- Email: [email protected]
Increase of Mitochondrial PLA2-Released Fatty Acids Is an Early Event in Tumor Necrosis Factor α-Treated WEHI-164 Cells
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
221, 531–538 (1996)
0631
Increase of Mitochondrial PLA2-Released Fatty Acids Is an Early Event in Tumor Necrosis Factor a-Treated WEHI-164 Cells Christiane Levrat and Pierre Louisot1 Department of Biochemistry, INSERM-CNRS U.189, BP12, 69921 Oullins Cedex, France Received March 4, 1996 We have previously reported that TNF induced changes in mitochondrial enzymes, one of which, succinatedehydrogenase, is specifically activated in various TNF-sensitive cell lines. In an attempt to further characterize the mechanism of trans-membrane signalling at the mitochondrial level, we have oriented our investigation to the study of phospholipase A2 activity localized in this organelle isolated from TNF-treated WEHI-164 cells. Under physiological conditions, this enzyme has a very low basal activity near the resting state, while under TNF treatment its activity is dramatically increased. This event is induced by TNF concentrations which also cause cytolysis; however, the activity of this enzyme is increased several hours before maximum cytotoxicity occurs. The activation of the mitochondrial phospholipase A2 is not arachidonoyl or fatty acid selective, as is the case for the cytosolic species. Phospholipase A2 and succinate-dehydrogenase display higher activities simultaneously under TNF treatment. This change in activity is linked to morphologic modifications. Mitochondria in particular display an orthodox state characterized by a large and clear matrix space and few crests. © 1996 Academic Press, Inc.
TNF-a (Tumor Necrosis Factor) is a multipotent cytokine capable of killing various tumor cells in vitro (1). The interaction of TNF with membrane receptors initiates a cascade of intracellular events that leads to cell destruction. Among the postreceptor events, the arachidonic acid release triggered by the activation of phospholipase A2 (PLA2) seems to be an essential step in the TNF cytotoxicity (2–4). It has been shown that PLA2 involved in the TNF-induced arachidonic acid release displays the characteristics of the arachidonoyl-selective cytosolic PLA2, this enzyme is characterized by a high molecular weight (group IV) (5–8). However Nakano et al. (9) provided some evidences for the participation of the low molecular weight PLA2 (group II) which are membrane-bound enzymes. They hydrolyze phosphatidylethanolamine much more effectively than phosphatidylcholine, have no substrate specificity for arachidonic acid or other polyunsaturated fatty acids but they have an absolute requirement for Ca2+ for expressing of full activity (10). TNF would increase the RNA level of this PLA2 group. This specie is ubiquitously distributed in the cell. In addition to its association to microsomes, golgi membranes, plasma membranes and lysosomal membranes (11) this enzyme is also associated with both outer and inner membranes of mitochondria (11–13) and has been purified to near homogeneity (14–16). The mitochondrial specie requires Ca2+ for its catalytic action which is characterized by two activity plateaus upon variations of the Ca2+ concentrations with KCa2+ values of 14mM and 2.4mM respectively (17–18). The participation of mitochondria in TNF-induced cytotoxicity was inferred from the observation that these organella present structural alterations in their morphology as a swollen appearance (19), metabolic disfunctions as impaired respiration (20–21), superoxide anion generation (22) and an early activation of succinate dehydrogenase (SDH) (23). The mechanisms that manage the mitochondrial alterations are unknown. A hypothetical role of the mitochondrial PLA2 could be expected taking into account that this enzyme is responsible for changes of the inner membrane permeability, and consequently alterations of the metabolism (24). The present study reports that 1 Address reprint requests to Pr. Pierre Louisot, Department of Biochemistry INSERM-CNRS U.189, BP12, 69921 Oullins Cedex, France. Fax: (33) 78-50-71-52, E-mail: [email protected].
531 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TNF induces the increase of the mitochondrial PLA2 activity at the early stage of the cytotoxic process. MATERIALS AND METHODS Reagents. Purified human r-TNFa (specific activity: 2×107 U/mg) was purchased from Genzyme (Boston USA). All chemicals and cell culture products were respectively purchased from Sigma (Saint-Louis, Mo) and Gibco BRL-Life technologies (France). Cell line. Murine fibrosarcoma WEHI-164 cells from ATCC (Rockeville, Mo) were cultured in RPMI 1640 supplemented with 10% foetal calf serum (SEPRACOR-France) and containing penicillin (100 U/ml) streptomycin (100 mg/ml), nonessential amino-acids (GIBCO-BRL). The cell line was routinely tested for mycoplasma. Cytotoxicity assay. Changes in cell viability were measured with cells plated in 24 well microtiter at 4.104 cells per well in 500ml of growth medium. 24h later when cells reached confluency, TNF additions were made ranging in concentrations from 10 to 100 ng/ml or as specified. Each treatment consisted of four replicated wells. Cells were incubated for 3h up to 48h as specified, and the cell viability was measured by the cristal violet method (25). The percent of dead cells is expressed as follows: (1-ODassay/OD control) × 100. Ultrastructural studies. Cells were fixed with 1% gluteraldehyde 0.1M phosphate buffer pH7.4, postfixed with 1% osmium tetroxide and dehydrated in graded ethanols. Thin sections were mounted on collodion carbon-coated nickel grids and stained with 3% uranyl acetate and lead citrate. They were examined in a CM 10 Philips electron microscope operated at 80kV. Preparation of mitochondria. Confluent cells (3×108) were washed twice with HBSS, then harvested and washed three times with PBS (0.14M NaCl, 0.01M KCl, 0.01M sodium phosphate pH 7.4). Cells were washed once with 0.25M sucrose 10mM Tris pH 7.4, 1mM EGTA 0.1% BSA (bovine serum albumin) then resuspended in the same medium and were broken with a Dounce homogenizer. The cell homogenate was centrifuged at 1500g for 10min at 4°C. The nuclear pellet was reextracted and the two supernatants were combined and centrifuged at 10 000g for 10min at 4°C and the mitochondria pellet obtained were layered on discontinuous gradients (38% metrizamide, 17% metrizamide, 6% Percoll) and centrifuged 15min at 50 500g. Pure mitochondria were recovered at the interface 17% metrizamide/35% metrizamide. Mitochondrial purity was assessed by measurements of marker enzyme activities and ultrastructural studies. Proteins were measured by the bicinchoninic acid assay (26) using bovine serum albumin as standard. Succinate dehydrogenase assay. The activity of SDH was measured by a modification of the phenazine metosulfate assay described by Ackrell et al., (27–28). Measurement of phospholipase A2 activity. PLA2 activity associated with mitochondria was determined by monitoring the release of polyunsaturated fatty acids from endogenous phospholipids. Mitochondria isolated from TNF-treated and untreated cells were delipidated according to the modified Folch’s procedure (24) in the presence of butylated hydroxytoluene. A known amount of heptadecanoic acid was present and carried through the subsequent procedures as a internal standard. Isolation and quantitative estimation of PLA2-released fatty acids. Free fatty acids were isolated by using prepacked aminopropyl bonded silica columns and eluted with a mixture of diethyl ether/acetic acid (29). Prior to gas chromatography analysis, free fatty acids were converted to methyl esters using 5% (V/V) sulfuric acid in methanol. Fatty acid methyl esters were taken to dryness under nitrogen and dissolved in isooctane for analysis by gas liquid chromatography on a Perkin Elmer Autosystem Gas Chromatography coupled with a PE Nelson 1020 integrator. The compounds were separated on a SP 2380 capillary column (0.32mm × 30m; Supelco). Runs were first temperature programmed from 50°C to 150°C at 32°C/min then to 215°C at 2°C/min; azote was used as the carrier gas. Compounds were identified by their retention times as compared to authentic standards. To correct for procedural losses and volume inconsistencies between samples, peak areas were normalized to the area of the methyl heptadecanoate internal standard peak. PLA2 activity was taken to be the sum of polyunsaturated fatty acids.
RESULTS TNF Induces an Increase of the Mitochondrial PLA2-Released Fatty Acid Rate in the WEHI-164 Cell Line WEHI-164 cells were cultured under identical conditions in the presence and absence of TNF 10ng/ml at 37°C. At various time intervals, cells were collected and disrupted and mitochondria were isolated according to the the protocole described in Materials and Methods, endogenous free fatty acids were analyzed by gas liquid chromatography. The qualitative and quantitative analysis of free fatty acids show that only the free polyunsaturated fatty acid level is deeply altered under TNF treatment which suggests a change in PLA2 activity. As can be seen in Figure 1, in the absence of treatment the rate of free polyunsaturated fatty acids is very low and does not signifi532
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Kinetics of the polyunsaturated fatty acid release. WEHI-164 cells were cultured in presence (assay) and absence (control) of TNF 10ng/ml. At different points, cells were collected and mitochondria were isolated as described in Materials and Methods; free fatty acids were analyzed by gas liquid chromatography. The values are the means ± SD (n 4 3). Control: untreated cells, assay: TNF-treated cells.
cantly change, while under TNF treatment it is dramatically increased. The fatty acid release starts after 2h of treatment and reaches a maximum after 4h TNF additions. This rapid change of PLA2 activity could be caused by the free fatty acids themselves. These latters are known to stimulate Na+/Ca2+ exchange which is most apparent at relatively low fatty acid levels (30). Their increase would in turn likely enhance membrane permeability to Ca2+ which might further activate PLA2. This explanation is the more likely as mitochondria display Na+ dependent Ca2+ release (31–32). The polyunsaturated fatty acid species consist of long chains from C18:2 to C22:6 (Figure 2), after 4h treatment the releases of C22:5 and C22:6 species are dramatically increased. We notice that there is not specific arachidonic acid or fatty acid release as it occurred in the cytosol. This result shows that the activation of the phospholipase itself seems to be more essential that the specificity of the fatty acids. However their high degrees of unsaturation (from 2 up to 6 double bonds) suggest
FIG. 2. Qualitative analysis of free polyunsaturated fatty acids. WEHI-164 cells were set up as described in Materials and Methods. At different points of TNF (10ng/ml) treatment 2h-3h-4h-5h-6h, mitochondrial free fatty acids were isolated as described and analyzed by gas liquid chromatography and identified by their retention times as compared to authentic standards. 533
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that these latters could participate to the oxidation process involved in the TNF cytotoxicity. The mitochondrial PLA2 as well as the cytosolic counterpart are early involved in the TNF cell signalling but their roles seem distinct in regard to their fatty acid specificities and the role that Ca2+ plays in their activities (33). Relative Kinetics of the PLA2 Activity and Cell Death In order to know whether the hydrolysis of polyunsaturated fatty acids in mitochondria is detected prior to cytolysis, parallel cultures were set up to measure these two processes in the presence or absence of 10ng/ml TNF after different periods of treatment. As can be seen in Figure 3, the release of polyunsaturated fatty acids becomes evident as early as 4h after TNF additions. This event precedes the cell lysis as measured by the cristal violet viability test. This latter starts to occur only about 9h after TNF additions, most cells are dead after 25h. The PLA2 activity is not measured later than 6h after TNF additions, from this period the lysis process starts to occurs and mitochondria have lost their structural integrities and cannot be isolated. Comparative Studies of SDH and PLA2 Activities Our preliminary studies about TNF action at the mitochondrial level have shown that TNF induced changes in mitochondrial enzyme activities (23). One enzyme, SDH was specifically activated in TNF-sensitive cells including U937 (human monocytic), WEHI-164 (murine fibrosarcoma) and ME-180 (human cervical carcinoma). SDH was activated by TNF concentrations which also cause cytolysis. In contrast TNF did not activate SDH in TNF-resistant variants derived from U937 and WEHI-164. The pattern analysis of SDH activity and cell viability respectively measured in mitochondria and cultured cells show that SDH activity increases with the length of the TNF treatment reaching a maximum (76% activity increase) at 5h and then decline. The peak of SDH activation occurs before that cell death being significant and reaching the maximum level. So PLA2 and SDH activations are early events which precede cell death. Their comparative kinetic studies (Figure 4) show that the modifications of their activities occur in the same time interval (2-6h TNF treatment), their maximal activities are slightly out of phase. Beyond 5h of TNF treatment, SDH activation as well as PLA2 activity strongly decrease suggesting that the cell lysis process has started. It is known that non-esterified long chain fatty acids have an inhibitory effect on the
FIG. 3. Kinetics of the polyunsaturated fatty acid release and cell death. Parallel culture of WEHI-164 cells were set up in the presence of TNF (10ng/ml). At different time points, cells were removed to determine mitochondrial polyunsaturated fatty acid release or cell death. Cell death was determined by the crystal violet cell viability test. 534
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Kinetics of the polyunsaturated fatty acid release and SDH activation. Parallel cultures of WEHI-164 cells were set up in the presence of TNF (10ng/ml). At different points, cells were removed to measure the polyunsaturated fatty acid release and SDH activation in isolated mitochondria as described in Materials and Methods.
respiratory chain especially on primary dehydrogenases, the best known example is the interconversion of the pyruvate dehydrogenase complex (34). In mitochondria, PLA2 is localized in the outer membrane contact sites (24), a micro-compartment of the outer membrane, and in the inner membrane (12) the permeability of which is controlled by this enzyme; so an alteration of its activity could affect the SDH activity. TNF Triggered Morphological Changes in Mitochondria Matthews, (19) first have reported that sensitive cells treated with TNF displayed abnormalities in their mitochondria. In order to show an expected relationship between morphological and functional changes induced by TNF in mitochondria, WEHI-164 cells were incubated with TNF 10ng/ml and examined in electron microscopy after different periods of treatment. As can be shown in Figure 5A, in the absence of treatment mitochondria show a typical structure characterized by the presence of numerous and tight crests. In contrast, as early as 3h after TNF additions (Figure 5B), mitochondria display a swollen appearance (orthodoxe state) characterized by a large and clear matrix space and few crests, these structural modifications increase with longer periods of treatment (Figures 5C-5D-5E). This configuration is specific of metabolic alterations, especially the development of an unspecific permeability to small molecules like ions (Ca2+) and low molecular proteins (35). This change of configuration has been reported to be a major consequence of calcium uptake (36) and so triggers required conditions to indirectly increase the activity of SDH. Comparatively the nuclear structure is well-preserved under these conditions which means that the lysis has not yet started, but 24h later the nuclear integrity is lost (data not shown). These results show that morphologic as well as metabolic changes in mitochondria are early events in the TNF-induced cytotoxicity process. DISCUSSION The first work of Matthews (19) on the involvement of mitochondria in TNF-treated cells only reported morphologic changes, the metabolic significance of these structural abnormalities was not investigated. By studying the time course of the fatty acid release in mitochondria isolated from TNF-treated WEHI-164 cells, we show that TNF induces the stimulation of the mitochondrial 535
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. Electron microscopy observations in situ of TNF-treated and untreated cells. WEHI-164 cells were incubated in RPMI 1640 plus 10% FCS (foetal calf serum) at 37°C without (A) or with TNF (10ng/ml) and examined as described. Lengths of TNF treatment: 3h (B)-4h (C)-5h (D)-6h (E). Bar: 0.5mm. 536
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
PLA2 at an early stage before that cell death occurs. It has been reported that the enhancement of this enzyme activity increased membrane permeability to small molecules especially to Ca2+ (29) and could lead to the cell damage. In addition mitochondria play a central role in the cell injury associated with many pathophysiological states (37). So, this enzyme could be implicated as a mediator of metabolic disfunctions induced by TNF. Its activity is regulated by calcium (17). While there are not direct evidences for a role of intracellular calcium influx triggered by TNF, this cation is essential for the TNF induced generation of superoxide anions. It has been demonstrated that blocking the uptake of Ca2+ by mitochondria decreased the cytotoxic action of TNF on L929 cells (22). The importance of mitochondrial PLA2 in the cytotoxic process is emphasized by the works of Hatch et al., (38) who have reported that its activity is dramatically enhanced in vivo by endotoxin. On the other hand this activity increase could be at the origin of the TNF-induced SDH activation observed under the same conditions (23). Our work in liver mitochondria (28) has reported that a Ca2+-stimulation of PLA2 provoked a stimulation of the SDH activity. This change of activity is linked to a Ca2+ flux in mitochondria mediated by a permeability alteration, this cation is not directly responsible for the SDH activation but rather be involved in the process which rules it. Consequently it would not be unlikely to consider that the stimulation of the PLA2 activity could be a key step in the mechanism of TNF trans-membrane signalling in mitochondria. In the other hand among the mitochondrial abnormalities, another important fact is the production of superoxide anions (22). The participation of PLA2 to this event could be mediated by SDH based on the fact that this dehydrogenase belongs to the site II of the respiratory chain which is responsible for the main production of free oxygen radicals (39–40). This hypothesis was indirectly confirmed by the works of Schulze-OSthoff et al., (21) and Taylor et al., (41). The first has shown that inhibition of the site II of the respiratory chain by the thenoyltrifluoroacetone markedly protected against TNF cytotoxicity. The results of the second reported the involvement of this respiratory chain complex in the enhancement of the mitochondrial oxidative stress caused by sepsis. Superoxide anions are converted in hydrogen peroxides by the MnSOD (manganese-superoxide dismutase) these compounds can generate hydroxyl radicals which in turn oxidize lipids in mitochondria. So lipid peroxidation could be one step of the TNF cytotoxic process. Two reports argue for this hypothesis. The enrichment of TNF sensitive WEHI-164 with n-3 and n-6 fatty acids has shown that TNF cytotoxic pathway involved directly or indirectly the metabolism of long chain polyunsaturated fatty acids (42). These authors (43) have brought some additional evidences by studying the protective effect of BHT (butylated hydroxitoluene) against TNF cytotoxicity. Based on these results the loss of electron transport along the respiratory chain observed under these conditions could be the consequence of the peroxidation of the lipids constitutive of the inner membrane. All these information lead to postulate that the mitochondrial PLA2 could be implicated in the TNF cell signalling pathway and responsible for the mitochondrial abnormalities. This hypothesis is based on the role of this enzyme in the regulation of the inner membrane permeability. However further studies would be needed to elucidate among the intracellular events induced by TNF which are responsible for the change of the PLA2 activity in mitochondria. ACKNOWLEDGMENTS Research was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), le Centre National de la Recherche Scientifique (CNRS), and the Université of Lyon (Lyon-Sud Medical School). The authors thank F. Lermé for carrying out electron microscopy experiments, and the Centre de Microscopie Electronique Appliqué à la Biologie et à la Géologie (Villeurbanne, France) and C. Rey for gas chromatography experiments.
REFERENCES 1. Carswell, E. A., Old, L. J., Kassel, R. L., Green, S., Fiore, N., and Williamson, B. (1975) Proc. Natl. Acad. Sci. USA 72, 3666–3670. 2. Fiers, W. (1991) FEBS Lett. 285, 199–212. 3. Camussi, G., Albano, E., Tetta, C., and Bussolino, F. (1991) Eur. J. Biochem. 202, 3–14. 537
Vol. 221, No. 3, 1996 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Vilcek, J., and Lee, T. H. (1991) J. Biol. Chem. 266, 7313–7316. Oka, S., and Arita, H. (1991) J. Biol. Chem. Biol. 266, 9956–9960. Goppelt-Strube, M., and Rehfeldt, W. (1992) Biochim. Biophys. Acta 1127, 163–167. Hoeck, W. G., Chakkodabylu, S. R., David, J. C., Fan, N., and Heller, R. A. (1993) Proc. Natl. Acad. Sci. 90, 4475– 4479. Hayakawa, M., Ishida, N., Takeuchi, K., Shibamoto, S., Hori, T., Oku, N., Ito, F., and Tsujimoto, M. (1993) J. Biol. Chem. 268, 11290–11295. Nakano, T., Ohara, O., Teraoka, N., and Arita, H. (1990) J. Biol. Chem. 265, 12745–12748. Kudo, I., Murakimi, M., Hara, S., and Inoue, K. (1993) Biochim. Biophys. Acta 117, 217–231. Nachbaur, J., Colbeau, A., and Vignais, P. M. (1972) Biochim. Biophys. Acta 274, 426–446. Waite, M., and Sisson, P. (1971) Biochemistry 10, 2377–2383. Zurini, M., Hugentobler, G., and Gazotti, P. (1981) Eur. J. Biochem. 119, 517–521. De Winter, J. M., Vianen, G. M., and van den Bosch, H. (1982) Biochim. Biophys. Acta 712, 332–341. Natori, Y., Karasawa, K., Arai, H., Tamori-Natori, Y., and Nojima, S. (1983) J. Biochem. (Tokyo) 93, 631–637. Aarsman, A. J., Neys, F., and van den Bosch, H. (1984) Biochim. Biophys. Acta. 792, 363–366. De Winter, J. M., Korpancova, J., and van den Bosch, H. (1984) Arch. Biochem. Biophys. 234, 243–252. Lenting, H. B. M., Neys, F. W., and van den Bosch, H. (1988) Biochim. Biophys. Acta 961, 129–138. Matthews, N. (1983) Br. J. Cancer 48, 405–410. Lancaster, J. R. J., Laster, M. S., Jr., and Gooding, L. R. (1989) FEBS lett. 248, 169–174. Schulze-Osthoff, K., Bakker, C. A., Vanhaesebroeck, B., and Beyaert, R. (1992) J. Biol. Chem. 267, 5317–5323. Hennet, T., Richter, C., and Peterhans, E. (1993) Biochem. J. 289, 587–592. Levrat, C., Larrick, J. W., and Wright, S. C. (1991) Life Sci. 49, 1731–1737. Levrat, C., and Louisot, P. (1992) Biochem. Biophys. Res. Commun. 183, 719–724. Heller, R. A., Song, K., Fan, F., and Chang, D. J. (1992) Cell 70, 47–56. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76–85. Ackrell, B. A. C., Kearney, E. B., and Singer, T. P. (1979) in Methods in Enzymology (Fleishcer, S., and Packer, J., Eds.), Vol. 53, pp. 466–495. Academic Press, Orlando, FL. Levrat, C., and Louisot, P. (1994) Biochem. Mol. Biol. Int. 34, 569–578. Pietsch, A., and Lorenz, R. L. (1993) Lipids 28, 945–947. Philipson, K. D., and Ward, R. (1985) J. Biol. Chem. 260, 9666–9671. Heffron, J. J. A., and Harris, E. J. (1981) Biochem. J. 194, 925–929. Golstone, T. P., and Crompton, M. (1982) Biochem. J. 204, 369–371. Dennis, E. A. (1994) J. Biol. Chem. 269, 13057–13060. Wojtczak, L., and Schönfeld, P. (1993) Biochim. Biophys. Acta. 1183, 41–57. Haworth, R. A., and Hunter, D. R. (1979) Arch Biochem Biophys. 195, 460–467. Hunter, R., Haworth, R. A., and Southard, J. H. (1976) J. Biol. Chem. 251, 5069–5077. Smith, M. W., Collan, Y., Kahnyg, M. W., and Trump, B. F. (1980) Biochim. Biophys. Acta 618, 192–201. Hatch, G. M., Vance, D. E., and Wilton, D. C. (1993) Biochem. J. 293, 143–150. Boveris, A., Oshina, N., and Chance, B. (1972) Biochem. J. 128, 617–630. Glinn, M., Ernster, L., and Lee, C. P. (1991) Arch. Biochem. Biophys. 290, 57–65. Taylor, D. E., Ghio, A. J., and Piantadosi, C. A. (1995) Arch. Biochem. Biophys. 316, 70–76. Brekke, O. L., Espevik, T., Bardal, T., and Bjerve, K. S. (1992) Cells. Lipids 27, 161–168. Brekke, O. L., Espevik, T., and Bjerve, K. S. (1994) Lipids 29, 91–102.
538
221, 531–538 (1996)
0631
Increase of Mitochondrial PLA2-Released Fatty Acids Is an Early Event in Tumor Necrosis Factor a-Treated WEHI-164 Cells Christiane Levrat and Pierre Louisot1 Department of Biochemistry, INSERM-CNRS U.189, BP12, 69921 Oullins Cedex, France Received March 4, 1996 We have previously reported that TNF induced changes in mitochondrial enzymes, one of which, succinatedehydrogenase, is specifically activated in various TNF-sensitive cell lines. In an attempt to further characterize the mechanism of trans-membrane signalling at the mitochondrial level, we have oriented our investigation to the study of phospholipase A2 activity localized in this organelle isolated from TNF-treated WEHI-164 cells. Under physiological conditions, this enzyme has a very low basal activity near the resting state, while under TNF treatment its activity is dramatically increased. This event is induced by TNF concentrations which also cause cytolysis; however, the activity of this enzyme is increased several hours before maximum cytotoxicity occurs. The activation of the mitochondrial phospholipase A2 is not arachidonoyl or fatty acid selective, as is the case for the cytosolic species. Phospholipase A2 and succinate-dehydrogenase display higher activities simultaneously under TNF treatment. This change in activity is linked to morphologic modifications. Mitochondria in particular display an orthodox state characterized by a large and clear matrix space and few crests. © 1996 Academic Press, Inc.
TNF-a (Tumor Necrosis Factor) is a multipotent cytokine capable of killing various tumor cells in vitro (1). The interaction of TNF with membrane receptors initiates a cascade of intracellular events that leads to cell destruction. Among the postreceptor events, the arachidonic acid release triggered by the activation of phospholipase A2 (PLA2) seems to be an essential step in the TNF cytotoxicity (2–4). It has been shown that PLA2 involved in the TNF-induced arachidonic acid release displays the characteristics of the arachidonoyl-selective cytosolic PLA2, this enzyme is characterized by a high molecular weight (group IV) (5–8). However Nakano et al. (9) provided some evidences for the participation of the low molecular weight PLA2 (group II) which are membrane-bound enzymes. They hydrolyze phosphatidylethanolamine much more effectively than phosphatidylcholine, have no substrate specificity for arachidonic acid or other polyunsaturated fatty acids but they have an absolute requirement for Ca2+ for expressing of full activity (10). TNF would increase the RNA level of this PLA2 group. This specie is ubiquitously distributed in the cell. In addition to its association to microsomes, golgi membranes, plasma membranes and lysosomal membranes (11) this enzyme is also associated with both outer and inner membranes of mitochondria (11–13) and has been purified to near homogeneity (14–16). The mitochondrial specie requires Ca2+ for its catalytic action which is characterized by two activity plateaus upon variations of the Ca2+ concentrations with KCa2+ values of 14mM and 2.4mM respectively (17–18). The participation of mitochondria in TNF-induced cytotoxicity was inferred from the observation that these organella present structural alterations in their morphology as a swollen appearance (19), metabolic disfunctions as impaired respiration (20–21), superoxide anion generation (22) and an early activation of succinate dehydrogenase (SDH) (23). The mechanisms that manage the mitochondrial alterations are unknown. A hypothetical role of the mitochondrial PLA2 could be expected taking into account that this enzyme is responsible for changes of the inner membrane permeability, and consequently alterations of the metabolism (24). The present study reports that 1 Address reprint requests to Pr. Pierre Louisot, Department of Biochemistry INSERM-CNRS U.189, BP12, 69921 Oullins Cedex, France. Fax: (33) 78-50-71-52, E-mail: [email protected].
531 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TNF induces the increase of the mitochondrial PLA2 activity at the early stage of the cytotoxic process. MATERIALS AND METHODS Reagents. Purified human r-TNFa (specific activity: 2×107 U/mg) was purchased from Genzyme (Boston USA). All chemicals and cell culture products were respectively purchased from Sigma (Saint-Louis, Mo) and Gibco BRL-Life technologies (France). Cell line. Murine fibrosarcoma WEHI-164 cells from ATCC (Rockeville, Mo) were cultured in RPMI 1640 supplemented with 10% foetal calf serum (SEPRACOR-France) and containing penicillin (100 U/ml) streptomycin (100 mg/ml), nonessential amino-acids (GIBCO-BRL). The cell line was routinely tested for mycoplasma. Cytotoxicity assay. Changes in cell viability were measured with cells plated in 24 well microtiter at 4.104 cells per well in 500ml of growth medium. 24h later when cells reached confluency, TNF additions were made ranging in concentrations from 10 to 100 ng/ml or as specified. Each treatment consisted of four replicated wells. Cells were incubated for 3h up to 48h as specified, and the cell viability was measured by the cristal violet method (25). The percent of dead cells is expressed as follows: (1-ODassay/OD control) × 100. Ultrastructural studies. Cells were fixed with 1% gluteraldehyde 0.1M phosphate buffer pH7.4, postfixed with 1% osmium tetroxide and dehydrated in graded ethanols. Thin sections were mounted on collodion carbon-coated nickel grids and stained with 3% uranyl acetate and lead citrate. They were examined in a CM 10 Philips electron microscope operated at 80kV. Preparation of mitochondria. Confluent cells (3×108) were washed twice with HBSS, then harvested and washed three times with PBS (0.14M NaCl, 0.01M KCl, 0.01M sodium phosphate pH 7.4). Cells were washed once with 0.25M sucrose 10mM Tris pH 7.4, 1mM EGTA 0.1% BSA (bovine serum albumin) then resuspended in the same medium and were broken with a Dounce homogenizer. The cell homogenate was centrifuged at 1500g for 10min at 4°C. The nuclear pellet was reextracted and the two supernatants were combined and centrifuged at 10 000g for 10min at 4°C and the mitochondria pellet obtained were layered on discontinuous gradients (38% metrizamide, 17% metrizamide, 6% Percoll) and centrifuged 15min at 50 500g. Pure mitochondria were recovered at the interface 17% metrizamide/35% metrizamide. Mitochondrial purity was assessed by measurements of marker enzyme activities and ultrastructural studies. Proteins were measured by the bicinchoninic acid assay (26) using bovine serum albumin as standard. Succinate dehydrogenase assay. The activity of SDH was measured by a modification of the phenazine metosulfate assay described by Ackrell et al., (27–28). Measurement of phospholipase A2 activity. PLA2 activity associated with mitochondria was determined by monitoring the release of polyunsaturated fatty acids from endogenous phospholipids. Mitochondria isolated from TNF-treated and untreated cells were delipidated according to the modified Folch’s procedure (24) in the presence of butylated hydroxytoluene. A known amount of heptadecanoic acid was present and carried through the subsequent procedures as a internal standard. Isolation and quantitative estimation of PLA2-released fatty acids. Free fatty acids were isolated by using prepacked aminopropyl bonded silica columns and eluted with a mixture of diethyl ether/acetic acid (29). Prior to gas chromatography analysis, free fatty acids were converted to methyl esters using 5% (V/V) sulfuric acid in methanol. Fatty acid methyl esters were taken to dryness under nitrogen and dissolved in isooctane for analysis by gas liquid chromatography on a Perkin Elmer Autosystem Gas Chromatography coupled with a PE Nelson 1020 integrator. The compounds were separated on a SP 2380 capillary column (0.32mm × 30m; Supelco). Runs were first temperature programmed from 50°C to 150°C at 32°C/min then to 215°C at 2°C/min; azote was used as the carrier gas. Compounds were identified by their retention times as compared to authentic standards. To correct for procedural losses and volume inconsistencies between samples, peak areas were normalized to the area of the methyl heptadecanoate internal standard peak. PLA2 activity was taken to be the sum of polyunsaturated fatty acids.
RESULTS TNF Induces an Increase of the Mitochondrial PLA2-Released Fatty Acid Rate in the WEHI-164 Cell Line WEHI-164 cells were cultured under identical conditions in the presence and absence of TNF 10ng/ml at 37°C. At various time intervals, cells were collected and disrupted and mitochondria were isolated according to the the protocole described in Materials and Methods, endogenous free fatty acids were analyzed by gas liquid chromatography. The qualitative and quantitative analysis of free fatty acids show that only the free polyunsaturated fatty acid level is deeply altered under TNF treatment which suggests a change in PLA2 activity. As can be seen in Figure 1, in the absence of treatment the rate of free polyunsaturated fatty acids is very low and does not signifi532
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Kinetics of the polyunsaturated fatty acid release. WEHI-164 cells were cultured in presence (assay) and absence (control) of TNF 10ng/ml. At different points, cells were collected and mitochondria were isolated as described in Materials and Methods; free fatty acids were analyzed by gas liquid chromatography. The values are the means ± SD (n 4 3). Control: untreated cells, assay: TNF-treated cells.
cantly change, while under TNF treatment it is dramatically increased. The fatty acid release starts after 2h of treatment and reaches a maximum after 4h TNF additions. This rapid change of PLA2 activity could be caused by the free fatty acids themselves. These latters are known to stimulate Na+/Ca2+ exchange which is most apparent at relatively low fatty acid levels (30). Their increase would in turn likely enhance membrane permeability to Ca2+ which might further activate PLA2. This explanation is the more likely as mitochondria display Na+ dependent Ca2+ release (31–32). The polyunsaturated fatty acid species consist of long chains from C18:2 to C22:6 (Figure 2), after 4h treatment the releases of C22:5 and C22:6 species are dramatically increased. We notice that there is not specific arachidonic acid or fatty acid release as it occurred in the cytosol. This result shows that the activation of the phospholipase itself seems to be more essential that the specificity of the fatty acids. However their high degrees of unsaturation (from 2 up to 6 double bonds) suggest
FIG. 2. Qualitative analysis of free polyunsaturated fatty acids. WEHI-164 cells were set up as described in Materials and Methods. At different points of TNF (10ng/ml) treatment 2h-3h-4h-5h-6h, mitochondrial free fatty acids were isolated as described and analyzed by gas liquid chromatography and identified by their retention times as compared to authentic standards. 533
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that these latters could participate to the oxidation process involved in the TNF cytotoxicity. The mitochondrial PLA2 as well as the cytosolic counterpart are early involved in the TNF cell signalling but their roles seem distinct in regard to their fatty acid specificities and the role that Ca2+ plays in their activities (33). Relative Kinetics of the PLA2 Activity and Cell Death In order to know whether the hydrolysis of polyunsaturated fatty acids in mitochondria is detected prior to cytolysis, parallel cultures were set up to measure these two processes in the presence or absence of 10ng/ml TNF after different periods of treatment. As can be seen in Figure 3, the release of polyunsaturated fatty acids becomes evident as early as 4h after TNF additions. This event precedes the cell lysis as measured by the cristal violet viability test. This latter starts to occur only about 9h after TNF additions, most cells are dead after 25h. The PLA2 activity is not measured later than 6h after TNF additions, from this period the lysis process starts to occurs and mitochondria have lost their structural integrities and cannot be isolated. Comparative Studies of SDH and PLA2 Activities Our preliminary studies about TNF action at the mitochondrial level have shown that TNF induced changes in mitochondrial enzyme activities (23). One enzyme, SDH was specifically activated in TNF-sensitive cells including U937 (human monocytic), WEHI-164 (murine fibrosarcoma) and ME-180 (human cervical carcinoma). SDH was activated by TNF concentrations which also cause cytolysis. In contrast TNF did not activate SDH in TNF-resistant variants derived from U937 and WEHI-164. The pattern analysis of SDH activity and cell viability respectively measured in mitochondria and cultured cells show that SDH activity increases with the length of the TNF treatment reaching a maximum (76% activity increase) at 5h and then decline. The peak of SDH activation occurs before that cell death being significant and reaching the maximum level. So PLA2 and SDH activations are early events which precede cell death. Their comparative kinetic studies (Figure 4) show that the modifications of their activities occur in the same time interval (2-6h TNF treatment), their maximal activities are slightly out of phase. Beyond 5h of TNF treatment, SDH activation as well as PLA2 activity strongly decrease suggesting that the cell lysis process has started. It is known that non-esterified long chain fatty acids have an inhibitory effect on the
FIG. 3. Kinetics of the polyunsaturated fatty acid release and cell death. Parallel culture of WEHI-164 cells were set up in the presence of TNF (10ng/ml). At different time points, cells were removed to determine mitochondrial polyunsaturated fatty acid release or cell death. Cell death was determined by the crystal violet cell viability test. 534
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Kinetics of the polyunsaturated fatty acid release and SDH activation. Parallel cultures of WEHI-164 cells were set up in the presence of TNF (10ng/ml). At different points, cells were removed to measure the polyunsaturated fatty acid release and SDH activation in isolated mitochondria as described in Materials and Methods.
respiratory chain especially on primary dehydrogenases, the best known example is the interconversion of the pyruvate dehydrogenase complex (34). In mitochondria, PLA2 is localized in the outer membrane contact sites (24), a micro-compartment of the outer membrane, and in the inner membrane (12) the permeability of which is controlled by this enzyme; so an alteration of its activity could affect the SDH activity. TNF Triggered Morphological Changes in Mitochondria Matthews, (19) first have reported that sensitive cells treated with TNF displayed abnormalities in their mitochondria. In order to show an expected relationship between morphological and functional changes induced by TNF in mitochondria, WEHI-164 cells were incubated with TNF 10ng/ml and examined in electron microscopy after different periods of treatment. As can be shown in Figure 5A, in the absence of treatment mitochondria show a typical structure characterized by the presence of numerous and tight crests. In contrast, as early as 3h after TNF additions (Figure 5B), mitochondria display a swollen appearance (orthodoxe state) characterized by a large and clear matrix space and few crests, these structural modifications increase with longer periods of treatment (Figures 5C-5D-5E). This configuration is specific of metabolic alterations, especially the development of an unspecific permeability to small molecules like ions (Ca2+) and low molecular proteins (35). This change of configuration has been reported to be a major consequence of calcium uptake (36) and so triggers required conditions to indirectly increase the activity of SDH. Comparatively the nuclear structure is well-preserved under these conditions which means that the lysis has not yet started, but 24h later the nuclear integrity is lost (data not shown). These results show that morphologic as well as metabolic changes in mitochondria are early events in the TNF-induced cytotoxicity process. DISCUSSION The first work of Matthews (19) on the involvement of mitochondria in TNF-treated cells only reported morphologic changes, the metabolic significance of these structural abnormalities was not investigated. By studying the time course of the fatty acid release in mitochondria isolated from TNF-treated WEHI-164 cells, we show that TNF induces the stimulation of the mitochondrial 535
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. Electron microscopy observations in situ of TNF-treated and untreated cells. WEHI-164 cells were incubated in RPMI 1640 plus 10% FCS (foetal calf serum) at 37°C without (A) or with TNF (10ng/ml) and examined as described. Lengths of TNF treatment: 3h (B)-4h (C)-5h (D)-6h (E). Bar: 0.5mm. 536
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
PLA2 at an early stage before that cell death occurs. It has been reported that the enhancement of this enzyme activity increased membrane permeability to small molecules especially to Ca2+ (29) and could lead to the cell damage. In addition mitochondria play a central role in the cell injury associated with many pathophysiological states (37). So, this enzyme could be implicated as a mediator of metabolic disfunctions induced by TNF. Its activity is regulated by calcium (17). While there are not direct evidences for a role of intracellular calcium influx triggered by TNF, this cation is essential for the TNF induced generation of superoxide anions. It has been demonstrated that blocking the uptake of Ca2+ by mitochondria decreased the cytotoxic action of TNF on L929 cells (22). The importance of mitochondrial PLA2 in the cytotoxic process is emphasized by the works of Hatch et al., (38) who have reported that its activity is dramatically enhanced in vivo by endotoxin. On the other hand this activity increase could be at the origin of the TNF-induced SDH activation observed under the same conditions (23). Our work in liver mitochondria (28) has reported that a Ca2+-stimulation of PLA2 provoked a stimulation of the SDH activity. This change of activity is linked to a Ca2+ flux in mitochondria mediated by a permeability alteration, this cation is not directly responsible for the SDH activation but rather be involved in the process which rules it. Consequently it would not be unlikely to consider that the stimulation of the PLA2 activity could be a key step in the mechanism of TNF trans-membrane signalling in mitochondria. In the other hand among the mitochondrial abnormalities, another important fact is the production of superoxide anions (22). The participation of PLA2 to this event could be mediated by SDH based on the fact that this dehydrogenase belongs to the site II of the respiratory chain which is responsible for the main production of free oxygen radicals (39–40). This hypothesis was indirectly confirmed by the works of Schulze-OSthoff et al., (21) and Taylor et al., (41). The first has shown that inhibition of the site II of the respiratory chain by the thenoyltrifluoroacetone markedly protected against TNF cytotoxicity. The results of the second reported the involvement of this respiratory chain complex in the enhancement of the mitochondrial oxidative stress caused by sepsis. Superoxide anions are converted in hydrogen peroxides by the MnSOD (manganese-superoxide dismutase) these compounds can generate hydroxyl radicals which in turn oxidize lipids in mitochondria. So lipid peroxidation could be one step of the TNF cytotoxic process. Two reports argue for this hypothesis. The enrichment of TNF sensitive WEHI-164 with n-3 and n-6 fatty acids has shown that TNF cytotoxic pathway involved directly or indirectly the metabolism of long chain polyunsaturated fatty acids (42). These authors (43) have brought some additional evidences by studying the protective effect of BHT (butylated hydroxitoluene) against TNF cytotoxicity. Based on these results the loss of electron transport along the respiratory chain observed under these conditions could be the consequence of the peroxidation of the lipids constitutive of the inner membrane. All these information lead to postulate that the mitochondrial PLA2 could be implicated in the TNF cell signalling pathway and responsible for the mitochondrial abnormalities. This hypothesis is based on the role of this enzyme in the regulation of the inner membrane permeability. However further studies would be needed to elucidate among the intracellular events induced by TNF which are responsible for the change of the PLA2 activity in mitochondria. ACKNOWLEDGMENTS Research was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), le Centre National de la Recherche Scientifique (CNRS), and the Université of Lyon (Lyon-Sud Medical School). The authors thank F. Lermé for carrying out electron microscopy experiments, and the Centre de Microscopie Electronique Appliqué à la Biologie et à la Géologie (Villeurbanne, France) and C. Rey for gas chromatography experiments.
REFERENCES 1. Carswell, E. A., Old, L. J., Kassel, R. L., Green, S., Fiore, N., and Williamson, B. (1975) Proc. Natl. Acad. Sci. USA 72, 3666–3670. 2. Fiers, W. (1991) FEBS Lett. 285, 199–212. 3. Camussi, G., Albano, E., Tetta, C., and Bussolino, F. (1991) Eur. J. Biochem. 202, 3–14. 537
Vol. 221, No. 3, 1996 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Vilcek, J., and Lee, T. H. (1991) J. Biol. Chem. 266, 7313–7316. Oka, S., and Arita, H. (1991) J. Biol. Chem. Biol. 266, 9956–9960. Goppelt-Strube, M., and Rehfeldt, W. (1992) Biochim. Biophys. Acta 1127, 163–167. Hoeck, W. G., Chakkodabylu, S. R., David, J. C., Fan, N., and Heller, R. A. (1993) Proc. Natl. Acad. Sci. 90, 4475– 4479. Hayakawa, M., Ishida, N., Takeuchi, K., Shibamoto, S., Hori, T., Oku, N., Ito, F., and Tsujimoto, M. (1993) J. Biol. Chem. 268, 11290–11295. Nakano, T., Ohara, O., Teraoka, N., and Arita, H. (1990) J. Biol. Chem. 265, 12745–12748. Kudo, I., Murakimi, M., Hara, S., and Inoue, K. (1993) Biochim. Biophys. Acta 117, 217–231. Nachbaur, J., Colbeau, A., and Vignais, P. M. (1972) Biochim. Biophys. Acta 274, 426–446. Waite, M., and Sisson, P. (1971) Biochemistry 10, 2377–2383. Zurini, M., Hugentobler, G., and Gazotti, P. (1981) Eur. J. Biochem. 119, 517–521. De Winter, J. M., Vianen, G. M., and van den Bosch, H. (1982) Biochim. Biophys. Acta 712, 332–341. Natori, Y., Karasawa, K., Arai, H., Tamori-Natori, Y., and Nojima, S. (1983) J. Biochem. (Tokyo) 93, 631–637. Aarsman, A. J., Neys, F., and van den Bosch, H. (1984) Biochim. Biophys. Acta. 792, 363–366. De Winter, J. M., Korpancova, J., and van den Bosch, H. (1984) Arch. Biochem. Biophys. 234, 243–252. Lenting, H. B. M., Neys, F. W., and van den Bosch, H. (1988) Biochim. Biophys. Acta 961, 129–138. Matthews, N. (1983) Br. J. Cancer 48, 405–410. Lancaster, J. R. J., Laster, M. S., Jr., and Gooding, L. R. (1989) FEBS lett. 248, 169–174. Schulze-Osthoff, K., Bakker, C. A., Vanhaesebroeck, B., and Beyaert, R. (1992) J. Biol. Chem. 267, 5317–5323. Hennet, T., Richter, C., and Peterhans, E. (1993) Biochem. J. 289, 587–592. Levrat, C., Larrick, J. W., and Wright, S. C. (1991) Life Sci. 49, 1731–1737. Levrat, C., and Louisot, P. (1992) Biochem. Biophys. Res. Commun. 183, 719–724. Heller, R. A., Song, K., Fan, F., and Chang, D. J. (1992) Cell 70, 47–56. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76–85. Ackrell, B. A. C., Kearney, E. B., and Singer, T. P. (1979) in Methods in Enzymology (Fleishcer, S., and Packer, J., Eds.), Vol. 53, pp. 466–495. Academic Press, Orlando, FL. Levrat, C., and Louisot, P. (1994) Biochem. Mol. Biol. Int. 34, 569–578. Pietsch, A., and Lorenz, R. L. (1993) Lipids 28, 945–947. Philipson, K. D., and Ward, R. (1985) J. Biol. Chem. 260, 9666–9671. Heffron, J. J. A., and Harris, E. J. (1981) Biochem. J. 194, 925–929. Golstone, T. P., and Crompton, M. (1982) Biochem. J. 204, 369–371. Dennis, E. A. (1994) J. Biol. Chem. 269, 13057–13060. Wojtczak, L., and Schönfeld, P. (1993) Biochim. Biophys. Acta. 1183, 41–57. Haworth, R. A., and Hunter, D. R. (1979) Arch Biochem Biophys. 195, 460–467. Hunter, R., Haworth, R. A., and Southard, J. H. (1976) J. Biol. Chem. 251, 5069–5077. Smith, M. W., Collan, Y., Kahnyg, M. W., and Trump, B. F. (1980) Biochim. Biophys. Acta 618, 192–201. Hatch, G. M., Vance, D. E., and Wilton, D. C. (1993) Biochem. J. 293, 143–150. Boveris, A., Oshina, N., and Chance, B. (1972) Biochem. J. 128, 617–630. Glinn, M., Ernster, L., and Lee, C. P. (1991) Arch. Biochem. Biophys. 290, 57–65. Taylor, D. E., Ghio, A. J., and Piantadosi, C. A. (1995) Arch. Biochem. Biophys. 316, 70–76. Brekke, O. L., Espevik, T., Bardal, T., and Bjerve, K. S. (1992) Cells. Lipids 27, 161–168. Brekke, O. L., Espevik, T., and Bjerve, K. S. (1994) Lipids 29, 91–102.
538