Cytochrome b(5) coexpression increases the CYP2E1-dependent mutagenicity of dialkylnitrosamines in methyltransferase-deficient strains of Salmonella typhimurium.

Addition of cytochrome b(5) to recombinant cytochrome P450 2E1 systems has been shown to enhance the metabolism of dialkylnitrosamines in vitro. To determine if this effect could be observed with recombinant expression systems in vivo, we have constructed mutagenicity tester strains that coexpress full-length human cytochrome P450 2E1 (CYP2E1), rat cytochrome P450 reductase, and human cytochrome b(5) in Salmonella typhimurium lacking ogt and ada methyltransferases (YG7104, ogt(-); and YG7108, ogt(-), ada(-)). These new recombinant strains exhibit a four- to five-fold greater mutagenic response to dimethylnitrosamine, diethylnitrosamine, and dipropylnitrosamine than strains that contain only CYP2E1 and reductase, and are over 100-fold more sensitive to nitrosamines than the parental strains in the presence of an exogenous activating system (S9 fraction). The four-fold increase in mutagenicity in the presence of cytochrome b(5) was consistent with increasing alkyl chain length up to dibutylnitrosamine, which was poorly activated by CYP2E1. The greatest enhancement was obtained with a tricistronic construct in which the b(5) cDNA preceded the P450 and reductase cDNAs; placing the b(5) cDNA after the reductase cDNA was substantially less effective. These new, highly sensitive strains may prove useful in the detection of nitrosamine contamination of food and environmental samples.

Garlic and garlic-derived compounds inhibit human squalene monooxygenase.

Although extracts of garlic inhibit cholesterol biosynthesis in cultured hepatocytes, the inhibitory components of garlic and the site or sites of inhibition in the cholesterol biosynthetic pathway have not been established. To elucidate potential mechanisms of inhibition, we examined the effect of fresh garlic extract and 16 water- or lipid-soluble compounds derived from garlic on purified recombinant human squalene monooxygenase. Squalene monooxygenase catalyzes the second and likely rate-limiting step in the downstream pathway for cholesterol biosynthesis. A 50% inhibitory concentration (IC(50)) of squalene epoxidation was achieved with 1 g/L of fresh garlic extract; of the 16 garlic compounds tested, only selenocystine (IC(50) = 65 micromol/L), S-allylcysteine (IC(50) = 110 micromol/L), alliin (IC(50) = 120 micromol/L), diallyl trisulfide (IC(50) = 195 micromol/L), and diallyl disulfide (IC(50) = 400 micromol/L) substantially inhibited the enzyme. Kinetic analysis showed that the inhibition by garlic and by these compounds was slow and irreversible, suggestive of covalent binding to the enzyme; the ability of thiol-containing compounds such as glutathione and 2,3-dimercaptopropanol to prevent and reverse the inhibition indicated that the garlic compounds were reacting with sulfhydryl groups on the protein. Dithiols were better reversal agents than monothiols, further suggesting that these inhibitors bind to the proposed vicinal sulfhydryls present on this enzyme. These results indicate that squalene monooxygenase may be one of the target enzymes through which garlic inhibits cholesterol biosynthesis.

Inhibition of human squalene monooxygenase by tellurium compounds: evidence of interaction with vicinal sulfhydryls.

Squalene monooxygenase is a flavin adenine dinucleotide-containing, microsomal enzyme that catalyzes the second step in the committed pathway for cholesterol biosynthesis. Feeding weanling rats a diet containing 1% elemental tellurium causes a transient, peripheral demyelination due to the disruption of cholesterol synthesis in Schwann cells secondary to inhibition of squalene monooxygenase. The tellurium species responsible for the inhibition is unknown, as is the mechanism of inhibition. To study the potential mechanisms of tellurium toxicity in humans, three likely in vivo metabolites of tellurium (tellurite, dimethyltellurium dichloride, and dimethyltelluride) were tested as inhibitors of purified human squalene monooxygenase. All three inhibitors reacted with the enzyme slowly and the resulting interaction was not freely reversible. The 50% inhibitory concentration for the methyltellurium compounds (approximately 100 nM) after a 30-min preincubation was 100-fold lower than that of tellurite, indicating a role for hydrophobicity in the enzyme-inhibitor interaction. The ability of glutathione and 2,3-dimercaptopropanol to prevent and reverse the inhibition indicated that the tellurium compounds were reacting with sulfhydryls on squalene monooxygenase, and the ability of phenylarsine oxide, which reacts specifically with vicinal sulfhydryls, to inhibit the enzyme indicated that these sulfhydryls are located proximal to one another on the enzyme. These results suggest that the unusual sensitivity of squalene monooxygenase to tellurium compounds is due to the binding of these compounds to vicinal cysteines, and that methylation of tellurium in vivo may enhance the toxicity of tellurium for this enzyme.

Mutagenicity of nitrosamines in methyltransferase-deficient strains of Salmonella typhimurium coexpressing human cytochrome P450 2E1 and reductase.

Although dialkylnitrosamines are environmentally significant carcinogens, the use of short-term bioassays to assess the mutagenic potential of these compounds is problematic. The Ames test, a mutagenicity assay based on the reversion of Salmonella typhimurium histidine auxotrophs, is the most widely used bioassay in genetic toxicology, but the traditional Ames tester strains are largely insensitive to dialkylnitrosamine mutagenicity. We have constructed two mutagenicity tester strains that co-express full-length human cytochrome P450 2E1 and P450 reductase in S. typhimurium lacking ogt and ada methyltransferases (YG7104ER, ogt- and YG7108ER, ogt-, ada-). These new strains are susceptible to dialkylnitrosamine mutagenicity in the absence of an exogenous metabolic activating system (S9 fraction). Mutagenicity is dependent upon the coexpression of P450 2E1 with P450 reductase and is similar to or greater than that obtained with the parental strains in the presence of S9 fraction from ethanol-induced rat liver. These strains were also sensitive to nitrosamines with longer alkyl side chains including diethylnitrosamine, dipropylnitrosamine and dibutylnitrosamine. Mutagenicity decreased with alkyl chain length, consistent with the stringency of the ada-encoded enzyme for methyl and ethyl DNA adducts. These new strains may prove useful in the evaluation of nitrosamine contamination of food and environmental samples.

Cloning, heterologous expression, and enzymological characterization of human squalene monooxygenase.

The cDNA for human squalene monooxygenase, a key enzyme in the committed pathway for cholesterol biosynthesis, was amplified from a human liver cDNA library and cloned, and the protein was expressed in Escherichia coli and purified. Kinetic analysis of the purified enzyme revealed an apparent K(m) for squalene of 7.7 microM and an apparent k(cat) of 1.1 min(-1). For FAD the apparent K(m) is 0.3 microM, consistent with a loosely bound flavin. The apparent K(m) for NADPH-cytochrome P450 reductase, the requisite electron transfer partner, is 14 nM. The amount of reductase needed for maximal activity is about threefold less than the amount of squalene monooxygenase present in the assay; thus, electron transfer to the monooxygenase is not likely to be rate limiting. Previous reports have implicated inhibition of this enzyme as the cause of a peripheral demyelination seen in weanling rats fed a diet containing tellurium. As no data were available for humans, the ability of a number of tellurium and related elemental compounds to inhibit the recombinant human enzyme was examined. Tellurite, tellurium dioxide, selenite, and selenium dioxide were inhibitory; the tellurium compounds were more potent than the selenium compounds, as indicated by their IC(50) values (17 and 37 microM, respectively). Kinetic analysis of the inhibition by tellurite suggests multiple sites of interaction with the enzyme in a noncompetitive manner with respect to squalene.

Coexpression of mammalian cytochrome P450 and reductase in Escherichia coli.

cDNAs for human cytochrome P450 2E1 and rat NADPH-cytochrome-P450 reductase were cloned separately and in tandem into bacterial expression vectors, and expression of the two proteins in Escherichia coli was monitored by immunoblotting, spectroscopy, and catalytic assays. The cDNAs were separated on the coexpression plasmid by 22 nucleotides, with the P450 cDNA preceding the reductase cDNA. P450 content in solubilized cell membranes, whether expressed alone or coexpressed with P450 reductase, was approximately 0.11 nmol/mg of protein, and approximately 0.8 nmol could be obtained per liter of culture. Reductase content was five- to sixfold greater than P450 content when coexpressed, but severalfold less than that obtained when expressed without the upstream P450 cDNA, indicating differences in both stability and translatability between the two proteins. Solubilized membranes from cells expressing both proteins catalyzed aniline hydroxylation, p-nitrophenol hydroxylation, and N-nitrosodimethylamine demethylation at rates equivalent to those obtained by combining P450 and reductase preparations; addition of purified reductase to these membranes did not augment the activity. However, in contrast to results obtained with P450 2E1 expressed in other heterologous systems, addition of rabbit liver cytochrome b5 to preparations catalyzing p-nitrophenol or N-nitrosodimethylamine oxidation did not increase turnover, and, although activity could be shown with unsolubilized membranes, oxidation of these substrates in vivo could not be demonstrated. Nonetheless, the ability to coexpress P450 and reductase in E. coli so as to generate a functional monooxygenase system in vitro enhances the utility of this organism for the expression and characterization of cloned P450 isoforms.

Correlation between codon usage, regional genomic nucleotide composition, and amino acid composition in the cytochrome P-450 gene superfamily.

The codon usage bias of 110 mammalian cytochrome P-450 genes has been determined and analyzed in relation to a variety of genetic, biochemical, and physiological parameters. In those P-450 genes exhibiting biased usage the preferred codons generally do not differ among the four species examined (rat, rabbit, man, and mouse) or from the predominantly used codons identified for all sequenced genes in a recent data base analysis (Wada et al. (1992) Nucleic Acids Res. 20 (Suppl.), 2111-2118). Codon usage bias does not correlate with evolutionary relationships, evolutionary age, or with the extent of evolutionary conservation of orthologous proteins; there is no obvious correlation with the level of expression of a given P-450, with its inducibility, nor with its physiologic role; and neither the preferred codons nor the degree of bias differ for P-450s expressed in different tissues. Codon usage bias does correlate with the C+G content at the codon third position, and thus preferred codons usually end in C or G; for those P-450s for which gene sequences are available this bias also correlates with the C + G content of the intronic and flanking regions of these genes. Moreover, a lesser increase in the C + G content at the codon first and second positions is also evident in genes located in regions of high C + G content; this leads to predictable differences in the amino acid compositions of P-450 enzymes that correlate with genomic nucleotide composition and the degree of bias in codon usage.

Mutagenesis at a highly conserved phenylalanine in cytochrome P450 2E1 affects heme incorporation and catalytic activity.

The phenylalanine corresponding to Phe-429 of rabbit cytochrome P450 2E1 is 1 of approximately 10 highly conserved residues in this superfamily of over 200 sequenced enzymes. This nearly invariant residue has been postulated to be involved in electron transfer between the heme of cytochrome P450cam and its redox partners [Stayton, P.S., Poulos, T. L., & Sligar, S.G. (1989) Biochemistry 28, 8201-8205]. To test this hypothesis, oligonucleotide-directed mutagenesis was used to replace this amino acid in rabbit P450 2E1 with aspartate, arginine, leucine, tryptophan, or tyrosine, and the mutant proteins were expressed in Escherichia coli. Although immunoblot analysis of whole cell lysates demonstrated that all P450 proteins (mutants and wild-type) were equally well expressed on a per cell basis, in solubilized membranes only the tryptophan and tyrosine mutants yielded ferrous-CO difference spectra characteristic of P450. The specific content (nanomoles per milligram of membrane protein) and yield per liter of the Trp mutant holoenzyme were approximately one-third those of the native enzyme, suggesting that heme incorporation was hindered by tryptophan at this position, whereas the specific content and yield per liter of the Tyr mutant were significantly greater than those of the native preparation. The stability of the Trp and Tyr mutants, as judged by thermal denaturation studies, was not different from that of the native enzyme. The Trp mutant had 38% of the aniline hydroxylase activity, 25% of the p-nitrophenol hydroxylase activity, and 39% of the N-nitrosodimethylamine demethylase activity of the native enzyme, demonstrating that this substitution also decreased catalytic activity.(ABSTRACT TRUNCATED AT 250 WORDS)

Purification and properties of a shortened form of cytochrome P-450 2E1: deletion of the NH2-terminal membrane-insertion signal peptide does not alter the catalytic activities.

As reported previously, alcohol-inducible cytochrome P-450 2E1 lacking the hydrophobic NH2-terminal segment is located primarily in the inner cell membrane when expressed in Escherichia coli and is active with a typical substrate. To study the catalytic properties in detail, we have purified the truncated P-450 lacking residues 3-29 to electrophoretic homogeneity from the solubilized bacterial membrane fraction in the presence of 4-methylpyrazole as a stabilizing agent. The resulting heme protein with a specific content of 15.8 nmol of P-450 per mg of protein has a reduced CO difference spectrum identical to that of the full-length enzyme, with a Soret maximum at 452 nm. The rates of catalysis of four reactions in the reconstituted enzyme system, including the oxygenation of ethanol to give acetaldehyde, the oxidative dealkylation of N-nitrosodiethylamine to give ethylene and acetaldehyde, and the ring hydroxylation of aniline and p-nitrophenol, are the same with the shortened and full-length enzymes. The apparent Km of p-nitrophenol is also the same, as is that for NADPH-cytochrome P-450 reductase and for cytochrome b5, which stimulates p-nitrocatechol formation about 3-fold. Moreover, the requirement for phosphatidylcholine for full catalytic activity is unchanged despite the absence of the NH2-terminal segment. Although this highly hydrophobic segment is believed to play a role in the intact cell as a membrane-insertion signal sequence, we conclude that it has no function in the catalytic activity of the cytochrome as an oxygenase, including interactions with the other components of the enzyme system.

Differences in the developmental expression of rabbit cytochromes P-450 2E1 and 2E2.

The alcohol-inducible CYP2E subfamily in rabbits contains two genes; CYP2E1 encodes the cytochrome earlier termed P-450 3a, and CYP2E2 encodes a cytochrome that is 97% identical in amino acid sequence to cytochrome P-450 (P-450) 2E1. In the present studies, the ontogenic expression of these two cytochromes was examined. In liver, P-450 2E2 mRNA is detectable immediately after birth and reaches slightly greater than the adult level at 2 weeks of age; in contrast, P-450 2E1 mRNA is not detectable until day 14 and increases rapidly to approximately twice the adult level at 5 weeks of age. P-450 2E protein is present in liver immediately after birth, coincident with the appearance of P-450 2E2 mRNA, peaks at 2 weeks, and then, despite the continued elevation in P-450 2E mRNA, decreases to the adult level at 5 weeks. In kidney, P-450 2E2 mRNA is not detectable at any age; P-450 2E1 mRNA, however, is present at 1 week, and the level increases to about half the adult level at 5 weeks of age. P-450 2E protein in this tissue is elevated at 2 weeks, relative to mRNA levels, and reaches approximately half the adult level at 5 weeks. The lack of close correlation between mRNA and protein levels in the liver and kidney of newborn rabbits indicates that the posttranscriptional control of P-450 2E enzyme levels that predominates in adult animals is also operative during the neonatal period. Monooxygenase activities with ethanol and p-nitrophenol as substrates reflect the developmental increase in P-450 2E protein, as well as the appearance and levels of spectrally detectable P-450, cytochrome b5, and NADPH-P-450 reductase in hepatic microsomes. The expression of P-450 2E2, but not P-450 2E1, in early neonates suggests that these two closely related cytochromes may have functional differences that are important during the first few weeks of life.

Alcohol-inducible cytochrome P-450IIE1 lacking the hydrophobic NH2-terminal segment retains catalytic activity and is membrane-bound when expressed in Escherichia coli.

We have expressed in Escherichia coli a cDNA encoding rabbit liver cytochrome P-450IIE1, the ethanol-inducible P-450. The expressed P-450 is located primarily in the bacterial inner cell membrane and comprises 3% of the E. coli total membrane protein. The partially purified cytochrome exhibits a reduced CO difference spectrum with a maximum at 452 nm, characteristic of P-450IIE1, and solubilized membranes or partially purified P-450 preparations reconstituted with NADPH-cytochrome P-450 reductase and phosphatidylcholine catalyze the deethylation of N-nitrosodiethylamine with a turnover number equal to that of purified liver P-450IIE1 (approximately 4.5 nmol/min/nmol of P-450). A modified IIE1 cDNA that encodes a protein lacking amino acids 3-29, a proposed membrane anchor for cytochrome P-450, was also expressed in E. coli and, unexpectedly, the shortened protein was also found to be predominantly located in the bacterial inner membrane rather than the cytosol. Like the full-length protein, this truncated cytochrome has a reduced CO difference spectrum characteristic of P-450IIE1 and is fully active in the deethylation of N-nitrosodiethylamine. These results demonstrate that the NH2-terminal hydrophobic segment is not solely responsible for attachment to the membrane and evidently is not required for proper protein folding or catalytic activity.

An unusual yet strongly conserved flavoprotein reductase in bacteria and mammals.

The recent determination of the amino acid sequences of the Bacillus megaterium cytochrome P-450 and the flavoprotein component of Salmonella typhimurium NADPH-sulfite reductase revealed that these enzymes contain a flavoprotein moiety remarkably similar to mammalian NADPH-cytochrome P-450 reductase. The presence of this oxidoreductase in these very different enzymes suggests that this flavoprotein arose early in evolution and was utilized as an enzymological building block. The multi-domain structure of the reductase further suggests that it arose through a fusion of genes encoding simple flavin electron-transport proteins.

cDNA and derived amino acid sequence of rabbit nasal cytochrome P450NMb (P450IIG1), a unique isozyme possibly involved in olfaction.

Olfactory-specific cytochrome P450NMb was previously purified to electrophoretic homogeneity from microsomes of rabbit nasal mucosa in this laboratory. In the present study, a cDNA library made from poly(A)+ RNA from rabbit nasal mucosa was screened with antibodies to this P450, and eight immunopositive clones were isolated and characterized. The sequence determined from two overlapping clones contained an open reading frame of 1446 nucleotides, with the predicted first 39 amino acids corresponding to residues 12 to 50 of purified NMb, except for position 46, where Leu was encoded instead of the Glu residue that was found earlier by Edman degradation analysis. The complete polypeptide, including residues 1 to 11, contains 494 amino acid residues and has a molecular weight of 56,640. Sequence comparisons indicated that NMb is more than 50% identical to members of the rabbit P450 gene II family, including IIB4, IIC3, IIC5, IIE1, and IIE2, and 83% identical to rat P450olf1 (IIG1). Hybridization of NMb to electrophoretically fractionated rabbit nasal poly(A)+ RNA revealed 3.6- and 2.1-kb species, but with a probe derived from the 3'-nontranslated portion of the cDNA only the 3.6-kb band was observed, suggesting the use of alternate polyadenylation sites or splicing. In agreement with the known tissue-specific distribution of NMb protein, NMb transcripts were found in olfactory mucosa, but not in liver, lung, intestine, or kidney. Genomic hybridization analysis indicated that there may be only one copy of the NMb gene present in the rabbit genome.

Expression of rabbit cytochrome P-450IIE2 in yeast and stabilization of the enzyme by 4-methylpyrazole.

A rabbit cytochrome P-450IIE2 full-length cDNA was cloned into a yeast episomal plasmid (YEp13) between the copper-responsive yeast metallothionein gene promoter (CUP1) and the iso-1-cytochrome c gene terminator (CYC1), and the cytochrome P-450 was expressed in Saccharomyces cerevisiae. The microsomal fraction prepared from copper-treated cells exhibited a ferrous carbonyl difference spectrum with an absorption maximum at 451 nm and contained approximately 0.07 nmol of P-450IIE2 per mg of protein. The P-450IIE2 protein expressed in yeast microsomes was catalytically competent as judged by the NADPH-dependent deethylation of N-nitrosodiethylamine and by the oxidation of butanol. Cholate solubilization and polyethylene glycol fractionation of yeast microsomal P-450IIE2 yielded a preparation with a markedly lower specific content than that of intact microsomes, but, when 4-methylpyrazole was included during solubilization, the holoenzyme was completely stabilized.

NADPH-cytochrome P-450 oxidoreductase gene organization correlates with structural domains of the protein.

cDNA clones to rat liver NADPH-cytochrome P-450 oxidoreductase were used to isolate genomic clones from a Wistar-Furth inbred rat genomic DNA library. Fifteen exons containing the coding region and 3'-nontranslated segment of the P-450 reductase gene were identified, spanning 20 kilobases of DNA contained in 3 lambda-Charon 35 clones. The organization of this single copy gene reveals a general correspondence between exons and structural domains of the protein, with the segment responsible for anchoring the reductase to the microsomal membrane and several segments involved in FMN, FAD, and NADPH binding encoded by discrete exons.

Induction and tissue-specific expression of rabbit cytochrome P450IIE1 and IIE2 genes.

Treatment of rabbits with a variety of dissimilar chemicals, including ethanol, acetone, and imidazole, results in elevated levels of hepatic and renal cytochrome P-450 form 3a, also designated P-450ALC or P-450IIE. The P450IIE1 subfamily in rabbits is composed of two genes that encode proteins with 97% sequence identity; the mRNAs from these genes can be distinguished by their differing electrophoretic mobilities. In the present studies, examination of the expression of these genes revealed that P450IIE1 (gene 1) mRNA is present in greatest abundance in the liver, is present in kidney and nasal mucosa at approximately 10% of the level in liver, and is present in lung at approximately 5% of the level in liver. P450IIE2 (gene 2) mRNA is present in liver and lung at approximately 50% of the level of gene 1 mRNA in these tissues but cannot be detected in kidney or nasal mucosa. Neither gene is expressed in testis, ovary, small intestine, or adrenal tissue. Treatment of rabbits with acetone or imidazole results in elevated levels of P-450 3a-immunoreactive protein in liver and kidney without concomitant increases in P450IIE gene mRNAs. Moreover, various lengths of ethanol treatment elevated the level of immunoreactive protein in liver and kidney, with a rapid reduction of gene 1 mRNA and, at 14 days, gene 2 mRNA to approximately 50% of control levels. In contrast to these chemical inducers of 3a, fasting for 48 hr significantly increases gene 1 and 2 mRNA in liver but does not increase the level of immunoreactive protein. These results indicate that the rabbit P450IIE genes are not coordinately expressed or regulated and, as found with the rat ortholog P-450j, chemical inducers of 3a evidently act through changes in the rate of synthesis or degradation of the enzyme, rather than through increased gene transcription.

Structural analysis of the FMN binding domain of NADPH-cytochrome P-450 oxidoreductase by site-directed mutagenesis.

Comparison of the amino acid sequence of rat liver NADPH-cytochrome P-450 oxidoreductase with that of flavoproteins of known three-dimensional structure suggested that residues Tyr-140 and Tyr-178 are involved in binding of FMN to the protein. To test this hypothesis, NADPH-cytochrome P-450 oxidoreductase was expressed in Escherichia coli using the expression-secretion vector pIN-III-ompA3, and site-directed mutagenesis was employed to selectively alter these residues and demonstrate that they are major determinants of the FMN-binding site. Bacterial expression produced a membrane-bound 80-kDa protein containing 1 mol each of FMN and FAD per mol of enzyme, which reduced cytochrome c at a rate of 51.5 mumol/min/mg of protein and had absorption spectra and kinetic properties very similar to those of the rat liver enzyme. Replacement of Tyr-178 with aspartate abolished FMN binding and cytochrome c reductase activity. Incubation with FMN increased catalytic activity to a maximum of 8.6 mumol/min/mg of protein. Replacement of Tyr-140 with aspartate did not eliminate FMN binding, but reduced cytochrome c reductase activity about 5-fold, suggesting that FMN may be bound in a conformation which does not permit efficient electron transfer. Substitution of phenylalanine at either position 140 or 178 had no effect on FMN content or catalytic activity. The FAD level in the Asp-178 mutant was also decreased, suggesting that FAD binding is dependent upon FMN; FAD incorporation may occur co-translationally and require prior formation of an intact FMN domain.