'IIII I|"‘. II III“ IIIIIU ”IIi' I I II." 3:33 III'3“‘I.I,3. W I! III 'IIIIIIIL 23331 I IIIIII III 1II‘II " tfiuI I I ‘3I‘IqII gird“. 3"" ;,: . Léig: II , IIIII’ I {HIP I I ‘ " IIIII'II“ 1?.7':"'"":3::-=' a" t A: fim‘fx Q '0 ”‘ J -. n. .25. . .. . .. ‘z—v— v3 {.— M m Q. a . M ”H .. .1,“ $.35 be "(I \ti 7% HI}; ‘I II III . I IIIWI I If" INIIII I3'..- I III. .. I 3 III I III “I I III fl ' I’ fill?! IIIIIIIII . I —~.“:‘___ fl- THESIS 1.13m 42 r ”l 2 .99???” t" v—vv fivW This is to certify that the dissertation entitled The Role of Chelated Iron in the Mechanism of Enzymatic Promotion of Lipid Peroxidation presented by Bruce A. Svingen has been accepted towards fulfillment of the requirements for Ph.D. , Biochemistry degree 1n Major professor Date August 18, 1981 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES .4“. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. The Role of Chelated Iron in the Mechanism of Enzymatic Promotion of Lipid Peroxidation by Bruce A. Svingen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY 1981 ABSTRACT The Role of Chelated Iron in the Mechanism of Enzymatic Promotion of Lipid Peroxidation by Bruce A. Svingen Ironqdependent peroxidation of microsomal phospholipids promoted by either NADPH-cytochrome P450 reductase or xanthine oxidase occurs by similar, if not identical, mechanisms. Lipid peroxidation occurs by two distinct sequential radical chain reaction initiating reactions, lipid hydroperoxide-independent and -dependent initiation. Lipid hydro- peroxide-independent initiation is characterized by the formation of low levels of lipid hydroperoxides in previously peroxide free lipids. Lipid hydroperoxide-independent initiation of lipid peroxidation may be promoted by the ADP-perferryl ion. In NADPH-dependent lipid peroxida- tion the ADP-perferryl ion is formed by the direct reduction of the ADP-ferric ion complex catalyzed by NADPH-cytochrome P450 reductase and the subsequent reaction of the ADP-ferrous ion complex with 02. In superoxide-dependent lipid peroxidation the ADP-perferryl ion is generated by the direct reaction of superoxide with the ADP-ferric ion complex. Lipid hydroperoxide-independent initiation of lipid peroxide is superoxide dismutase sensitive. Lipid hydroperoxide-independent ini- tiation apparently does not involve either singlet oxygen or the free hydroxyl radical. Lipid hydroperoxide-independent initiation accounts for approximately 10-15% of total peroxidative products formed during enzymatically promoted iron-dependent lipid peroxidation. Lipid hydroperoxide-dependent initiation of lipid peroxidation is dependent upon the presence of previously formed lipid hydroperoxides. Lipid hydroperoxide-dependent initiation is the metal catalyzed heterolysis of lipid hydroperoxides to form radical products. Lipid hydroperoxide-dependent Bruce A. Svingen. initiation can be efficiently promoted by either EDTArferrous ion or DTPArferrous ion complexes. In NADPH-dependent lipid peroxidation the reduced chelates are formed by direct enzymatic reduction of the ferric chelate. The NADPHrdependent reaction is thus not superoxide dismutase sensitive. In superoxide-dependent lipid peroxidation the formation of the ferrous ion- chelate complex is dependent upon the reaction of superoxide with the ferric ion-chelate complex and the reaction is, therefore, superoxide dismutase sen- sitive. The hydroxyl radical is not formed during the iron-promoted breakdown of lipid hydroperoxides. However small amounts of singlet oxygen are formed secondarily to lipid peroxy radical formation. Lipid hydroperoxide-dependent initiation of lipid peroxidation accounts for 80-90% of the total peroxidative products formed during enzymatically promoted iron-dependent lipid peroxida- tion. Lipid hydroperoxide-dependent initiation can also be promoted by ferric cytochrome P450. ii TABLE OF CONTENTS Page LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 1 Initiation of Lipid Peroxidation Summary . . . . . . . . . . . . . . . . . . . . . . . . 32 Introduction . . . . . . . . . . . . . . . . . . . . . . 32 Materials and Methods . . . . . . . . . . . . . . . . . 34 Results. . . . . . . . . . . . . . . . . . . . . . . . . 36 Discussion . . . . . . . . . . . . . . . . . . . . . . . 50 Chapter 2 Hydroxyl Radical-Dependent Initiation of Lipid Peroxidation Summary . . . . . . . . . . . . . . . . . . . . . . . . 58 Introduction . . . . . . . . . . . . . . . . . . . . . . 58 Materials and Methods . . . . . . . . . . . . . . . . . 61 Results . . . . . . . . . . . . . . . . . . . . . . . . 62 Discussion . . . . . . . . . . . . . . . . . . . . . . . 76 Chapter 3 Lipid Hydroperoxide-Dependent Initiation of NADPH— Dependent Peroxidation of Liver Microsomal Phospholipids Summary . . . . . . . . . . . . . . . . . . . . . . . . 79 Introduction . . . . . . . . . . . . . . . . . . . . . . 79 Materials and Methods . . . . . . . . . . . . . . . . . 82 Results . . . . . . . . . . . . . . . . . . . . . . . . 85 Discussion . . . . . . . . . . . . . . . . . . . . . . . 103 Chapter 4 Lipid Hydroperoxide-Dependent Initiation of 027-Promoted Peroxidation of Microsomal Phospholipids summary 0 o o o o o o o o o o o o o o o o o o o o o o o 115 IntrOdUCtion o o o o o o o o o o o o o o o o o o o o o o 115 Materials and Methods . . . . . . . . . . . . . . . . . 118 ReSUItS o o o o o o o o o o o o o o o o o o o o o o o o 121 DisCUSSion o o o o o o o o o o o o o o o o o o o o o o o 133 SUMMARY 0 o o o o o o o o o o o o o o o o o o o o o o o o o o o o o 145 REFERENCES 0 o o o o o o o o o o o o o o o o o o o o o o o o o o o 156 APPENDIX 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O 166 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 10 11 12 13 14 15 16 iii LIST OF TABLES Page Inhibition of Xanthine Oxidase Promoted Lipid Peroxidation by Superoxide Dismutase (SOD) and 2,5-Diphenylfuran (DPF). 37 Inhibition of Xanthine Oxidase Promoted Lipid Peroxidation by SOD O O O O O O O O O O O O O O O O O O O C C O O I 39 Inhibition of Xanthine Oxidase Promoted Lipid Peroxidation by DPF O O O O O O O O O O O O O O O O O O I O O I O 0 Inhibition of NADPH-Cytochrome P450 Reductase Promoted Lipid PerOXidation by SOD o o o o e o o o o o o o o 0 Inhibition of NADPH-Cytochrome P450 Reductase Promoted Lipid PerOXidation by DPF o o o o o o o o o o o o o o Involvement of Superoxide and Singlet Oxygen in NADPH- Cytochrome P450 Reductase-Dependent Lipid Peroxidation 40 41 43 44 Involvement of Superoxide and Singlet Oxygen in the Promo- tion of Lipid Peroxidation by Chelated Iron . . . . . 46 Lipid Hydroperoxide-Dependent Initiation of Lipid Peroxida- tion From LOOH Catalyzed by ADP and EDTA Chelated Iron . . 47 Involvement of Superoxide and Singlet Oxygen in the Promo- tion of Lipid Peroxidation in Partially Peroxidized Lipo- someS........................ The Effect of HO' Trapping Agents and Catalase on Fenton's Reagent-Promoted Lipid Peroxidation . . . . . . . . . Effect of Chelation on Fenton's Reagent Promoted Lipid PerOXidationoooooooooo00000000000 The Promotion of LOOHeDependent Initiation of Lipid Peroxidation by Iron Chelates . . . . . . . . . . . . The Promotion of LOOHrDegendent Initiation of Lipid Peroxidation by EDTA-Fe+ . . Promotion of LOOH—Dependent Initiation of NADPH-Dependent LipidPerOXidation.......o.......... NADPHrDependent Liposomal Peroxidation . . . . . . . . Promotion of LOOH—Dependent Initiation of NADPH-Dependent Lipid Peroxidation by EDTA-Fe+3 and DTPAr-Fe+3 . . . . 49 72 74 90 91 93 95 97 Table Table Table Table Table Table Table Table Table Table Table 17 18 19 20 21 22 23 24 25 26 27 iv LIST OF TABLES (Continued) Page The Effect of EDTA-Fe+3 on the SKF 5259A Inhibition of NADPH-Dependent Microsomal Lipid Peroxidation . . . . . . 99 The Effect of EDTA-Fe+3 on the Aminopyrine Inhibition of NADPH-Dependent Microsomal Lipid Peroxidation . . . . . . 100 Promotion of LOOHeDependent Initiation of Lipid Peroxida- tion by Ferric Cytochrome P450 . . . . . . . . . . . . . . 102 Promotion of LOOH—Dependent Initiation of NADPH-Dependent Liposomal Peroxidation by Ferric Cytochrome P450 . . . . . 104 The 027-Dependent Decomposition of Cumene Hydroperoxide in the Presence 0f ChelatEd Iron. a o o o o o o o o o o o o o 123 Iron Chelate Promotion of LOOH-Dependent Initiation of Lipidper0X1dationoooooo00000000000000125 Superoxide-Promoted LOOH—Dependent Initiation . . . . . . 127 Superoxide-Dependent Lipid Peroxidation . . . . . . . . . 130 Superoxide-Dependent LOOH—Dependent Initiation of Lipid Peroxidation Promoted by EDTAr-Fe+3 and DTPA-Fe+3 . . . . . 131 The Effect of SKF 5252A on Superoxide-Dependent Microsomal LipidPerOXidation.......o............132 Promotion of LOOHrDependent Initiation of OzT-Dependent Lipid Peroxidation by Ferric Cytochrome P450 . . . . . . . 134 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10 V LIST OF FIGURES Page Basic Scheme of Free Radical Lipid Peroxidation . . . . . 6 NADPH-Dependent and 027-Dependent Promotion of Peroxidation of Extracted Microsomal Lipid Vesicles . . . . . . . . . 57 Time Course of Fenton's Reagent-Promoted Lipid Peroxida- tion 0 O O O O O O I O O O O C O O O O O O O O O O O O O 64 The Dependence of Fenton's Reagent-Promoted Lipid Peroxidation on H202 Concentration . . . . . . . . . . . 67 The Effect of H202 to Ferrous Ion Concentration on Fenton's Reagent-Promoted Lipid Peroxidation . . . . . . 69 The pH Profile of Fenton's Reagent-Promoted Lipid Per0x1dation00000000000000.0000...71 The NADPH-Dependent Decomposition of Cumene Hydroperoxide in the Presence of Chelated Iron . . . . . . . . . . . . 88 Summary of the Proposed Scheme of NADPHeDependent Lipid Pe rOXidation O O O C I O O O O O O O O O O O O O O O O O 114 Schematic of the Reactions of OzT-Dependent Lipid Per0x1dationoooooooooooooooooo0.0.144 A Schematic of the Reactions of NADPH-Dependent and 02?- Dependent Lipid PerOXidation o o o o o o o o o o o o o o 153 ADP ADP-Fe+3 (Fe+2) BHT DMPO DPBF DPF DTPA DTPArFe+3 (Fe+2) EDTA EDTAeFe+3 (Fe+2) ESR Fe+20 1787-202, Fe+302' . HO' LH LOOH LOO' MDA 02: PUFA son TBA ABBREVIATIONS Adenosine 5'-diphosphate Ferric (ferrous) ion chelated by ADP Butylated hydroxytoluene 5,5-dimethyl-l-pyrroline-N-oxide Diphenylisobenzofuran 2,5-dipheny1furan Diethylenetriaminepentaacetic acid Ferric (ferrous) ion chelated by DTPA Ethylenediamintetraacetic acid Ferric (ferrous) ion chelated by EDTA Electron Spin Resonance Ferryl ion Perferryl ion Hydroxyl radical Lipid molecule Lipid alkyl radical Lipid alkoxy radical Lipid hydroperoxide Lipid peroxy radical Malondialdehyde Singlet oxygen Superoxide Polyunsaturated fatty acid Superoxide dismutase Thiobarbituric acid LITERATURE REVIEW Mostly through the excellent work of Gee, Bolland, Bateman, and co-workers, of the British Rubber Producers Research Association, lipid peroxidation has been shown to occur by a free radical mechanism [1-12]. Studies on both autoxidative and enzymatically-promoted lipid peroxida- tion by these investigators have established the basic reactions of lipid peroxidation. The reactions of free radical lipid peroxidation are outlined below (LH, polyunsaturated fatty acids (PUFA); LOOH, lipid hydroperoxide; L', lipid alkyl radical; LOO', lipid peroxy radical) [13,14]: Initiation: LH + 02 i_, _; free radicals (1) LOOH —_; free radicals (2) Propagation: L- + 02 ————-) L00- (3) L00- + LE ————9 LOOH + L° (4) Termination: L° + L° —} Stable (nonradical) products (5) L00' + LOO°-——————) Stable (nonradical) products (6) L° + L00' -—---9 Stable (nonradical) products (7) Reactions 1 and 2 represent two possible means for the initiation of lipid peroxidation. Reaction 1 is initiation independent of lipid hydroperoxides and reaction 2 is lipid hydroperoxide-dependent initi- ation. The relative importance of reactions 1 and 2 in the initiation of enzymatically promoted lipid peroxidation is matter of some contro- versy. _2_ Polyunsaturated fatty acids are essentially inert to reaction 1 as writ- ten. The kinetic inertness of PUFA to reaction 1 arises from the basic concepts of spin conservation during reaction. Spin conservation imposes restrictions on reaction mechanisms such that under normal cir- cumstances reactions occur only between reactants of the same spin multiplicity. In reaction 1, as written, the ground states of the reac- tants are of dissimilar multiplicity. The oxygen ground state is of triplet multiplicity and the PUFA ground state is of singlet multi- plicity. Thus the reaction is spin forbidden and unlikely to occur as written. On the basis of quantum theory it has been predicted that reaction 1 is very endothermic, has an extremely high activation energy and is unlikely to occur to any extent under normal biological condi- tions [15,16]. The kinetic inertness of lipid hydroperoxides to reac- tion 2, thermal homolysis, has been experimentally demonstrated [17,18]. Thermal homolysis of hydroperoxides is an extremely slow reaction under any but quite drastic conditions. Thus it is unlikely that, under bio- logical conditions, lipid peroxidation would be initiated by the unpro- moted form of either reaction 1 or 2. However, when either reaction 1 or 2 is promoted by a metal ion it becomes kinetically feasible as a mechanism for the initiation of lipid peroxidation. The participation of a metal in reaction I imparts radi- cal characteristics to the reaction and thus removes the spin restric- tions imposed on the unpromoted reaction [19]. The metal promotion of reaction 2, reductive activation, has been extensively studied and has been demonstrated to be quite a rapid reaction [17]. Oxidized heme pro- motion of reaction 2 has also been shown to be quite facile [17,20-24]. -3- The actual mechanism of initiation in enzyme-dependent lipid peroxida- tion will be dealt with in greater detail later. The relative contribu- tions of the two reactions to the overall rate of initiation can vary dramatically. In hydroperoxide-free lipids initiation must occur by some form of reaction 1. However, in lipids containing lipid hydro- peroxides the contributions of reaction 1 to initiation of lipid peroxi- dation are quickly overshadowed by those from reaction 2 [21,25]. In iron promoted peroxidation of hydroperoxide free lipid, kinetic analysis indicates that while the initial rate of peroxidation reflects the kine- tics of reaction 1, reaction 2 quickly supplants it [26,27]. After initiation has occurred the propagation cycle of the overall mechanism begins. By various methods it has been estimated that each free radical formed goes through 8 to 14 propagation cycles [28]. Because propagation can also produce a nonradical product necessary for initiation, lipid hydroperoxides, it is easy to envision how lipid per- oxidation can become a geometrically progressive reaction resulting in extensive membrane damage. As shown, termination of the propagatory chain can occur by several reactions. Although the products of the termination reactions are by definition not free radicals, this should not be construed to indicate that termination products are non-reactive. For example, reaction 6 has been shown to generate singlet oxygen [29]. If formed, singlet oxygen can react with PUFA to produce lipid dioxetenes and hydroperoxides [30]. The application of the basic tenets of free radical reaction mechanisms to the specific problem of PUFA peroxidation has resulted in the proposal of a general mechanism of lipid peroxidation as -4- schematically presented in Fig. 1 [31]. There are two possible initi- ation mechanisms. In peroxide free PUFA, initiation is proposed to occur by abstraction of a methylene hydrogen from a PUFA molecule generating a PUFA alkyl radical. Alternatively, in lipids which already contain peroxide material, such as PUFA hydroperoxides, initiation may predominantly occur by hydroperoxide breakdown to form either PUFA alkoxy or peroxy radicals. If initiation occurs by hydrogen abstraction to generate the PUFA alkyl radical the reaction then enters into the propagation stage of the mechanisms by the addition of dioxygen. The addition of oxygen is a diffusion limited process (k=109 - 1010 M"1 sec‘l) when the oxygen par- tial pressure of the reaction mixture is 100 mm of Hg or greater [18]. The partial pressure of oxygen has great significance for the termination reactions expressed. If the partial pressure is 100 mm or greater, only termination reactions involving the peroxy radicals need be considered. If the oxygen partial pressure is below 100 mm the termination reactions involving both peroxy and alkyl radicals must be considered. If ini- tiation occurs via generation of the PUFA peroxy radical, once again the classical reactions of propagation are directly entered into. If, however, initiation occurs by formation of the PUFA alkoxy radical, the next step is likely to be the abstraction of a methylene hydrogen from a neighboring PUFA generating a PUFA alcohol and a PUFA alkyl radical. The PUFA alkyl radical then enters the propagation stage of the mecha- nism as discussed above. The propagation of lipid peroxidation involves the abstraction of a PUFA methylene hydrogen by a PUFA peroxy radical. This reaction forms a PUFA hydroperoxide and a PUFA alkyl radical. The .COHumvaxoumm vHQHA Hmoawmm mmum mo mamnum oammm .H muswam \NHwNHON A./\.//\JW\/\ : 3” (LOOH) -0H PROPAGATWON AfoLflAAA (LO) +LH r” L. /\..fv:./\,_‘f‘-ff'v'\;.° TEqunATxn: MUM/WA : 3” " mobH) -H (LOO) F in.) \ (LH) -‘H AW 2" M/LH/ (L ) L O OH ' MW“ AOADADNNNA :. 3” +LH I /\_//\./\./V\/\/\ + O $g° L o 8 ALDEHYDES KETONES ALCOHOLS ETHERS Atms ETHANE PENTANE -7- PUFA alkyl radical then adds oxygen to reform a PUFA peroxy radical. A significant modification of the propagation reaction occurs when a PUFA peroxy radical internally cyclizes to form a PUFA endoperoxide radical. The PUFA endoperoxide radical can then subsequently add dioxygen to generate a PUFA endoperoxide peroxy radical. The PUFA endoperoxide peroxy radical can participate in the propagation scheme the same as any other PUFA peroxy radical. However, the significance of this internal endoperoxide formation with the addition of a B-oxygen function (the peroxy function) is that this endoperoxide appears to be the material detected in the thiobarbituric acid assay for the rate and extent of lipid peroxidation [32-36]. As previously mentioned the termination reactions are greatly altered by the oxygen partial pressure. Addition- ally, as the reaction mixture becomes more complex the termination reac- tion becomes more complex and can involve constitutents other than PUFA radicals. In biological membranes termination reactions will include reaction between PUFA radicals and membrane constituents such as a- tocopherol, cholesterol and sulfhydryl groups of glutathione and proteins. THE REQUIREMENT FOR IRON IN ENZYMATICALLY—PROMOTED LIPID PEROXIDATION Several metals which undergo univalent redox reactions can partici- pate in the promotion of autoxidative and enzymatic peroxidation of PUFA. Of these metals, the subgroup of cobalt, copper, iron and man- ganese are of biological significance. Of this group, iron has been found to be the most active promoter of lipid peroxidation both in_gi££9 and in gigg. The unique role of iron stems from two basic points. First, the iron concentration in animal tissues is higher than any other -8— member of this group. Second, certain biological chelators of iron enhance the ability of iron to function as a promoter of lipid peroxida- tion. This second point is supported by several lines of evidence. Tappel et a1. [20-23] showed that heme compounds, many of which are intimately associated with the phospholipids of biological membranes, are the best promoters of lipid peroxidation both in 31552.and in 2132, Also, iron is required for oxygen activation to initiate lipid hydro- peroxide formation upon which heme promotion of lipid peroxidation is dependent [37-40]. The role of metals in the promotion of autoxidation has been the subject of previous reviews [25,26,4l,42] and will not be dealt with here. This review will deal with the role of iron in the initiation of enzymatic lipid peroxidation. Many researchers investigating numerous and varied systems of enzy- matically promoted in litre lipid peroxidation have found that addition of iron or an iron-Chelate complex is required for peroxidation. This requirement was first demonstrated in NADPH-dependent microsomal lipid peroxidation. Ernster and co-workers [37,43-45] found that the addition of iron chelated by ADP was required for NADPH-dependent microsomal lipid peroxidation. Subsequently, several other investigators confirmed that the addition of complexed iron was required for promotion of NADPH- dependent microsomal lipid peroxidation [38,40,46-54]. It was found that ADP could be replaced by other nucleotides and pyrrophosphate [38,40]. Extensive experimentation showed that although other chelates could be used, activity was maximal with ADP. Pederson et a1. [55] found that the addition of EDTArchelated iron and ADP-chelated iron to microsomes promoting NADPH-dependent lipid peroxidation resulted in an -9- increase in the rate of peroxidation over that observed in the presence of only ADP-chelated iron. Pederson et al. also observed that EDTA? chelated iron alone could not function in the promotion of NADPH- dependent microsomal lipid peroxidation. However, Lai and Piette [56,57] found that it could. Further endeavors to elucidate the mecha- nism of microsomal lipid peroxidation led to the identification of NADPH-cytochrome P450 reductase as the enzyme responsible for the promo- tion of NADPH—dependent microsomal lipid peroxidation [55,58]. Pederson et al. [55,58] demonstrated that protease solubilized NADPH-cytochrome P450 reductase in the presence of NADPH and ADP-chelated iron could pro- mote the peroxidation of liposomes prepared from extracted microsomal lipid. These investigators also found that the promotion of lipid peroxidation in the above reaction mixture could be dramatically enhanced by the addition of EDTA-chelated iron. The enhancement observed in liposomes was much greater than that previously observed in microsomes. The findings of Pederson et a1. were confirmed by Sugioko and Nakano [59] and Noguchi and Nakano [60] utilizing a similar liposo- mal reaction mixture to study the reconstitution of NADPH-dependent lipid peroxidation. In contrast, Pospelova et al. [61] found that NADPH-dependent peroxidation of liposomes could be promoted by either ADP-chelated iron or EDTA-chelated iron. The choice of iron chelate apparently made little difference since the rate and extent of peroxida- tion were similar in both cases. Superoxide-dependent lipid peroxidation has also been shown to be promoted by iron complexes. Pederson and Aust [62] demonstrated that superoxide-dependent peroxidation of liposomes prepared from extracted -10- microsomal lipid was promoted by ADP-chelated iron. Again, as in NADPHrdependent liposomal peroxidation, the addition of EDTA-chelated iron together with ADP-chelated iron greatly enhanced the rates of peroxidation over those observed in the presence of ADP-chelated iron alone. These investigators found that EDTArchelated iron alone could not promote superoxide-dependent lipid peroxidation. The findings of Pederson and Aust [62] were confirmed by Svingen et al [63]. Superoxide-dependent peroxidation of mitochondrial membranes is also promoted by ADP-chelated iron [64]. THE ROLE OF IRON IN HYDROPEROXIDE-INDEPENDENT INITIATION OF LIPID PEROXIDATION Initiation of lipid peroxidation in hydroperoxide free PUFA is thought to occur by the abstraction of a methylene hydrogen from a PUFA molecule giving rise to a PUFA alkyl radical (see Fig. 1). Hydrogen abstraction occurs at the methylene carbon because the allylic position reduced the bond dissociation energy. The bond dissociation energy for methylene hydrogen is 41.4 Kcal per mole compared to 68.2 Kcal per mole for vinyl hydrogen and 1130.0 Kcal per mole for secondary hydrogen. Even though the methylene hydrogen can be considered partially activated it is obvious that a strong oxidizing agent must be involved in the abstraction. Of the oxidative reagents that have been proposed to be formed in biological systems, two are theoretically of the proper oxida- tive power to be involved in initiation of lipid peroxidation. The two oxidants most often proposed to be involved in initiation are the hydroxyl radical (H0°) and the perferryl ion (Fe+202). In this section of the review the experimental and theoretical support for the involve- ment of either H0: or Fe+202 in the initiation of lipid peroxidation -l 1- from the standpoint of iron-promotion of initiation will be examined. This section will be concluded with a theoretical examination of the identity of HO' and Fe+202. Are these two reactants perhaps members of the same group of iron activated oxygen intermediates and therefore perhaps indistinguishable? Initiation of lipid peroxidation by HO° abstraction of PUFA methy- lene hydrogen has been proposed by several research groups. In all instances HO' participation in initiation of lipid peroxidation has been proposed for superoxide-dependent lipid peroxidation. The basic mecha- nism proposed for formation of H0: in superoxide-dependent lipid peroxi- dation is based on a combination of the reactions that constitute Fenton's Reagent and the Habeereiss reaction [65]. The mechanism pro- posed is perhaps most adequately described as a superoxide driven, iron promoted Haber-Weiss reaction. The general scheme for HO° production is schematically shown below. 027 + Fe+3 —-——-) Fe” + 02 (8) 2027 + 2H+ -—--——-) H202 + 02 (9) Fe+2 + H202 -__.__., Fe+3 + OH“ + no- (10) Experimental evidence supporting the above reaction scheme is most often derived from the ability of HO° traps, catalase and superoxide dismutase to inhibit not only lipid peroxidation but also other reactions that are characteristic of HO°, H202 and superoxide. The three reactions were sequentially ordered on the basis of inhibition of lipid peroxidation and secondary reactions characteristic of superoxide, H202 and H0“. The participation of iron in the reaction sequence is based on its required presence for the occurrence of reactions characteristic of HO°. This -12- indicates that the unpromoted Haber-Weiss reaction does not occur to any significant extent. The requirement for iron reduction is shown by the inability of H202 and ferric ion to catalyze lipid peroxidation at neutral pH. Fong et a1. [39] proposed that NADPH-dependent lipid peroxidation promoted by NADPH-cytochrome P450 reductase in either microsomes or liposomes occurred via an iron promoted Haber-Weiss reaction. These authors found that lipid peroxidation, as measured by the release of acid phosphatase from lysosomes, required ADP-Fe+3 and was inhibited by superoxide dismutase, catalase and hydroxyl radical traps. From their data these authors proposed the following mechanism of HO- generation: 202 + NADPH -————) NADP+ + 11”" + 2 02* (11) 202* + 222* —-—-> H202 + 02 (9) H202 + 021' -————-> 02 + OH" + H0- (12) 02" + Fe‘*'3 --—""""9 Fe+2 + 02 (8) Fe” + H202 ——-—) Fe+3 + OH" + H0- (10) Reaction 11 would be catalyzed by the microsomal enzyme NADPH-cytochrome P450 reductase. The requirement for ADP was proposed to be for the solubility of iron in a neutral aqueous solution. The authors proposed that HO° was formed via reactions 11, 9, 8 and 10. Little if any HO° formation was proposed to occur via reactions 11, 9 and 12, the uncata- lyzed Haber-Weiss reaction. If HO° is generated it is most likely via the mechanism the authors propose since it is now widely accepted that the uncatalyzed Haber-Weiss reaction does not occur [66-68] and its par- ticipation in reactions of biological interest would be negligible -13- [69-71]. Further investigations into this reaction system by King et al. [72] supported the basic mechanism originally proposed by Fong et a1 [39]. There are some critical questions that must be addressed when the proposal of Fong et al. [39] is examined. First, if microsomes are heavily contaminated with catalase [73,74] how can the addition of exo- genous catalase inhibit a HZOZ-dependent reaction? It has been shown that the endogenous catalase content of microsomes is so great that unless azide is added to the microsomal suspension H202 production can- not be demonstrated [73,74]. The inhibition observed by Fong et al. upon addition of catalase is perhaps due to the presence of a stabiliz- ing antioxidant, such as thymol, in the commercial catalase preparation used. Most commercial catalase preparations contain a stabilizing anti- oxidant which must be removed prior to use. The stabilizing agents can often be quickly and conveniently removed by column chromatography of the commercial enzyme preparation over a desalting column. Additionally, other investigators have shown that stabilizer-free catalase does not inhibit but actually stimulates superoxide-dependent lipid peroxidation [64,75]. Second, Noguchi and Nakano [60] have demonstrated that the reduction of ADP-chelated iron by NADPHécytochrome P450 reductase is not superoxide-dependent and is not inhibited by superoxide dismutase. If this is true then the addition of superoxide dismutase to the reaction mixture should not inhibit but perhaps actually stimulate lipid peroxi- dation if it occurs by the proposed mechanism. Stimulation would arise from increased H202 production in the presence of superoxide dismutase while the rate of iron reduction would not be affected since this could be directly catalyzed by the reductase. Stimulation would occur only if -14- the rate limiting step is not the reduction of iron. Finally, the use of H0: traps to indicate the participation of H0' in a reaction is at best a tennable position. The inhibition of a reaction by H0° traps does not necessarily indicate that the observed reaction occurs via HO' but simply that an oxidant capable of oxidizing the trap participates in the reaction being studied. This subject will be more fully discussed later. Lai et al. [56,57,76] also proposed that NADPH-dependent microsomal lipid peroxidation was initated via HO°. These authors proposed that HO° was formed in essence by a superoxide driven, EDTA-chelated iron promoted Haber—Weiss reation. The mechanism of HO' generation proposed by Lai et a1. is the same as that proposed by Fong et al. [39], save for the iron chelator used and that the iron chelate must be added in the ferrous form. The mechanism of NADPH-dependent microsomal generation of HO° proposed by Lai et a1. is given below: 202 + NADPH 4, NADP+ + 11+ + 202? (11) 2022r + 211*" 2, H202 + 02 (9) 121312-1222+2 + H202—————) EDTA-Fe+3 + 011- + HO° (13) Reaction 11 is again catalyzed by NADPH-cytochrome P450 reductase. The mechanism suggested by Lai et al. is based on experimental evidence that: l) lipid peroxidation, as measured by malondialdehyde formation, can be inhibited by the addition of the spin trap 5,5-dimethyl-l- pyrroline—l-oxide (DMPO) [77] to a microsomal reaction mixture, 2) the electron spin resonance (ESR) signal generated by the reaction of DMPO with a free radical present in the microsomal reaction mixture is iden- tical to that formed when HO° and DMPO react; 3) spin adduct signal -15- intensity parallels the progress of malonaldehyde formation; 4) for- mation of the spin trap adduct signal is enhanced by the addition of superoxide dismutase to the reaction mixture; 5) addition of the HO° trap thiourea inhibits spin adduct formation; and 6) spin adduct signal formation requires EDTArFe+2. The inhibition of lipid peroxidation and ESR signal generation are proposed to be due to the following reaction: "t / H + H0' ——) "6 HE. IS: IV | l 0" o'_ (14) In examining the data that led Lai et al. to their final conclu- sions, a few critical pieces of information must be kept in mind. First, several researchers have previously demonstrated that NADPH* cytochrome P450 reductase in the presence of NADPH can reduce EDTAr chelated ferric ion [60,78-80]. Several researchers have also shown that superoxide, which may be produced by NADPH-cytochrome P450 reduc- tase during its catalytic cycle, can reduce EDTArchelated ferric ion [81-83]. The observed requirement for EDTArchelated ferrous ion for lipid peroxidation and spin adduct formation is thus curious. Second, other investigators [37-39,43-55] have been unable to demonstrate the promotion of NADPH—dependent microsomal lipid peroxidation in the pre- sence of EDTAriron alone. Third, the rates observed by Lai et al. are very low compared to the rates reported by others [37-39,43-55], almost at the level of background or autoxidative rates observed by others. Finally, spin traps such as that used by Lai et al., can enter into -l6- several side reactions in complex reaction mixtures. The use of proper controls must be stringently observed if reliable experimental data is to be obtained [84]. For example, in the presence of EDTA-chelated ferrous ion the DMPO-OH signal observed may have actually arisen from the breakdown of the DMPO-superoxide or DMPO-hydroperoxide spin adduct. The same would be true for DMPO adducts of lipid hydroperoxy radicals. These reactions are given below: M“ H ‘i‘ F312 -—--—-9 M‘ H s09;- “VI-Tea Me COB *H" W: N 0“ I (l). O’ (15) Or the hydroxyl radical spin adduct can be formed by the ferric ion oxi- dation of DMPO [85,86]: 4' F273 H + 13 Me / H ___; Me 7< )(OH + H + Fe He + H30 Me N N I t (16) This reaction also occurs with chelates of ferric ion. Thus it is obvious that the application of spin trapping techniques in a biological system can quickly become complex and that the use of properly designed control reactions is essential to the obtainment of valid experimental data. Perferryl ion (Fe+202 a Fe+3027)-promoted, hydroperoxide- independent initiation of lipid peroxidation has been proposed by investigators for both NADPH-dependent and superoxide-dependent lipid peroxidation. Perferryl ion-promoted initiation of NADPH-dependent microsomal lipid peroxidation was first proposed by Ernster and co- workers [37,43-45]. The proposal that the perferryl ion could promote -17- the initiation of lipid peroxidation was based on the similarities be- tween lipid peroxidation and the numerous biological oxidation systems studied by Mason in which the perferryl ion had been proposed to mediate the reaction [87,88]. Perferryl ion-promoted initiation of superoxide- dependent lipid peroxidation in both microsomes and liposomes was first proposed by Svingen et al. [63]. The mechanism of superoxide-dependent lipid peroxidation proposed by Svingen et al. was based on similarities to the NADPHrdependent reaction system. In both systems chelation of iron by ADP or a similar nucleotide or pyrrophosphate was required. Initiation of lipid peroxidation by an activated dioxygen- transition metal complex was first proposed by Heaton and Uri for a cobaltous-stearate reaction mixture that could promote PUFA peroxidation [89]. The close parallels between cobalt and iron chemistry led these authors to propose that a similar reaction occurred when the ferrous ion was substituted for the cobaltous ion. Experimentally this relationship has been borne out and much of the chemistry of the perferryl ion has been first investigated in a cobalt model system. As previously dis- cussed, reversible dioxygen binding by transition metals occurs with the reduced form of a transition metal that can undergo a one electron redox reaction. Dioxygen activation occurs because complexation with the transition metal imparts free radical characteristics to the oxygen molecule. This free radical characteristic of dioxygen circumvents the spin restrictions that are present for the reaction of ground state dioxygen with organic molecules, such as PUFA, and allows the reaction to occur. The reduction states of the dioxygen-ferrous ion complex can perhaps best be represented by the following set of equivalent struc- tures: -18- Fe+2 + 02 (——9 Fe+202 e—a Fe+302"' (17) The greatest contribution may come from the resonance form shown as the oxidized metal—superoxide complex. The formal charge on the iron atom of the perferryl ion is +6. Thus the perferryl ion is quite electro- negative and as expected it is predicted to be a strong oxidant. The reaction(s) by which activated dioxygen-ferrous ion complexes are pro- posed to initiate lipid peroxidation may be one of the following reac- tions [26]: Fe+302* + LH —-———> Fe+302H‘ + L- (18) Fe+302" + LH ———-+ Fe+2 + 110-2 + L- (19) Fe+302=' + LH ———-—-) Fe+3on + Lo- (20) Fe+30f + LH ———-:, Fe+3 - L‘ + H02° (21) At the present time there is no experimental evidence favoring one ini- tiation reaction over another, however, application of ESR spin trapping techniques may be useful in elucidating the relative importance of these reactions. Experimental evidence indicating that the initiation of lipid per- oxidation is promoted by the perferryl ion relies heavily on the observed requirement for ferrous ion and oxygen or ferric ion and super- oxide for the initiation of lipid peroxidation. Unfortunately, propo- sals for perferryl ion-promotion of initiation have often had to rely on corroborative negative data indicating that other intermediates, such as H0°, H202, superoxide alone or iron alone, cannot promote or do not par- ticipate in the promotion of initiation of lipid peroxidation. Thus, it often appears that the proposal for perferryl ion promotion of initi- ation of lipid peroxidation is used as a stop-gap measure filling a void -19- in our knowledge. However, the circumstantial evidence that the per- ferryl ion-promotes initiation of lipid peroxidation is perhaps admissable since to date the perferryl ion has not been isolated and circumstantial evidence is all that is available. The direct demonstra- tion of perferryl ion-promotion of initiation of lipid peroxidation must wait for further experimental and theoretical development. In their original paper, Hochstein and Ernster [37] did not observe a requirement for iron in the promotion of NADPH-dependent microsomal lipid peroxidation but rather an ADP requirement. However, subsequent investigations by these authors demonstrated that the commercial pre- paration of ADP used in their experiments was contaminated with iron and that both ADP and iron were required for the promotion of NADPH- dependent microsomal lipid peroxidation [43]. The perferryl ion was proposed to be formed in two steps. First, NADPH-dependent reduction of ADP-chelated ferric ion via a microsomal flavoprotein and second, the addition of dioxygen to the ADP-chelated ferrous ion complex. The requirement for ADP was proposed to be to chelate the ferric ion and keep it in solution at neutral pH. If it were not for chelation the ferric ion would precipitate as the hydroxide and the concentration of ferric ion would be drastically reduced. Pederson and Aust [55,58,75] also proposed that the perferryl ion promoted the initiation of NADPH-dependent lipid peroxidation. These investigators studied not only NADPHrdependent microsomal lipid peroxi- dation but also the reconstitution of microsomal lipid peroxidation in liposomes utilizing NADPH-cytochrome P450 reductase. In the microsomal reaction mixture these investigators observed the same requirements for promotion of lipid peroxidation as did Ernster and co-workers -20- [37,43-45]. Pederson and Aust also found that peroxidation could be further stimulated by the addition of EDTA-chelated ferric ion to a reaction mixture containing ADP-chelated iron. However, EDTA-chelated iron alone could not promote initiation of lipid peroxidation. These authors found that the NADPH-cytochrome P450 reductase promoted peroxi- dation of liposomes required both ADP-chelated ferric iron and EDTA- chelated ferric iron for maximal rates of peroxidation. The rates of peroxidation observed in the presence of ADP-chelated iron alone were very low as compared to NADPHrdependent ADP-chelated iron-promoted microsomal lipid peroxidation. No peroxidation was observed in the pre- sence of EDTA-chelated ferric iron alone. However, addition of EDTAr chelated ferric ion together with ADP-chelated iron greatly stimulated the rate of peroxidation. The proposal that the ADP-perferryl ion pro- moted initiation was based to some extent on negative data. Pederson and Aust could not demonstrate the participation of H202 or HO° in NADPH-dependent lipid peroxidation. The absolute requirement for ADP-chelated feric iron and NADPH, coupled with the known ability of NADPH-cytochrome P450 reductase to reduce ADP-chelated iron [60,79], indicated that ADP-chelated ferrous ion was probably directly involved in initiation. Other researchers had previously demonstrated that ferrous ion alone could initiate lipid peroxidation [90,91]. Because of the requirement for reduced iron in an oxygenated solution Pederson and Aust proposed that one of a variety of reactive intermediates formed between ferrous ion and oxygen was involved in the initiation of lipid peroxidation. One such reactive intermediate is the perferryl ion. It was previously demonstrated that these reactive intermediates could pro- mote reactions such as aromatic hydroxylations [92-94] and bio- and -21- chemiluminescence in aqueous solution [95,96]. Thus, these reactive intermediates may perhaps be capable of initiating lipid peroxidation. Pederson and Aust did not propose a function for EDTArchelated iron other than it replaced some microsomal component that was absent in liposomes prepared from the Folch lipid extract [97] of microsomes. There are some critical questions posed by the results of Pederson and Aust [55,59,62,75]. First, why are both ADP-chelated iron and EDTA- chelated iron required for lipid peroxidation? The answer is fairly obvious and is the one put forth by the authors themselves. The irons function to promote different reactions. The function each iron plays is dictated by the characteristics of its chelation. Chelation effects the redox potential of the iron. The redox potential of the iron not only affects the stability of the perferryl ion [98] but also the like- lihood that the iron will participate in such reactions as reductive activation of hydroperoxides [99,100]. Second, in their investigation of superoxide-dependent lipid peroxidation, they also found that both forms of iron chelates were necessary for the promotion of liposomal peroxidation [62,75]. What is the role for these iron chelates in the mechanism of superoxide-dependent lipid peroxidation? Pederson and Aust propose no role for these iron chelates and instead propose that lipid peroxidation is initiated via singlet oxygen addition to PUFA. These authors proposed that singlet oxygen was produced by superoxide dismu- tation. However, the conclusion that singlet oxygen is responsible for initiation is based on data showing that the singlet oxygen trap 2,5- diphenylisobenzofuran inhibited lipid peroxidation. This data may be in error as it has been suggested that 2,5-diphenylisobenzofuran is a free radical trap in addition to being a singlet oxygen trap [72]. -22- Svingen et al. [63,78] have proposed that the ADP-perferryl ion is involved in the initiation of both NADPH-dependent lipid peroxidation and superoxide-dependent lipid peroxidation. Using an iodometric method for the determination of lipid hydroperoxides [33] these authors demon- strated that lipid hydroperoxides could be generated in essentially lipid hydroperoxide free liposomes, prepared from extracted microsomal lipid, by ADP-chelated ferrous ion, by ADP-chelated ferric ion in the presence of NADPH and NADPH-cytochrome P450 reductase or by ADP-chelated ferric ion in the presence of a superoxide generating system, such as xanthine and xanthine oxidase. Lipid hydroperoxide formation could not be promoted by ferrous ion, ferric ion, ADP-ferric ion or by either EDTArchelated ferric or ferrous ion. Promotion of lipid hydroperoxide formation in the above reaction mixtures was essentially totally inhi- bited by superoxide dismutase but was not inhibited by catalase or H0- traps. Since superoxide alone has been shown to be unable to initiate lipid peroxidation [63,78] these authors, in light of the requirement for ADP-chelated iron concluded that promotion of initiation of lipid peroxidation occurred via the ADP-perferryl ion. The perferryl ion could be formed by reduction of ADP-chelated ferric ion followed by dioxygen addition or by the reaction of ADP-chelated ferric ion with superoxide. While EDTA—chelated ferric or ferrous ion could not promote initial formation of lipid hydroperoxides these authors found that addi- tion of EDTA-chelated ferrous ion, or EDTA-chelated ferric ion in the presence of reducing equivalents, could greatly stimulate ADP-perferryl ion promotion of lipid peroxidation. Svingen et al. proposed, as did Pederson and Aust [75] that the type of chelator used had a dramatic -23- effect on the participation of iron in lipid peroxidation. As pre- viously stated, chelation was proposed not only to facilitate iron par- ticipation in promotion of lipid peroxidation but to also moderate the role played by iron by changing the redox potential of iron. Chelation also has a dramatic effect on the stability of the perferryl ion not only from a redox standpoint but also from the standpoint of steric hin- derance towards further autoxidation [98,101-104]. The effect of chela- tion is cyclical in nature. The perferryl ion is most stable when che- lated by weak ligands such as phosphate anions or very strong chelators such as in the oxygen carrying hemoproteins. Between the two extremes of chelation there lies a whole spectrum of chelation effects with their corresponding effects on perferryl ion stability. Chelates that donate electron density to the ferrous ion strengthen the iron-oxygen bond in the perferryl ion and thus stabilize the perferryl ion. In the opposite sense, chelates that withdraw electron density from the ferrous ion reduce the stability of the perferryl ion. Applying this theory to the effects of chelation expressed in lipid peroxidation it can be seen that ADP, a relatively weak chelator which may increase electron density on the iron center and thus increases perferryl ion stability, gives an iron complex which is an active promoter of hydroperoxide-independent initiation of lipid peroxidation. 0n the other hand, chelation of ferrous ion by a stronger chelator, EDTA, would be expected to reduce the stability of the perferryl ion, if not forego its formation altogether. As predicted chelation of ferrous ion by EDTA renders the ferrous ion unable to promote hydroperoxide-independent initiation of lipid peroxidation. This theory is also supported by the demon- strated ability of free ferrous, where the ligand would be water, -24- to promote hydroperoxide independent initiation of lipid peroxidation [105]. The foundations of this theory will continue to be tested as more and more chelators of iron are examined for their effect on the ability of ferrous ions to promote lipid peroxidation. Like Pederson and Aust [55,58,62,75], Svingen et al. [63,78] found that both ADP-chelated and EDTArchelated iron were required for maximal rates of lipid peroxidation in liposomes. This was true for either NADPHrdependent or superoxide-dependent lipid peroxidation. If only the ADP-perferryl ion participates in the LOOH-independent initiation of lipid peroxidation the authors are faced with the question, what is the role played by EDTArchelated iron? Svingen et al. proposed that EDTA-chelated iron promoted the breakdown of initially formed lipid hydroperoxides. The EDTArchelated iron-promoted breakdown of lipid hydroperoxides is essentially the same as reductive activation of orga- nic hydroperoxides by ferrous ions [107-110]. Reductive activation of lipid hydroperoxides generates lipid alkoxy radicals and can be con- sidered hydroperoxide—independent initiation of lipid peroxidation. (Lipid hydroperoxide-dependent initiation of lipid peroxidation will be the subject of the final section of this review.) Finally, it has been proposed that the perferryl ion promotes lipid hydroperoxide formation in the lipoxygenase catalysis of lipid hydro- peroxide generation in PUFA [lll]. Nakano and Sugioka [111] have pro- posed that the perferryl ion promotes the abstraction of hydrogen from the methylene carbon of PUFA giving rise to a PUFA alkyl radical. This is essentially the same reaction mechanism proposed for the initiation of lipid peroxidation by the ADP-perferryl ion in lipid hydroperoxide free lipids, Fig. l. The generation of a PUFA alkyl radical during the -25- catalytic cycle of lipoxygenase has been demonstrated by ESR spin trapping techniques [112,113]. Continued investigation into the mecha- nisms of lipoxygenase catalysis may shed some light on the mechanism of ADP-perferryl ion promoted initiation of lipid peroxidation. It has already been shown that the lipoxygenase reaction is free radical in nature and is inhibited by superoxide dismutase but not by HO° or 102 traps. The mechanism for lipoxygenase thus appears to be similar to a controlled form of NADPH-dependent or superoxide-dependent ADP-perferryl ion-promoted initiation of lipid peroxidation. As can be seen from the above discussions there are two basic mechanisms that have been proposed for the initiation of iron promoted lipid peroxidation in hydroperoxide free lipids. Initiation via the perferryl ion and initiation via the hydroxyl radical. The two mecha- nisms of initiation have been proposed separately for the same reaction system, NADPH-dependent microsomal lipid peroxidation. Investigators have found strong experimental evidence to support both proposals. Such widely disparate conclusions drawn from the investigations of the same reaction mixture raise the question of whether grossly different experi- mental techniques are being employed? However, the answer to this question is an emphatic no. The possibility then arises that the researchers are perhaps looking at different aspects of the same reac- tion system. Perhaps the researchers should be examining the energetics of the initiation reaction and not concentrating so hard on giving a physical description to the intermediates involved. This approach is supported by experimental findings that the hydroxyl radical formed by Fenton's Reagent and the perferryl ion formed by ferrous ion autoxida— tion may not be as discrete entities as once believed. It is now -26- apparent that rather than discrete entities these two reactive inter- mediates are more likely to be part of a broad continuum of iron-oxygen reactive intermediates. Attempts at identification of the iron-oxygen intermediates may be more reflective of the experimental techniques employed than the actual intermediate. The experimentation of Walling et al. [114] and of Groves et a1. [115-117] indicates that free HO' may not be formed by the Fenton's Reagent under any but strongly acidic conditions. Rather some type of iron-hydroxyl radical intermediate is probably formed. It is only under conditions of low pH that Fenton's Reagent promoted reactions give pro- duct distributions truly characteristic of a free radical reaction. As the polarity and/or pH of the reaction mixture is altered the reaction products are less characteristic of a free radical mechanism and show the growing influence of a stereospecific reaction. The stereospecific nature of several Fenton's Reagent-promoted hydroxylations led Groves et al. [115-117] to propose that except under strong acid conditions the oxidative intermediate that promotes hydroxylation is the ferryl ion (Fe+20) and not HO'. The ferryl ion is known to be a strong oxidant and it has been proposed to be the promoter of several ferrous ion dependent oxidation reactions [118-121]. Ferrous ion dependent hydroxylations are essentially identical to Fenton's Reagent promoted hydroxylations run at neutral pH. The formation of ferryl ion from ferrous ion has been pro- posed to occur by the following mechanism. (Note that the first step of the mechanism involves formation of the perferryl ion) [122,123]. Fe+2 + 02 ; Fe+202 (22) Fe+202 + Fe+2 -——---) Fe+202Fe+2 (23) Fe+202Fe+2 4% 2 Fe+20 (24) -27- The ferryl ion could be formed in both the NADPH-dependent and superoxide-dependent lipid peroxidation reaction mixtures studied by Pederson et a1. [55,58,62,75], Svingen et al. [27,63,78] or McCay and co-workers [48,49,52,72]. LIPID HYDROPEROXIDE-DEPENDENT INITIATION OF LIPID PEROXIDATION Once low levels of lipid hydroperoxides are present in a lipid matrix the predominant mechanism of initiation involves their breakdown to form free radicals. I LOOH —-————-€) free radicals (2) Thus when lipid hydroperoxide formation promoted by the ADP-perferryl ion reaches a significant level its contribution to total peroxidation is quickly overshadowed by the contributions from lipid hydroperoxide- dependent mechanisms of initiation. The mechanisms of free radical generation from lipid hydroperoxides, especially by heavy metal promoted reactions, have been extensively studied and have been the subject of several reviews [124-126]. The predominance of metal promoted hydroperoxide-dependent initiation over oxygen activation mechanisms of initiation has led to special difficulties in the investigation of the mechanism of initiation of lipid peroxidation. The problem that arises for the biologist or biochemist is that during isolation of samples, and their subsequent handling, hydroperoxides are often formed in the mem- branes by autoxidation. Unless special precautions are taken to mini- mize or eliminate lipid hydroperoxide formation during sample manipula- tions one may draw the conclusion, and perhaps rightly so, that only lipid hydroperoxide-dependent initiation occurs in enzymatic lipid -28- peroxidation. Contamination of samples by lipid hydroperoxides can completely mask the activated oxygen mechanism of initiation. The mechanisms of hydroperoxide-dependent initiation of free radical reactions have been extensively discussed in detail [124-126]. Free radical generation from lipid hydroperoxides occurs by three general mechanisms: 1. Thermal or Unimolecular Homolysis: LOOH ———§ LO' + HO° (25) Hydroperoxides which are generated by autoxidation or other means undergo unimolecular homolysis at 37°C at an extremely slow rate. In fact it has been estimated that if H202 loss was by unimolecular homoly- sis only, its half life at body temperatures would be 1011 years [18]. Thus from all indications it is very doubtful that this mechanism of free radical generation contributes appreciably to lipid hydroperoxide- dependent initiation of lipid peroxidation. 2. Molecule Induced Homolysis (MIH) or Molecule Assisted Homolysis (MAH) - Bimolecular Homolysis: LOOH + LH -————§, LO° + L- + H20 (26) In this mechanism there is not only bond breakage but also bond formation, thus the energy of activation and the endothermicity of the reaction are both greatly reduced as compared to reaction 25. Reactions that occur via MIH are much faster than unimolecular reactions. The rate of MIH reactions are increased by polar solvents especially those which can hydrogen bond. However, the importance of free radical generation from lipid hydroperoxides via MIH mechanisms is likely to be insignificant compared to the rate of lipid peroxidation in biological -29- systems where the presence of heavy metals, most notably iron, can pro- mote reductive activation of lipid hydroperoxides. 3. Reductive Activation Fe” + LOOH —-> Fe+3 + LO° + 011- (27) Fe+3 + LOOH -——§ Fe” + L00° + 11+ (28) The metal promoted decomposition of organic hydroperoxides and peroxidic material has been extensively studied. The reductive activa- tion reaction with ferric iron (28) is a relatively slow reaction because it requires prior ionization of the hydroperoxides. Hydroper- oxides as a family have a pKa of 10.8 [125]. Free ferric ion is insoluble above pH 4. Since the two reactants have opposite pH maxima for acti- vity it is fairly obvious why this is a slow reaction [127,128]. The rate of the ferric ion reaction (28) can be greatly affected by chela- tion. This effect may be due to increased solubility at higher pHs, but the system is more complex. Chelation has been found to enhance, sup- press, or not affect the rate of reaction 28 depending on the nature of the chelate, the metal, and the chelate-metal complex formed [129]. For instance, it is well known that many ferric hemeproteins can promote the breakdown of lipid hydroperoxides in an extremely efficient manner. In general, the ferrous ion dependent reaction is much more rapid than the ferric ion reaction. However, as with the ferric ion reaction, the ferrous ion reaction is greatly affected by chelation [100]. In biolo- gical systems, either in_viyg_or in yitrg, it is likely that reductive activation is the predominant mechanism of free radical generation in lipid hydroperoxide-dependent initiation of lipid peroxidation. The ability of metal ions and heme compounds to catalyze the oxida- tion of unsaturated fatty acids has been extensively studied. Tappel -30- et al. [129-131] found that lipid peroxidation catalyzed by hematin com- pounds is a basic pathological reaction in_yizg_and a deteriorative reaction in vitrg. These investigators found that heme compounds are the most powerful catalysts of lipid peroxidation found in animal tissues. Additionally, hemeproteins are often found intimately asso- ciated with lipid membranes. Svingen et al. [63,78] demonstrated that in both NADPH-dependent and superoxide-dependent peroxidation of liposomes EDTA-chelated iron promoted reductive activation of lipid hydroperoxides. These investiga- tors found that EDTA—chelated iron promotion of lipid peroxidation was dependent upon the reduction of the ferric ion complex. The EDTA-ferric ion complex can be reduced either by superoxide or directly by NADPH- cytochrome P450 reductase in the presence of NADPH. EDTA-chelated iron promotion of lipid peroxidation requires lipid hydroperoxides and reduction of the iron complex. The EDTArferrous ion promoted lipid hydroperoxide-dependent initiation reaction was found to be free radical in nature, not to occur via a perferryl ion (activated dioxygen) inter- mediate, as it was not inhibited by superoxide dismutase, and to not involve HO°. Svingen et al. [63,78] demonstrated the integral relationship of ADP-perferryl ion hydroperoxide-independent initiation of lipid peroxida- tion and the EDTA-ferrous ion hydroperoxide-dependent initiation of lipid peroxidation. These authors found, as one might guess, that the latter was dependent on the former. Svingen et al. also demonstrated that hydroperoxide-dependent initiation accounts for greater than 90% of the total peroxidic products formed. Thus it is apparent that once a significant concentration of lipid hydroperoxides is initially formed -31- the predominant mechanism of initiation then becomes reductive activa- tion of lipid hydroperoxides. Svingen et al. [63,78] also directly demonstrated that in microso- mal lipid peroxidation, where only ADP-chelated iron is required, the microsomal cytochromes, most notably cytochrome P450, promote the heterolytic activation of ADP-perferryl ion generated lipid hydroperoxi- des. These investigators showed that inhibition of microsomal lipid peroxidation by non-antioxidant inhibitors of cytochrome P450 could be completely reversed by the addition of EDTArchelated iron. Since EDTA- chelated iron must be reduced before it will promote reductive activa- tion of lipid hydroperoxides it is apparent that the drug substrates inhibit lipid peroxidation by interacting with cytochrome P450 and not by competition for reducing equivalents [63]. The inhibition of cyto- chrome P450 promoted heterolytic activation of lipid hydroperoxides by oxidizable substrates of cytochrome P450 and the lipid hydroperoxide- dependent co-oxidation of those drug substrates demonstrates that inhibition is due to a peroxidase mechanism in which the substrate is oxidized and the lipid hydroperoxide is likely reduced to an alcohol. Reduced cytochrome P450 can also function as a peroxidase, reducing lipid hydroperoxides to lipid alcohols. -32- CHAPTER 1 INITIATION OF LIPID PEROXIDATION SUMMARY An investigation into the mechanism of lipid peroxidation promoted by xanthine oxidase showed a dependence upon superoxide, singlet oxygen and adenosine 5'-diphosphate chelated iron (ADP-Fe+3). In the absence of ADP-Fe+3 or in the presence of superoxide dismutase there is complete inhibition of enzymatically promoted peroxidation. Initiation of per- oxidation likely occurs through a complex formed by ADP, Fe+3 and superoxide. Use of the singlet oxygen trapping agent 2,5-diphenylfuran showed that singlet oxygen does not participate in the LOOH-independent initiation of peroxidation but rather in its LOOH-dependent initiation. The mechanism of NADPH-cytochrome P450 reductase-promoted and ADP-Fe+2-promoted lipid peroxidation parallel that of xanthine oxidase in that LOOH-independent initiation occurs through a superoxide dismu- tase-sensitive reaction and that singlet oxygen is present only during LOOH-independent initiation of lipid peroxidation. Superoxide dismutase sensitivity may result from the scavenging of 02' which disassociates from the reduced iron-oxygen complex leaving leaving oxidized metal: ADP-Fe+3+e" ADP-Fe+2 + 02 ADP-Fe+2-02 ADP-Fe+3-02'°' ADP-Fe+3 + 02" (29) INTRODUCTION The role that superoxide (02?) and singlet oxygen, 102(1 8) play in the peroxidation of unsaturated lipids is a subject of current intensive research. The number of biological systems known to produce 027 and catalyze lipid peroxidation is expanding. Included are the oxidation of -33- xanthine by xanthine oxidase (62), the cyclic reduction and oxidation of paraquat (methyl viologen) by microsomal NADPH-cytochrome P450 reductase and oxygen [132,133] and perhaps the same oxidation-reduction cycle of the anthraquinone structure of the anticancer drug adriamycin [134]. A mechanism of superoxide-dependent lipid peroxidation promoted by xanthine-xanthine oxidase has been proposed by Pederson and Aust [62] based on the dismutation of superoxide to give singlet oxygen according to the scheme [135]: 2021- + 2st 7‘ 102 + H202 (30) Singlet oxygen was proposed to be the direct initiator of lipid peroxi- dation by a concerted addition-abstraction reaction with the diene bonds of unsaturated lipid giving rise to lipid hydroperoxides [30]. Enzym- atic peroxidation required ADP—Fe+3 and was enhanced by the addition of Fe+3 chelated by EDTA. Initiation of xanthine-xanthine oxidase-promoted lipid peroxidation mediated through 102 was also proposed by Kellogg and Fridovich [136]. They reported that peroxidation was dependent upon 02' and H202 and pro- posed the following scheme for the generation of 102: 027 + 11202 -—-—-——-) OH“ + on- + 102 (31) Contrary to these two schemes, King et al. [72] found no evidence for the production of 102 by the xanthine-xanthine oxidase system in the absence of active lipid peroxidation. Indeed, whether the self dismuta- tion of 02' gives rise to 102 was questioned by Nilsson and Kearns [137]. King et al. [72] proposed that 102 was produced in both xanthine-xanthine oxidase and NADPH-cytochrome P450 reductase-promoted lipid peroxidation only after lipid peroxidation was initiated and that 102 was perhaps formed from the breakdown of lipid hydroperoxides (LOOH). -34- In this chapter a unified mechanism for lipid hydroperoxide- independent initiation of lipid peroxidation promoted by NADPH-cyto- chrome P450 reductase, and xanthine oxidase in the presence of ADP-Fe+3 and by ADP-Fe+2 alone is proposed. Lipid hydroperoxide-independent ini- tiation of lipid peroxidation is dependent on the production of an ADP-perferryl ion complex (ADP-Fe+2-02 i ADP-Fe+3—027). This initiation complex shows superoxide dismutase sensitivity. Singlet oxygen par- ticipation in lipid peroxidation arises from the breakdown of lipid hydroperoxides promoted by reduced iron chelates (ADP-Fe+2, EDTA-Fe+2). Once produced, 102 reacts with unsaturated lipid to produce lipid hydroperoxides. There are several LOOHrdependent reactions, only some of which produce 102. These reactions give rise to the observed sen- sitivity to 102 trapping agents, and account for the observed trapping agents' partial inhibition of lipid peroxidation. MATERIALS AND METHODS Chemicals: Milk xanthine oxidase, Sigma Type I (E.C. No. 1.2.3.2.), superoxide dismutase, bovine erythrocyte (E.C. No. l.15.l.l.), soybean lipoxygenase (E.C. No. 1.13.1.13), adenosine 5'-diphosphate (ADP), cytochrome c, Sigma Type IV, nicotinamide adenine dinucleotide phosphate, reduced form (NADPH) and xanthine were obtained from Sigma Chemical Company. 2,5-Diphenylfuran (DPF) was obtained from Eastman Organic Company. Trans-1,2-dibenzoyl ethylene was obtained from Aldrich Chemical Company. All other reagents used were of analytical grade. Microsomes and Microsomal Lipid: Microsomal phospholipid and NADPH-cytochrome P450 reductase were prepared by the methods of Pederson et al. [55]. NADPH-cytochrome P450 reductase specific activity, deter- mined by the method of Pederson et al. [55], was 45-52 units/mg. -35- Superoxide dismutase activity was measured by the method of McCord and Fridovich [138]. Lipid hydroperoxides were prepared by the action of soybean lipoxy- genase on liposomes of extracted microsomal lipid. The liposomes were prepared by sonication and diluted to a final concentration of 1.0 pmole lipid phosphate per ml in 0.05 M Tris-Cl pH 9.0 at 37°C. Sodium deoxy- cholate, 0.04% w/v, was included in the incubation system to enhance the reaction. The incubation was initiated by addition of lipoxygenase, at 100 pg per ml. The buffer was saturated with oxygen and the incubation was carried out under an oxygen atmosphere. At the end of a 45 minute incubation, the phospholipid hydroperoxides were extracted and stored by the method of Pederson and Aust [55]. Lipid hydroperoxide content was measured iodometrically by the method of Buege and Aust [33]. Iron Chelates: Chelated iron solutions were prepared in 0.05 M Tris-Cl pH 7.5 at 37°C. When preparing the ferrous chelates buffers were saturated with argon. The molar ratio of the ADP-iron solution was 17:1 while that of the EDTAriron was 1.1:1. Liposomes: Control liposomal incubation systems contained 1.0 pmole lipid phosphate per ml in 0.05 M Tris-Cl pH 7.5 at 37°C. When lipid peroxidation was promoted by xanthine oxidase, buffers were saturated with 02. Incubations were carried out at 37°C in a metabolic shaker bath under an air atmosphere. Thiobarbituric acid (TBA)-reactive material was determined by the method of Bernheim et al. [139]. To eli- minate nonenzymatic chromophore formation during the assay, 0.03 volume of 2% butylated hydroxytoluene (in ethanol) was added to the TBA reagent. Lipid hydroperoxides were assayed as previously indicated -36- [33]. Where indicated superoxide dismutase was added in 0.05 M Tris-Cl pH 7.5 at 37°C and DPF as a 0.02 ml of a solution prepared in acetone. Addition of 0.02 ml of acetone to any of the incubations had no effect on activity. Other Methods: The oxidation of DPF by 102 was confirmed by following the decrease in fluorescence of DPF at 368 nm, using an exci- tation wavelength of 333 nm. Confirmation of product formation was obtained by thin layer chromatography (TLC) of the chloroform extract of the incubation systems. TLC was performed on Silica Gel G plates deve- loped in the dark in a hexane-dioxane (3:1) solvent system. DPF was visualized by its fluorescence. The 102 oxidation product, cis-l, 2-dibenzoylethylene, was visualized by spraying the plate with 0.5% 2,4-dinitrophenyl hydrazine in 2 M HCl and identified by comparison to a cis-l,2-dibenzoylethylene standard prepared from the trans-1,2- dibenzoylethylene isomer by the method of Lutz and Wilder [140]. Total lipid phosphate was measured by the method of Bartlett [141]. Protein was determined by the method of Lowry et al. [142] standardized with bovine serum albumin (Pentex) using Elzcm at 280 nm equal to 6.6. RESULTS Xanthine Oxidase-Promoted Lipid Peroxidation Lipid peroxidation, assayed as TBA-reactive material, during xan- thine oxidation by xanthine oxidase in the presence of ADP-Fe+3 is inhi- bited 95% by superoxide dismutase (Table 1). This is in agreement with the results of Pederson and Aust [62] and King et a1 [72]. Lipid hydro- peroxide production was inhibited 76% by superoxide dismutase (Table l). -37- TABLE 1 INHIBITION OF XANTHINE OXIDASE—PROMOTED LIPID PEROXIDATION BY SUPEROXIDE DISMUTASE (SOD) AND 2,5-DIPHENYLFURAN (DPF) Control reaction mixtures were as described under "Methods" with the following additions: 1.7 mM ADP, 0.1 mM FeCl3, 0.33 mM xanthine and where indicated 1.0 units SOD/ml or 0.2 mM DPF. Incubations and assays were performed as des- cribed under "Methods". Results are reported as initial rates. nmoles MDA nmoles LOOH min ml"1 min ml-1 Control 0.02 1.8 +DPF 0.04 0.0 +SOD 0.01 0.9 +xanthine oxidase 0.68 6.6 +xanthine oxidase +DPF 0.60 4.5 +xanthine oxidase +SOD 0.03 1.6 -33- This suggests a dependence upon a superoxide dismutase-sensitive ini- tiation reaction in this system. Inhibition of xanthine oxidase-promoted lipid peroxidation by the singlet oxygen trapping agent DPF is considerably less than that observed with superoxide dismutase (Table 1). Inhibition of the for- mation of TBA-reactive material was 13% while the inhibition of LOOH production was 32%. That DPF can only partially inhibit xanthine- xanthine oxidase-promoted lipid peroxidation is substantiated by the observations of Pederson and Aust [62] and King et al. [72] that even at saturating concentrations of singlet oxygen trapping agents diphe- nylisobenzofuran (DPBF) or DPF respectively, lipid peroxidation is inhibited by only approximately 50%. Addition of EDTA-Fe+3 is necessary for maximal peroxidation in xanthine oxidase-promoted lipid peroxidation [62]. This addition apparently has no effect on the initiation mechanism of lipid peroxida- tion as evidenced by the consistent effects of DPF and superoxide dismutase on a system containing EDTA-Fe+3 (Tables 2 and 3). 2,5- Diphenylfuran inhibits by only 6% while superoxide dismutase essentially completely eliminates formation of TBA-reactive material. Lipid hydro- peroxide production in the presence of superoxide dismutase is reduced by 57%. In the presence of DPF, LOOH formation is reduced by 14%. NADPH-Cytochrome P450 Reductase-Promoted Lipid Peroxidation Promotion of lipid peroxidation by NADPH-cytochrome P450 reductase in the presence of NADPH and ADP-Fe+3 is shown in Table 4. Addition of superoxide dismutase reduces the NADPH-dependent formation of TBA- reactive material by more than 86%. This is in good agreement with the inhibition of NADPH-dependent lipid peroxidation in microsomes by -39- TABLE 2 INHIBITION OF XANTHINE OXIDASE PROMOTED LIPID PEROXIDATION BY SOD IN THE PRESENCE OF EDTAPFe+3 Control reaction mixtures were as described under "Methods". The following additions were made where indicated: 1.7 mM ADP, 0.11 mM EDTA, 0.2 mM FeCl3, 0.33 mM xanthine and 1.0 units SOD/m1. Incubations and assays were performed as in Table 1. nmoles MDA nmoles LOOH min ml.1 min m1-1 Control 0.05 2.4 +SOD 0.04 0.5 +xanthine oxidase 2.10 22.5 +xanthine oxidase +SOD 0.05 10.1 -40- TABLE 3 INHIBITION OF XANTHINE OXIDASE-PROMOTED LIPID PEROXIDATION BY DPF IN THE PRESENCE OF EDTA-Fe'i'3 Control reaction mixtures were as in Table 2. 0.2 mM DPF was added where indicated. Incubations and product assays were performed as in "Methods”. nmoles MDA nmoles LOOH min ml”1 min ml‘1 Control 0.05 1.0 +DPF 0.02 0.1 +xanthine oxidase 1.41 19.5 +xanthine oxidase +DPF 1.32 16.8 -41- TABLE 4 INHIBITION OF NADPH-CYTOCHROME P450 REDUCTASE PROMOTED LIPID PEROXIDATION BY SOD Control reaction mixtures were as described under ”Methods". The following additions were made where indicated: 0.1 units NADPH-cytochrome P450 reductase/ml, 1.7 mM ADP, 0.1 mM FeCl3 and where indicated 1.0 units/m1 SOD. The reactions were initiated by the addition of 0.1 mM NADPH. Incubations and assays were carried out as in "Methods". nmoles MDA nmoles LOOH min ml"1 min ml"1 Control 0.01 0.5 +SOD 0.01 0.1 +NADPH 0.15 1.8 +NADPH +SOD 0.02 0.1 -42- superoxide dismutase reported by King et al. [72]. Lipid hydroperoxide production is inhibited by superoxide dismutase by 94%. Addition of DPF inhibited the formation of TBArreactive material by 13% and LOOH by 21%, as shown in Table 5. As observed in xanthine oxidase-promoted initiation, DPF was a much less effective inhibitor than superoxide dismutase. As with the xanthine oxidase system, addition of EDTA-Fe+3 enhanced NADPH-cytochrome P450 reductase-promoted lipid peroxidation. The effects of adding DPF and superoxide dismutase to this system (Table 6) paralleled those found in the system lacking EDTA-Fe+3 (Table 4). Inhibition of the formation of TBAreactive material by superoxide dismutase was essentially complete. Lipid hydroperoxide formation was inhibited 92% by superoxide dismutase. 2,5-diphenylfuran inhibited the formation of TBArreactive material by 20% while inhibiting LOOH formation by 37%. The results with NADPH-cytochrome P450 reductase-promoted lipid peroxidation closely parallel those found in the xanthine oxidase system. This is true for the action of both DPF and superoxide dismu- tase. This suggests that the mechanism of these two systems is closer than previously proposed by Pederson et a1. [55], Pederson and Aust [62], King et al. [72] and Kellogg and Fridovich [136]. ADP-Fe+2 Promoted Lipid Peroxidation The two systems previously discussed both show an absolute require- ment for the ADP-Fe+3 and an enhancement of peroxidation upon the addi- tion of EDTA-Fe+3 [55,62]. These iron forms may exist in either the oxi- dized or reduced state. Addition of the two iron forms in either of these oxidation states showed that ADP-Fe+2 alone was capable of promoting -43- TABLE 5 INHIBITION OF NADPH-CYTOCHROME P450 REDUCTASE PROMOTED LIPID PEROXIDATION BY DPF Reaction mixtures were as described for Table 3. 0.2 mM DPF was added where indicated. Incubations and assays were per- formed as in ”Methods”. nmoles MDA nmoles LOOH min ml"1 min m1-1 Control 0.03 0.6 +DPF 0.01 0.5 +NADPH 0.15 1.9 +NADPH +DPF 0.13 1.5 -44- TABLE 6 INVOLVEMENT OF SUPEROXIDE AND SINGLET OXYGEN IN NADPH- CYTOCHROME P450 REDUCTASE-DEPENDENT LIPID PEROXIDATION Control mixtures were as described under "Methods". The following additions were made where indicated: 1.7 mM ADP, 0.11 mM EDTA, 0.2 mM FeC13, 0.1 unit NADPH-cytochrome P450 reductase/ml and 0.2 mM DPF or 1.0 unit SOD/ml as indicated. Incubations and assays were performed as in "Methods". nmoles MDA nmoles LOOH min ml"1 min m1—l Control 0.01 0.1 +DPF 0.01 0.1 +SOD 0.01 0.1 +NADPH 2.50 15.0 +NADPH +DPF 2.00 9.4 +NADPH +SOD 0.02 1.1 -45- peroxidation in unperoxidized liposomes (Table 7). The results of addi- tion of inhibitors are similar to those seen in the xanthine oxidase and NADPH-cytochrome P450 reductase promoted lipid peroxidation systems (Tables 1-6). Superoxide dismutase inhibits ADP-Fe+2-promoted for- mation of TBA-reactive material by 70% while DPF inhibits by only 17%. Superoxide dismutase inhibits LOOH formation by 78% and DPF inhibits LOOH formation by 33%. Lipid Hydroperoxide-Dependent Initiation of Lipid Peroxidation Lipid hydroperoxides can react with reduced metals, in this case iron, to form products such as lipid free radicals, alkoxy free radicals and lipid hydroperoxy free radicals capable of initiating lipid peroxi- dation. Oxidized metals react at a much slower rate and their reaction with LOOH is minimal. The results reported in Table 8 show that both ADP-Fe+2 and EDTA-Fe+2 can promote the LOOHrdependent initiation of peroxidation as detected by malondialdehyde (MDA) formation. However, EDTA-Fe+2 appears to be a better promoter since it not only promotes an increase in MDA but also in LOOH, which is both a reactant and product of the reaction. ADP-Fe+2, on the other hand, is a poor promoter of reductive activation as shown by the rapid loss of LOOH. This is in agreement with the findings of Pederson and Aust [55,62] and those reported here that in the peroxidation of extracted microsomal lipid ADP-iron alone is not sufficient for maximal rates of peroxidation. To maximize rates it was found that EDTA-Fe+3 must be added. The EDTAriron evidently functions in the ferrous form to promote the breakdown of LOOH formed by the action of ADP-Fe+2, Table 7. Our findings emphasize the central position of LOOH-dependent initiation reactions in the overall scheme of lipid peroxidation. -46- TABLE 7 INVOLVEMENT OF SUPEROXIDE AND SINGLET OXYGEN IN THE PROMOTION OF LIPID PEROXIDATION BY CHELATED IRON Control reaction mixtures were as described under "Methods". The following additions were made where indicated: 1.0 unit SOD/ml, 0.2 mM DPF, 1.7 mM ADP with 0.1 mM FeCl3, 1.7 mM ADP with 0.1 mM FeClZ. Reactions were initiated by addition of the appropriate iron forms. Incubations and assays were performed as described in "Methods". nmoles MDA nmoles LOOH min ml"1 min ml-1 Control 0.00 0.0 +DPF 0.00 0.0 +SOD 0.00 0.0 +EDTArFe+3 0.00 0.0 +EDTA-Fe+2 0.00 0.0 +ADP-Fe+3 0.00 0.0 +ADP-Fe+2 0.30 1.3 +ApP-Fe+2 +DPF 0.25 0.9 +ADP-Fe+2 +SOD 0.09 0.3 -47- TABLE 8 LIPID HYDROPEROXIDE DEPENDENT INITIATION OF LIPID PEROXIDATION FROM LOOH CATALYZED BY ADP AND EDTA CHELATED IRON Lipid hydroperoxides were generated as described under "Methods". Incubation mixtures contained 0.1 mole lipid hydroperoxide/ml in 0.05 M Tris-Cl pH 7.5 at 37°C (1.02 mole lipid hydroperoxides/ mole lipid phosphate). 0.1 mM Fe+2 was added as indicated either as ADP-Fe+2 (17:1 mole ratio) or EDTA-Fe+2 (1.1:1 mole ratio). Reactions were initated by addition of the appropriate iron forms. Lipid peroxidation was determined as in "Methods". nmoles MDA *nmoles LOOH min ml"1 min ml'1 Partially peroxidized liposomes No additions 0.02 + 6.6 +ADP-Fe+2 1 . 56 -18. 8 +EDTA-Fe+2 1. 16 + 6.4 *Initial rates of increase (+) or decrease (-) of LOOH during incubation -48- It has been suggested that 102 may also be a product of hydro- peroxide breakdown [143]. Singlet oxygen so produced could function in the reactions of lipid peroxidation by addition to lipid diene bonds producing LOOH. The results reported in Table 9 support the concept that 102 is a product of LOOH decomposition and is an intermediate in the reactions of lipid peroxidation. Using EDTA-Fe+2 which has been shown to be an efficient promoter of LOOH-dependent initiation of lipid peroxidation (Table 8) but incapable of LOOH-independent initiation of lipid peroxidation (Table 7) the participation of 102 in the ferrous complex promoted breakdown of LOOH was investigated using the 102 trapping agent DPF. Lipid hydroperoxide-dependent initiation, as determined by formation of TBA-reactive material, was inhibited 20% by DPF. In the presence of DPF, LOOH formation was inhibited and a loss of LOOH was observed. It is apparent that 102 is produced and reacts with PUFA. Addition of superoxide dismutase to the same EDTArFe+2-LOOH mix- ture had no effect on the formation TBA-reactive material but did inhi- bit LOOH formation to the point that a loss in LOOH content was observed. Thus 02* may participate in these reactions but to a lesser extent than does 102. The nature of the ADP-Fe+2 complex was examined by investigating the inhibition of its restricted abilities to promote LOOH-dependent initiation by DPF and SOD, Table 9. This is a more difficult situation to analyze since ADP-Fe+2 can promote LOOH-independent as well as LOOH- dependent initiation. DPF does not inhibit LOOH-independent initiation but does inhibit LOOH-dependent initiation to some extent as evidenced by a decrease in LOOH content. Addition of superoxide dismutase to the incubation mixture which will inhibit LOOH-independent initiation by TABLE 9 INVOLVEMENT OF SUPEROXIDE AND SINGLET OXYGEN IN THE PROMOTION -49- OF LIPID PEROXIDATION IN PARTIALLY PEROXIDIZED LIPOSOMES Incubation mixtures were as described in the legend to Table 8 with the following additions where indicated: 1.0 unit superoxide dismutase addition of the appropriate iron form. determined as in ”Methods”. /ml. Lipid peroxidation was 0.2 mM DPF or Reactions were initiated by nmoles MDA *nmoles LOOH min ml.1 min ml"1 Partially peroxidized liposomes: no additions 0.02 + 6.6 +DPF 0.04 - 7.0 +SOD 0.02 + 2.4 +EDTArFe+2 1.16 + 6.4 +EDTArFe+2 +DPF 0.96 - 4.8 +EDTArFe+2 +SOD 1.20 - 5.0 +ADP-Fe+2 1.56 -18.8 +ADP-Fe+2 +DPF 1.60 —28.0 +ADP-Fe+2 +SOD 1.24 -13.2 *Initial rates of increase (+) or decrease (-) of LOOH during incubation. _50- ADP-Fe‘l'2 gives a true indication of the LOOH-dependent initiating abilities of ADP-Fe+2. The addition of superoxide dismutase clearly shows that ADP-Fe+2 is a less efficient promoter of LOOH-dependent ini- tiation than is EDTArFe+2. The results also show that the limited LOOH- dependent initiation of peroxidation promoted by ADP-Fe‘l'2 is not inhi- bited by superoxide dismutase. Thus the active agent in LOOH-dependent initiation promted initiation by ADP-Fe+2 must be significantly differ- ent from that involved in LOOH-independent initiation. DISCUSSION Lipid peroxidation as determined by TBArreactive material and LOOH may be divided into two sequential parts. Such a division was first proposed by Tam and McCay [52] as the result of time course studies in microsomal lipid peroxidation. They observed a precursor-product rela- tionship between LOOH and the TBA—reactive material produced during peroxidation. We have been able to define and separate LOOH-independent and LOOH-dependent initiation of lipid peroxidation. Lipid hydro- peroxide independent initiation is the generation of LOOH accompanied by low levels of TBA-reactive material. The rate of formation of TBA- reactive material is lower than usually associated with enzymatic peroxidation, while LOOH formation may be quite rapid. Lipid hydroperoxide-dependent initiation is the breakdown of the initially formed LOOH into reactive intermediates capable of continuing the free- radical reactions of lipid peroxidation. Lipid hydroperoxide- independent initiation of peroxidation, as detected by LOOH formation, can be promoted by ADP-Fe+2, by either NADPH-cytochrome P450 reductase or xanthine oxidase activity in the presence of ADP-Fe'i'3 or by soybean lipoxygenase as demonstrated in this paper. Lipid hydroperoxides so -51- formed may be used in the examination of LOOH-dependent initiation of lipid peroxidation. Lipid hydroperoxide-dependent initiation of lipid peroxidation promoted by EDTArFe+2, which is incapable of LOOH- independent initiation (Table 7), was demonstrated in partially peroxi- dized lipids. EDTA-Fe+2 caused the rapid formation of TBA-reactive material from hydroperoxides while at the same time promoting the for- mation of LOOH, Table 8. The separation of LOOH-independent and LOOH- dependent initiation of lipid peroxidation allows a more definitive investigation into the mechanism of lipid peroxidation. Superoxide dismutase inhibits the formation of TBA-reactive material by xanthine oxidase promoted LOOH-independent initiation by greater than 90% while inhibiting the formation of LOOH by more than 75% (Table l). 2,5-Diphenylfuran, on the other hand, inhibits by only 13% and 32% respectively. Lipid hydroperoxide-independent initiation of lipid peroxidation promoted by xanthine oxidase must therefore be depen- dent upon a superoxide dismutase-sensitive intermediate. The xanthine oxidase system demonstrates an absolute requirement for ADP-Fe+2 [62]. Addition of EDTA-Fe+3 does not change the system's dependence upon 02', ADP-Fe+3 or its sensitivity to superoxide dismutase (Table 2). Singlet oxygen does not appear to be involved in LOOH-independent initiation. However, it is formed during active lipid peroxidation [72] and may result in some further peroxidation as evidenced by DPF inhibition. Lipid peroxidation promoted by NADPH-cytochrome P450 reductase also exhibits an absolute requirement for ADP-Fe+3 [58]. The inhibitory effects of superoxide dismutase and DPF on the NADPH-cytochrome P450 reductase-promotion of LOOH-independent initiation of lipid peroxidation (Tables 4 and 5) were similar to those observed for xanthine oxidase- -52- promoted LOOHeindependent initiation lipid peroxidation. Again, addition of EDTA-Fe+3 enhances peroxidation without changing the charac- teristics of the system (Table 6). This demonstrates that LOOH- independent initiation occurs through a superoxide dismutase-sensitive reaction and that 102 is not involved in LOOH-independent initiation. The requirement for ADP-Fe"3 and the effects of superoxide dismutase and DPF form a common denominator between the two systems. ADP-Fe+3 appears to function in like manner in both systems, even though the enzymes are quite different. Xanthine oxidase produces 02’ and NADPH-cytochrome P450 reductase participates in electron transport. Both systems are capable of the reduction of ADP-Fe+3 and it was shown that ADP-Fe+2 can promote LOOH-independent initiation lipid peroxidation, Table 7. Investigation of peroxidation promoted by ADP-Fe+2 showed that it closely parallels that promoted by both xanthine oxidase and NADPH- cytochrome P450 reductase. Inhibition by superoxide dismutase was nearly complete while DPF exhibited only partial inhibition. The LOOH-independent initiation mechanism appears to be common for these three systems and is dependent upon a superoxide dismutase- sensitive intermediate. The ability of ADP-Fe+2 alone to promote LOOH- independent initiation of lipid peroxidation indicates its direct par- ticipation in the initiation reaction. It has been proposed that LOOH- independent initiation of lipid peroxidation in the NADPH-cytochrome P450 reductase system occurs through an ADP-perferryl ion, ADP-Fe+2-02 [58]. The perferryl ion complex is generated by the reduction of ADP-Fe+3 to ADP-Fe+2 by NADPH-cytochrome P450 reductase and subsequent interaction with molecular oxygen. -53- The same initiation complex may be involved in xanthine oxidase- promoted LOOH-independent initiation. It was shown that 102 is not involved in the LOOH-independent initiation of peroxidation, Table l. The requirement for ADP-iron in xanthine oxidase-dependent lipid peroxi- dation [62] and the fact that 02' cannot directly react with polyunsatur- ated fatty acids leads to the proposal that the active initiation complex is formed by the reaction of 027 with ADP-Fe+3. The interaction of 02' with ADP-Fe+3 perhaps gives rise to the ADP-perferryl ion, the proposed promoter of LOOH-independent initation in NADPH-dependent peroxidation. The initiation complex could be produced via 02' during xanthine oxidation by xanthine oxidase by the scheme: 02* + ADP-Fe+3 (ADP-Fe+3-02"') (ADP-Fe+2-02) (32) This initiation complex would account for the requirement for ADP-Fe+3 in the xanthine oxidase and the NADPH-cytochrome P450 reductase systems and for the sensitivity of both to superoxide-dismutase. Dismutase activity with iron bound 02' has been proposed by Richter et al. [144]. Singlet oxygen is not involved in the LOOH-independent initiation of lipid peroxidation but is likely involved in the LOOH-dependent ini- tiation of lipid peroxidation based on the inhibition of peroxidation by DPF. During LOOH-dependent initiation 102 may be produced by the break- down of the initially formed lipid hydroperoxides. Howco et al. [143] have reported that the breakdown of linoleic acid hydroperoxides by hemeproteins involves the production of 102 in substantial quantities. Singlet oxygen is produced in the breakdown of LOOH as evidenced by the decreased formation of TBArreactive material from partially peroxidized lipids in the presence of DPF (Table 9). For EDTArFe+2, which is only -54- capable of promoting LOOH-dependent initiation, production of TBA- reactive material from partially peroxidized lipids was inhibited by 20% in the presence of DPF. Lipid hydroperoxide-dependent initiation of lipid peroxidation promoted by EDTA—Fe+2 was also slightly inhibited by superoxide dismutase as evidenced by a decrease in LOOH when superoxide dismutase is added. This indicates that 027 may play some role in LOOH- dependent initiation. That ADP—Fe+2 can promote LOOH-independent ini- tiation as well as LOOHedependent initiation is evidenced by the differ- ence in the formation of TBArreactive material from lipid hydroperoxides in the presence of superoxide dismutase or DPF (Table 9). The presence of DPF, which does not inhibit LOOH-independent initiation, does not affect the production of TBA-reactive material. Superoxide dismutase does inhibit LOOH-independent initiation and allows the determination of the LOOH-dependent initiation promoting abilities of ADP-Fe+2 (Table 9). This suggests that inhibition of LOOH-independent initiation by super- oxide dismutase occurs by the breakdown of the initiation complex (the ADP-perferryl ion), leaving ADP-Fe"'2 which can promote LOOH-dependent initiation. These findings are summarized in Figure 2. Lipid hydroperoxide independent initiation takes place through a common initiation complex, the ADP-perferryl ion. The initiation complex may be producing by complexing 02* with ADP-Fe+3 or by the reduction of ADP-Fe+3 to ADP-Fe+2 and subsequent interaction with oxygen. The initiation complex produces low levels of lipid hydroperoxides and some TBArreactive material upon reaction with unsaturated lipid. The initiation complex is sensitive to superoxide dismutase but not DPF. Lipid hydroperoxide-dependent -55- initiation involves the breakdown of the initial hydroperoxides into reactive intermediates of peroxidation. Since 102 is only one of several intermediates less than 50% inhibition by singlet oxygen trapping agents is to be expected. —56- .moHonm> wfiaan HmEomouoa: N wouomuuxw mo coaumwfixoumm mo cowuoEoum ucmwcmamal. o mam ucowcoaoolmma LOOH i DPF -58- CHAPTER 2 HYDROXYL RADICAL-DEPENDENT INITIATION OF LIPID PEROXIDATION SUMMARY Fenton's reagent, aqueous ferrous ion and H202, can promote the peroxidation of liposomes prepared from extracted microsomal lipid. Conditions for optimal rates of peroxidation were found to be an iron to H202 ratio of 2 to l and a reaction mixture pH of 7. Lipid peroxidation promoted by the Fenton's reagent was inhibited by mannitol, benzoate and thiourea. However, the observed inhibition may not directly demonstrate the participation of the hydroxyl radical in the promotion of lipid peroxidation since the addition of several buffers and metal chelators inhibited Fenton's reagent promoted lipid peroxidation to a similar extent. It appears that H0°-dependent initiation occurs only under well defined reaction conditions where there are no metal chelators or other complexeing substances. INTRODUCTION The hydroxyl radical has been proposed to initiate lipid peroxida- tion by the abstraction of methylene hydrogen from polyunsaturated fatty acids. The involvement of HO° in the initiation of enzyme-promoted lipid peroxidation has been suggested by several researchers. McCay et al. [39,48,49,52,72] and Piette et al. [56,57,76] have proposed that HO' initiates NADPHrdependent superoxide (02') mediated peroxidation of microsomal lipid. Fridovich et al. [136,145] have also proposed that HO° is involved in the initiation of OZT-dependent lipid peroxidation promoted by xanthine oxidase. Proposals that H0' is involved in lipid peroxidation is usually supported by data indicating that peroxidation -59- involves the sequential generation of 027, H202 and HO°. The sequence of intermediate generation has been demonstrated by the ability of H0° traps, catalase and superoxide dismutase to sequentially inhibit lipid peroxidation. There are two reaction mechanisms that are generally proposed as the means of H0° generation in biological systems. First, Fridovich et al. [136,145] have proposed that HO- is generated by an uncatalyzed Haber- Weiss reaction. 2082' + 211* ) H202 (9) 02' + H202 4:) OH‘ + 02 + H0° (12) However, current research indicates that the uncatalyzed Haber-Weiss reaction does not occur to any significant extent under biological con- ditions [66-71]. The second and perhaps the more feasible mechanism for the formation of H0° in a biological system is a reaction mechanism that may best be described as an iron—promoted Haber-Weiss reaction proposed by McCay et al. [39,48,49,52,72] and Piette et al. [56,57,76] for the H0° initiation of lipid peroxidation. 202? + 211* a) H202 + 02 (9) Fe” + H202 _ :y Fe+3 + OH“ + 110- (10) Both free iron and chelated iron have been proposed to promote the Haber4Weiss reaction. Recent experimental work by Walling et al. [114] and Groves et a1. [115-117] has yielded some valuable information about the course of the reaction as written above. Groves et a1. and Walling et a1. question the existence of the free H0° when generated by iron-promoted H202 breakdown. These authors have shown that the free H0° only exists -60- under acid conditions and in "pure" solution. When the pH of the solu- tion is raised to physiological pH or the solution deviates from pure water the free H0. is no longer present. Instead the oxidative moiety formed in ferrous ion promoted H202 breakdown appears to be a ferric ion hydroxyl radical complex. This ferric ion hydroxyl radical complex has been termed the ferryl ion based on the formal oxidation state of the iron (Fe+4). The ferryl ion is a powerful oxidant and the reactivity attributed to the H0: can also be attributed to the ferryl ion. The presence of the ferryl ion is indicated by data showing directive effects observed during ring hydroxylation. These directive effects are the antithesis of free H0° reactions. Thus in a biochemical system these authors would favor a proposal that initiation occurs via the ferryl ion. Indeed the initiation of oxidative free radical reactions by the ferryl ion has been proposed for many well known reactions [87,88]. The ferryl ion can also be formed from the perferryl ion as outlined below in the scheme for the autoxidation of ferrous ion [23,24, 122,123]. Fe+2 + 02 1% Fe+202 (22) Fe+202 + Fe+2 —— a Fe+202Fe+2 (33) Fe+202Fe+2 ; 2Fe+20 (34) Considering that Aust et al. [75] have proposed that both NADPH—depen- dent and 02‘-dependent lipid peroxidation are initiated by the perferryl ion and not H0°, equations 22,33,34 raise an interesting paradox. Are the perferryl ion, the ferryl ion and the H0° part of the same continuem of reactive intermediates and as such interchangeable. If this is so, then is experimental data supportive of one not supportive of all. Also, if the data is supportive of only one of the intermediates may not such -61- support be reflective of the experimental viewpoint rather than actual experimental fact. In this chapter the ability of Fenton's Reagent to promote the ini- tiation of lipid peroxidation and the effect of varying experimental parameters such as concentration of reactants and chelation of iron on lipid peroxidation will be examined. MATERIALS AND METHODS Chemicals: Catalase, Type I, was purchased as a lyopholized powder from Sigma Chemical Company. As a lyopholized powder the catalase con- tained no antioxidant preservations and required no further purifica- tion. Chelex 100, a cationic exchange resin, was purchased from Pharmacia Chemical Company. All other chemicals used were of analytical grade and were used without further purification. Microsomal lipid and liposomes: Total extraction of microsomal lipids was carried out by the method of Folch et al. [97]. Procedural precautions taken to minimize autoxidation of the lipid during handling and storage have been previously described [27]. Liposomes were pre- pared as described previously [55]. Aqueous solutions were prepared using distilled-deionized water that had been passed over a Chelex 100 column (50 x 2.5 cm). Distilled-deionized water was prepared from distilled water by passing it through a mixed-bed ion exchange resin (Boeringher Company). All buffer solutions were passed over a Chelex column a second time after pH adjustment. The pH of aqueous and buf- fered solutions was adjusted using 1M HCl and NaOH solutions that were passed over a Chelex 100 column after preparation. All aqueous solu- tions were shown to be free of contaminating iron [146]. -62- Reaction mixtures in which lipid peroxidation was promoted by Fenton's reagent contained 1.0 pmole lipid phosphate per ml (in the form of liposomes) and H202 and FeClz in the concentrations indicated in the appropriate figures and tables. Reaction mixtures were prepared using chelexed—distilled deionized 0.3 M NaCl at the pH indicated in the appropriate figures and tables. Reactions were initiated by the addition of FeClz. Other additions to the reaction mixtures were as indicated in the figures and tables. Incubations were carried out at 37°C under an air atmosphere in a Dubnoff metabolic shaking bath. Lipid peroxidation was measured by the formation of TBA-reactive material, malondialdehyde (MDA), and lipid hydroperoxides by the method of Buege and Aust [33]. Rates of MDA and lipid hydroperoxide formation reported are initial rates. Other methods: ADP and EDTArchelated ferric ion solutions were pre- pared in chelexed-distilled deionized water. The molar ratios of the chelator to ferric ion are indicated in the appropriate tables. The pH of the iron complex solutions was adjusted to pH 7.5 by methods described for aqueous solutions above. Ferrous ion solutions, whether free or chelexed, were prepared as above except the chelexed-distilled deionized water was degassed and purged with argon prior to use. This method of preparation minimized ferrous ion autoxidation prior to addi- tion to an air saturated reaction mixture. Total lipid phosphate was measured by the method of Bartlett et al. [141]. RESULTS Fenton's reagent can promote the peroxidation of liposomes prepared from extracted microsomal lipid (Fig. 3). The rate of peroxidation as measured by MDA formation is greater than that observed for either -63- :.mwo:umz: nova: oonwuumov mm onuowuma mums mzmmmm cam mcoaumnsocH .o.n ma.aomz z m.o ca Naoom SE ~.o cam Noam ZS H.o .He won oumnmmozm mace: o.H wocfimucoo monouxfie coauommm I coaumvfixouom VHQHA mouoEoumiuamwmom m.coucom mo mmusoo oEHH .m ouswfim + Fe 2 p — l5 iO 8.0 i 7.0 “ r 0. 6 O 0 5 4 ._E\oH mwfixouoaouw%m .AAVV maven ES N.o van 1 meme-=0 mum-cc:- i—i O 5 20 I IO Time (min) -89- The ability of these iron chelates to promote lipid peroxidation from LOOH was assessed by their addition to liposomes containing LOOH and measuring both MDA and LOOH formation (Table 12). Only EDTA-Fe+2 can promote LOOH—dependent initiation of lipid peroxidation. The initial rate of MDA formation, 1.16 nmoles/min/ml, reflects significant peroxi- dation since only 10% or less of total peroxidation is reflected by MDA formation [31,153]. The rate of LOOH formation, 9.3 nmoles/min/ml, also reflects significant peroxidation. It must be realized that during LOOH-dependent initiation LOOH are both reactant and product. For example, if LOOH breakdown occurs at a rate similar to CHP breakdown in the presence of EDTArFe+2, 7.0 nmoles/min/ml, an increase in LOOH con- tent of 9.3 nmoles/min/ml actually reflects a rate of LOOH formation of 16.3 nmoles/min/ml. The ability of ADP-Fe+2 to promote LOOHedependent initiation of lipid peroxidation is considerably less than that apparent in Table 12 since ADP--Fe+2 can also promote perferryl ion-dependent lipid peroxida- tion (Chapter 1). ADP-Fe+2 can promote LOOH-independent initiation of lipid peroxidation to the extent of 0.6 nmoles MDA/min/ml and 1.8 nmoles LOOH/min/ml (Chapter 1). If these rates of perferryl ion-dependent ini- tiation are subtracted from the values in Table l for ADP-Fe+2 promo- tion of LOOH-dependent initiation it is apparent that ADP-Fe+2 promo- tion of LOOHedependent initiation is minimal compared to EDTA-Fe+2 pro- motion of LOOHédependent initiation. The nature of the products formed during LOOHrdependent initiation is indicated by the results presented in Table 13. It appears that 02' may be a minor reactive intermediate formed during LOOH-dependent initiation as is reflected by the superoxide dismutase inhibition of -90- TABLE 12 THE PROMOTION OF LOOHFDEPENDENT INITIATION OF LIPID PEROXIDATION BY IRON CHELATES Reaction mixtures contained 1.0 umole lipid phosphate/ml (containing 0.1 umole LOOH/ml) in 0.05 M Tris-HCl, pH 7.5 at 37°C. The following additions were made to the reaction mixtures as indicated: 1.7 mM ADP and 0.1 mM FeCl3; 1.7 mM ADP and 0.1 mM FeClz; 0.11 mM EDTA and 0.1 mM FeCl3; 0.11 mM EDTA and 0.1 mM FeClz. Reactions were initiated by addition of the appropriate iron complex. Incuba- tions and assays were performed as described under ”Methods." nmoles MDA nmoles LOOH min ml"1 min m1“1 Control (No Additions) 0.01 0.7 +ADP-Fe+3 0.02 1.1 +EDTArFe+3 0.01 1.1 +ADP-Fe+2 0.60 1.4 +EDTArFe+2 1.16 9.3 -91- TABLE 13 THE PROMOTION OF LOOHrDEPENDENT INITIATION OF LIPID PEROXIDATION BY EDTArFe+2 Reaction mixtures contained 1.0 umole lipid phosphate/ml (containing 0.1 umole LOOH/umole lipid phosphate) in 0.05 M Tris-HCl, pH 7.5 at 37°C. The following additions were made as indicated: 0.11 mM EDTA and 0.1 mM FeCl3; 0.11 mM EDTA and 0.1 mM FeClz; 1.0 unit SOD/ml; 0.2 mM DPF; 1.0 mM BHT and 40 mM benzoate. Reactions were initiated by addition of the appropriate iron form. Incubations and reactions were performed as described under ”Methods.” nmoles MDA nmoles LOOH min ml"1 min ml"1 Control (No Additions) 0.01 0.7 +EDTArFe+3 0.16 1.3 +EDTArFe+2 1.31 10.2 +EDTArFe+2 +SOD 1.28 7.1 +EDTArFe+2 +DPF 1.16 7.2 +EDTArFe+2 +BHT 0.16 1.2 +EDTA-Fe+2 +Benzoate 1.26 10.3 -92- LOOH-dependent initiation. Superoxide dismutase inhibits MDA formation by 2% while inhibiting LOOH formation by 30%. Singlet oxygen may also be formed during LOOH-dependent initiation, as evidenced by the DPF inhibition of MDA formation of 11% and LOOH formation by 29%. (While the ability of DPF to act as a free radical scavenger clouds these results to a certain extent we have been unable to show that DPF inhi- bits ascorbate dependent peroxidation, a totally free radical mechanism, by more than 1% at the concentrations used for the experiments reported here.) The free radical nature of the LOOH-dependent initiation reac- tions are demonstrated by the inhibition of both MDA and LOOH formation by BHT. Addition of BHT inhibits both MDA and LOOH formation by 88%. The hydroxyl radical apparently does not participate in the LOOH- dependent initiation of lipid peroxidation as evidenced by the lack of inhibition of the reaction upon the addition of 40 mM benzoate (Table 13). Experiments with mannitol, gave similar results. These results are in agreement with the work of Tyler (64). The ability of EDTA-Fe+3 to promote LOOH-dependent initiation of lipid peroxidation when reduced by NADPH-cytochrome P450 reductase is demonstrated by the data presented in Table 14. Peroxidation, as detected by both MDA and LOOH formation, is significant only in the pre- sence of NADPH and EDTArFe+3. The addition of NADPH gives rise to the enzymatic formation of EDTArFe+2. The ability of NADPH-cytochrome P450 reductase to reduce EDTA—Fe+3 was demonstrated in a reaction mixture containing the reductase, NADPH and EDTA-Fe+3. NADPH oxidation occurred only in the presence of all three components. These results were substantiated by ferrous ion chromophore formation when an excess of bathophenanthroline was added to an anaerobic reaction mixture after a -93- TABLE 14 PROMOTION OF LOOH-DEPENDENT INITIATION OF NADPH-DEPENDENT LIPID PEROXIDATION Reaction mixtures contained 1.0 umole lipid phosphate/ml (con- taining 0.1 umole LOOH/umole lipid phosphate), 0.1 unit NADPH cytochrome P450 reductase/ml, 0.11 mM EDTA, 0.1 mM FeCl3 and 0.1 mM NADPH in 0.05 Tris-HCl, pH 7.5 at 37°C. The following addi- tions were made as indicated: 1.0 unit SOD/ml, 0.2 mM DPF, 40 mM benzoate and 1.0 mM BHT. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as described under ”Methods.” nmoles MDA nmoles LOOH min ml'1 min m1-1 Control (No Additions) 0.01 0.01 +NADPH 1.11 5.5 +NADPH +SOD 1.03 4.7 +NADPH +DPF 0.95 3.9 +NADPH +BHT 0.00 0.0 +NADPH +Benzoate 1.12 6.3 -94- five minute reaction time. It is apparent that the function of EDTA-Fe+3 in NADPH-dependent lipid peroxidation is to promote, in its reduced form, LOOH-dependent initiation of lipid peroxidation. Investigation into the nature of the reactive intermediates formed during enzymatic promotion of LOOHrdependent initiation (Table 14) yielded results equivalent to those obtained in the nonenzymatic reac- tion system (Table 13). The addition of superoxide dismutase inhibited MDA formation by 8% and LOOH formation by 15%. The addition of DPF inhibited MDA and LOOH formation by 13% and 28%, respectively. Again, BHT showed essentially complete inhibition of LOOHrdependent initiation. Benzoate did not inhibit NADPH-promoted LOOH-dependent initiation of lipid peroxidation, but actually enhanced peroxidation as detected by LOOH formation. It appears that the enzyme promoted LOOHrdependent ini- tiation reaction gives rise to radical intermediates of lipid peroxida- tion, however, H0; is not among them. In addition, 102 is apparently formed during LOOHrdependent initiation of lipid peroxidation. Lipid hydroperoxide-dependent initiation accounts for a significant portion of the MDA and LOOH formed during NADPH-dependent liposomal peroxidation (Table 15). In th presence of ADP-iron alone, the rate of MDA and LOOH formation in NADPH-dependent lipid peroxidation is 0.3 nmole/min/ml and 1.8 nmole/min/ml, respectively. When a promoter of LOOH—dependnt initiation, EDTArFe+3, is included, MDA and LOOH formation are increased ll-fold. NADPH-dependent lipid peroxidation in the reconstituted system, shown in Table 15, is characterized by the simulataneous occurrence of both LOOHrindependent initiation and LOOH-dependent initiation re- actions. Superoxide dismutase inhibits both MDA and LOOH formation by -95- TABLE 15 NADPH-DEPENDENT LIPOSOMAL PEROXIDATION Reaction mixtures contained 1.0 umole lipid phosphate/ml, 0.1 unit NADPH-cytochrome P450 reductase/ml, 1.7 mM ADP, 0.1 mM FeCl3 and 0.1 mM NADPH in 0.05 M Tris-H01, pH 7.5, at 37°C. The following additions were made where indicated: 0.11 mM EDTA and 0.1 mM FeCl3, 1.0 unit SOD/ml, 0.2 mM DPF, 1.0 mM BHT and 40 mM benzoate where indicated. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as under ”Methods.” nmoles MDA nmoles LOOH min ml‘1 min ml‘l Control (No Additions) 0.03 0.1 +NADPH 0.30 1.8 +EDTArFe+3 0.02 0.02 +NADPH +EDTArFe+3 3.40 19.5 +NADPH +EDTArFe+3 +300 0.50 3.5 +NADPH +EDTArFe+3 +DPF 3.10 16.3 +NADPH +EDTArFe+3 +BHT 0.00 0.0 +NADPH +EDTA-Fe+3 +Benzoate 3.45 20.1 -96- approximately 85% in a reconstituted reaction mixture promoting both forms of initiation. In contrast, in a similar system promoting only LOOH-dependent initiation (Table 14), superoxide dismutase inhibits MDA and LOOH formation by only 8% and 15%, respectively. Such a difference between the two reaction mixtures serves to emphasize the dependence of LOOH-dependent initation on perferryl ion-dependent LOOH-independent initiation. The addition of DPF to an NADPH-dependent reaction mixture promoting both forms of initiation (Table 15) inhibits MDA and LOOH for- mation by 9% and 16%, respectively. The inhibition observed is equiva- lent to that observed during the NADPH-dependent promotion of LOOH- dependent initiation alone (Table 14). These results show that 102 participates in only LOOH-dependent initiation and not in perferryl ion- dependent LOOH-independent initiation. The addition of BHT completely inhibits both MDA and LOOH formation demonstrating the radical nature of perferryl ion-dependent initiation. The radical nature of the LOOH- dependent initiation reaction was demonstrated in Table 14. Similar to the results in Table 14, no effect on the rate of lipid peroxidation upon the addition of benzoate was observed (Table 15) indicating that HO' does not participate in either form of initiation in NADPH-dependent liposomal peroxidation. The iron chelate diethylenetriamine pentaacetic acid (DPTA) has been used by some [154] to inhibit metal catalyzed lipid peroxidation. Considering that the structure of DTPA is similar to that of a dimer of EDTA, the ability of DPTA.-Fe"’3 to replace EDTA-Fe+3 in NADPH-dependent lipid peroxidation was examined (Table 16). The results demonstrated that DTPA-Fe+3 could effectively replace EDTArFe+3. Addition of DTPA- Fe+3 in the absence of ADP-Fe+3 showed that DTPA-Fe+3 could not function -97- TABLE 16 PROMOTION OF LOOHrDEPENDENT INITIATION OF NADPH-DEPENDENT LIPID PEROXIDATION BY EDTArFe+3 AND DTPA-Fe+3 Reaction mixtures contained 1.0 umole lipid phosphate/ml, 0.1 unit NADPH cytochrome P450 reductase/ml and 0.1 mM NADPH in 0.05 M Tris-HCl, pH 7.5 at 37°C. The following additions were made where indicated: 0.11 mM EDTA and 0.1 mM FeCl3; 0.11 mM DTPA and 0.1 mM FeCl3; 1.7 mM ADP and 0.1 mM FeCl3. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as described under ”Methods." nmoles MDA nmoles LOOH min ml'1 min ml‘l Control (No Additions) 0.01 0.2 +EDTArFe+3 0.01 0.3 +DTPA-Fe+3 0.01 0.2 +NADPH 0.02 0.4 +NADPH +EDTArFe+3 0.01 0.4 +NADPH +DTPA-Fe+3 0.01 0.2 +NADPH +ADP-Fe+3 0.35 1.9 +NADPH +ADP-Fe+3 3.46 19.6 +EDTArFe+3 +NADPH +ADP-Fe+3 3.19 18.9 +DTPA-Fe+3 -98- in the LOOHrindependent initiation of NADPH-dependent lipid peroxida- tion. The results indicate DTPA-Fe+3 functions as a promoter of LOOH-dependent initiation of lipid peroxidation. These findings may explain the kinetic findings of Thomas et al. [154] since they used DTPA to chelate contaminating iron in a system that is capable of reducing DTPArFe+3 to DTPArFe+2 via 02*. The involvement of cytochrome P450 in microsomal lipid peroxidation was investigated utilizing aminopyrine and SKF 525-A to inhibit NADPH- dependent microsomal lipid peroxidation. Neither compound nor their metabolic products were found to be antioxidants at the concentration used as determined by their inability to inhibit ascorbate dependent lipid peroxidation. The addition of 100 uM SKF 525€A to microsomes inhibited NADPH- dependent lipid peroxidation by 67% (Table 17). SKF SZSrA inhibits cytochrome P450 catalyzed reactions by preferentially binding to cytochrome P450 displacing other substrates [155]. The addition of EDTA-Fe+3, which has been shown to enhance NADPH-dependent microsomal lipid peroxidation [55,63], completely reversed the SKF 525-A inhibition of NADPH-dependent microsomal lipid peroxidation, indicating that SKF 525€A is specifically inhibiting an endogenous microsomal agent capable of promoting LOOHrdependent initiation. The specificity of SKF 525-A indicates that one of the endogenous promoters of LOOH-dependent ini- tiation in microsomes is cytochrome P450. Aminopyrine was proposed to inhibit NADPH-dependent microsomal lipid peroxidation by competing for reducing equivalents [156,157]. The addition of 5 mM aminopyrine to a microsomal reaction mixture inhibited NADPH-dependent lipid peroxidation by 57% (Table 18). The addition of -99- TABLE 17 THE EFFECT OF EDTA-Fe+3 ON THE SKF 525-A INHIBITION OF NADPH-DEPENDENT MICROSOMAL LIPID PEROXIDATION Reaction mixtures contained 0.5 mg microsomal protein/ml, 1.7 mM ADP, 0.1 mM FeCl3 and 0.1 mM NADPH in 0.05 M Tris-HCl, pH 7.5 at 37°C. The following additions were made as indicated: 100 pM SKF 525-A, 0.11 mM EDTA and 0.1 mM FeC13. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as described under "Methods.” nmoles MDA min ml'1 Control (-NADPH) 0.11 +NADPH 2.43 +NADPH +SKF 525eA 0.81 +NADPH +SKF 525-A 2.24 +E DTA-Fe+3 -100- TABLE 18 THE EFFECT OF EDTA-Fe+3 ON THE AMINOPYRINE INHIBITION OF NADPH- DEPENDENT MICROSOMAL LIPID PEROXIDATION Reaction mixtures contained 0.5 mg microsomal protein/ml, 1.7 mM ADP, 0.1 mM FeCl3 and 0.1 mM NADPH in 0.05 M Tris-HCl, pH 7.5 at 37°C. The following additions were made as indicated: 5.0 mM aminopyrine, 0.11 mM EDTA and 0.1 mM FeCl3. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as described under "Methods." nmoles MDA min ml"1 Control (-NADPH) 0.20 +NADPH 2.81 +NADPH +Aminopyrine 1.20 +NADPH +Aminopyrine 2.81 +E DTA-Fe+3 -101- EDTA-Fe+3 completely reversed aminopyrine inhibition. Since EDTA-Fe+3 must be reduced to be active in the promotion of LOOHrdependent ini- tiation of lipid peroxidation, drug substrate inhibition of NADPH- dependent microsomal lipid peroxidation apparently does not occur by competition for reducing equivalents. It would appear that drug substrates inhibit NADPH-dependent lipid peroxidation by interaction with cytochrome P450. Inhibition may be the result of cytochrome P450 peroxidase activity utilizing lipid hydroperoxides and oxidizable drugs as substrates in a manner analagous to the CHP dependent drug metabolism observed by others [158-160]. This data suggests that the endogenous promoter of LOOH-dependent initiation is cytochrome P450 as previously indicated by experiments with SKF 525-A. The ability of ferric cytochrome P450 to promote LOOH-dependent initiation of lipid peroxidation was investigated by addition of cytochrome P450 to a reaction mixture in which LOOH were generated 12 gi£2_by soybean lipoxygenase (Table 19). In detergent treated liposomes, lipoxygenase catalyzed initial rates of formation of 0.08 nmoles MDA/min/ml and 0.53 nmoles LOOH/min/ml. The addition of 0.3 nmol/m1 ferric cytochrome P450 to the reaction mixture resulted in an ll-fold increase in the rate of MDA formation and a 3-fold increase in the rate of LOOH formation. From this data it appears that cytochrome P450 is an excellent promoter of LOOH-dependent initiation. The data indicates that on a per mole basis cytochrome P450 is a beter promoter of LOOHrdependent initiation than is EDTArFe+2. However, as others have previously shown, cytochrome P450 is degraded during lipid peroxidation thereby limiting its promotional abilities [161,162]. -102- TABLE 19 PROMOTION OF LOOHFDEPENDENT INITIATION 0F LIPID PEROXIDATION BY FERRIC CYTOCHROME P450 Reaction mixtures contained 1.0 umole lipid phosphate/ml, 100 ug lipoxygenase/ml and 0.04% sodium deoxycholate in 0.05 M Tris-HCl, pH 7.5 at 37°C. Ferric cytochrome P450, 0.3 nmole/ml, was added where in- dicated. Reactions were initiated by the addition of lipoxygenase. Incubations and assays were performed as decribed under ”Methods.” nmoles MDA nmoles LOOH min ml“1 min ml‘l Control (-Lipoxygenase) 0.02 0.1 +ferric cytochrome P450 0.05 0.3 +lipoxygenase 0.08 0.5 +lipoxygenase + ferric cytochrome P450 0.88 1.4 -103- The complete reconstitution of NADPH-dependent microsomal lipid peroxidation in liposomes utilizing ferric cytochrome P450 as the pro- moter of LOOH-dependent initiation is shown in Table 20. In the absence of cytochrome P450, the ability of ADP-Fe+2 to initiate lipid peroxida- tion is evident (Chapter 1). The addition of 0.3 nmoles/ml cytochrome P450 as a promoter of LOOH-dependent initiation increases MDA and LOOH rates of formation by 8- and l6-fold, respectively. These results indi- cate the key role played by cytochrome P450 in NADPH-dependent lipid peroxidation. DISCUSSION The results presented here along with those previously reported (Chapter 1), suggest a mechanism of NADPHrdependent lipid peroxidation. Previously, it was shown that lipid peroxidation can be divided into two sequential series of radical reactions (Chapter 1). The first reaction is the ADP-perferryl ion promoted formation of LOOH. Perferryl ion- dependent initiation products were proposed to subsequentially undergo reductive activation reactions with EDTA-Fe+2 to generate radical inter- mediates of lipid peroxidation. Promotion of LOOHrdependent initiation resulted in the rapid formation of MDA and LOOH as is typical of iron— dependent enzymatic lipid peroxidation. In this chapter, it was demonstrated that LOOH-dependent initiation is promoted by EDTArFe+2, DTPA-Fe+2 or ferric cytochrome P450. The promotion of LOOH-dependent initiation is responsible for up to 90% of the peroxidation observed. NADPHrcytochrome P450 reductase promoted liposomal peroxidation requires both ADP-Fe+3 and EDTA-Fe+3 for maximum activity. In the enzy- matic reaction both the ferric and the ferrous forms of the iron chela- tes exists. It was previously established that only ADP-Fe+2 can -104- TABLE 20 PROMOTION OF LOOHrDEPENDENT INITIATION OF NADPH-DEPENDENT LIPOSOMAL PEROXIDATION BY FERRIC CYTOCHROME P450 Reaction mixtures contained 1.0 umole lipid phosphate/m1, 0.1 unit NADPH-cytochrome P450 reductase/ml, 0.1 mM NADPH, 1.7 mM ADP and 0.1 mM FeCl3 in 0.05 M Tris-HCl, pH 7.5 at 37°C. Ferric cytochrome P450, 0.3 nmole/ml, was added where indicated. Reactions were initiated by the addition of NADPH. Incubations and assays were performed as described under ”Methods." nmoles MDA nmoles LOOH min ml"1 min ml"1 Control (-NADPH) 0.01 0.1 +ferric cytochrome P450 0.04 0.1 +NADPH 0.21 0.4 +NADPH + ferric cytochrome P450 1.51 6.6 -105- initiate the formation of LOOH (Chapter 1). In LOOHrdependent ini- tiation the promoter must be able to efficiently promote the breakdown of LOOH to reactive intermediates of lipid peroxidation. Using CHP as a model organic hydroperoxide to assay promotion by the four possible iron complexes, it was apparent that only EDTA-Fe+2 could promote the effi- cient breakdown of the organic hydroperoxide (Figure 7). Thus, although EDTA-Fe+2 could not function in the perferryl ion promoted LOOH-inde- pendent initiation of lipid peroxidation it can promote LOOH-dependent initiation of lipid peroxidation. The ability of EDTA-Fe+2 to promote LOOH-dependent initiation of lipid peroxidation in partially peroxidized liposomes was directly demonstrated in Tables 12 and 13. Of the four iron chelates investi- gated in Figure 7, only EDTA—Fe+2 could efficiently promote the for- mation of MDA and additional LOOH from LOOH in a lipid matrix. Although the formation of MDA and LOOH upon addition of ADP-Fe+2 appears to be significant, it must be recognized that ADP-Fe+2 can promote perferryl ion-dependent initiation of lipid peroxidation in unperoxidized fatty acid (Chapter 1). Considering the ability of ADP-Fe+2 to promote per— ferryl ion-dependent initiation of lipid peroxidation at rates similar to those observed in Table 12, promotion of LOOHrdependent initiation by ADP-Fe+2 is apparently minimal compared to that of EDTA-Fe'l’2 which can- not promote perferryl ion-dependent initiation of lipid peroxidation. The findings reported here indicate that only EDTA-Fe+2 is efficient in the promotion of LOOH-dependent initiation of lipid peroxidation. The results presented in Table 13 showed that EDTA-Fe+2 promoted LOOHrdependent initiation was not mediated by a superoxide dismutase sensitive complex or 027, in contrast to ADP—Fe+2 promoted initiation. -lO6- Thus, the LOOH-dependent initiation reaction is apparently not promoted by an iron bound 02' since such a complex would likely be superoxide dismutase sensitive [144]. The LOOHedependent initiation reaction is likely promoted by EDTArFe+2. The minor inhibition by superoxide dismu- tase may reflect a small amount of 02* which is produced by air oxida- tion of the EDTA-Pei2 complex according to the following reaction: EDTA-Fe+2 + 02 EDTA-Fe+3 + 021' (37) Addition of superoxide dismutase would then shift the reaction equilibrium to the right, effectively decreasing the concentration of promoter. Tables 14 and 15 also suggest that minor amounts of 102 may be produced during the LOOHrdependent initiation reaction, perhaps via the reaction of two lipid hydroperoxide radicals [29,163]. The minor inhibition exhibited by DPF indicates that 102 is a product of perhaps only some of several possible reactions. The inhibition of peroxidation upon the addition of BHT to the reaction mixture indicates the radical nature of the reaction. The lack of inhibition upon the addition of benzoate to the reaction mixture indicates that H0° is not one of the radicals that mediate LOOHrdependent initiation of lipid peroxidation. Using NADPH-cytochrome P450 reductase to promote the formation of EDTArFe+2'iEH§i£3 (Table 14), it was demonstrated that the enzymatic reaction was identical to the nonenzymatic reaction (Table 13). Enzymatically promoted LOOH—dependent initiation was superoxide dismu- tase insensitive, showed some inhibition by DPF, showed no inhibition by benzoate, but was completely inhibited by BHT. Thus, the overall enzyme promoted reaction, mediated by EDTArFe+2, was radical in nature, however, minor amounts of 102 may be formed and participate in further peroxidation. -107- The effect of various inhibitors of lipid peroxidation on the total reconstitution of NADPH-dependent lipid peroxidation (Table 15) are con- sistent with the mechanism of initiation previously suggested (Chapter 1) and the mechanism of LOOHedependent initiation being proposed here. Superoxide dismutase almost completely inhibits lipid peroxidation because it inhibits ADP-perferryl ion catalyzed initiation, as pre- viously shown (Chapter 1). It was previously shown that superoxide dismutase has minimal effect on the LOOHrdependent initiaton reaction alone (Tables 13 and 14). The slight inhibition of peroxidation by the addition of DPF, indicates that DPF does not inhibit perferryl ion- dependent initiation, but does inhibit LOOH-dependent initiation to a minor extent (Tables 13 and 14) indicating that 102 is a minor product of the reaction. Butylated hydroxytoluene completely inhibits NADPH- dependent lipid peroxidation indicating that perferryl ion-dependent initiation is radical in nature. The radical nature of LOOH-dependent initiation was shown in Tables 13 and 14. The lack of benzoate inhibi- tion in NADPH-dependent lipid peroxidation indicates that H0° does not participate in either form of initiation. This data clearly shows that NADPHrdependent lipid peroxidation occurs in two successive radical steps each dependent upon metal catalysis. The significant contribution of EDTA-Fe+2 promoted LOOH-dependent initiation to the total quantity of products formed during NADPH- dependent lipid peroxidation has been shown previously (Chapter 1) and is demonstrated by the data presented in Table 15. The addition of EDTA-Fe+3 to a NADPH-dependent liposomal peroxidation mixture increased the rate of MDA and LOOH formation by ll-fold over the rates in the pre- sence of ADP-Fe+3 and NADPH alone. Thus, it appears that LOOH-dependent -108— initiation is a key reaction in lipid peroxidation accounting for more than 90% of total product formation. Table 16 demonstrates that EDTA is not unique among the chelators of iron in its ability to facilitate iron promoted LOOH-dependent ini- tiation of lipid peroxidation. DTPA-Fe+3 can replace EDTA-Fe+3 in NADPH-dependent lipid peroxidation in liposomes. ADP-Fe+3 is an effi- cient promoter of perferryl ion-dependent LOOH-independent initiation of lipid peroxidation but a very poor promoter of the breakdown of organic hydroperoxides (Figure 7) or the LOOH-dependent initiation of lipid peroxidation (Table 12). Chelation of iron by EDTA greatly stabilizes the ferric complex, lowering the reduction potential of the complex to 0.254 v [164,165] making the ferrous complex a relatively strong reducing agent. Chelation of iron by ADP does not stabilize the ferric complex to as great a degree as does EDTA. Such an increase in the reducing ability of iron upon complexation by EDTA should enhance the promotion of LOOH breakdown. The effect of iron chelation on the break- down of H202 by Fenton type reagents has been previously demonstrated [166,167]. If LOOH-dependent initiation is a Fenton type reaction, efficient promotion by EDTA-Fe+2 as compared to free iron would be pre- dicted and is consistent with the experimental data. This proposed function of iron chelation is supported by the ability of DTPA-Fe+3 to replace EDTArFe+3. DTPA chelation of iron is similar to that of EDTA [168]. The data suggests that the key to the formation of an efficient promoter of LOOH-dependent initiation may be the lowering of the iron reduction potential upon chelation. The ability of DTPA-Fe+3 to promote LOOH-dependent initiation may offer an alternative explanation to the experimental data reported -109- recently by Thomas et al. [154]. Thomas et al., proposed that 02*- dependent lipid peroxidation occurred via the reaction of 02' with LOOH as follows: 021' + LOOH % L0- + 02 + 011‘ (38) The authors found that peroxidation was dependent on LOOH formed by autoxidation. The reaction constant for reaction 38 was found to be 7 x 103 M"1 sec-1. However, the rate of self-dismutation of 02' at the pH of the investigation, 7.4, is at least 2 orders of magnitude greater than the rate constant for reaction 38. Thus, the importance of reaction 38 in OZT-dependent lipid peroxidation may be minimal. The observation that peroxidation was dependent upon preformed LOOH, correlates well with the mechanism of iron promotion of LOOHrdependent initiation of lipid peroxidation proposed here. The kinetics of the observed reaction [155] also agree with the kinetics of EDTA-Fe+2 or DTPA-Fe+2 promoted LOOH-dependent initiation, as reported here. That is, the rate of LOOH formation during LOOH-dependent initiation, as defined here, is approxi- mately equal to the rates of peroxidation observed by Thomas et al. These two observations together with the use by Thomas et al. of DTPA to chelate metals (most likely iron) in the phosphate buffer used in their reaction mixture suggest that they may have observed a metal chelated promoted LOOH-dependent initiation of lipid peroxidation. The chelate involved likely would be DTPArFe+3, which is reduced by 02' to DTPA-Fe+2 an active promoter of LOOHrdependent initiation as demonstrated in this paper. Pederson et al. [55,58] proposed that EDTA-Fe+3 replaced an endoge- nous microsomal agent that participated in NADPH-dependent microsomal lipid peroxidation. Table 17 shows the ability of EDTA-Fe+3 to reverse -110- the SKF 5259A inhibition of NADPH-dependent microsomal lipid peroxida- tion. At the concentration used, SKF 525-A acts by binding to cytochrome P450 and not by disruption of microsomal electron transport [155]. The reversal of SKF 525-A inhibition by addition of EDTA- Fe+3 indicates that cytochrome P450 may be one microsomal entity respon- sible for promotion of LOOHrdependent initiation in NADPH-dependent microsomal lipid peroxidation. The proposed role of cytochrome P450 in microsomal lipid peroxida- tion is supported by the EDTA-Fe+3 reversal of aminopyrine inhibition of NADPH-dependent microsomal lipid peroxidation (Table 18). The data indicates that drug substrates of the microsomal mixed-function oxidase system do not inhibit NADPH-dependent lipid peroxidation by competing for reducing equivalents as others have proposed [156,157]. If the inhibitory effect was 3 results of the competition for electrons, addi- tion of EDTA-Fe+3, which must be reduced to its active form, should not completely reverse the observed inhibition. The observations that hydrogen donors such as tetramethyl-p-phenylenediamine are oxidized in the presence of P450 and lipid hydroperoxides [168,169] and that CHP can support enzymatic oxidation of microsomal mixed-function oxidase drug substrates [152,158-160] indicate that drug substrates may inhibit lipid peroxidation by reducing LOOH via a cytochrome P450 peroxidase type mechanism. Such a peroxidase reaction would inhibit lipid peroxidation by competing with LOOHrdependent initiation for LOOH produced during perferryl ion-dependent initiation. These results also indicate that cytochrome P450 may function in a variety of roles during lipid peroxi- dation in addition to its participation of LOOH-dependent initiation. -111- The ability of ferric cytochrome P450 to promote LOOH-dependent ini- tiation of lipd peroxidation is clearly demonstrated in Tables 19 and 20. Addition of 0.3 nmoles/ml ferric cytochrome P450 to a reaction mix- ture in which LOOH are being generated ig_§i£2_by soybean lipoxygenase, resulted in an 11-fold increase in MDA formation and a 3-fold increase in LOOH formation (Table 19). The results in Table 20 demonstrate the role that ferric cytochrome P450 plays in LOOH-dependent initiation of NADPH- dependent lipid peroxidation. In the reaction mixture, LOOH are generated by the ADP—perferryl ion initiation complex. Addition of 0.3 nmoles/ml ferric cytochrome P450, which cannot be reduced by the pro- tease solubilized NADPH—cytochrome P450 reductase [80] results in an 8-fold increase in MDA formation and 15-fold increase in LOOH formation. Thus, ferric cytochrome P450 is apparently an excellent promoter of LOOHrdependent initiation of lipid peroxidation. As summarized in Figure 8, the results presented in this paper clearly demonstrate that NADPH-dependent lipid peroxidation occurs in two sequential steps. The first step, perferryl ion-dependent initiation, is promoted by the ADP-perferryl ion. The second step, LOOH-dependent initiation, is dependent upon the LOOH formed during per- ferryl ion-dependent initiation, and results in the rapid formation of reactive intermediates and products of lipid peroxidation. Lipid hydroperoxide-dependent initiation is the EDTArFe+2, DTPA-Fe+2 or ferric cytochrome P450 promoted breakdown of LOOH to form reactive inter- mediates of lipid peroxidation. Lipid hydroperoxideqdependent ini- tiation accounts for more than 90% of the products formed during lipid peroxidation and is radical in nature. Lipid hydroperoxide-dependent initiation is perhaps a Fenton type reaction resulting primarily in the -112- formation of lipid alkoxy radicals. Low levels of hydroperoxy radicals are also possibly formed as indicated by the presence of 102. -113- .cowumexouom UHQHA acmvcmmmnImmm (ADP - re+2 - 02) LH HLl PID HYDROPEROXI DE- Xanthine DEPENDENT INITIATION 0 EDTA - Fe+2 Ferricytochrome P 2 450 x0 ' ' , 0‘ *3 2 EDTA - Fe Degraded Cytochrome P450 Urate SOD BHT LH L-,LO-, Loo—+—>‘02—I—A—>LDDH DPF BHT LH END PRODUCTS -l45— SUMMARY The results presented in this thesis demonstrate that NADPH- dependent and OZT-dependent peroxidation of microsomal phospholipid occur by similar if not the same mechanism. Lipid peroxidation is pro- posed to consist of two sequential steps. The first step, LOOH— independent initiation, involves the formation of low levels of LOOH in previously unperoxidized lipid. The second step, LOOH-dependent ini- tiation is the breakdown of initially formed LOOH generating reactive intermediates of lipid peroxidation. LOOH-independent initiation is promoted by an ADP-ferrous ion dependent reaction. Nonenzymatic initiation of lipid peroxidation pro- moted by ADP--Fe'*'2 is SOD sensitive. Since 02* cannot directly promote LOOH—independent initiation of lipid peroxidation, the possible involve- ment of reactive oxygen intermediates, 102 and HO° formed by the nonen- zymatic dismutation of 02' and subsequent reactions in initiation was investigated. The ADP-Fe+2 LOOH-independent initiation reaction was only slightly inhibited by addition of DPF indicating that 102 was not involved in the reaction to any great extent. The NADPHrdependent LOOH- independent initiation of lipid peroxidation was found to be dependent upon ADP-Fe+3 indicating that initiation likely occurred by an ADP-Fe+2 mediated reaction. The NADPH-dependent LOOH-independent initiation of lipid peroxidation demonstrated SOD sensitivity but little participation by 102. NADPHrdependent LOOH-independent initiation was found to be radical in nature but not to involve H0-. Thus the LOOH-independent initiation of NADPH-dependent lipid peroxidation did not require the nonenzymatic dismutation of 02' to give rise to 102 or further reactions of the dismutation products to generate HO° as has been proposed by ~146- others. The requirement for ADP-iron in the LOOH-independent initiation and the sensitivity of their initiation to SOD suggest that NADPH- dependent initiation of lipid peroxidation may occur via an iron bound 02' complex or the ADP-perferryl ion. Superoxide dismutase dismutation of iron bound 027 has been suggested by others [144]. Superoxide- dependent LOOH-independent initiation of lipid peroxidation is also SOD sensitive and radical in nature. The 027-dependent LOOH-independent initiation of lipid peroxidation does not involve 102 or HO°. Again, since 02' is not reactive enough to directly promote LOOH-independent initiation of lipid peroxidation [39,64,72] and because possible reac- tive intermediates formed during and subsequent to the nonenzymatic dismutation of 02' do not participate in LOOH-independent initiation of 02'-dependent lipid peroxidation, it is here proposed that the ADP- perferryl ion catalyzes the LOOH-independent initiation of OZV-dependent lipid peroxidation. Thus, it appears that LOOH-independent initiation of lipid peroxidation in both NADPH-dependent and 027-dependent lipid peroxidation may occur via the ADP-perferryl ion. In NADPH-dependent lipid peroxidation the ADP-perferryl ion is likely formed by direct reduction of the ferric chelate and its subsequent reaction with 027. In OZV-dependent lipid peroxidation the ADP-perferryl ion is formed by the reaction of 02' with the ferric chelate. Lipid hydroperoxide-dependent initiation of lipid peroxidation can be promoted by EDTArFe+2, DTPArFe+2, and ferric cytochrome P450. The EDTA-Fe+2 promotion of LOOH-dependent initiation of lipid peroxidation showed only minor inhibition by SOD or DPF. These results indicate that 02' and 102 may participate in LOOH-dependent initiation to a limited extent. The NADPH-dependent LOOH-dependent initiation of lipid -l47- peroxidation requires EDTArFe+3. The NADPHrdependent LOOH-dependent iniiation reaction does not demonstrate SOD or benzoate sensitivity. Thus the enzymatic reduction of EDTA-Fe+3 does not occur via NADPH- cytochrome P450 reductase generated 027 and 027 is not a major par- ticipant in the LOOH-dependent initiation of NADPH-dependent lipid peroxidation. Also, the data indicates that H0° does not participate in LOOH—dependent initiation. The NADPHedependent LOOH-dependent ini— tiation reaction does show minor inhibition by DPF and significant inhi- bition by BHT. Thus the NADPH-dependent LOOHrdependent initiation of lipid peroxidation appears to generate minor amounts of 102 and to be essentially radical mediated. The minor amount of 102 generated during peroxidation could perhaps arise from the reaction of two lipid hydro- peroxy radicals [29,163]. The OZT-dependent LOOHrdependent initiation of liid peroxidation demonstrated many similarities to the NADPH- dependent reaction. The only significant difference being the expected SOD sensitivity. Superoxide-dependent LOOHedependent initiation requires EDTArFe+3. Superoxide-dependent LOOHrdependent initiation does not show benzoate sensitivity and thus does not involve H0°. Again, as for NADPH-dependent LOOHrdependent initiation, OZT-dependent initiation demonstrates minor inhibition upon the addition of DPF to the reaction mixture and total inhibition upon the addition of BHT. These results indicate that LOOH-dependent initiation is essentially radical in nature but may perhaps involve the secondary generation 102 as discussed above. The EDTA-Fe+2 promoted LOOH-dependent initiation of lipid peroxidation may be a Fentons-type reaction. Such a reaction would require the redox cycling of EDTAriron. Thus the observed enzymatic dependence of LOOH- dependent initiation of lipid peroxidation. The EDTA-ferric complex can -l48- be reduced either directly by NADPH-cytochrome P450 reductase or by reaction with 027. Since HO' apparently is not formed during LOOH- dependent initiation the most significant product generated during this reaction may be the lipid alkoxy radical (LO'). LOOH + EDTA-Fe+2 + L0° + 021‘ + EDTA-Fe+3 (41) Apparently the lipid hydroperoxy radical (L00°) is also formed as is evidenced by the participation of 102 in LOOHrdependent initiation. The investigation of NADPHedependent and OZF-dependent liposomal peroxidation, where LOOH-independent and -dependent initiation occurs simultaneously, yielded results consistent with the proposed two step mechanism of lipid peroxidation. Superoxide dismutase, which can inhi- bit LOOHrindependent initiation of both 027-dependent and NADPHedepen- dent lipid peroxidation, almost completely inhibited 027-dependent and NADPH-dependent liposomal peroxidation. These results also demonstrate the dependence of LOOHrdependent initiation upon LOOH-independent initiation. Butylated hydroxytoluene, which inhibits both LOOHrdepen- dent and -dependent initiation, totally inhibits the completely reconstituted liposomal reaction mixtures. Benzoate, which does not inhibit either LOOH-independent or -dependent initiation, does not inhi- bit either NADPH-dependent or 027-dependent liposomal peroxidation. Diphenylfuran, which only inhibits LOOH-dependent initiation to a minor extent, inhibits NADPH-dependent and OZT-dependent liposomal peroxida- tion to a similar extent. The ability to promote 02'-dependent and NADPH-dependent LOOH- dependent initiation of lipid peroxidation is not unique to the EDTA- iron complex but is also shown by the DTPA-iron complex. These multi- dentate chelators, EDTA and DTPA, form similar strong complexes with -l49- iron. The complexes formed, stabilize the ferric ion relative to the ferrous ion and lead to a lowering of the complexes reduction potential in respect to that of free iron. The complexation of iron by EDTA changes the redox potential of iron from 0.77 for free iron to 0.25 for EDTArcomplexed iron [164,165]. The EDTAr and DTPArferrous complexes are quite strong reducing agents. It may be this alteration in reduction potential that makes an iron complex an efficient promoter of LOOH- dependent initiation. The chelation of iron by phosphates, such as ADP, does not stabilize the ferric ion relative to the ferrous ion to as great an extent as does EDTA or DTPA and as we have shown ADP-Fe+2 does not promote the LOOHrdependent initiation of lipid peroxidation. The effect of chelation on the spin state of iron may also be a con- sideration in the formation of an effective propagating agent. The abi- lity of DTPA-iron to promote the LOOH-dependent initiation of both NADPHrdependent and OZT-dependent lipid peroxidation may offer an alter- nate explanation of the recently reported results of Thomas et al. [154]. These authors found that OZT-dependent lipid peroxidation was dependent on contaminating levels of LOOH in the lipid matrix being peroxidized. They proposed that 027-dependent lipid peroxidation was initiated by LO° formed via the following reaction, which is essentially an uncatalyzed Haber-Weiss reaction. 02' + LOOH + LO° + 0H" + 02 (37) The authors added DTPA to their reaction mixtures to chelate con- taminating iron and presumably inhibit iron-promoted lipid peroxidation. In light of the results presented here, we feel that Thomas et al. [154] may have been observing the OZV-dependent LOOH-dependent initiation of -150- lipid peroxidation promoted by DTPArFe+2. The rate of the 02F-dependent peroxidation reaction reported by Thomas et a1. is in agreement with the rate of DTPA-Fe+2 promoted propagation of lipid peroxidation reported here. NADPHrdependent and OZF-dependent microsomal lipid peroxidation also occur by the proposed two step mechanism of lipid peroxidation. Both NADPH-dependent and 027-dependent microsomal lipid peroxidation require ADP-Fe+3. By analogy to the previous investigation on NADPH-dependent and OZT-dependent liposomal peroxidation it is likely that ADP-Fe+3 participates in the LOOH-independent initiation of microsomal lipid peroxidation. The SOD sensitivity of NADPH-dependent microsomal lipid peroxidation observed by others [39,72] indicates the possible involve- ment of the ADP-perferryl ion in the LOOHrindependent initiation of microsomal lipid peroxidation. Lipid hydroperoxide-dependent initiation of both NADPH-dependent and 02'-dependent microsomal lipid peroxidation is apparently facilitated by cytochrome P450. NADPH-dependent microso- mal lipid peroxidation can be inhibited by both SKF 525-A and AP SKF 525+A, at the concentration used in these experiments, is proposed to inhibit only cytochrome P450 mediated reactions without having signifi- cant effect on other microsomal activities or constituents [155]. The addition of a promoter of LOOHrdependent initiation, EDTArFeT3, to SKF 5252A inhibited NADPHrdependent microsomal lipid peroxidation completely reversed the observed inhibition. This data indicates the significant role that may be played by cytochrome P450 in microsomal lipid peroxidation. The inhibition of NADPHedependent microsomal lipid peroxidation by drug substrates of the microsomal mixed-function oxidase system has classically been considered to be mediated by competition for -151- reducing equivalents between drug metabolism and lipid peroxidation [156,157]. However, we have shown that the addtion of a promoter of LOOH—dependent initiation, EDTArFe+3, to an AP inhibited NADPH- dependent microsomal lipid peroxidation reaction mixture completely reverses the observed inhibition. This data indicates that AP inhibi- tion of NADPH-dependent lipid peroxidation is not mediated by the com- petition for reducing equivalents but rather upon the interaction of AP with an endogenous microsomal promoter of LOOHvdependent initiation of lipid peroxidation. The ability of EDTA-Fe+3 to reverse both SKF 525-A and AP inhibition of NADPH-dependent microsomal lipid peroxidation indi- cates that the endogenous promoter may be cytochrome P450. That cytochrome P450 is the endogenous microsomal promoter of LOOH-dependent initiation of lipid peroxidation is further substantiated by the ability of EDTA—Fe+3 to reverse the SKF 5252A mediated inhibition of 02’- dependent microsomal lipid peroxidation. The ability of ferric cytochrome P450 to function as a promoter of LOOHrdependent initiation is shown by its promotion of lipid peroxidation from LOOH and by its ability to promote lipid peroxidation in a reaction mixture capable of initiating NADPH-dependent lipid peroxidation. The unified mechanisms of NADPH-dependent and OZF-dependent lipid peroxidation we are proposing is presented schematically in Figure 10. The mechanism of lipid peroxidation proposed here offers a plausable guide for initial investigations into the mechanism in_yizg_lipid peroxidation. Considering the concentation of nucleotides [184] and "free" iron [185] in the hepatic cell, the participation of nucleotide- iron complexes, such as ADP-iron, in the LOOH-independent initiation of -152- vfimfiu ucmwcoamal. N .COfiuwvaxoumm 0 wow ucovcmaooImmm