ABSTRACT PARAQUAT PERINATAL TOXICITY AND A PROPOSED MECHANISM OF ACTION INVOLVING LIPID PEROXIDATION BY James Stanley Bus Paraquat, l,l'—dimethyl-4,4'-bipyridylium dichloride, is a widely used herbicide effective against broad leaf weeds and~ grasses. The purpose of this investigation was to determine the effect of chronically administered paraquat on the postnatal develop- ment of mice and to determine in Vitro and in vivo the mechanism of paraquat toxicity in mammalian systems. Paraguat,in the drinking water of pregnant Swiss-Webster mice at 50 and 100 ppm at day 8 of gestation and with continued exposure of the pups to 42 days postnatally, did not at any time alter the average litter body weights compared to controls. One hundred ppm, but not 50 ppm, paraquat significantly increased postnatal mortality in a biphasic manner, with an initial 33% increase in mortality reached between days 7 and 28 postnatally. Mortality was 67% at 42 days after birth. Pulmonary lesions consisting of edema, thickening of alveolar septa, and alveolar consolidation were observed in the 100 ppm but not 50 ppm paraquat treated mice at 42 days postnatally. Liver and kidney were not affected. 5 James Stanley Bus Concentrations of paraquat in lungs of 100 ppm paraquat treated mice were less than 0.2 ug/g tissue at 42 days postnatally and non- detectable in lungs of 50 ppm paraquat treated mice. Both 50 and 100 ppm paraquat treated mice were more sensitive to 100% oxygen exposure at 42 days after birth than controls as determined by a significant reduction in the median time to death (LTSO) in oxygen. The LT for control animals in oxygen was 160 50 hours versus LTSO's of 108 and 40 hours for 50 and 100 ppm paraquat treatments, respectively. Sensitivity to 100% oxygen was detectable in 100 ppm but not 50 ppm paraquat treated mice at days 1 and 28 postnatally. In 42 day old mice, the LT after 3100 mg/kg ip 50 bromobenzene (LDBS) was 4.2 and 3.2 hours in 50 and 100 ppm paraquat treated mice, respectively, compared to the LT50 in controls of 20.0 hours. Under anaerobic conditions in vitro, paraquat was reduced to the blue-colored free radical by a single electron reduction reaction catalyzed by mouse lung microsomes and NADPH. The reaction was inhibited by antibody to NADPH-cytochrome c reductase, which suggested that microsomal NADPH-cytochrome c reductase catalyzed the paraquat reduction. Paraguat, when incubated aerobically with NADPH, NADPH-cytochrome c reductase and extracted microsomal lipid, initiated lipid peroxidation in a concentration dependent manner with a maximum 227% increase in malondialdehyde at 10-3M paraquat. Paraquat-induced in vitro lipid peroxidation was inhibited by the superoxide radical scavenger superoxide dismutase and a singlet oxygen trapping agent, 1,3-diphenylisobenzofuran. ‘Both agents James Stanley Bus together inhibited paraquat-induced lipid peroxidation to a greater degree than either agent alone, suggesting sequential intermediates of superoxide radicals and singlet oxygen in paraquat-induced lipid peroxidation. Paraquat-induced in vivo lipid peroxidation was studied by determination of the paraquat 7-day ip LD50 in mice fed diets deficient in the antioxidants selenium or vitamin E. Selenium and vitamin E deficiency significantly reduced the paraquat L050 to 10.4 and 9.2 mg/kg, respectively, compared to 30.0 mg/kg in labora— tory chow fed controls. Supplementation of selenium deficient diet with 0.1 and 2.0 ppm selenium and vitamin E deficient diet with 45 and 1500 mg vitamin E per kg diet returned the paraquat L050 to the control value. Liver glutathione peroxidase (a selenium dependent enzyme) activity in selenium deficient animals was 16% of control activity. Vitamin E deficiency was confirmed by the dialuric acid erythrocyte hemolysis test. The paraquat LDSO in mice was reduced to 9.4 mg/kg by pretreatment with diethyl maleate at 1.2 ml/kg ip, 30 minutes before paraquat. Diethyl maleate decreased the reduced glutathione (GSH) concentrations in liver and lung. GSH is a substrate for glutathione peroxidase, which detoxifies lipid hydroperoxides formed during membrane lipid peroxi- dation. Paraquat, 30 mg/kg ip, also significantly reduced liver concentrations of the water soluble antioxidant GSH 21% 24 and 36 hours after treatment and lung concentrations of lipid soluble anti- oxidants 53 to 59% between 12 and 96 hours after paraquat. Male Sprague-Dawley rats treated with 100 ppm paraquat in drinking water had a 68% increase in glucose-6-phosphate dehydrogenase James Stanley Bus and an 89% increase in glutathione reductase activity in lung tissue. Pretreatment of rats with 85% oxygen for 7 days, which induces tolerance to subsequent 100% oxygen exposure, increased the resistance of rats to a toxic (45 mg/kg ip) dose of paraquat, reflected by a significant increase in the paraquat LTSO to 50.0 hours as compared to 26.0 hours in nonpretreated rats. Mice pretreated with 0.1% w/v phenobarbital in the drinking water for 10 days and with continued exposure after paraquat injection also were protected against paraquat toxicity. The paraquat LD was 46.0 mg/kg with phenobarbital pretreatment and 50 30.0 mg/kg in controls. Phenobarbital may compete for electrons necessary for paraquat reduction. The in vitro and in vivo experiments suggest that lipid peroxidation of cellular membranes may be involved in paraquat toxicity. Paraquat may undergo a cyclic reduction-oxidation in vivo, with subsequent formation of superoxide radicals. Superoxide radicals can nonenzymatically dismutate to singlet oxygen, which reacts with unsaturated lipids associated with cell membranes to form lipid hydroperoxides. The lipid hydroperoxides can initiate the membrane damaging chain reaction process of lipid peroxidation. The result of such a process is loss of cellular function and integrity. PARAQUAT PERINATAL TOXICITY AND A PROPOSED MECHANISM OF ACTION INVOLVING LIPID PEROXIDATION BY James Stanley Bus A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pharmacology 1975 ACKNOWLEDGMENTS I would like to express my sincere appreciation to Dr. James E. Gibson for his continued guidance, constructive criticism and encouragement over the course of my graduate education. Special thanks are offered to Dr. Steven 1). Aust for his guidance in many aspects of the dissertation research. I also wish to acknowledge the contributions of committee members Drs. Theodore M. Brody and Jerry B. Book. I would like to thank Dr. Thomas C. Pederson and Mr. John A. Buege of the Department of Biochemistry for providing preparations of extracted microsomal lipid, antibody to NADPH~cytochrome c reductase, and superoxide dismutase; Dr. Duane Ullrey of the Animal Husbandry Nutrition Laboratory for providing dialuric acid; and Mr. John P. Hitchcock of the Animal Husbandry Nutrition Labora- tory for assaying diets for selenium. I also offer my appreciation to Mr. Stuart Z. Cagen and Mr. Mark K. Olgaard for their technical assistance. These studies were supported by NIH Research Grant E550060 NIH Training Grant GM01761 and by the Michigan Lung Association. ii TABLE OF CONTENTS ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . METHODS RESULTS Objectives. . . . . . . . . . . . . . . . . . . Chemistry of Lipid Peroxidation. . . . . . . . . Biochemical Consequences of Lipid Peroxidation . Biological Defenses Against Lipid Peroxidation . Examples of Oxidant Agents that May Induce Lipid Peroxidation . . . . . . . . . . . . . . Paraquat . . . . . . . . . . . . . . . . . . . . Purpose. . . . . . . . . . . . . . . . . . . . . Animals. . . . . . . . . . . . . . . . . . . . . Toxicity of Chronically Administered Paraquat in Developing Mice . . . . . . . . . . . . . . Paraquat In Vitro Oxidation-Reduction. . . . . . Paraquat-Induced In Vitro Lipid Peroxidation . . Tissue Malondialdehyde Concentrations in Mice after Acute Paraquat Treatment . . . . . . . . Effect of Nutritional Deficiencies on Paraquat— Induced Lethality. . . . . . . . . . . . . . . Tissue Reduced Glutathione and Lipid Soluble Antioxidant Concentrations after Acute Paraquat Treatment . . . . . . . . . . . . . . Interaction of Paraquat with the GSH Peroxidase System Enzymes . . . . . . . . . . . . . . . Interaction of Phenobarbital with Paraquat . . . Statistics . . . . . . . . . . . . . . . . . . . O O O O O O O I O O O O O O O O O I I O O O O O Toxicity of Chronically Administered Paraquat in Developing Mice. . . . . . . . . . . . . . . . Paraquat In Vitro Oxidation-Reduction. . . . . . Paraquat-Induced In Vitro Lipid Peroxidation . . Tissue Malondialdehyde Concentrations in Mice after Acute Paraquat Treatment . . . . . . . . iii Page ii vii UJOUJH 17 32 33 33 33 39 41 42 42 46 48 50 51 52 52 67 74 83 Page Effect of Nutritional Deficiencies on Paraquat-Induced Lethality . . . . . . . . . . . . . 84 Tissue Reduced Glutathione and Lipid Soluble Antioxidant Concentrations after Acute Paraquat Treatment . . . . . . . . . . . . . . . . . 96 Interaction of Paraquat with the GSH Peroxidase System Enzymes . . . . . . . . . . . . . . . . . . . 98 Interaction of Paraquat with Bromobenzene. . . . . . . 98 Interaction of Phenobarbital with Paraquat . . . . . . 102 DISCUSSION. . . . . . . . . . . . . . . . . . . . . . . . . . 109 Toxicity of Chronically Administered Paraquat in Developing Mice . . . . . . . . . . . . . . . . . 109 In Vitro Investigations of the Paraquat Mechanism of Action. . . . . . . . . . . . . . . . . . . . . . 115 In Vivo Investigations of the Paraquat Mechanism of Action. . . . . . . . . . . . . . . . . 120 Interaction of Paraquat Toxicity with Bromo- benzene and Phenobarbital. . . . . . . . . . . . . . 130 Summary and Conclusions. . . . . . . . . . . . . . . . 134 BIBLImRAPHY. O O O O O O O O O O I O O O O O O O O O O O O O 139 iv Table 10 ll 12 LIST OF TABLES Water consumption by developing mice treated with paraquat from day 28 to day 42 postnatally . . . . Effect of 100 ppm paraquat on postnatal mortality in mice when administered in the drinking water during selected periods of development . . . . . . Malondialdehyde concentrations in tissues of mice treated with 100 ppm paraquat from day 8 of gestation to 42 days postnatally . . . . . . . . . . . . . . Effect of paraquat in the water from day 8 of ges- tation to various days postnatally on the survival of mice exposed to 100% oxygen . . . . . . . . . . Effect of vitamin E on survival of one day old mice pretreated with 100 ppm paraquat on days 8-19 of gestation in 100% oxygen. . . . . . . . . . . . Effect of acute paraquat administration to one day old mice on survival in 100% oxygen. . . . . . . . Paraquat-induced in vitro lipid peroxidation . . . Inhibition of paraquat-induced in vitro lipid peroxidation by superoxide dismutase and 1,3- diphenylisobenzofuran. . . . . . . . . . . . . . . Malondialdehyde concentrations in tissues of mice after acute paraquat treatment . . . . . . . . . . Alteration of paraquat single dose 7-day intra- peritoneal L050 in mice by various selenium or vitamin E diets and diethyl maleate pretreatment . Liver and lung glutathione peroxidase activity in mice after 5 weeks exposure to selenium deficient, 0.1 ppm and 2.0 ppm selenium supplemented diets. . Reduction of reduced glutathione (GSH) in mouse lung and liver by diethyl maleate treatment. . . . Page 55 58 59 62 66 66 81 82 83 87 89 9O Table Page 13 Dialuric acid hemolysis of erythrocytes from mice fed various vitamin E diets. . . . . . . . . . . . . . 95 14 Body weights of mice fed various selenium or vitamin E diets for 5 weeks. . . . . . . . . . . . . . 95 15 Liver and lung reduced glutathione (GSH) after acute paraquat treatment in mice . . . . . . . . . . . 97 16 Lung and liver lipid soluble antioxidants after an acute dose of paraquat. . . . . . . . . . . . . . . 97 17 Activity of glutathione peroxidase system enzymes in rats after 100 ppm paraquat in the drinking water for 3 weeks. . . . . . . . . . . . . . . . . . . 101 18 Effect of 7-day 85% oxygen pretreatment on surVival of rats after a toxic dose of paraquat . . . . . . . . 101 19 Effect of paraquat in the water from day 8 of gestation to 42 days postnatally on the survival of mice treated with bromobenzene. . . . . . . . . . . 102 20 Liver GSH concentrations in mice administered paraquat in the water from day 8 of gestation to 42 days postnatally. . . . . . . . . . . . . . . . . . 103 21 Effect of phenobarbital (PB) pretreatment on the paraquat single dose 7-day intraperitoneal L050. . . . 103 vi Figure 10 11 LIST OF FIGURES Effect of 50 and 100 ppm paraquat in the water from day 8 of gestation to 42 days birth on the postnatal growth of mice. . Effect of 50 and 100 ppm paraquat in the water from day 8 of gestation to 42 days birth on the postnatal mortality of mice Lung tissue from 42 day old control mice treated with 100 ppm paraquat from day 8 tion to 42 days postnatally. . . . . . . Effect of 50 and 100 ppm paraquat in the Page drinking after . . . . . . . 54 drinking after 0 O O I O O O 57 and mice of gesta- . O O O O O O 61 drinking water of pregnant mice from day 8 of gestation to day 19 of gestation on the survival of one day old mice exposed to 100% oxygen for 120 hours. . . . . 65 Concentration of paraquat in fetal mouse at various times following 3.35 mg/kg ip paraquat on day 16 of gestation. . . . . organs 14C- . O O O O 0 O 69 Concentration of paraquat in one day old mouse organs at various times following 4.5 mg/kg sc 14C-paraquat o o o o o o o o o o o o o o O O I O O O O 71 Aerobic oxidation of NADPH by mouse lung micro- somes in the presence of paraquat. . . . . . . . . . . 73 Anaerobic reduction of paraquat in the presence of mouse lung microsomes and NADPH. . . . . O O O O O O O 76 Inhibition of the aerobic oxidation of NADPH catalyzed by mouse lung microsomes and paraquat by antibody to NADPH-cytochrome c reductase. . . . . . 78 Inhibition of the anaerobic reduction of paraquat catalyzed by mouse lung microsomes and NADPH by antibody to NADPH-cytochrome c reductase O O C O O O Q 80 Tissue malondialdehyde concentrations in mice after acute paraquat treatment . . . . . vii . . . . . . . 86 Figure Page 12 Lung tissue from selenium deficient and paraquat treated selenium deficient mice. . . . . . . . . . . . 92 13 Liver tissue from selenium deficient and paraquat treated selenium deficient mice. . . . . . . . . . . . 94 14 Effect of increasing doses of paraquat on lung lipid soluble antioxidants . . . . . . . . . . . . . . 100 15 Effect of 10-day phenobarbital pretreatment on elimination of 14C-paraquat from plasma and lung in mice. . . . . . . . . . . . . . . . . . . . . . . . 106 16 Effect of 10-day phenobarbital pretreatment on elimination of 14C-paraquat from liver and kidney in mice. . . . . . . . . . . . . . . . . . . . . . . . 108 17 PropOSed mechanism for the toxicity of paraquat based upon in vitro experiments. . . . . . . . . . . . 118 18 Proposed mechanism of action for the in vivo toxicity of paraquat . . . . . . . . . . . . . . . . . 124 viii INTRODUCTION Objectives The demand for increased agricultural productivity to support a growing world population has led to the introduction of many pesticidal agents over the last half century. Although many such agents have proved to be effective pest control agents, environ- mentalists and toxicologists have become increasingly concerned with the problem of pesticide residues in the soil, water, plants and in animals and humans. Of particular concern are the possible consequences to developing organisms of long term exposure to low levels of pesticides. Furthermore, as animals and man become increasingly exposed to a wide spectrum of drugs and chemicals, the potential for toxic interactions between such agents and pesticides in biological systems is significantly enhanced. Thus, the objectives of this research were twofold, both of which related to examination of the toxicity of the bipyridylium herbicide paraquat (1,1'-dimethyl-4,4'-bipyridylium dichloride). The first objective was to examine the effects of chronic ingestion of paraquat on the perinatal development of mice and to investigate the potential for the existence of subtle toxicity in these animals. The second objective was to determine interactions between acute and chronic paraquat treatment and various agents or experimental diets 2 that may influence the expression of toxicity. The agents that were examined were those that offered some potential for interaction in an environmental situation and that would yield insights into the mechanism of paraquat toxicity. The experimental data presented in this dissertation show that paraquat not only has direct toxic effects on postnatal development in mice, but also that more subtle paraquat toxicity is also present, which can be manifested when the mice are exposed to interacting substances such as oxygen. In studies of acute paraquat toxicity in adult mice, the toxicity also is shown to be enhanced in the presence of certain nutrient deficiency states in the animals. Thus, the results of these experiments provide evidence which allows further assessment of paraquat as an environmental hazard, particularly in relation to its effects on mammalian development and potential interactions with other environmental factors. Over the course of this investigation, however, a second body of information regarding paraquat toxicity became apparent, which may be of greater importance than the results directed to the assessment of the paraquat environmental hazard. This was the evidence obtained through experiments on paraquat interactions, that the mechanism of paraquat toxicity may be mediated through lipid peroxidation. The significance of suggesting a lipid peroxi- dative mechanism for paraquat toxicity is of threefold importance. First, the elucidation of a mechanism provides a basis for predicting possible interactions with environmental agents not investigated in this study. This is particularly beneficial in the assessment of paraquat as an environmental hazard. Second, a defined mechanism for paraquat toxicity provides useful information for implementation of a rational therapeutic approach in treating victims of acute paraquat poisoning. Finally, and perhaps most important, the lipid peroxidative mechanism of paraquat toxicity proposed in this study provides in vivo evidence for lipid peroxidation as a cause of tissue damage. Therefore it may serve as a model toxic mechanism for a number of other drugs and environmental agents. Furthermore, the techniques used in developing the mechanism for paraquat toxicity may be useful for investigating similar mechanisms of other agents. Chemistryiof Lipid Peroxidation Lipid peroxidation has been broadly defined by Tappel (1973) as the reaction of oxidative deterioration of polyunsaturated lipids. Peroxidation of lipids involves the direct reaction of oxygen with polyunsaturated lipids to form free radicals and semi- stable hydroperoxides, which then promote free radical chain oxida- tions (Holman, 1954; Barber and Bernheim, 1967; Tappel, 1973). The free radical chain reaction proceeds in three distinct steps (Pryor, 1973). First is the initiation process in which the radicals are generated. Second is a series of propagation reactions in which the number of free radicals is conserved. Finally, there is a series of termination reactions in which free radicals are destroyed. The 3 steps of lipid peroxidation are depicted below in a simpli- fied scheme (L = polyunsaturated lipid): Initiation: LH ; L' + (H') Propagation: L' + 02 ~ L02° LOZ' + LH > LOOH + L' Termination: L' + L' # nonradical products L' + L02' 4; nonradical products LOZ' + LOZ' > nonradical products Unsaturated fatty acids are susceptible to peroxidation as the presence of a double bond weakens the carbon-hydrogen bond on the carbon atom adjacent to the unsaturated carbon-carbon bond (Swern, 1961; Demopoulos, 1973a). Thus, these allylic hydrogens are partially "activated" and are liable to abstraction by small amounts of oxidants or initiators (initiation reaction). Molecular oxygen can abstract an allylic hydrogen, but it also must be "activated." The activated oxygen molecule, or singlet oxygen, is an electronically excited oxygen molecule in which a valence electron is shifted from its normal bonding orbital to an orbital of higher energy in which the electron spins are paired (Wilson and Hastings, 1970; Maugh, 1973). The possible role of singlet oxygen as an initiator of lipid peroxidation has been confirmed by several investigators (Howes and Steele, 1971; Dowty et a1., 1973; Pederson and Aust, 1973). In vivo, singlet oxygen appears to occur from nonenzymatic loss of an electron from the single electron reduced state of oxygen, the superoxide anion, O I. 2 (Stauff et a1., 1963; 5 Khan, 1970; Pederson and Aust, 1973), as depicted below: o°+o‘+2H+-———+Ho +10* 2 2 22 2 Superoxide, which may possibly initiate lipid peroxidation itself, is produced in many biological reactions such as autoxidations of reduced flavins, hydroquinones and catecholamines, and the aerobic actions of the enzymes xanthine oxidase, aldehyde oxidase and many flavin dehydrogenases (Fridovich, 1975). The lipid hydroperoxides (LOOH) that are generated in the pro- pagation step are unstable and decompose to form additional radical products. The decomposition of lipid hydroperoxides is catalyzed by trace amounts of transition metal ions (Holman, 1954; Heaton and Uri, 1961; Barber and Bernheim, 1967) and is shown in the following . + . . . reactions (Mn = tranSition metal ion): + . - + + L00H+Mn ——-+LO +oa +M‘n1) +1 + . + + LOOH+M(n ) -—+Loo +H +Mn The catalytic decomposition of lipid hydroperoxides along with prOpagation reactions previously described are therefore autocataly— tic, as more free radicals are the products the reactions. The autocatalytic reactions continue until substrate is depleted, sufficient termination reactions occur, or until intervention by antioxidants, which react with the free radicals and thereby inter- rupt the chain reaction process (Demopoulos, 1973b). 6 Biochemical Consequences of Lipid Peroxidation The membranes of the mitochondria and endoplasmic reticulum have a high proportion of unsaturated fatty acids (Rouser et a1., 1968). Consequently, these membranes are highly susceptible to lipid peroxidative damage. Furthermore, some of the most potent catalysts involved in lipid peroxidation, coordinated iron and hemeproteins, are found in close association to these membranes (Tappel, 1973). Thus, the occurrence of lipid peroxidation in biological membranes has profound effects due to the close rela- tionship of the unsaturated lipids with enzymes found in the membrane. Lipid peroxidative damage of mitochondrial membranes has been demonstrated by Hunter et a1. (1963) to correlate with swelling and lysis of the mitochondria. Alterations in the activity of NADH- cytochrome c reductase and the succinoxidase system of heart and liver mitochondria also are affected by lipid peroxidative damage (Tappel, 1965;Thppel.1973). Studies into the mechanism of ethanol induced hepatic damage have attributed lipid peroxidative damage in the mitochondria as a possible subcellular lesion (DiLuzio, 1973). As with the mitochondria, lipid peroxidative damage also affects microsomes. The most extensively studied agent regarding microsomal lipid peroxidation is carbon tetrachloride. Recknagel (1967) has reviewed the evidence supporting carbon tetrachloride induced lipid peroxidation. More recent evidence has indicated that carbon tetrachloride is metabolized to a trichloromethyl free radical which reacts with unsaturated membrane lipids to initiate 7 lipid peroxidation (Villarruel, 1973; Recknagel et a1., 1974). Carbon tetrachloride has also been shown to decrease microsomal drug metabolism and cytochrome P-450 levels, possibly through lipid peroxidation (Archakov and Kurzina, 1973). In other studies in vitro with isolated microsomes, stimulation of lipid peroxidation by addition of NADPH has been demonstrated to decrease the metabolism of acetanilide and pentobarbital (Jacobson et a1., 1973). Inhi- bition of metabolism was also correlated to a destruction of cytochrome P-450 (Jacobson et a1., 1973; Levin et a1., 1973). Similar observations regarding the in vitro inhibition of drug metabolism concurrent with lipid peroxidation have been made by other investigators, such as a decrease in morphine demethylation (Kamataki and Kitagawa, 1973) and a decrease in aminopyrine oxida- tion and aniline hydroxylation (Wills, 1969). It is evident from the previous paragraphs that peroxidation of unsaturated membrane lipids has far greater biochemical effects than merely deterioration of lipids. The intimate juxtaposition of unsaturated lipids and proteins in membranes allows for further interaction between the lipids and proteins. Several membrane proteins like succinic dehydrogenase and B-hydroxybutyrate dehy— drogenase may derive some of their structure from closely associated membrane lipids. In the presence of lipid peroxidation, and in particular the termination reactions in which two adjacent fatty acids are joined in abnormal bonds, the enzyme structure may be sufficiently altered to affect activity (Demopoulos, 1973a). The lipid peroxy radicals (LOO°) that are generated in the propagation 8 step may themselves abstract hydrogen atoms from neighboring pro- teins, resulting in protein cross-linking to form polymers (Tappel, 1965). This reaction sequence is shown schematically below (P = protein): Loo’ + P ———> LOOH + P(-H) P(-H) + P '—* P-P Studies by Chio and Tappel (1969) have demonstrated that sulfhydryl enzymes are most susceptible to inactivation by lipid peroxidation, which may occur either by an alteration of membrane structure or by protein-protein cross-linking. Thus, the biochemical consequences of lipid peroxidation at the membrane level are exceedingly complex and involve not only the unsaturated lipids but also the many dif- ferent proteins that are an integral part of membranes. Biological Defenses Against Lipid Peroxidation Protection Against Superoxide and Singlet Oxygen. Lipid peroxidation, or oxidative deterioration of unsaturated lipids, has been shown to be an exceedingly damaging biological process. Aerobic organisms are continually exposed to many potentially damaging oxidative reactions. As a result, aerobic organisms possess several different mechanisms which provide protection against uncontrolled free radical reactions. The most general defense of aerobic organisms against lipid peroxidation lies in the structural characteristics of cell membranes. The nature of membranes has been described as consisting of a 9 hydrophobic midzone between two hydrophilic surfaces. The hydro- phobic midzone appears to be richly penetrated by proteins (Rouser et a1., 1968). The unsaturated lipids are found for the most part in the hydrophobic inner layer. As many free radicals are cations or anions, they do not readily penetrate the hydrophobic layer where abstraction of allylic hydrogens can occur. Furthermore, the close juxtaposition of proteins and unsaturated lipids hinders free radical reactions. The proteins provide a spatial separation for the unsaturated lipids and thereby can prevent the spread of radical chain reactions through the membrane (Demopoulos, 1973b). However, in the case of oxygen, the structure of the membrane may be dis- advantageous. Oxygen is 7-8 times more soluble in nonpolar media and thereby has an affinity for the hydrophobic midzone of membranes, which is where the lipids most susceptible to oxidative damage are located (Demopoulos, 1973a). Because membrane structure alone is not adequate to prevent oxidation of unsaturated lipids, aerobic organisms also possess other defenses against lipid peroxidation. One specific defense, acting at the level of oxygen attack of unsaturated lipids, is the activity of the enzyme superoxide dismutase. Superoxide dismutase scavenges the toxic superoxide radical by catalyzing the reaction depicted below (McCord and Fridovich, 1969): + + + + 02 O2 2H -—+ 02 H202 Pederson and Aust (1973) have demonstrated that superoxide may initiate lipid peroxidation, as xanthine oxidase-induced peroxidation 10 of microsomal lipids was inhibited by superoxide dismutase. The biological importance of superoxide dismutase as a defense mechanism has been determined by a number of elegant studies con- ducted with bacteria. McCord et al. (1971) examined the distribu- tion of superoxide dismutase among 3 classes of microorganisms: aerobes which utilize oxygen in their metabolism almost exclusively, aerotolerant organisms which have an anaerobic metabolism even when grown in air, and finally strict anaerobes which cannot survive in oxygen. In all cases the aerobic organisms contained the highest activity of superoxide dismutase, followed by intermediate activity in the aerotolerant group. Strict anaerobes contained no superoxide dismutase, which may explain their inability to tolerate oxygen. In other studies (McCord et a1., 1973) a mutant strain of Escherichia coli was selected that had a temperature sensitive defect in its ability to grow aerobically over a temperature range of 30°C to 43.5°C. Thus, the mutant strain was less able to grow aerobically as the temperature was increased from 30°C compared to the wild type parent strain which could grow at all temperatures. The mutant strain was assayed for superoxide dismutase and found to have a progressive decrease in enzyme activity as the temperature increased. The enzyme activity in the parent strain was unaffected by temperature. Thus, in the mutant strain, the ability to grow aerobically as the temperature increased correlated with levels of superoxide dismutase. Further investigations have determined that superoxide dismu- tase is induced in bacteria (Gregory and Fridovich, l973a,b) and ll yeast (Gregory et a1., 1974) in response to exposure to elevated oxygen tension. The protective function of superoxide dismutase was also evident in these studies (Gregory and Fridovich, l973b) since E. coli grown under nitrogen were much less resistant to hyperbaric oxygen than E. coli grown under 2 atmospheres oxygen. This correlated with superoxide dismutase levels of 3.8 units/mg in the noninduced E. coli compared to 21 units/mg extract in the induced E. coli. The induction of superoxide dismutase in bacteria, therefore, offered resistance to exposure to elevated oxygen tensions. This resistance was presumably due to the enhanced ability of the bacteria to detoxify the superoxide radical. Induction of superoxide dismutase by elevated oxygen tensions is not limited to microorganisms, however. Rosenbaum et a1. (1969) demonstrated that exposure of rats to 85% oxygen for 7 days pro- longed the survival time compared to nonpretreated rats when these rats were transferred to 100% oxygen. Superoxide dismutase activity in the lungs of rats exposed to 85% oxygen for 7 days was increased 50% compared to controls. Furthermore, the rate of tolerance development to 100% oxygen for rats pretreated with 85% oxygen closely paralleled increases in pulmonary superoxide dismutase activity (Crapo and Tierney, 1974). Thus, rats show the same parallel in comparing superoxide dismutase activity with resistance to oxygen as do bacteria. Much of the work done to date regarding all aspects of super- oxide dismutase can be found in an excellent review by Fridovich (1974a). 12 As has been discussed previously, however, superoxide may not be the initiator of lipid peroxidation. Rather, the nonenzymatic dismutation product of superoxide, singlet oxygen, may be the actual initiator of lipid peroxidation (Pederson and Aust, 1973). Experiments conducted by Krinsky (1974a) have indicated a biochemical mechanism protective against the toxic effects of singlet oxygen. A mutant strain of bacteria deficient in carotenoid pigments was observed to be much more susceptible to killing by human polymorpho- nuclear (PMN) leukocytes than a comparable carotenoid-containing strain. PMN leukocytes have been proposed to destroy ingested bacteria by generation of singlet oxygen which disrupts bacterial membranes (Allen et a1., 1972; Maugh, 1973; Allen et a1., 1974). Carotenoid pigments have been demonstrated to be quenchers of singlet oxygen in vitro (Foote and Denny, 1968) and may function similarly in vivo (Krinsky, 1971; Krinsky, 1974b). Thus, the susceptibility of the mutant bacteria to destruction by PMN leuko- cytes may be due to an inability to quench singlet oxygen. There is some evidence to indiCate that carotenoids may also function in animals to quench singlet oxygen. Matthews (1964) was able to protect against the lethal photodynamic effects of intraperitoneally injected hematoporphyrin in mice when the animals were simultaneously injected with B-carotene. The protection afforded the mice to photodynamic lethality may be due to quenching of singlet oxygen, as many photooxidations proceed via a singlet oxygen mechanism (Foote, 1968). 13 Antioxidant Protection Against Lipid Peroxidation. Whereas superoxide dismutase and the carotenoid pigments appear to prevent the initiation of lipid peroxidation, organisms also possess bio- chemical defenses against the propagation of free radical reactions. This biochemical defense is centered around the endogenously occur- ring antioxidants. Two types of antioxidants function in vivo: the water soluble antioxidants which include ascorbic acid and glutathione and the lipid soluble antioxidants which mostly consist of the tocopherols (Demopoulos, l973b). In general, antioxidants function by allowing a hydrogen to be abstracted from themselves rather than from the allylic hydrogen of an unsaturated lipid and thus act by interrupting the free radical chain reactions. A simpli- fied scheme whereby vitamin E (a-tocopherol) acts as a chain breaker is depicted below (Tappel, 1972; aTH = a-tocopherol; aTQ = a-tocopherol quinone): Propagation: L' + 02 -————*-L02° L02°+LH——+Looa+L‘ Antioxidant action: L02° + aTH -————+ LOOH + aT' Termination: LOZ' + aT' -————+-LOOH + cTQ The antioxidant function of vitamin E became apparent when deficiency states of this vitamin were induced. It has been shown that liver mitochondria and microsomes isolated from rats fed vitamin E deficient diets have greater peroxidation rates compared to rats fed basal diets supplemented with vitamin E (Tappel and l4 Zalkin, 1959; Dillard and Tappel, 1971). Furthermore, evidence of in vivo lipid peroxidation, as indicated by the formation of fluorescent lipofuscin pigments in the tissues, has been found in rats fed vitamin E deficient diets supplemented with polyunsaturated lipids (Porta and Hartroft, 1969). Vitamin E deficiency has been reported to increase the susceptibility of mice to oxygen toxicity which was associated with a reduction in lung phospholipids (Mino, 1973). An increased susceptibility to another strong oxidant, ozone, also has been observed in vitamin E deficient animals (Goldstein et a1., 1970; Roehm et a1., 1972). In humans, erythrocytes from patients with low plasma toc0pherol levels have an increased suscep- tibility to lipid peroxidation (measured by increased malondialde- hyde levels) on exposure to hydrogen peroxide vapor. In vitro addition of toc0pherol reduced the lipid peroxidation (Tudhope and Hopkins, 1975). Dialuric acid has long been known to induce erythro- cyte hemolysis in vitamin E deficient animals both in vitro and in viva (Rose and Gyorgy, 1952). The hemolysis induced by dialuric acid has been shown to be associated with lipid peroxidation and was reduced by addition of vitamin E (Bunyan et a1., 1960; Tsen and Collier, 1960). Recently, Cohen and Heikkila (1974) have demon- strated that superoxide radical is a product of dialuric acid autoxidation and subsequently may initiate lipid peroxidation in the erythrocyte membrane. Vitamin E deficiency results in several nutritional disease states. The earliest recognized disease state associated with vitamin E deficiency was a degeneration of muscle tissue (Goettsch 15 and Pappenheimer, 1931). The degeneration of muscle tissue in vitamin E deficiency closely resembles the muscle degeneration associated with human myotonic muscular dystrophy. Consequently, the vitamin E deficient effect has been referred to as nutritional muscular dystrophy (Horwitt, 1965). However, vitamin B administra- tion does not alter the outcome of inherited muscular dystrophy (Tubis et a1., 1959). Other nutritional disease states that have been reported with vitamin E deficiency are necrotic liver destruc- tion in rats (Schwarz, 1965) and encephalomalacia in chicks (Dam et a1., 1957). Selenium and Enzymes of the Glutathione Peroxidase System. The antioxidant function of selenium was proposed from early studies which demonstrated that vitamin E deficiency syndromes such as nutritional muscular dystrophy were reversed by addition of small amounts of selenium to the diet (Bieri et a1., 1961; Scott, 1962; Scott, 1969). It has also been long known that the enzyme gluta- thione peroxidase could detoxify hydrogen peroxide and hydroperoxides (LOOH) in viva (Mitts and Randall, 1958; Cohen and Hochstein, 1963; Christopherson, 1969; O'Brien and Little, 1969). The unstable lipid hydroperoxides are enzymatically reduced by gluathione peroxidase to stable lipid alcohols, which results in termination of free radical chain reactions. Recently, Rotruck et a1. (1973) proposed that selenium was a necessary cofactor for glutathione peroxidase, as selenium deficient rats were unable to prevent hydrogen peroxide induced erythrocyte hemolysis. Purification of glutathione peroxidase from erythrocytes has demonstrated that the l6 enzyme consists of four subunits, with one gram-atom of selenium per subunit (Flohe et a1., 1973; Oh et a1., 1974). Other recent investigations have shown that glutathione peroxidase activity is directly related to the levels of dietary selenium in rats (Chow and Tappel, 1974; Hafeman et a1., 1974; Reddy and Tappel, 1974; Smith et a1., 1974; Tappel, 1974) and in chicks (Noguchi, 1973; Tappel, 1974). Thus, the majority of the antioxidant activity of selenium appears to be mediated through glutathione peroxidase activity. The importance of gluathione peroxidase activity in detoxi- fying lipid hydroperoxides has been demonstrated in studies in which animals were exposed to oxidant stress. It has been proposed that glutathione peroxidase may protect against lipid peroxidative damage since the activity of this enzyme was induced in rat lung after exposure to the oxidant gas, ozone (Chow and Tappel, 1972; Chow et a1., 1974). It was also observed by these investigators that the enzyme activity of glutathione reductase and glucose-6- phosphate dehydrogenase was induced in response to ozone exposure, in addition to the induction of glutathione peroxidase. The three enzymes were referred to as the glutathione peroxidase system and are proposed to function as a unit in combating lipid peroxidation as follows: the conversion of toxic lipid hydrOperoxides to lipid alcohols by glutathione peroxidase is linked to the activity of glutathione reductase and glucose-6-phosphate dehydrogenase which supply reducing equivalents in the form of reduced glutathione (GSH) and NADPH, respectively (Chow and Tappel, 1972). 17 Activity of the glutathione peroxidase system enzymes appears to be of clinical relevance in humans. Glutathione peroxidase activity is genetically determined in humans and consequently deficiency states of this enzyme in adults have been correlated to drug-induced hemolysis (Steinberg and Necheles, 1971) and to chronic hemolytic anemia (Necheles et a1., 1970). Hopkins and Tudhope (1974a) have observed that erythrocytes from glutathione peroxidase deficient individuals had an increased susceptibility to Heinz body formation when exposed to hydrogen peroxide vapor. These investigators also reported increased erythrocyte mechanical fragility with inhibition of glutathione peroxidase by menadione, gentisic acid or potassium chlorate (Hopkins and Tudhope, 1974b). Similar to individuals deficient in glutathione peroxidase, individuals with genetically related deficiencies in either glutathione reductase or glucose-6-phosphate dehydrogenase have been reported to be suscep- tible to erythrocyte hemolysis by a large number of drugs. The hemolysis has been proposed to be caused by an inability of the erythrocyte to detoxify peroxides that are formed in the cell as a consequence of the presence of the drug. The agents and mechanisms involved in erythrocyte hemolysis in enzyme deficient individuals has been reviewed by Beutler (1969). Examples of Oxidant Agents that May Induce Lipid Peroxidation Destruction of Bacteria by Polymorphonuclear Leukocytes (PMN Leukocytes). The importance of oxygen in bacterial killing by PMN leukocytes was first realized with the observation that a large 18 increase in oxygen uptake by PMN leukocytes occurred during phago- cytosis (Sbarra and Karnovsky, 1959). It has been estimated that up to 10% of the increase in oxygen consumption is subsequently excreted from PMN leukocytes as the highly toxic superoxide radical (Babior et a1., 1973). The superoxide radicals excreted by the PMN leukocytes may be responsible for phagocytic killing, as E. coli deficient in superoxide dismutase are particularly vulnerable to phagocytosis (Fridovich, 1974b). Furthermore, PMN leukocytes iso- lated from patients with chronic granulomatous disease have been demonstrated to produce only small amounts of superoxide compared to PMN leukocytes isolated from normal individuals (Curnutte et a1., 1974). Consequently, patients with such defective PMN leukocytes are not protected against certain species of bacteria such as Staphylococcus aureus. Recently Beckman et a1. (1973) found that cytosol superoxide dismutase was found in high concentrations in a number of human cell types except in the cytosol of PMN leuko- cytes. Fridovich (1974b) has speculated that PMN leukocytes lack the defense against the superoxide radical in order to be better able to kill bacteria through high levels of superoxide. The superoxide radical, however, is not the only potential bactericidal agent excreted by PMN leukocytes. Highly reactive singlet oxygen has also been demonstrated to be evolved from PMN leukocytes during phagocytosis (Allen et a1., 1972; Maugh, 1973; Allen et a1., 1974). That singlet oxygen may be a bactericidal agent is evident from a study in which mutant bacteria deficient in singlet oxygen trapping carotenoid pigments were found to be more l9 readily killed by PMN leukocytes than the carotenoid-containing parent strain (Krinsky, 1974a). Streptonigrin. Streptonigrin is a paraquinone antibiotic which has been shown to undergo a single electron reduction to a parahydroquinone in E. coli cultures (White and Dearman, 1965). The parahydroquinone can be spontaneously reoxidized back to the quinone by molecular oxygen with formation of superoxide radical, which may be the bactericidal agent (Misra and Fridovich, 1972). The generation of superoxide due to the cyclic reduction and reoxidation of streptonigrin may account for the decreased toxicity of streptonigrin to E. coli in the absence of oxygen and its enhanced toxicity in the presence of oxygen (White and White, 1968; White et a1., 1971). Further evidence that superoxide mediates streptonigrin toxicity is found in experiments in which E. coli containing high induced levels of superoxide dismutase were resistant to streptonigrin (Gregory and Fridovich, 1973b). Oxyge . Both animals and man develop pulmonary lesions after exposure to elevated oxygen tensions. Exposure of monkeys to 90— 100% oxygen resulted in sequential pulmonary changes characterized by interstitial edema, destruction of Type I alveolar cells, endo- thelial cell thickening, proliferation of Type II alveolar cells, and finally infiltration by fibroblasts. Rats were also observed to undergo similar pulmonary changes in 100% oxygen, although most died before reaching the proliferative stage (Kistler et a1., 1967; Kapanci et a1., 1969; Kaplan et a1., 1969; review by Winter and 20 Smith, 1972). Lipid peroxidative damage may result from oxygen exposure as vitamin E deficiency enhanced oxygen toxicity (Mino, 1973) and vitamin E administration protected against hyperbaric oxygen (Kann et a1., 1964). Tierney et a1. (1973) also observed that rats made tolerant to oxygen by exposure to 85% oxygen for 7 days had increased pulmonary levels of glucose—6-phosphate dehy- drogenase. The tolerance development may be in part associated with a response of the glutathione peroxidase system enzymes to lipid peroxidative damage. Ozone and Nitrogen Dioxide. Ozone and nitrogen dioxide are oxidant gases that are found in air pollutant gas mixtures. The pulmonary lesions resulting from exposure to ozone or nitrogen dioxide have been characterized as an initial destruction of epi- thelial Type I cells (Stephens et a1., 1972; Stephens et a1., 1974) followed by a proliferation of Type II pneumocytes (Evans et a1., 1972; Stephens et a1., 1974). Most of the lesions occur in the area of the terminal bronchiole and proximal alveoli (Stephens et a1., 1974). As with oxygen, exposure of animals to sublethal levels of ozone or nitrogen dioxide resulted in the development of tolerance to subsequent exposure to previously lethal doses of the gases (Stokinger and Scheel, 1962; Fairchild, 1967). The mechanism of tolerance development has been linked to induction of the glutathione peroxidase system which indicated lipid peroxidation as the mediator of toxicity of the two gases (Chow et a1., 1974). Other investiga- tors have also provided evidence that the in vivo pulmonary toxicity 21 of ozone and nitrogen dioxide may involve lipid peroxidation (Thomas et a1., 1968; Goldstein et a1., 1969; Fletcher and Tappel, 1973). Paraquat Historical Background and Uses. In the mid-1950's a series of bipyridylium compounds was found by Imperial Chemical Industries to possess herbicidal activity (Boon, 1965). The compound paraquat, l,l'-dimethyl-4,4'-bipyridylium dichloride, was introduced for commercial use in 1962 in England and in the United States in 1964 (Kimbrough, 1974) as a broad spectrum herbicide effective against both broad leaf weeds and grasses. Paraquat is marketed outside the United States as a 20% concentrate, Gramoxone, and in a 5% granular form, Weedol. It is available in the United States as a 29.1% solution, Orthoparaquat, a 42% solution, Ortho dualparaquat, and as a 0.44% solution, Ortho Spot. The broad spectrum herbicidal activity of paraquat has resulted in its wide use throughout the world. Paraquat has been used to kill weeds and grasses in orchards and between rows of crops and vineyards, to defoliate and desiccate crops for easier harvesting, to renew pastures overgrown with weeds and grasses, and for aquatic weed control. Calderbank (1968) has extensively reviewed the herbi- cidal uses of paraquat and other bipyridylium herbicides. Application of paraquat to crops appears to result in only low levels of residue contamination. Minute residues of paraquat have been detected in potatoes which were defoliated prior to harvest (Boon, 1967) while seed harvested from previously desiccated crops 22 contained paraquat residues which varied from nondetectable up to 10 ppm (Calderbank, 1968). Paraquat has been reported to be firmly bound when it contacts most types of soils and upon adsorption can be degraded by several microorganisms present in the soil. When paraquat is used for aquatic weed control, residues in the water decline fairly rapidly. However, up to 112 ppm paraquat has been found in aquatic weeds when the surrounding water was at an initial concentration of];ppm (reviewed by Calderbank, 1968). Human Toxicity of Paraquat. One reason for paraquat's rapid acceptance as an all-purpose herbicide was its apparent lack of any severe toxicity to workers or animals exposed to spray during field application. The only reports of toxicity to field workers exposed to paraquat spray were those of nail damage (Samman and Johnston, 1969; Swan, 1969; Hearn and Keir, 1971) and irritation of the nasal mucosa, skin and eyes (Swan, 1969). One case of severe eye injury resulted from splashes of the concentrate (Cant and Lewis, 1968). Soon after paraquat was introduced, however, cases of paraquat ingestion began to be reported. The major route of paraquat inges- tion is oral, either by accident or intention, and in many of these cases death was a result (Bullivant, 1966; Clark et a1., 1966; Duffy and Sullivan, 1968; Fennelly et a1., 1968; Oreopoulos et a1., 1968; Masterson and Roche, 1970; Beebeejuan et a1., 1971; Malone et a1., 1971; Copland et a1., 1974). In one suicide case, death resulted from a subcutaneous injection of 1 ml of 20% paraquat concentrate (Almog and Tal, 1967). Accidental ingestion of paraquat has also resulted in a number of deaths in children (Campbell, 23 1968; McDonagh and Martin, 1970). In one nonlethal case in a child, paraquat was detected in the urine after the herbicide was spilled over the skin, suggesting percutaneous absorption (McDonagh and Martin, 1970). A 1971 report indicated that the total number of deaths attributable to paraquat intoxication has reached 124 (Editorial, 1971). The fatal dose of paraquat in humans has been estimated to be from 4 mg/kg subcutaneously (Almog and Tal, 1967) to perhaps up to 50 mg/kg orally (Murray and Gibson, 1972). The overall mortality rate, encompassing the entire range of amounts ingested, appears to be approximately 33-50% (Editorial, 1971). The clinical symptoms that develop after ingestion of a toxic dose of paraquat proceed from an initial nausea and vomiting, ulcera- tion of the mouth and pharynx, oliguria, albuminuria and an increase in blood urea nitrogen, which then after a latent period progress to dyspnea, cyanosis and death due to massive pulmonary edema and interstitial fibrosis associated with a proliferation of alveolar epithelium (Bullivant, 1966; Almog and Tal, 1967; Campbell, 1968; Pasi and Hine, 1971; Toner et a1., 1971; Anonymous, 1972). In some cases both renal and centrilobular hepatic necrosis was observed (McDonagh and Martin, 1970). In most cases victims of paraquat poisoning survive for at least one week with some surviving up to four weeks after ingestion. The protracted development of lung lesions has been associated with a prolonged excretion of paraquat in the urine (Beebeejaun et a1., 1971). The treatment of individuals poisoned with paraquat has utilized a number of therapeutic approaches, all with varying 24 degrees of success. Adsorption of paraquat cation in the stomach with Fuller's earth, activated charcoal and bentonite gel has been recommended as a nonspecific antidote for paraquat (Browne, 1971; Anonymous, 1972). Hemodialysis (Grundles et a1., 1971; Galloway and Petrie, 1972), peritoneal dialysis (Fisher et a1., 1971) and forced diuresis (Kerr et a1., 1968; Beebeejaun, 1971; Tompsett, 1970) have all been used with variable results. As the pulmonary symptoms developed oxygen therapy has been used (Copland et a1., 1974), although evidence exists that oxygen may enhance paraquat pulmonary toxicity and is therefore contraindicated in paraquat poisoning (Flenley, 1971; Nienhaus and Ehrenfeld, 1971; Fisher et a1., l973a). Other investigators have recommended either a combination of steroids and cyclophosphamide (Malone et a1., 1971) or of steroids, d-propranolol and superoxide dismutase (Anonymous, 1973; Davies and Davies, 1974). In an extreme case a lung transplantation was attempted without success (Matthew et a1., 1968). Because of the wide variation in the amount of paraquat ingested among these cases, however, it is difficult to assess the effectiveness of any one treatment compared to another. Animal Toxicity of Paraquat. The toxicity of paraquat has been studied in several animal species including rats (Kimbrough and Gaines, 1970; Murray and Gibson, 1972; Robertson et a1., 1971; Sharp et a1., 1972), mice (Brook, 1971), rabbits (Butler and Kleinerman, 1971), guinea pigs and monkeys (Murray and Gibson, 1972). The pulmonary lesions in rats and monkeys have been observed to be similar to the lesions in humans with initial pulmonary edema, 25 congestion and intra-alveolar hemorrhage followed by the onset of interstitial fibrosis (Murray and Gibson, 1972). Examination of the ultrastructure of the pulmonary lesion in rats (Smith, 1971; Kimbrough and Linder, 1973) and mice (Brook, 1971) revealed early damage to Type I pneumocytes, increased numbers of fibroblasts, and a later proliferation of Type II pneumocytes. Vijeyaratnam and Corrin (1971) reported that such paraquat induced pulmonary damage resembled that of oxygen toxicity. Not all species are similarly affected by paraquat toxicity, however, as guinea pigs developed pulmonary edema which did not progress to interstitial fibrosis (Murray and Gibson, 1972) and rabbits did not develop any type of pulmonary lesions (Butler and Kelinerman, 1971). The toxicity of paraquat in tissue sites other than the lung has also been examined. Paraquat induced centrilobular and tubular necrosis of rat liver and kidney, respectively (Murray and Gibson, 1972). Necrosis of the cells of the proximal convoluted tubule in the kidneys of mice eXposed to paraquat was associated with the appearance of lipid lamellate cytosomes (Clark et a1., 1966; Fowler and Brooks, 1971). Atrophy of the thymus has been noted in rabbits (Butler and Kelinerman, 1971). Paraquat was also toxic in vitro to rat alveolar and peritoneal macrophages (Styles, 1974). Paraquat has been reported to induce a slight increase in costal cartilage malformations in rat embryos (Khera et a1., 1968) and to have a low teratogenic potential in mice (Bus et a1., 1975). Furthermore, Pasi et a1. (1974) observed that paraquat was not mutagenic in mice, although it had an antifertility effect. 26 Paraquat absorption is rapid and complete when administered subcutaneously in rats (Daniel and Gage, 1966). However, absorption after oral administration in rats appears to be poor as significant amounts of paraquat have been detected in the feces (Daniel and Gage, 1966; Murray and Gibson, 1974). The poor oral absorption of para- quat is reflected in the 6- to 8-fold increase in the oral LD in 50 rats compared to the parenteral LD (Calderbank, 1968; Howe and 50 Wright, 1965). Paraquat also has been reported to be sufficiently absorbed from the eye (Sinow and Wei, 1973) and skin (McElligott, 1972) to result in death in rabbits. Once absorbed, paraquat is distributed to most tissues except for the brain and spinal cord (Sharp et a1., 1972; Litchfield et a1., 1973; Murray and Gibson, 1974). Several investigators have reported that paraquat rapidly disappears from most tissues in rats and mice except for the lung, where retention of the herbicide occurs (Molnar and Hayes, 1971; Sharp et a1., 1972; Litchfield et a1., 1973; Illett et a1., 1974; Murray and Gibson, 1974). This observation has resulted in the suggestion that retention of para- quat in lung tissue may account for its pulmonary toxicity. No metabolism of paraquat has been demonstrated upon absorption (Murray and Gibson, 1974; Bus et a1., 1975) although paraquat may be metabolized by the gut bacteria after oral administration (Daniel and Gage, 1966). Paraquat is largely excreted by the kidney (Daniel and Gage, 1966; Murray and Gibson, 1974) and only small amounts are excreted in the bile (Daniel and Gage, 1966; Hughes et a1., 1973). Elimination of paraquat by the kidney 27 appears to be by an active secretory process in the proximal tubule (Baker, J. E., unpublished observation). A number of different antidotal regimens have been attempted in animals in order to suggest possible therapeutic approaches for treatment of paraquat poisoning in humans. Staiff et a1. (1973) reported that the Amberlite CG-120 resins, either 100-200 or 200-400 mesh, lowered the tissue levels of paraquat and may have provided protection against poisoning. Treatment of rats with a stomach wash followed by four administrations of bentonite plus purgatives (0.5 m1 castor oil, 250 mg magnesium sulfate/kg) at 2- to 3-hour intervals was effective as an antidote even if treatment was delayed up to 10 hours after paraquat administration (Rose, M. 8., personal communication). Administration of expectorants prevented paraquat induced decreases in lung surfactant although no measurements were made as to alterations in survival (Cambar and Aviado, 1970). Exposure of paraquat poisoned mice to lowered oxygen tensions (Rhodes, 1974) as well as administration of superoxide dismutase in rats (Autor, 1974) has also been able to alter paraquat lethality. Recently, repeated administration of propranolol has been shown to protect against paraquat lethality in rats (Maling et a1., 1975). Mechanism of Action of Paraquat. The mechanism of paraquat's herbicidal activity has been extensively studied. As early as 1933, it was shown that paraquat (which was used as an oxidation-reduction indicator with the name methyl viologen) was capable of undergoing a single electron reduction to a blue-colored free radical form with a redox potential of -446 mv (Michaelis and Hill, l933a,b). 28 This reduction reaction is depicted below: + - 3-+N/ \ \/ \N+-CH3fiCH3- PARAQUAT REDUCED PARAQUAT The free radical that is generated is apparently stabilized due to the complete delocalization of the added electron over the c0planar paraquat molecule (Boon, 1965). The ability of isolated plant chloroplasts to form the blue-colored paraquat free radical under anaerobic conditions has been demonstrated (Dodge, 1971). Anaerobic conditions were necessary in order to prevent the immediate reoxida- tion to the colorless parent compound. Early work by Mees (1960) showed that oxygen was necessary for the herbicidal activity of bipyridylium compounds. These studies demonstrated that such agents could not kill plant leaves when there was no oxygen present, despite the continuation of photosynthetic reactions capable of generating the free radical. Later experiments suggested that the paraquat free radical may transfer its extra electron to molecular oxygen to form superoxide radical (Farrington, 1973). There was evidence that the superoxide radical may persist long enough to diffuse to cell membranes where membrane damage may be initiated. Hydrogen peroxide has also been proposed as the membrane damaging agent (Conning et a1., 1969; Thorneley, 1974). The possibility that lipid peroxidation of cell membranes might be the damaging lesion caused by paraquat was indicated by experiments which found 29 increased levels of malondialdehyde (a by-product of lipid peroxida- tion) in plant leaves six hours after paraquat exposure (Dodge et a1., 1970). In animals, the pulmonary lesion caused by paraquat has been postulated to be a result of disruption of pulmonary surfactant, and thereby to closely resemble human respiratory distress syndrome (Manktelow, 1968; Robertson et a1., 1971). Both groups of investi- gators found decreased pulmonary surfactant activity associated with the formation of alveolar hyaline membranes after acute paraquat administration. Further investigations have shown that paraquat induced a specific decrease in the lecithin fraction of lung sur- factant, which led to the suggestion that the pulmonary atelectasis associated with paraquat poisoning may be due to an increase in alveolar surface forces (Fisher et a1., l973b; Malmqvist et a1., 1973). In other reports, however, no changes after paraquat were observed in pulmonary phospholipid fractions (Fletcher and Wyatt, 1970) or in the synthesis or destruction of lung dipalmitoyl lecithin (Fletcher and Wyatt, 1972) indicating that paraquat does not interfere with phospholipid metabolism. In 1968, Gage demonstrated the production of the blue-colored paraquat free radical when paraquat was incubated in vitro with rat liver microsomes and NADPH. It was further shown that the cyclic reduction-oxidation of paraquat was associated with an increase in malondialdehyde production in microsomal phospholipids. Recently, the link of oxygen to the herbicidal activity of paraquat has like- wise been demonstrated to the animal toxicity of paraquat. 30 Paraquat induced lethality in rats was significantly enhanced by simultaneous exposure to 100% oxygen (Fisher et al., 1973a) and was significantly decreased in mice with exposure to 10% oxygen (Rhodes, 1974). Davies and Davies (1974) demonstrated that paraquat incubated with rat liver microsomes and NADPH stimulated the oxida- tion of epinephrine to adrenochrome, which was an indication of the production of superoxide radicals. Furthermore, the oxidation of epinephrine was inhibited with the addition of superoxide dismu- tase. In vivo evidence for involvement of superoxide has been suggested from experiments in which paraquat induced lethality in rats was delayed by administration of superoxide dismutase, an enzyme which detoxifies superoxide radicals (Autor, 1974). The hypothesis that paraquat may cause pulmonary damage through oxidant reactions is further supported from observations in rats that para- quat stimulates the activity of lung glucose-G-phosphate dehydrogen- ase (Witschi and Kacew, 1974) and that sensitivity to paraquat is directly related to lung glucose-6-phosphate dehydrogenase activity (Ayers and Tierney, 1971). Glucose—6-phosphate dehydrogenase activity has been linked to the activity of glutathione reductase and glutathione peroxidase in combating oxidant stress (Chow and Tappel, 1974). Other biochemical parameters have also been measured in order to further investigate the mechanism of paraquat toxicity. Krieger et a1. (1973) demonstrated that paraquat inhibited the in vitro epoxidation of aldrin, causing a 50% inhibition at a concentration 7 x 10-4M. Because of paraquat's ability to accept electrons, the 31 inhibition was proposed to be due to an interruption of microsomal electron transport processes. Paraquat has been shown to decrease rat but not rabbit lung cytochrome P-450 concentrations and bromo- benzene metabolism tar rat lung microsomes (Ilett et a1., 1974). A significant increase in protein synthesis has been reported in rat lung, liver and kidney after a toxic dose of paraquat, while DNA synthesis was depressed in the three organs (Van Osten and Gibson, 1975). The biochemical changes showed a time correlation with histopathological changes induced by paraquat. Paraquat also increased plasma corticosteroids in rats for 24 hours after admin- istration while ACTH (adrenocorticotrophin) was elevated for only 4 hours, suggesting that paraquat may increase the response of the adrenal cortex to ACTH (Rose et a1., 1974a). Recently, experiments in vitro have shown that lung slices accumulated paraquat by an apparent energy-dependent process. The lung slices were able to accumulate paraquat but not an analog, diquat, which correlated to the in vivo retention of these compounds by rat lung (Rose et a1., 1974b). Uptake studies of paraquat into the isolated perfused rabbit lung did not appear to be by an energy- dependent mechanism. Little or no uptake of paraquat was observed initially and the subsequent uptake of paraquat in the perfused lung paralleled the development of edema, indicating that paraquat may be trapped in the damaged lung tissue (Orton et a1., 1973). How- ever, the rabbit has been described as unusually resistant to the development of the paraquat pulmonary lesion (Butler and Kleinerman, 1971) and thus account for the differences between the two studies. 32 P ose The purpose of this investigation was twofold: first, to determine the effects of chronically administered paraquat on the development of mice and to investigate subtle toxicities of paraquat that may exist in these animals; and second, to conduct experiments both in vitro and in vivo which will provide data for the formula- tion of a mechanism of toxicity of paraquat in mammalian systems. METHODS Animals Animals used in all experiments were purchased from Spartan Research Animals, Inc., Haslett, Michigan. Female Swiss-Webster mice, 30-35 grams body weight, were used as breeding stock to obtain timed pregnancies in developmental studies. Female Swiss-Webster mice, 25-30 grams body weight, and male Sprague-Dawley rats, 150- 200 grams body weight, were used in other experiments. Toxicity of Chronically Administered Paraquat in Developing Mice Mice were mated by placing 1 male in a cage of 5 females for 1 hour starting at 8 a.m. The day vaginal plugs were found was designated day 1 of gestation. Paraquat dichloride (paraquat concentrate, 240 mg cation/m1, Chevron Chemical Co., Richmond, California) was placed in the drinking water of pregnant females at concentrations of 0, 50, 100 and 150 ppm (parts per million, 1 mg paraquat per 1000 ml solution) beginning on day 8 of gesta- tion, which is the onset of organogenesis in mice (Rugh, 1968). Pregnant females were housed individually in clear plastic shoebox cages and allowed access to food and the paraquat treated water ad libitum. Following delivery of pups on day 20 of gestation, litters were normalized to 10 mice. The total number of live pups born in 33 34 each treatment group was recorded. Total litter body weights and mortality were recorded at weekly intervals from day 1 postnatally to termination of the experiment at 42 days postnatally. Water consumption was also measured up to 42 days postnatally. All litters were weaned on day 28 postnatally and segregated by sex. Exposure of mice to paraquat treated water was continuous from day 8 of gestation to 42 days postnatally. In other experiments, litters were either exposed to 100 ppm paraquat from day 8 of gestation to 28 days after birth (weaning) and then transferred to nonparaquat treated water until 42 days postnatally or placed on 100 ppm paraquat only between days 28 and 42 postnatally. Litter mortality rates were recorded at weekly intervals for each treatment group. The stability of the paraquat solutions was confirmed by colorimetric assay (Sharp et a1., 1972). Two milliliters of an appropriate dilution of the paraquat solutions was added to 0.5 ml of sodium dithionite (sodium hydrosulfite, 0.2% in 1N sodium hydroxide), mixed and the absorbance immediately read at 395 nm in a Beckman DB-GT spectrophotometer (Beckman Instruments, Fullerton, California). Paraquat dichloride (methyl viologen, Sigma Chemical Company, St. Louis, Missouri) was used to prepare the standard curve. No change in paraquat concentrations was observed for up to 4 weeks after preparation. Assay of Lung Tissue for Paraquat. Concentrations of paraquat in the lungs of 42 day old 50 and 100 ppm paraquat treated mice were determined by a colorimetric modification of the method of Ilett 35 et a1. (1974). Lung tissue was homogenized in 4 ml distilled water with a PolytronR homogenizer (Brinkman Instruments, Westbury, New York) followed by addition of 0.1 volume lON hydrochloric acid. The acid-precipitated homogenate was centrifuged at 5000 xg for 20 minutes and the resultant supernatant transferred to a Dowex ion—exchange column made from a 1.0 ml slurry of a 1:1 mix of water-Dowex 50W-X8 cation ion resin (100-200 mesh, Bio-Rad Laboratories, Richmond, California). The previous precipitate was washed once with 2.0 m1 1N hydrochloric acid, centrifuged, and the supernatant added to the ion—exchange column. The column was sub- sequently washed with 20.0 ml distilled water and the paraquat eluted with 5.6 ml of 5M ammonium chloride. The paraquat eluate was brought to 6.0 ml with 1N sodium hydroxide with a final eluate pH of 10. A 2.0 m1 aliquot of the eluate was assayed colorimetri- cally for paraquat by the method previously described. Assay of Tissue for Reduced Glutathione (GSH). Liver GSH was measured in 42 day old 50 and 100 ppm paraquat treated mice by the fluorometric method of Cohn and Lyle (1966). Approximately 100 mg of liver tissue was homogenized (Polytron) in 4.0 ml of cold 30 uM EDTA (Versene, Dow Chemical Co., Midland, Michigan) to which 1.0 m1 of cold 25% w/v metaphosphoric acid (HPOB) was added and followed by centrifugation in the cold at 5000 xg for 10 minutes. An 0.01 ml aliquot was removed from the supernatant and added to 0.99 ml of cold distilled water and 0.5 ml of 0.1M sodium phosphate buffer, pH 8. To the previous solution, 0.1 m1 of freshly prepared o-phthal- dialdehyde (0.1% w/v in absolute methanol, Sigma) was added, the 36 solution allowed to set at room temperature for 15-20 minutes, and then transferred to quartz cuvettes and the fluorescence measured at 420 nm resulting from activation at 350 nm in an AmincoR fluoro- meter (American Instrument Co., Silver Spring, Maryland). A standard curve was prepared from reagent GSH (Sigma). Assay_of Tissue for Malondialdehyde. Malondialdehyde was assayed in lung, liver and kidney tissue of 42 day old 100 ppm paraquat treated mice by a modificatiOn of the method of Placer et a1. (1966). Tissues were homogenized (Polytron) in 20 volumes of cold normal saline and centrifuged in the cold for 5 minutes at 400 xg. An 0.2 ml aliquot of the supernatant was added to a 25 m1 tube containing 1.3 ml of 0.2M TRIS-maleate buffer, pH 5.9, 1.5 ml of thiobarbituric acid reagent (0.8% w/v in 7% perchloric acid), and 0.15 ml ADP-Fe+++ (40 mM disodium adenosine diphosphate, Sigma, and 2.4 mM ferric ammonium sulphate) and placed in a boiling water bath for 10 minutes. Marbles were placed over the top of the tubes to prevent evaporation. After cooling for 10 minutes, 1.0 ml 1N sodium hydroxide and 3.0 ml of a 3:1 pyridine-butanol solution was added with a subsequent 15 second mix on a vortex blender. The absorbance of the resulting clear solution was read at 548 nm in a Spectronic 20R (Bausch and Lomb, Rochester, New York). Calculations of malondialdehyde concentrations were made using an extinction coefficient of 1.52 x 105. Histopathology. At 42 days, control, 50 and 100 ppm paraquat treated mice were sacrificed and lung, liver and kidneys fixed in 37 10% formalin. Five micron paraffin sections were prepared, stained with hematoxylin-eosin, and examined by light microscopy. Expgsure of Chronically Paraquat Treated Mice to Oxygen. Developing mice exposed to 0, 50 and 100 ppm paraquat beginning at day 8 of gestation were examined for sensitivity to 100% oxygen at 1 atmosphere on days 1, 28 and 42 postnatally. Oxygen exposure was accomplished by placing both control and treated mice in a clear plastic cage approximately 12 liters in volume (12x24x44 cm) with a plexiglass top and supplied with animal bedding, food and tap water. The one day old group was normalized to 10 newborns per mother and the mothers replaced after 48 hours of oxygen exposure with mothers which had recently littered to insure adequate nursing of the newborns. One hundred percent oxygen, which was humidified by flow over a water source prior to entry into the chamber, was maintained at a flow rate of 1.65 liters/min into the chamber. Oxygen concentrations in the chamber were measured to be 95 to 100% by a blood-gas analyzer (Radiometer, Copenhagen, Denmark). The sensitivity of the 28 and 42 day old paraquat treatment groups to 100% oxygen was measured by determination of the LTSO, or median time to death after initiation of oxygen treatment as calculated by the method of Litchfield (1949). The interaction of oxygen with newborn mice was determined by recording the number of mice dead after 120 hours of oxygen exposure. In other experiments, one day old mice obtained from timed pregnancies were injected with paraquat (methyliologen, Sigma), 5 mg/kg so, and placed in 100% oxygen along with saline injected 38 controls. Other paraquat injected one day old mice were exposed only to room air after injection. The number of mice that died in either oxygen or room air was recorded after 30 hours exposure. Other one day old mice obtained from dams treated with 100 ppm paraquat were injected with vitamin E (d-a-tocopherol in soybean oil, type I, Sigma), 25 IU per mouse sc, and placed in 100% oxygen along with other 100 ppm paraquat treated and nonparaquat exposed newborns injected with oil carrier. The number of mice dead after exposure to 100 hours of 100% oxygen was recorded. Interaction of Bromobenzene with Chronic Paraquat Treatment in Mice. The toxicity of bromobenzene in 42 day old 50 and 100 ppm paraquat treated mice was measured by determination of the bromoben- zene LTSO (Litchfield, 1949). Bromobenzene (Aldrich Chemical Co., Milwaukee, Wisconsin) dissolved in peanut oil was administered at 3100 mg/kg ip [LD85, based on the curve of the calculated LD50 of 1800 mg/kg (95% Confidence Interval, C.L.; 1333-2430)] and the time to death recorded. Distribution of 14C-Paraguat in Prenatal and Newborn Mice. Distribution of l4C-paraquat in organs of prenatal mice was examined after a 3.35 mg/kg ip dose. Paraquat (methyl-14C-paraquat, 36 mCi/ mmole, Amersham-Searle, Arlington Heights, Illinois) was diluted with nonradioactive paraquat (Sigma) for a final specific activity of 60 uCi/mg and administered to pregnant mice on day 16 of gesta- tion. At various times the dams were sacrificed and fetal tissues prepared for liquid scintillation counting by solubilization in 39 SolueneR (Amersham-Searle) followed by addition of 15 ml of toluene counting solution [5 g of 2,5—diphenyloxazole (PPO) and 200 mg of l,4-bis[2-(4-methy1-5-phenyloxazoly1]benzene (dimethyl POPOP) per liter of toluene. For the purpose of comparing the distribution in late fetal and neonatal animals, l4C-paraquat (36 mCi/mmole, Amersham- Searle, diluted with nonradioactive paraquat for a final specific activity of 10 uCi/mg) in one day old mouse organs was determined at various times after a 4.5 mg/kg sc dose of paraquat. Tissues were prepared for liquid scintillation counting as described above. Radioactivity was determined in a Packard 3380 liquid scintillation counter (Packard Instrument Company, Downers Grove, Illinois) and was converted to microgram levels of paraquat through use of external standard quench correction and assuming no in vivo metabolism of paraquat. Paraquat In Vitro Oxidation-Reduction Microsomes were prepared by cold homogenization of lung tissue from mice in 4 volumes of 1.15% potassium chloride containing 0.2% nicotinamide with 4 strokes of a loose fitting Teflon-glass Potter- type homogenizer (Gram, 1973). The homogenate was centrifuged at 15,000 xg for 20 minutes in the cold, the supernatant removed, poured through gauze and recentrifuged at 105,000 xg for 90 minutes. The supernatant was discarded and the microsomal pellet resuspended with 3 strokes of a Potter-type homogenizer in 0.05M tris buffer, pH 7.5, containing 50% glycerol. The final protein concentration was 4.6 mg/ml. The microsomal suspensions were divided into several 40 small vials, flushed with nitrogen, and stored at -20°C. Protein was assayed by the method of Lowry et a1. (1951). The aerobic oxidation of NADPH catalyzed by mouse lung micro- somes in the presence of paraquat was determined in a room tempera- ture incubation system containing 60 ug/ml microsomal protein, 1 x 10-4M NADPH (Sigma) and varying amounts of paraquat buffered in pH 7.5, 0.15M potassium phosphate. The incubation volume was 2.0 ml. Following the addition of NADPH, changes in optical density were recorded at 340 nm in a Coleman model 124 spectrophotometer (Hitachi, Ltd., Tokyo, Japan). The effect of addition of antibody to liver microsomal NADPH-cytochrome c reductase (IgG, prepared from serum of rabbits immunized with purified rat liver NADPH-cytochrome c reductase by the method of Pederson et a1., 1973) on the aerobic oxidation of NADPH was determined in an incubation system identical to the one just described. The paraquat concentration was 2.5 mM. The oxidation of NADPH was also recorded upon the addition of IgG pre-immune serum. In a related series of experiments, the anaerobic reduction of paraquat to the blue-colored paraquat radical catalyzed by mouse lung microsomes in the presence of NADPH was examined. Incubation conditions were identical to those in the aerobic experiments except that the incubations were carried out in an airtight cell flushed with oxygen-free nitrogen. The appearance of the paraquat free radical was measured at 395 nm in a Coleman 124 spectropho- tometer. The effect of addition of antibody to NADPH-cytochrome c reductase or pre-immune serum to the incubation system was also examined. 41 Paraquat-Induced In Vitro Lipid Peroxidation The ability of paraquat to initiate lipid peroxidation in vitro was investigated in an incubation system modified from Pederson and Aust (1973). Incubation mixtures contained 0.25M sodium chloride, 2.0 mM ADP, 0.12 mM ferric ammonium sulphate, 0.5 umole/ml lipid phosphorus, 0.2 mM NADPH, 60 ug protein/ml rat liver microsomal NADPH-cytochrome c reductase and paraquat in 0.25M TRIS buffer, pH 6.8. The total incubation volume was 5.0 ml. The incubation was conducted at 37°C under room air in an oscillating Dubnoff incubator. Aqueous suspensions of lipid phosphorus were prepared by anaerobic sonication of extracted microsomal lipid isolated from rats by the method of Pederson and Aust (1973). Sonication was accomplished by transferring an aliquot of the lipid stock solution to a plastic tube, removing the chloroform-methanol with a stream of nitrogen, addition of nitrogen saturated buffer, capping of the tube under nitrogen and placement of the tube in an ice-cold water bath near the tip of the sonifier (Branson, model 8125, Branson Instrument Co., Danbury, Connecticut) for 5 minutes at a sonication current of 10 amps. NADPH-cytochrome c reductase was prepared by the method of Pederson et a1. (1973). The incubation reaction was initiated by addition of NADPH after a 2 minute preincubation period at 37°C. At 1, 3, 5 and 8 minutes later, 1.0 m1 aliquots of the incubation reaction were removed to 2.0 m1 of thiobarbituric acid reagent (15% trichloroacetic acid, 0.375% thiobarbituric acid and 0.25N hydrochloric acid) to which 0.01 volume of 2% butylated hydroxytoluene in ethanol had been added immediately prior to use. 42 Malondialdehyde was determined by heating the mixture for 15 minutes in a boiling water bath to develop the color followed by cooling for 10 minutes. After centrifugation at 1000 xg, the absorbance of the supernatant of the assay mixture was determined at 535 nm in a Coleman Jr. Spectrophotometer (Coleman Instrument Co., Maywood, Illinois). A malondialdehyde standard curve was prepared from malondialdehyde tetraethylacetal (Aldrich Chemical Co.). The effect of superoxide dismutase (prepared from bovine erythrocytes by the method of Pederson and Aust, 1973) and 1,3- diphenylisobenzofuran (Aldrich Chemical Co.) on paraquat-induced lipid peroxidation was determined in incubations identical to those described above except for the addition of each agent separately or together to the reaction. Malondialdehyde was assayed in a similar fashion. Tissue Malondialdehyde Concentrations in Mice after Acute Paraquat Treatment Malondialdehyde concentrations were determined by the method previously outlined in "Assay of Tissue for Malondialdehyde" in mice after various doses of paraquat. Malondialdehyde was measured in mouse lung, liver and kidney either 3 and 7 days after 30 mg/kg ip (LDSO) paraquat or 4, 12 and 24 hours after an ip LD dose of 44 9o Ins/kg. Effect of Nutritional Deficiencies on Paraquat-Induced Lethalipy The 7-day intraperitoneal LD , or statistically calculated 50 dose lethal to 50% of the animals 7 days after a single injection 43 (Litchfield and Wilcoxin, 1949), was determined and used as a measure- ment of sensitivity to paraquat toxicity in mice. The paraquat (methyl viologen dichloride, Sigma) LD was determined in 4 groups 50 of mice: control mice which received laboratory animal chow (Wayne Lab Blox, Anderson Mills, Maumee, Ohio); mice fed diets deficient in selenium (General Biochemicals, Chagrin Falls, Ohio) or vitamin E (Draper diet, General Biochemicals) for 5 weeks; and mice treated with diethyl maleate (Aldrich Chemical Co.). 1.2 ml/kg in peanut oil, ip, 30 minutes before paraquat. The paraquat LD50 was also determined in other groups of mice fed either the basal selenium deficient diet supplemented with 0.1 and 2.0 ppm selenium (sodium selenite, Alfa Inorganics, Beverly, Massachusetts) or the basal vitamin E deficient diet supplemented with 45 and 1500 mg/kg diet vitamin E (a-tocopherol succinate, Sigma). In each group of mice at least 6 doses of paraquat were used to determine the LD 0’ with 5 6 mice per dose. The L0 in the 45 mg/kg vitamin E supplemented 50 group was determined using paraquat dichloride supplied by Dr. M. S. Rose, Imperial Chemical Industries, Ltd., Industrial Hygiene Research Laboratories, Alderley Park, Nr., Macclesfield Cheshire, SKlO 4TJ, England. The LD50 in the 1500 mg/kg vitamin E supplemented diet was determined using paraquat supplied by Schwarz—Mann, Orangeburg, New York. Mice fed selenium diets were maintained on double-distilled water for the 5 week period. Mouse body weights were recorded\at the onset and at the end of the 5 week feeding period. Supplemented diets were prepared by dissolving appropriate amounts of sodium selenite in water and vitamin E in ethanol such 44 that 10 ml of solution was added to each kilogram of diet during blending (5 minutes in a Blakeslee diet blender, Blakeslee, Inc., Chicago, Illinois) to give the designated dietary concentration. Assay of diets for selenium found 0.40 ppm selenium in Wayne Lab Blox, 0.01 ppm selenium in the selenium deficient diet and 0.098 ppm selenium and 1.938 ppm selenium in the 0.1 ppm and 2.0 ppm selenium supplemented diets, respectively. Glutathione (GSH) Peroxidase Assay. Liver and lung GSH peroxidase was assayed in mice fed selenium deficient and supple- mented diets for 5 weeks. Soluble liver and lung fractions con- taining this enzyme were prepared by homogenization (Polytron) of liver in 10 volumes and whole lung in 9 ml of cold 0.05M sodium phosphate buffer, pH 7.0. The homogenate was centrifuged at 750 xg for 10 minutes and then for 65 minutes at 100,000 xg (Chow and Tappel, 1974). The supernatant was assayed for GSH peroxidase by the method of Paglia and Valentine (1967). The assay of GSH peroxidase is coupled to the oxidation of NADPH in that GSH peroxidase reduces hydrogen peroxide with simultaneous formation of oxidized gluta- thione which, in the presence of an excess of GSH reductase, is reduced back to GSH with oxidation of NADPH. Varying amounts of soluble supernatant were incubated at room temperature with 0.10 ml 0.0084M NADPH, 0.01 ml GSH reductase (100 eu/mg protein in 1.0 ml, Sigma), 0.01 ml 1.125M sodium azide (to inhibit catalase, Sigma), 0.10 ml 0.15M reduced glutathione (Sigma), 0.10 ml 0.0022M hydrogen peroxide, and sufficient volume of 0.05M potassium phosphate buffer, pH 7.6, to bring the total incubation volume to 3.0 ml. The 45 reaction was started by addition of the hydrogen peroxide and the disappearance of NADPH followed on a Beckman DB-GT recording spec- trophotometer at 340 nm. Activity of the enzyme was expressed as nmoles of NADPH oxidized/minute/mg protein from calculations of NADPH concentrations based on an extinction coefficient for NADPH of 6.22 x 106 (Langdon, 1966). Protein was assayed by the method of Lowry (1951). Tissue Reduced Glutathione after Diethyl Maleate. GSH concen- trations were assayed in mouse lung and liver at various times after 1.2 ml/kg ip diethyl maleate by the method outlined pre- viously.in "Assay of Tissue for Reduced Glutathione." This was to confirm that diethyl maleate, which reduces tissue GSH by a conjugation reaction and subsequent excretion into the urine (Boyland and Chasseaud, 1970), was reducing tissue GSH at the dose used in the pretreatment regimen. Erythrocyte Hemolysis by Dialuric Acid. The vitamin E status of mice fed vitamin E deficient or supplemented diets for 5 weeks was assessed through use of the dialuric acid erythrocyte hemolysis assay (Friedman et a1., 1958). Blood was obtained by cardiac puncture from an ether anesthetized mouse and a 0.02 ml aliquot transferred to 5.0 ml of saline-phosphate buffer (1:1 mix of 0.1M dibasic sodium phosphate buffer, pH 7.4, and normal saline) in a 6 ml glass tube, mixed by inversion, and centrifuged at 500 xg for 10 minutes. The supernatant was removed and the erythrocytes resuspended in 4.5 ml of saline-phosphate buffer. Aliquots of 1.0 46 ml of suspended erythrocytes were transferred to each of 3 10 ml glass tubes containing the following: tube 1, 1.0 ml of a 0.01 mg/ ml solution of dialuric acid in saline phosphate buffer; tube 2, the same as tube 1; and tube 3, 1.0 m1 of saline-phosphate buffer. In the assay of the vitamin E supplemented mice, tubes 1 and 2 contained 0.05 mg/ml dialuric acid in saline-phosphate buffer. Duplicates were run with tube 1. All tubes were incubated at 37°C for one hour in a slowly oscillating Dubnoff incubator followed by incubation at room temperature for one hour. All tubes were inverted once every 15 minutes over the period of the incubation to insure the erythrocytes remained in suspension. Following the incubation, 5.0 ml saline-phosphate buffer was added to tubes 1 and 3 and distilled water to tube 2. Thus, tube 1 served as the variable hemolysis, tube 2 the total hemolysis, and tube 3 the basal hemolysis. The tubes were mixed gently, centrifuged, and the supernatant absorbance read at 415 nm in a Beckman DB-GT spectro- photometer. Erythrocyte hemolysis was calculated by dividing the difference in absorbance of tube 1 and 3 by the absorbance differ- ence of tube 2 and 3. Histopathology. Sections of lung and liver of mice fed selenium diets were prepared for light microscopic examination as previously described in "Histopathology." Tissue Reduced Glutathione and Lipid Soluble Antioxidant Concentrations after Acute Paraquat Treatment The concentrations of GSH in mouse lung and liver were determined at 12, 24, 36 and 48 hours after a 30 mg/kg ip dose 47 (LDSO) of paraquat (Sigma) by the method described in "Assay of Tissue for Reduced Glutathione." Lipid soluble antioxidants were measured in lung and liver of mice by a modification of the colorimetric method of Glavind (1963). The principle of the assay is based on the decolorization of the violet-colored free radical, 1,1-diphenyl-Z-picrylhydrazyl (DPPH, Sigma) upon acceptance of an electron from antioxidants in an assay solution. Lipid soluble antioxidants were measured at various times after a 30 mg/kg ip (LDSO) dose of paraquat and 12 hours after ip doses of 25 mg/kg (LD ), 30 mg/kg (LDSO) and 44 mg/kg (LD9 ). 25 0 Wet lung weights were recorded at the time of sacrifice. In other experiments, percent water content of control and paraquat treated lungs was determined after oven drying. Whole lungs or 300-400 mg of liver were removed to ice-cold saline and subsequently homogenized (Polytron) in 10 ml nitrogen saturated chloroform- methanol (2:1) in the presence of 200-300 mg of anhydrous sodium sulfate. The homogenates were filtered with suction and the fil- trate evaporated under nitrogen to a volume proportional to the tissue weight. The homogenates and solutions were kept on ice until the reaction with DPPH, which was done at room temperature. One milliliter of 8 mg% DPPH in chloroform was added to 3.0 ml of tissue extract or solvent blank and after a 20 minute room tempera- ture incubation, the absorbance read at 517 nm in a Beckman DB-GT spectrophotometer. The sample was then decolorized by addition of 2 drops of 0.5% pyrogallol (Sigma) in absolute ethanol and the absorbance redetermined at 517 nm. The difference in the initial 48 absorbance and the decolorized absorbance of the sample were desig- nated 65' The difference in the initial and decolorized absorbance of the solvent reference was designated Gr' The decrease in absorb- ance induced by the sample was calculated from the difference of Gr-Gs. Calculations of tissue lipid soluble antioxidants were based on a standard curve prepared with vitamin E standards (d-a— toc0pherol, type I, Sigma). Addition of known amounts of vitamin E to homogenates and subsequent assay resulted in a vitamin E recovery of 101.0 :_3.l%. Interaction of Paraquat with the GSH Peroxidase System Enzymes Effect of Chronic Paraquat Treatment on Enzyme Activity. Male rats were allowed access to 100 ppm paraquat ad libitum for 3 weeks in the drinking water, with housing in clear plastic cages in groups of four. Rats were killed at the end of the 3 week exposure period by a blow to the head. Lungs were perfused in situ with ice-cold saline, removed, and homogenized in the appropriate buffer. Liver was removed with perfusion and homogenized. GSH peroxidase was assayed in lung and liver tissue by the method described previously in "Glutathione (GSH) Peroxidase Assay." Assays for GSH reductase and glucose-6-phosphate dehydrogenase are described below. GSH Reductase Assay. Rat lung and liver soluble fractions were prepared as described in "Glutathione (GSH) Peroxidase Assay." GSH reductase was assayed by the method of Racker (1955) with an incuba— tion system containing 0.10 ml soluble fraction, 0.10 ml 0.0084M 49 NADPH, 0.10 ml 3% bovine serum albumin (Sigma), 0.10 ml 6% (w/v) oxidized glutathione (Sigma) and 2.60 ml 0.05M dibasic potassium phosphate buffer, pH 7.6, for a total 3.0 ml volume. The incuba- tion was conducted at room temperature and initiated by addition of oxidized glutathione with rapid mixing by inversion. The oxidation of NADPH was followed at 340 nm in a Beckman DB-GT recording spectrophotometer. The activity of the enzyme was expressed as nmoles NADPH oxidized/minute/mg protein using an extinction coef- ficient for NADPH of 6.22 x 106. Protein was assayed by the method of Lowry (1951). Glucose-6-Phosphate Dehydrogenase Assay. Rat soluble lung fractions were prepared as described in "Glutathione (GSH) Peroxi- dase Assay." G1ucose-6-phosphate dehydrogenase was measured by the method of Langdon (1966) in a room temperature incubation system containing 0.10 ml 1M TRIS-hydrochloride buffer, pH 7.5, 0.10 ml 2 x 10-2M glucose-G-phosphate (Sigma), 0.10 ml 2 x 10-3M NADP (Sigma), 0.10 ml 0.2M magnesium chloride, 0.10 ml soluble fraction and a sufficient volume of distilled water to yield a final incuba- tion volume of 3.0 ml. The reaction was started by the addition of glucose-6-phosphate and the formation of NADPH recorded at 340 nm in a Beckman DB-GT recording spectrophotometer. The enzyme activity was expressed as nmoles NADP reduced/minute/mg protein using 6.22 x 106 as the extinction coefficient. Protein was assayed by the method of Lowry (1951). Cross-Tolerance of Oxygen with Paraquat. Male rats were placed in 85% oxygen for 7 days. Oxygen exposure was accomplished by 50 placing 4 rats in the oxygen chamber described in "Interaction of Oxygen with Paraquat Treated Mice" in which an 85% oxygen concentra- tion was achieved by mixing 99.9% pure oxygen (2.78 liters/minute) and 90% pure nitrogen (0.47 liters/minute) through gas flowmeters. The chamber air was determined to be 86% oxygen by analysis in a blood-gas analyzer (Radiometer). The plastic chamber was cleaned and fresh food added every 36 hours. Control rats were exposed to room air. Following 7 days of exposure, control and 85% oxygen exposed rats were injected with 45 mg/kg ip paraquat (Schwarz-Mann) and the time at death recorded. The paraquat LT was calculated 50 by the method of Litchfield (1949). Interaction of Phenobarbital with Paraquat Mice were maintained in clear plastic shoebox cages with access to food and 0.1% phenobarbital water (0.1% w/v sodium phenobarbital, pH 7.5, Mallinckrodt, St. Louis, Missouri) ad libitum for 10 days. The paraquat (Schwarz-Mann) LD (Litchfield 50 and Wilcoxin, 1949) was determined both immediately after the 10 day exposure with the mice allowed continued access to the pheno- barbital and when the mice were transferred to tap water for 24 hours before paraquat administration with subsequent continued access to tap water. In other experiments, the paraquat LDso was determined in mice pretreated with 50 mg/kg ip phenobarbital 30 minutes before paraquat. Distribution of l4C-Paraquat in Phenobarbital Treated Mice. l4C-Paraquat, 30 mg/kg ip (Amersham-Searle, specific activity of 51 injected solution 1.3 uCi/mg), was administered to mice pretreated for 10 days with 0.1% phenobarbital in the drinking water. At various times after administration, blood was obtained from ether anesthetized mice by cardiac puncture and plasma separated by centrifugation. Simultaneously, samples of lung, liver and kidney were obtained. All samples and plasma were prepared for liquid scintillation counting as described in "Distribution of 14C-Paraquat in Prenatal and Newborn Mice" and counted in a Packard 3380 liquid scintillation counter with quench correction. Statistics Statistical evaluation of the data was by the Student's Eftest or analysis of variance (completely randomized design) with dif- ferences among means analyzed by the least significant difference method (Sokal and Rohlf, 1969). The level of significance was chosen as p<.05. RESULTS Toxicity of Chronically Administered ?” Paraquat in Developing Mice Paraquat, when placed in the drinking water of pregnant mice V-r- m t'Iv .7 l and continued at 50 and 100 ppm from day 8 of gestation to 42 days postnatally, did not alter the average litter body weight compared l “ r.\vV_ to controls at any time during the experiment (Figure l). The average litter body weight on day 1 after birth was 1.69 :_0.15 g, 1.69 i 0.09 g and 1.95 :_0.20 g for the controls, 50 and 100 ppm paraquat treated groups, respectively.. At 42 days of age the weights were, respectively, 28.8 :_0.4 g, 30.3 :_0.8 g and 26.9 i 1.7 9. Pregnant mice receiving 150 ppm paraquat died before delivery of newborns occurred (death usually by day 16 of gestation). Total water consumption was not significantly altered among the 0, 50 and 100 ppm paraquat treated groups when measured over the 14 day period from day 28 postnatally (weaning) to 42 days postnatally (Table l). The average daily dose of paraquat was calculated assuming an average body weight of 25 g and was 16 and 27 mg/kg/day for the 50 and 100 ppm paraquat treated mice, respectively. One hundred ppm paraquat significantly increased postnatal mortality compared to 50 ppm paraquat and controls from day 7 after birth, when mortality was 33.3 :_8.8%, up to a 42 day mortality of 52 53 Figure 1. Effect of 50 and 100 ppm paraquat in the drinking water from day 8 of gestation to 42 days after birth on the post- natal growth of mice. Each point is the mean mouse body weight within litters for three to four litters. There were 10 mice/ litter on day 1. 54 N? an ...—.83 cur—1 “>10 mu pd #— 0", 2.... 02 .I.... it 8 III 8‘pzou I .8388 I 8 :9 (mm!) mm moi mum 3 8 Figure 1 55 Table 1. Water consumption by developing mice treated with para- quat from day 28 to day 42 postnatally Water co?sumption 2 Total consumption Consumption/day Approximate dose Treatment (ml/mouse) (ml/mouse) (mg/kg/day) Control 110 L". 5 7.8 --- 50 ppm. 118 :_12 8.4 16 100 ppm. 95 :_ 4 6.8 27 l Mean :_S.E. for 3-4 litters; measured from day 28 to 42 postnatally. Assume average body weight of 25 g. 66.7 :_3.3% (Figure 2). Mortality among the 50 ppm paraquat group and controls was not significantly different and remained below 7% over the course of the experiment. Mortality induced by 100 ppm paraquat appeared biphasic, with an initial rapid increase to 33.3 :_8.8% during the first 7 days after birth, which then plateaued up to day 21 when mortality was 36.7 :_8.8%. The plateau phase was followed by a second increase prior to weaning on day 28, reaching a final mortality of 66.7 :_3.3% on day 35. Paraquat treatment did not affect the number of live pups born among the treatment groups, with 12 :_1, l3 :_1, and 11 3.2 pups born per litter to the control, 50 ppm and 100 ppm treatment groups, respectively. Experiments in which developing mice were exposed to 100 ppm paraquat from day 8 of gestation to 28 days postnatally and then transferred to tap water until 42 days after birth resulted in a 56 Figure 2. Effect of 50 and 100 ppm paraquat in the drinking; water from day 8 of gestation to 42 days after birth on the post—- natal mortality of mice. Each point is the mean percent mortalit:§r within litters for three to four litters. There were 10 mice/ litter on day l. 57 {IIIIi E3 umhu< “>10 N? 3 ON _ N S. h — Cl °|°|°\ ooooo ocvo o..====== O..........-... 0 .......=.-.. O .......-=..._ 06%. ‘\° \.. \. II\0 o lla'°\ ..\ 00\ .v \ it 8.. Qualo \00 E 8 033380 615.209 0'0 OIIII o 8 8 MI'IVLIOW mu WV 8 Figure 2 58 mortality of 26.7 i 12.0% by day 28 (Table 2) . This mortality level was not significantly different from the 28 day mortality (46.7 i 12.0%) of mice continuously exposed to 100 ppm paraquat throughout Table 2. Effect of 100 ppm paraquat on postnatal mortality in mice when administered in the drinking water during selected periods of development n“: “T V Percent postnatal mortality Period of paraquat exposure Day 28 Day 42 Day 8 of gestation to day 42 46.7 i 12.0 66.7 i 3.3 , postnatally :_ Day 8 of gestation to day 28 26.7 i 12.0 26.7 i 12.0 postnatally2 Day 28 to day 42 postnatally3 0.0 i 0.0 46.7 i 14.5 1Mean i 8.23. of 3 litters, 10 mice/litter or day 1 post- natally. 2Tap water from day 28 to day 42 postnatally. 3Tap water from day 8 of gestation to day 28 postnatally. development. Furthermore, transfer to tap water resulted in a 42 day mortality that was unchanged from day 28 and thus eliminated the second rapid increase in 100 ppm paraquat mortality depicted in Figure 2 (Table 2) . Exposure of developing mice to 100 ppm para- quat only from days 28 to 42 postnatally resulted in 46.7 _~I_-_ 14.5% mortality, which was not significantly different from the 30% increase in mortality observed in the second 100 ppm paraquat mortality phase of Figure 2 (Table 2) . 59 Histopathology. Lung sections from 42 day old 100 ppm para- quat treated mice showed extensive alveolar consolidation and collapse, and areas of thickening of intra-alveolar septa. Edema fluid was seen in a few alveoli (Figure 3). Lung sections from 50 ppm and control mice did not show any significant pathological changes. Examination of liver and kidney sections of all treat- ment groups also did not reveal any significant pathological changes. Tissue Malondialdehyde of 42 Day Old 100 ppm Paraquat Treated ‘Migg, Assay of lung, liver and kidney tissue for malondialdehyde revealed no significant differences in the lipid peroxidation by- product compared to controls, when measured at 42 days postnatally (Table 3). Malondialdehyde levels were approximately equal in lung Table 3. Malondialdehyde concentrations in tissues of mice treated with 100 ppm paraquat from day 8 of gestation to 42 days .\ ———.- u—‘y. _ "l‘l‘kfi'. ‘1': . postnatally Malondialdehyde concentrations1 (nmoles MDA/g wet weight) Tissue Control 100 ppm paraquat Lung 374 ;_I-_ 38 368 i 19 Liver 402 _-!_-_ 60 328 -_+_ 24 Kidney 707 i 39 776 i 17 lMean -_i-_ S.E. of 4 determinations. and liver tissue while that in kidney tissue was approximately twice the levels observed in the other two tissues examined. is. 60 Figure 3. Lung tissue from 42 day old control mice and mice treated with 100 ppm paraquat from day 8 of gestation to 42 days postnatally. Upper panel is from control mice; lower panel from paraquat treated mice. Original magnification, x100. Figure 3 62 Interaction of Oxygen with Chronic Paraquat Treatment. Paraquat significantly enhanced the sensitivity to oxygen toxicity of 42 day old mice which received 50 and 100 ppm paraquat throughout develop- ment (Table 4). The LT for 50 and 100 ppm paraquat treated mice 50 was 108 and 40 hours, respectively, compared to the control LT50 of Table 4. Effect of paraquat in the water from day 8 of gestation to various days postnatally on the survival of mice exposed to 100% oxygen Begin 02 LTSO 95% Potency Treatment exposure (day) (hours) C.L. ratio Control 28 181 (156-210) --- 50 ppm 28 180 (142-229) 1.0 100 ppm 28 121 (108-135) 1.51 Control 42 160 (126-203) --- 50 ppm 42 108 (81-144) 1.51 100 ppm 42 4o (30- 54) 4.01 1Significantly different from respective control, p<.05. 160 hours. The LT50 curves for both treatment groups were parallel to each other and controls, with calculated potency ratios of 1.5 for 50 ppm paraquat versus controls and 4.0 for 100 ppm paraquat versus controls. The concentration of paraquat in lung tissue of the 100 ppm paraquat treated mice was less than 0.2 ug/g lung tissue (wet weight) and was below detectable levels in 50 ppm paraquat lung tissue. 63 At 28 days after birth, an increased sensitivity to 100% oxygen exposure was detected only in 100 ppm paraquat treated mice (Table 4). The LT of 100 ppm paraquat treated mice was significantly 50 reduced to 121 hours, compared to an LT of 180 hours for 50 ppm 50 paraquat mice and 181 hours for control mice. A potency ratio of 1.5 was calculated for 100 ppm paraquat mice versus controls. One day old mice, which were exposed to paraquat only by pre— natal placental transfer, also were sensitized to oxygen toxicity (Figure 4). Mortality of one day old mice whose mothers received 100 ppm paraquat was significantly increased to 53.5 :_7.0% after 120 hours oxygen exposure compared to 25.8 :_5.1% and 24.2 :_3.2% mortality in 50 ppm paraquat treated and control mice, respectively. In other one day old mice whose mothers received 100 ppm para- quat in the drinking water from day 8 of gestation to day 19 of gestation, administration of subcutaneous vitamin E resulted in a nonsignificant but definite trend towards decrease in oxygen induced lethality (Table 5). After 100 hours of oxygen exposure, mortality in the nonparaquat treated control mice was 25.0%, in the 100 ppm paraquat treated one day old mice, 71.4%, and 27.3% in 100 ppm paraquat mice which received subcutaneous vitamin E. One day old mice whose mothers received nontreated tap water during gestation were significantly sensitized to exposure to a 100% oxygen environment when the newborns were injected with paraquat, 5 mg/kg sc, immediately before entering the oxygen chamber (Table 6). Mortality after 30 hours of 100% oxygen was zero in water injected control mice but was 71.4% in paraquat treated newborns. 64 Figure 4. Effect of 50 and 100 ppm paraquat in the drinking; water of pregnant mice from day 8 of gestation to day 19 of gestation on the survival of one day old mice exposed to 100% oxygen for 120 hours. Each treatment represents the mean percerrt; mortality for three to four litters with 10 mice/litter on day 1.- * indicates significantly different from control and 50 ppm paraquat, p<.05. PERCEN'I’ DSAD 100 80 20 Control 65 50mm Figure 4 100 ppm 66 Table 5. Effect of vitamin E on survival of one day old mice pre- treated with 100 ppm paraquat on days 8-19 of gestation in 100% oxygen No. mice dead Percent Treatment No. mice after 100 hr of 02 dead Control 20 5 25.0 100 ppm 7 5 71.4 I 100 ppm + 11 3 27.3 vit. E2 1100% oxygen at 1 atm. EW‘ 7; . 2d-o-Tocopherol in soybean oil (type I, Sigma), 25 Inter- national Units per mouse, sc. Table 6. Effect of acute paraquat administration to one day old mice on survival in 100% oxygen No. mice dead 1 Percent Treatment No. mice after 30 hr 02 dead Control 32 0 o 2 3 Paraquat 28 20 71.4 + 02 2 Paraquat 26 0 0 + air 1100% oxygen at 1 atm. 25 mg/kg, sc. 3 Significantly different from controls, p<.05. 67 Newborn mice injected with paraquat, 5 mg/kg so, and allowed to remain in room air also had no mortality 30 hours later. Distribution of 14C-Paraguat in Prenatal and Newborn Mice. The elimination of radioactivity from fetal mouse organs following maternal administration of 14C-paraquat at 3.35 mg/kg ip on day 16 of gestation was similar in fetal lung, liver and kidney (Figure 5) . No significant retention or elevation of paraquat occurred in fetal mouse lung tissue compared to fetal liver and kidney for 72 hours after paraquat administration. In one day old newborn mice given paraquat, however, radioactivity in neonatal lung was significantly elevated compared to liver and kidney from 8 hours to 96 hours after paraquat administration (Figure 6). Furthermore, detectable levels of paraquat in neonatal lung were observed up to 96 hours after treatment, which was 24 hours after paraquat could no longer be detected in neonatal liver and kidney. Paraquat In Vitro Oxidation-Reduction Mouse lung microsomes catalyzed the oxidation of 6 nmoles of NADPH per minute per mg protein. In the presence of 1.0 mM para- quat this value increased to 62 nmoles of NADPH oxidized per minute per mg protein. Under aerobic conditions, NADPH oxidation rates (measured at 340 nm) increased linearly with increasing paraquat concentrations, with a maximal oxidation rate reached at 1.5 mM paraquat (Figure 7). Under anaerobic conditions, in which the appearance of the blue-colored paraquat radical was measured at 395 nm, paraquat reduction by mouse lung microsomes and NADPH also 68 I 2 Figure 5. Concentration of paraquat in fetal mouse organs at various times following 3.35 mg/kg ip 14C-paraquat on day 16 of gestation. Each point represents the mean tissue concentra- tion obtained from 3 fetal litters, one pooled litter/determinai:jL<=>11«t 29h<¢hmv=tn< h§=z< a... 8. cm. on. on. 2. ON 0 333333333 0 32‘ 0333330 :55. 35.55.9805 rap» on ... ollollololoL oo— Figure 9 79 Figure 10. Inhibition of the anaerobic reduction of paraquat catalyzed by mouse lung microsomes and NADPH by antibody to NADPH-cytochrome c reductase. The incubation contained 1 x 10'4M NADPH, 60 ug/ml lung microsomal protein, 2.5 mM paraquat, and varying amounts of either NADPH-cytochrome c reductase antibody or antibody pre-immune serum. 100 80 8 mam o: CONTROL o—o—o—O—O 80 o——-0 Pro-immune serum 0 ululuuuuo M'M ""00!annuuuuuuuuuluau-nu O IIIIIINIIIIIOIIIIIIIIIIIIIIIII O .30 mg ANTIBODY. Figure 10 .l‘e--. 81 * Table 7. Paraquat-induced in vitro lipid peroxidation Paraquat concentra- tion per incubation Malondialdehyde formedl Percent increase in malondialdehyde mixture (nanomoles/min/ml) formed 0 0.37 .01 0 -6 10 M 0.43 .03 16.2 10'5M 0. 60 . 022 62 . 2 10’4M 1.21 .092 227.0 It! . 0 * Incubation mixtures contained 0.25M NaCl, 2.0 mM ADP, 0.12 mM Fe(NH4)2(SO4)2, 0.5 moles/ml lipid phosphorus, 0.2 mM NADPH, 60 ug/ml liver microsomal NADPH-cytochrome c reductase, and paraquat in 0.25M TRIS buffer, pH 6.8. in an oscillating Dubnoff incubator under air. volume was 5.0 ml. 1 Mean :_S.E. of 3 determinations. Incubations were conducted at 37°C Total incubation 2Significantly different from no paraquat, p<.05. 82 Table 8. Inhibition of paraquat-induced in vitro lipid peroxidation by superoxide dismutase and l,3-diphenylisobenzofuran 4 Malondialdehyde Percent of 10- M formed (nanomoles/ paraquat incu- Incubation min/ml) bation I . -4 10 M paraquat 0.84 :_.09 --- Plus superoxide dismutase: 2 20 uM 0.60 1.062 71.4 r 60 HM 0.28 :_.04 33.3 E E Plus 1,3-diphenyliso- v benzofuran: 2.0 uM 0.72 1.022 85.7 10.0 uM 0.45 :_.03 53.6 Plus superoxide dismutase 2 3 (20 um) and 1.3-dipheny1- 0.11 :_.02 ' 13.1 isobenzofuran (10.0 uM) * Incubation conditions were identical to those given in Table 7 except as described above. 1Mean :_S.E. of 3 determinations corrected for no paraquat control. 2Significantly different from lO-4M paraquat, p<.05. 3Not significantly different from basal rate, p>.05. 83 was greater than that achieved when either agent was incubated alone (Table 8). Tissue Malondialdehyde Concentrations in Mice after Acute Paraquat Treatment Administration of an LD50 dose of paraquat, 30 mg/kg ip, in mice did not induce any significant alterations in malondialdehyde concentrations in lung, liver and kidney tissues when measured 3 and 7 days after paraquat administration (Table 9). Malondialde- hyde concentrations in the kidney tissue of these mice were elevated compared to lung and liver, which was similar to the observations in 42 day old mice found in Table 3. Table 9. Malondialdehyde concentrations in tissues of mice after acute paraquat treatment Malondialdehyde (nmole MDA/g wet weight)1 3 days after paraquatzi 7 days after paraquat2 Tissue Control Treated Control Treated Lung 483 i 55 420 i 34 420 i 40 581 i 145 Liver 477 i 77 523 _+_ 90 443 i 106 546 _+_ 38 Kidney '-- --- 632 + 102 759 + 87 1Mean :_S.E. of 3 determinations. 2Paraquat, 30 mg/kg ip (LDSO). When paraquat was administered at the LD dose of 44 mg/kg ip, 90 however, malondialdehyde concentrations were significantly elevated in liver tissue at 12 and 24 hours after paraquat administration to l '7- mm? 1 : l www.1- 84 371 1.34 and 278 :_32% of controls, respectively (Figure 11). No change in malondialdehyde was observed at 4 hours in liver or at 4, 12 and 24 hours after paraquat in lung and kidney. The 12 and 24 hour groups of paraquat treated mice were extremely ill at the time of sacrifice and no attempts at eating or drinking were observed from a short time after injection until sacrifice. Effect of Nutritional Deficiencies on Paraquat-Induced Lethality '. LL. '4' .il g... Mice fed diets deficient in selenium or vitamin E for 5 weeks S or pretreated with the tissue GSH reducing agent, diethyl maleate, resulted in paraquat LD '5 of 10.4 mg/kg in selenium deficiency, # 50 9.2 mg/kg in vitamin E deficiency, and 9.4 mg/kg with diethyl maleate pretreatment (Table 10). The paraquat LD50 in each of the 3 nutritionally deficient states was significantly decreased come pared to the 30 mg/kg control LD (mice fed Wayne Lab Blox). The 50 log-probability plots of the dose-lethality data for the various treatments yielded curves that were parallel to that of control. Supplementation of the basal selenium deficient diet with 0.1 and 2.0 ppm selenium (as sodium selenite) returned the paraquat L050 to the control value (Table 10). Addition of 1500 mg/kg vitamin E to the basal vitamin E deficient diet also returned the paraquat LDso to the control L950 (Table 10). The LDSO in the 1500 mg/kg vitamin E supplemented group was determined with paraquat supplied by Schwarz-Mann, while the other previously described LDSO's were determined with paraquat from Sigma. There were apparently no dif- ferences in the paraquat from the two suppliers as the Schwarz- Mann control L050 was 30.0 mg/kg (26.3-34.2) and the two control 85 Figure 11. Tissue malondialdehyde concentrations in mice after acute paraquat treatment. Paraquat was injected at 44 mg/kg ip (LD90). Each bar represents the mean :_S.E. of 4 determinations. An asterisk indicates significantly different from control, p<.05. 86 0 o o o o o o 0.0 Qo.0.0.0.0.0.9.0.0.o.o.9.0.04o.04910404940494040co49404040404ooooocoooooooorororoworo. cocooooooooooooo 00000000000 0000000000000000000 wowowowowowowowoooo 44000909292! Wouooooooooooooooooooooo ...T 400 000000 000000000000.0009000000090»...oononouonououonouonouououenouowowowowowowowowowow .0.0.0.0.0.0.0.0.0.0,0 0 0 0 0 O .6 m9>2u0a<320¢<2 0.00000000000 ’.- O at ..OflhZOu 00000090000000.0000. 000000000000 09000 000000000000000000000000000000000000000 0 0 0 0 0 0 00000000000000000 0 0 0 0 0 0 0 0 0 . 0 0 0 0 0 0 0 0 0 0 0 00000000000000000000000D0D0D0D0Dt ’0) b. ’D0b0b0kbp0b050)’ 0 0.0.0.000 0.0.0 0.0.0.0.0 66.0.0.0...064040 O 0000“. ooooonoooooooooo W0W0W0W0W0W0W0W0‘0‘0‘0‘0‘000000000 0 00.000.000.00000 50 uOZ.>.v.bA 0 . 4 4 4 4 4 4. - 00000000004004 4 4 404040404 «... 4 4..4.4«ow.».003.».«o«.«¢o».«.«.«03.30».«.«o».«.«.«.».u.«.w« o wWWWEaQWwQQQQQQwQQQXQQWQEQQQQQQQEEQEQEQQQ. p.’.’.’.’ ’ R.’ R.’ R R R R R R R F.’.}.>.}.D.D.D.D.b.v .......... ... ... . . O A .. E3 0 00 0 00 .020 CONTROL 0 00 0’0 fi 3 2 a I 6.. a J 2.. a 0 I. L ... 50:3 :3 Eu\muz