Role of Cytochrome P450 in the Metabolism and Toxicity of Hydroperoxides in Isolated Rat Hepatocytes

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 323, No. 2, November 10, pp. 463–470, 1995

Role of Cytochrome P450 in the Metabolism and Toxicity of Hydroperoxides in Isolated Rat Hepatocytes1 John A. Thompson,2 Kathleen M. Schullek, Sherri B. Turnipseed,3 and David Ross Department of Pharmaceutical Sciences, Box C238, University of Colorado Health Sciences Center, Denver, Colorado 80262

Received July 14, 1995

The contributions of cytochromes P450 (P450) to the metabolism and toxicity of hydroperoxides in freshly isolated rat hepatocytes were investigated utilizing 2,6 - di - tert - butyl - 4 - hydroperoxy - 4 - methyl - 2,5 - cyclohexadienone (BHTOOH). This hydroperoxide was rapidly degraded in cell suspensions, and the products were identical to those determined previously with subcellular preparations of ferric P450. With 250 mM BHTOOH, the ratio of glutathione peroxidase-mediated:P450-mediated metabolism was estimated to be about 3:1. Surprisingly, BHTOOH was found to be a more potent cytotoxin than cumyl hydroperoxide (CuOOH), despite the fact that it caused substantially less lipid peroxidation than the latter. P450 inhibition enhanced the toxicity of BHTOOH, but lowered the toxicity of CuOOH. These data demonstrate that intracellular ferric P450 can compete with glutathione peroxidase to reduce hydroperoxides by 1- and 2-electron processes. If the alkoxy radical from homolytic cleavage of the O–O bond can undergo facile intramolecular reactions to nontoxic products, as with BHTOOH, the role of P450 is detoxification. On the other hand, if the alkoxy radical preferentially attacks membrane lipids, as with CuOOH, P450 contributes to lipid peroxidation and toxicity. It was determined that the levels of glutathione, protein thiols, and ATP decreased in parallel with BHTOOH-induced cell death, but no conclusions are possible concerning mechanisms underlying the relatively potent toxicity of BHTOOH. Toxicity may be related to the high lipophilicity of this hydroperoxide which, presumably, facilitates its passage into cells and distribution to various intracellular sites. BHTOOH appears to be an excellent model compound for investigating mechanisms of hydroperoxide-medi1

This work was supported by National Institutes of Health Grant CA41248. 2 To whom correspondence and reprint requests should be addressed. Fax: (303) 270-6281. 3 Present address: Animal Drugs Research Center, Food and Drug Administration, Denver Federal Center, Denver, CO 80225-0087.

ated cytotoxicity which do not involve lipid peroxidation. q 1995 Academic Press, Inc. Key Words: cytochrome P450; hydroperoxides; hepatocytes; glutathione; lipid peroxidation; cytotoxicity.

Mechanisms of oxidative stress-related cell death have been investigated with model systems consisting of isolated or cultured rat hepatocytes and free radicalgenerating compounds such as quinones (1) and hydroperoxides (2). Numerous biochemical perturbations occur, including oxidation of GSH4 and NADPH (3), lipid peroxidation (3, 4), mitochondrial damage (5), protein S-thiolation (2, 6), and disruption of calcium homeostasis (7). Cell death has been attributed mainly to lipid peroxidation and mitochondrial damage (8), but due to numerous interrelated biochemical changes accompanying hydroperoxide exposure, it has been difficult to discriminate between the events which fatally injure cells and those which occur merely as a consequence of irreversible damage (8–11 and references therein). Despite the fact that Gpx protects cells by catalyzing the 2-electron reductions of hydroperoxides to alcohols, intracellular nonheme, nonferritin iron contributes to toxicity through 1-electron reductions of hydroperoxides resulting in alkoxy radical formation (3, 5, 12, 13). Subcellular preparations of P450 readily reduce hydroperoxides (14–16). Despite the fact that these are the most abundant hemeproteins in the hepatocyte capable of reducing organic hydroperoxides, very little is known about the role of P450 in the metabolism and toxicity of these compounds in intact cells. Many of the studies on hydroperoxide-induced toxicity have been 4

Abbreviations used: P450, cytochromes P450; Gpx, glutathione peroxidases; BHTOOH, 2,6-di-tert-butyl-4-methyl-4-hydroperoxy2,5-cyclohexadienone; CuOOH, cumyl hydroperoxide; tBuOOH, tertbutyl hydroperoxide; GSH, reduced glutathione; GSSG, oxidized glutathione; ABT, 1-aminobenzotriazole; DPPD, diphenylphenylene diamine; MDA, malondialdehyde. 463

0003-9861/95 $12.00 Copyright q 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.

/ m4345$9173

10-12-95 10:27:51

arca

AP: Archives

464

THOMPSON ET AL.

conducted with cultured rat hepatocytes (4, 5, 17), but it is unlikely that P450 contributes significantly to hydroperoxide reduction in cultured cells due to rapid decreases in P450 levels unless the cells are maintained under special conditions (18, 19). It has been demonstrated that the ferric form of P450 catalyzes a Fenton-type reaction involving 1-electron reductions of O–O bonds (20). As shown in Eq. [1] with PFe representing the P450 heme, this process generates an alkoxy radical and returns the heme to its ferric oxidation state. The ferric form of P450 catalyzes both heterolytic and homolytic processes represented by Eqs. [2] and [3], respectively (16, 21, 22). The terminal oxygen of the peroxy group is transferred to iron producing ferryl–oxo complexes analogous to Compound I (Eq. [2]) or Compound II (Eq. [3]) of peroxidases (14). The relative contributions of heterolysis and homolysis may be influenced by the P450 apoprotein and the structure of the hydroperoxide (23). ROOH / PFeII / H/ r ROr / PFeIII / H2O [1] ROOH / PFeIII r ROH / Pr/ FeIVO III

ROOH / PFe

r ROr / PFe OH IV

[2] [3]

The potential exists for P450s to contribute to the cytotoxicity of hydroperoxides through the production of alkoxy radicals. This possibility is supported by the finding that metyrapone, a P450 inhibitor, suppresses lipid peroxidation and delays cell death in isolated rat hepatocytes treated with tBuOOH (2). In the present report, the role of intracellular P450 in hydroperoxide reduction was investigated by incubating freshly isolated rat hepatocytes with BHTOOH, a para-peroxyquinol derived from butylated hydroxytoluene (21). This and related peroxyquinols are useful probes of hemeprotein-catalyzed O–O bond cleavage mechanisms because the resulting products are reliable indicators of the pathways involved in their formation (15, 23, 24). As summarized in Fig. 1, heterolysis by ferric P450 generates BHTOH (reaction a), some of which is hydroxylated to BHT(OH)2 by the Compound I form of P450 (reaction b) before migrating away from the heme (21). Homolysis (reaction c) produces BHTOr which readily rearranges via a putative carbon-centered radical intermediate (reaction e) to form the products RP5 and RP7. Only very small amounts of DBBQ are produced, demonstrating that rearrangement is substantially more facile than b-scission (reaction d) (15). Rearrangement of BHTOr also dominates over intermolecular reactions with lipids, as very little peroxidation occurred when BHTOOH was incubated with liver microsomes (25). This result contrasts with the extensive peroxidation of microsomal lipids generated by tBuOOH and CuOOH, as expected for alkoxy radi-

/ m4345$9173

10-12-95 10:27:51

arca

cals which cannot undergo facile rearrangements. Another important advantage of BHTOOH as a probe for P450 activity is the fact that this peroxy compound does not destroy the P450 heme nearly as extensively as the other two hydroperoxides (26). The present studies demonstrate that P450 competes with Gpx for hydroperoxide reduction in isolated rat hepatocytes, and that the P450 reactions occur predominantly with the ferric form of the heme. It was determined also that P450-dependent metabolism contributes to the detoxification of BHTOOH but enhances the toxicity of CuOOH. These and other results demonstrate significant differences in the mechanisms by which peroxyquinols and alkyl hydroperoxides affect isolated hepatocytes. MATERIALS AND METHODS Materials. Bovine serum albumin (fraction V), 5,5*-dithio-bis-(2nitrobenzoic acid), NADPH, 2-thiobarbituric acid, metyrapone, and Hepes were obtained from Sigma (St. Louis, MO), and aminobenzotriazole was provided by Dr. P. R. Ortiz de Montellano, University of California (San Francisco, CA). Collagenase and an ATP bioluminescence kit were obtained from Boehringer Mannheim (Indianapolis, IN). GSH, GSSG, DBBQ, 2,4-di-tert-butylphenol, CuOOH, and DPPD were purchased from Aldrich (Milwaukee, WI). The syntheses of BHTOOH, [18O2]BHTOOH, and BHTOH were described previously (15, 21). Hepatocyte isolation and incubations. Cells were prepared from the livers of 180–250 g male Sprague–Dawley rats (Sasco, Omaha, NE) by collagenase perfusion as described (27). Incubations were performed at 377C using 106 cells/ml in Krebs–Henseleit buffer supplemented with 12 mM Hepes, pH 7.4. Cell viability at the beginning of each experiment was at least 85% as determined by the exclusion of trypan blue (27). Hydroperoxides and BHTOH were added to the incubations as solutions in dimethyl sulfoxide (10 ml/ml of incubate). Control incubations contained the vehicle only. Cytotoxicities produced in hepatocyte incubations with hydroperoxides added in methanolic solution were indistinguishable from those utilizing dimethyl sulfoxide as the vehicle. Analytical procedures. For analysis of BHTOOH metabolites, 1.0ml aliquots of the incubation mixture were added to cold, distilled diethyl ether (2.5 ml) containing 22 mg of the internal standard 2,4di-tert-butylphenol. Compounds were extracted into the organic phase, and analyzed with a Beckman Model 332 HPLC system as described (21) using a 4.6 1 250-mm Ultrasphere ODS column (Beckman, Fullerton, CA) coupled to a Hewlett Packard (Palo Alto, CA) 1040M photodiode array detector. Labeled products were isolated with a semipreparative (10 1 250 mm) Beckman Ultrasphere ODS column. The isotopic composition of BHT(OH)2 , as its trimethylsilyl derivative, and RP5 was determined with a Hewlett Packard 5988 gas chromatograph–mass spectrometer operated in the selected ion mode as described (21). Lipid peroxidation was estimated by determining the formation of malondialdehyde with thiobarbituric acid, and measuring the absorbance of the complex at 535 nm (28). Protein thiols were measured using acid precipitation and 5,5*-dithio-bis-(2nitrobenzoic acid) according to DiMonte et al. (29). Data were expressed as nmol thiols/mg protein. Protein was determined using the BCA assay reagent (30) from Pierce (Rockford, IL), and GSH and GSSG concentrations were measured by the HPLC method of Reed et al. (31). NADPH was measured using a modification of published methods (32, 33), and involved adding 100 ml of 0.5 M potassium hydroxide containing 50% (v/v) ethanol and 35% (w/v) cesium chloride to 1.0-ml cell suspensions. These samples were immediately

AP: Archives

METABOLISM AND TOXICITY OF HYDROPEROXIDES IN HEPATOCYTES

465

FIG. 1. Scheme for the cytochrome P450-catalyzed reduction of BHTOOH, where PFeIII represents the ferric heme. The abbreviations used in the text are shown.

cooled on ice and centrifuged, and 100-ml aliquots were analyzed with a 4.6 1 250-mm Supelcosil LC-18-T column (Supelco, Bellefonte, PA) on an LKB 2249 HPLC with an LKB 2141 uv monitor (Pharmacia Biotech, Piscataway, NJ) set to 340 nm. The mobile phase consisted of 0.1 M KH2PO4 , pH 6.0, containing 8% methanol at a flow rate of 1.3 ml/min. Quantitation of hepatic NADPH was based on standard curves constructed with NADPH over the concentration range 0.5 to 5.0 mM. ATP levels were determined using a modification of a published method (34). Cell incubates (1.0 ml) were treated with 200 ml of 15% trichloroacetic acid containing 4 mM EDTA. These samples were centrifuged and an aliquot (10 ml) from each was diluted with 500 ml of a solution containing 33 mM Na2AsO4 , 3.3 mM KH2PO4 , and 7.0 mM MgSO4 (pH 7.6). The luciferase reagent (100 ml) was added to 100 ml of this solution, luminescence measured, and the quantitation was conducted utilizing a standard calibration curve prepared over the range 0 to 40 mM ATP.

was generated slowly over the entire 90-min incubation and its formation was accompanied by a small decrease in BHTOH (Fig. 3C). Samples of the diol isolated from incubations with [18O2]BHTOOH were analyzed by mass spectrometry, and the results are shown in Table I. Almost all BHT(OH)2 produced within the first 2.5 min contained two atoms of oxygen-18, a result similar

RESULTS

Metabolism. Metabolites formed by incubating BHTOOH with freshly isolated rat hepatocytes were analyzed by HPLC (Fig. 2) and characterized by comparing their elution characteristics and uv spectra with products identified in earlier studies (15, 21). The compounds present at various incubation times demonstrated nearly complete consumption of 250 mM BHTOOH within 5 min, and hydroperoxide disappearance coincided with the formation of BHTOH, RP5, and RP7 (Figs. 3A and 3B). On the other hand, BHT(OH)2

/ m4345$9173

10-12-95 10:27:51

arca

FIG. 2. HPLC analysis of metabolites formed by incubating 250 mM BHTOOH with isolated rat hepatocytes (106 cells/ml) for 5 min. Incubation and chromatographic conditions are described under Materials and Methods. IS Å internal standard.

AP: Archives

466

THOMPSON ET AL.

FIG. 4. Hepatocyte viability with 200, 250, and 300 mM BHTOOH, and 350, 400, and 450 mM CuOOH as a function of incubation time. Incubations were conducted with isolated rat hepatocytes (106 cells/ ml) and BHTOOH or CuOOH. Controls did not contain a hydroperoxide. Cell viability was determined at the times shown by the exclusion of trypan blue.

FIG. 3. Time courses for BHTOOH degradation and metabolite formation in hepatocyte suspensions (106 cells/ml). Data for plots A, B, and C were obtained with 250 mM BHTOOH, and the data for plot D were obtained with 100 mM BHTOOH.

to that obtained with liver microsomes and consistent with the reaction sequence shown in Fig. 1 (reactions a and b) involving retention of the terminal oxygen of the hydroperoxy group (21). Analysis of BHT(OH)2 isolated after 30 min, however, demonstrated that 71%

TABLE I

Retention of Oxygen-18 in Metabolites Derived from [18O2]BHTOOH Isotope compositiona

Metabolite BHT(OH)2 RP5

Incubation time (min) 2.5 30 2.5 30

18

O,16O (%)

10.1 71.0 57.8 60.2

18

O,18O (%)

89.9 29.0 42.2 39.8

Note. [18O2]BHTOOH (250 mM) was incubated with hepatocytes (106 cells/ml), metabolites were isolated by HPLC, and isotope compositions were determined by GC/MS as described under Materials and Methods. a Values were calculated from ratios of the ion currents corresponding to fragment ions containing 18O / 16O or 18O / 18O, and are the means of three determinations corrected to 100% 18O2 enrichment of the labeled hydroperoxide. Standard deviations were in the range of 3 to 5 % of the means.

/ m4345$9173

10-12-95 10:27:51

arca

of the diol contained oxygen-16 in the side chain, presumably introduced via the NADPH/O2-supported hydroxylation of BHTOH. The oxygen-18 content of RP5 isolated from cell incubations containing the labeled hydroperoxide also was measured, and the results (Table I) further indicate the involvement of P450 in BHTOOH degradation. This rearranged product is believed to arise either by a ‘‘rebound’’ mechanism between the carbon-centered radical and the Compound II complex of P450 leading to oxygen-18 incorporation in the side chain (Fig. 1, reaction f), or oxidation of the radical and attack by water on the carbocation (reactions g and h) (15). The relative contributions of these competing pathways in hepatocytes are similar to those found in liver microsomes, i.e., 40% of the RP5 arises from radical rebound in the former system and 32% in the latter. The disappearance of 250 mM BHTOOH over the first 5 min was accompanied by the formation of approximately 25% of the total BHT(OH)2 present in hepatocyte incubations after 90 min (Figs. 3A and 3B). With 100 mM BHTOOH, 5% of the total diol and all of the RP7 (the only homolytic product detected due to the low uv response of RP5) were generated during the initial stages of the incubation, i.e., the period of hydroperoxide consumption (Fig. 3D). These results demonstrate that, even under conditions where GSH is not depleted (see below) and Gpx reduces a greater percentage of the hydroperoxide, P450 is still able to contribute to BHTOOH metabolism. Cytotoxicity. Data in Fig. 4 compare the cytotoxicity of BHTOOH with that of CuOOH, an alkyl hydroperoxide more frequently employed in studies of oxidative damage (35). These results demonstrate that the former is significantly more toxic than the latter. Reaction with BHTOOH virtually depleted intracellular GSH

AP: Archives

467

METABOLISM AND TOXICITY OF HYDROPEROXIDES IN HEPATOCYTES TABLE III

Effects of BHTOOH on Protein Thiols and ATP in Isolated Hepatocytes Protein thiols

FIG. 5. Time course of GSH and GSSG levels in hepatocytes incubated with 250 mM BHTOOH.

within the first 0.5 min and this was accompanied by a rise in GSSG levels (Fig. 5). Most of the initial GSH was restored within 20 min and recovery was followed by a slow decline in GSH levels which paralleled cell death (data not shown). With 100 mM BHTOOH, a nonlethal dose, the GSH concentration was rapidly lowered but not fully depleted, and its recovery to about 60% of normal concentrations occurred within 1 min. NADPH also declined rapidly, falling to 56% of initial levels within the first minute before a subsequent rebound (Table II). Decreases in the concentrations of protein thiols and ATP, however, occurred slowly and in parallel with cell death (Table III). The data summarized in Table IV demonstrate that BHTOOH caused substantially less lipid peroxidation than CuOOH, in agreement with previous results utilizing liver microsomes (25). CuOOH-induced peroxidation was eliminated and cell death was substantially reduced when

nmol/mg protein

%

0a 5 15 30 60 90

56 { 4 ndb 53 { 5 50 { 6 46 { 7 42 { 6

100 — 95 89 82 75

100 73 60 47 33 —

TABLE IV

Lipid peroxidation (nmol MDA/million cells)

mM { { { { { {

0.5 0.4 0.3 0.4 0.4 0.4

%

Hydroperoxide

100 66 56 54 72 86

BHTOOH

Note. Incubations were conducted with freshly isolated hepatocytes 106 cells/ml) and 250 mM BHTOOH. Values are means { SD of at least three cell preparations. a NADPH levels did not change during 15 min in control incubations (vehicle only).

/ m4345$9173

4 6 4 3 4

incubations were conducted in the presence of the antioxidant DPPD. As expected, DPPD had no significant effect on the cytotoxicity of BHTOOH. Influence of P450 inhibitors on metabolism and toxicity. The role of P450 in hydroperoxide-induced cytotoxicity was investigated with metyrapone, a competitive inhibitor, and ABT, a suicide inhibitor of the enzyme (36). Cells were preincubated with an inhibitor for 15 min prior to adding the hydroperoxide. Product analyses conducted after incubating BHTOOH with hepatocytes for 30 min in the presence of an inhibitor demonstrated a 40 to 70% decrease in the formation of

NADPH

5.0 3.3 2.8 2.7 3.6 4.3

30 { 22 { 18 { 14 { 10 { nd

Effects of the Antioxidant DPPD on Hydroperoxide-Induced Lipid Peroxidation and Toxicity

TABLE II

0a 0.5 1.0 2.5 5.0 15

%

mM

Note. Freshly isolated hepatocytes (106 cells/ml) were incubated with 250 mM BHTOOH, aliquots were removed at the indicated times, and analyses were conducted as described under materials and Methods. Values are means { SD of at least three cell preparations. a The concentrations of protein thiols in control incubations (containing the vehicle only) decreased by 6% over the 90-min incubation period while ATP decreased by 3% over 60 min. b Value not determined.

Effects of BHTOOH on Hepatocyte NADPH Levels

Incubation time (min)

ATP

Incubation time (min)

10-12-95 10:27:51

arca

Cytotoxicity (% viable)

Time (min)

0DPPD

/DPPD

0DPPD

/DPPD

30 60 30 60

1.5 1.3 3.1 3.4

0.5a 0.5a 0.6a 0.6a

73 52 12 4

71 54 65 55

CuOOH

Note. Incubations were conducted with isolated hepatocytes (106/ ml), 250 mM BHTOOH or 400 mM CuOOH, and some contained 20 mM DPPD added 5 min before the hydroperoxide. Cell viability throughout the 60-min incubation period in the absence of a hydroperoxide was 86 { 4%. a Indistinguishable from background levels of MDA which averaged 0.6 { 0.2 nmol/ml.

AP: Archives

468

THOMPSON ET AL.

FIG. 6. Influences of P450 inhibitors on hydroperoxide-induced cytotoxicity. (A) BHTOOH (250 mM, solid line) was incubated with hepatocytes (106 cells/ml) containing no inhibitor (l), 0.5 mM metyrapone (j), or 0.5 mM ABT (m) for the times indicated. Control incubations (dotted line) were identical except that BHTOOH was omitted. Following 15-min incubations with an inhibitor, BHTOOH was added at the 0-min point and cell viability was determined as described in the legend to Fig. 4. (B) CuOOH (400 mM, solid line) was incubated as above with no additions (l), or with 0.5 mM metyrapone (j). Control incubations (dotted line) contained no CuOOH or inhibitor.

all metabolites except BHTOH. Twelve to 16% greater amounts of alcohol were produced due, presumably, to additional contributions from Gpx. Neither inhibitor alone affected cell viability, but both enhanced the cytotoxicity of BHTOOH substantially (Fig. 6A). Experiments conducted as described for Fig. 4, but with BHTOH substituted for BHTOOH, demonstrated that the alcohol is markedly less toxic than the hydroperoxide (data not shown). Increased formation of BHTOH, therefore, does not explain the effects of P450 inhibitors on BHTOOH toxicity. Effects of metyrapone on CuOOH cytotoxicity were examined and the results (Fig. 6B) demonstrate that, in contrast to its effects on BHTOOH toxicity, metyrapone protects cells from the toxic effects of CuOOH. This result is consistent with the effects of metyrapone on tBuOOH-induced toxicity determined previously (2). DISCUSSION

The metabolites formed by incubating BHTOOH with suspensions of freshly isolated rat hepatocytes demonstrate that Gpx and P450 both participate in hydroperoxide reduction. As expected for the Gpx component, rapid consumption of BHTOOH was accompanied by the formation of BHTOH and GSSG. The contribution of P450 is indicated by the fact that BHT(OH)2 , RP5, and RP7 are produced also. The formation of BHT(OH)2 does not occur with any of the other heme-containing compounds examined, including myoglobin, hemoglobin (24), horseradish peroxidase (21), microperoxidase, and hematin.5 A high level 5

Thompson, J. A., and Schullek, K. M., unpublished results.

/ m4345$9173

10-12-95 10:27:51

arca

of oxygen-18 retention in the fraction of diol formed simultaneously with the disappearance of [18O2]BHTOOH is consistent with P450-catalyzed isomerization. This process involves transfer of the terminal oxygen of the hydroperoxy group to a hydrocarbon substituent within the same molecule (Fig. 1, reactions a and b) (21, 37, 38). Although the rearranged products RP5 and RP7 can be generated by other hemeproteins (24), the oxygen-18 content of RP5 from hepatocytes is similar to that observed with microsomal P450 (15). According to the scheme shown in Fig. 1, retention of this isotope in RP5 is dependent on the competition between reactions f and g. These processes should be influenced by the reactivity of the oxygenated heme intermediate, so the isotopic content of RP5 is expected to be sensitive to structural differences among hemeproteins, including the nature of the proximal ligand (39). Preliminary data reveal significant differences in the oxygen-18 content of RP5 when ferric P450s were compared to hemoglobin and myoglobin because the radical recombination pathway (reaction f) is substantially more efficient in the case of P450 (see footnote 5). The relative contributions of Gpx and P450 cannot be determined accurately because BHTOH is produced by both enzymes, and GSH is continuously regenerated by GSSG reductase. An estimate is available, however, by comparing the ratio of BHT(OH)2 :BHTOH present at the 5-min point in cell incubations with the ratio obtained from rat liver microsomes (23). This comparison assumes that the second stage of the P450-catalyzed isomerization (i.e., Fig. 1, reaction b) occurs to the same extent in both systems. The BHT(OH)2 :BHTOH ratio from hepatocytes is approximately 25–30% of that determined with microsomes, demonstrating that cellular P450 plays a major roll in BHTOOH reduction. The data summarized above strongly support hydroperoxide reduction by ferric rather than ferrous P450. In its ferrous oxidation state, the enzyme would reduce BHTOOH by one electron resulting in homolysis and rearrangement to the carbon-centered radical (Eq. [1]). It is highly unlikely, however, that RP5 or RP7 would be produced in this case because no ferryl complex is present to catalyze the final stages of product formation. Previous studies have shown that BHTOr generated within the P450 active site undergoes intramolecular reactions rather than hydrogen abstraction from lipids (15, 25). The modest amount of lipid peroxidation occurring in cells treated with BHTOOH is more readily explained by nonheme, nonferritin iron-catalyzed homolysis occurring at sites where alkoxy radical attack on lipids can compete with rearrangement (4). Reduction of BHTOOH by the ferric form of the enzyme may be due to inefficient NADPH-P450 reductase activity because of rapid decreases in NADPH levels (Table II). Additionally, the strong affinity of this lipophilic hydroperoxide for the P450 active site may facilitate

AP: Archives

METABOLISM AND TOXICITY OF HYDROPEROXIDES IN HEPATOCYTES

interaction between the heme iron and the terminal oxygen of the hydroperoxy group and enhance electron transfer from the ferric heme. Significant differences were found in the responses of hepatocytes to BHTOOH compared with CuOOH, and the effects of the latter are similar to those of tBuOOH determined in previous studies (2–4). Higher concentrations of CuOOH and tBuOOH are required to destroy hepatocytes and both induce substantial amounts of lipid peroxidation which can be inhibited by DPPD. When peroxidation is suppressed, hepatocytes remain intact for longer periods of time, thereby implicating peroxidative events in cell death (2, 4, 13). The failure of BHTOOH to induce substantial levels of lipid peroxidation in liver microsomes (25) or hepatocytes (Table IV) is explained by the alternative reaction pathway available to BHTOr, namely rearrangements to stable products. Rearrangement is more facile than bscission (Fig. 1, reaction d) as the latter requires extrusion of the relatively unstable methyl radical (40), so very little DBBQ is produced (15). The small amount of lipid peroxidation resulting from BHTOOH treatment was suppressed by DPPD, but this had no effect on cell death. The potent cytotoxicity of BHTOOH in the absence of substantial lipid peroxidation demonstrates notable differences between this peroxyquinol and alkyl hydroperoxides with regard to mechanisms of cell destruction. Evidence that P450 affects hydroperoxide-mediated cytotoxicity was obtained using enzyme inhibitors. The cytotoxicity of CuOOH examined here and tBuOOH in previous studies (2) was lowered in the presence of the competitive inhibitor metyrapone. In contrast, the cytotoxicity of BHTOOH was substantially enhanced by metyrapone (Fig. 6). Metabolite analysis confirmed that metyrapone inhibited BHTOOH degradation suggesting that P450 protects cells from the lethal effects of this hydroperoxide by converting it to nontoxic products. This conclusion is strengthened by the finding that the suicide inhibitor ABT produced the same effects as metyrapone. Contributions of P450 to hydroperoxide-induced cytotoxicity are due most likely to the homolytic scission of O–O bonds and reactions of the resulting alkoxy radicals. The principal reaction pathways for alkoxy radicals derived from CuOOH and tBuOOH are b-scission and hydrogen abstraction. The former process is not highly favorable, as it requires extrusion of an unstabilized methyl radical, so hydrogen abstraction from lipids can compete with b-scission in those cases (15, 41). In contrast to the results discussed above, P450 protected cells from the lethal effects of BHTOOH, partly because homolysis leads to rearranged products instead of lipid degradation. Inhibiting P450, therefore, may enable the processing of BHTOOH along more toxic pathways. For example, BHTOOH-induced toxic-

/ m4345$9173

10-12-95 10:27:51

arca

469

ity in a cell line derived from murine keratinocytes is believed to involve para-quinone methide formation through (i) oxidation of the peroxyquinol to its peoxy radical, (ii) loss of O2 (Eq. [4]), and (iii) disproportionation of the resulting phenoxy radial (42). BHTOOH r BHTOOr r BHTr / O2

[4]

Quinone methides are electrophilic and can covalently modify proteins. This process occurs in cells containing little or no P450, so it is feasible that it also occurs in hepatocytes treated with P450 inhibitors. The mechanisms of BHTOOH-induced cytotoxicity cannot be discerned from data available at the present time, but could be related to its high lipophilicity which should facilitate absorption into cells and distribution to various intracellular sites including mitochondria. Mitochondrial damage has been implicated previously in tBuOOH-induced cell death (5), and mitochondrial effects of BHTOOH were demonstrated in the present work by losses of ATP in parallel with cell death. BHTOOH appears to be an excellent model compound for further studies into mechanisms of hydroperoxideinduced cell death which do not involve lipid peroxidation. In conclusion, the results summarized in this paper demonstrate that hepatocyte P450, in its ferric oxidation state, competes with Gpx for the reduction of organic hydroperoxides. Both heterolytic and homolytic cleavages of O–O bonds occur as expected from studies with purified P450. Marked differences in mechanisms of cytotoxicity of CuOOH versus BHTOOH were observed with lipid peroxidation contributing to CuOOHbut not to BHTOOH-induced toxicity. In cases where the alkoxy radical from homolysis attacks lipids, P450 contributes to cytotoxicity. If the alkoxy radical can undergo facile intramolecular reactions to produce stable products, the role of P450 is detoxification. REFERENCES 1. Ross, D., Thor, H., Threadgill, M. D., Sandy, M. S., Smith, M. T., Moldeus, P., and Orrenius, S. (1986) Arch. Biochem. Biophys. 248, 460–466. 2. Jewell, S. A., Di Monte, D., Richelmi, P., Bellomo, G., and Orrenius, S. (1986) J. Biochem. Toxicol. 1, 13–22. 3. Tribble, D. L., Jones, D. P., and Edmondson, D. E. (1988) Mol. Pharmacol. 34, 423–420. 4. Masaki, N., Kyle, M., and Farber, J. L. (1989) Archiv. Biochem. Biophys. 269, 390–399. 5. Masaki, N., Kyle, M. E., Serroni, A., and Farber, J. L. (1989) Arch. Biochem. Biophys. 270, 672–680. 6. Chai, Y.-C., Hendrich, S., and Thomas, J. A. (1994) Arch. Biochem. Biophys. 310, 264–272. 7. Nicotera, P., McConkey, D. J., Svensson, S.-A., Bellomo, G., and Orrenius, S. (1988) Toxicology 52, 55–63. 8. Farber, J. L. (1994) Environ. Health Prospect. 102(Suppl. 10), 17–24.

AP: Archives

470

THOMPSON ET AL.

9. Boobis, A. R., Fawthrop, D. J., and Davies, D. S. (1989) Trends Pharmacol. Sci. 10, 275–280. 10. Nicotera, P., McConkey, D. J., Dypbukt, J. M., Jones, D. P., and Orrenius, S. (1989) Drug Metab. Rev. 20, 193–201. 11. Reed, D. J. (1990) Annu. Rev. Pharmacol. Toxicol. 30, 603–631. 12. Shertzer, H. G., Bannenberg, G. L., and Moldeus, P. (1992) Biochem. Pharmacol. 44, 1367–1373. 13. Sakaida, I., Kyle, M. E., and Farber, J. L. (1990) Mol. Pharmacol. 37, 435–442. 14. Marnett, L. J., Weller, P., and Battista, J. R. (1986) in Cytochrome P450, Structure, Mechanism & Biochemistry (Ortiz de Montellano, P. R., Ed.), pp. 29–76, Plenum, New York. 15. Wand, M. D., and Thompson, J. A. (1986) J. Biol. Chem. 261, 14049–14056. 16. Ortiz de Montellano, P. R., and Watanabe, M. D. (1987) Mol. Pharmacol. 31, 213–219. 17. Adamson, G. M., and Harman, A. W. (1989) Biochem. Pharmacol. 38, 3323–3330. 18. Steward, A. R., Dannan, G. A., Guzelian, P. S., and Guengerich, F. P. (1985) Mol. Pharmacol. 27, 125–132. 19. Paine, A. J. (1990) Chem.-Biol. Interact. 74, 1–31. 20. Vaz, A. D. N., and Coon, M. J. (1987) Proc. Natl. Acad. Sci. USA 84, 1172–1176. 21. Thompson, J. A., and Wand, M. D. (1985) J. Biol. Chem. 260, 10637–10644. 22. Shimizu, T., Murakami, Y., and Hatano, M. (1994) J. Biol. Chem. 269, 13296–13304.15. 23. Correia, M. A., Yao, K., Allentoff, A. J., Wrighton, S. A., and Thomposon, J. A. (1995) Arch. Biochem. Biophys. 317, 471–478. 24. Allentoff, A. J., Bolton, J. L., Wilks, A., Thompson, J. A., and Ortiz de Montellano, P. R. (1992) J. Am. Chem. Soc. 114, 9744– 9749. 25. Thompson, J. A., Bolton, J. L., and Malkinson, A. M. (1991) Exp. Lung Res. 17, 439–453.

/ m4345$9173

10-12-95 10:27:51

arca

26. Wand, M. D., and Thompson, J. A. (1986) Adv. Exp. Med. Biol. 197, 337–346. 27. Moldeus, P., Hogberg, J., and Orrenius, S. (1978) Methods Enzymol. 52, 60–71. 28. Klingenberg, M. (1970) in Methoden der Enzymatischen Analyze (Bergmeyer, H. V., Ed.), p. 175, Verlag Chemie, Weinheim. 29. DiMonte, D., Bellomo, G., Thor, H., Nicotera, P., and Orrenius, S. (1984) Arch. Biochem. Biophys. 235, 343–350. 30. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76–85. 31. Reed, D. J., Babson, J. R., Beatty, P. W., Brodie, A. E., Ellis, W. W., and Potter, D. W. (1980) Anal. Biochem. 106, 55–62. 32. Jones, D. P. (1981) J. Chromatogr. 225, 446–449. 33. Stocchi, V., Cucchiarini, L., Magnani, M., Chiarantini, L., Palma, P., and Crescentini, G. (1985) Anal. Biochem. 146, 118–124. 34. Stubberfield, C. R., and Cohen, G. M. (1988) Biochem. Pharmacol. 37, 3967–3974. 35. Hill, K. E., and Burk, R. F. (1984) Toxicol. Appl. Pharmacol. 72, 32–39. 36. Ortiz de Montellano, P. R., and Reich, N. O. (1986) in Cytochrome P450, Structure, Mechanism & Biochemistry (Ortiz de Montellano, P. R., Ed.), pp. 273–314, Plenum, New York. 37. Larroque, C., and van Lier, J. E. (1986) J. Biol. Chem. 261, 1083–1087. 38. Fish, K. M., Avaria, G. E., and Groves, J. T. (1988) in Microsomes and Drug Oxidations (Miners, J., Birkett, D. J., Dew, and McManus, Eds.), pp. 176–183, Taylor and Francis, New York. 39. Ortiz de Montellano, P. R. (1992) Annu. Rev. Pharmacol. Toxicol. 32, 89–107. 40. Walling, C., and Padwa, A. (1963) J. Am. Chem. Soc. 85, 1593– 1601. 41. Yumibe, N. P., and Thompson, J. A. (1988) Chem. Res. Toxicol. 1, 385–390. 42. Guyton, K. Z., Thompson, J. A., and Kensler, T. W. (1993) Chem. Res. Toxicol. 6, 731–738.

AP: Archives