- Email: [email protected]
Gliotoxin induces Mg2+ efflux from intact brain mitochondria
Neurochemistry International 45 (2004) 759–764
Gliotoxin induces Mg2+ efflux from intact brain mitochondria Mauro Salvi, Aleksandra Bozac, Antonio Toninello∗ Dipartimento di Chimica Biologica, Universita’ di Padova and Istituto di Neuroscienze del C.N.R., Unita’ per lo Studio delle Biomembrane, Via G. Colombo 3, 35121 Padua, Italy Received 4 September 2003; received in revised form 28 November 2003; accepted 14 January 2004 Available online 8 April 2004
Abstract Gliotoxin (GT) is a hydrophobic fungal metabolite of the epipolythiodioxopiperazine group which reacts with membrane thiols. When added to a suspension of energized brain mitochondria, it induces matrix swelling of low amplitude, collapse of membrane potential (Ψ ), and efflux of endogenous cations such as Ca2+ and Mg2+ , typical events of mitochondrial permeability transition (MPT) induction. These effects are due to opening of the membrane transition pore. The addition of cyclosporin A (CsA) or ADP slightly reduces membrane potential collapse, matrix swelling and Ca2+ efflux; Mg2+ efflux is not affected at all. The presence of exogenous Mg2+ or spermine completely preserve mitochondria against Ψ collapse, matrix swelling and Ca2+ release. Instead, Mg2+ efflux is only slightly affected by spermine. Our results demonstrate that, besides inducing MPT, gliotoxin activates a specific Mg2+ efflux system from brain mitochondria. © 2004 Elsevier Ltd. All rights reserved. Keywords: Gliotoxin; Rat brain mitochondria; Mg2+ ; Mitochondrial permeability transition
1. Introduction Gliotoxin (GT) is a hydrophobic fungal metabolite of the epipolythiodioxopiperazine family. It has a quinoid moiety and a disulfide bridge across the piperazine ring. It is capable of forming mixed disulphides with free thiol groups, and several reports suggest that the main toxic effects of GT in cells are due to its interaction with enzymes by mixed disulphide formation (Mason and Kidd, 1951; Middleton, 1974). Another property of GT is its ability to trigger a redox cycle in the presence of an appropriate reducing agent, generating reactive oxygen species (for a review, see Waring and Beaver, 1996). GT induces specific Ca2+ release from intact rat liver mitochondria. This stimulation is suggested to be due to the reaction of GT with thiols critically involved in the specific Ca2+ release pathway (Schweizer and Richter, 1994). GT also induces apoptotic death in numerous cell types (Waring and Beaver, 1996), and one report seems to demonstrate the involvement of the MPT pore (Kweon et al., 2003). The MPT pore is a regulated membrane channel which alAbbreviations: AdNT, adenine nucleotide translocase; MPT, mitochondrial permeability transition; RBM, rat brain mitochondria; GT, gliotoxin ∗ Corresponding author. Tel.: +39-049-827-6134; fax: +39-049-827-6133. E-mail address: [email protected] (A. Toninello). 0197-0186/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuint.2004.01.001
lows free diffusion of all solutes of <1500 Da and which is associated with mitochondrial swelling, membrane potential collapse, efflux of endogenous Mg2+ , rapid depletion of mitochondrial Ca2+ stores, and uncoupling of oxidative phosphorylation. The MPT is induced by supraphysiological levels of Ca2+ and by a large number of compounds, usually called inducers, and is selectively inhibited by the immunosuppressant cyclosporin A (CsA) (for a review, see Zoratti and Szabo, 1995; Crompton et al., 1999; Kim et al., 2003). The opening of the MPT pore can also be inhibited in liver and heart mitochondria by some physiological agents such as ADP, reduced pyridine nucleotides, and Mg2+ . Naturally occurring polyamines also have a similar effect, and spermine in particular has been demonstrated to be a strong inhibitory agent of this phenomenon (Tassani et al., 1995). It has been reported that MPT can also be induced in isolated rat brain mitochondria (RBM) (Kristal and Dubinsky, 1997; Maciel et al., 2001; Kristian et al., 2002). However, brain mitochondria exhibit unique MTP characteristics, including relative resistance to inhibition by CsA. Instead, Mg2+ and ADP strongly inhibit the phenomenon, to a level comparable with their efficacy in liver. It has also been found that dithiotreitol and N-ethylmaleimide are partial inhibitors, and ruthenium red almost completely blocks MPT induction (Kristal and Dubinsky, 1997; Kristian et al.,
760
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
2002). Increased reactive oxygen species are observed with the opening of the MPT pore in brain mitochondria induced by Ca2+ , and Ca2+ -induced brain mitochondrial MPT is significantly inhibited by catalase (Maciel et al., 2001). The aim of this investigation is to study the effect of GT on the permeability of the inner membrane of rat brain mitochondria, in order to highlight some important peculiarities of this type of organelle.
2. Experimental procedures 2.1. Mitochondrial preparations Rat brain mitochondria were purified by the Ficoll gradient method, according to Nicholls (1978), with some modifications (Salvi et al., 2003). Briefly, rat brain (cerebral cortex) was homogenized in isolation medium (320 mM sucrose, 5 mM Hepes, 0.5 mM EDTA, pH 7.4; 0.3% BSA was added during homogenisation and the first step of purification) and subjected to centrifugation (900 × g) for 5 min. The supernatant was centrifuged at 17,000×g for 10 min, to precipitate crude mitochondrial pellets. The pellets were resuspended in isolation medium plus 1 mM ATP and layered on top of a discontinuous gradient, composed of 2 ml of isolation medium containing 16% (w/v) Ficoll, 2 ml of isolation medium containing 14% (w/v) Ficoll, 3 ml of isolation medium containing 12% (w/v) Ficoll, and 3 ml of isolation medium containing 7% (w/v) Ficoll. The gradient was centrifuged for 30 min at 75,000×g. Mitochondrial pellets were suspended in isolation medium and centrifuged for 10 min at 800 × g. Again the pellets were suspended in isolation medium without EDTA. Protein content was measured by the biuret method with bovine serum albumin as a standard (Gornall et al., 1949). 2.2. Standard incubation procedures Mitochondria (1 mg protein per ml) were incubated in a water-jacketed cell at 20 ◦ C under continuous stirring. The standard medium contained 200 mM sucrose, 10 mM Hepes (pH 7.4), 5 mM succinate and 1.25 M rotenone. Variations and/or other additions are given with each experiment. 2.3. Mitochondrial swelling Mitochondrial swelling was determined by the change in the apparent absorbance of mitochondrial suspensions at 540 nm in a 3-ml cuvette, with a Kontron Uvikon mod. 922 spectrophotometer equipped with thermostatic control and magnetic stirrer. Mitochondria were suspended as indicated above and, upon stabilisation of the absorbance trace, swelling was assessed after additions of other compounds as described in the figure legends.
2.4. Membrane potential measurements The membrane potential was calculated on the basis of the movements of the lipid-soluble cation tetraphenylphosphonium (TPP+ ) through the inner membrane with a specific electrode for TPP+ prepared according to published procedures (Kamo et al., 1979) and an Ag/AgCl reference electrode. TPP+ was included at a concentration of 2 M in order to achieve high sensitivity in measurements and to avoid toxic effects on the proton ATPase and calcium flux (Jensen and Gunter, 1984; Karadjov et al., 1986). The measurement was started by adding mitochondria to the incubation medium supplemented with other compounds, as indicated in the appropriate legends, after TPP+ calibration. The membrane potential was calculated from the Nernst equation and corrected for non-specific intramitochondrial binding of TPP+ using the equation: Ψ = (Ψ electrode − 66.16 mV)/0.92 (Jensen et al., 1986). 2.5. Cation efflux Ca2+ and Mg2+ release was estimated by atomic absorption spectroscopy of the supernatant fraction. Mitochondria (1 mg/ml) were incubated in standard medium containing other compounds, as indicated in the specific figure legends. The suspension was stirred continuously and, at intervals of time, 1-ml portions of the suspension were withdrawn and centrifuged for 2 min in an Eppendorf bench centrifuge (model 5415C) at 15,000 rpm, this speed being sufficient to sediment at least 98% of the mitochondria. The supernatant fluids were removed and their Ca2+ and Mg2+ contents determined with a Perkin-Elmer 1100B spectrometer.
3. Results The results (Fig. 1) show that addition of 25 M Ca2+ to brain mitochondria incubated in standard medium does not induce any change in the apparent absorbance of the suspension (see trace control). Addition of 100 M GT, 3 min after that of Ca2+ , induces an absorbance decrease of about 0.1 units, indicative of mitochondrial swelling. When compared with that obtainable in liver mitochondria, a decrease in absorbance of about 1 unit (data not shown) means that the swelling of RBM may be considered of low amplitude. The presence of CsA or ADP, alone or in combination, only partially reduces the extent of the swelling, whereas the polyamine spermine not only blocks the phenomenon but induces significant shrinkage; Mg2+ shows strong inhibition (Fig. 1). Fig. 2 shows the effects of GT on the electrical potential (Ψ ) of Ca2+ -loaded mitochondria. As shown in this figure, our brain mitochondrial preparations, when incubated in standard medium, have a Ψ value of about 160 mV. After Ψ reaches steady state, addition of 25 M Ca2+ leads to
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
Fig. 1. Mitochondrial swelling induced by gliotoxin in isolated RBM. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 25 M Ca2+ in conditions described in Section 2. Then 100 M GT, 1 mM ADP, 1 M CsA, 100 M spermine and 1 mM Mg2+ were added. Dashed line: RBM incubated in absence of GT (control). Downward deflection indicates mitochondrial swelling. Experiments were performed five times with comparable results.
sudden transient depolarization of the membrane (followed by prompt repolarization) (trace control). When the Ca2+ pulse is 25 M, the new steady state is virtually identical to that preceding Ca2+ loading. With high Ca2+ pulses, the new steady state gradually falls (data not shown). Addition of GT induces an almost complete drop in Ψ , which is partially prevented by CsA or ADP. Addition of ADP, quickly followed by CsA after the drop in Ψ , is completely ineffective. If ADP and CsA fail to completely prevent or to restore Ψ collapse by GT (see Fig. 2), the presence of spermine in the incubation medium, as demonstrated by the results of Fig. 3A, at 0.1 mM concentration, completely prevents a Ψ drop by GT in the presence of Ca2+ . Indeed, if the polyamine is added after Ψ collapse, sudden repo-
761
larization to normal values is observed. As reported in Fig. 3B, similar results are obtained if spermine is substituted by 1 mM Mg2+ , since this cation can protect against Ψ collapse and restore it after collapse. Addition of ADP after repolarisation by spermine (Fig. 3A) or Mg2+ (Fig. 3B) also induces the characteristic state 3/state 4 Ψ transition, indicative of the maintenance of the bioenergetic capacity of mitochondria. Fig. 4A shows the effect of GT on Ca2+ movements. When RBM are suspended in the incubation medium containing 25 M Ca2+ , they accumulate the cation and retain it until anaerobiosis (see trace control). Addition of GT after 3 min induces gradual release of Ca2+ , which is complete after 20 min. The presence of ADP or CsA in the incubation medium provokes reduced inhibition of Ca2+ efflux; spermine or Mg2+ completely block this GT-induced release. Generally, when mitochondria are suspended in an incubation medium, they instantaneously release an aliquot of Mg2+ from the intermembrane space. In this case, RBM, as shown in Fig. 4B, release about 3 nmol/mg of Mg2+ protein, in all experimental conditions. Addition of Ca2+ induces a slight gradual Mg2+ efflux (trace control) which is also observable in the presence of ADP, CsA or spermine. This efflux of about 0.5 nmol/mg is very probably due to the interaction of Ca2+ during its cycling across the inner membrane with membrane binding sites having the same affinity for both cations. Subsequent addition of GT provokes a rapid efflux of a further amount of 3.5–4 nmol/mg prot of Mg2+ , for a total amount of about 6 nmol/mg prot, corresponding to 80% of endogenous Mg2+ in RBM. The presence in the medium of ADP or CsA has no effect on this efflux. Spermine is also practically ineffective, demonstrating only very slight inhibition. It should be emphasized that, in the absence of Ca2+ , addition of GT does not induce Mg2+ efflux (see control curve “-Ca2+ ”).
Fig. 2. Collapse of membrane potential induced by gliotoxin. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 2 M TPP+ for Ψ measurement. Then 25 M Ca2+ and 100 M gliotoxin (GT) were added where indicated. When present, 1 mM ADP, 1 M CsA, E: electrode potential. Five additional experiments exhibited same trend.
762
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
Fig. 3. Spermine (A) and Mg2+ (B) protect against and restore collapse of membrane potential induced by gliotoxin. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 2 M TPP+ for Ψ measurement. Then 25 M Ca2+ , 100 M gliotoxin (GT), and 500 M ADP were added where indicated. When present or added, 100 M spermine, and 1 mM Mg2+ . Five additional experiments exhibited same trend.
Fig. 4. Effect of gliotoxin on endogenous Ca2+ (A) and Mg2+ (B). RBM (1 mg protein per ml) were incubated in standard medium supplemented with 25 M Ca2+ in conditions described in Section 2. GT was added at 100 M were indicated. When present, 1 mM ADP, 1 M CsA, 100 M spermine, and 1 mM Mg2+ . Experiments were performed five times with comparable results. The figure reports two controls, one performed in the absence of GT and presence of Ca2+ (-GT), as in the other figures, and another in the absence of Ca2+ and presence of GT (-Ca2+ ).
All the above results are reproduced when the standard sucrose medium is substituted with a potassium-based medium.
4. Discussion The results presented here show that treatment of RBM with 100 M GT results in matrix swelling of low amplitude,
Ψ collapse, and release of pre-accumulated calcium and endogenous Mg2+ . All these events except Mg2+ release are partially inhibited by CsA, a ligand of mitochondrial cyclophilin and ADP, whereas Mg2+ and spermine completely preserve mitochondrial integrity and bioenergetic functions. Partial inhibition by CsA is most probably due to the small amount of cyclophilin present in RBM, as previously observed (Kristal and Dubinsky, 1997). The explanation of the reduced effect of ADP is more complex. Adenine
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
nucleotide translocase (AdNT), the main protein involved in the MPT (Halestrap and Brennerb, 2003), presents different tissue-specific isoforms (Walker and Runswick, 1993), by which pore opening can be modulated in different ways in each type of cell. In brain, the reduced effect of ADP is probably due to decreased affinity of its GT-induced binding to AdNT. This would not be observable in other tissues. All these experimental observations lead to the conclusion that GT behaves as a typical MPT inducer. However, this result is not surprising, because other reports of the effects of GT on cells indicate that its possible mechanism of action is through opening of the MPT pore. In particular, it has been reported that GT induces apoptosis of activated human hepatic stellate cells, with induction of MPT (Kweon et al., 2003), although no report mentions the direct action of GT on isolated mitochondria. The present study demonstrates that, in isolated brain mitochondria, GT directly targets these organelles and that it is directly responsible for MPT induction. The main result of this report is that GT may specifically activate a Mg2+ efflux system from RBM while mitochondrial integrity is preserved (high membrane potential, no swelling, retention of other ions). It is well documented that the concentration of Mg2+ in cells undergoes rapid changes following a variety of hormonal and non-hormonal stimuli, and that it controls a variety of metabolic pathways (Romani and Scarpa, 2000; Wolf et al., 2003). In particular, a decrease in matrix Mg2+ has direct effects on mitochondrial respiration by stimulating succinate and glutamate dehydrogenase (Panov and Scarpa, 1996). Mitochondria may both take up and extrude Mg2+ by respiratory-dependent reactions, but the actual pathways for uptake and release remain obscure. While uptake of Mg2+ is due to an electrophoretic response to the Ψ component of the electrochemical gradient, efflux seems to depend on the pH component (for reviews, see Jung and Brierley, 1994). Scarpa’s group reports that cAMP is a messenger for major mobilization of Mg2+ in hepatocytes and that the major targets for cAMP effect are mitochondria. This effect is specific, because it is not reproduced by other nucleotides. Evidence is presented suggesting that the AdNT is the target of cAMP-dependent Mg2+ efflux and that cAMP interferes with the normal function of translocase (Romani et al., 1991). Our report presents convincing evidence of a new efflux pathway for Mg2+ induced specifically by GT. GT has the potential both to form hydrogen peroxide and to act as a thiol-modifying reagent. These peculiarities permit it to behave as a MPT inducer and a promoter of Mg2+ efflux in mitochondria maintained in an energized state by spermine. Our hypothesis is that the Mg2+ efflux system has a double regulation site: critical thiol(s) binding GT, and a calcium binding site. The latter assumption is supported by the fact that, in the absence of Ca2+ , GT does not induce Mg2+ efflux from RBM, as it is unable to induce MPT (data not shown).
763
Very recent results have shown that some pro-oxidants, including gliotoxin, induce an efflux of Mg2+ from energized liver mitochondria, the mechanism of which resembles that reported in this study.1 This observation means that it may be considered as a general phenomenon occurring in mitochondria, termed “low conductance state of MPT” (Ichas et al., 1997). However, it should be emphasized that the observation is aspecific regarding the functions of brain mitochondria. Mobilization of mitochondrial Mg2+ towards the external compartment, with a consequent rise in cytoplasmic Mg2+ level, can indeed serve “in vivo” as a regulator of channel-related neuronal events (Shi et al., 2002). In particular, high intracellular Mg2+ levels act as endogenous regulators of the N-methyl-d-aspartate (NMDA) (Zhu and Auerbach, 2001) and Ca2+ channels (Yamoaka et al., 2002). As NMDA and Ca2+ channels are implicated in neuronal damage (Wolf et al., 2003), mobilization of Mg2+ from energized mitochondria by gliotoxin may exert a protective action against neurodamaging effects. In this regard, it has been reported that high levels of Mg2+ , by blocking the NMDA receptor, ameliorate post-trauma neuronal damage (Zhang et al., 1996).
Acknowledgements The authors are grateful to Mario Mancon for skilled technical assistance.
References Crompton, M., Virgji, S., Doyle, V., Johnson, N., Ward, J.M., 1999. The mitochondrial permeability transition pore. Biochem. Soc. Symp. 66, 167–179. Gornall, A.G., Bardawill, C.J., David, M.M., 1949. Determination of serum proteins by means of the biuret method. J. Biol. Chem. 177, 751–766. Halestrap, A.P., Brennerb, C., 2003. The adenine nucleotide translocase: a central component of the mitochondrial permeability transition pore and key player in cell death. Curr. Med. Chem. 10, 1507–1525. Ichas, F., Jouaville, L.S., Mazat, J.P., 1997. Mitochondria are excitable organelles capable of generating and conveying electrical and calcium signals. Cell 89, 1145–1153. Jensen, B.D., Gunter, T.E., 1984. The use of tetraphenylphosphonium (TPP+ ) to measure membrane potential in mitochondria: membrane binding and respiratory effects. Biophys. J. 45, 92. Jensen, B.D., Gunter, K.K., Gunter, T.E., 1986. The efficiencies of the component steps of oxidative phosphorylation. II. Experimental determination of the efficiencies in mitochondria and examination of the equivalence of membrane potential and pH gradient in phosphorylation. Arch. Biochem. Biophys. 248, 305–323. Jung, D.W., Brierley, G.P., 1994. Magnesium transport by mitochondria. J. Bioenerg. Biomembr. 26, 527–535. Kamo, N., Muratsugu, M., Hongoh, R., Kobatake, Y., 1979. Membrane potential of mitochondria measured with an electrode sensitive to tetraphenylphosphonium and relationship between proton 1
M. Salvi and A. Toninello (unpublished experiments).
764
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
electrochemical potential and phosphorylation potential in steady state. J. Membr. Biol. 49, 105–121. Karadjov, J.S., Kudzina, L.Y., Zinchenko, V.P., 1986. TPP+ inhibits Na+ -stimulated Ca2+ efflux from brain mitochondria. Cell Calcium 7, 115–119. Kim, J.S., He, L., Lemasters, J.J., 2003. Mitochondrial permeability transition: a common pathway to necrosis and apoptosis. Biochem. Biophys. Res. Commun. 304, 463–470. Kristal, B., Dubinsky, J.M., 1997. Mitochondrial permeability transition in the central nervous system: induction by calcium cycling-dependent and -independent pathways. J. Neurochem. 69, 524–538. Kristian, T., Weatherby, T.M., Bates, T.E., Fiskum, G., 2002. Heterogeneity of the calcium-induced permeability transition in isolated non-synaptic brain mitochondria. J. Neurochem. 83, 1297–1308. Kweon, Y.O., Paik, Y.H., Schnabl, B., Qian, T., Lemasters, J.J., Brenner, D.A., 2003. Gliotoxin-mediated apoptosis of activated human hepatic stellate cells. J. Hepatol. 39, 38–46. Maciel, E.N., Vercesi, A.E., Casthilho, R.F., 2001. Oxidative stress in Ca2+ -induced membrane permeability transition in brain mitochondria. J. Neurochem. 79, 1237–1245. Mason, J.W., Kidd, J.G., 1951. Effects of gliotoxin and other sulphur containing compounds on tumor cells in vitro, with observations on mechanisms of action of gliotoxin. J. Immunol. 66, 99–106. Middleton, M.C., 1974. The involvement of the disulphide group of sporidesmin in the action of the toxin on the swelling and respiration of liver mitochondria. Biochem. Pharmacol. 23, 811–820. Nicholls, D.G., 1978. Calcium transport and proton electrochemical potential gradient in mitochondria from guinea-pig cerebral cortex and rat heart. Biochem. J. 170, 511–522. Panov, A., Scarpa, A., 1996. Mg2+ control of respiration in isolated rat liver mitochondria. Biochemistry 35, 12849–12856. Romani, A., Dowell, E., Scarpa, A., 1991. Cyclic AMP-induced Mg2+ release from rat liver hepatocytes, permeabilized hepatocytes, and isolated mitochondria. J. Biol. Chem. 266, 24376–24384.
Romani, A.M., Scarpa, A., 2000. Regulation of cellular magnesium. Front. Biosci. 5, 720–734. Salvi, M., Brunati, A.M., Bordin, L., La Rocca, N., Clari, G., Toninello, A., 2003. Characterization and location of Src-dependent tyrosine phosphorylation in rat brain mitochondria. Biochim. Biophys. Acta 1589, 181–195. Schweizer, M., Richter, C., 1994. Gliotoxin stimulates release from intact rat liver mitochondria. Biochemistry 33, 13401–13405. Shi, J., Krishnamoorthy, G., Yang, Y., Hu, L., Chaturvedi, N., Harilal, D., Quing, J., Cui, J., 2002. Mechanism of magnesium activation of calcium-activated potassium channels. Nature 418, 876–880. Tassani, V., Biban, C., Toninello, A., Siliprandi, D., 1995. Inhibition of mitochondrial permeability transition by polyamines and magnesium: importance of the number and distribution of electric charges. Biochem. Biophys. Res. Commun. 207, 661–667. Walker, J.E., Runswick, M.J., 1993. The mitochondrial transport protein superfamily. J. Bioenerg. Biomembr. 5, 435–446. Waring, P., Beaver, J., 1996. Gliotoxin and related epipolythiodioxopiperazines. Gen. Pharma. 27, 1311–1316. Wolf, F.I., Torsello, A., Fasanella, S., Cittadini, A., 2003. Cell physiology of magnesium. Mol. Aspects Med. 24, 11–26. Yamoaka, K., Yuki, T., Kawase, K., Munemori, M., Seyama, I., 2002. Temperature-sensitive intracellular Mg2+ block of L-type Ca-channels in cardiac myocytes. Am. J. Physiol. Heart Circ. Physiol. 82, H1092– H1101. Zoratti, M., Szabo, I., 1995. The mitochondrial permeability transition. Biochim. Biophys. Acta 1241, 139–176. Zhang, L., Rzigalinski, B.A., Ellins, E.F., Satin, L.S., 1996. Reduction of voltage-dependent Mg2+ blockade of NMDA current in mechanically injured neurons. Science 274, 1921–1923. Zhu, Y., Auerbach, A., 2001. Na+ occupancy and Mg2+ block of the N-methyl-D-aspartate receptor channel. J. Gen. Physiol. 117, 275– 286.
Gliotoxin induces Mg2+ efflux from intact brain mitochondria Mauro Salvi, Aleksandra Bozac, Antonio Toninello∗ Dipartimento di Chimica Biologica, Universita’ di Padova and Istituto di Neuroscienze del C.N.R., Unita’ per lo Studio delle Biomembrane, Via G. Colombo 3, 35121 Padua, Italy Received 4 September 2003; received in revised form 28 November 2003; accepted 14 January 2004 Available online 8 April 2004
Abstract Gliotoxin (GT) is a hydrophobic fungal metabolite of the epipolythiodioxopiperazine group which reacts with membrane thiols. When added to a suspension of energized brain mitochondria, it induces matrix swelling of low amplitude, collapse of membrane potential (Ψ ), and efflux of endogenous cations such as Ca2+ and Mg2+ , typical events of mitochondrial permeability transition (MPT) induction. These effects are due to opening of the membrane transition pore. The addition of cyclosporin A (CsA) or ADP slightly reduces membrane potential collapse, matrix swelling and Ca2+ efflux; Mg2+ efflux is not affected at all. The presence of exogenous Mg2+ or spermine completely preserve mitochondria against Ψ collapse, matrix swelling and Ca2+ release. Instead, Mg2+ efflux is only slightly affected by spermine. Our results demonstrate that, besides inducing MPT, gliotoxin activates a specific Mg2+ efflux system from brain mitochondria. © 2004 Elsevier Ltd. All rights reserved. Keywords: Gliotoxin; Rat brain mitochondria; Mg2+ ; Mitochondrial permeability transition
1. Introduction Gliotoxin (GT) is a hydrophobic fungal metabolite of the epipolythiodioxopiperazine family. It has a quinoid moiety and a disulfide bridge across the piperazine ring. It is capable of forming mixed disulphides with free thiol groups, and several reports suggest that the main toxic effects of GT in cells are due to its interaction with enzymes by mixed disulphide formation (Mason and Kidd, 1951; Middleton, 1974). Another property of GT is its ability to trigger a redox cycle in the presence of an appropriate reducing agent, generating reactive oxygen species (for a review, see Waring and Beaver, 1996). GT induces specific Ca2+ release from intact rat liver mitochondria. This stimulation is suggested to be due to the reaction of GT with thiols critically involved in the specific Ca2+ release pathway (Schweizer and Richter, 1994). GT also induces apoptotic death in numerous cell types (Waring and Beaver, 1996), and one report seems to demonstrate the involvement of the MPT pore (Kweon et al., 2003). The MPT pore is a regulated membrane channel which alAbbreviations: AdNT, adenine nucleotide translocase; MPT, mitochondrial permeability transition; RBM, rat brain mitochondria; GT, gliotoxin ∗ Corresponding author. Tel.: +39-049-827-6134; fax: +39-049-827-6133. E-mail address: [email protected] (A. Toninello). 0197-0186/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuint.2004.01.001
lows free diffusion of all solutes of <1500 Da and which is associated with mitochondrial swelling, membrane potential collapse, efflux of endogenous Mg2+ , rapid depletion of mitochondrial Ca2+ stores, and uncoupling of oxidative phosphorylation. The MPT is induced by supraphysiological levels of Ca2+ and by a large number of compounds, usually called inducers, and is selectively inhibited by the immunosuppressant cyclosporin A (CsA) (for a review, see Zoratti and Szabo, 1995; Crompton et al., 1999; Kim et al., 2003). The opening of the MPT pore can also be inhibited in liver and heart mitochondria by some physiological agents such as ADP, reduced pyridine nucleotides, and Mg2+ . Naturally occurring polyamines also have a similar effect, and spermine in particular has been demonstrated to be a strong inhibitory agent of this phenomenon (Tassani et al., 1995). It has been reported that MPT can also be induced in isolated rat brain mitochondria (RBM) (Kristal and Dubinsky, 1997; Maciel et al., 2001; Kristian et al., 2002). However, brain mitochondria exhibit unique MTP characteristics, including relative resistance to inhibition by CsA. Instead, Mg2+ and ADP strongly inhibit the phenomenon, to a level comparable with their efficacy in liver. It has also been found that dithiotreitol and N-ethylmaleimide are partial inhibitors, and ruthenium red almost completely blocks MPT induction (Kristal and Dubinsky, 1997; Kristian et al.,
760
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
2002). Increased reactive oxygen species are observed with the opening of the MPT pore in brain mitochondria induced by Ca2+ , and Ca2+ -induced brain mitochondrial MPT is significantly inhibited by catalase (Maciel et al., 2001). The aim of this investigation is to study the effect of GT on the permeability of the inner membrane of rat brain mitochondria, in order to highlight some important peculiarities of this type of organelle.
2. Experimental procedures 2.1. Mitochondrial preparations Rat brain mitochondria were purified by the Ficoll gradient method, according to Nicholls (1978), with some modifications (Salvi et al., 2003). Briefly, rat brain (cerebral cortex) was homogenized in isolation medium (320 mM sucrose, 5 mM Hepes, 0.5 mM EDTA, pH 7.4; 0.3% BSA was added during homogenisation and the first step of purification) and subjected to centrifugation (900 × g) for 5 min. The supernatant was centrifuged at 17,000×g for 10 min, to precipitate crude mitochondrial pellets. The pellets were resuspended in isolation medium plus 1 mM ATP and layered on top of a discontinuous gradient, composed of 2 ml of isolation medium containing 16% (w/v) Ficoll, 2 ml of isolation medium containing 14% (w/v) Ficoll, 3 ml of isolation medium containing 12% (w/v) Ficoll, and 3 ml of isolation medium containing 7% (w/v) Ficoll. The gradient was centrifuged for 30 min at 75,000×g. Mitochondrial pellets were suspended in isolation medium and centrifuged for 10 min at 800 × g. Again the pellets were suspended in isolation medium without EDTA. Protein content was measured by the biuret method with bovine serum albumin as a standard (Gornall et al., 1949). 2.2. Standard incubation procedures Mitochondria (1 mg protein per ml) were incubated in a water-jacketed cell at 20 ◦ C under continuous stirring. The standard medium contained 200 mM sucrose, 10 mM Hepes (pH 7.4), 5 mM succinate and 1.25 M rotenone. Variations and/or other additions are given with each experiment. 2.3. Mitochondrial swelling Mitochondrial swelling was determined by the change in the apparent absorbance of mitochondrial suspensions at 540 nm in a 3-ml cuvette, with a Kontron Uvikon mod. 922 spectrophotometer equipped with thermostatic control and magnetic stirrer. Mitochondria were suspended as indicated above and, upon stabilisation of the absorbance trace, swelling was assessed after additions of other compounds as described in the figure legends.
2.4. Membrane potential measurements The membrane potential was calculated on the basis of the movements of the lipid-soluble cation tetraphenylphosphonium (TPP+ ) through the inner membrane with a specific electrode for TPP+ prepared according to published procedures (Kamo et al., 1979) and an Ag/AgCl reference electrode. TPP+ was included at a concentration of 2 M in order to achieve high sensitivity in measurements and to avoid toxic effects on the proton ATPase and calcium flux (Jensen and Gunter, 1984; Karadjov et al., 1986). The measurement was started by adding mitochondria to the incubation medium supplemented with other compounds, as indicated in the appropriate legends, after TPP+ calibration. The membrane potential was calculated from the Nernst equation and corrected for non-specific intramitochondrial binding of TPP+ using the equation: Ψ = (Ψ electrode − 66.16 mV)/0.92 (Jensen et al., 1986). 2.5. Cation efflux Ca2+ and Mg2+ release was estimated by atomic absorption spectroscopy of the supernatant fraction. Mitochondria (1 mg/ml) were incubated in standard medium containing other compounds, as indicated in the specific figure legends. The suspension was stirred continuously and, at intervals of time, 1-ml portions of the suspension were withdrawn and centrifuged for 2 min in an Eppendorf bench centrifuge (model 5415C) at 15,000 rpm, this speed being sufficient to sediment at least 98% of the mitochondria. The supernatant fluids were removed and their Ca2+ and Mg2+ contents determined with a Perkin-Elmer 1100B spectrometer.
3. Results The results (Fig. 1) show that addition of 25 M Ca2+ to brain mitochondria incubated in standard medium does not induce any change in the apparent absorbance of the suspension (see trace control). Addition of 100 M GT, 3 min after that of Ca2+ , induces an absorbance decrease of about 0.1 units, indicative of mitochondrial swelling. When compared with that obtainable in liver mitochondria, a decrease in absorbance of about 1 unit (data not shown) means that the swelling of RBM may be considered of low amplitude. The presence of CsA or ADP, alone or in combination, only partially reduces the extent of the swelling, whereas the polyamine spermine not only blocks the phenomenon but induces significant shrinkage; Mg2+ shows strong inhibition (Fig. 1). Fig. 2 shows the effects of GT on the electrical potential (Ψ ) of Ca2+ -loaded mitochondria. As shown in this figure, our brain mitochondrial preparations, when incubated in standard medium, have a Ψ value of about 160 mV. After Ψ reaches steady state, addition of 25 M Ca2+ leads to
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
Fig. 1. Mitochondrial swelling induced by gliotoxin in isolated RBM. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 25 M Ca2+ in conditions described in Section 2. Then 100 M GT, 1 mM ADP, 1 M CsA, 100 M spermine and 1 mM Mg2+ were added. Dashed line: RBM incubated in absence of GT (control). Downward deflection indicates mitochondrial swelling. Experiments were performed five times with comparable results.
sudden transient depolarization of the membrane (followed by prompt repolarization) (trace control). When the Ca2+ pulse is 25 M, the new steady state is virtually identical to that preceding Ca2+ loading. With high Ca2+ pulses, the new steady state gradually falls (data not shown). Addition of GT induces an almost complete drop in Ψ , which is partially prevented by CsA or ADP. Addition of ADP, quickly followed by CsA after the drop in Ψ , is completely ineffective. If ADP and CsA fail to completely prevent or to restore Ψ collapse by GT (see Fig. 2), the presence of spermine in the incubation medium, as demonstrated by the results of Fig. 3A, at 0.1 mM concentration, completely prevents a Ψ drop by GT in the presence of Ca2+ . Indeed, if the polyamine is added after Ψ collapse, sudden repo-
761
larization to normal values is observed. As reported in Fig. 3B, similar results are obtained if spermine is substituted by 1 mM Mg2+ , since this cation can protect against Ψ collapse and restore it after collapse. Addition of ADP after repolarisation by spermine (Fig. 3A) or Mg2+ (Fig. 3B) also induces the characteristic state 3/state 4 Ψ transition, indicative of the maintenance of the bioenergetic capacity of mitochondria. Fig. 4A shows the effect of GT on Ca2+ movements. When RBM are suspended in the incubation medium containing 25 M Ca2+ , they accumulate the cation and retain it until anaerobiosis (see trace control). Addition of GT after 3 min induces gradual release of Ca2+ , which is complete after 20 min. The presence of ADP or CsA in the incubation medium provokes reduced inhibition of Ca2+ efflux; spermine or Mg2+ completely block this GT-induced release. Generally, when mitochondria are suspended in an incubation medium, they instantaneously release an aliquot of Mg2+ from the intermembrane space. In this case, RBM, as shown in Fig. 4B, release about 3 nmol/mg of Mg2+ protein, in all experimental conditions. Addition of Ca2+ induces a slight gradual Mg2+ efflux (trace control) which is also observable in the presence of ADP, CsA or spermine. This efflux of about 0.5 nmol/mg is very probably due to the interaction of Ca2+ during its cycling across the inner membrane with membrane binding sites having the same affinity for both cations. Subsequent addition of GT provokes a rapid efflux of a further amount of 3.5–4 nmol/mg prot of Mg2+ , for a total amount of about 6 nmol/mg prot, corresponding to 80% of endogenous Mg2+ in RBM. The presence in the medium of ADP or CsA has no effect on this efflux. Spermine is also practically ineffective, demonstrating only very slight inhibition. It should be emphasized that, in the absence of Ca2+ , addition of GT does not induce Mg2+ efflux (see control curve “-Ca2+ ”).
Fig. 2. Collapse of membrane potential induced by gliotoxin. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 2 M TPP+ for Ψ measurement. Then 25 M Ca2+ and 100 M gliotoxin (GT) were added where indicated. When present, 1 mM ADP, 1 M CsA, E: electrode potential. Five additional experiments exhibited same trend.
762
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
Fig. 3. Spermine (A) and Mg2+ (B) protect against and restore collapse of membrane potential induced by gliotoxin. RBM (1 mg protein per ml) were incubated in standard medium supplemented with 2 M TPP+ for Ψ measurement. Then 25 M Ca2+ , 100 M gliotoxin (GT), and 500 M ADP were added where indicated. When present or added, 100 M spermine, and 1 mM Mg2+ . Five additional experiments exhibited same trend.
Fig. 4. Effect of gliotoxin on endogenous Ca2+ (A) and Mg2+ (B). RBM (1 mg protein per ml) were incubated in standard medium supplemented with 25 M Ca2+ in conditions described in Section 2. GT was added at 100 M were indicated. When present, 1 mM ADP, 1 M CsA, 100 M spermine, and 1 mM Mg2+ . Experiments were performed five times with comparable results. The figure reports two controls, one performed in the absence of GT and presence of Ca2+ (-GT), as in the other figures, and another in the absence of Ca2+ and presence of GT (-Ca2+ ).
All the above results are reproduced when the standard sucrose medium is substituted with a potassium-based medium.
4. Discussion The results presented here show that treatment of RBM with 100 M GT results in matrix swelling of low amplitude,
Ψ collapse, and release of pre-accumulated calcium and endogenous Mg2+ . All these events except Mg2+ release are partially inhibited by CsA, a ligand of mitochondrial cyclophilin and ADP, whereas Mg2+ and spermine completely preserve mitochondrial integrity and bioenergetic functions. Partial inhibition by CsA is most probably due to the small amount of cyclophilin present in RBM, as previously observed (Kristal and Dubinsky, 1997). The explanation of the reduced effect of ADP is more complex. Adenine
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
nucleotide translocase (AdNT), the main protein involved in the MPT (Halestrap and Brennerb, 2003), presents different tissue-specific isoforms (Walker and Runswick, 1993), by which pore opening can be modulated in different ways in each type of cell. In brain, the reduced effect of ADP is probably due to decreased affinity of its GT-induced binding to AdNT. This would not be observable in other tissues. All these experimental observations lead to the conclusion that GT behaves as a typical MPT inducer. However, this result is not surprising, because other reports of the effects of GT on cells indicate that its possible mechanism of action is through opening of the MPT pore. In particular, it has been reported that GT induces apoptosis of activated human hepatic stellate cells, with induction of MPT (Kweon et al., 2003), although no report mentions the direct action of GT on isolated mitochondria. The present study demonstrates that, in isolated brain mitochondria, GT directly targets these organelles and that it is directly responsible for MPT induction. The main result of this report is that GT may specifically activate a Mg2+ efflux system from RBM while mitochondrial integrity is preserved (high membrane potential, no swelling, retention of other ions). It is well documented that the concentration of Mg2+ in cells undergoes rapid changes following a variety of hormonal and non-hormonal stimuli, and that it controls a variety of metabolic pathways (Romani and Scarpa, 2000; Wolf et al., 2003). In particular, a decrease in matrix Mg2+ has direct effects on mitochondrial respiration by stimulating succinate and glutamate dehydrogenase (Panov and Scarpa, 1996). Mitochondria may both take up and extrude Mg2+ by respiratory-dependent reactions, but the actual pathways for uptake and release remain obscure. While uptake of Mg2+ is due to an electrophoretic response to the Ψ component of the electrochemical gradient, efflux seems to depend on the pH component (for reviews, see Jung and Brierley, 1994). Scarpa’s group reports that cAMP is a messenger for major mobilization of Mg2+ in hepatocytes and that the major targets for cAMP effect are mitochondria. This effect is specific, because it is not reproduced by other nucleotides. Evidence is presented suggesting that the AdNT is the target of cAMP-dependent Mg2+ efflux and that cAMP interferes with the normal function of translocase (Romani et al., 1991). Our report presents convincing evidence of a new efflux pathway for Mg2+ induced specifically by GT. GT has the potential both to form hydrogen peroxide and to act as a thiol-modifying reagent. These peculiarities permit it to behave as a MPT inducer and a promoter of Mg2+ efflux in mitochondria maintained in an energized state by spermine. Our hypothesis is that the Mg2+ efflux system has a double regulation site: critical thiol(s) binding GT, and a calcium binding site. The latter assumption is supported by the fact that, in the absence of Ca2+ , GT does not induce Mg2+ efflux from RBM, as it is unable to induce MPT (data not shown).
763
Very recent results have shown that some pro-oxidants, including gliotoxin, induce an efflux of Mg2+ from energized liver mitochondria, the mechanism of which resembles that reported in this study.1 This observation means that it may be considered as a general phenomenon occurring in mitochondria, termed “low conductance state of MPT” (Ichas et al., 1997). However, it should be emphasized that the observation is aspecific regarding the functions of brain mitochondria. Mobilization of mitochondrial Mg2+ towards the external compartment, with a consequent rise in cytoplasmic Mg2+ level, can indeed serve “in vivo” as a regulator of channel-related neuronal events (Shi et al., 2002). In particular, high intracellular Mg2+ levels act as endogenous regulators of the N-methyl-d-aspartate (NMDA) (Zhu and Auerbach, 2001) and Ca2+ channels (Yamoaka et al., 2002). As NMDA and Ca2+ channels are implicated in neuronal damage (Wolf et al., 2003), mobilization of Mg2+ from energized mitochondria by gliotoxin may exert a protective action against neurodamaging effects. In this regard, it has been reported that high levels of Mg2+ , by blocking the NMDA receptor, ameliorate post-trauma neuronal damage (Zhang et al., 1996).
Acknowledgements The authors are grateful to Mario Mancon for skilled technical assistance.
References Crompton, M., Virgji, S., Doyle, V., Johnson, N., Ward, J.M., 1999. The mitochondrial permeability transition pore. Biochem. Soc. Symp. 66, 167–179. Gornall, A.G., Bardawill, C.J., David, M.M., 1949. Determination of serum proteins by means of the biuret method. J. Biol. Chem. 177, 751–766. Halestrap, A.P., Brennerb, C., 2003. The adenine nucleotide translocase: a central component of the mitochondrial permeability transition pore and key player in cell death. Curr. Med. Chem. 10, 1507–1525. Ichas, F., Jouaville, L.S., Mazat, J.P., 1997. Mitochondria are excitable organelles capable of generating and conveying electrical and calcium signals. Cell 89, 1145–1153. Jensen, B.D., Gunter, T.E., 1984. The use of tetraphenylphosphonium (TPP+ ) to measure membrane potential in mitochondria: membrane binding and respiratory effects. Biophys. J. 45, 92. Jensen, B.D., Gunter, K.K., Gunter, T.E., 1986. The efficiencies of the component steps of oxidative phosphorylation. II. Experimental determination of the efficiencies in mitochondria and examination of the equivalence of membrane potential and pH gradient in phosphorylation. Arch. Biochem. Biophys. 248, 305–323. Jung, D.W., Brierley, G.P., 1994. Magnesium transport by mitochondria. J. Bioenerg. Biomembr. 26, 527–535. Kamo, N., Muratsugu, M., Hongoh, R., Kobatake, Y., 1979. Membrane potential of mitochondria measured with an electrode sensitive to tetraphenylphosphonium and relationship between proton 1
M. Salvi and A. Toninello (unpublished experiments).
764
M. Salvi et al. / Neurochemistry International 45 (2004) 759–764
electrochemical potential and phosphorylation potential in steady state. J. Membr. Biol. 49, 105–121. Karadjov, J.S., Kudzina, L.Y., Zinchenko, V.P., 1986. TPP+ inhibits Na+ -stimulated Ca2+ efflux from brain mitochondria. Cell Calcium 7, 115–119. Kim, J.S., He, L., Lemasters, J.J., 2003. Mitochondrial permeability transition: a common pathway to necrosis and apoptosis. Biochem. Biophys. Res. Commun. 304, 463–470. Kristal, B., Dubinsky, J.M., 1997. Mitochondrial permeability transition in the central nervous system: induction by calcium cycling-dependent and -independent pathways. J. Neurochem. 69, 524–538. Kristian, T., Weatherby, T.M., Bates, T.E., Fiskum, G., 2002. Heterogeneity of the calcium-induced permeability transition in isolated non-synaptic brain mitochondria. J. Neurochem. 83, 1297–1308. Kweon, Y.O., Paik, Y.H., Schnabl, B., Qian, T., Lemasters, J.J., Brenner, D.A., 2003. Gliotoxin-mediated apoptosis of activated human hepatic stellate cells. J. Hepatol. 39, 38–46. Maciel, E.N., Vercesi, A.E., Casthilho, R.F., 2001. Oxidative stress in Ca2+ -induced membrane permeability transition in brain mitochondria. J. Neurochem. 79, 1237–1245. Mason, J.W., Kidd, J.G., 1951. Effects of gliotoxin and other sulphur containing compounds on tumor cells in vitro, with observations on mechanisms of action of gliotoxin. J. Immunol. 66, 99–106. Middleton, M.C., 1974. The involvement of the disulphide group of sporidesmin in the action of the toxin on the swelling and respiration of liver mitochondria. Biochem. Pharmacol. 23, 811–820. Nicholls, D.G., 1978. Calcium transport and proton electrochemical potential gradient in mitochondria from guinea-pig cerebral cortex and rat heart. Biochem. J. 170, 511–522. Panov, A., Scarpa, A., 1996. Mg2+ control of respiration in isolated rat liver mitochondria. Biochemistry 35, 12849–12856. Romani, A., Dowell, E., Scarpa, A., 1991. Cyclic AMP-induced Mg2+ release from rat liver hepatocytes, permeabilized hepatocytes, and isolated mitochondria. J. Biol. Chem. 266, 24376–24384.
Romani, A.M., Scarpa, A., 2000. Regulation of cellular magnesium. Front. Biosci. 5, 720–734. Salvi, M., Brunati, A.M., Bordin, L., La Rocca, N., Clari, G., Toninello, A., 2003. Characterization and location of Src-dependent tyrosine phosphorylation in rat brain mitochondria. Biochim. Biophys. Acta 1589, 181–195. Schweizer, M., Richter, C., 1994. Gliotoxin stimulates release from intact rat liver mitochondria. Biochemistry 33, 13401–13405. Shi, J., Krishnamoorthy, G., Yang, Y., Hu, L., Chaturvedi, N., Harilal, D., Quing, J., Cui, J., 2002. Mechanism of magnesium activation of calcium-activated potassium channels. Nature 418, 876–880. Tassani, V., Biban, C., Toninello, A., Siliprandi, D., 1995. Inhibition of mitochondrial permeability transition by polyamines and magnesium: importance of the number and distribution of electric charges. Biochem. Biophys. Res. Commun. 207, 661–667. Walker, J.E., Runswick, M.J., 1993. The mitochondrial transport protein superfamily. J. Bioenerg. Biomembr. 5, 435–446. Waring, P., Beaver, J., 1996. Gliotoxin and related epipolythiodioxopiperazines. Gen. Pharma. 27, 1311–1316. Wolf, F.I., Torsello, A., Fasanella, S., Cittadini, A., 2003. Cell physiology of magnesium. Mol. Aspects Med. 24, 11–26. Yamoaka, K., Yuki, T., Kawase, K., Munemori, M., Seyama, I., 2002. Temperature-sensitive intracellular Mg2+ block of L-type Ca-channels in cardiac myocytes. Am. J. Physiol. Heart Circ. Physiol. 82, H1092– H1101. Zoratti, M., Szabo, I., 1995. The mitochondrial permeability transition. Biochim. Biophys. Acta 1241, 139–176. Zhang, L., Rzigalinski, B.A., Ellins, E.F., Satin, L.S., 1996. Reduction of voltage-dependent Mg2+ blockade of NMDA current in mechanically injured neurons. Science 274, 1921–1923. Zhu, Y., Auerbach, A., 2001. Na+ occupancy and Mg2+ block of the N-methyl-D-aspartate receptor channel. J. Gen. Physiol. 117, 275– 286.