Of several furanocoumarins [5-methoxypsoralen (5-MOP), 8-methoxypsoralen (8-MOP), 5-hydroxypsoralen (5-OH-P), 8-hydroxypsoralen (8-OH-P), 4',5'-dihydro-8-MOP (DH-8-MOP), and psoralen (P)] tested as mechanism-based inactivators (MBIs) of purified reconstituted cytochrome P450 (P450) 2B1, 8-MOP was found to be the most potent (KI, kinact, and partition ratio of 2.9 microM, 0.34 min-1, and 1.3, respectively). The inactivation was not prevented by reactive oxygen species scavengers or nucleophilic trapping agents and proceeded with a decrease in P450 spectral content. Liquid chromatography (LC) separation of the reconstituted enzyme mixture, followed by liquid scintillation counting, indicated that [14C]-8-MOP binding was specific to the apoprotein of P450 2B1 with a binding stoichiometry of 0.7:1. The major metabolites formed by P450 2B1 from the furanocoumarins that were MBIs were characterized by LC electrospray ionization tandem mass spectrometry (ESI-MS/MS) as dihydro diols. Results from H218O incorporation experiments supported initial oxidation of 8-MOP and P to an epoxide which can react with some nucleophilic active site residue and inactivate the enzyme or partition to a dihydro diol metabolite by hydrolytic ring opening. On the other hand, 5-MOP was converted to an epoxide or gamma-keto enal intermediate prior to inactivation or dihydro diol formation. Comparison of the ESI mass spectra of P450 2B1 and furanocoumarin exposed P450 2B1, indicated a mass difference consistent with the covalent addition of a furanoepoxide to P450 2B1.

Acetaminophen (APAP), a widely used analgesic and antipyretic agent, is bioactivated by cytochromes P450 to cause severe hepatotoxicity. APAP is oxidized by two pathways to form a toxic intermediate, N-acetyl-p-benzoquinone imine (NAPQI), and a nontoxic catechol metabolite, 3-hydroxy-APAP (3-OH-APAP). We investigated the role of P450 2E1 and 2A6 in APAP oxidation by using baculovirus-expressed and highly purified forms of human P450 2E1 and 2A6. An electrochemical HPLC assay was developed to quantify both oxidative metabolites simultaneously. For the first time, it was demonstrated that human P450 2E1 selectively oxidized APAP to NAPQI (assayed as its glutathione conjugate, GS-APAP), whereas human P450 2A6 selectively oxidized APAP to 3-OH-APAP. At 1 mM APAP, the relative ratio for the formation of GS-APAP vs 3-OH-APAP with human P450 2E1 was approximately 6:1, whereas the ratio with human P450 2A6 was 1:3. Apparent Km and Vmax values for the formation of GS-APAP by human P450 2E1 were 1.3 mM and 6.9 nmol/min/nmol of P450, respectively, whereas they were 4.6 mM and 7.9 nmol/min/nmol of P450 for P450 2A6. Apparent Km and Vmax values for the formation of 3-OH-APAP by human P450 2E1 were 4.0 mM and 2.5 nmol/min/nmol of P450, respectively, whereas they were 2.2 mM and 14.2 nmol/min/nmol of P450, respectively, for P450 2A6. Thus, although at toxic doses of APAP P450 2E1 is the more efficient catalyst for the formation of the toxic metabolite NAPQI, P450 2A6 also can contribute significantly to NAPQI production.

Several furanocoumarins were tested for their ability to inhibit human P450 2A6 activity. The metabolites and conjugates formed from these furanocoumarins after incubation with reconstituted purified P450 2A6 in the absence and presence of exogenous nucleophiles were characterized by UV and LC/ESI-MS/MS analysis. The results suggest initial oxidation to form a furanoepoxide followed by hydrolytic attack, or attack of exogenous nucleophiles, to form dihydrofuranocoumarin products. Initial epoxidation is confirmed by the finding that a single 18O atom is incorporated into the 8-methoxypsoralen (8-MOP) and psoralen (P) dihydrodiol metabolites when the incubations are performed in the presence of H218O. In contrast, 19% of the dihydrodiol formed from 5-methoxypsoralen (5-MOP) involves incorporation of two 18O atoms, implicating a gamma-ketoenal intermediate in the formation of this metabolite. Thus, the structure of the reactive intermediate(s) formed is dictated by the intrinsic electronic properties of the parent compound. After exposure to [14C]-8-MOP and [14C]-5-MOP, SDS-PAGE and HPLC experiments, followed by radiometric detection, indicated that both P450 2A6 and P450 reductase were covalently modified in the purified system. In contrast, only P450 2A6 was covalently modified in a lymphablastoid cell line (GENTEST). With the purified system, partition ratios were higher (1.5-3.9X), and the ability to scavenge reactive intermediates with exogenous nucleophiles was greater. These results suggest that relative to the cell system, more reactive intermediates can escape, rather than bind to, the active site of purified reconstituted P450 2A6.

A general scheme for the purification of baculovirus-expressed cytochrome P450s (P450s) from the crude insect cell pastes has been designed which renders the P450s suitable for analysis by high-performance liquid chromatography (HPLC) electrospray ionization mass spectrometry (ESI-MS). An HPLC/ESI-MS procedure has been developed to analyze small amounts of intact purified P450 (P450s cam-HT, 1A1, 1A2, 2A6, 2B1, 2C9, 2C9 C175R, 3A4, 3A4-HT) and rat NADPH cytochrome P450 reductase (P450 reductase). The experimentally determined and predicted (based on the amino acid sequences) molecular masses (MMs) of the various proteins had identical rank orders. For each individual protein, the difference between the experimentally determined (+/-SD, based on experiments performed on at least 3 different days) and predicted MMs ranged from 0.002 to 0.035%. Each experimentally determined MM had a standard deviation of less than 0.09% (based on the charge state distribution). Application of this HPLC/ESI-MS technique made the detection of the covalent modification to P450 2C9 following mechanism-based inactivation by tienilic acid possible. In the absence of glutathione, three P450 2C9 species were detected that produced ESI mass spectra corresponding to native P450 2C9 and both a monoadduct and a diadduct of tienilic acid to P450 2C9. In the presence of glutathione, only native P450 2C9 and the monoadduct were detected. Based on the observed mass shifts for the P450 2C9/tienilic acid adducts, a mechanism for the inactivation of P450 2C9 by tienilic acid is proposed.