Cytochrome P-450 LM2 Research Articles

Hydroxyl radical-mediated, cytochrome P-450-dependent metabolic activation of benzene in microsomes and reconstituted enzyme systems from rabbit liver.

The mechanism of benzene oxygenation in liver microsomes and in reconstituted enzyme systems from rabbit liver was investigated. It was found that the NADPH-dependent transformation of benzene to water-soluble metabolites and to phenol catalyzed by cytochrome P-450 LM2 in membrane vesicles was inhibited by catalase, horseradish peroxidase, superoxide dismutase, and hydroxyl radical scavengers such as mannitol, dimethyl sulfoxide, and catechol, indicating the participation of hydrogen peroxide, superoxide anions, and hydroxyl radicals in the process. The cytochrome P-450 LM2-dependent, hydroxyl radical-mediated destruction of deoxyribose was inhibited concomitantly to the benzene oxidation. Also the microsomal benzene metabolism, which did not exhibit Michaelis-Menten kinetics, was effectively inhibited by six different hydroxyl radical scavengers. Biphenyl was formed in the reconstituted system, indicating the cytochrome P-450-dependent production of a hydroxycyclohexadienyl radical as a consequence of interactions between hydroxyl radicals and benzene. The formation of benzene metabolites covalently bound to protein was efficiently inhibited by radical scavengers but not by epoxide hydrolase. The results indicate that the microsomal cytochrome P-450-dependent oxidation of benzene is mediated by hydroxyl radicals formed in a modified Haber-Weiss reaction between hydrogen peroxide and superoxide anions and suggest that any cellular superoxide-generating system may be sufficient for the metabolic activation of benzene and structurally related compounds.

Fluorescent energy transfer measurements on fluorescein isothiocyanate modified cytochrome P-450 LM2

The distance between the heme iron and the N-terminus of cytochrome P-450 LM2 was determined by fluorescence energy transfer measurements. Fluorescein isothiocyanate which was covalently bound to the N-terminal methionine was used as donor chromophor. The R o value between fluorescein isothiocyanate and the heme was calculated to be 3.98 nm. The distance between the nitrogen of the N-terminal methionine and the heme was estimated with 2.84±0.23 nm excluding most likely the N-terminal amino acid of cytochrome P-450 LM2 to participate in the electron transfer to the heme iron. A cytochrome P-450 LM2 membrane model is proposed.

Reverse-phase high-performance liquid chromatography of hydrophobic proteins and fragments thereof

Reverse-phase high-performance liquid chromatography (HPLC) resolution and recovery of cytochrome P-450 and bovine rhodopsin, both integral membrane proteins, and large peptides derived from P-450 LM 2 were enhanced by utilizing ternary solvents. Surprisingly, most test materials eluted later in the gradient when using mixtures of acetonitrile and propanol in the mobile phase compared to using either solvent alone. Of the supports tested, the best recovery of hydrophobic cytochrome P-450 LM 4 was experienced on the less retentive CN-bonded phase. Two alternate solvents for HPLC of polypeptides are proposed: (1) 0.02–0.1 m hexafluoroacetone/NH 3, pH 7.2 for highly acidic peptides; and (2) 6 m formic acid/0.13 m trimethylamine, pH 1.5, vs 4 m formic acid/0.09 m trimethylamine in propanol for relatively insoluble peptides. Anomalous side reactions between formic acid and peptides can cause HPLC peak broadening, increased retention, and decreased resolution. These deleterious effects are thought to be due in part to formyl esterification of serine and threonine residues and appear to be reversible by aminoethanol treatment.

The complete amino acid sequence of rabbit phenobarbital-induced liver microsomal cytochrome P-450.

The complete amino acid sequence of the major phenobarbital-induced cytochrome P-450 (P-450LM2) from rabbit liver microsomes has been determined. The protein contains 489 amino acid residues in a single polypeptide chain and has Mr = 55,464. The sequence was compared with the amino acid sequence of P-450CAM and the nucleotide sequence of cDNA obtained from phenobarbital-induced rat liver cytochrome P-450 mRNA. These comparisons suggest that, despite functional similarities, the structural homology between microbial and microsomal cytochromes P-450 is limited to a single 8-residue region, and, in contrast, the structure of inducible microsomal cytochrome P-450 isozymes is highly conserved among mammalian species. Furthermore, we propose that the thiolate heme ligand of cytochrome P-450 is contributed by a cysteinyl residue near the COOH terminus, position 434 in the rabbit P-450LM2 sequence, based on the homology in this region with P-450CAM. The NH2 terminus of the protein from residues 1-310 is characterized by 8 hydrophobic segments 18-23 residues long, each of which is terminated by a cluster of charged amino acid residues. Residues 320-443 comprise a hydrophilic region which contains the putative heme binding cysteinyl residue as well as segments of homology with a constitutive rabbit cytochrome P-450 isozyme. The sequence data suggest that cytochrome P-450LM2 contains multiple transmembranous segments as well as a hydrophilic cytoplasmic domain. The hydrophilic domain contains regions of homology with several other cytochromes P-450, and thus appears to have an essential role in the biological function of the protein.

Identification and partial purification of hamster microsomal cytochrome P-450 isoenzymes

We have identified and partially purified three forms of cytochrome P-450 from hamster liver microsomes. Phenobarbital (PB) treatment induced three major polypeptides with relative mobilities ( M r ) of 47,000, 50,000 and 51,500. The 47,000 polypeptide was assigned as epoxide hydrolase, since it was also enhanced by trans-stilbene oxide (TSO) treatment. Two polypeptides ( M r = 48,500 and 53,500) were induced by both 3-methylcholanthrene (3-MC) and β-naphthoflavone (BNF) treatments. Treatment with Aroclor 1254 induced three polypeptides ( M r = 48,500, 50,000 and 53,500), indicating the induction of both drug- and carcinogen-inducible cytochrome P-450s. Liver microsomal benzo[ a]pyrene hydroxylase activity was not affected significantly by any of these inducers. In contrast, it was induced 2- to 3-fold in lung microsomes by 3-MC, BNF or Aroclor 1254 treatment. Benzphetamine N-demethylase and 7-ethoxycoumarin O-deethylase activities, expressed as nmoles of product formed per min per mg of liver microsomal protein, were increased 3- to 4-fold by either PB or Aroclor treatment. The activity of 7-ethoxycoumarin O-deethylase was the only one enhanced significantly by 3-methylcholanthrene or β-naphthoflavone treatment in liver microsomes. Pregnenolone-16-α-carbonitrile (PCN) and TSO did not alter any of these activities. The major polypeptides induced by PB ( M r = 50,000) and 3-MC ( M r = 48,500 and 53,500 respectively) were partially purified, to a specific content of 6–10 nmoles P-450/mg of protein and were active in catalyzing N-demethylation of benzphetamine, hydroxylation of benzo[ a]pyrene, and O-deethylation of 7-ethoxycoumarin with different substrate specificity. None of these isoenzymes immuno-cross-reacted with antibodies prepared against rabbit cytochrome P-450 LM2 or P-450 LM4.

Kinetics and mechanism of CO binding to cytochromes P-450LM2 and P-450LM4. Effect of phospholipid, nonionic detergent, and substrate binding.

The kinetics of carbon monoxide binding to purified, soluble cytochrome P-450LM2 from liver microsomes of phenobarbital-induced rabbits and P-450LM4 from liver microsomes of 5,6-benzoflavone-induced rabbits was determined in the presence of catalytically essential phospholipid, d-benzphetamine or nonionic detergent (0.1% Tween 80, 0.1% Triton N-101, or 0.3% n-octylglucoside). The kinetics under these conditions was consistent with a two-step binding mechanism consisting of bimolecular association followed by a unimolecular rearrangement step. However, when d-benzphetamine was added to P-450LM2 in the presence of sonicated dilauroylglyceryl-3-phosphorylcholine (0.15 mM) or any of the above detergents, CO binding followed simple second order kinetics with a bimolecular rate constant of approximately 3 x 10 M-1 s-1 at 20 degrees C. P-450LM4-CO binding kinetics was not affected by concurrent addition of phospholipid and d-benzphetamine. These results imply that simultaneous binding of amphiphile and substrate induces a conformation of reduced P-450LM2 which is different from that found when either substance is bound alone.

A form of rabbit liver cytochrome P-450 that catalyzes the 7α-hydroxylation of cholesterol

A form of cytochrome P-450 which comigrates with cytochrome P-450 LM4 (molecular weight, 55000) on SDS-polyacrylamide gel was purified from liver microsomes of cholestyramine-treated rabbits. This form of cytochrome P-450 catalyzed the 7α-hydroxylation of cholesterol with an activity of 37.5 pmol/min per nmol cytochrome P-450 in the reconstituted enzyme system containing cytochrome P-450 and NADPH-cytochrome P-450 reductase. The substrate specificity of this form of cytochrome P-450 was compared with cytochrome P-450 LM4 isolated from phenobarbital- and β-naphthoflavone-treated rabbit liver microsomes. The latter two isoenzymes do not catalyze 7α-hydroxylation of cholesterol, but are more active in O-deethylation of 7-ethoxycoumarin and p-nitrophenetole. Ouchterlony double diffusion revealed cross-reactivity between anti- P-450 LM4 (phenobarbital) IgG and cytochrome P-450 isolated from cholestyramine- or β-naphthoflavone-treated rabbit liver microsomes. A two-dimensional iodinated tryptic peptide fingerprint indicated only minor structural differences among these three cytochrome P-450 LM4 preparations.

Oxidation-reduction potential of soluble and membrane-bound rabbit liver microsomal cytochrome P-450 LM2.

The redox midpoint potentials of rabbit liver microsomal cytochromes P-450 and of soluble and membrane-bound rabbit liver microsomal cytochrome P-450 LM2 were determined using EPR-spectroscopy and absorption difference spectrometry with NADPH or dithionite as reductants. Using EPR, a redox midpoint potential of -0.36 V was obtained both for the low spin and the high spin components of microsomal cytochrome P-450. Spectrophotometrical determinations yielded very similar values: -0.37 V and -0.34 V for the low and high spin signals, respectively. Soluble cytochrome P-450 LM2 had a midpoint potential of -0.32 V. This redox potential was not significantly affected by incorporation of the protein into an artificial membrane structure or, furthermore, by the presence of cytochrome b5 the same membrane.

Spin equilibrium relaxation kinetics of cytochrome P450 LM2

The relaxation kinetics of the temperature-dependent spin equilibrium of cytochrome P450 LM2 has been resolved by means of a laser temperature-jump device. The Nd laser system in the Q-switched mode delivers 3 J/25 ns pulses which were Raman shifted to 1.89 μm by use of a high pressure H 2-gas cell. The LM2 equilibrium relaxation proceeds in the ns-range, the rate constants of the on- and off-reactions amount to about 10 6 s −1 in correspondence with respective processes of other hemoproteins and protein-free model compounds. Benzphetamine increases the on-rate constant (low-spin → high-spin) about 5 fold, the off-process is nearly unchanged.

Relationship between state of aggregation and catalytic activity for cytochrome P-450LM2 and NADPH-cytochrome P-450 reductase.

The detergent 1-O-n-octyl-beta-D-glucopyranoside (octylglucoside) was found to replace the phospholipid requirement in the demethylation of benzphetamine by cytochrome P-450LM2 and NADPH-cytochrome P-450 reductase purified from phenobarbital-treated rabbit liver. At low enzyme concentration (0.1 microM) in the absence of glycerol and phosphate, the maximum rate of benzphetamine-specific NADPH oxidation was approximately 35% of that observed in the presence of dilauroylglyceryl-3-phosphoryl choline. At higher enzyme concentration (2.5 microM) and in the presence of 0.15 M phosphate, 20% glycerol, octylglucoside was as effective as phospholipid in stimulating the production of formaldehyde from benzphetamine. The detergent concentration required for maximal enzymatic activity was 2.5-4.0 g/liter, depending on the cytochrome preparation used. At higher octylglucoside concentrations (5-7 g/liter), activity decreased to zero, although neither enzyme appeared to be irreversibly denatured at these detergent concentrations. Sedimentation equilibrium experiments with P-450LM2 alone or in the presence of equimolar reductase showed that increasing octylglucoside levels promoted disaggregation of the cytochrome. Pentamers and hexamers predominated at detergent concentrations where maximal activity was observed, while higher levels of detergent where activity was absent produced cytochrome dimers and, ultimately, monomers. The reductase was monomeric at detergent levels between at least 3 and 7 g/liter. Moreover, both gel filtration and sedimentation equilibrium experiments demonstrated that a stable complex between P-450LM2 and its reductase was not formed at octylglucoside concentrations where high activity was evident. These results are consistent with a model of P-450/reductase interaction in which functional aggregates of three to six cytochrome polypeptides move laterally in the microsomal membrane and interact with the reductase by random collision.

Hydroxylations in biosynthesis of bile acids. Isolation of subfractions with different substrate specificity from cytochrome P-450LM4.

Chromatography of electrophoretically homogeneous cytochrome P-450LM4 from cholestyramine-treated rabbits on octylamine-Sepharose resulted in the isolation of two subfractions, cytochrome P-450LM4 I and cytochrome P-450LM4 II, with different catalytic properties. The original cytochrome P-450LM4 fraction catalyzed 7 alpha-hydroxylation of cholesterol, 12 alpha-hydroxylation of 5 beta-cholestane-3 alpha, 7 alpha-diol, 25-hydroxylation of 5 beta-cholestane-3 alpha, 7 alpha-diol and 5 beta-cholestane-3 alpha, 7 alpha, 12 alpha-triol, 6 beta-hydroxylation of testosterone, and demethylation of ethylmorphine. Cytochrome P-450LM4 I was inactive in cholesterol 7 alpha-hydroxylation, but catalyzed the other hydroxylations. Cytochrome P-450LM4 II catalyzed efficient cholesterol 7 alpha-hydroxylation. It also catalyzed the other hydroxylations, although at lower rates than cytochrome P-450LM4 I. Emulgen inhibited all steroid hydroxylase activities in cytochrome P-450LM4 II except the cholesterol 7 alpha-hydroxylase activity. Cytochrome P-450LM4 I and cytochrome P-450LM4 II showed the same apparent molecular weight and spectral properties as the original cytochrome P-450LM4 fraction. The two subfractions differed in amino acid composition. They produced similar but not identical one-dimensional peptide maps upon limited proteolysis with papain, chymotrypsin, and trypsin. The results show that cytochrome P-450LM4 from cholestyramine-treated rabbits contains at least two species with different amino acid compositions and different substrate specificities toward C27-steroids involved in biosynthesis of bile acids.

Studies on the active center of cytochrome P-450 using a spin-labeled type I substrate analogue

The binding of an active-site-targeted spin-labeled compound to cytochrome P-450 showing type I spectral characteristics in its optical difference spectrum was studied by optical and EPR spectroscopy and compared with that of other spin-labeled substrate analogues studied previously. The spin-labeled aminoisoindoline (1,1,3,3-tetraethyl-2-oxy-5-aminoisoindoline) binds with high affinity to the isozyme cytochrome P-450 LM2 (number of binding sites n ≈ 1, apparent dissociation constant approx. 15 μM). The analysis of the EPR binding data by Scatchard plots reveals additional unspecific binding sites at the enzyme ( n ≈ 5; 250 μM). Competitive binding studies with real type I substrates show the binding of the spin-labeled compound at the substrate-specific binding site of the enzyme. The bound probe is characterized by the paramagnetic nitroxide group being more than 7 Å, from the heme iron as well as from the protein surface. The last conclusion was derived from studying the accessibility of the bound probe to ferricyanide ions.

Hydrodynamic properties of monomeric cytochromes P-450 LM2 and P-450 LM4 in [formula omitted

Sedimentation equilibrium and sedimentation velocity measurements were carried out on cytochrome P-450 LM2 from phenobarbital-treated rabbit liver and on cytochrome P-450 LM4 from 5,6-benzoflavone-treated rabbit liver in the presence of the nonionic detergent 1-O- n - octyl-β- D- glucopyranoside . P-450 LM2 was monomeric with a molecular weight of 48,800 and a Stokes radius of 3.1 nm in 7 g/l detergent and P-450 LM4 was monomeric with a molecular weight of 49,800 and a Stokes radius of 2.6 nm at 5 g/l detergent. Both particles were spherical in shape under these conditions. Neither cytochrome was irreversibly denatured at these detergent concentrations as indicated by the ability to form substantial amounts (>60%) of the CO adduct with an absorption maximum at 451 nm (P-450 LM2) or 448 nm (P-450 LM4) when diluted into detergent-free buffer containing CO and sodium dithionite.

Hydroxylations in Biosynthesis of Bile Acids. Cytochrome P-450 LM4 and 12alpha-Hydroxylation of 5beta-Cholestane-3alpha,7alpha-diol

12alpha-Hydroxylation of 5beta-cholestane-3alpha, 7alpha-diol was studied in reconstituted systems consisting of electrophoretically homogeneous cytochrome P-450 LM4 fractions and NADPH-cytochrome P-450 reductase from rabbit liver microsomes. Cytochrome P-450 LM4 fractions were prepared from untreated, phenobarbital-treated, beta-naphthoflavone-treated and starved rabbits. The purified cytochromes catalyzed 12alpha-hydroxylation more efficiently than the corresponding microsomes. In the reconstituted systems, carbon monoxide inhibited 12alpha-hydroxylation by 50-80%. The rate of 12alpha-hydroxylation was three to four times higher with cytochrome P-450 LM4 fractions from starved rabbits than with cytochrome P-450 LM4 fractions from untreated, phenobarbital-treated and beta-naphthoflavone-treated animals. Amino acid analyses, peptide mapping experiments as well as absorption spectral and circular dichroism spectral analyses revealed physical differences between cytochrome P-450 LM4 fractions from starved animals and preparations from phenobarbital-treated animals. The results indicate the presence of a cytochrome P-450 species in the cytochrome P-450 LM4 fraction specific for 12alpha-hydroxylation.

The quantitation of liver cytochrome P450— LM2 mRNA in rabbits of different ages and after phenobarbital treatment

Using very young (neonate or 0 day), 50-, 100-, 300- and 830-day-old rabbits we have studied changes in liver drug-metabolizing activities. Total liver microsomal cytochrome P450 content, benzphetamine N-demethylase and 7-ethoxycoumarin 0-deethylase activities all decline with increasing age. When the rabbits of different age groups were pretreated with phenobarbital, liver microsomal enzyme activities increased in each group as compared with its control, but the total induced activities still decreased with increasing age. Using a specific assay for translatable cytochrome P450-LM 2 mRNA, we show that the age-dependent decrease in control and phenobarbital-induced enzyme activities is due to a decrease in the levels of translatable mRNA specific for cytochrome P450-LM 2. This is the first report on an age-dependent decrease in the level of a specific translatable mRNA. We do not know whether this decrease is due to decreased mRNA synthesis, or increased mRNA degradation.

Association of cytochrome b5 and cytochrome P-450 reductase with cytochrome P-450 in the membrane of reconstituted vesicles.

A protein-protein association of cytochrome P-450 LM2 with NADPH-cytochrome P-450 reductase, with cytochrome b5, and with both proteins was demonstrated in reconstituted phospholipid vesicles by magnetic circular dichroism difference spectra. A 23% decrease in the absolute intensity of the Soret band of the magnetic CD spectrum of cytochrome P-450 was observed when it was reconstituted with reductase. A difference spectrum corresponding to a 7% decrease in absolute intensity was obtained when cytochrome b5 was incorporated into vesicles that already contained cytochrome P-450 and cytochrome P-450 reductase compared to a decrease of 13% in absolute intensity when cytochrome b5 was incorporated into vesicles that contained only cytochrome P-450. The use of the magnetic circular dichroism confirmed that protein-protein associations that have been detected by absorption spectroscopy between purified and detergent-solubilized proteins also exist in membranes. High ionic strength was shown to interrupt direct electron flow from cytochrome P-450 reductase to cytochrome P-450 but not the electron flow from reductase through cytochrome b5 to cytochrome P-450. Upon incorporation of cytochrome b5 into cytochrome P-450- and cytochrome P-450 reductase-containing vesicles, an increase of benzphetamine N-demethylation activity was observed. The magnitude of this increase was numerically identical to the residual activity of the reconstituted vesicles measured in the presence of 0.3 M KCl. It is concluded that there is a requirement for at least one charge pairing for electron transfer from reductase to cytochrome P-450. These observations are combined in a proposed mechanism of coupled reversible association reactions in the membrane.

Lipid-protein interactions as determinants of activation or inhibition by cytochrome b5 of cytochrome P-450-mediated oxidations.

Activation or inhibition by cytochrome b5 of benzphetamine N-demethylation was studied in micelle-reconstituted systems containing cytochrome P-450 LM2, NADPH-cytochrome P-450 reductase, and dilauroyl-phosphatidylcholine. The effects of cytochrome b5 were critically dependent on both protein:protein and lipid:protein ratios. A 200% stimulation of N-demethylation by cytochrome b5 was obtained at cytochrome P-450 reductase:cytochrome P-450 ratios similar to those in microsomes, compared to only a 20% stimulation at a ratio of 1:1. At lipid:protein ratios less than 50:1, the addition of cytochrome b5 caused significant inhibition of benzphetamine N-demethylation. Such an inhibition could be partially reversed by increasing phospholipid content of micelles and was not seen in vesicle-reconstituted systems at cytochrome b5:cytochrome P-450 ratios of 1:1 or lower. At high cytochrome P-450 reductase:cytochrome P-450 ratios, addition of cytochrome b5 did not alter the efficiency (80%) with which NADPH was utilized: however, at ratios similar to those in microsomes, an increase in efficiency from 42% to 80% was observed. The function of cytochrome b5 was interpreted in terms of a model in which inhibition of cytochrome P-450-mediated reactions results from changes in phospholipid-protein interactions and activation occurs via facilitation of electron transfer between NADPH-cytochrome P-450 reductase and cytochrome P-450 in the membrane.

Insights into the active site of the cytochrome P-450 haemoprotein family--a unifying concept based on structural considerations.

1. Cytochromes P-450CAM of Pseudomonas putida and P-450LM-2 of rabbit liver, and to a lesser extent also adrenocortical mitochondrial cytochromes P-450(11) beta and P-450SCC, were investigated by active site-targeted reagents and by immunochemical techniques. The results of these studies and of the alignment of the amino acid sequences of cytochromes P-450CAM and of phenobarbital-induced rat-liver P-450LM support the following conclusions. 2. Cytochrome P-450 haemoproteins follow a common architectural design, most readily apparent in the surroundings of the haem group. Sequence homology and associated folding leads to immunochemical similarities revealed by radioimmune assay. 3. The haem domain is partially exposed on the surface of the protein. This part appears to be tightly structured from several distinct portions of the polypeptide chain which carry the haem catalytic site and the substrate binding site, respectively. 4. After photocovalent labelling with substrate- or inhibitor-derived arylazido reagents, a substantial portion of the haem domain may be excised by BrCN or formic acid and purified without loss of haem ('haemopeptide'). A major antibody-binding site is associated with this fragment. 5. Exploration of the haem pocket of cytochrome P-450CAM, using bromoacetyl and mercapto derivatives of its substrate, camphor, revealed an activated SH group responsible for binding and precision alignment of camphor toward the haem iron, presumably by forming a transient thiohemiketal bond. This cysteine may also function in translocation of nascent product for facilitated release. Specific terpenoid substrates of other cytochrome P-450 haemoproteins may be bound in a similar manner. 6. From radiolabelled peptide studies the substrate-binding cysteine was identified as residue 56 in the amino acid sequence of cytochrome P-450CAM. The haemchelating cysteine occupies position 134 or 146 in cytochrome P-450CAM and position 152 or 173 in cytochrome P-450LM, respectively.

Cytochrome P-450 rotamers control mixed-function oxygenation in reconstituted membranes. Rotational diffusion studied by delayed fluorescence depolarization.

1. Rotational diffusion of purified rabbit-liver microsomal cytochrome P-450 LM2 in reconstituted lipid-vesicle membranes was investigated by measurement of time-dependent polarized emission of delayed fluorescence. 2. Cytochrome P-450 labelled with diiodofluorescein iodoacetamide exhibited strict uniaxial rotation about the normal to the membrane. 3. Benzphetamine retards rotation, while reduction of the cytochrome P-450 substrate-complex accelerates rotation. 4. A model is proposed in which cytochrome P-450 forms disc-shaped rotamers immersed in the bilayer membrane to a depth which is varied by substrate-induced and redox state-dependent conformational changes. The model describes a regulating mechanism of the electron transfer-controlling mixed-function oxygenation.

On the significance of the cytochrome P-450-dependent hydroxyl radical-mediated oxygenation mechanism.

1. Reconstituted membrane vesicles containing purified preparations of cytochrome P-450 LM2 and NADPH-cytochrome P-450 reductase effectively destroyed 2-deoxy-D-ribose in an NADPH-dependent process. 2. The destruction was mediated by hydroxyl radicals formed in an iron-catalysed Haber-Weiss reaction between superoxide anions and hydrogen peroxide liberated from the haemoprotein. 3. Administration of ethanol or benzene to rabbits, compounds known to be oxygenated by the hydroxyl radical-dependent mechanism, resulted in induction of a species of cytochrome P-450 effective in the radical-dependent metabolism of both chemicals. 4. Benzene treatment of rabbits also resulted in an enhanced hydroxyl radical-dependent metabolism of ethanol and benzene in liver microsomes. 5. It is suggested that, for certain substrates, hydroxyl radical-mediated cytochrome P-450-dependent oxygenation reactions are of importance for the microsomal metabolism of these compounds. 6. It is speculated that radical-producing species of cytochrome P-450 may contribute to hydroxyl radical-mediated cell damage.