Prototype Systems Containing Human Cytochrome P450 for High-Throughput Real-Time Detection of DNA Damage by Compounds That Form DNA-Reactive Metabolites.

The formation of reactive metabolites through biotransformation is the suspected cause of many adverse drug reactions. Testing for the propensity of a drug to form reactive metabolites has increasingly become an integral part of lead-optimization strategy in drug discovery. DNA reactivity is one undesirable facet of a drug or its metabolites and can lead to increased risk of cancer and reproductive toxicity. Many drugs are metabolized by cytochromes P450 in the liver and other tissues, and these reactions can generate hard electrophiles. These hard electrophilic reactive metabolites may react with DNA and may be detected in standard in vitro genotoxicity assays; however, the majority of these assays fall short due to the use of animal-derived organ extracts that inadequately represent human metabolism. The current study describes the development of bacterial systems that efficiently detect DNA-damaging electrophilic reactive metabolites generated by human P450 biotransformation. These assays use a GFP reporter system that detects DNA damage through induction of the SOS response and a GFP reporter to control for cytotoxicity. Two human CYP1A2-competent prototypes presented here have appropriate characteristics for the detection of DNA-damaging reactive metabolites in a high-throughput manner. The advantages of this approach include a short assay time (120-180 min) with real-time measurement, sensitivity to small amounts of compound, and adaptability to a microplate format. These systems are suitable for high-throughput assays and can serve as prototypes for the development of future enhanced versions.

Comparison of the 17 alpha-hydroxylase/C17,20-lyase activities of porcine, guinea pig and bovine P450c17 using purified recombinant fusion proteins containing P450c17 linked to NADPH-P450 reductase.

The cDNAs for cytochrome P450c17 (P450c17) of three species, pig, guinea pig, and cow, representing three families of mammals (suidae, procaviidae, and bovidae, respectively) were each engineered into an expression plasmid (pCWori+). The P450c17 domain of the coding sequence was connected to a truncated form of rat NADPH-P450 reductase by a linker sequence encoding two amino acids (SerThr). These fusion proteins were expressed in E. coli and purified for use in enzymatic assays to determine similarities and differences in 17 alpha-hydroxylase and lyase activities. The fusion proteins were found to catalyze both the 17 alpha-hydroxylation of progesterone (P4) and pregnenolone (P5) to 17 alpha-hydroxylated P4 and P5 (17 alpha-OH P4 and 17 alpha-OH P5) followed by the C17,20-lyase reaction for the conversion of these C(21)-17 alpha-hydroxylated steroids to C(19)-steroids (the C17,20-lyase reaction). These in vitro studies show that (a) porcine P450c17 possesses cytochrome b(5) (b(5))-stimulated C17,20-lyase activity that converts 17 alpha OH-P4 to androstenedione (AD) but also converts 17 alpha-OHP5 to dehydroepiandrosterone (DHEA); (b) guinea pig P450c17 possesses a b(5)-stimulated C17,20-lyase activity that converts 17 alpha-OH P4 to AD but does not convert 17 alpha-OH P5 to DHEA., and (c) bovine P450c17 possesses a b(5)-stimulated C17,20-lyase activity that converts 17 alpha-OH P5 to DHEA but does not convert 17 alpha-OH P4 to AD. Thus, the P450c17 of each species differs in its ability to catalyze in vitro the conversion of C(21)-steroids to C(19)-steroids. In addition, each P450c17 is capable of catalyzing additional hydroxylation reactions leading to low levels of 2 alpha-, 6 beta-, 16- and 21-hydroxy-metabolites. Porcine P450c17 also catalyzes the b(5)-dependent synthesis of andien-beta (androsta-5,16-dien-3beta-ol) from P5. When the amino acid sequences of the three P450c17s were aligned there was an approximate 50% variation in the alignment identity (227 differences in the sequences of 509 amino acids). Alignment did not permit the assignment of specific amino acids or domains to the observed differences in enzymatic activities.

The effect of reciprocal active site mutations in human cytochromes P450 1A1 and 1A2 on alkoxyresorufin metabolism.

Five reciprocal active site mutants of P450 1A1 and 1A2 and an additional mutant, Val/Leu-382 --> Ala, were constructed, expressed in Escherichia coli, and purified by Ni-NTA affinity chromatography. In nearly every case, the residue replacement led to loss of 7-methoxy- and 7-ethoxyresorufin O-dealkylase activity compared to the wild-type enzymes, except for the P450 1A1 S122T mutation which increased both activities. Mutations at position 382 in both P450 1A1 and 1A2 shifted substrate specificity from one enzyme to another, confirming the importance of this residue. Changes in activity of P450 1A enzymes upon amino acid replacement were, in general, consistent with molecular dynamics analyses of substrate motion in the active site of homology models.

Microsomal cytochrome P450 levels and activities of isolated rat livers perfused with albumin.

PURPOSE: We recently showed that the perfusion of isolated rat livers with perfusates containing bovine serum albumin (BSA) would significantly stimulate the release of tumor necrosis factor (TNF)-alpha. Here, we hypothesize that BSA-induced increase in the release of TNF-alpha, and possibly other cytokines, would affect cytochrome P450 (CYP)-mediated drug metabolism. METHODS: Rat livers were perfused ex vivo for 1, 2, or 3 h with a physiologic buffer containing or lacking 1% BSA (n = 4-5/group). At the end of perfusion, liver microsomes were prepared and analyzed for their total CYP, CYP2E1, CYP3A2, and CYP2C11 protein contents and the activities of cytochrome c reductase, CYP2E1, CYP3A2, CYP2C11, CYP2E1, CYP2D1, CYP1A1, and CYP2B1/2. In addition, the concentrations of various cytokines and nitric oxide were quantified in the outlet perfusate. RESULTS: In the absence of BSA, the perfusate levels of all measured cytokines and nitric oxide were low. However, when the perfusate contained BSA, the levels of TNF-alpha, interleukin-6, and nitric oxide increased significantly (p < 0.005). Perfusion of the livers for 3 h with the BSA-containing perfusate resulted in significant (p < 0.05) decreases in the total CYP (41%), CYP2E1 (59%), CYP3A2 (68%), and CYP2C11 (50%) protein contents and activities of cytochrome c reductase (31%), CYP2E1 (66%), CYP3A2 (54%), and CYP2G11 (51%). In contrast, perfusion of livers for 1 or 2 h with the BSA perfusate did not have any significant effect on CYP-mediated metabolism. The CYP1A2, CYP2D1, and CYP2B1/2 activities were not affected by BSA, regardless of perfusion time. CONCLUSION: Addition of BSA to perfusates, which is a routine practice in isolated rat liver studies, can reduce CYP-mediated drug metabolism by a mechanism independent of protein-binding effect.

Characterization of substrate binding to cytochrome P450 1A1 using molecular modeling and kinetic analyses: case of residue 382.

Key residue Val-382 in P450 1A1 has been predicted to interact with the alkoxy chain of resorufin derivatives. Therefore, we undertook a detailed analysis of substrate mobility in the active site of the P450 1A1 homology model and assessed the effect of mutations at position 382. Dynamic trajectories of 7-methoxy-, 7-ethoxy-, and 7-pentoxyresorufin indicated that 7-ethoxyresorufin would be oxidized most efficiently by the wild-type enzyme. The Val-382-->Ala mutation would increase the O-dealkylation of 7-pentoxyresorufin but decrease the oxidation of other substrates. In the case of the V382L mutant, the large bulk of Leu would block alkoxyresorufins from productive binding orientations leading to lowered activities. Binding free energy calculations for three substrates with 1A1 WT and two mutants indicated that binding constants would be similar for all enzyme-substrate combinations. Modeling predictions were tested experimentally. The plasmid containing the cDNA for human P450 1A1 modified for bacterial expression was altered to include a C-terminal PCR-generated six histidine domain to facilitate enzyme purification. The V382A and V382L mutants were constructed by site-directed mutagenesis and Escherichia coli-expressed enzymes purified using Ni-NTA affinity chromatography. The activity of the WT 1A1 was highest toward 7-ethoxyresorufin and lowest toward 7-pentoxyresorufin. Both mutants displayed a decrease in V(max) with 7-methoxy- and 7-ethoxyresorufin, whereas for the V382A mutant, V(max) with 7-pentoxyresorufin was increased. No significant changes in K(m) were observed relative to the wild-type enzyme. The experimental results are thus in good agreement with modeling predictions.