Progressive apoptotic cell death triggered by transient oxidative insult ...

Am J Physiol Heart Circ Physiol 286: H2169–H2182, 2004. First published January 22, 2004; 10.1152/ajpheart.00199.2003.

Progressive apoptotic cell death triggered by transient oxidative insult in H9c2 rat ventricular cells: a novel pattern of apoptosis and the mechanisms Hong Han,1 Hong Long,1 Huizhen Wang,1 Jingxiong Wang,1,2 Yiqiang Zhang,1,2 and Zhiguo Wang1,2 1

Research Center, Montreal Heart Institute, Montreal, Quebec H1T 1C8; and Department of Medicine, University of Montreal, Montreal, Quebec, Canada H3C 3J7

2

Submitted 6 March 2003; accepted in final form 14 January 2004

Han, Hong, Hong Long, Huizhen Wang, Jingxiong Wang, Yiqiang Zhang, and Zhiguo Wang. Progressive apoptotic cell death triggered by transient oxidative insult in H9c2 rat ventricular cells: a novel pattern of apoptosis and the mechanisms. Am J Physiol Heart Circ Physiol 286: H2169 –H2182, 2004. First published January 22, 2004; 10.1152/ajpheart.00199.2003.—Many pathophysiological processes are associated with oxidative stress and progressive cell death. Oxidative stress is an apoptotic inducer that is known to cause rapid cell death. Here we show that a brief oxidative insult (5-min exposure to 400 ␮M H2O2), although it did not kill H9c2 rat ventricular cells during the exposure, triggered an intracellular death cascade leading to delayed time-dependent cell death starting from 1 h after the insult had been withdrawn, and this post-H2O2 cell death cumulated gradually, reaching a maximum level 8 h after H2O2 withdrawal. By comparison, sustained exposure to H2O2 caused complete cell death within a narrow time frame (2 h). The time-dependent post-H2O2 cell death was typical of apoptosis, both morphologically (cell shrinkage and nuclear condensation) and biochemically (DNA fragmentation, extracellular exposure of phosphatidylserines, and caspase-3 activation). A dichlorofluorescein fluorescent signal showed a time-dependent endogenous increase of reactive oxygen species (ROS) production, which was almost abolished by inhibition of the mitochondrial electron transport chain. Application of antioxidants (vitamin E or DTT) before H2O2 addition or after H2O2 withdrawal prevented the H2O2-triggered progressive ROS production and apoptosis. Sequential appearance of events associated with activation of the mitochondrial death pathway was found, including progressive dissipation of mitochondrial membrane potential, cytochrome c release, and late activation of caspase-3. In conclusion, transient oxidative stress triggers an intrinsic program leading to self-sustained apoptosis in H9c2 cells via cumulative production of mitochondrial ROS and subsequent activation of the mitochondrial death pathway. This pattern of apoptosis may contribute to the progressive and long-lasting cell loss in some degenerative diseases.

APOPTOSIS, A PROGRAMMED CELL death or cell suicide, is believed to be an important mechanism, other than necrosis, of cell elimination. Progressive cell death due to apoptosis is seen in many conditions, both physiologically and pathologically. Physiologically, progressive apoptotic cell death plays a fundamental role in developmental morphogenesis, cell homeostasis, and tissue modeling in organisms. Pathologically, progressive apoptotic cell death is critical in determining the progression of many diseases. Emerging evidence suggests that the progressive cell death due to apoptosis is responsible for the anatomic remodeling of some tissues and organs in their

pathological processes, such as heart failure (23, 25, 33), atherosclerotic arteries and hypertensive vessels (17, 25, 26), and Alzheimer disease (7, 31, 34). These conditions are characterized in common by degenerative changes of the tissues, with gradual elimination of cells or progressive apoptotic cell death. However, it is not known whether progressive apoptotic cell death requires the continuous presence of apoptotic inducers or whether the cells possess intrinsic mechanisms to sustain the apoptosis triggered by transient stress. Moreover, although multiple signaling pathways have been identified to be involved in regulation of apoptosis, it is not known which pathway is responsible for the progressive apoptotic cell death or which cellular events maintain the progressive cell death. Oxidative stress is a well-known factor promoting apoptosis (4, 26, 32) and has been implicated in the pathogenesis of a variety of diseases, including heart failure, atherosclerosis, and Alzheimer disease (8, 16, 22, 27, 30). H2O2 is one of the most frequently and widely used oxidants for the study of apoptotic cell death. However, sustained exposure of cells to H2O2 was used to investigate apoptosis in all relevant previous studies. In the body, transient exposure of cells to H2O2 is a frequently encountered situation, as in the case of acute myocardial ischemia and reperfusion. Transient H2O2 challenge can produce preconditioning cytoprotection against subsequent lethal stresses (18) but can also result in long-term damage to the cells. Many previous studies on H2O2 induction of apoptosis actually employed a long-term exposure paradigm (from several hours to several days). Actually, it is more important and relevant to study the effect of transient H2O2 insult on apoptotic cell death. However, it is not known whether a transient exposure to H2O2 can also induce apoptosis. The objectives of this study were to investigate whether the cardiac H9c2 cells have the intrinsic mechanism to sustain apoptosis once it is initiated by transient oxidative stress and, if so, to delineate the underlying signaling cascades. Our data demonstrated that a transient H2O2 insult (5-min exposure to 400 ␮M H2O2) did not cause cell death in its presence but triggered a delayed time-dependent increase in apoptosis after H2O2 had been withdrawn. For simplicity, we call this timedependent increase in apoptosis in the absence of continuous presence of environmental cues progressive apoptosis. The self-cumulating property of apoptosis was attributable to the time-dependent increase in production of mitochondrial reactive oxygen species (ROS), which then led to activation of the mitochondrial death pathway.

Address for reprint requests and other correspondence: Z. Wang, Research Center, Montreal Heart Institute, 5000 Belanger East, Montreal, PQ, Canada H1T 1C8 (E-mail: [email protected]).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

mitochondria; hydrogen peroxide; antioxidants

http://www.ajpheart.org

0363-6135/04 $5.00 Copyright © 2004 the American Physiological Society

H2169

H2170

MITOCHONDRIAL ROS AND PROGRESSIVE APOPTOTIC CELL DEATH

METHODS

Cell culture. The rat embryonic ventricular myocardial cell line H9c2 was purchased from American Type Culture Collection (Manassas, VA). Cells were cultured in Dulbecco’s modified Eagle’s medium containing 1,500 mg/l NaHCO2, supplemented with 10% heat-inactivated FBS, 50 U/ml penicillin, and 50 ␮g/ml streptomycin. Cells were passaged regularly and subcultured to ⬃90% confluence before experimental procedures. For experiments involving immunostaining, cells were plated on precoated glass coverslips (laminin, 10 ␮g/ml, 300 ␮l each) in 12-well plates at 6 ⫻ 104 per well 1 day before experimentation. For DNA fragmentation ELISA, cells were plated in 24-well plates at 3 ⫻ 104 per well. For Western blot, cells were plated on 10-cm petri dishes. Experimental protocols. To trigger apoptosis, the cells were first incubated in culture medium containing 400 ␮M H2O2 for 5 min and then grown in fresh normal medium for up to 10 h. Apoptotic cell death was monitored immediately after H2O2 withdrawal (time 0), thereafter hourly to 6 h, and finally at 8 and 10 h after H2O2 withdrawal. Other measurements and assays were performed at 15, 30, 60, and 300 min after H2O2 withdrawal. For experiments involving vitamin E (␣-tocopherol) or dithiothreitol (DTT), the agent was added to the culture immediately after H2O2 withdrawal or 1 h before addition of H2O2. H2O2 was purchased from Sigma Chemical, and the 1,000⫻ stocks were freshly prepared in sterile water. Vitamin E and rotenone (a specific inhibitor of mitochondrial electron transport complex I) were also from purchased from Sigma Chemical and dissolved in ethanol and DMSO, respectively. DTT, also purchased from Sigma Chemical, was dissolved in water. Quantification of cell viability by 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyl tetrazolium bromide. Cell Proliferation Kit I [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT); Roche Molecular Biochemicals, Laval, PQ, Canada] was used to determine cell viability. The manufacturer’s protocols were followed. Briefly, after treatment (H2O2 or drugs), the MTT labeling reagent (5 mg/ml MTT in PBS) was added to the culture at different times to a final concentration of 0.5 mg/ml, and the culture was incubated for 4 h. Then the solubilization solution was added to the culture medium to solubilize the purple crystals at a 1:1 ratio, and the reaction was incubated overnight. On the next day, absorbance of the sample was read using a microplate reader (Rainbow, Tecan) at 580 nm. ELISA. The Cell Death Detection ELISA kit (Boehringer Mannheim) was employed to quantify DNA fragmentation on the basis of antibody detection of histone and fragmented DNA. The procedures were the same as those previously described in detail (18). TdT-mediated dUTP nick end labeling. DNA fragmentation of individual cells was detected in situ by TdT-mediated dUTP nick end labeling (TUNEL) with the In Situ Cell Death Detection Kit, Fluorescein (Boehringer Mannheim). The detailed procedures are described elsewhere (18). Annexin V. The TACS annexin V-biotin apoptosis detection kit was used to detect phosphatidylserines exposed to the outer surface of the cytoplasmic membrane of the cells undergoing apoptosis (18). Intracellular ROS measurement. 5(6)-Chloromethyl-2⬘,7⬘-dichlorodihydrofluorescein diacetate-acetyl ester (CM-H2DCFDA; Molecular Probes) is an ROS-sensitive probe that can be used to detect oxidative activity in living cells. It passively diffuses into cells, where its acetate groups are cleaved by intracellular esterases, releasing the corresponding dichlorodihydrofluorescein derivative. Its thiol-reactive chloromethyl group reacts with intracellular GSH and other thiols. Subsequent oxidation yields a fluorescent adduct that is trapped inside the cell. When it is excited at 480 nm, its emissions at 505–530 nm can be captured. CM-H2DCFDA is prepared in DMSO immediately before loading. H9c2 cells or L cells cultured on glass coverslips in 12-well plates were loaded with the dye (10 ␮g/ml) immediately after H2O2 had been removed (time 0) and at several times after H2O2 AJP-Heart Circ Physiol • VOL

withdrawal. The cultures were then incubated in the dark at 37°C for 30 min. After they were washed, the cells were allowed to recover at 37°C for 20 min and then fixed in 3% paraformaldehyde in PBS for 20 min. The coverslips were mounted and examined immediately under a laser scanning confocal microscope (LSM 5100, Zeiss). Control cells (vesicle treatment) were also loaded with the dye and treated in the same way in parallel. Rotenone (5 ␮M; Sigma Chemical) was added to the culture 10 min before H2O2 exposure to preinhibit the mitochondrial electron transport chain and kept in the medium throughout the experiments. Vitamin E (50 ␮M; Sigma Chemical) was added to the culture immediately after H2O2 withdrawal. Mitochondria staining. Mitochondria were identified by staining with MitoTracker green probes (Molecular Probes). H9c2 cells and L cells cultured on coverslips were fixed with 3% paraformaldehyde in PBS and then permeabilized in 0.5% Triton X-100 in PBS. After they were blocked in 5% BSA, the cells were stained with the probes diluted in 2.5% BSA at 200 nM for 1 h. The coverslips were then mounted on the slides and examined under a confocal microscope. Measurement of mitochondrial membrane potential. JC-1 dye (Molecular Probes) exhibits potential-dependent accumulation in mitochondria and, thus, was employed to detect the change in mitochondrial membrane potential (⌬⌿m). JC-1 dye aggregates at high ⌬⌿m and can be excited at 488 nm. The emission shifts from green (525 nm) to red (590 nm) when J aggregates form. JC-1 dye was added to the culture medium at 10 ␮g/ml and incubated for 15 min at 37°C. The medium was then removed, and the cells were washed three times with PBS. After being mounted on the slides, the cells were examined immediately under a confocal microscope (Zeiss). Immunostaining of cleaved caspase-3 and caspase-7, cytochrome c, and phosphorylated JNK. The procedures were as previously described (47). Control cells and cells treated with H2O2 were fixed with freshly prepared 3% paraformaldehyde in PBS for 20 min at 4°C. After they were washed with PBS, the cells were permeabilized in 0.5% Triton-PBS for 5 min and blocked in 3% BSA. The cells were incubated overnight at 4°C with primary antibodies diluted in 3% BSA and then incubated with FITC-conjugated secondary antibodies at room temperature for 2 h. For caspase staining samples, 1 ␮g/ml propidium iodide (PI) was mixed with the secondary antibody to counterstain the nucleus. The coverslip was mounted on a slide with 10 ␮l of antifading solution, and the cells were examined under a confocal microscope. Rabbit anticleaved caspase-3, caspase-7, and antiphosphorylated stress-activated protein kinase (JNK) antibodies were purchased from Cell Signaling Technology (Beverly, MA). Sheep anticytochrome c antibody was purchased from Sigma Chemical. Donkey anti-rabbit FITC-conjugated and donkey anti-sheep FITC-conjugated antibodies were purchased from Jackson ImmunoResearch (Baltimore, MD). Caspase-3 activity assay. At 4 h after transient H2O2 or H2O2vitamin E treatment, the H9c2 cells were scraped off the plates in ice-cold PBS. The cells from each 10-cm plate were then lysed in 100 ␮l of ice-cold cell lysis buffer {50 mM HEPES, pH 7.4, 0.1% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), 1 mM DTT, and 0.1 mM EDTA} for 5 min and then ruptured by 20 passages through a 26-gauge needle. The lysate was centrifuged at 12,000 g for 10 min at 4°C. For assay, 10 ␮l of the cytosolic preparation (supernatant) were added into 80 ␮l of assay buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 0.1% CHAPS, 10 mM DTT, 0.1 mM EDTA, and 10% glycerol) in each well of a 96-well dark plate. The reaction was allowed to warm up to 37°C before 10 ␮l of the fluorescence substrate acetyl-Asp-Glu-Val-Asp (DEVD)-7amino-4-trifluoromethylcoumarin (2 mM) were added. The absorbance [optical density (OD), 425 nm excitation and 515 nm emission] was read in a Fusion Microplate reader (Packard, Bioscience) at 0 min and every 30 min for 90 min. To confirm the specificity of caspase-3 activity, 20 ␮l of 0.1 mM DEVD-aldehyde (a caspase inhibitor) were added to the H2O2-treated sample reaction mix. The OD values after 286 • JUNE 2004 •

www.ajpheart.org


subtraction of blank were used to calculate the ratios of H2O2 to control and vitamin E ⫹ H2O2 to control, and the values were normalized to the protein concentration. Western blotting analysis for cytochrome c. The immunoblotting procedures were carried out as described in detail elsewhere (18). At 5 h after H2O2 withdrawal, cells were scraped off the plate in ice-cold PBS and spun down at 300 g for 2 min. Buffer A (in mM: 20 HEPES-KOH, pH 7.5, 10 KCl, 1.5 MgCl2, 1 DTT, 1 EDTA, 1 EGTA, 1 PMSF, and 250 sucrose) was added to the cell pellet, and the cells were homogenized on ice and then centrifuged at 650 g for 10 min to pellet the cell debris and unbroken cells. The supernatant was further centrifuged at 12,000 g for 15 min to pellet the mitochondria-rich fraction. The supernatant, after collection of the mitochondria-rich fraction, was subjected to ultracentrifugation at 100,000 g for 1 h to purify the cytosolic fraction. Protein content was determined with a Bio-Rad protein assay reagent (Bio-Rad, Mississauga, ON, Canada) using BSA as a standard. The extracted proteins (40 ␮g) were fractionated by SDS-PAGE (15% polyacrylamide gels), transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA), blocked in 5% milk, incubated with mouse anticytochrome c antibody (clone 7H8.2C12, BD PharMingen) at 1 ␮g/ml in 2.5% BSA, and incubated again with anti-mouse, HRP-conjugated secondary antibody and HRP-conjugated antibiotin antibody to detect the marker. Bound antibodies were detected with Chemiluminescence Reagent Plus (Perkin Elmer Life Sciences, Boston, MA), exposed to an X-film, and quantified by Quantity One software. GSH measurement. Intracellular GSH was measured using a GSH assay kit (Calbiochem, San Diego, CA), with some modifications. H9c2 cells were plated at 8 ⫻ 105 per 10-cm plate 24 h before experiments. Cells were transiently exposed to H2O2 for 5 min and then cultured in growth medium for 2 or 5 h after H2O2 withdrawal. In parallel, another group of cells were exposed to H2O2 at the same concentration for 2 or 5 h. At each time point, five plates of cells were harvested in ice-cold PBS and then centrifuged at 3,000 rpm for 5 min. After they were washed once with PBS, cells were resuspended in freshly prepared 5% metaphosphoric acid (50 ␮l/plate cells) and then sonicated on ice for 10 strokes. The cell debris was removed by centrifugation at 3,000 rpm for 10 min; 50 ␮l of supernatant were used for each assay. The assay was assembled in a 96-well plate in the following order: 50 ␮l of supernatant ⫹ 40 ␮l of buffer 3 ⫹ 5 ␮l of reagent 1 ⫹ 5 ␮l of reagent 2. The mixture was incubated at 25°C for 10 min in the dark, and then the OD values were read in a microplate reader at 400 nm. The measured values were normalized to control. Data analysis. Group data are expressed as means ⫾ SE. Comparisons among groups were made by ANOVA (F test). Bonferroni’s adjusted t-tests were used for multiple group comparisons, and a paired or an unpaired t-test was used for single comparison. A two-tailed P ⬍ 0.05 was taken to indicate a statistically significant difference. RESULTS

Apoptosis of H9c2 rat ventricular cells triggered by brief oxidative stress. Cell death induced by a brief exposure (5 min) to H2O2 (400 ␮M) was monitored over time up to 10 h after H2O2 withdrawal. No obvious morphological changes in the H9c2 cells were seen within 1 h after withdrawal of H2O2 (Fig. 1B). Cell shrinkage (the morphological hallmark of apoptosis), however, began to develop thereafter in H2O2-treated cells; cell shrinkage was not seen in nontreated cells (Fig. 1A). To quantify cell death, the round and detached cells were counted as dead cells. Nuclear condensation was consistently seen and is better viewed in Figs. 7–9. The proportion of post-H2O2 cell death was increased with time and reached a maximum level (⬃47%) at 8 h after H2O2 withdrawal (Fig. 1, B and D). Correspondingly, MTT assay showed a time-dependent decline AJP-Heart Circ Physiol • VOL

H2171

in cell viability in H2O2-treated cells starting at 1 h and reaching a plateau level at 8 h after H2O2 withdrawal (Fig. 1E). On the contrary, there were minimal decreases in cell viability in control cells (Fig. 1, A and D). Also, in sharp contrast to the progressive cell death triggered by transient exposure to H2O2, the cells in the continuous presence of H2O2 (400 ␮M) at the same concentration in the culture medium died within a rather narrow time window. Within 1 h after incubation with H2O2, cell death was ⬃20%, and by 2 h cell death was ⬃100% (Fig. 1D). To clarify that the post-H2O2 cell death was due to apoptosis, we used TUNEL and ELISA methods to detect chromosomal and DNA fragmentation, the biochemical hallmark of apoptosis. When assayed 5 h after H2O2 withdrawal in H2O2treated H9c2 cells, significant numbers of nuclei were positive for TUNEL staining, whereas the nontreated cells were virtually TUNEL negative (Fig. 2A). Similarly, OD from ELISA experiments for quantification of DNA fragmentation was substantially increased in H2O2-treated compared with nontreated cells, starting from 1 h after H2O2 withdrawal and reaching a maximum level (⬃16-fold increase over control) 5 h after H2O2 withdrawal (Fig. 2B). By comparison, the cells in the continuous presence of H2O2 demonstrated a sharp increase in DNA fragmentation (OD value) within 2 h to a maximum degree (Fig. 2B). Because cell shrinkage occurred as early as 1 h after H2O2 withdrawal, we sought to detect earlier events of apoptosis before morphological alterations. Early in the apoptotic process, phospholipid asymmetry is disrupted, leading to the exposure of phosphatidylserine on the outer leaflet of the cytoplasmic membrane. Annexin V is an anticoagulant protein that preferentially binds to negatively charged phospholipids and can be used to identify cells in the early stages of apoptosis. We used conjugated annexin V (green fluorescence) together with PI (red fluorescence), which enters the cells when membranes are disrupted or the permeability is increased during necrosis or at the late stages of apoptosis, to identify early apoptosis. Figure 2C displays examples of annexin V staining. Virtually no cells from the control group or the H2O2 group immediately after H2O2 withdrawal were stained, whereas at 60 min after H2O2 withdrawal 28% of the cells displayed green annexin V staining without red PI nucleus staining. At 2 h after H2O2 withdrawal, 36% of the cells showed positive annexin V staining and only 5% of the cells showed PI nucleus staining. To quantify the number of early apoptotic cells, those that stained green only (no red) and those that stained red (with and without green) were counted separately, and the data are summarized in Fig. 2C (right). To obtain a rough idea of the mode (apoptosis or necrosis) of cell death with transient vs. sustained exposure to H2O2, we carried out experiments with trypan blue staining, which has been used to identify necrotic cells. The results are presented in Fig. 2D, where the mean percentages of the cells stained with trypan blue measured 2 h after H2O2 withdrawal or in the continuous presence of H2O2 are shown. Approximately 10% of the cells with transient exposure to H2O2 and ⬃100% of the cells with sustained exposure to H2O2 were stained by trypan blue. Role of mitochondrial ROS in post-H2O2 apoptosis. One question that needs to be answered is how H2O2, which does not kill the cells during its presence, causes time-dependent 286 • JUNE 2004 •

www.ajpheart.org

H2172


Fig. 1. Cell death triggered by brief oxidative insult. A: H9c2 cells, under a light microscope (magnification ⫻100), serving as a control (no H2O2 treatment) for data in B. B and C: morphological changes in H9c2 cells treated with 400 ␮M H2O2 transiently (5 min, B) or in the continuous presence of 400 ␮M H2O2 (C). Note post-H2O2 time-dependent cell death. D: percentage (mean ⫾ SE) of dead cells averaged from 20 fields of the type shown in A–C. E: cell viability (n ⫽ 5 independent samples) determined by MTT assay. Note decline of survival rate as a function of time after H2O2 withdrawal.

apoptosis after its withdrawal. One possibility is that residue H2O2 entered the cells during the 5-min exposure. We tested this notion by measuring the intracellular ROS level using CM-H2DCFDA fluorescence dye. It is expected that the residue H2O2, if any, would decline with time, depending on scavenging by cellular antioxidants. However, our data demonstrated the opposite: a time-dependent increase in ROS level (Fig. 3A). Immediately after H2O2 withdrawal, the intracellular ROS level was only slightly higher than the basal ROS level in the cells without exposure to H2O2, suggesting that residue H2O2 alone could hardly account for the time-dependent cell death. The time-dependent increase in ROS level started 5 min after H2O2 withdrawal and continued until 2 h to a maximum level without further changes up to 8 h. Evidently, the initial H2O2 exposure triggered the intracellular production of ROS. The mitochondrion is the most likely site for massive production of ROS. To test this notion, we further measured intracellular ROS after mitochondrial production of ROS was blocked by pretreatment of the cells with rotenone, an inhibitor of the mitochondrial electron transport chain. Under this condition, intracellular ROS production was markedly suppressed and AJP-Heart Circ Physiol • VOL

showed a time-dependent decline (Fig. 3B), similar to that in control untreated cells. The intensity of fluorescent staining for ROS was analyzed using the LSM program suite, and the data were normalized to control (without H2O2 treatment) values and to H2O2-treated cells involving rotenone treatment. The intracellular ROS levels increased with time. At 5 min and 2 h after H2O2 withdrawal, the ROS levels were 2- and 3.5-fold higher than the control level, respectively. In the presence of rotenone, ROS levels were reduced to ⬃50% of levels in cells treated with H2O2 alone. We previously showed that brief exposures of mouse fibroblast L cells to H2O2 do not induce cell death (18), as seen in H9c2 cells; instead, the treatment results in preconditioning cytoprotection. To further verify the role of mitochondrial ROS in mediating post-H2O2 apoptosis in H9c2 cells, we compared the cellular density of mitochondria between H9c2 and L cells using MitoTracker green staining and cell death by ELISA. MitoTracker green dye is potential insensitive, and, once bound to mitochondria, it does not readily dissociate from these organelles, irrespective of the magnitude of ⌬⌿m. After staining with MitoTracker green, the mitochondria in control 286 • JUNE 2004 •

www.ajpheart.org


cells showed filament-like staining of ⬃1-␮m thickness that was distributed in the cytoplasm (Fig. 4). The intensity of staining was considerably greater in H9c2 cells than in L cells, suggesting a higher density of mitochondria in the former than

H2173

in the latter. Correspondingly, 5-min exposure to H2O2 triggered significant cell death 10 h after H2O2 withdrawal in H9c2 cells, whereas virtually no cell death was detected in L cells under the same conditions. Furthermore, the ROS level was

Fig. 2. Apoptosis determined by TdT-mediated dUTP nick end labeling (TUNEL), ELISA, and annexin V methods and necrosis determined by trypan blue staining. A: TUNEL staining performed 5 h after H2O2 withdrawal. Nuclear staining is superimposed on the phase contrast image of the same scope to view the contour of the cells. Mean (n ⫽ 3) percent TUNELpositive cells were counted from 20 fields for each group. ***P ⬍ 0.001, 5 h vs. 0 h. B: mean optical density (OD) values from ELISA (n ⫽ 6) measured at various times after H2O2 withdrawal, indicating degree of DNA fragmentation. Left: actual OD values; right: OD values normalized to time 0 or immediately after H2O2 withdrawal to show fold increases in OD over time. C: annexin V (cytoplasmic membrane) and propidium iodide (PI, nuclei) staining at different times after H2O2 withdrawal. Mean percent annexin V- and PI-stained cells were obtained from a total of 3 experiments. *P ⬍ 0.05; ***P ⬍ 0.001 vs. 0 h. D: percentage of trypan blue staining of cells with transient (H2O2-T) or sustained (H2O2-S) exposure to 400 ␮M H2O2. Values are means, determined 2 h after addition of H2O2 for H2O2-S group and after H2O2 withdrawal for H2O2-T group, from 5 batches of cells for each group. *P ⬍ 0.05; ***P ⬍ 0.001 vs. control (Ctl). ⫹⫹⫹P ⬍ 0.001 vs. H2O2-T (t-test).

AJP-Heart Circ Physiol • VOL

286 • JUNE 2004 •

www.ajpheart.org

H2174


Fig. 3. Intracellular reactive oxygen species (ROS) level after brief exposure to H2O2. A: confocal images of dichlorofluorescein (DCF) staining for ROS at various time points after H2O2 withdrawal in the absence of rotenone (Rot, top) and averaged (n ⫽ 120 cells) fluorescence intensity normalized to control (Ctl) cells (without H2O2 treatment, bottom). B: ROS production after treatment with 0.5 ␮M rotenone, an inhibitor of mitochondrial electron transport chain (top) and mean data (n ⫽ 80 cells) of DCF fluorescence intensity as an indication of ROS levels in cells treated with and without rotenone (bottom). *P ⬍ 0.05; ***P ⬍ 0.01 vs. H2O2 alone (unpaired t-test). ⫹P ⬍ 0.05 (F test), indicating significant time dependence of ROS alterations.

consistently lower in L cells and did not show time-dependent increases after H2O2 withdrawal (Fig. 5). On the contrary, a time-dependent decline in the ROS level was seen, indicating minimal production of endogenous ROS in L cells. If the post-H2O2 apoptosis in H9c2 cells was consequent to ROS production in mitochondria, then rescue from cell death by antioxidants is expected. This was indeed confirmed by the experiments using 100 ␮M vitamin E, which has been shown to inhibit mitochondrial ROS generation (13), and 50 ␮M DTT. Vitamin E or DTT was applied to the cell culture immediately after removal of H2O2, and cell death was mon-

itored over 10 h. Our data clearly showed that vitamin E and DTT effectively rescued the cells from apoptosis (Fig. 6, A and C). To investigate whether pretreatment with vitamin E or DTT could prevent apoptotic cell death, we conducted another series of experiments in which we preincubated the cells with vitamin E or DTT for 1 h before adding H2O2 for 5 min. ELISA was performed 5 h after H2O2 withdrawal. Pretreatment of cells with vitamin E or DTT also significantly prevented cell death (Fig. 6C). By comparison, 100 ␮M vitamin E reduced cell death (by MTT) and DNA fragmentation (determined by ELISA) by only 20% in the cells with sustained H2O2 treat-

Fig. 4. Number of mitochondria and post-H2O2 cell death in H9c2 cells (A) and L cells (B). Presence of mitochondria was identified by MitoTracker green staining (left), and cell viability was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay (right; mean data, n ⫽ 5 for H9c2 cells and n ⫽ 3 for L cells). Staining is much stronger in H9c2 cells than in L cells, suggesting that there are more mitochondria in H9c2 cells. Correspondingly, although post-H2O2 cell death is prominent in H9c2 cells, there is little cell death in L cells 10 h after brief H2O2 treatment. **P ⬍ 0.01 vs. 0 h.


286 • JUNE 2004 •

www.ajpheart.org


H2175

Fig. 5. ROS levels in H9c2 cells (A) and L cells (B). Note time-dependent increase and decline of DCF fluorescence intensity in H9c2 and L cells, respectively. Similar results were obtained in a total of 4 batches of cells.

ment. However, ⬃40% reduction of cell death and DNA fragmentation was achieved when vitamin E concentration was elevated to 1 mM. We further studied the time-dependent effects of vitamin E on DNA fragmentation by applying 100 ␮M vitamin E at different times after H2O2 withdrawal. The optimal effect of vitamin E was found within 30 min after H2O2 withdrawal, and the beneficial effect decreased with delayed application of vitamin E (Fig. 6D). Vitamin E applied 3 h after H2O2 withdrawal nearly lost its ability to protect cells from apoptosis. Consistently, vitamin E also dramatically diminished intracellular ROS production (Fig. 7A); the fluorescent intensity of the staining was reduced to ⬍50% of the ROS level in cells without vitamin E treatment. The reduction was greater with longer time after H2O2 withdrawal. Because in our experiments the number of apoptotic cells increased with time after H2O2 withdrawal, we quantified the number of cells with higher levels of ROS relative to control untreated cells. The cells with intensity of ROS staining ⱖ1.5-fold greater than the basal level from control untreated cells were considered ROS-positive cells and taken into account for quantification, and the percent-

age of ROS-positive cells was calculated. Comparisons were made among four groups: control untreated, transient H2O2, H2O2-rotenone, and H2O2-vitamin E (rotenone or vitamin E was added immediately after H2O2 withdrawal and was continuously present in the culture medium). The number of ROS-positive cells showed biphasic increases with time after H2O2 withdrawal (Fig. 7C): the first rapidly increasing phase was completed within the first 10 min, and the second slowly increasing phase lasted ⱖ2 h. When treated with rotenone or vitamin E, the number of ROS-positive cells decreased dramatically compared with the cells treated with H2O2 alone. Particularly, the second slowly increasing phase was abolished in the presence of rotenone or vitamin E. The number of ROS-positive cells corresponded well with the number of apoptotic cells of the same group. There is a possibility that the observed increase in intracellular ROS level was due to a loss of the antioxidant buffering capacity against endogenous ROS production. To test this notion, we compared intracellular GSH levels between cells with transient exposure or sustained exposure to H2O2 (400 ␮M) at 2 and 5 h. In cells exposed to H2O2 for 5 min, there is

Fig. 6. Effects of vitamin E (VitE) and dithiothreitol (DTT) on post-H2O2 cell viability and DNA fragmentation in H9c2 cells. A and B: percent cell survival determined by MTT assay (n ⫽ 4) and degrees of DNA fragmentation as indicated by OD values determined by ELISA (n ⫽ 5), respectively. Ctl, control cells without H2O2 treatment; H2O2, cells treated with H2O2 for 5 min; VitE, vitamin E at various concentrations in ␮M, applied on removal of H2O2. C: effects of vitamin E and DTT applied before (VitE ⫹ H2O2 and DTT ⫹ H2O2) and after (H2O2 ⫹ VitE and H2O2 ⫹ DTT) exposure to H2O2. D: effects of 100 ␮M vitamin E on DNA fragmentation at 0 –180 min after H2O2 withdrawal. Averaged data are shown from 4 batches of cells for control, 8 for transient H2O2 treatment alone, and 6, 4, 4, 6, and 6 for vitamin E addition at 0, 30, 60, 120, and 180 min, respectively. *P ⬍ 0.05; **P ⬍ 0.01; ***P ⬍ 0.001 vs. H2O2 alone.


286 • JUNE 2004 •

www.ajpheart.org

H2176


Fig. 7. Effects of vitamin E (an antioxidant) and rotenone (Rot, a specific inhibitor of mitochondrial electron transport complex I) on intracellular level of ROS determined by CM-H2DCFDA fluorescence dye. A: confocal microscopic images showing intracellular ROS levels in H9c2 cells treated with (bottom) and without (top) 50 ␮M vitamin E. Similar results were obtained in a total of 4 batches of cells. B: mean data (n ⫽ 120 cells) for fluorescent intensity normalized to values from control cells without H2O2 treatment. C: percentage of cells with positive staining for ROS. Cells with fluorescence intensity stronger than control cells are taken into account. Ctl, control cells without H2O2 challenge; H2O2, cells treated with H2O2 for 5 min; H2O2 ⫹ Rot, cells treated with rotenone after H2O2 withdrawal; H2O2 ⫹ VitE, cells treated with 50 ␮M vitamin E after H2O2 withdrawal. D: intracellular levels of GSH in cells with transient and sustained exposures to H2O2 at 2 and 5 h after H2O2 withdrawal. Mean normalized OD values are shown from 10 measurements for control cells (Ctl), 13 measurements 2 h after H2O2 withdrawal, and 8 measurements 5 h after H2O2 withdrawal and 2 and 5 h after sustained incubation with H2O2. *P ⬍ 0.05; ***P ⬍ 0.001 vs. H2O2 alone (unpaired t-test). ⫹P ⬍ 0.05 (F test), indicating significant time dependence of ROS alterations.

only ⬃23% reduction at 2 h and ⬃15% reduction at 5 h after H2O2 withdrawal (Fig. 7D). By comparison, sustained H2O2 treatment caused more dramatic decreases in GSH level. The GSH level dropped to ⬃46% of the control level at 2 h and to only ⬃42% of the control level at 5 h. Changes in ⌬⌿m. H2O2 is known to stimulate activation of several signaling pathways leading to apoptosis. The mitochondrial death pathway is one of the important pathways (1, 29, 36). Mitochondrial function is highly susceptible to oxidative damage. We investigated whether in our model the postH2O2 apoptosis was attributable to activation of the signaling cascade of the mitochondrial death pathway. We conducted a series of experiments to monitor the alterations of mitochonAJP-Heart Circ Physiol • VOL

drial functions. ⌬⌿m is known to be a critical factor determining the integrity of mitochondria, and loss of ⌬⌿m can lead to release of cytochrome c from mitochondria, which in turn activates downstream caspases to execute apoptosis or, in some cases, itself suffices to activate the apoptotic pathways to commit the cell to die (14, 29, 36, 37). ⌬⌿m was detected using JC-1 dye. Untreated control cells that were stained with the fluorescent dye JC-1 exhibited numerous, brightly stained mitochondria that emitted redorange fluorescence (Fig. 8), representing J aggregates that accumulate at normally hyperpolarized membrane potential. Cells after transient H2O2 treatment exhibited fewer orange J aggregates, indicating gradual dissipation of ⌬⌿m. The earliest 286 • JUNE 2004 •

www.ajpheart.org


H2177

Fig. 8. Measurement of mitochondrial membrane potential (⌬⌿m) by JC-1 staining. “J”-shaped and red-stained bars (J aggregates) gradually dissipate with time after H2O2 withdrawal. Similar data were obtained from another 3 experiments.

change in JC-1 staining was detected at 30 min after H2O2 withdrawal. From 3 h after H2O2 withdrawal, the J aggregates in these cells were barely visible, indicating complete collapse of ⌬⌿m. Meanwhile, some green fluorescence of monomers appeared in the cytoplasm, representing the depolarized mitochondria. Immunocytochemical localization of cytochrome c. Cytochrome c release from mitochondria is a critical step toward apoptosis (24, 36, 46). We therefore analyzed cytochrome c localization using the cytochrome c antibody. Immediately after H2O2 exposure, cytochrome c staining was seen as defined rod-shaped structures surrounding the nuclei (Fig. 9). The staining became diffused and aggregated into puncta throughout the cytosol of the cell starting at 30 min after H2O2 withdrawal, indicating release of cytochrome c into the cytoplasm. Cytochrome c release from mitochondria was also confirmed by Western blot analysis of relative cytochrome c concentrations between cytosolic and mitochondria-rich fractions (Fig. 9C). Cytochrome c in the cytosolic fraction was markedly increased by ⬃2.5-fold 5 h after H2O2 withdrawal, and the increase was abolished by 100 ␮M vitamin E. In the absence of H2O2, the ratio of cytochrome c in the cytosol to cytochrome c in the mitochondrion was 0.45, and after transient exposure to H2O2 the ratio increased to 0.93, indicating a release of cytochrome c from mitochondrion to cytosol. Activation of caspase-3 and caspase-7. Caspase-3 and caspase-7 are key executioners of apoptosis, and caspase-3 activities are known to be associated with activation of the mitochondrial death pathway (5, 46). Activities of caspase-3 and caspase-7 were determined by immunostaining with the antibodies raised against the cleaved form (activated) of the enzymes, together with PI staining to view the nuclei. Figure AJP-Heart Circ Physiol • VOL

10 shows the results obtained 5 h after H2O2 or vehicle (control) withdrawal. Obviously, there was little immunoreactivity (green staining) for both caspases in control cells in which the spindle-shaped cell contour and oval nuclei with well-defined nuclear membrane (red staining) were clearly seen. In H2O2-treated cells, however, there was striking staining for caspase-3 throughout the cytoplasm in cells showing condensed nuclei revealed by PI staining (Fig. 10A). By comparison, the intensity of caspase-7 staining (Fig. 10B) was much weaker, although in these cells the same degree of nucleus condensation was seen as in those cells showing strong caspase-3 staining. To quantify the activation of caspase-3, the activities of caspase-3 were measured using enzyme activity assay. Transient H2O2 stress stimulated an increase in the activity of caspase-3 by nearly fourfold, as indicated by the increased OD values, and vitamin E greatly prevented the increase (Fig. 10C). In the presence of the caspase inhibitor DEVD-aldehyde in the reaction mix, the caspase-3 activities decreased by 93%, indicating that the increased activity resulted from caspase-3 activation with minimal contributions from other caspase species. Activation of JNK. There is a possibility that in our model the mitochondrial death pathway was triggered by activation of JNK (40, 41). To test this point, we analyzed JNK activities by immunostaining with the antibody against the phosphorylated (activated) form of JNK. JNK activity demonstrated a transient increase after H2O2 withdrawal (Fig. 11A). The activity was barely detectable at time 0 (immediately after H2O2 withdrawal) but appeared at 15 min, further increased to a maximum level at 30 min, and began to decline thereafter. By 3 h, JNK activity completely dissipated. By comparison, in cells 286 • JUNE 2004 •

www.ajpheart.org

H2178


Fig. 9. A: immunocytochemical localization of cytochrome c. Cytochrome c staining is shown at 0 –5 h after H2O2 withdrawal. B: amplified images from areas defined by dashed-line boxes in A showing a better view of cytochrome c staining. Cytochrome c was initially (0 h) distributed in the bar-shaped structures (representing mitochondria) surrounding the nucleus, later (after 30 min) became diffused throughout the cytosol, and finally aggregated into spots. Similar data were obtained from a total of 3 experiments. C: Western blot analysis of cytochrome c distribution in cytosolic (Cyto) and mitochondria (Mito)-rich fractions. Left: vitamin E abolished the increase in cytosolic cytochrome c induced by a transient H2O2 insult; right: redistribution of cytochrome c between cytosol and mitochondrion was induced by a transient H2O2 insult. Each bar represents a pooled sample from 4 independent batches of cells.

continuously exposed to 400 mM H2O2 for 2 h, JNK activity was constantly higher than in control untreated cells over the period of 2 h (Fig. 11B). DISCUSSION

We have shown that a brief oxidative insult (5-min exposure to H2O2), although it did not kill H9c2 cells during its presence, triggered an intracellular death cascade leading to remarkable time-dependent cell death after the insult had been withdrawn. By comparison, sustained exposure of cells to H2O2 at the same concentration caused abrupt cell death within a short time (1 h). The transient H2O2-induced cell death was typical of apoptosis, both morphologically (cell shrinkage and nuclear condensation) and biochemically (DNA fragmentation, extracellular exposure of phosphatidylserines, and caspase-3 AJP-Heart Circ Physiol • VOL

activation), with an indication of some necrotic cells (trypan blue staining). The cells in the continuous presence of H2O2 displayed ⬃100% trypan blue staining with significant DNA fragmentation, suggesting mixed necrosis and apoptosis. The progressive apoptosis induced by transient oxidative insult was likely due to a time-dependent increase in ROS production in mitochondria, which acts as a signaling molecule to trigger the apoptotic cascade or damages the integrity of the mitochondrial membrane to activate the death pathway and maintain the time-dependent cell death. Apoptosis is usually studied by exposing cells to an environment containing proapoptotic factors until cell death is seen. The cell death process in the continuous presence of apoptotic inducers appears to be abrupt. For example, in the present study, cell death reached ⬃100% within 2 h of expo286 • JUNE 2004 •

www.ajpheart.org


H2179

Fig. 10. A and B: immunocytochemical analysis of activities of caspase-3 and caspase-7, respectively. Images show double staining using antibody against cleaved caspase (C-caspase) and PI to view nuclei in the same cells. Cleaved caspase-3 and caspase-7, indicating activities of these enzymes, are stained green; PI staining of nuclei (red) was performed to identify dead cells as indicated by condensed nuclei. Cells cultured in normal medium for 5 h after H2O2 (or vehicle for control) withdrawal were used for the assay. Caspase-3 activation is more pronounced than caspase-7 activation (which is virtually absent), although the same degree of nuclear condensation (PI staining) is seen. C: caspase-3 activity assay. Data were collected 4 h after transient H2O2 (400 ␮M; 5 min) insults or after transient H2O2 insults in the continuous presence of 100 ␮M vitamin E. Each bar represents mean of 3 independent batches of cells. ***P ⬍ 0.001 vs. control (Ctl). ⫹⫹P ⬍ 0.01 vs. H2O2 alone.

sure to H2O2. By comparison, cell death in response to transient exposure to the same concentration (400 ␮M) of H2O2 was merely detectable within 1 h after H2O2 withdrawal. However, cell death, once triggered, increased gradually with time and reached the maximum level 8 h after H2O2 removal. The time-dependent increase in the number of cells that underwent apoptosis after removal of the apoptotic inducer demonstrated in the present study indicates that cells have the intrinsic program to dictate progressive cell death in the absence of continuous environmental cues. Our findings might bear some important pathophysiological implications. In many AJP-Heart Circ Physiol • VOL

situations, the tissues and cells of the human body may be exposed to transient proapoptotic insults, such as transient ischemia, hypoxia, and increase in intracellular ROS. In the case of progressive cardiac hypertrophy and heart failure, cell death, mainly apoptosis, is a progressive condition that contributes importantly to a process called continuous anatomic remodeling and usually continues even though the original stimuli have been removed or have disappeared. Sometimes the initial insult is mild, which causes no immediate cardiac dysfunction (e.g., a small myocardial infarction), but a continuous remodeling process is triggered (49). In fibrillating and 286 • JUNE 2004 •

www.ajpheart.org

H2180


Fig. 11. Immunocytochemical analysis of JNK activities. Post-H2O2 time-dependent changes in JNK activity are indicated by staining with phospho-JNK antibody. Note transient increase in JNK activity after H2O2 withdrawal. The same results were obtained from a total of 4 experiments.

dilated atria, apoptotic death of myocytes with myolysis contributes to cellular remodeling (2). Similarly, progressive cell loss in specific neuronal populations is also a pathological hallmark of neurodegenerative diseases (31). The progressive apoptosis described in this study is somewhat reminiscent, in respect to the cell death pattern among many other pathological processes, of anatomic remodeling at the tissue/organ level. There is a growing body of evidence indicating that oxidative stress is enhanced in failing hearts and may contribute to the structural and functional changes that lead to disease progression (22). Transient metabolic stress, such as acute ischemia, is a frequently occurring condition. Maulik et al. (28) and Dumont et al. (15) showed that myocardial ischemia per se caused little apoptosis, and apoptotic cells were seen only in ⬎60 min of reperfusion. Whether these observations are attributable to progressive apoptosis as revealed in this study merits future studies. However, our data were obtained from dedifferentiated cells and may not be applied directly to nonproliferating cells such as cardiomyocytes and neurons. The mitochondrial electron transport system consumes ⬎85% of the amount of oxygen used by a mammalian cell, ⬃5% of which is converted to superoxide, H2O2, and other ROS under physiological conditions (44, 45). Increased production of ROS by mitochondria has been seen under pathological situations. Sublethal levels of ROS may produce preconditioning cytoprotection against subsequent oxidative stress (18, 43). However, when the cellular level of ROS overwhelms the antioxidant capacity of the cell, damage to cellular macromolecules such as lipid, protein, and DNA may occur. In many situations, cellular ROS levels may fluctuate according to the balance between ROS production and antioxidant regeneration. In the present study, 400 ␮M H2O2 was used to initiate the progressive apoptotic cell death. This concentration of H2O2 is presumably lethal. However, the exposure time was brief, only 5 min, and during this period and the following 1 h after H2O2 withdrawal, no cell death was seen. This implies that the initial H2O2 treatment is insufficient to kill the cells but serves to AJP-Heart Circ Physiol • VOL

trigger the intracellular production of ROS in mitochondria, which in turn activates the mitochondrial death pathway leading to apoptotic cell death. Thus ROS production plays a central role in the observed progressive apoptosis. This is further supported by our experiments with application of vitamin E or DTT, which prevented the time-dependent increase in ROS level and the progressive apoptotic cell death. The protective effect of 100 ␮M vitamin E was maximal when applied within 30 min after H2O2 withdrawal and was weakened with delayed application. Noticeably, even with delayed application 2 h after H2O2 withdrawal, vitamin E still could significantly rescue the cells from apoptosis. Intriguingly, Vanden Hoek et al. (43) and our own laboratory (18) previously showed that brief exposure of cells to H2O2 induced preconditioning cytoprotection against subsequent lethal stimuli, instead of progressive cell death. The disparity between our present and previous studies can be explained on the basis of different numbers of mitochondria in H9c2 cells and L cells: the former has a greater number of mitochondria and can produce more ROS than the latter. The discrepancy between the present study and the study reported by Vanden Hoek et al. is most likely because the concentration of H2O2 used in their study was only 15 ␮M and in our study it was 400 ␮M. Although 15 ␮M H2O2 is within the physiological range of extracellular and intracellular H2O2, 400 ␮M H2O2 may correspond to pathological levels in some cells (39, 49). Thus whether ROS produces beneficial or injurious effects critically depends on concentration, exposure time, and intrinsic properties of cells; excessive generation of ROS can lead to progressive apoptotic cell death. Our data provide several lines of evidence in support of the idea that the mitochondrial death pathway is critical for progressive apoptosis in H9c2 cells. First, a clear sequence of events was associated with mitochondrial function: ROS production 3 ⌬⌿m disruption 3 cytochrome c release 3 caspase-3 activation 3 cell death. The earliest change identified was significant ROS production as early as 10 min after H2O2 withdrawal. After ROS production, ⌬⌿m dissipation was 286 • JUNE 2004 •

www.ajpheart.org


initiated 15 min after H2O2 withdrawal, but not until 30 min was significant breakdown of ⌬⌿m seen. Cytochrome c release began at 30 min, and significant caspase-3 activation was detected 3 h later. All these changes occurred ahead of significant morphological alterations and DNA fragmentation of the cells. Second, caspase-7, which is not associated with the mitochondrial death pathway (5, 46), was activated only slightly and to a far less extent than caspase-3. Third, L cells, which have fewer mitochondria than H9c2 cells, did not show post-H2O2 apoptosis under the same experimental conditions. Fourth, vitamin E efficiently reduced the intracellular ROS level and prevented the post-H2O2 apoptotic cell death. DTT, another antioxidant, also produced effects similar to those produced by vitamin E. Finally, the results from our GSH experiments suggest that the post-H2O2 apoptotic cell death was not due to depletion of antioxidant buffering capacity of the cells, because 2 h after transient exposure to H2O2, when the amount of cell death became significant, intracellular GSH remained as high as 77% of the control level, and 5 h after H2O2 withdrawal, when DNA fragmentation nearly reached the maximum degree, GSH was increased to 85% of the control level. By comparison, with sustained exposure, GSH decreased to less than half of the control level at 2 and 5 h after incubation with H2O2. Evidently, although the endogenous antioxidant buffering capacity was significantly impaired in cells with sustained exposure to H2O2, it was well reserved in cells after transient exposure to the same concentration of H2O2. Collectively, it is conceivable that ROS production in mitochondria was triggered by the initial H2O2 insult. The increased ROS can, on one hand, further induce more ROS production (ROS-induced ROS production) (50) in the same cells and can also diffuse to the neighboring cells and cause ROS production in these cells. Thus the time-dependent increase in apoptosis in our model is likely due to a timedependent increase in mitochondrial ROS production that further leads to activation of the mitochondrial death pathway. The time-dependent augmentation of mitochondrial ROS production is an initial step in this self-destructive process. Although the importance of mitochondria in determining the fate of cells (life or death) has been well established (5, 14, 24, 29, 36, 44, 46), our study is the first to demonstrate that dysfunction of mitochondria is not merely pivotal in committing cells to die but is also critical in maintaining progressive apoptosis without the continuous presence of environmental cues. It appears that disruption of ⌬⌿m is a critical step toward the post-H2O2 apoptosis in our model and is actually considered by many investigators to constitute a point of no return in the cell death process (5, 14, 24, 29, 36, 44, 46). Yet it is unclear how changes in ⌬⌿m were initiated. One possible mechanism is activation of the apoptosis signal-regulating kinase-1-JNK-baxmitochondrion pathway (40, 41), provided that all these components are present in H9c2 cells and activate in the sequential order in response to H2O2 stress. It is possible that transient activation of JNK in our model plays a part in activation of the mitochondrial death pathway. Indeed, Aoki et al. (6) recently demonstrated an important role of JNK in activation of mitochondrial apoptosis machinery in adult rat ventricular myocytes. Alternatively, ⌬⌿m dissipation could be attributed to the altered function of mitochondrial ATP-sensitive K⫹ channels; several recent studies have documented the importance of mitochondrial ATP-sensitive K⫹ channels in maintaining ⌬⌿m in cardiac myocytes (3, 20). Finally, AJP-Heart Circ Physiol • VOL

H2181

it is also likely that the mechanism underlying ⌬⌿m breakdown is a consequence of a direct damaging effect on the mitochondrial membrane caused by ROS produced in mitochondria. The data from our experiments with vitamin E, DTT, rotenone, and L cells provided strong evidence in support of this notion. Alternatively, the delayed cell death after H2O2 withdrawal could be explained by the possibility that the endogenous antioxidant reserve protects the cells against the initial oxidative insult at the early stage. Thereafter, the sustained endogenous ROS production after H2O2 withdrawal eventually builds up and overwhelms or depletes the endogenous antioxidant reserve, leading to progressive apoptotic cell death. However, our data argue against this view; the intracellular ROS reached a maximum level within 2 h, and the GSH level remained as high as ⬃80% of the control level, whereas cell death reached a maximum level at 8 h after H2O2 withdrawal. This finding favors the hypothesis that ROS acts as a trigger of the death-signaling pathways. In addition to activation of the mitochondrial death pathway, endogenous production of ROS may activate other apoptosisregulating signaling pathways to a certain extent, such as p38 MAPK, ERK, and PKC, which may also contribute to this delayed, time-dependent apoptotic cell death. There are numerous reports regarding the involvement of these signaling kinases in apoptosis (9, 19, 35). However, in this study, we did not find significant activation of these kinases, and inhibitors of these kinases failed to alter the progressive apoptosis (data not shown). Although we cannot rule out the participation of these signaling pathways, we are convinced that endogenous production of ROS in mitochondria and activation of the mitochondrial death pathway played a major role in our progressive apoptosis model. Because massive production of ROS is the underlying mechanism, use of antioxidants would be an efficient way to protect cells from apoptosis induced by transient oxidative insult. ACKNOWLEDGMENTS The authors thank XiaoFan Yang for excellent technical support and Louis Villeneuve for assistance in confocal microscope examination. GRANTS This work was supported in part by the Canadian Institutes of Health Research, the Heart and Stroke Foundation of Quebec, and the Fonds de la Recherche de l’Institut de Cardiologie de Montreal. Z. Wang is a research scholar of the Fonds de Recherche en Sante´ de Quebec. H. Wang is a research fellow of the Canadian Institutes of Health Research, and H. Han is a research fellow of the Heart and Stroke Foundation of Canada. REFERENCES 1. Aikawa R, Komuro I, Yamazaki T, Zou Y, Kudoh S, Tanaka M, Shiojima I, Hiroi Y, and Yazaki Y. Oxidative stress activates extracellular signal-regulated kinases through Src and Ras in cultured cardiac myocytes of neonatal rats. J Clin Invest 100: 1813–1821, 1997. 2. Aime-Sempe C, Folliguet T, Rucker-Martin C, Krajewska M, Krajewska S, Heimburger M, Aubier M, Mercadier JJ, Reed JC, and Hatem SN. Myocardial cell death in fibrillating and dilated human right atria. J Am Coll Cardiol 34: 1577–1586, 1999. 3. Akao M, Ohler A, O’Rourke B, and Marban E. Mitochondrial ATPsensitive potassium channels inhibit apoptosis induced by oxidative stress in cardiac cells. Circ Res 88: 1267–1275, 2001. 4. Antunes F, Cadenas E, and Brunk UT. Apoptosis induced by exposure to a low steady-state concentration of H2O2 is a consequence of lysosomal rupture. Biochem J 365: 549 –555, 2001. 5. Anuradha CD, Kanno S, and Hirano S. Oxidative damage to mitochondria is a preliminary step to caspase-3 activation in fluoride-induced apoptosis in HL-60 cells. Free Radic Biol Med 31: 367–373, 2001. 286 • JUNE 2004 •

www.ajpheart.org

H2182


6. Aoki H, Kang PM, Hampe J, Yoshimura K, Noma T, Matsuzaki M, and Izumo S. Direct activation of mitochondrial apoptosis machinery by c-Jun N-terminal kinase in adult cardiac myocytes. J Biol Chem 277: 10244 –10250, 2002. 7. Behl CJ. Apoptosis and Alzheimer’s disease. J Neural Transm 107: 1325–1344, 2000. 8. Bennett MR. Reactive oxygen species and death: oxidative DNA damage in atherosclerosis. Circ Res 88: 648 – 650, 2001. 9. Birkenkamp KU, Tuyt LM, Lummen C, Wierenga AT, Kruijer W, and Vellenga E. The p38 MAP kinase inhibitor SB203580 enhances nuclear factor-␬B transcriptional activity by a non-specific effect upon the ERK pathway. Br J Pharmacol 131: 99 –107, 2000. 10. Bogoyevitch MA, Ng DCH, Court NW, Draper KA, Dhillon A, and Abas L. Intact mitochondrial electron transport function is essential for signaling by hydrogen peroxide in cardiac myocytes. J Mol Cell Cardiol 32: 1469 –1480, 2000. 11. Bychkov R, Pieper K, Ried C, Milosheva M, Bychkov E, Luft FC, and Haller H. Hydrogen peroxide, potassium currents, and membrane potential in human endothelial cells. Circulation 99: 1719 –1725, 1999. 12. Chen YR, Wang X, Templeton D, Davis RJ, and Tan TH. The role of c-Jun N-terminal kinase (JNK) in apoptosis induced by ultraviolet C and ␥ radiation. Duration of JNK activation may determine cell death and proliferation. J Biol Chem 271: 31929 –31936, 1996. 13. Chow CK, Ibrahim W, Wei Z, and Chan AC. Vitamin E regulates mitochondrial hydrogen peroxide generation. Free Radic Biol Med 27: 580 –587, 1999. 14. Di Lisa F, Blank PS, Colonna R, Gambassi G, Silverman HS, Stern MD, and Hansford RG. Mitochondrial membrane potential in single living adult rat cardiac myocytes exposed to anoxia or metabolic inhibition. J Physiol 486: 1–13, 1995. 15. Dumont EA, Hofstra L, van Heerde WL, van den Eijnde S, Doevendans PA, DeMuinck E, Daemen MA, Smits JF, Frederik P, Wellens HJ, Daemen M, and Reutelingsperger CP. Cardiomyocyte death induced by myocardial ischemia and reperfusion: measurement with recombinant human annexin-V in a mouse model. Circulation 102: 1564–1568, 2000. 16. Ferrari R, Agnoletti L, Comini L, Gaia G, Bachetti T, Cargnoni A, Ceconi C, Curello S, and Visioli O. Oxidative stress during myocardial ischemia and heart failure. Eur Heart J 19: B2–B11, 1998. 17. Hamet P, deBlois D, Dam TV, Richard L, Teiger E, Tea BS, Orlov SN, and Tremblay J. Apoptosis and vascular wall remodeling in hypertension. Can J Physiol Pharmacol 74: 850 – 861, 1996. 18. Han H, Wang H, Long H, Nattel S, and Wang Z. Oxidative preconditioning and apoptosis in L-cells: roles of protein kinase B and mitogenactivated protein kinases. J Biol Chem 276: 26357–26364, 2001. 19. Harada J and Sugimoto M. An inhibitor of p38 and JNK MAP kinases prevents activation of caspase and apoptosis of cultured cerebellar granule neurons. Jpn J Pharmacol 79: 369 –378, 1999. 20. Holmuhamedov EL, Jovanovic S, Dzeja P, Jovanovic S, and Terzic A. Mitochondrial ATP-sensitive K⫹ channels modulate cardiac mitochondrial function. Am J Physiol Heart Circ Physiol 275: H1567–H1576, 1998. 21. Hong F, Kwon SJ, Jhun BS, Kim SS, Ha J, Kim SJ, Sohn NW, Kang C, and Kang I. Insulin-like growth factor-1 protects H9c2 cardiac myoblasts from oxidative stress-induced apoptosis via phosphatidylinositol 3-kinase and extracellular signal-regulated kinase pathways. Life Sci 68: 1096 –1105, 2001. 22. Ide T, Tsutsui H, Kinugawa S, Suematsu N, Hayashidani S, Ichikawa K, Utsumi H, Machida Y, Egashira K, and Takeshita A. Direct evidence for increased hydroxyl radicals originating from superoxide in the failing myocardium. Circ Res 86: 152–157, 2000. 23. Jaffe R, Flugelman MY, Halon DA, and Lewis BS. Ventricular remodeling: from bedside to molecule. Adv Exp Med Biol 430: 257–266, 1997. 24. Kluck RM, Bossy-Wetzel E, Green DR, and Newmeyer DD. The release of cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis. Science 275: 1132–1136, 1997. 25. Li Z, Bing OH, Long X, Robinson KG, and Lakatta EG. Increased cardiomyocyte apoptosis during the transition to heart failure in the spontaneously hypertensive rat. Am J Physiol Heart Circ Physiol 272: H2313–H2319, 1997. 26. Li P, Dietz R, and von Harsdorf R. Differential effect of hydrogen peroxide and superoxide anion on apoptosis and proliferation of vascular smooth muscle cells. Circulation 96: 3602–3609, 1997. 27. Markesbery WR. Oxidative stress hypothesis in Alzheimer’s disease. Free Radic Biol Med 23: 134 –147, 1997.


28. Maulik N, Yoshida T, and Das DK. Oxidative stress developed during the reperfusion of ischemic myocardium induces apoptosis. Free Radic Biol Med 24: 869 – 875, 1998. 29. Mignotte B and Vayssiere JL. Mitochondria and apoptosis. Eur J Biochem 252: 1–15, 1998. 30. Nunomura A, Perry G, Aliev G, Hirai K, Takeda A, Balraj EK, Jones PK, Ghanbari H, Wataya T, Shimohama S, Chiba S, Atwood CS, Petersen RB, and Smith MA. Oxidative damage is the earliest event in Alzheimer disease. J Neuropathol Exp Neurol 60: 759 –767, 2001. 31. Offen D, Elkon H, and Melamed H. Apoptosis as a general cell death pathway in neurodegenerative diseases. J Neural Transm Suppl 58: 153– 166, 2000. 32. Ren JG, Xia HL, Just T, and Dai YR. Hydroxyl radical-induced apoptosis in human tumor cells is associated with telomere shortening but not telomerase inhibition and caspase activation. FEBS Lett 488: 123–132, 2001. 33. Sabbah HN, Sharov VG, and Goldstein S. Programmed cell death in the progression of heart failure. Ann Med 30: S33–S38, 1998. 34. Shimohama S. Apoptosis in Alzheimer’s disease—an update. Apoptosis 5: 9 –16, 2000. 35. Stevenson MA, Pollock SS, Coleman CN, and Calderwood SK. X-irradiation, phorbol esters, and H2O2 stimulate mitogen-activated protein kinase activity in NIH-3T3 cells through the formation of reactive oxygen intermediates. Cancer Res 54: 12–15, 1994. 36. Susin SA, Lorenzo HK, Zamzami N, Marzo I, Snow BE, Brothers GM, Mangion J, Jacotot E, Costantini P, and Loeffler M. Molecular characterization of mitochondrial apoptosis-inducing factor. Nature 397: 441– 446, 1999. 37. Tada-Oikawa S, Oikawa S, Kawanishi M, Yamada M, and Kawanishi S. Generation of hydrogen peroxides precedes loss of mitochondrial membrane potential during DNA alkylation-induced apoptosis. FEBS Lett 442: 65– 69, 1999. 38. Tan S, Sagara Y, Liu Y, Maher P, and Schuber D. The regulation of reactive oxygen species production during programmed cell death. J Cell Biol 141: 1423–1432, 1998. 39. Thannickal VJ and Fanburg BL. Reactive oxygen species in cell signaling. Am J Physiol Lung Cell Mol Physiol 279: L1005–L1028, 2000. 40. Tibbles LA and Woodgett JR. The stress-activated protein kinase pathways. Cell Mol Life Sci 55: 1230 –1254, 1999. 41. Tobiume K, Matsuzawa A, Takahashi T, Nishitoh H, Morita K, Takeda K, Minowa O, Miyazono K, Noda T, and Ichijo H. ASK1 is required for sustained activations of JNK/p38 MAP kinases and apoptosis. EMBO Rep 2: 222–228, 2001. 42. Turner NA, Xia F, Azhar G, Zhang X, Liu L, and Wei JY. Oxidative stress induces DNA fragmentation and caspase activation via the c-Jun NH2-terminal kinase pathway in H9c2 cardiac muscle cells. J Mol Cell Cardiol 30: 1789 –1801, 1998. 43. Vanden Hoek TL, Becker LB, Shao Z, Li C, and Schumacker PT. Reactive oxygen species release from mitochondria during brief hypoxia induces preconditioning in cardiomyocytes. J Biol Chem 273: 18092– 18098, 1998. 44. Vanden Hoek TL, Shao Z, Li C, Schumacker PT, and Becker LB. Mitochondrial electron transport can become a significant source of oxidative injury in cardiomyocytes. J Mol Cell Cardiol 29: 2441–2450, 1997. 45. Vrablic AS, Albright CD, Craciunescu CN, Salganik RI, and Zeisel SH. Altered mitochondrial function and overgeneration of reactive oxygen species precede the induction of apoptosis by 1-O-octadecyl-2-methylrac-glycero-3-phosphocholine in p53-defective hepatocytes. FASEB J 15: 1739 –1744, 2001. 46. Wang GW, Klein JB, and Kang YJ. Metallothionein inhibits doxorubicin-induced mitochondrial cytochrome c release and caspase-3 activation in cardiomyocytes. J Pharmacol Exp Ther 298: 461– 468, 2001. 47. Wang H, Zhang Y, Han H, Cao L, Wang J, Long H, Nattel S, and Wang Z. HERG K⫹ channel: a regulator of tumor cell apoptosis and proliferation. Cancer Res 62: 4843– 4848, 2002. 48. Wang X, McCullough KD, Franke TF, and Holbrook NJ. Epidermal growth factor receptor-dependent Akt activation by oxidative stress enhances cell survival. J Biol Chem 275: 14624 –14631, 2000. 49. Willenheimer R. Left ventricular remodelling and dysfunction. Can the process be prevented? Int J Cardiol 72: 143–150, 2000. 50. Zorov DB, Filburn CR, Klotz LO, Zweier JL, and Sollott SJ. Reactive oxygen species (ROS)-induced ROS release: a new phenomenon accompanying induction of the mitochondrial permeability transition in cardiac myocytes. J Exp Med 192: 1001–1014, 2000.

286 • JUNE 2004 •

www.ajpheart.org