Enzymes belonging to the cytochrome P450 3A (CYP3A) subfamily play an important role in the metabolism of endogenous substances and xenobiotics, including pharmaceuticals. Xenobiotics can alter CYP3A expression and activity, and therefore, changes in CYP3A activity may serve as a biomarker of xenobiotic exposure. To determine changes in CYP3A enzyme activity for environmental risk assessment of xenobiotics including pharmaceuticals, high-throughput assays are needed, but these are missing for fish cells to date. Here, we report on the development of a fluorescent-based CYP3A high-throughput assay for four fish cell lines cultivated in 96-well plates based on 7-benzyloxy-4-trifluoromethylcoumarin as a CYP3A substrate. We show that human CYP3A substrate BFC is catalyzed by fish CYP3A enzymes to a fluorescent product. Its formation is dependent on cell numbers and incubation time. Furthermore, we demonstrate that with this new CYP3A assay induction and inhibition of enzyme activity by pharmaceuticals can be determined. This new cell-based assay is suitable for detection of alteration in CYP3A enzyme activity in large-scale experiments for screening of pharmaceuticals occurring in the environment.
Despite the occurrence of pharmaceuticals in the environment, little is known about their potential effects on ecosystems . Some pharmaceuticals have been investigated for their effects on aquatic organisms such as fish [2,3], but it remains unclear whether dependable biomarkers of exposure and effects exist for human pharmaceuticals. Here, we propose the cytochrome P450 3A (CYP3A) enzyme as a potential candidate for exposure assessment in fish.
The CYP3A family is the largest subfamily of CYPs found in the liver and small intestine of mammals and fish [4–6]. Cytochrome P450 3A plays an important role in the metabolism of endogenous substances and xenobiotics including pharmaceuticals. Induction of CYP3A expression by rifampicin and dexamethasone has been shown to occur in fish . Cytochrome P450 3A was identified in several teleost species (e.g., CYP3A27 and CYP3A45 in rainbow trout [5,8], CYP3A38 and CYP3A40 in medaka [9,10], CYP3A30 and CYP3A56 in killifish , and CYP3A65 in Danio rerio ).
Because xenobiotics can influence CYP3A expression and activity, a change in CYP3A expression could serve as an indicator of exposure of an organism to them and provide information on their metabolism. There are several requirements for assays to be suitable as potential biomarkers. They should have a toxicological basis and short handling time, be low in costs, and, preferably, have a high-throughput potential. Alterations of CYP3A activity, respectively, mRNA level, are known as a suitable biomarker used for assessing contaminant exposure and effects.
Cytochrome P450 3A enzymes of mammals and fish exhibit similar catalytic properties due to structural similarities . In fish, they are regulated by the same receptors as in mammals, namely, through the pregnane X receptor (PXR) [13–15]. The CYP3A expression can also be modulated by the constitutive androstane receptor, but the constitutive androstane receptor is restricted to mammals [16,17]. Due to these similarities, specific probes to determine CYP3A activity used in humans are supposed to be also specific for CYP3A activity in fish. Up to now, several substrates and their metabolites, respectively, have been used to determine CYP3A activity in mammals and in fish. The most widely used substrates to test CYP3A activity are testosterone  and erythromycin (erythromycin N-demethylase activity) [19,20]. Assays using these substrates have the disadvantage that they require time-intensive analytical tools such as high-performance liquid chromatography coupled with UV spectrophotometric detection or coupled with mass spectroscopy to investigate the metabolites. In addition, O-alkyl derivatives of resorufin, fluorescein, 7-hydroxycoumarins, and 6-hydroxyquinolines have been examined as substrates in mammals . Commercial kits (P450-Glo™ Assays, Promega) offer another possibility to measure CYP activity based on a luminescent method. These biochemical assays using purified components or subcellular fractions are designed to measure the activities of P450 enzymes from recombinant and native sources and can also be used for cell-based CYP450 induction assays. The assays contain different human substrates for different CYP450 enzymes such as CYP3A, CYP2C, etc.
Currently, only assays with testosterone hydroxylase, erythromycin N-demethylase, 7-benzyloxy-4-trifluoromethyl-coumarin (BFC), or aminopyrine N-demethylase as a substrate have been used to show CYP3A activity in microsomal preparations in fish [22–24]. Thus far, there have been no whole-cell-based assays in microtiter plates for measuring CYP3A activity. Absorbance- or fluorescence-based assays in microtiter plates that do not require metabolite separation by chemical analytical techniques represent important progress and should be developed for large-scale monitoring of xenobiotics including pharmaceuticals.
Here, we present a CYP3A assay based on fluorescence measurement in a 96-well plate reader using fish cell lines suitable for large-scale measurements of CYP3A activity. Previous studies demonstrated that CYP1A can be induced by xenobiotics in Poeciliopsis lucida hepatoma cell-line 1 (PLHC-1) cells [25,26], zebrafish liver (ZFL) cells , and rainbow trout gonadal cell-line 2 (RTG-2) cells . However, data on induction or inhibition of CYP3A enzyme activity and mRNA expression in fish are currently missing. The aim of the present study was to develop a versatile high-throughput assay for CYP3A activity in several fish cell lines, ZFL, RTG-2, and PLHC-1, using BFC as a substrate for potential use in studies of effects of pharmaceuticals, metabolism, and biomonitoring. We also analyzed whether pharmaceuticals can act as inducers or inhibitors of CYP3A-related enzyme activities in fish cell lines. Such information could facilitate the establishment of a new biomarker tool for pharmaceuticals, which can lead to induction or inhibition of CYP3A activity.
MATERIALS AND METHODS
Dimethylsulfoxide (DMSO), acetonitrile (ACN), rifampicin, and BFC were purchased from Sigma-Aldrich. Phosphate-buffered saline and diazepam were purchased from Roche Diagnostics.
Stock solutions of all pharmaceuticals were prepared in DMSO or ACN at a concentration of 100 mM if not otherwise stated. For the assay, stock solutions were diluted in cell culture medium resulting in a maximal solvent concentration of 0.1%. Further concentrations were prepared by serial dilution.
ZFL. Danio rerio liver cells were obtained from the American Type Culture Collection (ATCC No. CRL-2643). The medium consisted of 50% Leibovitz's L-15 medium (LuBioScience), 35% Dulbecco's modified Eagle's medium with 4.5 g/L glucose (LuBioScience), and 15% Ham's F12 (LuBioScience) (all without sodium bicarbonate) supplemented with 0.15 g/L sodium bicarbonate, 15 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 0.01 mg/ml insulin, 50 ng/ml epidermal growth factor, and 5% heat-inactivated fetal bovine serum (all from Sigma-Aldrich). The ZFL cells were grown at 28°C in a humidified incubator. Cells were split usually every 7 d and subcultured at split ratios of approximately 1:3.
PLHC-1. Poeciliopsis lucida hepatoma cell-line 1 cells were grown in DMEM/F12 (LuBioScience) supplemented with 5% fetal bovine serum (FBS; Sigma-Aldrich) in a humidified incubator with 5% CO2 at 30°C. Cells were split usually every 4 d and subcultured at split ratios of approximately 1:6.
FHM. Pimephales promelas (fathead minnow) (FHM) cells were grown in DMEM/F12 (LuBioScience) supplemented with 5% FBS (Sigma-Aldrich) and 20 mM HEPES at pH 7.2 (Sigma-Aldrich) at room temperature (21 ± 2°C). Cells were split usually every 7 d and subcultured at split ratios of approximately 1:3.
RTG-2. Oncorhynchus mykiss (rainbow trout) gonadal cell-line 2 (RTG-2) was obtained from the American Type Culture Collection (ATCC No. CCL-55). They were grown in DMEM/F12 supplemented with 10% FBS, 20 mM HEPES at pH 7.2 (Sigma-Aldrich) at room temperature (21 ± 2°C). Cells were split usually every 7 d and subcultured at split ratios of approximately 1:3.
Huh7. The human hepatoma cells Huh7 were grown in DMEM/F12 with GlutaMAX™ (LuBioScience) supplemented with 5% FBS in a humidified incubator with 5% CO2 at 37°C. Cells were split usually every 4 d and subcultured at split ratios of approximately 1:6.
Determination of CYP3A65 catalytic activity
Assays performed in Eppendorf tubes. Approximately 2.5 million cells of each cell line were transferred to an Eppendorf tube and incubated in phenol-red-free medium for 5 h. Phenol-red-free medium was used to prevent the color of phenol red from interfering with spectrophotometric and fluorescence measurements in assays containing 50 μM BFC (Sigma-Aldrich) or ACN, respectively. 7-Benzyloxy-4-trifluoromethyl-coumarin is metabolized by human CYP3A4 to give highly fluorescent 7-hydroxy-4-(trifluoromethyl)coumarin that can be detected readily spectrofluorometrically . After 5 h, the cells were centrifuged for 4 min at 2,000 g at 4°C. The supernatant was collected and transferred into a black 96-well plate (Huber). To stop the reaction, a stop solution (80% ACN and 20% 0.5 M Tris base) equal to 40% of the reaction volume was added. The plate was read on a GENios Tecan reader (Tecan) using an excitation wavelength of 410 nm and an emission wavelength of 530 nm. Each experiment was repeated three times.
Assays in 24-, 48-, and 96-well plates. Different numbers of cells were seeded in 24-, 48-, and 96-well plates and grown in culture medium. After 24 h, the medium was removed, and the cells were incubated in phenol-red-free medium containing 50 μM BFC or ACN, respectively, for 5 h. After this exposure time, the supernatant was transferred into a black 96-well plate and mixed with the stop solution, and the experiment was performed as described above.
Exposure to pharmaceuticals. Cells (∼100,000 cells per well) were seeded in 96-well plates and grown in culture medium. After 24 h, the medium was removed, and cells were incubated in phenol-red-free medium containing the test pharmaceutical or the solvent control, respectively. The cells were incubated for an additional 24 h, and the CYP3A-assay was performed as described above. To control for equal numbers of cells in each well, the cells were removed from the well with trypsin and incubated with lysis buffer (50 mM TrisHCl, 150 mM NaCl, 2 mM ethylenediaminetetraacetic acid, 1% Triton X-100, and 0.1% sodium dodecyl sulfate) to obtain whole cell extracts. From each whole cell extract, the protein content was determined with a Bradford assay (Bio-Rad Laboratories). Only wells with equal protein concentrations were analyzed for CYP3A activity.
Statistical analysis. Data were graphically illustrated with GraphPad® Prism 4 (GraphPad Software). Data distribution for normality was tested by Kolmogorov-Smirnov test, and the homogeneity of variance was tested by Levene's test. Differences between treatments were assessed by analysis of variance followed by a Bonferroni (Levene's test p > 0.05) test to compare the treatment means with the respective controls. In correlation tests, the different treatment groups were obtained by Pearson coefficients r. The results are given as mean ± standard deviation. Differences were considered significant at p ≤ 0.05. All computations were performed with SPSS® 16.0 for Windows.
Basic CYP3A activity in different fish cell lines
In the first set of experiments, we determined the CYP3A activity in vitro according to the protocol of Mensah-Osman et al. . Our aim was to evaluate whether the human CYP3A substrate (BFC) is also a substrate for fish CYP3A enzymes. To control for known CYP3A activity, the human hepatoma cell line Huh7 was used. Figure 1 demonstrates that all evaluated fish cell lines showed basic CYP3A activity. The PLHC-1 and FHM cell lines displayed a more than 1.5-fold higher CYP3A activity than Huh7. The ZFL cells showed the lowest CYP3A activity, but the detected values were clearly above the background signals.
Treatment of cells with different drugs at different concentrations at the same time for monitoring the CYP3A activity is not feasible, so far. For this reason, we changed the existing protocol to a more efficient experimental system that allows the processing of multiple samples at the same time. For this procedure, approximately 100,000 cells per well were seeded in a 96-well plate. After cultivation of the cells for 24 h, they were incubated with the CYP3A substrate directly in the wells instead of trypsinizing the cells and transferring them into a 1.5-ml tube for incubation with the substrate. Hence, product formation was assessed with intact cells. Figure 2A illustrates the two different procedures schematically (left panel: assay performed in 1.5-ml tubes; right panel: assay performed in microtiter plate). Figure 2B shows that CYP3A activity was detected when performing the assay in the wells. This is more elegant than performing the assay in a 1.5-ml tube. Because of fewer handling steps, more samples can be investigated in a shorter time, and there are fewer sources of potential handling mistakes. Through the use of 96-well plates, the amounts of reagents used such as pharmaceuticals, cell media, and buffers can be reduced. Without trypsinization of the cells, the risk of inducing cellular stress responses that could alter the outcome of the assay is reduced. Hence, the results are more reproducible. The measured basic CYP3A activities in the different cell lines obtained by the new cell assay were comparable to the activities obtained with the former protocol. Again, PLHC-1 and FHM cells showed the highest activity, and ZFL cells showed again much lower activity. From these experiments, we conclude that it is possible to perform the CYP3A activity assay directly in the culture dish without trypsinizing the cells.
Dependency of CYP3A activity on different conditions
To evaluate and validate the newly established CYP3A assay, different numbers of ZFL cells were seeded in 96- and 48-well plates. After 24 h of cultivation, the CYP3A activity assay was performed in the culture plates. Figure 3A demonstrates that increasing conversion of the CYP3A substrate into the fluorescent product was detected with increasing numbers of cells. This clearly confirms that the detected fluorescence is based on CYP3A enzyme activity of the cells and not due to background fluorescence. Additionally, zebrafish cells grown overnight in a 96-well plate were incubated with CYP3A substrate, and the enzyme reaction was stopped at different time points. As shown in Figure 3B, there was a clear correlation between incubation time and amount of measured fluorescence. This result confirms again that the measured fluorescence signals originate from cellular enzyme activity.
Dependency of CYP3A activity on BFC concentration
As another control experiment to control the functionality of the CYP3A assay, we incubated ZFL and FHM cells with different concentrations of BFC (10, 20, 30, 40, and 50 μM). There was a clear correlation between measured fluorescence and concentration of BFC. The highest fluorescent signal was measured with 50 μM BFC (Fig. 3C and D). This BFC concentration was used for all of the other experiments. As a proof of function of the enzyme in this new assay, the assay was stopped by protein precipitation with trichloroacetic acid. No conversion of the substrate was detectable when incubating cells with 10% trichloroacetic acid for 1 h prior to the incubation with BFC. These results clearly show that only living cells and active enzyme, respectively, can cause substrate conversion (data not shown).
Induction and inhibition of CYP3A activity
To confirm the functionality of CYP3A enzymes in fish cells and to demonstrate the usefulness of the new CYP3A cellular assay, we evaluated the induction or inhibition potential of CYP3A enzyme activity by model compounds. First, we tested the influence of the solvents used on CYP3A activity. To this end, Huh7 cells and PLHC-1 cells, grown overnight in 96-well plates, were left untreated or treated for 24 h with 0.1% ACN and 0.1% DMSO prior to incubation with BFC. As shown in Figure 3E and F, the solvents had no effects on CYP3A activity. After confirming that solvents do not alter CYP3A activity, we treated the cells with the model compounds. For this purpose, PLHC-1, RTG-2, FHM, and control Huh7 cells, all grown separately in a 96-well plate, were treated for 24 h with 10, 25, and 50 μM rifampicin before performing the CYP3A activity assay in the plate. Treatment of Huh7 cells with rifampicin resulted in an increase in enzyme activity up to around 250% compared with that of solvent control cells treated with DMSO (Fig. 4A). The treatment of PLHC1, RTG-2, and FHM with rifampicin also resulted in an induction of CYP3A enzyme activity (Fig. 4B-D). In all three cell lines, the induction was up to 150%.
In addition to the induction of CYP3A activity, the inhibition of CYP3A activity was evaluated by treatment of PLHC-1 cells and Huh7 cells as controls with the known human CYP3A inhibitor diazepam. PLHC-1 and Huh7 cells grown in a 96-well plate were treated for 24 h with diazepam followed by CYP3A activity measurement. As shown in Figure 5, the CYP3A activity in PLHC-1 cells was inhibited up to 50%; in Huh7 cells, the CYP3A activity was inhibited to 60 to 70%.
The aim of our present study was to develop a high-throughput procedure that allows the analysis of the effects of many different environmental chemicals and pharmaceuticals on fish CYP3A activity at the same time. The basic principle of existing CYP3A activity assays is to incubate CYP3A enzymes with a specific substrate for a few hours. During this time, the substrate is converted by the CYP3A enzymes into a product that is determined afterward. Different CYP3A substrates exist for the incubation step, but all of them are substrates for human CYP3A. In the present study, we decided to work with the human CYP3A substrate BFC, which is catalyzed by CYP3A and small contribution of CYP1A2 [30,31], that is converted into a fluorescent product by CYP3A. We first wanted to know whether this human CYP3A substrate is also metabolized by CYP3A enzymes of different fish. For this reason, we adapted the protocol from Mensah-Osman et al.  to evaluate whether the human-specific CYP3A substrate BFC [30–32] can also be catalyzed through fish CYP3A enzymes into the fluorescent product, as in carp . As shown in Figure 1, fish CYP3A enzymes indeed can catalyze BFC. We demonstrate this in four cell lines from different tissues and different fish families. Surprisingly, the basal catalytic activities of PLHC-1 and FHM were even higher than that of the human Huh7, which was used as a control. The ZFL and RTG-2 cells showed lower enzyme activities than the human cells, but the detected signals were above the background level.
After demonstrating that the human substrate is also a substrate for fish cells, we modified the existing protocol so that the whole procedure can be completely performed on microtiter plates. The cells were seeded on a 96-well plate, and after recovery overnight, the substrate BFC was added directly to the well. After incubation, the supernatant of each well was transferred to a new 96-well plate, loaded already with stop buffer, and the plate was read on a spectrofluorophotometer. Figure 2 demonstrates the same pattern of CYP3A activity as with the original protocol. Again, PLHC-1 and FHM cells showed higher CYP3A activity than Huh7 cells, and ZFL and RTG-2 cells again showed the lowest CYP3A activity.
To verify that we actually measured CYP3A activity, we performed a number of control experiments to evaluate the effects of different cell numbers, different incubation times with the substrate, and different concentrations of the substrate. Figure 3A shows a clear correlation between the number of cells and measured fluorescence. There is also a good correlation between measured fluorescence and incubation time (Fig. 3B): The longer the incubation time was, the higher the amount of fluorescence. The best results were obtained when the incubation time with the substrate was 5 h. This confirms that an enzyme activity was measured and not an artificial signal. We could also demonstrate a correlation between the concentration of BFC used and the measured fluorescence after incubation (Fig. 3C and D). The highest fluorescence signal was detected when the cells were incubated with 50 μM BFC. These results confirm again that we measured enzyme activity and not an artificial signal. Because we want to use this assay to evaluate the effects of different environmental pollutants and pharmaceuticals on CYP3A activity, we performed control experiments with the solvents that will be used later to dissolve the compounds. For this purpose, cells were incubated with 0.1% solvent (DMSO or ACN) for 24 h, and the CYP3A activity was measured afterward. As shown in Figure 3E and F, there is no influence of the solvents on CYP3A activity. With all of these control experiments, we were able to set up the optimal conditions, namely, 5 h of incubation with 50 μM BFC, for the CYP3A assay that were used later for induction and inhibition experiments.
To strengthen our newly developed assay, the effect of inducers and inhibitors on CYP3A activity was assessed. The CYP3A inducer rifampicin led to induction of CYP3A activity in PLHC-1 (Fig. 4B), RTG-2 (Fig. 4C), and FHM (Fig. 4D) cells, being significant already at 10 μM rifampicin. Compared with human control cells that show induction of enzyme activity up to 350%, the fish CYP3A enzymes which show an induction up to 150% are obviously less inducible than the human CYP3A. In addition to the induction of CYP3A activity, our new cell assay is also suitable to detect an inhibition of enzyme activity by pharmaceuticals. A clear dose-dependent inhibition of the enzymatic activity is found with diazepam (Fig. 5). Therefore, our cell assay is not only suitable for determining induction of CYP3A activity but also for detecting its inhibition. Our results clearly show that our modified protocol can be used to detect alteration in CYP3A activity by chemicals.
The overall advantage of our assay lies in the fact that the whole procedure can be performed on microtiter plates. Specifically, different compounds or different concentrations of one compound can be investigated on one plate at the same time; most of the steps can be performed using multichannel pipettes, which reduces handling time and minimizes pipetting mistakes and moreover has the potential for adaption for a pipetting robot; and the assay is not expensive compared with existing CYP3A kits on the market. This assay can be used as a first screening step in the environmental risk assessment of pharmaceuticals. With this method, pharmaceuticals or chemicals can be screened easily with regard to their abilities to alter CYP3A activity. Identified compounds can be examined further, including the quantification of CYP3A activity.
Another interesting potential of this assay is the comparison of different sensitivities of different fish species to environmental contaminants in regard to CYP3A activity alterations. The CYP3A activity and expression are regulated by PXR [13,14]. The PXR is a member of the nuclear hormone receptor superfamily. Pregnane X receptor functions as a ligand-activated transcription factor and regulates the metabolism, transport, and excretion of exogenous compounds, steroid hormones, vitamins, bile salts, and xenobiotics. Pregnane X receptor genes have been cloned and functionally characterized from a variety of vertebrate species, including human, rhesus monkey, mouse, rat, rabbit, dog, pig, chicken, frog, and zebrafish . Like other nuclear receptors, PXRs have a modular structure with two major domains: an N-terminal DNA-binding domain and a larger C-terminal ligand-binding domain (LBD). The PXR LBD is unusually divergent across species, compared with other nuclear receptors, with only 50% sequence identity between mammalian and nonmammalian PXR sequences; other nuclear receptors tend to have corresponding sequence identities at least 10 to 20% higher. Even the PXR DNA-binding domain, which is more highly conserved across species than the LBD, shows more cross-species sequence diversity than other nuclear receptors [16,17]. Due to this diversity, we would expect different sensitivities of different fish species to pharmaceuticals in regard to the alteration of CYP3A activity.
Future steps will be the use of this novel assay to screen pharmaceuticals for their abilities to alter CYP3A activity in different fish cell lines. This screening will help to identify compounds of ecotoxicological interest that subsequently will be investigated further.
We thank M. Heim (University Hospital Basel, Basel, Switzerland) for providing Huh7 cells, T. Wahli (University of Berne, Berne, Switzerland) for FHM cells, L.E. Hightower for PLHC-1 cells, and F. Hoffmann-La Roche Ltd. (Basel, Switzerland) for providing diazepam. We thank Andreas Hartmann and Birgit Hörger (Novartis International AG, Basel, Switzerland) and Jürg O. Straub (F. Hoffmann-La Roche Ltd., Basel, Switzerland) for providing pharmaceuticals, support, and reading the manuscript. This study was funded by the Bundesamt für Berufsbildung und Technologie (BBT), Kommission für Technologie und Innovation (KTI-Project 7114.2 LSPP-LS to K. Fent), Novartis International AG, F. Hoffmann-La Roche Ltd., and Förderverein Fachhochschule Nordwestschweiz Solothurn. V. Christen and D. Oggier contributed equally to this work.