Protein engineering of cytochromes P-450

The cytochromes P-450 are an immensely important superfamily of heme-containing enzymes. They catalyze the monooxygenation of an enormous range of substrates. In bacteria, cytochromes P-450 are known to catalyze the hydroxylation of environmentally significant substrates such as camphor, phenolic compounds and many herbicides. In eukaryotes, these enzymes perform key roles in the synthesis and interconversion of steroids, while in mammals hepatic cytochromes P-450 are vital for the detoxification of many drugs. As such, the cytochromes P-450 are of considerable interest in medicine and biotechnology and are obvious targets for protein engineering. The purpose of this article is to illustrate the ways in which protein engineering has been used to investigate and modify the properties of cytochromes P-450. Illustrative examples include: the manipulation of substrate selectivity and regiospecificity, the alteration of membrane binding properties, and probing the route of electron transfer.     

Mouse liver cytochrome P-450 (P-450IIIAM1): its cDNA cloning and inducibility by dexamethasone.

A full-length cDNA complementary to mouse liver mRNA coding for one of the cytochromes P-450 (P-450) in the P-450IIIA family, namely P-450IIIM1, was isolated and completely sequenced. The sequence of this cDNA clone, pMDex13, revealed that it encoded a polypeptide of 504 deduced amino acid residues (Mr = 57,853). The deduced amino acid sequence showed 87.3 and 84.9% identity with rat P-450IIIA1 and P-450IIIA2, respectively. The NH2-terminal 24 amino acid sequences of P-450IIIAM1 were completely identical with purified mouse P-450UT protein. RNA blot analysis showed that mRNA content of hepatic P-450IIIAM1 was remarkably increased by treatment of mice with dexamethasone.     

Presence of cytochrome P-450- and cytochrome P-448-dependent monooxygenase functions in hepatoma cell lines

Cell lines derived from Reuber H-4-II-E hepatoma cells and their hybrids that differ in the expression of liver-specific functions are shown to contain different forms of monooxygenases. According to 1) the specificity toward the substrates benzo(a)pyrene, aldrin and chenodexycholic acid, 2) the kinetics of the epoxidation of aldrin, 3) the response to inducers, such as benz(a)anthracene and dexamethasone, and 4) the $\text{in}$$\text{vitro}$ modifier 7,8-benzoflavone, the monooxygenases predominating in differentiated cell lines belong to the cytochrome P-450-dependent enzyme(s), those in the less differentiated lines belong to the cytochrome P-448-dependent form(s).     

Oxidation of uroporphyrinogen by methylcholanthrene-induced cytochrome P-450. Essential role of cytochrome P-450d.

We have previously shown that uroporphyrinogen is oxidized to uroporphyrin by microsomes (microsomal fractions) from 3-methylcholanthrene-pretreated chick embryo liver [Sinclair, Lambrecht & Sinclair (1987) Biochem. Biophys. Res. Commun. 146, 1324-1329]. We report here that a specific antibody to chick liver methylcholanthrene-induced cytochrome P-450 (P-450) inhibited both uroporphyrinogen oxidation and ethoxyresorufin O-de-ethylation in chick-embryo liver microsomes. 3-Methylcholanthrene-pretreatment of rats and mice markedly increased uroporphyrinogen oxidation in hepatic microsomes as well as P-450-mediated ethoxyresorufin de-ethylation. In rodent microsomes, uroporphyrinogen oxidation required the addition of NADPH, whereas chick liver microsomes required both NADPH and 3,3',4,4'-tetrachlorobiphenyl. Treatment of rats with methylcholanthrene, hexachlorobenzene and o-aminoazotoluene increased uroporphyrinogen oxidation and P-450d, whereas phenobarbital did not increase either. The contribution of hepatic P-450c and P-450d to uroporphyrinogen oxidation and ethoxyresorufin O-de-ethylation in methylcholanthrene-induced microsomes was assessed by using specific antibodies to P-450c and P-450d. Uroporphyrinogen oxidation by methylcholanthrene-induced rat liver microsomes was inhibited up to 75% by specific antibodies to P-450d, but not by specific antibodies to P-450c. In contrast, ethoxyresorufin de-ethylation was inhibited only 20% by anti-P450d but 70% by anti-P450c. Methylcholanthrene-induced kidney microsomes which contain P-450c but non P-450d did not oxidize uroporphyrinogen. These data indicate that hepatic P-450d catalyses uroporphyrinogen oxidation. We suggest that the P-450d-catalysed oxidation of uroporphyrinogen has a role in the uroporphyria caused by hexachlorobenzene and other compounds.     

The cDNA and protein sequence of a phenobarbital-induced chicken cytochrome P-450

Several cDNA clones complementary to a chicken phenobarbital-inducible cytochrome P-450 have been isolated and sequenced, representing the first non-mammalian eukaryotic cytochrome P-450 sequence to be analyzed. The cDNA clones hybridized to two mRNAs of 3.5 and 2.5 kilobases in length, but further analysis indicated that the clones were derived from the larger mRNA. The sequence contains a 5'-noncoding region of 39 nucleotides and an open reading frame of 1473 nucleotides. The remainder of the sequence is due to the 3'-noncoding region and poly(A) tail. The open reading frame encodes a protein of 491 amino acids with a molecular weight of 56,196. The chicken cytochrome P-450 shows an overall homology of 45-54% compared with the mammalian phenobarbital-induced cytochrome P-450s. The degree of homology is not uniform, with some short regions showing much greater levels of sequence conservation. In particular, the chicken cytochrome P-450 contains the conserved cysteinyl domain near the carboxyl terminus, found in all cytochrome P-450s and which is thought to be involved in heme binding. Using the chicken sequence, a more accurate estimate of the evolutionary rates of cytochrome P-450s has been made. It is suggested that the phenobarbital-, 3-methylcholanthrene, and pregnenolone 16 alpha-carbonitrile-induced cytochrome P-450 gene families diverged from a common ancestral gene 600 million years ago. Furthermore the phenobarbital-inducible gene apparently underwent gene duplication events at about the time of the divergence of the chicken and mammalian lineages. The results imply that most mammals should have at least four rather distantly related phenobarbital-inducible gene subfamilies.     

Hamster cytochrome P-450 IA gene family, P-450IA1 and P-450IA2 in lung and liver: cDNA cloning and sequence analysis.

Two cDNA clones, 2C19 and 4C1, were isolated from a lung cDNA library of 3-methylcholanthrene (MC)-treated hamster by using rat P-450c cDNA as a probe. The cDNA determined from 2C19 and 4C1 was 2,916 bp long and contained an entire coding region for 524 amino acids with a molecular weight of 59,408. The deduced amino acid sequence showed a 85% identity with that of rat P-450c indicating 2C19 and 4C1 encode the hamster P-450IA1 protein. Another cDNA clone, designated H28, was isolated from a MC-induced hamster liver cDNA library by using the hamster lung 2C19 or 4C1 cDNA clone as a probe. H28 was 1,876 bp long and encoded a polypeptide of 513 amino acids with a molecular weight of 58,079. The N-terminal 20 residues deduced from nucleotide sequence of H28 were identical to those determined by sequence analysis of purified hamster hepatic P-450MCI. The high similarity of the nucleotide and deduced amino acid sequences between H28 and P-450IA2 of other species indicated that H28 encoded a P-450 protein which belongs to the P-450IA2 family. Northern blot analysis revealed that the mRNAs for hamster P-450IA1 and IA2 were about 2.9 and 1.9 kb long, respectively. Hamster P-450IA1 mRNA was induced to the same level in lungs as in livers by MC treatment, whereas hamster P-450IA2 mRNA was induced and expressed only in hamster liver.     

cDNA cloning of cytochrome P-450 related to P-450p-2 from the cDNA library of human placenta. Gene structure and expression

We have isolated and analyzed cDNA (designated P-450HP cDNA) clones from a human placenta cDNA library, using the cDNA for rabbit pulmonary cytochrome P-450p-2, a prostaglandin omega-hydroxylase, as a hybridization probe. The cDNA obtained encoded a polypeptide comprising 511 amino acids with a calculated molecular mass of 58987 Da, and the amino acid sequence similarity with P-450p-2 and rat liver laurate omega-hydroxylase (P-450LA omega) was only about 50%. RNA blot analysis showed that the mRNA hybridizable with the human P-450HP cDNA was inducibly expressed 3-5-fold in rabbit small intestine and lung by gestation, but the expression remained constant in rabbit liver and kidney. This mode of expression was quite different from that of P-450p-2 and P-450LA omega. Interestingly, the mRNA hybridized with the cDNA of P-450HP was found to be expressed in all the human tumor tissues so far examined, in sharp contrast with the facts that almost all the other species of P-450s are known to disappear in the tumor tissues. Taken together, the deduced hemoprotein termed P-450HP dose not seem to be the human counterpart of rabbit P-450p-2 or rat P-450LA omega, and is presumably a new member of the P-450 family including P-450p-2 and P-450LA omega. Furthermore, the corresponding genomic DNA was also cloned and analyzed. The gene of P-450HP spanned 18.8 kb and was separated into 11 exons by 10 introns whose locations were completely different from those of P-450 genes so far determined.     

Maximizing the expression of mammalian cytochrome P-450 monooxygenase activities in yeast cells

Cytochrome P-450s constitute a superfamily of mono-oxygenases which require the association with specific redox enzymes bound to the endoplasmic reticulum membrane for their activity. Conditions for the functional expression of these mammalian enzymes in yeast cells and the respective merits and limitations of currently used P-450 expression systems, are considered. The dependence of the mouse P-450 IA1 specific activity on the cytochrome expression level in yeast microsomes is studied and results demonstrate that the low amounts of endogenous NADPH-cytochrome P-450 reductase and cytochrome b5 which are naturally present, are limiting for the heterologous monooxygenase activities. The sequences encoding human liver cytochrome b5, the native and a modified form of the yeast NADPH-cytochrome P-450 reductase were cloned by making use of PCR techniques, over-expressed in yeast as functional forms, and characterized. New vectors allowing a high level of mammalian P-450 expression upon induction were also constructed and tested. A strategy for the construction of a co-expression system allowing maximal activity of mammalian cytochrome P-450s is discussed.     

The formation of cytochrome P-450 from cytochrome P-420 is promoted by spermine

This paper is concerned with camphor-bound bacterial cytochrome P-450 and processes that alter its spin-state equilibrium and influence its transition to the nonactive form, cytochrome P-420, as well as its renaturation to the native camphor-bound cytochrome P-450. Spermine, a polycation carrying a charge of 4 +, and potassium, a monovalent cation, were shown to differently cause an increase of high-spin content of camphor-bound cytochrome P-450. The spermine-induced spin transition saturates around 75% of the high spin; a further addition of KCl to the spermine-containing sample shifted the spin state to 95% of the high spin. The volume change of these spin transitions as measured by the use of high pressure indicated an excess of -40 mL/mol for the sample containing potassium as compared to that containing spermine. These results suggest that the proposed privileged site for potassium has not been occupied by spermine and that pressure forces both the camphor and the potassium ion from its sites, allowing solvent movement into the protein as well as ordering of solvent by the excluded camphor and potassium. Cytochrome P-420 was produced from cytochrome P-450 by hydrostatic pressure in the presence of potassium, spermine, and cysteine. Potassium cation shows a bigger effect on the stability of cytochrome P-450 than spermine or cysteine, as revealed by a higher value of the pressure of half-inactivation, P1/2, and a bigger inactivation volume change. However, potassium cation did not promote renaturation of cytochrome P-420 to cytochrome P-450 while the presence of spermine did.(ABSTRACT TRUNCATED AT 250 WORDS)     

NADPH: cytochrome P-450 reductase in olfactory epithelium. Relevance to cytochrome P-450-dependent reactions.

The presence of a very active cytochrome P-450-dependent drug-metabolizing system in the olfactory epithelium has been confirmed by using 7-ethoxycoumarin, 7-ethoxyresorufin, hexobarbitone and aniline as substrates, and the reasons for the marked activity of the cytochrome P-450 in this tissue have been investigated. The spectral interaction of hexobarbitone and aniline with hepatic and olfactory microsomes has been examined. By this criterion there was no evidence for marked differences in the spin state of the cytochromes of the two tissues, or for the olfactory epithelium containing a greater amount of cytochrome capable of binding hexobarbitone, a very actively metabolized substrate. Rates of NADPH and NADH: cytochrome c reductase activity were found to be higher in the olfactory epithelium than in the liver, and direct evidence was obtained for a greater amount of the NADPH-dependent flavoprotein in the olfactory microsomes. Investigation of male rats and male and female mice, as well as male hamsters, demonstrated that, in all cases, the cytochrome P-450 levels of the olfactory epithelium were lower than those of the liver, while the 7-ethoxycoumarin de-ethylase and NADPH:cytochrome c reductase activities were higher. A correlation was found between 7-ethoxycoumarin de-ethylase and NADPH:cytochrome c reductase activities for both tissues in all species examined. The ratio of reductase to cytochrome P-450 was found to be considerably higher in the olfactory epithelium (1:2-1:3) than in the liver (1:11-1:15), regardless of the species examined, suggesting that facilitated electron flow may contribute significantly to the cytochrome P-450 catalytic turnover in the olfactory tissue.     

Oxidation of substituted N,N-dimethylanilines by cytochrome P-450: estimation of the effective oxidation-reduction potential of cytochrome P-450

Rates of N-demethylation of N,N-dimethylaniline and of eight meta- or para-substituted N,N-dimethylanilines by rat liver cytochrome P-450PB-B (P-450) were determined under conditions where oxidation was supported by iodosylbenzene or NADPH-P-450 reductase. The rates of dimethylaniline oxidation were found to correlate with the substrate oxidation-reduction potential within each series of substrates supported by a particular oxygen activation protocol; the kcat for each substrate studied was approximately 20-fold faster in the iodosylbenzene-supported system relative to the NADPH-P-450 reductase supported system. Since the N-demethylation of amines is believed to proceed via an initial electron-transfer step, a kinetic scheme for P-450 was proposed that enabled evaluation of the data according to theoretical treatments that correlate rates of electron transfer with extrakinetic parameters. In these analyses, the data could be fitted to the Rehm-Weller and Agmon-Levine equations, providing lambda values (for the energetics of enzyme-substrate reorganization) of 22-26 kcal mol-1 and apparent E1/2 (oxidation-reduction potentials) of 1.7-2.0 V (vs saturated calomel) for the oxidized enzyme. The apparent E1/2 for the enzyme is composed of contributions from the intrinsic potential of the active prosthetic core of the enzyme, the Fe = O - porphyrin species, and a coulombic factor that is a function of the charge-separated radical anion/radical cation pair produced upon electron transfer.(ABSTRACT TRUNCATED AT 250 WORDS)     

Protein components of a cytochrome P-450 linalool 8-methyl hydroxylase

The cytochrome P-450 heme-thiolate monooxygenases that hydroxylate monoterpene hydrocarbon groups are effective models for the cytochrome P-450 family. We have purified and characterized the three proteins from a P-450-dependent linalool 8-methyl hydroxylase in Pseudomonas putida (incognita) strain PpG777. The proteins resemble the camphor 5-exohydroxylase components in chemical and physical properties; however, they show neither immunological cross-reactivity nor catalytic activity in heterogenous recombination. These two systems provide an excellent model to probe more deeply the heme-thiolate reaction center, molecular domains of substrate specificity, redox-pair interactions, and the regulation of the reaction cycle.     

Denaturation of cytochrome P-450 by cyclophosphamide metabolites

Cyclophosphamide (CP) metabolites, acrolein and 4-hydroxy-CP, were found to denature rat liver microsomal cytochrome P-450, whereas another metabolite, phosphoramide mustard, CP $\text{per}\text{se}$ or its analog Ifosfamide had no effect. The denaturation produced by CP metabolites could be blocked by cysteine, suggesting an interaction between CP metabolite(s) and sulfhydryl groups in cysteine and probably in cytochrome P-450. These studies might explain the biochemical basis of the specific depression of various microsomal mixed function oxygenase activities produced by high doses of CP.     

Isolation and characterization of a cDNA clone from Catharanthus roseus encoding NADPH:cytochrome P-450 reductase, an enzyme essential for reactions catalysed by cytochrome P-450 mono-oxygenases in plants

The membrane-bound flavoprotein NADPH:cytochrome P-450 (cytochrome c) reductase, that functions in electron transfer to cytochrome P-450 mono-oxygenases, was purified from a cell suspension culture of the higher plant Catheranthus roseus . Anti-serum raised against the purified protein was found to inhibit NADPH:cytochrome c reductase activity as well as the activities of the cytochrome P-450 enzymes geraniol 10-hydroxylase and trans -cinnamate 4-hydroxylase, which are involved in alkaloid biosynthesis and phenylpropanoid biosynthesis, respectively. Immunoscreening of a C. roseus cDNA expression library resulted in the isolation of a partial NADPH: cytochrome P-450 reductase cDNA clone, which was identified on the basis of sequence homology with NADPH:cytochrome P-450 reductases from yeast and animal species. The identity of the cDNA was confirmed by expression in Escherichia coli as a functional protein capable of NADPH-dependent reduction of cytochrome c and neotetrazolium, two in vitro substrates for the reductase. The N-terminal sequence of the reductase, which was not present in the cDNA clone, was determined from a genomic NADPH: cytochrome P-450 reductase clone. It was demonstrated that the reductase probably is encoded by a single copy gene. A sequence comparison of this plant NADPH:cytochrome P-450 reductase with the corresponding enzymes from yeast and animal species showed that functional domains involved in binding of the cofactors FMN, FAD and NADPH are highly conserved between all kingdoms. In C. roseus cell cultures a rapid increase of the reductase steady state mRNA level was observed after the addition of fungal elicitor preparations that are known to induce cytochrome P-450-dependent biosynthetic pathways.     

Rotation of cytochrome P-450. II. Specific interactions of cytochrome P-450 with NADPH-cytochrome P-450 reductase in phospholipid vesicles

Purified rat liver microsomal cytochrome P-450 and NADPH-cytochrome P-450 reductase were co-reconstituted in phosphatidylcholine-phosphatidylethanolamine-phosphatidylserine vesicles using a cholate dialysis technique. The co-reconstitution of the enzymes was demonstrated in proteoliposomes fractionated by centrifugation in a glycerol gradient. The proteoliposomes catalyzed the N-demethylation of a variety of substrates. Rotational diffusion of cytochrome P-450 was measured by detecting the decay of absorption anisotropy r(t), after photolysis of the heme.CO complex by a vertically polarized laser flash. The rotational mobility of cytochrome P-450, when reconstituted alone, was found to be dependent on the lipid to protein ratio by weight (L/P450) (Kawato, S., Gut, J., Cherry, R. J., Winterhalter, K. H., and Richter, C. (1982) J. Biol. Chem. 257, 7023-7029). About 35% of cytochrome P-450 was immobilized and the rest was rotating with a mean rotational relaxation time phi 1 of about 95 mus in L/P450 = 1 vesicle. In L/P450 = 10 vesicles, about 10% of P-450 was immobile and the rest was rotating with phi 1 congruent to 55 mus. Co-reconstitution of equimolar amounts of NADPH-cytochrome P-450 reductase into the above vesicles results in completely mobile cytochrome P-450 with a phi 1 congruent to 40 mus. Only a small decrease in the immobile fraction of cytochrome P-450 is observed when the molar ratio of cytochrome P-450 to the reductase is 5. The results suggest the formation of a monomolecular 1:1 complex between cytochrome P-450 and NADPH-cytochrome P-450 reductase in the liposomes.     

Induction of cytochrome P-450-dependent drug metabolism by juvenile hormone mimics in mammalian cell culture

The monooxygenase activity of fetal hepatocytes in culture shows a differential response toward juvenile hormone I and analogs. Juvenile hormone I, R-20458, and Methoprene increase the deethyiation of 7-ethoxyresorufin while not affecting or even inhibiting the N-demethylation of p-chloro-N-methylaniline. RO-203600, a 1,3-benzodioxole-containing analog, increases both the deethylase and the N-demethylase, whereas Hydroprene does not affect either activity. The inductive effect with juvenile hormone I is obtained with exposure periods of at least 30 min and is maximum when the concentration of the hormone is 14 μM in the medium. This amount results in the covalent binding to cellular macromolecules of 1.3 × 19^?18 moles/cell. The induction requires continuous protein synthesis but RNA synthesis only for a short initial period. It is concluded that juvenile hormone and mimics induce specific cytochrome P-450 species in fetal liver cells even if the culture conditions are not optimal. The toxicological implications of these results are briefly discussed.