Vital imaging of H9c2 myoblasts exposed to tert-butylhydroperoxide – characterization of morphological features of cell death

The myoblast cell line H9c2, derived from embryonic rat heart [1], has been used as an in vitro model for both skeletal and cardiac muscle. H9c2 cells show electrophysiological and biochemical properties of both skeletal and cardiac tissues, including depolarization in response to acetylcholine [1], and rapid activation of calcium currents through L-Type channels [2–4]. An interesting feature of this cell line is its ability to differentiate from mono-nucleated myoblasts to myotubes upon reduction of serum concentration [2]. Accompanying myotube formation is the expression of myogenic transcription factors [5], calcium channel proteins [6], and the LIM protein FHL2 [7]. During the differentiation process, cells retain several elements of the electrical and hormonal signaling pathway of cardiac cells [1] and have therefore become an accepted in vitro model to study the effects of ischemia and diabetes on the heart [8–10].

L'Ecuyer et al. (2001) [11] demonstrated that H9c2 cells can be used to study free radical production, and can be engineered to express foreign genes at controllable levels, making them a suitable system to study molecular responses to oxidative damage. Previous works in this area include studies of oxidative cell damage caused by doxorubicin [11–13], ischemia and reperfusion [14], hydrogen peroxide [15, 16] and peroxynitrite [17].

Tert-butyl hydroperoxide (tBHP) is one of the most common pro-oxidant agents used to evaluate the effects of oxidative stress [18], and it has been found to induce cell death in several cell lines such as U937 [19], HepG2 [20], and primary cultures of cardiomyocytes [21] and hepatocytes [18, 22]. According to the cell line used, induction of cell death by tBHP is characterized by cytochrome c release [23], increased expression of p53 [24], induction of the mitochondrial permeability transition pore [25], TUNEL-positive labeling, phosphatidylserine (PS) exposure and caspase 3 activation [26].

However, little is currently known regarding the specific behavioral responses of H9c2 cells exposed to pro-oxidants. Therefore, the objective of the present work was to use tBHP to study the morphological events accompanying pro-oxidant damage in H9c2 myoblasts, with a special focus on mitochondrial changes. Furthermore, the morphological alterations resulting from tBHP treatment were correlated with well established markers of apoptosis and necrosis, including chromatin condensation, PS exposure in the outer leaflet of the cell membrane, BAX translocation, and loss of plasma membrane integrity. This study provides novel information regarding the behavioral responses of a muscle cell line to pro-oxidant damage. Furthermore, the data obtained permits the extension of our understanding of the conditions that drive cells from apoptosis to necrosis, which are distinct methods of cell death that embody dissimilar medical consequences for the affected target organ.

Before analyzing changes in cellular morphology induced by tBHP, the pro-oxidant action of this compound was confirmed in H9c2 cells. Increased intracellular oxidative stress was detected by the oxidation of dichlorofluorescein (DCF) in cells that were exposed to 50 μM tBHP for one hour (Figure 1). Most cells suffering from high oxidative stress remained well-spread (compare Fig. 1A and 1C); however, small membrane blebs started to become evident at the cell periphery. Quantitation of mean DCF fluorescence intensity in control and treated cells was also performed (Figure 1 – bottom panel). The results showed a statistically significant increase in DCF oxidation, confirming increased intracellular oxidative stress after 1 hour treatment with 50 μM tBHP.

By measuring cell numbers with the sulforhodamine B dye method [27], the protective effect of antioxidants N-acetylcysteine and Trolox (a water-soluble vitamin E derivative) was evaluated against tBHP cytotoxicity (Figure 2, top panel). The results showed a significant decrease in H9c2 cell number after 6 hours of treatment with 50 and 100 μM tBHP, compared with the control. The decrease in cell numbers due to tBHP induced cell death was prevented by N-acetylcysteine and Trolox (Figure 2, top). In order to analyze if the cytotoxic effect of tBHP was related to the induction of the mitochondrial permeability transition (MPT), an oxidative stress-dependent phenomenon characterized by the loss of the impermeability of the mitochondrial inner membrane, we also analyzed the effect of cyclosporin-A (a well-known MPT inhibitor). As presented in Figure 2, bottom panel, cyclosporin-A did not prevent the toxicity of tBHP.

Morphological alterations of H9C2 cells after treatment with tBHP were analyzed by epifluorescence microscopy (Fig. 3). Control (untreated) cells are well spread, strongly labeled by calcein, and exhibit filamentous mitochondria distributed throughout the cytoplasm (Fig. 3B). However, after tBHP treatment, mitochondria lost their filamentous shape and displayed several morphological alterations. After 45 min exposure to 80 μM tBHP, mitochondria were concentrated close to the nuclear region and in some cells, the amount of polarized mitochondria (mitochondria with negative-inside electric potential) decreased (Fig. 3C). At 90 minutes, the mitochondrial network in some cells showed signs of degradation, being converted from a filament-like form to small round mitochondria (Fig. 3D and 3F). Exposure of H9c2 cells to 25 μM tBHP for three hours induced some cells to round up and exhibit increased membrane blebbing (Fig. 3G). Morphologically, cells treated with 25 μM tBHP display peripheral membrane blebbing, followed by rounding up of cells as membrane blebbing continues.

Interestingly, examination of triple-labeled cells (Hoechst, calcein and TMRM) by confocal microscopy, demonstrated that mitochondria could remain polarized inside apoptotic bodies (Fig 3H, arrow).

To verify if the morphological changes resulting from tBHP treatment were accompanied by changes in mitochondrial-associated apoptotic proteins, H9c2 cells were fixed and labeled with an antibody to the pro-apoptotic protein Bax. After 1 hour treatment with 50 μM tBHP, we observed the formation of Bax aggregates close to the mitochondrial network (figure 4, panel E and J). Bax clusters were also observed in some cells after treatment with 100 μM tBHP (figure 4, panel H). Examination of cells at high magnification indicates that the Bax aggregates are located both in and around clusters of mitochondria (figure 4, panel J and K). Statistical analyses of cell fluorescence intensities for Hoechst, Bax immunolabeling and Mitotracker Red (Figure 4, bottom panel) were also performed. The results show a statistically significant increase for both Hoechst and Bax fluorescence after treatment with 100 μM tBHP, as opposed to a decrease in the fluorescence intensity of Mitotracker Red. Bax clusters located in and around mitochondria were also observed after treatment with 50 μM tBHP.

To examine whether phosphatidylserine (PS) exposure occurs in H9c2 cells treated with tBHP, control and oxidant-treated cultures were dual-labeled with annexin V and propidium iodide (PI). Because PI is excluded from cells with intact plasma membranes, it was used to determine whether annexin V labeling was due to externalization of PS, or instead resulted from a loss of membrane integrity (Fig. 5). Control cells were neither PI nor Annexin V positive, while treatment with 50 μM tBHP for 1 hour caused some cells to become Annexin V positive (0 vs. 53.3 ± 17.6%, p < 0.05 vs control). Although an increase in double PI/Annexin V labeling (necrotic cells) was observed (0 vs. 16.7 ± 12.0%), the result was not statistically significant (p > 0.05 vs control). In contrast, one hour treatment with 100 μM tBHP caused a significant increase in the labeling of both PI and Annexin V (0 vs. 58.4 ± 5.7%, p < 0.05 vs control). No differences in Annexin V labeling were found between the 100 μM tBHP treatment groups and controls (0 vs. 8.3 ± 8.3%, p > 0.05 vs control). The results demonstrated that 100 μM tBHP could lead to both necrosis and accelerated apoptosis after just 1 hour of treatment.

Nuclear morphologic changes typical of apoptosis were also detected by epifluorescence observation of tBHP-treated cells stained with Hoechst 33342 (Fig. 6, upper panel). After treatment with different concentrations of tBHP (25–100 μM) during 3 hours, we detected a dose-dependent increase in the number of nuclei showing condensed chromatin. However, a statistically significant difference was only found for 100 μM tBHP (Fig. 6, lower panel). On the other hand, 6 hours of tBHP exposure induced a statistically significant increase in the number of apoptotic nuclei for all concentrations tested (Fig 6, lower panel). No differences between control and treated cells were observed after 1 hour of treatment (data not shown).

Finally, we were interested in following by real-time epifluorescence microscopy the transition between apoptosis and necrosis after treatment with 100 μM tBHP. In order to analyze the transition, the loss of calcein fluorescence and an increase in nuclear-derived ethidium homodimer fluorescence were detected after treatment with tBHP. Figure 7 shows sequential images, beginning at 120 min. of tBHP exposure, from one field of view that contained one blebbing cell and one well-spread cell. The well-spread cell (white arrow) slightly shrank in size without significant membrane blebbing and suffered necrosis, as observed by the loss of calcein fluorescence, increase in ethidium homodimer red fluorescence and visible extrusion of cytosolic contents. The blebbing cell (yellow arrow) shrank in size, but ultimately also underwent necrosis, as observed by the loss of calcein fluorescence, gain in ethidium homodimer red fluorescence and visible extrusion of cytosolic contents. Notably, the well-spread cell maintained the majority of its basic structure unaltered while the blebbing cell lost its original form and rounded up. Strikingly, at 150 min., no fluorescence whatsoever was observed coming from the cells. Presumably, this represents a lag between loss of calcein fluorescence from the cytosol and permeabilization of the plasma membrane, allowing entry of ethidium homodimer.

Tert-butyl hydroperoxide (tBHP) is a membrane-permeant pro-oxidant agent used in several cell lines as a model to study the effects of oxidative stress on cellular function and cell death pathways [18, 19, 21, 22]. Once inside cells, tBHP generates tert-butoxy radicals inducing several physiological alterations with consequent loss of cell viability via apoptosis or necrosis [28]. The type of effect is dependent on the cell line or primary culture type, tBHP concentration, and exposure period. Lipid peroxidation [28], depletion of intracellular reduced glutathione [29], modification of protein thiols [30] and cytosolic calcium deregulation [20] are some of the most common alterations.

In cardiomyocytes, tBHP induced loss of cell shape, depletion of ATP, and formation of adenosine [21]. Other studies demonstrated that treatment with 1 mM tBHP lead to a rise in intracellular calcium concentration, hyper-contracture and loss of membrane integrity in cardiac myocytes isolated from rat ventricles [31]. To the best of our knowledge, however, morphological alterations in cardiomyocytes or myoblasts during tBHP-induced cell death has not been reported.

In the present study, tBHP was used as a model compound to characterize morphologic changes in H9c2 cells resulting from oxidative stress. It was found that 1 hour treatment with 50 μM tBHP induces an increase in the oxidation of the probe DCF, which indicates the presence of intracellular oxidative stress. The data also indicates that after 6 hours treatment with 50 and 100 μM tBHP, cell death and detachment occurred. These effects were prevented by N-acetylcysteine (NAC) and Trolox. NAC is the acetylated form of the amino acid L-cysteine and a source of sulfhydryl (SH) groups. In the body, NAC is converted into metabolites capable of stimulating glutathione (GSH) synthesis and can also act directly as a free radical scavenger, due to its nucleophilic and antioxidant properties [32, 33]. Trolox is a water-soluble analogue of vitamin E lacking the phytyl chain, with strong antioxidant properties [34]. As reported in Figure 2 (top panel) both antioxidant compounds prevented the cytotoxic effect induced by tBHP, supporting the hypothesis that ROS production in H9c2 cells is responsible for the cell death that occurs following tBHP treatment. Moreover, vital imaging studies demonstrate that some features of apoptosis, including cell rounding, membrane blebbing, and chromatin condensation, occur in cells undergoing tBHP-induced oxidative stress. However, cells also remain calcein-positive, and mitochondrial membrane potential can persist for quite some time in these apoptotic cells.

Apoptosis is a cell death program that is dependent on ATP, most of which is normally produced by mitochondria under aerobic conditions. Although tBHP-treated apoptotic cells displayed significant changes in mitochondrial morphology, many altered mitochondrial remained polarized. On the other hand, most necrotic cells lacked polarized mitochondria. The results suggest that functional mitochondria are necessary for tBHP-induced apoptosis in H9c2 cells. Nevertheless, although morphologically compromised, the mitochondrial network retains sufficient function to continue producing enough ATP to drive apoptosis in oxidatively damaged cells. Transformation of filamentous mitochondria to small spherical forms, as observed in this study, has been termed the "thread-grain transition" and has been proposed to represent a mechanism to isolate a damaged part of the mitochondrial system from the rest of the mitochondrial network [35]. According to this hypothesis, thread-grain transitions represent an obligatory step in mitochondrial-mediated apoptosis. In addition to thread-grain transitions, we noted that polarized mitochondria became concentrated near the nuclear region upon treatment with tBHP. Skulachev et al. [35] proposed that small mitochondria around the nucleus may serve to more rapidly direct some apoptotic proteins to their nuclear targets. Alternatively, concentrating polarized mitochondria around the nucleus could be an important mechanism to supply the energy needed by the nucleus during the apoptotic program.

Immunocytochemistry of H9c2 cells revealed the formation of Bax aggregates near the mitochondrial network after tBHP treatment (Figure 4). This labeling appeared to be most pronounced in areas of the mitochondrial network that displayed the weakest Mitotracker Red labeling. Mitotracker Red is a widely used mitochondrial marker, which is membrane potential dependent (according to the manufacturer). In these experiments, control and tBHP-treated cultures were processed at the same time and under identical conditions for Mitotracker, Bax and Hoechst labeling. Both control and experimental samples were photographed in the same session using identical microscope and camera settings. Visual examination showed consistent differences in fluorescence intensities of these probes between control and tBHP-treated groups, and statistical analyses of fluorescence values from digitized images demonstrated a significant inverse relationship between Mitotracker Red and Bax immunolabeling (Figure 6, lower panel). We also observed a direct association between Bax immunolabeling and Hoechst staining, showing that increased Bax labeling accompanies the stronger Hoechst labeling, associated with chromatin condensation in apoptotic tBHP-treated H9c2 cells.

Bax has been suggested to be translocated from the cytosol of cardiomyocytes in two distinct phases [36]. The first phase involves the Bax-induced release of cytochrome c, and the second phase involves packaging of Bax monomers close to mitochondria. Our results are consistent with this model in that we observed a translocation of Bax around mitochondria upon treatment with tBHP, which is further evidence that tBHP is inducing apoptosis. Considering that Bax translocation to mitochondria can be accompanied by morphological alterations in these organelles, we hypothesized that the mitochondrial permeability transition (MPT) could be involved as well. It has been reported by other authors that the mitochondrial permeability transition is an oxidative stress-dependent mechanism [37]. As tBHP induces oxidative stress in H9c2 cells, we also suspected that the MPT could be another reason for tBHP-induced H9c2 cytotoxicity. In order to test this hypothesis, we examined whether tBHP cytotoxicity could be prevented with cyclosporin-A, a MPT inhibitor. The results (Figure 2, bottom panel) showed that cyclosporin-A did not prevent cell death induced by tBHP, indicating that the MPT does not appear to be cause for the cytotoxic effect induced by tBHP under our experimental conditions.

Changes in plasma membrane asymmetry are one of the earliest features of cells undergoing apoptosis. In apoptotic cells, phosphatidylserine (PS) is translocated from the inner to the outer leaflet of plasma membrane, which serves as a recognition signal for macrophages [38]. Annexin V binds PS, and is commonly used to detect externalized PS in apoptotic cells. However, because Annexin V can also label non-externalized PS in necrotic cells with compromised plasma membranes, propidium iodide (PI) is commonly used together with annexin V to identify and distinguish necrotic from apoptotic cells [39]. Our results demonstrate that tBHP-treatment is able to induce the externalization of PS, as evidenced by the presence of annexin V-positive but PI-negative cells. Additionally, treatment with lower concentrations of tBHP for short periods of time induced an increase in the number of nuclei showing condensed chromatin, characteristic of apoptosis (Figure 6).

Cell membrane behavior after treatment with tBHP was also evaluated by DIC imaging (Figures 3, 5 and 7). tBHP-treated H9c2 cells shrank in size while undergoing membrane blebbing and releasing apoptotic bodies. Nevertheless, during this period, cell membrane integrity is maintained as seen by the maintenance of calcein fluorescence inside the cells and PI exclusion (Figure 7). In many cells, this apoptotic period was followed by necrosis, and release of cellular contents.

In conclusion, exposure of H9c2 cells to tBHP results in: 1) an increase in ROS production, 2) cell rounding and membrane blebbing, 3) alterations in the mitochondrial network, including peri-nuclear clustering and thread-grain transitions, 4) the appearance of immunoreactive Bax deposits around depolarized mitochondria, 5) translocation of PS to the outer leaflet of the plasma membrane, and 6) chromatin condensation and nuclear shrinkage. All of these changes are suggestive of apoptosis; however, depending on the concentration and exposure time, H9c2 cells can also undergo necrosis, as evidenced by a rapid loss of plasma membrane integrity. Our results demonstrate the causal relationship between increased oxidative stress and cell death and emphasize the importance of vital imaging of intact cells to more completely understand the morphological and functional changes that occur during the induction of different modes of cell death by toxic chemicals.