THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 266, 15, pp. …15091 . 15092 Perhydroxyl Radical Initiated...

THE JOURNAL (c) 1991 by The American Society for Biochemistry and

OF BIOLOGICAL CHEMISTRY Molecular Biology, Inc.

Vol. 266, No. 23, Issue of August 15, pp. 15091-15098,1991 Printed in U. S. A.

Perhydroxyl Radical (HOO') Initiated Lipid Peroxidation T H E ROLE OF FATTY ACID HYDROPEROXIDES*

(Received for publication, March 13, 1991)

John AikensS and Thomas A. DixSt11 From the Departments of $Chemistry and §Biological Chemistry, The University of California, Irvine, California 9271 7

It is demonstrated that the perhydroxyl radical (HOO', the conjugate acid of superoxide (OF)), initiates fatty acid peroxidation (a model for biological lipid peroxidation) by two parallel pathways: fatty acid hydroperoxide (LOOH)-independent and LOOH-dependent. Previous workers (Gebicki, J. M., and Bielski, B. H. J. (1981) J. Am. Chem. SOC. 103,7020-7025) demonstrated that HOO', generated by pulse radiolysis, initiates peroxidation in ethanol/water fatty acid dispersions by abstraction of the bis-allylic hydrogen atom from a polyunsaturated fatty acid. Addition of O2 to the fatty acid radicals forms peroxyl radicals (LOO'S), the chain-propagating species of lipid peroxidation. In this work it is demonstrated that HOO', generated either chemically (KOz) or enzymatically (xanthine oxidase), is a good initiator of fatty acid peroxidation in linoleic acid ethanoltwater dispersions; 0; serves only as the source of HOO', and HOO' initiation can be observed at physiologically relevant pH values. In contrast to the previous results, the initiating effectiveness of HOO' is related directly to the initial concentration of LOOHs in the lipids to be peroxidized. This defines a LOOH-dependent mechanism for fatty acid peroxidation initiation by HOO', which parallels the previously established LOOH-independent pathway. Since the LOOH-dependent pathway is much more facile than the LOOH-independent pathway, LOOH is the kinetically preferred site of HOO' attack in these systems. Experiments comparing HOO'/LOOH-dependent fatty acid peroxidation with transition metal- and peroxyl radical-initiated peroxidation rule out the participation of the latter two species as initiators, which defines the HOO'/LOOH initiation system as mechanistically unique. LOOH product studies are consistent with either a direct or indirect hydrogen atom transfer between LOOH and HOO' to yield LOO's, which propagate peroxidation. The LOOH-dependent pathway of HO0'-initiated fatty acid peroxidation may be relevant to mechanisms of lipid peroxidation initiation in vivo.

A significant component of oxidant-associated mammalian

* This work was supported by United States Public Health Service Grant GM40338. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

7 To whom correspondence should be addressed Dept. of Chem- istry, The University of California, Irvine, CA 92717. Tel.: 714-856- 5455.

physiology and pathology involves the generation and reaction of superoxide (0;)' (1-4). The biological importance of 0; is underscored by noting the ubiquity of the superoxide dismu- tases (5) and genetic abnormalities which result in the absence of phagocyte 0;-generating activity by NAD(P)H oxidases (6-8). However, the established chemistry of 0; does not obviously define it as an important biological oxidant; the only efficient non-metal-dependent reactions unequivocally established for 0; in water are dismutation and proton abstraction (9, 10). 0: is apparently active because it is con- verted to more potent oxidants, such as the hydroxyl radical (HO'). While certainly of great importance, HO' does not fill all required functions of biological oxidants; in particular, its extreme reactivity with virtually any organic molecule pre- cludes diffusion from the site of generation which may limit its role in the multicompartment environment of cells and tissues (11). These observations and many others clearly indicate that defining molecular mechanisms of 0; reactivity remains of major importance.

To define these interactions, fatty acid peroxidation studied in vitro is a relevant model: the chemistry of the peroxidative process is identical to in uiuo lipid peroxidation (implicated as both a cause and effect of free radical-associated pathological conditions) (12, 13). Fatty acid peroxidation is initiated when a radical is generated; usually, but not exclusively, this occurs by abstraction of the bis-allylic hydrogen atom of a polyunsaturated fatty acid carbon chain. Once a radical is generated, propagation chain reactions result in the overall oxidation of polyunsaturated fatty acids to fatty acid hydroperoxides (LOOHs); thus, one initiating oxidant is amplified to result in the oxidation of many fatty acid residues. The role of LOOHs in cellular oxidant damage is also of consid- erable interest. LOOHs are minor constituents of cellular membranes, are intermediates and final products of prosta- glandin and leukotriene biosynthesis, and can he decomposed by transition metals to form other oxygen radicals; thus, the fates of LOOHs serve as an example of secondary processes that may be the cause of biological effects associated with lipid peroxidation.

The abbreviations used are: OF, superoxide; HO ' , hydroxyl radical; LOOH, polyunsaturated fatty acid hydroperoxide (generic); HOO', perhydroxyl radical; ROO', peroxyl radical (generic); KO2, potassium superoxide; AAPH, 2,2'-azobis(2-amidinopropane)hydrochloride; HPLC, high pressure liquid chromatography; 13-LOOH, mixture of 13-hydroperoxyoctadeca-9,11-dienoic acid (major component) and 9-hydroperoxyoctadeca-10,12-dienoic acid (minor component); 18:3-13-LOOH, mixture of 13-hydroperoxyoctadeca-9,12,15- trienoic acid (major component) and 9-hydroperoxyoctadeca- 10,12,15-trienoic acid (minor component); 13-LOH, mixture of methyl 13-hydroxyoctadeca-9,1l-dienoate (major component) and 9- hydroxyoctadeca-l0,12-dienoate (minor component); 18:1-LOOH, an equimolar mixture of 9-hydroperoxyoctadec-10-enoic acid and 10- hydroperoxyoctadec-8-enoic acid; LOO', LOOH-derived peroxyl radical.

15091

15092 Perhydroxyl Radical Initiated Lipid Peroxidation

Another minor cellular constituent that may play an important role in oxidant damage is the perhydroxyl radical (HOO'),' the conjugate acid of 0; (14, 15). Because the pKo of HOO' is 4.88 (16), its biological presence in tandem with 0; (albeit at relatively low concentrations) is assured; in addition, 0; spontaneously and enzymatically dismutates through the intermediacy of HOO' (10, 17). Although the chemistry of HOO ' in aqueous media is not well established, peroxyl radical (ROO') reactivity serves as the most relevant model (18). In addition to its reactivity, other properties of HOO' make it an attractive candidate for an important biological oxidant: it will react selectively (versus indiscrimi- nately) with organic molecules while its lack of charge potentially makes it more soluble in key sites of oxidant reactivity (14, 15). Notably, the pK, of HOO ' in aprotic solvents, as a model for the membrane bilayer lipid environment, has been estimated as 23 (9); thus, 0; formed in a biological membrane presumably will exist exclusively as HOO' . Accordingly, HOO' may contribute to the biological effects associated with 0; generation.

In this paper, we confirm previous results (19, 20) demonstrating that HOO * is active while 0; is inactive at initiating fatty acid peroxidation in chemically defined fatty acid dispersions. We extend these results by demonstrating that the initiating effects of HOO * can be observed at physiologically relevant pH values, long propagation chain lengths per initiating oxidant can be established, and the effects of HOO' are independent of oxidant source. An exploration of the param- eters of HOO' -dependent initiation reveals that the initiation mechanism in our systems is unique; specifically, the presence of LOOHs are required and metals are not involved. As a whole, this study provides support for the concept of HOO' as a reactive derivative of 0; that may participate in fatty acid peroxidation initiation, and implicates LOOHs as possible participants in the process. Thus, this study may have implications for mechanisms of oxidant damage in vivo.

EXPERIMENTAL PROCEDURES

Materials

Rossville undenatured ethyl alcohol was purchased from Gold Shield Chemical Co. Potassium superoxide (KO,), 2[3]-t-butyl-4- hydroxyanisole (butylated hydroxyanisole) and EDTA were from Aldrich. Linoleic and oleic acid were from Nu-Chek Prep. Lipoxidase (soybean type I), cytochrome c (type 111), xanthine oxidase (type I), superoxide dismutase, and linolenic acid were from Sigma. 2,2'- Azobis(2-amidinopropane)hydrochloride (AAPH) was obtained from Polysciences, Inc. (Warrington, PA), desferal methanesulfonate from Ciba-Geigy, and ["C]linoleic acid from Du Pont-New England Nu- clear. Distilled water was filtered through a Milli-Q (Millipore) water purification system before use. Diethyl ether used in the purification of fatty acids and fatty acid hydroperoxides was freshly distilled from NaH to remove antioxidant additives and peroxide contaminants. All other solvents and chemicals were obtained at the highest available purity from standard chemical supply firms.

Instrumentation All spectrophotometric experiments were performed on a Perkin-

Elmer Lambda 4 spectrophotometer interfaced with an ESC personal computer. The base-line sensitivity of the spectrophotometer, as evaluated for a 1.0 mM solution of linoleic acid (5% LOOH) in 80/20 ethanol/water (sample and reference) at 234 nm, is (+.)0.001 A , whereas the base-line drift is <0.001 Almin. Data were acquired and analyzed using UVS Spectroscopy Software from Softways (Moreno Valley, CA). HPLC was performed with a Waters system interfaced with an IBM AT personal computer running Maxima software (Waters); all columns were from Waters.

' While the IUPAC name for HOO' is the hydrodioxyl radical, we will use the common name, perhydroxyl radical.

Preparation and Manipulation of Fatty Acids and Derivatives

Linoleic acid was purified by flash column chromatography on silica gel (230-400 mesh) under N, using hexane/diethyl ether (BO/ 20) as the mobile phase. Fractions were analyzed by thin layer chromatography on silica gel eluted with hexane/diethyl ether ( l / l ) , and peroxide-containing spots visualized with N,N-dimethyl-l,.l- phenylenediamine dihydrochloride (21). The peroxide contamination after purification was assayed spectrophotometrically (22); typically, linoleic acid purified by this method still contained quantifiable levels of peroxide. 13-Hydroperoxyoctadeca-9,11-dienoic acid (13-LOOH) and 13-hydroperoxyoctadeca-9,12,15-trienoic acid (183-13-LOOH), were prepared enzymatically from linoleic and linolenic acid, respectively, by the method of Funk et al. (22). Purification of all fatty acid hydroperoxides was accomplished by flash column chromatography on silica gel as described above. After purification and removal of solvent, the resulting colorless oils were diluted in methanol and stored under N, at -20 "C. The isomeric purity of 13-LOOH was established by the gas chromatography-mass spectrometry procedure of Dix and Marnett (23); typically, about 10% of the purified fraction was the 9-isomer. For hydroperoxide product studies, greater than 99% isomerically pure 13-hydroperoxyoctadeca-9-cis-11-trans-dienoic acid (13-cis,tram-LOOH) could be obtained by straight phase HPLC separation using the conditions of Porter and Wujek (24) (conditions described under "13-LOOH Product Studies"). Fatty esters were prepared by treatment of the free acid with diazomethane (23). Methyl 13-hydroxyoctadeca-9,ll-dienoate (13-LOH) was prepared by NaBH4 reduction of methyl 13-LOOH (25); 10% of the fraction after reduction was the 9-isomer. 18:l-LOOH (an equimolar mixture of 9-hydroperoxyoctadec-10-enoic acid and 10-hydroperoy- octadec-8-enoic acid) was prepared by singlet oxygen oxidation of oleic acid (26). A three-necked round bottom flask was charged with a 60-m195/5 CC14/MeOH solution containing oleic acid (4.18 g, 14.8 mmol) and methylene blue (60 mg, 0.1 mmol). 0, was bubbled into the stirred reaction, and the flask was exposed to a high intensity UV lamp overnight. Solvent was removed in vacuo, and the hydroperoxides were purified by flash column chromatography (hexane/diethyl ether 1/1) and stored as described above.

KO,: Determination of Active 0;

The amount of active 0; in KO, is much less than stoichiometric; for example, the amount previously had been noted as 8% (27). Accordingly, the amount of active 0; was assayed immediately prior to each experiment by two different methods. First, KO, was placed in a vial under N, and mixed with an 80/20 EtOH/H,O solution containing 0.01 M KOH to yield a 1 X M KO, solution (measured pH 12.5). An aliquot of this solution was placed in a cuvette and the decay of 0; recorded at 250 nm (Ac = 2134 M" cm" (10)). The decay of 0; was second order with an observed rate of 11 f 1 M" s-', in agreement with the established rate of decay for 0; under these conditions (10). Calculations using data derived from the decay curve demonstrated that, in different sets of experiments, the 0; in solution comprised between 7-13% of the stoichiometric amount of KO, added; the percentage in a given experiment was batch- and humidity- dependent and decreased with the shelf life of the KO,. Second, the 0; concentration was confirmed in each experiment using the cytochrome c assay (28). KO, was added to a final concentration of 1 X

M to a solution of cytochrome c (1 X 10" M) in phosphate buffer (0.05 M, pH 8.5). Cytochrome c reduction was determined by monitoring the increase in absorbance at 550 nm (Ac = 2.1 X lo4 M" cm"); the amount of 0; in the KO,, determined by this method, was identical to that obtained in the spectrophotometric experiments. No other oxidants are present in the KO, because all effects attributable to 0; in solution were not observed if the 0; dismutated before use. All reported 0; concentrations are based on the active 0; in the KO, and are accurate to *IO%.

Xanthine Oxidase Determination of 0; Production Production of 0; by xanthine oxidase was determined in a modi-

fication of the procedures of McCord and Fridovich (29). Reactions contained 50 PM cytochrome c and 48 mM freshly distilled acetaldehyde dissolved in 50 mM Tris. C1 buffer at pH 8.0 and were initiated by the addition of 0.01 unit of xanthine oxidase. The reduction of cytochrome c was monitored by the change in absorbance at 550 nm (Ac = 2.1 X lo4 M" cm") over 300 s; this reduction was inhibited in a dose-dependent fashion by the addition of superoxide dismutase. These conditions yielded a 1.5 X M.S" flux of 0;. Commercial

Perhydroxyl Radical Initiated Lipid Peroxidation 15093 preparations of xanthine oxidase can contain iron bound exogenously to the enzyme (30) which was assayed according to Carter (31). Metal contamination was removed by overnight dialysis of all xanthine oxidase preparations with 1 mM desferal (32); the activity of the enzyme was unaffected by dialysis.

General Reaction Procedures KO, Initiation-All reactions contained 1.0 M fatty acid dissolved

in 80/20 EtOH/H,O and were stirred rapidly in air at 25 "C to oxygenate the solution during the course of the reaction. Acidic solutions (pH.,, = 1.8), neutral solutions (pH.,, = 7.0), and basic solutions (pHaP, = 11.8) used in the KO, initiation studies also contained 0.1 N HBSO,, 0.992 M KOH, and 1.1 M KOH, respectively; the pH.,, values were measured in all experiments with a standard pH electrode before the addition of 0;. The high concentrations of KOH in neutral and basic solutions were avoided when the fatty esters were used. To initiate the reaction, 100 pl of a 0; solution in KOH was added dropwise to the stirred reaction mixture; control experiments demonstrated that the amount of 0, added did not change the pH.,, of the solution. UV analysis was performed by removing 100-pl aliquots from the reaction mixture and diluting to 10 ml both before the reaction and at different times following initiation; a 1-ml sample of the diluent then was analyzed. From the UV data, the concentration of hydroperoxide could be calculated using the molar extinction coefficient for conjugated, bk-unsaturated LOOHs at 234 nm (2.5 X lo' M" cm")(22). Scanning from 220 to 250 nm demonstrated formation of the LOOH chromophore.

Xanthine Oxidase Turnover-dependent Initiation-Lipid peroxidation initiated by xanthine oxidase turnover was performed according to procedures modified from Thomas et al. (33). Reactions contained 1 mM linoleic acid dispersed in 50 mM Tris. C1 buffer, pH 8.0, and included 20 PM desferal, 48 mM acetaldehyde, and (in certain experiments) LOOHs. Reactions also contained 10% ethanol to sol- ubilize the fatty acids; in control experiments, the presence of ethanol had no effect on the activity of xanthine oxidase. Reactions were initiated by addition of 0.01 unit of desferal-dialyzed xanthine oxidase at room temperature and continuously monitored by UV spectroscopy for 20 min at 234 nm to follow the formation of LOOHs, as described above.

AAPH Initiation-Reactions initiated by AAPH were performed under the conditions of the KOn incubations except that 1.1 mg of AAPH (10 mM concentration in solution) was substituted for 0; and the reaction temperature maintained at 30 "C (34). Because the large absorbance of AAPH at 234 nm interferes with the UV assay, a HPLC assay was substituted to quantify LOOH formation. Aliquots (10 PI) of the 1.0-ml reaction were analyzed every hour by HPLC separation on a Waters pBondapak C18 column eluted with CH&N/ H,O/AcOH (95/5/0.1) (flow rate, 2.0 ml.min-'1; the LOOHs eluted as a peak at a retention time of approximately 18 min. Products were quantified by computer integration of peak area in reference to standard integration curves developed using the same separation protocol.

Evaluation of Trace Transition Metal Contamination Control reactions were assayed spectrophotometrically at 510 nm

in the presence of 1,lO-orthophenanthroline to evaluate the presence of trace levels of transition metals (Fe"+ and Cu2+). Although reactions contained <1 pM metal (i.e. below the sensitivity limit of the assay), desferal (20 mM) and/or EDTA (10 mM) were added in control reactions to sequester any residual Fe3'. Experiments designed to compare metal- uersus HO0'-initiated fatty acid peroxidation were run under conditions described for the KO, experiments but included Fe(NH4),(S04),.6H20 (10 p ~ ) .

13-LOOH Product Studies The products of HOO' reacting with LOOH were investigated

using xanthine oxidase as the HOO' source (as described above), except that 1.0 mM 13-cis,trans-LOOH (HPLC purified, as above) was substituted for the linoleic acid and the xanthine oxidase concentration was increased to 0.1 unit. Reactions were allowed to run 2 h at room temperature and adjusted to pH 3 with 10% HC1. The solutions were then extracted three times with diethyl ether, organic fractions combined, dried (sodium sulfate), and reduced to LOHs with triphenylphosphine (24). The HPLC separation method of Por- ter and Wujek (24) permits base-line separation of the four LOH isomers on a straight-phase pPorasil column eluted with hexane/ isopropyl alcohol (995/5) at a 1.5 ml. min" flow rate. Products were

quantified by UV detection and computer integration of relative peak areas in controls versus reactions; the 13-cis,trans-, 13-trans,trans-, g-truns,cis-, and 9-trans,trans-LOOHs eluted at approximately 13, 15, 18, and 20 min, respectively. Reactions investigating UV-trans- parent products used [14CJ13-LOOH as a substrate and were analyzed on a Novapack reversed-phase radial-compression column (23). The LOHs eluted at approximately 28 min, while oxygenated derivatives (trihydroxy and epoxyhydroxy fatty acids) eluted at approximately 17 and 26 min, respectively (23). Quantitation was by liquid scintillation counting of 1-min fractions of the eluent collected continuously post-column.

RESULTS

H 0 0 ' - and Not 0;-initiated Fatty Acid Peroxidation in Linoleic Acid Dispersions-Bielski and co-workers (19) were the first to demonstrate, using pulse radiolysis oxidant-generating methods, that HOO' and not 0; is capable of initiating peroxidation in unsaturated fatty acids dispersed in acidic solutions of ethanol/water. To explore this system further, we initially chose a different HOO'/O; generating method, KOn, which, when added to protic solvents, generates a defined concentration of HOO' and/or 0; depending on the pH of the solution. Fatty acid peroxidation initiating activities of HOO' and 0; could thus be distinguished by changing the pHapp of the reaction (Table I). Reactions were performed by adding 0; (as ethanolic KOn) to 1.0 M linoleic acid (containing 5% linoleic acid hydroperoxide isomers (LOOHs)) dispersed in 80/20 EtOH/H20, and assayed by spectrophotometrically monitoring the formation of additional LOOH at 234 nm after appropriate dilution. At pH,,, 1.8, 0; exists essentially completely as its conjugate acid, HOO '. Lipid peroxidation was initiated to the extent of approximately 50 LOOHs formed per HOO' added in these experiments. The chain length per initiation event was certainly much higher; initiation com- peted directly with the spontaneous dismutation of HOO'/ O;, which has a bimolecular rate constant of 8.3 X lo5 M"S"

at this pH in 100% water (10). At the pH,,, in which all of the oxidant is 0; (11.8), no fatty acid peroxidation was initiated. Because the dismutation rate of 0; at this pH is effectively zero (lo), the lipids were in contact with a much higher concentration of oxidant in comparison to the pH,,, 1.8 experiments. At pH.,, 7.0, a small, reproducible amount of fatty acid peroxidation was initiated, which is attributable to the small concentration of HOO' (-1% of the added 0;) at this pH; chain lengths of 250 LOOH formed per HOO ' added could be estimated. The greater initiating ability of HOO' at pH,,, 7.0 uersus 1.8 (1% oxidant initiates approximately 5% the level of fatty acid peroxidation) was of interest,

TABLE I Lipid peroxidation-initiating actiuities of HOO' and 0;

Reactions contained 1 M linoleic acid (including 5.2 k 0.2% LOOH) or methyl linoleate (including 5.2 & 0.4% LOOH dissolved in 80/20 EtOH/H,O and were initiated by the addition of 0.18 mM oxidant, as described under "Experimental Procedures." The extent of fatty acid peroxidation was quantified as total LOOH formation (mean of five experiments, standard deviations are indicated).

PH.,," Oxidant ALOOHb ALOOH' rnM mM

1.8 HOO' 7.0 0, + 1% HOO'd 0.4 ? 0.03 0.4 k 0.05

8.7 f 0.1 8.5 k 0.2

11.8 0; 0 0 " The apparent pH of the media.

Formation of LOOH using linoleic acid as the reactant. Formation of LOOH using methyl linoleate as the reactant. Approximate HOO ' at pH 7.0.


but not easily e~plained.~ Parallel experiments (Table I), in which methyl linoleate was substituted for linoleic acid, gave identical results. This demonstrated that the initiating abili- ties of HOO' and 0; were not a function of changes in the fatty acid environment because the pH,,, was varied. To summarize, our results parallel those reported previously (19) in that HOO' but not 0: initiated fatty acid peroxidation in linoleic acid dispersions. In addition, the initiation of long oxidation chains was demonstrated (the longest previously reported were around three)(l8), and initiation attributable to HOO' could be observed at a physiologically relevant pH.

0; (HOO') Produced by Xanthine Oxidase Also Initiated Lipid Peroxidation-Xanthine oxidase is used routinely as a source of 0; (10); Pryor and co-workers (26, 33) previously had observed LOOH-dependent fatty acid peroxidation initiated by oxidants released during xanthine oxidase turnover. The experimental conditions employed in our experiments were different from those using KO, as the oxidant source: the fatty acid concentration was 1.0 mM (including 0.5% LOOH) and 10% ethanol in water was the solvent. (The ethanol solubilized the lipids but did not affect the activity of xanthine oxidase at its pH optimum of 8.0.) In the presence of xanthine oxidase and a substrate (acetaldehyde), a flux of 1.5 X M.s" 0; was produced and linoleic acid was peroxidized at a rate of 6.2 X 10"' M.s". No fatty acid peroxidation was observed in the absence of either xanthine oxidase or acetaldehyde, or if the reaction included either butylated hydroxyanisole (0.1 mM) or superoxide dismutase (600 units/ml). Comparing the rates of HOO' production to fatty acid peroxidation established a chain length of approximately 4 LOOHs formed per HOO' added in these experiments (at pH 8.0, 1:lOOO of the 0; is HOO'), assuming once again that all of the HOO' is reacting with the lipids. The shorter chain lengths in these experiments versus the KO2 experiments were expected the concentration of fatty acid in these experiments was 1.0 mM (1:lOOO the concentration of the KO2 experiments). Thus, at neutral pH values, xanthine oxidase was a better source of 0; (HOO') than KOz because a greater total amount of fatty acid peroxidation was observed with much less concentrated fatty acid. This was attributable to 0; (HOO') being generated continuously at a much lower concentration flux and a higher total concentration than by KOs addition; the lower steady state concentration of HOO' favored reaction with the lipids versus bimolecular dismutation (35). Attempts to run the reactions at pH 7.0, to permit a more direct comparison of the KOz and xanthine oxidase systems, were plagued with fatty acid insolubility problems (addition of higher amounts of ethanol inhibited xanthine oxidase). However, the change in pH of the xanthine oxidase experiments from pH 7.0 to 8.0 was irrelevant with respect to relative HOO' formation in comparison to the KO2 experiments; both the steady state amount of HOO' in the 0; produced and the 02 dismutation rate are 10 times less at pH 8.0 versus 7.0. Controls to evaluate possible metal ion contam-

The observed rate of dismutation of O;/HOO' at pH 1.8 and 7.0 turns out to be almost identical. A t pH.,, 1.8, the dismutation rate is described by 2 HOO' + H202 + O2 (rate = kdiss [H00'12), while at pH.,, 7.0, the dismutation rate is described by HOO' + 0; + H20 --* H,02 + O2 + OH- (rate' = & [HOO'][O;] . Since kdiss= 8.3 X lo5 l/mol.s (lo), k&,, = 9.7 X lo7 l/mol.s (lo), [HOO'] = 0.18 mM at pH.,, 1.8, 1.8 pM at pH.,, 7.0, and [O;] = 0.18 mM; rate = rate'. Thus, the rate of reaction of HOO' with the lipids should be identical if the rate constants (originally measured in H20)(10) are appropriate for the fatty acid/ethanol/water mixtures. Alternatively, various propagation and/or termination rate constants may be different a t the different pH,,, values. Competing pathways for fatty acid peroxidation initiation at pH.,, 7.0 are ruled out by various controls (see text).

ination in the xanthine oxidase incubations are presented below.

HOO' -initiated Lipid Peroxidation Was Linearly Dependent on the Initial Concentration of LOOH-All polyunsaturated lipids are contaminated with LOOHs in the absence of ex- haustive purification (36). For example, after purification by column chromatography and storage under Nz, our linoleic acid samples still contained 0.5-1.5 mol % LOOHs. To evaluate a possible role of LOOHs in HOO * -initiated fatty acid peroxidation, an increasing mole percent of 13-LOOH was included in the fatty acid dispersions prior to oxidant addition; both KOz and xanthine oxidase were used as the oxidant source in different sets of experiments. Fig. 1 plots the amount of new fatty acid hydroperoxide generated with HOO' initiation (KO2 source, 1.0 M fatty acid, 80% ethanol in water at pH,,, 1.8) versus the initial mole percent of 13-LOOH present in the fatty acid dispersions. The extent of fatty acid peroxidation was linearly (R = 0.99) dependent on the initial concentration of 13-LOOH in the fatty acid dispersion. Fig. 2 is identical to Fig. 1 except that xanthine oxidase was the source of HOO * (1.0 mM fatty acid, 10% ethanol in water at pH 8.0); again, a linear (R = 0.99) dependence on the LOOH concentration was established. Thus, under both experimental conditions, the second order reaction of HOO' with 13-LOOH to initiate fatty acid peroxidation was competing with the second order dismutation of HOO': the initial presence of 13-LOOH

z

s E X 0

U

v

n.n . I _.i 0 .0 1 .0 2.0 3 .0 4.0

mol% 13-LOOH FIG. 1. HO0'-initiated lipid peroxidation (KO2 oxidant

source) dependence on [ 13-LOOH]i.itia. Reactions contained 1 M total fatty acid (linoleic acid and 13-LOOH) in 80/20 EtOH/H,O at pH.,, 1.8 and were initiated by the addition of 0.042 mM HOO' (as KOs). Each point is the average of three experiments; standard deviations never exceeded 10%.

4.0

h r E x 2.0 E v

0 0 d

0.0 0.0 1.0 2.0 3 . 0 4.0 5 .0

mol% 13-LOOH FIG. 2. HO0'-initiated lipid peroxidation (xanthine oxidase

oxidant source): dependence on [ 13-LOOH]i,iti,l. Reactions contained 10 mM total fatty acid (linoleic acid and 13-LOOH) in 10/90 EtOH/H20 at pH 8.0 and were initiated by the addition of 0.01 unit of xanthine oxidase and 48 mM acetaldehyde. Each point is the average of three experiments; standard deviations never exceeded 10%.

Perhydroxyl Radical Initiated Lipid Peroxidation 15095

was required for observing HO0'-initiated fatty acid peroxidation in these systems.

One concern in these experiments was that the LOOH dependence could have been occurring in the propagation versus initiation phase of fatty acid peroxidation. If this was the case, LOOH dependence should be independent of the initiating oxidant. Thus, AAPH was used to generate peroxyl radicals (ROO ' s, whose reactivity parallels fatty acid peroxyl radicals (34)) under KO2 initiating conditions (1.0 M linoleic acid and LOOH dispersed in 80:20 ethanol/water). Changing the amount of 13-LOOH in the fatty acid mixture from 0.5 to 5 mol % failed to change the rate of fatty acid peroxidation initiated by a defined flux of ROO'; therefore the LOOH dependence occurring in the HOO . -initiated experiments must be in the initiation phase. Solubility problems precluded performing this experiment under xanthine oxidase conditions (AAPH is not sufficiently soluble in 10/90 ethanol/ water).

Transition metal-dependent fatty acid peroxidation can also show a dependence on LOOH concentration (37); thus, it was crucial to ensure that there are no residual metal catalysts present in the incubations that could be responsible for the fatty acid peroxidation initiation attributed to HOO ' . LOOHs could be involved in metal-dependent fatty acid peroxidation by being reduced to alkoxyl radicals or oxidized to peroxyl radicals (38):

LOOH + metal"' + LO' + metal("+"+ + -OH (1)

LOOH + metal"+ -+ LOO' + metal(""'+ + H+ (2)

A series of experiments both ruled out metal participation in these reactions and defined the HO0'-initiated, LOOH-dependent reaction as mechanistically unique. First, in both the KO, or xanthine oxidase initiating systems, reactions in which metal sequestering compounds (desferal and/or EDTA) were included failed to exhibit a change in the amount of H O 0 ' - dependent fatty acid peroxidation in the presence of a defined concentration of LOOH. Furthermore, the kinetics of metal- catalyzed and HOO * -initiated (KO, as oxidant source) fatty acid peroxidation in the presence of 5% 13-LOOH are quite different, as demonstrated by the results presented in Fig. 3 which compares the rate of LOOH-dependent initiation of fatty acid peroxidation by iron, HOO '/iron, and HOO' alone. Iron-initiated fatty acid peroxidation exhibited a monophasic increase in fatty acid peroxidation throughout the course of the incubation as the metal was apparently cycled through

3.0

Time (min) FIG. 3. HOO' uersus Fe2+ initiation of 13-LOOH-dependent

lipid peroxidation. Reactions contained 0.95 M linoleic acid and 0.05 M 13-LOOH ( I M total fatty acid) in 80/20 EtOH/H20 at pH,,, 1.8. Reactions were initiated by: A, Fez+ (IO PM) and KO, (0.042 mM HOO' 1; B, Fez+ (IO p ~ ) only; C, KO, (0.042 mM HOO') only; D, control (no initiators). Each point is the average of three experiments; standard deviations never exceeded 10%.

the reactions 1 and 2 (38). The HOO'liron-initiated peroxidation exhibited biphasic kinetics: an initial rapid burst of fatty acid oxidation was followed by a second phase in which oxidation continued at a slower rate. HOO . -initiated peroxidation exhibited only the initial burst phase. The biphasic kinetic behavior in the iron/HOO' system can be explained by initial HOO' reaction with 13-LOOH followed by a slower metal-dependent process. HOO ' -initiated fatty acid peroxidation was monophasic; once the HOO' is exhausted, there is no pathway to recycle the system to generate active initiator. Addition of iron to xanthine oxidase initiated systems simply increased the rate of fatty acid peroxidation, as a metal-dependent initiation process was added to the HOO'/ LOOH metal-independent initiation.

LOOH Structure-Activity and Metabolism Studies-Exper- iments designed to seek insight into the role of 13-LOOH in HOO' fatty acid peroxidation initiation were performed in both the KO,- and xanthine oxidase-initiated systems. First, because HOO' potentially could react at multiple sites of 13- LOOH, 13-LOH (the reduced form of 13-LOOH) was substituted to determine if other sites were involved in the reactive role implied for the hydroperoxide functionality. 13-LOH did not increase the amount of fatty acid peroxidation initiated in either system, which defined the central role of the hydroperoxide functional group in HOO' initiation. Second, the degree of unsaturation in the hydrocarbon chain of LOOH was varied systematically to address the possible structural role of unsaturation in hydroperoxide reactivity with HOO' . Fatty acid hydroperoxides containing either one (18:l- LOOH), two (13-LOOH), or three (183-LOOH) sites of unsaturation were equally able to increase the amount of HOO ' -initiated fatty acid peroxidation in both chemical and enzy- matic generating systems; significantly, t-butyl hydroperoxide and H202 did not function in this role. The latter experiment ruled out H202, a product of H00'/02 dismutation and a possible KOe contaminant (27), as a direct participant or oxidant precursor in HOO ' initiation.

Two general types of LOOH reactions could be operating in this system: peroxide hydrogen atom transfer to generate peroxyl radicals or one-electron reduction of the peroxide oxygen-oxygen bond to generate alkoxyl radicals. These pos- sibilities could be evaluated with 13-LOOH product studies (Fig. 4): if an indirect or direct hydrogen atom transfer mechanism was operating, the 13-LOO' could rearrange to form 9-LOO' (and subsequently 9-LOOHs after hydrogen abstraction; 9-LOHs after reduction) which could be detected and quantified by straight-phase HPLC (34), whereas homolytic peroxide bond cleavage would generate alkoxyl radicals and a

H-abstraction / 1 - 1 / I I

Peroxide Bond Cleavage "-1 I. cyclize

2. + 0 2 , +H' 3. reduce

FIG. 4. 13-LOOH: a molecular probe for mechanisms of by- droperoxide metabolism. R = -(CH,),COOH, R' = -(CH,),CH:+ Only the major, most diagnostic product is shown (in squares) for each pathway.


different series of rearrangement/hydrolysis products (for example, epoxy alcohols) detectable by reversed-phase HPLC (23). For these experiments, xanthine oxidase was chosen as the oxidant source because larger total amounts of HOO' could be produced in a given experiment and, since the flux of HOO ' released from the enzyme is small (approximately, 90 nM/min), HOO' reacted almost quantitatively with the LOOHs instead of dismutating (35). Accordingly, 11 WM HOO' (resulting from the generation of 11 mM 0; by xanthine oxidase over a 2-h time span) was reacted directly with HPLC-purified 13-cis-trans-LOOH (100 p ~ ) under conditions otherwise identical to the fatty acid peroxidation studies except that linoleic acid was omitted. The fatty acid products then were reduced and analyzed separately by straight- and reversed-phase HPLC. Fig. 5 shows the straight-phase separation (UV detection) of the four LOH isomers (13-cis,transs-, 13-trans,trans-, 9-trans,cis-, and 9-trans,trans-LOH) that could arise from formation, rearrangement, and reduction of 13-cis,trans-L00'; a control (Trace B ) contained small amounts of the three rearranged isomers (whose hydroperoxide precursors slowly form during storage of purified 13- cis,trans-LOOH). Reaction of HOO' with 13-cis,trans-LOOH led to exclusive formation of 3% (+0.5%) 9-trans,trans-LOOH (Trace A ) ; control experiments (lacking either xanthine oxidase or acetaldehyde) demonstrated no rearrangement over the base-line amounts. The 9-trans,trans-LOOH was the major product expected from rearrangement of 13-cis-trans- LOOH; the 13-trans,trans- and g-trans,cis-LOOHs are formed in secondary (slower) rearrangements (24). The 3% rearrangement was diagnostic of a minimal 6% peroxyl radical formation- the bis-allyl radical intermediate collapses back to the g-trans,trans-LOO' and 13-cis,trans-L00' with equal facility (24). Fig. 6 shows the reversed-phase HPLC separation of LOHs and oxygenated products from experiments identical to those of Fig. 5, except that I4C-labeled 13-LOOH was used as the substrate to enable quantitation of the UV-

li 1 1

4

12.0

Time (min) FIG. 5. HPLC assay for 13-cis,truns-LOOH rearrangement

promoted by HOO'. Reactions contained 100 PM HPLC purified 13-cis,trans-LOOH dissolved in 10% ethanol in 50 mM Tris.Cl buffer, pH 8.0, and were initiated by the addition of xanthine oxidase (0.1 unit) and acetaldehyde to 48 mM. Total time of incubation was 2 h. After workup and reduction, the LOHs were separated on a straight- phase (silica) HPLC column; peaks 1,2,3, and 4 are 13-cis,trans-, 13- trans,trans-, 9-trans,cis-, and g-trans,trans-LOHs, respectively (UV detection at 234 nm). Trace A is the reaction while Trace B is control (omit acetaldehyde). All experiments were performed in triplicate although each figure tracing is from a single experiment. The 13- cis,trans-LOH peaks were set off scale in order to see the other three LOHs, this peak was approximately 3% smaller in the reaction uersus control indicating that the only reaction was to form the 9- trans,trans-LOH (peak 4 ) .

60 -

40 -

20 -

o r n 10 20 30

Time(min)

FIG. 6. HPLC assay for alkoxyl radical promoted metabolism of 13-cis,truns-LOOH by HOO'. Reactions were performed identically to those described in Fig. 5 except that %labeled 13- LOOH was used as the hydroperoxide substrate to allow quantitation of UV invisible peaks after sample reduction by fraction collection and scintillation counting. Separation was by reversed-phase HPLC on C18-silica. Trace A is the reaction, Trace B the control (omit acetaldehyde), and Trace C substitutes hematin (10 p ~ ) for xanthine oxidase and acetaldehyde. Epoxy-hydroxy and trihydroxy fatty acids elute at 24-28 and 16-18 min retention time, respectively (note Trace C) while 13-LOH and 9-LOH co-elute at 29-30 min. Experiments were performed in triplicate.

r O2

FIG. 7. Parallel mechanisms for lipid peroxidation initiation by HOO': LOOH-independent (A) and LOOH-dependent ( B ) .

invisible products formed from alkoxyl radical rearrangements (37). Hematin-catalyzed breakdown of 13- LOOH occurs through an alkoxyl radical intermediate (23); when performed under conditions otherwise identical to the xanthine oxidase experiments, hydroxyepoxy (elution time, 24-28 min) and trihydroxy (elution time, 16-18 min) fatty acid derivatives were observed (Fig. 6, black squares) with only a small amount of 13-LOH (from 13-LOOH reduction) remaining (elution time, 29-30 min). In contrast, no oxygenated products above background were observed in either controls (diamonds) or in HOO' reactions with 13-LOOH (white squares); the only compound observed is the 13-LOOH reduction product, 13-LOH. (9-LOH, resulting from the rearrangement and reduction of 13-LOOH, co-eluted with 13- LOH.) In summary, these experiments demonstrated that peroxyl, not alkoxyl, radicals are the major products of the (direct or indirect) reaction of HOO' with 13-LOOH.

DISCUSSION

We will discuss our results in the context of Fig. 7. Bielski and co-workers (19,20) employedpulse-radiolysis studies first to demonstrate that HOO' and not 0; initiates fatty acid peroxidation by abstracting a hydrogen atom from the bis- allylic methylene of unsaturated fatty acids (Fig. 7A). We supported these findings by demonstrating HOO' -dependent fatty acid peroxidation initiation under a wide variety of conditions: notably, with two different oxidant sources and at neutral pH values that did not involve 0; in any metabolic role other than as the HOO' source. A series of control

Perhydroxyl Radical Initiated Lipid Peroxidation 15097

"?' "[ "+?' 1 __ "-7' OOH 00H".OOH 00.

+ HOOH

FIG. 8. Possible mechanism of HO0'-initiated, LOOH-dependent lipid peroxidation.

experiments ruled out the participation of metal-dependent redox processes (in particular, addition of iron to the incubations completely changed the kinetics of LOOH generation) and demonstrated that the KO, and xanthine oxidase systems share unique mechanistic characteristics. Specifically, in both systems, the rate of HOO ' -initiated fatty acid peroxidation was linearly related to the initial concentration of 13-LOOH; the 13-LOOH must be involved in the fatty acid peroxidation initiation process, because LOOHs were products (not reac- tants) in fatty acid peroxidation propagation reactions and peroxyl radical chain transfer reactions (see below) did not increase the concentration of peroxyl radicals. Therefore, the LOOH-dependent process for HOO * -dependent initiation of fatty acid peroxidation paralleled, and was mechanistically exclusive of, the LOOH-independent process (Fig. 7B).4

The LOOH-dependent initiation was clearly a much more facile process than bis-allylic hydrogen atom abstraction in these systems because the overall amount of fatty acid peroxidation initiated was a direct function of the initial LOOH concentration. Bimolecular rate constants of 0.3-1.2 X lo3 M" s" for the reaction of HOO' with the bis-allylic hydrogens of linoleic acid in 70:30 EtOH/H,O have been reported (18, 19); these rate constants thus serve as a lower limit for the reaction of LOOH with HOO ' (assuming their applicability to somewhat modified experimental conditions). In our experiments, HOO' reacting with LOOH uersus linoleic acid was the major (and possibly exclusive) mechanism by which fatty acid peroxidation was initiated; we were not able to establish conditions where only the LOOH-independent process was operating. These observations have general significance for fatty acid peroxidation studies; in vitro, all polyunsaturated lipids contain detectable levels of LOOHs unless exhaustively purified (36) while, in uiuo, LOOHs are common intermediates and products in a variety of cellular processes (2).

Inferences about the mechanistic details of the HO0' - initiated process will now be made. Mechanistically, LOOH is participating in HOO ' -initiated fatty acid peroxidation by providing a kinetically preferred reaction site, the hydroperoxide. We can focus on the hydroperoxide in mechanistic discussions because equimolar substitution of 13-LOH, a compound substituting an alcohol for the hydroperoxide but otherwise structurally identical to 13-LOOH, failed to increase the amount of fatty acid peroxidation initiated by HOO '. In addition, no evidence for metal ion or LOOH-derived alkoxy1 radical participation in the reaction could be obtained. The only LOOH reaction was the rearrangement of 13-cis,trans- LOOH to 9-trans,truns-LOOH; Porter and co-workers (12) and Chan et al. (39) have demonstrated that this rearrangement occurs through the intermediacy of peroxyl radicals (LOO .SI. Accordingly, our mechanistic hypothesis for the role of LOOHs in HOO' -initiated fatty acid peroxidation is summarized in Fig. 8: HOO ' reacts to abstract the hydroperoxide hydrogen atom from LOOH to generate peroxyl radicals. The literature provides support for the proposed

It is important to realize that the LOOH-independent initiation is a "true" (first-chain) initiation process, whereas the LOOH-dependent initiation is actually a stimulation of the peroxidative process. See discussion in Ref. 2, pp. 196-200.

initiation reaction. The chain transfer reaction of Fig. 8, while not previously reported for HOO ' , is quite facile for peroxyl radicals reacting with secondary hydroperoxides; bimolecular rate constants of -lo3 M" s" (40) are representative. Chenier and Howard (41) demonstrated that the energy of activation for hydrogen atom abstraction during the peroxyl radical chain transfer reaction is surprisingly low due to initial formation of a hydrogen-bonded complex between the peroxyl radical and hydroperoxide; this explains why cleavage of the hydroperoxide 0-H bond (bond strength: 90 kcal/mo1)(42) was favored kinetically over abstraction of a bis-allylic hydrogen (76 kcal/mo1)(42). The lack of fatty acid peroxidation sensitization observed with t-butyl hydroperoxide and H202 are also not at odds with the proposed mechanism. Hendry et al. (40) demonstrated that the rate constant for t-butyl peroxyl radical abstraction of the bis-allylic hydrogen of 1,4-cyclo hexanediene is about 18 times slower than for abstraction by LOO'; thus, if generated by reaction with HOO', the t-butyl peroxyl radical would be less effective at propagating fatty acid peroxidation than 13-LOO'. If HOO ' abstraction of a hydrogen atom from H2O2 occurs, no net radical propagation chains are generated.

The direct reaction of HOO ' and LOOH to generate peroxyl radicals is consistent with both experimental results and the literature; however, it remains to be proven: peroxyl radical rearrangements could have been occurring in propagation rather than initiation reactions. While no other 13-LOOH- derived products were observed in the reaction of HOO ' with 13-LOOH, the ratio of propagation to initiation reactions can be quite high. If the hydrogen atom transfer mechanism is not correct, HOO ' must have been reacting directly with LOOH to generate a radical species that is not the peroxyl radical followed by reaction of the latter with another LOOH to generate a peroxyl radical. The overall result of either mechanism adequately explains the role of LOOHs at sensi- tizing HOO ' -dependent fatty acid peroxidation: a short-lived radical (HOO', which is lost through dismutation) was con- verted (either directly or through a series of reactions) to a longer lived peroxyl radical (LOO'), the major chain propa- gator of fatty acid peroxidation.

Some differences between this work and the literature need to be addressed. The inability of Bielski and co-workers (43) to see HOO' reactivity with 13-LOOH in competition with HOO' dismutation (kdiss = 1.3 X lo6) may impose an upper limit on the rate constant for the chain transfer reaction (Fig. 8); alternatively, the differences between these studies and ours may have been simply due to differences in the experimental systems. Notably, our systems contained different amounts of lipids and ethanol/water percentages, which could affect the ability of HOO ' and LOOH to form the initial hydrogen-bonded complex expected to be necessary for the chain transfer r e a ~ t i o n . ~ Previous observations of LOOH- dependent fatty acid peroxidation sensitizations (2, 44) may be attributable to either HO0'-dependent initiations or the operation of metal-catalyzed Fenton or Haber-Weiss reactions. Under the conditions of our experiments, 0; did not react either with linoleic acid or LOOH to initiate fatty acid peroxidation. Many investigators have provided evidence both supporting (26, 44-46) and refuting (43, 47-49) the ability of 0 2 to react directly with bis-unsaturated fatty acids and/or LOOHs. While some of the differences may be a function of the solvents and hydroperoxides employed, the operation of HOO . -dependent initiations in the systems exhibiting fatty

' Recent results from our laboratory have demonstrated that hydrogen bond strengths are enormously solvent-dependent. Beeson, C., Kevorkian, G. , and Dix, T . A., unpublished experiments.


acid peroxidation is a distinct possibility. The fatty acid peroxidation initiating ability of HOO'

established in previous work (19) and amplified in this paper may have biological significance: HOO * could be an initiating oxidant in lipid peroxidation events associated with 0; generation. A diagnostic criteria for HOO' participation in biological oxidant damage, if the proper controls are performed, is exacerbation of the effects by the presence of LOOHs. Biologically, LOOHs are commonly formed in a number of physiological and pathological processes suggesting that LOOH amplification of HOO ' oxidant effects could also be occurring in vivo. Intriguing physiological conditions in which HOO * /LOOH could be active include heart ischemia/reper- fusion damage associated with xanthine oxidase, oxygen radicals, and lipid peroxidation (50), the phagocyte oxidative burst (6-8) especially when operating under acidic conditions (acidic phagosome or stomach environment), and lipophilic biological environments such as the lipid membrane (14, 15) that potentially favor both access and reactivity of HOO' versus 0;. Our laboratory is expanding this study toward the evaluation of possible synergistic roles of HOO' and LOOH in biological oxidant damage.

1.

2.

3. 4.

5.

6. 7.

8. 9.

10.

11. 12. 13.

14.

15.

16.

17.

18. 19.

REFERENCES Pryor, W. A. (ed) (1976-1984) Free Radicals i n Biology, Vols.

Halliwell, B., and Gutteridge, J. M. C. (1989) Free Radicals i n

Fridovich, I. (1986) Arch. Biochem. Biophys. 247, 1-11 Halliwell, B. (ed) (1988) Oxygen Radicals and Tissue Injury:

Oberley, L. E. (ed) (1985) Superoxide Dismutase, CRC Press,

Weiss, S. J., and LoBuglio, A. F. (1982) Lab. Inuest. 47, 5-18 Fantone, J. C., and Ward, P. A. (1982) A m . J . Pathol. 107, 397-

Babior, B. M. (1984) J. Clin. Inuest. 73, 599-601 Sawyer, D. T., and Valentine, J. S. (1981) Acc. Chem. Res. 14,

Bielski, B. H. J., Cabelli, D. E., Ravindra, L. A., and Ross, A. B.

Pryor, W. A. (1986) Annu. Reu. Physiol. 48, 657-667 Porter, N. A. (1986) Acc. Chem. Res. 19, 262-268 Kagan, V. E. (ed) (1988) Lipid Peroxidation in Biomembranes,

CRC Press, Boca Raton, FL Fridovich, I. (1988) in Oxygen Radicals in the Pathophysiology of

Heart Disease (Singal, P. K., ed) pp. 1-11, Kluwer Academic Publishers, Boston

Sawyer, D. T., McDowell, M. S., and Yamaguchi, K. S. (1988) Chem. Res. Toxicol. 1,97-100

Behar, D., Czapski, G., Rabini, J., Dorfman, L. M., and Schwarz, H. A. (1970) J . Phys. Chem. 74, 3209-3213

Nakamura, S., and Yamazaki, I. (1969) Biochim. Biophys. Acta

Ingold, K. U. (1969) Acc. Chem. Res. 2 , 1-9 Gebicki, J . M., and Bielski, B. H. J. (1981) J. Am. Chem. SOC.

I-VI, Academic Press, New York

Biology and Medicine, 2nd Ed., Clarendon Press, Oxford

Proceedings of an Upjohn Symposium, Allen Press, Kansas

Boca Raton, FL

418

393-400

(1985) J. Phys. Chem. Ref. Data 14, 1041

189,29-37

103,7020-7022

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31. 32.

33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

Bielski, B. H. J., Arudi, R. L., and Sutherland, M. W. (1983) J.

Milas, N. A., and Belic, I. (1959) J. Am. Chem. SOC. 8 1 , 3359-

Funk, M. O., Isaacs, R., and Porter, N. A. (1976) Lipids 11,113-

Dix, T. A., and Marnett, L. J. (1985) J. Biol. Chem. 260, 5351-

Porter, N. A., and Wujek, D. G. (1984) J. Am. Chem. SOC. 106,

Chan, H. W.-S., Levett, G., and Matthew, J. A. (1979) Chem.

Thomas, M. J., Mehl, K. S., and Pryor, W. A. (1978) Biochem.

Lokesh, E. R., and Cunningham, M. L. (1986) Toxicol. Lett. 34,

Sawyer, D. T., Calderwood, T. S., Yamaguchi, K., and Angelis,

McCord, J. M., andFridovich, I. (1969) J. Biol. Chem. 244,6049-

Vile, G. F., and Winterbourn, C. C. (1986) J. Free Radicals Biol.

Carter, P. (1971) Anal. Biochem. 40, 450-458 Britigan, B. E., Pou, S., Rosen, G. M., Lilleg, D. M., and Buettner,

G. R. (1990) J. Biol. Chem. 265, 17533-17538 Thomas, M. J., Mehl, K. S., and Pryor, W. A. (1982) J. Biol.

Chem. 257,8343-8347 Porter, N. A., Lehman, L. S., Weber, B. A., and Smith, K. J.

(1981) J. Am. Chem. SOC. 103,6447-6455 Brawn, K., and Fridovich, I. (1980) Acta Physiol. Scand. 492, 9-

18 Arudi, R. L., Sutherland, M. W., and Bielski, B. H. J. (1983) J.

Lipid Res. 24, 485-488 Sosnovsky, G., and Rawlinson, D. J. (1971) in Organic Peroxides

(Swern, D., ed) Vol. 11, pp. 153-268, John Wiley and Sons, New York

Biol. Chem. 258,4759-4761

3361

117

5357

2626-2629

Phys. Lipids 2 4 , 245-256

Biophys. Res. Commun. 83,927-932

75-84

C. T. (1973) Inorg. Chem. 22, 2577-2583

6055

& Med. 2,393-396

Gardner, H. W. (1989) Free Radical Biol. & Med. 7,65-86 Chan, H. W. S., Levett, G., and Matthew, J. A. (1979) Chem.

Phys. Lipids. 24, 245-256 Hendry, D. G., Mill, T., Piszkiewicz, L., Howard, J. A,, and

Eigenmann, H. K. (1974) J. Phys. Chem. Ref. Data 3,937-978 Chenier, J . H. B., and Howard, J. A. (1975) Can. J. Chem. 53,

Benson, S. W., and Shaw, R. (1971) in Organic Peroxides (Swern, D., ed) Vol. 1, pp. 132-136, John Wiley and Sons, New York, NY

Thomas, M. J., Sutherland, M. W., Arudi, R. L., and Bielski, B. H. J. (1984) Arch. Biochem. Biophys. 233, 772-775

Thomas, M. J., Mehl, K. S., and Pryor, W. A. (1982) J. Biol. Chem. 257,8343-8347

Sutherland, M. W., and Gebicki, J. M. (1982) Arch. Biochem.

623-627

~ i ~ p h y ~ . 214,1-i1 Peters. J . W.. and Foote. C. S. (1976) J. Am. Chem. SOC. 98,873- . .

875 Gibian, M. J., and Ungermann, T. (1979) J. Am. Chem. SOC. 101,

Bors, W., Michel, G., and Saran, M. (1979) FEBS Lett . 107,

Stanley, J. P. (1980) J. Org. Chem. 45, 1413-1418 Singal, P. K. (ed) (1988) Oxygen Radicals in the Pathophysiology

1291-1293

403-406

of Heart Disease, Kluwer Academic Publishers, Boston

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 266, 15, pp. …15091 . 15092 Perhydroxyl Radical Initiated...

Documents

Transcript of THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 266, 15, pp. …15091 . 15092 Perhydroxyl Radical Initiated...