Cytochrome P-450-catalysed monoterpenoid oxidation in catmint (Nepeta racemosa) and avocado (Persea americana); evidence for related enzymes with different activities

A cytochrome P-450 present in ripening avocado (Persea americana) fruit mesocarp (CYTP71A1) had previously been shown to metabolize the monoterpenoids nerol and geraniol (Hallahan et al. (1992) Plant Physiol. 98, 1290-1297). Using DNA encoding CYP71A1 as a hybridization probe, we have shown by Southern analysis that a related gene is present in the catmint, Nepeta racemosa. RNA blot analysis, together with Western analysis of catmint leaf polypeptides using avocado cyt P-450 antiserum, showed that a closely related gene is expressed in catmint leaves. Cytochrome P-450 in catmint microsomes catalysed the specific hydroxylation of nerol and geraniol at C-10, whereas avocado CYP71A1, in either avocado microsomes or heterologously expressed in yeast, catalysed 2,3- or 6,7-epoxidation of these substrates. These results suggest that orthologous genes of the CYP71 family are expressed in these two plant species, but catalyse dissimilar reactions with monoterpenoid substrates.     

Hydroxylation of geraniol and nerol by a monooxygenase from Vinca rosea

A microsomal mixed function oxidase isolated from V. rosea seedlings was shown to catalyze the hydroxylation of the monoterpene alcohols, geraniol and nerol, to their corresponding 10-hydroxy derivatives. Hydroxylase activity was dependent upon NADPH and oxygen and was associated with the 100,000 X g pellet which exhibited a characteristic reduced P-450-CO binding spectra. Light reversible inhibition by CO as well as differential sensitivity to other inhibitors established the hydroxylase as a cytochrome P-450 type. Cis-trans isomerase activity was not observed in this preparation. Both geraniol and nerol were shown to be hydroxylated almost exclusively at the C-10 methyl group.     

Characterization of a cytochrome P-450 dependent monoterpene hydroxylase from the higher plant Vinca rosea.

A monooxygenase isolated from 5-day old etiolated Vinca rosea seedlings was shown to catalyze the hydroxylation of the monoterpene alcohols, geraniol and nerol, to their corresponding 10-hydroxy derivatives. Hydroxylase activity was inpendent upon NADPH (neither NADH nor combination of NADH, NADP+ and ATP served as substitutes) and O2. Geraniol hydroxylation was enhanced by dithiothreitol (monothiols were less effective) and inhibited by phospholipases, thiol reagents, metyrapone, and cytochrome c, as well as other inhibitors of cytochrome P-450 systems. Geraniol was hydroxylated at a faster rate than nerol, but the alcohols possessed similar apparent Km values. The membrane-bound hydroxylase was solubilized by treatment with sodium cholate, Renex-30, or Lubrol-WX. Cholate-treated enzyme was resolved by DEAE-cellulose chromatography and reconstitution of the hydroxylase was effected utilizing different fractions containing cytochrome P-450, a NADPH-cytochrome c reductase, and lipid.     

Cytochrome P-450 from the Mesocarp of Avocado (Persea americana)

The microsomal fraction from the mesocarp of avocado (Persea americana) is one of few identified rich sources of plant cytochrome P-450. Cytochrome P-450 from this tissue has been solubilized and purified. Enzymatic assays (p-chloro-N-methylaniline demethylase) and spectroscopic observations of substrate binding suggest a low spin form of the cytochrome, resembling that in the microsomal membrane, can be recovered. However, this preparation of native protein is a mixture of nearly equal proportions of two cytochrome P-450 polypeptides that have been resolved only under denaturing conditions. Overall similarities between these polypeptides include indistinguishable amino acid compositions, similar trypsin digest patterns, and cross reactivity with the same antibody. The amino terminal sequences of both polypeptides are identical, with the exception that one of them lacks a methionine residue at the amino terminus. This sequence exhibits some similarities with the membrane targeting signal found at the amino terminus of most mammalian cytochromes P-450.     

Evidence for geraniol as an obligatory precursor of isothujone

Isothujone (trans-thujan-3-one) was formed from MVA-[¹⁴C, ³H] in Tanacetum vulgare with retention of the pro-(4R) hydrogen of precursor, but with loss of the pro-(4S) hydrogen and of one hydrogen from C-5. Cell-free extracts could not sustain the formation of isothujone from MVA but yielded geraniol and nerol (3,7-dimethylocta-trans-2,6-dien-1-ol and its cis isomer) with retention of the pro-(4R) and loss of the pro-(4S) hydrogen in each case: no hydrogen was lost from C-5 of MVA in formation of geraniol, but one such atom was lost in the formation of nerol. These results support the sequence: geraniol → nerol → isothujone: in which the first two compounds (or their biogenetic equivalents) are interconverted by a redox process involving their derived aldehydes. They are not consistent with a direct pathway to nerol from C₅ intermediates or with routes involving cyclisation of linalol (3,7-dimethylocta-1,6-dien-3-ol) formed directly from the C₅ compounds or from geraniol. The cell-free preparations could not interconvert geraniol and nerol, their phosphates or pyrophosphates. This may be due to the inability of a prenyltransferase-isomerase multi-enzyme system to accept exogenously-supplied intermediates under these (in vitro) conditions.     

Electron paramagnetic resonance studies on hepatic microsomal cytochrome P-450 from a marine teleost fish

Hepatic microsomal cytochrome P-450 from fish (Stenotomus versicolor), untreated or treated with 3-methylcholanthrene, 5, 6-benzoflavone, or tricaine methanesulfonate, exhibited an absorption maximum at 450 nm when reduced and ligated to CO. Microsomes from all groups exhibited EPR spectra with g values near 2.4, 2.24 and 1.9, yielding crystal field parameters similar to those for cytochrome P-450 from a variety of other sources. Treatment with 3-methylcholanthrene or 5, 6-benzoflavone resulted in elevated levels of aryl hydrocarbon (benzo[a]pyrene) hydroxylase activity yet produced no apparent change in the levels or optical properties of CO-ligated cytochrome P-450. Tricaine methanesulfonate, a common fish anaesthetic, caused a decrease in the levels of fish cytochrome P-450.     

Purification of rat liver microsomal cytochrome P-450b without the use of nonionic detergent

Sodium cholate, Emulgen 911, and (3-[(-cholamidopropyl)-dimethyl- ammonio]-1-propanesulfonate) (CHAPS) were selected to examine the effects of ionic, nonionic, and zwitterionic detergents on testosterone hydroxylation catalyzed by four purified isozymes of rat liver microsomal cytochrome P-450, namely P-450a, P-450b, P-450c, and P-450h, in reconstituted systems containing optimal amounts of dilauroylphosphatidylcholine and saturating amounts of NADPH- cytochrome P-450 reductase (reductase). The major phenobarbital-inducible form of rat liver microsomal cytochrome P-450, designated P-450b, was extremely sensitive to the inhibitory effects of Emulgen 911, which is used in several procedures to purify this and other forms of cytochrome P-450. In contrast, sodium cholate and CHAPS had little effect on the catalytic activity of cytochrome P-450b, even at ten times the concentration of Emulgen 911 effecting 50% inhibition (IC-50). By substituting the zwitterionic detergent CHAPS for Emulgen 911, we purified cytochrome P-450b without the use of nonionic detergent. The protein is designated cytochrome P-450b^* to distinguish it from cytochrome P-450b purified with the use of Emulgen 911. NADPH-cytochrome P-450 reductase was also purified both with and without the use of nonionic detergent. The absolute spectra of cytochrome P-450b and P-450b^* were indistinguishable, as were the carbon monoxide (CO)- and metyrapone-difference spectra of the dithionite-reduced hemoproteins. When reconstituted with NADPH-cytochrome P-450 reductase and dilauroylphosphatidylcholine, cytochromes P-450b and P-450b^* catalyzed the N-demethylation of benzphetamine and aminopyrine, the 4-hydroxylation of aniline, the O-dealkylation of 7-ethoxycoumarin, the 3-hydroxylation of hexobarbital, and the 6-hydroxylation of zoxazolamine. Both hemo-proteins catalyzed the 16α- and 16β-hydroxylation of testosterone, as well as the 17-oxidation of testosterone to androstenedione. Both hemoproteins were poor catalysts of erythromycin demethylation and benzo[a]pyrene 3-/9-hydroxylation. The rate of biotransformation catalyzed by cytochrome P-450b^* was up to 50% greater than the rate catalyzed by cytochrome P-450b when reconstituted with either reductase or reductase^*. The activity of cytochrome P-450b and P-450b^* increased up to 50% when reconstituted with reductase^* instead of reductase. In addition to establishing the feasibility of purifying an isozyme of rat liver microsomal cytochrome P-450 without the use of nonionic detergent, these results indicate that the catalytic activity of cytochrome P-450 is not unduly compromised by residual contamination with the nonionic detergent Emulgen 911.     

Aryl hydrocarbon hydroxylase in larvae of the southern armyworm (Spodoptera eridania)

Aryl hydrocarbon (benzo[a]pyrene) hydroxylase (AHH) activity in midgut microsomes of southern armyworm (Spodoptera eridania) larvae was induced 11-fold and 5.6-fold respectively following three days of feeding on diets containing pentamethylbenzene or naphthalene (both 0.2% w/v). β-Naphthoflavone and Aroclor 1254 were less effective inducers of AHH activity, phenobarbital was only slightly active and 3-methylcholanthrene caused a decrease in enzyme activity. AHH activity in microsomes from untreated and induced larvae was susceptible to inhibition by α-naphthoflavone, 1-phenylimidazole and piperonyl butoxide. Equilibrium dialysis studies with 1-(4′-³H-phenyl)imidazole showed that control and induced armyworm midgut microsomes contained a class of cytochrome(s) P-450 with a uniformly high affinity for phenylimidazole. It is concluded that AHH activity in the armyworm is catalyzed by a class of cytochrome(s) P-450 with characteristics intermediate between mammalian cytochrome(s) P-450 and P-448.     

Induction of a cytochrome P-450-dependent fatty acid monooxygenase in Bacillus megaterium by a barbiturate analog, 1-[2-phenylbutyryl]-3-methylurea

Summary In previous publications from our laboratory, we reported that a soluble, cytochrome P-450-dependent fatty acid monooxygenase from Bacillus megaterium ATCC 14581 can be induced by phenobarbital and a variety of other barbiturates. The tested barbiturates showed an excellent correlation between increasing lipophilicity and increasing inducer potency (Kim BH, Fulco AJ; Biochem Biophys Res Commun 116: 843–850, 1983). The only exception proved to be mephobarbital (N-methylphenobarbital) which, although more lipophilic than phenobarbital, is not an inducer of fatty acid monooxygenase activity. We have now found that 1-[2-phenylbutyryl]-3-methylurea (PBMU), an acylurea that can be derived from mephobarbital by hydrolytic cleavage of the barbiturate ring, is an excellent inducer of this activity. Paradoxically, the addition of mephobarbital to the bacterial growth medium containing PBMU significantly enhances the apparent potency of the acylurea to induce fatty acid monooxygenase activity as measured in cell-free extracts. When cell-free extracts of cells grown separately in PBMU or mephobarbital are mixed no enhancement of activity is seen. This finding suggests that the effect of mephobarbital is to somehow increase the efficiency of PBMU as an inducer of the P-450-dependent fatty acid monooxygenase rather than to induce an activator of this enzyme or a rate-limiting component of the monooxygenase system. Finally, both mephobarbital and PBMU induce the synthesis of total cytochrome P-450 in B. megaterium although PBMU is a much more potent P-450 inducer. For cytochrome P-450 induction, however, there is no synergistic or even additive effect when mephobarbital and PBMU are used together in the bacterial growth medium.Abbreviations PBMU 1-[2-phenylbutyryl]-3-methylurea - M.P. melting point     

Human cytochrome P-450 PB-1: A multigene family involved in mephenytoin and steroid oxidations that maps to chromosome 10

The cytochrome P-450 monooxygenase system possesses catalytic activity toward many exogenous compounds (e.g., drugs, insecticides, and polycyclic aromatic hydrocarbons) and endogenous compounds (e.g., steroids, fatty acids, and prostaglandins). Multiple forms of cytochrome P-450 with different substrate specificities have been isolated. In the present paper we report the isolation and sequence of a cDNA clone for the human hepatic cytochrome P-450 responsible for mephenytoin (an anticonvulsant) oxidation. The mephenytoin cytochrome P-450 is analogous to the rat cytochrome P-450 form termed PB-1 (family P450C2C). We also report that human PB-1 is encoded by one of a small family of related genes all of which map to human chromosome 10q24.1-10q24.3. The endogenous role of this enzyme appears to be in steroid oxidations. This cytochrome P-450 family does not correspond to any of the hepatic cytochrome P-450 gene families previously mapped in humans.     

Cytochrome P-450 heme and the regulation of hepatic heme oxygenase activity

The degradation of cytochrome P-450 heme in the liver has been studied by a new approach. In rats, hepatic heme was labeled by administration of a tracer pulse of [5-¹⁴C]δ-aminolevulinic acid (ALA), and its degradation was analyzed in terms of labeled carbon monoxide (¹⁴CO) excretion, which is a specific degradation product of the labeled heme. Within minutes after administration of [5-¹⁴C]ALA, 14CO was detectable and increased after 2 h to an “early peak,” reflecting the elimination of labeled heme from a rapidly turning over pool in the liver. Beyond the early peak, the rate of ¹⁴CO production decreased in a log-linear manner, consistent with the degradation of heme in stable hepatic hemoproteins. From the rate at which ¹⁴CO production declined during this phase, from the predominant labeling of cytochrome P-450 heme by the administered [5-¹⁴C]ALA and from the known turnover characteristics of this hemoprotein in the liver, it could be inferred that production of ¹⁴CO—between 16 and 30 h after administration of labeled ALA—largely reflected degradation of cytochrome P-450 heme. This approach, which permits serial measurements in a single animal, was used to study the effect on cytochrome P-450 heme of administered heme or endotoxin, both of which are potent stimulators of hepatic heme oxygenase activity. Both of these substances caused marked acceleration of the degradation of cytochrome P-450 heme, the effect occurring over the same dose range as that for stimulation of hepatic heme oxygenase. The findings suggest that stimulation of this enzyme activity in the liver is closely related to the rate of degradation of cytochrome P-450 heme.     

Biosynthesis of geraniol and nerol in cell-free extracts of Tanacetum vulgare

Cell-free extracts from leaves of Tanacetum vulgare synthesised geraniol and nerol (3,7-dimethylocta-trans-2-ene-1-ol and its cis isomer) in up to 11·9 and 2·4% total yields from IPP-[4-¹⁴C] and MVA-[2-¹⁴C] respectively. Optimum preparations were obtained from plant material just before the onset of flowering. The ratio of the monoterpenols varied 28-fold for different preparations under conditions where these products or their phosphate esters were not interconverted. Similar extracts incorporated α-terpineol-[¹⁴C] and terpinen-4-ol-[¹⁴C] (p-menth-1-en-8- and -4-ol respectively) in 0·05 to 2·2% yields into a compound tentatively identified as isothujone (trans-thujan-3-one), and preparations from flowerheads converted IPP-[4-¹⁴C] in 2·7% yield into geranyl and neryl β-d-glucosides. Inhibitors of IPP-isomerase had little effect on the incorporation of IPP into the monoterpenols in cell-free systems from which endogenous compounds of low molecular-weight had been removed. The inference that a pool of protein-bonded DMAPP or its biogenetic equivalent was present was supported by the demonstration that geraniol and nerol biosynthesised in the absence of the inhibitors were predominantly (65 to 100%) labelled in the moiety derived from IPP.     

Purification, composition, and some properties of rat liver carbamyl phosphate synthetase (ammonia).

Hepatic microsomes prepared from vitamin E deficient and supplemented rats were analyzed for cytochrome P-450 content and drug metabolizing activity. Reduced levels of benzo[α]pyrene hydroxylase and ethylmorphine N-demethylase activities were observed in microsomes derived from rats fed a diet deficient in vitamin E compared to those of control rats. NADPH-mediated destruction of P-450, and pentobarbital and zoxazolamine sleeping times were similar in the two groups. Induction with 3-methylcholanthrene raised the levels of benzo[α]pyrene hydroxylase activity of both supplemented and deficient rats to the same absolute levels. No differences were noted in cytochrome P-450 or P-448 content between control and tocopherol deficient rats, nor did the activity of liver catalase differ between the two dietary groups. Thus, these studies did not demonstrate any impairment of heme protein synthesis in vitamin E deficient rats.     

The role of cytochrome P-450 forms in 2-aminoanthracene and benz[alpha]pyrene mutagenesis.

The role of four forms of cytochrome P-450 from rabbit liver in the metabolic activation of two suspected carcinogens, 2-aminoanthracene and benz[α]pyrene, was investigated with a S. typhimurium tester strain, TA 98. Each of the forms, 2,3,4 and 6 was reconstituted with NADPH-cytochrome P-450 reductase and lipid, and assay conditions were established such that the cytochrome P-450 concentration was rate-limiting. Under these conditions, cytochrome P-450 form 4, but not the other forms, converted 2-aminoanthracene into a potent mutagen. In contrast, form 6 was the only form which metabolized benz[α]pyrene to a mutagen. These results indicate that specific cytochrome P-450 forms preferentially activate particular mutagens.     

7 Alpha-hydroxylation of cholesterol by reconstituted systems from rat liver microsomes

A reconstituted system from rat liver microsomes, consisting of partially purified fractions of cytochrome P-450 and NADPH-cytochrome P-450 reductase was shown to catalyze 7α-hydroxylation of cholesterol in the presence of NADPH and a synthetic phosphatidylcholine. The rate of 7α-hydroxylation of added [4-¹⁴C] cholesterol was linear with the concentration of cytochrome P-450 and increased with the concentration of NADPH-cytochrome P-450 reductase up to a certain level and then remained constant. Omission of phosphatidylcholine resulted only in a 20% decrease in cholesterol 7α-hydroxylase activity of the system. The rate of 7α-hydroxylation was 2–3 times higher in reconstituted systems with cytochrome P-450 from cholestyramine-treated rats than in those with cytochrome P-450 from untreated rats.     

Zwitterionic detergent mediated interaction of purified cytochrome P-450LM4 from 5,6-benzoflavone-treated rabbits with MADPH-cytochrome P-450 reductase

Hydroxylation of acetanilide catalyzed by purified cytochrome P-450LM4 and NADPH-cytochrome P-450 reductase was reconstituted with the zwitterionic detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS). The optimum rate of production of 4-hydroxyacetanilide was observed between 3 and 7 mM CHAPS and was about half that with 0.05 mM dilauroylglyceryl-3-phosphocholine (di-12-GPC). At higher detergent concentrations, hydroxylase activity decreased until at 15-20 mM CHAPS the system was inactive. The effect of CHAPS on the state of aggregation of P-450LM4 and on interaction between the cytochrome and P-450 reductase alone and under turnover conditions was investigated by ultracentrifugation. At 4 mM CHAPS, P-450LM4 was hexameric to heptameric (Mr 369,000). Neither reductase nor reductase plus acetanilide and NADPH altered the state of P-450LM4 aggregation, suggesting that a stable 1:1 P-450/reductase complex did not form under turnover conditions. Replacing CHAPS with 0.05 mM di-12-GPC resulted in formation of heterogeneous P-450 oligomers (Mr greater than 480,000). At CHAPS concentrations where substrate hydroxylation did not occur (15 and 22 mM), P-450LM4 was shown by sedimentation equilibrium measurements to be dimeric and monomeric, respectively. P-450 reductase was shown to reduce monomeric P-450LM4 in the presence of NADPH. Thus, the dependence of hydroxylase activity on [CHAPS] may be related to the state of aggregation of the cytochrome. An apparent correlation between P-450 aggregation state and NADPH-supported hydroxylation was also observed with phenobarbital-inducible P-450LM2 in the presence of detergents [Dean, W.L., & Gray, R.D. (1982) J. Biol. Chem. 257, 14679-14685; Wagner, S.L., Dean, W.L., & Gray, R.D. (1984) J. Biol. Chem. 259, 2390-2395].     

Ferredoxins from two sulfonylurea herbicide monooxygenase systems in Streptomyces griseolus.

We have purified and characterized two ferredoxins, designated Fd-1 and Fd-2, from the soluble protein fraction of sulfonylurea herbicide induced Streptomyces griseolus. These cells have previously been shown to contain two inducible cytochromes P-450, P-450SU1 (CYP105A1) and P-450SU2 (CYP105B1), responsible for herbicide metabolism [O'Keefe, D. P., Romesser, J. A., & Leto, K. J. (1988) Arch. Microbiol. 149, 406-412]. Although Fd-2 is more effective, either ferredoxin can restore sulfonylurea monooxygenase activity to an aerobic mixture of NADPH, spinach ferredoxin:NADP oxidoreductase, purified cytochrome P-450SU1, and herbicide substrate. The gene for Fd-1 is located in the genome just downstream of the gene for cytochrome P-450SU1; the gene for Fd-2 follows the gene for P-450SU2. The deduced amino acid sequences of the two ferredoxins show that, if monomeric, each has a molecular mass of approximately 7 kDa, and alignment of the two sequences demonstrates that they are approximately 52% positionally identical. The spectroscopic properties and iron and acid-labile sulfide contents of both ferredoxins suggest that, as isolated, each contains a single [3Fe-4S] cluster. The presence of only three cysteines in Fd-1 and comparisons with three [4Fe-4S] ferredoxins with high sequence similarity suggest that both Fd-1 and Fd-2 have an alanine in the position where these [4Fe-4S] proteins have a fourth cysteine ligand to the cluster. Transformation of Streptomyces lividans, a strain unable to metabolize sulfonylureas, with DNA encoding both P-450SU1 and Fd-1 results in cells capable of herbicide metabolism. S. lividans transformants encoding only cytochrome P-450SU1 do not metabolize herbicide.(ABSTRACT TRUNCATED AT 250 WORDS)     

Substrate-binding region of cytochrome P-450scc (P-450 XIA1). Identification and primary structure of the cholesterol binding region in cytochrome P-450scc

Cytochrome P-450_scc (P-450 XIA1) from bovine adrenocortical mitochondria was investigated using a suicide substrate: [¹⁴C]methoxychlor. [¹⁴C]Methoxychlor irreversibly abolished the activity of the side-chain cleavage enzyme for cholesterol (P-450_scc) and the inactivation was prevented in the presence of cholesterol. The binding of [¹⁴C]methoxychlor and cytochrome P-450_scc occurred in a molar ratio of 1:1 and the cholesterol-induced difference spectrum of cytochrome P-450_scc was similar with the methoxychlor-induced difference spectrum. [¹⁴C]Methoxychlor-binding peptides were purified from tryptic-digested cytochrome P-450_scc modified with [¹⁴C]methoxychlor. Determination of the sequence of the amino-acid residues of a [¹⁴C]methoxychlor-binding peptide allowed identification of the peptide comprising the amino-terminal amino-acid residues 8 to 28.     

Involvement of FMN and phenobarbital cytochrome P-450 in stimulating a one-electron reductive denitrosation of 1-(2-chloroethyl)-3-(cyclohexyl)-1-nitrosourea catalyzed by NADPH-cytochrome P-450 reductase

Purified hepatic NADPH-cytochrome P-450 reductase, which was reconstituted with dilauroylphosphatidylcholine, catalyzed a one-electron reductive denitrosation of 1-(2-[14C]-chloroethyl)-3-(cyclohexyl)-1-nitrosourea ([14C]CCNU) to give 1-(2-[14C]-chloroethyl)-3-(cyclohexyl)urea at the expense of NADPH. Ambient oxygen or anoxic conditions did not alter the rates of [14C]CCNU denitrosation catalyzed by NADPH-cytochrome P-450 reductase with NADPH. Electron equivalents for reduction could be supplied by NADPH or sodium dithionite. However, the turnover number with NADPH was slightly greater than with sodium dithionite. Enzymatic denitrosation with sodium dithionite or NADPH was observed in anaerobic incubation mixtures which contained NADPH-cytochrome P-450 reductase with or without cytochrome P-450 purified from livers of phenobarbital (PB)-treated rats; PB cytochrome P-450 alone did not support catalysis. PB cytochrome P-450 stimulated reductase activity at molar concentrations approximately equal to or less than NADPH-cytochrome P-450 reductase concentration, but PB cytochrome P-450 concentrations greater than NADPH-cytochrome P-450 reductase inhibited catalytic denitrosation. Cytochrome c, FMN, and riboflavin demonstrated different degrees of stimulation of NADPH-cytochrome P-450 reductase-dependent denitrosation. Of the flavins tested, FMN demonstrated greater stimulation than riboflavin and FAD had no observable effect. A 3-fold stimulation by FMN was not observed in the absence of NADPH-cytochrome P-450 reductase. These studies provided evidence which establish NADPH-cytochrome P-450 reductase rather than PB cytochrome P-450 as the enzyme in the hepatic endoplasmic reticulum responsible for CCNU reductive metabolism.