Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Toxicity or therapeutic failure has long been recognized as a product of drug-drug interactions when polytherapy is instituted. With increasing use of polytherapy, as in the treatment of acquired immunodeficiency syndrome and its associated opportunistic infections, there is a need for rapid, simple, and cost-effective in vitro methods to predict clinically significant drug-drug interactions. Once validated, such methods can be used to identify and prioritize the further study of drug-drug interactions that are most likely to be significant in the clinic.

As part of an ongoing series of studies on drug-drug interactions associated with anti-acquired immunodeficiency syndrome and anti-opportunistic infection (OI) drugs (Palamada et al., 1995), we sought to determine the capacity of these drugs to inhibit the microsomal cytochrome P450 monooxygenase (P450³) enzymes most commonly involved in xenobiotic metabolism. To do so, we first evaluated the selectivity of various in vitro probes and the suitability of various organic solvents for the measurement of human cytochrome P450 activities in human liver microsomes. We chose to focus our studies on the P450 enzymes because these enzymes are involved in the clearance of the vast majority of the anti-OI drugs. In humans, these enzymes are present at the highest concentration in the liver but are also present in other tissues such as the intestinal mucosa. They exist as a superfamily of isozymes, of which there are more than 37 individual members in humans (Nelson et al., 1996). However, relatively few isozymes are responsible for most metabolic biotransformations involving the clearance of drugs from the body. These are P450s 1A2, 2A6, 2C8/9, 2C19, 2D6, 2E1, and 3A4/5, which collectively account for more than 70% of all P450 isozymes present in the human liver (Shimada et al., 1994).

We selected our assays based on previous accounts of the selectivity of each drug probe for each specific P450 isozyme of interest. We also considered the ease with which each assay could be implemented on a simple HPLC system with either UV or fluorescence detection. As a result, we chose caffeine N3-demethylation as a probe for CYP1A2 (Butler et al., 1989; Fuhr et al., 1992; Gu et al., 1992; Tassaneeyakul et al., 1992), coumarin 7-hydroxylation for CYP2A6 (Maurice et al., 1991; Miles et al., 1990; Yamano et al., 1990), tolbutamide 4-methyl-hydroxylation for CYP2C8/9 (Miners et al., 1988; Brian et al., 1989; Relling et al., 1990; Doecke et al., 1991; Srivastava et al., 1991), S-mephenytoin 4-hydroxylation for CYP2C19 (Srivastava et al., 1991; Goldstein et al., 1991), dextromethorphan O-demethylation for CYP2D6 (Kupfer et al., 1986; Dayer et al., 1991), chlorzoxazone 6-hydroxylation for CYP2E1 (Peter et al., 1990), and dextromethorphan N-demethylation for CYP3A4/5 (Jacqz-Aigrain et al., 1993; Gorski et al., 1994). The concentration of each probe drug used in each assay was chosen with the aim that the assay would be isozyme specific while providing enough sensitivity in the assay to allow facile detection of the isozyme-specific metabolite. Because it was our aim to have a battery of isozyme-specific assays and isozyme-selective inhibitors that were internally consistent, we tested each assay with one or more selective inhibitors. In addition, as many drugs are poorly soluble in water or buffer, solvents are frequently used to aid drug solubility in in vitro drug metabolism experiments. With the exception of the cytochrome P450 2E1 (Peter et al., 1990), there is little data regarding the consequences of the use of organic solvents on the cytochromes P450 even though concentrations of 1% solvent have been reported in previous inhibitor selectivity studies (Newton et al., 1995). Therefore, we also assessed the effect of the organic solvents (1% v/v) methanol, dimethylformamide (DMF), isopropanol, DMSO, acetone, and acetonitrile in these microsomal P450 assays.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. Caffeine, chlorzoxazone, coumarin, dextromethorphan, diethyldithiocarbamate (DDC), 8-methoxypsoralen, NADPH, 4-nitrophenol, quinidine, tolbutamide, troleandomycin (TAO), and umbelliferone (7-hydroxycoumarin) were purchased from Sigma. Furafylline, 4-hydroxy-S-mephenytoin, sulfaphenazole, and 4-methylhydroxytolbutamide were purchased from RBI (Natick, MA). Dextrophan tartrate, 3-hydroxymorphinan, and 3-methoxymorphinan were provided by Hoffmann-La Roche (Nutley, NJ). S-Mephenytoin was provided by Dr. Rene H. Levy, and 6-hydroxychlorzoxazone was provided by Dr. R. M. Peter. Other reagents were purchased from Sigma or JT Baker (Phillipsburg, NJ), and HPLC grade solvents were purchased from Baxter Healthcare (Muskegon, MI). Solvents for the inhibition experiments were purchased from Sigma (isopropanol, molecular biology grade 99+%, dimethyl sulfoxide, ACS reagent and N, N-dimethylformamide, HPLC grade 99.9+%), Baxter (methanol, B & J Brand), JT Baker (acetone, 99.4+%), and Fisher Scientific (acetonitrile, HPLC grade 99.9+%).

Isolation of Human Liver Microsomes. Human liver tissue was procured, prepared, and stored as described previously (Rettie et al., 1989). Human liver microsomes were prepared by differential centrifugation as described previously (Rettie et al., 1989) with the following modifications. Microsomes were prepared from 10-15 g of thawed tissue, which was homogenized in 5 volumes of 250 mM sucrose, 100 mM potassium phosphate, pH 7.4, 1mM EDTA, using a Virtis Vertishear probe homogenizer (The Virtis Company, Inc., Gardiner, NY) at 50% speed for 2 × 30 sec. The final microsome pellet was resuspended in homogenization buffer (without glycerol), and protein concentrations were estimated using the Bradford assay with bovine serum albumin standard as commercially available from Bio-Rad. The microsomes from three human livers, designated HL139 (15-year-old female with history of phenytoin therapy), HL140 (63-year-old male with a history of alcohol use), and HL141 (59-year-old male with a history of alcohol use who quit 14 years prior to tissue collection) were used in this study.

Microsomal Incubations for the Determination of Cytochrome P450 Monooxygenase Activities. Microsomal incubations were carried out in a total volume of 100 µl and in the presence of 1mM NADPH (freshly prepared). Microsomes (25 µl in 250 mM sucrose, 100 mM potassium phosphate and 1mM EDTA, pH7.4) were preincubated for 5 min at 37° with 100 mM potassium phosphate, 1mM EDTA buffer, pH 7.4 (55 µl) and substrate (10 µl) prepared in water with minimal use of methanol or DMF [caffeine (water), coumarin (0.01% methanol), tolbutamide (0.2% DMF), S-mephenytoin (0.1-0.375% methanol), dextromethorphan (2D6, 0.0025% methanol; 3A4/5, 0.25% methanol), and chlorzoxazone (final equivalent of aqueous 0.06 mM KOH)]. The reactions were started by the addition of 10 µl of freshly prepared 10 mM NADPH. Reactions were terminated by the addition of 10 µl of 2 M HCl and cooling on ice (5-60 min). Samples were then centrifuged (20,000g, 5 min, 4°), and the supernatant was injected directly onto the HPLC. The time of incubation and concentration of microsomal protein were chosen for each substrate such that rates of metabolite formation were demonstrated to be linear (table 1). Unless otherwise stated, the concentration of the substrate, the incubation time, and the protein concentration used in each assay were as specified in table 1. All microsomal incubations were conducted in triplicate. Controls included incubations without microsomes, without NADPH, or without substrate.

Determination of Michaelis-Menten Parameters for Cytochrome P450 Isozyme-Specific Activities. Enzyme velocity experiments were carried out as described above over a range of concentrations of the P450 substrates caffeine (50-30,000 µM), coumarin (0.25-10 µM), tolbutamide (12.5-800 µM), S-mephenytoin (10-150 µM), dextromethorphan-2D6 (1-50 µM), chlorzoxazone (5-500 µM), and dextromethorphan-3A4/5 (50-1200 µM). The parameters K_m and V_max, with standard errors, were estimated by fitting the Michaelis-Menten equation to the data using nonlinear regression analysis (PCnonlin). Initial estimates for nonlinear regression were chosen based on substrate concentration (S) verses reaction velocity (V) plots and Eadie-Hofstee plots (V/S verses V).

Inhibition of Specific P450 Isozymes by Selected Solvents. To investigate the inhibitory capacity of the solvents acetone, DMF, DMSO, isopropanol, methanol, and acetonitrile, microsomal incubations were conducted as above, except 10 µl of 10% solvent (v/v) in water (or water as control) was added to each preincubation (1% v/v final concentration in assay). The volume of 100 mM phosphate buffer with 1 mM EDTA was adjusted to 45 µl to maintain the final total volume of 100 µl.

Inhibition of Specific P450 Isozymes by Selected Compounds. The concentration of each inhibitor was selected such that the inhibition of each target P450 activity was estimated to be greater than or equal to 90% based on previous studies (Broly et al., 1989; Chang et al., 1994; Chiba et al., 1993; Kunze and Trager, 1993; Maenpaa et al., 1993; Miners et al., 1988; Peter et al., 1990; Tassaneeyakul et al., 1993). To investigate the inhibitory capacity of 8-methoxypsoralen (5 µM), sulfaphenazole (2.16 µM), S-mephenytoin (360 µM), quinidine (0.45 µM), 4-nitrophenol (540 µM), and chlorzoxazone (900 µM), microsomal incubations were conducted as above, except 10 µl of each inhibitor (at 10 times the final concentration) was added to each preincubation. The volume of 100 mM phosphate buffer with 1 mM EDTA was adjusted to 45 µl to maintain the final total volume of 100 µl. Furafylline (10 µl of 300 µM in 5% solvent⁴) was preincubated with microsomes (25 µl) and phosphate buffer (45 µl) for 3 min as described above except without substrate. NADPH (10 µl of 10 mM in water) was then added and incubated for a further 3 min prior to the addition of enzyme substrate (10 µl) to start the reaction. TAO (25 µl of 200 µM in 5% solvent⁴) was incubated with microsomes (125 µl), phosphate buffer (75 µl), and NADPH (25 µl of 10 mM in water) for a 30-min inactivation period at 37°. A 50 µl aliquot of this TAO/microsome incubation was added to a mixture of substrate (10 µl), NADPH (10 µl of 10 mM in water), and phosphate buffer (30 µl) to start the reaction. DDC was used in the same way as TAO, except DDC was soluble in water (25 mM), the concentration in the inactivation incubation was 25 µM, and the inactivation took place over 15 min. The concentration of solvent⁴ (v/v) in the final incubation was as follows: furafylline (0.5%), 8-methoxypsoralen (0.02%), coumarin (0.05%), sulfaphenazole (0.2%), S-mephenytoin (0.5%), quinidine (<0.01%), DDC (0%), chlorzoxazone (0%), p-nitrophenol (0.5%), and TAO (0.25%). Control incubations were conducted with the appropriate concentration of solvent.⁴

Analysis of Microsomal Incubates by HPLC. Isocratic HPLC was sufficient for all the separations described using a Waters 501 or 510 HPLC pump with either a Waters 712 WISP autosampler or Gilson model 323 bio sample injector. Reversed phase HPLC columns (Microsorb MV, Rainin Instrument Company, CA) were used with a 0.2 µm precolumn filter (Upchurch Scientific, Oak Harbor, WA) for all the assays described below. The mobile phase conditions used to separate each substrate from the metabolite of interest without interference resulting from the other components or metabolites in the assay were as follows. The product of N3-demethylation of caffeine, 1,7-dimethylxanthine (17X), was separated from caffeine, 1,7-dimethyluric acid, 1,3-dimethylxanthine, and 3,7-dimethylxanthine using a C18 column (4.6 × 250 mm, 5-µm particle size) and methanol:water:acetic acid:triethylamine 12:88:1:0.02 (v/v) as mobile phase at 1 ml·min^-1. A linear gradient to methanol:water 50:50 (v/v) over 1 min held for 1 min with a linear gradient back to mobile phase over 1 min was used to hasten the elution of the substrate, caffeine, after the elution of 17X. The product of 7-hydroxylation of coumarin was separated from coumarin using a C18 column (4.6 × 100 mm, 3-µm particle size) and methanol:20mM sodium phosphate buffer (pH 4.4) 27:73 (v/v) as mobile phase at 0.8 ml·min^-1. The product of the hydroxylation of tolbutamide was separated from tolbutamide using a C8 column (4.6 X 250 mm, 5-µm particle size) and acetonitrile:water:acetic acid:triethylamine 35:65:1:0.02 (v/v) as mobile phase at 1ml·min^-1. The tolbutamide HPLC assay conditions were also used for the separation of 4-hydroxy-S-mephenytoin from S-mephenytoin and the separation of the dextromethorphan metabolites, 3-hydroxymorphinan, dextrophan, and 3-methoxymorphinan from dextromethorphan. Chlorzoxazone was separated from 6-hydroxychlorzoxazone using a C8 column (4.6 × 250 mm, 5-µm particle size) and acetonitrile:water:acetic acid:triethylamine 30:70:1:0.02 (v/v) at 1ml·min^-1. Analytes were detected using variable wavelength absorbance (Waters 481 or 484 detector) for caffeine (270 nm), coumarin (324 nm), tolbutamide (240 nm), S-mephenytoin (240 nm), and chlorzoxazone (297nm). Dextromethorphan and its metabolites were detected by fluorescence (Hewlett Packard HP1046A programmable fluorescence detector) using excitation and emission wavelengths of 235 nm and 310 nm, respectively, with a 280 nm filter. In each case, data were recorded using Waters Maxima HPLC analysis software. Quantitation of metabolites in unknown samples was possible by comparison of peak area with calibration lines generated by injections of authentic standards.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

We present data suggesting that a system of seven selective P450 assays and several selective inhibitors are suitable for use in in vitro metabolism studies directed at the prediction of metabolic drug-drug interactions involving P450 isozymes. We have determined suitable incubation conditions for the following enzyme activities: caffeine N3-demethylation (1A2), coumarin 7-hydroxylation (2A6), tolbutamide 4-methyl-hydroxylation (2C8/9), S-mephenytoin 4-hydroxylation (2C19), dextromethorphan O-demethylation (2D6), chlorzoxazone 6-hydroxylation (2E1), and dextromethorphan N-demethylation (3A4/5). Each enzyme activity was characterized in terms of linearity with respect to time and with respect to protein concentration. The reaction times, protein concentration used, and the Michaelis-Menten parameters determined for each isozyme activity in the human liver microsomes are summarized in table 1. The effects of six organic solvents at a final concentration of 1% (v/v) on each enzyme activity compared with activity in the absence of the additional solvent were then determined, and these results are summarized in fig. 1. Using this information, we then investigated the selectivity of the inhibitors, furafylline (1A2), 8-methoxypsoralen (2A6), coumarin (2A6), sulfaphenazole (2C9), S-mephenytoin (2C19), quinidine (2D6), DDC (2E1), p-nitrophenol (2E1), chlorzoxazone (2E1), and TAO (3A4) on each P450 activity. These results are summarized in fig. 2.

View larger version (49K): [in this window] [in a new window]   Fig. 1.   The effect of several organic solvents on the activity of P450-mediated biotransformations. The effect of acetone, DMF, DMSO, isopropanol, methanol, and acetonitrile on caffeine N3-demethylation (A), coumarin 7-hydroxylation (B), tolbutamide 4-methyl-hydroxylation (C), S-mephenytoin 4-hydroxylation (D), dextromethorphan O-demethylation (E), chlorzoxazone 6-hydroxylation (F), and dextromethorphan N-demethylation (G) was determined by coincubation of each solvent (1% v/v) with human liver microsomes, 1 mM NADPH, and caffeine (1000 µM), coumarin (5 µM), tolbutamide (100 µM), S-mephenytoin (75 µM), dextromethorphan (5 µM), chlorzoxazone (50 µM), or dextromethorphan (500 µM), respectively. All activities were measured under initial rate conditions and compared with incubations in the absence of additional solvent. Each bar represents the mean of experiments conducted on microsomes prepared from three different livers that have been expressed individually as a percent of control activity. The error bars indicate standard deviations. Due to substantial activation of caffeine N3-demethylation, data for this activity are presented on an expanded y axis compared with the other solvent inhibition experiments.

View larger version (54K): [in this window] [in a new window]   Fig. 2.   The effect of several selective P450 inhibitors on the activity of P450-mediated drug biotransformations. The effect of inhibitors furafylline (30 µM preincubation for 3 min), 8-methoxypsoralen (5 µM), sulfaphenazole (2.16 µM), S-mephenytoin (360 µM), quinidine (0.45 µM), DDC (25 µM preincubation for 15 min), p-nitrophenol (540 µM), and TAO (20 µM preincubation for 30 min) on caffeine N3-demethylation (A), coumarin 7-hydroxylation (B), tolbutamide 4-methyl-hydroxylation (C), S-mephenytoin 4-hydroxylation (D), dextromethorphan O-demethylation (E), chlorzoxazone 6-hydroxylation (F), and dextromethorphan N-demethylation (G) was determined by coincubation of each inhibitor (as described in Materials and Methods) with human liver microsomes, 1 mM NADPH, and caffeine (1000 µM), coumarin (5 µM), tolbutamide (100 µM), S-mephenytoin (75 µM), dextromethorphan (5 µM), chlorzoxazone (50 µM), or dextromethorphan (500 µM), respectively. All activities were measured under initial rate conditions and compared with control incubations in the absence of inhibitor. Each bar represents the mean of experiments conducted on microsomes prepared from three different livers that have been expressed individually as a percent of control activity. The error bars indicate standard deviations. NA, not applicable or not determined due to interference of the assay by inhibitor or metabolites of the inhibitor; ND, not detectable. * denotes a large standard deviation due to an unrepresentative response of dextromethorphan O-demethylation activity to both chlorzoxazone and p-nitrophenol in HL140 or inter-liver variability in the inhibition of dextromethorphan N-demethylation by troleandomycin.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In vitro metabolic probes for each of the principle drug-metabolizing P450 microsomal oxidases have been chosen so that at a suitable concentration its conversion to the metabolite of interest is mediated by a single form, or in some cases two closely related forms, of these microsomal enzymes. Where possible, we have chosen substrate concentrations close to that of the K_m for the P450 isozyme of interest for subsequent inhibition studies. This is in contrast to the evaluation by Newton and colleagues (1995) where the concentration of the P450 substrates were 5-10 times the K_m as in, for example, phenacetin (100 µM), tolbutamide (500 µM), chlorzoxazone (500 µM), bufuralol (100 µM), and testosterone (250 µM). Although it can be argued that concentrations approaching V_max (as used by Newton) offer certain advantages, such as increased sensitivity of assay and reduced effect of K_m on apparent velocity, V_max concentrations are less sensitive to inhibition by both specific and inadvertant inhibitors (mechanism-based inhibitors excepted). In this study, a concentration approximately equal to the K_m was chosen as a practical compromise between sensitivity of the assay for measurement of the metabolite (higher concentrations allow for more reliable detection of metabolite) and both selectivity of the P450 assay for a single isozyme and sensitivity of the enzyme velocity toward inhibition by selected compounds (concentrations below K_m preferred). Each specific P450 activity (and its selective inhibitor) is discussed in turn.

Caffeine N3-Demethylation (1A2). Caffeine N3-demethylation is a relevant in vitro measure of cytochrome P450 1A2 activity (Butler et al., 1989; Gu et al., 1992; Tassaneeyakul et al., 1992). We observed that caffeine N3-demethylation (50-30,000 µM caffeine) followed biphasic kinetics in human liver microsomes, which is in agreement with several previous studies (Campbell et al., 1987; Fuhr et al., 1992; Gu et al., 1992; Tassaneeyakul et al., 1992). The K_m estimates for the low affinity component were 20.5 mM and 22.3 mM for microsomes prepared from HL139 and HL141, respectively. Furthermore, this low affinity component was effectively abolished by preincubation of the human liver microsomes with TAO (as described in Materials and Methods), which is in agreement with a previous study indicating that P450 3A4 is responsible for this low affinity activity (Tassaneeyakul et al., 1992). Using a two-enzyme Michaelis-Menten model, we calculated that this low affinity enzyme (3A4) would be responsible for less than 5% of caffeine N3-demethylation at a caffeine concentration of 1000 µM, which is the estimated K_m of the high affinity enzyme (table 1). The preincubation of the human liver microsomes with furafylline, a specific mechanism-based inhibitor of 1A2 (Kunze and Trager, 1993; Newton et al., 1995; Sesardic et al., 1990; Tassaneeyakul et al., 1994), confirmed this prediction (fig. 2A). In addition, this result indicated a minimal role for the related isozyme 1A1 in caffeine N3-demethylation in the three livers tested, as 1A1 is less sensitive to inhibition by furafylline (Tassaneeyakul et al., 1994). Furthermore, furafylline preincubation did not inhibit any of the other six P450 activities investigated here by more than 10% compared with controls (fig. 2, B-G) and thus proved to be a specific inhibitor for 1A2 under the conditions used. This is in contrast to a previous study where incubations with higher, and presumably less selective, concentrations of furafylline (200 µM) indicated that furafylline was not a specific inhibitor of 1A2 (Ono et al., 1996).

Coumarin 7-Hydroxylation (2A6). Coumarin 7-hydroxylation is catalyzed by CYP2A6 (Maurice et al., 1991; Miles et al., 1990; Yamano et al., 1990). This reaction followed monophasic kinetics consistent with the involvement of a single enzyme (0.25-10 µM coumarin). The mean estimates of K_m and V_max for coumarin 7-hydroxylation (table 1) were similar to the 2.3 µM K_m and the 0.03-1.35 nmol/min/mg protein V_max estimated in human liver microsomes by Yamano and colleagues (1990). Due to limitations in assay sensitivity, we used 5 µM coumarin for examining the effect of solvents and inhibitors on coumarin 7-hydroxylation activity so that we could reliably detect at least 10% of the control activity.

Tolbutamide 4-Methyl-Hydroxylation (2C8/9). The hydroxylation of tolbutamide at the 4-methyl position is mediated by cytochrome P450s of the 2C subfamily (Miners et al., 1998). Specifically, tolbutamide has been shown to be a substrate for the expressed human P450s 2C8, 2C9, and 2C10 (Brian et al., 1989; Relling et al., 1990; Srivastava et al., 1991; Veronese et al., 1993). Tolbutamide hydroxylation in human liver microsomes prepared from HL139-141 (12.5-800 µM tolbutamide) was consistent with the involvement of a single enzyme or more than one enzyme with a similar K_m as previously indicated by Miners and colleagues (1988). Our estimates of K_m (133-348 µM) and V_max (0.387-0.434 nmol/min/mg protein) are also in agreement with this previous study.

S-Mephenytoin 4-Hydroxylation (2C19). The 4-hydroxylation of S-mephenytoin is mediated by cytochrome P450 2C19 (Goldstein et al., 1994). Substrate concentration velocity experiments indicated that S-mephenytoin 4-hydroxylation (10-150 µM S-mephenytoin) was mediated by a single enzyme in human liver microsomes prepared from HL139 and HL141. Our estimates of K_m were 29.5 and 75.8 µM, and our estimates of V_max were 0.066 and 0.055 nmol/min/mg protein, respectively, for the two livers tested. We did not determine these values for HL140 due to low turnover of S-mephenytoin in this liver microsome preparation. S-Mephenytoin 4-hydroxylation was inhibited by the solvents (1% v/v) isopropanol>DMF>DMSO>acetone> methanol>acetonitrile (fig. 1D). Methanol, which inhibited this activity less than 20% compared with the control (75 µM S-mephenytoin), was chosen where solvents were necessary for further experiments. In the absence of any putative specific inhibitors of P450 2C19 other than S-mephenytoin itself and because a direct measurement of inhibition of S-mephenytoin 4-hydroxylation by itself is not possible, we assumed that K_m was equal to K_i for this enzyme and chose 360 µM S-mephenytoin as our 2C19 inhibitor. This concentration was predicted, using the Michaelis-Menten equation in the presence of a competitive inhibitor, to inhibit 2C19 activity by 85 and 65% in HL139 and HL141, respectively, if the substrate was present at its K_m concentration. S-Mephenytoin (360 µM) inhibited P450 2C9 by 22% compared with control but inhibited the other P450-specific assays (1A2, 2A6, 2D6, 2E1, and 3A4/5) only marginally (less than 10% inhibition) or not at all.

Dextromethorphan O-Demethylation (2D6). Dextromethorphan O-demethylation activity correlates well with debrisoquin 4-hydroxylase (Kupfer et al., 1986; Schmid et al., 1985) and is a good in vitro probe activity for P450 2D6 (Gorski et al., 1994; Jacqz-Aigrain et al., 1993). Substrate concentration velocity experiments indicated that dextromethorphan O-demethylation activity (1-50 µM) was mediated by a single enzyme in human liver microsomes prepared from HL139 through 141. Our estimates of K_m ranged from 2.2 to 8.5 µM and V_max from 0.097 to 0.566 nmol/min/mg protein, which were consistent with several previous studies [K_m, 7.5 µM; V_max, 0.07 nmol/min/mg protein (Jacqz-Aigrain et al., 1993); K_m, 3.4 µM; V_max, 0.17 nmol/min/mg protein (Dayer et al., 1989); K_m, 4.6 µM; V_max, 0.07 nmol/min/mg protein (Broly et al., 1989)]. Dextromethorphan O-demethylation activity was inhibited by the solvents (1% v/v) DMF>acetonitrile>acetone (fig. 1E) and moderately increased by the solvents (1% v/v) isopropanol >DMSO. Methanol at 1% (v/v) had, on average, no effect on dextromethorphan O-demethylation (5 µM substrate) and was used in subsequent experiments. Quinidine is a potent and selective inhibitor of P450 2D6 (Broly et al., 1989). Based on a previously published K_i of 0.025 µM (Broly et al., 1989), we used 0.45 µM (18 times K_i) for our inhibition experiments. As predicted for a 2D6 activity with substrate concentration at the K_m, we observed 90% inhibition of dextromethorphan O-demethylation (fig. 2E). Quinidine (0.45 µM) was also a selective inhibitor of 2D6 inhibiting the P450s 1A2, 2A6, 2C9, 2C19, 2E1, and 3A4/5 marginally (less than 20%) or not at all in our three liver microsome preparations.

Chlorzoxazone 6-Hydroxylation (2E1). Chlorzoxazone 6-hydroxylation activity is a relevant measure of P450 2E1 activity (Peter et al., 1990). Substrate concentration velocity experiments indicated that this activity (5-500 µM chlorzoxazone) was mediated by a single enzyme in human liver microsomes prepared from HL139 through 141. Our estimates of K_m ranged from 51 to 64 µM and V_max from 2.75 to 5.59 nmol/min/mg protein, which are consistent with previous studies [K_m, 22-49 µM; V_max, 1.1-5.9 nmol/min/mg protein (Peter et al., 1990)]. Chlorzoxazone 6-hydroxylation activity was inhibited by the solvents (1% v/v) isopropanol>DMSO>DMF>acetone> methanol>acetonitrile (fig. 1F). Methanol (1% v/v), which inhibited this activity approximately 55% compared with the control (50 µM substrate), was used in subsequent experiments as methanol (0.1% v/v) only marginally inhibited chlorzoxazone 6-hydroxylation (10%) compared with the control (data not shown). However, given the apparent tolerance of chlorzoxazone 6-hydroxylation activity to acetonitrile, this solvent may prove a useful alternative to methanol in future studies involving P450 2E1.

Dextromethorphan N-Demethylation (3A4/5). Dextromethorphan N-demethylation activity is a selective in vitro probe activity for P450 3A4/5 (Gorski et al., 1994; Jacqz-Aigrain et al., 1993). Substrate concentration velocity experiments indicated that this activity (50-1200 µM dextromethorphan) was mediated by a single enzyme in human liver microsomes prepared from HL139 through HL141. Our estimates of K_m ranged from 133 to 369 µM and V_max from 0.33 to 2.89 nmol/min/mg protein, which compared with previous studies [K_m, 520-710 µM; V_max, 0.375-0.812 nmol/min/mg protein (Gorski et al., 1994); 450-830 µM, 0.078-147 nmol/min/mg protein (Jacqz-Aigrain et al., 1993)]. Dextromethorphan N-demethylation activity was inhibited by the solvents (1% v/v) isopropanol>acetone= DMF = methanol>acetonitrile>DMSO (fig. 1G). Methanol (1% v/v), which inhibited this activity less than 20% compared with the control (500 µM substrate), was used in subsequent experiments.