the melting temperatures of the amplicons relative to their associated primer-dimers (see Results). Nucleotide positions of the oligonucleotides used in these studies were as follows: CYP1A1 (GenBank accession no. X00469) forward primer (+69 to +94): 5'-GATCATGCCTTCTGTGTATGGATTCC-3', reverse primer (+523 to +500): 5'-TGGAGAAACTCTTCAGCGCATTCT-3'. The primers for CYP2B1/2 (CYP2B1 GenBank accession no. J00719) were as follows: forward primer (+687 to +711): 5'-CTCCAAAAACCTCCAGGAAATCCTC-3', reverse primer (+1047 to +1023): 5'-GTGGATAACTGCATCAGTGTATGGC-3'. Note that these primers are 100% homologous to both CYP2B1 and CYP2B2 and will amplify the closely related isoforms with near-equal efficiency during real-time RT-PCR. The primers for CYP3A1 (GenBank accession no. X64401) were as follows: forward primer (+41 to +66): 5'-GAGGAGTAA TTTGCTGACAGAACCTGC-3', reverse primer (+189 to +167): 5'-CCAGGAAT CCCCTGTTTCTTGAA-3'.

LightCycler Real-Time RT-PCR. Each RNA sample (30-100 ng) was analyzed by real-time RT-PCR in a final reaction buffer volume of 20 µl containing SYBR-Green I dye, 6 mM MgCl₂, and 0.5 µM 5' and 3' primers. Reactions were transferred to glass capillaries and analyzed in the LightCycler instrument according to the manufacturer's instructions using fluorescence detection for SYBR-Green I with an excitation wavelength of 470 nm and an emission wavelength of 530 nm. An initial RT step occurred for 10 min at 55°C, and subsequent PCR proceeded as follows: denaturation 10 s at 95°C; annealing 10 s at 66°C; extension 20 s at 72°C. As a modification from the vendor-supplied protocol, fluorescence detection occurred after each cycle at 87°C during the ramp from extension temperature to denaturation. The total run time for a 45-cycle amplification was ~42 min.

In Vitro RNA Transcription. T7 sites (5'-GCGCTAATACGACTCACTATAGG GAGA-3') were added to the 5' oligonucleotides used for real-time RT-PCR analysis of the CYPs and used to amplify double-stranded PCR templates for RNA synthesis. In vitro transcription was performed using the MegaShortscript kit (Ambion, Austin, TX) and the relative purities of the in vitro-synthesized RNA transcripts were assessed by gel and capillary electrophoresis. Control RNA transcripts were stored at 80°C until further use.

mRNA Invader Assay. Hepatocyte total RNA samples (1-100 ng) were analyzed with the mRNA Invader assays (Third Wave Technologies) for rat CYP1A1, CYP2B1, and CYP3A1 according to the manufacturer's instructions. Primary reactions were incubated for 90 min at 60°C and secondary reactions were incubated for another 90 min at 60°C. Following the second incubation 100 µl of stop buffer (10 mM Tris, pH 8.0, 10 mM EDTA) was added to each well and the entire volume was transferred to a 96-well Costar microtiter plate. Fluorescence in each well was read on a Wallac Victor II plate reader using fluorescein filters (excitation 485 nm, emission 530 nm) and net fluorescence in samples was calculated by subtracting the fluorescence from a no RNA control. All assays were performed in triplicate. For kinetic traces, samples were treated as described above except primary reactions were incubated in glass capillaries, mixed with secondary reaction mixes, and monitored continuously in the LightCycler for the duration of the secondary reaction.

LC/MS/MS-Based CYP Enzyme Assays. At the end of the incubation period, parallel cultures of hepatocytes in 96-well plates were aspirated and washed in Hepatocyte Incubation medium without phenol red (catalog no. Z90009; In Vitro Technologies, Baltimore, MD) containing 5% fetal calf serum. Hepatocytes were then resuspended in Hepatocyte Incubation medium (0.1 ml) containing either 5 µM phenacetin (for CYP1A-catalyzed phenacetin O-deethylation) or 50 µM testosterone (for CYP2B-catalyzed testosterone 16-hydroxylation or for CYP3A-catalyzed testosterone 6-hydroxylation). Cells were returned to the 37°C incubator with 5% CO₂ for 2 h. After 2 h the media were collected and placed on ice. The remaining cells were trypsinized with 0.05 ml of 1× Trypsin-EDTA (Invitrogen) for 10 min and the resulting detached cells were added to the media. Samples were immediately frozen on dry ice and then shipped for processing at Pennsylvania Biolabs, Inc. (King of Prussia, PA) After thawing, the samples were immediately sonicated to disrupt the cells and were mixed with 2 volumes of cold acetonitrile. The supernatant was evaporated to dryness, extracted, and reconstituted in a methanol/water mixture. LC/MS/MS analysis was performed on an HP1100 (Hewlett Packard, Meriden, CT) LC system and a Quattro Ultima mass spectrometer (Micromass, Inc., Beverly, MA). Samples containing acetaminophen (metabolite of phenacetin O-dealkylation) were analyzed using electrospray ionization in the negative ion mode. Atmospheric pressure chemical ionization in positive ion mode was carried out to analyze the concentrations of 6-hydroxytestosterone and 16-hydroxytestosterone. Calibration standard curves and QC samples (four or five levels) of each metabolite were prepared in control hepatocyte lysates. The CYP450-dependent enzymatic activity is expressed in picomoles of metabolite per minute per 10⁶ cells and as percentage of vehicle control.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Observed Induction of CYP Enzyme Activity by LC/MS/MS-Based Assays in Hepatocytes. To initially confirm that the hepatocytes would be useful for evaluating the two mRNA quantitation methodologies, cultures (in serum-free media with ECM overlay) were first incubated with each of the inducers and then analyzed for elevations in CYP enzymatic activity. Consistent with previous reports on conditions most similar to those used in the present study (Madan et al., 1999), phenacetin O-deethylation (monitored as a marker of CYP1A activity in these studies) was induced maximally by 10 µM NF; CYP2B-catalyzed testosterone 16-hydroxylation was induced by low (100 µM) PB; and CYP3A-catalyzed testosterone 6-hydroxylation was induced by 100 µM HC (Fig. 2).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Induction of CYP enzyme activity in cultured rat hepatocytes. Rat hepatocytes were cultured in serum-free Williams' E medium with an ECM overlay (for analysis of CYP1A, CYP3A) or in serum-free Williams' E medium containing 100 nM DEX with an ECM overlay (for analysis of CYP2B). At 48-h postseeding, inducers at the following concentrations were added and incubated an additional 72 h: -NF, 10 µM; PB, 100 µM; and HC, 100 µM. Following induction hepatocytes were incubated with substrates for 2 h (phenacetin for CYP1A, testosterone for CYP2B and CYP3A). Untreated (UNT) hepatocytes served as the control. Sonication-disrupted cells and media were combined, and the products of CYP metabolism were analyzed by LC/MS/MS to determine rates of CYP activity.

Optimization of CYP Detection with Real-Time RT-PCR Using SYBR-Green I. In subsequent studies with real-time RT-PCR, the dye SYBR-Green I was used to detect double-stranded DNA accumulation after each cycle of amplification. The use of SYBR-Green I is attractive because it eliminates the need for redesigning gene-specific fluorescent dye-labeled oligonucleotides for each CYP to be analyzed. However, a limitation of this system is that SYBR-Green I will undergo excitation-induced fluorescence when bound to any double-stranded DNA: whether the target amplicon of interest, a closely related isoform, or a nonspecific double-stranded "primer-dimer", a truncated PCR product resulting from the nonspecific extension of primers annealed to primers during the latter phases of exponential PCR.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Melting peak analysis for optimized CYP amplicons. A, real-time quantitative RT-PCR in H2O (nonspecific, red trace) or rat liver RNA (positive control, blue trace) for CYPs. B, following the final cycle of amplification in the initial real-time RT-PCR reactions, all fluorescence values in each sample were set to 100%, and products were subjected to a 4°C/min heating ramp. Each reaction was monitored continuously for SYBR-Green I fluorescence. A plot of fluorescence versus temperature generated an initial melting curve analysis. C, by following the rate of change in fluorescence with respect to temperature, melting peaks were constructed and melting temperatures (Tm, the temperatures at which the rates of loss of fluorescence were greatest) of the amplicons and associated primer dimers were determined. The traces from primer sets generating optimal profiles for each CYP are presented.

Linearity and Dynamic Range of Real-Time RT-PCR and mRNA Invader Assay. To compare the abilities of each method to measure CYP mRNA in vitro, the purity of a full-length transcript for the CYP2B1 isoform (Third Wave Technologies) was verified by capillary electrophoresis and then serially diluted for analysis by both methods. Real-time RT-PCR traces for each dilution of the CYP2B1 transcript are depicted in Fig. 4A. As expected, increasing concentrations of the CYP2B1 transcript attain a threshold fluorescence at earlier cycle numbers.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Real-time RT-PCR and mRNA Invader assay kinetics for calibration curves of CYP2B1. Serial dilutions of the full-length in vitro transcript for CYP2B1 were subjected to analysis by real-time RT-PCR (A) and the mRNA Invader assay (B). Fluorescence in both assays was monitored continuously by the LightCycler instrument. The resulting data for each method were plotted in C as cycle threshold versus attomolar CYP2B1 transcript for real-time RT-PCR and in D as fluorescence versus attomolar CYP2B1 transcript for the mRNA Invader assay.

5

Effects of Inducers on CYP mRNA Measured by Both Methods. To examine each method's ability to measure CYP induction in a cellular system, hepatocytes were cultured under various conditions and then treated with inducers to determine levels of CYP1A1, 2B1/2, and 3A1 mRNAs by both methods. In these experiments, serum and ECM overlay were varied in an attempt to optimize the induction of each CYP as previously reported by other laboratories (Sidhu et al., 1993; Brown et al., 1995; Silva et al., 1998; LeCluyse et al., 1999). Hepatocytes were prepared and seeded onto Biocoat collagen-coated dishes (Falcon) in Williams' medium E under four conditions: 1) in the presence of both serum and an ECM overlay; 2) in the presence of an ECM overlay in serum-free media; 3) in the presence of serum without an overlay; and 4) in the presence of an ECM overlay in serum-free media containing 100 nM dexamethasone (DEX) (Fig. 5).

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Fold inductions of CYP1A1, CYP2B1(/2) and CYP3A1 under various culture conditions measured by real-time RT-PCR and the mRNA Invader assay. Hepatocytes (n = 3 preparations) were cultured under the various conditions described in Fig. 1 and then exposed to 10 µM -NF, 100 µM PB, or 100 µM HC for 48 h. RNA was purified and samples were analyzed by both methods as described. The fold inductions measured by real-time RT-PCR (dark blue) and the mRNA Invader assay (light blue) are presented.

Probing the Basis for Fold Inductions Observed Using Real-Time RT-PCR. The results for the -NF induction of CYP1A1 in three replicate hepatocyte preparations grown under different culture conditions are presented in Fig. 6. Quantitation of the attomolar CYP1A1 transcript per microgram of RNA in each sample demonstrates the basis for the varying fold inductions observed in the different culture systems. For instance, the greatest fold induction of CYP1A1 is in the presence of fetal bovine serum (lowest basal levels) with an ECM overlay (maximum inducible levels). Inclusion of DEX in the media maintained the basal content of CYP1A1 at a much higher level, resulting in an overall diminished fold-induction following treatment with -NF. This is in contrast to CYP2B1/2, where basal levels were not modulated by the culture conditions and only the maximal PB-inducible CYP2B1/2 levels were enhanced by DEX.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Basal and induced levels of CYP1A1 measured in multiple replicate experiments under various culture conditions by real-time RT-PCR. Three different preparations of rat hepatocytes were cultured as described in Fig. 1 and then incubated with 10 µM -NF for 48 h. Cycle thresholds for each of the samples were determined by real-time RT-PCR and compared with a standard curve for the CYP1A1 transcript to quantitate attomolar CYP1A1 present in each sample. Results by real-time RT-PCR for untreated (dark blue) and -NF treated (light blue) RNA samples are presented.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Reduced fold induction of CYP3A1 in DEX-free and 100 nM DEX-containing Williams' E medium. Hepatocytes were cultured in serum-free Williams' E medium in the absence or presence of 100 nM DEX and exposed to 100 µM HC for 48 h. RNA was prepared and subjected to real-time RT-PCR. A, large fold induction of CYP3A1 by HC in DEX-free Williams' E medium. B, elevated basal levels and reduced HC-inducible levels of CYP3A1 in Williams' E medium containing 100 nM DEX.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present studies were performed to evaluate the feasibility of using two relatively new methodologies to measure CYP induction in cultured rat hepatocytes. Both real-time RT-PCR and the mRNA Invader assay possess the sensitivity (only 1-100 ng of RNA required) necessary to screen the increasing numbers of pharmaceutical entities entering preclinical development. While real-time RT-PCR is currently limited to a low-to-moderate throughput, the mRNA Invader assay is especially well suited to a high-throughput screening format since the method is isothermal and requires no temperature cycling. Such a higher throughput technology could also be applied to screening larger numbers of compounds earlier in the drug discovery phase. An additional important benefit of the low sample requirement for either of these methods is the utilization of 96- and/or 384-well plate formats, which in turn greatly reduces the amount of pharmaceutical compound necessary for screening.

In these studies both the LightCycler real-time RT-PCR and mRNA Invader assays used purified RNA as the input material. Equivalent RNA amounts were normalized following A₂₆₀/A₂₈₀ measurements before analysis. While automated sample preparation systems and 96-well-based RNA purification columns are available, analysis would benefit greatly by adaptation to measure RNA directly in cell lysates, obviating the need for RNA purification. mRNA Invader assays have recently been successfully performed on crude lysates in fibroblasts (Kwiatkowski et al., 1999), but attempts in this laboratory to optimize quantification of RNA from lysates of ECM-overlay primary hepatocytes using the mRNA Invader assay have met with only limited success to date.

The close correlation between the fold inductions measured by both methods in these studies is encouraging. Others have recently used real-time RT-PCR with the TaqMan methodology for the quantification of CYPs in mouse and human hepatocytes with great success (Bowen et al., 2000; Pan et al., 2000). The results reported here are in close correlation with previous studies in rat hepatocytes using semiquantitative RT-PCR (Morris and Davila, 1996; Davila and Morris, 1999) and confirm the utility of real-time RT-PCR approaches (in this case, using the LightCycler technology) for the analysis of CYP mRNA by real-time RT-PCR in rat hepatocyte cultures as well.

While real-time RT-PCR using SYBR-Green I detection was found to be the more sensitive method, the mRNA Invader assay possesses superior specificity. The nonspecific nature of intercalation of SYBR-Green I dye into double-stranded DNA means that more than one mRNA may be detected by real-time RT-PCR experiments designed to detect a single isoform. Although the CYP1A1 and CYP3A1 primer pairs designed in this study were specific for their respective isoforms, the primers for CYP2B1 PCR amplification could not be optimized to discriminate between CYP2B1 and CYP2B2 (nucleotide identity ~96%). Thus, the real-time RT-PCR assays in this study using SYBR-Green I detection were unable to distinguish between the CYP2B1 and CYP2B2 products generated during PCR. It is unlikely that even internal fluorescent gene-specific oligonucleotides, which normally provide greater specificity than SYBR-Green I detection during real-time RT-PCR, could distinguish between the CYP2B1 and CYP2B2 isoforms. Thus, to date, it appears nearly impossible to discriminate between the CYP2B1/2 isoforms using PCR-based methods.

In constrast, since the requisite cleavage event in the mRNA Invader assay primarily relies on the base pairing of only a single nucleotide between the mRNA-specific detection oligonucleotide and the target (Fig. 1), isoforms of even the most closely related mRNAs are discriminated by the RNA invasive cleavage assay with great certainty. This has recently been conclusively demonstrated with respect to the CYP2B1 and CYP2B2 isoforms (Eis et al., 2001). This capability of the RNA invasive cleavage assay is particularly relevant for analysis of CYP isoforms, a superfamily of genes with unique members differing by as little as less than 5% in many cases (Nelson et al., 1996). It is not unreasonable to expect that in the cases of as-yet-undiscovered inducers, various closely related CYP isoforms from the same subfamily may be induced to different extents depending on the nuclear/orphan receptor heterodimers involved; on the exact nucleotide sequence of the response elements; and the overall topology of the promoter in question (Waxman, 1999). In such cases, the mRNA Invader assay should prove invaluable for rapidly assessing whether a given drug treatment differentially induces CYP isoforms from the same subfamily.

Using both real-time RT-PCR and the mRNA Invader assay, robust inductions of CYP1A1 and CYP3A1 by their respective inducers were preserved throughout a wide range of culture conditions, and the greatest fold inductions were observed in the presence of both serum and ECM overlay. It has long been known that DEX supplementation preserves the liver phenotype of hepatocytes in culture (Dich et al., 1988). Since the main objective of these studies was to maximize conditions for induction, DEX was initially omitted from the media formulation in an attempt to generate greater fold inductions of the various CYPs for the purpose of high-throughput-based screening methods. Hepatocytes cultured in the absence of DEX gave much higher fold inductions for CYP1A1 and CYP3A1 in the presence of their respective inducers. However, consistent with multiple previous reports, CYP2B1 was only weakly induced by PB alone but was strongly induced by PB in the presence of DEX (Waxman et al., 1990; Sidhu et al., 1993; Silva et al., 1998). Supplementation with DEX was found to maintain CYP3A1 at high basal levels, thereby diminishing effects of potential CYP3A1 inducers (20-50-fold compared with ~250-fold in the absence of DEX). This was expected, since it was previously observed that 100 nM DEX in Williams' E medium elevates CYP3A1 levels (Sidhu and Omiecinski, 1995). Under the conditions used in this study, culturing hepatocytes in both the presence and absence of DEX is preferred to maximize the fold inductions for all three CYPs in the interest of rapid screening detection.

In summary, both real-time RT-PCR and the mRNA Invader assay are effective new methodologies for screening CYP mRNA induction in primary rat hepatocytes in culture. While real-time RT-PCR appears to be the more sensitive method, the RNA invasive cleavage assay affords greater specificity. Both assays permit large numbers of compounds to be screened from a single hepatocyte preparation in 96-well plate format and require much lower amounts of active pharmaceutical compound than other methods for induction.