BIOLOGICAL TRACE ELEMENT RESEARCH 4, 303-317 (1982)
Anomalous Antioxidant Effects in Seleniumand Vitamin E-Deficient Liver Mitochondria? W E R N E R m . B A U M G A R T N E R * AND VIRGINIA m . H I L L
Nuclear Medicine and Research Services, Veterans Administration Wadsworth Medical Center, Los Angeles, California 90073, USA Received April 22, 1982; Accepted July 28, 1982
Abstract In this study, we investigate the mechanisms of two anomalous protective effects of exogenous vitamin E that had previously been postulated to involve either a specific antioxidant effect or a non-antioxidant function of the vitamin. These atypical vitamin E effects were observed during the prevention of NAD-induced respiratory decline occurring in homogenates and mitochondria prepared from vitamin E- and seleniumdeficient rat liver. The study showed neither hypothesis to be true; rather, the two effects, one in homogenates and the other in isolated mitochondria, were explained by other mechanisms. The protective effect against respiratory decline in homogenates was found to result from interference in the thiobarbituric acid assay for lipid peroxidation by ethanol (the conventional solvent for vitamin E addition). With other noninterfering solvents, inhibition of lipid peroxidation by vitamin E, in contrast to previous studies, correlated perfectly with prevention of respiratory decline. The atypical vitamin E effect occurring in isolated mitochondria--and consisting of a requirement for cytosol proteins for the prevention of respiratory decline by exogenous vitamin E was found to be caused by the prevention of adverse glass effects and not by the action of vitamin E-specific binding proteins. Frequent failures in the combined protective effect of vitamin E and cytosol, which had been a major complication of respiratory decline studies, were found to be caused by phospholipase activity generated during isolation procedures. Irreversible deactivation of respiratory enzymes by lipid peroxidation was found not to be involved in the respiratory decline mechanism. Index Entries: Antioxidant effects, in Se- and vitamin E-deficient mitochondria; vitamin E-deficient mitochondria, antioxidant effects in; selenium-deficient mitochondria, antioxidant effects in; liver mitochondria, antioxidant effects in; mitochondria, antioxidant effects in liver.
Introduction Respiratory Decline (RD) is an in vitro phenomenon that occurs 7-10 days before the onset of dietary liver necrosis in rats fed vitamin E- and selenium-deficient diets tin memoriam: Klaus Schwarz, MD, 1914-1978. 303
9 1982 by The Humana Press Inc. All rights of any nature whatsoever reserved. Q16"~t0~4/~2/12flO-0303503.00
304
BAUMGARTNER AND HILL
(1). Caused by exogenous nicotinamide adenine dinucleotide (NAD), RD is an abrupt decline in mitochondrial succinic oxidase activity after an initial period of normal respiration. With liver homogenates (minus nuclear fraction), RD can be prevented by addition of exogenous vitamin E. Corwin found, however, that this protection occurred under conditions where the vitamin had no significant inhibitory effect on random lipid peroxidation (2). This unusual result was interpreted by Schwarz (3) to mean that lipid peroxidation was not the cause of RD and that the true biological role of the vitamin was its electron transport-promoting effect and not its antioxidant properties. Our own studies of RD with isolated mitochondria had also revealed the absence of a correlation between RD and lipid peroxidation (4, 5). Here, however, the converse situation to that reported by Corwin was observed--namely, that RD could not be prevented by exogenous vitamin E on its own in spite of the vitamin having effectively suppressed lipid peroxidation. Protection could be achieved only when vitamin E was added to the reaction mixture in the presence of cytosol proteins. Since cytosol had recently been reported to contain vitamin E-specific binding proteins facilitating the incorporation of the vitamin into mitochondrial membrane structures (6), we adopted the hypothesis that the atypical vitamin E effects observed by us and Corwin were due not to an electron transport function of the vitamin, but rather to the operation of a specific antioxidant effect. A specific antioxidant effect has been defined (5) as a radical scavenger action operating at a specific site resulting in the control of normal metabolic processes proceeding by radical mechanisms. In the case of RD, such a specific antioxidant effect is postulated to be directed against a "leak" of electron transport radicals causing RD by irreversibly damaging key enzyme systems at specific mitochondrial sites. The production of hydrogen peroxide during mitochondrial respiration suggests the existence of such radical leaks (7). Radical leaks have also been postulated in aging processes (8). Our initial search for direct evidence for this hypothesis had been thwarted by unexpected technical difficulties (5). These initial studies also suggested that Corwin's atypical vitamin E effect, but not the one reported by us with the isolated mitochondria, could have been caused by an artifact in the TBA assay for lipid peroxidation (5, 9). In the present study we have further investigated the possible involvement of the TBA assay artifact and specific antioxidant effects in the atypical protective actions of vitamin E in homogenates and isolated mitochondria.
Materials and Methods Animals and Diets
Fisher 344 male weanling rats, weighing 25-35 g (Charles River Laboratories, Wilmington, MA), were maintained on a diet of vitamin E and selenium-deficient Torula yeast (5) and distilled water. To avoid any uncontrolled contact with selenium, rats were housed in metal cages fitted with metal screens, but without wood
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
305
shavings in the lower pans. Animals were sacrificed after 4-6 weeks on this diet, i.e., during the occurrence of RD in isolated liver homogenate as measured by the Warburg respirometer assay.
Subcellular Fractions Liver homogenates (minus nuclear fraction), mitochondrial, and microsomal fractions were prepared by conventional methods (5), except for two important modifications, the most important of which was the exclusion of EDTA (ethylenediamine tetraacetic acid) from the isolation medium of 0.25M sucrose (Sigma, Grade 1) and 0.01M Tris buffer, pH 7.6. EDTA had to be excluded as this would have masked the protective effect of exogenous vitamin E against RD by its inhibitory effect on in vitro, metal-catalyzed lipid peroxidation. The other modification was in the number of times the mitochondrial pellet was washed. Normally, this is an arbitrarily determined fixed number. In our case, the number of washes ranged from zero to four washes for the reasons indicated in the various experiments. Washing consisted simply of the gentle resuspension of the mitochondrial pellet in fresh aliquots of isolation medium by means of a slow-turning Potter-Elvehjem homogenizing pestle followed by recentrifugation at 6000g. The final suspension of mitochondria was adjusted such that 0.1 mL of the suspension in 2.9 mL of 0.25M sucrose gave an absorbance value of 0.50 at 600 nm (1 cm light path at the standard distance from the phototube of a Gilford spectrophotometer).
Assays Respiratory decline in succinic oxidase activity in homogenates or mitochondria was measured with a Warburg respirometer in the following reaction medium: 90 ~mol PO4 (added as sodium phosphate buffer, pH 7.6), 15 ixmol MgCI2, 3 Ixmol ADP, 3 ~mol ATP, 0.1 mg cytochrome c, 60 ~xmol sodium succinate, and 0.1M Tris buffer to bring to a final volume of 3.5 mL including the 0.5 mL of mitochondria or homogenate. For reactions where RD was to be induced, NAD (6 ixmol) was added to the medium. Mitochondrial swelling occurring during the reaction was monitored by light scatter measurements. Lipid peroxidation was measured by the thiobarbituric acid (TBA) assay as described previously (5).
Vitamin E Preparation A constant amount (25 mg/mL) of DL-~-tocopherol (Sigma Chemical Co.), in its pure state or dissolved in acetone (Fisher, spectranalyzed), was added to different suspending media: cytosol, bovine serum albumin solution (10%), isotonic sucrose, isotonic KC1, mitochondria, or microsomes. When pure (undissolved) vitamin E oil was added to the above media, it was dispersed by sonication; when dissolved in acetone, vitamin E was dispersed by stirring with a magnetic stirrer, set on near maximum speed (Magnestir, Curtin Scientific Co.) Acetone was used in place of ethanol (the conventional solvent for vitamin E addition), since we had shown previously that ethanol can give rise to false lipid peroxidation values by the TBA test (5).
306
BAUMGARTNER AND HILL
All vitamin E emulsions were found to be equally stable, showing very little breaking of emulsions during 24 h at room temperature. The droplet size ( - 10 ~m in diameter) was also very similar in all emulsions. Aliquots (0.2 mL) of these emulsions were added with stirring to 3.0 mL of homogenate and 0.5 mL of homogenate was used for each RD reaction.
Testing of Vitamin E Preparations The efficacies of the different vitamin E preparations against RD and lipid peroxidation were tested by performing Warburg assays with homogenate at 37~ in the presence of NAD. The vitamin E preparations were added to the homogenates prior to commencing the succinic oxidase reaction. After 120 min of respiration reaction media were tested for lipid peroxidation and mitochondrial swelling. The occurrence of RD was determined from the kinetics of the reaction.
Activation Energies of Respiration and RD Activation energies for the respiration reactions and for the rate-determining reactions involved in RD in homogenates and in isolated mitochondria were determined by measuring reaction rates as a function of temperature (at 35.0, 25.0, and 19.5~ Since the concentrations of all reagents were kept constant in these experiments a change in reaction rate could be used as a measure of a change in the rate constant. Respiration rates were expressed as change in oxygen pressure per unit time (dp/dt). Rates of the processes involved in RD were measured as the reciprocal of the time required for the onset of RD. Reaction rates at 35~ were assigned a value of 1.0, and reaction rates at lower temperatures were given relative values which would then be between 0 and 1.0. Activation energies were calculated from the Arrhenius equation log Kr = log A - E/2.3R • I/T i.e., from the slope of the plot of log Kr vs log 1/T, where T is the reaction temperature expressed in absolute units.
Glass and Phospholipase-lnduced Mitochondrial Swelling and Reversibility of RD These experiments were performed with isolated mitochondria by measuring succinic oxidase activity with a Warburg respirometer under the various conditions described in the text.
Results We first tested the hypothesis that the requirement for cytosol in the protective action of exogenous vitamin E against RD in isolated mitochondria results from the action of vitamin E-specific binding proteins. Since testing this hypothesis directly
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
307
was complicated by a serious lack of reproducibility of the protective vitamin E effect in isolated mitochondria, we used an indirect approach, which involved using the highly reproducible homogenate test system. Because of important mechanistic implications, we also investigated the reason for the difference in reliability of the vitamin E protective effect with isolated mitochondria and homogenates.
Respiratory Decline in Homogenates The indirect approach consisted of preparing a number of different vitamin E emulsions (see method section), some containing cytosol proteins and others not, and comparing the effectiveness of these preparations in preventing lipid peroxidation and RD in the liver homogenate test system. In spite of similar physical properties, we found that the various vitamin E preparations had drastically different effects on lipid peroxidation and on RD (Table 1). An acetone solution of vitamin E added directly to the homogenate test system was as effective in preventing lipid peroxidation and RD as the solution added first to cytosol and then to the homogenate (Table 1, Expts. 1-3). In contrast, emulsions prepared in the absence of cytosol (vitamin E suspended only in sucrose or in KCI) were completely ineffective against lipid peroxidation as well as against RD (Expts. 4-6). This was the case irrespective of whether the vitamin E emulsion had been prepared from an acetone solution or from the pure vitamin with sonication. In Expt. 7 we tested whether cytosol proteins acted as carriers of vitamin E or whether the proteins bestowed certain surface properties to the vitamin E droplets (zeta potential, liposome effects) facilitating the incorporation of the droplets into mitochondrial membranes. We tested for this possibility by passing the vitamin E-cytosol emulsion through a 1-1xm-pore-size filter. The 10 p~m droplets of vitamin E were very effectively retained by the filter as indicated by the drastic reduction in light scatter of the filtrate (0.23 OD reduced to 0.01 OD). The only vitamin E found in the filtrate was that bound to cytosol proteins and only this was found to be effective against lipid peroxidation and RD. This result explains the observed ineffectiveness of the other protein-free vitamin E emulsions (Expts. 4-6). In Expt. 8, the vitamin E emulsion was prepared by adding vitamin E oil directly to cytosol and dispersing this with sonication. This preparation was also found to be ineffective against lipid peroxidation and RD, indicating that the vitamin must be added to cytosol in the dissolved state in order to effectively interact (bind) with cytosol proteins. This was further indicated by the results of Expt. 9 in which an acetone solution of vitamin E was first dispersed in sucrose; this sucrose-vitamin E emulsion was then added with vigorous stirring to cytosol. Once again, the resuiting vitamin E preparation was ineffective, indicating that no exchange of vitamin from droplets to cytosol proteins occurs. Experiment 10, in which an acetone solution of vitamin E was added directly to a mitochondrial suspension, showed that the vitamin can be incorporated directly into mitochondrial membranes without the intercession of cytosol proteins. Similarly, we found that vitamin E dissolved in acetone could interact directly with microsomal membranes (Expt. 11). Expts. 10 and 11 differed from the other experiments listed in Table 1 in that the antioxidant properties of the mitochondrial and
308
BAUMGARTNER
b
~x 0
8
4
O ~ ~ m m ~ m m ~
0
20
mmp~
40 60 80 TIME (min)
~
120
Fig. 2. Respiratory activity of mitochondria used in studying the reversibility of respiratory decline: o, control reaction containing no respiratory decline-producing exogenous NAD; o, mitochondria undergoing respiratory decline; m, mitochondria reisolated from exogenous NAD after experiencing respiratory decline, reacting in an NAD-free medium in the absence of exogenous cytochrome C; A, mitochondria, reioslated from exogenous NAD after experiencing respiratory decline, reacting in an NAD--free medium in the presence of exogenous cytochrome C; A, instantaneous respiratory decline of mitochondria which were preswollen by repeated freezing and thawing.
resuspended in sucrose, and tested for succinic oxidase activity in the absence of exogenous NAD. Very little respiratory activity was detected in such mitochondria (Fig. 2). However, the same result was obtained with mitochondria that had been recovered from control reactions containing no NAD and which therefore had not undergone RD. Activity was restored to mitochondria from both reactions by addition of cytochrome C (Fig. 2). Mitochondrial swelling, occurring during the initial Warburg assay, had led to the solubilization of membrane-bound cytochrome C. Separation of the mitochondria from the reaction medium had resulted in the loss of cytochrome C. Thus, irreversible radical damage of respiratory enzyme systems is not involved in the mechanism of RD.
ANTIOX1DANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
315
That swelling leading to RD can also be produced by processes other than lipid peroxidation, glass, and phospholipase activity was demonstrated with mitochondria in which permeability changes had been introduced by repeated freezing and thawing prior to contact with NAD. As expected, such mitochondria underwent RD immediately on exposure to NAD rather than after the usual 40-50 min, attesting further to the importance of permeability changes in the RD mechanism (Fig. 2). As expected, RD in frozen and thawed mitochondria could not be prevented by the combined action of vitamin E, cytosol, and nupercaine. In the absence of exogenous NAD, frozen and thawed mitochondria exhibited normal respiration rates.
Discussion In contrast to Corwin's original observations (2), we found in the present study that the prevention of RD in homogenates is perfectly correlated with the prevention of random lipid peroxidation. From our present and earlier studies (5), and Corwin's use of ethanol for addition of exogenous vitamin E, it is clear that this difference in results is caused by the influence of the TBA assay artifact in Corwin's study. Thus, RD in homogenates can now be explained in terms of conventional antioxidant theory. The nature of the second atypical vitamin E effect occurring with isolated mitochondria was also clarified. Our work with the different vitamin E preparations clearly demonstrated that the effect of cytosol was not a result of the action of a specific antioxidant effect conferred upon vitamin E by specific binding proteins. This latter conclusion is based not so much on our observation that albumin-bound vitamin E was equally effective in preventing RD as vitamin E bound to cytosol, since this result could also have arisen through possible exchange reactions between these two vitamin E carriers, as it is on our ability to obtain active preparations of vitamin E merely by adding an acetone solution of the vitamin directly to mitochondria. In reaching this conclusion, we do not wish to imply that vitamin E-specific carrier proteins do not exist, nor that such proteins could not incorporate the vitamin more efficiently or even more specifically into certain membrane compartments, but rather that such specific effects are not required for the prevention of RD. In spite of our negative results, we feel that the regulation of normal metabolic processes involving radical mechanisms by specific antioxidant effects is still a viable concept. Such regulatory processes may be involved in prostaglandin synthesis (14), mixed-function oxidase activity (15), and certain immune functions (16). The theory also has the heuristic advantage of bridging the two opposing schools of thought on the action of antioxidants, one school holding that antioxidants act merely as scavengers of undesirable radical reactions (17), and the other that in vivo radical scavenger actions of antioxidants are relatively unimportant compared to their hypothesized actions as critical components of certain enzyme systems
(18-22). The fact that the stable vitamin E emulsions prepared without solvents were not effective antioxidants is also of some practical interest. For example, such prepara-
316
BAUMGARTNER AND HILL
tions were used by Packer and Smith in their initial (but still not reproduced) study of beneficial vitamin E effects on clonal senescence (23, 24). Our results suggest that their initial success with vitamin E against clonal senescence may have been caused by poorly controlled vitamin E preparations or by the presence of an unusual serum component facilitating the formation of an active vitamin E preparation under conditions normally resulting in inactive preparations. The lack of evidence for a specific antioxidant effect in the prevention of RD led us to adopt a new hypothesis for the atypical vitamin E effect in isolated mitochondria. Our new hypothesis that RD could be triggered by swelling resuiting from a variety of causes was confirmed by the observation that cytosol served to reduce glass-induced swelling of mitochondria, whereas phospholipaseinduced swelling, initiated by washing of mitochondria during the isolation, accounted for the frequent failures in the combined protective effects of exogenous vitamin E and cytosol. Finally, the reversibility of RD demonstrated that radical damage of key respiratory enzymes is not involved in the mechanism of RD. With these facts in hand the following mechanism can be proposed for RD: mitochondrial swelling from the various causes identified above allows exogenous NAD to enter certain mitochondrial membrane compartments from which it is normally excluded by membrane barrier effects. The immediate cause of RD is not radical damage, but NAD-catalyzed oxalacetic acid feedback inhibition of succinic dehydrogenase activity (25) regulated by ATP-dependent compartmentalization effects (26) and/or changes in the membrane-bound NAD/NADH ratio (27). The initiating and rate-determining step in RD, i.e., swelling, differs in homogenates and isolated mitochondria. Only lipid peroxidation-induced swelling is involved in RD occurring in homogenates, and only this is correlated with the development of liver necrosis resulting from ingestion of a vitamin E- and selenium-deficient diet.
Acknowledgments The study was supported from a grant from the National Institutes of Health (5R01-CA-14221). The expert assistance of Ajit Arora, PhD, MD, and Charles Porter is gratefully acknowledged.
References 1. 2. 3. 4.
L. M. Corwin and K. Schwarz, J. Biol. Chem. 235, 3387 (1960). L. M. Corwin, Arch. Biochem. Biophys. 97, 51 (1962). K. Schwarz, Fed. Proc. 24, 58 (1965). K. Schwarz and W. A. Baumgartner, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, p. 317. 5. W. A. Baumgartner, V. A. Hill, and E. T. Wright, Mech. Ageing Develop. 8, 311 (1978). 6. O. V. Rajaram, P. Fatterpaker, and A. Screenivasan, Biochem. Biophys. Res. Commun. 52, 459 (1973). 7. A. Boveris, N. Oshino, and B. Cance, Biochem. J. 128, 617 (1972).
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
317
D. Harman, J. Am. Geriat. Soc. 20, 145 (1972). W. A. Baumgartner, N. Baker, V. A. Hill, and E. T. Wright, Lipids 10, 311 (1975). L. Weiss, Exp. Cell. Res. 53, 603 (1968). L. Weiss and R. F. Woodbridge, Fed. Proc. 26, 88 (1967). D.L. Cinti, P. Moldeus, and J. B. Schenkman, Biochem. Biophys. Res. Commun. 47, 1028 (1972). J. Scarpa, Eur. J. Biochem. 27, 401 (1972). W. C. Hope, C. Dalton, L. T. Machlin, R. J. Filipski, and F. M. Vane, Prostaglandins 10, 557 (1975). M. P. Carpenter, Ann. NYAcad. Sci. 203, 81 (1972). W. A. Baumgartner, in Trace Metals in Health and Disease, N. Kharasch, Raven, 1979, p. 287. A. L. Tappel, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, 369. J. Green, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, 293. P. P. Naire, Ann. NY Acad. Sci. 203, 53 (1972). R. E. Olson, Am. J. Clin. Nutr. 27, 1117 (1974). A. T. Diplock, Am. J. Clin. Nutr. 27, 995 (1974). J. A. Lucy, Ann. NY Acad. Sci. 203, 4 (1972). L. Packer and R. Smith, Proc. Natl. Acad. Sci. USA 71, 4763 (1974). L. Packer and R. Smith, Proc. Natl. Acad. Sci. USA 74, 1640 (1977). L. M. Corwin, J. Biol. Chem. 240, 38 (1965). J. M. Tager, in Regulation of Metabolic Processes in Mitochondria, J. M. Tager, S. Papa, E. Quagliariello, and E. C. Slater, eds., 1966, pp. 202-217. T. P. Singer, E. B. Kearney, and W. C. Kenney, in Advances in Enzymology, vol. 27, A. Meister, ed., Academic Press, New York, 1973, 189.
Anomalous Antioxidant Effects in Seleniumand Vitamin E-Deficient Liver Mitochondria? W E R N E R m . B A U M G A R T N E R * AND VIRGINIA m . H I L L
Nuclear Medicine and Research Services, Veterans Administration Wadsworth Medical Center, Los Angeles, California 90073, USA Received April 22, 1982; Accepted July 28, 1982
Abstract In this study, we investigate the mechanisms of two anomalous protective effects of exogenous vitamin E that had previously been postulated to involve either a specific antioxidant effect or a non-antioxidant function of the vitamin. These atypical vitamin E effects were observed during the prevention of NAD-induced respiratory decline occurring in homogenates and mitochondria prepared from vitamin E- and seleniumdeficient rat liver. The study showed neither hypothesis to be true; rather, the two effects, one in homogenates and the other in isolated mitochondria, were explained by other mechanisms. The protective effect against respiratory decline in homogenates was found to result from interference in the thiobarbituric acid assay for lipid peroxidation by ethanol (the conventional solvent for vitamin E addition). With other noninterfering solvents, inhibition of lipid peroxidation by vitamin E, in contrast to previous studies, correlated perfectly with prevention of respiratory decline. The atypical vitamin E effect occurring in isolated mitochondria--and consisting of a requirement for cytosol proteins for the prevention of respiratory decline by exogenous vitamin E was found to be caused by the prevention of adverse glass effects and not by the action of vitamin E-specific binding proteins. Frequent failures in the combined protective effect of vitamin E and cytosol, which had been a major complication of respiratory decline studies, were found to be caused by phospholipase activity generated during isolation procedures. Irreversible deactivation of respiratory enzymes by lipid peroxidation was found not to be involved in the respiratory decline mechanism. Index Entries: Antioxidant effects, in Se- and vitamin E-deficient mitochondria; vitamin E-deficient mitochondria, antioxidant effects in; selenium-deficient mitochondria, antioxidant effects in; liver mitochondria, antioxidant effects in; mitochondria, antioxidant effects in liver.
Introduction Respiratory Decline (RD) is an in vitro phenomenon that occurs 7-10 days before the onset of dietary liver necrosis in rats fed vitamin E- and selenium-deficient diets tin memoriam: Klaus Schwarz, MD, 1914-1978. 303
9 1982 by The Humana Press Inc. All rights of any nature whatsoever reserved. Q16"~t0~4/~2/12flO-0303503.00
304
BAUMGARTNER AND HILL
(1). Caused by exogenous nicotinamide adenine dinucleotide (NAD), RD is an abrupt decline in mitochondrial succinic oxidase activity after an initial period of normal respiration. With liver homogenates (minus nuclear fraction), RD can be prevented by addition of exogenous vitamin E. Corwin found, however, that this protection occurred under conditions where the vitamin had no significant inhibitory effect on random lipid peroxidation (2). This unusual result was interpreted by Schwarz (3) to mean that lipid peroxidation was not the cause of RD and that the true biological role of the vitamin was its electron transport-promoting effect and not its antioxidant properties. Our own studies of RD with isolated mitochondria had also revealed the absence of a correlation between RD and lipid peroxidation (4, 5). Here, however, the converse situation to that reported by Corwin was observed--namely, that RD could not be prevented by exogenous vitamin E on its own in spite of the vitamin having effectively suppressed lipid peroxidation. Protection could be achieved only when vitamin E was added to the reaction mixture in the presence of cytosol proteins. Since cytosol had recently been reported to contain vitamin E-specific binding proteins facilitating the incorporation of the vitamin into mitochondrial membrane structures (6), we adopted the hypothesis that the atypical vitamin E effects observed by us and Corwin were due not to an electron transport function of the vitamin, but rather to the operation of a specific antioxidant effect. A specific antioxidant effect has been defined (5) as a radical scavenger action operating at a specific site resulting in the control of normal metabolic processes proceeding by radical mechanisms. In the case of RD, such a specific antioxidant effect is postulated to be directed against a "leak" of electron transport radicals causing RD by irreversibly damaging key enzyme systems at specific mitochondrial sites. The production of hydrogen peroxide during mitochondrial respiration suggests the existence of such radical leaks (7). Radical leaks have also been postulated in aging processes (8). Our initial search for direct evidence for this hypothesis had been thwarted by unexpected technical difficulties (5). These initial studies also suggested that Corwin's atypical vitamin E effect, but not the one reported by us with the isolated mitochondria, could have been caused by an artifact in the TBA assay for lipid peroxidation (5, 9). In the present study we have further investigated the possible involvement of the TBA assay artifact and specific antioxidant effects in the atypical protective actions of vitamin E in homogenates and isolated mitochondria.
Materials and Methods Animals and Diets
Fisher 344 male weanling rats, weighing 25-35 g (Charles River Laboratories, Wilmington, MA), were maintained on a diet of vitamin E and selenium-deficient Torula yeast (5) and distilled water. To avoid any uncontrolled contact with selenium, rats were housed in metal cages fitted with metal screens, but without wood
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
305
shavings in the lower pans. Animals were sacrificed after 4-6 weeks on this diet, i.e., during the occurrence of RD in isolated liver homogenate as measured by the Warburg respirometer assay.
Subcellular Fractions Liver homogenates (minus nuclear fraction), mitochondrial, and microsomal fractions were prepared by conventional methods (5), except for two important modifications, the most important of which was the exclusion of EDTA (ethylenediamine tetraacetic acid) from the isolation medium of 0.25M sucrose (Sigma, Grade 1) and 0.01M Tris buffer, pH 7.6. EDTA had to be excluded as this would have masked the protective effect of exogenous vitamin E against RD by its inhibitory effect on in vitro, metal-catalyzed lipid peroxidation. The other modification was in the number of times the mitochondrial pellet was washed. Normally, this is an arbitrarily determined fixed number. In our case, the number of washes ranged from zero to four washes for the reasons indicated in the various experiments. Washing consisted simply of the gentle resuspension of the mitochondrial pellet in fresh aliquots of isolation medium by means of a slow-turning Potter-Elvehjem homogenizing pestle followed by recentrifugation at 6000g. The final suspension of mitochondria was adjusted such that 0.1 mL of the suspension in 2.9 mL of 0.25M sucrose gave an absorbance value of 0.50 at 600 nm (1 cm light path at the standard distance from the phototube of a Gilford spectrophotometer).
Assays Respiratory decline in succinic oxidase activity in homogenates or mitochondria was measured with a Warburg respirometer in the following reaction medium: 90 ~mol PO4 (added as sodium phosphate buffer, pH 7.6), 15 ixmol MgCI2, 3 Ixmol ADP, 3 ~mol ATP, 0.1 mg cytochrome c, 60 ~xmol sodium succinate, and 0.1M Tris buffer to bring to a final volume of 3.5 mL including the 0.5 mL of mitochondria or homogenate. For reactions where RD was to be induced, NAD (6 ixmol) was added to the medium. Mitochondrial swelling occurring during the reaction was monitored by light scatter measurements. Lipid peroxidation was measured by the thiobarbituric acid (TBA) assay as described previously (5).
Vitamin E Preparation A constant amount (25 mg/mL) of DL-~-tocopherol (Sigma Chemical Co.), in its pure state or dissolved in acetone (Fisher, spectranalyzed), was added to different suspending media: cytosol, bovine serum albumin solution (10%), isotonic sucrose, isotonic KC1, mitochondria, or microsomes. When pure (undissolved) vitamin E oil was added to the above media, it was dispersed by sonication; when dissolved in acetone, vitamin E was dispersed by stirring with a magnetic stirrer, set on near maximum speed (Magnestir, Curtin Scientific Co.) Acetone was used in place of ethanol (the conventional solvent for vitamin E addition), since we had shown previously that ethanol can give rise to false lipid peroxidation values by the TBA test (5).
306
BAUMGARTNER AND HILL
All vitamin E emulsions were found to be equally stable, showing very little breaking of emulsions during 24 h at room temperature. The droplet size ( - 10 ~m in diameter) was also very similar in all emulsions. Aliquots (0.2 mL) of these emulsions were added with stirring to 3.0 mL of homogenate and 0.5 mL of homogenate was used for each RD reaction.
Testing of Vitamin E Preparations The efficacies of the different vitamin E preparations against RD and lipid peroxidation were tested by performing Warburg assays with homogenate at 37~ in the presence of NAD. The vitamin E preparations were added to the homogenates prior to commencing the succinic oxidase reaction. After 120 min of respiration reaction media were tested for lipid peroxidation and mitochondrial swelling. The occurrence of RD was determined from the kinetics of the reaction.
Activation Energies of Respiration and RD Activation energies for the respiration reactions and for the rate-determining reactions involved in RD in homogenates and in isolated mitochondria were determined by measuring reaction rates as a function of temperature (at 35.0, 25.0, and 19.5~ Since the concentrations of all reagents were kept constant in these experiments a change in reaction rate could be used as a measure of a change in the rate constant. Respiration rates were expressed as change in oxygen pressure per unit time (dp/dt). Rates of the processes involved in RD were measured as the reciprocal of the time required for the onset of RD. Reaction rates at 35~ were assigned a value of 1.0, and reaction rates at lower temperatures were given relative values which would then be between 0 and 1.0. Activation energies were calculated from the Arrhenius equation log Kr = log A - E/2.3R • I/T i.e., from the slope of the plot of log Kr vs log 1/T, where T is the reaction temperature expressed in absolute units.
Glass and Phospholipase-lnduced Mitochondrial Swelling and Reversibility of RD These experiments were performed with isolated mitochondria by measuring succinic oxidase activity with a Warburg respirometer under the various conditions described in the text.
Results We first tested the hypothesis that the requirement for cytosol in the protective action of exogenous vitamin E against RD in isolated mitochondria results from the action of vitamin E-specific binding proteins. Since testing this hypothesis directly
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
307
was complicated by a serious lack of reproducibility of the protective vitamin E effect in isolated mitochondria, we used an indirect approach, which involved using the highly reproducible homogenate test system. Because of important mechanistic implications, we also investigated the reason for the difference in reliability of the vitamin E protective effect with isolated mitochondria and homogenates.
Respiratory Decline in Homogenates The indirect approach consisted of preparing a number of different vitamin E emulsions (see method section), some containing cytosol proteins and others not, and comparing the effectiveness of these preparations in preventing lipid peroxidation and RD in the liver homogenate test system. In spite of similar physical properties, we found that the various vitamin E preparations had drastically different effects on lipid peroxidation and on RD (Table 1). An acetone solution of vitamin E added directly to the homogenate test system was as effective in preventing lipid peroxidation and RD as the solution added first to cytosol and then to the homogenate (Table 1, Expts. 1-3). In contrast, emulsions prepared in the absence of cytosol (vitamin E suspended only in sucrose or in KCI) were completely ineffective against lipid peroxidation as well as against RD (Expts. 4-6). This was the case irrespective of whether the vitamin E emulsion had been prepared from an acetone solution or from the pure vitamin with sonication. In Expt. 7 we tested whether cytosol proteins acted as carriers of vitamin E or whether the proteins bestowed certain surface properties to the vitamin E droplets (zeta potential, liposome effects) facilitating the incorporation of the droplets into mitochondrial membranes. We tested for this possibility by passing the vitamin E-cytosol emulsion through a 1-1xm-pore-size filter. The 10 p~m droplets of vitamin E were very effectively retained by the filter as indicated by the drastic reduction in light scatter of the filtrate (0.23 OD reduced to 0.01 OD). The only vitamin E found in the filtrate was that bound to cytosol proteins and only this was found to be effective against lipid peroxidation and RD. This result explains the observed ineffectiveness of the other protein-free vitamin E emulsions (Expts. 4-6). In Expt. 8, the vitamin E emulsion was prepared by adding vitamin E oil directly to cytosol and dispersing this with sonication. This preparation was also found to be ineffective against lipid peroxidation and RD, indicating that the vitamin must be added to cytosol in the dissolved state in order to effectively interact (bind) with cytosol proteins. This was further indicated by the results of Expt. 9 in which an acetone solution of vitamin E was first dispersed in sucrose; this sucrose-vitamin E emulsion was then added with vigorous stirring to cytosol. Once again, the resuiting vitamin E preparation was ineffective, indicating that no exchange of vitamin from droplets to cytosol proteins occurs. Experiment 10, in which an acetone solution of vitamin E was added directly to a mitochondrial suspension, showed that the vitamin can be incorporated directly into mitochondrial membranes without the intercession of cytosol proteins. Similarly, we found that vitamin E dissolved in acetone could interact directly with microsomal membranes (Expt. 11). Expts. 10 and 11 differed from the other experiments listed in Table 1 in that the antioxidant properties of the mitochondrial and
308
BAUMGARTNER
b
~x 0
8
4
O ~ ~ m m ~ m m ~
0
20
mmp~
40 60 80 TIME (min)
~
120
Fig. 2. Respiratory activity of mitochondria used in studying the reversibility of respiratory decline: o, control reaction containing no respiratory decline-producing exogenous NAD; o, mitochondria undergoing respiratory decline; m, mitochondria reisolated from exogenous NAD after experiencing respiratory decline, reacting in an NAD-free medium in the absence of exogenous cytochrome C; A, mitochondria, reioslated from exogenous NAD after experiencing respiratory decline, reacting in an NAD--free medium in the presence of exogenous cytochrome C; A, instantaneous respiratory decline of mitochondria which were preswollen by repeated freezing and thawing.
resuspended in sucrose, and tested for succinic oxidase activity in the absence of exogenous NAD. Very little respiratory activity was detected in such mitochondria (Fig. 2). However, the same result was obtained with mitochondria that had been recovered from control reactions containing no NAD and which therefore had not undergone RD. Activity was restored to mitochondria from both reactions by addition of cytochrome C (Fig. 2). Mitochondrial swelling, occurring during the initial Warburg assay, had led to the solubilization of membrane-bound cytochrome C. Separation of the mitochondria from the reaction medium had resulted in the loss of cytochrome C. Thus, irreversible radical damage of respiratory enzyme systems is not involved in the mechanism of RD.
ANTIOX1DANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
315
That swelling leading to RD can also be produced by processes other than lipid peroxidation, glass, and phospholipase activity was demonstrated with mitochondria in which permeability changes had been introduced by repeated freezing and thawing prior to contact with NAD. As expected, such mitochondria underwent RD immediately on exposure to NAD rather than after the usual 40-50 min, attesting further to the importance of permeability changes in the RD mechanism (Fig. 2). As expected, RD in frozen and thawed mitochondria could not be prevented by the combined action of vitamin E, cytosol, and nupercaine. In the absence of exogenous NAD, frozen and thawed mitochondria exhibited normal respiration rates.
Discussion In contrast to Corwin's original observations (2), we found in the present study that the prevention of RD in homogenates is perfectly correlated with the prevention of random lipid peroxidation. From our present and earlier studies (5), and Corwin's use of ethanol for addition of exogenous vitamin E, it is clear that this difference in results is caused by the influence of the TBA assay artifact in Corwin's study. Thus, RD in homogenates can now be explained in terms of conventional antioxidant theory. The nature of the second atypical vitamin E effect occurring with isolated mitochondria was also clarified. Our work with the different vitamin E preparations clearly demonstrated that the effect of cytosol was not a result of the action of a specific antioxidant effect conferred upon vitamin E by specific binding proteins. This latter conclusion is based not so much on our observation that albumin-bound vitamin E was equally effective in preventing RD as vitamin E bound to cytosol, since this result could also have arisen through possible exchange reactions between these two vitamin E carriers, as it is on our ability to obtain active preparations of vitamin E merely by adding an acetone solution of the vitamin directly to mitochondria. In reaching this conclusion, we do not wish to imply that vitamin E-specific carrier proteins do not exist, nor that such proteins could not incorporate the vitamin more efficiently or even more specifically into certain membrane compartments, but rather that such specific effects are not required for the prevention of RD. In spite of our negative results, we feel that the regulation of normal metabolic processes involving radical mechanisms by specific antioxidant effects is still a viable concept. Such regulatory processes may be involved in prostaglandin synthesis (14), mixed-function oxidase activity (15), and certain immune functions (16). The theory also has the heuristic advantage of bridging the two opposing schools of thought on the action of antioxidants, one school holding that antioxidants act merely as scavengers of undesirable radical reactions (17), and the other that in vivo radical scavenger actions of antioxidants are relatively unimportant compared to their hypothesized actions as critical components of certain enzyme systems
(18-22). The fact that the stable vitamin E emulsions prepared without solvents were not effective antioxidants is also of some practical interest. For example, such prepara-
316
BAUMGARTNER AND HILL
tions were used by Packer and Smith in their initial (but still not reproduced) study of beneficial vitamin E effects on clonal senescence (23, 24). Our results suggest that their initial success with vitamin E against clonal senescence may have been caused by poorly controlled vitamin E preparations or by the presence of an unusual serum component facilitating the formation of an active vitamin E preparation under conditions normally resulting in inactive preparations. The lack of evidence for a specific antioxidant effect in the prevention of RD led us to adopt a new hypothesis for the atypical vitamin E effect in isolated mitochondria. Our new hypothesis that RD could be triggered by swelling resuiting from a variety of causes was confirmed by the observation that cytosol served to reduce glass-induced swelling of mitochondria, whereas phospholipaseinduced swelling, initiated by washing of mitochondria during the isolation, accounted for the frequent failures in the combined protective effects of exogenous vitamin E and cytosol. Finally, the reversibility of RD demonstrated that radical damage of key respiratory enzymes is not involved in the mechanism of RD. With these facts in hand the following mechanism can be proposed for RD: mitochondrial swelling from the various causes identified above allows exogenous NAD to enter certain mitochondrial membrane compartments from which it is normally excluded by membrane barrier effects. The immediate cause of RD is not radical damage, but NAD-catalyzed oxalacetic acid feedback inhibition of succinic dehydrogenase activity (25) regulated by ATP-dependent compartmentalization effects (26) and/or changes in the membrane-bound NAD/NADH ratio (27). The initiating and rate-determining step in RD, i.e., swelling, differs in homogenates and isolated mitochondria. Only lipid peroxidation-induced swelling is involved in RD occurring in homogenates, and only this is correlated with the development of liver necrosis resulting from ingestion of a vitamin E- and selenium-deficient diet.
Acknowledgments The study was supported from a grant from the National Institutes of Health (5R01-CA-14221). The expert assistance of Ajit Arora, PhD, MD, and Charles Porter is gratefully acknowledged.
References 1. 2. 3. 4.
L. M. Corwin and K. Schwarz, J. Biol. Chem. 235, 3387 (1960). L. M. Corwin, Arch. Biochem. Biophys. 97, 51 (1962). K. Schwarz, Fed. Proc. 24, 58 (1965). K. Schwarz and W. A. Baumgartner, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, p. 317. 5. W. A. Baumgartner, V. A. Hill, and E. T. Wright, Mech. Ageing Develop. 8, 311 (1978). 6. O. V. Rajaram, P. Fatterpaker, and A. Screenivasan, Biochem. Biophys. Res. Commun. 52, 459 (1973). 7. A. Boveris, N. Oshino, and B. Cance, Biochem. J. 128, 617 (1972).
ANTIOXIDANT EFFECTS IN SE- AND VITAMIN E-DEFICIENT MITOCHONDRIA
8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
317
D. Harman, J. Am. Geriat. Soc. 20, 145 (1972). W. A. Baumgartner, N. Baker, V. A. Hill, and E. T. Wright, Lipids 10, 311 (1975). L. Weiss, Exp. Cell. Res. 53, 603 (1968). L. Weiss and R. F. Woodbridge, Fed. Proc. 26, 88 (1967). D.L. Cinti, P. Moldeus, and J. B. Schenkman, Biochem. Biophys. Res. Commun. 47, 1028 (1972). J. Scarpa, Eur. J. Biochem. 27, 401 (1972). W. C. Hope, C. Dalton, L. T. Machlin, R. J. Filipski, and F. M. Vane, Prostaglandins 10, 557 (1975). M. P. Carpenter, Ann. NYAcad. Sci. 203, 81 (1972). W. A. Baumgartner, in Trace Metals in Health and Disease, N. Kharasch, Raven, 1979, p. 287. A. L. Tappel, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, 369. J. Green, in The Fat Soluble Vitamins, H. F. DeLuca and J. W. Suttie, eds., University of Wisconsin Press, Madison, 1969, 293. P. P. Naire, Ann. NY Acad. Sci. 203, 53 (1972). R. E. Olson, Am. J. Clin. Nutr. 27, 1117 (1974). A. T. Diplock, Am. J. Clin. Nutr. 27, 995 (1974). J. A. Lucy, Ann. NY Acad. Sci. 203, 4 (1972). L. Packer and R. Smith, Proc. Natl. Acad. Sci. USA 71, 4763 (1974). L. Packer and R. Smith, Proc. Natl. Acad. Sci. USA 74, 1640 (1977). L. M. Corwin, J. Biol. Chem. 240, 38 (1965). J. M. Tager, in Regulation of Metabolic Processes in Mitochondria, J. M. Tager, S. Papa, E. Quagliariello, and E. C. Slater, eds., 1966, pp. 202-217. T. P. Singer, E. B. Kearney, and W. C. Kenney, in Advances in Enzymology, vol. 27, A. Meister, ed., Academic Press, New York, 1973, 189.
