XENOBIOTICA,
1991, VOL.21,
NO.
8, 1033-1040
Review The potential of antioxidant enzymes as pharmacological agents in vivo J. F. T U R R E N S Department of Biomedical Sciences, College of Allied Health, University of South Alabama, Mobile, Alabama 36688, USA
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
Received 26 March 1990; accepted 12 August 1990 1. Oxygen radicals have been associated with a number of unrelated pathological processes including ageing, radiation sickness, inflammation, oxygen toxicity, reoxygenation of ischaemic tissues, etc. The partial reduction of oxygen to superoxide anion (0;)and H,O, leads to the formation of more deleterious species such as hydroxyl radical (OH.) starting a chain reaction ultimately causing lipid peroxidation and cell death.
2. T o prevent the increased steady-state concentration of oxygen radicals many researchers have designed potential treatments including the i.v. injection of antioxidant enzymes or .enzyme derivatives with longer half-life in circulation (i.e. enzymes encapsulated in liposomes or covalently modified). 3. Tissue distribution and half-life in circulation depend upon the type of enzyme being used as well as whether the enzyme is or is not in its native form. 4. This review comments on some of the scenarios where these enzymes have been utilized, and discusses relevant problems of stability of different enzymes in circulation.
Production of oxidants in biological systems T h e partial reduction of oxygen to intermediates such as superoxide anion (0;) and hydroxyl radical (OH-) by biological systems was first proposed by Gerschman et al. (1954). This proposal was considered only an interesting hypothesis and it was almost forgotten for 15 years, until McCord and Fridovich (1969) discovered the enzyme superoxide dismutase (SOD). T h e identification of an enzyme suggested that its substrate had to be a normal cell metabolite and, very soon, hundreds of reports were published identifying a variety of sources of 0; and/or H 2 0 2 ,including free enzymes and multi-enzymic systems (Chance et al. 1979). There are several intracellular processes leading to the formation of partially reduced oxidative species. T h e simplest reaction involves the reduction of an oxygen molecule by a single electron producing O l , which in turn disproportionates to H,O, and 0, either spontaneously or in a reaction catalysed by the enzyme SOD (McCord and Fridovich 1969). Two components of the mitochondria1 respiratory chain (NADH dehydrogenase and ubisemiquinone) are examples of redox systems directly producing 0 2 (Turrens and Boveris 1980, Turrens et al. 1982, 1985). Alternatively, H 2 0 , may be generated in a single step through concerted transfer of two electrons to an oxygen molecule, as in the reaction catalysed by the enzyme glucose oxidase (Muller 1987). There are also enzymes such as xanthine oxidase capable of forming simultaneously 0; and H,O, in different proportions depending on the p H at which the reaction is carried out (Fridovich 1970). Superoxide anion formation may start a cascade of oxidative reactions, not as an oxidant itself but rather as a reductant. Inside the cell superoxide anion may reduce iron bound to ferritin which then is released from the protein in the ferrous state. Reduced iron may then react with H 2 0 2in a Fenton-like reaction, producing OH.. This species may in turn abstract a hydrogen from any biological molecule including 0049-8254/91 $ 3 4 0
0 1991 Taylor & Francis Ltd.
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
1034
J . F. Turrens
nucleotides, amino acids and unsaturated lipids. T h e latter reaction is the first in a series of reactions leading to lipid peroxidation. Not only can the combination of H,O, and 0; be deleterious for biological systems, but each species can eventually be involved in toxic reactions in the absence of the other. Hydroxyl radical may also be generated from the reaction between H,O, and a semiquinone, in a process that does not require 0; (Koppenol and Butler 1985). On the other hand, 0, may directly inactivate several enzymes, including catalase and glutathione peroxidase, which are needed to eliminate H,O, from the intracellular milieu (Kono and Fridovich 1982, Blum and Fridovich 1985). Cells are well protected against oxidative stress by antioxidant enzymes (SOD, catalase and glutathione peroxidase) and also by other non-protein antioxidants such as vitamin E and carotenes (Chance et al. 1979). T h e extracellular environment has much lower antioxidant defences: only a single enzyme known as extracellular SOD (EC-SOD) has been purified from most extracellular fluids and is present in relatively low concentrations (Marklund et al. 1982). There may be a strong evolutionary reason for keeping a low antioxidant activity in the extracellular space, since it is in this compartment that polymorphonuclear cells form the first line of defence against infectious organisms, producing 0, and H,O,. Thus, increased extracellular antioxidant activity would ultimately be deleterious for the host, as it would also decrease the natural defences against pathogens, favouring opportunistic infections. Under pathological conditions, however, there may be a need for increased extracellular antioxidant defences. For example, in adult respiratory distress syndrome (ARDS) polymorphonuclear cells become activated in the intravascular space, damaging the endothelium, increasing capillary permeability and causing oedema and swelling (Elliot et al. 1985). Also in many autoimmune diseases, the deposition of immune complexes leads to complement activation, producing chemoattractants (C5a) that bring polymorphonuclear cells to the area. In these cases neutrophils mediate the inflammatory process (Oldham et al. 1988). Intracellular oxidative injury may also lead to migration and activation of polymorphonuclear cells. As the intracellular steady-state lipid peroxidation increases, cells release partially oxidized molecules that act as chemoattractants for polymorphonuclear cells (Petrone et al. 1980), which in turn migrate to the area, become activated magnifying the injury and ultimately destroy the tissue. This occurs during lung oxygen toxicity (Crapo et al. 1983), secondary to the increased generation of oxidative species by lung endothelial and epithelial cells. This inflammatory response is responsible in part for the increased pulmonary oedema accompanying lung oxygen toxicity. It is in these cases that an increased intravascular antioxidant activity may be beneficial to the tissue. Oxidative injury is the common denominator in several diseases, including lung oxygen toxicity, hyperbaric injury to the central nervous sysystem, intoxication with xenobiotics (i.e. paraquat or carbon tetrachloride), radiation sickness, and more recently, reoxygenation of ischaemic organs (Chance et al. 1979, McCord 1985). Although oxygen radicals are involved in these pathological processes, they may be generated through different processes and in several cell compartments (i.e. cytoplasm, mitochondria, cell membrane or into the capillary lumen). Thus, a major problem for developing an effective antioxidant therapy is how to target antioxidant enzymes to the appropriate compartment where the oxidative species are being generated.
1035
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
Antioxidant enzymes in vivo
Intravenous injection of native or modified antioxidant enzymes Several research groups have considered the idea of using antioxidant enzymes as pharmacological agents, to specifically modulate the steady-state concentration of 0, or H 2 0 2 .The first problem with this approach is that the antioxidant enzymes most commonly used (Cu,Zn-SOD) and catalase have an extremely short half-life in circulation (6-8min) (McCord and Wong 1979, Turrens et al. 1984)). Other enzymes, including Mn-SOD and SOD derivatives (i.e. polyethylene glycol-bound SOD (PEG-SOD)) appear to have much longer half-lives (6-8 h (McCord and Wong 1979, Baret et al. 1984)), thus being more suitable for long-term experiments (table 1). These two long-lasting forms of SOD behave differently in circulation: while circulating Mn-SOD equilibrates relatively fast with the lymph (10-20 minutes), PEG-SOD remains in circulation for much longer times (Omar and McCord 1990). This brings into consideration a second important problem, not usually taken into account: the i.v. injection of different forms of a given antioxidant enzyme may target different compartments; protection will eventually depend on SOD reaching the same compartment where 0, is generated. For example, while Mn-SOD equilibrates with the lymph in a few minutes, CuZn-SOD does so in 1 h. Thus, if the goal is targeting the interstitium using CuZn-SOD, the enzyme will have to be injected frequently, as with such a short half-life (table l), it may be cleared out of circulation before it reaches a significant concentration in the interstitium. Targeting the intracellular space T h e protective effect of most soluble SODS in circulation must be exerted outside the cells (where intracellular 0, has limited permeability (Lynch and Fridovich 1978)), and could be efficacious during inflammatory processes, in which Table 1. Half-life and tissue distribution of different antioxidant enzymes injected intravenously. Enzyme form
Half-life in circulation
Native Cu, Zn SOD
Compartment
References
6 8min
Plasma and interstitial space
(Baret et al. 1984, Turrens et at. 1984, Omar and McCord 1990)
Mn-SOD
7h
Plasma and interstitial space
McCord and Wong 1979, Baret et al. 1984
PEG-Cu, Zn-SOD
37h
Mostly plasma but may penetrate inside endothelial cells
(Beckman et al. 1988, White et al., 1989)
Remains in the plasma or attached to endothelial cells
(Karlsson and Marklund 1987)
EC-SOD
Liposome-entrapped SOD or catalase
Approx. 5h
Delivered to the endothelial cells
(Turrens et al. 1984)
Native catalase
20 min
Not determined
(Turrens et ol., 1984, White et al., 1989)
PEG-catalase
18h
Not determined
(White et al., 1989)
J . F. Turrens
1036
0s
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
the source of is located on the outer surface of the polymorphonuclear cell membrane. T h e native forms of catalase or SOD will not enter the intracellular environment unless they become phagocytosed by endothelial cells, and even then they will be degraded in the lysosomes. T h e only soluble form of antioxidant enzymes that has been reported to penetrate inside endothelial cells in culture is PEG-SOD (Beckman et al. 1988), through a process that probably combines interaction of PEG groups with the cell membrane, endocytosis and protection from proteolysis by PEG groups associated with the enzymes. Although free SOD and catalase are not protective against hyperoxic injury to the lung, when conjugated with PEG they reach the intracellular milieu, and become protective (White et al.
1989). Antioxidant enzymes may also be delivered to the intracellular milieu by encapsulating them into liposomes (Freeman et al. 1983, Turrens et al. 1984, Freeman et al. 1985). These phospholipid vesicles can be prepared by reverse phase evaporation, in which an aqueous solution of the enzyme is mixed with lipid (phospholipids and cholesterol) dissolved in an organic phase (Turrens et al. 1984). T h e organic solvent is then evaporated and the lipids form vesicles containing the enzyme in the aqueous phase inside. This method gives reproducible preparations with a large extent of encapsulation (4040%). T h e liposomes are relatively large (the size may be standardized by passing the suspension through polycarbonate filters of defined size), and some of them become trapped in the lung capillary bed upon i.v. injection. Using this method, it is possible to deliver up to 2%of the liposomes to the lung, increasing several-fold the intracellular content of both SOD and catalase, after a single injection of liposomes containing either enzyme (Turrens et al. 1984). In addition, the half-life of either enzyme in circulation was increased from 6 min to several hours, as encapsulated enzymes are not cleared out of circulation as quickly as free enzymes. Liposome-encapsulated antioxidant enzymes protected cultured cells (Freeman et al. 1983) and rats (Turrens et al. 1984) exposed to 100%oxygen. In rats, four i.v. injections of liposomes every 12 h increased their survival from 70 h to 120 h (Turrens et al. 1984; table 2), and in later experiments indefinitely (Freeman et al. 1985). A similar treatment protected rats from hyperbaric injury to the brain (Yusa et al. 1984). In every case, administration of free antioxidant enzymes was not protective (table 2), indicating that in this case, the source of oxidative species is mostly located inside the endothelial cells. Notice that although injection of Table 2. Effect of different antioxidanttreatmentson survival of rats exposed to 100%oxygen, and in the accumulation of pleural fluid determined after autopsy. The data was taken from Turrens et al. (1984).
Treatmentt
Pleural effusion (ml)
Survival time (h)
Saline Free catalase plus SOD Empty liposomes Catalase liposomes SOD liposomes Liposomes containing SOD plus catalase
9 8 + 0 4 (24) 9.2k1.2 (8) 9.4k0.7 (8) 49*1.4$ (6) 90+05 (7) 0 5 +0.3$ (11)
6 9 5 k 1 . 5 (21) 71.4k1.7 (11) 71.2k1.5 (10) 744+7.3 (7) 7 2 4 f 1 . 6 (7) 118.1 +9.9$ (12)
t Rats were injected intravenously with a total volume of 2 ml saline or a suspension of liposomes (6G90 pmol DPPC) containing either saline, catalase (8-9.8 x lo4 U), or SOD (2-3 x l o 3 U) suspended in saline. Free catalase ( 5 x lo4 U/ml) and SOD (2 x lo3 U/ml) were injected in 2 ml saline. $These results were significantly different from all others in the same group (P
1991, VOL.21,
NO.
8, 1033-1040
Review The potential of antioxidant enzymes as pharmacological agents in vivo J. F. T U R R E N S Department of Biomedical Sciences, College of Allied Health, University of South Alabama, Mobile, Alabama 36688, USA
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
Received 26 March 1990; accepted 12 August 1990 1. Oxygen radicals have been associated with a number of unrelated pathological processes including ageing, radiation sickness, inflammation, oxygen toxicity, reoxygenation of ischaemic tissues, etc. The partial reduction of oxygen to superoxide anion (0;)and H,O, leads to the formation of more deleterious species such as hydroxyl radical (OH.) starting a chain reaction ultimately causing lipid peroxidation and cell death.
2. T o prevent the increased steady-state concentration of oxygen radicals many researchers have designed potential treatments including the i.v. injection of antioxidant enzymes or .enzyme derivatives with longer half-life in circulation (i.e. enzymes encapsulated in liposomes or covalently modified). 3. Tissue distribution and half-life in circulation depend upon the type of enzyme being used as well as whether the enzyme is or is not in its native form. 4. This review comments on some of the scenarios where these enzymes have been utilized, and discusses relevant problems of stability of different enzymes in circulation.
Production of oxidants in biological systems T h e partial reduction of oxygen to intermediates such as superoxide anion (0;) and hydroxyl radical (OH-) by biological systems was first proposed by Gerschman et al. (1954). This proposal was considered only an interesting hypothesis and it was almost forgotten for 15 years, until McCord and Fridovich (1969) discovered the enzyme superoxide dismutase (SOD). T h e identification of an enzyme suggested that its substrate had to be a normal cell metabolite and, very soon, hundreds of reports were published identifying a variety of sources of 0; and/or H 2 0 2 ,including free enzymes and multi-enzymic systems (Chance et al. 1979). There are several intracellular processes leading to the formation of partially reduced oxidative species. T h e simplest reaction involves the reduction of an oxygen molecule by a single electron producing O l , which in turn disproportionates to H,O, and 0, either spontaneously or in a reaction catalysed by the enzyme SOD (McCord and Fridovich 1969). Two components of the mitochondria1 respiratory chain (NADH dehydrogenase and ubisemiquinone) are examples of redox systems directly producing 0 2 (Turrens and Boveris 1980, Turrens et al. 1982, 1985). Alternatively, H 2 0 , may be generated in a single step through concerted transfer of two electrons to an oxygen molecule, as in the reaction catalysed by the enzyme glucose oxidase (Muller 1987). There are also enzymes such as xanthine oxidase capable of forming simultaneously 0; and H,O, in different proportions depending on the p H at which the reaction is carried out (Fridovich 1970). Superoxide anion formation may start a cascade of oxidative reactions, not as an oxidant itself but rather as a reductant. Inside the cell superoxide anion may reduce iron bound to ferritin which then is released from the protein in the ferrous state. Reduced iron may then react with H 2 0 2in a Fenton-like reaction, producing OH.. This species may in turn abstract a hydrogen from any biological molecule including 0049-8254/91 $ 3 4 0
0 1991 Taylor & Francis Ltd.
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
1034
J . F. Turrens
nucleotides, amino acids and unsaturated lipids. T h e latter reaction is the first in a series of reactions leading to lipid peroxidation. Not only can the combination of H,O, and 0; be deleterious for biological systems, but each species can eventually be involved in toxic reactions in the absence of the other. Hydroxyl radical may also be generated from the reaction between H,O, and a semiquinone, in a process that does not require 0; (Koppenol and Butler 1985). On the other hand, 0, may directly inactivate several enzymes, including catalase and glutathione peroxidase, which are needed to eliminate H,O, from the intracellular milieu (Kono and Fridovich 1982, Blum and Fridovich 1985). Cells are well protected against oxidative stress by antioxidant enzymes (SOD, catalase and glutathione peroxidase) and also by other non-protein antioxidants such as vitamin E and carotenes (Chance et al. 1979). T h e extracellular environment has much lower antioxidant defences: only a single enzyme known as extracellular SOD (EC-SOD) has been purified from most extracellular fluids and is present in relatively low concentrations (Marklund et al. 1982). There may be a strong evolutionary reason for keeping a low antioxidant activity in the extracellular space, since it is in this compartment that polymorphonuclear cells form the first line of defence against infectious organisms, producing 0, and H,O,. Thus, increased extracellular antioxidant activity would ultimately be deleterious for the host, as it would also decrease the natural defences against pathogens, favouring opportunistic infections. Under pathological conditions, however, there may be a need for increased extracellular antioxidant defences. For example, in adult respiratory distress syndrome (ARDS) polymorphonuclear cells become activated in the intravascular space, damaging the endothelium, increasing capillary permeability and causing oedema and swelling (Elliot et al. 1985). Also in many autoimmune diseases, the deposition of immune complexes leads to complement activation, producing chemoattractants (C5a) that bring polymorphonuclear cells to the area. In these cases neutrophils mediate the inflammatory process (Oldham et al. 1988). Intracellular oxidative injury may also lead to migration and activation of polymorphonuclear cells. As the intracellular steady-state lipid peroxidation increases, cells release partially oxidized molecules that act as chemoattractants for polymorphonuclear cells (Petrone et al. 1980), which in turn migrate to the area, become activated magnifying the injury and ultimately destroy the tissue. This occurs during lung oxygen toxicity (Crapo et al. 1983), secondary to the increased generation of oxidative species by lung endothelial and epithelial cells. This inflammatory response is responsible in part for the increased pulmonary oedema accompanying lung oxygen toxicity. It is in these cases that an increased intravascular antioxidant activity may be beneficial to the tissue. Oxidative injury is the common denominator in several diseases, including lung oxygen toxicity, hyperbaric injury to the central nervous sysystem, intoxication with xenobiotics (i.e. paraquat or carbon tetrachloride), radiation sickness, and more recently, reoxygenation of ischaemic organs (Chance et al. 1979, McCord 1985). Although oxygen radicals are involved in these pathological processes, they may be generated through different processes and in several cell compartments (i.e. cytoplasm, mitochondria, cell membrane or into the capillary lumen). Thus, a major problem for developing an effective antioxidant therapy is how to target antioxidant enzymes to the appropriate compartment where the oxidative species are being generated.
1035
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
Antioxidant enzymes in vivo
Intravenous injection of native or modified antioxidant enzymes Several research groups have considered the idea of using antioxidant enzymes as pharmacological agents, to specifically modulate the steady-state concentration of 0, or H 2 0 2 .The first problem with this approach is that the antioxidant enzymes most commonly used (Cu,Zn-SOD) and catalase have an extremely short half-life in circulation (6-8min) (McCord and Wong 1979, Turrens et al. 1984)). Other enzymes, including Mn-SOD and SOD derivatives (i.e. polyethylene glycol-bound SOD (PEG-SOD)) appear to have much longer half-lives (6-8 h (McCord and Wong 1979, Baret et al. 1984)), thus being more suitable for long-term experiments (table 1). These two long-lasting forms of SOD behave differently in circulation: while circulating Mn-SOD equilibrates relatively fast with the lymph (10-20 minutes), PEG-SOD remains in circulation for much longer times (Omar and McCord 1990). This brings into consideration a second important problem, not usually taken into account: the i.v. injection of different forms of a given antioxidant enzyme may target different compartments; protection will eventually depend on SOD reaching the same compartment where 0, is generated. For example, while Mn-SOD equilibrates with the lymph in a few minutes, CuZn-SOD does so in 1 h. Thus, if the goal is targeting the interstitium using CuZn-SOD, the enzyme will have to be injected frequently, as with such a short half-life (table l), it may be cleared out of circulation before it reaches a significant concentration in the interstitium. Targeting the intracellular space T h e protective effect of most soluble SODS in circulation must be exerted outside the cells (where intracellular 0, has limited permeability (Lynch and Fridovich 1978)), and could be efficacious during inflammatory processes, in which Table 1. Half-life and tissue distribution of different antioxidant enzymes injected intravenously. Enzyme form
Half-life in circulation
Native Cu, Zn SOD
Compartment
References
6 8min
Plasma and interstitial space
(Baret et al. 1984, Turrens et at. 1984, Omar and McCord 1990)
Mn-SOD
7h
Plasma and interstitial space
McCord and Wong 1979, Baret et al. 1984
PEG-Cu, Zn-SOD
37h
Mostly plasma but may penetrate inside endothelial cells
(Beckman et al. 1988, White et al., 1989)
Remains in the plasma or attached to endothelial cells
(Karlsson and Marklund 1987)
EC-SOD
Liposome-entrapped SOD or catalase
Approx. 5h
Delivered to the endothelial cells
(Turrens et al. 1984)
Native catalase
20 min
Not determined
(Turrens et ol., 1984, White et al., 1989)
PEG-catalase
18h
Not determined
(White et al., 1989)
J . F. Turrens
1036
0s
Xenobiotica Downloaded from informahealthcare.com by Freie Universitaet Berlin on 05/02/15 For personal use only.
the source of is located on the outer surface of the polymorphonuclear cell membrane. T h e native forms of catalase or SOD will not enter the intracellular environment unless they become phagocytosed by endothelial cells, and even then they will be degraded in the lysosomes. T h e only soluble form of antioxidant enzymes that has been reported to penetrate inside endothelial cells in culture is PEG-SOD (Beckman et al. 1988), through a process that probably combines interaction of PEG groups with the cell membrane, endocytosis and protection from proteolysis by PEG groups associated with the enzymes. Although free SOD and catalase are not protective against hyperoxic injury to the lung, when conjugated with PEG they reach the intracellular milieu, and become protective (White et al.
1989). Antioxidant enzymes may also be delivered to the intracellular milieu by encapsulating them into liposomes (Freeman et al. 1983, Turrens et al. 1984, Freeman et al. 1985). These phospholipid vesicles can be prepared by reverse phase evaporation, in which an aqueous solution of the enzyme is mixed with lipid (phospholipids and cholesterol) dissolved in an organic phase (Turrens et al. 1984). T h e organic solvent is then evaporated and the lipids form vesicles containing the enzyme in the aqueous phase inside. This method gives reproducible preparations with a large extent of encapsulation (4040%). T h e liposomes are relatively large (the size may be standardized by passing the suspension through polycarbonate filters of defined size), and some of them become trapped in the lung capillary bed upon i.v. injection. Using this method, it is possible to deliver up to 2%of the liposomes to the lung, increasing several-fold the intracellular content of both SOD and catalase, after a single injection of liposomes containing either enzyme (Turrens et al. 1984). In addition, the half-life of either enzyme in circulation was increased from 6 min to several hours, as encapsulated enzymes are not cleared out of circulation as quickly as free enzymes. Liposome-encapsulated antioxidant enzymes protected cultured cells (Freeman et al. 1983) and rats (Turrens et al. 1984) exposed to 100%oxygen. In rats, four i.v. injections of liposomes every 12 h increased their survival from 70 h to 120 h (Turrens et al. 1984; table 2), and in later experiments indefinitely (Freeman et al. 1985). A similar treatment protected rats from hyperbaric injury to the brain (Yusa et al. 1984). In every case, administration of free antioxidant enzymes was not protective (table 2), indicating that in this case, the source of oxidative species is mostly located inside the endothelial cells. Notice that although injection of Table 2. Effect of different antioxidanttreatmentson survival of rats exposed to 100%oxygen, and in the accumulation of pleural fluid determined after autopsy. The data was taken from Turrens et al. (1984).
Treatmentt
Pleural effusion (ml)
Survival time (h)
Saline Free catalase plus SOD Empty liposomes Catalase liposomes SOD liposomes Liposomes containing SOD plus catalase
9 8 + 0 4 (24) 9.2k1.2 (8) 9.4k0.7 (8) 49*1.4$ (6) 90+05 (7) 0 5 +0.3$ (11)
6 9 5 k 1 . 5 (21) 71.4k1.7 (11) 71.2k1.5 (10) 744+7.3 (7) 7 2 4 f 1 . 6 (7) 118.1 +9.9$ (12)
t Rats were injected intravenously with a total volume of 2 ml saline or a suspension of liposomes (6G90 pmol DPPC) containing either saline, catalase (8-9.8 x lo4 U), or SOD (2-3 x l o 3 U) suspended in saline. Free catalase ( 5 x lo4 U/ml) and SOD (2 x lo3 U/ml) were injected in 2 ml saline. $These results were significantly different from all others in the same group (P
