Portal de Periódicos da CAPES

Apoptosis driven by IP3-linked mitochondrial calcium signals

1999; Springer Nature; Volume: 18; Issue: 22 Linguagem: Inglês

10.1093/emboj/18.22.6349

1460-2075

Gábor Szalai, Rajeshwari Krishnamurthy, György Hajnóczky,

Mitochondrial Function and Pathology

Article15 November 1999free access Apoptosis driven by IP3-linked mitochondrial calcium signals Gábor Szalai Gábor Szalai Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author Rajeshwari Krishnamurthy Rajeshwari Krishnamurthy Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author György Hajnóczky Corresponding Author György Hajnóczky Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author Gábor Szalai Gábor Szalai Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author Rajeshwari Krishnamurthy Rajeshwari Krishnamurthy Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author György Hajnóczky Corresponding Author György Hajnóczky Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA Search for more papers by this author Author Information Gábor Szalai1, Rajeshwari Krishnamurthy1 and György Hajnóczky 1 1Department of Pathology, Anatomy and Cell Biology, Room 253 JAH, Thomas Jefferson University, Philadelphia, PA, 19107 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:6349-6361https://doi.org/10.1093/emboj/18.22.6349 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Increases of mitochondrial matrix [Ca2+] ([Ca2+]m) evoked by calcium mobilizing agonists play a fundamental role in the physiological control of cellular energy metabolism. Here, we report that apoptotic stimuli induce a switch in mitochondrial calcium signalling at the beginning of the apoptotic process by facilitating Ca2+-induced opening of the mitochondrial permeability transition pore (PTP). Thus [Ca2+]m signals evoked by addition of large Ca2+ pulses or, unexpectedly, by IP3-mediated cytosolic [Ca2+] spikes trigger mitochondrial permeability transition and, in turn, cytochrome c release. IP3-induced opening of PTP is dependent on a privileged Ca2+ signal transmission from IP3 receptors to mitochondria. After the decay of Ca2+ spikes, resealing of PTP occurs allowing mitochondrial metabolism to recover, whereas activation of caspases is triggered by cytochrome c released to the cytosol. This organization provides an efficient mechanism to establish caspase activation while mitochondrial metabolism is maintained to meet ATP requirements of apoptotic cell death. Introduction Increases of [Ca2+] regulate a diverse range of cellular processes from fertilization to death (for review see Clapham, 1995; Thomas et al., 1996; Berridge, 1997; Berridge et al., 1998; Putney, 1998). The control of each process may also utilize multiple Ca2+-regulated elements that sometimes appear to work in different directions. Regulation of cell death by Ca2+ is particularly complex. Large and sustained elevations of cytosolic [Ca2+] ([Ca2+]c) can switch on a number of mechanisms that lead to the disintegration of cells, necrosis or to the ordered form of cell death, apoptosis (Ankarcrona et al., 1995; Khan et al., 1996; McConkey, 1996; Bian et al., 1997; Lemasters et al., 1998; Nicholls and Budd, 1998; Nicotera and Orrenius, 1998; Wang et al., 1999). Alternatively, calcium ions have also been implicated in mechanisms that exert protection against cell death (Bian et al., 1997; He et al., 1997; Zhang et al., 1998). It is not yet understood how Ca2+ turns from a signal for life to a signal for death and how selective activation of apoptotic or necrotic death pathways is ensured by calcium signals. Evidence is emerging that spatial, temporal and amplitude patterns of calcium signals are important in encoding the specificity of cellular responses (for review see Clapham, 1995; Thomas et al., 1996; Berridge, 1997; Berridge et al., 1998; Putney, 1998). Recent studies have demonstrated that mitochondria may discriminate between different spatial and temporal patterns of cytosolic Ca2+ signals due to a local Ca2+ transfer between IP3-regulated intracellular Ca2+ release sites and mitochondrial Ca2+ uptake sites (Rizzuto et al., 1993, 1994, 1998; Pralong et al., 1994; Hajnóczky et al., 1995; Csordás et al., 1999). At the same time, mitochondria were discovered to play an important role in apoptosis by releasing apoptotic factors (cytochrome c: Newmeyer et al., 1994; Liu et al., 1996; Kluck et al., 1997; Yang et al., 1997; and apoptosis-inducing factor: Susin et al., 1996, 1999). Furthermore, a major target of mitochondrial [Ca2+] regulation, the permeability transition pore (PTP), was shown to be involved in the process of releasing apoptotic factors (Petit et al., 1996; Bradham et al., 1998; Marzo et al., 1998; Pastorino et al., 1998). These results prompted us to put forward the hypothesis that the modulation of PTP by physiological Ca2+ signals may provide a means to utilize the spatial and temporal features of calcium signalling in the control of apoptotic cell death. The central component of the apoptotic machinery is a proteolytic system that involves a family of proteases called caspases (Thornberry et al., 1992; Yuan et al., 1993). In cell-free systems induction of caspase activation (Liu et al., 1996) and caspase-mediated apoptosis (Newmeyer et al., 1994) were found to be dependent on the presence of cytochrome c released from mitochondria. Release of cytochrome c from mitochondria occurs during apoptosis and this process is inhibited by the presence of Bcl-2 on these organelles (Kluck et al., 1997; Yang et al., 1997). Cytosolic cytochrome c is an essential component of the apoptosome, which is composed of cytochrome c, Apaf-1 and procaspase-9 (Li et al., 1997). This yields activation of caspase-9, which then processes and activates other caspases to orchestrate the biochemical execution of cells. Activation of caspases and caspase-dependent dismantling of the cell require ATP and so maintained ATP production by mitochondrial metabolism may be important to support execution of the steps of the apoptotic program distal to cytochrome c release. Opening of PTP has been implicated in the release of cytochrome c and apoptosis-inducing factor from mitochondria (for recent review see Skulachev, 1996; Reed, 1997; Green and Kroemer, 1998; Green and Reed, 1998; Duchen, 1999) and in the effect exerted by proapoptotic (e.g. BAX) and anti-apoptotic members of the Bcl-2 family on the release (Hockenbery et al., 1990; Oltvai et al., 1993; Marzo et al., 1998; Pastorino et al., 1998). Opening of PTP is controlled by Ca2+, pH, adenine nucleotides, free radicals and mitochondrial membrane potential (ΔΨm) (for review see Bernardi and Petronilli, 1996), but it has not been elucidated how these factors are involved in PTP-dependent cytochrome c release. Recent data on mitochondrial transport of Ca2+ and solutes are in support of the idea that in health, increases in mitochondrial matrix [Ca2+] ([Ca2+]m) induced by Ca2+ mobilizing agonists do not evoke PTP opening associated with cytochrome c release (Rizzuto et al., 1994; Ichas et al., 1997; Ichas and Mazat, 1998). We postulated that in cells exposed to proapoptotic stimuli (e.g. ceramide, staurosporine) the Ca2+ sensitivity of PTP increases and, in turn, IP3-mediated [Ca2+]m spikes result in PTP opening and cytochrome c release. Since Ca2+ spikes are established by synchronized activation and deactivation of Ca2+ release channels (Hajnóczky and Thomas, 1997), the rise of [Ca2+] rapidly decays and so resealing of PTP may occur. Provided that residual cytochrome c is sufficient to support the electron transport, recovery of mitochondrial metabolism and ATP production may follow closing of the PTP. Here we provide evidence that Ca2+ pulsing or IP3-induced Ca2+ spikes yield PTP opening, cytochrome c release, caspase activation and nuclear apoptosis in cells exposed to proapoptotic stimuli, but not in naive cells. Our data demonstrate that IP3 receptors utilize very small amounts of Ca2+ to evoke PTP opening and establish a transient opening of PTP followed by recovery of mitochondrial energy metabolism. Restoration of mitochondrial ATP production may account for the absence of necrotic cell death. Thus, IP3-mediated [Ca2+] spikes can serve as an efficient and selective activator of the mitochondrial phase of the apoptotic machinery. Results and discussion Ca2+-induced activation of PTP in permeabilized cells exposed to apoptotic agents In order to study activation of PTP by [Ca2+]m signals, we first established simultaneous fluorometric measurements of ΔΨm and [Ca2+]c in suspensions of permeabilized HepG2 cells. Figure 1A shows that addition of Ca2+ pulses (40 μM CaCl2 each) to the permeabilized cells yielded increases of [Ca2+]c which were transient (left panel, black trace). Endoplasmic reticulum (ER) and mitochondrial Ca2+ stores co-operate in intracellular Ca2+ accumulation, but in these experiments the ER Ca2+ATPase inhibitor, thapsigargin (Tg) (2 μM), was added to permeabilized cells to abolish ER Ca2+ uptake. Under this condition the mitochondrial Ca2+ store is expected to play the major role in Ca2+ uptake. This was indeed shown to be the case as the falling phase of the [Ca2+]c increases was reversed after addition of a Ca2+ ionophore, ionomycin or uncoupler (Figure 1A, left panel) and it was prevented by inhibitors of mitochondrial Ca2+ uptake sites, ruthenium red (RuRed; Figure 1A, right panel) or Ru360 (2 μM, n = 3). Figure 1A shows that relatively small and reversible decreases of ΔΨm were associated with the falling phase of cytosolic [Ca2+] signals evoked by Ca2+ addition whereas complete dissipation of ΔΨm was brought about by uncoupler (Figure 1A, left, black trace). In the presence of RuRed Ca2+ pulses failed to evoke depolarizations, showing that Ca2+-induced depolarizations are distal to mitochondrial Ca2+ uptake. Depolarizations can be explained by the fact that mitochondrial Ca2+ uptake occurs at the expense of ΔΨm. Activation of the PTP by accumulated Ca2+ may also yield loss of ΔΨm but Ca2+-induced depolarization in naive cells was not sensitive to inhibitors of PTP (Figure 1C, see below). Figure 1.Ca2+-induced ΔΨm and [Ca2+]c responses in permeabilized cells exposed to proapoptotic stimuli. Simultaneous measurements of ΔΨm and [Ca2+]c were carried out in suspensions of permeabilized HepG2 cells using a membrane potential probe, JC1, and a Ca2+ tracer, fura2FF, respectively. (A) Effect of C2 on Ca2+-induced mitochondrial depolarization and mitochondrial Ca2+ uptake. C2 (40 μM, purple) or solvent, DMSO (black), were added 180 s prior to Ca2+ pulsing to the permeabilized cells supplemented with succinate (2 mM). The same measurements were carried out in the absence (left panel) and presence of ruthenium red (2 μM RuRed added 60 s prior to Ca2+, right panel). Other additions: CaCl2, 40 μM CaCl2 each; Iono, 10 μM ionomycin; Unc, 1 μM FCCP + 2.5 μg/ml oligomycin. Since RuRed resulted in a large, artifactual rise in RJC1 a compressed scale shows ΔΨm after addition of RuRed. Traces are shown from separate incubations using the same cell preparation (n = 3). (B) Time course of C2 effect on rates of Ca2+-induced depolarization and Ca2+ uptake. The effect of 10 min pretreatment with C2-dihydroceramide (iC2) is also shown (blue symbols). Experiments were carried out as shown in (A) and the rates of depolarization and Ca2+ uptake were calculated using the responses given to the third Ca2+ pulse. Rates represent the mean ± SEM of values from five or six separate experiments. (C) Effect of CSA on Ca2+-induced mitochondrial depolarization and Ca2+ uptake. Measurement was carried out as described for (A) in permeabilized cells exposed to 40 μM iC2 (blue), 40 μM C2 (purple) or DMSO (black) for 180 s prior to Ca2+ addition. Incubations were carried out in the absence (left) and presence (right) of CSA (1 μM added 240 s prior to Ca2+). Traces are shown from separate incubations using the same cell preparation (n = 5). (D) Effect of staurosporine (Stauro, 1 μM) on Ca2+-induced mitochondrial depolarization and Ca2+ uptake. Pulses of CaCl2 (10 μM each) were added to naive (black), C2-pretreated (180 s, purple) and Stauro-pretreated (60 min, red) cells supplemented with malate (1 mM)/glutamate (5 mM) and oligomycin (2.5 μg/ml). In the presence of oligomycin mitochondria could not utilize ATP added to the medium and Ca2+ release associated with Ca2+-induced dissipation of ΔΨm was augmented. Traces from three separate incubations using the same cell preparation (n = 4). Download figure Download PowerPoint In order to study whether apoptotic stimuli affect [Ca2+]m and ΔΨm responses given to Ca2+ addition, cells were exposed to ceramide (C2) or staurosporine (Obeid et al., 1993; Bertrand et al., 1994). These agents have been described to involve mitochondria in the apoptotic program and to exert their effect in subcellular models as well (e.g. Bossy-Wetzel et al., 1998; Garcia-Ruiz et al., 1997; Gudz et al., 1997; Yang et al., 1997). Preincubation of the permeabilized cells with C2 for 180 s did not change the steady state [Ca2+]c or ΔΨm ([Ca2+]c: 0.75 ± 0.02 and 0.76 ± 0.01 ratio units; ΔΨm: 2.26 ± 0.17 and 2.26 ± 0.2 ratio units in cells incubated in the absence and presence of C2, n = 6) but elicited a large decrease of the Ca2+-induced mitochondrial Ca2+ uptake and potentiated the Ca2+-induced fall of ΔΨm (Figure 1A left, purple traces). The enhancement of Ca2+-induced depolarization by C2 was very small at the first addition of Ca2+ but it became large during the second and third pulses (Figure 1A). No depolarization was observed in C2-pretreated cells if mitochondrial Ca2+ uptake was inhibited by RuRed (Figure 1A right, purple traces). These results show that mitochondrial Ca2+ uptake caused larger depolarizations in C2-treated cells. Furthermore, this difference between naive and C2-treated cells is not due to stimulation of mitochondrial Ca2+ uptake by C2. Preincubation of the permeabilized cells with C2 for 10 s was sufficient to alter Ca2+-induced [Ca2+]c and ΔΨm responses (rate of [Ca2+]c decay measured after the second Ca2+ pulse: 69.7 ± 3.7% of control, P <0.001; rate of depolarization: 184.3 ± 26.5% of control, P <0.01, n = 6) and maximal effect of C2 was reached within 10 min of preincubation (Figure 1B). Figure 1 also shows that C2-dihydroceramide (iC2), a structural analog of ceramide that does not cause apoptosis, failed to affect Ca2+-induced mitochondrial Ca2+ uptake and loss of ΔΨm (Figure 1C), whereas a chemically unrelated activator of the apoptotic program, staurosporine, exerted C2-like effects (Figure 1D). Taken together, these data indicate that exposure of permeabilized cells to proapoptotic stimuli alters mitochondrial Ca2+ handling. Although prolonged treatment with apoptotic agents is known to yield mitochondrial damage including release of cytochrome c (e.g. Bossy-Wetzel et al., 1998; Garcia-Ruiz et al., 1997; Yang et al., 1997), our data suggest that an effect of these agents on mitochondrial Ca2+ regulation occurs very early. Impaired mitochondrial Ca2+ handling and large depolarizations could be due to a primary effect of apoptotic agents on mitochondrial metabolism, yielding decreased H+ extrusion. Notably, inhibitory effects of C2 on mitochondrial electron transport chain complex III have been reported (Garcia-Ruiz et al., 1997; Gudz et al., 1997). This mechanism could contribute to the effects of C2 and staurosporine that we observed in permeabilized cells supplemented with succinate (complex II substrate, Figure 1A–C) or malate/glutamate (complex I substrate, Figure 1D). To prevent the effects of C2 that are dependent on respiratory chain activity, in subsequent experiments endogenous substrates were depleted by addition of ADP (100 μM) and then the cells were provided only with ATP and ATP regenerating system. Under this condition generation of ΔΨm may be due to proton extrusion in the reverse mode of the ATP synthase. As expected, addition of an ATP synthase inhibitor, oligomycin (2.5 μg/ml), caused dissipation of ΔΨm in permeabilized cells provided only with ATP and ATP regenerating system (half-life of depolarization was 33 ± 4 s, n = 2) whereas no oligomycin-induced depolarization was observed when mitochondria were energized with succinate or malate/glutamate (n = 4). Since C2 potentiated the Ca2+-induced depolarization and inhibited Ca2+-induced mitochondrial Ca2+ uptake in cells provided only with ATP (rate of depolarization: 218 ± 5% of control, P <0.001; rate of [Ca2+]c decay: 34 ± 5% of control, P <0.001, n = 3), inhibition of the electron transport chain can not account for these effects. Alternatively, apoptotic agents may facilitate activation of PTP complex by accumulated Ca2+. In contrast to the Ca2+-induced depolarizations obtained in naive cells that were synchronized to the rapid falling phase of the [Ca2+]c response and were reversed subsequently, the depolarizations in cells exposed to apoptotic agents carried on when [Ca2+]c displayed a slow decrease or continuous increase (Figure 1, black versus purple or red traces). This result is consistent with the idea that Ca2+ led to activation of large pores, which allowed dissipation of the ionic gradients. Activation of PTP by [Ca2+]m can be inhibited by cyclosporin A (CSA), a drug causing dissociation of cyclophilin D from PTP on the luminal side (Broekemeier et al., 1989; Halestrap and Davidson, 1990). Figure 1C shows that CSA (1 μM) did not exert a major effect on Ca2+-induced changes of ΔΨm and [Ca2+]c in naive cells (compare black traces, left and right panels) but abolished the effects of C2 (purple traces, left and right panels). FK506 (10 μM), which affects extramitochondrial immunophilins but fails to affect mitochondrial cyclophilin D, did not decrease the effect of C2 (0 and 8% inhibition of depolarization in two experiments). An additional tool to target the permeability pore is bongkrekic acid, which binds to the adenine nucleotide translocator and, in turn, interferes with activation of the PTP complex. Since ADP supports the binding and inhibitory effect of bongkrekic acid (Klingenberg et al., 1983; Brustovetsky and Klingenberg, 1996), the ATP regenerating system was omitted from the medium to prevent extramitochondrial phosphorylation of ADP. To our surprise, bongkrekic acid (300 nM–1 μM added 5 min before C2) enhanced Ca2+-induced depolarization in naive cells (ΔΨm after Ca2+ pulses: 50.9% of control) but this drug also inhibited Ca2+-induced depolarization in C2-treated cells (ΔΨm after Ca2+ pulses: 155.5% of control, n = 2). Measurements of [Ca2+]c showed that bongkrekic acid facilitated the decay of the Ca2+-induced [Ca2+]c increases in C2-treated cells, but not in naive cells (n = 2), suggesting that sequestration of Ca2+ in the mitochondria was supported by the drug. Taken together, these data provide evidence that activation of PTP by accumulated Ca2+ accounts for the altered ΔΨm and [Ca2+]c responses of permeabilized cells exposed to C2. Using the three-pulses protocol of Ca2+ addition shown in Figure 1, 30–40 μM pulses of CaCl2 were required to obtain CSA-sensitive depolarization in C2-pretreated cells supplemented with succinate and ATP (Figure 1A–C). Opening of PTP required lower doses of Ca2+ (10–20 μM pulses of CaCl2, Figure 1D) when electrons were provided to complex I (using malate/glutamate as substrate) rather than to complex II (using succinate as substrate) as described recently by Fontaine et al. (1998). Nevertheless, elevations of [Ca2+]c evoked by addition of Ca2+ pulses were in the range 5–20 μM, which are substantially higher than the global increases of [Ca2+]c in cells stimulated with Ca2+ mobilizing stimuli. Recent reports suggest a model wherein activation of caspases is proximal to mitochondrial depolarization and cytochrome c release. Bossy-Wetzel et al. (1998) demonstrated that activation of a DEVD-specific caspase occurs before the dissipation of ΔΨm. Most recently, activation of death receptors was described to yield activation of caspase-8, which cleaves BID and, subsequently, truncated BID causes cytochrome c release from the mitochondria (Li et al., 1998; Luo et al., 1998). In our experiments washout of the cytosol prior to C2 addition did not prevent Ca2+-induced PTP opening (an example is shown in Figure 3A), providing evidence that soluble cytosolic factors do not play an obligatory role in Ca2+-induced PTP opening in cells pretreated with C2. Experiments were also carried out in the presence of caspase inhibitors, DEVD-CHO (50 μM), which inhibits most caspases including caspase-3, caspase-8 and caspase-9, or zVAD–FMK (50 μM), which inhibits all 10 known caspases. Caspase inhibitors failed to inhibit Ca2+-induced depolarization (C2 + DEVD-CHO: 112%; C2 + zVAD–FMK: 100 and 106% of C2 without caspase inhibitors, n = 2) or to decrease the duration of [Ca2+]c transients induced by Ca2+ pulsing in C2-pretreated cells (C2 + DEVD-CHO: 104%; C2 + zVAD–FMK: 92 and 103% of C2 without caspase inhibitors, n = 2). These results suggest that activation of caspases is not required for facilitation of Ca2+-induced opening of PTP by proapoptotic stimuli in this system. Figure 2.Reversibility of Ca2+-induced PTP opening in permeabilized cells exposed to C2. Experimental conditions were as in Figure 1(A). In the experiments shown in (A) five pulses of 30 μM CaCl2 were added at 25 s intervals, in (B) 120 μM CaCl2 was added continuously (75 s). Other additions: RuRed, 2 μM ruthenium red; C2, 40 μM C2-ceramide; EGTA, 200 μM EGTA; Unc, 1 μM FCCP + 2.5 μg/ml oligomycin. Data are representative of five (A) and three experiments (B). Download figure Download PowerPoint Another fundamental question of the mechanism underlying Ca2+-induced opening of PTP in cells exposed to apoptotic agents is whether resealing of PTP occurs after decay of the [Ca2+] rise. In living cells orchestrated Ca2+ transport through intracellular membranes and plasma membrane is responsible for deactivation of [Ca2+]c transients. Since Ca2+ accumulation by the mitochondria represents the only mechanism to decrease [Ca2+]c in permeabilized cells treated with Tg, to bring about the decay of the Ca2+ response, a Ca2+ chelator, EGTA (Figure 2A), or an inhibitor of mitochondrial Ca2+ accumulation, RuRed (Figure 2B) were added. Mitochondrial membrane potential that had been dissipated in response to Ca2+ was rapidly regenerated under both conditions, suggesting that PTP became closed and mitochondrial metabolism recovered. Repolarization of the mitochondria was not inhibited by oligomycin (rate of repolarization was 0.39 ± 0.06 ratio units/min, n = 3 and 0.39 ± 0.06 ratio units/min, n = 5 in cells incubated in the absence and presence of oligomycin), providing evidence that proton extrusion was not due to the reversed mode of the ATPase but to activation of the electron transport chain. These results show that a transient exposure to high Ca2+ leads to a transient activation of PTP in permeabilized cells exposed to apoptotic agents. Figure 3.Release of cytochrome c and activation of caspases evoked by Ca2+ pulses in permeabilized cells exposed to proapoptotic stimuli. (A–C) Cytochrome c release associated with Ca2+-induced opening of PTP. Measurements of ΔΨm in suspensions of washed permeabilized cells (A), immunoblots (B) and cumulative data of cytochrome c measurements (C) in ‘cytosol’ samples generated at the end of the ΔΨm recordings shown in (A). Additions: C2, 40 μM C2-ceramide; Cyt c, 400 nM cytochrome c; CSA, 1 μM cyclosporin A; CaCl2, 10 μM CaCl2 each. Data are from experiments carried out in two separate cell preparations and each represents the mean of duplicates. (D and E) Caspase activation associated with Ca2+-induced opening of PTP. Fluorometric assay of DEVD–AMC cleavage in cytosol extracts prepared at the end of measurements of ΔΨm (shown in Figure 1D). Data in (D) are representative of three different experiments, whereas DEVD–AMC cleavage normalized to the activity obtained with C2+Ca2+ is shown as mean ± SEM in (E) (n = 3–4). The inset shows the dose–response for inhibition of DEVD–AMC cleavage by DEVD-CHO. Download figure Download PowerPoint Calcium-induced activation of cytochrome c release and caspase activation in permeabilized cells exposed to apoptotic stimuli The next set of experiments was carried out to determine whether activation of PTP by [Ca2+]m yields cytochrome c release and caspase activation in C2-pretreated cells. To remove cytochrome c present in the cytosol before Ca2+ pulsing, the cells were washed after permeabilization. First, we repeated the measurements of ΔΨm shown in Figure 1 using washed permeabilized cells. Figure 3A shows that Ca2+ addition induced extensive depolarization in C2-pretreated cells that was prevented by CSA. This result shows that washout of the cytosol did not prevent activation of PTP by Ca2+ pulses in permeabilized cells exposed to C2. It is noteworthy that addition of cytochrome c (400 nM) to the permeabilized cells was not found to evoke depolarization, showing that the PTP opening in response to Ca2+ is not due to cytochrome c release. At the end of each fluorescence measurement samples of the cytosolic fraction (14 000 g supernatant) were prepared to determine cytochrome c release corresponding to the ΔΨm responses. In order to prevent Ca2+ from affecting cytochrome c release during isolation of the cytosol, the permeabilized cells were supplemented with Ca2+ chelator (1 mM EGTA) after exposure to Ca2+ pulses. EGTA resulted in resealing of PTP, and in turn, recovery of ΔΨm, as shown in Figure 2A. Figure 3B and C shows data from Western blot analysis of the cytosol samples. The cytochrome c band was identified using molecular weight markers and also with exogenous cytochrome c standard added to the permeabilized cells during the fluorescence measurement (Figure 3B). Cytosolic cytochrome c was present in the same amount in samples derived from suspensions of naive cells exposed to Ca2+ or from C2-pretreated cells without Ca2+ addition, whereas a several-fold larger level was detected in samples derived from C2-pretreated cells exposed to Ca2+ (Figure 3B and C). Cytochrome c release was inhibited by pretreatment with CSA (Figure 3C). Cytochrome c is known to activate caspases in the cytosol (Kluck et al., 1997; Yang et al., 1997). A fluorometric assay (Nicholson et al., 1995; Hampton et al., 1998) was used to determine caspase activity in cytosol fractions prepared from suspensions of permeabilized cells exposed to C2 and Ca2+ as shown in Figure 1. Figure 3D shows the time course of DEVD–AMC cleavage and the inhibitory effect exerted by the caspase inhibitor, DEVD-CHO (IC50 ∼ 1 nM). Addition of exogenous cytochrome c (400 nM) to the cell suspension brought about a large increase in DEVD–AMC cleavage (Figure 3D and E). This result reconfirmed the observation that cytochrome c evokes caspase activation. Exposure of C2-pretreated permeabilized cells to Ca2+ pulses yielded stimulation of cytosolic DEVD–AMC cleavage (Figure 3D and E, P <0.001, n = 4). This effect was as large as the effect of exogenous cytochrome c (Figure 3D and E). Taken together, these experiments demonstrate that Ca2+-induced opening of PTP in C2-pretreated cells was associated with cytochrome c release and released cytochrome c led to activation of caspases. As noted above (see Figure 2), regeneration of ΔΨm follows the decay of the [Ca2+] rise, suggesting that release of cytochrome c does not prevent repolarization of the mitochondria. Thus, depolarizations preceding cytochrome c release may be hardly detectable at the level of the population average in cell populations that exhibit asynchronous [Ca2+]c transients and PTP openings. Furthermore, it has been unresolved how ATP requirements of the energy-dependent apoptotic program would be satisfied if cytochrome c release was associated with a compromised metabolic function of the mitochondria. In addition to the impaired ATP production, opening of PTP would lead to mitochondrial ATP consumption due to the reverse operation of the F1F0-ATPase. Our findings that mitochondria reseal after the decay of the [Ca2+] rise and that the residual cytochrome c content of the mitochondria is sufficient to maintain mitochondrial metabolism support the idea that mitochondria can provide fuels for the steps of the apoptotic program distal to cytochrome c release. As such, [Ca2+] spikes may serve as an optimal signal to trigger the mitochondrial phase of apoptosis. IP3-induced activation of PTP in C2-treated permeabilized cells As reported previously in other cell types (e.g. Rizzuto et al., 1993, 1994, 1998; Hajnóczky et al., 1995; Csordás et al., 1999), large and rapid increases of [Ca2+]m were found in HepG2 cells in association with [Ca2+]c signals elicited by activation of the IP3-linked calcium signaling pathway through P2y receptors with ATP (10–200 μM, data not shown). To determine whether the cells exposed to proapoptotic stimuli respond to IP3-mediated [Ca2+]m signals by mitochondrial permeability transition and cytochrome c release, experiments were carried out to test IP3-induced ΔΨm responses in suspensions of naive and C2-treated permeabilized cells, respectively. Thapsigargin was omitt