Insufficiency of Pro-heparin-binding Epidermal Growth Factor-like Growth Factor Shedding Enhances Hypoxic Cell Death in H9c2 Cardiomyoblasts via the Activation of Caspase-3 and c-Jun N-terminal Kinase*

Heparin-binding epidermal growth factor-like growth factor (HB-EGF) is a cardiogenic and cardiohypertrophic growth factor. ProHB-EGF, a product of the Hb-egf gene and the precursor of HB-EGF, is anchored to the plasma membrane. Its ectodomain region is shed by a disintegrin and metalloproteases (ADAMs) when activated by various stimulations. It has been reported that an uncleavable mutant of Hb-egf, uc-Hb-egf, produces uc-proHB-EGF, which is not cleaved by ADAMs and causes dilation of the heart in knock-in mice. This suggests that the shedding of proHB-EGF is essential for the development and survival of cardiomyocytes: however, the molecular mechanism involved has remained unclear. In this study, we investigated the relationship between uc-proHB-EGF expression and cardiomyocyte survival. Human uc-proHB-EGF was adenovirally introduced into the rat cardiomyoblast cell line H9c2, and the cells were cultured under normoxic and hypoxic conditions. Uc-proHB-EGF-expressing H9c2 cells underwent apoptosis under normoxic conditions, which distinctly increased under hypoxic conditions. Furthermore, we observed an increased Caspase-3 activity, reactive oxygen species accumulation, and an increased c-Jun N-terminal kinase (JNK) activity in the uc-proHB-EGF-expressing H9c2 cells. Treatment of the uc-proHB-EGF transfectants with inhibitors of Caspase-3, reactive oxygen species, and JNK, namely, Z-VAD-fmk, N-acetylcysteine, and SP600125, respectively, significantly reduced hypoxic cell death. These data indicate that insufficiency of proHB-EGF shedding under hypoxic stress leads to cardiomyocyte apoptosis via Caspase-3- and JNK-dependent pathways.

One of the cardiogenic and cardiohypertrophic growth factors is the heparin-binding epidermal growth factor-like growth factor (HB-EGF), 2 which is a member of the EGF family (1,2). An up-regulated Hb-egf expression is observed under conditions of tissue damage, hypoxia, and oxidative stress. The preventive roles of HB-EGF against hypoxia-induced apoptosis have been focused upon with regard to various organs, including the intestine (3)(4)(5), kidney (6,7), skeletal muscle (8), brain (9,10), placenta (11), and heart (12). Furthermore, it has also been demonstrated that the expression levels of Hb-egf and EGF receptor (EGFRs) mRNAs significantly increased in tissues surrounding myocardial infarction lesions (13,14).
HB-EGF is synthesized as a type I transmembrane protein (proHB-EGF) (1). Following ectodomain shedding, which is the cleavage of the juxtamembrane domain from the membraneanchored protein, by a disintegrin and metalloprotease (ADAM) and matrix metalloprotease, a soluble form of Hb-egf product (HB-EGF) is released into the extracellular space and binds to the EGFRs (15,16). The EGFRs bound to HB-EGF become phosphorylated and activate downstream signaling of the receptors such as mitogen-activated protein kinase and phosphatidylinositol 3-kinase. On the other hand, the C-terminal fragment of proHB-EGF (HB-EGF-CTF) translocates to the inner nuclear membrane and regulates the transcriptional activity by releasing transcriptional repressors such as promyelocytic leukemia zinc finger protein and B-cell lymphoma/leukemia-6 protein (Bcl-6) from DNA (17)(18)(19).
ProHB-EGF shedding itself has been known to play an important role in proliferation, differentiation, and survival of various types of cells. It has been reported that mice expressing an uncleavable proHB-EGF (uc-proHB-EGF) mutant (HB uc/uc ) develop severe cardiac dilation, which is not observed in proHB-EGF null mice (20,21). Histological examination of the hearts of the HB uc/uc mice revealed loss of cardiomyocytes, which is similar to that found in ischemic or idiopathic dilated cardiomyopathy. It has also been demonstrated that proHB-EGF shedding is required for human trophoblast survival under hypoxic conditions (11). However, the molecular mechanism underlying tissue disturbance that depends on impaired proHB-EGF shedding remains unclear.
In this study, we investigated the role of proHB-EGF shedding in the survival of a rat cardiomyoblast cell line, H9c2. Overexpression of uc-proHB-EGF or abrogation of proHB-EGF shedding by a metalloprotease inhibitor enhanced cell death both under normoxic and hypoxic conditions. However, the effect was considerably more under hypoxic conditions. Furthermore, we found that uc-proHB-EGF caused the activation of Caspase-3 and c-Jun N-terminal kinase (JNK) as well as the accumulation of reactive oxygen species (ROS). Our findings would be helpful in understanding the molecular mechanisms involved in the survival of cardiomyoblasts under hypoxic conditions and determination of the local HB-EGF concentration in coronary arteries may be useful in developing a novel prognostic marker in ischemic heart disease.
Adenovirus Construction and Infection-Adenovirus vectors carrying genes encoding green fluorescent protein (GFP), LacZ, wild-type proHB-EGF (wt-proHB-EGF), and uc-proHB-EGF were constructed using an adenovirus expression kit (Takara Bio Inc.). The cells were infected with purified, concentrated, and titer-checked viruses at a multiplicity of infection of 40 for 6 h before subjection to hypoxic conditions. Simulated Ischemia Model-Hypoxic conditions were created using a chamber with Anaero-Pack (Mitsubishi Gas Chemical), which catalytically reduced O 2 concentration to 1% within 60 min at 37°C. After the adenoviral infection, the cells were cultured in the hypoxic chamber for the indicated time periods.
Trypan Blue Dye Exclusion Assay-Monolayer cells were rinsed with phosphate-buffered saline and treated with trypsin and EDTA. They were then immediately stained with 0.5% typan blue. Cell viability was determined by counting the unstained cells under a microscope.
Terminal Deoxynucleotidyl Transferase-mediated dUTP Nick End-labeling (TUNEL) Assay-TUNEL staining was performed using a commercially available kit (ApopTag fluorescein in situ apoptosis detection kit; Chemicon) according to the manufacturer's instructions. The extent of binding of a fluorescent antibody to digoxigenin in the positive cells was determined by observing under a fluorescent microscope.
Measurement of Caspase-3 Activity-The activity of Caspase-3 was measured by the cleavage of a Caspase-3 substrate, FAM-DEVD-fmk, using a commercially available kit (Apo LOGIX Carboxyfluorescein Caspase detection kit; Cell Technology Inc.).
Fluorescein Isothiocyanate (FITC)-labeled Annexin V (Annexin V-FITC) Assay-Apoptosis was detected using a commercially available kit (Annexin V-FITC kit; Beckman Coulter) in accordance with the protocol provided by the manufacturer. Detection of Annexin V positive cells was performed by an Olympus fluorescent microscopy or FACSCalibur (BD Bioscience).
Live Cell Imaging-The cells were cultured in glass-bottomed dishes and observed using BioStation IM (Nikon). The phase-contrast images were acquired every 20 min for 60 h.
Evaluation of ROS Accumulation-Oxidative stress was evaluated by measuring the intracellular generation of H 2 O 2 . The cells were seeded in 96-well plates. After subjection to hypoxic conditions for 24 h, the cells were incubated in the dark for 30 min with 20 M DCFDA. Dichlorofluorescein (DCF) fluorescence was measured using a multiplate fluorometer (Wallac ARVO SX; PerkinElmer Life Sciences) at excitation and emission wavelengths of 488 and 530 nm, respectively. The DCF images were also obtained by a fluorescence microscopy (Olympus).
RT 2 Profiler PCR Array System-The expression of 84 apoptosis-related genes in a rat were examined using the RT 2 Profiler PCR array (SuperArray Bioscience). Total RNA was isolated from the GFP-or uc-proHB-EGF-expressed H9c2 cells (GFP/H9c2 or uc-proHB-EGF/H9c2) by using the TRIzol reagent (Invitrogen) and RNeasy mini kit (Qiagen). cDNA was synthesized from 1 g of RNA using a SuperScript RT II enzyme (Invitrogen). PCR was performed with the RT 2 Profiler PCR array system according to the manufacturer's instructions using ABI 7300 (Applied Biosystems). The expression levels of the mRNA of each gene in uc-proHB-EGF/H9c2 were normal-ized using the expression of Rplp1, Hprt1, Rpl13a, Ldha, and Actb, considered the housekeeping genes and then compared with the data obtained from GFP/H9c2. The results were confirmed by quantitative reverse transcriptase-PCR performed using individual RNA samples from the cells in each group by LightCycler (Roche Applied Science). The primers used for real-time PCR are listed in supplemental Table S1.
Gene Knockdown Assay-Small interference RNA (siRNA) against Hb-egf, JNK, and Caspase-3 were purchased from Applied Biosystems Inc. The siRNA was transfected by Code-Breaker TM siRNA transfection reagent (Promega) at 40 nM following the instruction manual, and the transfectants were infected by the adenoviruses at 24 h after transfection. Six hours later, the cells were incubated under hypoxia for 36 h. The knockdown efficiency of each was validated by a TaqMan gene expression assay (Applied Biosystems) or Western blotting. The siRNA used for a knockdown assay are listed in supplemental Table S1.
Statistical Analysis-All assays were independently performed 3 times. The results are represented as mean Ϯ S.E. The 2 groups were compared using the Student's t test. Analysis of variance with Bonferroni post hoc test was used for multiple comparisons. p Ͻ 0.05 was considered statistically significant. The Uc-proHB-EGF/H9c2 Apoptosis Was Enhanced under Hypoxic Conditions-Hypoxia is a pathophysiological condition frequently occurring in cardiomyocytes. To investigate the effect of uc-proHB-EGF on cell death under hypoxic stress, we cultured uc-proHB-EGF/H9c2 under hypoxic conditions. Apoptosis developed more significantly and promptly in uc-proHB-EGF/H9c2 than in the cells of other groups (Fig. 1, A,  C, and D). The TUNEL and Annexin V-FITC assays also revealed a significant enhancement of apoptosis in uc-proHB-EGF/H9c2 at 36 h after hypoxia (Fig. 1, E and F). Furthermore, the uc-proHB-EGF-enhanced cell death was not recovered by co-expression of wt-proHB-EGF (Fig. 1G). Next, we performed the assay with H9c2 cells treated with siRNA against Hb-egf. The expression of Hb-egf in knockdown cells decreased to apparently 30% in the mRNA level (box in Fig. 1H). However, knockdown of Hb-egf had no effect on hypoxia-induced apoptosis (Fig. 1H). These data indicate that proHB-EGF shedding but not its expression is associated with survival of cardiomyocytes under hypoxia.

Overexpression of Uc-proHB-EGF in H9c2 Cells Increased Cell Death under Normoxic Conditions-To
The Released HB-EGF Did Not Regulate H9c2 Cell Viability under Hypoxia-To confirm that sufficient proHB-EGF shedding is important to maintain cell viability under hypoxic conditions, we used an ADAM inhibitor, KB-R7785, in cell culture. KB-R7785 has been established as an effective inhibitor of proHB-EGF shedding in vivo and in vitro (23). To examine whether KB-R7785 also works as an inhibitor of proHB-EGF shedding in H9c2 cells, we performed an alkaline phosphatase-proHB-EGF shedding assay with a stable H9c2 transfectant with alkaline phosphatase-tagged proHB-EGF. The result showed that shedding of proHB-EGF by phorbol 12-myristate 13-acetate, a strong inducer of proHB-EGF shedding (24), was completely blocked by KB-R7785 (supplemental Fig. S2). On the other hand, KB-R10082, an inactive analog of KB-R7785, had no effect on the shedding. We sequentially investigated effects of these compounds on apoptosis of H9c2 cells under hypoxia. In the presence of KB-R7785, the viabilities of mock/ H9c2, GFP/H9c2, and wt-proHB-EGF/H9c2 significantly decreased under hypoxic conditions. However, the viability of uc-proHB-EGF/H9c2 remained unaffected (Fig. 2, A and B). KB-R10082 had no apparent effect on the apoptosis induction under hypoxic conditions. ProHB-EGF shedding results in paracrine and/or autocrine manners of HB-EGF. To investigate whether the enhancement of apoptosis is due to the reduced production of HB-EGF in uc-proHB-EGF/H9c2, recombinant human HB-EGF was added to the cell culture. Recombinant HB-EGF could not adequately recover the viability of uc-proHB-EGF/H9c2, although tyrosine phosphorylation of EGFR by recombinant HB-EGF was observed in uc-proHB-EGF/H9c2, as well as in GFP/H9c2 or wt-proHB-EGF/H9c2 (Fig. 2, C-E).
Caspase-3-dependent Pathway Played a Part in Cell Death by Uc-proHB-EGF Expression-The activity of a unique class of cysteine proteases called Caspases is involved in effecting cell death. To investigate the involvement of Caspases-dependent pathways in the uc-proHB-EGF/H9c2 apoptosis, we examined the activity of Caspase-3, the final executor of apoptosis, in uc-proHB-EGF/H9c2 exposed to hypoxic conditions. Caspase-3 activity increased more significantly in uc-proHB-EGF/H9c2 than in the cells of other groups (Fig. 3, A and B). Moreover, we observed the inhibitory effect of a pan-Caspase inhibitor, Z-VAD-fmk, on uc-proHB-EGF/H9c2 apoptosis. Although Z-VAD-fmk strongly inhibited Caspase-3 activity, it failed to completely recover the cells from apoptosis (Fig. 3,  A-D). These results suggest that not only Caspase-dependent but also Caspase-independent pathways are involved in this effecting cell death.

ROS Accumulated in Uc-proHB-EGF/H9c2 under Hypoxic Conditions, and a ROS Scavenger Reduced the Enhanced Cell
Death-Hypoxia increases mitochondrial ROS generation at Complexes I and III (25)(26)(27)(28). Moreover, ROS has been known as a proapoptotic factor that activates stress-activated protein kinases. We evaluated the accumulation of ROS in each of the gene transfectants under hypoxic conditions. The results showed that the level of ROS accumulation was higher in uc-proHB-EGF/H9c2 than in the cells of the other groups ( Fig.   Uc-proHB- A and B). Furthermore, we investigated the effect of a ROS scavenger on uc-proHB-EGF/H9c2 apoptosis under hypoxic conditions. Pretreatment with antioxidant NAC significantly reduced the uc-proHB-EGF/H9c2 apoptosis under hypoxic conditions (Fig. 4, C and D). These data suggest that apoptosis in uc-proHB-EGF/H9c2 under hypoxia are also due to the excessive accumulation of ROS.

Activation of Stress-activated Kinase JNK Signaling Was Involved in Uc-proHB-EGF/H9c2 Apoptosis under Hypoxic
Conditions-During apoptosis, ROS stimulates stress-activated protein kinases such as JNK and p38 via apoptosis signal-regulating kinase 1 (ASK1) (29). To investigate whether the activities of stress signaling pathways are up-regulated in uc-proHB-EGF/H9c2, an immunoblot analysis was performed with an anti-phosphorylated JNK or an anti-phosphorylated p38 antibody. Until 12 h of subjection to hypoxic conditions, no difference was observed in the extent of JNK and p38 phosphorylation among the cells of all groups. However, at 18 h, JNK  activation markedly increased in uc-proHB-EGF/H9c2 (Fig.  5A). Moreover, the inhibition of JNK activity by SP600125, a JNK inhibitor, abrogated uc-proHB-EGF/H9c2 apoptosis under hypoxic conditions (Fig. 5, B-D). Activation of Akt and ERK1/2 has also been known to participate in cell survival. We investigated any changes of their activation in transfectants under hypoxic conditions. However, no remarkable difference was observed in the activities of both Akt and ERK1/2 between the wt-proHB-EGF and uc-proHB/H9c2 (Fig. 5A). Furthermore, we investigated whether combinational treatments of a pan-Caspase inhibitor, a ROS scavenger, and a JNK inhibitor considerably affect the enhanced uc-proHB-EGF/H9c2 apoptosis under hypoxic conditions. A treatment of Z-VAD-fmk and SP600125 or NAC strongly inhibited uc-proHB-EGF/H9c2 cell apoptosis as compared with each single treatment (Fig. 5E). However, the effect of a treatment of these three inhibitors together on apoptosis was similar to that with Z-VAD-fmk and SP600125. On the other hand, these inhibitors did not affect apoptosis of wt-proHB-EGF/H9c2 (data not shown). To confirm that Caspase-3 and JNK are responsible for the apoptosis, a gene knockdown assay was also performed. The protein levels of Caspase-3 or JNK were significantly reduced in each siRNAtreated cells (supplemental Fig. S3A). We found that knockdown of Caspase-3 or JNK significantly rescued apoptosis of uc-proHB-EGF/H9c2 (Fig. 5F). These data show that both Caspase-and JNK-dependent pathways, including stimulation by ROS accumulation, mainly work in the apoptosis of uc-proHB-EGF/H9c2.

Expression of XIAP and Bcl-XL Decreased and That of Bax Increased in Uc-proHB-EGF/H9c2 under Hypoxic Conditions-
It has been recently shown that HB-EGF-CTF, a product by proHB-EGF shedding, regulates gene transcription via the release of transcriptional repressors from DNA (18,19). In this study, we examined whether uc-proHB-EGF affects gene expression under hypoxic conditions. By screening apoptosisrelated gene expression using RT 2 Profiler PCR array (Super-Array Bioscience), we found that XIAP and Bcl-XL mRNA decreased, and the expression of Bax increased significantly in uc-proHB-EGF/H9c2 after 12 h of subjection to hypoxic conditions (data not shown). Moreover, when quantitative reverse transcriptase-PCR was independently performed using RNA in each group, the expression of XIAP and Bcl-XL significantly decreased to 20 and 50%, respectively, in uc-proHB-EGF/H9c2 as compared with other groups (Fig.  6A, the left and middle panels). However, the protein level of XIAP, but not Bcl-XL was clearly reduced in uc-proHB-EGF/ H9c2 at 24 h after subjection to hypoxic conditions (Fig. 6B). In contrast, uc-proHB-EGF/H9c2 provided a 2-fold increase in expression of the Bax gene as compared with other groups (Fig. 6A, right panel). The protein level of Bax in uc-proHB- EGF/H9c2 also increased from 12 h under hypoxia (Fig. 6B). These data suggest that the expression pattern of antiapoptotic and proapoptotic genes is altered such that they affect survival in uc-proHB-EGF/H9c2.

DISCUSSION
In this report, we found that overexpression of uc-proHB-EGF caused an enhancement of H9c2 cell apoptosis under hypoxic conditions. Furthermore, metalloproteases inhibitors also enhanced hypoxic cell death of H9c2 cells (Fig. 2, A and B). On the other hand, knockdown of Hb-egf did not enhance the cell death of H9c2 cells under hypoxia (Fig. 1H). Previous reports also showed that the cardiomyocytes in Hb-egf-null mice are not led to cell death, unlike those in uc-proHB-EGF knock-in mice (20,21). In human trophoblasts, proHB-EGF shedding has been known to be required for survival under hypoxic conditions (11). Therefore, the events of proHB-EGF shedding but not the presence of proHB-EGF are essential for cardiomyocyte survival under hypoxia.
We have previously shown that uc-proHB-EGF, which functions as a dominant negative gene product, is involved in the proliferation of mouse embryonic fibroblasts (30). In this study, we also found co-expression of wt-proHB-EGF did not rescue the cell death of uc-proHB-EGF/H9c2 (Fig. 1G). Uc-proHB-EGF mutant is speculated to be useful for analyzing the various effects of a dominant negative modulator on proHB-EGF shedding.
Both HB-EGF and HB-EGF-CTF are released from proHB-EGF by ectodomain shedding. In our study, HB-EGF could not restore the cells from uc-proHB-EGF-enhanced apoptosis (Fig.  2, C and D). These data indicate that the production of HB-EGF is not involved in the uc-proHB-EGF/H9c2 apoptosis under hypoxic conditions. On the other hand, some novel functions of HB-EGF-CTF have recently been highlighted. After proHB-EGF shedding, HB-EGF-CTF is generated and retrogradely translocated to the inner nuclear membrane (17). HB-EGF-CTF has also been known to regulate the transcriptional level of cyclin A and cyclin D2 by binding to and releasing promyelocytic leukemia zinc finger or Bcl-6 transcriptional repressor from each promoter (18,19). We have also reported that the induction of c-Myc by growth factors is reduced in uc-proHB-EGF-expressing mouse embryonic fibroblasts (30). We thought that the enhancement of the apoptosis may also be due to the events occurring via inadequate transcriptional regulation by insufficient production of HB-EGF-CTF under hypoxia (supplemental Fig. S3B). In this study, we performed a gene expression analysis with uc-proHB-EGF/H9c2 under hypoxic conditions, and the expression of XIAP was observed to be downregulated (Fig. 6A). The level of XIAP protein also decreased in uc-proHB-EGF/H9c2 (Fig. 6B). XIAP is a direct inhibitor of Caspases (31). It has recently been reported that XIAP binds to transforming growth factor-␤-activated kinase 1 (TAK1) and inhibits the kinase activity of TAK1 toward JNK (32,33). In this report, the activation of both Caspase-3 and JNK significantly increased in uc-proHB-EGF/H9c2 (Figs. 3, A and B,  and 5A). The down-regulation of XIAP could cause an increase in Caspase-3 and JNK activities during hypoxic death of uc-HB-EGF/H9c2. We also detected a down-regulation of Bcl-XL and an up-regulation of Bax in uc-HB-EGF/ H9c2 under hypoxic conditions (Fig. 6A). Bcl-XL and Bax are members of the Bcl-2 family. Bax increases the permeability of the mitochondrial outer membrane and facilitates the release of cytochrome c. The released cytochrome c activates Caspase-9, an activator of Caspase-3. Bcl-XL binds to Bax to inhibit apoptosis (34). The expression pattern of the Bcl-2 family members observed in uc-proHB-EGF/H9c2 indicates that the cells might be in a proapoptotic state due to increased mitochondrial membrane permeability. We advocate that impairment shedding of proHB-EGF in hypoxia reduces a threshold level of Caspase-3 and JNK activation via gene regulation of XIAP and the Bcl-2 family members, and leads to cell death (Fig. 7). The cytoplasmic domain of proHB-EGF has also been reported to bind Bcl-2-associated athanogene (BAG-1), and the interaction of BAG-1 with a proHB-EGF molecule leads to decreased cell adhesion and increased cell viability with etoposide treatment (35). BAG-1 functions as an antiapoptotic factor when associated with Bcl-2 (36,37). BAG-1 also associates with heat shock protein-70, which is a heat shock transcription factor-1-induced cell protector (38). However, it still remains unknown whether HB-EGF-CTF forms a complex with BAG-1 and regulates the antiapoptotic functions. Further examinations are required to elucidate the relationship between the involvement of uc-proHB-EGF and BAG-1 in hypoxic cell death. We unveiled that ROS accumulated in uc-proHB-EGF/H9c2 under hypoxic conditions (Fig. 4, A and B) and that a ROS scavenger rescued uc-proHB-EGF/H9c2 from the apoptosis under hypoxic conditions. BAG-1 has been also known to act as an antioxidant (37). The abovementioned data suggest that proHB-EGF shedding could also FIGURE 6. The expression of antiapoptotic and proapoptotic molecules was altered in uc-proHB-EGF/H9c2. A, the expression levels of XIAP, Bcl-XL, and Bax mRNAs in GFP/H9c2, wt proHB-EGF/H9c2, and uc-proHB-EGF/H9c2 cells were assessed by quantitative real-time PCR. Each sample was collected at 12 h after subjection of the cells to hypoxic conditions. B, the levels of XIAP, Bcl-XL, and Bax protein are represented at each indicated time after hypoxia by Western blot analysis. ␤-Actin serves as a loading control. FIGURE 7. Representation of a molecular mechanism underlying cardiomyocyte death due to insufficient proHB-EGF shedding under hypoxia.
control BAG-1 to regulate redox reactions by retaining a reducing intracellular environment.
As a conflicting data, Krieg et al. (39) reported that proHB-EGF shedding generated ROS following EGF receptor transactivation. They used chemical components such as acetylcholine to mimic an ischemic preconditioning state. We used rat-derived cells in this study, whereas Krieg et al. (39) studied with rabbit-derived cells. These differences could lead to the distinct results. However, our data are consistent with the in vivo data (20,21). It has been recently reported that oxidative stress increases in patients with idiopathic cardiomyopathy or idiopathic dilated cardiomyopathy (40). A uc-proHB-EGF knock-in mouse exhibited a similar condition in idiopathic cardiomyopathy or idiopathic dilated cardiomyopathy (21). Insufficient shedding of proHB-EGF could play a role in increasing oxidative stress in these disorders.
It has been established that the accumulation of ROS induces JNK activation via the activation of ASK1 (29,41) and ASK1 knock-out neonatal cardiomyocytes are resistant to H 2 O 2 -induced apoptosis (42). On the basis of the above reports, it may be assumed that ROS-ASK1 signaling stimulated JNK in our study. On the other hand, the combination of NAC and SP600125 was more effective in the survival of uc-proHB-EGF/ H9c2 cells than a treatment of NAC alone (Fig. 5E). This indicates that there are other stimulators of JNK activation besides ROS.
In this study, we demonstrated that uc-proHB-EGF enhanced induction of apoptosis under hypoxic conditions. Thus, we can assume that disruption of proHB-EGF shedding could lead to extensive cardiomyocyte apoptosis and poor prognosis in patients with ischemic heart disease, who are exposed to excessive hypoxic stress. This raises the possibilities that the amount of HB-EGF in the heart is a marker of prognosis in patients with ischemic heart disease and that improving insufficient shedding of proHB-EGF might develop as a new therapeutic strategy for cardiac disorder.