Selection of human cytochrome P450 1A2 mutants with enhanced catalytic activity for heterocyclic amine N-hydroxylation

Cytochrome P450 (P450) 1A2 is the major enzyme involved in the metabolism of 2-amino-3,5-dimethylimidazo[4,5-f]quinoline (MeIQ) and other heterocyclic arylamines and their bioactivation to mutagens. Random mutant libraries of human P450 1A2, in which mutations were made throughout the entire open reading frame, were screened with Escherichia coli DJ3109pNM12, a strain designed to bioactivate MeIQ and detect mutagenicity of the products. Mutant clones with enhanced activity were confirmed using quantitative measurement of MeIQ N-hydroxylation. Three consecutive rounds of random mutagenesis and screening were performed and yielded a highly improved P450 1A2 mutant, SF513 (E225N/Q258H/G437D), with >10-fold increased MeIQ activation based on the E. coli genotoxicity assay and 12-fold enhanced catalytic efficiency (k(cat)/K(m)) in steady-state N-hydroxylation assays done with isolated membrane fractions. SF513 displayed selectively enhanced activity for MeIQ compared to other heterocyclic arylamines. The enhanced catalytic activity was not attributed to changes in any of several individual steps examined, including substrate binding, total NADPH oxidation, or H(2)O(2) formation. Homology modeling based on an X-ray structure of rabbit P450 2C5 suggested that the E225N and Q258H mutations are located in the F-helix and G-helix, respectively, and that the G437D mutation is in the "meander" region, apparently rather distant from the substrate. In summary, the approach generated a mutant enzyme with selectively elevated activity for a single substrate, even to the extent of a difference of a single methyl group, and several mutations had interacting roles in the development of the selected mutant protein.     

Inter-individual differences in the metabolism of environmental toxicants: cytochrome P450 1A2 as a prototype.

Cytochrome P450 (P450) 1A2 provides an interesting paradigm for inter-individual differences in the metabolism of pro-carcinogens. The enzyme is known to vary 40-fold among individuals and may contribute to cancers caused by heterocyclic amines and other chemicals. Rat and human P450 1A2 are known to be 75% identical and were compared for several catalytic activities. The human enzyme was an order of magnitude more efficient in the N-hydroxylation of two heterocyclic amines. Further, the levels of P450 1A2 expressed in human livers show a 40-fold variation, with some as high as 0.25 nmol P450 1A2 per milligram microsomal protein. Some human liver samples are more active (than those isolated from polychlorinated biphenyl-treated rats) in the activation of heterocyclic amines. A bacterial genotoxicity assay has been developed in which human P450 1A2 and NADPH-P450 reductase are expressed within Escherichia coli and bacterial mutants can be assayed using reversion to lac prototrophy. A random mutagenesis strategy for human P450 1A2 has been developed and used to examine the changes in catalytic activity seen with many single-amino acid substitutions. These results may be of relevance in consideration of genetic polymorphisms. Further, the findings pose a challenge to molecular epidemiology effort in that results with one substrate do not necessarily predict those for others. Some dinitropyrenes are P450 1A2 substrates but others are not. 6-Nitrochrysene can be activated by human P450 1A2 but the (mono) nitropyrenes examined were not; these were oxidized by P450 3A4 instead.     

Analysis of coumarin 7-hydroxylation activity of cytochrome P450 2A6 using random mutagenesis

Cytochrome P450 (P450) 2A6 is an important human enzyme involved in the metabolism of many xenobiotic chemicals including coumarin, indole, nicotine, and carcinogenic nitrosamines. A combination of random mutagenesis and high-throughput screening was used in the analysis of P450 2A6, utilizing a fluorescent coumarin 7-hydroxylation assay. The steady-state kinetic parameters (k(cat) and Km) for coumarin 7-hydroxylation by wild-type P450 2A6 and 35 selected mutants were measured and indicated that mutants throughout the coding region can have effects on activity. Five mutants showing decreased catalytic efficiency (k(cat)/Km) were further analyzed for substrate selectivity and binding affinities and showed reduced catalytic activities for 7-methoxycoumarin O-demethylation, tert-butyl methyl ether O-demethylation, and indole 3-hydroxylation. All mutants except one (K476E) showed decreased coumarin binding affinities (and also higher Km values), indicating that this is a major basis for the decreased enzymatic activities. A recent x-ray crystal structure of P450 2A6 bound to coumarin (Yano, J. K., Hsu, M. H., Griffin, K. J., Stout, C. D., and Johnson, E. F. (2005) Nat. Struct. Mol. Biol. 12, 822-823) indicates that the recovered A481T and N297S mutations appear to be close to coumarin, suggesting direct perturbation of substrate interaction. The decreased enzymatic activity of the K476E mutant was associated with decreases both in NADPH oxidation and the reduction rate of the ferric P450 2A6-coumarin complex. The attenuation is caused in part to lower binding affinity for NADPH-P450 reductase, but the K476E mutant did not achieve the wild-type coumarin 7-hydroxylation activity even at high reductase concentrations.     

Enhancement of 7-methoxyresorufin O-demethylation activity of human cytochrome P450 1A2 by molecular breeding

Alkylresorufins are model substrates for cytochrome P450 (P450) 1A2. The ability of human P450 1A2 to catalyze 7-methoxyresorufin O-demethylation was improved by screening of random mutant libraries (expressed in Escherichia coli) on the basis of 7-methoxyresorufin O-demethylation. After three rounds of mutagenesis and screening, the triple mutant E163K/V193M/K170Q yielded a kcat > five times faster than wild type P450 1A2 in steady-state kinetic analysis using either isolated membrane fractions or purified, reconstituted enzymes. The enhanced catalytic activity was not attributed to changes in substrate affinity. The kinetic hydrogen isotope effect of the triple mutant did not change from wild type enzyme and suggests that C-H bond cleavage is rate-limiting in both enzymes. Homology modeling, based on an X-ray structure of rabbit P450 2C5, suggests that the locations of mutated residues are not close to the substrate binding site and therefore that structural elements outside of this site play roles in changing the catalytic activity. This approach has potential value in understanding P450 1A2 and generating engineered enzymes with enhanced catalytic activity.     

Modification of a catalytically important residue of indoleglycerol-phosphate synthase from Escherichia coli

The active-site residues of indoleglycerol-phosphate synthase from Escherichia coli were tentatively localized by comparing crystallographic data with the amino acid identities among the known indoleglycerol-phosphate synthase sequences. To test the validity of the resulting model of catalysis one of the residues in the presumptive active site, Lys 55, was changed to serine using oligonucleotide-directed mutagenesis. The specificity constant kcat/Km of the mutant is 3 x 10(4)-times lower than that of the wild-type enzyme, due to a 60-fold decrease in kcat and a 450-fold increase in Km. This finding shows that Lys 55 is important for both catalysis and substrate binding.     

The mechanism of ATP inhibition of wild type and mutant phosphofructo-1-kinase from Escherichia coli.

Escherichia coli 6-phosphofructo-1-kinase was inhibited by high concentrations of ATP at alkaline pH. The mechanism of the inhibition was studied with two mutants generated by site-directed mutagenesis; I126A, with a Km for fructose-6-P that was more than two orders of magnitude higher than that of wild type but with minimal changes in kcat and Km for ATP, and R72H, with little change in substrate half-saturation concentrations but with a kcat that was 300-fold lower that of wild type enzyme. ATP and fructose-6-P interacted in a mutually antagonistic manner; that is ATP decreased the apparent affinity for fructose-6-P and vice versa. The half-saturation concentrations for both substrates, most strikingly fructose-6-P, increased with increasing pH while the kcat increased. Studies with I126A suggested that ATP inhibition was not dependent on a dissociable group with a pK in the alkaline range and that the inhibition was not caused by abortive binding of substrate to the wrong substrate site. Inhibition was not the result of differential affinity of ATP for the R and T states of the enzyme. The low kcat mutant, R72H, did not display ATP inhibition. These data indicate that ATP inhibition results from substrate antagonism coupled with a steady state random mechanism wherein the high rate of catalysis does not permit equilibration of substrates.     

Functional characterization of four allelic variants of human cytochrome P450 1A2

Human cytochrome P450 1A2 catalyzes important reactions in xenobiotic metabolism, including the N-hydroxylation of carcinogenic aromatic amines. In 2001, Chevalier et al. reported four new P450 1A2 sequence variants in the human population. We have now expressed these variants in Escherichia coli and measured protein expression (optical spectroscopy of holoenzyme and immunoblotting) and bioactivation of IQ (2-amino-3-methylimidazo[4,5-f]quinoline) and MeIQ (2-amino-2,4-dimethylimidazo[4,5-f]quinoline) in the lacZ reversion mutagenicity test. Enzyme kinetic analyses were performed for N-hydroxylation of five heterocyclic amine substrates and for O-deethylation of phenacetin. The most drastic effect was that of the R431W substitution: no holoenzyme was detectable. This residue is located in the "meander" peptide region and earlier site-directed mutagenesis studies demonstrated that it is critical for maintenance of protein tertiary structure. The other three variants had subtly different catalytic activities compared to the wild-type enzyme.     

Plasmid-mediated expression of the UmuDC mutagenesis proteins in an Escherichia coli strain engineered for human cytochrome P450 1A2-catalyzed activation of aromatic amines.

The mutagenic actions of many chemicals depend on the activities of bacterial "mutagenesis proteins", which allow replicative bypass of DNA lesions. Genes encoding these proteins occur on bacterial chromosomes and plasmids, often in the form of an operon (such as umuDC or mucAB) encoding two proteins. Many bacterial strains used in mutagenicity testing carry mutagenesis protein genes borne on plasmids, such as pKM101. Our objective was to introduce mutagenesis protein function into Escherichia coli strain DJ4309. This strain expresses recombinant human cytochrome P450 1A2 and NADPH-P450 reductase and carries out the metabolic conversion of aromatic and heterocyclic amines into DNA-reactive mutagens. We discovered that many mutagenesis-protein plasmids severely inhibit the response of strain DJ4309 to 2-amino-3,4-dimethylimid-azo[4,5-f]quinoline (MeIQ), a typical heterocyclic amine mutagen. Among many plasmids examined, one, pGY8294, a pSC101 derivative carrying the umuDC operon, did not inhibit MeIQ mutagenesis. Strain DJ4309 pGY8294 expresses active mutagenesis proteins, as shown by its response to mutagens such as 1-nitropyrene and 4-nitroquinoline 1-oxide (4-NQO), and is as sensitive as the parent strain DJ4309 to P450-dependent mutagens, such as MeIQ and 1-aminopyrene.     

Unique properties of purified, Escherichia coli-expressed constitutive cytochrome P4504A5.

Cytochromes P450 of the 4A family metabolize a variety of fatty acids, prostaglandins, and eicosanoids mainly at the terminal carbon (omega-hydroxylation) and, to a lesser extent, at the penultimate carbon [(omega-1)-hydroxylation]. In the present study, cytochrome P4504A5 (4A5) has been successfully expressed in Escherichia coli, with an average yield of enzyme of approximately 80 nmol/liter of cells. Spectroscopic characterization of the purified enzyme, using electron paramagnetic resonance and absolute and substrate-perturbed optical difference spectroscopy, showed that the heme of resting 4A5 is primarily low spin, but is converted primarily to high spin by substrate binding. The kcat and Km values for laurate omega-hydroxylation were 41 min-1 and 8.5 microM, respectively, in the absence of cytochrome b5, and 138 min-1 and 38 microM, respectively, in the presence of cytochrome b5. Hydroxylation of palmitate was dependent on the presence of cytochrome b5; kcat and Km values were 48 min-1 and 122 microM, respectively. Hydroxylation of arachidonic acid was barely detectable and was unchanged by the addition of cytochrome b5.     

Specificity determinants of rat tissue kallikrein probed by site-directed mutagenesis.

Site-specific mutagenesis was employed to study structure-function relationships at the substrate binding site of rat tissue kallikrein. Four kallikrein mutants, the Pro219 deletion (P219del), the 34-38 loop Tyr-Tyr-Phe-Gly to Ile-Asn mutation [YYFG(34-38)IN], the Trp215----Gly exchange (W215G) and the double mutant with Tyr99----His and Trp215----Gly exchange (Y99H:W215G) were created by site-directed mutagenesis to probe their function in substrate binding. The mutant proteins were expressed in Escherichia coli at high levels and analyzed by Western blot. These mutant enzymes were purified to apparent homogeneity. Each migrated as a single band on SDS-PAGE, with slightly lower molecular mass (36 kDa) than that of the native enzyme, (38 kDa) because of their lack of glycosylation. The recombinant kallikreins are immunologically identical to the native enzyme, displaying parallelism with the native enzyme in a direct radioimmunoassay for rat tissue kallikrein. Kinetic analyses of Km and kcat using fluorogenic peptide substrates support the hypothesis that the Tyr99-Trp215 interaction is a major determinant for hydrophobic P2 specificity. The results suggest an important role for the 34-38 loop in hydrophobic P3 affinity and further show that Pro219 is essential to substrate binding and efficient catalysis of tissue kallikrein.     

Metabolic conversion of IQ and MeIQ to bacterial mutagens

The metabolic conversion of 2-amino-3-methyl- and 2-amino-3,4-dimethyl-imidazo[4,5-f]quinoline (IQ and MeIQ respectively) to bacterial mutagens was studied using a bacterial mutation assay. Studies were performed using S9 fractions derived from either corn oil (uninduced) or Aroclor-1254-treated Sprague-Dawley rats. Aroclor 1254 treatment lowered the S9 protein concentration required for optimum levels of mutagenesis, enhanced the numbers of mutants observed and altered the effects of metabolic inhibitors and cofactors added to the assay. Studies with uninduced preparations revealed that IQ and MeIQ exhibited similar responses to the effects of metabolic inhibitors and cofactors involved in detoxication reactions. Both IQ and MeIQ activation appeared to be inhibited by the biogenic amines tryptamine and tyramine and inactivated by conjugation with either acetyl coenzyme A or glutathione.     

Kinetics and crystal structure of a mutant Escherichia coli alkaline phosphatase (Asp-369-->Asn): a mechanism involving one zinc per active site.

Using site-directed mutagenesis, an aspartate side chain involved in binding metal ions in the active site of Escherichia coli alkaline phosphatase (Asp-369) was replaced, alternately, by asparagine (D369N) and by alanine (D369A). The purified mutant enzymes showed reduced turnover rates (kcat) and increased Michaelis constants (Km). The kcat for the D369A enzyme was 5,000-fold lower than the value for the wild-type enzyme. The D369N enzyme required Zn2+ in millimolar concentrations to become fully active; even under these conditions the kcat measured for hydrolysis of p-nitrophenol phosphate was 2 orders of magnitude lower than for the wild-type enzyme. Thus the kcat/Km ratios showed that catalysis is 50 times less efficient when the carboxylate side chain of Asp-369 is replaced by the corresponding amide; and activity is reduced to near nonenzymic levels when the carboxylate is replaced by a methyl group. The crystal structure of D369N, solved to 2.5 A resolution with an R-factor of 0.189, showed vacancies at 2 of the 3 metal binding sites. On the basis of the kinetic results and the refined X-ray coordinates, a reaction mechanism is proposed for phosphate ester hydrolysis by the D369N enzyme involving only 1 metal with the possible assistance of a histidine side chain.     

The C1-C2 interface residue lysine 50 of pig kidney fructose-1, 6-bisphosphatase has a crucial role in the cooperative signal transmission of the AMP inhibition.

To understand the mechanism of signal propagation involved in the cooperative AMP inhibition of the homotetrameric enzyme pig-kidney fructose-1,6-bisphosphatase, Arg49 and Lys50 residues located at the C1-C2 interface of this enzyme were replaced using site-directed mutagenesis. The mutant enzymes Lys50Ala, Lys50Gln, Arg49Ala and Arg49Gln were expressed in Escherichia coli, purified to homogeneity and the initial rate kinetics were compared with the wild-type recombinant enzyme. The mutants exhibited kcat, Km and I50 values for fructose-2,6-bisphosphate that were similar to those of the wild-type enzyme. The kinetic mechanism of AMP inhibition with respect to Mg2+ was changed from competitive (wild-type) to noncompetitive in the mutant enzymes. The Lys50Ala and Lys50Gln mutants showed a biphasic behavior towards AMP, with total loss of cooperativity. In addition, in these mutants the mechanism of AMP inhibition with respect to fructose-1,6-bisphosphate changed from noncompetitive (wild-type) to uncompetitive. In contrast, AMP inhibition was strongly altered in Arg49Ala and Arg49Gln enzymes; the mutants had > 1000-fold lower AMP affinity relative to the wild-type enzyme and exhibited no AMP cooperativity. These studies strongly indicate that the C1-C2 interface is critical for propagation of the cooperative signal between the AMP sites on the different subunits and also in the mechanism of allosteric inhibition of the enzyme by AMP.     

Characterization of a cellobiose phosphorylase from a hyperthermophilic eubacterium,Thermotoga maritima MSB8

The cepA putative gene encoding a cellobiose phosphorylase of Thermotoga maritima MSB8 was cloned, expressed in Escherichia coli BL21-codonplus-RIL and characterized in detail. The maximal enzyme activity was observed at pH 6.2 and 80 degrees C. The energy of activation was 74 kJ/mol. The enzyme was stable for 30 min at 70 degrees C in the pH range of 6-8. The enzyme phosphorolyzed cellobiose in an random-ordered bi bi mechanism with the random binding of cellobiose and phosphate followed by the ordered release of D-glucose and alpha-D-glucose-1-phosphate. The Km for cellobiose and phosphate were 0.29 and 0.15 mM respectively, and the kcat was 5.4 s(-1). In the synthetic reaction, D-glucose, D-mannose, 2-deoxy-D-glucose, D-glucosamine, D-xylose, and 6-deoxy-D-glucose were found to act as glucosyl acceptors. Methyl-beta-D-glucoside also acted as a substrate for the enzyme and is reported here for the first time as a substrate for cellobiose phosphorylases. D-Xylose had the highest (40 s(-1)) kcat followed by 6-deoxy-D-glucose (17 s(-1)) and 2-deoxy-D-glucose (16 s(-1)). The natural substrate, D-glucose with the kcat of 8.0 s(-1) had the highest (1.1 x 10(4) M(-1) s(-1)) kcat/Km compared with other glucosyl acceptors. D-Glucose, a substrate of cellobiose phosphorylase, acted as a competitive inhibitor of the other substrate, alpha-D-glucose-1-phosphate, at higher concentrations.     

Alteration of aspartate 101 in the active site of Escherichia coli alkaline phosphatase enhances the catalytic activity

The function of aspartic acid residue 101 in the active site of Escherichia coli alkaline phosphatase was investigated by site-specific mutagenesis. A mutant version of alkaline phosphatase was constructed with alanine in place of aspartic acid at position 101. When kinetic measurements are carried out in the presence of a phosphate acceptor, 1.0 M Tris, pH 8.0, both the kcat and the Km for the mutant enzyme increase by approximately 2-fold, resulting in almost no change in the kcat/Km ratio. Under conditions of no external phosphate acceptor and pH 8.0, both the kcat and the Km for the mutant enzyme decrease by approximately 2-fold, again resulting in almost no change in the kcat/Km ratio. The kcat for the hydrolysis of 4-methyl-umbelliferyl phosphate and p-nitrophenyl phosphate are nearly identical for both the wild-type and mutant enzymes, as is the Ki for inorganic phosphate. The replacement of aspartic acid 101 by alanine does have a significant effect on the activity of the enzyme as a function of pH, especially in the presence of a phosphate acceptor. At pH 9.4 the mutant enzyme exhibits 3-fold higher activity than the wild-type. The mutant enzyme also exhibits a substantial decrease in thermal stability: it is half inactivated by treatment at 49 degrees C for 15 min compared to 71 degrees C for the wild-type enzyme. The data reported here suggest that this amino acid substitution alters the rates of steps after the formation of the phospho-enzyme intermediate.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization of mitochondrial thioredoxin reductase from C. elegans

Thioredoxin reductase catalyzes the NADPH-dependent reduction of the catalytic disulfide bond of thioredoxin. In mammals and other higher eukaryotes, thioredoxin reductases contain the rare amino acid selenocysteine at the active site. The mitochondrial enzyme from Caenorhabditis elegans, however, contains a cysteine residue in place of selenocysteine. The mitochondrial C. elegans thioredoxin reductase was cloned from an expressed sequence tag and then produced in Escherichia coli as an intein-fusion protein. The purified recombinant enzyme has a kcat of 610 min(-1) and a Km of 610 microM using E. coli thioredoxin as substrate. The reported kcat is 25% of the kcat of the mammalian enzyme and is 43-fold higher than a cysteine mutant of mammalian thioredoxin reductase. The enzyme would reduce selenocysteine, but not hydrogen peroxide or insulin. The flanking glycine residues of the GCCG motif were mutated to serine. The mutants improved substrate binding, but decreased the catalytic rate.     

Identification of a glycine-rich sequence as an NAD(P)H-binding site and tyrosine 128 as a dicumarol-binding site in rat liver NAD(P)H:quinone oxidoreductase by site-directed mutagenesis.

Site-directed mutagenesis was utilized to identify binding sites for NAD(P)H and dicumarol in rat liver NAD(P)H:quinone oxidoreductase (NQOR, EC 1.6.99.2). The mutant cDNA clones were generated by a procedure based on the polymerase chain reaction and were expressed in Escherichia coli. The mutant enzymes were purified to apparent homogeneity as judged by SDS-polyacrylamide gel electrophoresis and were found to contain 2 FADs/enzyme molecule identical with that of the wild-type NQOR. Purified mutant enzymes Y128D, G150F, G150V, S151F, and Y155D showed dramatic decreases in activities in the reduction of dichlorophenolindophenol in comparison with the activities of the wild-type enzyme, whereas the activities of F124L, T127V, T127E, Y128V, Y128F, S151A, and Y155V were similar to those of NQOR. Enzyme kinetic analysis revealed that the Km values of T127E, Y128D, G150F, G150V, S151F, and Y155D were, respectively, 4-, 2-, 13-, 5-, 26-, and 19-fold higher than the Km of NQOR for NADPH, and were, respectively, 2-, 3-, 7-, 3-, 20-, and 11-fold higher than that of NQOR for NADH. The kcat values of Y128D, G150F, and G150V were also much lower than those of NQOR, but the kcat values of other mutants were similar to those of the wild-type enzyme. The Km values of the mutants for dichlorophenolindophenol were the same or slightly higher than that of NQOR. The apparent inhibition constants (Ki) for dicumarol on Y128V and F124L were elevated 12 and 8 times, respectively. Similar, but smaller, changes on Ki for 4-hydroxycoumarin were also observed. This study demonstrated that residues Gly150, Ser151, and Tyr155 in the glycine-rich region of NQOR are essential for NADPH and NADH binding and Tyr128 is important for dicumarol binding. Based on the results of the study, it is proposed that the glycine-rich region of the enzyme, along with other residues around the region, forms a beta sheet-turn-alpha helix structure important for the binding of the pyrophosphate group of NADPH and NADH.     

Alteration of catalytic properties of chymosin by site-directed mutagenesis

Artificial mutations of chymosin by recombinant DNA techniques were generated to analyze the structure--function relationship in this characteristic aspartic proteinase. In order to prepare the mutant enzymes in their active form, we established procedures for purification of correctly refolded prochymosin from inclusion bodies produced in Escherichia coli transformants and for its subsequent activation. Mutagenesis by linker insertion into cDNA produced several mutants with an altered ratio of milk clotting activity to proteolytic activity and a different extent of stability. In addition to these mutants, several mutants with a single amino acid exchange were also constructed by site-directed mutagenesis and kinetic parameters of these mutant enzymes were determined by using synthetic hexa- and octa-peptides as substrates. Exchange of Tyr75 on the flap of the enzyme to Phe caused a marked change of substrate specificity due to the change of kcat or Km, depending on the substrate used. Exchange of Val110 and Phe111 also caused a change of kinetic parameters, which indicates functional involvement of these hydrophobic residues in both the catalytic function and substrate binding. The mutant Lys220----Leu showed a marked shift of the optimum pH to the acidic side for hydrolysis of acid-denatured haemoglobin along with a distinct increase in kcat for the octa-peptide in a wide pH range.     

Cytochrome P450c17-expressing Escherichia coli as a first-step screening system for 17alpha-hydroxylase-C17,20-lyase inhibitors

We have designed and synthesized a number of cytochrome P450 17alpha-hydroxylase-C17,20-lyase (P450c17) inhibitors with the aim of inhibiting androgen synthesis. To select the most potent inhibitors, we initially used human testicular microsomes, which have a high level of expression of this enzyme. However, due to lack of availability of human tissue and variability among the samples, we utilized recombinant human enzyme expressed in Escherichia coli. We designed a simple and economical protocol based on the report that recombinant bovine P450c17 can be functionally active in live bacteria. In the assay we report here, we substituted high-performance liquid chromatography product isolation with a rapid biochemical acetic acid releasing assay and utilized intact P450c17-expressing E. coli for the source of the enzyme. Enzymatic parameters of the bacterial system (Km = 5.1 x 10(-7) M, Vmax = 15.0 pmol/min/mg) were similar to those of human testicular microsomes (Km = 4.8 x 10(-7) M, Vmax = 40.0 pmol/min/mg), and our compounds displayed a similar pattern of inhibition in both systems. This new system is a fast, reliable, and reproducible method for screening P450c17 inhibitors. Furthermore, it eliminates our dependence on human tissue and potential data fluctuations caused by variations in enzymatic activity between donors.