Introduction

Top Abstract Introduction Materials & Methods Results & Discussion References

CYP3A4¹ is one of the most abundant P450 enzymes expressed in human liver, accounting for approximately 30% of the total microsomal cytochrome P450 (Gonzalez, 1990; Guengerich, 1992; Parkinson, 1996; Shimada et al., 1994; Wrighton and Stevens, 1992). All human livers seem to express CYP3A4, although the levels vary widely (10-fold or more) among individuals. In addition to CYP3A4, 10-30% of human livers express the structurally and functionally related P450 enzyme, CYP3A5. Some adult human livers also express CYP3A3, which seems to be an allelic variant of CYP3A4, or CYP3A7, which is otherwise considered a fetal member of the CYP3A subfamily (Gonzalez, 1990; Guengerich, 1992; Parkinson, 1996; Wrighton and Stevens, 1992). CYP3A4 and, to a lesser extent, CYP3A5 are responsible for the metabolism of a wide variety of xenobiotics (such as drugs, pesticides, and chemical carcinogens) and endobiotics (such as steroid hormones) (Gonzalez, 1990; Guengerich, 1992; Parkinson, 1996; Wrighton and Stevens, 1992).

Induction of CYP3A4 and perhaps more importantly inhibition of CYP3A4 are important causes of drug-drug interactions (Guengerich, 1992; Parkinson, 1996; Wrighton and Stevens, 1992). For this reason, drugs and new chemical entities are now being screened for their ability to inhibit CYP3A4, based on a variety of HPLC-based assays including the 6-hydroxylation of testosterone (Waxman et al., 1988, 1991), the hydroxylation and N-dealkylation of terfenadine (Rodrigues et al., 1995; Yun et al., 1993), the 1'- and 4'-hydroxylation of midazolam (Kronbach et al., 1989), the oxidation of nifedipine (Guengerich et al., 1986), the M1-, M17-, and M21-oxidation of cyclosporin (Combalbert et al., 1989; Kronbach et al., 1988), and the 10-hydroxylation of R-warfarin (Rettie et al., 1992). In the present study, we describe a non-HPLC assay of CYP3A4/5 activity based on the release of tritium (with formation of tritiated water) when [1,2,6,7-³H]testosterone is incubated with human liver microsomes and NADPH. The tritium-release assay is simple and relatively rapid and should, therefore, prove useful for screening new chemical entities as potential inhibitors of CYP3A4/5. Although the use of a radioactive substrate might be viewed a negative aspect of the new procedure, it has the added advantage that test articles will not interfere with the determination of CYP3A4/5 activity, which is occasionally a problem with HPLC analysis.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results & Discussion References

Chemicals and Reagents. [1,2,6,7-³H]Testosterone (specific activity, 92 Ci/mmol) and [³H]water (specific activity, 18 mCi/mol) were purchased from Amersham International (Arlington Heights, IL). Testosterone, erythromycin, activated charcoal (mesh 250-350), terfenadine, and metoprolol tartrate were purchased from Sigma. Testosterone was purified by preparative reversed phase HPLC to remove traces of androstenedione and other contaminants. The N-dealkylated metabolite of terfenadine, azacyclonol, was manufactured by Janssen Chimica but purchased from Spectrum Chemical Co. (Houston, TX). Hydroxyterfenadine (MDL 17,523) was obtained from Marion Merrell Dow (now Hoechst Marion Roussel, Inc.). Testosterone metabolites were obtained from sources described by Sonderfan et al. (1988). The steroid 5-reductase inhibitor 17-N,N-diethylcarbamoyl-4-methyl-4-aza-5-androstan-3-one (4MA) was a gift from Dr. G. H. Rasmussen of Merck, Sharp, and Dohme (Rahway, NJ).

Testosterone Oxidation: HPLC Analysis. Rates of testosterone oxidation were determined by the reversed phase HPLC method described by Pearce et al. (1996).

Testosterone Oxidation: Tritium-Release Assay. The method for determining rates of testosterone oxidation based on the release of tritium (with formation of tritiated water) from [1,2,6,7-³H]testosterone was patterned after the tritium-release assay for [4,6-3H]zoxazolamine described by Tomaszewski et al. (1976). Briefly, liver microsomes (0.2 mg) were incubated at 37±1°C in 500-µl incubation mixtures containing potassium phosphate buffer (50 mM, pH 7.4), MgCl₂ (3 mM), EDTA (1 mM, pH 7.4), NADP (1 mM), glucose 6-phosphate (5 mM), glucose 6-phosphate dehydrogenase (1 unit/ml), testosterone (14-400 µM), and the steroid 5-reductase inhibitor, 4-MA (1 µM), at the final concentrations indicated. Testosterone (0.7-25 mM) and 4-MA (1 mM) were added to each 500-µl incubation in 10 µl of methanol and 0.5 µl of acetone, respectively. Each sample contained 80 nCi of [1,2,6,7-³H]testosterone, in addition to the nonradioactive testosterone, which was present at final concentrations ranging from 14 to 400 µM. Reactions were started by addition of the NADPH-generating system. Incubations were stopped, typically after 8 min, with 1 ml of water containing ~75 mg of activated charcoal (the charcoal suspension was constantly stirred until all reactions had been terminated). The samples were vigorously mixed on a batch vortexer, and the charcoal was pelleted by centrifugation (2,500g for 15 min at 2-8°C). A 750-µl aliquot (i.e. 50%) of the aqueous phase was transferred to a scintillation vial containing 5 ml of biodegradable scintillation cocktail (Econo-Safe, Research Products International, Mount Prospect, IL), and the amount of radioactivity was determined by scintillation spectrometry with a Beckman LS6500 multi-purpose scintillation counter. Zero-time incubations served as blanks. All samples and standards were incubated in duplicate or triplicate. The experimental conditions were selected such that the amount of testosterone metabolized was <15%. Rates of testosterone 6-hydroxylation were determined from the empirical observation that 57 dpm of [³H]water in the aqueous phase was equivalent to 1.0 nmol of 6-hydroxytestosterone when the final concentration of testosterone was 250 µM and the specific activity was 0.6 Ci/mol. The basis for this relationship is described under Results and Discussion.

Terfenadine Metabolism. The in vitro metabolism of terfenadine was carried out in 500-µl incubation mixtures (final volume) containing human liver microsomes (50 µg of protein), potassium phosphate buffer (50 mM, pH 7.4), MgCl₂ (3 mM), EDTA (1 mM), NADP (1 mM), glucose-6-phosphate (5 mM), glucose-6-phosphate dehydrogenase (1 unit/ml), and terfenadine (4 µM) at the final concentrations indicated. Terfenadine was dissolved in methanol (20 mM) and added to each 500-µl incubation mixture in 0.1 µl. Reactions were started by addition of the NADPH-generating system and were stopped after a 4-min incubation at 37 ± 1°C by addition of an equal volume of methanol containing 1.0 µM metoprolol (internal standard). Precipitated protein was removed by centrifugation (~2,000g for 5-15 min at 4°C). An aliquot (typically 200 µl) of the clear supernatant fraction was analyzed by HPLC as described below. Zero-time incubations served as blanks.

Analysis of Kinetic Constants. The apparent kinetic constants (V_max, K_m, and K_i) were determined with an enzyme kinetics program from Trinity Software (Campton, NH, version 1.4.1), which weights data toward the higher reaction rates, which occur at high concentrations of substrate and/or low concentrations of inhibitor (weighting factor = 4).

	Results and Discussion

Top Abstract Introduction Materials & Methods Results & Discussion References

The purpose of this study was to develop a non-HPLC assay, based on the release of tritium during the oxidation of [1,2,6,7-³H]testosterone, that would allow rapid determination of the CYP3A activity in human liver microsomes. The ideal substrate would have been testosterone labeled with tritium in only the 6-position, which is the major site of testosterone oxidation by CYP3A4 (Waxman et al., 1988, 1991). However, [6-³H]testosterone is not commercially available and would likely be difficult to synthesize. Korzekwa et al. (1990) attempted to label specifically the 6-position of testosterone with deuterium and observed that 74% of the deuterium was incorporated into the 6-position. The [³H]testosterone used in these studies is commercially available and is labeled in the 1-, 1-, 2-, 2-, 6-, 6-, 7-, and 7-position. Most of the tritium is located in the 7- (25%) and 6- (20.1%) position. Only ~6% of the tritium is located in the 2- and 6-positions (the relative distribution between these two sites is not known). Although the distribution of tritium in [1,2,6,7-³H]testosterone is far from ideal, there were two reasons to believe that this substrate would serve its intended purpose.

First, CYP3A4 seems to be the only enzyme in human liver microsomes capable of hydroxylating testosterone at the 1, 2, 6, or 7-position. As shown in fig. 1, 6-hydroxytestosterone accounted for 75-80% of all metabolites detected by HPLC when testosterone was incubated with multiple samples of liver microsomes, despite large differences in CYP3A activity. Other metabolites known to be formed by CYP3A4/5 (namely 6-dehydrotestosterone and 1/-, 2-, and 12/18-hydroxytestosterone) accounted for an additional 10.5 to 12.0% of all metabolites (data not shown). None of the samples of human liver microsomes catalyzed the 2-, 6-, or 7-hydroxylation of testosterone, which would likely be formed by P450 enzymes other than CYP3A4 based on its predilection for abstracting hydrogen atoms in the -configuration. Oxidation at other sites, including the formation of androstenedione and trace amounts of 16-hydroxytestosterone, was not considered relevant (in terms of interfering with the assay of CYP3A activity) because little or no tritium label was present in the 16- or 17-position.

View larger version (44K): [in this window] [in a new window]   Fig. 1.   Testosterone oxidation by human liver microsomes. Human liver microsomes (0.1 mg) from 12 individuals (numbered 14-25) and a pool of microsomes from 7 individuals were incubated with testosterone (250 µM) for 8 min, and the pathways of oxidation were determined by HPLC as described in Materials and Methods. The right column (labeled %6) depicts the amount of 6-hydroxytestosterone as a percentage of the total testosterone metabolites formed. Data are averages of duplicate determinations.

Second, the 6-hydroxylation of testosterone was shown by Bjorkhem (1972) to proceed with no isotope effect (K_H/K_3H = 1.0), so the tritium label itself should not bias the results. In contrast to the sample of tritium-labeled testosterone used by Bjorkhem, the percentage of tritium in the 6-position of the sample of [1,2,6,7-³H]testosterone purchased from Amersham was too low to examine a possible isotope effect.

Inasmuch as the [1,2,6,7-³H]testosterone used in these studies contained only 6% of the tritium in the 2- and 6-positions, it was necessary to remove all of the radioactive substrate to avoid unacceptably high levels of background radioactivity. As shown in table 1, addition of ~75 mg of activated charcoal quantitatively (>99.95%) adsorbed [³H]testosterone such that the radioactivity in the aqueous phase decreased from >170,000 dpm to 30-50 dpm, which was then only about twice the level of background radioactivity. Under these conditions, no tritiated water adsorbed to charcoal (results not shown). No radioactivity (after blank correction) was recovered in the aqueous phase when human liver microsomes were incubated at 37°C with [1,2,6,7-³H]testosterone in the absence of NADPH (results not shown). However, when human liver microsomes (0.2 mg of protein) were incubated at 37°C for 8 min with [1,2,6,7-³H]testosterone (final concentration, 250 µM; specific activity, 0.6 Ci/mol) in the presence of NADPH, the amount of radioactivity in the aqueous phase increased 2-16-fold depending on the levels of CYP3A4/5 (discussed later).

Complete (100%) metabolism of [1,2,6,7-³H]testosterone by human liver microsomes would result in only a fraction of the tritium being released from the substrate. Testosterone is hydroxylated mainly in the 6-position, but this site contains no more than 6% of the total amount of tritium (the actual value is probably less than 6% because this represents the amount of tritium in both the 2- and 6-positions). Experiments were conducted to determine the relationship between the amount of radioactivity and the amount of 6-hydroxytestosterone formed. As shown in table 1, when samples of human liver microsomes were incubated with 250 µM testosterone (specific activity, 0.64 Ci/mol), the amount of [³H]water formed was proportional to the amount of 6-hydroxytestosterone formed. Formation of 1 nmol of 6-hydroxytestosterone was associated with the formation of approximately 57 dpm of [³H]water. The total amount of testosterone in the 500-µl incubation mixture was 125 nmol, and 6-hydroxytestosterone represented 75-80% of the total metabolites formed (fig. 1). Therefore, formation of 1 nmol of 6-hydroxytestosterone represented metabolism of of the substrate, thus complete metabolism of the substrate would be expected to produce 7,125 dpm of [³H]water. This amount of tritiated water (i.e. 7,125 dpm) represents only 4.0% of the total amount of radioactivity added to each incubation mixture (each incubation contained 80 nCi of [1,2,6,7-³H]testosterone, which is equivalent to 178,000 dpm). This suggests that approximately 4% of the tritium in [1,2,6,7-³H]testosterone is located in the 6-position, which is consistent with the manufacturer's claim that the 2- and 6-positions combined account for ~6% of the tritium in [1,2,6,7-³H]testosterone. The amount of [1,2,6,7-³H]testosterone added to each 500-µl incubation mixture (80 nCi) was kept constant, but the amount of unlabeled testosterone was varied to achieve different substrate concentrations. When the amount of unlabeled testosterone was changed, the specific activity of tritiated testosterone in the assay changed accordingly, as did the relationship between formation of [³H]water and 6-hydroxytestosterone. For example, when the concentration of testosterone was decreased from 250 µM to 125 µM with 80 nCi of [1,2,6,7-³H]testosterone added to each 500-µl incubation, the specific activity of tritiated testosterone doubled (from 0.64 to 1.28 Ci/mol), and formation of 57 dpm of [³H]water required formation of half as much 6-hydroxytestosterone (i.e. 0.5 nmol instead of 1 nmol).

Sensitivity of HPLC and Non-HPLC Assays. It would seem from the results shown in table 1 that the tritium-release assay is considerably less sensitive than the HPLC assay for measuring rates of testosterone 6-hydroxylation by human liver microsomes. In the case of sample 19, the signal-to-noise ratio of the tritium-release assay was only 3.8 (186 dpm vs. 49 dpm), whereas the signal-to-noise ratio for the HPLC assay was almost 100 (4,600 pmol of metabolite formed vs. a limit of quantitation of ~50 pmol/incubation). However, the sensitivity of the tritium-release assay can be improved simply by increasing the specific activity of the [1,2,6,7-³H]testosterone, as illustrated in fig. 2. When the specific activity of [1,2,6,7-³H]testosterone was increased, the amount of tritium released during the formation of 6-hydroxytestosterone increased accordingly. For example, the signal increased 10-fold when the specific activity of testosterone was increased from 20 to 200 Ci/mol (fig. 2). However, the background radioactivity (i.e. the amount of radioactivity in the zero-time incubations) increased only 4-fold (from 79 to 313 dpm). Consequently, both the signal and the signal-to-noise ratio increased as the specific activity of [1,2,6,7-³H]testosterone was increased. Although beneficial from a sensitivity point of view, adding a large amount of [1,2,6,7-³H]testosterone to each microsomal incubation has the disadvantage of increasing the cost of the assay and increasing the amount of radioactive waste generated. The experiments described hereafter involved 500-µl incubation mixtures that contained 80 nCi of [1,2,6,7-³H]testosterone plus 7-200 nmol of nonradioactive testosterone; hence, the specific activity ranged from 0.4 to 11.4 Ci/mol.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Effect of increasing the specific activity of [3H]testosterone on the amount of [3H]water formed by human liver microsomes. Human liver microsomes (0.2 mg) from a pooled sample were incubated with testosterone (12.5 µM, 80-800 nCi/incubation), and rates of testosterone 6-hydroxylation were determined by the non-HPLC assay as described in Materials and Methods. Values are the amount of radioactivity (mean ± standard deviation) in 50% of the aqueous phase after subtraction of the blank (zero time) values, which were 79 ± 8 dpm for 20 Ci/mol, 91 ± 2 dpm for 40 Ci/mol, 142 ± 11 dpm for 80 Ci/mol, and 313 ± 21 dpm for 200 Ci/mol.

Effect of Time and Protein Concentration. When human liver microsomes were incubated with [1,2,6,7-³H]testosterone and NADPH, tritium release was directly proportional to incubation time for 30 min (at a protein concentration of 0.2 mg/incubation) and directly proportional to protein concentration (up to 0.4 mg/incubation) during a 30-min incubation, as shown in fig. 3. Most incubations contained 0.1 or 0.2 mg of microsomal protein and were terminated after 8 min. Inter-assay variability and intra-assay variability were no more than 10%.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Effect of incubation time and protein concentration on the 6-hydroxylation of [3H]testosterone by human liver microsomes. Human liver microsomes (0.05 to 0.8 mg) from a pooled sample were incubated with [1,2,6,7-3H]testosterone (250 µM, 80 nCi/incubation) for different times (10-120 min with 0.2 mg protein/incubation) or at different protein concentrations (0.05-0.8 mg/incubation for 30 min). Rates of testosterone 6-hydroxylation were determined by the tritium-release assay as described in Materials and Methods. Data are averages of duplicate determinations.

cDNA-Expressed P450 Enzymes. A panel of recombinant human P450 enzymes was incubated with [1,2,6,7-³H]testosterone and NADPH, and rates of testosterone 6-hydroxylation were determined from the amount of tritiated water formed. As expected, cDNA-expressed CYP3A4 was considerably more active than any of the other recombinant P450 enzymes examined at catalyzing tritium release from [1,2,6,7-³H]testosterone, as shown in fig. 4. When the amount of each P450 enzyme in human liver microsomes is taken into account, CYP3A4 would be predicted to be the principal testosterone-hydroxylating enzyme, based on its high catalytic activity (fig. 4) and its abundance in human liver microsomes (Shimada et al., 1994). It should be noted that this panel of recombinant enzymes did not include CYP3A5, which would also be expected to catalyze a high rate of tritium release from [1,2,6,7-³H]testosterone (Wrighton et al., 1990).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Rates of testosterone 6-hydroxylation based on the formation of [3H]water from [1,2,6,7-3H]testosterone by microsomes containing cDNA-expressed human P450 enzymes. Microsomes (0.5 mg) prepared from human B-lymphoblastoid cells containing various cDNA-expressed P450 enzymes were incubated with [1,2,6,7-3H]testosterone (250 µM, 80 nCi/incubation), and rates of testosterone 6-hydroxylation were determined from the amount of [3H]water formed, as described in Materials and Methods. Control microsomes were from B-lymphoblastoid cells transfected with the vector alone. Data are averages of duplicate determinations.

Human Liver Microsomes. The rate of testosterone hydroxylation by 12 individual samples of human liver microsomes was determined both by the HPLC assay for testosterone 6-hydroxylation and by the tritium-release assay. As shown in fig. 5, the sample-to-sample variation in testosterone 6-hydroxylation determined by HPLC analysis correlated extremely well with that determined by the tritium-release assay (r = 0.98). In addition, rates of tritium release from [1,2,6,7-³H]testosterone correlated well (r  0.90) with the sample-to-sample variation in the rates of terfenadine hydroxylation and N-dealkylation, both of which are catalyzed by CYP3A4 (Rodrigues et al., 1995; Yun et al., 1993), as shown in fig. 6.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Sample-to-sample variation in testosterone 6-hydroxylation by human liver microsomes: comparison of the tritium-release assay with the HPLC assay. Human liver microsomes (0.1 or 0.2 mg) from 12 donors and a pooled sample were incubated for 8 min with nonradioactive testosterone (250 µM) or [3H]testosterone (250 µM, 80 nCi/incubation), and rates of formation of 6-hydroxytestosterone were determined by HPLC analysis or by the tritium-release assay, respectively, as described in Materials and Methods. Data are averages of duplicate determinations.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Sample-to-sample variation in terfenadine metabolism by human liver microsomes and testosterone metabolism determined by the tritium-release assay .  Human liver microsomes (0.2 mg) from 12 donors and a pooled sample were incubated with [3H]testosterone (250 µM, 80 nCi/incubation) for 8 min, and rates of testosterone 6-hydroxylation were determined with the tritium-release assay, as described in Materials and Methods. Rates of terfenadine metabolism were determined by HPLC analysis, as described in Materials and Methods. Data are averages of duplicate determinations.

Effects of Substrate Concentration. A pooled sample of human liver microsomes was incubated with a wide range of substrate concentrations (14-400 µM [1,2,6,7-³H]testosterone, and the rates of formation of tritiated water and 6-hydroxytestosterone were determined by the tritium-release assay and by HPLC analysis, respectively. Regardless of the analytical method, human liver microsomes catalyzed the hydroxylation of testosterone with an apparent K_m of 50-60 µM and V_max of 4500-4900 pmol/mg/min, as shown in fig. 7. Because a constant amount of radioactive testosterone (80 nCi) was added to each incubated mixture, the sensitivity of the tritium-release assay did not decrease as the substrate concentration decreased (in fact, it actually increased). In contrast, the sensitivity of the HPLC assay decreased with decreasing substrate concentration, although 6-hydroxytestosterone could still be readily detected even at the lowest concentration of testosterone tested.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effect of substrate concentration on the rate of testosterone oxidation determined by tritium-release and HPLC assay. Human liver microsomes (0.2 mg) from a pooled sample were incubated with testosterone (14-400 µM), and rates of formation of 6-hydroxytestosterone were determined by the HPLC assay. Rates of testosterone 6-hydroxylation (14-400 µM, 80 nCi/incubation) were determined with the non-HPLC assay as described in Materials and Methods. Data are averages of duplicate determinations.

Inhibition by Erythromycin. One application of the tritium-release assay is to screen drugs and new chemical entities as inhibitors of CYP3A4. Therefore, the amount of tritiated water formed when human liver microsomes were incubated with [1,2,6,7-³H]testosterone was examined in the presence and absence of erythromycin, a known inhibitor of CYP3A4. As expected, erythromycin inhibited the release of tritium from [1,2,6,7-³H]testosterone, as shown in the Dixon plot in fig. 8. In this experiment, human liver microsomes were incubated with erythromycin and [1,2,6,7-³H]testosterone simultaneously, so there was little opportunity for erythromycin to function as a mechanism-based inhibitor. Consequently, erythromycin inhibited testosterone oxidation competitively with a K_i value of 130 µM.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Determination of the Ki for erythromycin inhibition of testosterone oxidation by human liver microsomes. Human liver microsomes (0.2 mg) from a pooled sample were incubated with testosterone (12.5, 50, or 200 µM; 80 nCi/incubation) in the presence and absence of erythromycin (200, 400, or 600 µM), and rates of testosterone 6-hydroxylation were determined by the tritium-release assay as described in Materials and Methods. Data are averages of duplicate determinations.

Conclusion. The results of this study suggest that the formation of tritiated water from [1,2,6,7-³H]testosterone may be used as a selective probe of CYP3A4/5 activity in human liver microsomes. The tritium-release assay described in this paper is simple and rapid. Compared with the conventional HPLC method, the tritium-release assay is particularly suitable for the rapid screening of chemicals as potential inhibitors of CYP3A4/5. The tritium-release assay circumvents the occasional problem of test articles interfering with the detection of metabolites by HPLC.