Introduction

Abstract Introduction Results & Discussion References

When drug-induced hepatic porphyrin accumulation is found during routine animal toxicology testing, the question arises whether development of a drug should be abandoned or whether the porphyrinogenic effect of the drug is species-dependent, thus allowing development of the drug to proceed. The rat and chick embryo have been the major species used to test drugs for porphyrin-inducing activity (1), and the porphyrinogenic effects of several xenobiotics have been shown to depend on their ability to cause mechanism-based inactivation of P450¹ isozymes (2-4). Upon biotransformation by certain P450 isozymes, alkyl radicals are released from these xenobiotics at the active site of P450, resulting in heme N-alkylation and inactivation of the P450 isozyme (2, 3). The N-alkylheme dissociates from the apoprotein of P450 and iron is released, yielding N-alkylPP, which is a potent inhibitor of ferrochelatase (EC 4.99.1.1) (5, 6), the terminal enzyme in heme biosynthesis. As a result of hepatic ferrochelatase inhibition, a large amount of protoporphyrin IX accumulates in the liver. The porphyrinogenicity of three prototype porphyrinogenic compoundsnamely, DDEP, TTMS (7-12), and AIA (1)has been documented in a variety of animal species. Moreover, in rat and chick embryos, the porphyrinogenicity of these xenobiotics is directly related to their ability to form N-alkylPP after mechanism-based inactivation of P450 isozymes (1). Although the pattern of P450 isozymes differs substantially between species (13), some of the P450 isozymes, targeted for mechanism-based inactivation by porphyrinogenic xenobiotics in animal species, are present as closely similar orthologs in human liver. If the hepatic P450 isozymes targeted for mechanism-based inactivation by porphyrinogenic xenobiotics in animal species are also targeted in human liver, it would be reasonable to deduce that the xenobiotics would also exhibit porphyrinogenic properties in humans. Thus, the objective of this study was to compare the human P450 isozymes, targeted for mechanism-based inactivation by the prototypical porphyrinogenic xenobiotics, DDEP, TTMS, and AIA with P450 isozymes previously shown to be targeted by these drugs in rat and chick embryo livers. To study the in vitro effects of these compounds on individual human P450 isozymes, microsomes prepared from human lymphoblastoid cell lines (AHH-1 TK+/) that expressed single cDNA-expressed human P450s were used.

Materials and Methods

Source of Chemicals. TTMS and DDEP were purchased from Colour Your Enzyme (Bath, Ontario, Canada). AIA was obtained as a gift from Hoffman-La Roche Ltd. (Vaudreuil, Quebec, Canada). Resorufin and 7-ethoxyresorufin were purchased from Molecular Probes, Inc. (Eugene, OR). [4-¹⁴C]Testosterone was from Dupont Canada, Inc. (Mississauga, Ontario, Canada). Diclofenac, dextromethorphan, testosterone, 2-hydroxytestosterone, 6-hydroxytestosterone, NADPH, and cytochrome c were purchased from Sigma Chemical Co. (St. Louis, MO). 4-Hydroxydiclofenac was purchased from Gentest Corp. (Woburn, MA).

In Vitro Incubation of Microsomes with Porphyrinogenic Xenobiotics. Microsomes prepared from human lymphoblastoid cell lines (AHH-1 TK+/)possessing single cDNA-expressed human P450 isozymes [1A1, 1A2, 2C9-Arg₁₄₄ (2C9), 2D6-Val₃₇₄ (2D6), or 3A4] or only cDNA-expressed human P450 oxidoreductasewere obtained from Gentest Corp. The 2C9, 2D6, and 3A4 microsomes were prepared from cell lines coexpressing P450 and oxidoreductase cDNA. Control microsomes from both untransfected human lymphoblastoid cells (AHH-1 TK+/), and from human lymphoblastoid (AHH-1 TK+/) cells that were transfected with the plasmid vector, which was devoid of cDNA for human P450, were also purchased from Gentest Corp. Microsomes were stored at 70°C for no more than 3 months before use.

Enzyme Assays. EROD Activity. EROD activity was used to quantitate the functional activity of human P4501A1 and 1A2 samples, as previously described (14). Microsomal pellets were rehomogenized in 0.1 M K₂HPO₄ buffer (pH 7.4) to yield a protein concentration of ~0.1 mg/ml. The reaction was initiated by addition of 1 ml of microsomes to 1 ml of prewarmed (37°C) cofactor solution (containing 10 µM 7-ethoxyresorufin, 4 mM NADPH, and 6.6 mM MgCl₂). Formation of resorufin was monitored fluorometrically over a 2-min time period and quantitated by reference to a standard curve of fluorescence vs. resorufin (0-0.2 nmol/ml).

Diclofenac 4-Hydroxylase Activity. Diclofenac 4-hydroxylation was used to assess the functional activity of human P4502C9-Arg₁₄₄ samples, as previously described (15). Microsomal pellets were rehomogenized in 0.1 M Tris buffer (pH 7.5) to yield a protein concentration of ~0.8 mg/ml. The reaction was initiated by the addition of 0.5 ml of the microsomal suspension to 0.5 ml of prewarmed (37°C) cofactor solution (containing 0.2 mM diclofenac, 4 mM NADPH, and 3.3 mM MgCl₂), and the reaction was allowed to proceed at 37°C for 20 min. After 20 min, the reaction was terminated by addition of 0.2 ml of 94% acetonitrile/6% glacial acetic acid to the reaction mixture and placing the samples on ice. Diclofenac and its reaction product, 4-hydroxydiclofenac, were separated by HPLC, and detected by absorbance at 280 nm. 4-Hydroxydiclofenac was quantitated by comparison to a standard curve of absorbance vs. 4-hydroxydiclofenac (0-20 nmol/ml).

Dextromethorphan O-demethylase Activity. Dextromethorphan O-demethylase activity was used to assess the functional activity of human P4502D6-Val₃₇₄ samples. Microsomal pellets were rehomogenized in 0.1 M K₂HPO₄ buffer (pH 7.4) to yield a protein concentration of ~0.8 mg/ml. The reaction was initiated by addition of 1 ml of microsomes to 1 ml of prewarmed (37°C) cofactor solution (containing 40 µM dextromethorphan, 4 mM NADPH, and 6.6 mM MgCl₂), and the reaction was allowed to proceed for 20 min. After 20 min, the reaction was terminated by addition of 2 ml of ice-cold 0.1 M sodium acetate buffer (pH 5.0). The samples were extracted, and dextromethorphan and its reaction product, dextrorphan, were measured by HPLC using fluorescence detection, according to the method of Jacqz-Aigrain et al. (16), with modifications as described by Jurima-Romet et al. (17). Dextrorphan was quantitated by comparison to a standard curve of fluorescence vs. dextrorphan (1-80 nmol/ml).

Testosterone 6-Hydroxylase Activity. Testosterone 6-hydroxylation was used to assess the functional activity of human P4503A4 samples (18, 19). Microsomal pellets were rehomogenized in 0.1 M K₂HPO₄ buffer (pH 7.4) to yield a protein concentration of ~0.8 mg/ml. The microsomal suspension (0.1 ml) was added to 0.38 ml of prewarmed (37°C) cofactor solution containing NADPH and MgCl₂. Thirty seconds later, the reaction was initiated by removing an 80 µl aliquot from this mixture, and adding it to 20 µl of [4-¹⁴C]testosterone (5.33 MBq) in phosphate buffer. After an additional 3-min incubation, the reaction was quenched with 50 µl of tetrahydrofuran, and the samples were placed on ice. The final incubation conditions were: 0.1 mg/ml microsomal protein, 25 µM testosterone, 1 mM NADPH, and 15 mM MgCl₂ in 0.1 M K₂HPO₄ buffer (pH 7.4). Testosterone was separated from its reaction product, 6-hydroxytestosterone by TLC; after localization by autoradiography, the radioactive areas were quantitated by liquid scintillation counting.

Cytochrome c Reductase Activity. Microsomal pellets were rehomogenized in 50 mM K₂HPO₄/KH₂PO₄ (KPi) buffer (pH 7.7) containing 0.1 mM EDTA, and the reduction of cytochrome c was used to assess the functional activity of human P450 oxidoreductase samples, as previously described (20).

Protein Determination. Quantitation of protein content for all experiments was performed according to the method of Lowry et al. (21).

Statistical Analysis. For all experiments, a repeated-measures design one-way ANOVA was used to determine if means differed significantly (p  0.05). If a significant F ratio was observed in the ANOVA at the 0.05 level, a Newman-Keuls post-hoc test was used to indicate the means that differed significantly from each other.

	Results and Discussion

Abstract Introduction Results & Discussion References

Human hepatic P450s 1A2, 2C9, 2D6, and 3A4 are among the most abundantly expressed P450 isozymes of human liver, and are responsible for the biotransformation of most drugs that undergo P450-mediated metabolism in human liver (22). Although several rat and chick embryo hepatic P450s have been identified as targets for mechanism-based inactivation by DDEP and TTMS (7-12), and several rat hepatic P450s have been identified as targets for mechanism-based inactivation by AIA (23-25), little is known regarding which human hepatic P450 isozymes are targeted for inactivation by these compounds.

In the current study, the effect of in vitro incubation of DDEP, TTMS, or AIA on catalytic activities for cDNA-expressed human P450s 1A1, 1A2, 2C9, 2D6, and 3A4 was examined, as was the effect on cDNA-expressed human P450 oxidoreductase. Microsomes prepared from the human lymphoblastoid AHH-1 TK+/ cell line, containing single cDNA-expressed P450s were used for this purpose. Control microsomes were either from untransfected AHH-1 TK+/ cells, or from AHH-1 TK+/ cells that were transfected with the plasmid vector, which was devoid of the cDNA for human P450s. In control microsomes, we were unable to detect any marker enzyme activity, namely EROD, diclofenac hydroxylase, dextromethorphan O-demethylase, or testosterone 6-hydroxylase. Neither DDEP, TTMS, nor AIA significantly affected the rate of cytochrome c reduction after incubation with these microsomes in the presence of NADPH (results not shown). Therefore, the decreased rate of catalytic activity observed in microsomes possessing cDNA-expressed human P450s after the incubation of porphyrinogenic compounds and NADPH is the result of mechanism-based inactivation of the P450 isozymes, rather than inactivation of P450 oxidoreductase.

When 45 µM DDEP or TTMS, or AIA (45, 450, or 1,000 µM) were incubated in vitro with microsomes possessing either cDNA-expressed human P450s 1A1 or 1A2, only DDEP and TTMS were observed to cause an NADPH-dependent decrease in EROD activity measured in these microsomes (figs. 1 and 2). AIA was observed to be inactive. The inactivation of cDNA-expressed human P4501A1 and 1A2 isozymes by DDEP and TTMS in the current study corresponds to a similar inactivation of P4501A1 in rat liver (7, 10, 11), and a P4501A isozyme in chick embryo liver (11, 26). The fact that AIA did not cause mechanism-based inactivation of either cDNA-expressed human P4501A1 and 1A2 isozymes corresponds to observations previously made with the corresponding rat liver isozymes (25). Thus, the results previously observed with DDEP, TTMS, and AIA in rats, and DDEP and TTMS in the chick embryo agreed with results obtained with cDNA-expressed human P4501A isozymes.

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Effect of DDEP, TTMS, and AIA on EROD activity in microsomes containing cDNA-expressed human P4501A1. Microsomes were incubated for 30 min with the respective xenobiotic or xenobiotic vehicle to serve as controls, in the presence or absence of NADPH, reharvested by ultracentrifugation, and then assayed for enzyme activity. Each bar represents the mean (±SD) of five experiments. * Significantly different (p  0.05) from control ( xenobiotic; + NADPH) and incubation in the presence of the xenobiotic in the absence of NADPH (+ xenobiotic;  NADPH), as determined by a repeated-measures design one-way ANOVA with Newman-Keuls test.

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Effect of DDEP, TTMS, and AIA on EROD activity in microsomes containing cDNA-expressed human P4501A2. Microsomes were incubated for 30 min with the respective xenobiotic or xenobiotic vehicle to serve as controls, in the presence or absence of NADPH, reharvested by ultracentrifugation, and then assayed for enzyme activity. Each bar represents the mean (±SD) of five experiments. * Significantly different (p  0.05) from control ( xenobiotic; + NADPH) and incubation in the presence of the xenobiotic in the absence of NADPH (+ xenobiotic;  NADPH), as determined by a repeated-measures design one-way ANOVA with Newman-Keuls test.

When DDEP and TTMS were incubated in vitro with hepatic microsomes from phenobarbital-pretreated rats, P450s 2C6 and 2C11 were found to be targeted for mechanism-based inactivation (7-10, 12). When AIA was administered in vivo to phenobarbital-treated rats, inactivation of P450s 2C6 and 2C11 was reported (24). In vitro mechanism-based inactivation of cDNA-expressed human P4502C9 was observed after incubation with 45 and 450 µM DDEP in the current study, using diclofenac 4-hydroxylase activity to monitor the functional activity of this isozyme (fig. 3). Neither TTMS nor AIA (45 and 450 µM) caused an NADPH-dependent inactivation in the functional activity of this isozyme. Because no P4502C-like isozymes have been characterized in chick embryo liver, the effect of these porphyrinogenic compounds on possible P4502C-like isozymes in chick embryo liver has not been elucidated. Thus, no comparison can be made between the inactivation of human and possible chick embryo hepatic P4502C isozymes by DDEP, TTMS, or AIA.

View larger version (39K): [in this window] [in a new window]   Fig. 3.   Effect of DDEP, TTMS, and AIA on diclofenac 4-hydroxylase activity in microsomes containing cDNA-expressed human P4502C9-Arg144. Microsomes were incubated for 30 min with the respective xenobiotic or xenobiotic vehicle to serve as controls, in the presence or absence of NADPH, reharvested by ultracentrifugation, and then assayed for enzyme activity. Each bar represents the mean (±SD) of three experiments. * Significantly different (p  0.05) from control ( xenobiotic; + NADPH) and incubation in the presence of the xenobiotic in the absence of NADPH (+ xenobiotic;  NADPH). + Significantly different (p  0.05) from control ( xenobiotic;  NADPH), as determined by a repeated-measures design one-way ANOVA with Newman-Keuls test.

Tephly et al. (7) found that incubation of rat hepatic microsomes with DDEP did not affect rat hepatic P4502D1. As yet, no P4502D-like isozymes have been characterized in chick embryo liver. When 45 and 450 µM DDEP, TTMS, or AIA were incubated in vitro with microsomes possessing cDNA-expressed human P4502D6, no NADPH-dependent effect on dextromethorphan O-demethylase activity, a diagnostic marker for human P4502D6 was observed (fig. 4). On the other hand, dextrorphan formation by cDNA expressed P4502D6 microsomes was inhibited by quinidine (5 µM), the prototype P4502D6 inhibitor. We conclude that cDNA-expressed human P4502D6 is not targeted for in vitro mechanism-based inactivation by these porphyrinogenic xenobiotics.

View larger version (42K): [in this window] [in a new window]   Fig. 4.   Effect of DDEP, TTMS, and AIA on dextromethorphan O-demethylase activity in microsomes containing cDNA-expressed human P4502D6-Val374. Microsomes were incubated for 30 min with the respective xenobiotic or xenobiotic vehicle to serve as controls, in the presence or absence of NADPH, reharvested by ultracentrifugation, and then assayed for enzyme activity. Each bar represents the mean (±SD) of three experiments.

P4503A isozymes in rat hepatic microsomes and P4503A-like isozyme(s) in chick embryo hepatic microsomes have been identified as targets for in vitro mechanism-based inactivation by DDEP and TTMS (7-12). When AIA was administered in vivo to phenobarbital-treated rats, inactivation of P4503A was reported (24). Moreover, some evidence for inactivation of P4503A was obtained in in vitro studies using rat liver microsomes from dexamethasone-treated rats (24). In the current study, 45 µM DDEP or TTMS, or AIA (45, 450, or 1,000 µM) were incubated, in the presence of NADPH, with microsomes possessing cDNA-expressed human P4503A4, and the effect on testosterone 6-hydroxylase activity was measured. It was found that both DDEP and TTMS, but not AIA, caused significant NADPH-dependent inactivation of testosterone 6-hydroxylase activity after incubation with these microsomes, indicating that cDNA-expressed human P4503A4 is targeted for mechanism-based inactivation by DDEP and TTMS (fig. 5). Thus, the results previously observed with DDEP and TTMS, but not with AIA, in animal species, agreed with results found with cDNA-expressed human P4503A4.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Effect of DDEP, TTMS, and AIA on testosterone 6-hydroxylase activity in microsomes containing cDNA-expressed human P4503A4. Microsomes were incubated for 30 min with the respective xenobiotic or xenobiotic vehicle to serve as controls, in the presence or absence of NADPH, reharvested by ultracentrifugation, and then assayed for enzyme activity. Each bar represents the mean (±SD) of five experiments. * Significantly different (p  0.05) from control ( xenobiotic; + NADPH) and incubation in the presence of the xenobiotic in the absence of NADPH (+ xenobiotic;  NADPH), as determined by a repeated-measures design one-way ANOVA with Newman-Keuls test.

It is of interest that rat hepatic P450s 1A and 3A and chick embryo hepatic P450s 1A and 3A-like isozymes (7-11, 23, 26), as well as cDNA-expressed human P450s 1A1, 1A2, and 3A4 undergo in vitro mechanism-based inactivation by DDEP and TTMS. DDEP, but not TTMS, caused mechanism-based inactivation of cDNA-expressed human P4502C9. With respect to DDEP and TTMS, a comparison of the human P450 isozyme data with results obtained in rat and chick embryo liver show a close correspondence between results observed with the P450s 1A and 3A, but not the P4502C subfamily. In vivo administration of AIA to phenobarbital-treated rats results in the inactivation of P450s 2A1, 2C6, 2C11, and 3A in addition to P4502B1 (24), and there is some evidence for inactivation of P4503A in the in vitro studies from dexamethasone-treated rats (24). Clearly, there is no correspondence between results obtained with AIA between animal and human P450 isozymes. The previously described results argue for the importance of supplementing studies of mechanism-based inactivation of xenobiotics in animal models with similar studies on cDNA-expressed human P450 isozymes.

Bocker and Guengerich (27) have demonstrated that after incubation of human liver microsomes with DDEP (200 µM), phenacetin O-deethylase, hexobarbital 3-hydroxylase, and nifedipine oxidase activities are significantly decreased in an NADPH-dependent manner. These activities are predominantly associated with human P450s 1A2, 2C9, and 3A4, respectively (28). However, the inactivation of human P4501A2 (phenacetin O-deethylase) and P4502C9 (hexobarbital 3-hydroxylase) activities was implied to occur through migration of alkyl radicals formed at the active site of P4503A4 (29). Thus, DDEP was suggested to cause mechanism-based inactivation of human P4503A4, and the effect of DDEP on human P450s 1A2 and 2C9 was suggested to be indirect. However, the in vitro inactivation of cDNA-expressed human P4501A2 and P4502C9 in the current studies show that these isozymes undergo direct inactivation by DDEP, as opposed to indirect inactivation.

The advantage of using expression systems in our studies is that the effect of porphyrinogenic xenobiotics on individual P450 isozymes can be determined, without interference by other P450 isozymes that may obscure observation of mechanism-based inactivation. However, results obtained with these expression systems may not reflect the overall in vivo effects of these compounds. Therefore, when discussing the mechanism-based inactivation of cDNA-expressed human P450s, it is with the understanding that differences may exist with the in vivo metabolism of such compounds, which is dependent on both the specific activity and relative abundance of each isozyme. A logical follow up study to the present is to study mechanism-based inactivation of P450 isozymes with DDEP, TTMS, and AIA in human liver microsomes that are currently commercially available.

It is necessary to emphasize that mechanism-based inactivation of P450 isozymes by xenobiotics can, in addition to alkylating the nitrogen atoms of the heme moiety with formation of ferrochelatase-inhibitory N-alkylPPs, also proceed along alternative pathways (30). Thus, mechanism-based inactivation may cause alkylation of the P450 isozyme apoprotein or activation of the prosthetic heme so that it forms a covalent bond to the P450 apoprotein. Because hepatic porphyrin accumulation will result only from N-alkylation of the heme moiety, it is important to determine which of the P450 isozymes targeted for mechanism-based inactivation results in N-alkylPP formation. This is the direction in which our future research is proceeding.