Prostaglandins
and Medicine 1: 55-70, 1978.
THE ROLE OF IRON IN PROSTAGLANDIN SYNTHESIS: FERROUS IRON MEDIATED OXIDATION OF ARACHIDONIC ACID G.H.R. Rao, J.M. Gerrard, J.W., Eaton, J.G. White. University of Minnesota Health Sciences Center, Box 198 Mayo Memorial Building, Minneapolis, Minnesota 55455 (reprint requests to GHRR) ABSTRACT Arachidonic acid (AA) is the essential substrate for production of platelet endoperoxides and thromboxanes. Iron or heme is an essential cofactor for the peroxidase, lipoxygenase and cycle-oxygenase enzymes involved in formation of these products. The present study has examined the direct interactions between iron and arachidonic acid. Iron caused the oxidation of AA into more polar products which could be detected by UV absorbtion at 232 nM or the thiobarbituric acid (TBA) reaction. High pressure liquid chromatography, them-ionization and electron-impact mass spectrometry and nuclear magnetic resonance spectroscopy suggest that the major product was a hydroperoxide of AA. Ferrous iron (Fe++) and oxygen were absolute requirements. Fe++ was converted to the ferric iron (Fe+++) state during oxidation of AA, but Fe+++ could not substitute for Fe++. No other enzymes, cofactors or ions were involved. Conversion of AA to a hydroperoxide by Fe++ was inhibited by the antioxidant, 2, (3)-Tert-butyl-4-hydroxyanisole,the radical scavenger, nitroblue tetrazolium, and iron chelating agents, including EDTA, imidazole and dihydroxybenzoic acid. The reaction was not affected by superoxide dismutase, catalase or aspirin. These findings and preliminary studies of the Fe* induced oxidation product of AA as a substrate for prosta, glandin synthesis and inhibitor of prostacyclin production indicate the critical role of Fe++ in AA activation. INTRODUCTION Arachidonic acid (AA) is a fatty acid substrate for platelet lipoxy enase and cycle-oxygenase (l-4). Its oxidation yields the endoperoxides 95-9), thromboxanes (10-12) and other products important to the function of platelets in hemostasis (13-19). Several authors have suggested that conversion of AA involves participation of free radicals or development of a free radical state by the fatty acid, but no definitive requirement for hydroperoxy radical, superoxide anion, singlet oxygen, hydroxyl free radical or hydrogen peroxide has been demonstrated (20-27). Our interest in the primary phase of AA peroxidation and free radical generation was stimulated by the observation that a peroxidase might be involved because 3-Amino-1,2,4-triazole, an inhibitor of peroxidase, blocked platelet endoperoxide and throm-
55
boxane synthesis, secretion and aggregation (28). The peroxidase enzyme contains iron, and heme groups or iron have been linked to the activity of lipoxygenase (29-30), cycle-oxygenase (31,32) and thromboxane synthetase (11), though the precise relationships of iron to the enzyme reactions is not yet clear. Since heme proteins are known to be effective catalysts for the oxidation of polyunsaturated fatty acids (33), it seemed worthwhile to examine the interactions between AA and iron. The results of this study demonstrate that ferrous iron is required for AA oxidation and that one product of their interaction is a fatty acid peroxide with potent biological activities.
MATERIALS AND METHODS Chemicals and Reagents (l-14C) arachidonic acid (55 mCi/mmol, New England Nuclear, Boston) and unsaturated fatty acids (> 99% pure) obtained from Nu Chek Prep (Elysian, Mn) were used to study the oxidation process in the test system developed in this investigation. Various inhibitors such as 3-Amino-1,2,4-triazole (AMT), 2,6-Ditertbutyl-P-cresol (BHT), 2, (3) Tert-butyl-4-hydroxyanisole (BHA), Acetylsalicylic acid (ASA), indomethacin (INM), 0-phenanthroline (OP), bovine catalase (BS), imidazole (IMD), 1,3_diphenylisobenzofuran(DPBF) and 3,3'-Diaminobenzidine (DAB), were obtained from Sigma Chemical Company, St. Louis, MO. Superoxide dismutase was obtained from Truett Laboratories, Dallas, Texas. Ferrous sulfate was obtained from Mallinckrodt, St. Louis, MO. and Ferric sulfate'from Allied Chemicals, New Jersey. 2-thiobarbituric acid (TBA) was purchased from Eastman Kodak, Rochester, New York, and 2,3Dihydroxybenzoic acid (DHB) from Aldrich Chemical Company, Wisconsin. Ferrous Iron Mediated AA Oxidation Fatty acid oxidation was measured in two different ways: a) formation of conjugated intermediates and the generation of hydroperoxides was followed by the increase in UV absorbance at 232 nm according to the method of Tappel et al (33). The typical reaction mixture consisted of an ml of deionized water, 160 nanomoles of AA and 120 nanomoles of ferrous sulfate. Absorption of the products formed during the reaction was recorded at selected intervals in a Beckman spectrophotometer b) the same reaction mixtures were analyzed for the presence of thiobarbituric acid reactive compounds by a modified method of Smith et al (34). One hundred microliters of bovine serum albumin were added to each sample to bind unaltered AA and protein precipitation was achieved by reacting with 20% trichloroacetic acid. Protein free extracts were reacted with TBA, and the complex formed measured at 532 nm in a Beckman spectrophotometer. Participation of Ferrous Iron in AA Oxidation To study the role of iron in the oxidation process, the change in the iron redox state as a result of AA oxidation was followed by measuring the unreacted ferrous iron in the reaction mixture using orthopenathroline (OP). OP when complexed with ferrous iron give a brown color which could be quantitated by measuring the absorption at 515 nm. Formation of ferric iron as a result of AA oxidation was confirmed by the method of Wagner et a1 using the Ferrous thiocyanate calorimetric technique (35). 56
Participation of Oxygen in the Ferrous Iron Mediated AA Oxidation Ferrous iron induced oxidation of AA in the presence of inhibitors was followed by quantitating the amount of TBA reacting compound formed. Inhibitors such as antioxidants, iron chelators, aspirin, indomethacin, superoxide dismutase and catalase were used in the test system to study the mechanism of action. The concentrations of each inhibitor are presented in the tables. Characterization of the Product of AA Oxidation In order to isolate and characterize the products of AA oxidation, 250 vgs of AA was reacted with an equal concentration of ferrous iron for 3 minutes at room temperature, acidified with 0.1 n HCl (pH 3.5) and double extracted into diethyl ether. Small samples of the extract were analyzed by high pressure liquid chromatography using a Waters Model 204 LC system equipped with a 254 nm UV detector. A one meter C-18 Bondapak column was used and the solvent system consisted of acetonitrile and water (40:60). Unknown products present in the ether extract were methylated by using the diazomethane method and the methylated derivatives were subjected to them-ionization (C-I) and electron impact (E-I) mass spectrometry by the direct inlet technique. Samples were also prepared for nuclear resonance spectroscopy. About 100 mgs of AA was subjected to the action of ferrous iron in D20 and after the reaction, the oxidized product was extracted into deuterated chloroform and analyzed by NMR spectroscopy. Co-oxygenation of Diphenylisobenzofuran DPBF in acetone, when added to the reaction vessel containing water, has a yellow color with maximum absorption at 420 nm. The change in absorption reflects the oxidation of DPBF. Oxidation of DPBF was followed with a Beckman spectrophotometer using the method of Marnett et al (36). RESULTS Effect of Iron on the Oxidation of Arachidonic Acid Addition of ferrous iron to the reaction mixture containing AA (160 nanomoles/ ml) suspended in water, resulted in the formation of products with strong UV absorption at 232 nm. The reaction was rapid and maximum conversion of AA occurred in less than 90 seconds (Figure 1). With no Fe* in the reaction mixture, there was no detectable autoxidation or UV induced oxidation of AA during the reaction period. To confirm these results, an alternate measure of lipid peroxidation, the thiobarbituric acid (TBA) reaction, was employed. Results of these studies showed (Figure 2) that the generation of TBA reacting material followed a similar time course to generation of the UV absorbing component. With either assay, maximum lipid peroxidation was obtained by adding 160 nanomoles to 120 nanomoles/ml of ferrous iron. Increasing the concentration of AA above this level or incubation for longer than 3 minutes did not increase the amount of AA oxidized. Ferric iron could not substitute for ferrous iron in the oxidation of AA. Generation of superoxide using xanthine-xanthine oxidase also did not cause any lipid peroxidation, suggesting that the ferrous iron was not acting simply by producing superoxide in the medium. 57
AA+ Fe++/f2Oi AA+ Fe++f60/
AA+ Fe+’ f3OJ
AA+
Fe++ /IS/
AA+ Fe++/ZS/ AA+ Fe++ f3.81 AA
Q
Fe++/fZW
180
1
Time lsecondsl
Figure 1. Time dependent oxidation of arachidonic acid (AA) induced by ferrous iron. The oxidation of AA (160 nmoles/ml) was followed by measuring the increase in the UV absorbance of the product formed. Reaction was initiated by the addition of known amounts of ferrous iron (3.8 to 120 nmoles/ml) to the cuvette containing 1 ml of water and AA. All the reactions were conducted at room temperature. 1.0
Xdtihe-
OO
1 60
xonihine ox@se c AA
120
180
T i me /seconds)
Figure 2. Time dependent formation of TBA reacting compounds on addition of ferrous iron to arachidonic acid (AA). The formation of lipid peroxides was followed using the TBA reaction (LP-TBA complex). Samples of AA (160 nmoles/ ml) exposed to ferrous iron (120 nmoles/ml) for varying time intervals were analyzed for TBA reacting compounds. Generation of lipid peroxides as measured in this reaction was similar to the time course of LP production shown in Fig. 1, measured by using UV absorption at 232 nm. Ferric iron could not substitute for ferrous iron in this reaction. Superoxide generated by xanthine-xanthine oxidase had no effect on AA oxidation. 58
Conversion of Ferrous Iron to Ferric Iron Transformation of ferrous iron to ferric iron during AA oxidation was followed by using orthophenanthroline,which complexes specifically with ferrous iron. The complex formed was measured at 515 nm. The results indicate that for each nanomole of AA added, approximately one nanomole of ferrous iron was converted to the ferric form (Figure 3). Generation of ferric iron as a result of AA oxidation was confirmed by the method of Wagner et al (35). Role of Oxygen in Ferrous Iron Mediated AA Oxidation After establishing the absolute requirement of ferrous iron for the oxidation of AA, studies were initiated to explore the role of oxygen in the ferrous iron mediated oxidation process. Test reactions were run in a tonometer with air and with degassed components under nitrogen saturation. As shown in Figure 4, the reaction proceeded normally with air, whereas in the nitrogen atmosphere, no conversion of AA could be observed. Experiments conducted under the same conditions as described above using orthophenathroline to detect the amount of ferrous iron in the reaction mixture revealed that ferrous iron was oxidized when AA was converted in air, but under nitrogen saturation no oxidation of iron could be detected. Ferrous Iron Induced Oxidation of Other Fatty Acids The ability of ferrous iron to oxidize various fatty acids was followed by measuring the amount of UV absorbing compounds formed in the reaction mixture. Results indicate (Fig. 5) that 5,8,11,14 eicosatetranenoic acid, 4, 7,10,13,16,19 docosahexaenoic acid and 6,9,12 octadecatrienoic acid served as better substrates for this reaction, whereas 8,11,14-eicosatrienoicacid was less effective than AA, and monosaturated as well as other fatty acids tested were poor substrates. Effect of Inhibitors on the Oxidation of AA by Ferrous Iron Using the TBA reaction for detection of lipid peroxides as the test system, the effect of various inhibitors was evaluated in order to understand their mechanism of action. Data presented in Table 1 shows that superoxide dismutase and catalase had no inhibitory effect, suggesting that this reaction is not mediated by superoxide anion or hydrogen peroxide or products of their interaction, such as singlet oxygen or hydroxyl free radical. The antioxidant, BHA was a potent inhibitor and probably acts as a competitive species for the oxidative process. EDTA showed a strong inhibitory action by effectively removing iron from the system through chelation. Diaminobenzedine, a strong reducing agent, prevented the conversion of iron from the ferrous to the ferric state, thereby interferinq with the participation of ferrous iron in the oxidation. Iron chelators, imidazole and dihydroxybenzoid acid, when reacted with AA first, gave only a moderate degree of inhibition (33-37%). On the other hand, when these rea ents were reacted with iron first, they gave a high degree of inhibition 483-93%). Aminotriazole did not strongly inhibit AA oxidation. Cycle-oxygenase inhibitors, aspirin and indomethacin, had no inhibitory effect on this system. To further characterize the mechanism of action of imidazole, unutilized ferrous iron was quantitated with orthopenanthroline using various concen59
1.2
5.0
20
AA
nmoles /ml
80
Arachidonic acid (AA) oxidation and transformation of ferrous iron errous iron mediated oxidation of AA was followed by measuring the lipid peroxides (LP) produced as an increase in UV absorption at 232 nm. Each sample contained 120 nmoles/ml of ferrous iron and varying concentrations of AA (1.2 to 160 nmoles/ml). Transformation of ferrous iron was followed by reaction of the unutilized iron with orthophenanthroline (OP) and measuring the colored complex at 515 nm (Fe++-OP complex). AA oxidation was accompanied by equimolar oxidation of ferrous iron to ferric iron.
P=
1.0 Y 3 0.8
8 B
0.6
t 9
0.4
;
B c In 0.2
0
z s 8
Role of oxygen in ferrous iron mediated oxidation of arachidonic To study the role of oxygen in ferrous iron mediated oxidation of AA, reactions were conducted in a tonometer. AA (160 nmoles/ml) was reacted with ferrous iron (120 nmoles/ml) in the presence of oxygen (A) and under nit, rogen saturation (B). After three minutes samples were analyzed for the presence of peroxides using the TBA method or for the ferrous iron using the OP method. In the presence of oxygen significant quantities of peroxides were generated. Under nitrogen saturation no lipid peroxides formed and iron was not oxidized to ferric form. Reaction was studied at only one time point (3 minutes). 60
m P
0.033 + 0.002
EDTA (1 x lo-*M)
2.5
94.7
95.4
90.7
37.7
-
33.5
-
-
11.0
-
% Inhibition
.003
.004
0.042 + 0.001
0.039 2 0.013
0.071 5 0.002
0.102 -+
0.545 + 0.073
0.044 +
0.765 + 0.02
0.856 + 0.01
0.764 + 0.006
0.645 + 0.03
0.715 + 0.06
OD at 532 nm*
9.7
94.1
94.5
90.0
83.8
13.5
93.8
-
-
-
% Inhibition
Inhibitors Reacted with Iron
AC10 (AA)
Effect of inhibitors on the ferrous iron mediated oxidation of arachidonic acid (AA). Using the TBA reaction to detect the lipid peroxides formed, the effect of various inhibitors on ferrous iron mediated oxidation of AA was studied. Each sample contained 160 nmoles/ml of AA and 120 nmoles/ml of ferrous iron. Inhibitors were reacted either for 10 minutes with AA before the addition of iron or with iron for 10 minutes before addition of AA. All iron chelators showed a high degree of inhibition. Aspirin, indomethacin, SOD, and catalase had no inhibitory effect and AMT had little influence on the reaction.
and the standard error (n = 6)
0.029 + 0.005
Butylated hydroxyanisole (1 x lo-*M)
Mean
0.058 + 0.004
Diaminobenzidine (1 x lo-*M)
*
0.392 2 0.007
Dihydroxybenzoic acid (1 x lo-*M)
0.621 + 0.035
Catalase (100 vg/ml)
0.630 + 0.016
0.667 t 0.012
Superoxide dismutase (100 US/ml)
Aminotriazole (1 x lo-*M)
0.657 f. 0.016
Indomethacin (1 x lo-5M)
0.419 + 0.014
0.561 + 0.016
Aspirin (1 x 10m4M)
Imidazole (1 x lo-*M)
0.630 2 0.020
AA (160 nmoles/ml) + Fe++ (120 nmoles/ml)
OD at 532 nm*
Inhibitors Reacted with Arachidonic Acid
EFFECT OF INHIBITORS 0N THE FERROUS IRON (Fe++) MEDIATED OXIDATION 0F A~~ti100NIc
Table 1.
trations of imidazole. Results indicate (Figure 6) that the imidazole bound to ferrous iron, making it unavailable to the orthophenanthroline, and inhibited AA oxidation in a concentration dependent fashion. Characterization of the Oxidation Product In addition to the studies on UV absorption of the products formed, the reaction mixture was double extracted into diethyl ether under acidic conditions and run in a high pressure liquid chromatograph system. The results showed a major peak for the oxidized AA and some minor peaks for unknown components. The products of the ether extract of the reaction mixture were methylated using diazomethane. Methylated compounds were analyzed using mass spectrometry by the direct inlet probe method. Analysis of the methylated derivative of arachidonic acid used showed the typical mass spectrum of methyl-AA with M = 318. There were no high mass peaks above 320 indicating an absence of autoxidized products. On the other hand, the ferrous iron oxidized samples showed a peak at 348 (M + 30), indicating an addition of 02 associated with a loss of two H, and another peak at 378 (M + 60) indicating addition of 02 molecules with loss of four H. Ozidized AA was also extracted into deuterated chloroform and subjected to 13C-NMR spectroscopy. This revealed losses of one or more double bonds from the oxidized sample, whereas the spectrum of the control sample run under similar conditions exactly matched the spectrum of authentic AA. Oxygenation of 1,3-Diphenylisobenzofuran(DPBF) As a further proof for the generation of AA peroxides, co-oxygenation of DPBF was studied. Since the substrate for this reaction was made up in acetone, it was necessary to eliminate the effect of acetone on the test system. Stock solutions of 100 mM and 10 mM DPBF were prepared and the amount of acetone used in any reaction never exceeded 10 ul/ml. Results shown in figure 7 demonstrate loss of absorbance indicating the oxidation of DPBF, starting from one minute after the introduction of AA and iron into the system. When Fe++ or AA were used alone no oxidation of DPBF was observed. When the reactions were run with the antioxidant, BHA, or the iron chelator, DAB, there was no co-oxygenation, indicating that AA oxidation promoted the co-oxygenation of DPBF in this reaction. The delay of 60 seconds in the reaction could be explained by our earlier findings that the peak of AA oxidation is at about 60 seconds. Co-oxygenation of DPBF was not inhibited by aspirin and, as shown previously, aspirin does not block the oxidation of AA mediated by ferrous iron. The product of DPBF oxygenation was isolated as described by Marnett et al and identified as 0-dibenzoylbenzene by mass spectroscopy [prominent ions at me/e 386 (M) 209 (M-77) and 105](36). DISCUSSION Based on the concept that iron is intimately involved in the synthesis of endo. peroxides and thromboxanes by blood platelets (11,29-32), the present investigation has explored the interaction of iron and arachidonic acid (AA), the precursor of prostaglandin products. Results of the study indicate that iron is a potent catalyst for the oxidation of AA. The reaction, followed by the increase in UV absorbtion at 232 nm, occurred rapidly, reaching a peak in less than 90 seconds. An alternate measure of lipid peroxidation, the TBA reaction 62
ET*120:1/ 00
f/8:3/
0.6 z 3 6
0.6
Et=
tm:3/
Et,,
120:3J
t ii ‘i
0.4
Ho 117../J I/ 04 f/8:11 ED 120:2/
8
EA (22:
0.2
I
I
I
60
120
180
Time
fseccanfs)
Fatty acid specificity of ferrous iron mediated oxidation. To stu y t e effect of ferrous iron on the oxidation of various unsaturated fatty acids, individual substrates (160 nmoles/ml) were exposed to ferrous iron (120 nmoles/ml) for three minutes and the UV absorbing products generated were measured at 232 nm. Degree of unsaturation and the proximity of the double bond to the carboxyterminal favored the oxidation of substrates by ferrous iron. ET4 = 5,8,11,14 Eicosatetraenoic acid (arachidonic acid); DH = 4,7,10,13,16,19 Docosahexaenoic acid; OD = 6,9,12 Qctadecatrienoic acid; ET3A = 8,11,14 Eicosatrienoic acid; ET3B = 11,14,17 Eicosatrienoic acid; HD = 10,Heptadecaenoic acid; EA = Erucic acid, OA = Oleic acid; ED = 11,14 Eicosadienoic acid.
-F--i
lmidozole
/mMI
Fjgure 6. Concentration dependent inhibition of arachidonic acid (AA) oxidation by imidazole. Using the TBA reaction to detect the formation of lipid peroxides, the effect of various concentrations of imidazole on the ferrous iron mediated AA oxidation was studied. Imidatole showed concentration dependent inhibition of AA oxidation. With the same concentrations of imidazole reactions were analyzed for the amount of unreacted ferrous iron with the orthophenanthrolinemethod. Oxidation of AA followed closely the oxidation of ferrous iron. O-O oxidized AA-TBA complex; Ferrous iron-Orthophenanthroline complex. 63
/with c @ z c g * s
DA8 or WA
I
0.8
0.6
0.4
8 0.2
00-
60
120 Time
lea
/saw&/
Figure 7. Co-oxygenation of diphenylisobenzofuran (DPBF) mediated by the oxidation of arachidonic acid (AA). The effect of transformation of AA on the oxidation of DPBF was followed by the decrease in absorbance at 420 nm. DPBF in acetone was added (0.05 mM) to a cuvette containing AA (160 nmoles/ml) and the reaction was initiated by the addition of ferrous iron (120 nmoles/ml). Addition of AA or iron alone to the substrate had no effect (O-O). When AA and iron were added together loss of absorbance due to the oxidation of DPBF could be followed (X-X). Aspirin had no inhibitory effect (h -----II), whereas DAB and BHA blocked the reaction (-----).
64
gave a similar result. The degree of AA peroxidation varied with the concentration of both reactants. When Fe+ concentration was maintained at 120 nanomoles/ml, the extent of oxidation depended on the amount of AA added until maximum conversion was reached at 160 nanomoles AA. Addition of more AA or prolonging the reaction time had no effect on the amount of AA oxidized. The reaction was dependent of the valence of iron. Ferrous iron was required and the ferric form was inert in catalyzing the oxidation of AA. Studies employing orthophenanthroline (OP) demonstrated that ferrous iron was oxidized to the ferric state during the reaction. The reaction between iron and AA was dependent on the presence of oxygen. Experiments carried out in a tonometer replacing air with a nitrogen atmosphere, revealed no conversion of AA due to interaction with iron under anaerobic conditions. The nature of the products formed during oxidation of AA by ferrous iron was examined in several ways. Maximum UV absorbtion of the reaction products occurred at 232 nm, and Tappel has demonstrated that lipid peroxides have a higher molar extinction coefficient at this wave length than other products of autoxidation formed during the process (37). The identification of the product by GC-MS was difficult as the substance was unstable and did not survive the gas chromatograph column. However, using the direct inlet probe method, evidence for addition of one and two oxygen molecules to the arachidonic acid was obtained together with loss of either 2 or 4 hydrogens. In AA, with four double bonds, the carbon atoms at 4,7,10 and 13 are "omethylene" carbons with weak bonds to their hydrogen atoms (38). The carbons in these positions are partially activated and the molecular oxygen in association with Fe++ probably abstracts the hydrogens from these positions, with addition of an oxygen molecule or molecules. The resulting product could be a single component, but is more likely a mixture of several peroxides, hydroperoxides and perhaps cyclic peroxides. Peroxides and cyclic peroxides would be expected to break down to malondialdehyde, or a malondialdehyde-like compound during processing for the TBA reaction (39), accounting for the positive TBA reaction seen. Although our interest was primarily on the interactions between iron and AA, we also examined the oxidation of other fatty acids by ferrous iron. Monounsaturated fatty acids yielded little oxidation product when exposed to ferrous iron. Several unsaturated fatty acids other than AA were good substrates and the amount of oxidation in the presence of iron depended on the degree of unsaturation. Thus, 4,7,10,13,16,19 docosahexaenoic acid, for example, formed more oxidation product when exposed to ferrous iron than AA under similir conditions. The proximity of the double bond of the carboxy terminal was also important. Oxidation of fatty acids with double bonds from 4-11 yielded more product than those having double bonds beyond carbon 10. Investigations into the influences of inhibitors and possible promoters on the ferrous iron induced oxidation of AA revealed additional features of the reaction. Although superoxide anion supposedly initiates the peroxidation of polyunsaturated fatty acids, it had no effect on AA alone. Also, superoxide dismutase failed to inhibit the ferrous iron-AA interaction. Harbor and Weiss have proposed a mechanism in which hydrogen peroxide mediates the change in the redox state of iron (40). Catalase, however, had no inhibitory effect on the ferrous iron induced oxidation of AA, suggesting that H202 65
probably has no significant role in this reaction. Tappel (41) found evidence to show that bonding of hydroxyl ion to the sixth bond position of iron in hematin compounds at pH 7.0 influences their peroxidative activity. His work on the effect of increased hydroxyl ion on hemoglobin catalyzed linoleate oxidation showed that the rate of peroxidation increased 2 fold. He proposed a scheme in which fatty acid and hematin form a complex mediated by hydroxyl ion which is reversible. Such a complex may very well form with ferrous iron and AA in our test system, or iron may directly interact with AA in the presence of oxygen to form a complex. The effects of other inhibitory agents on the ferrous iron mediated conversion of AA confirmed the importance of the valence state of iron and revealed insights into the mechanism through which antioxidants may block the process of oxidation. EDTA effectively removed ferrous iron from the reaction through chelation, and DAB, a strong reducing agent, blocked conversion of ferrous iron to the ferric state, an essential step in the reaction. The inhibitory effects of imidiazole and DHB were more complex. When reacted with AA first, they were far less inhibitory than when incubated with iron prior to initiation of the reaction. Thus, they inhibit the reaction by binding ferrous iron, but the complex takes more time to form than chelation of ferrous iron by EDTA. BHA, on the other hand, was a potent inhibitor, but did not influence the state of iron. Rather BHA appeared to act as a competitive substrate for the oxidation process, in essence replacing AA. The ability of the peroxidase inhibitor, aminotriazole, to block platelet prostaglandin synthesis, secretion, and aggregation first stimulated our interest in the role of iron in the peroxidation of AA to endoperoxides and thromboxanes (28). However, AMT had only a slight inhibitory effect on the ferrous iron induced oxidation of AA. We further explored the effect of AMT using diaminobenzidine (DAB) which reduced ferric iron to ferrous iron. In the presence of DAB, ferric iron was found to give a positive reaction with orthophenanthroline, indicating ferrous iron was now present. However, if ferric iron was reacted first with AMT before exposing the ferric iron to DAB, the orthophenanthraline reaction was markedly inhibited. Thus, the failure of AMT to inhibit ferrous iron induced oxidation of AA is understandable, because it binds ferric but not ferrous iron. In biological systems, however, the redox state of iron is closely regulated to serve basic cellular functions (41). Oscillations in valence are essential for the activity of iron in a variety of reactions. As a result, agents such as AMT, which trap iron in the ferric state effectively block oxidation and peroxidation reactions requiring the ferrous form, though the inhibition may require prolonged incubation. During the process of prostaglandin synthesis in platelets, hydroperoxides are generated (9). Marnett et al (36) on the basis of their studies suggested that co-oxygenation of substrates including DPBF, was mediated by hydroperoxide intermediates formed during prostaglandin synthesis because PGQ, which has a 15 hydroperoxy group, was effective but P6H2 was not. Our finding in the present investigation that ferrous iron mediated oxidation of AA can induce co-oxygenation of DPBF offers further evidence for 66
the generation of hydroperoxides in the system and supports the findings of Marnett et al (36) relating hydroperoxide intermediates to co-oxygenatton reactions. The results of the present investigation have a number of important implications. Several workers have pointed out previously the importance of heme groups to the activity of cycle-oxygenase and other enzymes involved in synthesis of endoperoxides, thromboxanes and additional products of prostaglandin synthesis (11,29-32,42). Examination of the interaction between AA and iron in the present study suggests that iron is not only a critical factor in the oxidation, but that it must be in the ferrous form. This finding is consistent with the electron paramagnetic resonance (EPR). Studies on lipoxygenase show that in the native enzyme iron is present in EPR silent ferrous state (43). In addition,it has been shown that the cycle-oxygenase activity of sheep vesicular gland microsomes is inhibited by 0-phenthroline, which is capable of reversibly inhibiting enzymes by either preventing pyridine nucleotide binding or by directly binding ferrous iron (44). Evaluation of the effects of inhibitors in our studies has demonstrated that the mechanism by which they block AA oxidation relates to their ability to bind ferrous iron or prevent its conversion to the ferric state. Lipid oxidation or peroxidation is a major problem in food preservation and aging. A search for effective antioxidants based on their interaction with iron may result in the design of more effective agents. Thus, single agents or a combination which binds ferrous or ferric iron or keeps the iron reduced and competes with AA and other fatty acids as a substrate for oxygenation could be valuable. The biological implications of the present investigations are also important. Transformation of AA to a more polar product, such as a peroxide or hydroperoxide, may facilitate its movement in cells to sites of specific enzymes. Preliminary studies suggest that the fatty acid peroxide generated during the oxidation of AA by ferrous iron can block the formation of prostaglandin (PG12) in vessel walls (45). Thus, the interaction of AA with ferrous iron may be one of the critical steps involved in the prostaglandin synthetic pathway and products formed as a direct result of ferrous iron induced oxidation of AA may serve to regulate the activity of specific enzymes such as prostacyclin synthetase (46,47). Further investigations currently in progress are attempting to determine if the interaction between ferrous iron and AA is a primary step of prostaglandin synthesis in cells. ACKNOWLEDGEMENTS We thank Dr. Thomas P. Krick, Department of Biochemistry, University of Minnesota for his technical assistance in mass spectroscopy. Supported by USPHS grants HL-11880, AM-06317, HL-06314, CA-12607, CA-08832, CA-11996, GM-AM-22167, HL-20695, HL-16833 and a grant from the Leukemia Task Force. REFERENCES 1.
Bergstrom S, Danielsson H, Samuelsson B. The enzymatic formation of prostaglandin Ep from arachidonic acid. Biochim Biophys 90:207, 1964.
2.
Bills TK, Smith JB, Silver MJ. Metabolism of (14C) arachidonic acid by human platelets. Biochim Biophys Acta 424:303, 1976. 67
3.
Nugteren DH. Arachidonate lipoxygenase in blood platelets. Biochim Biophys Acta 380:299, 1975.
4.
Smith WL, Lands WEM. Oxygenation of polyunsaturated fatty acids during prostaglandin synthesis by sheep vesicular gland. Biochemistry 11: 3276, 1972.
5.
Van Dorp DA, Beerthuis RK, Nugteren DM and Vonkeman H. The biosynthesis of prostaglandins. Biochim Biophys Acta 90:204, 1964.
6.
Hamberg M, Smauelsson B. Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Nat1 Acad Sci USA 70:899, 1973.
7.
Hamberg M, Samuelsson B. Prostaglandin endoperoxides. Novel transformations of arachidonic acid in human platelets. Proc Nat1 Acad Sci USA 71:3400, 1974.
8.
Hamberg M, Svensson J, Wakabayashi T, Samuelsson B. Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Nat1 Acad Sci USA 71:345, 1974a.
9.
Hamberg M, Svensson J, Samuelsson B. Prostaglandin endoperoxides. A new concept concerning the mode of action and release of prostaglandins. Proc Nat1 Acad Sci USA 71:3824, 1974b.
10.
Needleman P,Minkes M, Raz A. Thromboxanes: Selective biosynthesis and distinct biological properties. Science 193:163, 1976.
11.
Ho PPK, Walters P, Sullivan HL. Biosynthesis of thromboxane B2: assay, isolation, and properties of the enzyme system in human platelets. Prostaglandins 12:951, 1976.
12.
Moncada S, Needleman P, Bunting S, Vane JR. Prostaglandin endoperoxide and thromboxane generating systems and their selective inhibition. Prostaglandins 12:323, 1976.
13.
Willis AL. Isolation of a chemical trigger for thrombosis. Prostaglandins 5:1, 1974.
14.
Willis AL, Vane FM, Kuhn DC, Scott CG, Petrin M. An endoperoxide aggregator (LASS), formed in platelets in response to thrombotic stimuli: Purification, identification and unique biological significance. Prosta, glandins 8:453, 1974.
15.
Hamberg M, Svensson J, Wakabayashi T, Samuelsson B. Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Nat1 Acad Sci USA 71:345, 1974.
16.
Labile agGerrard JM, White JG, Rao GHR, Krivit W, Witkop CJ Jr. gregation stimulating substance (LASS). The factor from storage pool deficient platelets correcting defective aggregation and release of aspirin treated normal platelets. Br J Haematol 29:657, 1975.
17.
Gerrard JM, White JG. The influence of prostaglandin endoperoxides on platelet ultrastructure. Am J Path01 80:189, 1975.
18.
Hamberg M, Svensson J, Samuelsson B. Thromboxanes: A new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Nat1 Acad Sci USA 72:2994, 1975.
68
19.
Malmsten C, Hamberg M, Svensson J, Samuelsson B. Physiological role of an endoperoxide in human platelets: Hemostatic defect due to platelet cycle-oxygenase deficiency. Proc Nat1 Acad Sci USA 72:1446, 1975.
20.
Panganamala RV, Brownlee NR, Sprecher H, Cornwell DG. Evaluation of superoxide anion and singlet oxygen in the biosynthesis of prostaglandins from eicosa-8,11,14-Trienoicacid. Prostaglandins 7:21, 1974.
21.
Panganamala RV, Sharma HH, Sprecher H, Geer JC, Cornwell DG. A suggested role for hydrogen peroxide in the biosynthesis of prostaglandins. Prosta. glandins 8:3, 1974.
22.
Panganamala RV, Sharma HH, Heikkila RE, Geer JC, Cornwell DG. Role of hydroxyl radical scavengers dimethyl sulfoxide, alcohols and methional in the inhibition of prostaglandin biosyntheses. Prostaglandins 11:599, 1976.
23.
O'Brien PJ, Rahimtula A. The possible involvement of a peroxidase in prostaglandin biosyntheses. Biochem Biophys Res Conm 70:832, 1976.
24.
O'Brien PJ, Hawco FJ, Piatt JF, Rahimtula AD. Enzymic singlet oxygen formation from peroxides in biological systems and its effects. Comm Int Conf Singlet oxygen related spec in Chem Biol August 21-26, 1977 Pinawa, Manitoba, Canada.
25.
Rao GHR, Gerrard JM, Eaton J, White JG: Detection of arachidonic acid (AA) free radical formation, an essential step in platelet prostaglandin (PG) synthesis. Blood 48:977, 1976.
26.
Rao GHR, White JG. Effects of vitamin E (VE) and nitroblue tetrazolium (NBT) on platelet ultrastructure and aggregation. Fed Proc 36:350, 1977.
27.
Rao GHR, Gerrard JM, Eaton JW, White JG. Arachidonic acid peroxidation. prostaglandin synthesis and platelet function. (In press) Comn Int Conf Singlet oxygen and related species in chemistry and biology. August 21-26, 1977 Pinawa, Manitoba, Canada.
28.
Gerrard JM, White JG, Rao GHR, Townsend D. Localization of platelet prostaglandin production in the platelet dense tubular system. Am J Path01 83:283, 1976.
29.
Ho PPK, Walters P, Sullivan HR. A particulate arachidonate lipoxygenase in human blood platelets. Biochem Biophys Res Comm 76:398, 1977.
30.
Bild GS, Bhat SG, Ramadoss CS, Axelrod B. Biosynthesis of a prostaglandin by a plant enzyme. J Biol Chem 253:21, 1978.
31.
Yoshimoto A, Ito H, Tomita K. Cofactor requirements of the enzyme synthesizing prostaglandin in bovine seminal vesicles. J Biol Chem 68~487, 1970.
32.
Miyamoto T, Ogino N, Yamamoto S, Hayaishi 0. Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J Biol Chem 251:2629, 1976.
33.
Tappel AL, Boxer PD, Lundberg WO. The reaction mechanism of soybean lipoxidase. J Biol Chem 199:267, 1952.
34.
Smith JB, Ingerman CM, Silver MJ. Malondialdehyde formation as an indicator of prostaglandin production by human platelets. J Lab Clin Med 88:167, 1976. 69
35.
Wagner CO, Clever HL, Peters ED. Evaluation of the ferrous thiocyanate calorimetric method. Ind Eng Chem 19:980, 1947.
36.
Marnett LJ, Wlodawer P, Samuelsson B. Co-oxygenation of organic substrates by the prostaglandin synthetase of sheep vesicular gland. J Biol Chem 250:8510, 1975.
37.
Tappel AL. The mechanism of the oxidation of unsaturated fatty acids catalyzed by hematin compounds. Arch Biochem and Biophys 44:378, 1953.
38.
Demopoulos HB. 1859, 1973.
39.
Pryor WA, Stanley JP, Blair E. Autoxidation of fatty acids. II. A suggested mechanism for the formation of TBA-reactive materials from prostaglandin like endoperoxides. Lipids 11:370, 1976.
40.
Harber F, Weiss J. The catalytic decomposition of hydrogen peroxide by iron salts. Proc Roy Sot Lond Ser A 147:332, 1934.
41.
Tappel AL. Unsaturated lipid oxidation catalyzed by hematin compounds. J Biol Chem 217:721, 1955.
42.
Hemler M, Lands WEM. Purification of the cycle-oxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyme. J Biol Chem 251:5575, 1976.
43.
Egmond MR, Finazzi-Agro A, Fasella PM, Veldink GA, Vliegenthart JFG. Changes in the fluorescence and absorbance of lipoxygenase-1 induced by 13-Ls-hydroperoxylinoleicacid and linoleic acid. Biochim et Biophys Acta 397:43, 1975.
44.
Smith WL, Lands WEM. Stimulation and blockade of prostaglandin biosynthesis. J Biol Chem 246:6700, 1971.
45.
Rao GHR, Gerrard JM, Eaton JW, Clawson CC, White JG. The role of iron in prostaglandin biosynthesis: Ferrous iron mediated oxidation of arachidonic acid. Fed Proc 37:249A, 1978.
46.
Moncada S, Gryglewski RJ, Bunting S, Vane JR. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature (London) 263:663, 1976.
47.
Moncada S. A lipid peroxideinhibitsthe enzyme in blood vessel microsomes that generates from prostaglandin endoperoxides, the substance (Prostaglandin X) which prevents platelet aggregation. Prostaglandins 12:715, 1976.
The basis of free radical pathology. Fed Proc 32:
and Medicine 1: 55-70, 1978.
THE ROLE OF IRON IN PROSTAGLANDIN SYNTHESIS: FERROUS IRON MEDIATED OXIDATION OF ARACHIDONIC ACID G.H.R. Rao, J.M. Gerrard, J.W., Eaton, J.G. White. University of Minnesota Health Sciences Center, Box 198 Mayo Memorial Building, Minneapolis, Minnesota 55455 (reprint requests to GHRR) ABSTRACT Arachidonic acid (AA) is the essential substrate for production of platelet endoperoxides and thromboxanes. Iron or heme is an essential cofactor for the peroxidase, lipoxygenase and cycle-oxygenase enzymes involved in formation of these products. The present study has examined the direct interactions between iron and arachidonic acid. Iron caused the oxidation of AA into more polar products which could be detected by UV absorbtion at 232 nM or the thiobarbituric acid (TBA) reaction. High pressure liquid chromatography, them-ionization and electron-impact mass spectrometry and nuclear magnetic resonance spectroscopy suggest that the major product was a hydroperoxide of AA. Ferrous iron (Fe++) and oxygen were absolute requirements. Fe++ was converted to the ferric iron (Fe+++) state during oxidation of AA, but Fe+++ could not substitute for Fe++. No other enzymes, cofactors or ions were involved. Conversion of AA to a hydroperoxide by Fe++ was inhibited by the antioxidant, 2, (3)-Tert-butyl-4-hydroxyanisole,the radical scavenger, nitroblue tetrazolium, and iron chelating agents, including EDTA, imidazole and dihydroxybenzoic acid. The reaction was not affected by superoxide dismutase, catalase or aspirin. These findings and preliminary studies of the Fe* induced oxidation product of AA as a substrate for prosta, glandin synthesis and inhibitor of prostacyclin production indicate the critical role of Fe++ in AA activation. INTRODUCTION Arachidonic acid (AA) is a fatty acid substrate for platelet lipoxy enase and cycle-oxygenase (l-4). Its oxidation yields the endoperoxides 95-9), thromboxanes (10-12) and other products important to the function of platelets in hemostasis (13-19). Several authors have suggested that conversion of AA involves participation of free radicals or development of a free radical state by the fatty acid, but no definitive requirement for hydroperoxy radical, superoxide anion, singlet oxygen, hydroxyl free radical or hydrogen peroxide has been demonstrated (20-27). Our interest in the primary phase of AA peroxidation and free radical generation was stimulated by the observation that a peroxidase might be involved because 3-Amino-1,2,4-triazole, an inhibitor of peroxidase, blocked platelet endoperoxide and throm-
55
boxane synthesis, secretion and aggregation (28). The peroxidase enzyme contains iron, and heme groups or iron have been linked to the activity of lipoxygenase (29-30), cycle-oxygenase (31,32) and thromboxane synthetase (11), though the precise relationships of iron to the enzyme reactions is not yet clear. Since heme proteins are known to be effective catalysts for the oxidation of polyunsaturated fatty acids (33), it seemed worthwhile to examine the interactions between AA and iron. The results of this study demonstrate that ferrous iron is required for AA oxidation and that one product of their interaction is a fatty acid peroxide with potent biological activities.
MATERIALS AND METHODS Chemicals and Reagents (l-14C) arachidonic acid (55 mCi/mmol, New England Nuclear, Boston) and unsaturated fatty acids (> 99% pure) obtained from Nu Chek Prep (Elysian, Mn) were used to study the oxidation process in the test system developed in this investigation. Various inhibitors such as 3-Amino-1,2,4-triazole (AMT), 2,6-Ditertbutyl-P-cresol (BHT), 2, (3) Tert-butyl-4-hydroxyanisole (BHA), Acetylsalicylic acid (ASA), indomethacin (INM), 0-phenanthroline (OP), bovine catalase (BS), imidazole (IMD), 1,3_diphenylisobenzofuran(DPBF) and 3,3'-Diaminobenzidine (DAB), were obtained from Sigma Chemical Company, St. Louis, MO. Superoxide dismutase was obtained from Truett Laboratories, Dallas, Texas. Ferrous sulfate was obtained from Mallinckrodt, St. Louis, MO. and Ferric sulfate'from Allied Chemicals, New Jersey. 2-thiobarbituric acid (TBA) was purchased from Eastman Kodak, Rochester, New York, and 2,3Dihydroxybenzoic acid (DHB) from Aldrich Chemical Company, Wisconsin. Ferrous Iron Mediated AA Oxidation Fatty acid oxidation was measured in two different ways: a) formation of conjugated intermediates and the generation of hydroperoxides was followed by the increase in UV absorbance at 232 nm according to the method of Tappel et al (33). The typical reaction mixture consisted of an ml of deionized water, 160 nanomoles of AA and 120 nanomoles of ferrous sulfate. Absorption of the products formed during the reaction was recorded at selected intervals in a Beckman spectrophotometer b) the same reaction mixtures were analyzed for the presence of thiobarbituric acid reactive compounds by a modified method of Smith et al (34). One hundred microliters of bovine serum albumin were added to each sample to bind unaltered AA and protein precipitation was achieved by reacting with 20% trichloroacetic acid. Protein free extracts were reacted with TBA, and the complex formed measured at 532 nm in a Beckman spectrophotometer. Participation of Ferrous Iron in AA Oxidation To study the role of iron in the oxidation process, the change in the iron redox state as a result of AA oxidation was followed by measuring the unreacted ferrous iron in the reaction mixture using orthopenathroline (OP). OP when complexed with ferrous iron give a brown color which could be quantitated by measuring the absorption at 515 nm. Formation of ferric iron as a result of AA oxidation was confirmed by the method of Wagner et a1 using the Ferrous thiocyanate calorimetric technique (35). 56
Participation of Oxygen in the Ferrous Iron Mediated AA Oxidation Ferrous iron induced oxidation of AA in the presence of inhibitors was followed by quantitating the amount of TBA reacting compound formed. Inhibitors such as antioxidants, iron chelators, aspirin, indomethacin, superoxide dismutase and catalase were used in the test system to study the mechanism of action. The concentrations of each inhibitor are presented in the tables. Characterization of the Product of AA Oxidation In order to isolate and characterize the products of AA oxidation, 250 vgs of AA was reacted with an equal concentration of ferrous iron for 3 minutes at room temperature, acidified with 0.1 n HCl (pH 3.5) and double extracted into diethyl ether. Small samples of the extract were analyzed by high pressure liquid chromatography using a Waters Model 204 LC system equipped with a 254 nm UV detector. A one meter C-18 Bondapak column was used and the solvent system consisted of acetonitrile and water (40:60). Unknown products present in the ether extract were methylated by using the diazomethane method and the methylated derivatives were subjected to them-ionization (C-I) and electron impact (E-I) mass spectrometry by the direct inlet technique. Samples were also prepared for nuclear resonance spectroscopy. About 100 mgs of AA was subjected to the action of ferrous iron in D20 and after the reaction, the oxidized product was extracted into deuterated chloroform and analyzed by NMR spectroscopy. Co-oxygenation of Diphenylisobenzofuran DPBF in acetone, when added to the reaction vessel containing water, has a yellow color with maximum absorption at 420 nm. The change in absorption reflects the oxidation of DPBF. Oxidation of DPBF was followed with a Beckman spectrophotometer using the method of Marnett et al (36). RESULTS Effect of Iron on the Oxidation of Arachidonic Acid Addition of ferrous iron to the reaction mixture containing AA (160 nanomoles/ ml) suspended in water, resulted in the formation of products with strong UV absorption at 232 nm. The reaction was rapid and maximum conversion of AA occurred in less than 90 seconds (Figure 1). With no Fe* in the reaction mixture, there was no detectable autoxidation or UV induced oxidation of AA during the reaction period. To confirm these results, an alternate measure of lipid peroxidation, the thiobarbituric acid (TBA) reaction, was employed. Results of these studies showed (Figure 2) that the generation of TBA reacting material followed a similar time course to generation of the UV absorbing component. With either assay, maximum lipid peroxidation was obtained by adding 160 nanomoles to 120 nanomoles/ml of ferrous iron. Increasing the concentration of AA above this level or incubation for longer than 3 minutes did not increase the amount of AA oxidized. Ferric iron could not substitute for ferrous iron in the oxidation of AA. Generation of superoxide using xanthine-xanthine oxidase also did not cause any lipid peroxidation, suggesting that the ferrous iron was not acting simply by producing superoxide in the medium. 57
AA+ Fe++/f2Oi AA+ Fe++f60/
AA+ Fe+’ f3OJ
AA+
Fe++ /IS/
AA+ Fe++/ZS/ AA+ Fe++ f3.81 AA
Q
Fe++/fZW
180
1
Time lsecondsl
Figure 1. Time dependent oxidation of arachidonic acid (AA) induced by ferrous iron. The oxidation of AA (160 nmoles/ml) was followed by measuring the increase in the UV absorbance of the product formed. Reaction was initiated by the addition of known amounts of ferrous iron (3.8 to 120 nmoles/ml) to the cuvette containing 1 ml of water and AA. All the reactions were conducted at room temperature. 1.0
Xdtihe-
OO
1 60
xonihine ox@se c AA
120
180
T i me /seconds)
Figure 2. Time dependent formation of TBA reacting compounds on addition of ferrous iron to arachidonic acid (AA). The formation of lipid peroxides was followed using the TBA reaction (LP-TBA complex). Samples of AA (160 nmoles/ ml) exposed to ferrous iron (120 nmoles/ml) for varying time intervals were analyzed for TBA reacting compounds. Generation of lipid peroxides as measured in this reaction was similar to the time course of LP production shown in Fig. 1, measured by using UV absorption at 232 nm. Ferric iron could not substitute for ferrous iron in this reaction. Superoxide generated by xanthine-xanthine oxidase had no effect on AA oxidation. 58
Conversion of Ferrous Iron to Ferric Iron Transformation of ferrous iron to ferric iron during AA oxidation was followed by using orthophenanthroline,which complexes specifically with ferrous iron. The complex formed was measured at 515 nm. The results indicate that for each nanomole of AA added, approximately one nanomole of ferrous iron was converted to the ferric form (Figure 3). Generation of ferric iron as a result of AA oxidation was confirmed by the method of Wagner et al (35). Role of Oxygen in Ferrous Iron Mediated AA Oxidation After establishing the absolute requirement of ferrous iron for the oxidation of AA, studies were initiated to explore the role of oxygen in the ferrous iron mediated oxidation process. Test reactions were run in a tonometer with air and with degassed components under nitrogen saturation. As shown in Figure 4, the reaction proceeded normally with air, whereas in the nitrogen atmosphere, no conversion of AA could be observed. Experiments conducted under the same conditions as described above using orthophenathroline to detect the amount of ferrous iron in the reaction mixture revealed that ferrous iron was oxidized when AA was converted in air, but under nitrogen saturation no oxidation of iron could be detected. Ferrous Iron Induced Oxidation of Other Fatty Acids The ability of ferrous iron to oxidize various fatty acids was followed by measuring the amount of UV absorbing compounds formed in the reaction mixture. Results indicate (Fig. 5) that 5,8,11,14 eicosatetranenoic acid, 4, 7,10,13,16,19 docosahexaenoic acid and 6,9,12 octadecatrienoic acid served as better substrates for this reaction, whereas 8,11,14-eicosatrienoicacid was less effective than AA, and monosaturated as well as other fatty acids tested were poor substrates. Effect of Inhibitors on the Oxidation of AA by Ferrous Iron Using the TBA reaction for detection of lipid peroxides as the test system, the effect of various inhibitors was evaluated in order to understand their mechanism of action. Data presented in Table 1 shows that superoxide dismutase and catalase had no inhibitory effect, suggesting that this reaction is not mediated by superoxide anion or hydrogen peroxide or products of their interaction, such as singlet oxygen or hydroxyl free radical. The antioxidant, BHA was a potent inhibitor and probably acts as a competitive species for the oxidative process. EDTA showed a strong inhibitory action by effectively removing iron from the system through chelation. Diaminobenzedine, a strong reducing agent, prevented the conversion of iron from the ferrous to the ferric state, thereby interferinq with the participation of ferrous iron in the oxidation. Iron chelators, imidazole and dihydroxybenzoid acid, when reacted with AA first, gave only a moderate degree of inhibition (33-37%). On the other hand, when these rea ents were reacted with iron first, they gave a high degree of inhibition 483-93%). Aminotriazole did not strongly inhibit AA oxidation. Cycle-oxygenase inhibitors, aspirin and indomethacin, had no inhibitory effect on this system. To further characterize the mechanism of action of imidazole, unutilized ferrous iron was quantitated with orthopenanthroline using various concen59
1.2
5.0
20
AA
nmoles /ml
80
Arachidonic acid (AA) oxidation and transformation of ferrous iron errous iron mediated oxidation of AA was followed by measuring the lipid peroxides (LP) produced as an increase in UV absorption at 232 nm. Each sample contained 120 nmoles/ml of ferrous iron and varying concentrations of AA (1.2 to 160 nmoles/ml). Transformation of ferrous iron was followed by reaction of the unutilized iron with orthophenanthroline (OP) and measuring the colored complex at 515 nm (Fe++-OP complex). AA oxidation was accompanied by equimolar oxidation of ferrous iron to ferric iron.
P=
1.0 Y 3 0.8
8 B
0.6
t 9
0.4
;
B c In 0.2
0
z s 8
Role of oxygen in ferrous iron mediated oxidation of arachidonic To study the role of oxygen in ferrous iron mediated oxidation of AA, reactions were conducted in a tonometer. AA (160 nmoles/ml) was reacted with ferrous iron (120 nmoles/ml) in the presence of oxygen (A) and under nit, rogen saturation (B). After three minutes samples were analyzed for the presence of peroxides using the TBA method or for the ferrous iron using the OP method. In the presence of oxygen significant quantities of peroxides were generated. Under nitrogen saturation no lipid peroxides formed and iron was not oxidized to ferric form. Reaction was studied at only one time point (3 minutes). 60
m P
0.033 + 0.002
EDTA (1 x lo-*M)
2.5
94.7
95.4
90.7
37.7
-
33.5
-
-
11.0
-
% Inhibition
.003
.004
0.042 + 0.001
0.039 2 0.013
0.071 5 0.002
0.102 -+
0.545 + 0.073
0.044 +
0.765 + 0.02
0.856 + 0.01
0.764 + 0.006
0.645 + 0.03
0.715 + 0.06
OD at 532 nm*
9.7
94.1
94.5
90.0
83.8
13.5
93.8
-
-
-
% Inhibition
Inhibitors Reacted with Iron
AC10 (AA)
Effect of inhibitors on the ferrous iron mediated oxidation of arachidonic acid (AA). Using the TBA reaction to detect the lipid peroxides formed, the effect of various inhibitors on ferrous iron mediated oxidation of AA was studied. Each sample contained 160 nmoles/ml of AA and 120 nmoles/ml of ferrous iron. Inhibitors were reacted either for 10 minutes with AA before the addition of iron or with iron for 10 minutes before addition of AA. All iron chelators showed a high degree of inhibition. Aspirin, indomethacin, SOD, and catalase had no inhibitory effect and AMT had little influence on the reaction.
and the standard error (n = 6)
0.029 + 0.005
Butylated hydroxyanisole (1 x lo-*M)
Mean
0.058 + 0.004
Diaminobenzidine (1 x lo-*M)
*
0.392 2 0.007
Dihydroxybenzoic acid (1 x lo-*M)
0.621 + 0.035
Catalase (100 vg/ml)
0.630 + 0.016
0.667 t 0.012
Superoxide dismutase (100 US/ml)
Aminotriazole (1 x lo-*M)
0.657 f. 0.016
Indomethacin (1 x lo-5M)
0.419 + 0.014
0.561 + 0.016
Aspirin (1 x 10m4M)
Imidazole (1 x lo-*M)
0.630 2 0.020
AA (160 nmoles/ml) + Fe++ (120 nmoles/ml)
OD at 532 nm*
Inhibitors Reacted with Arachidonic Acid
EFFECT OF INHIBITORS 0N THE FERROUS IRON (Fe++) MEDIATED OXIDATION 0F A~~ti100NIc
Table 1.
trations of imidazole. Results indicate (Figure 6) that the imidazole bound to ferrous iron, making it unavailable to the orthophenanthroline, and inhibited AA oxidation in a concentration dependent fashion. Characterization of the Oxidation Product In addition to the studies on UV absorption of the products formed, the reaction mixture was double extracted into diethyl ether under acidic conditions and run in a high pressure liquid chromatograph system. The results showed a major peak for the oxidized AA and some minor peaks for unknown components. The products of the ether extract of the reaction mixture were methylated using diazomethane. Methylated compounds were analyzed using mass spectrometry by the direct inlet probe method. Analysis of the methylated derivative of arachidonic acid used showed the typical mass spectrum of methyl-AA with M = 318. There were no high mass peaks above 320 indicating an absence of autoxidized products. On the other hand, the ferrous iron oxidized samples showed a peak at 348 (M + 30), indicating an addition of 02 associated with a loss of two H, and another peak at 378 (M + 60) indicating addition of 02 molecules with loss of four H. Ozidized AA was also extracted into deuterated chloroform and subjected to 13C-NMR spectroscopy. This revealed losses of one or more double bonds from the oxidized sample, whereas the spectrum of the control sample run under similar conditions exactly matched the spectrum of authentic AA. Oxygenation of 1,3-Diphenylisobenzofuran(DPBF) As a further proof for the generation of AA peroxides, co-oxygenation of DPBF was studied. Since the substrate for this reaction was made up in acetone, it was necessary to eliminate the effect of acetone on the test system. Stock solutions of 100 mM and 10 mM DPBF were prepared and the amount of acetone used in any reaction never exceeded 10 ul/ml. Results shown in figure 7 demonstrate loss of absorbance indicating the oxidation of DPBF, starting from one minute after the introduction of AA and iron into the system. When Fe++ or AA were used alone no oxidation of DPBF was observed. When the reactions were run with the antioxidant, BHA, or the iron chelator, DAB, there was no co-oxygenation, indicating that AA oxidation promoted the co-oxygenation of DPBF in this reaction. The delay of 60 seconds in the reaction could be explained by our earlier findings that the peak of AA oxidation is at about 60 seconds. Co-oxygenation of DPBF was not inhibited by aspirin and, as shown previously, aspirin does not block the oxidation of AA mediated by ferrous iron. The product of DPBF oxygenation was isolated as described by Marnett et al and identified as 0-dibenzoylbenzene by mass spectroscopy [prominent ions at me/e 386 (M) 209 (M-77) and 105](36). DISCUSSION Based on the concept that iron is intimately involved in the synthesis of endo. peroxides and thromboxanes by blood platelets (11,29-32), the present investigation has explored the interaction of iron and arachidonic acid (AA), the precursor of prostaglandin products. Results of the study indicate that iron is a potent catalyst for the oxidation of AA. The reaction, followed by the increase in UV absorbtion at 232 nm, occurred rapidly, reaching a peak in less than 90 seconds. An alternate measure of lipid peroxidation, the TBA reaction 62
ET*120:1/ 00
f/8:3/
0.6 z 3 6
0.6
Et=
tm:3/
Et,,
120:3J
t ii ‘i
0.4
Ho 117../J I/ 04 f/8:11 ED 120:2/
8
EA (22:
0.2
I
I
I
60
120
180
Time
fseccanfs)
Fatty acid specificity of ferrous iron mediated oxidation. To stu y t e effect of ferrous iron on the oxidation of various unsaturated fatty acids, individual substrates (160 nmoles/ml) were exposed to ferrous iron (120 nmoles/ml) for three minutes and the UV absorbing products generated were measured at 232 nm. Degree of unsaturation and the proximity of the double bond to the carboxyterminal favored the oxidation of substrates by ferrous iron. ET4 = 5,8,11,14 Eicosatetraenoic acid (arachidonic acid); DH = 4,7,10,13,16,19 Docosahexaenoic acid; OD = 6,9,12 Qctadecatrienoic acid; ET3A = 8,11,14 Eicosatrienoic acid; ET3B = 11,14,17 Eicosatrienoic acid; HD = 10,Heptadecaenoic acid; EA = Erucic acid, OA = Oleic acid; ED = 11,14 Eicosadienoic acid.
-F--i
lmidozole
/mMI
Fjgure 6. Concentration dependent inhibition of arachidonic acid (AA) oxidation by imidazole. Using the TBA reaction to detect the formation of lipid peroxides, the effect of various concentrations of imidazole on the ferrous iron mediated AA oxidation was studied. Imidatole showed concentration dependent inhibition of AA oxidation. With the same concentrations of imidazole reactions were analyzed for the amount of unreacted ferrous iron with the orthophenanthrolinemethod. Oxidation of AA followed closely the oxidation of ferrous iron. O-O oxidized AA-TBA complex; Ferrous iron-Orthophenanthroline complex. 63
/with c @ z c g * s
DA8 or WA
I
0.8
0.6
0.4
8 0.2
00-
60
120 Time
lea
/saw&/
Figure 7. Co-oxygenation of diphenylisobenzofuran (DPBF) mediated by the oxidation of arachidonic acid (AA). The effect of transformation of AA on the oxidation of DPBF was followed by the decrease in absorbance at 420 nm. DPBF in acetone was added (0.05 mM) to a cuvette containing AA (160 nmoles/ml) and the reaction was initiated by the addition of ferrous iron (120 nmoles/ml). Addition of AA or iron alone to the substrate had no effect (O-O). When AA and iron were added together loss of absorbance due to the oxidation of DPBF could be followed (X-X). Aspirin had no inhibitory effect (h -----II), whereas DAB and BHA blocked the reaction (-----).
64
gave a similar result. The degree of AA peroxidation varied with the concentration of both reactants. When Fe+ concentration was maintained at 120 nanomoles/ml, the extent of oxidation depended on the amount of AA added until maximum conversion was reached at 160 nanomoles AA. Addition of more AA or prolonging the reaction time had no effect on the amount of AA oxidized. The reaction was dependent of the valence of iron. Ferrous iron was required and the ferric form was inert in catalyzing the oxidation of AA. Studies employing orthophenanthroline (OP) demonstrated that ferrous iron was oxidized to the ferric state during the reaction. The reaction between iron and AA was dependent on the presence of oxygen. Experiments carried out in a tonometer replacing air with a nitrogen atmosphere, revealed no conversion of AA due to interaction with iron under anaerobic conditions. The nature of the products formed during oxidation of AA by ferrous iron was examined in several ways. Maximum UV absorbtion of the reaction products occurred at 232 nm, and Tappel has demonstrated that lipid peroxides have a higher molar extinction coefficient at this wave length than other products of autoxidation formed during the process (37). The identification of the product by GC-MS was difficult as the substance was unstable and did not survive the gas chromatograph column. However, using the direct inlet probe method, evidence for addition of one and two oxygen molecules to the arachidonic acid was obtained together with loss of either 2 or 4 hydrogens. In AA, with four double bonds, the carbon atoms at 4,7,10 and 13 are "omethylene" carbons with weak bonds to their hydrogen atoms (38). The carbons in these positions are partially activated and the molecular oxygen in association with Fe++ probably abstracts the hydrogens from these positions, with addition of an oxygen molecule or molecules. The resulting product could be a single component, but is more likely a mixture of several peroxides, hydroperoxides and perhaps cyclic peroxides. Peroxides and cyclic peroxides would be expected to break down to malondialdehyde, or a malondialdehyde-like compound during processing for the TBA reaction (39), accounting for the positive TBA reaction seen. Although our interest was primarily on the interactions between iron and AA, we also examined the oxidation of other fatty acids by ferrous iron. Monounsaturated fatty acids yielded little oxidation product when exposed to ferrous iron. Several unsaturated fatty acids other than AA were good substrates and the amount of oxidation in the presence of iron depended on the degree of unsaturation. Thus, 4,7,10,13,16,19 docosahexaenoic acid, for example, formed more oxidation product when exposed to ferrous iron than AA under similir conditions. The proximity of the double bond of the carboxy terminal was also important. Oxidation of fatty acids with double bonds from 4-11 yielded more product than those having double bonds beyond carbon 10. Investigations into the influences of inhibitors and possible promoters on the ferrous iron induced oxidation of AA revealed additional features of the reaction. Although superoxide anion supposedly initiates the peroxidation of polyunsaturated fatty acids, it had no effect on AA alone. Also, superoxide dismutase failed to inhibit the ferrous iron-AA interaction. Harbor and Weiss have proposed a mechanism in which hydrogen peroxide mediates the change in the redox state of iron (40). Catalase, however, had no inhibitory effect on the ferrous iron induced oxidation of AA, suggesting that H202 65
probably has no significant role in this reaction. Tappel (41) found evidence to show that bonding of hydroxyl ion to the sixth bond position of iron in hematin compounds at pH 7.0 influences their peroxidative activity. His work on the effect of increased hydroxyl ion on hemoglobin catalyzed linoleate oxidation showed that the rate of peroxidation increased 2 fold. He proposed a scheme in which fatty acid and hematin form a complex mediated by hydroxyl ion which is reversible. Such a complex may very well form with ferrous iron and AA in our test system, or iron may directly interact with AA in the presence of oxygen to form a complex. The effects of other inhibitory agents on the ferrous iron mediated conversion of AA confirmed the importance of the valence state of iron and revealed insights into the mechanism through which antioxidants may block the process of oxidation. EDTA effectively removed ferrous iron from the reaction through chelation, and DAB, a strong reducing agent, blocked conversion of ferrous iron to the ferric state, an essential step in the reaction. The inhibitory effects of imidiazole and DHB were more complex. When reacted with AA first, they were far less inhibitory than when incubated with iron prior to initiation of the reaction. Thus, they inhibit the reaction by binding ferrous iron, but the complex takes more time to form than chelation of ferrous iron by EDTA. BHA, on the other hand, was a potent inhibitor, but did not influence the state of iron. Rather BHA appeared to act as a competitive substrate for the oxidation process, in essence replacing AA. The ability of the peroxidase inhibitor, aminotriazole, to block platelet prostaglandin synthesis, secretion, and aggregation first stimulated our interest in the role of iron in the peroxidation of AA to endoperoxides and thromboxanes (28). However, AMT had only a slight inhibitory effect on the ferrous iron induced oxidation of AA. We further explored the effect of AMT using diaminobenzidine (DAB) which reduced ferric iron to ferrous iron. In the presence of DAB, ferric iron was found to give a positive reaction with orthophenanthroline, indicating ferrous iron was now present. However, if ferric iron was reacted first with AMT before exposing the ferric iron to DAB, the orthophenanthraline reaction was markedly inhibited. Thus, the failure of AMT to inhibit ferrous iron induced oxidation of AA is understandable, because it binds ferric but not ferrous iron. In biological systems, however, the redox state of iron is closely regulated to serve basic cellular functions (41). Oscillations in valence are essential for the activity of iron in a variety of reactions. As a result, agents such as AMT, which trap iron in the ferric state effectively block oxidation and peroxidation reactions requiring the ferrous form, though the inhibition may require prolonged incubation. During the process of prostaglandin synthesis in platelets, hydroperoxides are generated (9). Marnett et al (36) on the basis of their studies suggested that co-oxygenation of substrates including DPBF, was mediated by hydroperoxide intermediates formed during prostaglandin synthesis because PGQ, which has a 15 hydroperoxy group, was effective but P6H2 was not. Our finding in the present investigation that ferrous iron mediated oxidation of AA can induce co-oxygenation of DPBF offers further evidence for 66
the generation of hydroperoxides in the system and supports the findings of Marnett et al (36) relating hydroperoxide intermediates to co-oxygenatton reactions. The results of the present investigation have a number of important implications. Several workers have pointed out previously the importance of heme groups to the activity of cycle-oxygenase and other enzymes involved in synthesis of endoperoxides, thromboxanes and additional products of prostaglandin synthesis (11,29-32,42). Examination of the interaction between AA and iron in the present study suggests that iron is not only a critical factor in the oxidation, but that it must be in the ferrous form. This finding is consistent with the electron paramagnetic resonance (EPR). Studies on lipoxygenase show that in the native enzyme iron is present in EPR silent ferrous state (43). In addition,it has been shown that the cycle-oxygenase activity of sheep vesicular gland microsomes is inhibited by 0-phenthroline, which is capable of reversibly inhibiting enzymes by either preventing pyridine nucleotide binding or by directly binding ferrous iron (44). Evaluation of the effects of inhibitors in our studies has demonstrated that the mechanism by which they block AA oxidation relates to their ability to bind ferrous iron or prevent its conversion to the ferric state. Lipid oxidation or peroxidation is a major problem in food preservation and aging. A search for effective antioxidants based on their interaction with iron may result in the design of more effective agents. Thus, single agents or a combination which binds ferrous or ferric iron or keeps the iron reduced and competes with AA and other fatty acids as a substrate for oxygenation could be valuable. The biological implications of the present investigations are also important. Transformation of AA to a more polar product, such as a peroxide or hydroperoxide, may facilitate its movement in cells to sites of specific enzymes. Preliminary studies suggest that the fatty acid peroxide generated during the oxidation of AA by ferrous iron can block the formation of prostaglandin (PG12) in vessel walls (45). Thus, the interaction of AA with ferrous iron may be one of the critical steps involved in the prostaglandin synthetic pathway and products formed as a direct result of ferrous iron induced oxidation of AA may serve to regulate the activity of specific enzymes such as prostacyclin synthetase (46,47). Further investigations currently in progress are attempting to determine if the interaction between ferrous iron and AA is a primary step of prostaglandin synthesis in cells. ACKNOWLEDGEMENTS We thank Dr. Thomas P. Krick, Department of Biochemistry, University of Minnesota for his technical assistance in mass spectroscopy. Supported by USPHS grants HL-11880, AM-06317, HL-06314, CA-12607, CA-08832, CA-11996, GM-AM-22167, HL-20695, HL-16833 and a grant from the Leukemia Task Force. REFERENCES 1.
Bergstrom S, Danielsson H, Samuelsson B. The enzymatic formation of prostaglandin Ep from arachidonic acid. Biochim Biophys 90:207, 1964.
2.
Bills TK, Smith JB, Silver MJ. Metabolism of (14C) arachidonic acid by human platelets. Biochim Biophys Acta 424:303, 1976. 67
3.
Nugteren DH. Arachidonate lipoxygenase in blood platelets. Biochim Biophys Acta 380:299, 1975.
4.
Smith WL, Lands WEM. Oxygenation of polyunsaturated fatty acids during prostaglandin synthesis by sheep vesicular gland. Biochemistry 11: 3276, 1972.
5.
Van Dorp DA, Beerthuis RK, Nugteren DM and Vonkeman H. The biosynthesis of prostaglandins. Biochim Biophys Acta 90:204, 1964.
6.
Hamberg M, Smauelsson B. Detection and isolation of an endoperoxide intermediate in prostaglandin biosynthesis. Proc Nat1 Acad Sci USA 70:899, 1973.
7.
Hamberg M, Samuelsson B. Prostaglandin endoperoxides. Novel transformations of arachidonic acid in human platelets. Proc Nat1 Acad Sci USA 71:3400, 1974.
8.
Hamberg M, Svensson J, Wakabayashi T, Samuelsson B. Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Nat1 Acad Sci USA 71:345, 1974a.
9.
Hamberg M, Svensson J, Samuelsson B. Prostaglandin endoperoxides. A new concept concerning the mode of action and release of prostaglandins. Proc Nat1 Acad Sci USA 71:3824, 1974b.
10.
Needleman P,Minkes M, Raz A. Thromboxanes: Selective biosynthesis and distinct biological properties. Science 193:163, 1976.
11.
Ho PPK, Walters P, Sullivan HL. Biosynthesis of thromboxane B2: assay, isolation, and properties of the enzyme system in human platelets. Prostaglandins 12:951, 1976.
12.
Moncada S, Needleman P, Bunting S, Vane JR. Prostaglandin endoperoxide and thromboxane generating systems and their selective inhibition. Prostaglandins 12:323, 1976.
13.
Willis AL. Isolation of a chemical trigger for thrombosis. Prostaglandins 5:1, 1974.
14.
Willis AL, Vane FM, Kuhn DC, Scott CG, Petrin M. An endoperoxide aggregator (LASS), formed in platelets in response to thrombotic stimuli: Purification, identification and unique biological significance. Prosta, glandins 8:453, 1974.
15.
Hamberg M, Svensson J, Wakabayashi T, Samuelsson B. Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc Nat1 Acad Sci USA 71:345, 1974.
16.
Labile agGerrard JM, White JG, Rao GHR, Krivit W, Witkop CJ Jr. gregation stimulating substance (LASS). The factor from storage pool deficient platelets correcting defective aggregation and release of aspirin treated normal platelets. Br J Haematol 29:657, 1975.
17.
Gerrard JM, White JG. The influence of prostaglandin endoperoxides on platelet ultrastructure. Am J Path01 80:189, 1975.
18.
Hamberg M, Svensson J, Samuelsson B. Thromboxanes: A new group of biologically active compounds derived from prostaglandin endoperoxides. Proc Nat1 Acad Sci USA 72:2994, 1975.
68
19.
Malmsten C, Hamberg M, Svensson J, Samuelsson B. Physiological role of an endoperoxide in human platelets: Hemostatic defect due to platelet cycle-oxygenase deficiency. Proc Nat1 Acad Sci USA 72:1446, 1975.
20.
Panganamala RV, Brownlee NR, Sprecher H, Cornwell DG. Evaluation of superoxide anion and singlet oxygen in the biosynthesis of prostaglandins from eicosa-8,11,14-Trienoicacid. Prostaglandins 7:21, 1974.
21.
Panganamala RV, Sharma HH, Sprecher H, Geer JC, Cornwell DG. A suggested role for hydrogen peroxide in the biosynthesis of prostaglandins. Prosta. glandins 8:3, 1974.
22.
Panganamala RV, Sharma HH, Heikkila RE, Geer JC, Cornwell DG. Role of hydroxyl radical scavengers dimethyl sulfoxide, alcohols and methional in the inhibition of prostaglandin biosyntheses. Prostaglandins 11:599, 1976.
23.
O'Brien PJ, Rahimtula A. The possible involvement of a peroxidase in prostaglandin biosyntheses. Biochem Biophys Res Conm 70:832, 1976.
24.
O'Brien PJ, Hawco FJ, Piatt JF, Rahimtula AD. Enzymic singlet oxygen formation from peroxides in biological systems and its effects. Comm Int Conf Singlet oxygen related spec in Chem Biol August 21-26, 1977 Pinawa, Manitoba, Canada.
25.
Rao GHR, Gerrard JM, Eaton J, White JG: Detection of arachidonic acid (AA) free radical formation, an essential step in platelet prostaglandin (PG) synthesis. Blood 48:977, 1976.
26.
Rao GHR, White JG. Effects of vitamin E (VE) and nitroblue tetrazolium (NBT) on platelet ultrastructure and aggregation. Fed Proc 36:350, 1977.
27.
Rao GHR, Gerrard JM, Eaton JW, White JG. Arachidonic acid peroxidation. prostaglandin synthesis and platelet function. (In press) Comn Int Conf Singlet oxygen and related species in chemistry and biology. August 21-26, 1977 Pinawa, Manitoba, Canada.
28.
Gerrard JM, White JG, Rao GHR, Townsend D. Localization of platelet prostaglandin production in the platelet dense tubular system. Am J Path01 83:283, 1976.
29.
Ho PPK, Walters P, Sullivan HR. A particulate arachidonate lipoxygenase in human blood platelets. Biochem Biophys Res Comm 76:398, 1977.
30.
Bild GS, Bhat SG, Ramadoss CS, Axelrod B. Biosynthesis of a prostaglandin by a plant enzyme. J Biol Chem 253:21, 1978.
31.
Yoshimoto A, Ito H, Tomita K. Cofactor requirements of the enzyme synthesizing prostaglandin in bovine seminal vesicles. J Biol Chem 68~487, 1970.
32.
Miyamoto T, Ogino N, Yamamoto S, Hayaishi 0. Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J Biol Chem 251:2629, 1976.
33.
Tappel AL, Boxer PD, Lundberg WO. The reaction mechanism of soybean lipoxidase. J Biol Chem 199:267, 1952.
34.
Smith JB, Ingerman CM, Silver MJ. Malondialdehyde formation as an indicator of prostaglandin production by human platelets. J Lab Clin Med 88:167, 1976. 69
35.
Wagner CO, Clever HL, Peters ED. Evaluation of the ferrous thiocyanate calorimetric method. Ind Eng Chem 19:980, 1947.
36.
Marnett LJ, Wlodawer P, Samuelsson B. Co-oxygenation of organic substrates by the prostaglandin synthetase of sheep vesicular gland. J Biol Chem 250:8510, 1975.
37.
Tappel AL. The mechanism of the oxidation of unsaturated fatty acids catalyzed by hematin compounds. Arch Biochem and Biophys 44:378, 1953.
38.
Demopoulos HB. 1859, 1973.
39.
Pryor WA, Stanley JP, Blair E. Autoxidation of fatty acids. II. A suggested mechanism for the formation of TBA-reactive materials from prostaglandin like endoperoxides. Lipids 11:370, 1976.
40.
Harber F, Weiss J. The catalytic decomposition of hydrogen peroxide by iron salts. Proc Roy Sot Lond Ser A 147:332, 1934.
41.
Tappel AL. Unsaturated lipid oxidation catalyzed by hematin compounds. J Biol Chem 217:721, 1955.
42.
Hemler M, Lands WEM. Purification of the cycle-oxygenase that forms prostaglandins. Demonstration of two forms of iron in the holoenzyme. J Biol Chem 251:5575, 1976.
43.
Egmond MR, Finazzi-Agro A, Fasella PM, Veldink GA, Vliegenthart JFG. Changes in the fluorescence and absorbance of lipoxygenase-1 induced by 13-Ls-hydroperoxylinoleicacid and linoleic acid. Biochim et Biophys Acta 397:43, 1975.
44.
Smith WL, Lands WEM. Stimulation and blockade of prostaglandin biosynthesis. J Biol Chem 246:6700, 1971.
45.
Rao GHR, Gerrard JM, Eaton JW, Clawson CC, White JG. The role of iron in prostaglandin biosynthesis: Ferrous iron mediated oxidation of arachidonic acid. Fed Proc 37:249A, 1978.
46.
Moncada S, Gryglewski RJ, Bunting S, Vane JR. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature (London) 263:663, 1976.
47.
Moncada S. A lipid peroxideinhibitsthe enzyme in blood vessel microsomes that generates from prostaglandin endoperoxides, the substance (Prostaglandin X) which prevents platelet aggregation. Prostaglandins 12:715, 1976.
The basis of free radical pathology. Fed Proc 32:
