Suppression of stress kinase JNK is involved in HSP72-mediated protection of myogenic cells from transient energy deprivation. HSP72 alleviates the stewss-induced inhibition of JNK dephosphorylation.

Since protection of cells from stress-induced apoptosis by the heat shock protein Hsp72 involves suppression of stress kinase JNK, we suggested that Hsp72-mediated JNK inhibition might also be critical for myocardial protection from ischemia/reperfusion. Transient energy deprivation of H9c2 myogenic cells, used as an in vitro model of myocardial ischemia, led to cell death that had morphological features of apoptosis and necrosis and was independent of caspases. Surprisingly, this unusual type of cell death was regulated by JNK and ERK kinases. In fact, specific inhibition of JNK increased cell survival; specific inhibition of ERKs enhanced deleterious consequences of energy deprivation, whereas inhibition of p38 kinase had no effect. Hsp72 suppressed activation of JNK and did not increase ERK activity, suggesting that inhibition of JNK is the important component of Hsp72-mediated protection. Upon transient energy deprivation, activation of JNK proceeds via two distinct pathways, stimulation of JNK phosphorylation by a protein kinase SEK1 and inhibition of JNK dephosphorylation. Remarkably, in cells exposed to transient energy deprivation, Hsp72 enhanced the rate of JNK dephosphorylation but did not affect SEK1 activity. Therefore, it appears that Hsp72 specifically down-regulates JNK by accelerating its dephosphorylation, which reduces the susceptibility of cardiac cells to simulated ischemia/reperfusion.

Following ischemia/reperfusion of the myocardium, cardiomyocytes die via necrosis and apoptosis. Necrosis normally occurs in the myocardial zones subjected to severe ischemia and is characterized by cell swelling and rapid efflux of cytosolic constituents due to damage of plasma membrane (1,2). Standard assays for evaluation of ischemic myocardial damage are based on detection of cell necrosis (e.g. release of cytosolic enzymes such as creatine kinase); however, a significant fraction of cardiac cells exposed to ischemia/reperfusion undergoes apoptosis (see Ref. 3 for review). Apoptosis usually occurs in the zones of milder ischemia. It does not lead to plasma membrane damage, which allows tissues to avoid inflammation (3,4).
Apoptosis is characterized by activation of caspases, nuclear condensation, and DNA degradation. Many studies have recently demonstrated that apoptotic death of cardiomyocytes caused by ischemia/reperfusion contributes significantly to the development of infarct as well as to loss of cells surrounding the infarct area (1)(2)(3). In addition, a large fraction of dying cells manifest features of both apoptosis and necrosis, i.e. both plasma membrane damage and nuclear condensation are observed (e.g. Refs. 5 and 6).
The major heat shock protein Hsp72 can protect myocardium against ischemia (see Refs. 7 and 8 for review). For example, in vitro studies showed that expression of recombinant Hsp72 in H9c2 myogenic cells as well as in neonatal cardiomyocytes reduced necrosis under ischemic conditions (9 -11). Moreover, in vivo studies with transgenic mice that overexpress Hsp72 in myocardium demonstrated a significant reduction in the size and creatine kinase efflux of the infarct as well as improvement of contractile recovery following ischemia reperfusion (12)(13)(14). Furthermore, endogenous Hsp72, when induced by mild hyperthermia, confers resistance to ischemia (15,16). Recently, new data emerged that Hsp72 also prevents myocardial apoptosis following ischemia. For example, both induction of Hsp72 in H9c2 myogenic cells by cardiotrophin-1 (17) or overexpression of Hsp72 in neonatal cardiac myocytes via adenovirus vector protected cells from ischemia-induced apoptosis (18).
It has been established that Hsp72 and other members of the Hsp70 protein family function as molecular chaperones reducing aggregation of certain proteins in heat-shocked or energydeprived cells (see Refs. 7 and 19 for review). On the other hand, the anti-apoptotic effect of Hsp72 in heat-stressed cells can be related to suppression of the stress kinase c-Jun Nterminal kinase (JNK) 1 (20,21). Studies from many groups have established that the stress-activated protein kinase JNK plays a crucial role in induction of apoptosis by a variety of stresses (e.g. heat shock, ceramide, and radiation) (22)(23)(24). Therefore, we suggested that suppression of JNK activity by Hsp72 might also be critical for the anti-apoptotic action of this protein in ischemic myocardium (25).
Although ischemia reperfusion is a powerful activator of JNK and p38, a homologous stress kinase in the myocardium (see Ref. 26 for review), involvement of these kinases in ischemia-induced cell death remains to be elucidated. Surprisingly, unlike many other cell types, selective activation of either JNK or p38 alone in cardiomyocytes does not cause cell death but leads to hypertrophy (27)(28)(29). However, simultaneous activation of JNK and p38 causes programmed death of these cells (29). Activation of JNK and p38 precedes nucleosomal DNA fragmentation (a hallmark of apoptosis) (30), which is consistent with a causal relationship of stress kinase activation and apoptosis. Furthermore, blockade of p38 by a specific inhibitor, SB 203580, has recently been shown to reduce ischemia-induced myocardial apoptosis as well as necrosis (31,32). In contrast to JNK and p38, activation of another kinase, ERK, appears to protect myocardium against ischemia-induced apoptosis and functional injury (33).
There are two major factors contributing to cell death upon ischemia reperfusion of the myocardium, energy deprivation and oxidative stress (see Refs. 7 and 34 for review). Ischemia/ reperfusion can be simulated in vitro by transient energy deprivation of cells. Indeed, transient energy deprivation alone was shown to cause both stress kinase activation (35) and apoptosis (36). Here, employing this model, we tested the hypothesis that the protective effect of Hsp72 in energy deprivation-induced cell death is due to suppression of stress kinase activation.

EXPERIMENTAL PROCEDURES
Cell Culture-H9c2 rat myogenic cells (ATCC) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Cells were grown at 37°C in an atmosphere of 5% CO 2 to 60 -85% confluence.
Plasmids and Transient Transfection of Cells-For transient transfection of cells plasmids expressing JNK1 tagged with HA epitope (JNK1-HA) and dominant negative mutant of SEK1 fused in frame to GST, GST-SEK1 (K/R) was used. These plasmids were kindly provided by Dr. J. Avruch (Massachusetts General Hospital). EGFP plasmid was kindly provided by Dr. J. Kordowska (Boston Biomedical Research Institute). H9c2 cells were transfected using GenePORTER transfection reagent (Gene Therapy Systems, Inc, San Diego, CA). Maximal transfection efficiency (10 -20%) with minimal toxicity was achieved when cells were grown to 50 -70% of confluency, and 1 g of plasmid DNA plus 6 l of GenePORTER in 1 ml of Opti-MEM (Life Technologies, Inc.) was added to each 35-mm plate. Corresponding vectors of plasmids were used for mock transfection and as a carrier when the amount of transfected plasmid of interest was lower than 1 g. After overnight incubation with transfection reagents and plasmids, H9c2 cells were washed and left in growth medium for an additional 24 -30 h to reach maximal expression of plasmid DNA.
Adenovirally Based Expression of Hsp72-A recombinant adenovirus vector expressing Hsp72 (AdTR5/Hsp70-GFP) was constructed by cloning a dicistronic transcription unit encoding human Hsp72 and the Aequorea victoria green fluorescent protein gene, separated by an encephalomyocarditis virus internal ribosome entry site from pTR-DC/ HSP70-GFP (37) into an adenovirus transfer vector. Expression of this transcription unit is controlled by the tetracycline-regulated transactivator protein tTA (38) which we expressed from the separate recombinant adenovirus (AdCMV/tTA,) (39). Twenty four hours after infection, the medium was changed, and the cells were left for 12 more hours. Cell Viability Quantification-The percentage of dead cells after ATP depletion was quantified using a fluorescence microscope after staining with acridine orange (3 g/ml) and propidium iodide (10 g/ml) as described (21). The effect of transfection of dnSEK1 on cell viability was assessed by counting the number of viable transfected (i.e. EGFPexpressing) cells by fluorescence microscopy. In each experiment, at least 10 microscopic fields per plate (300 -500 cells) were counted, and the experiments were repeated 3-4 times.
Preparation of Cytosol Extracts and Analysis of Kinase Activity-Cells on 35-mm dishes were lysed in 200 l of a buffer (40 mM HEPES, pH 7.5; 50 mM KCl; 1% Triton X-100; 2 mM dithiothreitol; 1 mM Na 3 VO 4 ; 50 mM ␤-glycerophosphate; 50 mM NaF; 5 mM EDTA; 5 mM EGTA; 1 mM phenylmethylsulfonyl fluoride; 1 mM benzamidine; 5 g/ml each of leupeptin, pepstatin A, and aprotinin) and aspirated into a microcentrifuge tube. The lysates were clarified by centrifugation in a microcentrifuge at 12,000 ϫ g for 5 min. All the procedures were performed at 4°C. Protein concentration was measured in the supernatants by the Bio-Rad protein assay reagent, and they were diluted with a lysis buffer to achieve equal protein concentrations in all samples.
To measure total JNK activity, 5 l of extracts were added to a reaction mixture (20 l final volume) containing 40 mM HEPES, pH 7.5; 1 mM Na 3 VO 4 ; 25 mM ␤-glycerophosphate; 10 mM MgCl 2 ; 20 M ATP; 15 Ci of [␥Ϫ 32 P]ATP; and 40 ng of recombinant c-Jun-GST. The reaction was allowed to proceed for 30 min at 30°C and then was stopped by addition of 10 l of loading SDS-PAGE buffer. Samples were separated by SDS-PAGE and transferred to nitrocellulose, and the membranes were exposed to a Molecular Imager for quantification. Subsequently, membranes were immunoblotted with a JNK1 antibody to verify equivalence in protein loading. Another assay for JNK activity allowed us to measure separately the activity of two major JNK isoforms, JNK1 and JNK2 (46 and 54 kDa, respectively), using a JNK antibody specific to the activated (phosphorylated) form of JNK.
To measure the activity of transfected JNK1-HA, it was immunoprecipitated for 2 h at 4°C from extracts with an antibody against the HA epitope. Precipitates were washed 3 times with a kinase reaction buffer as follows: 25 mM HEPES, pH 7.4, 10 mM MgCl 2 , 2 mM dithiothreitol, 25 mM ␤-glycerophosphate, 2 mM Na 3 VO 4 . Phosphorylation of GST-c-Jun by immunoprecipitated JNK was carried out in the same buffer in the presence of 20 M ATP and 25 Ci of [␥Ϫ 32 P]ATP at 37°C for 10 min. The samples were then subjected to SDS-PAGE followed by transfer to a nitrocellulose membrane and autoradiography. This membrane was later used for an immunoblot with anti-JNK1 antibody to ensure that equal amounts of the kinase were immunoprecipitated.
Since phosphorylation of p38 is not suitable for measuring the effect of SB 203580 on p38 activity (40), another assay of p38 activity was used. A substrate of p38, MAPKAP-2 kinase, was immunoprecipitated with MAPKAP-2 antibodies for 2 h at 4°C and washed 3 times with the kinase buffer, and MAPKAP-2 kinase activity was measured using Hsp27 (StressGene) as a substrate (40).
Caspase-3 Activation-Cell lysates were prepared as described above and subjected to SDS-PAGE and Western blotting with antibodies to caspase-3 which recognize both inactive procaspase-3 and activated (cleaved) form.
Measurement of ATP Content-Cellular ATP content was measured by the luciferin-luciferase method, using ATP bioluminescence somatic cell assay kit (Sigma). Bioluminescence was determined by scintillation counting (41). The cytosolic ATP content was normalized to total cytosolic protein that was measured with the Bio-Rad protein assay reagent.

Activation of Kinases and Cell Death Caused by Transient
Energy Deprivation-Initially, we characterized the mode of H9c2 myogenic cell death after transient energy deprivation. To deplete cellular ATP, cells were incubated in glucose-free medium in the presence of a mitochondrial inhibitor, rotenone. This treatment caused a rapid decline in ATP level (to 7% of the initial level within 30 min and down to 3% after 1 h). Subsequent transfer of cells to a complete medium restored the ATP level to 60% of the initial level after 1 h of recovery (Fig. 1A). The fraction of dead cells was evaluated by standard assays (staining with the permeable dye acridine orange (AO) and the impermeable dye propidium iodide (PI)). Energy deprivation for less than 3 h did not cause immediate changes in staining.
However, cells began to die after 3 h of recovery, and cell dying progressed over the next 24 h of recovery. Loss of cell viability measured after 24 h of recovery closely correlated with the duration of energy deprivation. Cell viability was not affected for up to 1.5 h of energy deprivation; 2 h of ATP depletion caused death in 12% of the cells, and this fraction sharply increased after 2.5 h of ATP depletion (Fig. 1B). Surprisingly, under these conditions, whereas a small fraction of dead cells demonstrated typical features of apoptosis, i.e. condensed nuclei and intact plasma membrane (lack of PI staining) (Fig. 2B) (seen, for example, with staurosporine, Fig. 2D, or tumor necrosis factor (not shown)), the majority of the dead cells manifested features of both apoptosis and necrosis, i.e. condensed nuclei and permeability to PI (Fig. 2B). This cell death process was not associated with activation of the caspase-3 as it was seen with staurosporine or tumor necrosis factor (Fig. 2E) and was not prevented by pancaspase inhibitor benzyloxycarbonyl-VAD-formylmethyl ketone (data not shown). Therefore, transient energy deprivation causes unusual caspase-independent cell death with certain morphological features of apoptosis. Dead cells with this morphology started to appear at 3 h of recovery, while typical apoptotic cells emerged at a later time. Thus, the morphology of PI-permeable fraction of H9c2 cells following energy deprivation does not represent a late stage of apoptosis. It is noteworthy that energy deprivation for more than 3 h caused typical necrotic death, i.e. cell detachment and plasma membrane damage without nuclear condensation (Fig. 2C).
During ATP depletion, activities of stress kinases decreased to almost an undetectable level (because JNK and p38 could not be phosphorylated by upstream kinases without ATP), whereas restoration of ATP levels increased activities of JNK and p38 more than 5-10 times over the basal levels in untreated cells (Fig. 3A). Maximal activation of the kinases was seen at 15 min after removal of the mitochondrial inhibitor and wide arrows indicate apoptotic cells (condensed nuclei without PI staining). Note that in B regular arrows indicate cells that are permeable to PI and have condensed nuclei, i.e. cells that manifest features of both apoptosis and necrosis. E, lack of activation of the caspase 3 in cells exposed to transient energy depletion. H9c2 cells were exposed to ATP depletion (rot) for 2 h and 2.5 h followed by 24-h recovery, staurosporine (str) for 6 h, and TNF (10 ng/ml) ϩ emetine (10 g/ml) for 6 h (TNF). Formation of the active caspase 3 (fragment) was detected by the immunoblot. procas-3, procaspase-3; con, control. then slowly declined (Fig. 3B). Along with JNK and p38 stress kinases, transient ATP depletion was a powerful activator of homologous MAP kinase, Erk1/2 (Fig. 3C).
The extent of activation of stress kinases appeared to be dependent on the duration of ATP depletion. Indeed, upon transferring cells to complete medium following 0.5 h of ATP depletion, little activation of JNK and p38 was observed (not shown). Slightly stronger but still transient JNK activation was observed after 1 h of ATP depletion (Fig. 3D). Prolongation of ATP depletion up to 2.5 h increased both the level of initial JNK activation and the duration of JNK activity (Fig. 3D); a similar effect was seen with p38 kinase (Fig. 3E). Thus, the longer the duration of ATP depletion, the stronger activation of the stress kinases and the higher level of cell death.
JNK Activation Is Essential for Cell Death Caused by Transient ATP Depletion-The observed correlation suggested that unusual caspase-independent cell death under transient ATP depletion may be affected by JNK, p38, or Erk1/2, similar to what was seen with apoptosis (see Introduction). To address directly the role of JNK in such cell death, we transfected H9c2 with a dominant-negative form of the JNK-regulating kinase SEK1 (dnSEK1) (see e.g. Ref. 23). To follow JNK activity in transfected cells only, the cells were co-transfected with the plasmid expressing JNK1 tagged with the hemagglutinin epitope (JNK1-HA). JNK1-HA activity was assayed after immunoprecipitation with anti-HA antibodies using c-Jun as a substrate. As seen in Fig. 4A, expression of dnSEK1 strongly reduced JNK activation in response to transient energy deprivation. To assess the extent of cell death in the cells transfected with dnSEK1, the cells were co-transfected with a plasmid encoding green fluorescent protein (EGFP). Transfected H9c2 cells were subjected to transient ATP depletion for 2 h, and 16 h later the number of the survived cells (fluorescent with noncondensed nuclei) was assessed with a fluorescent microscope. Expression of dnSEK1 led to about a 2-fold increase in cell survival (Fig. 4C), suggesting that JNK positively regulates the cell death process.
To assess the effect of JNK inhibition on cell death by an alternative approach, we used H9c2 cells constitutively expressing dominant-negative JNK1 mutant (42). Survival of non-transfected (naive) cells after energy deprivation was higher than that of Mock-transfected cells, probably due to a sensitizing effect of transfection (Fig. 4C). In line with the above experiment, inhibition of JNK activity by the dnJNK1 mutant (Fig. 4B) also led to significant enhancement of cell survival following energy deprivation (Fig. 4C). Thus, JNK activation appears to play an important role in induction of cell death by transient energy deprivation.
To elucidate the role of p38 kinase in cell death, we used a specific inhibitor of this kinase, SB 203580, at concentrations ranging from 5 to 20 M (31,32). This inhibitor at 10 M concentration reduced p38 activity by 70%, without affecting JNK activity (not shown). SB 203580 failed to reduce the caspase-independent cell death following ATP depletion/repletion in H9c2 (not shown), indicating that in these cells p38 is not involved in such cell death.
To assess the involvement of Erk1/2 in cell death, we used PD98059, a specific inhibitor of Erk1/2 signaling pathway. At 50 M concentration, PD98059 almost completely prevented Erk1/2 activation (not shown) and dramatically enhanced cell FIG. 3. Activation of JNK, p38, and ERK kinases in cells following ATP depletion. A, JNK and p38 activities in H9c2 cells after 120 min of ATP depletion (ϪATP) and at 15 and 30 min of ATP recovery (ϩATP). JNK activity was measured by an in vitro kinase assay using c-Jun as a substrate, and p38 activity was measured using an antibody to phosphorylated (active) p38 (see "Experimental Procedures" for details). Lane C, control. B, time course of JNK and p38 activities in H9c2 cells during ATP depletion/repletion. JNK1, JNK2, and p38 activities were measured by immunoblotting with antibodies to phosphorylated (active) JNK and p38, respectively (see "Experimental Procedures" for details). Arrow indicates transfer to complete medium. C, Erk1/2 activity in H9c2 cells after 2 h of ATP depletion and recovery for 15, 30, or 60 min. Erk1/2 activity was measured by immunoblotting with antibody to phosphorylated (active) Erk1/2. D and E, extent of JNK (D) and p38 kinase (E) activations in H9c2 cells depends on duration of ATP depletion. Time of recovery indicates the time after transfer to complete medium. JNK and p38 activities were measured as in A.
Hsp72 Protects from Ischemia/Reperfusion by Suppression of JNK death (Fig. 4D), indicating that Erk1/2 activation is involved in protection of H9c2 cells from caspase-independent death under transient energy deprivation.
Hsp72 Expression Decreases Stress Kinase Activation and Cell Death after ATP Depletion-Upon establishing a role of JNK in cell death following transient ATP depletion, we investigated whether the protective function of Hsp72 is related to JNK suppression as seen with heat-induced apoptosis (20,21). To express Hsp72, H9c2 cells were infected with an adenovirus encoding human Hsp72 under the control of the tetracyclineregulated promoter (tTA). In this expression system Hsp72 and GFP are encoded by a dicistronic transcription unit so accumulation of GFP serves as a marker for Hsp72 synthesis. Almost 100% of the infected cells cultured without tetracycline for 24 h became brightly fluorescent, indicating that almost all the cells expressed Hsp72. In contrast, in control cultures infected with virus and incubated in the presence of tetracycline, no GFP fluorescence was observed. Expression of Hsp72 in the absence, but not in the presence, of tetracycline was confirmed by immunoblotting of cell lysates with anti-Hsp72 antibody (Fig. 5A).
We found that expression of Hsp72 significantly inhibited JNK activation after ATP depletion/repletion (Fig. 5B); accordingly, under these conditions cell death was decreased more than 2-fold (Fig. 5D). To increase Hsp72 levels in cells by a more physiologically relevant way, cells were pretreated with mild heat shock (43°C for 30 min) followed by recovery for 16 h (Fig. 5A). When these cells were then exposed to ATP depletion/ repletion, we observed strong suppression of JNK activation (Fig. 5B) and an increase in survival (Fig. 5D). Therefore, an increase in Hsp72 levels in myogenic cells can suppress JNK activation and cell death following ischemic stress. Activation of another stress kinase, p38, was also inhibited by Hsp72 expressed either by adenovirus or mild heat shock pretreatment (Fig. 5C). To test whether cell protective activity of Hsp72 may be related to enhancement of Erk1/2, we measured activity of this kinase in Hsp72-expressing cells after ATP deletion/ repletion. Neither expression of recombinant Hsp72 nor pretreatment with mild heat shock enhanced Erk1/2 activity (not shown), indicating that JNK represents the major target of Hsp72 in the suppression of caspase-independent cell death upon transient energy deprivation.
Hsp72 Expression Alleviates Inhibition of JNK Dephosphorylation Caused by ATP Depletion/Repletion-TNF, interleukin-1, or UV irradiation activates JNK via protein kinase cascade MEKK-SEK (see Ref. 43 for example). We have previously found that, in contrast to UV irradiation or TNF, heat shock and certain other stresses do not activate SEK1 over its basal level; instead, they activate JNK by inhibiting JNK dephosphorylation (44). Here we investigated which of these pathways of JNK regulation is activated by the ATP depletion/repletion.
The level of phosphorylated (active) JNK is determined by the rate of JNK phosphorylation by upstream kinases (e.g. SEK1) and the rate of JNK dephosphorylation by phosphatase(s). Therefore, measurement of decline of the phosphorylated form of JNK after inhibition of upstream kinase cascade reflects the rate of JNK dephosphorylation. Previously, to inhibit upstream kinases we successfully utilized staurosporine that in JNK-activating cascade inhibits only SEK1 (44). H9c2 cells were exposed to ATP depletion/repletion, and after addition of staurosporine samples were taken, and the levels of phosphorylated JNK were measured by Western blot with anti-phospho-JNK antibody. ATP depletion/repletion inhibited JNK dephosphorylation more than 3-fold compared with that in unstressed cells (Fig. 6, A and B). Activation of SEK1 was also measured by Western blot with anti-phospho-SEK1 antibody that recognizes the active form of SEK1 specifically. ATP de- pletion/repletion strongly activated SEK1 (Fig. 6C). Therefore, ATP depletion/repletion stimulates JNK via two pathways, activation of an upstream kinase cascade and inhibition of JNK dephosphorylation.
Previously, we reported that Hsp72 expression attenuates JNK phosphatase inhibition following heat shock, thus suppressing JNK activation (44). Therefore, we studied how Hsp72 expression affects JNK dephosphorylation after ischemic stress. Expression of Hsp72 by adenovirus vector significantly (about 2-fold) accelerated JNK dephosphorylation following ATP depletion/repletion (Fig. 6, A and B). By contrast, expression of Hsp72 did not affect phosphorylation and activation of SEK1 (Fig. 6C). Therefore, JNK phosphatase(s) appear to be the target of Hsp72 action in ATP depletion/repletion. DISCUSSION It is now well established that ischemia/reperfusion of the myocardium can induce both necrosis and apoptosis (see Ref. 3 for review). However, the signal transduction pathways leading to ischemia-induced cell death are not yet clear. During the last few years it was demonstrated that initiation of apoptosis after many stresses (e.g. heat shock, oxidative stress, ceramide, radiation, etc.) is mediated by the stress kinase JNK (23,24). Here, employing ATP depletion/repletion as an in vitro model of ischemia/reperfusion, we demonstrated that JNK also plays a critical role in ischemia-induced death of H9c2 myogenic cells, which manifests features of both apoptosis and necrosis (Fig. 2). This conclusion is based on our finding that reduction of JNK activation after energy deprivation by expression of dominant-negative SEK or JNK1 increased cell survival (Fig.  4). Accordingly, the extent of ischemia-induced death correlated with the level of JNK activation (Fig. 3).
It has been observed previously that inhibition of p38 by SB 203580 in isolated cardiomyocytes and the myocardium sup- Hsp72 expression on JNK dephosphorylation. H9c2 cells were subjected to ATP depletion for 2 h and then transferred to complete medium for 15 min to activate JNK (indicated as ϪATP). After this treatment staurosporine was added to the cells to block JNK activity, and the rate of JNK dephosphorylation was assayed at different time points. Levels of phosphorylated JNK1 (A) and JNK2 (B) were measured as in Fig. 4 with antibodies to phosphorylated JNK (active JNK). To investigate the effect of an Hsp72 on JNK dephosphorylation, cells were infected with an Hsp72-expressing adenovirus as in Fig. 4 and then exposed to ATP-depleting conditions and treated as described above. C, effect of transient energy deprivation and Hsp72 expression on the upstream kinase cascade. Control or Hsp72-expressing H9c2 cells (infected with adenovirus as above) were exposed to ATP depletion/repletion as above, and the levels of SEK1 phosphorylation were measured by immunoblotting with an anti-phospho-SEK1 antibody. Lane C, control. pressed apoptosis and necrosis under ischemic conditions (31,32). However, we did not observe a protective effect of SB 203580 in H9c2 (see "Results") exposed to transient energy deprivation. That is in line with another study where inhibition of p38 in H9c2 cells failed to suppress apoptosis after oxidative stress (an important factor in reperfusion injury), whereas JNK suppression had a protective effect (42). Besides cell specificity, the conditions used to cause stresses may be very important. Myocardial ischemia/reperfusion is a more complicated phenomenon than transient energy deprivation in vitro, which we used in this study as a simplified model. Besides ATP loss, some other factors such as hypoxia, oxidative stress, ionic imbalance, etc. may be involved in the apoptotic process in reperfused myocardium. Therefore, it is likely that in myocardial cell death under genuine ischemia/reperfusion treatment, p38 may be involved in addition to JNK. Indeed, activation of p38 alone by its upstream regulators in the absence of any stress was insufficient to induce cardiomyocyte death, whereas simultaneous activation of JNK and p38 led to cytotoxicity (29). In any case, our finding that the activity of p38 is downregulated by Hsp72 similar to JNK activity may be relevant to Hsp72-mediated cardioprotection. In contrast to JNK and p38, Erk1/2 activation following ATP depletion/repletion of H9c2 cells apparently plays a protective role (Fig. 4D), similar to ischemic myocardium (33). Therefore, energy deprivation appears to activate both cell death and protective programs, and interplay between these programs determines the cell fate.
It has been suggested that ATP depletion as well as ischemia can cause protein damage. For example, a reduction of ATP levels was shown to induce marked aggregation of intracellular proteins such as actin, myosin, vinculin, and some others (see Refs. 45 and 46 and Ref. 7 for a review). Furthermore, similar to other protein-damaging stresses, ischemia reperfusion in vivo as well as transient ATP depletion in vitro are powerful activators of heat shock gene transcription through stimulation of HSF (7). Therefore, ischemia reperfusion appears to be a protein-damaging insult. Our results indicate that ATP depletion, similar to other protein-damaging insults such as heat shock, ethanol, and oxidative stress (44), activates JNK by inhibiting JNK phosphatase(s) activity. Unlike these stresses, however, energy deprivation also stimulates the upstream kinase SEK1, acting similar to cytokines, UV, and osmotic stress. Thus, both inhibition of JNK phosphatase and activation of upstream kinases (i.e. SEK1) are involved in the elevation of JNK activity under transient energy deprivation.
Clarification of pathways of JNK activation by transient energy deprivation allows us to investigate the mechanism of Hsp72-mediated suppression of JNK. We found that Hsp72 can alleviate the inhibition of JNK phosphatase caused by transient energy deprivation but does not affect activation of SEK1. Therefore, JNK phosphatase appears to be an important target of Hsp72 action in ischemic myocardium. Based on these data, we suggest that Hsp72 can reduce myocardial cell death after ischemia/reperfusion via reduction of JNK (and, possibly, p38) activity. This novel function of Hsp72 may be critical for its well known role in myocardial protection.