Oxidative and reductive metabolism by cytochrome

P450 2E1DENNIS R. KOOP

Department of Pharmacology, Oregon Health Sciences University, Portland, Oregon 97201, USA

724 0892-6638192/0006-724/$01 .50. © FASEB

ABSTRACT We are constantly exposed to many poten-tially toxic chemicals. Most require metabolic activationto species responsible for cell injury. Althoughcytochrome P450 2E1 is only one of many different formsof cytochrome P450 that catalyze these reactions, it hasan important role in human health as a result of beingreadily induced by acute and chronic alcohol ingestion.The enzyme efficiently catalyzes the low Km metabolismof compounds commonly used as solvents in industry andat home as well as components found in cigarette smoke,many of which are established carcinogens andhepatotoxins. As a result, there is the potential for in-creased risk to low level exposure to such chemicals whilecytochrome P450 2E1 is induced. Many substrates havebeen identified for cytochrome P450 2E1. Of the 52 sub-strates for the enzyme identified in this review, the de-methylation of N,N-dimethylnitrosamine and the hydrox-ylation of p-nitrophenol and chlorzoxazone are the mosteffective for monitoring the level of this enzyme. In addi-tion to oxidative reactions, cytochrome P450 2E1 is alsoan efficient catalyst of reductive reactions. CC14-inducedhepatotoxicity is one of the best-documented cases for theparticipation of cytochrome P450 2E1 in a toxicologicallyimportant reductive reaction. The reduction of oxygen tosuperoxide and peroxide are also important reductivereactions of the enzyme and could be important in lipidperoxidation. However, the role of this reaction in vivoremains controversial. - Koop, D. R. Oxidative andreductive metabolism by cytochrome P450 2E1. FASEBJ.6: 724-730; 1992.

Key Words: P450 2E1 ethanol-inducible . lipid peroxida-lion - hepatotoxicity

IT IS WELL DOCUMENTED THAT chronic ethanol ingestion cancause hepatotoxicity that predominates in the centrilobularregion of the liver (1-3). In addition to the toxic effects ofethanol itself, ethanol also potentiates the toxicity of otherchemicals including benzene, alkylnitrosamines, halogenatedalkanes, and acetaminophen (3-5). Acute ethanol intake caninhibit metabolism and thus actually protect against toxicity.

However, even low doses of ethanol will induce cytochromeP450 2EP (6) and when ethanol is cleared after acute treat-ment, potentiation of toxicity can be observed.

Most chemicals that cause hepatotoxicity must be bioacti-vated to reactive species that initiate cell damage (5). This re-

quirement for bioactivation, which occurs predominately viathe liver microsomal cytochrome P450-dependent mixed-function oxidase system, led to speculation that the induc-tion of an ethanol-inducible form of P450 (P450 2El) was in-volved in the bioactivation reactions. An ethanol-inducibleform of P450 was purified from rabbits (7) and has beencharacterized from many species, including rats, mice, ham-sters, and humans, and it was demonstrated that the purified

enzyme will catalyze bioactivation reactions (1, 3, 4). Addi-tional observations support the involvement of P450 2E1 inethanol-potentiated hepatotoxicity: other compounds thatinduce P450 2E1, such as acetone and other short-chain alco-hols, also potentiate the same hepatotoxicity (4, 5). P450 2E1is located in the cell layers near the terminal hepatic vein andis induced in the same region by all inducers examined (2).The centrilobular region of the liver is most susceptible tochemical toxins that are substrates for P450 2E1 (1-5).

The number of compounds that have been identified assubstrates for P450 2E1 has increased significantly since asimilar compilation was done in 1986 (4). The list includes

chemicals that are potential mechanism-based inhibitors forthe enzyme as well as those compounds that provide a selec-tive measure of P450 2E1 in hepatic microsomes, cell suspen-sions, and, potentially, in vivo. In addition to compoundsthat are substrates for oxidative metabolism by P450 2E1,substrates for reductive reactions will be reviewed and thepotential role of these pathways in toxicity will be discussed.

CRITERIA FOR THE PARTICIPATION OF P450 2E1

Multiple criteria are required to demonstrate a role for P4502E1 in the metabolism of a xenobiotic. The experimental ap-proach uses both direct and indirect experiments. Indirectexperiments involve demonstrating that the activity is indu-cible by compounds known to induce P450 2E1 such as ace-tone, ethanol, and isoniazid (4, 6). All these compounds areeasily administered in the drinking water and levels of induc-tion from two- to eightfold are often observed after treatmentfrom 1 to 7 days. As shown by intraperitoneal administra-tion, pyridine and pyrazole are also potent inducers of theenzyme (4, 8)’. A correlation between an increase in the ac-tivity being measured and treatment with acetone is notdefinitive as acetone also induces P450 2B1 in rats (9). Theactivity measured should correlate with the level of P450 2E1present in various microsomal preparations. The concentra-

tion of microsomal P450 2E1 should vary over as wide arange as possible and not cluster around a single value. Thisis readily accomplished in animal models, as the dose andtime of treatment can be manipulated, but it is more difficultwith samples of human liver. The absolute concentration ofP450 2E1 need not be established for these types of experi-ments. The relative concentration can be easily obtained byimmunoblot analysis of microsomal samples. With the in-creased sensitivity of staining procedures currently available,it is possible to obtain excellent immunoblot data on as littleas 0.05 tg of microsomal protein or 2 eg of total cellhomogenates (10). It is useful to also quantify other forms ofP450 in the same microsomal preparations to ensure that theactivity does not correlate with other P450 forms present.

‘Abbreviation: P450 2E1, cytochrome P450 2E1.

P450 2E1-DEPENDENT METABOLISM 725

The activity under investigation can be correlated with themetabolism of other substrates for which P450 2E1 has beenshown to be the principal catalyst. These include N,N-

dimethylnitrosamine demethylation at substrate concentra-tions equal to or below 1 mM (11), chlorzoxazone6-hydroxylation (12), and p-nitrophenol 2-hydroxylation (13).The activity used for comparison depends on a number offactors and will be discussed in more detail below.

Selective inhibitors are an effective means to establish theparticipation of P450 2E1 in substrate metabolism. This ap-proach is currently difficult to apply in that a specific inhibi-tor of P450 2E1 has not been described. As described below,some compounds appear to be mechanism-based inhibitorsof the enzyme but their specificity has not been completelyestablished. The use of reversible inhibitors is less effective,and in most cases the chemicals represent alternative sub-strates and/or ligands for other P450s as well. As the inhibi-tion depends on the binding constant to P450 2E1 and toother forms of P450, some selectivity may be observed, butis often difficult to interpret. Specific inhibitory antibody toP450 2E1 is an effective and direct approach to identify a roleof the enzyme in catalysis. This approach has been used ex-tensively with both polyclonal and monoclonal antibodies.

The most direct experimental approach to identify cataly-sis by P450 2E1 involves measuring the activity with thepurified enzyme in a reconstituted system containingNADPH cytochrome P450 oxidoreductase and cytochromeb5. The inclusion of cytochrome b5 is important as thishemoprotein can significantly alter the activity toward manysubstrates by affecting both Vm, and Km (11, 13, 14). Thsapproach can be misleading with some forms of P450 be-cause the purified enzyme is not always catalytically compe-tent. There are no reports of purified P450 2E1 from rats,rabbits, hamsters, or humans being inactive.

Contamination of a P450 2El preparation with a very smallamount of another form of P450 with high activity couldlead to the false identification of an activity being associatedwith P450 2E1. This can be avoided by the expression of theenzyme in either bacterial or mammalian cells where P4502E1 is the only P450 expressed (15). Under these conditions,if metabolism is observed, it can be attributed to the expres-sion of P450 2El. In these types of experiments there mustbe ample P450 oxidoreductase expressed. As a result of theimportant role that cytochrome b5 can have in P450 2E1function, the coexpression of this hemoprotein should also beconsidered. This is especially important as cytochrome b5can decrease the apparent Km, and thus specificity formetabolism can be established at low substrate concentra-tions.

OXIDATIVE SUBSTRATES FOR P450 2E1

Many compounds have been identified as substrates for P4502E1, as summarized in Table 1. The compounds are, ingeneral, all small, relatively polar compounds. The list in-

cludes many solvents used extensively in industry such asbenzene, chloroform, and trichloroethylene. As a result, thebioactivation of these types of protoxicants by P450 2E1places particular emphasis on this form of P450 in humanhealth. The enzyme is readily inducible in humans (15, 20,22) and there can be significant exposure to such substratesin the workplace. Chronic ethanol ingestion is not requiredfor induction of P450 2E1; significant increases in the en-zyme can be observed after a single dose of ethanol.

It is important to determine the catalytic activity of thehuman ortholog of P450 2E1, as in some cases significantdifferences in the activity of a structural ortholog have beenreported. It should be emphasized that with respect to P4502E1, all compounds that have been shown to be substrates inanimal models are also substrates for the human ortholog.Human P450 2E1 catalyzes the demethylation of AN-dimethylnitrosamine (15) and the activation of acetaminophen(20). In a more recent study, Guengerich et al. (22) describedthe metabolism of many low molecular weight cancer sus-pects by human P450 2E1. In this study, antibody to humanP450 2El was not as effective an inhibitor of the oxidationof most of the halogenated alkanes when compared with sub-strates such as chlorzoxazone (22). The metabolism of CC14was inhibited but the results presented were from a single ex-periment, and potential variation in the analysis cannot bedismissed. Whether the difference in inhibition by the anti-body suggests a different mode of metabolism by the enzymeremains to be determined. If the antibody inhibits the trans-fer of electrons from P450 oxidoreductase to the P450 or ac-cess to the active site, the antibody should be equally effec-tive with all substrates for the enzyme. The putativemechanism-based inhibitor diethyldithiocarbamate (22, 38)was not a very effective inhibitor of P450 2E1-dependentCC14 metabolism (which is reductive), but was very effectiveagainst the other halogenated alkanes tested (22). The resultssuggest that the inactivated enzyme may not catalyze oxygenactivation but can still effectively transfer a single electron toCd4.

The potential for common organic solvents to act as sub-strates for P450 2El has important implications when otherpotential substrates are tested in metabolic assays. Solventeffects can be quite dramatic as relatively low K1 values werereported (i.e., DMSO, 390 tiM; 2-mercaptoethanol, 20 fsM;

dimethylformamide, 90 sM; ethyl acetate, 220 tM) (39).Glycerol has a much higher K1 (53 mM) and Km (18 mM)(28), but this level of glycerol can be reached with the recon-stituted enzyme when 20% (v/v) glycerol (2.74 M) is used instorage buffers. The inhibition observed will depend greatlyon the stock concentration of the purified P450 and P450 ox-idoreductase. For example, if both enzymes are stored as 30iM stocks in 20% glycerol and are used at final concentra-tions of 0.1 and 0.3 aiM, the final glycerol concentration willbe 36 mM in a 1.0 ml reaction mixture.

The P450-dependent oxidation of glycerol to formalde-hyde (28) suggests other vicinal diols as potential substratesfor P450 2E1. These compounds include propanediol andbutanediol, the blood levels of which are increased by alcoholtreatment. Benzene is oxidized to muconic acid in vivo andtraits, trnns-munconaldehyde is thought to be an intermediatein this biotransformation (40). The pathway for muconalde-hyde formation has not been established. Latriano et al. (40)reported that this reactive dialdehyde was produced in vitrowith microsomes from benzene-pretreated mice. Benzenetreatment induces P450 2E1 (4), which is the principalcatalyst of the oxidation of benzene to phenol presumablyvia the intermediate benzeneoxide (18, 19). As a result, thetrans-dihydrodiol may be increased under these conditionsdepending on the relative concentration of epoxide hydro-lase. By a reaction analogous to the oxidation of glycerol, theformation of muconaldehyde from the dihydrodiol may bepossible (Fig. 1). In addition, a radical-mediated pathwayfor the formation of muconaldehyde was suggested (40). Asdiscussed below, P450 2E1 very efficiently supports hydroxylradical-mediated types of reactions (41, 42) and P450 2E1

TABLE 1. Substrates rnetabolized by cytoc/trorne P450 2E1

Substrate Product Measured Reference

Aromatic compoundsPyridine Pyridine N-oxide 163-Hydroxypyridine 2 ,5-Dihydroxypyridine 17

p-Nitrophenol 4-Nitrocatechol 13

Benzene Phenol 18

Phenol Hydroquinone; catechol 19Acetaminophen Glutathione conjugates 20Pyrazole 4-Hydroxypyrazole 21Chlorzoxazone 6-Hydroxychlorzoxazone 12Styrene Glutathione conjugate 22

Aniline p-Aminophenol 4

Halogenated alkanes and alkenes/alkanesChloroform Glutathione conjugate 22, 23Pentane product not measured 24

Chloromethane Formaldehyde 22Dibromoethane Glutathione conjugate 22Dichloromethane Glutathione conjugates 221,2-Dichloropropane Glutathione conjugate 22Ethyl carbamate 1,M-Ethenoadenosine 221,1,1 -Trichloroethane 1,1,1 -Trichloro-2-hydroxyethane 22Trichloroethylene Chloral 22Ethylene dibromide 1,M-Ethenoadenosine 22Ethylene dichloride 1 ,N6-Ethenoadenosine 22

Vinyl chloride 1 ,M-Ethenoadenosine 22

Vinyl bromide 1 ,M-Ethenoadenosine 22

Vinyl carbamate 1,N6-Ethenoadenosine 22

Enflurane Fluoride 25Halothane Trifluoroacetic acid 261,1,1 ,2-Tetrafluoroethane Fluoride 27

Alcohols/ketones/nitrilesEthanol Acetaldehyde 4Propanol Propionaldehyde 4Isopropanol Acetone 4

Butanol Butyraldehyde 4

Pentanol Valeraldehyde 4Glycerol Formaldehyde 28Acetol Methylglyoxal 4

Acetone Acetol 4Acetonitrile (+ catalase) Cyanide 29Acrylonitrile 1,M-Ethenoadenosine 22

Nitrosamines/azocompoundsN N-Dimethylnitrosamine Formaldehyde/nitrite 11Azoxymethane Azoxymethanol 30Methylazoxymethanol Methanol/formic acid 30N N-Diethylnitrosamine Acetaldehyde 11N-Nitrosopyrrolidine 4-Hydroxybutyraldehyde 31N-Nitroso-2,6-dimethylmorpholine N-Nitroso-(2-hydroxypropyl)-(2-oxopropyl)amine 4

EthersDiethyl ether Acetaldehyde 32

Methyl i-butyl ether Formaldehyde/t-butanol 33

Reductive substratesCarbon tetrachloride Lipid peroxidation/chloroform 22, 34Chromium (Cr”t) Product not measured 3513-Hydroperoxy-9,ll-octadecadienoic acid Pentane 3615-Hydroperoxy-5 ,8, 11,1 3-eicosatetraenic acid Pentane 36

Cumyl hydroperoxide Methane/acetophenone 36t-Butylhydroperoxide Methane/acetone 36Oxygen Superoxide/peroxide/water 37

726 Vol. 6 January 1992 The FASEB Journal KOOP

OH

‘OH

NO2

P4fl 7F1.DFPFNDENT METABOLISM 727

0

H-1.._H

P450 2E1? H H71C-H

NADPH, 02 II0

trans-i 2-Dihydrobenzene- 1 2-dial trans.trans-Uuconaldehyde

Figure 1. Proposed formation of muconaldehyde. A potential path-way for the formation of trans,trans-muconaldehyde from irans-1,2-

dihydrobenzene-1,2-diol via a P450 2E1-mediated reaction isshown.

may participate in this manner in muconaldehyde formationin lieu of a direct oxidative mechanism.

Of the many oxidatively metabolized substrates shown inTable 1, the most extensively investigated has been the P4502E1-dependent demethylation of N,N-dimethylnitrosamine(reviewed by Yang et al. in ref 11). This nitrosamine was oneof the first to be shown to be selectively metabolized by P4502E1 at low concentrations of substrate (11). As with ethanoloxidation (43), there does not appear to be any stereoselec-tivity (i.e., for the (Z)- vs. (E)-isomers) in the initial oxidationreaction, which suggests very little restriction on the sub-strate in the active site of the enzyme (44). Yang et al. (11)proposed an active site model based on the structure of

‘450cam that permits the oxidation of small molecules suchas those shown in Table 1. When P450 2E1 is presented witha nitrosamine such as N-methyl-N-butylnitrosamine, themethyl group is preferentially oxidized (as determined by therelease of formaldehyde instead of butyraldehyde) whereasother isozymes, such as P450 2B1, preferentially oxidize thebutyl group (11). P450 2E1 will oxidize the methyl group ofall nitrosamines that contain one methyl group and anothersubstituent such as propyl or benzyl, but the selectivity forcatalysis by P450 2E1 is decreased significantly (4, 11).

METABOLIC MARKERS FOR P450 2E1

It would be extremely desirable to have a specific assay forP450 2E1 that could be used to identify the presence of thefunctional enzyme as immunoblot analysis detects both apo-and holoenzyme. The demethylation of N,N-dimethylnitro-samine at a concentration of less than 1 mM and the hydrox-ylation of p-nitrophenol and chlorzoxazone are the bestmetabolic markers for the presence of P450 2E1. The struc-ture of each of these substrates and the position of hydroxyla-tion are shown in Fig. 2. The choice of which marker to usedepends on several factors. The measurement of p-nitrophenol hydroxylation is simple and rapid, and the onlyinstrumentation required is a spectrophotometer. The de-methylation of N-methylnitrosamines is also relatively sim-ple and the formaldehyde produced is measured colorimetri-cally. Both of these assays are limited by the sensitivity of themethods and can detect around 400 pmol of product. Thenitrosamines are established carcinogens. As a result, specialcare must be used when handling these compounds, androutine use is more difficult and waste disposal isproblematic. The sensitivity of these assays can be greatlyenhanced with radioactive substrates, but again, this makesroutine use more difficult and disposal is a problem. In con-trast, the 6-hydroxylation of chlorzoxazone is about 10-foldmore sensitive than the colorimetric assays. The substrate isan approved drug and not toxic at low doses. The6-hydroxylase assay requires high-pressure liquid chro-matography although a simple isocratic solvent system was

described (12). A major advantage of chlorzoxazonemetabolism is that the hydroxylation has the potential to beused as a noninvasive probe for the presence of P450 2El inanimal models and humans, provided that the in vivo rateof metabolism is dependent on the concentration of P4502El (i.e., the drug has low intrinsic clearance) under all con-ditions. Lucas et al. (45) recently concluded that of 12 differ-ent activities, the demethylation of N,N-dimethylnitrosaminewas one of the best indicators of P450 2El induction. Thehydroxylation of benzene and p-nitrophenol were also goodindicators of P450 2E1 but were not as specific (45).

MECHANISM-BASED INHIBITORS OF P450 2E1

Although microsomal studies can implicate P450 2El incatalysis in vitro, the availability of a specific inhibitor for theenzyme for the assessment of in vivo function would be ex-tremely valuable. Mechanism-based inhibitors offer thegreatest potential for specificity as catalysis is required for in-hibition (46). There has been some progress toward this endin the past few years. Four compounds shown in Fig. S havebeen suggested to be mechanism-based inhibitors for P4502E1.

3-Amino-1,2,4-triazole exhibited a number of kineticcriteria consistent with mechanism-based inhibition of P4502E1 (47). The inhibition was time- and NADPH-dependent.The first order loss of p-nitrophenol hydroxylase activity wassaturable, irreversible, and insensitive to exogenousnucleophiles, and had a pH dependence similar to otherreactions catalyzed by P450 2E1. The mechanism of the in-activation was not established. Heme was not destroyed asdetermined by the reduced pyridine hemochrome but car-bon monoxide binding was inhibited. Radiolabel from theC-S position of 3-amino-1,2,4-triazole was not incorporatedinto the protein. The compound is a well establishedmechanism-based inhibitor of catalase and the specificity forvarious P450 forms needs to be established (47).

Gannett Ct al. (48) described the irreversible inactivationof rat liver P450 2E1 by the natural product from red pep-pers, dihydrocapsaicin. NADPH-dependent covalent bind-ing of dihydrocapsaicin to P450 2E1 was demonstrated byimmunoaffinity purification of the inactivated enzyme, butthe stoichiometry was not established. Based on elec-trochemical studies, the authors suggested that P450 2E1catalyzed the 1-electron oxidation of dihydrocapsaicin to thephenoxy radical that reacted with P450 2E1 (48). BecauseP450 2E1 is responsible for the bioactivation of many car-cinogens, the inhibition of P450 2El by compounds such asdihydrocapsaicin may partially explain inhibition of tumori-genesis by some foods.

0

0H CH3-CH3

CI N

Chlorzoxazone N-Nitrosodimethylamine p-Nitroplienol

Figure 2. Substrate probes for P450 2El. The structures of sub-strates useful to monitor the presence of P450 2El are shown. Theposition of hydroxylation is indicated by the arrow for eachstructure.

NH2 (-CH2CH2N=C=S

N’ Phenethyl isothiocyanate

3-Amino- 1,2,4-triazole

CH3CH2M 0

CH3CH2- ii

728 Vol. 6 january 1992 The FASEBjournal KOOP

Diethyldithiocarbamate Dihydrocapsaicm

Figure 3. Mechanism-based inhibitors of P450 2E1. The structuresof putative mechanism-based inhibitors of P450 2E1 discussed inthe text are shown.

Ishizaki et al. (49) recently reported that phenethylisothiocyanate, another dietary constituent found in crucifer-ous vegetables such as cabbage and brussel sprouts, was aneffective inhibitor of the microsomal demethylation of !‘1N-dimethylnitrosamine. A K1 of 1 eM for the reversible inhibi-tion was reported. There was also a time- and NADPH-dependent inactivation of the demethylation reaction whichsuggests mechanism-based inhibition. There was asignificant time-dependent loss of activity in the absence ofNADPH, which is not surprising due to the reactivity of theisothiocyanate group. The specificity and mechanism of themechanism-based inactivation require further investigation.

Disulfiram inhibited the carcinogenicity of 1,2-dimethyl-hydrazine (50) and the hepatotoxicity of CHC13, CC!4,acetaminophen, and Ps N-dimethylnitrosamine (38).Disulfiram is reduced to diethyldithiocarbamate, which wasan effective inhibitor of P450 2E1 in rat (38) and human (22)microsomes. Brady et al. (38) reported a time- and NADPH-dependent loss of rat liver microsomal N,N-dimethylnitro-samine demethylation. However, the specificity and mechan-ism were not addressed although the authors suggested thepotential for an initial oxidation of diethyldithiocarbamateby the NADPH-dependent flavin containing monoox-ygenase in microsomes. Guengerich et al. (22) reported thepreliminary characterization of diethyldithiocarbamate as amechanism-based inhibitor of human P450 2E1. Diethyl-dithiocarbamate caused a time- and NADPH-dependentloss of both N,N-dimethylnitrosamine demethylation andchlorzoxazone hydroxylation. In competition assays using300 eM diethyldithiocarbamate, which inhibited about80-85% of the N,N-dimethylnitrosamine demethylation andchlorzoxazone hydroxylation, (± )-mephenytoin 4’-hydroxyla-tion was inhibited about 40% whereas tolbutamide methylhydroxylation and (±)-bufuralol 1-hydroxylation were in-hibited only about 10% (22). These results suggest somespecificity for the diethyldithiocarbamate inhibition. If thecompound were used in preincubation reactions at low con-centrations it might irreversibly inactivate all of the P4502E1 with little effect on other isoforms.

REDUCTIVE REACTIONS OF P450 2E1

The oxidative metabolism of organic substrates by P450 iswell recognized. In addition, P450 also catalyzes reductivereactions. In most cases these reactions are competitive withoxygen, and so are generally thought to occur only under

very low oxygen tensions. P450 2E1 is an effective catalyst ofreductive reactions, and those that have been clearly estab-lished for this enzyme are also shown in Table 1.

One of the best-documented cases of ethanol-potentiatedtoxicity after both acute and chronic ethanol ingestion isCC14-induced liver damage (1, 2, 5). Although many investi-gators have demonstrated that compounds that induce P4502E1 also potentiate CCI4-induced hepatic injury, Johanssonand Ingelman-Sundberg (34) directly demonstrated that thepurified rabbit ortholog of P450 2E1 was 100-fold more ac-tive than rabbit P450 1A2 and P450 2B4 in initiating CC!4-dependent lipid peroxidation. They also demonstrated thatunder anaerobic conditions, microsomes from imidazole-treated rabbits were about threefold more effective thanmicrosomes from untreated rabbits in the production ofCHC13 from CCI4 (34). Antibody to P450 2E1 inhibited theNADPH-dependent lipid peroxidation of hepatic micro-somes from acetone-treated rats (51) and from humans (52).Although inhibition of the NADPH-dependent lipid peroxi-dation could be attributed to a decrease in the level of reac-tive oxygen produced and not to an inhibition of CCI4 reduc-tion to the trichloromethyl radical, the formation of chloro-form from CC14 suggests P450 2E1-dependent metabolism ofCCJ4.

The mechanism of CCI4 reduction is not clear. Isoniazid(1 mM) inhibited 70% of the NADPH oxidation and perox-ide production by purified rat P450 2E1 (53). However, thesame concentration of isoniazid inhibited CC!4 metabolismto chloroform by only about 22% (53). In contrast, Lindroset al. (54) reported that 2 mM isoniazid inhibited CC14induced hepatocyte damage in perivenular cells by about80%, a value consistent with the inhibition of microsomallipid peroxidation by isoniazid (53). The formation of trich-loromethyl radicals from CC!4 enhanced by ethanol treat-ment was reported by Reinke et al. (55). The effect ofcytochrome b5 on the reductive reactions has not been care-fully examined. It is important to understand the role ofcytochrome b5 in the reductive reactions because it has beenproposed that this cytochrome can function as a source of thesecond electron in the catalytic cycle of P450 (56). Ifcytochrome b5 efficiently donates an electron to the ferrous-oxygen enzyme, then one might predict that the release ofsuperoxide by autooxidation would decrease. The ferricperoxide intermediate formed would rapidly cleave to givewater and an active oxygen intermediate that would bereduced by two electrons to give a second molecule of water(37, 56). If the peroxide is produced by dismutation of su-peroxide, then cytochrome b5 would decrease peroxideproduction. Gorsky et a!. (37) reported that P450 2E1 ex-hibited the greatest 4-electron oxidase activity of six purifiedrabbit forms of P450.

Vaz et a!. (36) described the efficient reduction ofhydroperoxides by P450 2E1. These reactions were con-ducted under anaerobic conditions, and produced hydrocar-bons and either aldehydes or ketones from the originalhydroperoxide. Lipid hydroperoxides were proposed as thephysiological substrates, and the nearly stoichiometric for-mation of pentane and NADPH oxidation for the reactionwith 13-hydroperoxy-9,11-octadecadienoic acid was reported(36). Other organic hydroperoxides such as cumylhydroperoxide and tertiary butyl hydroperoxide were alsoeffective substrates. The physiological significance of thisreaction has not been demonstrated, but P450 2E1 appearsquite capable of initiating lipid peroxidation (51, 57). Thus,although initially involved in the formation of lipidhydroperoxides, P450 2E1 may continue to function in thereduction of hydroperoxides as the oxygen concentration is

PAcn )t1flCPCNifl1N.JT tAITARflI cM 729

depleted in the centrilobular region of the liver. This couldcontribute to the degeneration of the lipid bilayer. Tereliusand Ingelman-Sundberg (24) demonstrated that pentane isan effective substrate for P450 2E1 with an apparent Km of

about 9 tIM. Thus, as pentane is formed, P450 2E1 will beable to oxidize the compound if the oxygen tension is in-creased. Whereas the oxidation of other alkanes has not beenreported, it is likely they would also be substrates for P450

2El.Also included in the list of substrates for reductive reac-

tions is molecular oxygen. The NADPH-dependent reduc-tion of oxygen by cytochrome P4SO in the presence and ab-sence of substrates to peroxide is well documented for allforms of the enzyme. This reduction reaction - referred to asthe oxidase activity of P450-is probably as important, if notmore so, than the other reductive reactions listed in Table 1.Peroxide formation is thought to occur by the release andsubsequent dismutation of superoxide, but the direct releaseof peroxide from the 2-electron reduced enzyme is possible(37, 56).

The reduction of oxygen may be especially important forP4SO 2E1. This form of P450 is isolated in the high spin state(7) and if the enzyme is also high spin in the endoplasmic

reticulum, it may be readily reduced by NADPH

cytochrome P450 oxidoreductase in the absence of substrate.Compared with other forms of P450, P450 2E1 exhibits ahigher rate of oxidase activity when purified (37), and micro-somes from animals pretreated with inducers of P450 2E1also exhibit much greater rates of NADPH oxidation thanmicrosomes from untreated animals (41, 42, 51, 57). En-hanced oxidase activity would result in the increased produc-tion of both superoxide and hydrogen peroxide, which in thepresence of chelated iron can produce reactive hydroxyl radi-cals (41, 42). P450 2E1 exhibits a unique ability to potentiateiron-catalyzed Fenton chemistry in a reconstituted system,and increased rates of hydroxyl-radical mediated metabolismof substrates such as ethanol and dimethylsulfoxide were

reported for microsomes isolated from animals pretreated toinduce P450 2E1 (41, 42, 51, 53, 57). The increased capacity

to produce active oxygen species is manifested by an in-creased rate of microsomal lipid peroxidation by microsomesor liposomes enriched in P450 2E1 (51, 53, 57). Antibody toP450 2E1 inhibited about 65% of microsomal peroxideproduction while almost completely inhibiting NADPH-dependent lipid peroxidation (51). These results suggest that

although there are other microsomal sources of peroxide, theparticipation of P450 2El is necessary for lipid peroxidationto be observed. -

However, the role of the oxidase reaction in vivo remainscontroversial. There is little experimental data to directly testthe hypothesis that the peroxide produced in the cytosol isgenerated from the reduction of oxygen to superoxide and/orperoxide by cytochrome P450. Krieter et al. (58) monitoredthe biliary efflux of oxidized glutathione (OSSO) as a meas-ure of peroxide formation in the presence and absence ofaminopyrine, a substrate for some forms of P450 but not forP450 2E1. The efflux of GSSG was very low in the absenceof substrate. In the presence of aminopyrine, the OSSGreleased was not dependent on glutathione peroxidase andperoxide as a selenium-deficient diet had no effect on GSSGrelease (58). In these experiments, the results suggested that

the P450 system produced no peroxide in the absence of ahydroxylatable substrate, and the peroxide that wasproduced in the presence of aminopyrine was also not at-tributable to peroxide release. Studies in vitro indicate thatthe extent of uncoupling is dependent on the P450 form (37).Although the results of Krieter et a!. (58) suggest that the

P450 population present in untreated and phenobarbital-treated rat liver does not produce peroxide, this may not bethe case when P450 2El is induced.

It has been suggested that the induction of P450 2E1 in thecentrilobular region of the liver may participate in the accen-tuation of the oxygen gradient between the perivenous andcentrilobular regions of the liver after ethanol treatment (1,2). If the presence of P450 2E1 is the only reason for thislower oxygen tension, then induction of P450 2E1 by othertreatments and also with single nontoxic doses of ethanolshould also increase hypoxia in the centrilobular region. It

would be interesting to determine whether there is enhancedOSSO efflux into the bile under conditions of optimal P4502El induction. The hypothesis that increased oxidase ac-tivity of P450 2El in the centrilobular region results in in-creased oxygen consumption while also increasing oxygenradical production that participates in lipid peroxidation isattractive. This would lead to the regioselective hepatotoxic-ity observed after ethanol treatment. However, it remains tobe demonstrated that the enzyme is uncoupled in vivo.

Work in the author’s laboratory was supported by USPHS grantAA-08608 from the National Institute on Alcohol Abuse and Alco-holism.

REFERENCES

I. Lieber, C. S. (1984) Alcohol and the liver:1984 update. Hepalology 4,1243-1260

2. Lieber, C. S. (1990) Mechanism of ethanol induced hepatic injury.Phar-macol. Ther. 46, 1-41

3. Garro, A. J., and Lieber, C. S. (1990) Alcohol and cancer. Annu. Rev.Pharmacol. Toxicol. 30, 219-249

4. Koop, D. R., and Coon, M. J. (1986) Ethanol oxidation and toxicity:

role of alcohol P-450 oxygenase. Alcoholism: Gun. Exp. Res. 10, 44s-49s

5. Zimmerman, H. J. (1986) Effectsof alcohol on other hepatotoxins. Alco-holism: Gun. Exp. Res. 10, 3-15

6. Koop, D. R., and Tierney, D. J. (1990) Multiple mechanisms in the

regulation of ethanol-inducible cytochrome P45011E1. BioEssays 12,429-435

7. Koop, D. R., Morgan, E. T., Tarr, C. E., and Coon, M. J. (1982)Purification and characterization of a unique isozyme of cytochrome

P-450 from livermicrosomes of ethanol-treated rabbits.J. Biol. Chem.

257, 8472-84808. Kaul, K. L.,and Novak, R. F.(1987) Inhibition and induction of rabbit

livermicrosomal cytochrome P-450 by pyridine.J. Pharmacol. Exp. Ther.43, 384-390

9. Johansson, I.,Ekstrom, G., Scholte, B., Puzycki, D., Jornvall, H., and

Ingelman-Sundberg, M. (1988) Ethanol-, fasting-, and acetone-

inducible cytochromes P-450 in rat liver:regulation and characteristics

of enzymes belonging to the JIB and lIE gene subfamilies. Biochemistry

27, 1925-1934

10. Koop, D. R., Chernosky, A., and Brass, E. P. (1991) Identificationandinduction of P450 2E1 in rat Kupifer cells.J. Pharma.col. Exp. Ther. 258,1072-1076

11. Yang, C. S., Yoo, J. -S. I-i.,Ishizaki, H., and Hong, J. (1990)

Cytochrome P45011E1: roles in nitrosamine metabolism and mechan-

isms of regulation. Drug Metab. Rev. 22, 147-159

12. Peter, R., Bocker, R., Beaune, P. H., Iwasaki, M., Guengerich, F. P.,

and Yang, C. 5. (1990) Hydroxylation of chlorzoxazone as a specificprobe for human liver cytochrome P-450lIEl. Chem. Res. Toxicol. 3,566-573

13. Koop, D. R. (1986) Hydroxylation of p-nitrophenol by rabbit ethanol-

inducible cytochrome P-450 isozyme 3a. Mol. P/,arrnacol. 29, 399-40414. Patten, C. J., Ning, S. M., Lu, A. Y. H., and Yang, C. 5. (1986)

Acetone-inducible cytochrome P-450: purification, catalytic activity,

and interaction with cytochrome b5. Arch. Biochem. Biophys. 251, 629-63815. Umeno, M., McBride, 0. W., Yang, C. S.,Gelboin, H. V., and Gonza-

lez, F. J. (1988) Human ethanol-inducible P450I1E1: complete gene se-

quence, promoter, characterization,chromosome mapping, and cDNA-

directed expression. Biochemistry 27, 9006-9013

16. Kim, S. G., Williams, D. E.,Schuetz, E. G., Guzelian, P. 5.,and Novak,R. F. (1988) Pyridine induction of cytochrome P-450 in the rat:role of

P-450j (alcohol-inducible form) in pyridine N-oxidation. J. Pharmacol.Exp. Ther. 246, 1175-1182

17. Kim, S. C., and Novak, R. F. (1990) Role of P45011E1 in the

710 Vol 1 bn,iirs, lqQ7 Th lActt (n,,rr,I

metabolism of 3-hydroxypyridine, a constituent of tobacco smoke: redoxcycling and DNA strand scission by the metabolite 2,5-dihydroxypyridine.Cancer Res. 50, 5333-5339

18. Johansson, I., and Ingelman-Sundberg, M. (1988) Benzene metabolismby ethanol-, acetone-, and benzene-inducible cytochrome P-450 (IlEl)in rat and rabbit liver microsomes. Cancer Res. 48, 5387-5390

19. Koop, D. R., Laethem, C. L., and Schnier, G. G. (1989) Identificationof ethanol-inducible P450 isozyme 3a (P45OIIE1) as a benzene andphenol hydroxylase. Toxicol. Appi. Pharmacol. 98, 2 78-288

20. Raucy,J., Fernandes, P., Black, M., Yang, S. L., and Koop, D. R. (1987)Identification of a human liver cytochrome P-450 exhibiting catalyticand immunochemical similarities to cytochrome P-450 3a of rabbitliver.Biochem. P/zarmacol. 36, 2921-2926

21. Clejan, L. A., and Cederbaum, A. I. (1990) Oxidation of pyrazole byreconstituted systems containing cytochrome P-450 IIEJ. Biochirn. Bio-phys. Acta 1034, 233-237

22. Guengerich, F. P., Kim, D. -H., and Iwasaki, M. (1991)Role of humancytochrome P-450 IlEl in the oxidation of many low molecular weightcancer suspects. C/tern. Res. Toxicol. 4, 168-179

23. Brady, J. F, Li, D., Ishizaki, H., Lee, M., Ning, S. M., Xiao, F., andYang, C. S. (1989) Induction of cytochromes P45011E1 and P45011B1 bysecondary ketones and the role of P45OIIEI in chloroform metabolism.Toxicol. AppI. P/zarmacol. 100, 342-349

24. Terelius, Y., and Ingelman.Sundberg, M. (1986) Metabolism of n-pentane by ethanol-inducible cytochrome P-450 in liver microsomesand reconstituted membranes. Eur. j Biochem. 161, 303-308

25. Hoffman, J., Konopka, K., Buckhorn, C., Koop, D. R., and Waskell,L. (1989) Ethanol-inducible cytochrome P450 in rabbits metabolizesenflurane. Br j Anaesth. 63, 103-108

26. Gruenke, L. D., Konopka, K., Koop, D. R., and Waskell, L. (1988)Characterization of halothane oxidation by hepatic rnicrosomes andpurified cytochrome P-450 using gas chromatographic mass spectro-metric assay. J. Pharmacol. Exp. Tiler 246, 454-459

27. Olson, M.J., Kim, S. G., Reidy, C. A., Johnson, J. T., and Novak, R. F.(1990) Oxidation of 1,l,l,2-tetrafluoroethane in rat liver microsomes is

catalyzed primarily by cytochrome P-4501IE1. Drug Metab. Dispos. 19,298-303

28. Winters, D. K., Clejan, L. A., and Cederbaum, A. 1. (1988) Oxidationof glycerol to formaldehyde by rat liver microsomes. Biochern. Biophys.Res. Commun. 153, 612-617

29. Feierman, D. E., and Cederbaum, A. L. (1989) Role of cytochromeP-450 IIEI and catalase in the oxidation of acetonitrile to cyanide. C/tern.Res. 7?,xicol. 2, 359-366

30. Sohn, 0. S., Ishizaki, H., Yang, C. S., and Fiala, E. S. (1991) Metabolismof azoxymethane, methylazoxymethanol and N-nitrosodimethylamineby cytochrome P45011E1. Carcinogenesis 12, 127-131

31. McCoy, G. D., and Koop, D. R. (1988) Reconstitution of rabbit livermicrosomal N-nitrosopyrrolidine a-hydroxylase activity. Cancer Res. 48,398 7-3992

32. Brady, J. F., Lee, M. J., Li, M., Ishizaki, H., and Yang, C. S. (1987)Diethyl ether as a substrate for acetone/ethanol-inducible cytochromeP-450 and as an inducer for cytochrome(s) P-450. Mol. Pharrnacol. 33,148-154

33. Brady, J. F., Xiao, F., Ning, S. M., and Yang, C. S. (1990) Metabolismof methyl tertiary-butyl etherby rat hepatic microsomes. Arch, Toxicol. 64,157- 160

34. Johansson, I., and Ingelman-Sundberg, M. (1985) Carbon tetrachloride-induced lipid peroxidation dependent on an ethanol-inducible form ofrabbit liver microsomal cytochrome P-450. FEBS Let!. 183, 265-269

35. Mikalsen, A., Alexander, J., Wallin, H., Ingelman-Sundberg, M., and

Andersen, R. A. (1991) Reductive metabolism and protein binding ofchromium (VI) by P450 protein enzymes. Carcinogenesis 12, 825-831

36. Vaz, A. D. N., Roberts, E. S., and Coon, M. J. (1990) Reductive jI-scission of the hydroperoxides of fatty acids and xenobiotics: role ofalcohol-inducible cytochrome P-450. Proc. NatI. Acad. Sci. USA 87,5499-5503

37. Gorsky, L. D., Koop, D. R., and Coon, M. J. (1984) On the stoichiome-

try of the oxidase and monooxygenase reactions catalyzed by livermicrosomal cytochrome P-450.j Biol. C/tern. 259, 6812-6817

38. Brady, J. F., Xiao, F., Wang, M.-H., Li, Y., Ning, S. M., Gapac, J. M.,and Yang, C. S. (1991) Effects of disulfiram on hepatic P45OIIE1, othermicrosomal enzymes, and hepatotoxicity in rats. Toxicol. Appl. Pharmacol.

108, 366-37339. Yoo, J. -S. H., Cheung, R. J., Patten, C. J., Wade, D., and Yang, C. S.

(1987) Nature of N-nitrosodimethylamine demethylase and its inhibi-tors. Cancer Res. 47, 3378-3383

40. Latriano, L., Goldstein, B. D., and Wjtz, G. (1986) Formation ofmuconaldehyde, an open-ring metabolite of benzene, in mouse livermicrosomes: an additional pathway for toxic metabolites. Proc. Nati.Acad. Sci. USA 83, 8356-8360

41. Dicker, E., and Cederbaum, A. I. (1987) Hydroxyl radical generationby microsomes after chronic ethanol consumption. Alcoholism: Clin. Exp.Res. 11, 309-314

42. Feierman, D. E., Winston, G. W., and Cederbaum, A. I. (1985) Ethanoloxidation by hydroxyl radicals: role of iron chelates, superoxide, andhydrogen peroxide. Alcoholism: Cue. Exp. Res. 9, 95-102

43. Ekstrom, C., Norsten, C., Cronholm, T., and Ingelman-Sundberg, M.(1987) Cytochrome P-450 dependent ethanol oxidation. Kinetic isotopeeffects and absence of stereoselectivity. Biochemistry 26, 7348-7354

44. Keefer, L. K., Kroeger-Koepke, M. B., Ishizaki, H., Michejda, C. J.,Saavedra, J. E., Hrabie, J. A., Yang, C. S., and Roller, P. P. (1990)Stereoselectivity in the microsomal conversion of N-nitrosodimethylamineto formaldehyde. Chem. Res. Toxicol. 3, 540-544

45. Lucas, D., Berthou, F., Dreano, Y., Floch, H. H., and Menez, J. F.(1990) Ethanol-inducible cytochrome P-450: assessment of substrates’specific chemical probes in rat liver microsomes. Alcoholism: Clin. Exp.Res. 14, 590-594

46. Rando, R. R. (1984) Mechanism-based enzyme inactivators. P/zarma#{128}ol.

Rev. 36, 111-14247. Koop, D. R. (1990) Inhibition of ethanol-inducible cytochrome

P45011E1 by 3-amino-l,2,4-triazole. C/tern. Res. Toxicol. 3, 377-383

48. Gannett, P. M., Iversen, P., and Lawson, T. (1990) The mechanism ofinhibition of cytochrome P45OIIEI by dihydrocapsaicin. Bioorg. C/tern.

18, 185-19849. lshizaki, H., Brady, J. F., Ning, S. M., and Yang, C. S. (1990) Effect of

phenethyl isothiocyanate on microsomal N-nitrosodimethylaminemetabolism and other monooxygenase activities. Xenobiotica 20, 255-264

50. Fiala, E. 5. (1981) Inhibition of carcinogen metabolism and action bydisulfiram, pyrazole, and related compounds. In Inhibition of Tumor In-duction and Development (Zadeck, M. S., and Lipkin, M., eds) pp. 23-69,Plenum, New York

51. Ekstrom, G., and Ingelman-Sundberg, M. (1989) Rat livermicrosomal

NADPH-supported oxidase activity and lipid peroxidation dependenton ethanol-inducible cytochrome P-450 (P-450lIE1). Bloc/tern. Pharmacol.38, 1313-1319

52. Ekstrom, G., von Bahr, C., and Ingelman-Sundberg, M. (1989) Humanliver microsomal cytochrome P.450IIEl. Immunological evaluation ofits contribution to microsomal ethanol oxidation, carbon tetrachloridereduction and NADPH oxidase activity. Biochern. P/ta rmacol. 38,689-693

53. Persson, J. 0., Terelius, Y., and Ingelman-Sundberg, M. (1990)Cytochrome P.450.dependent formation of reactive oxygen radicals:isozyme-specific inhibition of P-450-mediated reduction of oxygen andcarbon tetrachloride. Xenobio!ica 20, 887-900

54. Lindros, K. 0., Cai, Y., and Penttila, K. E. (1990) Role of ethanol-inducible cytochrome P.450 IIEI in carbon tetrachloride-induceddamage to centrilobular hepatocytes from ethanol-treated rats. He/rntol-ogy 12, 1092-1097

55. Reinke, L. A., Lai, E. K., and McCay, P. B. (1988) Ethanol feedingstimulates trichloromethyl radical formation from carbon tetrachloridein liver. Xenobioiica 18, 1311-1318

56. Guengerich, F. P. (1991)Reactions and significance of cytochrome P-450enzymes. j Biol. C/tern. 266, 10019-10022

57. Krikun, C., and Cederbaum, A. I. (1986) Effectof chronic ethanol con-sumption on microsomal lipid peroxidation: role of iron and compari-son between controls. FEBS Let!. 208, 292-296

58. Krieter, P. A., Ziegler, D. M., Hill, K. E., and Burk, R. F. (1985)Studies on the biliary efflux of GSH from rat liver due to themetabolism of aminopyrine. Biochern. Pharrnacol. 34, 955-960