Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Bupropion hydrochloride is an antidepressant and a non-nicotine aid to smoking cessation that acts by weakly inhibiting the reuptake of dopamine and norepinephrine (Cooper et al., 1994). It is prescribed instead of other antidepressants to patients who have failed to respond to or have not tolerated other agents (Walker et al., 1993). There is evidence that bupropion and selective serotonin reuptake inhibitor (SSRI)¹ combination therapy is more effective for treatment of refractory depression than the use of either agent alone (Bodkin et al., 1997; Nelson, 1998). In addition, bupropion may be used to treat attention-deficit/hyperactivity disorder when other agents are not effective (Cantwell, 1998).

Recently the FDA approved the use of sustained release bupropion (Zyban; Glaxo Wellcome, Research Triangle Park, NC) as an anti-smoking agent. In a clinical trial, bupropion was more effective than placebo in smoking cessation (Hurt et al., 1997). In another trial, there was a significant improvement in cessation of smoking with bupropion alone or bupropion combined with the nicotine patch compared with the nicotine patch alone or placebo (Jorenby et al., 1999). Since there are 50 million smokers in the United States and almost half are trying to quit each year, the use of bupropion as an anti-smoking agent is expected to increase (Jorenby et al., 1999).

Bupropion may induce seizures with high doses or in patients with a predisposition to seizures (Davidson, 1989). There is a 0.1% incidence of seizures with doses up to 300 mg per day of sustained release bupropion; this incidence increases to 0.4% with doses up to 450 mg per day of the immediate release formulation (Dunner et al., 1998). It has been hypothesized that the seizures may be due to high concentrations of bupropion or a metabolite (Preskorn, 1991).

In humans, bupropion is extensively metabolized to three principal metabolites (Fig. 1): hydroxybupropion (morphinol), erythrohydrobupropion, and threohydrobupropion (Schroeder, 1983; Golden et al., 1988; Preskorn, 1991). The pharmacologically active metabolite hydroxybupropion appears to be the major metabolite, since plasma levels of hydroxybupropion greatly exceed those of the parent drug (Golden et al., 1988). Product labeling information and data reported in abstract form indicate that the cytochrome P450 (CYP) enzyme system, especially CYP2B6, has an important role in bupropion hydroxylation (Wurm et al., 1996; Faucette et al., 2000; Lindley et al., 2000). Product labeling also indicates that bupropion or hydroxybupropion inhibits CYP2D6.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   The chemical structures of bupropion and its major in vivo metabolites: hydroxybupropion, threohydrobupropion, and erythrohydrobupropion.

The present study has examined the in vitro hydroxylation of bupropion by the CYP enzyme system. CYP2B6 is identified as having the major role in hydroxybupropion formation. In addition, we investigated the possibility of CYP2D6 inhibition by bupropion or hydroxybupropion. Finally, we studied the in vitro effects of several newer antidepressants on bupropion hydroxylation.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Bupropion hydrochloride and hydroxybupropion were kindly provided by Glaxo-Wellcome (Research Triangle Park, NC), and trazodone was provided by Mead Johnson (Evansville, IN). S-Mephenytoin, nirvanol, inhibitory antibody to CYP2B6 (catalog no. A326), and the Western blotting anti-peptide antibody to CYP2B6 (catalog no. A143) were purchased from Gentest Corp. (Woburn, MA). Chemical inhibitors, antidepressants and their metabolites, and other chemical reagents were provided by their manufacturers or purchased from commercial sources. NADP⁺, isocitrate dehydrogenase, DL-isocitrate, and 50 mM potassium phosphate buffer (pH 7.5) were purchased from Sigma (St. Louis, MO).

Incubations Using Human Liver Microsomes. Solutions of bupropion hydrochloride, hydroxybupropion, chemical inhibitors, and other antidepressants were prepared in methanol. Varying amounts of bupropion were added to the incubation tubes to yield final concentrations that ranged from 0 to 1000 µM. The solvent was evaporated to dryness in a 45°C vacuum oven before the addition of cofactors. Samples using human liver microsomes were incubated in duplicate. Incubation mixtures contained 50 mM potassium phosphate buffer (pH 7.5), 0.5 mM NADP⁺, an isocitrate/isocitric dehydrogenase regenerating system, and 5 mM MgCl₂. The samples were preincubated in a water bath at 37°C for 2 to 3 min. Reactions were initiated by the addition of microsomal protein, and the final volume was 0.25 ml. A protein concentration of 0.25 mg/ml was used for human liver microsomal incubations. Incubations were performed in a shaking water bath for 20 min at 37°C and terminated by addition of 50 µl of 1 N HCl. Trazodone (10-25 µl of 125 µM) was added as an internal standard. The mixture was vortex mixed and spun at 16,000g for 10 min. Supernatants were injected into the HPLC system for analysis.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   A representative HPLC chromatogram of in vitro biotransformation of bupropion. Hbup, hydroxybupropion; Bup, bupropion; IS, internal standard (trazodone).

Incubations Using Microsomes Containing cDNA-Expressed CYPs. A screen of microsomes from human lymphoblastoid cells expressing CYP 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, or transfected with vector, was performed using the incubation method described above. Bupropion concentrations were 50 and 500 µM, and cDNA-expressed microsomal protein concentration was 1 mg/ml. Hydroxybupropion formation rates were normalized to picomoles of CYP. The cDNA-expressed microsomal incubations were performed with little agitation and mild shaking, as recommended by the supplier. A kinetic curve was generated by incubating microsomes (1 mg/ml) containing CYP2B6 with varying concentrations of bupropion (0-1000 µM). CYP2B6 was also incubated with bupropion (50 µM) alone or with coaddition of ketoconazole (2.5 µM).

Inhibition of Human Liver Microsomes with CYP2B6 Inhibitory Antibody. Inhibition studies with the CYP2B6 inhibitory antibody were performed by preincubating human liver microsomal protein (62-70 µg) with antibody (15-µl total volume) for 15 min on ice. To determine the optimal amount of antibody to achieve maximal inhibition, the concentration of bupropion (100 µM) and the amount of protein (70 µg) were held constant while the amount of antibody was varied (0-20 µg). Twenty micrograms of CYP2B6 inhibitory antibody and a bupropion concentration of 50 µM were used in the subsequent CYP2B6 inhibitory antibody studies. The reaction was started with a 235-µl addition of a mixture of substrate, cofactor, buffer, and bupropion. The incubations were performed with minimal agitation for 20 min, and 50 µl 1 N HCl and internal standard (trazodone) were added. Samples were processed as before for HPLC analysis.

Western Blotting. Microsomal protein [10-50 µg of human liver microsomes, and 0.05-2.5 pmol of lymphoblast-expressed CYP2B6 (Gentest Corp.)] was denatured for 5 min in 100 mM Tris buffer containing 10% glycerol, 2% -mercaptoethanol, 2% SDS, and 5 µg/ml pyronin Y (pH 6.8) at 100°C. Protein was separated by SDS-polyacrylamide gel electrophoresis in precast 7.5% polyacrylamide gels (Bio-Rad, Hercules, CA). Samples were run at 100 V for 1 h in 25 mM Tris buffer/0.192 M glycine/0.1% SDS buffer. Then samples were transferred to Immobilon-P (polyvinylidene difluoride membrane) (Millipore, Bedford, MA) for 1 h at 100 V in 25 mM Tris buffer/20% methanol. Blots were blocked with 0.5% powdered nonfat milk in TBS-Tween (0.15 M NaCl, 0.04 M Tris-HCl, pH 7.7, and 0.06% Tween 20) for 1 h at room temperature. Blots were then incubated with a 1:500 dilution of a polyclonal anti-peptide CYP2B6 antibody (Stresser and Kupfer, 1999) (Gentest Corp.) in TBS-Tween containing 0.1% BSA for 1 h at room temperature. After washing in TBS-Tween, blots were incubated with a 1:500 dilution of horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (Gentest Corp.) in 0.5% powdered nonfat milk in TBS-Tween (Gentest Corp.) for 1 h at room temperature. Blots were rinsed with TBS-Tween, and the Super Signal Cl-HRP Substrate System (Pierce) was used for enhanced chemiluminescence detection. Blots were exposed to film (Fig. 6A). Quantitation of CYP2B6 content was completed via computer image analysis (NIH Image 1.62 software). A standard curve of pixel area × density versus picomoles of CYP2B6 was created and fit to the equation y = m × ln(x) + A using nonlinear least-squares regression (Fig. 6A).

Data Analysis. The formation of hydroxybupropion by human liver microsomes and cDNA-expressed CYP2B6 were consistent with a one-enzyme Michaelis-Menten model. Data points were fitted to this equation using nonlinear regression (Sigma Plot software; SPSS Inc., Chicago, IL), yielding values of V_max and K_m.

(1)

(2)

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Estimated V_max and K_m values for hydroxybupropion formation by human liver microsomes (Fig. 3A) are shown in Table 1. Bupropion hydroxylation was reduced to 62% of control by 1.0 µM ketoconazole and 51% of control by 2.5 µM ketoconazole. Omeprazole, sulfaphenazole, and quinidine had minimal effect (Fig. 4A). Incubation of heterologously expressed CYP2B6 with 2.5 µM ketoconazole reduced reaction velocities to 68% of control at 50 µM bupropion. This indicates that the modest degree of inhibition by ketoconazole in liver microsomes is attributable to its effect on CYP2B6.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   The velocity of formation of hydroxybupropion versus the concentration of bupropion (µM) for four different human livers (A) (velocity units = pmol of hydroxybupropion formed/min/mg of protein) and microsomes from lymphoblastoid cells expressing CYP2B6 (B) (velocity units = pmol of hydroxybupropion formed/min/pmol of CYP2B6). Symbols are experimental data points. Lines are fitted functions described by a Michaelis-Menten equation. See Table 1 for the kinetic parameters.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   A, the inhibition of bupropion hydroxylation activity in human liver microsomes from four different donors; B, percentage of control activity of bupropion hydroxylation versus amount of inhibitory antibody (µg) using 70 µg of human liver microsomal protein. A, activity values from incubations with inhibitor were divided by activity values from control incubations (no inhibitor) and are represented as a percentage of control activity (mean ± S.E.). The bupropion concentration was 50 µM. [KET, ketoconazole, 1.0 µM (solid bar) and 2.5 µM (striped bar); QUIN, quinidine, 5 µM; OME, omeprazole, 25 µM; SPA, sulfaphenazole, 10 µM; and anti-2B6, inhibitory antibody to CYP2B6.] B, percentage of control activity of bupropion hydroxylation versus amount of inhibitory antibody (µg) using 70 µg of human liver microsomal protein. A bupropion concentration of 100 µM was used.

According to the supplier, the anti-CYP2B6 antibody does not inhibit human CYPs 1A1, 1A2, 1B1, 2A6, 2C8, 2C9, 2C18, 2C19, 2D6, 2E1, or 3A. We determined the optimal amount of antibody to achieve maximal inhibition of CYP2B6 (Fig. 4B). The anti-CYP2B6 inhibitory antibody produced almost complete inhibition of hydroxybupropion formation (Fig. 4A).

Among cDNA-expressed CYPs, hydroxybupropion was formed only by CYP2B6 at 50 µM bupropion. At 500 µM bupropion, hydroxybupropion was formed by both CYP2B6 and CYP2E1, although the formation rate with CYP2B6 was nearly 80-fold greater than that with CYP2E1. CYP 1A2, 2A6, 2C8, 2C9, 2C19, 2D6, 3A4 or microsomes from vector-transfected cells showed no detectable activity with a limit of detection of 5 ng per sample. The formation of hydroxybupropion by heterologously expressed CYP2B6 was consistent with single-enzyme Michaelis-Menten kinetics (Fig. 3B; Table 1).

In a random sampling of 12 liver samples, CYP2B6 content ranged from 1.5 to 148.4 pmol of CYP2B6/mg of protein, with a median value of 44.6 pmol of CYP2B6/mg of protein (Fig. 5A). Velocities of bupropion hydroxylation in this subset were significantly correlated with immunochemically determined CYP2B6 content (r = 0.96) (Fig. 5B). Bupropion hydroxylation velocity among several different human livers samples (Fig. 6) was significantly correlated with velocities of N-demethylation of S-mephenytoin, a relatively specific CYP2B6-mediated reaction, in the same samples (r = 0.94) (Fig. 7). With the high outlying value excluded, the correlation coefficient increases to 0.97. 

View larger version (32K): [in this window] [in a new window]   Fig. 5.   A, immunoquantification of CYP2B6 content (representative blots); B, correlation plot of the velocity of bupropion hydroxylation for 12 different human livers (250 µM bupropion) with the amount of CYP2B6 as determined by immunoquantification using the same 12 livers. A, lanes 1 to 5 were loaded with 0.1, 0.25, 0.5, 1.0, and 2.5 pmol of CYP2B6, respectively. Lanes 6 to 9 were loaded with 10 µg of microsomal protein from four different human livers with relatively high bupropion hydroxylation activity. Lanes 14 to 19 were loaded with 2.5, 1.0, 0.5, 0.25, 0.1, and 0.05 pmol of CYP2B6, respectively. Lanes 10 to 13 were loaded with 50 µg of human microsomal protein from four different human livers with relatively low bupropion hydroxylation activity. The apparent altered mobility of CYP2B6 in microsomal samples in lanes 10 to 13 when compared with lanes 14 to 19 may be due to the larger amount of total protein loaded onto lanes 10 through 13. B, correlation plot of the velocity of bupropion hydroxylation for 12 different human livers (250 µM bupropion) with the amount of CYP2B6 as determined by immunoquantification using the same 12 livers.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Rates of bupropion hydroxylation in vitro by human liver microsomes from 36 different donors at 250 µM bupropion. Velocity units are picomoles of hydroxybupropion formed per minute per milligram of protein.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Correlation plot of the velocity of bupropion hydroxylation versus the velocity of N-demethylation of S-mephenytoin for 32 different human livers of the same liver panel.

Mean IC₅₀ values for bupropion and hydroxybupropion versus dextrorphan formation from dextromethorphan (25 µM) were 58 and 74 µM (Fig. 8; Table 2). Among antidepressants tested as potential inhibitors of bupropion hydroxylation, paroxetine was the most potent inhibitor (IC₅₀ = 1.6 µM). Sertraline, norfluoxetine, and fluvoxamine also had significant inhibitory potency, while desmethylsertraline, fluoxetine, and nefazodone were less active as inhibitors (Fig. 9, A and B; Table 3). Other antidepressants tested at 100-µM concentrations were weak inhibitors. Bupropion hydroxylation velocity was reduced to 92 ± 2% of control by venlafaxine, 60 ± 5% of control by O-desmethylvenlafaxine, 81 ± 1% by citalopram, and 68 ± 1% by desmethylcitalopram.

View larger version (14K): [in this window] [in a new window]   Fig. 8.   Percentage of control activity versus concentration of inhibitor (bupropion or hydroxybupropion) for CYP2D6 index reaction (dextrorphan formation from dextromethorphan). Hbup, hydroxybupropion; Bup, bupropion. See Table 2 for IC50 values.

View larger version (20K): [in this window] [in a new window]   Fig. 9.   A, percentage of control activity for bupropion hydroxylation versus concentration of inhibitor; B, percentage of control activity for bupropion hydroxylation versus concentration of inhibitor. F, fluoxetine; NorF, norfluoxetine; Nef, nefazodone; Fluv, fluvoxamine; S, sertraline; DMS, desmethylsertraline; P, paroxetine. See Table 3 for IC50 values.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

CYP2B6 is the primary enzyme mediating the formation of hydroxybupropion from bupropion in human liver microsomes. CYP2E1 may make a very small contribution at high concentrations of bupropion, but this contribution is unlikely to be of clinical importance. A C_max of 0.6 µM was reported after a single 150-mg tablet of sustained-release bupropion hydrochloride (Hsyu et al., 1997), and at this concentration, CYP2B6 would be the dominant enzyme mediating hydroxylation. The mean K_m for hydroxybupropion formation in liver microsomes is 89 µM, which is close to the K_m for hydroxybupropion formation by cDNA-expressed CYP2B6 (85 µM). Bupropion hydroxylation and S-mephenytoin N-demethylation activities among individual liver samples were highly correlated, as were immunoquantified CYP2B6 in human livers and bupropion hydroxylation. Our results are consistent with results reported in abstract form, indicating that bupropion hydroxylation is a valid CYP2B6 probe (Lindley et al., 2000). These authors also found a high correlation between bupropion hydroxylation activity and CYP2B6 content (r² = 0.99) and between bupropion hydroxylation and S-mephenytoin N-demethylation activities (r² = 0.98).

The manufacturer of bupropion found cDNA-expressed CYP3A4 to form detectable levels of hydroxybupropion (Wurm et al., 1996). Since CYP3A4 is more abundant than CYP2B6 in human livers, a minor role of CYP3A4 in bupropion hydroxylation could be clinically significant. However, in agreement with our results, Faucette et al. (2000) found that CYP3A4 has no significant contribution to bupropion hydroxylation because of the poor correlation of hydroxylation activity with both immunoquantified CYP3A4 content and with testosterone 6 -hydroxylation activity.

Since the anti-CYP2B6 antibody inhibits bupropion hydroxylation almost completely at a substrate concentration less than the K_m, use of this antibody probably represents the most valid and specific approach for studies requiring inhibition of CYP2B6 activity. Although orphenadrine has been proposed as a chemical inhibitor of CYP2B6, Guo et al. (1997) showed orphenadrine is nonspecific and also inhibits CYP2D6, CYP1A2, CYP2A6, CYP3A4, and CYP2C19 at high concentrations. Omeprazole, sulfaphenazole, and quinidine produced minimal inhibition of bupropion hydroxylation. Ketoconazole at 1.0 and 2.5 µM produced measurable inhibition of the reaction in both liver microsomes and heterologously expressed CYP2B6, confirming that ketoconazole is not fully specific for CYP3A.

There are many substrates biotransformed partially by CYP2B6 in vitro, but few relatively specific CYP2B6 substrates have been identified, since the role of this enzyme in drug metabolism is not fully characterized (Mimura et al., 1993; Ekins and Wrighton, 1999; Gervot et al., 1999; Hanna et al., 2000). Identified substrates include S-mephobarbital (Kobayashi et al., 1999), S-mephenytoin (Ko et al., 1998), cyclophosphamide (Chang et al., 1993), and RP73401 (Stevens et al., 1997; Domanski et al., 1999). At the present time, it is not known whether CYP2B6 is involved in the biotransformation of these substrates in vivo. Heyn et al. (1996) identified N-demethylation of S-mephenytoin as a reasonable probe for CYP2B6 activity. Ko et al. (1998) showed this reaction also has a high-affinity/low-capacity component mediated by CYP2C9 and demonstrated it must be used at high concentrations to be used as a CYP2B6 probe. Recently, Kobayashi et al. (1999) showed that N-demethylation of S-mephobarbital is mediated mainly by CYP2B6, by using chemical inhibition and cDNA-expressed enzymes. They used 100 and 300 µM orphenadrine to inhibit CYP2B6 activity to 47 and 29% of control activity.

Earlier immunoquantification studies suggested that CYP2B6 comprised less than 1% of total hepatic P450, and CYP2B6 was concluded to be a minor CYP isoform (Shimada et al., 1994). Using more specific antibodies to CYP2B6, several research groups have immunoquantified CYP2B6 in panels of livers and observed highly variable expression levels. Code et al. (1997) detected CYP2B6 in 12 of 17 human livers, ranging from 0 to 74 pmol of CYP2B6/mg of protein. Ekins et al. (1998) detected 0.7 to 7.1 pmol of CYP2B6/mg of protein in a human liver panel. Stresser and Kupfer (1999) developed a polyclonal antibody that recognized 20 residues of human CYP2B6 and detected 2 to 82 pmol of CYP2B6/mg of protein in a human liver panel. Immunoquantification results from these groups suggest that CYP2B6 may not be a minor human hepatic enzyme. Highly variable expression levels suggest variability of bupropion metabolism. Since high concentrations of bupropion and its metabolites are associated with toxicity (Preskorn, 1991), a very low or very high amount of CYP2B6 may increase the risk of toxicity.

We used the same polyclonal antibody as did Stresser and Kupfer (1999) and have likewise found high variability of CYP2B6 content in 12 different human livers (2.5-148.4 pmol of CYP2B6/mg of protein). Previous results from a study of variability of propofol hydroxylation in our laboratory support data that CYP2B6 expression is highly variable (Court et al., 2000). CYP2B6 was found to have a significant role in propofol hydroxylation. Inhibition by anti-CYP2B6 antibody and good correlations between propofol hydroxylation and CYP2B6 marker activities indicate that CYP2B6 is responsible for the variability in propofol hydroxylation activity.

The interaction of other drugs with CYP2B6 has not been thoroughly investigated. Newer antidepressants, including SSRIs, may inhibit the activity of human cytochromes, but interactions with CYP2B6 are not established. We observed that several SSRIs (sertraline, paroxetine, norfluoxetine, and fluvoxamine) have low IC₅₀ values for inhibition of bupropion hydroxylation (Table 3). Although the clinical significance of these IC₅₀ values is not established, paroxetine appears to have the highest probability of interference with CYP2B6. Clinical monitoring during combined use of these SSRIs with bupropion is necessary, since elevated bupropion plasma levels may be associated with central nervous system toxicity.

Product labeling for bupropion indicates that CYP2D6 is inhibited by bupropion or hydroxybupropion. The present study indicates that bupropion and hydroxybupropion have relatively low inhibitory potential of CYP2D6 in vitro, with IC₅₀ values of 58 and 74 µM, respectively. A case report has suggested the inhibition of CYP2D6 by bupropion in a patient whose desipramine levels were increased after combination therapy of imipramine and bupropion (Shad and Preskorn, 1997). Pollock et al. (1996) reported that debrisoquine metabolic ratios in three patients before and after at least 2 months of bupropion treatment did not change importantly. Furthermore, the plasma concentration/dose ratio for bupropion was not substantially different between CYP2D6 extensive and poor metabolizers. It was concluded that bupropion does not inhibit CYP2D6 in vivo and that bupropion itself is not likely to be a substrate for CYP2D6.

The present study has focused on the biotransformation of bupropion to hydroxybupropion, but we did not detect formation of two other metabolites, threohydrobupropion and erythrohydrobupropion. According to the manufacturer, threohydrobupropion formation was detected in vitro using human liver microsomes, but no threohydrobupropion was detected using individually expressed CYP isoenzymes (Wurm et al., 1996). Erythrohydrobupropion formation was not detected in vitro (Wurm et al., 1996), and we did not detect either of these metabolic products in the present study. Hydroxybupropion shows stronger anti-tetrabenazine activity and has a lower LD₅₀ value than the erythro and threo metabolites, suggesting that hydroxybupropion is the most important active metabolite in vivo (Schroeder, 1983). Since plasma hydroxybupropion levels are usually higher than bupropion, it has been suggested that this metabolite may be responsible for toxicity (Laizure et al., 1985). In vivo plasma levels of both erythrohydrobupropion and threohydrobupropion also exceed those of bupropion itself. However, the mechanism of formation of these metabolites remains uncertain.

Since bupropion has been approved as an anti-smoking agent, its use is expected to increase. In addition, bupropion may be prescribed in place of other antidepressants since it has less likelihood of sexual side effects, and combination of bupropion with SSRIs is used to treat refractory depression. The risk of drug interactions or central nervous system toxicity associated with bupropion may be of clinical importance and may correlate with high bupropion or metabolite levels. Therefore, understanding the metabolism of bupropion and in vitro interactions with other xenobiotics may give insight into the risk of adverse effects.