Reconstituted microsomal lipid peroxidation: ADP-Fe3+-dependent peroxidation of phospholipid vesicles containing NADPH-cytochrome P450 reductase and cytochrome P450

Free Radical Biology & Medicine, Vol. 4, pp. 269-277, 1988 0891-5849/88 $3.00 + .00 Printed in the USA. All rights reserved. © 1988 Pergamon Press plc

Original Contribution

RECONSTITUTED MICROSOMAL LIPID PEROXIDATION: ADP-FE 3÷- DEPENDENT PEROXIDATION OF PHOSPHOLIPID VESICLES

CONTAINING NADPH-CYTOCHROME P450 REDUCTASE AND CYTOCHROME P450

LEE A. MOREHOUSE* and STEVEN D. A U S T t

Center for Active Oxygen in Biology and Medicine, Department of Biochemistry, Michigan State University, East Lansing, MI 48824

(Received 21 April 1987; Revised 22 June 1987; Accepted 7 July 1987)

Abstract--A reconstituted lipid peroxidation system consisting of rat liver microsomal NADPH-cytochrome P450 reductase and cytochrome P450 incorporated into phospholipid vesicles was developed and characterized. Per- oxidation of the vesicles required NADPH and ADP-Fe 3+, just as in the NADPH-dependent peroxidation of microsomes. The peroxidation of the vesicles was inhibited 30-50% by superoxide dismutase, depending upon their cytochrome P450 content: those with higher cytochrome P450 contents exhibited greater rates of malondialdehyde formation which were less sensitive to inhibition by superoxide dismutase. When cytochrome P450 was incorporated into vesicles, EDTA-Fe 3÷ was not required for lipid peroxidation, distinguishing this system from the one previously described by Pederson and Aust [Biochem. Biophys. Res. Comm. 48,789; 1972]. Since at least 50% of the malondialdehyde formation in the vesicular system was not inhibited by superoxide dismutase, alter- native means of iron reduction (O2:-independent) were examined. It was found that rat liver microsomes or a reconstituted mixed function oxidase system consisting of NADPH-cytochrome P450 reductase and cytochrome P450 in dilauroylphosphatidylcholine micelles reduced ADP-Fe 3÷ under anerobic conditions.

Keywords--Microsomal lipid peroxidation, Iron reduction, NADPH-cytochrome P450 reductase, Cytochrome P450, Phospholipid vesicles, Malondialdehyde, Superoxide dismutase, ADP-Fe 3+

INTRODUCTION

Microsomal lipid peroxidation was first described by Hochstein and Ernster who demonstrated the requirement for NADPH and ADP, and the enzymatic nature of the process.1 Their subsequent study of the perox- idative process also identified the necessity for iron 2 which had been a contaminant of their original ADP solutions. That the enzymatic nature of the process was responsible for ADP-Fe 3 ÷ reduction became evident in studies in which reductants such as ascorbate could substitute for NADPH, 2 or where ferrous addition to microsomes resulted in lipid peroxidation without the inclusion of reducing agents. 2-~

Pederson and Aust further characterized the en- zymic nature of lipid peroxidation by demonstrating that NADPH-cytochrome P450 reductase (reductase)

*Current Address: Pfizer Central Research, Groton, CT 06340. tAddress to: Biotechnology Center, Utah State University, Logan, UT 84322-4430.

was the enzyme linking NADPH oxidation to ADP- Fe3+-dependent peroxidation of microsomal membranes. 4 They also developed a reconstituted lipid peroxidation system consisting of phospholipid liposomes, purified reductase, NADPH, and ferric chelates. In their reconstituted system, a second ferric chelate, EDTA-Fe 3+ (in addition to ADP-Fe 3÷) was required. Other investigators have since confil"med these observations. 5

These observations that only ADP-Fe 3+ is necessary to promote NADPH-dependent peroxidation of microsomes, while both EDTA-Fe 3 + and ADP-Fe 3 ÷ are needed to promote peroxidation in a reconsituted system suggest that there may be microsomal component(s) that directly reduce ADP-Fe 3÷ for which EDTA-Fe 3+ can substitute in the reconstituted lipid peroxidation system. Several lines of evidence would suggest that cytochrome P450 isozymes might be those microsomal components. Cytochrome P450 is a terminal electron acceptor for the NADPH-dependent electron transport

269

270 L.A. MOREHOUSE and S. D. AUST

chain, and inhibitors and substrates of cytochrome P450 have been reported to inhibit microsomal lipid peroxidation. 6'7 Besides forming a competent electron transport chain capable of mixed function oxidase activity, coupling of cytochrome P450 with the reductase is proposed to result in increased 027 production. 8-1° Many investigators have reported that the NADPH- dependent iron reduction and lipid peroxidation are Oz ~ dependent processes. ~- 13

Only one investigation has attempted to address in a definitive manner the potential of cytochrome P450 to participate in the initiation phases of lipid peroxidation. Elkstrom and Ingelman-Sundberg reported 14 that phospholipid vesicles containing cytochrome P450 and reductase underwent NADPH-dependent peroxidation to a greater extent than vesicles containing only the reductase. Unfortunately, they did not demonstrate that the peroxidation process required exogenously added iron, a characteristic of most model lipid peroxidation systems. Therefore, it is conceivable that the cytochrome P450-dependent peroxidation they observed was due to the cleavage of preformed lipid hydroperoxides, a process termed lipid hydroperoxide-dependent lipid peroxidation. 15 Hemoproteins such as cytochrome P450 and the other transition metal complexes readily cleave lipid hydroperoxides to radical species 16 which are capable of propagating lipid peroxidation, is presumably by abstracting methylene hydrogens from polyunsaturated fatty acids. The following investigation was conducted to ascertain whether the addition of cytochrome P450 to the reconstituted system would result in the reduction of ADP-Fe 3 + and the initiation of lipid peroxidation.

MATERIALS AND METHODS

Chemicals

Butylated hydroxytolulene, 2-thiobarbituric acid (TBA), cytochrome c (Type VI), superoxide dismutase (SOD) (from bovine erythrocytes), NADPH, ADP, die- thylenetriaminepentaacetic acid (DTPA), dilauroylphosphatidylcholine (DLPC), and o-phenanthroline were obtained from Sigma (St. Louis, MO). Azolectin (a commercial preparation of phospholipids isolated from soybeans) was purchased from Associated Concen- trates (Long Island, NY). Catalase was a product of Cooper Biomedical (Malvern, PA) and EDTA was purchased from Mallinckrodt Chemical Co. (St. Louis, MO). All other reagents were of analytical grade and used without further purification. Solutions used in lipid peroxidation or iron reduction experiments were pretreated with Chelex 100 (Bio-Rad Laboratories, Richmond, CA) to remove contaminating transition metal ions. Catalase and SOD preparations were chro-

matographed on Sephadex G-25 to remove low molecular weight contaminants and reassayed by standard procedures before use.

Preparation of Microsomes

Rat liver microsomes were isolated from male Spra- gue-Dawley rats (250-300 g) (Charles River, Portage, MI) by the method of Pederson and Aust.17 All buffers used for the isolation of microsomes were ice-cold and exhaustively purged with argon prior to use in order to minimize the oxidation of microsomal lipids, and all steps were done at 4°C unless otherwise indicated. Microsomal pellets were resuspended in 10 mM EDTA, 150 mM KC1, pH 7.0 and recentrifuged at 150,000 x g for 60 min. Pellets were resuspended in 20 mM Tris pH 7.5 containing 0.15 M KCI and chromatographed on a Sepharose CL-2B column (2.5 x 30 cm). Mi- crosomes eluting at the void volume of the column were pooled and centrifuged at 105,000 x g for 60 min. The resulting pellet was used immediately or suspended in 50 mM Tris pH 7.5 containing 50% glycerol and stored at -20°C.

Purification of Enzymes

Isolation of cytochrome P450b and the reductase was from liver microsomes of rats pretreated with 0.1% sodium phenobarbital in their drinking water for 10 days prior to sacrifice. Microsomes were solubilized with 1.5% Emulgen 911 and subjected to DEAE Seph- adex A25 chromatography by the procedure of Dignam and Strobel.18 The cytochrome P450 isozymes present in the void-volume were precipitated with 18% poly- ethylene glycol and centrifuged at 105,000 x g for 1 h. The precipitate was suspended in a small volume of 10 mM KHEPO4, pH 7.4, 20% glycerol, 0.2% Emul- gen 911, 0.5% sodium cholate and 0.1 mM EDTA. Cytochrome P450b was further purified using the procedure of Waxman and Walsh.19 Briefly, this consists of DEAE-cellulose chromatography as initially described by West et al. z° with subsequent chromatography of the partially purified protein on hydroxyapa- tite and CM-Sepharose. Resulting preparations were subjected to SDS-PAGE and judged to be homogene- ous. These preparations had specific contents ranging from 10-14 nmol/mg protein.

NADPH-cytochrome P450 reductase was eluted from the DEAE Sephadex A25 column with buffer containing 0.3 M KC1; and diluted 3-fold with 20% glycerol and 0.1% Emulgen 911 and subjected to ADP-Agarose affinity chromatography as originally described.21 Spe- cific activities ranged from 40 to 60 units/rag protein.

ADP-Fe3+-dependent lipid peroxidation 271

Preparation of Phospholipid Vesicles

A 10% solution of asolectin in Argon-purged chloro- form:methanol (2:1 v/V) was placed in a capped vial and kept overnight at - 20°C. The clear, upper phase containing the phospholipids was decanted and used for vesicle preparation. Aliquots of the clear supernatant were placed in test tubes and dried under a stream of argon. Tris buffer (50 mM, pH 7.5) that had been pretreated with Chelex 100 was added along with reductase and various concentrations of cytochrome P450 isozymes to give a final volume of 3 ml. The reductase was preincubated with cytochrome P450 prior to their addition to the lipid. The head space of the test tube was purged with argon and the tube was capped. The tube was immediately placed into a 150 ml beaker containing ice-cold water and the solution was indi- rectly sonicated using a Branson Sonifier probe at 60% power output (immersed in the beaker) until the lipid was completely resuspended (approximately 5 min). The tube containing the phospholipid vesicles and enzymes was then placed in a Branson Bath Sonicator and sonicated for 1 h with ice replenished periodically in order to maintain a temperature of 0-5°C. The sonicated solution was centrifuged at 105,000 x g for 1 h at 4°C and the turbid supernatant that contained the bulk of cytochrome P450 content and reductase activity was used for lipid peroxidation assays.

Iron Reduction Assays

Iron reduction in incubations containing microsomes or purified enzymes was monitored spectropho- tometrically at 510 nm, observing the absorbance of the o-phenanthroline-Fe 2÷ complex. Aliquots of anaerobic incubations (0.5 or 1.0 ml) were quenched in 0.2% o-phenanthroline (1 ml) and the protein precipitated by the addition of 0.5 ml of 20% TCA. The o- phenanthroline-Fe 2÷ chelate was extracted from the acidified mixture with 2 ml of n-amyl alcohol and the absorbance of the organic phase was measured. Stan- dard curves were prepared using known ferric chelate concentrations reduced with excess thioglycolate.

Lipid Peroxidation Assays

Stock solutions of ADP-Fe 3 ÷ and NADPH, prepared in 50 mM Tris, pH 7.5, were used to constitute lipid peroxidation mixtures that were incubated in a shaking water bath maintained at 37°C. Final concentrations of the solutions are given in the figure and table legends. Lipid peroxidation incubation mixtures also contained either phospholipid vesicles prepared as previously described, 4 or azolectin vesicles containing cyto-

chrome P450 and reductase. When phospholipid liposomes were used, aliquots of concentrated solutions of purified reductase and cytochrome P450 were also added to the incubations as described in the table legends. Lipid peroxidation was initiated by the addition of NADPH. At specific times, aliquots of the reaction mixture were withdrawn and assayed for malondialdehyde (MDA) by the TBA assay. 22

RESULTS

Incorporation of Cytochrome P 450 and NADPH-Cytochrome P450 Reductase Into Phospholipid Vesicles

Cytochrome P450 and reductase were incorporated into azolectin vesicles using an indirect sonication procedure in the absence of detergents. Preparations of vesicles containing the incorporated proteins were subjected to gel permeation chromatography on Sepharose CL-4B. The resulting fractions were analyzed to phosphate, reductase activity, and cytochrome P450 content. An elution profile is shown in Figure 1. The majority of phosphate, reductase and cytochrome P450 eluted at the void volume of the column, indicating that the proteins had been incorporated into the vesicles. Treatment of these vesicles with bromelain or substitution of the protease-solubilized reductase (lack- ing its membrane-binding domain) for the detergent- solubilized enzyme resulted in elution profiles in which the reductase activity eluted in the partially-included volume of the column (data not shown).

NADPH-Dependent Peroxidation of Phospholipid Vesicles

Azolectin vesicles containing reductase and cytochrome P450 isozymes were incubated with iron chelates and/or NADPH in order to assess whether these vesicles containing cytochrome P450 and the reductase would promote ADP-Fe 3 ÷-dependent peroxidation like that observed with rat liver microsomes. The addition of both ADP-Fe 3+ and NADPH to vesicles resulted in significant MDA formation (Table 1). The addition of SOD to the vesicles in the presence of ADP-Fe 3÷ and NADPH resulted in a 42% inhibition of MDA formation. As the concentration of vesicles incubated with ADP-Fe 3 + and NADPH was increased, the corresponding rates of MDA formation also increased (Fig. 2). Also, vesicle containing more reductase and/or cytochrome P450 per ixmol lipid phosphate exhibited greater rates of peroxidation (data not shown).

That this MDA formation was dependent upon the ADP-Fe 3 + promoted initiation reactions was indicated


6

5

o -6 4 - 0 E

~ 3

IJ Q.

I ! ! i I • ! I !

I0 20 50 40

Volume (ml)

0 4

E "6

03

o tt3 (1.

o2 ~ o

0.1

Fig. 1. Elution profile of NADPH-cytochrome P450 reductase and cytochrome P450-containing phospholipid vesicles subjected to gel filtration. A 1 ml aliquot of vesicles was chromatographed on a 1 × 30 cm Sepharose CL-4B column. Fractions (1 ml) were collected and assayed for reductase activity (O), cytochrome P450 content (11) and phosphate (A)-

by several pieces of data. First, lipid hydroperoxide- dependent lipid peroxidation has been reported to be

promoted by iron chelates such as EDTA-Fe 3 ÷ or hemo-

proteins such as cytochrome P450, and only back-

ground rates of peroxidation occurred in the presence of EDTA-Fe 3÷ (Table 1). Secondly, as the concentra-

tion of ADP-Fe 3÷ in the incubations was increased,

the corresponding rates of lipid peroxidation also rose

proportionally (Fig. 3). By altering the concentrat ion of cytochrome P450

added to the azolectin prior to vesicle formation, ves-

icles having different cytochrome P450 contents were

prepared. Vesicles with greater cytochrome P450 con-

Table 1. Requirements for the NADPH-Dependent Peroxidation of NADPH-Cytochrome P450 Reductase and Cytochrome P450-

Containing Phospholipid Vesicles

MDA (nmol/15 min/ml)

Vesicles 0.2 + ADP-FE 3+ 0.5 + NADPH 0.4 +EDTA-Fe 3+ + NADPH 0.2

Complete System 7.4 + SOD (250 u/ml) 4.3 + EDTA-Fe 3÷ 31.1

Note: The complete system contained vesicles (0.43 u/ml reductase, 0.33 nmoltml cytochrome P450, 3.7 mol lipid phosphate/ml), ADP-Fe 3. (0.75 mM ADP, 0.15 mM FeC13) and NADPH (0.2 raM) in 50 mM Tris, pH 7.5. EDTA- Fe 3+ was 0.11 mM EDTA and 0.1 mM FeC13.

tents peroxidized more rapidly (Fig. 4). The addition

of SOD to these incubations resulted in a 3 0 - 5 0 %

inhibi t ion of lipid peroxidation depending on the cy-

tochrome P450 content. The greater the cytochrome

P450 content in the vesicles, the less the rate of per-

oxidation was inhibited by SOD (Fig. 4).

t w w • !i Vesicles (roll

Fig. 2. Effect of vesicle concentration on the peroxidation of phospholipid vesicles. Various concentrations of phospholipid vesicles (4.2 units reductase; 3.3 nmol cytochrome P450; 37 nmol lipid phosphate) were incubated with ADP-Fe 3+ (0.75 mM ADP, 0.15 mM FeCI3) and NADPH (0.2 mM) in 50 mM Tris, pH 7.5, at 37°C. The final incubation volumes were 2.5 ml.

ADP-Fe~÷-dependent lipid peroxidation 273

I(

c

~E 4

| I

' ' o ' . z ' o . 4

ADP-Fe 3÷ (raM)

Fig. 3. Effect of ADP-Fe 3÷ concentration on the peroxidation of phospholipid vesicles. Phospholipid vesicles (0.4 units reductase activity/ml, 0.25 nmol cytochrome P450/ml; 3.5 nmol lipid phosphate/ml) were incubated with NADPH (0.2 mM) and various concentrations of ADP-Fe 3+ (prepared as a 5 : 1 chelate) in 50 mM Tris, pH 7.5.

Reduction of ADP-Fe 3 + by the Microsomal Electron Transport System

Although it has been repor ted that rat l iver microsomes and reconst i tu ted mixed function oxidase systems generate 02 ~, NADPH-dependent microsomal lipid

peroxida t ion has not a lways been repor ted to be inhib-

i ted by SOD. In fact , in some studies SOD was repor ted to have vir tual ly no effect on l ipid peroxida t ion , 23-25

leading us to suggest that the predominant means of iron reduct ion in these systems was direct . 23 Previous

studies had demonst ra ted that ferric chelates such as fer r icyanide , EDTA-Fe 3 ÷ and DTPA-Fe 3 ÷ s t imulated

microsomal N A D P H oxidat ion by virtue o f being electron acceptors for the reductase.23'26 In one of these

studies, it was shown that EDTA-Fe 3÷ was reduced by the reductase under anaerobic conditions but ADP-Fe 3 ÷

was not. The reductase was shown to generate very litt le if any 02 ~,23'27 SO these observat ions are consis-

tent with not only the inabi l i ty of A D P - F e 3 ÷ to promote the peroxida t ion of phosphol ip id l iposomes in systems

containing only the reductase , but also the addi t ional requirement of E D T A -Fe 3÷. However , the incubat ion

of rat l iver microsomes with A D P - F e 3 ÷ and N A D P H results in l ipid peroxida t ion , inferring that A D P - F e 3÷ was reduced. The results shown in Table 2 demonst ra te that A D P - F e 3÷ can be reduced by rat l iver microsomes under anaerobic condi t ions . The rate of A D P - F e 3÷ reduction was only 15-20% that of EDTA-Fe 3÷ or DTPA- Fe 3+, ferric chelates that are reduced by the purif ied reductase. 23 The rate of A D P - F e 3÷ reduct ion by mi-

crosomes (at saturat ing N A D P H concentrat ions) was l inear ly dependent upon the concentra t ion of micro-

somal protein present in the incubat ion (Fig. 5).

Although the purified reductase does not reduce ADP- Fe 3+ anaerobica l ly , 23 incubat ion of the reductase with

the major phenobarb i ta l - inducib le i sozyme of cytochrome P450 in DLPC micel les (the standard conditions for reconstituting mixed-function oxidase activity 2s)

5

0 g I I I I I I I I

8 " 0.05 0.1 O J5 02 Cytochrome P450 (nmol/ml)

Fig. 4. Effect of increasing cytochrome P450 content on the peroxidation of phospholipid vesicles. Vesicle preparations having different cytochrome P450 contents were incubated with ADP-Fe 3÷ (0.5 mM ADP, 0.1 mM FeCI3) and NADPH (0.2 mM) in 50 mM Tris, pH 7.5. The reductase activity in the incubations was 0.33 --- 0.02 units/ml and the phospholipid content was 1.93 ± 0.4 p.mol lipid phosphate/ml. When added, SOD was 240 u/ml.

274 L . A . MOREHOUSE and S. D. AUST

Table 2. Rates of Reduction of Ferric Chelates by Rat Liver Microsomes under Anaerobic Conditions

Chelates nmol reduced/min/ml

EDTA-Fe 3 * 69.3 DTPA-Fe 3. 56.2 ADP-Fe 3÷ 10.2

Note: Anaerobic incubations contained NADPH (1 mM), ferric chelate (0.55 mM chelator: 0.5 mM FeCI3; except ADP- Fe 3÷ was 2.5 mM ADP and 0.5 mM FeC13) and rat liver microsomes (0.5 mg/ml) in chelexed 0.05 M NaCI, pH 7.0, at 37°C.

Table 3. Rates of Reduction of Ferric Chelates by the Reconstituted Mixed Function Oxidase System under

Anaerobic Conditions

Chelates nmol reduced /min /ml

EDTA-Fe3 + 117.0 DTPA-Fe 3+ 120.8 ADP-Fe 3 ÷ 6.8

Note: Anaerobic incubations contained NADPH (l raM), ferric chelate (0.55 mM chelator: 0.5 mM FeC13; except ADP- Fe 3÷ was 2.5 mM ADP, 0.5 mM FeCI3), and a reconstituted MFO system (0.8 nmol cytochrome P450b, 0.25 nmol reductase, and 48 nmol DLPC in 0.05 M NaCI, pH 7, at 37°C).

formed an electron transport chain capable of reducing ADP-Fe 3÷ anaerobically (Table 3). As with the intact microsomal system, reduction of ADP-Fe 3÷ occurred considerably more slowly than did the reduction of EDTA-Fe 3+ or DTPA-Fe 3÷.

Effect of Cytochrome P450 on NADPH-Cytochrome P450 Reductase-Dependent Peroxidation of Liposomes

The original reconstituted microsomal lipid peroxidation system of Pederson and Aust 4 required both ADP-Fe 3 + and EDTA-Fe 3 +. The results described above provide a consistent explanation for these findings: the reductase does not reduce ADP-Fe 3 + directly and therefore requires either cytochrome P450 or EDTA-Fe 3+ in order to reduce ADP-Fe 3 ÷ as the initial step towards the formation of an initiating species of lipid peroxidation. Therefore, cytochrome P450 was added to the original reconstituted system to determine if it could

substitute for EDTA-Fe 3+. Cytochrome P450 was reconstituted with the reductase in DLPC micelles and incubated with phospholipid liposomes in the presence of EDTA-Fe 3 ÷ and/or ADP-Fe 3 ÷ and the resultant rates of lipid peroxidation were monitored. No peroxidation occurred in the presence of ADP-Fe 3÷ and NADPH; both ADP-Fe 3÷ and EDTA-Fe 3÷ were necessary for MDA formation (Table 4) just as they were in the reconstituted lipid peroxidation system 4 that contained no cytochrome P450. Thus, cytochrome P450 present in DLPC micelles was unable to support the ADP-Fe 3÷ promoted peroxidation of liposomes, despite that the results of iron reduction experiments indicated that ADP-Fe 3÷ was being reduced. At first consideration, these results might appear to be inconsistent with the iron chelate requirements for peroxidation in the vesicular system. However, it may be that the site of ADP- Fe 3÷ reduction in the DLPC micelles was too remote from the more readily peroxidizable phospholipid liposomes, and therefore did not allow for sufficient

¢ -

-533 E t - -

._~ 29 O

E ~z5

, 2 1

C~3

0

r • • • • ,

J

d | | | Oi | * 1 0.2 0.3 0.4 $ 0 6 07

Microsomol Protein (mo/ml)

t Fig. 5. Effect of microsomal protein concentrat ion on NADPH-dependent ADP-Fe 3+ reduction. Incubat ions contained ADP- Fe 3* (17 mM ADP: 1 mM FeC13) and NADPH (1 mM) with various concentrat ions of microsomes ( isolated from rats pretreated with phenobarbi tal) in 50 mM NaC1, pH 7.0.

ADP-Fe3+-dependent l ipid peroxidat ion 275

Table 4. Requirements for the NADPH-Dependent Peroxidation of Phospholipid Liposomes Promoted by the Reconstituted Mixed

Function Oxidase System

MDA (nmol /min /ml )

Reconstituted MFO 0.05 + EDTA-Fe 3 + 0.00 + ADP-Fe 3 ÷ 0.04 +EDTA-Fe 3÷ + ADP-Fe 3÷ 1.32

Note: The reconstituted MFO system contained reductase (0.1 u/ml), cytochrome P450b (0.02 nmol/ml), DLPC (20 ~g/ml), phospholipid liposomes (1 I~mol lipid phosphate/ml), and NADPH (0.1 mM) in 0.05 M NaCI, pH 7.0. Cytochrome P450 isozyme was preincubated with the reduetase and DLPC for 5 min at room temperature, then 1 h on ice before use. Ferric chelate concentrations were: EDTA-Fe 3÷ (0.11 mM EDTA, 01 mM FeC13) and ADP-Fe 3+ (0.5 mM ADP, 0.1 mM FeC13).

initiation reactions to yield measurable amounts of TBA-reactive products.

D I S C U S S I O N

The results of the present study demonstrate that cytochrome P450 is required for the reduction of ADP- Fe 3÷ and the subsequent formation of an initiating species of microsomal lipid peroxidation. Coreconsti- tution of NADPH-cytochrome P450 reductase and cytochrome P450 in azolectin vesicles formed a competent electron transport system capable of initiating peroxidation in the presence of ADP-Fe 3 ÷ and NADPH. Increased rates of lipid peroxidation occurred in the presence of increased ADP-Fe 3 ÷ concentration or when vesicles having greater cytochrome P450 contents were used.

The conclusion that cytochrome P450 was required for NADPH-dependent lipid peroxidation was reached previously by Elkstrom and Ingelman-Sundberg. 14 However, since their experiments were performed in the absence of added iron chelates, it is not clear ex- actly how peroxidation was initiated. Trace levels of transition metals in their reagents may have been the catalysts for the formation of an initiating species of lipid peroxidation. Since the results presented here indicated that the rate of vesicular lipid peroxidation was proportional to the ADP-Fe 3÷ concentration, the rel- atively low rates of peroxidation they reported would be consistent with the presence of low concentrations of contaminating iron. Moreover, their observations that micromolar concentrations of the iron chelators EDTA, DTPA, and desferrioxamine inhibited peroxidation is also consistent with this interpretation. Heme proteins including cytochrome P450 and various iron chelates are capable of degrading lipid hydroperoxides. Since it is clearly difficult to demonstrate that a lipid preparation is totally free of hydroperoxides, it is dif-

ficult to exclude the iron chelate-catalyzed decomposition of pre-existing lipid hydroperoxides as the prin- ciple pathway of MDA formation in these experiments. However, the iron chelate requirements for the peroxidation of liposomes containing 10% phospholipid hydroperoxides (generated by the action of soybean lipoxidase on microsomal phospholipids) have previously been compared with those required for the peroxidation of liposomes prepared from freshly isolated phospholipids. 15 EDTA_Fe 3 + addition to partially peroxidized liposomes resulted in significant MDA formation whereas its addition to liposomes prepared from freshly isolated phospholipids resulted in no measurable MDA formation in the presence of EDTA-Fe 3÷ (Tables 1 and 4) indicates that the lipid preparations do not contain significant levels of hydroperoxides, and strongly suggests that MDA formation arises via ADP-Fe 3+-dependent initiation reactions. Although we cannot completely discount the possibility that there may be trace concentrations of hydroperoxides, the data presented in Tables 1 and 4 indicate that they do not have a significant contribution to MDA formation observed in the presence of ADP-iron.

The ability of the microsomal MFO to reduce ADP- Fe 3+ and the ability of the reductase to reduce EDTA- Fe 3+ but not ADP-Fe 3+23 are in agreement with the iron chelate requirements for the initiation of lipid peroxidation in microsomal and reconstituted systems. ADP-Fe 3 + has been shown to be the only iron chelate required for the NADPH-dependent peroxidation of microsomes. 2 In this study, microsomes were shown to reduce ADP-Fe 3+ anaerobically, albeit slowly, and the rate of iron reduction was proportional to the microsomal protein content. However, in the reconstituted lipid peroxidation system of Pederson and Aust where no cytochrome P450 had been included, no peroxidation occurred unless EDTA-Fe 3+ was also added to the incubation. 4 It is now known that the purified reductase does not reduce ADP-Fe 3÷ directly, 23 so EDTA-Fe 3÷, serving as an electron acceptor for the reductase, was proposed to serve as an intermediate electron transfer agent affecting the reduction of ADP- Fe3÷ .4 In the present reconstituted lipid peroxidation system, reduction of ADP-Fe 3 ÷ was probably rate-lim- iting as evidenced by the linear relationship between ADP-Fe 3 + concentration and MDA formation and the stimulation of the ADP-Fe3+-dependent rate of peroxidation by EDTA-Fe 3÷ .

That iron must be reduced in order to promote microsomal lipid peroxidation is the characteristic of vir- tually all mechanisms proposed for the initiation of lipid peroxidation. However, the mechanism by which iron is reduced has been a subject of controversy. Re- searchers have theorized that reduction of ferric che-


lates is via 02 ~ since microsomal lipid peroxidation was inhibited to varying degrees by SOD. "-13 Ac- cordingly, several investigators have shown that microsomes and to a lesser extent the purified reductase generate some 0 2 : . 23,29,30 Other investigators have maintained that iron reduction is direct or at least independent o f 027 since SOD did not inhibit peroxidation in their systems. 23-25

The results o f this study do not settle this issue. Peroxidation of azolectin vesicles containing the reductase and cytochrome P450 was inhibited to varying degrees by SOD. However , the higher the cytochrome P450 content in the vesicles was, the less was MDA formation inhibited by SOD. This suggests that different results obtained by investigators might be due to different cytochrome P450 contents in the microsomes, in addition to other differences in experimental conditions such as presence of various iron chelates. However , the results indicate that reduction of iron chelates by the microsomal electron transport system can proceed anaerobically and thus independently o f 02 ~ . Therefore, it is not necessary to propose that SOD is not accessible to the site of 02 = generation in order to account for the inability of SOD to inhibit peroxidation; ~3,31 it may simply be that the majority of iron reduction in certain systems is O2: -independent.

However , the reduction of ADP-Fe 3÷ (either via O2~-mediated or direct reduction) does not necessarily result in lipid peroxidation as evidenced by data presented in this investigation. The addition of cytochrome P450 and purified reductase in DLPC micelles to phospholipid l iposomes did not result in ADP-Fe 3÷ -dependent peroxidation, even though measurable rates of ADP-Fe 3 ÷ reduction occurred in similar incubations under anaerobic conditions. It is not clear if an initiator of peroxidation was in fact produced in this liposomal system, but when the enzymes are reconstituted in vesicles capable of forming TBA-reactive substances, peroxidation does occur. These results would suggest a slightly different type of site-specific mechanism for initiation of peroxidation than has previously been proposed. 32 We would suggest that the reduction of ADP- Fe 3+ may occur at discrete sites in the phospholipid vesicles where cytochrome P450 is incorporated. In juxtaposition with phospholipids containing polyunsaturated fatty acids, the NADPH-dependent reduction of ADP-Fe 3÷ and subsequent iron-oxygen chemistry results in the formation of a species that can abstract methylene hydrogens from polyunsaturated fatty acids and initiate lipid peroxidation.

It is not known whether ADP-Fe 3 ÷ reduction is fa- cilitated by a few, many or all cytochrome P450 isozymes. The potential for inducing cytochrome P450

isozymes with an increased propensity for iron reduction may have toxicological significance. The rates of peroxidation of microsomes isolated from animals treated with various cytochrome P450 inducers do not appear to vary significantly in our hands (data not shown), perhaps suggesting that iron chelate reduction may be a property of at least several isozymes. However , to ascertain this it will be necessary to incorporate various isozymes into the vesicular reconstituted system.

Acknowledgements--The expert secretarial assistance of Teresa Vollmer and Terri Maughan in the preparation of this manuscript is gratefully acknowledged. This research was supported by grants GM 33443 and HL 33543 from the National Institutes of Health.

REFERENCES

1. Hochstein, P.;Ernster, L. ADP-activatedlipidperoxidation cou- pled to the TPNH oxidase system of microsomes. Biochem. Biophys. Res. Comm. 12: 388-394; 1963.

2. Hochstein, P.; Nordenbrand, K.; Ernster, L. Evidence for the involvement of iron in the ADP-activated peroxidation of lipid in microsomes and membranes. Biochem. Biophys. Res. Comm. 14: 223-238; 1964.

3. BeloffoChain, A.; Serlupi-Cresenzi, G.; Cantanzaro, R.; Venet- tacci, D.; Balliano, M. Influence of iron on oxidation of NADPH in rat liver microsomes. Biochem. Biophys. Acta 97: 416-421; 1965.

4. Pederson, T.C.; Aust, S.D. NADPH-dependent lipid peroxidation catalyzed by purified NADPH-cytochrome c reductase from rat liver microsomes. Biochem. Biophys. Res. Comm. 48: 789-795; 1972.

5. Noguchi, T.; Nakano, M. Effect of ferrous ions on microsomal phospholipid peroxidation and related light emission. Biochem. Biophys. Acta 368: 446-455; 1974.

6. Hirokata, Y.; Shigematsu, A.; Omura, T. Immunochemical study on the pathway of electron flow in reduced nicotinamide adenine dinucleotide-dependent microsomal lipid peroxidation. J. Biochem. 83: 431-440; 1978.

7. Orrenius, S.; Dallner, G.; Ernster, L. Inhibition of the TPNH- linked lipid peroxidation of liver microsomes by drugs undergo- ing oxidative demethylation. Biochem. Biophys. Res. Comm. 14: 329-334; 1964.

8. Kuthan, H.; Tsuji, H.; Graf, H.; Ullrich, V.; Werringloer, J.; Estabrook, R.W. Generation of superoxide anion as a source of hydrogen peroxide in a reconstituted monooxygenase system. FEBS Lett. 91: 343-345; 1978.

9. Soodaera, S.K.; Skotzelyas, E.D.; Zhukov, A.A.; Archakov, A.I. Comparative studies of superoxide radical generation in microsomes and reconstituted monooxygenase systems. In: Hie- tanen, E.; Laitinen, M.; Hanninen, O., editors Cytochrome P 450, Biochemistry, Biophysics, and Environmental Implications. Amsterdam: Elsevier; 1982: 615-618.

10. Parkinson, A.; Thomas, P.E.; Ryan, D.E.; Gorsky, L.D.: Shirely, J.E.; Sayer, J.M.; Jerina, D.M.; Levin, W. Mechanism of in- activation of rat liver microsomal cytochrome P450c by 2-bromo- 4'-nitroacetophenone. J. Biol. Chem. 261:11487-11495; 1986.

11. Fong, R.L.; McCay, P.B.; Poyer, J.L.; Keele, B.B.; Misra, H. Evidence that peroxidation of lysosomal membranes is initiated by hydroxyl free radicals produced during flavin enzyme activity. J. Biol. Chem. 248: 7792-7797; 1973.

12. Lai, C.S.; Piette, L.H. Spin-trapping studies of hydroxyl radical production involved in lipid peroxidation. Arch. Biochem. Bio- phys. 190: 27-38; 1978.

13. Koster, J.F.; Slee, R.G. Lipid peroxidation of rat liver microsomes. Biochem. Biophys. Acta 620: 489-499; 1980.

14. Elkstrom, G.; Ingelman-Sundberg, M. Cytochrome P450-de-

ADP-Fe3÷-dependent lipid peroxidation 277

pendent lipid peroxidation in reconstituted membrane vesicles. Biochem. Pharmacol. 33: 2523-2525; 1984.

15. Svingen, B.A.; Buege, J.A.; O'Neal, F.O.; Aust, S.D. The mechanism of NADPH-dependent lipid peroxidation. J. Biol. Chem. 254: 5892-5899; 1979.

16. O'Brien, P.J. Intracellular mechanisms for the decomposition of a lipid peroxide. I. Decomposition of a lipid peroxide by metals ions, heme compounds, and nucleophiles. Can. J. Biochem. 47: 485-492; 1969.

17. Pederson, T.C.; Aust, S.D. Aminopyrine demethylase: Kinetic evidence for multiple microsomal activities. Biochem. Phar- macol. 19: 2221-2230; 1970.

18. Dignam, J.D.; Strobel, H.W. NADPH-cytochrome P450 reductase from rat liver: Purification by affinity chromatography and characterization. Biochemistry 16d: l116-1123; 1977.

19. Waxman, D.J.; Walsh, C. Phenobarbital-induced rat liver cytochrome P450: Purification and characterization of two closely related isozymic forms. J. Biol. Chem. 257: 10446-10457; 1982.

20. West, S.B.; Huang, M.T.; Miwa, G.T.; Lu, A.Y.H. A simple and rapid procedure for the purification of phenobarbital-in- ducible cytochrome P450 from rat liver microsomes. Arch. Biochem. Biophys. 193: 42-50; 1979.

21. Yasukochi, Y.; Masters, B.S.S. Some properties of a detergent- solubilized NADPH-cytochrome c (cytochrome P450) reductase purified by biospecific affinity chromatography. J. Biol. Chem. 251: 5337-5344; 1976.

22. Buege, J.A.; Aust, S.D. Microsomal lipid peroxidation. In: Fleischer, S.; Packer, L., editors. Methods in Enzymology, Vol- ume 52. New York: Academic Press; 1978: 302-310.

23. Morehouse, L.A.; Thomas, C.E.; Aust, S.D. Superoxide generation by NADPH-cytochrome P450 reductase: The effect of iron chelators and the role of superoxide in microsomal lipid peroxidatiou. Arch. Biochem. Biophys. 232: 366-377; 1984.

24. Kornbrust, D.J.; Mavis, R.D. Microsomal lipid peroxidation.

I. Characterization of the role of iron and NADPH. Mol. Phar- macol. 17: 400-407; 1980.

25. Noguchi, T.; Nakano, M. Effect of ferrous ions on microsomal phospholipid peroxidation and related light emission. Biochem. Biophys. Acta 368: 446-455; 1974.

26. Masters, B.S.S.; Bilimore, M.H.; Kamin, H. The mechanism of 1- and 2-electron transfers catalyzed by reduced triphospho- pyridine nucleotide-cytochrome c reductase. J. Biol. Chem. 240: 4081-4088; 1965.

27. Bilimoria, M.H.; Kamin, H. The effect of high salt concentrations upon cytochrome c, cytochrome b5 and iron-EDTA reductase activities of liver microsomal NADPH-cytochrome c reductase. Ann. N.Y. Acad. Sci. 212: 428-448; 1973.

28. Strobel, H.W.; Lu, A.Y.H.; Heidema, J.; Coon, M.J. Phos- phatidylcholine requirement in the enzymatic reduction of hemo- protein P450 and fatty acid, hydrocarbon and drug hydroxyla- tion. J. Biol. Chem. 245: 4851-5854; 1970.

29. Kharasch, E.D.; Novak, R.F. Bis(alkylamino)anthracenedione antineoplastic agent metabolic activation by NADPH-cytochrome P450 reductase and NADH dehydrogenase: Diminished activity relative to anthracyclines. Arch. Biochem. Biophys. 224: 682-694; 1983.

30. Kuthan, H.; Ullrich, V. A quantitative test for superoxide radicals produced in biological systems. Biochem. J. 203: 551- 558; 1982.

31. Mimnaugh, E.G.; Trush, M.A. Superoxide anion-dependency of NADPH-dependent rat liver microsomal lipid peroxidation as demonstrated by the inhibition of peroxidation by superoxide dismutase. In: Cohen, G.; Greenwald, R.A., editors. Oxy Rad- icals and Their Scavenger Systems. Volume 1: Molecular As- pects. New York: Elsevier; 1983: 300-303.

32. Samuni, A.; Chevion, M.; Czapski, G. Unusual copper-induced sensitization of the biological damages due to superoxide radicals. J. Biol. Chem. 256: 12632-12635; 1982.

Documents

Reconstituted microsomal lipid peroxidation: ADP-Fe3+-dependent peroxidation of phospholipid vesicles containing NADPH-cytochrome P450 reductase and cytochrome P450