Cytochrome P450-Mediated Metabolism of Haloperidol and Reduced

Jun 13, 2006 - and Queensland Centre for Schizophrenia Research, Saint Lucia, Brisbane, Queensland, Australia 4072. ReceiVed January 14, 2006...
0 downloads 0 Views 123KB Size
Download PDF
914

Chem. Res. Toxicol. 2006, 19, 914-920

Cytochrome P450-Mediated Metabolism of Haloperidol and Reduced Haloperidol to Pyridinium Metabolites Kathryn M. Avent,†,§ J. J. DeVoss,‡ and Elizabeth M. J. Gillam*,† School of Biomedical Sciences, and School of Molecular and Microbial Sciences, The UniVersity of Queensland, and Queensland Centre for Schizophrenia Research, Saint Lucia, Brisbane, Queensland, Australia 4072 ReceiVed January 14, 2006

Haloperidol (HP) has been reported to undergo cytochrome P450 (P450)-mediated metabolism to potentially neurotoxic pyridinium metabolites; however, the chemical pathways and specific enzymes involved in these reactions remain to be identified. The aims of the current study were to (i) fully identify the cytochrome P450 enzymes capable of metabolizing HP to the pyridinium metabolite, 4-(4chlorophenyl)-1-(4-fluorophenyl)-4-oxobutylpyridinium (HPP+), and reduced HP (RHP) to 4-(4chlorophenyl)-1-(4-fluorophenyl)-4-hydroxybutylpyridinium (RHPP+); and (ii) determine whether 4-(4chlorophenyl)-1-(4-fluorophenyl)-4-oxobutyl-1,2,3,6-tetrahydropyridine (HPTP) and 4-(4-chlorophenyl)1-(4-fluorophenyl)-4-hydroxybutyl-1,2,3,6-tetrahydropyridine (RHPTP) were metabolic intermediates in these pathways. In vitro studies were conducted using human liver microsomal preparations and recombinant human cytochrome P450 enzymes (P450s 1A1, 1A2, 1B1, 2A6, 2B6, 2C9, 2C19 2D6, 2E1, 3A4, 3A5, and 3A7) expressed in bicistronic format with human NADPH cytochrome P450 reductase in Escherichia coli membranes. Pyridinium formation from HP and RHP was highly correlated across liver preparations, suggesting the same enzyme or enzymes were responsible for both reactions. Cytochrome P450s 3A4, 3A5, and 3A7 were the only recombinant enzymes which demonstrated significant catalytic activity under optimized conditions, although trace levels of activity could be catalyzed by NADPHP450 reductase alone. NADPH-P450 reductase-mediated activity was inhibited by reduced glutathione but not catalase or superoxide dismutase, suggesting O2-dependent oxidation. No evidence was obtained to support the contention that HPTP and RHPTP are intermediates in these pathways. Km values for HPP+ (34 ( 5 µM) and RHPP+ (64 ( 4 µM) formation by recombinant P450 3A4 agreed well with those obtained with human liver microsomes, consistent with P450 3A4 being the major catalyst of pyridinium metabolite formation in human liver. Introduction Haloperidol, 4-[4-(4-chlorophenyl)-4-hydroxy-1-piperidinyl]1-(4′-fluorophenyl)-1-butanone (HP),1 has been shown to be converted to 4-(4-chlorophenyl)-1-(4-fluorophenyl)-4-oxobutylpyridinium (HPP+), a structural analogue of the neurotoxic agent MPP+ (1-methyl-4-phenylpyridinium) (1). The extrapyramidal movement disorders associated with HP treatment in some patients resemble the parkinsonian symptoms caused by MPP+, leading to the hypothesis that HPP+ or other structurally related metabolites of HP are responsible for its toxicity by a * To whom correspondence should be addressed at School of Biomedical Sciences, The University of Queensland, St. Lucia, Brisbane, 4072 Queensland, Australia. Fax, 61-7-3365-1766; phone, 61-7-3365-1410/3506; Ee-mail, [email protected]. † School of Biomedical Sciences, The University of Queensland. ‡ School of Molecular and Microbial Sciences, The University of Queensland. § Queensland Centre for Schizophrenia Research. 1 Abbreviations: HAC, chlorohaloperidol; HL, human liver; hNPR, human NADPH-cytochrome P450 reductase; HP, haloperidol, 4-[4-(4chlorophenyl)-4-hydroxy-1-piperidinyl]-1-(4′-fluorophenyl)-1-butanone; HPDP+, 4-(4-chlorophenyl)-1-(4-fluorophenyl)-4-oxobutyl-2,3-dihydropyridinium; HPP+, 4-(4-chlorophenyl)-1-(4-fluorophenyl)-4-oxobutylpyridinium; HPTP, 4-(4-chlorophenyl)-1-(4-fluorophenyl)-4-oxobutyl-1,2,3,6-tetrahydropyridine; MPDP+, 1-methyl-4-phenyl-2,3-dihydropyridinium; MPP+, 1-methyl-4-phenylpyridinium; MPTP, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine; P450, cytochrome P450; PPIS, pyridinium internal standard, 4-phenyl-1,4-fluorophenyl-4-oxobutylpyridinium chloride; RHP, reduced haloperidol; RHPP+, reduced HPP+, 4-(4-chlorophenyl)-1-(4-fluorophenyl)4-hydroxybutylpyridinium; RHPTP, 4-(4-chlorophenyl)-1-(4-fluorophenyl)4-hydroxybutyl-1,2,3,6-tetrahydropyridine.

similar mechanism to MPP+ (1). Metabolic studies have since reported the presence of HPP+ in the urine, plasma, liver, and brain of HP-treated rats and mice (1-3). Further studies identified HPP+ and its reduced analogue, 4-(4-chlorophenyl)1-(4-fluorophenyl)-4-hydroxybutylpyridinium (reduced HPP+; RHPP+), in the plasma and urine of HP-treated patients where they represented approximately 3% of the daily urinary excretion of the HP dose (4, 5). The identification of HPP+ and RHPP+ in the brains of baboons treated with HPTP (4-(4-chlorophenyl)1-(4-fluorophenyl)-4-oxobutyl-1,2,3,6-tetrahydropyyridine) and in HP-treated patients has prompted further investigation into the enzymology of their formation (6, 7). Two pathways have been proposed to account for the biotransformation of HP to HPP+ (and similarly for RHP to RHPP+), as illustrated in Scheme 1. Either (1) ring R-carbon oxidation yields the iminium intermediate and is converted to HPDP+, directly or via its enamine conjugate base or (2) initial dehydration yields the tetrahydropyridine intermediate (HPTP), which undergoes R-carbon oxidation to form HPDP+ (4-(4chlorophenyl)-1-(4-fluorophenyl)-4-oxobutyl-2,3-dihydropyridinium). In both instances, HPDP+ would then spontaneously oxidize to HPP+ (1, 8). The oxidation of HP to HPP+ has been shown to be catalyzed by cytochrome P450 enzymes using liver microsomes (4, 9, 10-15) recombinant P450 3A4 and 3A5 (9, 14, 16, 17) and recombinant P450 2D6 (16). Although HPTP was reportedly found in metabolic incubations of HP with nonhuman liver

10.1021/tx0600090 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/13/2006

Metabolism of Haloperidol to Pyridiniums

Chem. Res. Toxicol., Vol. 19, No. 7, 2006 915

Scheme 1. Proposed Catalytic Pathways for the Cytochrome P450-Mediated Oxidation of HP to HPP+ a

a Either (1) ring R-carbon oxidation yields the iminium intermediate and is converted to HPDP+, directly or via its enamine conjugate base, or (2) initial dehydration yields the tetrahydropyridine intermediate, HPTP, which undergoes R-carbon oxidation to form HPDP+. In both instances, HPDP+ would then spontaneously oxidize to HPP+. A similar catalytic pathway is proposed for the biotransformation of RHP to RHPP+. Adapted from ref 8.

microsomal preparations (9-13) and recombinant human cytochrome P450 enzymes (9, 16) in which HPP+ was formed, the presence of HPTP as an intermediate in HP metabolism remains to be demonstrated with human liver microsomal preparations. The overall aim of the studies presented here was to identify which cytochrome P450 enzymes catalyze the metabolism of HP and RHP to HPP+ and RHPP+, respectively, using human liver microsomal preparations and a panel of recombinant P450s expressed in bicistronic format with hNPR in bacterial membranes, and to determine whether HPTP and RHPTP (4-(4chlorophenyl)-1-(4-fluorophenyl)-4-hydroxybutyl-1,2,3,6-tetrahydropyridine) are intermediates in these pathways.

Experimental Procedures Caution: Haloperidol has neurotoxic effects and should be handled with care. Chemicals and Enzyme Preparations. HP chloride and racemic RHP were donated by Janssen-Cilag (Beerse, Belgium). HPTP, HPP+ (4), racemic RHPP+ chloride (2), RHPTP (8), and the pyridinium internal standard, 4-phenyl-1,4-fluorophenyl-4-oxobutylpyridinium chloride (PPIS), were synthesized by Dr D. W. Eyles (Department of Medicine, University of Queensland) (5). Catalase and superoxide dismutase were purchased from Sigma-Aldrich (St. Louis, MO). All other solvents and chemicals were purchased from local suppliers at the highest quality available. Livers were obtained from male Wistar strain rats (8 weeks, 150-200 g; Central Animal Breeding Facility, The University of Queensland, St Lucia, Australia) and from organ donors according to procedures approved by The University of Queensland and affiliated hospital ethics committees. Samples of human liver were frozen in liquid nitrogen and stored at -70 °C prior to use. Bicistronic constructs expressing both P450 and hNPR were prepared previously in this laboratory or in the laboratory of Prof. F. P. Guengerich (Vanderbilt University, Nashville, TN) and have been reported in other studies (P450s 1A1, 1A2, 2D6, 2E1, and 3A4 (18); 1B1 (19); 2A6 (20); 2B6 (21); 2C9, 2C19 (22); and 3A5, 3A7 (23). The pGro7 plasmid (24) was kindly donated by H. Yanagi (HSP Research Institute, Shimogyo-ku, Kyoto, Japan). Preparation and Characterization of Liver Microsomes and Recombinant P450s. Liver microsomes were prepared as described previously (25). Human P450 enzymes were expressed in bacteria in bicistronic format with hNPR as described previously (18, 22, 23), with the expression of P450 forms 1A1, 2A6, 2B6, 2D6, and 3A7 augmented by coexpression of bacterial chaperones encoded by the plasmid pGro7 (26). Reduced-CO (Fe2+-CO) versus reduced (Fe2+) difference spectra were measured in microsomes or membrane fractions according to the method of Omura and Sato (27). NPR activity was measured by its NADPH-cytochrome c reductase activity, as described by Guengerich (25).

Enzyme Incubations. Incubations routinely contained 100 mM potassium phosphate buffer (pH 7.4), 300 mM KCl, 0.15 µM P450 in either human liver microsomes or P450/hNPR membranes, and substrate added from a methanolic stock (0.8% (v/v) methanol, final concentration) and were initiated by the addition of an NADPH regenerating system (1 mM NADP+, 25 mM glucose-6-phosphate, and 0.5 U/mL glucose-6-phosphate dehydrogenase, final concentrations). Aliquots were removed at appropriate times and added to an equal volume of 0.1% (v/v) acetic acid in acetonitrile to which had been added two internal standards, 4-[4-(p-chlorophenyl)-4hydroxypiperidino]-4′-chlorobutyrophenone (chlorohaloperidol; HAC) and the pyridinium internal standard, PPIS. Samples were mixed and centrifuged at 16 000g for 10 min, then 10-50 µL aliquots of the supernatant were analyzed by HPLC. Experiments using recombinant P450s 2C9, 2C19, 3A4, and 3A5 were supplemented with hNPR in bacterial membranes such that the ratio of hNPR/ P450 was 5:1 in the case of P450 2C forms and 2:1 for P450s 3A4 and 3A5. Variations to these standard conditions, such as the inclusion of KCl, MgCl2, EDTA, or supplementation with hNPRcontaining bacterial membranes, were as indicated for individual experiments. Final concentrations of methanol were maintained at 0.8% (v/v) unless otherwise indicated. Negative controls, lacking substrate, NADPH-generating system, or P450, were included in each assay. P450-deficient controls included membranes isolated from cells transformed either with the expression vector alone (pCW′) or with the monocistronic expression vector for hNPR (pCW′/hNPR). Kinetic Studies. Product formation from HP and RHP was measured over 30-90 min and using P450 concentrations of 0.1 µM for human liver microsomes and 0.04 µM for recombinant P450 3A forms, conditions under which product formation was linear with respect to time and P450 concentration at 1 µM HP or RHP, the lowest substrate concentration to be used in kinetic analysis (determined by the limit of sensitivity for product formation). The rate of product formation from HP (1-100 µM) and RHP (1-750 µM) was determined in microsomal preparations from three human livers (HL23, HL24, HL26) and using P450 3A4 and 3A5, under standard assay conditions. HPLC Analysis. The published HPLC assay of Eyles et al. (5) was modified to include parallel electrochemical and fluorescence detection enabling concurrent quantification of HP, RHP, HPTP, RHPTP, HPP+, and RHPP+. Metabolites were separated using a Waters C18 Nova-Pak column (4 µm, 4.6 × 150 mm) and a 28% acetonitrile:72% 60 mM ammonium acetate buffer (pH 4.5) mobile phase delivered at a flow rate of 1.5 mL/min and temperature of 40 °C. The flow was split evenly between the electrochemical detector (applied potential of +0.97 V) and the fluorescence detector (λex ) 304 nm, λem ) 374 nm). The limits of sensitivity were 10 pmol/mL for HP, HPTP, and RHPTP; 20 pmol/mL for RHP; 2 pmol/mL for HPP+; and 0.6 pmol/mL for RHPP+. Data Analysis. Kinetic parameters were determined by fitting the data to three models routinely used for the analysis of P450

916 Chem. Res. Toxicol., Vol. 19, No. 7, 2006

AVent et al.

Figure 1. Relationship between HPP+ formation from HP and RHPP+ formation from RHP across six human liver microsomal preparations. HP or RHP (10 µM) was incubated with human liver microsomes (0.15 µM P450) at 37 °C in the presence and absence of an NADPHgenerating system. Aliquots were quenched at 0 and 90 min and analyzed for HPP+ or RHPP by HPLC. There was no evidence of HPP+ or RHPP formation in the absence of NADPH or at time zero (data not shown). Each point represents the mean of three determinations for an individual microsomal preparation. The formation of HPP+ and RHPP+ from HP and RHP, respectively, was highly correlated; r2 ) 0.95, p < 0.0001.

enzyme kinetics (28). Km and Vmax values were determined by fitting the data to the Michaelis-Menten equation for a one-enzyme model using a nonlinear least-squares regression analysis program (GraphPad Prism version 3.00 for Windows, GraphPad Software, San Diego, CA, www.graphpad.com). Data were then reanalyzed using the Hill equation and an equation describing uncompetitive substrate inhibition (UCSI). F-tests were conducted to determine whether fits to the Hill or USCI equations were significantly better (i.e., P < 0.05) than for the Michaelis-Menten equation.

Results HPP+

RHPP+

or formation was time- and NADPH-dependent for all human liver preparations studied and varied 3- to 4-fold across the six preparations under these conditions. The conversion of HP and RHP to HPP+ and RHPP+, respectively, was highly correlated (r2 ) 0.95, p < 0.0001) across the six liver preparations studied (Figure 1). In preliminary assays, recombinant P450 3A4 demonstrated apparent pyridinium formation from 100 µM HP or RHP that was significantly greater than the background seen in the absence of NADPH or that shown by P450-deficient controls. Assay conditions were partially optimized for this form. The effect of varying KCl (0-300 mM), MgCl2 (0, 10 mM), and EDTA (0, 1 mM) concentrations in incubations was assessed. Only KCl significantly enhanced activity (p < 0.001 and p < 0.001 Student’s t-test for HPP+ and RHPP+ formation, respectively; data not shown), so 300 mM KCl was included in all further incubations. When standard assay conditions; human liver microsomes; and recombinant P450s 1A1, 1B1 3A4, 3A5, and 3A7 were used, all enzyme preparations demonstrated time- and NADPHdependent HPP+ formation from 100 µM HP that was significantly greater than the apparent HPP+ formation in P450deficient controls (pCW and hNPR; Figure 2A). Trace levels of HPP+ were seen in hNPR controls. Similarly, human liver microsomes and several recombinant P450s (1A1, 1B1, 2A6, 2C9, 2C19, 2D6, 3A4, 3A5, and 3A7) demonstrated RHPP+ formation from 375 µM RHP, and apparent RHPP+ formation was also seen in hNPR controls (Figure 2B). The incubation of 100 µM HP and 375 µM RHP with membranes containing hNPR alone for 90 min resulted in pyridinium formation which increased in proportion to hNPR

Figure 2. Preliminary screen of HP and RHP oxidation to pyridinium metabolites by human liver microsomes and recombinant human P450s expressed in Escherichia coli membranes with hNPR. HP (100 µM; panel A) or RHP (375 µM; panel B) was incubated with rP450/hNPR membranes (0.15 µM P450) at 37 °C in the presence (solid and crosshatched bars) and absence (open bars) of an NADPH generating system. Samples were collected at 0 (crosshatched bars) and 90 min (solid and open bars) and analyzed for HPP+ (A) or RHPP+ (B) by HPLC. Human liver microsomes (0.15 µM P450) were used as a positive control. P450-deficient controls consisted of membranes from E. coli transformed with expression vector pCW alone, or with the monocistronic expression vector for hNPR (pCW/hNPR). Incubations of P450s 2C9, 2C19, 3A4, and 3A5 were supplemented with membranes containing expressed hNPR such that final hNPR concentrations were 5 times or twice the P450 concentration for P450 2C forms and P450s 3A4 and 3A5, respectively (the hNPR level in 3A7/hNPR membranes was already 6.7 times the P450 level; therefore, no additional hNPR was added to incubations with P450 3A7). Preparations that showed significant HPP+ formation compared to the relevant NADPH-deficient control are indicated by asterisks: *p < 0.05; **p < 0.01; ***p < 0.001. Results are presented as mean ( SD of three experiments.

concentration (0.05-1.0 µM) (Figure 3A). Pyridinium formation was significantly higher in the presence of 1.0 µM hNPR than in the presence of pCW membranes at the same protein concentration (HPP+, p < 0.05; RHPP+, p < 0.002) (Figure 3A). At 1.0 µM hNPR, pyridinium formation was inhibited by glutathione (0.05-50 mM) (Figure 3B), but not by catalase or superoxide dismutase (each used at 250 or 500 U/mL; results not shown). When the hNPR concentration was normalized to 1.0 µM for all P450s (hNPR/P450 ) 6.66:1), only P450s 3A4, 3A5, and 3A7 demonstrated significant pyridinium formation from HP and RHP compared to NADPH- and P450-deficient controls (Figure 4). Tables 1 and 2 summarize the kinetic constants obtained using nonlinear regression analysis for HP and RHP metabolism to the respective pyridinium metabolites by HL23, HL24, HL26, and recombinant P450 3A4. Levels of activity with P450s 3A5 and 3A7 were too low to allow accurate determination of full kinetic parameters. Substrate saturation curves are shown in

Metabolism of Haloperidol to Pyridiniums

Chem. Res. Toxicol., Vol. 19, No. 7, 2006 917

Figure 4. HP and RHP oxidation to pyridinium metabolites by recombinant human P450s with normalization of hNPR concentrations. HP (100 µM; panel A) or RHP (375 µM; panel B) was incubated at 37 °C with rP450/hNPR membranes (0.15 µM P450), supplemented with membranes containing expressed hNPR such that final hNPR concentrations were 1.0 µM (hNPR/P450 6.7:1) in the presence of an NADPHgenerating system. Samples were collected at 90 min and analyzed for HPP+ (A) or RHPP+ (B) by HPLC. P450-deficient controls consisted of membranes from E. coli transformed with expression vector pCW alone or the expression vector with hNPR (pCW/hNPR). P450s which showed significant HPP+ formation with respect to the relevant hNPRonly control are indicated by asterisks: *p < 0.05; **p < 0.01. Results are presented as mean ( SD of three experiments. Figure 3. The effect of hNPR and GSH on pyridinium metabolite formation. (A) Dependence of pyridinium formation from HP and RHP on the concentration of hNPR membranes. HP (100 µM) or RHP (375 µM) was incubated with different concentrations of E. coli membranes containing hNPR (0.05-1.0 µM hNPR) for 90 min in the presence of an NADPH-generating system. HPP+ (squares) and RHPP+ (triangles) formation was measured by HPLC analysis. Solid and dashed lines represent the amounts of HPP+ and RHPP+, respectively, formed in the presence of E. coli membranes containing pCW added to the same protein concentration as 1.0 µM hNPR. Results are presented as the average of two determinations. (B) Effect of glutathione on pyridinium formation from HP and RHP by hNPR membranes. HP (100 µM) or RHP (375 µM) was incubated with E. coli membranes containing 1.0 µM hNPR and glutathione (0.05-50 mM) for 90 min in the presence of an NADPH generating system. HPP+ (squares) and RHPP+ (triangles) formation was measured by HPLC. Results are presented as the average of two determinations.

Figures 5 and 6 for one representative liver microsomal preparation and P450 3A4. No evidence for HPTP or RHPTP formation was obtained from the incubation of HP or RHP with rat or human liver microsomal incubations or any recombinant P450, under standard assay conditions. Representative chromatograms demonstrating the time- and NADPH-dependent formation of HPP+ from HP and RHPP+ from RHP by recombinant P450 3A4 are presented in Figure 7, and the absence of concurrent HPTP or RHPTP formation is presented in Figure 8. The sensitivity limit for HPP and RHPTP detection in these incubations was 10 nM, whereas the peaks of HPP+ and RHPP+ shown in Figure 7 corresponded to concentrations in incubations of 11 and 12.5 µM, respectively.

Discussion The time- and NADPH-dependent metabolism of HP and RHP to pyridinium metabolites evident in these studies is consistent with cytochrome P450 involvement, as previously

Table 1. Kinetic Parameters for HP Conversion to HPP+ by Human Liver Microsomes and Recombinant P450 Forms enzyme preparation

Vmax, (pmol/min)/pmol P450 ((pmol/min)/mg protein)

Km, µM

Kia, µM

HL23 HL24 HL26 P450 3A4

0.54 ( 0.03 (220 ( 10) 0.46 ( 0.01 (270 ( 60) 0.77 ( 0.53 (370 ( 250) 1.1 ( 0.1

44 ( 6 48 ( 14 190 ( 150 34 ( 5

150 ( 80 31 ( 28

a Inhibition constant for data showing statistically significant evidence for substrate inhibition.

Table 2. Kinetic Parameters for RHP Conversion to RHPP+ by Human Liver Microsomes and Recombinant P450 Forms enzyme preparation

Vmax, (pmol/min)/pmol P450 ((pmol/min)/mg protein)

Km, µM

HL23 HL24 HL26 P450 3A4

0.26 ( 0.01 (107 ( 1) 0.27 ( 0.01 (159 ( 4) 0.14 ( 0.01 (67 ( 2) 2.5 ( 0.1 (460 ( 10)

50 ( 2 40 ( 3 49 ( 3 64 ( 4

Kia, µM

1900 ( 300

a Inhibition constant for data showing statistically significant evidence for substrate inhibition.

reported (4, 14, 15). In the current investigation, the conversion of HP and RHP to HPP+ and RHPP+, respectively, was highly correlated across the six liver preparations studied, suggesting that the same enzyme or enzymes may be responsible for both of these reactions. Studies by Usuki et al. (14), Fang et al. (9, 16), and Kalgutkar et al. (17) published during the course of this work suggested that the enzyme principally responsible for the oxidative metabolism of HP might be P450 3A4, with some contribution by P450 3A5 (17). Preliminary studies reported here, using a wider panel of recombinant P450s expressed in bicistronic format with hNPR in bacterial membranes, further

918 Chem. Res. Toxicol., Vol. 19, No. 7, 2006

AVent et al.

Figure 5. Substrate concentration dependence of HP conversion to HPP+ by one representative human liver microsomal preparation and recombinant P450 3A4. HP (0-100 µM) was incubated for 30 min at 37 °C in the presence of an NADPH-generating system with either human liver microsomes (0.1 µM P450, HL24) or recombinant P450 3A4/hNPR membranes (0.04 µM P450) supplemented with membranes containing recombinant hNPR such that the final hNPR concentrations were twice the P450 concentration. The fitted curves correspond to the kinetic analyses presented in Table 1. Results are presented as individual measurements from three independent experiments.

Figure 6. Substrate concentration dependence of RHP conversion to RHPP+ by one representative human liver microsomal preparation and recombinant P450 3A4. RHP (0-750 µM) was incubated for 30 min at 37 °C in the presence of an NADPH-generating system with either human liver microsomes (0.1 µM P450 HL24) or recombinant P450 3A4/hNPR membranes (0.04 µM P450) supplemented with membranes containing recombinant hNPR such that the final hNPR concentrations were twice the P450 concentration. The fitted curves correspond to the kinetic analyses presented in Table 2. Results are presented as individual measurements from three independent experiments.

supported these findings with P450 3A4 showing the most activity toward pyridinium formation at the highest substrate concentrations that were able to be achieved in vitro (Figure 2A). Lower levels of apparent activity were observed with P450s 3A5 and 3A7. Similarly, P450 3A4 showed maximal activity toward RHP conversion to RHPP+ in a preliminary screen at the maximum achievable substrate concentrations (Figure 2B). P450 3A5 and 3A7 showed lower levels of activity, while some apparent activity was seen with other forms, as well as hNPR alone. The rate of conversion of HP and RHP to pyridiniums by recombinant P450 3A4 was comparable to that by liver microsomes at equivalent P450 concentrations assuming an average P450 3A4 content of human liver microsomes of ∼ 28% total P450 (29). These preliminary findings that multiple recombinant P450s were capable of catalyzing the conversion of HP to HPP+ and RHP to RHPP+ appear to be in partial agreement with the findings of Fang et al. (9) and Kalgutkar et al. (17), who reported that P450s 3A4 and 3A5 catalyze HP and RHP oxidation to the respective pyridinium metabolites. However, in the present study, no evidence was obtained that P450 1A1 catalyzed HP oxidation to HPP+ above background levels. The reason for this discrepancy is unclear, but may relate to differences in P450/

Figure 7. Representative chromatograms illustrating the time and NADPH-dependent formation of HPP+ and RHPP+ from HP and RHP by rP450 3A4/hNPR membranes. A mixture of the pure compounds indicated (A) and extracts from incubation mixtures of (B) HP (100 µM) and (C) RHP (375 µM) with rP450 3A4/hNPR were subjected to HPLC analysis with fluorescence detection. Metabolites were resolved using a Waters C18 Nova-Pak column (4 µm, 4.6 × 150 mm) and a 28% acetonitrile, 72% 60 mM ammonium acetate buffer (pH 4.5) mobile phase at a flow rate of 1.5 mL/min and temperature of 40 °C. The fluorescence detector excitation and emission wavelengths were 304 and 374 nm, respectively. Chromatograms represent (B) HPP+ and (C) RHPP+ formed (a) in the absence of NADPH and in the presence of NADPH at (b) 0 min and (c) 90 min.

reductase ratios or other factors. Notably, the background apparent pyridinium formation observed in the present study in the absence of P450 suggests that a small degree of P450independent oxidation of HP and RHP is occurring in membranes. In the presence of NADPH, there was a small amount of pyridinium formed in membranes isolated from cells transformed with the empty expression vector alone, but this represented less than 20% of that formed in membranes containing hNPR, suggesting a role for hNPR in this apparent activity. The possibility of low-level catalysis by some bacterial oxidase, which is capable of using hNPR as an electron donor, is unlikely but cannot be excluded. Alternatively, a wellestablished side reaction of NPR, in the presence of NADPH, is the reduction of O2 to generate reactive oxygen species (ROS) (30, 31). The susceptibility of HP to ROS-catalyzed oxidation, due to the lone pair of electrons on the nitrogen, is demonstrated by its H2O2-mediated conversion to HP-N-oxide (11). The decrease in pyridinium formation in the presence of GSH supports the involvement of ROS in the apparent pyridinium formation, and the limited effects of catalase and superoxide dismutase suggest that this ROS is neither H2O2 nor superoxide (32). The ROS concerned may be a lipid peroxide, which would readily be formed by NADPH-dependent oxidation of membrane lipids in the immediate vicinity of the membrane-associated hNPR. Such lipid peroxides would be resistant to the effects of the water-soluble catalase and SOD. Initiation of lipid peroxidation by, for example, hydroxyl radical, superoxide, or hydrogen peroxide generated by hNPR would not be effectively blocked by these enzymes within the immediate proximity of hNPR. The possibility of P450-independent, hNPR-mediated

Metabolism of Haloperidol to Pyridiniums

Figure 8. Representative chromatograms illustrating the absence of HPTP and RHPTP in incubations of HP and RHP with human liver microsomes. A mixture of the pure compounds indicated (A) and incubation mixtures of (B) HP (100 µM) and (C) RHP (375 µM) were subjected to HPLC analysis with electrochemical detection. Metabolites were resolved using a Waters C18 Nova-Pak column (4 µm, 4.6 × 150 mm) and a 28% acetonitrile, 72% 60 mM ammonium acetate buffer (pH 4.5) mobile phase at a flow rate of 1.5 mL/min and temperature of 40 °C. The electrochemical detector applied potential was set to +0.97 V. Chromatograms show extracts from incubations with human liver microsomes (a) in the absence of NADPH and in the presence of NADPH at (b) 0 min and (c) 90 min. No HPTP or RHPTP could be detected in these incubations (limit of sensitivity 10 pmol/mL).

catalysis of HP or RHP oxidation was not considered in previous studies (9, 16). Although HPTP has been reported as an intermediate in HP metabolism using nonhuman liver microsomal preparations (911), the results of the present study using human liver microsomes do not support this finding, since no HPTP could be seen in incubations using HP as a substrate in which abundant HPP+ was generated, either in liver microsomes or recombinant P450 3A4 incubations. Similarly, there was no evidence for the formation of RHPTP as an intermediate in the biotransformation of RHP to RHPP+. The failure to detect HPTP in the current study supports the proposal that the biotransformation of HP to HPP+ proceeds without HPTP formation via the first of the proposed pathways (Scheme 1). It has been suggested (6, 9) that the presence of HPTP may result from disproportionation of the inherently unstable HPDP+ to HPTP and HPP+, as was reported for MPDP+ (1-methyl-4phenyl-2,3-dihydropyridinium) (33, 34). The disproportionation of MPDP+ to equal amounts of MPTP (1-methyl-4-phenyl1,2,3,6-tetrahydropyridine) and MPP+ was greater in mitochondrial incubation reaction mixtures than in pH 7.4 buffer (3335) and at substrate concentrations above 100 µM (34). In contrast to MPDP+, HPDP+ underwent stoichiometric disproportionation at 2 and 22 µM (36). Thus, it is possible that the presence of HPTP in previous reports may be an artifact of HPDP+ disproportionation, which did not occur under the conditions used in the present study. Moreover, the ability of Zn2+ to act as a Lewis-acid may have promoted the acidcatalyzed dehydration of HP to HPTP, when ZnSO4 was included as a protein precipitant prior to the HPLC/MS confirmation of structure (9-11). Overall, the data presented here supports the proposal that HP biotransformation to HPP+ proceeds without the metabolic formation of HPTP.

Chem. Res. Toxicol., Vol. 19, No. 7, 2006 919

Nonlinear regression analysis of the kinetics of pyridinium formation from HP and RHP by P450 3A4 generated Km and Vmax values (Tables 1 and 2, HP and RHP, respectively), which agreed well with those obtained for HP and RHP oxidation by human liver microsomes in both this study and other studies published during the course of the present work (14, 17). Vmax values were enhanced severalfold, consistent with an enrichment of the key enzymatic catalyst in P450 3A4 incubations compared to human liver microsomes. Statistical evidence for substrate inhibition was found in two human liver preparations used for HP oxidation and one for RHP oxidation. However, the data from recombinant P450 3A4 incubations conformed to the simpler Michaelis-Menten kinetic model. The reason for this discrepancy is not clear but may indicate subtle differences between the microsomal and bacterial membrane environment or reflect the contribution in microsomes of a second P450 enzyme. While the data presented here confirms that P450 3A4 is the primary enzyme responsible for pyridinium formation, it also provides evidence of the potential catalytic activity of P450s 3A5 and 3A7 in HP and RHP oxidation. The relevance of P450s 3A5 and 3A7 in HP metabolism will be dependent on their level of expression in individuals and their affinities for HP and RHP, as well as the circulating concentrations of these compounds. Unfortunately, the much lower activity shown by these forms compared to P450 3A4 precluded a thorough characterization of their kinetics. The limited data obtained for P450 3A5, however, suggested this form showed a generally higher Km for HP and RHP than P450 3A4 (in line with results published by Kalgutkar et al. (17)). In summary, the results presented here support the finding that P450 3A4 is the primary P450 involved in pyridinium formation from HP and RHP and provide evidence that P450 3A5 and 3A7 are also capable of catalyzing these reactions. No evidence was found to suggest that HPTP and RHPTP are intermediates in these reactions.

References (1) Subramanyam, B., Rollema, H., Woolf, T., and Castagnoli, N., Jr. (1990) Identification of a potentially neurotoxic pyridinium metabolite of haloperidol in rats. Biochem. Biophys. Res. Commun. 166, 238244. (2) Van der Schyf, C. J., Castagnoli, K., Usuki, E., Fouda, H. G., Rimoldi, J. M., and Castagnoli, N., Jr. (1994) Metabolic studies on haloperidol and its tetrahydropyridine analog in C57BL/6 mice. Chem. Res. Toxicol. 7, 281-285. (3) Igarashi, K., and Castagnoli, N., Jr. (1992) Determination of the pyridinium metabolite derived from haloperidol in brain tissue, plasma and urine by high-performance liquid chromatography with fluorescence detection. J. Chromatogr., Biomed. Appl. 579, 277-283. (4) Subramanyam, B., Pond, S. M., Eyles, D. W., Whiteford, H. A., Fouda, H. G., and Castagnoli, N., Jr. (1991) Identification of potentially neurotoxic metabolite in the urine of schizophrenic patients treated with haloperidol. Biochem. Biophys. Res. Commun. 181, 573-578. (5) Eyles, D. W., McLennan, H. R., Jones, A., McGrath, J. J., Stedman, T. J., and Pond, S. M. (1994) Quantitative analysis of two pyridinium metabolites of haloperidol in patients with schizophrenia. Clin. Pharmacol. Ther. 56, 512-520. (6) Avent, K. M. (2001) In Cytochrome P450-mediated metabolism of haloperidol and reduced haloperidol. Ph.D. Thesis, p 230, The University of Queensland, Brisbane, Australia. (7) Eyles, D. W., Avent, K. M., Stedman, T. J., and Pond, S. M. (1997) Two pyridinium metabolites of haloperidol are present in the brain of patients at post-mortem. Life Sci. 60, 529-534. (8) Avent, K. M., Riker, R. R., Fraser, G. L., Van der Schyf, C. J., Castagnoli, N., Jr., and Pond, S. M. (1997) Metabolism of haloperidol to pyridinium species in patients receiving high doses intravenously: Is HPTP an intermediate? Life Sci. 61, 2383-2390. (9) Fang, J., and Gorrod, J. (1991) Dehydration is the first step in the bioactivation of haloperidol to its pyridinium metabolite. Toxicol. Lett. 59, 117-123.

920 Chem. Res. Toxicol., Vol. 19, No. 7, 2006 (10) Fang, J., and Gorrod, J. (1992) Metabolism of haloperidol by mouse hepatic microsomal preparations. Med. Sci. Res. 20, 175-177. (11) Gorrod, J., and Fang, J. (1993) On the metabolism of haloperidol. Xenobiotica 23, 495-508. (12) Fang, J., and Gorrod, J. (1993) High-performance liquid chromatographic method for the detection and quantitation of haloperidol and seven of its metabolites in microsomal preparations. J. Chromatogr. 614, 267-273. (13) Igarashi, K., Kasuya, F., Fukui, M., Usuki, E., and Castagnoli, N., Jr. (1995) Studies on the metabolsim of haloperidol (HP): The role of CYP3A in the production of the neurotoxic pyridinium metabolite HPP+ found in rat brain following ip administration of HP. Life Sci. 57, 2439-2466. (14) Usuki, E., Pearce, R., Parkinson, A., and Castagnoli, N., Jr. (1996) Studies on the conversion of haloperidol and its tetrahydropyridine dehydration product to potentially neurotoxic pyridinium metabolites by human liver microsomes. Chem. Res. Toxicol. 9, 800-806. (15) Eyles, D. W., McGrath, J. J., and Pond, S. M. (1996) Formation of pyridinium species of haloperidol in human liver and brain. Psychopharmacology 125, 214-219. (16) Fang, J., Baker, G. B., Silverstone, P. H., and Coutts, R. T. (1997) Involvement of CYP3A4 and CYP2D6 in the metabolism of haloperidol. Cell. Mol. Neurobiol. 17, 227-233. (17) Kalgutkar, A. S., Taylor, T. J., Venkatarakrishnan, K., and Isin, E. M. (2004) Assessment of the contributions of CYP3A4 and CYP3A5 in the metabolism of the antipsychotic agent haloperidol to its potentially neurotoxic pyridinium metabolite and effect of antidepressants on the bioactivation pathway. Drug Metab. Dispos. 31, 243249. (18) Parikh, A., Gillam, E. M. J., and Guengerich, F. P. (1997) Drug metabolism by Escherichia coli expressing human cytochromes P450. Nat. Biotechnol. 15, 784-788. (19) Shimada, T., Wunsch, R. M., Hanna, I. H., Sutter, T. R., Guengerich, F. P., and Gillam, E. M. J. (1998) Recombinant human cytochrome P450 1B1 expression in Escherichia coli. Arch. Biochem. Biophys. 357, 111-120. (20) Gillam, E. M. J., Aguinaldo, A. M., Notley, L. M., Kim, D., Mundkowski, R. G., Volkov, A. A., Arnold, F. H., Soucek, P., DeVoss, J. J., and Guengerich, F. P. (1999) Formation of indigo by recombinant mammalian cytochrome P450s. Biochem. Biophys. Res. Commun. 265, 467-472. (21) Gillam, E. M. J., Notley, L. M., Cai, H., DeVoss, J. J., and Guengerich, F. P. (2000) Oxidation of indole by cytochrome P450 enzymes. Biochemistry 39, 13817-13824. (22) Cuttle, L., Munns, A. J., Hogg, N. A., Scott, J. R., Hooper, W. D., Dickinson, R. G., and Gillam, E. M. J. (2000) Phenytoin metabolism by human cytochrome P450: Involvement of P450 3A and 2C forms in secondary metabolism and drug-protein adduct formation. Drug Metab. Dispos. 28, 945-950. (23) Gillam, E. M. J., Ueng, Y.-F., Wunsch, R. M., Shimada, T., Kamataki, T., and Guengerich, F. P. (1997) Expression of cytochrome P450 3A7

AVent et al.

(24)

(25) (26) (27) (28) (29) (30) (31)

(32)

(33) (34)

(35) (36)

in Escherichia coli: Effects of 5′ modification and catalytic characterization of recombinant enzyme expressed in bicistronic format with NADPH-cytochrome P450 reductase. Arch. Biochem. Biophys. 346, 81-90. Nishihara, K., Kanemori, M., Kitagawa, M., Yanagi, H., and Yura, T. (1998) Chaperone coexpression plasmids: Differential and synergistic roles of DnaK-DnaJ-GrpE and GroEL-GroES in assisting folding of an allergen of Japanese Cedar pollen, Cryj2, in Escherichia coli. Appl. EnViron. Microbiol. 64, 1694-1699. Guengerich, F. P. (1994) Analysis and characterization of enzymes. In Principles and Methods of Toxicology 3rd ed. (Hayes, A. W., Ed.), pp 1259-1313, Raven Press, Ltd., New York. Notley, L. M., de Wolf, C. J. F., Wunsch, R. M., Lancaster, R. G., and Gillam, E. M. J. (2002) Bioactivation of tamoxifen by recombinant human cytochrome P450 enzymes. Chem. Res. Toxicol. 15, 614-622. Omura, T., and Sato, R. (1964) The carbon monoxide-binding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J. Biol. Chem. 239, 2370-2378. Hutzler, J. M., and Tracy, T. S. (2002) Atypical kinetic profiles in drug metabolism reactions. Drug Metab. Dispos. 30, 355-362. Guengerich, F. P. (2005) Human cytochrome P450 enzymes. In Cytochrome P450 3rd ed. (Ortiz de Montellano, P. R., Ed.), Kluwer Academic/Plenum Press, New York. Lai, C. S., Grover, T. A., and Peitte, L. H. (1979) Hydroxyl radical production in a purified NADPH-cytochrome c (P450) reductase system. Arch. Biochem. Biophys. 193, 373-378. Winston, G. W., and Cederbaum, A. I. (1983) NADPH-dependent production of oxy radicals by purified components of the rat liver mixed function oxidase system. I. Oxidation of hydroxyl radical scavenging agents. J. Biol. Chem. 258, 1508-1513. Baez, S., and Seqgura-Aguilar, J. (1995) Effect of superoxide dismutase and catalase during reduction of adrenochrome by DT-diaphorase and NADPH-cytochrome P450 reductase. Biochem. Mol. Med. 56, 3744. Castagnoli, N., Jr., Chiba, K., and Trevor, A. J. (1985) Potential bioactivation pathways for the neurotoxin 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine (MPTP). Life Sci. 36, 225-230. Peterson, L. A., Caldera, P. S., Trevor, A., Chiba, K., and Castagnoli, N., Jr. (1985) Studies on the 1-methyl-4-phenyl-2,3-dihydropyridinium species 2,3-MPDP+, the monoamine oxidase catalyzed oxidation product of the nigrostriatal toxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). J. Med. Chem. 28, 1432-1436. Gessner, W., Brossi, A., Shen, R., and Abell, C. W. (1985) Further insight into the mode of action of the neurotoxin 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP). FEBS Lett. 183, 345-348. Subramanyam, B., Woolf, T., and Castagnoli, N., Jr. (1991) Studies on the in vitro conversion of haloperidol to a potentially neurotoxic pyridinium metabolite. Chem. Res. Toxicol. 4, 123-128.

TX0600090