Eur. J. Biochem. 264, 687±701 (1999) q FEBS 1999
R E V I E W A RT I C L E
Mitochondria and cell death Mechanistic aspects and methodological issues Paolo Bernardi1, Luca Scorrano1, Raffaele Colonna1, Valeria Petronilli1 and Fabio Di Lisa2 CNR Unit for the Study of Biomembranes and Departments of 1Biomedical Sciences and 2Biological Chemistry, University of Padova, Padova, Italy
Mitochondria are involved in cell death for reasons that go beyond ATP supply. A recent advance has been the discovery that mitochondria contain and release proteins that are involved in the apoptotic cascade, like cytochrome c and apoptosis inducing factor. The involvement of mitochondria in cell death, and its being cause or consequence, remain issues that are extremely complex to address in situ. The response of mitochondria may critically depend on the type of stimulus, on its intensity, and on the specific mitochondrial function that has been primarily perturbed. On the other hand, the outcome also depends on the integration of mitochondrial responses that cannot be dissected easily. Here, we try to identify the mechanistic aspects of mitochondrial involvement in cell death as can be derived from our current understanding of mitochondrial physiology, with special emphasis on the permeability transition and its consequences (like onset of swelling, cytochrome c release and respiratory inhibition); and to critically evaluate methods that are widely used to monitor mitochondrial function in situ. Keywords: mitochondria; cell death; apoptosis; necrosis; permeability transition; calcium; caspases; fluorescence; probes; cytochrome c. Mitochondria are central to the life of eukaryotic cells. In recent years, it has become clear that mitochondria also play a key role in the pathways to cell death [1±8]. This role of mitochondria cannot be explained by a mere `loss of function' resulting in an energetic deficit. Rather, it is increasingly recognized as an `active' process mediated by regulated effector mechanisms in a wide variety of conditions. A large body of observations from the fields of immunology, toxicology, oncology, neurology and cardiology testifies about the centrality of this issue. However, the general reader is probably puzzled by the growing number of conflicting reports, and by the diverging conclusions reached by different groups when mechanistic issues are addressed and/ or cause-relationship effects are discussed. A clear example is represented by the question of cytochrome c release, and of how this event relates to mitochondrial respiration, maintenance of the mitochondrial membrane potential difference (Dcm) and volume homeostasis. In our opinion, part of the problem resides in the intrinsic difficulties one faces when measuring `mitochondrial function' in intact cells, an issue that is not always adequately appreciated. Mitochondrial responses in vivo may be extremely Correspondence to P. Bernardi/F. Di Lisa, CNR Unit for the Study of Biomembranes, Viale Giuseppe Colombo 3, I-35121 Padova, Italy. Fax: + 39 049827 6361, E-mail: [email protected]
or [email protected]
Abbreviations: Dcm, mitochondrial membrane potential difference; m Ä H, mitochondrial proton electrochemical gradient; Dcp, plasma membrane potential difference; MTP, mitochondrial permeability transition pore; PT, permeability transition; Cs, cyclosporin; PN, pyridine nucleotides; ANT, adenine nucleotide translocase; AIF, apoptosis inducing factor; [Ca2+]m, mitochondrial matrix free Ca2+ concentration; [Ca2+]c, cytosolic free Ca2+ concentration; DiOC6(3), 3,3 0 -dihexyloxacarbocyanine iodide; CMTMRos, chloromethyltetramethyl rosamine; FCCP, carbonylcyanide-p-trifluoromethoxyphenyl hydrazone; MDR, multidrug resistance pump; JC-1, 5,5 0 ,6,6 0 -tetrachloro-1,1 0 ,3,3 0 -tetraethylbenzimidazolocarbocyanine iodide; TMRM, tetramethylrhodamine methyl ester; AM, acetoxymethyl. (Received 8 April 1999, accepted 19 July 1999)
hard to decipher, and a large discrepancy currently exists between our understanding of mitochondrial function in the isolated organelle and in the intact cell. However, an acceptable description of in vivo events must always take into account facts and concepts established in decades of in vitro studies. Also, it should be realized that discrepancies may arise from inapparent differences in the experimental conditions, but also from artifacts or misinterpretation of the in situ experiments. As should become clear later in the review, particular care should be exerted in studies with fluorescent probes. These are easily detected by cytofluorimetric and imaging techniques, but often have not been adequately characterized for phototoxicity and for directs effects on mitochondrial function that may preclude or seriously limit their use with living cells. Major goals of this review are to identify and discuss the mechanistic aspects of mitochondrial involvement in cell death as can be derived from our current understanding of mitochondrial physiology; and to critically evaluate the key issues emerged from recent studies, with the objective of enucleating points of controversy and of identifying outstanding problems. This review is not intended to provide an extensive coverage of the literature, which can be found in recent reviews [9±24].
THE MEMBRANE POTENTIAL, M I TO C H O N D R I A L C AT I O N T R A N S P O R T A N D M AT R I X S W E L L I N G The initial event of energy conservation is charge separation at the inner mitochondrial membrane. Electrons deriving from oxidation of substrates are funneled to oxygen through the redox carriers of the respiratory chain, and this process is coupled to H+ ejection on the redox pumps at complexes I, III and IV. As the passive permeability to H+ and to cations and anions is generally low, H+ ejection results in the establishment of a H+ electrochemical gradient, mÄH [25,26]. The magnitude
688 P. Bernardi et al. (Eur. J. Biochem. 264)
of the proton electrochemical gradient is about 2220 mV, and under physiological conditions most of the gradient is in the form of a Dcm . Monovalent cations The inside-negative Dcm represents the major driving force for mitochondrial cation accumulation. As the cytosolic concentrations of K+ and Na+ are about 150 and 5 mm, respectively, for a Dcm of 2180 mV equilibrium matrix [K+] and [Na+] should be 150 m and 5 m, respectively. This is never achieved for two main reasons: (a) the high permeability of the inner membrane to water, which prevents the buildup of a cation concentration gradient and therefore maintains the matrix isoosmolar relative to the intermembrane and extramitochondrial spaces despite cation uptake, a point that is not always appreciated ; because of this, any net uptake of cation salts down the electrochemical gradient, no matter how slow, would eventually lead to swelling with inner membrane unfolding and outer membrane rupture; (b) the operation of the electroneutral H+-Na+ and H+-K+ exchangers, which catalyze cation efflux and thus prevent thermodynamic equilibration of the gradients; because of this, the cation accumulation ratio is kinetically modulated by the relative rates of uptake via the K+ and Na+ channels and efflux via the H+-Na+ and H+-K+ exchangers [29,30]. These fluxes are tightly regulated to achieve volume homeostasis without major energy drain. A detailed coverage of these transport pathways, which is beyond the scope of this review, can be found in . Calcium As cytosolic free [Ca2+] ([Ca2+]c) oscillates between about 0.1 and 1 mm, for a Dcm of 2180 mV equilibrium matrix free [Ca2+] ([Ca2+]m) should be 0.1±1.0 m, i.e. at least 100 000 to 1 000 000-fold higher than the actual values measured in isolated mitochondria and in intact cells . Also in this case displacement from thermodynamic equilibrium is due to the fact that Ca2+ distribution represents a steady state where electrophoretic Ca2+ uptake via the Ca2+ uniporter and/or the rapid uptake mode  is matched by Ca2+ efflux on two separate pathways, the `Na+-dependent Ca2+ efflux pathway' and the `Na+-independent Ca2+ efflux pathway' [34,35]. As the maximal rate of Ca2+ efflux via these pathways is much lower than that of Ca2+ uptake, mitochondria are exposed to the hazards of Ca2+ overload when [Ca2+]c rises steadily, or when the frequency of its oscillations increases. Due to the low [Ca2+]c and [Ca2+]m, and to the presence of Pi and other Ca2+-complexing anions and proteins in the matrix, Ca2+ uptake is unlikely to be a direct cause of mitochondrial swelling. In vitro at least, Ca2+-dependent swelling is rather due to opening of a high-conductance channel, the mitochondrial permeability transition pore (MTP) which is briefly covered below. At variance from swelling that depends on K+ uptake, swelling caused by MTP opening occurs in fully depolarized mitochondria, and it demands a concentration gradient of solute(s) between the intermembrane and matrix spaces.
THE PERMEABILITY TRANSITION The permeability transition (PT) is a sudden increase of the inner membrane permeability to solutes with molecular mass below approximately 1500 Da [36±38]. This phenomenon is most easily observed after the matrix accumulation of Ca2+, and it is widely believed to be caused by opening of a regulated
q FEBS 1999
channel. This channel, MTP, can be defined as a voltagedependent, cyclosporin (Cs) A-sensitive, high-conductance inner membrane channel. In the fully open state, the apparent pore diameter is about 3 nm  and the pore open±closed transitions are highly regulated by multiple effectors at sites that are summarized in Table 1. Many of these sites can be affected by conditions and mediators implied in a variety of models of cell death, and evidence exists to suggest that the PT may be an early event in committment to apoptosis, particularly in the immune system . Regulation of the MTP has recently been covered in detail elsewhere . Here, we will rather focus on its consequences because these are central to understanding its potential role in cell death. The only primary consequence of MTP opening in vitro is membrane depolarization, and common PT assays in intact cells are indeed based on the fluorescence changes of probes that are accumulated by mitochondria in reponse to the Dcm. The theoretical and practical aspects of the use of probes, and their limits, will be discussed in some detail below. What needs to be stressed here is that although MTP opening is always followed by depolarization, depolarization is not always caused by MTP opening. It is unfortunate that many otherwise excellent studies consider depolarization in situ as an unequivocal evidence that a PT has occurred, which can be an obvious source of misunderstandings when the experimental results are interpreted in mechanistic terms. Pore opening can have consequences on respiration that depend on the substrates being oxidized. With pyridine nucleotide (PN)-linked substrates MTP opening is followed by respiratory inhibition because matrix PN are lost, an experimental finding that preceded definition of the Ca2+-dependent permeability change as a PT . With complex II-linked substrates the PT is rather followed by uncoupling. The consequences of a PT on respiration in vivo (and on the related issue of production of reactive oxygen species) therefore depend on whether, and to what extent, PN depletion and subsequent hydrolysis by outer membrane glycohydrolase take place . Irrespective of whether respiration is inhibited or stimulated, collapse of the mÄH caused by the PT should curtail ATP synthesis as long as the pore is open. Together with increased hydrolysis by the mitochondrial ATPase, this should contribute to cellular ATP depletion (reviewed in ). Pore opening is followed by equilibration of ionic gradients and of species that have a molecular mass lower than about
Table 1. Modulators of the mitochondrial permeability transition. Open probability Control point
Voltage Matrix pH Surface potential Matrix Me2+ site External Me2+ site ANT Cyclophilin D Dithiols/glutathione Pyridine nucleotides Quinones
      [44,45] [46,47] [46,47] [48,49]
Depolarization Alkalinization More negative Ca2+ ± `c' conformation ± Oxidation Oxidation ±
Hyperpolarization Acidification More positive Mg2+, Sr2+, Mn2+ Ca2+, Mg2+ `m' conformation CsA Reduction Reduction Ubiquinone 0 Decylubiquinone
q FEBS 1999
1500 Da, and this represents the basis for the common swelling assays of MTP opening in vitro. It must be stressed, however, that swelling can be easily prevented by low concentrations of pore-impermeant solutes (such as 25 mm poly(ethylene glycol) 3400) ; and that for short open times, solute equilibration (and therefore swelling) may not occur at all [55±57]. Thus, whether swelling, outer membrane rupture and cytochrome c release necessarily follow a PT in vivo remains difficult to predict.
TWO MODES OF MITOCHONDRIAL SWELLING In summary, we can safely state that mitochondrial swelling is always an osmotic process that results from net solute and water diffusion towards the matrix. This may occur according to two basic mechanisms: (a) Electrophoretic (energy-dependent) uptake of monovalent cations, K+ in particular. This type of swelling occurs without loss of energy other than that spent for the process of cation accumulation. The mÄH is largely regenerated by respiration, the inner membrane permeability remains low, and the swollen mitochondria retain a high Dcm and a high level of coupling. Indeed, following energy-dependent swelling due to accumulation of K+ in the presence of valinomycin, mitochondria can utilize the K+ gradient to synthesize ATP . This type of swelling coincides with the `high energy' swelling defined by Azzone and Azzi in 1965 . (b) Passive diffusion of species down their concentration gradients following a decrease of the permeability barrier. This type of swelling can be caused by MTP opening, occurs in deenergized mitochondria (`low energy' swelling ), and is followed by repolarization upon pore closure and reenergization with ATP or respiration. Shrinkage and volume recovery are only possible when the species responsible for swelling can be extruded by endogenous transport pathways. This is the case for K+ and Na+ but not for sugars like sucrose, which is the standard osmotic support for studies with isolated mitochondria, a choice that contributed to convey the misleading impression that pore-mediated swelling is an irreversible process. Whether, and when, mitochondrial swelling occurs in the progression towards cell death, whether swelling occurs with or without energy conservation and whether outer membrane rupture is an essential requisite for release of intermembrane proteins involved in caspase and nuclease activation appear to be central issues that need clarification in order to assess the role of mitochondria in cell death.
M I T O C H O N D R I A L U LT R A S T R U C T U R E AND RELEASE OF INTERMEMBRANE PROTEINS A series of observations on thick sections (200±2000 nm or greater) of mitochondria subjected to tomographic reconstruction after high-voltage electron microscopy has revealed an ultrastructure that differs significantly from textbook pictures based on interpretation of 2D images from thin sections (50±80 nm) of fixed and stained samples [60,61]. The major point is that the intercristal compartments are pleiomorphic structures that communicate with the peripheral (intermembrane) space, and sometimes between themselves, through narrow tubular regions . These findings suggest that the intercristal and intermembrane spaces could be functionally
Mitochondria and cell death (Eur. J. Biochem. 264) 689
distinct. This is consistent with several earlier observations concerning the status of mitochondrial cytochrome c. The mitochondrial outer membrane possesses a rotenoneinsensitive NADH cytochrome b5 reductase that is able to reduce cytochrome c in isolated mitochondria . Only a small fraction of cytochrome c can be reduced, which is then reoxidized by cytochrome oxidase and thus catalyzes a shuttle of electrons between the outer and inner mitochondrial membranes . The fraction of cytochrome c that can be reduced by NADH via the outer membrane pathway (i.e. in the presence of rotenone and antimycin A) varies from a minimum of 3% in low ionic strength media to a maximum of 10±15% above 80 mm KCl . These findings define two pools of cytochrome c: (a) a small pool of intermembrane cytochrome c, which can be reduced both by the bc1 complex at the inner membrane and by the NADH-cytochrome b5 reductase at the outer membrane; and (b) a larger pool of cytochrome c, which can only be reduced by the bc1 complex and would be localized inside the intercristal compartments . In this scenario permeabilization of the outer membrane per se can only account for the release of about 10±15% of the total cytochrome c, which could have a minimal impact on respiration and therefore on maintenance of the Dcm. Swelling of the matrix would be required to make more cytochrome c available for release into the cytosol, with concomitant inhibition of respiration and depolarization (Fig. 1). Swelling should not be necessarily intended as a massive, irreversible and synchronous process, as swelling can reversibly occur in a variable fraction of mitochondria. Because apoptosis inducing factor (AIF, see below) and caspases activated by cytochrome c can feed back into the mitochondria to amplify the process, AIF and cytochrome c release could result into a self-amplifying (or self-limiting) process depending on the state of the cell death signalling machinery at the moment of their release. A rigorous assessment of the mechanism(s) of cytochrome c and AIF release, and of its relationship with mitochondrial depolarization, swelling and outer membrane rupture or permeabilization thus demands a careful measurement of several variables. We shall return to this problem after a discussion of the methods that are most widely used to assess mitochondrial function in the living cell.
A N A LY S I S O F M I T O C H O N D R I A L FUNCTION IN SITU Monitoring mitochondrial function in situ is today performed essentially by means of fluorescence techniques. So far, most techniques that have been developed are for the study of Dcm and intramitochondrial [Ca2+] ([Ca2+]m), two parameters which are crucial in defining the role of mitochondria in cell death. These powerful methodologies are generating results which could not have been predicted on the basis of preexisting concepts. These advances, which are largely related to the role of mitochondria in apoptosis, have considerably expanded the scope of mitochondrial research thus contributing to an improved understanding of mitochondrial function in situ. On the other hand, a number of reports have been published that either openly contrast with established concepts of mitochondrial physiology, or reach conclusions that are undermined by a lack of appreciation of the limits of currently available methods. Far from presenting a probe directory, the following paragraphs are intended to review the properties of fluorescent probes with particular emphasis on potential sources of artifacts that may prevent progress in this complex field of research.
690 P. Bernardi et al. (Eur. J. Biochem. 264)
MEASUREMENTS OF THE MITOCHONDRIAL MEMBRANE POTENTIAL At the single cell level, Dcm can be measured by using cationic fluorescent probes in conjunction with microscopy techniques. In the epifluorescence mode the emission is detected by a photomultiplier tube as the total signal from the entire field of observation. In more complex setups the fluorescence image of the cell is acquired and digitized by means of video cameras and dedicated softwares that allow analysis of differences even among the mitochondria of a single cell. Although at high magnification (of the order of 1000) it is possible to analyse single mitochondria , the prolonged exposure time required to obtain such a resolution dramatically affects both cell
q FEBS 1999
viability and mitochondrial function. This problem is worsened in the case of confocal microscopy, where cells are exposed to the enormous power of laser beams. Single cell studies are unique in that they allow the continuous monitoring of the relationships between various parameters, such as, for example, Dcm and [Ca2+]c [1,65±71]. These procedures, however, are limited by definition to the number of cells that can be analyzed at a time. When Dcm has to be assessed in a large cell population, electrodes or radioactive tracers can be utilized to determine the distribution of lipophilic cations between cells and suspending buffers. Radioactive cations, such as[3H]triphenylmethylphosphonium, have also been used to measure Dcm in tissues [72,73]. However, as these distribution analyses require several million cells, flow cytometry is increasingly used to detect the
Fig. 1. Mitochondrial ultrastructure and release of cytochrome c. The scheme in panel A depicts an intramitochondrial compartment created by infoldings of the inner membrane (i.m.). This intercristal space communicates with the intermembrane space through a narrow tubular region, as can be deduced by tomographic reconstructions of thick sections after high-voltage electron microscopy [60±62]. Only a small fraction of cytochrome c is located in the intermembrane space, where it can accept electrons from either the outer membrane (o.m.) rotenone-insensitive NADH-cytochrome b5 reductase or from the inner membrane complex III. Proton pumping creates a m Ä H, which is represented here in the form of a Dcm and denoted by signs. Note that H+ pumping also occurs towards the compartments where most of cytochrome c is located. Panel B depicts the consequences of MTP opening, based on the assumption that a solute gradient exists between the intermembrane and matrix spaces. The membrane potential collapses, and net solute influx is accompanied by water resulting in the buildup of hydrostatic pressure pushing the matrix towards the intermembrane space. This may result in more cytochrome c gaining access to the intermembrane space, possibly through widening of the tubular connecting regions (which remains entirely hypothetical). The increased pressure eventually causes rupture of the outer membrane, with release of cytochrome c and other intermembrane soluble proteins (including AIF which is not depicted here) in the extramitochondrial space (panel C).
q FEBS 1999
fluorescence changes of a relatively small number of cells (103 ±105) loaded with the same probes utilized in fluorescence microscopy studies. The ease with which one can get impressive images of mitochondria in a living cell tends to dim the attention for potential sources of artifacts, which always have to be taken into account when fluorescence emissions are related to mitochondrial function. The following list highlights common problems which are frequently overlooked in studies of mitochondrial function in situ. Cellular uptake and extrusion Once inside the cell, lipophilic cations are electrophoresed into mitochondria but their rate of uptake from the extracellular medium depends on the plasma membrane potential (Dcp). Therefore, the intracellular distribution of the probe depends on both Dcm and Dcp, and on the time of incubation . It has been demonstrated that depending on Dcp values, time of incubation, absolute dye concentrations and cell/dye ratios the fluorescence emitted by the cyanine derivative 3,3 0 -dihexyloxacarbocyanine iodide [DiOC6(3)] can be significantly affected
Mitochondria and cell death (Eur. J. Biochem. 264) 691
by both Dcm and Dcp, the maximal response of fluorescence changes to Dcm changes being observed at the lowest dye/cell ratio and for probe concentrations lower than 40 nm . At higher concentrations the probe may be more responsive to changes of Dcp than of Dcm, and it must be noted that a relevant part of the studies where changes of Dcm have been related to apoptosis has been performed by flow cytometry using DiOC6(3) at concentrations higher than 100 nm. Unfortunately, the possibility that apoptogenic substances or conditions can affect Dcp has not been investigated in the same models, although one major hallmark of apoptosis is the flipping of phosphatidylserine from the inner to the outer leaflet of the plasma membrane , which might in turn affect its electrical properties. A further problem is that this cyanine derivative binds to the endoplasmic reticulum as well , a property that is indeed commercially advertised for the staining of this organelle. The signal coming from mitochondria and the endoplasmic reticulum may therefore be difficult to sort unless cells are imaged. The most important problem with the vast majority of fluorescent probes for Dcm, however, is that their cellular accumulation can be drastically reduced because of efficient
Fig. 2. Inhibition of the MDR pump improves mitochondrial loading with TMRM. Fluorescence images of MH1C1 cells loaded with 100 nm tetramethylrhodamine methyl ester. The low level of mitochondrial fluorescence observed in control cells (panel A) is dramatically increased by treatment with CsH (panel B), verapamil (panel C) or CsA (panel D). Bar, 20 mm.
692 P. Bernardi et al. (Eur. J. Biochem. 264)
extrusion by the multidrug-resistance pump (MDR). Expression of this glycoprotein is indeed routinely investigated by monitoring the decreased uptake of Rhodamine derivatives ; and the increased mitochondrial signal observed with Dcm probes after the addition of CsA could be largely due to an effect on the MDR pump. This is illustrated in Fig. 2, which shows that the efficiency of loading of MH1C1 cells with TMRM is dramatically increased by verapamil, CsH or CsA, which all inhibit the MDR [78,79], while only CsA inhibits the MTP. A lack of effect of verapamil or CsH on mitochondrial fluorescence is therefore required before one can suspect, but not prove, that the effect of CsA is on MTP, an essential control that is almost invariably missing. Binding and quenching The extent of accumulation of lipophilic cations within mitochondria depends on both their initial extramitochondrial concentration and the magnitude of Dcm. Very high intramitochondrial concentrations can in turn result in extensive self-quenching of the dye fluorescence. Indeed, all the methods for Dcm estimation based on the fluorescence changes of cationic probes in mitochondrial suspensions are based on the magnitude of quenching of the total fluorescence. Conversely, in both cell imaging and flow cytometry Dcm is estimated from the intensity of the cell fluorescence and not from the quenching of the total suspension fluorescence. Therefore, to maximize the response of the cell fluorescence to the magnitude of Dcm it is necessary to minimize the quenching of the dye associated with mitochondria. It has also to be pointed out that profound differences exist among the various fluorescent probes concerning their ability to bind to mitochondrial and other cellular membranes independent of Dcm. The fraction of the dye which actually responds to Dcm changes should always be calibrated based on the fluorescence changes induced by uncouplers. Toxicity and bleaching Owing to the combined driving forces of Dcm and Dcp, lipophilic cations are accumulated within the mitochondrial matrix at concentrations exceeding those present in the extracellular milieu by approximately four orders of magnitude. Thus, phototoxic effects are likely to be elicited from the fluorescent molecules, causing in turn deleterious consequences. The imidazolic ring of histidyl residues of the adenine nucleotide translocase (ANT) is highly susceptible to photodamage, particularly by singlet oxygen [80,81]. Also the sensitivity of MTP to various control factors, including Ca2+, is altered by singlet oxygen produced by irradiation of hematoporphyrin-loaded mitochondria . In addition, and probably independent of photodynamic effects, several mitochondrial functions are inhibited by high concentrations of any potentiometric fluorescent probe available. Among the mitochondrial targets, complex I shows a high susceptibility to inhibition . Because of their (photo)toxic effects, potentiometric fluorescent probes are in fact under investigation as potential therapeutic agents against various types of tumors [83±85], which appear to accumulate these molecules more efficiently than normal, non transformed cells. In this context it is worth mentioning that merocyanine, one of the first Dcm probes in isolated mitochondria, is a powerful inducer of apoptosis [86,87], a finding that does support a role for mitochondria in the execution of the cell death program. Thus, if dye concentrations and exposure times are not thoroughly tested,
q FEBS 1999
the fluorescent probes could contribute to cause or potentiate rather than simply measure the changes of Dcm. The ensuing alterations of mitochondrial function and structure could result in an increased production of oxyradicals . It would then become difficult to ascertain whether a decrease in cellular fluorescence depends on a fall of Dcm or rather on the bleaching caused by oxyradicals , an issue that is rarely given due attention. The problem of `fixable' probes Studies of mitochondrial involvement in cell death often demand the assessment of a series of parameters that may require cell fixation, which is obviously not compatible with the maintenance of Dcm and leads to immediate release of the Dcm probes previously accumulated by energized mitochondria. To circumvent this problem, Kroemer and coworkers introduced the use of chloromethyl derivatives of the fluorescent cationic rosamine probes, which are marketed by Molecular Probes (Eugene, OR, USA) under the trademark of `mitotrackers'. The rationale of this approach is that the positively charged probe will be accumulated by energized mitochondria in response to Dcm, followed by binding to mitochondrial SH groups via the probe chloromethyl moiety. At this point, the covalently bound probe should not be released despite deenergization, and therefore stably `mark' the Dcm existing prior to disruption of membrane integrity. Macho et al. indeed showed that accumulation of chloromethyltetramethyl rosamine (CMTMRos) by thymocytes can be partly prevented by the uncoupler carbonylcyanide-p-trifluoromethoxyphenyl hydrazone (FCCP). After cell fixation, the ratio of CMTMRos fluorescence of untreated versus FCCP-pretreated cells under optimal conditions was measurable but extremely small . It is obvious, however, that the probe cannot reliably measure a decrease of Dcm once it has bound matrix and membrane SH groups. This easily explains why release of cytochrome c, caspase activation and poly(ADP-ribose) polymerase cleavage occurred prior to any detectable change of cellular CMTMRos fluorescence in staurosporine-induced apoptosis, a finding that has been taken to mean that mitochondria maintained a high Dcm despite release of substantial amounts of cytochrome c . To make matters worse, our recent studies demonstrate that CMTMRos is a powerful inducer of the mitochondrial PT and an inhibitor of respiratory complex I . Thus, swelling due to a PT remains a plausible mechanism for cytochrome c release in staurosporine-induced apoptosis because CMTMRos may cause a PT-dependent depolarization that cannot be detected by its own fluorescence changes. Signal calibration At variance from the case of isolated mitochondria, calibration of the fluorescence signal with the magnitude of Dcm cannot be achieved in situ . A further problem posed by the available potentiometric fluorescent dyes is that they belong to the single excitation-single emission type of molecules. Thus, a decrease of the signal due to mitochondrial depolarization is indistinguishable from a Dcm-independent leak of the dye into the extracellular medium, especially when prolonged incubations are required. These problems preclude a meaningful comparison between the fluorescent intensities observed in different cells. The cyanine derivative, 5,5 0 ,6,6 0 -tetrachloro-1,1 0 ,3,3 0 -tetraethylbenzimidazolo carbocyanine iodide (JC-1), is frequently (but quite improperly) referred to as a `ratiometric' probe for
q FEBS 1999
Mitochondria and cell death (Eur. J. Biochem. 264) 693
Dcm [68,91,92]. JC-1 displays two major emission peaks (597 and 539 nm with excitation at 490 nm), which correspond to the multimeric and monomeric forms of the dye, respectively . The red fluorescence of the multimer is observed in aqueous solution at high ionic strengths, whereas the green fluorescence of the monomer is present when the probe is located in hydrophobic environments . In isolated mitochondria both the monomer and the multimer emissions can be observed. The green monomer emission was shown to be responsive to values of Dcm below 2140 mV, a range of membrane potentials that hardly modified the multimer emission . Conversely, the red multimer emission was shown to respond to higher (more negative) values of Dcm. In energized cells loaded with JC-1 mitochondria display a bright red fluorescence, which is decreased by deenergization. At the same time, deenergization causes an increase of the green fluorescence reflecting an increase of the monomer concentration, but this is not limited to the mitochondrial membranes. Owing to the hydrophobic interactions of the monomer released from the matrix, the green fluorescence spreads all over cellular structures (Fig. 3). Thus, only the multimer form measures the Dcm-dependent mitochondrial accumulation, whereas the green fluorescence depends on passive binding of JC-1 to any cellular membrane; and comparing the ratios of the two emissions of JC-1 is a questionable practice that relates phenomena occurring in different cellular regions. Interactions between fluorescent molecules When simultaneous imaging of two mitochondrial probes is required, the possible interference between probes should always be checked in solution, although the results in situ still remain difficult to predict. For instance, we showed that intramitochondrial calcein emission is quenched by exposing hepatocytes and hepatoma cells to high concentrations of tetramethylrhodamine methyl ester (TMRM) . From the above points it emerges that the ideal Dcm probe: (a) should accumulate within mitochondria only in response to Dcm; (b) should not be a substrate of the MDR pump; (c) should not be toxic; (d) should not bind passively to mitochondria or to other intracellular organelles; and (e) should not cause photodynamic effects or other forms of photodamage. Although such an ideal probe is not available, reliable results can already be achieved by the correct choice of probe; by carefully adjusting the dye/cell ratio to maximize the Dcm over Dcp response; by checking the contribution of MDR activity to probe loading with verapamil; by minimizing the illumination intensity; and by performing proper controls for phototoxicity and other basic effects on mitochondrial function (such as respiratory activity and maintenance of the membrane potential). Among the many commercial molecules our preference goes to the rhodamine group, in particular TMRM, which in the low nanomolar range exclusively stains the mitochondria and is not retained by the cell upon Dcm collapse.
MEASUREMENTS OF MITOCHONDRIAL FREE [Ca2+] The first technique for monitoring [Ca2+]m in intact cells capitalized on the fact that cell loading with fluorescent [Ca2+] indicators also caused probe localization within mitochondria . In the so called Mn2+-quenching technique the incubation of rat cardiomyocytes with Mn2+ results in the disappearance of the cytosolic signal, leaving the mitochondrial fluorescence of Indo-1 almost unaffected . By developing this procedure
Fig. 3. Effect of mitochondrial uncoupling on cellular JC-1 fluorescence. NRK cells were loaded with JC-1 as previously described , and imaged by confocal microscopy with excitation wavelenght at 488 nm. Emissions at 525 nm (green) and 585 nm (red) were collected simultaneously by using two separate channels on the detector assembly. The red fluorescence reflects the multimer form of the dye and is localized within mitochondria under control conditions (A). The addition of FCCP (B) induced a prompt disappearance of the red fluorescence along with the diffusion of the green fluorescence corresponding to the monomer form of JC-1. Bar, 10 mm. For further explanation see text.
Stern, Silverman and Coworkers could study the relationship between the value of [Ca2+]m attained during anoxia and the death or survival of individual cardiomyocytes upon reoxygenation . This allowed the definition of a threshold of 250 nm [Ca2+]m as the `point of no return' for cardiomyocyte survival. The reliability of the quenching technique requires the total disappearance of the cytosolic signal, which is demonstrated by the absence of Ca2+ transients in contracting cardiomyocytes . However, it might be more difficult to rule out a residual cytosolic contribution in other cell types and, in any case, the cation used for quenching can interfere with mitochondrial Ca2+ homeostasis.
694 P. Bernardi et al. (Eur. J. Biochem. 264)
A different procedure to monitor [Ca2+]m in intact cells in culture was developed by targeting aequorin within mitochondria . The chemiluminescent signal of aequorin, a photoprotein which emits light upon Ca2+ binding, is generally too low to be recorded at the single cell level, although this limitation was recently overridden by using improved detection methods . Nevertheless, the application of this technique to cell suspensions allowed the demonstration of the functional coupling between endoplasmic reticulum and mitochondria . A third approach towards monitoring of [Ca2+]m was made possible by the introduction of a rhodamine derivative, Rhod 2 . Its acetoxymethyl (AM) ester is the only cell-permeant Ca2+ indicator that has a net positive charge, a property that promotes its sequestration into mitochondria. The utilization of dihydro-Rhod 2-AM was reported to increase the selectivity of mitochondrial loading . In this case, the indicator becomes fluorescent only upon its oxidation, which should occur preferentially within the mitochondria. The ease of loading and the lack of interference with mitochondrial Ca2+ movements rapidly made Rhod 2 the most utilized probe for [Ca2+]m. However, the following points should be emphasized: (a) unexplained differences in the indicator batches affect the reproducibilty of the observations, as noted in several laboratories including ours; (b) Rhod 2 is not a ratiometric probe, and fluorescence changes could be produced by factors other than variations in [Ca2+]m; (c) a variable contribution by the cytosolic compartment can affect the mitochondrial signal; as an example, in a recent and elegant study pictures showing an exclusive mitochondrial localization of Rhod 2 coexisted with images showing a diffuse pattern of fluorescence in untreated cells ; and (d) the intracellular distribution can be affected by MTP opening; indeed, Rhod 2 is released through the MTP in isolated mitochondria (V. Petronilli, unpublished observation) and CsA blocks the disappearance of mitochondrial fluorescence in Rhod 2 loaded cardiomyocytes ; this hampers the use of Rhod 2 for investigating the relationships between [Ca2+]m changes and MTP opening. Despite these limitations, relevant results have been obtained with these methods. In particular, it was shown that matrix Ca2+ can undergo rapid changes , and that mitochondria can modulate the frequency and the amplitude of cytosolic Ca2+ oscillations in living cells [103±105]. These findings revived interest in the role of mitochondria in cellular Ca2+ homeostasis, which had been neglected for almost 20 years (reviewed in ). One relevant achievement of these in situ measurements has been the demonstration that under resting conditions, [Ca2+]m in cardiomyocytes is lower than [Ca2+]c [95,106], a finding that rules out the possibility that mitochondria may cause cell damage by releasing Ca2+. A link between Ca2+ overload and mitochondrial dysfunction has long been proposed, but its relevance in vivo remains hard to predict. Mitochondria can tolerate transient increases of [Ca2+]m exceeding 10 mm , while increases of [Ca2+]c in the micromolar range resulted in only slight variations of Dcm, which were perfectly reversible . On the other hand, it must be pointed out that also a decrease in Ca2+ availability could be harmful for mitochondria, as mitochondrial Ca2+ uptake in situ is accompanied by stimulation of ATP synthesis [107±109]. Indeed, high-frequency shortening of cardiomyocytes is accompanied by an increase of [Ca2+]m that stimulates energy production, and the lack of this [Ca2+]m response plays a pivotal role in the evolution towards cardiomyocyte failure and death in the syrian cardiomyopathic hamster .
q FEBS 1999
Taken together, these observations indicate that valuable information on the involvement of mitochondria in cell death may be obtained from measurements of [Ca2+]m, and that the relationship between onset of apoptosis and changes of [Ca2+]m should be investigated more thoroughly.
MEASUREMENTS OF THE PERMEABILITY TRANSITION The role of MTP in cell physiology and pathology is still a matter of debate also because adequate methods to probe MTP directly in intact cells are lacking. The evidence for its occurrence in vivo is largely based on either indirect methods, such as measurements of mitochondrial depolarization, or on pharmacological tools, such as the effects of in vitro pore inducers or inhibitors (such as CsA). The shortcomings are obvious, as depolarization can be caused by a variety of events (most notably, increased ATP demand) while CsA also interferes with calcineurin-dependent signaling, which like MTP has a prominent Ca2+-dependence . Even MeVal-4-Cs, a CsA derivative which inhibits the pore but not calcineurin , still inhibits all cellular cyclophilins and therefore is not selective for the MTP. In addition, the indirect approaches used so far do not allow to address the cellular modulation of MTP, leaving crucial issues unsolved, such as the cause-effect relationship between pore opening and Dcm decrease . These considerations explain the intrinsic interest in developing unequivocal tools to measure MTP opening in living cells. In principle, MTP can be probed directly by a molecule which is excluded (or retained) by mitochondria when the pore is closed, and is taken up by (or released from) mitochondria when the pore opens. Even with such a probe, however, selective inhibition by CsA would still be necessary to prove MTP involvement. Suitable molecules should meet the following minimal requirements: (a) molecular mass lower than 1.5 kDa; (b) little or no hydrophobicity; and (c) lack of utilization as a substrate by mitochondrial enzymes. Among the available fluorescent molecules, calcein has been selected as the probe of choice to detect pore opening with imaging techniques . Although it is cell impermeant, calcein can be easily loaded into cells by using its AM ester form which is nonfluorescent. Once inside the cell the probe is deesterified and trapped in its so-called free form, which is fluorescent ± a strategy commonly used to load cells with fluorescent probes. The initial strategy in calcein utilization was based on the finding of fluorescence voids corresponding to mitochondria in hepatocytes incubated with calcein-AM ester in the presence of TMRM. The voids were interpreted as mitochondria excluding calcein, and their filling with calcein was presented as a procedure to monitor MTP in intact cells . However, we found that mitochondria are easily filled with calcein upon the incubation of many cell types with its AM ester. In addition, calcein fluorescence is quenched by TMRM within mitochondria, indicating that the voids were more likely to be produced by the intramitochondrial colocalization of the two dyes . More importantly, we found that the cytosolic signal can be quenched by incubating calcein-loaded cells with Co2+. This procedure results in the appearance of mitochondria as glowing bodies over a dark background, and allows to study MTP opening as the decrease of mitochondrially associated calcein fluorescence as a function of time. These studies revealed a spontaneous, slow decrease of mitochondrial calcein fluorescence that was completely prevented by CsA, suggesting that MTP fluctuates rapidly between open and
q FEBS 1999
closed states in intact cells . It is noteworthy that no fluorescence changes could be detected by Dcm probes during the CsA-sensitive decrease of calcein fluorescence. Although MTP opening must result in Dcm collapse, the response time of available techniques is not fast enough to detect Dcm changes in the millisecond range, which is well above the open-closed transitions of MTP . MTP openings of short duration can rather be studied by the calcein loading-Co2+ quenching method , a technique that should help characterize conditions associated with MTP opening in situ, and to define the causal relationships between pore opening, collapse of Dcm, perturbation of intracellular and mitochondrial Ca2+ homeostasis, and release of apoptogenic proteins.
M I T O C H O N D R I A L P R O T E I N S I N V O LV E D I N C E L L D E AT H The involvement of mitochondria in cell death has been investigated for many years, particularly in relation to Ca2+ homeostasis. The idea that opening of the MTP could be a factor in ischemia-reperfusion and toxic damage, put forward in the 1980s [115,116], is finding support from recent in vivo studies [117±120]. There is little doubt, however, that the recent impulse to mitochondrial studies in the context of cell death came with the identification of mitochondrial proteins that participate in modulating the execution phase of apoptosis. These proteins can be grouped in two classes: (a) pro- and antiapoptotic members of the Bcl-2 family that largely localize to the outer mitochondrial membrane; and (b) proteins that may be released during apoptosis, like AIF and, quite unexpectedly, cytochrome c. Bcl-2 family This class of proteins includes both anti-apoptotic (such as Bcl-2, Bcl-XL, Bcl-W and Mcl-1) and proapoptotic members (such as Bax, Bak and Bok). The two classes share high sequence homology except for the BH3 domain, which is present only in the proapoptotic proteins. Proteins of this family associate into dimers, and the cellular response to the death signals depends on the ratio of pro-to antiapoptotic molecules (reviewed in ). Studies of subcellular distribution indicate that they localize both to the cytosol and to intracellular membranes, including the nuclear envelope and the outer mitochondrial membrane . The consequences of localization to the outer mitochondrial membrane appear to be twofold: (a) they provide docking sites for other proteins involved in the death cascade, including the kinase Raf and the phosphatase calcineurin ; as these bound proteins remain enzymatically active, they can affect the formation of heterodimers between Bcl-2 and other members of the superfamily that lack the membrane insertion C-terminal domain, and only heterodimerize in the dephosphorylated form ; (b) they appear to affect onset of the PT and/or release of AIF  and cytochrome c ; although the mechanism remains controversial (the relevant literature will be discussed in some detail below) it is intriguing that proapoptotic members (Bax) favor the PT and cytochrome c release while antiapoptotic members (Bcl-2) make the PT more difficult, at least with some inducers; it has been suggested that these effects may be related to the mechanical properties conferred upon the outer membrane by the presence of Bcl-2 family proteins (reviewed in ). It has been shown that Bcl-2 type proteins can form ionic channels that are anion-selective for proapoptotic proteins and cation-selective for antiapoptotic proteins (reviewed in ).
Mitochondria and cell death (Eur. J. Biochem. 264) 695
The link between this channel activity, which has been observed under rather extreme conditions of pH in vitro, and the role in cell death remains unclear. The outer membrane location precludes direct effects on the Dcm and/or on the permeability of the inner membrane, an issue that is not always appreciated. In summary, we think that the reported effects of Bcl-2 family members on the probability of MTP opening and on cytochrome c release still await a mechanistic explanation. Cytochrome c In an effort to purify components required for in vitro activation of caspase 3, Liu et al. identified a fraction containing cytochrome c . These Authors showed that in the presence of dATP (mM) or ATP (mM) cytochrome c was able to activate procaspase 9, followed by recruitment and activation of procaspase 3, the effector protease that cleaves lamin, poly(ADP-ribose) polymerase and fodrin . Considerable evidence exists to show that cytochrome c release to the cytosol takes place in a variety of models of apoptosis, yet a number of problems are apparent when the mechanistic aspects of cytochrome c release are analyzed. In a classical paper, Jacobs and Sanadi showed that cytochrome c can be released by suspending mitochondria in a hypotonic medium, followed by a wash of the membranes in an isotonic saline medium . The hypotonic treatment was required to force mitochondrial swelling and outer membrane rupture, while the salt wash was necessary to detach cytochrome c, which interacts with complexes III and IV mainly by virtue of electrostatic forces. Respiration was inhibited after these treatments, but it could be restored by the addition of exogenous cytochrome c . Because of these observations, cytochrome c depletion has long been known to be a consequence of mitochondrial swelling in saltcontaining media, irrespective of the cause of swelling; and studies of ion transport in saline media are (or should be) routinely performed in cytochrome c-supplemented mitochondria [128,129]. It should be appreciated that cytochrome c release requires both swelling and a high ionic strength. Swelling in hypotonic, sucrose-based media will not release cytochrome c, which under these conditions remains bound to the inner membrane despite outer membrane rupture; and incubation in isotonic saline media alone will not release cytochrome c because the outer membrane does not allow its diffusion. These requirements for cytochrome c release in vitro should be kept in mind, and can probably explain some discrepancies in the field. Indeed, apparent cytochrome c release could be caused by the techniques used to disrupt cells prior to organelle separation, particularly if homogenization is carried out in salt-containing media. Indeed, it has been reported that cytochrome c release could not be detected in the course of apoptosis induced by Fas ligation when Jurkat cells were disrupted by nitrogen cavitation rather than mechanical homogenization . This observation should induce some caution when the evidence for cytochrome c release is based solely on cell fractionation rather than on in situ methods. AIF After in vivo treatment of mice with dexamethazone, a subpopulation of lymphoid cells maintaining a lower Dcm, as assessed with a fluorescent probe, could be sorted out which would then undergo apoptosis in culture. Addition of CsA prevented the fluorescence decrease and delayed apoptosis, suggesting PT involvement in the apoptotic cascade . A
696 P. Bernardi et al. (Eur. J. Biochem. 264)
correlation has been subsequently established between onset of the PT in vitro and appearance of the nuclear signs of apoptosis in a reconstituted system where nuclei were incubated with isolated mitochondria. Nuclear degradation was observed when a PT had occurred, while nuclei remained intact when mitochondria did not undergo a PT irrespective of the methods used to induce or inhibit the pore [131,132]. These observations strengthened the general theory of the mitochondrial control of apoptosis, and the proposal that the PT is a key event in the effector phase of programmed cell death . The link between opening of the PT and nuclear degradation has been identified by the Kroemer group in a protein, AIF, which is associated with markers of the mitochondrial intermembrane space. AIF is a protease acting through proteolytic activation of a nuclear endonuclease; and it is inhibited by N-benzyloxycarbonyl-ValAla-Asp-fluoromethylketone (a caspase inhibitor) but not by specific inhibitors of known Ca2+, serine, or cysteine proteases including caspase-3, all of which are involved in the apoptotic cascade . Cloning has revealed that AIF is a flavoprotein with predicted mass of 57 kDa which displays a striking homology with bacterial ferredoxin and NADH oxidoreductases. The protein contains FAD, which is required for the oxidoreductase but not for the apoptogenic activity . This list of proteins is by no means conclusive, as mitochondria also contain other proteases, including procaspase-3 , caspase-9  and calpain-like species  that might participate in a positive feedback loop, or even trigger mitochondrial apoptotic responses that may include MTP opening.
CYTOCHROME c RELEASE, Bax, Dcm AND THE PERMEABILITY TRANSITION Right after the discovery that cytochrome c release was an early event in apoptosis, researchers have tried to define the pathways for cytochrome c release and its relationships with mitochondrial function, in particular with outer membrane integrity, maintenance of Dcm, and occurrence of a PT. Before the analysis of experimental observations, however, a few key points need to be discussed. The first issue is the relationship between respiration and Dcm, a problem that has been thoroughly investigated in Bioenergetics. This relationship is complex and nonlinear [136±138], which in turn poses obvious questions: how much of cytochrome c must be released before a measurable decrease of Dcm can be expected? Is cytochrome c release mediated by a small fraction of mitochondria, initially at least, or rather by all mitochondria in the cell? These questions demand on one hand a quantitative assessment of cytochrome c release (i.e. a determination of both released and mitochondrially associated cytochrome c), and on the other an assessment of probe sensitivity both to homogeneous and non homogeneous Dcm changes. Finally, and most importantly, mitochondria can maintain a high Dcm by hydrolyzing glycolytic ATP even when respiration is completely blocked, a condition where no relationship can be expected between maintenance of Dcm and cytochrome c release. A careful analysis of the relevant literature suggests that these problems have not always been given due consideration; and that some conclusions need to be reassessed. The widespread conviction that apoptotic cytochrome c release precedes rather than follows mitochondrial depolarization, and therefore that it cannot be caused by a PT, originated from two studies reported in 1997 [125,139]. Wang et al. investigated the model of staurosporine-induced apoptosis and measured Dcm
q FEBS 1999
by rhodamine 123 staining, followed by visualization of mitochondria by laser confocal microscopy. They found that staurosporine did cause mitochondrial depolarization, but release of cytochrome c occurred before a detectable decrease of Dcm . These experiments are not conclusive for three main reasons: (a) cytochrome c release was assessed after mechanical disruption of the cells, and it could have occurred during homogenization because the buffer contained enough salts to release cytochrome c; (b) although the rhodamine 123 signal responded to a high concentration of uncoupler, no reliable calibration is possible in these protocols, and the occurrence of a depolarization smaller than that caused by FCCP (or occurring in just a fraction of the mitochondria) cannot be excluded; and (c) in these protocols Dcm could have been effectively maintained by the mitochondrial hydrolysis of glycolytic ATP even if release of all of cytochrome c had caused complete respiratory inhibition. A more recent study where Dcm and a cytochrome c/green fluorescent protein chimera were simultaneously imaged indeed suggests that after staurosporine treatment mitochondrial depolarization occurs at the same time as cytochrome c release . In the study by Newmeyer et al. experiments were carried out in a cell-free apoptotic system where cytochrome c is spontaneously released from mitochondria. Despite the release of cytochrome c changes of Dcm were not observed in assays based on retention of DiOC6(3) that were otherwise fully sensitive to the addition of a protonophore . These results are easily explained by the fact that 2 mm ATP and an ATP-regenerating system based on phosphocreatine and creatine kinase were present in the assay. Indeed, the Dcm generated by ATP hydrolysis is obviously unaffected by electron flow while it can still be dissipated by protonophoric uncouplers. The idea that cytochrome c release occurs from fully energized mitochondria has been considerably reinforced by recent studies centered on Bid, a BH3 domain-containing protein that interacts with Bcl-2 and Bax. After cleavage by caspase-8 activated by stimulation of cell surface death receptors, the C-terminal portion of cytosolic Bid is able to insert into the mitochondrial outer membrane [141±143]. This Bid fragment appears to cause cytochrome c release directly (i.e. in the absence of mitochondrial swelling) in a process that is antagonized by overexpression of Bcl-2, and was presumed to occur in the absence of mitochondrial depolarization . This latter conclusion needs a reassessment, however, as this study used CMTMRos, which causes respiratory inhibition and induction of the PT while being largely insensitive to changes of Dcm in living cells . Irrespective of the issue of mechanism, a number of studies have shown that mitochondrial overexpression of Bcl-2 decreases release of cytochrome c and of AIF, the latter effect having been ascribed to inhibition of the PT . Conversely, it has been reported that mitochondrial overexpression of the proapoptotic Bax facilitates cell death because it facilitates the mitochondrial release of cytochrome c. Whether the latter effect is due to a PT is perhaps the single most controversial issue of recent literature. Pastorino et al. have reported that overexpression of Bax in a Jurkat T cell line by stable transfection caused mitochondrial depolarization, cytochrome c release, caspase 3 activation, and cell death that could be completely prevented by a combination of CsA and of the phospholipase A2 inhibitor, aristolochic acid . Similar results were reported by Bradham et al. in a model where stimulation of the tumor necrosis factor a receptor induced mitochondrial depolarization, cytochrome c release,
q FEBS 1999
caspase activation and cell death, all events that could be prevented by CsA . These findings led to the conclusion that MTP opening was responsible for cytochrome c release and the subsequent downstream events. In a similar study where Bax was overexpressed in HeLa and COS cells, and cytochrome c release was assessed both in situ and in isolated mitochondria, Eskes et al. reached just the opposite conclusion, as Bax-induced release of cytochome c was not prevented by CsA either in intact cells or in isolated mitochondria . After 15 h of treatment CsA did not prevent release of cytochrome c, but this is not necessarily an argument against involvement of the PT at earlier time points, or even later in the cell death sequence. It is well established that inhibition by CsA is transient in long time-frame experiments even with isolated mitochondria , and Eskes et al.  did not assess whether release of cytochrome c by Bax was accompanied (and therefore potentially caused) by swelling. Indeed, whether CsA inhibits the PT in vitro is not easy to predict, and it remains possible that Bax overexpression affected the PT sensitivity to CsA directly, or by varying the retention of factors that are essential for MTP inhibition by CsA like ADP . In summary, we feel that involvement of the PT in cytochrome c release cannot be ruled out at present. It remains quite possible that different mechanisms (both PT-dependent and PT-independent) may be operating in different experimental systems, or at different time points; or that subtle, inapparent differences exist that will eventually explain what today appear to be unresolved experimental discrepancies. In conclusion, we believe that the renewed enthusiasm about mitochondria in cell death is fully justified. The key role of mitochondria in energy production, the mitochondrial localization of pro- and anti-apoptotic proteins, and the prominent position of mitochondria in cellular Ca2+ homeostasis all suggest potential points of regulation. We hope that this review will help address the mechanistic issues that still need to be solved.
ACKNOWLEDGEMENTS Research in our laboratories is supported by the Ministero per l'Universita 0 e la Ricerca Scientifica e Tecnologica, Projects `Bioenergetics and Membrane Transport' (P. B.) and `Favouring myocardium viability to necrosis'. A challenge in ischemic heart disease: molecular mechanisms and clinical relevance' (F. D. L.); the Consiglio Nazionale delle Ricerche (Dotazione Centro Biomembrane and Grant n. 98.00413.CT04 to P. B.); Telethon-Italy Grant No. 1141 (P. B.); the National Institutes of Health (USA) Grant No. 1R21 Gm58792 (P. B.); and the Giovanni ArmeniseHarvard Foundation (P. B.).
Mitochondria and cell death (Eur. J. Biochem. 264) 697
9. 10. 11.
15. 16. 17. 18.
REFERENCES 1. Lemasters, J.J., DiGiuseppi, J., Nieminen, A.L. & Herman, B. (1987) Blebbing, free Ca2+ and mitochondrial membrane potential preceding cell death in hepatocytes. Nature 325, 78±81. 2. Hockenbery, D.M., NunÄez, G., Milliman, C., Schreiber, R.D. & Korsmeyer, S.J. (1990) Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334±336. 3. Pastorino, J.G., Snyder, J.W., Serroni, A., Hoek, J.B. & Farber, J.L. (1993) Cyclosporin and carnitine prevent the anoxic death of cultured hepatocytes by inhibiting the mitochondrial permeability transition. J. Biol. Chem. 268, 13791±13798. 4. Imberti, R., Nieminen, A.-L., Herman, B. & Lemasters, J.J. (1993) Mitochondrial and glycolytic dysfunction in lethal injury to hepatocytes by t-butylhydroperoxide: protection by fructose,
23. 24. 25.
cyclosporin A and trifluoperazine. J. Pharmacol. Exp. Ther. 265, 392±400. Ankarcrona, M., Dypbukt, J.M., Bonfoco, E., Zhivotovsky, B., Orrenius, S., Lipton, S.A. & Nicotera, P. (1995) Glutamate-induced neuronal death: a succession of necrosis or apoptosis depending on mitochondrial function. Neuron 15, 961±973. Zamzami, N., Marchetti, P., Castedo, M., Decaudin, D., Macho, A., Hirsch, T., Susin, S.A., Petit, P.X., Mignotte, B. & Kroemer, G. (1995) Sequential reduction of mitochondrial transmembrane potential and generation of reactive oxygen species in early programmed cell death. J. Exp. Med. 182, 367±377. Liu, X., Kim, C.N., Yang, J., Jemmerson, R. & Wang, X. (1996) Induction of apoptotic program in cell-free extracts: requirement for dATP and cytochrome c. Cell 86, 147±157. Zou, H., Henzel, W.J., Liu, X., Lutschg, A. & Wang, X. (1997) Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 90, 405±413. Mignotte, B. & Vayssiere, J.L. (1998) Mitochondria and apoptosis. Eur. J. Biochem. 252, 1±15. Bernardi, P. (1998) Mitochondria in cell death (preface). Biochim. Biophys. Acta 1366, 1±2. Ichas, F. & Mazat, J.-P. (1998) From calcium signalling to cell death: two conformations for the mitochondrial permeability transition pore. Biochim. Biophys. Acta 1366, 33±50. Lenaz, G. (1998) Role of mitochondria in oxidative stress and aging. Biochim. Biophys. Acta 1366, 53±67. Di Lisa, F., MenaboÁ, R., Canton, M. & Petronilli, V. (1998) The role of mitochondria in the salvage and the injury of the ischemic myocardium. Biochim. Biophys. Acta 1366, 69±78. Halestrap, A.P., Kerr, P.M., Javadov, S. & Woodfield, K.-Y. (1998) Elucidating the molecular mechanism of the permeability transition pore and its role in reperfusion injury. Biochim. Biophys. Acta 1366, 79±94. Nicholls, D.G. & Budd, S.L. (1998) Mitochondria and glutamate neurotoxicity. Biochim. Biophys. Acta 1366, 97±112. Montal, M. (1998) Mitochondria, glutamate neurotoxicity and the death cascade. Biochim. Biophys. Acta 1366, 113±126. Reed, J.C., Jurgensmeier, J. & Matsuyama, S. (1998) Bcl-2 family proteins and mitochondria. Biochim. Biophys. Acta 1366, 127±137. Cai, J., Yang, J. & Jones, D.P. (1998) Mitochondrial control of apoptosis: the role of cytochrome c. Biochim. Biophys. Acta 1366, 139±149. Susin, S.A., Zamzami, N. & Kroemer, G. (1998) Mitochondria as regulators of apoptosis: doubts no more. Biochim. Biophys. Acta 1366, 151±165. Gores, G.J., Miyoshi, H., Botla, R., Aguilar, H.I. & Bronk, S.F. (1998) Induction of the mitochondrial permeability transition as a mechanism of liver injury during cholestasis: a potential role for mitochondrial proteases. Biochim. Biophys. Acta 1366, 167±175. Lemasters, J.J., Nieminen, A.L., Qian, T., Trost, L., Elmore, S.P., Nishimura, Y., Crowe, R.A., Cascio, W.E., Bradham, C.A., Brenner, D.A. & Herman, B. (1998) The mitochondrial permeability transition in cell death: a common mechanism in necrosis, apoptosis and autophagy. Biochim. Biophys. Acta 1366, 177±196. Di Mauro, S., Bonilla, E., Davidson, M., Hirano, M. & Schon, E.A. (1998) Mitochondria in neuromuscular disorders. Biochim. Biophys. Acta 1366, 199±210. Beal, M.F. (1998) Mitochondrial dysfunction in neurodegenerative diseases. Biochim. Biophys. Acta 1366, 211±223. Schapira, A.H.V. (1998) Mitochondrial dysfunction in neurodegenerative disorders. Biochim. Biophys. Acta 1366, 225±233. Mitchell, P. (1966) Chemiosmotic Coupling in Oxidative and Photosynthetic Phosphorylation. Glynn Research, Bodmin, Cornwall, England. Mitchell, P. (1979) Keilin's respiratory chain concept and its chemiosmotic consequences. Science 206, 1148±1159. Azzone, G.F., Pietrobon, D. & Zoratti, M. (1984) Determination of the
698 P. Bernardi et al. (Eur. J. Biochem. 264)
30. 31. 32.
proton electrochemical gradient across biological membranes. Curr. Top. Bioenerg. 13, 1±77. Green, D.R. & Reed, J.C. (1998) Mitochondria and apoptosis. Science 281, 1309±1312. Brierley, G.P., Baysal, K. & Jung, D.W. (1994) Cation transport systems in mitochondria: Na+ and K+ uniports and exchangers. J. Bioenerg. Biomembr. 26, 519±526. Garlid, K.D. (1996) Cation transport in mitochondria ± the potassium cycle. Biochim. Biophys. Acta 1275, 123±126. Bernardi, P. (1999) Mitochondrial transport of cations: channels, exchangers and permeability transition. Physiol. Rev. 79 (in press). Pozzan, T., Bragadin, M. & Azzone, G.F. (1977) Disequilibrium between steady-state Ca2+ accumulation ratio and membrane potential in mitochondria. Pathway and role of Ca2+ efflux. Biochemistry 16, 5618±5625. Sparagna, G.C., Gunter, K.K., Sheu, S.S. & Gunter, T.E. (1995) Mitochondrial calcium uptake from physiological-type pulses of calcium. A description of the rapid uptake mode. J. Biol. Chem. 270, 27510±27515. Gunter, T.E., Gunter, K.K., Sheu, S.S. & Gavin, C.E. (1994) Mitochondrial calcium transport: physiological and pathological relevance. Am. J. Physiol. 267, C313±C339. Gunter, T.E., Buntinas, L. & Gunter, K.K. (1998) The Ca2+ transport mechanisms of mitochondria and Ca2+ uptake from physiologicaltype Ca2+ transients. Biochim. Biophys. Acta 1399, 5±15. Hunter, D.R. & Haworth, R.A. (1979) The Ca2+-induced membrane transition in mitochondria. I. The protective mechanisms. Arch. Biochem. Biophys. 195, 453±459. Haworth, R.A. & Hunter, D.R. (1979) The Ca2+-induced membrane transition in mitochondria. II. Nature of the Ca2+ trigger site. Arch. Biochem. Biophys. 195, 460±467. Hunter, D.R. & Haworth, R.A. (1979) The Ca2+-induced membrane transition in mitochondria. III. Transitional Ca2+ release. Arch. Biochem. Biophys. 195, 468±477. Massari, S. & Azzone, G.F. (1972) The equivalent pore radius of intact and damaged mitochondria and the mechanism of active shrinkage. Biochim. Biophys. Acta 283, 23±29. Bernardi, P. (1992) Modulation of the mitochondrial cyclosporin A-sensitive permeability transition pore by the proton electrochemical gradient. Evidence that the pore can be opened by membrane depolarization. J. Biol. Chem. 267, 8834±8839. Bernardi, P., Vassanelli, S., Veronese, P., Colonna, R., Szabo, I. & Zoratti, M. (1992) Modulation of the mitochondrial permeability transition pore. Effect of protons and divalent cations. J. Biol. Chem. 267, 2934±2939. Broekemeier, K.M. & Pfeiffer, D.R. (1995) Inhibition of the mitochondrial permeability transition by cyclosporin A during long time frame experiments: relationship between pore opening and the activity of mitochondrial phospholipases. Biochemistry 34, 16440±16449. Bernardi, P., Veronese, P. & Petronilli, V. (1993) Modulation of the mitochondrial cyclosporin A-sensitive permeability transition pore. I. Evidence for two separate Me2+ binding sites with opposing effects on the pore open probability. J. Biol. Chem. 268, 1005±1010. Connern, C.P. & Halestrap, A.P. (1996) Chaotropic agents and increased matrix volume enhance binding of mitochondrial cyclophilin to the inner mitochondrial membrane and sensitize the mitochondrial permeability transition to [Ca2+]. Biochemistry 35, 8172±8180. Nicolli, A., Basso, E., Petronilli, V., Wenger, R.M. & Bernardi, P. (1996) Interactions of cyclophilin with the mitochondrial inner membrane and regulation of the permeability transition pore, a cyclosporin A-sensitive channel. J. Biol. Chem. 271, 2185±2192. Chernyak, B.V. & Bernardi, P. (1996) The mitochondrial permeability transition pore is modulated by oxidative agents through both pyridine nucleotides and glutathione at two separate sites. Eur. J. Biochem. 238, 623±630. Costantini, P., Chernyak, B.V., Petronilli, V. & Bernardi, P. (1996) Modulation of the mitochondrial permeability transition pore by
q FEBS 1999
pyridine nucleotides and dithiol oxidation at two separate sites. J. Biol. Chem. 271, 6746±6751. Fontaine, E., Eriksson, O., Ichas, F. & Bernardi, P. (1998) Regulation of the permeability transition pore in skeletal muscle mitochondria. Modulation by electron flow through the respiratory chain complex I. J. Biol. Chem. 273, 12662±12668. Fontaine, E., Ichas, F. & Bernardi, P. (1998) A ubiquinone-binding site regulates the mitochondrial permeability transition pore. J. Biol. Chem. 273, 25734±25740. Kroemer, G., Dallaporta, B. & Resche-Rigon, M. (1998) The mitochondrial death/life regulator in apoptosis and necrosis. Annu. Rev. Physiol. 60, 619±642. Vinogradov, A., Scarpa, A. & Chance, B. (1972) Calcium and pyridine nucleotide interaction in mitochondrial membranes. Arch. Biochem. Biophys. 152, 646±654. Scott Boyer, C., Moore, G.A. & Moldeus, P. (1993) Submitochondrial localization of the NAD+ glycohydrolase. Implications for the role of pyridine nucleotide hydrolysis in mitochondrial calcium fluxes. J. Biol. Chem. 268, 4016±4020. Di Lisa, F. & Bernardi, P. (1998) Mitochondrial function as a determinant of recovery or death in cell response to injury. Mol. Cell Biochem. 184, 379±391. Pfeiffer, D.R., Gudz, T.I., Novgorodov, S.A. & Erdahl, W.L. (1995) The peptide mastoparan is a potent facilitator of the mitochondrial permeability transition. J. Biol. Chem. 270, 4923±4932. 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. HuÈser, J., Rechenmacher, C.E. & Blatter, L.A. (1998) Imaging the permeability pore transition in single mitochondria. Biophys. J. 74, 2129±2137. Petronilli, V., Miotto, G., Canton, M., Colonna, R., Bernardi, P. & Di Lisa, F. (1999) Transient and long-lasting openings of the mitochondrial permeability transition pore can be monitored directly in intact cells by changes of mitochondrial calcein fluorescence. Biophys. J. 76, 725±734. Rossi, E. & Azzone, G.F. (1970) The mechanism of ion translocation in mitochondria. 3. Coupling of K+ efflux with ATP synthesis. Eur. J. Biochem. 12, 319±327. Azzone, G.F. & Azzi, A. (1965) Volume changes in liver mitochondria. Proc. Natl Acad. Sci. USA 53, 1084±1089. Mannella, C.A., Marko, M., Penczek, P., Barnard, D. & Frank, J. (1994) The internal compartmentation of rat-liver mitochondria: tomographic study using the high-voltage transmission electron microscope. Microsc. Res. Technical 27, 278±283. Perkins, G., Renken, C., Martone, M.E., Young, S.J., Ellisman, M. & Frey, T. (1997) Electron tomography of neuronal mitochondria: three-dimensional structure and organization of cristae and membrane contacts. J. Struct. Biol. 119, 260±272. Mannella, C.A., Marko, M. & Buttle, K. (1997) Reconsidering mitochondrial structure: new views of an old organelle. Trends. Biochem. Sci. 22, 37±38. Sottocasa, G.L., Kuylenstierna, B., Ernster, L. & Bergstrand, A. (1967) An electron-transport system associated with the outer membrane of liver mitochondria. J. Cell Biol. 32, 415±438. Bernardi, P. & Azzone, G.F. (1981) Cytochrome c as an electron shuttle between the outer and inner mitochondrial membranes. J. Biol. Chem. 256, 7187±7192. Loew, L.M., Carrington, W., Tuft, R.A. & Fay, F.S. (1994) Physiological cytosolic Ca2+ transients evoke concurrent mitochondrial depolarizations. Proc. Natl Acad. Sci. USA 91, 12579±12583. Zoeteweij, J.P., van de Water, B., de Bont, H.J., Mulder, G.J. & Nagelkerke, J.F. (1993) Calcium-induced cytotoxicity in hepatocytes after exposure to extracellular ATP is dependent on inorganic phosphate. Effects on mitochondrial calcium. J. Biol. Chem. 268, 3384±3388. Schinder, A.F., Olson, E.C., Spitzer, N.C. & Montal, M. (1996) Mitochondrial dysfunction is a primary event in glutamate neurotoxicity. J. Neurosci. 16, 6125±6133.
q FEBS 1999 68. White, R.J. & Reynolds, I.J. (1996) Mitochondrial depolarization in glutamate-stimulated neurons: an early signal specific to excitotoxin exposure. J. Neurosci. 16, 5688±5697. 69. Budd, S.L. & Nicholls, D.G. (1996) A reevaluation of the role of mitochondria in neuronal Ca2+ homeostasis. J. Neurochem. 66, 403±411. 70. Leyssens, A., Nowicky, A.V., Patterson, L., Crompton, M. & Duchen, M.R. (1996) The relationship between mitochondrial state, ATP hydrolysis, [Mg2+]i and [Ca2+]i studied in isolated rat cardiomyocytes. J. Physiol. (London) 496, 111±128. 71. Delcamp, T.J., Dales, C., Ralenkotter, L., Cole, P.S. & Hadley, R.W. (1998) Intramitochondrial [Ca2+] and membrane potential in ventricular myocytes exposed to anoxia-reoxygenation. Am. J. Physiol. 275, H484±H494. 72. Hoek, J.B., Nicholls, D.G. & Williamson, J.R. (1980) Determination of the mitochondrial protonmotive force in isolated hepatocytes. J. Biol. Chem. 255, 1458±1464. 73. Wan, B., Doumen, C., Duszynski, J., Salama, G. & LaNoue, K.F. (1993) A method of determining electrical potential gradient across mitochondrial membrane in perfused rat hearts. Am. J. Physiol. 265, H445±H452. 74. Rottenberg, H. & Wu, S. (1998) Quantitative assay by flow cytometry of the mitochondrial membrane potential in intact cells. Biochim. Biophys. Acta 1404, 393±404. 75. Fadok, V.A., Voelker, D.R., Campbell, P.A., Cohen, J.J., Bratton, D.L. & Henson, P.M. (1992) Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148, 2207±2216. 76. Terasaki, M. (1989) Fluorescent labeling of endoplasmic reticulum. Methods Cell Biol. 29, 125±135. 77. Petriz, J. & Garcia-Lopez, J. (1997) Flow cytometric analysis of P-glycoprotein function using rhodamine 123. Leukemia 11, 1124±1130. 78. Eytan, G.D., Regev, R., Oren, G., Hurwitz, C.D. & Assaraf, Y.G. (1997) Efficiency of P-glycoprotein-mediated exclusion of rhodamine dyes from multidrug-resistant cells is determined by their passive transmembrane movement rate. Eur. J. Biochem. 248, 104±112. 79. Litman, T., Zeuthen, T., Skovsgaard, T. & Stein, W.D. (1997) Structure-activity relationships of P-glycoprotein interacting drugs: kinetic characterization of their effects on ATPase activity. Biochim. Biophys. Acta 1361, 159±168. 80. Atlante, A., Passarella, S., Quagliariello, E., Moreno, G. & Salet, C. (1989) Haematoporphyrin derivative (Photofrin II) photosensitization of isolated mitochondria: inhibition of ADP/ATP translocator. J. Photochem. Photobiol. B4, 35±46. 81. Salet, C. & Moreno, G. (1990) Photosensitization of mitochondria. Molecular and cellular aspects. J. Photochem. Photobiol. B5, 133±150. 82. Salet, C., Moreno, G., Ricchelli, F. & Bernardi, P. (1997) Singlet oxygen produced by photodynamic action causes inactivation of the mitochondrial permeability transition pore. J. Biol. Chem. 272, 21938±21943. 83. Bernal, S.D., Shapiro, H.M. & Chen, L.B. (1982) Monitoring the effect of anti-cancer drugs on L1210 cells by a mitochondrial probe, rhodamine-123. Int. J. Cancer 30, 219±224. 84. Davis, S., Weiss, M.J., Wong, J.R., Lampidis, T.J. & Chen, L.B. (1985) Mitochondrial and plasma membrane potentials cause unusual accumulation and retention of rhodamine 123 by human breast adenocarcinoma-derived MCF-7 cells. J. Biol. Chem. 260, 13844±13850. 85. Arcadi, J.A. (1998) The effect of rhodamine-123 on 3 prostate tumors from the rat. J. Urol. 160, 2402±2406. 86. Gulliya, K.S., Sharma, R., Liu, H.W., Arnold, L. & Matthews, J.L. (1995) Relationship of mitochondrial function and cellular adenosine triphosphate levels to pMC540 and merodantoin cytotoxicity in MCF-7 human breast cancer cells. Anticancer Drugs 6, 545±552. 87. Rottenberg, H. & Wu, S. (1997) Mitochondrial dysfunction in
Mitochondria and cell death (Eur. J. Biochem. 264) 699
lymphocytes from old mice: enhanced activation of the permeability transition. Biochem. Biophys. Res. Commun. 240, 68±74. Macho, A., Decaudin, D., Castedo, M., Hirsch, T., Susin, S.A., Zamzami, N. & Kroemer, G. (1996) Chloromethyl-X-Rosamine is an aldehyde-fixable potential-sensitive fluorochrome for the detection of early apoptosis. Cytometry 25, 333±340. Bossy-Wetzel, E., Newmeyer, D.D. & Green, D.R. (1998) Mitochondrial cytochrome c release in apoptosis occurs upstream of DEVD-specific caspase activation and independently of mitochondrial transmembrane depolarization. EMBO J. 17, 37±49. Scorrano, L., Petronilli, V., Colonna, R., Di Lisa, F. & Bernardi, P. (1999) Chloromethyltetramethylrosamine (Mitotracker OrangeTM) induces the mitochondrial permeability transition and inhibits respiratory complex I. Implications for the mechanism of cytochrome c release. J. Biol. Chem. 274, in press. Reers, M., Smith, T.W. & Chen, L.B. (1991) J-aggregate formation of a carbocyanine as a quantitative fluorescent indicator of membrane potential. Biochemistry 30, 4480±4486. Cossarizza, A., Ceccarelli, D. & Masini, A. (1996) Functional heterogeneity of an isolated mitochondrial population revealed by cytofluorometric analysis at the single organelle level. Exp. Cell Res. 222, 84±94. Di Lisa, F., Blank, P.S., Colonna, R., Gambassi, G., Silverman, H.S., Stern, M.D. & Hansford, R.G. (1995) Mitochondrial membrane potential in single living adult rat cardiac myocytes exposed to anoxia or metabolic inhibition. J. Physiol. (London) 486, 1±13. Miyata, H., Silverman, H.S., Sollott, S.J., Lakatta, E.G., Stern, M.D. & Hansford, R.G. (1991) Measurement of mitochondrial free Ca2+ concentration in living single rat cardiac myocytes. Am. J. Physiol. 261, H1223±H1234. Miyata, H., Lakatta, E.G., Stern, M.D. & Silverman, H.S. (1992) Relation of mitochondrial and cytosolic free calcium to cardiac myocyte recovery after exposure to anoxia. Circ. Res. 71, 605±613. Rizzuto, R., Simpson, A.W., Brini, M. & Pozzan, T. (1992) Rapid changes of mitochondrial Ca2+ revealed by specifically targeted recombinant aequorin. Nature 358, 325±327. Rutter, G.A., Burnett, P., Rizzuto, R., Brini, M., Murgia, M., Pozzan, T., Tavare, J.M. & Denton, R.M. (1996) Subcellular imaging of intramitochondrial Ca2+ with recombinant targeted aequorin: significance for the regulation of pyruvate dehydrogenase activity. Proc. Natl Acad. Sci. USA 93, 5489±5494. Rizzuto, R., Pinton, P., Carrington, W., Fay, F.S., Fogarty, K.E., Lifshitz, L.M., Tuft, R.A. & Pozzan, T. (1998) Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses. Science 280, 1763±1766. Minta, A., Kao, J.P. & Tsien, R.Y. (1989) Fluorescent indicators for cytosolic calcium based on rhodamine and fluorescein chromophores. J. Biol. Chem. 264, 8171±8178. Hajnoczky, G., Robb-Gaspers, L.D., Seitz, M.B. & Thomas, A.P. (1995) Decoding of cytosolic calcium oscillations in the mitochondria. Cell 82, 415±424. Boitier, E., Rea, R. & Duchen, M.R. (1999) Mitochondria exert a negative feedback on the propagation of intracellular Ca2+ waves in rat cortical astrocytes. J. Cell Biol. 145, 795±808. Bowser, D.N., Minamikawa, T., Nagley, P. & Williams, D.A. (1998) Role of mitochondria in calcium regulation of spontaneously contracting cardiac muscle cells. Biophys. J. 75, 2004±2014. Werth, J.L. & Thayer, S.A. (1994) Mitochondria buffer physiological calcium loads in cultured rat dorsal root ganglion neurons. J. Neurosci. 14, 348±356. Jouaville, L.S., Ichas, F., Holmuhamedov, E.L., Camacho, P. & Lechleiter, J.D. (1995) Synchronization of calcium waves by mitochondrial substrates in Xenopus laevis oocytes. Nature 377, 438±441. Babcock, D.F., Herrington, J., Goodwin, P.C., Park, Y.B. & Hille, B. (1997) Mitochondrial participation in the intracellular Ca2+ network. J. Cell Biol. 136, 833±844. Di Lisa, F., Gambassi, G., Spurgeon, H. & Hansford, R.G. (1993)
700 P. Bernardi et al. (Eur. J. Biochem. 264)
Intramitochondrial free calcium in cardiac myocytes in relation to dehydrogenase activation. Cardiovasc. Res. 27, 1840±1844. McCormack, J.G., Halestrap, A.P. & Denton, R.M. (1990) Role of calcium ions in regulation of mammalian intramitochondrial metabolism. Physiol. Rev. 70, 391±425. Hansford, R.G. (1994) Physiological role of mitochondrial Ca2+ transport. J. Bioenerg. Biomembr. 26, 495±508. Robb-Gaspers, L.D., Rutter, G.A., Burnett, P., Hajnoczky, G., Denton, R.M. & Thomas, A.P. (1998) Coupling between cytosolic and mitochondrial calcium oscillations: role in the regulation of hepatic metabolism. Biochim. Biophys. Acta 1366, 17±32. Di Lisa, F., Fan, C.Z., Gambassi, G., Hogue, B.A., Kudryashova, I. & Hansford, R.G. (1993) Altered pyruvate dehydrogenase control and mitochondrial free Ca2+ in hearts of cardiomyopathic hamsters. Am. J. Physiol. 264, H2188±H2197. Bennett, W.M. & Norman, D.J. (1986) Action and toxicity of cyclosporine. Annu. Rev. Med. 37, 215±224. Petronilli, V., Cola, C., Massari, S., Colonna, R. & Bernardi, P. (1993) Physiological effectors modify voltage sensing by the cyclosporin A-sensitive permeability transition pore of mitochondria. J. Biol. Chem. 268, 21939±21945. Nieminen, A.L., Saylor, A.K., Tesfai, S.A., Herman, B. & Lemasters, J.J. (1995) Contribution of the mitochondrial permeability transition to lethal injury after exposure of hepatocytes to t-butylhydroperoxide. Biochem. J. 307, 99±106. Szabo, I. & Zoratti, M. (1991) The giant channel of the inner mitochondrial membrane is inhibited by cyclosporin A. J. Biol. Chem. 266, 3376±3379. Crompton, M., Costi, A. & Hayat, L. (1987) Evidence for the presence of a reversible Ca2+-dependent pore activated by oxidative stress in heart mitochondria. Biochem. J. 245, 915±918. Crompton, M. & Costi, A. (1988) Kinetic evidence for a heart mitochondrial pore activated by Ca2+, inorganic phosphate and oxidative stress. A potential mechanism for mitochondrial dysfunction during cellular Ca2+ overload. Eur. J. Biochem. 178, 489±501. Folbergrova, J., Li, P.A., Uchino, H., Smith, M.L. & SiesjoÈ, B.K. (1997) Changes in the bioenergetic state of rat hippocampus during 2.5 min of ischemia, and prevention of cell damage by cyclosporin A in hyperglycemic subjects. Exp. Brain Res. 114, 44±50. Li, P.A., Uchino, H., Elmer, E. & SiesjoÈ, B.K. (1997) Amelioration by cyclosporin A of brain damage following 5 or 10 min of ischemia in rats subjected to preischemic hyperglycemia. Brain Res. 753, 133±140. Friberg, H., Ferrand-Drake, M., Bengtsson, F., Halestrap, A.P. & Wieloch, T. (1998) Cyclosporin A, but not FK 506, protects mitochondria and neurons against hypoglycemic damage and implicates the mitochondrial permeability transition in cell death. J. Neurosci. 18, 5151±5159. Kondo, Y., Asanuma, M., Iwata, E., Kondo, F., Miyazaki, I. & Ogawa, N. (1999) Early treatment with cyclosporin A ameliorates the reduction of muscarinic acetylcholine receptors in gerbil hippocampus after transient forebrain ischemia. Neurochem. Res. 24, 9±13. Reed, J.C. (1997) Double identity for proteins of the Bcl-2 family. Nature 387, 773±776. Riparbelli, M.G., Callaini, G., Tripodi, S.A., Cintorino, M., Tosi, P. & Dallai, R. (1995) Localization of the Bcl-2 protein to the outer mitochondrial membrane by electron microscopy. Exp. Cell Res. 221, 363±369. Wang, H.G., Rapp, U.R. & Reed, J.C. (1996) Bcl-2 targets the protein kinase Raf-1 to mitochondria. Cell 87, 629±638. Susin, S.A., Zamzami, N., Castedo, M., Hirsch, T., Marchetti, P., Macho, A., Daugas, E., Geuskens, M. & Kroemer, G. (1996) Bcl-2 inhibits the mitochondrial release of an apoptogenic protease. J. Exp. Med. 184, 1331±1341. Yang, J., Liu, X., Bhalla, K., Kim, C.N., Ibrado, A.M., Cai, J., Peng, T.I., Jones, D.P. & Wang, X. (1997) Prevention of apoptosis by Bcl-2: release of cytochrome c from mitochondria blocked. Science 275, 1129±1132.
q FEBS 1999 126. Schendel, S., Montal, M. & Reed, J.C. (1998) Bcl-2 family proteins as ion channels. Cell Death Diff. 5, 372±380. 127. Jacobs, E.E. & Sanadi, D.R. (1960) The reversible removal of cytochrome c from mitochondria. J. Biol. Chem. 235, 531±534. 128. Bernardi, P., Pozzan, M. & Azzone, G.F. (1982) Mitochondrial oscillation and activation of H+/cation exchange. J. Bioenerg. Biomembr. 14, 387±403. 129. Bernardi, P., Angrilli, A., Ambrosin, V. & Azzone, G.F. (1989) Activation of latent K+ uniport in mitochondria treated with the ionophore A23187. J. Biol. Chem. 264, 18902±18906. 130. Adachi, S., Gottlieb, R.A. & Babior, B.M. (1998) Lack of release of cytochrome C from mitochondria into cytosol early in the course of Fas-mediated apoptosis of Jurkat cells. J. Biol. Chem. 273, 19892±19894. 131. Zamzami, N., Marchetti, P., Castedo, M., Hirsch, T., Susin, S.A., Masse, B. & Kroemer, G. (1996) Inhibitors of permeability transition interfere with the disruption of the mitochondrial transmembrane potential during apoptosis. FEBS Lett. 384, 53±57. 132. Zamzami, N., Susin, S.A., Marchetti, P., Hirsch, T., Gomez Monterrey, I., Castedo, M. & Kroemer, G. (1996) Mitochondrial control of nuclear apoptosis. J. Exp. Med. 183, 1533±1544. 133. Susin, S.A., Lorenzo, H.K., Zamzami, N., Marzo, I., Snow, B.E., Brothers, G.M., Mangion, J., Jacotot, E., Costantini, P., Loeffler, M., Larochette, N., Goodlett, D.R., Aebersold, R., Siderovski, D.P., Penninger, J.M. & Kroemer, G. (1999) Molecular characterization of mitochondrial apoptosis-inducing factor. Nature 397, 441±446. 134. Mancini, M., Nicholson, D.W., Roy, S., Thornberry, N.A., Peterson, E.P., Casciola, R.L. & Rosen, A. (1998) The caspase-3 precursor has a cytosolic and mitochondrial distribution: implications for apoptotic signaling. J. Cell Biol. 140, 1485±1495. 135. Krajewski, S., Krajewska, M., Ellerby, L.M., Welsh, K., Xie, Z., Deveraux, Q.L., Salvesen, G.S., Bredesen, D.E., Rosenthal, R.E., Fiskum, G. & Reed, J.C. (1999) Release of caspase-9 from mitochondria during neuronal apoptosis and cerebral ischemia. Proc. Natl Acad. Sci. USA 96, 5752±5757. 136. Nicholls, D.G. (1974) The influence of respiration and ATP hydrolysis on the proton-electrochemical gradient across the inner membrane of rat-liver mitochondria as determined by ion distribution. Eur. J. Biochem. 50, 305±315. 137. Pietrobon, D., Azzone, G.F. & Walz, D. (1981) Effect of funiculosin and antimycin A on the redox-driven H+-pumps in mitochondria: on the nature of `leaks'. Eur. J. Biochem. 117, 389±394. 138. Luvisetto, S., Schmehl, I., Intravaia, E., Conti, E. & Azzone, G.F. (1992) Mechanism of loss of thermodynamic control in mitochondria due to hyperthyroidism and temperature. J. Biol. Chem. 267, 15348±15355. 139. Kluck, R.M., Bossy-Wetzel, E., Green, D.R. & Newmeyer, D.D. (1997) The release of cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis. Science 275, 1132±1136. 140. Heiskanen, K.M., Bhat, M.B., Wang, H.-W., Ma, J. & Nieminen, A.-L. (1999) Mitochondrial depolarization accompanies cytochrome c release during apoptosis in PC6 cells. J. Biol. Chem. 274, 5654±5658. 141. Li, H., Zhu, H., Xu, C.J. & Yuan, J. (1998) Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell 94, 491±501. 142. Luo, X., Budihardjo, I., Zou, H., Slaughter, C. & Wang, X. (1998) Bid, a Bcl2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors. Cell 94, 481±490. 143. Gross, A., Yin, X.M., Wang, K., Wei, M.C., Jockel, J., Milliman, C., Erdjument-Bromage, H., Tempst, P. & Korsmeyer, S.J. (1999) Caspase cleaved BID targets mitochondria and is required for cytochrome c release, while BCL-XL prevents this release but not tumor necrosis factor-R1/Fas death. J. Biol. Chem. 274, 1156±1163. 144. Pastorino, J.G., Chen, S.T., Tafani, M., Snyder, J.W. & Farber, J.L. (1998) The overexpression of Bax produces cell death upon induction of the mitochondrial permeability transition. J. Biol. Chem. 273, 7770±7775.
q FEBS 1999 145. Bradham, C.A., Qian, T., Streetz, K., Trautwein, C., Brenner, D.A. & Lemasters, J.J. (1998) The mitochondrial permeability transition is required for tumor necrosis factor alpha-mediated apoptosis and cytochrome c release. Mol. Cell. Biol. 18, 6353±6364. 146. Eskes, R., Antonsson, B., Osen-Sand, A., Montessuit, S., Richter, C., Sadoul, R., Mazzei, G., Nichols, A. & Martinou, J.-C. (1998)
Mitochondria and cell death (Eur. J. Biochem. 264) 701 Bax-induced cytochrome c release from mitochondria is independent of the permeability transition pore but highly dependent on Mg2+ ions. J. Cell Biol. 143, 217±224. 147. Novgorodov, S.A., Gudz, T.I., Milgrom, Y.M. & Brierley, G.P. (1992) The permeability transition in heart mitochondria is regulated synergistically by ADP and cyclosporin A. J. Biol. Chem. 267, 16274±16282.