Mechanistically Distinct Steps in the Mitochondrial Death Pathway Triggered by Oxidative Stress in Cardiac Myocytes
Oxidative stress plays an important role in the pathogenesis of cardiovascular diseases. In the present study, we characterize three distinct phases of the H2O2-induced response, which leads to loss of mitochondrial membrane potential (ΔΨm) and subsequent cell death in cultured cardiac myocytes. (1) Priming: After H2O2 exposure (100 μmol/L), cells maintain a constant ΔΨm for the cell-to-cell specific latency but at the same time undergo progressive changes in inner mitochondrial membrane structure (swelling and loss of cristae by electron microscopy). An increase of matrix calcium is required, but not sufficient, for this process. (2) Depolarization: Priming is followed by sudden depolarization of ΔΨm, which is mediated by mitochondrial permeability transition pore opening, as evidenced by the concomitant release of calcein from mitochondria. This process is rapid (<4 minutes), complete, and irreversible. The duration of depolarization is constant and does not depend on the length of the priming process in any given cell. (3) Fragmentation: Along with massive mitochondrial swelling and release of cytochrome c into the cytoplasm, cells undergo surface membrane alterations, such as exposure of phosphatidylserine and eventual loss of membrane integrity and cellular fragmentation. Thus, oxidant stress elicits reproducible and stereotyped responses in cardiac cells. The priming phase, during which mitochondria undergo major ultrastructural alterations but remain functional, represents a particularly attractive target for intervention in the prevention of cell death.
Oxidative stress triggers the mitochondrial death pathway, thereby causing a variety of human diseases. It plays a central role in the pathogenesis of a number of cardiovascular diseases, such as ischemic heart disease, heart failure, and atherosclerosis. The mitochondrial death pathway sequentially features the loss of mitochondrial membrane potential (ΔΨm), the release of toxic proteins into the cytoplasm, and caspase activation.1,2 The loss of ΔΨm plays a key role in the cascade: several lines of evidence indicate that mitochondrial depolarization is closely associated with dispersion of cytochrome c and hence apoptosis.3,4 The prevailing concept is that ΔΨm is dissipated by the opening of a nonspecific large-conductance channel in the inner membrane known as the “permeability transition pore” (PTP).5 Calcium overload and oxidant stress favor PTP opening,6,7 and drugs that block PTP inhibit both necrotic and apoptotic cell death,8–10 but little more is known with certainty. We have recently reported that activation of mitochondrial ATP-sensitive potassium (mitoKATP) channels can inhibit early steps in the mitochondrial apoptotic pathway triggered by oxidative stress.11,12 More detailed understanding of the mechanisms of mitoKATP-mediated protection requires dissection of the mechanisms by which ΔΨm loss is triggered and mediated.
In the present study, we find that oxidant stress produces a stereotyped progression of cellular changes in cardiac myocytes. Upon exposure to oxidant stress, major changes take place in mitochondria, and we call this phase “priming”: mitochondria undergo progressive morphological changes that require matrix calcium overload, but ΔΨm remains unchanged. Next follows a sudden dissipation of ΔΨm mediated by opening of PTP (“depolarization” phase). Finally, cells undergo surface membrane alterations and eventually break up into smaller fragments (“fragmentation” phase). Precise characterization of the sequence of events is of crucial importance in the establishment of therapeutic strategies against diseases in which oxidative stress is involved.
Materials and Methods
All procedures were performed in accordance with The Johns Hopkins University animal care guidelines, which conform to the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health.
Chemicals and Reagents
All of the chemicals and reagents were purchased from Sigma, unless otherwise stated.
Primary Culture of Neonatal Rat Cardiac Ventricular Myocytes
Cardiac ventricular myocytes were prepared from 1- to 2-day-old Sprague-Dawley rats (Zivic Laboratories Inc, Pittsburgh, Pa) and cultured as previously described.11
Loading of Cells With Fluorescent Indicators
To monitor ΔΨm, cells were loaded with tetramethylrhodamine ethyl ester (TMRE) (Molecular Probes) at 100 nmol/L, 37°C, 15 minutes.
To measure a change in mitochondrial permeability, cells were loaded with calcein acetomethoxy ester (calcein AM) (Molecular Probes), as previously described13; briefly, cells were loaded for 15 minutes with 2 μmol/L calcein AM at room temperature in the presence of 4 mmol/L cobalt chloride to quench cytosolic and nuclear calcein loading.
To monitor mitochondrial matrix calcium level ([Ca2+]m), cells were loaded with 2 μmol/L rhod-2 AM (Molecular Probes) at 4°C for 1 hour, followed by incubation at 37°C for 30 minutes, as previously described14; this two-step loading facilitates the selective accumulation of rhod-2 in mitochondrial matrix. To analyze [Ca2+]m level with minimal contamination by cytosolic rhod-2 signals, regions of interest were drawn over individual mitochondria (defined by punctate regions of calcein accumulation) for quantitative image analysis.
The detection of apoptotic cells and necrotic cells was carried out by staining with annexin V and propidium iodide (PI), according to manufacturer’s instructions (Vibrant apoptosis kit No. 2; Molecular Probes).
Construction of Adenoviral Vector Carrying Green Fluorescent Protein (GFP)-Tagged Cytochrome c
Full-length cDNA sequence of GFP-tagged cytochrome c (CcGFP)3 was cloned into the adenovirus shuttle vector (AdCMV-CcGFP), which is driven by a cytomegalovirus promoter. Detailed methods of adenovirus vector construction have been described previously.15,16 Briefly, adenovirus vectors were generated using Cre-lox recombination of purified Ψ5 viral DNA and shuttle vector AdCMV-CcGFP. The recombinant vectors were expanded and purified using cesium chloride gradients, yielding concentrations of about 6×1010 plaque-forming units (PFUs) per milliliter.
In Vitro Transduction
Neonatal rat ventricular myocytes were transduced with the adenovirus AdCMV-CcGFP on day 6 in culture. The multiplicity of infection that gave the best results was ≈4. Transductions were carried out in culture medium (Opti-MEM; Life Technologies) for 1 hour at 37°C. After the 1-hour transduction time, the myocytes were washed with virus-free medium and incubated at 37°C and 5% CO2. The experiments were carried out at least 48 hours after transduction.
Cells plated on glass-bottom dishes were loaded with fluorescent probes as described above. After washing the dyes, cells were placed in phenol red-free DMEM supplemented with 25 mmol/L HEPES and were kept at constant temperature using a heater platform on a microscope stage. Cells were illuminated (488-nm line and 568-nm line of a krypton/argon laser), and images were taken by confocal microscopy (UltraVIEW; PerkinElmer). Time-lapse confocal imaging was carried out with various intervals between frames.
Transmission Electron Microscopy
Cells were fixed in 2% glutaraldehyde and subjected to postfixation with 1% osmium tetroxide in 0.1 mol/L cacodylate. Formvar-coated copper grids were stained with 2% uranyl acetate, which was followed by subsequent dehydration with series of ethanol. The samples were embedded with epoxy resin and then polymerized in a dry oven at 60°C. Ultrathin sections (70-nm thick) from the embedded samples were imaged by use of a Philips CM 120 transmission electron microscope.
Quantitative image analysis was performed using image analysis software (ImageJ).17
Fluorescence-Activated Cell Sorter (FACS) Analysis
For FACS analysis of ΔΨm, TMRE-loaded cells were harvested by trypsinization at the end of experimental protocols and analyzed by FACScan (20 000 cells/sample) (Becton Dickinson). The fluorescence intensity of TMRE was monitored at 582 nm (FL-2 channel). FACS data were analyzed using WinMDI software.18
Whole lysates of the cells were obtained by directly lysing the cells with Laemmli buffer (BioRad), and dissolved proteins were subjected to immunoblot. Subcellular fractions of cells were prepared as described elsewhere.4 Briefly, cells were washed in PBS and incubated for 30 minutes on ice in lysis buffer (68 mmol/L sucrose, 200 mmol/L mannitol, 50 mmol/L KCl, 1 mmol/L EGTA, 1 mmol/L EDTA, 1 mmol/L DTT, and 1× Complete protease inhibitor [Roche]). Cells were then homogenized with an ultrasound sonicator and centrifuged at 4°C (800g) to remove nuclei, unbroken cells, and debris. The supernatant was centrifuged at 14 000g for 15 minutes. The supernatant (cytosol) and pellets (mitochondria) were stored at −70°C for immunoblot analysis. Immunoblot was performed as previously described.11 Primary antibodies against GFP and cytochrome c oxidase subunit IV (COX IV) were purchased from Clontech and Molecular Probes, respectively.
All quantitative data are presented as mean±SEM. Multiple comparisons among groups were carried out by one-way ANOVA with Fisher’s least significant difference as the post hoc test, unless otherwise stated. A level of P<0.05 was accepted as statistically significant.
We first aimed to characterize the process of ΔΨm change and specifically to determine whether PTP opening is involved. Myocytes were coloaded with a red fluorescent ΔΨm indicator, TMRE, and the green fluorescent indicator calcein. Fluorescent calcein concentrates in the mitochondrial matrix, such that the loss of punctate calcein fluorescence can be used to detect the opening of the PTP in situ.13 As shown in Figure 1A, calcein fluorescence colocalized with the mitochondrial marker TMRE, confirming the mitochondrial accumulation of calcein in myocytes under basal conditions. During oxidant stress, time-lapse confocal microscopy enabled us to track the subcellular localization of the two fluorescent indicators (Figure 1B and online Movie 1, available in the data supplement at http://www.circresaha.org). After a certain waiting time or latency (see frames 20 to 38 minutes in Figure 1B), during which no remarkable changes in fluorescence were observed, cardiac cells underwent a rapid loss of ΔΨm within a short period of time (≈4 minutes). Importantly, the kinetics of ΔΨm loss were synchronous with calcein release at the level of individual mitochondria (eg, arrows in Figure 1B, 30-minute frame).
From these image sequences, we quantitatively assessed the kinetics of ΔΨm loss and PTP opening in individual cells. Regions of interest were drawn over part of an individual cell (a typical example is shown in Figure 1B, 20-minute frame), and fluorescence signals within these regions were collected over time. ΔΨm was monitored by mean TMRE brightness within the regions. To assess calcein localization, we measured the standard deviation (SD) of the intensity of all the green pixels within the same regions of interest. SD is high when fluorescence is punctate and low when fluorescence is diffuse. Figure 1C shows the time courses of fluorescence obtained from a representative cell. Exposure to H2O2 did not immediately result in ΔΨm loss; instead, there was a variable latency (38 minutes in this particular cell), after which ΔΨm started to decrease. Once it began, this rapid loss of ΔΨm was complete within 4 minutes. The time course of ΔΨm dissipation paralleled the loss of calcein localization (ie, the decline in green fluorescence SD), whereas the mean calcein fluorescence remained constant, indicating that the changes reflect redistribution of calcein within the cell rather than leakage to the outside. Figure 1D examines the temporal relationships among duration and latency of ΔΨm loss and calcein redistribution. The left panel plots the time required for TMRE loss in each cell against the time required for calcein release in the same cells; “duration” was defined as the time required for 50% loss of TMRE fluorescence or green SD. The tight positive correlation between these two parameters (r=0.758, P<0.03) supports the notion that loss of ΔΨm coincides with PTP opening. In contrast, the right panel of Figure 1D shows that the durations of TMRE loss did not vary with latency: ΔΨm dissipation occurred over <3 minutes, regardless of the delay with which it began.
Figure 1E summarizes the dependence on temperature of latency and duration of ΔΨm loss. The latency was significantly shorter at 37°C compared with room temperature (25.0±1.6 minutes at 37°C, 42.8±2.4 minutes at room temperature; P<0.01, n=8, paired t test), whereas the duration of ΔΨm loss was temperature-independent (1.94±0.15 minutes at 37°C, 2.13±0.18 minutes at room temperature; NS, n=8). Thus, latency is highly sensitive to temperature, whereas duration is not.
Taken together, these data reveal that oxidant stress elicits a reproducible and stereotyped response consisting of rapid ΔΨm loss after a variable latency. The ΔΨm dissipation occurs hand in hand with the redistribution of calcein from mitochondria to cytosol, signifying that PTP opening underlies ΔΨm loss. These results begin to suggest two phases in the response to oxidant stress: first, priming, in which myocytes prepare for ΔΨm loss, and second, depolarization, in which ΔΨm dissipation occurs rapidly.
If the multistep concept is genuine, one might expect a shift in the fraction of myocytes exhibiting specific phenotypic properties (eg, fluorescence or morphology) over time. To examine large populations of myocytes, we used FACS analysis. Figure 2A shows representative histograms of TMRE fluorescence. Based on the distributions of TMRE fluorescence in control cells and H2O2-treated cells, we defined three populations (Figure 2A: populations I, II, and III). Population I consisted of cells with intact ΔΨm, II of cells with dissipated ΔΨm, and III of cells or cell fragments with very low fluorescence. The majority of cells in the control group belonged to I (viable cells). Exposure to H2O2 shifted the predominant population from I to II and III. A mitochondrial uncoupler, carbonylcyanide p-(trifluoromethoxy) phenylhydrazone (FCCP; 1 μmol/L) induced immediate and complete dissipation of ΔΨm to produce a single log-normal distribution comparable to II, confirming that II consists of intact cells with fully depolarized ΔΨm. The even lower fluorescence in III thus arises from fragments of dead cells. The existence of population III points to a third process, that of fragmentation, after priming and depolarization.
We examined the concentration-response relationship of H2O2. Figure 2B, top panel, plots the percentage of cells that retain a high TMRE fluorescence. The lower concentration of H2O2 resulted in fewer cells losing TMRE fluorescence; nevertheless, those cells that did not lose ΔΨm underwent the same three-step progression in an all-or-none manner, since the positions of the peaks in all populations were the same regardless of H2O2 concentration (data not shown). Furthermore, to see whether the priming process is irreversible, we washed out H2O2 before loss of ΔΨm (Figure 2B, bottom panel). Cells were exposed to 100 μmol/L H2O2 for the indicated duration followed by washout (total period for all groups was 60 minutes). The results suggest that the cells were committed to an irreversible process as early as 5 minutes after H2O2 exposure.
Figure 2C shows density plots of TMRE fluorescence versus side-scatter (SSC), an index of cellular morphology, at various times after the addition of 100 μmol/L H2O2. Corresponding populations were identified and outlined in gray (populations I, II, and III). In the first 20 to 30 minutes of exposure to H2O2, cells moved downward within population I: SSC decreased while TMRE fluorescence was maintained. Afterward, cells started to shift into populations II and III, retaining a low side scatter while progressively losing TMRE fluorescence. These data reveal that the priming phase is not entirely silent: rather, it is accompanied by distinctive changes of cell morphology as reported by the fall of SSC. Thereafter, the predominant changes are in the level of TMRE fluorescence rather than in SSC. No decline in SSC was observed with FCCP-induced ΔΨm dissipation (data not shown), suggesting that the changes in SSC are specific to the response to oxidants.
Because a decrease in SSC may indicate mitochondrial swelling,19 we examined mitochondrial morphology with transmission electron microscopy (Figure 2D). Normal mitochondria were elongated and had numerous narrow pleomorphic cristae evident as electron-transparent areas in a contiguous electron-dense matrix (Figure 2D, 0 minutes). H2O2-treated cells exhibited a series of morphological changes in mitochondria, notably progressive swelling and loss of cristae. These changes were observed even during the first 10 to 20 minutes (eg, Figure 2D, 20 minutes), while ΔΨm was still preserved, and were eventually followed by massive swelling and rupture of mitochondria. Quantitative assessment of mitochondrial morphology (Figure 2E) reveals progressive increases in cross-sectional area (filled circles, top panel) and circularity (open circles). The latter parameter, defined as 4π · [area/(perimeter)2], quantifies the roundness of mitochondria. During H2O2 exposure, mitochondria approach a perfect circularity value of 1.0. Meanwhile, the number of cristae per 100 nm falls sharply and progressively (Figure 2E, bottom panel). Thus, the observed decrease in SSC most likely represents progressive swelling and loss of cristae in mitochondria.
Dysregulation of [Ca2+]m is recognized as a prelude to PTP opening.6,7 We therefore explored the possibility that changes in [Ca2+]m may underlie the priming process. Cells coloaded with calcein and a [Ca2+]m indicator, rhod-2, were subjected to time-lapse confocal microscopy. Upon exposure to H2O2, mitochondrial rhod-2 fluorescence started to increase gradually, with a more precipitous rise after 28 to 30 minutes (Figure 3A, online Movie 2). Several minutes later (Figure 3A, 34 to 38 minutes), calcein was released from the matrix, indicative of PTP opening (see a typical mitochondrion indicated by a white arrow in 34-minute frame). Figure 3B shows the time course of the changes of fluorescence in this representative cell. Thus, priming is accompanied by progressive [Ca2+]m overload. Addition of the Ca2+ chelator BAPTA (50 μmol/L) loaded intracellularly prevented not only the [Ca2+]m overload (Figures 3C and 3D, left panel) but also the decrease in SSC during priming (Figure 3D, right panel, open squares). Similarly, mitochondrial Ca2+ uniporter blocker ruthenium red (RuR; 50 μmol/L) partially prevented [Ca2+]m overload (Figure 3D, left panel) as well as the decrease in SSC (Figure 3D, right panel, open triangles), further supporting the involvement of [Ca2+]m as a critical mediator of the SSC changes. The Ca2+ dependence of priming was further confirmed by experiments in which cells were studied at low external [Ca2+]. This resulted in attenuated [Ca2+]m overload after H2O2 exposure (Figure 3E, left panel) and relatively preserved SSC levels (Figure 3E, right panel). To determine whether cellular calcium overload suffices to dissipate ΔΨm, we tested the effect of ouabain, an inhibitor of the Na+/K+ ATPase. Ouabain (500 μmol/L) elicited a similar level of [Ca2+]m overload to that seen with H2O2 (Figure 3F, left panel) but did not result in a decrease of SSC (Figure 3F, right panel). Moreover, ouabain caused neither ΔΨm depolarization (Figure 3G) nor mitochondrial swelling in electron microscopy (data not shown). We conclude that [Ca2+]m overload is necessary but not sufficient for the priming process.
Next, we studied changes in markers of apoptosis and necrosis. Apoptotic cells undergo translocation of phosphatidylserine from the inner to the outer leaflet of the surface membrane, which can be detected by the measurement of annexin V labeled with a green fluorophore. Meanwhile, PI, a red fluorescent DNA binding dye, is impermeant in viable and apoptotic cells but can stain necrotic cells that lose surface membrane integrity. Figure 4A shows confocal images of cells stained with annexin V (green) and PI (red). Exposure to 100 μmol/L H2O2 for 4 hours resulted in the appearance of both annexin V-positive PI-negative (apoptotic) cells (white arrowheads) and PI-positive (necrotic) cells (white arrows). The time courses of annexin V and PI fluorescence (Figure 4B) show that phosphatidylserine exposure appeared after ≈60 minutes of H2O2 exposure, followed by the loss of membrane integrity, as reported by the increase of PI fluorescence. These data reveal that both apoptotic and necrotic cell death occur under our experimental conditions.
Release of cytochrome c from mitochondria to the cytoplasm is a crucial step in the apoptotic pathway of a cell2 and is suggested to determine the fate of a cell in an all-or-none fashion.4 Cytochrome c release is reportedly synchronous with ΔΨm dissipation,20 but the precise timing and mechanism of the release in cardiac myocytes is unknown. To examine the temporal and spatial relationship between cytochrome c release and ΔΨm depolarization in primary cultured cardiac myocytes, we have monitored both events in real time. To visualize cytochrome c release, we made an adenovirus construct that contains CcGFP. As shown in Figure 5A, CcGFP was predominantly expressed in the mitochondria and colocalized with TMRE, whereas GFP distributed throughout the cytosol. Selective accumulation of transduced CcGFP in mitochondria was further confirmed by immunoblot, showing only a trace amount of expression in the cytosolic fraction (Figure 5B). Cells loaded with TMRE and CcGFP were subjected to time-lapse confocal microscopy (Figure 5C, online Movie 3). In contrast with the rapid drop of TMRE fluorescence (see frames 40 to 44 minutes in Figure 5C), CcGFP release was not apparent within the same time period. However, after ΔΨm depolarization, CcGFP was slowly released from mitochondria to the cytoplasm, as shown by a progressively more diffuse pattern of CcGFP fluorescence over time (see frames 42 to 52 minutes in the bottom row of Figure 5C). Quantitative analysis of these image sequences is shown in Figure 5D. Similar to the analysis in Figure 1, the SD of CcGFP was used as a measure of cytochrome c release; the data indicate the relatively slow kinetics of CcGFP release (with duration of ≈10 minutes), after the loss of ΔΨm. We confirmed by immunoblot that the behavior of CcGFP release is comparable to that of the release of endogenous cytochrome c (Figure 5E). Cytochrome c release became evident after 20 to 30 minutes of H2O2 exposure. Given that this experiment was carried out at 37°C, this time course is comparable to 40 to 60 minutes of H2O2 exposure in Figures 5C and 5D (priming takes almost twice as long at room temperature than at 37°C, as shown in Figure 1E). Intracellularly loaded BAPTA suppressed cytochrome c release (Figure 5E), indicating that [Ca2+]m overload was required upstream of this process.
Priming, Depolarization, and Fragmentation: Mechanistically Distinct Steps in the Mitochondrial Death Pathway
Based on our observations regarding the response to H2O2, we propose a model as follows. H2O2-exposed cells undergo at least three mechanistically distinct steps in the process of ΔΨm loss and subsequent cell death: priming, depolarization, and fragmentation.
After H2O2 exposure, cells maintained an almost constant level of ΔΨm for a certain latency period (Figures 1B and 1C). As shown in Figure 2C, priming is associated with a decrease in SSC within population I, without a corresponding change in ΔΨm. A decrease in spectrophotometric light scattering is classically equated with mitochondrial swelling.21 Likewise, in the FACS analysis, a decrease in SSC has been taken to reflect mitochondrial swelling.19 We demonstrated striking changes in the ultrastructure of mitochondria (swelling and loss of cristae) during the priming period (Figures 2D and 2E). In parallel with those changes in mitochondrial ultrastructure, priming was accompanied by progressive [Ca2+]m overload (Figures 3B and 3C). Both the progressive [Ca2+]m overload and the decrease in SSC were completely blocked by BAPTA (Figures 3C and 3D). However, [Ca2+]m overload elicited by a nonoxidative stress (ouabain) did not induce a decrease in SSC (Figure 3F) or ΔΨm depolarization (Figure 3G). Taken together, [Ca2+]m overload is required, but not sufficient for the SSC changes during priming, which reflects morphological alterations in mitochondria as part of the earliest response to oxidant stress. In agreement with our observations, Scorrano et al22 recently demonstrated dynamic remodeling of mitochondrial cristae during early apoptosis, before ΔΨm loss or cytochrome c release. Mannella et al23 suggested that the inner membrane is not comprised of baffle-like foldings of one continuous surface, but of pleomorphic compartments connecting internal tubular regions (cristae) to each other and to the membrane periphery. This may explain why the loss or disruption of cristae did not affect ΔΨm during priming. Finally, we showed that the duration of priming was quite sensitive to temperature (Figure 1E, left panel), hinting that this process is rate-limited by temperature-sensitive enzymatic and/or metabolic processes.
Once priming is completed, cells underwent irreversible dissipation of ΔΨm, which was accompanied by the opening of PTP (Figures 1B and 1C). This rapid and complete ΔΨm loss corresponds to the transition from population I to II (Figure 2C). Once initiated, the duration of depolarization was always rapid (≈3 minutes, Figure 1D, left panel) regardless of the cell-to-cell variability in latency during the priming phase (Figure 1D, right panel). Moreover, the duration of depolarization was temperature-independent (Figure 1E, right panel), suggesting that this process is not enzyme-associated but rather mediated by a passive mechanism, consistent with the concept of the PTP as a large-conductance nonspecific pore.
Along with more prominent matrix swelling (Figure 2E) and resultant cytochrome c release into the cytoplasm (Figure 5), cells with fully depolarized ΔΨm undergo surface membrane alterations (Figure 4) and eventually lose cellular integrity and become broken and fragmented (population III in Figure 2, a fact that we confirmed by electron microscopy [data not shown]).
Temporal Relationship Between Cytochrome c Release and ΔΨm Depolarization
The relationship between ΔΨm depolarization and cytochrome c release was also investigated in the present study. Green and colleagues showed that the kinetics of cytochrome c release in HeLa cells is rapid (within ≈5 minutes), irreversible, and complete4 and that cytochrome c release precedes ΔΨm depolarization.19 In contrast with their results, the kinetics of cytochrome c release in cardiac myocytes were slower (≈10 minutes) and not complete, and cytochrome c release followed, rather than preceded, ΔΨm depolarization (Figures 5C and 5D). Our data suggest that massive swelling and subsequent rupture of the outer membrane after ΔΨm depolarization result in the passive release of cytochrome c into the cytoplasm. Similar to the progressive changes in mitochondria during priming, cytochrome c release requires calcium overload (Figure 5E). Cytochrome c release in our cardiac myocyte model appears to be merely a downstream consequence of mitochondrial swelling and not a determinant of cell fate.
In the present study, we used H2O2-induced cell death in neonatal cardiac myocytes as a surrogate for cellular injury by ischemia/reperfusion. Although this is a well-accepted model system for studying cell death in a number of cardiovascular diseases, additional studies are needed to determine whether this three-phase progression applies to other initiators of cell death.
We took advantage of several novel approaches to reveal distinct events in the oxidant-induced mitochondrial death pathway. The stereotyped progression of events revealed here (priming, depolarization, and fragmentation) represents a valuable conceptual framework for examining the action of various cardioprotective agents. From first principles, intervention at the level of priming would be particular attractive, in that subsequent catastrophic events would never be initiated. A comparison study,24 which used the framework developed here, examined the effect of pharmacological agents.
This study was supported by the NIH (R37 HL36957) and a Banyu Fellowship in Cardiovascular Medicine (M.A.). E.M. holds the Michel Mirowski, MD Professorship of Cardiology. We thank J. Miake, A. Sidor, C. Cooke, and S.P. Jones for their technical assistance and helpful discussions. We also thank Dr Nieminen (Case Western Reserve University, Cleveland, Ohio) for providing full-length cDNA of CcGFP.
This manuscript was sent to Richard A. Walsh, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
Original received September 11, 2002; revision received November 27, 2002; accepted December 2, 2002.
Green DR, Reed JC. Mitochondria and apoptosis. Science. 1998; 281: 1309–1312.
Heiskanen KM, Bhat MB, Wang HW, Ma J, Nieminen AL. Mitochondrial depolarization accompanies cytochrome c release during apoptosis in PC6 cells. J Biol Chem. 1999; 274: 5654–5658.
Crompton M. The mitochondrial permeability transition pore and its role in cell death. Biochem J. 1999; 341: 233–249.
Crompton M, Costi A, Hayat L. Evidence for the presence of a reversible Ca2+-dependent pore activated by oxidative stress in heart mitochondria. Biochem J. 1987; 245: 915–918.
Lemasters JJ, Nieminen AL, Qian T, Trost LC, Elmore SP, Nishimura Y, Crowe RA, Cascio WE, Bradham CA, Brenner DA, Herman B. The mitochondrial permeability transition in cell death: a common mechanism in necrosis, apoptosis and autophagy. Biochim Biophys Acta. 1998; 1366: 177–196.
Akao M, Ohler A, O’Rourke B, Marbán E. Mitochondrial ATP-sensitive potassium channels inhibit apoptosis induced by oxidative stress in cardiac cells. Circ Res. 2001; 88: 1267–1275.
Petronilli V, Miotto G, Canton M, Brini M, Colonna R, Bernardi P, Di Lisa F. Transient and long-lasting openings of the mitochondrial permeability transition pore can be monitored directly in intact cells by changes in mitochondrial calcein fluorescence. Biophys J. 1999; 76: 725–734.
Hardy S, Kitamura M, Harris-Stansil T, Dai Y, Phipps ML. Construction of adenovirus vectors through Cre-lox recombination. J Virol. 1997; 71: 1842–1849.
Johns DC, Nuss HB, Marbán E. Suppression of neuronal and cardiac transient outward currents by viral gene transfer of dominant-negative Kv4.2 constructs. J Biol Chem. 1997; 272: 31598–31603.
Image Processing and Analysis in Java. ImageJ [software]. Available at: http://rsb.info.nih.gov/ij./
The Scripps Research Institute. Flow Cytometry Core Facility Cell Sorting and Analysis. WinMDI [software]. Available at: http://facs.scripps.edu.
Waterhouse NJ, Goldstein JC, von Ahsen O, Schuler M, Newmeyer DD, Green DR. Cytochrome c maintains mitochondrial transmembrane potential and ATP generation after outer mitochondrial membrane permeabilization during the apoptotic process. J Cell Biol. 2001; 153: 319–328.
Akao M, O’Rourke B, Kusuoka H, Teshima Y, Jones SP, Marbán E. Differential actions of cardioprotective agents on the mitochondrial death pathway. Circ Res. 2003; 92: 195–202.