Oxidation of arachidonic acid in micelles by superoxide and hydrogen peroxide.

Arachidonic acid was co-oxidized by xanthine oxidase. Both superoxide radical and hydrogen peroxide were required for oxidation, as shown by essentially complete inhibition caused by superoxide dismutase or by catalase. Pure arachidonate, free of lipid hydroperoxides, was susceptible to this co-oxidation, and the presence of lipid hydroperoxides did not accelerate the process. The role of trace metals was indicated by the stimulatory effect of EDTA-Fe and by the inhibitory effect of diethylenetriamine pentaacetate. Initiation of arachidonate co-oxidation was due to a potent oxidant generated by the interaction of H2O2 and O2- in the presence of Fe, rather than to either O2- or H2O2 per se. Hence, mannitol, a scavenger of OH ., but not of O2- or H2O2, also inhibited oxidation. Arachidonic acid autoxidation, a much slower process than xanthine oxidase co-oxidation, was barely detectable on the time scale of these observations. Unlike the co-oxidation, autoxidation was autocatalytic and therefore accelerated by hydroperoxide products. Marked quantitative differences in the distribution of isomeric hydroperoxide products of enzymic co-oxidation, as compared to the autoxidation, were noted and their significance was discussed.

Oxidation of polyunsaturated lipids usually proceeds by a free radical chain mechanism, yielding a mixture of hydroperoxide, carbonyl, and hydrocarbon products. Because this process is the basis of the oxidative polymerization of drying oils and of the rancidification of foods, it has been well studied and has been the subject of several reviews (1-5). Expiration of hydrocarbons, such as ethane and pentane, demonstrates that lipid oxidation can occur in vivo (6-10). Polyunsaturated lipids are particularly abundant in biological membranes and their oxidation constitutes an obvious threat to the integrity of such membranes. Moreover, this lipid peroxidation in a variety of biological systems generates peroxide products with intriguing and unique properties.
Hydroperoxides derived from arachidonic acid, for example, function as modulators of the enzymes involved in pros~glandin biosynthesis (11) and are potent chemotactic agents for neutrophils (12). A lipid hydroperoxide produced from arachidonic acid, 5-hydroperoxyeicosatetraenoic acid, is a proposed intermediate in the biosynthesis of leukotrienes (13-15).
Liver microsomes, which catalyze a NADPH-dependent hydroxylation of a variety of compounds, cause an iron-dependent and NA~PH-dependent oxidation of lipids (16) * This work was supported by Research Grant HL 22219 from the National Institutes of Health. 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.
$ Recipient of a Research Career Developnlent Award, 1577-1582. which is inhibited by hydroxylatable compounds (17). Superoxide dismutase was seen to inhibit this microsomal lipid oxidation, suggesting the involvement of 02- (18). Superoxide dismutase ~i b i t i o n of lipid oxidation has been noted under a variety of circumstances (19-33) . Yet there has been serious disagreement. Kellogg and Fridovich (23) noted that the xanthine oxidase reaction, a known source of 0 2 -and H202, caused the oxidation of linolenate. Superoxide dismutase or catalase inhibited this lipid oxidation, indicating a requirement for both 02and H202 (23).
Essentially identical results were subsequently noted with phospholipid vesicles and with intact erythrocytes (28) and with vesicles formed from washed erythrocyte stoma (34). These data were interpreted on the basis of an interaction of 0 2 -with H202, to generate OH., or a comparably potent oxidant, which then attacked the polyunsaturated lipid. This interaction had been proposed much earlier (35) and was later modified to include catalysis by iron compounds (36, 37). Thomas et al. (32) reinvestigated this process by studying the oxidation of linoleate by the xanthine oxidase reaction. They reported, in contradiction to the earlier work (23, 28), that linoleate oxidation by 0 2 -was dependent upon the extent of prior accumulation of lipid hydroperoxide and concluded that the proximal oxidant was generated by interaction of 0 2 -with lipid hydroperoxide, not with H202.
It seemed essential that these apparent discrepancies be clarified in order to provide a fiim foundation of fact upon which further progress could be based. We now report an investigation of the oxidation of arachidonate by an enzymic source of 02plus H202 under conditions allowing continuous monitoring of the rate of accumulation of conjugated diene products.

MATERIALS AND METHODS
Xanthine oxidase was prepared from unpasteurized cream (38) and was assayed by ultraviolet s~ctrophotometry 139). It was stored at -20°C in 50 mM potassium phosphate, 0.1 mM EDTA at pH 7.8. Superoxide dismutase was isolated from bovine erythrocytes and was assayed in terms of its ability to inhibit the reduction of cytochrome c by xanthine oxidase (40). Bovine liver catalase was purchased from the Worthington Biochemical Corporation and horse heart cytochrome c, type 111, from Sigma. Arachidonic acid was from Nu-Chek Preparations, Inc. and sodium octyl sulfate was from Schwartz/Mann. Acetaldehyde from Eastman was freshly distilled before use. All other compounds used were obtained from commercial sources at the highest available level of purity.
In order to permit continuous spectrophotometric monitoring Of its oxidation, arachidonic acid was dispersed in the form of mixed micehs with sodium octyl sulfate. Sodium octyl sulfate was chosen for this purpose, rather than longer chain alkyl sulfonates such as sodium dodecyl sulfate, because it has less of a tendency to denature proteins. ' We found that sodium octyl sulfate at 10 mM did not interfere with the xanthine oxidase reaction, but at 100 mM it acted as a Strong oxidation inhibitor. Optically clear stock dispersions of arachidonate (6.0 mM) were prepared by adding it to a solution containing 0.11 M Dr. J. A. Reynolds, personal commun~cation. sodium octyl sulfate, 50 m~ sodium pyrophosphate, and 0.1 m M EDTA at pH 9.0. Mixing in a Vortex blender at 25°C was followed by 15-fold dilution into the sodium pyrophosphate-buffered reaction mixtures. Potassium salts caused precipitation of the octyl sulfonate and were avoided.
Reaction mixtures usually contained 50 m M sodium pyrophosphate, 6.0 mM sodium octyl sulfate, 0.33 mM arachidonate, 0.1 mM EDTA, 20 m M acetaldehyde, and 1.7 X 10." M xanthine oxidase at pH 9.0 and 38°C. Where indicated, H402 was also present at 0.05 mM. Reactions were initiated, after thermal equilibration, by adding the xanthine oxidase. Accumulation of conjugated diene products was monitored at 233 nm in a Gilford model 2000 recording spectrophotometer, equipped with a thermostatted cell compartment. Temperature equilibration prior to initiation of the reaction was critical, since the micellar configuration was responsive to temperature (41) and caused an apparent temperature-dependent change in absorbance. Preparative scale oxidations were performed in partially filled Erlenmeyer flasks continually agitated in a water bath at 38°C. At intervals, 3.0ml samples of such large scale reactions were temporarily transferred into cuvettes and their absorbance was recorded at 233 nm.
The correspondence between increased absorbance at 233 nm and the formation of conjugated diene hydroperoxides was verified by HPLC,' TLC, and repeated scanning of the absorption spectrum with a Hitachi 100-80 recording spectrophotometer. Samples were prepared for HPLC and TLC by acidification with H2S04, followed by extraction with chloroform/methanol. Extracts were then reduced to dryness in uacuo, and redissolved in organic solvent. Silica TLC plates were developed with hexane:isopropyl alcoho1:acetic acid (200: 45:5) (42) and spots were visualized by spraying with 4% KI in acetic acid followed 4 min later with 1% aqueous starch (23). Fatty acid hydroperoxides were separated by normal phase HPLC using a Whatman Magnum 9 10-pm silica column (9 mm, inner diameter, X 50 cm). The solvent used was acetic acid:isopropyl alcoho1:hexane (2:14:984), and the flow rate was 3 ml/min, with detection by uv at 235 nm (43).

RESULTS
Co-oxidation ofArachidonate byxanthine Oxidase-Xanthine oxidase, while catalyzing the aerobic oxidation of acetaldehyde, caused a co-oxidation of arachidonate. This cooxidation, followed in terms of absorbance increase at 233 nm, was entirely dependent upon the simultaneous presence of xanthine oxidase, acetaldehyde, and arachidonate. Omission of any one of these components eliminated detectable changes in absorbance. We may conclude that spontaneous autoxidation of arachidonate was not significant on the time scale and under the conditions of these observations and that the conversion of acetaldehyde to acetate did not contribute noticeably to the rates followed at 233 nm.
There was an autocatalytic aspect to this co-oxidation of arachidonate. Thus, as shown by Line 1 in Fig. 1, the absorbance a t 233 nm increased exponentially during the first 15 min of the reaction. It seemed likely that this reflected the progressive accumulation of H202. H202, when added at the outset, did decrease the exponential behavior, while increasing the initial rate. Indeed as shown by Line 4, Fig. 1, 0.033 mM Hz02 gave a completely linear co-oxidation of arachidonate. H202, by itself, did not cause any oxidation of arachidonate and omission of xanthine oxidase or acetaldehyde completely eliminated increases in absorbance at 233 nm, even though 0.05 mM H a 0 2 and the other components of the reaction mixture were present. The effect of varying the concentration of acetaldehyde on the co-oxidation of arachidonate was explored. Typical saturation behavior was seen with a halfmaximal rate achieved at approximately 4 mM acetaldehyde. The effect of temperature was also explored. The rate of cooxidation of arachidonate vaned as would be expected on the basis of the Arhennius equation with an energy of activation of 10.7 kcal/mol. The Arhennius plot was linear over the temperature range 25-45°C with a correlation coefficient of '' The abbreviations used are: HPLC, high pressure liquid chromatography; DTPA, diethylenetriaminepentaacetic acid.  1 (left). The effect of Hz02 on the enzymic co-oxidation of arachidonate. Reaction mixtures contained 0.33 mM arachidonate, 6.0 m M sodium octyl sulfate, 0.1 m~ EDTA, 50 mM sodium pyrophosphate, 20 n m acetaldehyde, 8.3 nM xanthine oxidase, and the following amounts of HZ02: Line 1, none; Line 2, 5.6 p~; Line 3, 11 p~; Line 4, 33 pM; Line 5, 67 pM; and Line 6, 130 pM. Incubation was at 38°C and pH 9.0. FIG. 2 (right). The effect of varying the concentration of xanthine oxidase. Reaction mixtures were as described in the legend to Fig. 1, but without added H202 and with the following concentrations of xanthine oxidase: Line 1 , 17 nM; Line 2, 8.2 nM; Line 3, 4.2 nM; and Line 4, 2.1 nM. In the absence of xanthine oxidase there was no detectable increase in absorbance at 233 nm over the period of observation.
-," 0.99. Increasing the concentration of xanthine oxidase increased the rate of co-oxidation of arachidonate. This is shown by the data in Fig. 2. It was often convenient, in subsequent experiments, to eliminate the exponential behavior evident in Fig. 2 by adding 0.05 mM H202.
Catalase Inhibition of Arachidonate Co-oxidation-It seemed likely that the exponential increase in the rate of cooxidation of arachidonate reflected accumulation of HZOZ. This was supported by the effect of exogenous H202 as shown in Fig. 1. If H202 was essential for the co-oxidation of arachidonate, then catalase should inhibit this reaction. The results in Fig. 3 demonstrate that this was the case. Catalase inhibited 50% at 0.16 pg/ml and 95% at 4.8 pg/ml. When catalase was added to a reaction containing 0.05 mM exogenous HzOr, the rate of arachidonate co-oxidation fell, over the course of a few minutes, as the level of HzOa was decreased to a lower steady state by the balance between the production of H 2 0 a by xanthine oxidase and its removal by catalase. Catalase inhibited to an identical extent as the experiment with exogenous H202 (see above), when the only source of H202 was the xanthine oxidase reaction itself and when it was applied after a linear rate had been achieved.
Heat denaturation of catalase markedly diminished, but did not entirely eliminate, its ability to inhibit arachidonate cooxidation. This is probably a reflection of the known peroxidatic activity of ferrihemes (44).
Superoxide Dismutase Inhibition of Arachidonate Co-oxidation-Superoxide dismutase powerfully inhibited the cooxidation of arachidonate and did so whether added at the outset or after the reaction was well underway. In the lat,ter circumstance, addition of superoxide dismutase caused a cessation of arachidonate co-oxidation within the few seconds required for its admixture with the other reaction components. As shown in Fig. 4, 5.5 n g / d caused 50% inhibition and 1.0 p g / d caused complete inhibition. Since the molecular weight of superoxide dismutase is 32,000 (35) we estimate that 1.7 x 10"' M was sufficient to cause 50% inhibition. Superoxide dismutase, inactivated by heating a t 1OO"C, was entirely without effect when tested at 0.48 pg/ml.
Since superoxide dismutase scavenges 02-, but not H202, while catalase scavenges Hz02 but not 02-, and since either superoxide dismutase or catalase could cause virtually complete inhibition of arachidonate co-oxidation by xanthine oxidase, it follows that both Oeand HZ02 are required for this co-oxidation.  FIG. 3 (left). Inhibition of arachidonate co-oxidation by catalase. Reaction mixtures were as described in the legend to Fig. 1, with 50 PM H202 added in all cases and with the indicated levels of catalase added about 5 min after the reaction was initiated with xanthine oxidase. The linear rates achieved after addition of catalase were used to calculate per cent inhibition. Identical linear rates were attained when no exogenous Hz02 was added. Catalase caused 50% inhibition at 0.16 Pg/ml. Inhibition of Arachidonate Co-oxidation by MannitoZ-Oxidations exhibiting a dual dependence upon both 02and H202 have repeatedly been observed to be inhibited by scavengers of OH. (35,(45)(46)(47)(48)(49)(50). We tested the effects of mannitol, which reacts very rapidly with OH. (51), but which scavenges neither 02nor H202. Mannitol at 20 mM caused 32% inhibition and at 40 mM caused 50% inhibition of arachidonate cooxidation. We conclude that 02and Hz02, generated by the xanthine oxidase reaction, interacted in a way which produced a potent oxidant which could be scavenged by mannitol and which was responsible for initiating the oxidation of arachidonate.
The Role of Metals-The simplest explanation for the production of a potent oxidant from 0 2 -plus H202 would be the reduction of H202 to OH-+ OH. by 0 2 -. This reaction, frequently referred to as the Haber-Weiss reaction, has been explored and its rate found to be very low (52). Others have, however, found evidence for catalysis of the Haber-Weiss reaction by iron complexes (53,54). We explored the possibility of such catalysis by testing the effects of added EDTA-Fe(II), which at the pH of the reaction mixtures would oxidize to EDTA-Fe(II1). Fig. 5 demonstrates that micromolar levels of EDTA-Fe(I1) markedly stimulated the co-oxidation of arachidonate by the xanthine oxidase reaction. Arachidonate co-oxidation, in the presence of exogenous 25 PM EDTA-Fe(II), remained fully susceptible to inhibition by superoxide dismutase or by catalase and fully dependent on the xanthine oxidase reaction.
Ferrous salts plus H202 do constitute a source of OH. by the Fenton reaction (55), and high levels of EDTA-Fe(I1) (50 PM) did cause arachidonate oxidation in reaction mixtures containing 0.05 mM H202 and lacking xanthine oxidase. However, even with 50 mM EDTA-Fe(II), the addition of xanthine oxidase increased arachidonate oxidation 8.3-fold. Commercially available EDTA and sodium pyrophosphate contain impurities of iron, and we suspect that the co-oxidation of arachidonate seen in the absence of exogenous EDTA-Fe(I1) was dependent upon this endogenous Fe. Indeed, adding EDTA, per se, increased arachidonate co-oxidation to an extent which suggested a 0.5% impurity of Fe in the EDTA.
If the co-oxidation of arachidonate by xanthine oxidase was entirely dependent upon trace impurities of iron, as well as upon 02and H202, then a chelating agent which inactivated the iron impurities should eliminate this co-oxidation. Previous studies have suggested that DTPA, unlike EDTA, does *,mi\ Fe* l P M Reactions were run in a final volume of 3.0 ml incubated at 38°C and pH 9.0. Superoxide dismutase caused 50% inhibition at 5.5 ng/ml. 5 (right). Stimulation of arachidonate co-oxidation by Fe-EDTA. Reaction mixtures contained 0.33 mM arachidonate, 6.0 mM sodium octyl sulfate, 0.2 mM EDTA, 50 PM Hz&, 50 mM sodium pyrophosphate, 20 m~ acetaldehyde, 17 nM xanthine oxidase, and the indicated concentrations of added FeS04. It should be noted that the rate in the presence of 0.2 m~ EDTA was 46% greater than when 0.1 mM EDTA was used. This could reflect an iron impurity in the EDTA. When xanthine oxidase was omitted, the rate was that indicated by 0, and when both xanthine oxidase and H20~ were omitted the rate was that indicated by A.

FIG.
inactivate iron salts (56). When 0.1 mM EDTA was replaced by 0.1 mM DTPA, the rate of arachidonate co-oxidation was inhibited 46%, and when 0.2 mM EDTA was replaced by 0.2 mM DTPA this inhibition increased to 75%. This effect of DTPA, as well as the effects of superoxide dismutase, catalase, and mannitol, suggest that OH., or a species of comparable reactivity, generated by an iron-catalyzed interaction of 02and H202, was the root cause of arachidonate co-oxidation.

Does Lipid Hydroperoxide Play a Role in Arachidonate
Co-onidation?-The arachidonic acid used in the studies described above was devoid of hydroperoxide by the criterion of TLC. Furthermore, quantitative HPLC demonstrated that its content of hydroperoxide was less than 0.005%. Nevertheless, since Thomas et al. (32) reported an essential role of hydroperoxides in the co-oxidation of arachidonate by xanthine oxidase, we explored the effect of arachidonic hydroperoxides. Neat arachidonic acid was air-oxidized under the influence of ultraviolet irradiation. The extent of accumulation of hydroperoxide was estimated from its absorbance at 233 nm, applying a molar extinction coefficient of 25,000 (57). This arachidonic acid was then dispersed with sodium octyl sulfate, and its susceptibility to co-oxidation by xanthine oxidase was measured. The results are shown in Table I. It is clear that hydroperoxide did not increase suceptibility to co-oxidation, but rather decreased it substantially. This lack of a stimulatory effect of hydroperoxides was expected. Thus, when Hz02 was present at the outset, the co-oxidation of arachidonate by xanthine oxidase was a linear function of time, until slowed by inactivation of xanthine oxidase after -45 min. Had hydroperoxides been a required reactant, we should have observed

TABLE I
Effects of arachidonate hydroperoxides on eo-oxidation of arachidonate Reaction mixtures contained 0.33 mM arachidonate whose content of peroxide is indicated as a percentage of the total arachidonate. Other components were 20 m~ acetaldehyde, 0.1 mM EDTA, 6.0 mM sodium octyl sulfate, 50 m~ sodium pyrophosphate, and 17 nM xanthine oxidase. Incubation was at 38°C and pH 9.0.

Initial oeroxide
Linear rate  6. HPLC chromatogram of arachidonate co-oxidation and autoxidation. a, HPLC chromatogram of arachidonic acid-free radical autoxidation. Autoxidation was carried out in the bulk phase for 1 day at room temperature. HPLC conditions are as described in the text. b, HPLC chromatogram of xanthine oxidase initiated co-oxidation of arachidonic acid. Conditions of oxidation and chromatography are as described in the text. The substitution of hydroperoxide on the fatty acid is noted in both a and b.
an exponential increase of rate with increased time of reaction, due to the effect of accumulated peroxide products.
While the co-oxidation of arachidonate by xanthine oxidase was not increased by the presence of hydroperoxides the au~xidation of arachidonate, in the absence of xanthine oxidase, did show such an augmentation. This autoxidation was very much slower than the enzymic co-oxidation, but it could he ohserved over long periods of incubation. Thus, in the usual reaction mixture, such as the one described in the legend of Fig. 2 but lacking xanthine oxidase, we noted an autoxidation of 2.9 nM/min, averaged over the first 1.2 h, and of 7.2 nM/min, averaged over the next hour. In contrast, arachidonate co-oxidation in the complete reaction mixture containing xanthine oxidase was 110 m/min averaged over the fiist 1.2 h, and had decreased to 87 nM/min over the next hour.
Superoxide dismutase at 0.5 p g / d was sufficient to cause 5ocTo inhibition of this autocatalytic autoxidation, implying a role for 02-. Product Analysis-TLC and HPLC were used to examine and identify the conjugated diene hydroperoxide products of arachidonate co-oxidation by xanthine oxidase. In complete reaction mixtures, such as those described in the legend to Fig. 4, but lacking superoxide dismutase, 3.9% conversion of arachidonate to hydroperoxide was achieved in 130 min of incubation. TLC confmed the presence of several peroxide products. HPLC with effluent monitoring at 235 nm showed the six hydroperoxide isomers usually produced by the autoxidation of arachidonate (43).
Conjugated diene hy~operoxides substituted at positions 15,12,11,9,8, and 5 of the 20-carbon chain were observed. In addition to these major products having trans,cis-diene stereochemistry, minor products tentatively assigned the truns,trans-diene conf~guration were detected. As in bulk phase autoxidation of arachidonate, the 15-hydroperoxide and 5-substituted hydroperoxide (truns,cis stereochemistry) were the major products formed in xanthine oxidase co-oxidation. There were, however, striking quantitative differences in the isomeric hydroperoxides observed in autoxidation and enzymic co-oxidation. The ratio of the 5 15 hydroperoxides found from autoxidation was generdy less than 0.6, while the same ratio observed m the xanthine oxidase-catalyzed reaction was greater than 1.1 (four experiments). In addition, the relative smounts of the 12-, 11-, 9-, and ~hy~roperoxides formed in the enzymic oxidation were significantly less than was found in autoxidation (see Fig. 6). In fact, the 12-, 11-, 9-, and 8hydroperoxides amounted to less than 5% of the hydroperoxide mixture in the xanthine oxidase oxidation while in bulk phase arachidonic acid oxidation, these hydroperoxides amounted to 30% of the hydroperoxides formed.

DISCUSSION
The co-oxidation of arachidonate, to conjugated dienoic hydroperoxides, by xanthine oxidase was dependent upon both 02-and H20s. Thus, catalase, a selective Hz02 scavenger, caused virtually complete inhibition of oxidation, as did superoxide dismutase, a selective 0 2 -scavenger. This co-oxidation exhibited an initial exponential phase due to the accumulation of H202. Addition of Hz02 gave an enzymic cooxidation which was linear within the period of observation. The species actually initiating arachidonate oxidation appeared to be OH., or something of comparable reactivity, since mannitol, which scavenges OH., but not 0 2 -or € 3 2 0 2 , inhibited the enzymic co-oxidation. These results are most readily accomodated by proposing an interaction of 0 2 -and H20z which generated OH. or its equivalent. Endogenous iron salts and complexes were important for this interaction as shown by the accelerating effect of EDTA-Fe and by the inhibition by DTPA. Several reaction schemes can be proposed for this iron-catalyzed production of a potent oxidant from 0 2 -plus HaOa. Thus: Fe"+ + H0OH-t (Fe-OOH?' + H' (Fe"+OH)2+ + 02". 0 2 + OH-+ (FeO)" (111) (FeO)'2" + H' 4 (Fe-OH):'" --t Fe"' i OH.
In Schemes I and 11, hydroxyl radical is the oxidant, while in Scheme 111, either hydroxy or (FeO)'" could play this role. Spin-trapping studies have suggested hydroxyl radical as the oxidant in comparable systems (58, 59) yet (Fe0)2' has also been shown to be capable of hydrogen atom abstraction from inactivated organic compounds (60,61).

Oxidation of Arachidonate by 02plus H 2 0 2
Pryor and coworkers reported that the co-oxidation of linoleate by xanthine oxidase was inhibitable by superoxide dismutase and was dependent upon the presence of hydroperoxides (32). They accomodated their results by proposing the reaction 0 2 -+ ROOH -+ O2 + OH-+ R-0.. We, in contrast, have noted no dependence of the enzymic co-oxidation upon the presence of hydroperoxides. Indeed peroxides, when introduced by prior autoxidation (Table I), slightly diminished the rate of enzymic co-oxidation. Autoxidation of arachidonate was, however, autocatalytic and thus was probably accelerated by lipid hydroperoxides.
An explanation of the apparent discrepancy between our results and those of the Pryor group may now be proposed. While the latter reported a single LOOH-dependent pathway for oxidation of arachidonic acid involving 0 2 -in the xanthine oxidase system, two distinct oxidative processes were at work in our system: a major co-oxidation and a minor autoxidation (seen only in the absence of the xanthine oxidase reaction). Unlike the co-oxidative pathway, the pathway for autoxidation gave an autocatalytic rate, implying a dependence on peroxides. 0 2 -was also implicated in this autoxidation, since superoxide dismutase at 0.5 p g / d inhibited 50%.
Hence, the oxidation reported by Pryor's group appears equivalent to the minor autoxidative pathway of arachidonate observed in our system, rather than to its enzymic co-oxidation. Their reaction systems contained 1.72 M ethanol to solubilize the fatty acid, as well as 0.1 m~ DTPA as a chelator.
Ethanol is an excellent scavenger of OH -, and has been seen to inhibit the co-oxidation fo methianal to ethylene by the xanthine oxidase system (35). DTPA at 0.1 mM would partially convert endogenous Fe into a chelate inactive with respect to catalyzing the interaction of 02with H202 (50% inhibition of co-oxidation in our system). These components of the reaction mixtures used by Pryor's group must account for the differences between the results reported by them, and those obtained by us.
The enhancement of 5-hydroperoxy-eicosatetraenoic acid in the enzymic co-oxidation described here over the amounts of this isomer found in the bulk phase oxidation of arachidonic acid is of chemical and biological interest.
The finding that 5-and 15-hydroperoxy-eicosatetraenoic acid are major products in superoxide/hydrogen peroxidemediated arachidonic acid oxidation relates to anaphylaxis and the inflammatory response. A superoxide-dependent plasma chemotactic factor has been suggested to be a lipid component bound to serum albumin (62) and in vitro incubation of purified arachidonic acid with a superoxide-generating system results in the formation of products that are strongly chemotactic (63). Our experiments provide a direct chemical link between these superoxide-mediated oxidations and arachidonic acid hydroperoxides, compounds known to be strongly chemotactic (12).
We also note that 5-hydroperoxy-eicosatetraenoic acid is the proposed intermediate in the biosynthesis of the slow reacting substance of anaphylaxis (14) and, as such, it plays a crucial role in the allergic response. 5-hydroperoxy-eicosatetraenoic acid is the major hydroperoxide product formed in superoxide/hydrogen peroxide-mediated arachidonic acid oxidation and incorporation of 5-hydro~roxy-eicosat~traenoic acid, so generated, into the slow reacting substance of anaphylaxis biosynthetic pathway may have profound biological effects.