an investigation into the role of hydroxyl radical in xanthine oxidase-dependent lipid peroxidation

10
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 216, No. 1, June, pp. 142-151, 1982 An Investigation into the Role of Hydroxyl Radical in Xanthine Oxidase-Dependent Lipid Peroxidation MING TIEN, BRUCE A. SVINGEN, AND STEVEN D. AUST’ Lkpartment of Biochemistry, Michigan State University, East Lansing, Michigan .68824 Received October 14,1981, and in revised form February 6,1982 A model lipid peroxidation system dependent upon the hydroxyl radical, generated by Fenton’s reagent, was compared to another model system dependent upon the en- zymatic generation of superoxide by xanthine oxidase. Peroxidation was studied in detergent-dispersed linoleic acid and in phospholipid liposomes. Hydroxyl radical gen- eration by Fenton’s reagent (FeClz + HzOz) in the presence of phospholipid liposomes resulted in lipid peroxidation as evidenced by malondialdehyde and lipid hydroperoxide formation. Catalase, mannitol, and Tris-Cl were capable of inhibiting activity. The addition of EDTA resulted in complete inhibition of activity when the concentration of EDTA exceeded the concentration of Fe 2+. The addition of ADP resulted in slight inhibition of activity, however, the activity was less sensitive to inhibition by mannitol. At an ADP to Fe2+ molar ratio of 10 to 1, 10 m&i mannitol caused 25% inhibition of activity. Lipid peroxidation dependent on the enzymatic generation of superoxide by xanthine oxidase was studied in liposomes and in detergent-dispersed linoleate. No activity was observed in the absence of added iron. Activity and the apparent mech- anism of initiation was dependent upon iron chelation. The addition of EDTA-chelated iron to the detergent-dispersed linoleate system resulted in lipid peroxidation as evi- denced by diene conjugation. This activity was inhibited by catalase and hydroxyl radical trapping agents. In contrast, no activity was observed with phospholipid li- posomes when iron was chelated with EDTA. The peroxidation of liposomes required ADP-chelated iron and activity was stimulated upon the addition of EDTA-chelated iron. The peroxidation of detergent-dispersed linoleate was also enhanced by ADP- chelated iron. Again, this peroxidation in the presence of ADP-chelated iron was not sensitive to catalase or hydroxyl radical trapping agents. It is proposed that initiation of superoxide-dependent lipid peroxidation in the presence of EDTA-chelated iron oc- curs via the hydroxyl radical. However, in the presence of ADP-chelated iron, the participation of the free hydroxyl radical is minimal. The one-electron reduction of molecular oxygen to superoxide (0;) has been dem- onstrated to occur during many biochem- ical reactions, including the autoxidation of reduced flavins (1) and ferridoxins (2) and in certain enzymatic reactions (3, 4). Superoxide dismutase (SOD)2 rapidly dis- r To whom correspondence should be addressed. ’ Abbreviations used: SOD, superoxide dismutase; BHT, butylated hydroxytoluene; MDA, malondialde- hyde; LOOH, lipid hydroperoxides; DMPO, 5,5-di- mutates 0; to H202 and ground-state ox- ygen (Reaction [l]) (5). Hydrogen peroxide is in turn metabolized by catalase and glu- tathione peroxidase (6,7). The deleterious effects of 0~ and H202 may arise in cases where the production of these forms of oxygen exceeds their rates of metabolism. This may occur under conditions of excess oxygen (8) or upon exposure to xenobiotics methyl-1-pyrrolline-N-oxide; HPLC, high-pressure liquid chromatography. 9003-9361/82/070142-10$62.00/0 Copyright Q 1982 by Academic Press, Inc. All rights of reproduction in any form reserved. 142

Upload: ming-tien

Post on 19-Oct-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 216, No. 1, June, pp. 142-151, 1982

An Investigation into the Role of Hydroxyl Radical in Xanthine Oxidase-Dependent Lipid Peroxidation

MING TIEN, BRUCE A. SVINGEN, AND STEVEN D. AUST’

Lkpartment of Biochemistry, Michigan State University, East Lansing, Michigan .68824

Received October 14,1981, and in revised form February 6,1982

A model lipid peroxidation system dependent upon the hydroxyl radical, generated by Fenton’s reagent, was compared to another model system dependent upon the en- zymatic generation of superoxide by xanthine oxidase. Peroxidation was studied in detergent-dispersed linoleic acid and in phospholipid liposomes. Hydroxyl radical gen- eration by Fenton’s reagent (FeClz + HzOz) in the presence of phospholipid liposomes resulted in lipid peroxidation as evidenced by malondialdehyde and lipid hydroperoxide formation. Catalase, mannitol, and Tris-Cl were capable of inhibiting activity. The addition of EDTA resulted in complete inhibition of activity when the concentration of EDTA exceeded the concentration of Fe 2+. The addition of ADP resulted in slight inhibition of activity, however, the activity was less sensitive to inhibition by mannitol. At an ADP to Fe2+ molar ratio of 10 to 1, 10 m&i mannitol caused 25% inhibition of activity. Lipid peroxidation dependent on the enzymatic generation of superoxide by xanthine oxidase was studied in liposomes and in detergent-dispersed linoleate. No activity was observed in the absence of added iron. Activity and the apparent mech- anism of initiation was dependent upon iron chelation. The addition of EDTA-chelated iron to the detergent-dispersed linoleate system resulted in lipid peroxidation as evi- denced by diene conjugation. This activity was inhibited by catalase and hydroxyl radical trapping agents. In contrast, no activity was observed with phospholipid li- posomes when iron was chelated with EDTA. The peroxidation of liposomes required ADP-chelated iron and activity was stimulated upon the addition of EDTA-chelated iron. The peroxidation of detergent-dispersed linoleate was also enhanced by ADP- chelated iron. Again, this peroxidation in the presence of ADP-chelated iron was not sensitive to catalase or hydroxyl radical trapping agents. It is proposed that initiation of superoxide-dependent lipid peroxidation in the presence of EDTA-chelated iron oc- curs via the hydroxyl radical. However, in the presence of ADP-chelated iron, the participation of the free hydroxyl radical is minimal.

The one-electron reduction of molecular oxygen to superoxide (0;) has been dem- onstrated to occur during many biochem- ical reactions, including the autoxidation of reduced flavins (1) and ferridoxins (2) and in certain enzymatic reactions (3, 4). Superoxide dismutase (SOD)2 rapidly dis-

r To whom correspondence should be addressed. ’ Abbreviations used: SOD, superoxide dismutase;

BHT, butylated hydroxytoluene; MDA, malondialde- hyde; LOOH, lipid hydroperoxides; DMPO, 5,5-di-

mutates 0; to H202 and ground-state ox- ygen (Reaction [l]) (5). Hydrogen peroxide is in turn metabolized by catalase and glu- tathione peroxidase (6,7). The deleterious effects of 0~ and H202 may arise in cases where the production of these forms of oxygen exceeds their rates of metabolism. This may occur under conditions of excess oxygen (8) or upon exposure to xenobiotics

methyl-1-pyrrolline-N-oxide; HPLC, high-pressure liquid chromatography.

9003-9361/82/070142-10$62.00/0 Copyright Q 1982 by Academic Press, Inc. All rights of reproduction in any form reserved.

142

HYDROXYL RADICAL-DEPENDENT LIPID PEROXIDATION 143

which undergo cyclic reduction and autox- idation, such as paraquat (9), 6-hydroxy- dopamine (lo), or adriamycin (11, 12). Lipid peroxidation has been suggested to result from excess 0; production (9-12). However, the reactive form(s) of oxygen involved in OH-dependent lipid peroxida- tion has not been unequivocally estab- lished.

Superoxide has been demonstrated not to be exceptionally reactive with a large number of organic substrates, including polyunsaturated fatty acids (13, 14). The reduction of transition metals such as iron by 05 has been demonstrated (15) and the reaction of ferrous iron with HzOz, pro- duced by 0s dismutation, is known to pro- duce the very reactive hydroxyl radical (. OH) (16). The sequence of reactions de- scribed (Reactions [2], [3]) is called the iron-catalyzed Haber-Weiss reaction:

2H+ + 205 - HzOz + 02, PI 0; + Fe3+ - Fe’+ + O2 , PI

Fe’+ + HzOz - Fe3+ + OH- + . OH. [3]

Thus, the O;-dependent production of + OH and its involvement in lipid peroxi- dation has been proposed by several in- vestigators both in the case of xanthine oxidase-promoted lipid peroxidation (14, 17-19) and for microsomal NADPH-de- pendent lipid peroxidation (14, 20, 21). In this report Fe’+ and H202, termed the Fen- ton’s reagent (Reaction [3]) has been uti- lized to initiate lipid peroxidation. The response of this . OH-dependent lipid peroxidation system to the presence of iron chelation agents, and to known in- hibitors of . OH-dependent reactions has been characterized and compared to the results of similar studies utilizing a lipid peroxidation system dependent upon the superoxide anion.

MATERIALS AND METHODS

Materials, Cyctochrome c (Type VI), 2-thiobarbi- turic acid, NADPH, ADP, BHT, mannitol, Lubrol, and sodium benzoate were purchased from Sigma Chemical Company. Linoleic acid was obtained from Nu Chek Prep, Elysian, Minnesota. All other chem- icals used were of analytical grade. All buffers and

reagents were passed through Chelex 100 (Bio-Rad Laboratories) ion-exchange resin to free them of con- taminants.

Enqmes. Bovine erythrocyte SOD (2969 unita/mg protein), beef liver catalase (30,996 unita/mg pro- tein), and xanthine oxidase (0.9 unit/mg protein) were obtained from Sigma Chemical Company. Gel filtration chromatography on Sephadex G-25 was used to remove the antioxidant thymol from catalase and the ammonium sulfate from xanthine oxidase. After chromatography, xanthine oxidase activity was measured by aerobic reduction of cytochrome c (22) and a unit of activity is defined as 1 pmol cy- tochrome c reduced/min. SOD activity was measured by the method of McCord and Fridovich (22). Cata- lase activity was measured by the procedure of Holmes and Masters (23) and a unit of activity is defined as 1 rmol HzOz decomposed/min.

hficrosomal lipid. Male Sprague-Dawley rats (250- 274 g) were obtained from Spartan Research Animals (Haslett, Mich.). Liver microsomes were isolated by the method of Pederson et al. (24). Microsomal lipid was extracted from freshly isolated microsomes by the method of Folch et aE. (25). All solvents used for extractions were purged with argon and all steps were performed at 4°C to minimize autoxidation of unsaturated lipids. Extracted lipid was stored in ar- gon-saturated CHCls:CHzOH (2:l) at -20°C. Lipid phosphate determinations were performed by the method of Bartlett (26).

Reaction mixtures. Liposomes were prepared by sonication of the extracted microsomal lipid in ar- gon-saturated distilled deionized water at 4°C (24). Liposomal peroxidation reactions initiated by the Fenton’s reagent were accomplished by the addition of FeCle to incubations containing HzOc and lipo- somes (1 pmol lipid phosphate/ml) in 36 mM NaCl, pH 7.5. The concentrations of FeClz and HeOz are as specified in the figures and tables. Other additions or deletions are as specified in the legends to the figures and tables. These reactions were initiated by the addition of FeClz. Xanthine oxidase-dependent peroxidation of liposomes was performed by incu- bating liposomes (1 pmol lipid phosphate/ml) with xanthine oxidase (0.1 unit/ml), EDTA-Fe’+ (0.11 mM

EDTA, 0.1 mM FeCL), ADP-Fe’+ (1.7 mM ADP, 0.1 mM FeC13) and 0.33 mM xanthine in 30 mM NaC1, pH 7.5. Reactions were initiated by the addition of xan- thine oxidase. Other additions or deletions are as specified in the table and figure legends. Incubations were done in a metabolic shaking water bath at 37°C under an air atmosphere. Although the reaction mix- tures were unbuffered, the pH did not change during the course of the reactions. At 0, 3, and 6 min, xan- thine oxidase-dependent lipid peroxidation reaction mixtures were sampled for MDA and LOOH content (27). The rates of peroxidations were linear within this time span. The MDA content was determined by

144 TIEN, SVINGEN, AND AUST

the thiobarituric acid test (27). To prevent further peroxidation of lipid during the assay procedure for MDA, 0.03 vol of 2% BHT in ethanol was added to the thiobarbituric acid reagent (27). The LOOH con- tent was determined iodometrically (27).

Linoleate stock solutions were made by suspending linoleic acid (100 mg) into argon-purged Chelexed water by adding 5-10 drops of 6 N NaOH to yield a clear solution. The pH of this solution was then slowly lowered to 7.5 by the addition of 6 N HCl. The resultant micelle solution was taken up to a final volume of 50 ml and used immediately. Lipid per- oxidation of detergent-dispersed linoleate was ac- complished by incubating sodium linoleate (5.7 mM) with xanthine oxidase (6.6 munit/ml), Lubrol (l%), 35 rnM acetaldehyde, EDTA-Fe’+ (0.11 mM EDTA, 0.1 mM FeClv), and ADP-Fe*+ (0.5 mM ADP, 0.1 mM FeC&) in 30 mu NaCl, pH 7.5, at 3’7’C, under air. Other additions or deletions are as indicated in the table legends. Reactions were initiated by the addi- tion of xanthine oxidase. These incubations were car- ried out in a cuvette in a Cary 219 spectrophotometer. Diene conjugation during lipid peroxidation was con- tinuously monitored by the absorbance change at 234 nm (23). The rate of diene conjugation was linear for up to approximately 5 min, after which a decrease in rate was observed. All rates were determined from the initial velocity of the reaction. The data shown are results of representative experiments.

Otti methods ADP-chelated and EDTA-chelated iron solutions were prepared by the addition of either FeCla or FeCla to chelate solutions adjusted to pH 7.5. Due to the pH change from the addition of FeC&, the pH of these chelate solutions was readjusted to

FIG. 1. Time course of hydroxyl radical-dependent lipid peroxidation. Reaction mixtures contained li- posomes (1 pmol lipid phosphate/ml), 0.1 rnre HaOa, and 0.2 rnlld FeCla in 30 ml NaCl. pH 7.5, at 37°C. Incubation and assay conditions are described under Materials and Methods.

6 t

FIG. 2. Effect of hydrogen peroxide concentration on hydroxyl radical-dependent lipid peroxidation. Reaction mixtures contained liposomes (1 pmol lipid phosphate/ml), 0.2 mM FeC&, and the specified amount of HeOr in 30 mM NaCl, pH 7.5, at 37°C. Incubation and assay conditions are described under Materials and Methods.

7.5. No changes in pH were observed from the ad- dition of FeCla to the chelate solutions. Water used in the preparation of the ferrous solutions was argon purged to minimize Fee+ autoxidation. Since most buffers are iron chelators or reactive with hydroxyl radical, the reaction mixtures were not buffered. Con- sequently, the pH of all reagents was carefully ad- justed to 7.5 prior to use.

RESULTS

Hgdmxyl Radicul-Dependent tipid

The reaction of H202 with Fe2+ (Fenton’s reagent) has been demonstrated to pro- duce the . OH (15). The formation of this radical has been demonstrated with EPR spin-trapping techniques utilizing DMPO (29). The EPR spectrum has a 1:2:2:1 signal intensity pattern, a g value of 2.006 and hyperfme splitting constants of AN = 14.95 G and AH = 14.95 G. To confirm the for- mation of the l OH in our system, this spectrum was reproduced by the reaction of 0.15 mM Fe2+ and 0.1 InM H202 in the presence of 60 mM DMPO (not shown). The generation of the . OH in the presence of liposomes resulted in lipid peroxidation as evidenced by a rapid rate of LOOH and MDA formation (Fig. 1). The rate of MDA

HYDROXYL RADICAL-DEPENDENT LIPID PEROXIDATION 145

[FeCI~ hM)

FIG. 3. Effect of FeCle concentration on hydroxyl radical-dependent lipid peroxidation. Reaction mix- tures contained liposomes (1 pmol lipid phosphate/ ml), 0.1 mb% HsO, and the specific amount of FeCle in 30 mM NaCI, pH 7.5, at 37°C. Incubation and assay conditions are described under Materials and Meth- ods.

formation was constant at 6.24 nmol/min/ ml up to 2 min after initiation of the re- action with Fe2’. The LOOH content, de- termined iodometrically (27) increased at a much greater rate (approximately 150 nmol/min/ml), but the rate of LOOH for- mation decreased after the first minute of the reaction. No detectable MDA or LOOH formation resulted from the separate ad- dition of FeC12 or H202 to liposomes.

The effect of H202 concentration on the rate of lipid peroxidation is illustrated in Fig. 2. At 0.2 mM FeC12, the rate of MDA formation reached a maximum when H202 concentration was about 0.1 mM. Concen- trations of Hz02 greater than 0.1 mM caused a decrease in the rate of lipid per- oxidation. At a constant H202 concentra- tion of 0.1 mM, increasing FeC12 concen- trations to about 0.1 mM resulted in a linear increase in the rate of MDA for- mation (Fig. 3). Higher concentrations of FeC12 had no effect on the rate of lipid peroxidation.

Inhibition by mannitol or benzoate has been used as a criterion to assess . OH in- volvement in reaction mechanisms (14,17- 19). As shown in Table I, peroxidation of liposomes by Fenton’s reagent could be effectively inihibited by these a OH traps.

At 10 InM, mannitol inhibited activity by 95% while 10 mM benzoate caused 88% inhibition. Tris-Cl is also an effective . OH trap (30) and 10 mbi Tris-Cl caused 92% inhibition of activity. Addition of 1 unit/ml catalase inhibited activity 93% (Table I) while boiled catalase had no ef- fect.

Most model lipid peroxidation systems employ iron chelates. In addition, biolog- ical systems are abundant in substances which can chelate iron. For these reasons, the effect of iron chelation on *OH-de- pendent lipid peroxidation was examined. ADP and EDTA were examined because both have been implicated in Fenton-type reactions to yield the . OH (14,16,31). The chelation of Fe2’ by EDTA in Fenton’s reagent resulted in inhibition of MDA for- mation (Fig. 4). Complete inhibition oc- curred when the ratio of EDTA to Fe2$ was equimolar. The chelation of Fe’+ by ADP resulted in slight inhibition of activ- ity (Fig. 5). Activity in the absence of ADP could be completely inhibited by mannitol but as the concentration of ADP in- creased, the amount of activity inhibited by mannitol decreased, reaching 25% in- hibition at an ADP to Fe’+ molar ratio of 10 to 1.

TABLE I

HYDROXYL RADICAL DEPENDENT PEROXIDATION OF PHOSPHOLIPID LIPOKIMES

Addition

None Fe2+

Hz02

Fez+ H 0 2 2

Fe2+’ H202, benzoate Fe’+: Hz02 mannitol Fe2+, H20d Tris-Cl Fe*+, HrOs, catalase

MDA (nmol/min/mI)

0.01 0.15 0.10 3.45 0.41 0.22 0.29 0.25

Note. Liposomes (1 pmol lipid phosphate/ml) were incubated in 30 mild NaCl, pH 7.5, at 37°C. The fol- lowing additions were made as indicated: 0.2 rnlu FeC12, 0.1 mM HrO, 1 unit catalase/ml, and 10 rnhf benzoate, mannitol, or Tris-Cl. Incubation and assay conditions are described under Materials and Meth- ods.

146 TIEN, SVINGEN, AND AUST

Xanthine Oxiclase-Dependent Pw~tim of Lipom?nes

The requirements for xanthine oxidase- dependent peroxidation of liposomes are shown in Table II. In the absence of iron, the generation of Op and HzO, by the ac- tion of xanthine oxidase on xanthine did not result in lipid peroxidation as assayed by LOOH and MDA formation. The ad- dition of chelated iron to this reaction mixture caused a linear increase in both MDA and LOOH content. However, this increase was exceedingly sensitive to the iron chelators. Relatively low rates of lipid peroxidation were obtained by the addi- tion of Fe3+ chelated by EDTA (Table II). The inclusion of Fe3+ chelated by ADP re- sulted in low levels of MDA and LOOH formation. Much greater rates of lipid per- oxidation (2.02 nmol MDA/min/ml and 15.3 nmol LOOH/min/ml) occurred when both ADP-Fe3+ and EDTA-Fe3+ were in- cluded in the reaction mixture. The effect of unchelated Fe3+ on xanthine oxidase- dependent peroxidation of liposomes was also examined, however, no activity was observed.

The involvement of the . OH in xanthine

0 0 0.2 0.4 06 0.8 lo 12

EDTA/Fe* mob ratio

FIG. 4. Inhibition of hydroxyl radical-dependent lipid per-oxidation by EDTA. Reaction mixtures con- tained liposomes (1 pmol lipid phosphate/ml) and 0.1 maa HaOx in 30 mM NaCl, pH 7.5, at 3’7°C. EDTA concentration is expressed as molar ratios to FeCla. The FeClz concentrations were either 0.05 mbi (open circles) or 0.1 mM (closed circles). Incubation and assay conditions are described under Materials and Methods.

ADPI& mdor mtio

FIG. 5. Effect of ADP on hydroxyl radical-depen- dent lipid peroxidation. Reaction mixtures contained liposomes (1 pmol lipid phosphate/ml), 0.1 mM H,Or, 0.2 mbd FeCla, and the specified amount of ADP in 30 mM NaCl, pH 7.5, at 3’7°C. ADP concentration is ex- pressed as molar ratios of FeClr. Incubations were performed in the presence (open circles) and absence (closed circles) of 10 mM mannitol. The precentage inhibition by mannitol as a function of ADP is plot- ted as open triangles. Incubation and assay condi- tions are described under Materials and Methods.

oxidase-dependent peroxidation of lipo- somes was investigated by the criteria utilized in investigating peroxidation of liposomes initiated by the Fenton’s re- agent. Mannitol (10 mM) and Tris-Cl (50 mM), both very effective inhibitors of lipid peroxidation initiated by the Fenton’s re- agent, were without effect in lipid perox- idation promoted by xanthine oxidase (Table III). SOD (1 unit/ml) was an effec- tive inhibitor of lipid peroxidation, pre- sumably by blocking the reduction of iron by 0;. However, catalase (1 unit/ml) did not inhibit the rate of lipid peroxidation, suggesting that the involvement of Hz02 in xanthine oxidase-dependent lipid per- oxidation is minimal.

The direct addition of HzOz caused a sharp decrease in the rates of MDA and LOOH formation (Fig. 6). Xanthine oxi- dase activity, as measured by uric acid for- mation, was also decreased by HzOz, how- ever, this decrease was minimal and did not correlate with the decrease in the rate of lipid peroxidation. Hydrogen peroxide had no effect on the assays for MDA or LOOH (not shown).

HYDROXYL RADICAL-DEPENDENT LIPID PEROXIDATION 147

TABLE II

EFFECT OF IRON CHELATES ON XANTHINE OXIDASE-DEPENDENT PEROXIDATION

OF PHOSPHOLIPID LIPOSOMES

nmol/min/ml

Addition MDA LOOH

None 0.02 0.1 ADP-Fe=‘, EDTA-Fe’+ 0.02 0.4 Xanthine oxidase 0.02 0.6 Xanthine oxidase, ADP-Fe8+ 0.29 2.2 Xanthine oxidase, EDTA-Fe’+ 0.18 1.0 Xanthine oxidase, ADP-Fe’+, EDTA-Fe*+ 2.02 15.3

Note. Reaction mixtures contained lipnsomes (1 rmol lipid phosphate/ml) and 0.33 mM xanthine in 30 mM NaCl, pH 7.5, at 3’7°C. The following additions were made as indicated: ADP-Fe*+, (1.7 mM ADP, 0.1 my Fe&), EDTA-Fea+ (0.11 mbi EDTA, 0.1 mnd FeCl& and 0.1 unit xanthine oxidase/ml. Incubation and assay conditions are described under Mate- rials and Methods.

Xanthine O&&se-Dependent Peroxidation of Detergent-Dispersed Linokate

In a recent report, Fridovich and Porter (19) demonstrated that EDTA-Fe3+ facil- itated the xanthine oxidase-dependent peroxidation of detergent-dispersed ar- achidonate. Our results with liposomes in- dicated that EDTA-Fe3+ was not effective in catalyzing xanthine oxidase-dependent lipid peroxidation. For these reasons, xan- thine oxidase-dependent lipid peroxida- tion was also studied using detergent-dis- persed linoleate as the substrate. The extent of lipid peroxidation was monitored by the increase in absorbance at 234 nm due to conjugated diene formation. The formation of conjugated diene hydroper- oxides, and the corresponding increase in absorbance at 234 nm was verified by HPLC analysis of the products. The ex- tracted products of xanthine oxidase-de- pendent peroxidation of linoleate were shown to have similar retention times as linoleate hydroperoxides generated by soybean lipoxidase (not shown). Due to the intense ultraviolet absorbance of xanthine and ADP, modification of the experimen- tal conditions was necessary to monitor diene conjugation. Acetaldehyde was uti-

lized as the substrate for xanthine oxidase and the ADP concentration of ADP-Fe3+ solutions was lowered from 1.7 to 0.5 mM.

Further studies with xanthine oxidase-de- pendent peroxidation of liposomes dem- onstrated that this change in ADP con- centration did not affect the rate of peroxidation (not shown).

Table IV shows the effect of different iron chelates on the peroxidation of de- tergent-dispersed linoleate. Again, no ac- tivity was observed in the absence of added iron. In contrast, to the results with liposomes, the addition of EDTA-Fe3+ re- sulted in a linear increase in diene con- jugation of dispersed linoleate (AAW = O.O70/min/ml). The rate of diene con- jugation was slightly higher in the pres- ence of ADP-Fe3+ (ti= = O.O85/min/ml) while the addition of both iron complexes resulted in maximal rates of lipid perox- idation (AAW = O.O92/min/ml). The ef- fects of SOD, catalase, and various . OH trapping agents on xanthine oxidase-de- pendent lipid peroxidation were studied in the presence of EDTA-Fe3+, ADP-Fe3+, or EDTA-Fe3+ plus ADP-Fe3+. The results are summarized in Table V. It is clear that

TABLE III

EFFECT OF VARIOUS SCAVENGERS ON XANTHINE OXIDASE-DEPENDENT PEROXIDATION OF

PHOSPHOLIPID LIPOSOMES

nmol/min/ml

Addition MDA LOOH

None 2.02 15.3 Catalase 2.23 17.3 SOD 0.03 5.0 Mannitol 2.32 14.5

Tris-Cl 2.25 17.1

Note. The complete system contained liposomes (1 pmol lipid phosphate/ml), ADP-Fe*+ (1.7 mM ADP, 0.1 mM FeCls), EDTA-Fe3’ (0.11 mM EDTA, 0.1 mM

FeC&), 0.33 mM xanthine, and 0.1 unit xanthine ox- idase/ml in 30 mM NaCI, pH 7.5, at 37’C. The fol- lowing additions were made as indicated: 1 unit ca- talaselml, 1 unit SOD/ml, 10 rnn! mannitol, and 50

mM Tris-Cl. Incubation and assayed conditions are described under Materials and Methods.

143 TIEN. SVINGEN. AND AUST

SOD (10 units/ml) was a potent inhibitor of lipid peroxidation in all three systems. Essentially no inhibition was observed with catalase (1,5 and 10 units/ml), man- nitol (10 and 50 mM), or ethanol (10 and 50 DIM) on lipid peroxidation in the pres- ence of ADP-Fe3+ while significant inhi- bition occurred in the presence of EDTA- Fe3+. These agents caused only modest inhibition of lipid peroxidation in the presence of both iron complexes.

DISCUSSION

The results of these experiments dem- onstrate that peroxidation of liposomes can be initiated by . OH generated by Fen- ton’s reagent. EPR spin-trapping experi- ments with DMPO confirmed the forma- tion of - OH by Fe2+ and H,Oz. The addition of mannitol or benzoate effectively inhib- ited lipid peroxidation of liposomes initi- ated by Fenton’s reagent. The involvement of H202 was illustrated by inhibition of lipid peroxidation by catalase. The rate of lipid peroxidation was highly sensitive to both FeCl, and H20z concentrations. High concentrations of H202 caused a slight in- hibition of lipid peroxidation. This may be due to the dual role of H202 as a substrate

I 1

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.9

[H&d (m)

FIG. 6. Inhibition of xanthine oxidase-dependent lipid peroxidation by HzOe. Reaction mixtures con- tained liposomes (1 rmol lipid phosphate/ml), ADP- FeS+ (1.7 mM ADP, 0.1 mbf FeC18), EDTA-Fe’+ (0.11 mM EDTA, 0.1 mM FeCl& 0.33 m&f xanthine, 0.1 unit xanthine oxidase/ml, and the specified amount of HeOz in 30 mM NaCl, pH 7.5, at 3’7’C. Incubation and assay conditions are described under Materials and Methods.

TABLE IV

EFFECT OF IRON CHELATES ON XANTHINE OXIRASE- DEPENDENT PEROXIDATION OF DETRRGENT-

DISPERSED LINOLEATE

Addition M&min/ml

None <0.001 ADP-Fe?+ <o.cm EDTA-Fe”+ <0.001 ADP-Fe”+, EDTA-Fe”+ <O.OOl Xanthine oxidaae <O.Ool Xanthine oxidaae, ADP-Fe* 0.085 Xanthine oxidase, EDTA-Fee+ 0.070 Xanthine oxidaae, ADP-Fe”, EDTA-FL?+ 0.082

Note. Reaction mixtures contained 5.7 mM sodium lino- ieate, 1% Lubroi, and 35 mM acetaldehyde in 30 rnlu NaCl, pH 7.5. The following additions were made ae indicated: ADP-Fee+ (0.5 mM ADP, 0.1 rnM FeCI.&, EDTA-Fee+ (0.11 rnM EDTA, 0.1 mM Fe&), and xanthine oxidaae (6.6 X lo-‘unit/ ml). Diene conjugation was assessed by continuous monitor of increase in 234-nm absorbance. Incubation and assay con- ditions are described under Materials and Methods.

for . OH formation and also as a . OH trap. The rate of constant for . OH react- ing with H202 is 4.5 X 10” M-’ s-l (32):

- OH + Hz02 --) Hz0 + 0; + H+. [4]

Thus, at high concentrations, H202 may be competing with polyunsaturated lipid for the *OH resulting in decreased rates of lipid peroxidation. Ferrous ion has also been demonstrated to be highly reactive with . OH, resulting in a termination re- action (32). Although this reaction,

Fe2+ + * OH - Fe3+ + OH-, r51 proceeds with a second-order rate con- stant of 3 X 10’ M-’ s-’ (32), high Fe2+ con- centrations did not inhibit lipid peroxi- dation under the experimental conditions employed. The participation of Fe” in sec- ondary initiation reactions with LOOH, thus augmenting the rates of lipid per- oxidation (33), may account for the lack of inhibition observed at high FeC12 con- centrations.

The effect of iron chelation by EDTA or ADP on the Fenton’s reagent was exam- ined since both these chelates have been used to study xanthine oxidase-dependent lipid peroxidation (14,19, 31). EDTA che- lation of Fe3+ has been reported to facil-

HYDROXYL RADICAL-DEPENDENT LIPID PEROXIDATION 149

TABLE V

EFFECT OF VARIOUS SCAVENGERS ON XANTHINE OXIDASE-DEPENDENT PEROXIDATION OF DETERGENT-DISPERSED LINOLEATE

Percentage inhibition

Addition

None 10 units/ml SOD 1 unit/ml Catalase 5 units/ml Catalase 10 units/ml Catalase 10 m?d Mannitol 50 rnM Mannitol 10 mM Ethanol 40 mM Ethanol

EDTA-Fe3+ ADP-Fe3+

- -

(0.070) (0.085) 91 90 20 3 37 3 40 0 19 0 31 0 26 1 37 2

EDTA-Fe3+ ADP-Fea+ , -

(0.092) 8.3

0 3 4 8 9 4 9

Note. The effect of SOD, catalase, and various * OH trapping agents were studied on lipid peroxidation in the presence of EDTA-Fe ‘+ ADP-Fe’+, or EDTA-Fee’ plus ADP-Fea+. The iron chelate concentrations were , 0.11 mM EDTA, 0.1 mM FeC13, and 0.5 mM ADP, 0.1 m&f FeCls. All reaction mixtures contained 5.7 rnrd sodium linoleate, 1% Lubrol, 35 mM acetaldehyde, and 6.6 X 10v3 unit/ml xanthine oxidase in 30 mM NaCl, pH 7.5. Additions were made as indicated. Rates are expressed as the percentage inhibition. The value in parenthesis represent the initial velocities of the reaction expressed in PA,/min/ml. Diene conjugation was assessed by continuous monitor of increase in 234-nm absorbance. Incubation and assay conditions are described under Materials and Methods.

itate catalysis of the Haber-Weiss reac- tion (1534). Catalysis of the Haber-Weiss reaction involves initial reduction of Fe3+ by 05 followed by reduction of Hz02 by the reduced iron. Our results show that the latter reaction, the Fenton’s-type reaction, is inhibited by EDTA chelation of Fe2+. Further studies on the autoxidation of Fez+ (measured by oxygen consumption using a Clark electrode (indicated that EDTA chelation resulted in a rapid au- toxidation of Fe’+. Thus the inhibition of * OH-dependent lipid peroxidation ob- served by EDTA chelation of Fe*+ is most likely the result of a decrease concentra- tion of Fe*+ due to autoxidation. The ob- served enhancement of the iron catalyzed Haber-Weiss reaction by EDTA (15, 34) is probably due to facilitation of the re- duction of Fe3’ by 0;.

The inclusion of ADP in the Fenton’s reagent caused a slight decrease in the rate of lipid peroxidation. This slight de- crease could also be attributed to autoxi- dation of Fe2+ facilitated by ADP. Oxygen consumption experiments indicated that ADP chelation also enhanced the rate of

Fe’+ autoxidation, however, these rates were much slower than those observed with EDTA (not shown). Although purine nucleotides have been shown to react with the . OH (35), 10 mM ADP only caused a 10% decrease in activity. In addition, lipid peroxidation initiated by FeCl, and H20e in the presence of ADP was less sensitive to mannitol inhibition. These results sug- gest that ADP chelation of Fe’+ results in lipid peroxidation that is not dependent upon the . OH. However, it is possible that ADP reacts with the *OH to yield a rad- ical product that does not react with man- nitol but is capable of initiating lipid per- oxidation. These possibilities are currently under investigation.

In accord with the report of Lai and Piette (31), our results indicate that in the presence of EDTA-Fe3’, initiation of xan- thine oxidase-dependent lipid peroxida- tion occurs via the *OH. In a ho- mogeneous system of detergent-dispersed linoleate, the EDTA-Fe’+-dependent gen- eration of the . OH resulted in lipid per- oxidation as evidenced by diene conjuga- tion. The inhibition of activity by SOD,

150 TIEN, SVINGEN, AND ATJST

catalase, mannitol, and ethanol suggests that initiation occurs via the iron-cata- lyzed Haber-Weiss reaction. However, in liposomes, the generation of 0; and H202 by the action of xanthine oxidase and xan- thine in the presence of EDTA-Fe3+ did not result in lipid peroxidation. This dif- ference in activity would suggest a cor- relation between proximity of . OH gen- eration and its lipid peroxidation-initiating activity. In the case of liposomes, l OH generation would probably occur in the aqueous phase. It is unlikely that the highly reactive *OH would diffuse away from the site of formation to liposomes to initiate lipid peroxidation (16). Dispersion of the lipid with detergent to yield a ho- mogeneous phase would bring the lipid in closer proximity to the site of . OH for- mation, thus resulting in lipid peroxida- tion.

In contrast to the results with EDTA- Fe3+, . OH participation in xanthine oxi- dase-dependent lipid peroxidation appears minimal in the presence of ADP-Fe3+. Cat- alase and *OH trapping agents, at con- centrations which caused complete in- hibiton of Fenton’s reagent-dependent peroxidation of liposomes, were without effect on xanthine oxidase-dependent per- oxidation of liposomes in the presence of ADP-Fe3+. However, SOD was an effective inhibitor of xanthine oxidase-dependent peroxidation of liposomes causing a 97% decrease in the rate of MDA formation and a 67% decrease in the rate of LOOH formation. The addition of H202 to xan- thine oxidase-dependent peroxidation of liposomes actually resulted in decreased rates of peroxidation. In the detergent- dispersed linoleate system, the concentra- tions of catalase and . OH trapping agents which caused significant inhibi- tion of EDTA-Fe3+-dependent peroxidation were not effective in the presence of ADP- Fe3+. In liposomes and detergent-dis- persed linoleate, the addition of both EDTA-Fe3+ and ADP-Fe3+ resulted in the maximal rate of lipid peroxidation. Again, catalase and -OH trapping agents were without effect. In could be argued that ADP-Fe3+ localizes the generation of the . OH closer to the lipid such that water-

soluble trapping agents cannot effectively scavenge the *OH. This would suggest an affinity of ADP for the lipid phase. How- ever, preliminary experiments where li- posomes and ADP-Fe3+ were incubated prior to filtration to separate the lipo- somes from the aqueous phase showed no evidence for enrichment of iron in the lipid phase as assayed by atomic plasma emis- sion spectroscopy. Moreover, detergent dispersion of the lipid phase into a ho- mogeneous system did not render ADP- Fe3+-dependent peroxidation more sus- ceptable to catalase or . OH trapping agents.

Our results on xanthine oxidase-depen- dent lipid peroxidation in the presence of ADP-Fe3+ are in agreement with the re- sults of Svingen et al. (36) and Tyler (13). Svingen and co-workers proposed that ini- tiation of lipid peroxidation occurred via an ADP-iron-oxygen complex. Although we have no direct evidence for such a com- plex, our results support this possibility. In addition, these workers proposed that enhancement of lipid peroxidation by EDTA-Fe3+ is a result of EDTA-Fe3+ (upon reduction to EDTA-Fe2+ by OH)-catalyzed decomposition of LOOH resulting in accelerated peroxidation through lipid hydroperoxide-dependent initiation reac- tions. Although our results and the results of others (19, 31) suggest that the role of EDTA-Fe’+ in xanthine oxidase-depen- dent lipid peroxidation is catalysis of the Haber-Weiss reaction, it is probable that once lipid hydroperoxides are formed, EDTA-Fe2+ acts to reductively cleave these lipid hydroperoxides as it can with H202. This may account for the high rates of lipid peroxidation observed in the pres- ence of both EDTA-Fe’+ and ADP-Fe3+.

Fong and co-workers (14) also studied xanthine oxidase-dependent lipid peroxi- dation in the presence of ADP-Fe3+. These workers reported that SOD stimulated while catalase and . OH trapping agent inhibited O;-induced lysis of lysosomes. Aside from different methods used for assessing lipid peroxidation, we have no apparent explanation for the discrepancy in results. Our results with a model system of ADP-Fe3+ and liposomes, which we

HYDROXYL RADICAL-DEPENDENT LIPID PEROXIDATION 151

would anticipate to have more biological significance than EDTA-Fe3+ with deter- gent-dispersed linoleate, indicate that 05 -dependent lipid peroxidation proceeds through a mechanism not dependent upon the *OH. Although the exact nature of this initiating complex remains unknown, thermodynamic considerations require that this initiation complex be of oxidizing strength equivalent to that of the . OH. However, it is apparent that the genera- tion of free . OH is not required for its lipid peroxidation-initiating activity.

ACKNOWLEDGMENTS

We would like to thank Dr. John Bucher for as- sistance in preparation of this manuscript. The teeh-

nical assistance of John Vitkuske is greatly appre-

ciated. We would also like to acknowledge the secretarial assistance of Cathy M. Custer. Supported by NSF Grant PCM 79-15328, Michigan Agricultral

Experiment Station Journal Article 10019.

REFERENCES

1. MCCORD, J. M., AND FRIDOVICH, I. (1970). J. Bid

Chem 245,1374-1377.

24. PEDERSON, T. C., BUEGE, J. A., AND AUST, S. D. (1973) J. Biol. Chem 248, 7134-7141.

25. FOLCH, J., LEES, M., AND SLOANE-STANLEY, G. H. (1957) J. Biol. C&m. 226,497-509.

26. BARTLETT, G. R. (1959) J. Biol Chem 234, 466- 468.

2. MISRA, H. P., AND FRIDOVICH, I. (1971) J. Bi01

Chem 246, 68866890. 3. FRIDOVICH, I., AND HANDLER, P. (1958) J. Bid

Chem 233,1581-1585.

27. BUEGE, J. A., AND AUST, S. D. (1978) in Methods in Enzymology, (Fleischer, S., and Packer, L.,

eds.) Vol. 52, Part C, pp. 302-310, Academic

Press, New York. 4. BRADY, F. O., FORMAN, H. J., AND FERGELSON, P.

(1971) J. Bid Chem. 246, ‘7119-7124. 5. MCCORD, J. M., AND FRIDOVICH, I. (1968) J. Bid

Chem 243,5753-5760.

6. SIES, H., GERSTENECKER, C., MENZEL, H., AND FLOHE, L. (1972). FEBS Lett. 27,171-175.

7. SIES, H., GERSTENECKER, C., SUMMER, K. H.,

MENZEL, H., AND FLOHE, L. (1974) in Gluta-

thione (Flohe, L., Benohr, H. Ch., Sies, H., Waller, H. D., and Wendel, A., eds.), pp. 261-

276, Thieme, Stuttgart.

28. ADAMS, E. G., AND WARDMAN, P. (1977) in Free

Radicals in Biology (Pryor, W. A., ed.), Vol. 3,

pp. 53-95, Academic Press, New York. 29. FINKELSTEIN, E., ROSEN, G. M., AND RAUCKMAN,

E. J. (1980) Arch Biochem Biophys. 299,1-16.

30. TIEN, M., SVINGEN, B. A., AND AUST, S. D. (1981) Fed Proc. 40,179-182.

31. LAI, C. S., AND PIETTE, L. (1978). Arch B&hem Biophys. 190,27-38.

8. KISTLER, G. S., CALDWELL, P. R. B., AND WEIBEL,

E. R. (1967) J. Cell. BioL 32,605-628. 9. Bus, J. S., AUST, S. D., AND GIBSON, J. E. (1974)

Biochem Biqhys Res. Commun 58.749-755. 10. HEIKKILA, R. E., AND COHEN, G. (1973) Science

181.456-457.

32. ANBAR, M., AND NETTA, P. (1967) Znt. J. Appl Radiat. Zsotq. 18,493-523.

33. SOSNOVSKY, G., AND RAWLINSON, D. J. (1971) in, Organic Peroxides (Swern, D., ed.), Vol. II, pp.

153-268, Wiley-Interscience, New York.

34. HALLIWELL, B. (1978) FEBS L&t. 92,321-326. 35. SCHOLES, G., WARD, J. F., AND WEISS, J. (1960)

J. Mol. Biol 2.379-391. 11. GOODMAN, J., AND HOCHSTEIN, P. (1977) Biochem 36. SVINGEN, B. A., O’NEAL, F. O., AND AUST, S. D.

Biophys. Res. Cmmun 77,797-803. (1978) Photo&em Photobiol 28, 803-810.

12. MYERS, C. E., MCGUIRE, W. P., LISS, R. H., IFRIM,

I., GROTZINGER, K., AND YOUNG, R. C. (1977)

Science 197, 165-167.

13. TYLER, D. D. (1975) FEBS Leti 51,180-183.

14. FONG, K., MCCAY, P. B., POYER, J. L., KEELE, B. B., AND MISRA, H. (1973) J. Bid Cbm. 248,

7792-7797. 15. ILAN, Y. A., AND CZAPSKI, G. (1977)Biochim. Bb

phys. Acta 498,368-394.

16. WALLING, C. (1975) Ace. Chem Res. 8. 125-131. 17. KELLOGG, E. W., III, AND FRIDOVICH, I. (1975) J.

Bid Chem 250,8812-8817.

18. KELLOGG, E. W., III, AND FRIDOVICH, I. (1977) J. BioL Chem 252,6721-6728.

19. FRIDOVICH, S. E., AND PORTER, N. A. (1981) J.

Bid Chem. 256, 260-265. 20. KOSTER, J. F., AND SLEE, R. G. (1980) Biochim

Biuphys. Actu 620,489-499. 21. LAI, C. S., AND PIEWE, L. H. (1977) Biochem

Biophys. Res. Commun 78,51-59.

22. MCCORD, J. M., AND FRIDOVICH, I. (1969) J. Bid Chem. 244, 6049-6055.

23. HOLMES, R. S., AND MASTERS, C. J. (1970) FEBS Lett. 11, 45-48.