MSU LIBRARIES m y 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. SUPEROXIDE DISMUTASE AS A PROTECTIVE AGENT AGAINST OXYGEN TOXICITY IN THE RETINA OF THE RAINBOW TROUT (SALMO GAIRDNERI) By Paul Emile Desrochers A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Physiology 1982 ABSTRACT SUPEROXIDE DISMUTASE As A PROTECTIVE AGENT AGAINST OXYGEN TOXICITY IN THE RETINA OF THE RAINBOW TROUT (SALMO GAIRDNERI) By Paul Emile Desrochers The toxic effects of hyperoxia have been attributed to reactive metabolites which inactivate enzymes and disrupt membranes. The ocular tissues of Salmo gairdneri are exposed to O2 tensions ten times those of arterial levels, yet maintain resistance to oxygen toxicity. Superoxide dismutases (SOD) provide primary protection against superoxide radicals. It was postulated that SOD was prevalent in the rainbow trout retina. Teleost retinal SOD levels were compared to those of the frog (Rana pipiens) and rat (Long-Evans strain) which have ocular p02 levels less than arterial. Liver and brain tissues were also examined. Crude tissue homogenates were assayed by the hydroxylamine oxidation method, and a parallel line analysis used. to detect endogenous interfering substances. The results were correlated with estimates of superoxide production, based on measurements of O2 consumption. Retinal SOD activity was the same in the trout and frog, but highest in the rat. The ratio, SOD activity/superoxide radical, indicated excess SOD protection in the rainbow trout retina. This thesis is dedicated to my parents, Roger and Juliette Desrochers and to my loving, patient wife Denise. All have inspired, encouraged, and supported me in the pursuit of my goals. I would also like to dedicate this to my newborn daughter Renee and hope that I can provide her with the same educational opportunities afforded me. 11 ACKNOWLEDGMENTS I would like to express my sincere gratitude to Dr. Jack R. Hoffert for his guidance and encouragement throughout this study. His patience and willingness to help have had a great impact on the results of this work. I would also like to thank Esther Brenke for her expert technical assistance. Her perseverance during the development of this assay was truly remarkable. She is also to be commended for the typing of this thesis. I appreciate the advice and guidance of my committee members, Dr. Paul O. Fromm and Dr. William L. Frantz. Their criticisms and suggestions have helped shape the final form of this thesis. A sincere thank you is expressed to Dr. Nicholas J. Rencricca of the University of Lowell for introducing me to research in physiology. The friendships that have developed with these people and other members and cohorts of the fish lab will be cherished. I am also grateful to Dr. Hoffert for financial support through grant No. EY-OOOO9 from the National Eye Institute. iii TABLE OF CONTENTS Page LIST OF TABLESOOOOOOOOOO..0...O0....OOOOOIOOOOOOOOOOOOOOOOO000...... Vi LIST OF FIGURES..................................................... vii INTRODUCTION........................................................ 1 LITERATURE REVIEW................................................... 4 Oxygen Toxicity................................................. 4 Ocular Oxygen Toxicity.......................................... 5 Cellular Oxygen Toxicity........................................ 10 Free Radical Chemistry.......................................... 12 Formation of Superoxide in Biological Systems................... 13 Other Damaging Species of Oxygen................................ 15 Protection Against Oxygen Toxicity.............................. 20 Enzymatic Scavenging of Superoxide......................... 20 Enzymatic Scavenging of Hydrogen Peroxide.................. 21 Non Enzymatic Scavengers................................... 22 The Significance of the Superoxide Dismutases................... 25 Superoxide Dismutase Activity Levels............................ 27 Superoxide Dismutase Assays..................................... 27 Superoxide Dismutase Assay Used in This Study................... 31 Resistance Of Teleost Ocular Tissues to Oxygen Toxicity......... 33 MATERIALS AND METHODS............................................... 34 Experimental Animals............................................ 34 Superoxide Dismutase Studies.................................... 35 Tissue PreparationOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 35 Crude Tissue Homogenate Preparation........................ 36 Assay ProcedureOOOOOOOOOOOOOOOOOOOOOOOOIOI.OOOOIOOOOOOOOOOO 38 iv Statistical AnalySiSo O O O O O O O O O O O I 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 O O ExpreSSion Of ACtiVity. C O I O O O O O O O O O O O O O O C O O O O O O O O O O O O O O O O O O O O O O 0 oxygen consmption Studies 0 I O O O O O O O O O O O I O O O O O O O O O O O O O O I O O I O O O O O 0 Tissue Preparation. 0 O O O O O I 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 O I O 0 Dry weight Determinations...000......OOOOOOOOOOOOOOOOOOOOO. Assay ProcedureOOOOOOO0.000......0.0.0....OOOOOIOOOOIOOOOOO RESULTS............................................................. Plotting Routine................................................ Assay Verification.............................................. Assay Sensitivity............................................... Preliminary Tests to Establish Experimental Protocol............ Enzyme Recovery................................................. SOD Results..................................................... Comparison of Tissue SOD Activities Within Species......... Comparison of Tissue SOD Activities Between Species........ Oxygen Consumption Studies...................................... Tissue SOD Activity vs. Tissue Oxygen Consumption............... DISCUSSION.......................................................... Validity of the Assay........................................... Advantages of this Assay........................................ Superoxide Dismutase Activity Levels............................ SUMMARY AND CONCLUSIONS............................................. RECOMMENDATIONS..................................................... LITERATURE CITED.................................................... APPENDIX I: Superoxide Dismutase Assay.............................. APPENDIX II: Lowry Protein Determination............................ APPENDIX III: Composition of Krebs-Ringer-Phosphate Medium.......... APPENDIX IV: Individual Experimental Results........................ 40 4O 41 41 41 41 43 43 54 54 57 59 63 63 69 69 73 75 75 76 77 81 84 86 93 96 97 9e LIST OF TABLES TABLE Page 1. Superoxide dismutase levels for liver, brain and eyeOO0.0.0...O0..0.00..0.0000000000000000000000...000...... 28 2. Comparison of treatments for rainbow trout retina.............. 58 3. Comparison of treatments for BESOD standard.................... 60 4. Superoxide dismutase activities in the liver, brain and retina from a teleost, amphibian and mammal................ 68 5. Oxygen consumption levels of retina, brain and liver from a teleost, amphibian and mammal..................... 72 6. Estimation of amount of protection in the liver, retina and brain from a teleost, amphibian and mammal.......... 74 7. Comparison of treatments for rainbow trout retina - individual experimental data................................... 98 8. Comparison of treatments for BESOD standard - individual experimental data................................... 99 vi LIST OF FIGURES Figures Page 1. A plot of the change in the optical density from the standard blank with increasing concentrations Of bOVine erythrocyte SOD.0..0....OOOOOOOOOOOOOOOOOOOOO0.0..0... 45 2. A plot of the change in the optical density from the standard blank vs. the coded logarithm of the bovine superoxide dismutase concentrations...................... 47 3. A plot of the percent inhibition of maximal color production by the oxidation of hydroxylamine vs. the coded 1n dose of BESOD...................................... 49 4. A plot of the probit response vs. coded ln dose Of BESOD...0.0.0.0....OOOOOOOOOOOOOCIOOOO0.0.0.0...0.0.0.0000... 51 5. A plot of the probit response vs. the coded 1n dose 0f BESODOOOOO0.0000000000000000000000000000000000000000000000000 53 6. A plot of the probit response vs. the coded ln dose Of BESOD.0.0.0000000000000000000000000000000000000IOOOO000...... 56 7. A representative plot of the probit response vs. the coded ln dose of BESOD standards after various treatments.’OOOOOOOOOOOOOOOOO0.0000000000IOOOOOOOOOOOOOO0.0.000. 62 8. A graph of the parallel line statistical approach for assaying superoxide dismutase activity...................... 65 9. Comparison of superoxide dismutase activities from various tissues in the rainbow trout, frog and rat.............. 67 10. Comparison of superoxide dismutase activities in the retina, liver and brain from various experimental animals....... 71 vii INTRODUCTION Three separate findings have had a considerable impact on research in the area of oxygen toxicity. First, was the realization that exces- sive oxygen administration to premature infants caused retrolental fibr0plasia (Shahinian and Malachowski, 1978). A high incidence of blindness among these infants led to a reduction in both the fraction of oxygen used and exposure time. The second important finding was the recognition of the similarity between oxygen toxicity and radiation injury (Gerschman et al., 1954a). A common mechanism was suspected i.e., the formation of free radical oxygen intermediates. Oxygen at high levels was subsequently shown to cause considerable damage not only to the vasculature of the premature retina, but to the retina itself‘ as well as 13ther ‘tissues. Most toxicity studies have centered around the lung but other investigations have been conducted on the retina and visual cell dysfunctions have been noted. Electroretinograms (ERG) of isolated retinas subjected to hyperoxic conditions have indicated a direct effect of oxygen on retinal cells (Ubels, Hoffert and Fromm, 1977). Exposure to as low as 55-60% 02 at ambient pressure for 7 days has been reported to result in visual cell death (Noell, 1962). Oxygen free radicals are believed to be the mediators of oxygen toxicity, inflicting damage through enzyme inactivation and disruption of lipid membranes. In fact, in current theory the free 02 radical is l increasingly being considered as a causal mechanism in aging and inflammation as well as in cancer. The free radical theory of oxygen toxicity has derived much of its strength from the third important finding, the identification of the enzyme superoxide dismutase (McCord and Fridovich, 1969). This enzyme scavenges the superoxide radical, a species considered of primary importance in oxygen toxicity because of its postulated reactivity and ability to form other deleterious oxygen intermediates. The swimbladders of certain fish are recognized as being resistant to oxygen toxicity. The retina of the rainbow trout (Salmo gairdneri) is also able to withstand high oxygen tensions, yet has gone largely unnoticed. The posterior portion of the retina in this species is continuously exposed to oxygen tensions on the order of 400 mmHg (Fairbanks, Hoffert and Fromm, 1969). This high p02 is generated by a countercurrent oxygen multiplier in the choroidal £232 mirabile. The p02 tensions generated around this tissue should cause oxygen toxicity, since exposure of mammalian species to levels of approximately 450 mmHg result in damage (Noell, 1962). Investigations on the retina in this species however, have repeatedly revealed its resistance to 1oxygen toxicity. It was postulated that some mechanism must be present in the retina to guard against the toxic effects of oxygen, the most likely candidate being superoxide dismutase. Assaying for superoxide dismutase activity in a crude homogenate is not simple since most of the assays have been develOped for use on a purified enzyme. In addition many are time consuming, lack sensitivity, are subject to interference by exogenous substances and are not suitable to multiple assaying. The goal of this study was: 1) 2) 3) 4) 5) To develo;> a superoxide dismutase assay ‘which incorporated benefits of previous assays. To apply the parallel line analysis to the assay responses to test for interfering substances. To measure the superoxide dismutase activity in the rainbow trout retina and compare it to retinal levels in a mammal (rat) and amphibian (frog). To measure SOD activity in the brain of these three species to determine if the activity was similar to that found in the retina. Liver tissues were also studied to compare the activ- ities to non-neural tissue. To measure oxygen consumption levels in these tissues in order to estimate superoxide anion production and obtain some estim- ation of the SOD protective capacity of these tissues. LITERATURE REVIEW Oxygen Toxicity The toxic effects of oxygen are related to the partial pressure of the gas, not the percent of oxygen being respired. Oxygen toxicity is a pathological condition which develops when the alveolar p02 is appre- ciably higher than that fbund when normal air is breathed. The gross manifestations of oxygen toxicity have been observed chiefly in the lung under the conditions of hyperoxia, a likely site, since the lung is in direct contact with the highest partial pressures of inspired oxygen. The pathological condition of oxygen toxicity in the lung has recently been reviewed (Deneke and Fanburg, 1980; Frank and Massaro, 1980), and includes edema, alveolar hemorrhage, fibrosis, inflammation, and func- tional impairment. In addition, toxicity has been demonstrated in the cardiovascular and central nervous systems, as well as causing red cell lysis and alterations in the metabolic state of the whole animal. Exposure to hyperbaric oxygen (HBO) hastens the toxic effects, most likely as a result of the higher arterial p0 Onset of pulmonary 20 damage is faster at 2 atm inspired pO2 than at 1 atmosphere. The major cardiovascular effect of hyperbaric oxygen is vasoconstriction. Hyper- baric oxygen exposure in humans has been found to decrease cerebral blood flow (Lambertsen et al., 1953). Animal studies in the rat indicated HBO exposure led to systemic hypertension producing left 5 ventricular failure and pulmonary edema. The increased pulmonary venous pressure ferced excess fluid and erythrocytes into the alveolar spaces from the capillaries (Wood et al., 1972). These investigators also reported a 15.7% decrease in the myocardial ATP levels of animals exposed to 100% 02, suggesting that pulmonary edema and pleural effusion was secondary to cardiac failure caused by a decreased ATP production. The greatest danger of exposure to hyperbaric oxygen is the poten— tial for central nervous system (CNS) damage, which is expressed as a generalized convulsive state. Oxygen toxicity of the CNS occurs in subjects breathing 100% 02 at 2 atm or more. In each case the severity of oxygen toxicity is dependent on the pressure and duration of exposure. Ocular Oxygen Toxicity Interest in the adverse effects of oxygen on the retina flourished in the early 1950’s when it was established that hyperoxia was the cause of retrolental fibroplasia (RLF) in premature infants. The vasculature to the retina does not fully mature until after full term delivery. As a result the immature vessels are susceptible to RLF. The primary effect of hyperoxia on these vessels is a vasoconstriction but continued exposure results in capillary endothelial damage and occlusion of the immature regions of the retinal vasculature. The secondary effects of RLF occur after the infant is removed to ambient air, when proliferation of the remaining vessels occurs, with an extrusion into the vitreous (Payne and Patz, 1979). This proliferation often results in vitreal hemorrhaging with scar tissue formation. The retina detaches then billows out toward the posterior surface of the lens causing blindness. 6 An analogous response of the retina to hyperoxia was found in the kitten. When exposed to 60-80% 0 at atmospheric pressure the ingrowing 2 vessels of the developing retina were obliterated (Ashton, Ward and Serpell, 1953). They concluded the extent of vessel obliteration was directly proportional to three factors: 1) the degree of immaturity of the retinal vasculature, 2) the duration of hyperoxic exposure, and 3) the concentration of oxygen used. Vessel obliteration was fbllowed by vessel growth with hemorrhaging when the kittens were returned to room air. Retinal detachment was observed 30 days after exposure to hyperoxia. Similar work was reported in newborn mice (Gerschman et al., 1954b), where eye lesions in the fOrm of retinal atrophy and changes in vascu- larization were attributed to continuous exposure to 70% 02. Subsequent work by Ashton (1968) implied oxygen acted on the endo- thelium in the immature retina. Comparative studies of oxygen toxicity suggested that destruction of the immature vessels by hyperoxia was a general biological phenomenon. It was proposed that the vaso- proliferation following hyperoxic exposure was the result of vaso-obliterative conditions. Uveal and retinal detachments were also observed in dogs exposed to 1 atm of 100% 02 for 48 to 50 hours (Yanoff, Miller and Waldhausen, 1970). In accord with the findings of Ashton, the toxic effect of oxygen on the choriocapillary endothelial cells was believed responsible for fluid leakage from the choroid into the subretinal space causing detachment. In 1955, NOell demonstrated that high oxygen tensions actually led to degeneration of the visual cells. Rabbits were exposed to oxygen 7 concentrations from 60-100% at ambient pressure, and a marked atten— uation of the electroretinogram (ERG) occurred progressively earlier with increasing 02 concentrations. The ERG was used as an indicator of visual cell viability. As expected with a decreased ERG amplitude, retinal deterioration was evident with degeneration of more than 70% of the visual cells after exposure to 100% 02 for 48 hours. More complete studies by Noell (1962) on adult rabbits demonstrated that exposure to 100% 02 at 3 atm for 4 h caused selective death of the visual cell layer while the outer nuclear layer diminished in thickness and in number of nuclei. The inner layers of the retina were un- affected. The b-wave of the ERG was used as an index for quantifying the effect of oxygen on the retinal layers. A larger reduction in the b-wave amplitude occurred at higher ambient pressures of 100% 02. Retinal examination revealed that slightly less than half of the rabbits exposed to 55-60% 02 at ambient pressure for 7 days demonstrated visual cell death. It was concluded that prolonged exposure to oxygen des- troyed visual cells without necessarily affecting the animal in other ways. The toxic effects of hyperbaric oxygen on the electroretinogram have also been studied in vitro (Ubels, Hoffert and Fromm, 1977). Hyperbaric oxygen was toxic to frog and rat retinas as demonstrated by attenuation of the ERG. The greatest effect on the frog ERG was seen with the b and c-waves, which were markedly reduced in the 6 h experimental period. The authors speculated that oxygen in this species, was more toxic to the pigmented epithelium and the neurons of the Muller cell layer than to the visual cells. This speculation was based upon knowledge of the cellular origin of the various components of the ERG. 8 The effect of HBO on the rat ERG was more rapid, with severe atten- uation occurring in a matter of 90-105 minutes. Abolishment of the a-wave was thought to be due to an effect on the visual cell layer. The accompanying decrease in b-wave amplitude was explained by the synaptic organization of the retina, where the response of the b-wave is depen- dent on the a-wave. This work was of particular importance since it showed that oxygen administration in vitro, had a direct effect on the electrical activity of the retina, apart from its effects on the respiratory and cardiovascular systems. Noell’s findings were confirmed by Bresnick (1970), who studied electron micrographs of the adult rabbit retina after the animals were exposed to 100% O2 fbr 24-40 hours. The earliest changes occurred in the inner segments and nuclei of the visual cells, followed by dis- organization of the entire visual cell. Swelling within the membranes of various cellular' organelles was noted and it was proposed that excessive oxygen or an oxygen metabolite was responsible fOr the disruption of the lipoprotein membranes of these organelles. Bresnick compared pigmented and albino rabbits to determine if the pigmented rabbits were less susceptible to oxygen toxicity, since melanin pigment was thought to be a good acceptor of free radicals and might act as an antioxidant. However damage to the retinal visual cells was identical in both species. In an effort to clarify how oxygen damages the visual cells, Shaw and Icon (1970) studied the effect of oxygen on enzymes essential in visual cell metabolism. Adult white rabbits were exposed to pure 02 at 600 or 760 mm Hg for a prolonged period of time. In support of previous investigators, they described severe damage to the visual cells after 9 oxygen exposure, eventually leading to visual cell death within 48 h of 100% 02 at ambient pressure. This was evidenced by the disappearance of nuclei in the outer nuclear layer. The b-wave of the ERG was markedly depressed prior to any histological evidence of ‘visual cell death. Significant losses in glucose-6-phosphate dehydrogenase activity and succinic dehydrogenase after oxygen exposure, were evident after visual -cell death. However, a downward trend in succinic dehydrogenase activity was seen earlier, within 24 h of 02 exposure at 760 mm Hg. This trend was observed prior to any marked ERG changes. The .authors con- cluded that the irreversible drOp in succinic dehydrogenase may be responsible fOr the early retinal changes after oxygen exposure. In vitro studies on retinal metabolism (Baeyens, Hoffert and Fromm, 1973) have also been undertaken, with the assumption that inhibition of the oxidative enzymes would result in decreased oxygen consumption. Studies on dog retinas incubated in oxygen at 37 C, at 1,470 mm Hg revealed a 37% drOp in oxygen consumption after a 24 h exposure. No change in oxygen consumption was observed fer frog retinas at 15 C under similar exposure conditions. A decrease in oxygen consumption was reported as the result of inactivation of critical enzymes of carbo- hydrate metabolism. The same group of investigators went on to study the influence of hyperbaric oxygen. on lactate dehydrogenase (LDH) activity (Baeyens, Hoffert and Fromm, 1974) in the retina. They described an inhibition of LDH activity in both frog and dog retinas after a 24 h oxygen exposure of 1,470 mm Hg at 15 C. Upon electrophoretic examination of the dog LDH isozymes it was suggested that oxygen exerted its deleterious effect at the active site of the molecule or an associated allosteric site. 10 Studies have also been conducted on the enzyme Na+-K+ ATPase (Ubels and Hoffert, 1981) in the retina, to determine if it is affected by hyperbaric oxygenation. It was hypothesized that since the ability of a cell to respond to a stimulus was dependent on the presence of a trans- membrane ionic potential, perhaps the toxic effect of oxygen was due to an inhibition of the processes responsible for maintaining transmembrane potentials in the retina. They cite that Na+-K+ ATPase is an essential enzyme for retinal function. Exposure of rat retinas to HBO at 3,800 mm Hg or 11,600 mm Hg for 4 h caused a significant 50—66% decrease in enzyme activity. It was concluded that this inhibition. may be a contributing factor in the toxicity of HBO to the mammalian retina. Cellular Oxygen Toxicity The harmful effects of oxygen have been known to occur at the cellular level. In addition to visual cell damage, oxygen has a more generalized effect on cells from other tissues. Decreased metabolism has been shown in various tissue slices from the rat after exposure to hyperbaric oxygenation (Stadie, Riggs and Haugaard, 1945). Studies on oxygen consumption of liver and brain slices from the rainbow trout, frog and dog verified metabolic inhibition after incubation in oxygen for 24 h at 1,470 mm Hg (Baeyens, Hoffert and Fromm, 1973). Inhibition of specific enzymes after oxygen exposure has been reported by several investigators. Haugaard (1946) established a link between the inactivation. of enzymes containing essential sulfhydryl groups and oxygen. He concluded that 02 inactivation was a property of enzymes incorporating essential sulfhydryl groups and that sulfhydryl oxidations converted the enzymes to an inactive form. 11 Other investigators have also described enzyme inactivation by oxygen. Horn and Haugaard (1966), reported 1 atm oxygen rapidly inhibited glycolysis in rat heart homogenates, directly through a reversible oxidation of sulfhydryl groups in glyceraldehyde-3-phosphate dehydrogenase. A comparative study on lactate dehydrogenase from the rainbow trout, frog and dog, revealed decreased LDH activity in liver and brain homogenates after hyperbaric oxygen exposure at 1,470 mm Hg for 24 h (Baeyens, Hoffert and From, 1974). Studies by Brown et al., (1979) revealed that inhibition of quinolinate phosphoribosyl trans- ferase by hyperoxia led to a rapid onset of serious metabolic defects. This enzyme is responsible for de novo biosynthesis of nicotinamide adenine dinucleotide (NAD), 95% of which is catabolized in numerous biochemical reactions. As a result, inhibition of this enzyme along with subsequent depletion of intermediates which contribute to the NAD pool via salvage pathways, would lead to cell death. Haugaard (1968) has provided an excellent review of the cellular mechanisms of oxygen toxicity. In addition to the actions of oxygen on enzyme sulfhydryl groups, he discusses its actions on nonprotein sulfhydryls, some of which are essential in maintaining enzyme sulfhydryls in their reduced state and in preventing the formation of mixed disulfides, known to inhibit hexokinase and interfere with mito- chondrial oxidation and phosphorylation. He has also discussed the role of oxygen in lipid peroxidation, with the assumption that free radicals formed from autoxidation were involved in the destructive effects on the cellular components. Further investigations of the deleterious effects of hyperbaric oxygen have been conducted at the subcellular level using rat heart 12 mitochondria (Nohl, Hegner and Summer, 1981). An increased accumulation of lipid peroxides was reported, as well as alterations in the fatty acid composition of the inner mitochondrial membrane. The findings point to the reactive oxygen metabolites as the species responsible for initiating lipid autoxidation. The effects of oxygen poisoning were recognized to be very similar to damage from radiation injury (Gerschman et al., 1954a) and led to the hypothesis that cellular damage to enzymes and alteration of cell function were brought about by a common mechanism, i.e., the formation of free radical oxygen intermediates. Gerschman also reported that agents which protected against the damaging effects of x-ray radiation were valuable in protecting against oxygen toxicity. It is now widely believed that the highly reactive metabolites of oxygen. are responsible for’ the diminished cellular function, partly evidenced by inactivation of enzymes and disruption of lipid membranes. Free Radical Chemistry Free radical fOrmation from oxygen may be predicted on the basis of inherent properties of the molecule. Molecular oxygen contains two unpaired electrons with parallel spins. As a result, any reducing agent which can donate a pair of electrons to oxygen must allow for spin inversion of one of the electrons. This occurs because it is magne- tically unfeasible for two electrons with parallel spins to occupy the same orbital. However spin inversion is unlikely since it is a rather slow process in comparison to the collisional properties of the molecules (Fridovich, 1977). Spin restriction can be avoided in two ways. One mechanism allows for the combination of oxygen either with transition metals containing l3 unpaired electrons or organic substances such as flavins, all of which allow for a multivalent reduction without the release of oxygen inter- mediates. An alternative is to add electrons to oxygen singly. It is the latter univalent reduction reaction which results in the formation of the superoxide anion radical (0;). Consequently due to spin restric- tion, the univalent reduction of oxygen is favored and commonly occurs in biological systems (Fridovich, 1977). Formation of Superoxide in Biological Systems It has been prOposed that various components of the electron trans- port system in the mitochondria are responsible for the formation of superoxide anions. Likely sites include the cytochrome molecules since they are one electron carriers. Flohe et al., (1977) have presented evidence suggesting that mitochondrial H202 is primarily, if not exclu- sively, produced from superoxide radicals. In: an attempt to describe the site of formation, previous experimental work on H 02 production in 2 mitochondria was referenced. Eflectron transport chain inhibitors were used on uncoupled intact mitochondria and on sub-mitochondrial parti- cles, in order to determine the locale of H202 formation. It was soon recognized that cytochrome b566 when fully reduced, readily reacted with O2 and was capable of producing H202. Since cytochrome b566 is a single electron carrier, the product of its reaction with oxygen is the super- oxide anion. This reaction however should not be considered a major contributor to the superoxide anion pool since cytochrome b566 must be in a reduced state. In the presence of ADP cytochrome b566 is completely oxidized and only gets reduced in the presence of ATP (Flohe et al., 1977). Therefore the ATP/ADP ratio must be extremely high before cytochrome b 66 will significantly contribute O2 anions. 5 14 A second site may involve the cytochrome oxidase enzyme. It had been proposed (Fridovich and Handler, 1961; Chance and Leigh, 1977) that the superoxide anion may be formed from this respiratory chain enzyme involved in the reduction of oxygen to water. In addition to a cytochrome b-type pigment and the cytochrome oxidase enzyme as possible generators of superoxide anions, an iron- sulphur component of NADH dehydrogenase has also been considered. Tyler (1975) reported that NADH dehydrogenase was a more likely site of super- oxide formation since the effects of this reactive species have been restricted to the NADH branch of the respiratory chain. In. work using submitochondrial particles, Tyler (1975) provided evidence which suggested that superoxide anions were produced during respiratory chain activity. The specific site of formation, based on studies of NADH oxidase activity, was thought to be an iron-sulphur component of the NADH dehydrogenase enzyme. The generation of superoxide anions from this component in the incubation media was believed to react with and inhibit NADH dehydrogenase activity and result in inhibition of the entire NADH oxidase system. No loss of activity was seen when the submitochondrial particles were incubated anaerobically, indicating that oxygen was necessary for inactivation. Speculation then arose that inactivation was due to the oxidation or some other modification of the type III thiol groups of NADH dehydrogenase by the superoxide radical. Tyler (1975) concluded that the NADH dehydrogenase complex in the electron transport chain was capable of the one electron reduction of oxygen to 0;, consistent with the E; value of the 02/0; couple of -O.3 V reported by several investigators. 15 Since succinate dehydrogenation can add electrons to the respiratory chain, succinate oxidase activity was also monitored. Electrons entering the transport chain from the succinate pathway, bypass the NADH dehydro- genase step, i.e., the 0g generating step. Related experiments conveyed the inability of succinate oxidation to cause inactivation of the NADH dehydrogenase enzyme. Tyler interpreted this to mean that formation of 0; from cytochrome oxidase was not expected to result in an interruption of electron transport chain activity. In addition to some of the electron transport components as possible generators of superoxide anions, there are 'numerous other reactions within the cellular milieu which are capable of releasing superoxide anions. Several enzymatic oxidation reactions (e.g. aldehyde oxidase and xanthine oxidase) result in the release of superoxide. The genera- tion of 0; also occurs via autoxidation of epinephrine, leucoflavin, hydroquinones, reduced ferridoxins, and hemoglobin. Whole cells such as phagocytic cells are also capable of releasing 0; anions (Fridovich, 1977). Other Damaging Species of Oxygen Damage to cellular components after oxygen exposure has often been associated with the superoxide radical, which has been implicated in lipid peroxidation, inactivation. of sulfhydryl enzymes and membrane disruption. There have been accounts in the literature however, which attribute this damage to other reactive oxygen species. The complete reduction of molecular oxygen to water requires 4 electrons. Were this reduction process to take place in univalent steps the_reactive inter— mediates formed in succession would include 02, H202 and the OH' (Fridovich, 1975). 16 According to Fridovich (1977), whenever 0; is generated in an aqueous system, hydrogen peroxide will soon also be formed. This hydrogen peroxide fOrmation results from the spontaneous dismutation of 2 2 and H202. dismutation is always quite rapid. As stated earlier H202 production is 0;, since 0 is not stable relative to the products of this dismutation, notably 0 Over the biochemical pH range this spontaneous quite common within the mitochondria. Several cellular oxidases are capable of generating H202 by a transfer of two electrons onto each oxygen molecule (Halliwell, 1979). Hydrogen peroxide has often been regarded as the species responsible for oxygen toxicity, since high concentrations will kill most cells. In fact, in a recent report (Simon, Scoggin and Patterson, 1981), hydrogen peroxide has been implicated as the agent responsible for killing human fibroblasts in vitro. These findings were based on studies which used appropriate oxygen radical scavengers and quantified their ability, if any, to protect the cells from oxidative damage. In most aerobic organisms, H202 is not considered to be the primary cause of oxygen toxicity, since it is at such low steady state concen- trations in animal cells and its production does not increase subsequent to hyperbaric oxygen exposure. Halliwell (1979) further reports that moderate levels of H202 are tolerable to aerobic cells but the concen- trations must not be allowed to increase to high levels. Hydrogen peroxide itself is not a highly reactive species, but in the presence of transition metal ions it is able to decompose to the hydroxyl radical (OH'), the most potent oxidant known which will attack and damage almost every molecule fOund in living cells (Fridovich, 1977; Halliwell, 1979). 17 Much debate has ensued over whether this species can be fermed directly from the presence of superoxide radicals and hydrogen peroxide. In 1934, Haber and Weiss prOposed a chain radical reaction for the decomposition of H202 by iron salts with the production of oxygen. The reaction: 02 + H202 ----- > 02 implied that the simultaneous presence of superoxide and hydrOgen + OH” + OH' (1) peroxide would give rise to hydroxyl radicals. This reaction however has never been clearly demonstrated (Cohen, 1977). Pure H202 and 0; will not react together at significant rates to form the hydroxyl radical in vitro unless catalyzed by complexes of iron or some other transition metal (Fridovich, 1977; Halliwell, 1979). The following reactions have been proposed for hydroxyl radical production in vivo: Fe(III) + o; ----- > Fe(II) + 02 (2) Fe(II) + H202 ----- > Fe(III) + 0H“ + OH' (3) Hydroxyl radicals can. also be generated in 'vitro by the action of ionizing radiation on water and via a photolytic decompostion of H202 with ultraviolet radiation (Fee and Valentine, 1977). However hydroxyl radicals have not been directly observed in vivo. Their involvement in biological destruction has only been inferred from studies involving "so-called selective radical scavengers" (Willson, 1979). Experiments have been conducted using hydroxyl radical scavengers as well as scavengers for O2 and H202, as a means of determining whether OH' radicals are produced from this reaction. From these indirect methods various conclusions have arisen that: (1) when 0; and H202 are 18 present simultaneously a strong oxidant is generated with properties associated with the known chemistry of OH', and (2) the "strong oxidant" can be intercepted by OH' scavengers with approximately the same reaction rate constants as fOund fOr the reactions between each scav- enger and the hydroxyl radical. Superoxide dismutase and catalase were also effective, most probably by the removal of one of the reactants. Thus it was concluded that the "strong oxidant" generated was identical to the hydroxyl radical (Cohen, 1977). Although there are numerous biochemical accounts in the literature which suggest that the Haber-Weiss reaction is an important reaction in biological systems, Fee and Valentine (1977), are quick to point out that all the chemical evidence indicates that superoxide in an aqueous solution will not directly bring about the reductive cleavage of H202, repudiating the suggestions of its importance in biological systems. Another reactive oxygen species which has been questionable in biological systems is the oxygen molecule in its excited state, other- wise known as singlet oxygen (102). This species is formed when an electron undergoes spin inversion and is excited to a higher energy state. The lifetime of singlet oxygen is relatively long since the transition from excited to ground state is forbidden (Fee and Valentine, 1977). Singlet oxygen has been shown to undergo oxidative reactions with various cellular components including bilirubin, amino acids, cholesterol and nucleic acid bases (Lindig and Rodgers, 1981). However its actual formation in biological systems has been challenged. Kellogg and Fridovich had proposed the fbrmation of singlet oxygen for the initiation of lipid peroxidation, from the Haber-Weiss reaction: - " o 1 02 + H202 ----- > OH + OH + 02 . (4) 19 Effective scavenging of 05 and H202 as well as very efficient protection from known singlet scavengers led to this prOposal (Fridovich, 1977)- Singlet oxygen formation had also been predicted by Pederson and Aust (1973) by the non-enzymatic dismutation of superoxide radicals: 20‘ + 2H+ ----- > 1O + H O (5) 2 2 2 2 However numerous investigators have questioned the validity of these reactions, as well as the specificity of the singlet oxygen scavengers (Packer et al., 1981). Their concerns have been expressed and reviewed (Fee and Valentine, 1977; Svingen, O’Neal and Aust, 1978). Michelson (in Hill, 1979) has also cast doubts on the formation of singlet oxygen, based on evidence provided from emission wavelength studies which indi- cated that singlet oxygen formation does not occur by spontaneous dis- mutation of superoxide radicals. In addition he speculated that results from lipid peroxide studies could have been due to a number of reac- tions, and the only way singlet oxygen could be formed was by photo- chemical reactions. Other powerful oxidizing agents which have rapidly gained more attention are the metal-oxygen complexes, (e.g., [Mn02]+ and [FeIIOZ]-), whose reactivity is comparable to the hydroxyl radical. These complexes are formed from reactions of 0; with metal ions. The best biological source of the ferryl intermediate ([Fe1102]') is most likely cytochrome P450 (Hill, 1979). In summary, although superoxide itself is a destructive species within the cell, further danger exists from the proposed reductions it may undergo to fOrm OH' radicals, H202 and singlet oxygen. 20 Protection Against Oxygen Toxicity Enzymatic Scavenging of Superoxide Biological interest in oxygen free radical chemistry was spurred by the discovery of an enzyme that scavenged the superoxide radical (McCord and Fridovich, 1969). This enzyme limited the buildup of superoxide radicals by converting them to hydrogen peroxide plus oxygen. SOD O; + 0; + 2H+ ..... > H202 + 02 (6) As a result it was named superoxide dismutase (SOD). The fact that the 1sec-1) in the rate of this reaction is normally very high (8x104 M- absence of superoxide dismutase, could only lend credence to the suppo- sition that the 0; radical is a highly damaging species. The family of superoxide dismutases are said to be common among respiring cells. There are three types of superoxide dismutases which have been catagorized. All three catalyze the same reaction with com- parable efficiencyx The iron-containing (Fe-SOD; MW=39,000) and manganese-containing (Mn-SOD; MW=40,000) enzymes are very similar in amino acid sequence and are commonly found in bacteria. The enzyme which contains both COpper and zinc (CuZn-SOD; MW=32,600) is routinely associated with the cytosol of eukaryotes, but has also been isolated from the intermembranal space of the mitochondria. The Mn-SOD form is also found in eukaryotes, primarily within the mitochondrial matrix, differing only in size with a molecular weight of 80,000. These last two forms of SOD are distinguishable by their sensitivities to cyanide and treatment by a chloroform-ethanol mixture (Fridovich, 1978). The CuZn-SOD form is inhibited by cyanide but supposedly stable to a 21 chloroform-ethanol extraction. However, the Hn-SOD form is resistant to cyanide yet destroyed by the chloroform-ethanol mixture. The mechanism of action fer all of the superoxide dismutases can be written: E-Men + 0; ----- > E-Nen’1 + 02 (7) E-Men'1 + o; +211+ ----- > E-Men + H202 (a) where E denotes the enzyme and Me the metal ion. The metal ion involved in the reduction and reoxidation of the CuZn-SOD form is the copper ion, which alternates between the cupric and cuprous states. The zinc is thought to play a structural role lending stability to the molecule. The metal ion in the Mn-SOD and Fe-SOD forms will alternate between the trivalent and divalent states. The superoxide dismutases are extremely -1, which is efficient with a rate of reaction approximating 2X109 M-1sec very close to a diffusion-limited process. Only a brief discussion of the enzyme has been presented here, since more extensive reviews are available in the literature (Fridovich, 1974, 1975, 1978, 1979; Michelson, McCord and Fridovich, 1977). Enzymatic Scavenging of Hydrogen Peroxide Superoxide and the superoxide dismutases have gained the most atten- tion with respect to oxygen toxicity, however other scavengers of oxygen intermediates must also be considered. The dismutation of superoxide radicals as well as the action of cellular oxidases will result in hydrOgen peroxide fbrmation. As previously mentioned, hydrogen peroxide is not very reactive, but does have the capacity to produce hydroxyl radicals. Accumulation of hydrogen peroxide is prevented by two related enzymes, the catalases and 22 the peroxidases. Both catalyze the reduction of H202 to water. The enzymes differ in that catalases can use H202 itself as the source of electrons for the reduction, whereas peroxidases use some other reduc- tant. The respective reactions are given below: Catalase H202 + H202 ---------- > 2H20 + O2 (9) Peroxidase H202 + H R ------------ > 2H20 + R (10) 2 The physiological reductant may be glutathione, ascorbate or cytochrome c, and these would be continuously renewed by the reduced forms of nic- otinamide adenine dinucleotide (NADH) and nicotinamide adenine dinucleo- tide phosphate (NADPH) (Eldred, 1979). Neither enzyme: is as ‘widely distributed as the superoxide dis- mutases. In humans, catalase is found in high levels in erythrocytes, liver and kidney. Very little is found in other tissues, which rely on the catalase of the circulating RBC’s to remove and decompose the H202 they produce (Fridovich, 1977). The enzymes however may work in con- junction, in that a lack of catalase may be compensated for by high peroxidase levels. In animal cells, glutathione peroxidase is probably the more significant enzyme. ZLt is widely distributed and is effective at low concentrations of H O 2 2. Not only can it reduce H202, but it is also capable of reducing lipid peroxides to hydroxyacids. It therefore is able to terminate the chain reaction of lipid peroxidation, limiting membrane destruction (Fridovich, 1977; Halliwell, 1979). Non-Enzymatic Scavengers The superoxide dismutases, catalases and peroxidases are the first line of defense against the toxic intermediates of oxygen. Secondary 23 defenses include the antioxidants ascorbic acid and vitamin E as well as thiol compounds and numerous other free radical scavengers. The mechanism of ascorbate scavenging involves hydrogen abstraction by the superoxide anion (Nishikimi, 1975) (AH2 = ascorbic acid, AH' = semidehydroascorbate): 0; + AH2 + H+ ----- > H202 + AH' (11) Bielski, Richter and Chan (1975) reported that the ascorbate free radical was a relatively nonreactive species, unable to reduce 02 to 0;. The ascorbate radical could then be further oxidized by 0; to dehydro- ascorbic acid or would undergo spontaneous dispr0portionation, thereby terminating the prOpagation of free radical reactions (Nishikimi and Yagi, 1977; Bielski et al., 1975). Green. and. O’Brien (1973) speculated that ascorbate may also be effective in removing pmtentially harmful lipid peroxides by utilizing the reducing capabilities of NADH. It was prOposed that the reaction of ascorbate with a lipid peroxide would give rise to the semidehydro- ascorbate radical. NADH would then be oxidized by the ascorbate radical resulting in the regeneration of ascorbate. Localization of the semi- dehydroascorbate reductase was found likely to be along the electron transport chain and within the outer membrane of the mitochondrion (Green and O’Brien, 1973). The reaction rate of 0; with ascorbate is slow in comparison to the rate for SOD with 0- but the overall rate of reaction of 02 with 2’ ascorbate is comparable at very high ascorbate concentrations. Elevated levels of ascorbate however, especially in the presence of 24 iron (III) salts may be harmful. The ascorbate acts to reduce Fe3+ to F62+z 3+ 2+ 2Fe + ascorbate ----- > 2Fe + dehydroascorbate (12) The Fe2+ can then give rise to 0; and OH' radicals: Fe2+ + O < ----- > Fe2+ - O < ----- > Fe3+ + o‘ (13) 2 2 2 Fe2+ + H202 """ > Fe3+ + OH' + OH- (14) (Wong et al., 1981) Vitamin E is a natural antioxidant and inhibitor of the autoxidation of polyunsaturated fatty acids (Fridovich, 1977). On the basis of its chemical properties and autoxidation mechanisms, Tappel (1972) reported that a-tocopherol (vitamin E) might be expected to react as a chain- breaking antioxidant in the inhibition of lipid free radical peroxida- tion. The prOpagation reactions for lipid peroxidation include (L' = lipid free radical; LOé = lipid peroxidase radical; LH = polyunsaturated lipid; LOOH = lipid hydroperoxide): L‘ + 02 ----- > LOé (15) Log + LH ----- > LOOH + L' (16) Inhibition would occur by abstraction of hydrogen from OL-tocOpherol (mun: LOé +cx-TH ----- > LOOH + o-T' (17) This antioxidant protection. by ‘vitamin E has been supported. by numerous investigators noting increased lipid peroxidation by vitamin E deficient animals as well as protection against 02 induced lipid perox- idation (Haugaard, 1968). However, this theory has also come under some criticism (Green, 1972). Leung, Vang and Mavis (1981), have provided evidence for a cooPera- tive interaction between vitamin E and ascorbic acid for the suppression 25 of lipid peroxidation. Vitamin E enhanced the efficiency of ascorbate in providing extended antioxidant protection in a model membrane system. It was hypothesized that the COOperative effects could occur in vivo. Ascorbate was thought to provide antioxidant reserve while vitamin E, with its lipid solubility characteristics, could function as a specific highly efficient antioxidant (Leung et al., 1981). Thiol (SH) compounds also exert a protective action against oxygen toxicity by undergoing hydrogen abstraction, causing reactivation of enzymes inhibited by oxygen. The thieyl radicals (~S’) could combine to form disulfide bonds (-S-S-). One of the more important SH compounds is reduced glutathione. The primary function proposed for reduced gluta- thione is the maintenence of enzyme sulfhydryl groups in their reduced state. Glutathione may also act as a scavenger of free radicals (GSH = reduced glutathione, GS' = glutathieyl radical, GSSG = glutathione dimer): 2GSH + 02 ----- > 2GS + H202 (1s) 2Gs' ————— > GSSG (19) Since glutathione is a stronger reducing agent than ascorbic acid, it may also play a role in keeping ascorbate in a reduced state in the cell (Haugaard, 1968). Haugaard (1968) and Eldred (1979) have provided excellent reviews of the numerous compounds which will scavenge free radicals and limit the lipid peroxidative chain reactions. The Significance of the Superoxide Dismutases The free radical theory of oxygen toxicity and the importance of the superoxide dismutases has received a tremendous amount of support 26 throughout the literature. As Fridovich (1977) has contended: "The very existence of enzymes which catalytically scavenge O- is an indication of the potential cytogoxicity of this radical. How else can one explain the necessity of enzymes to speed what is spontaneously a rapid reaction?" As indicated, superoxide is capable of causing biological damage, whether by direct or indirect means. The essentiality of the superoxide dismutases is supported by several types of evidence. It was expected that since superoxide dismutase protected against the products of univalent oxygen reduction, it would be deemed essential only for oxygen metabolizing cells. An early survey of the superoxide dismutase activity in a wide range of bacteria supported this hypoth- esis. Aerobic organisms which almost exclusively utilize oxygen in their metabolism were found to have the highest SOD activity. Aero- tolerant organisms, anaerobes which can grow in oxygen but never develOp a respiratory chain, contained intermediate levels of SOD. Strict anaerobes which are intolerant to oxygen, did not possess superoxide dismutase activity (Fridovich, 1975). Subsequent studies have revealed that the correlation between SOD and oxygen tolerance is not perfect since SOD has been found in lesser amounts in some strict anaerobes (Fridovich, 1979). Exposure of some aerotolerant organisms to oxygen resulted in increased production of superoxide dismutase which was correlated with an increased resistance to oxygen toxicity. Transfer of E. coli cells from an anaerobic to aerobic environment induced the synthesis of Mn-SOD which resulted in a resistance to oxygen toxicity (Fridovich, 1978). Superoxide dismutase has also been induced in vivo in rat lungs after an 27 exposure to 85% oxygen for at least 7 days. This increase in SOD activity was believed responsible for an increased tolerance to 100% O2 (Crapo, 1977). Although there have been numerous accounts in the literature per- taining to the importance of the enzyme, questions still arise concern- 2 systems (Sawyer and Valentine, 1981) as well as the scavenging capabil- ing the actual reactivity and production of the O anion in biological ities of the superoxide dismutases (Fee, personal communication). Superoxide Dismutase Activity Levels Various assays fer superoxide dismutase have been developed with the aim of designing one which was sensitive, specific and capable of hand- ling large sample numbers. Superoxide dismutase activity has been determined in many tissues, but due to a limited space only the liter- ature values fer tissues used in this study have been reported (Table 1). A more complete survey of SOD activities has been conducted by Eldred (1979). A comparison of SOD activities however is extremely difficult. No standardization for the expression of enzyme activity has been estab- lished, e.g., units of activity per milligram DNA, per mg protein, per mg tissue, or per gram of hemoglobin, etc. In addition, the definition of one unit of SOD activity is usually the amount necessary to inhibit reaction rates by 50%. Since the reaction kinetics vary with each assay, a restriction is placed on widespread comparisons. The deter- mination of activity is therefore intrinsic to the individual assay. Superoxide Dismutase Assays The problem with the assay procedures is the lack of specificity of the superoxide anion. Although SOD is specific for 02, the superoxide 28 Rhesus Monkey Human 21.5 U/mg protein 26.2 U/mg protein Table 1. Superoxide dismutase levels for liver, brain and eye. Tissue Level Reference Liver: Carp 215 U/g wet tissue Matkovics et al., 1977. Carp 569 U/g tissue Mazeaud, Maral and Michelson, 1979. Saithe 460 U/g fresh tissue Aksnes and Njaa, 1981. Mackerel 716 U/g fresh tissue Aksnes and Njaa, 1981. Frog 1050 U/g wet tissue Matkovics et al., 1977. Deer Mouse 19.3 U/mg protein Tolmasoff, Ono and Cutler, 1980. Mouse 20.8 U/mg protein Van Balgooy and Roberts, 1979. Rat 27.3 U/mg protein Van Balgooy and Roberts, 1979. Rat 4000 U/g wet tissue Matkovics et al., 1977. Rhesus Monkey 21.0 U/mg protein Tolmasoff et al., 1980. Human 38.0 U/mg protein Tolmasoff et al., 1980. Human 1680 U/mg protein Westman and Marklund, 1981. Brain: Carp 184 U/g wet tissue Matkovics et al., 1977. Carp 25 U/g tissue Mazeaud et al., 1979. Saithe 778 U/g fresh tissue Aksnes and Njaa, 1981. Trout 15.2 U/mg protein Van Balgooy and Roberts, 1979. Frog 13.0 U/mg protein Van Balgooy and Roberts, 1979. Frog 832 U/g wet tissue Matkovics et al., 1977. Deer Mouse 16.2 U/mg protein Tolmasoff et al., 1980. Mouse 17.6 U/mg protein Van Balgooy and Roberts, 1979. Rat 18.8 U/mg protein Van Balgooy and Roberts, 1979. Rat 240 U/g wet tissue Matkovics et al., 1977. Rabbit 17.3 U/mg protein Van Balgooy and Roberts, 1979. Tolmasoff et al., 1980. Tolmasoff et al., 1980. Table 1. (cont'd.). Tissue Level Brain: Human Gray Matter 710 U/mg protein White Matter 760 U/mg protein .2121 Bovine Iris 30.3 U/mg protein Iris 5.83* Lens Nucleus 0.1 U/mg protein Lens Cortex 0.2 U/mg protein Lens 1.15* Sclera 27.8 U/mg protein Sclera 4.86* Vitreous 15.8 U/mg protein Choroid 20.6 U/mg protein Retina 27.8 U/mg protein Retina 4.08. Cornea 0 U/mg protein Corneal 4.25* Epithelium e Equivalent microgram bovine (x10-2). 29 Reference Westman and Marklund, 1981. Westman and Marklund, 1981. Crouch, Priest and Duke, 1978. Bensinger and Johnson, 1981. Crouch et Crouch et Bensinger Crouch et Bensinger Crouch et Crouch et Crouch et Bensinger Crouch et Bensinger erythrocyte al., 1978. al., 1978. and Johnson, al., 1978. and Johnson, al., 1978. al., 1978. al., 1978. and Johnson, al., 1978. and Johnson, enzyme to equal 1 1981. 1981. 1981. 1981. unit 30 anion is not specific to SOD. As previously suggested, the relative instability of 0; makes assaying SOD activity less than straightforward (McCord, Crapo and Fridovich, 1977). Most of the SOD assays have had to rely on an enzyme catalyzed alteration in the steady-state concentration of superoxide. Therefore the commonly used assays consist of two components, a superoxide generator and a superoxide detector. The generator will produce the radical at a controlled, constant rate. In the absence of SOD the superoxide radical accumulates to a steady-state, where the rate of production of the superoxide anion is equal to the rate of reaction with the detector. When SOD is present, it competes with the detector for superoxide, lowering the steady-state concentration of the radical. As a result, there is a decreased rate of occurrence of the detection phenomenon (McCord et al., 1977). Hence the indirect assays rely on a competition between SOD and the detector. The majority of the assays were developed for use on the purified enzyme. Unfortunately, few of the assays have been modified for use with crude tissue homogenates. Many substances are present within crude homogenates that are capable of interfering with the assay reactions. As previously mentioned, the SOD-like activity of alternative superoxide scavengers (e.g. ascorbate, glutatione and vitamin E) may mask the true SOD activity. The autoxidation of ferrous ions in the presence of high ascorbate levels or in phosphate buffer may also affect the true SOD activity by increasing the steady-state level of 05 (Michelson, 1977a). Altered SOD activity may also result from substances which might act directly upon the indicator, such as the reduction of cytochrome c by ferrous ions, FMN, riboflavin and NADH (Michelson, 1977b). Numerous 31 other small molecules which may affect the assays have also been discussed (McCord et al., 1977; Michelson, 1977b). Another problem with crude tissue homogenates is erythrocytic contamination. Superoxide dismutase activity of the erythrocytes may comprise a substantial fraction of the total activity in homogenates of blood filled organs. Hemoglobin is also known to interfere with many of the spectrophotometric assays and should be removed prior to the determination. McCord and Fridovich (1969) employed the Tsuchihashi fractionation (a mixture of ethanol and chloroform) for the precipi- tation of hemoglobin in the first assay for superoxide dismutase. This method has been adapted for use with other assays and appears to be the best. The major problem of a chloroform-ethanol extraction is the denaturation of the Mn-SOD form (Fridovich, 1978). This would lead to an underestimation of the total SOD activity in tissue samples. The reader is referred to the fOIlowing (McCord et al., 1977; Misra and Fridovich, 1977; and Eldred, 1979), fer more complete discussions of the pros and cons of the commonly used superoxide dismutase assays. Superoxide Dismutase Assay Used in This Study Elstner’ and. Heupel (1976) have developed a superoxide dismutase assay which is specific, sensitive and inexpensive to run. This assay was later modified by Rao et al. (1978) for use on animal tissues. A xanthine-xanthine oxidase system is used to generate OE radicals which oxidize hydroxylamine to fbrm nitrites at neutral pH. These couple to sulfanilic acid andci-naphthylamine to produce an azo compound which can be measured spectrOphotometrically at a wavelength of 530 nm. Super- oxide dismutase activity can be assayed on the basis of its ability to inhibit the fOrmation of nitrites from hydroxylamine. Competition for 32 the superoxide radicals by SOD limits the amount of color produced. The superoxide dismutase assay as described by Elstner and Heupel (1976) used a-naphthylamine, a known class A carcinogen. Rather than face the inherent risks of using this compound, N,N-Dimethyl- 1-naphthylamine was substituted in its place. An added advantage reported by Germuth (1929) was that the dimethylated compound was capable of producing more distinct coloration in solutions containing small amounts of nitrite ions and that the color possessed the property of stability and permanence. This assay was selected for a number of reasons, primarily because of its reported sensitivity and specificity. It is also considered more sensitive than the standard cytochrome 0 reduction assay (Lengfelder and Elstner, 1979). The autoxidation of hydroxylamine may have been per- ceived as a potential problem. However, Kono (1978) has investigated the autoxidative prOperties of hydroxylamine and has reported the absence of hydroxylamine autoxidation below pH 8.0. The assay is also particularly useful for multiple sample deter- minations since it is set up as an "endpoint" assay. With no SOD present, maximum color is produced within 20 minutes. In the presence of SOD some of the 0; anions are scavenged resulting in less color when the reaction is completed. An added advantage of the assay over most is that the superoxide formed functions as an oxidant, not a reductant. Therefore this assay may be less susceptible to specific interference by endogenous reduc- tants which might act on a detector substance that must undergo a reduction (McCord et al., 1977). No attempt was made to purify the crude tissue homogenates, however a parallel-line analysis of variance 33 was applied to the data in order to reveal the presence of interfering substances (Eldred and Heffert, 1981). .A more detailed discussion of this statistical approach is given in the Results section. Resistance of Teleost Ocular Tissues to Oxygen Toxicity The teleost eye possesses an oxygen countercurrent multiplier, the choroidal rete mirabile, which is capable of generating oxygen tensions in excess of 400 mmHg at the posterior surface of the retina (Wittenberg and Wittenberg, 1962; Fairbanks, Hoffert and Fromm, 1969). The trout retina is dependent on this high p02 for normal visual func- tioning (Fonner, Hoffert and From, 1973). Oxygen tensions in this range cause oxygen toxicity in most species. Surprisingly, the rainbow trout retina has repeatedly been shown to be resistant to oxygen 'toxicity. The metabolic enzymes of the trout retina are not destroyed by hyperoxia as evidenced by an increase in oxygen consumption (Baeyens et al., 1973). The enzymes lactate dehydrOgenase and Na+-K+ ATPase have also been shown to be resistant to hyperbaric conditions (Baeyens et al., 1974; Ubels and Hoffert, 1981). ID: fact, lactate dehydrogenase activity was enhanced by hyperbaric oxygen. ElectroretinOgraphic studies have indicated that the teleost retina is not only well adapted to a p02 of 400 mmHg but is also resistant to oxygen tensions much greater than those fOund in vivo (Ubels, Hoffert and Fromm, 1977). The neural retina in this species is avascular and therefore depen- dent on the diffusion of oxygen down a gradient from the rete. The evidence summarized here would suggest that the rainbow trout retina has developed some mechanism to protect against the deleterious forms of oxygen. MATERIALS AND METHODS Experimental Animals SuperOxide dismutase (SOD) activity was measured in variOus tissues frOm selected teleOsts, amphibians and mammals. The SOD activity in whOle retinas fr‘om rainbOw trOut (Salmo gairdneri) were cOmpared to neural tissue (brain) and nOn-neural tissue (liver). These were cen- trasted to the respective activities in the nOrthern grass fr‘og (3322 pipiens), and rat (Leng-Evans strain). Rainbow trout, 200-400 g, (Midwest Fish Farming Enterprises, Inc., Harrison, MI) were held in fiberglass tanks at 9:1 C, with a c‘ontinubus flow Of decthrinated, aerated water. Twice a week the tanks were cleaned and the fish fed. The photOperiOd was 16 h light, 8 h dark. NOrthern grass frOgs, 15-25 g, (Nasco BiOIOgic Co., Ft. AtkinsOn, WI) were maintained in a refrigerated rOOm at 12:1 C, in a mOist aquarium which was cleaned when necessary. Feeding was not required because Of the cbld envirOnmental temperatures. The photOperiOd was 12 h light, 12 h dark. long Evans rats, 250-400 g, (Charles River Breeding LabOra- tOries, Wilmingtbn, MA) were hOused under natural photOperiOd at rOOm temperature. The rats were fed ad libitum and the cages cleaned twice a week. The LOng Evans strain was selected fer the superOxide dismutase studies because Of their pigmented eyes. The melanin within the pigment epithelium protects the retina frOm the high levels Of light nOrmally encOuntered by the albino Sprague-Dawley strain. This high incidence of 34 35 light may induce the production Of Oxygen free radicals resulting in retinal damage. Sprague-Dawley rats were used fOr the oxygen cOnsump- tion studies instead Of the LOng-Evans rats because they were readily available and it was assumed Oxygen consumption levels between the two strains wOuld not differ significantly. SuperOxide Dismutase Studies Tissue Preparation Preliminary experiments revealed hemOglobin interfered with the indirect hydrOxylamine Oxidation. assay' fOr' superOxide’ dismutase. A. series Of tests were cOnducted to determine the best methOd tO eliminate hemOglObin. WhOle body perfusiOn with isotOnic saline eliminated the hemOglObin problem and ensured the preparatiOn Of erythrOcyte-free tissue homogenates. This decreased the likelihOOd Of cOntaminatiOn by erythrocytic superOxide dismutase. The trOut were anesthetized with tricaine methane sulfonate (MS-222; Finquel, Ayerst LabOratOries, Inc., N.Y., NY) and placed in a trOugh ventral side up. They received cOntinuOus ventilatiOn frOm water passing Over their gills at a rate Of 4.2 L/min. The heart was expOsed by a midsagittal incision along the ventral side. An injectiOn Of 0.1 ml Sodium ‘Heparin (1000 units/ml), (Uijhn. Co., Kalamazoo, MI) ‘was intrOduced to the circulatiOn via the ventricle to prevent clotting. The ventricle 'was then. cut, a cannula (PE-50) inserted past the 'bulbus arteriOsus and held in place with a ligature. The frogs were anesthetized with 148-222 and pinned to a cutting bOard. The thOracic cavity was Opened and the fascia, muscle and sternum trimmed away to expOse the heart. SOdium Heparin (1000 units/ml) was injected into the ventricle which was then cut and a 36 cannula (PE-50) inserted directly. In both species the atria were cut to prevent a buildup Of pressure. IsOtOnic saline was perfused thrOugh the tissues Of both species at a rate of 6 ml/min fOr apprOximately 6-10 minutes. PerfusiOn was cOn- tinued until the tissues tOOk On a pallOr appearance. The animals were killed by cervical sectiOn and single pithed. The rats were anesthetized with diethyl ether (Mallinckrodt, Inc., ,Paris, KY). A jugular vein was isOlated and cannulated with PE-50 tubing, and both femOral arteries were severed. Sodium Heparin (50 units/100 ml) was added to the perfusate and perfusiOn thrOugh the jugular was initiated at 6 ml/min fOr apprOximately 10-18 minutes. PerfusiOn was halted when the liver turned a pale brOwn cOlOr. In each instance, Once the perfusiOn was stopped the tissues were immediately excised. Liver samples were easily Obtained with tissue forceps and iris scissOrs. Heavy scissOrs were used tO cut thrOugh the cranial bOnes to reach the brain. The eyes were remOved by severing the extraOcular muscles and optic nerve. All tissue samples were placed On weighing trays cOntaining 0.9% NaCl and kept On ice. Crude Tissue HOmOgenate Preparation The enucleated eyes frOm all animals were trimmed free Of muscle and weighed. An incisiOn was made at the cOrneal limbus and the iris and cOrnea remOved after cutting alOng the limbus. The lens was lifted frOm the eyecup with a fOrceps. The retina was scOOped away frOm the sclera, blotted to remOve large pOrtiOns of vitreOus, then placed in a weighing tray. The vitreOus adheres to the rainbOw trOut retina and is difficult to remove entirely withOut damaging the retina. TherefOre the retinal 37 samples cOntained neural retina, pigment epithelium, chOriOcapillaris and acne vitreOus. The isOlated retinal samples were weighed and placed in cOld, isOtOnic saline (500 mg:1 m1) and kept On ice in 12 x 75 mm pOlyprOpylene tubes (Bio-Rad LabOratOries, RichmOnd, CA). Liver and brain tissues were trimmed to rOughly 150-300 mg samples, blotted dry, weighed, then diluted (500 mg:1 ml) in pOlyprOpylene tubes with isOtOnic saline (4 C) and kept On ice. It was necessary to pOOl brain and retinal tissues frOm two frOgs as well as retinal tissues frOm two rats in order to have enOugh hOmOgenate to assay. When assaying these retinal samples, OccasiOnally the volume of 0.9% NaCl was increased by 0.2 ml in Order to have sufficient hOmOgenate. The tissue samples in the pOlyprOpylene tubes were then hOmOgenized Over ice using a sOnifier cell disruptOr fOr 1 min at a pOwer setting Of 20 watts (Heat Systems CO., Melville, L.I., NY, MOdel W—185-C, fitted with a Micro Tip, BransOn Instruments, Inc., Danbury, CT). TO further disrupt the cell membranes, tissue hOmogenates were twice frOzen in a dry ice-ethanol bath. Thawing was dOne in a 50 C water bath (Lab-Line Instruments, Inc., Meerse Park, IL). The hOmOgenates were then centri- fuged at 13,200 g fOr 30 min at 4 c (SOrvall, Inc., Model RC2-B, 33-34 rOtOr, NewtOwn, CT). The supernatant (S1) cOntaining the sOluble enzyme was then pipetted Off, placed in a clean pOlyprOpylene tube and re- centrifuged using the abOve settings fOr an additiOnal 20 minutes. This supernatant (S2) was diluted with sOdium phOsphate buffered saline (pH 7.8), in Order that the final cOncentratiOn wOuld fall in the range Of the standard curve. (SuperOxide Dismutase Assay, Elstner and Heupel, 1976; Appendix I). 38 Assay PrOcedure Standard bOvine erythrOcyte superOxide dismutase (BESOD), E.C. No.1.15.1.1 was Obtained frOm Sigma Chemical CO., St. LOuis, M0 (#3-8254; Lot 16C-8030; 2900 U/mg protein; 2880 U/mg sOlid; assayed as per McCOrd and FridOvich, 1969). Standard superOxide dismutase sOlu- tions were prepared by adding distilled water to 10.1 mg BESOD to make 500 ml. This sOlutiOn was divided intO 5 ml aliquots, frOzen Over dry ice, lyOphilized and stOred desiccated at 0 C until used. Each vial therefOre cOntained 290 SOD units as assayed by Sigma. Two standard curves were run fOr each tissue determinatiOn. Standards were prepared frOm the lyOphilized aliquots Of BESOD by diluting with 1 ml distilled water. FrOm this sOlutiOn two 300 pl samples were remOved using a MicrO/PettOr (Scientific Manufacturing Industries, Inc., Berkeley, CA) and placed in separate pOlyprOpylene tubes. One Of the 300 ul samples was then diluted to 1.5 ml with phOs- phate buffered saline and was referred to as the Original standard (OS). The secOnd 300 pl sample was diluted with 600 pl 0.9% NaCl and treated the same as tissue samples beginning with the sOnificatiOn step. After treatment phOsphate buffered saline was added to bring the volume up to 1.5 ml. This standard was referred to as the prOcessed standard (PS) and was used fOr all tissue cOmparisOns. In additiOn, cOmparisOn Of these standards would prOvide an indication Of enzyme recOvery. Tissue superOxide dismutase activity was assayed On the basis Of its ability to inhibit the formation of nitrites frOm hydrOxylamine at rOOm temperature (Elstner and Heupel, 1976). A xanthine-xanthine Oxidase system generated the superoxide radicals (0;). The 02 Oxidized hydroxylamine to nitrites at neutral pH. The nitrites were coupled to 39 sulfanilic acid and N,N-dimethyl-1-naphthylamine, which produced an azo compound such that high levels of 0; produced a dark violet color. Competition for the superoxide radicals by SOD limited the amount Of color produced. A dose response curve for each standard and tissue sample was gener- ated over at least 4 concentrations with duplicate samples. In 16x125 mm disposable culture tubes (Scientific Products, Division Of American Hospital Supply Corporation, McGraw Park, IL) the apprOpriate standard or tissue dose was diluted to a volume Of 1.7 ml with phosphate buffered saline (pH 7.8). The assay reagents were added bringing the final vol- ume of the incubation media to 4.15 ml. After the color develOped it was necessary to include a 10 min centrifugation at 2500 rpm in order to remove the turbidity (International Centrifuge CO., MOdel SVB, Size 1, rotor-240, Boston, MA). The homogenate was then decanted into a disposable cuvette (Kartell Disposable Cuvettes, Markson Science, Inc., Del Mar, CA). Percent transmission values were determined spectro- photometrically at a wave length of 530 nm (Beckman Model DB-G Grating Spectrophotometer, Beckman Instruments, Inc., Fullerton, CA). Two blanks are used in this assay. The reagent blank contained all of the reagents except xanthine and superoxide dismutase and was used to set the percent transmission ($T) scale. The standard blank lacked only superoxide dismutase and was used as the baseline for percent inhibition calculations. Preliminary data were subjected to several curve fitting routines as described in the Hewlett-Packard Model-65 STAT PAC1, in order to determine the most applicable transformation. The coefficient of determination (r2) was used as a measure of the "goodness of fit" of 40 the regression line. As pmeviously mentioned the crude tissue homog- enates were subjected to a parallel line ANOVA to check for interfering substances. The applicability’ of these statistics is dependent on a linear response. The transformations of the data with their effects on the curve are described more fully in the Results section. Statistical Analysis The variation about the dose response lines was estimated using the ANOVA approach to regression analysis as described by Neter and Wasserman (1974). Prior to running the ANOVA all lines were tested to ensure linear regression functions and significant positive lepes (i.e. F-tests). The regression lines were then examined for the presence of outliers according to Gill. (1978). If an outlier was detected it was eliminated and a new regression line formed. Comparison of the standard and sample regression lines followed from the analysis of variance. The error variances of the lines were tested for homogeneity in order that the comparison for parallelism be appropriate. For a valid assay, the crude homogenates should differ from the standards only in dilution and should be parallel. Expression of Activity One unit of SOD activity was defined as that amount of standard or tissue homogenate which inhibited maximum color formation by 50%. The volume of standard added at the 50% inhibition point was termed the ED50(1). The concentrations of the standards (SOD units/pl) could than be obtained from the reciprocal of the EDSO' The potency of the tissue homogenate was determined from the calculation ED50(1)/ED5O(2), where ED50(2) represented the volume of tissue homogenate added to get 50% 41 inhibition. The SOD concentration of the tissue homogenate in units/pl is the product of the standard concentration and the potency. In order to make comparisons between samples, the amount of wet tissue weight at ED50(2) was determined as well as the tissue protein concentration using the method of Lowry (Oyama and Eagle, 1956; Appendix II). The SOD activity could then. be expressed as units/mg 'wet tissue ‘weight or units/mg tissue protein. Oxygen Consumption Studies Tissue Preparation Retinas, livers and brains from the teleost, amphibian and mammal were removed in the manner previously described excluding the perfusion step. Liver and brain samples were cut into 0.2 mm slices using a MOdel 7120A Stadie-Riggs Tissue Slicer (Arthur H. Thomas CO., Philadelphia, PA), blotted dry and weighed. Retinas were isolated, blotted to remove as much vitreous as possible, then weighed and assayed whole. In the case of the frog and rat two retinas were pooled. Dry Weight Determinations Percent dry weight was measured for liver and brain from the three species. Samples were trimmed, blotted dry and placed on preweighed aluminum planchets. They were then weighed and kept in a drying oven (Blue M Electric 0o., Blue Island, IL) at 103 0. After 24 h the plan- chets were reweighed and percent dry weight determined and averaged for the respective organs. Assaerrocedure Oxygen consumption levels were determined on the tissues using a YSI Model 53 Biological Oxygen Monitoring System (Yellow Springs Instrument CO., Yellow Springs, OH) attached to a two channel strip chart recorder 42 (Linear Instruments MOdel 486, Linear Instruments Corp., Irvine, CA). The samples were placed in 3 ml of modified Krebs-Ringer-Phosphate solution containing glucose and saturated with room air (Appendix III). Temperature of the media was kept stable at 15:0.2 C or 37:0.2 C with a Constant Temperature ‘Water 'Bath (Haake, Inc., Model D1, Temperature Control Equipment Div., Saddle Brook, NJ). Oxygen uptake by the tissues was measured by recording a predetermined linear (e.g. 90%-->70%) change in p0 over time. 2 Calculations of oxygen consumption were based on the solubility coefficient of oxygen in Ringer solution (0.0340 ml 02/ml fluid at 1 atm of the gas at 15 C; 0.0239 ml OZ/ml fluid at 1 atm at 37 C; Umbreit, Burris and Stauffer, 1964). After oxygen consumption determinations were made, the samples were removed and sonified as previously stated. The homogenates were then centrifuged for 10 min at 2500 rpm (Inter- national Centrifuge Co., Model SVB, Size 1, rotor-240, Boston, MA). Protein determinations (Appendix II) were done on the supernatant by the method of Lowry (Oyama and Eagle, 1956). Oxygen consumption of the tis- sues was expressed as microliters of oxygen consumed per hour per milli- gram protein at 1 atm, and all values were corrected to standard temper- ature and pressure dry (STPD). RESULTS Plotting Routine Percent transmission values for each standard and tissue sample were converted to optical density and the difference from the standard blank (AOD) computed. A dose response curve was generated over at least four concentrations of SOD. The AOD was plotted on the ordinate as a func— tion of the dose of superoxide dismutase. From several curve fitting routines a logarithmic curve fit was confirmed (Figure 1). Logarithmic transformation of the dose altered the curve such that it fit a straight line in the form y = mx + b (Figure 2). The assay response may also be considered as the percent inhibition of maximum color produced. There- fore the difference in optical density caused by the addition of SOD could also be expressed as percent inhibition vs. ln-dose SOD. The result again is a linear ln-dose response curve shown in Figure 3. Percentage values however may have introduced a statistical bias, this was avoided by converting the percentage values to probit response using readily available statistical tables to normalize the data (Figure 4). In order to use the parallel line assay, it was first necessary to determine the linear portion of the dose response curve. A concentrated SOD solution was serially diluted and assayed. The dose response curves from two experiments were combined and the results shown in Figure 5. The linear portion of the curve was found between 4.0 and 6.0 probits or 43 .GOflpmppsmocoo some as mGOflpOOHHmon 03» you swamp on» sea: convoam ops spam .oom ovhoounphpo 0=H>on mo mcoflpmuuqoocoo mowmwouocw £33 £539 opwoswam one sop.“ .3356 Hwoflmo oz» Gun 03:30 93 no pain < .F opsmwm 45 H ouswwm :53 N¢ndé com o. m m - m 0 ¢ n N _ P p P d d J ’ ‘ ow as ou 0.390 one. :3 384 com ~23? 3.. .w¢._ V o O m 1 . O and. u. l o. ..._- .83 .6.» sad Pm.“ 46 .. onsmfim scum convoamop was spun .mcofiuwnusooaoo omw958mfic ovwxopomsm osfi>on 0:» mo snpwuwmoa vocoo .m> xcwan vpmcsmpm 0:» Beam minnow 13390 on» Ga omsmno on» mo no.3 .4 .m opswflm 47 N ouzwam $5: Nutn. com mmoo c. awooo =E\D Nvmdv crow 2.. to m o m w _ o9. - . o.o . o... . ow . ”To :3 8.34 m230> 6.. :5 v O 0 m .fmw.— .mml emu 48 .— opsmwm 90pm coupoamop ops spam .nommm he owes ca cocoo on» .m> oswsloNopch: mo :ofluwcwxo on» ma sowposcopm poaoo 3:2.”me mo :owfipwnsfi unmouom 0:» mo vein 4 .m opsmwm 49; :63 Nut. 3 oowflmoo c. ouooo m whamfih 2.5: «3.9 o_ as 2.. e m e a a _ ow... .on. at . o.~ o._ o :3 38¢ u:a..o> l. sou 40¢ madam. : 5.00 :00 e3. NOLLIBIHNI 96 50 .vumccwpm mom Ofipmoonspmpo oeseop or» see noesm as Aomnmv soaescsrca mow pm omoo pcoam>wswo one .mocfla cognac no steam who me>pOpqfl monocwmsoo Rmm one .F 85mg 895 8030.30.“ 0.3 span .9035 no ones :a cocoo .m> oncommop panonm on» mo uoag < .¢ opsmfim 51 q muswfim :5: Nuts now $8 5368 :53 «one 2 mm a o m c n m _ om... .ow . owm . o.~ o._ . new: .23 entomoulvm .3322 m230> A $24.. . ow \ ?\ I I 1 I I I 1 2 I I I i E .5 ID SllBOHd .xbm L66 52 .oameHHmmw no: noooumgw HOOHpmfipwpm was» scans. pom omens .psoswa on» washes op coma mos nmwpw mane .mommm mo soapoppcoosoo some new monas> mo swamp ms» psomoumop mumn Hmogpo> .mommm mo omov ca 80800 on» .m> oncogmop panopm 0:» mo poam < .m opsmfim 53 m mpswwm :63 2E5 buy now mmoo :. owooo hwodv 000. 00. o. _ .To.n m i _ .5. sod 40.0 SAIQOBd 54 over a range of 15% to 85% inhibition. Points outside that linear zone were eliminated. Assay Verification The value taken from the curve in Figure 4 represents the equivalent dose necessary to inhibit 50% of the maximum color formed (EDSO) and is given in terms of 111 of stock BESOD added to the reaction vessel. If the volume in which the standard was diluted (1.7 ml) is taken into con- sideration, the activity can be reported as Sigma units. A value of 2.2 Sigma units was reported as that amount necessary to bring about 50% inhibition. A more complete interpretation of these results is included in the Discussion section. Assay Sensitivity The sensitivity of the assay was appraised by testing the response in the presence of a suspected interfering substance. Preliminary data collected in our laboratory suggested that ascorbate was present in detectable amounts in the tissues studied. Ascorbic acid could react competitively with SOD for the superoxide anion or possibly generate 0; anions at high concentrations. A vial of superoxide dismutase standard was prepared and two 300 ifil samples of that standard were contaminated with ascorbic acid to final concentrations of 2.4 mM and 4.8 mM ascorbate. Catalase (22 g/ml) was also included in the reaction mixture to prevent the ascorbate from being oxidized by H202 which is formed in the xanthine-xanthine oxidase reaction (Figure 6). Deviations from linearity were not observed for the lower dose of ascorbate. However the higher concentration of ascorbate clearly altered the response of the "contaminated standard." This implied that 55 .opspxfla :ofipooon on» :H novsaonfl ome was AHs\w mmv _ omwamumo .mpwnuoomo mo mHo>mH m50flpw> spa: cepwcfiewpnoo who: vaccwpm 93 mo mposvaaw 039 .mommm mo omow :H coooo on» .m> oncommop wagons on» no poam < .w opswflm 56 o ouswwm :53 ma. new moon e. 358 :53 «end. m... 0.. on u” m. m w m m _ I q I I I I I OI¢ I oIN o‘— on. oo. 8 A. L.IJo... ..o.m ”.2883 252205 c 55.68... Eons; . 2.35.5 com o .oo scaly. SllBOHd 57 significant levels of interfering substances would be detected by the parallel line ANOVA. Other factors were also found to interfere with the superoxide dismutase assay. On observation, nonperfused, untreated, crude tissue homogenates of rainbow trout retinas were tinted red. When assayed these same tissue homogenates often yielded negative SIOpes such that some percent transmission values read lower than the standard blank. The presence of hemoglobin appeared responsible for the aberrant lines. This led to a series of experiments designed to establish a protocol for eliminating the hemoglobin. Preliminagy Tests to Establish Experimental Protocol Two groups of rainbow trout were used; prior to enucleation one group was saline perfused one was not. Retinal homogenates from each of these groups were further divided. One group was vigorously shaken with (0.5 v/v) chloroform prior to the second centrifugation at 13,200 g, and one did not receive the chloroform treatment. The statistical param- eters generated from the ANOVAs for f0ur treatment groups were compared (Table 2). 'When the retina ‘was neither 'perfused nor treated ‘with chloroform (Group 1), the superoxide dismutase activity could not be accurately determined. A The presence of hemoglobin altered the tissue extract regression line such that none were parallel to the processed standards and in some cases the lines did not even register a signif- icant positive s10pe. Regression lines for tissues from the other three groups were parallel to their respective standard, and their regression parameters did not differ significantly, implying neither procedure altered the assay response. Although there were no significant dif- ferences in the activity (units/mg protein), there was a trend for the 58 Table 2. Comparison of treatments for rainbow trout retina. TISSUE - ED50 UNITS/mg UNITS/mg TREATMENT INTERCEPT SLOPE (1;) TISSUE PROTEIN GROUP 1 4.25 0.177 131.39 0.84 74.83 ‘ 10.39 10.103 191.71 10.76 159.52 GROUP 2 2.65‘ 0.538’ 48.20 0.12 38.75 10.31 10.043 15.09 10.01 12.56 GROUP 3 2.73’ 0.575‘ 50.34 0.13 40.13 10.16 10.030 15.26 10.02 16.75 GROUP 4 2.89' 0.549’ 47.68 0.14 28.21 10.13 10.027 16.27 10.02 17.38 N = 4; (X1SE) 'l' Significantly different from Group 1 at a = 0.05 using Students 11-156813. GROUP 1 8 Retina; nonperfused, without chloroform treatment GROUP 2 8 Retina; nonperfused, with chloroform treatment GROUP 3 = Retina; perfused, with chloroform treatment GROUP 4 Retina; perfused, without chloroform treatment 59 nonperfused, chloroform-treated retinas to have a higher activity. Perhaps chloroform extracted some of the tissue proteins which would lead to an inaccurate measurement of SOD protection/mg tissue protein. In conjunction with these experiments standard samples (Processed Standards) were treated in. a manner similar to the tissues (i.e., diluted with saline and sonified) to test whether this treatment had any effect on enzyme recovery (Table 3). Statistical parameters indicated that this process had no effect, e.g., the s10pes and Y-intercepts were not significantly different from those for Group 1 implying full re- covery of the standard. The standards treated with chloroform (Group 2) also did not Show a significant change in the slope but the concentra- tion and ED50 response for this group were significantly different from those not treated with chloroform (Groups 1 and 3). Recovery of the enzyme was only half of that obtained when chloroform was not used. Figure 7 shows representative lines from the three groups with their respective ED5O responses. Based on this information and the previous results, it was decided that a whole body perfusion without chloroform treatment was the desired protocol. This eliminated the hemoglobin interference and ensured the measurement of endogenous tissue superoxide dismutase activity. Enzyme Recovery All assays fOr SOD involved the determination of activity from the BESOD standards (original and processed) and each of the tissue homog- enates. Although 100% recovery of the original standard was implied (Table 3) rather large variations were noted for the processed stan- dards. Therefore, two standards (OS and PS) were run fer each assay to account for any loss of activity from enzyme treatment. Recovery of 60 Table 3. Comparison of treatments for BESOD standard. STANDARD ED50 SOD 00Nc.(x10‘2) TREATMENT INTERCEPT SLOPE (ul) (Units/ul) % RECOVERY GROUP 1 2.90 0.586 36.13 2.8 10.09 10.014 12.94 10.2 GROUP 2 2.50 0.580 76.32“ 1.4' 48.5 10.17 10.028 110.58 10.2 13.0 GROUP 3 2.83 0.604 37.73 2.8 100.0 10.13 10.038 14.96 10.3 114.8 N - 4; (X1SE) 8 Significantly different from Groups 1 AND 3 at a = 0.05 using Students t-test. GROUP 1 = Original BESOD standard (0S) GROUP 2 . Processed BESOD standard; with chloroform treatment (PS-C) GROUP 3 = Processed BESOD standard; without chloroform treatment (PS) 61 .unoccspm some pom empuomop omHo was mosam.’ 0m mm 04230930.“ 0:9 .ucospwou» shomopoano saw: cpscsmpm 80300099 I 01mm was cpovompm 6630093 A mm .2823»... anaemic a mo .mucospoop» moons; poems menoccspm momma mo once :a cocoo oz» .m> omqommou $8.93 93 mo pea o>33qomonmou 4 .> enamfim 62 com 0.00b0 ’ # 1 09m 8 . ow - a 36.89.." 3%.. .88 0:0... 00 wmoo :. omaoo s odour. n? m N. . D L I C J . 0.? ON I 1. - ".auowboom W 40.? 81.18086 .. 0.9 63 BESOD activity frOm the processed standards fOr all experiments averaged 88%, a respectable figure yet significantly less than full recOvery. The 12% loss in BESOD activity is of impOrtance only if One wishes to express the activity in terms of Sigma units. It was cOnsidered neces- sary to run both standards (OS and PS) with each assay to accOunt for dilution error or any particular daily treatment effects and data fOr crude tissue homOgenates were cOmpared to those for prOcessed standards for parallelism. SOD Results Comparison of Tissue SOD Activities Within Species SuperOxide dismutase activities were measured in crude retinal homogenates of the rainbow trout. A comparison Of the retinal activity was made to neural tissue (brain) and nOn-neural tissue (liver). The parallel line statistical apprOach is depicted in Figure 8 which shows a representative rainbow trOut retina and the processed BESOD standard. The activities were expressed as SOD units/mg tissue protein (Figure 9; Table 4a). No difference in activity was fOund between the retina and brain. The liver SOD activity however, was apprOximately three times that fOund in the neural tissues. The SOD distribution in the tissues of the frOg and LOng-Evans rat were also determined (Figure 9; Table 4s). No significant differences were found between the retinal and brain SOD .activities in either species. The highest activity was recOrded in the rat liver which was found to be significantly abOve the levels in the neural tissues. Liver SOD activity in the frOg was significantly higher than that fOund in the retina, but was not statistically greater than the activity in the frOg brain. 64 .mom we omoo :H oocoo .m> newnopm mm conmwum one noncommop one .cpwcswpm aommm commoooum 0:» ca conmmsoo ma weapon poop» sonoflwp o>flpwpcomonmou < .hufi>wpow omo»:8mfic ocwxopomsm mnwhwmmw pom nowonmmm Hoowpmfiuopm mafia Hoaawpwm 0:» mo sampm < .m ousmwm 65 ror.'log~..l{ 1.. .. . com 38 e. 858 s as... f a n o 68 S. moo. cm or 8 o. \\ \t 258.63 £233 000m 000m .ssnw Qo 00.0.. N. 3.6. m. -0.-. O--------.------d 83 I: 5.0 SllSOUd 66 .AU Nummv oHSpouomsov aoop no can ouo: whoom< .mooonpoopom a“ ao>fim ouo vomomoo hopes: one .mpofiuo choosopm H onoos opo :Hopopm ws\opfias mom on coupomou mosao> .vop coo mosh .wsou» sonoflop on» a“ monomfip osowno> scum oofiugfloo ooopssoflc ogxoaomsm mo soofiuomsoo .m ousmfim 67 .3. at .0. h<¢ 00cm .3. 8.. a. m ouswwm #30:... Mew 2.0 EN. 0 .. 00 .400. i000 .. 000 2.3.0 on mw>3 .5 42.5: w... .000 “108°“. WI'IWO 00$ 68 Table 4. Superoxide dismutase activities in the liver, brain and retina from a teleost, amphibian and mammal. a. LIVER BRAIN RETINA TROUT 78.28 1 6.75( 7) 26.33p;__2.81(10) 25.75#;7p1.89(23) FROG 48.85 1 5.96(10) 28.97 11 8.60( 4) 18.64 1 2.51( 5) RAT 451.72 1 63.81( 7) 106.6];123.04( 5) 69.80 1 11.66( 6) b. RAT TROUT FROG RETINA 69.80 1 11.66( 6) 25.7511 1.89(23) 18.64 1 2.51( 5) LIVER 451.72 1 63.81( 7) 78.28 1 6.75( 7) 48.85 15.96(10) BRAIN 106.68 1 23.04( 5) 26.30 1 2.81(10) 28.97 1_ 8.60( 4) Means not underscored by the same line (within groups) are significantly different (pg 0.05). Enzyme activity expressed as units of SOD/mg tissue protein. 69 COmparison of Tissue SOD Activities Between Species Trout retinal SOD activity was lOwer than that in the liver and similar to the activity in the brain. It was pOstulated hOwever, that it wOuld be significantly higher than the retinal levels in the frOg and especially the rat, since both Of these species have, to varying degrees, been shown to be susceptible to Oxygen t0xicity. COntrary to what was expected, there was no difference between frog and trout retinal SOD activity, and the activity in the rat retina, rather than being lower than the other two species, was significantly higher. The same was true for the SOD activity in the brain and liver; no differ- ences were found between the frOg and the trOut tissues, yet the activity for rat tissues was cOnsiderably higher (Table 4b; Figure 10). Oxygen ConsumptiOn Studies The high SOD activity in the tissues of the LOng-Evans rat appeared equated to the metabolic demands of this species. At lOwer envirOn- mental temperatures pOikilOthermic animals decrease their metabOlic demands. Oxygen consumptiOn wOuld be expected to be lOwer for the poikilotherms than fOr the homeotherm. The Oxygen cOnsumptiOn Of tis- sues was measured and the data are shOwn in Table 5a. NO difference was fOund in the Oxygen cOnsumptiOn of the retinas and brains within any Of the species. In trOut and frogs but not the rat, the Oxygen cOnsumptiOn of the neural tissues were significantly greater than that Of the liver. When the Oxygen cOnsumptiOn level Of each tissue was cOmpared amOng species it was fOund that liver oxygen consumptiOn levels were no dif- ferent in any Of the species. The levels in the neural tissues Of the trOut and frOg also did not shOw any significant differences. High Oxygen cOnsumption was fOund for the neural tissues Of the rat, such 70 .oomonpcouog ca co>fim ouo coeoooo popes: one .3 mummv opgopomeop soon no eoposcooo ono: memom< .Ammumv :Hopopm me\oufias now no comoopmxo opo woo m opswfie :H copuomon omega mo oeoo one. ohm mood; .oaosflso Hopcosfipomxo 983.2; Eon.“ :Hopn coo no.3.” .ocflou one ca moguggoo omoussofic ocwxopomso mo comfiuomaoo .o_ opsmfim 71 2.5.3 ...AQ.. .va A0: mu>_.._ ..At. 8: At 8H osswaa (Zr—hm ..o. s. as .. 0 ..on S 0 ..oo. o n mw L... n +00¢m b o W 400?“... ._.