Execution of apoptosis signal-regulating kinase 1 (ASK1)-induced apoptosis by the mitochondria-dependent caspase activation.

ASK1 activates JNK and p38 mitogen-activated protein kinases and constitutes a pivotal signaling pathway in cytokine- and stress-induced apoptosis. However, little is known about the mechanism of how ASK1 executes apoptosis. Here we investigated the roles of caspases and mitochondria in ASK1-induced apoptosis. We found that benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone (zVAD-fmk), a broad-spectrum caspase inhibitor, mostly inhibited ASK1-induced cell death, suggesting that caspases are required for ASK1-induced apoptosis. Overexpression of ASK1DeltaN, a constitutively active mutant of ASK1, induced cytochrome c release from mitochondria and activation of caspase-9 and caspase-3 but not of caspase-8-like proteases. Consistently, caspase-8-deficient (Casp8 (-/-)) cells were sensitive to ASK1-induced caspase-3 activation and apoptosis, suggesting that caspase-8 is dispensable for ASK1-induced apoptosis, whereas ASK1 failed to activate caspase-3 in caspase-9-dificient (Casp9 (-/-)) cells. Moreover, mitochondrial cytochrome c release, which was not inhibited by zVAD-fmk, preceded the onset of caspase-3 activation and cell death induced by ASK1. ASK1 thus appears to execute apoptosis mainly by the mitochondria-dependent caspase activation.

Apoptosis is a highly regulated process essential for the development and homeostasis of multicellular organisms (1)(2)(3)(4). Genetic studies of apoptosis in the nematode Caenorhabditis elegans identify crucial regulators of apoptosis (5). ced-3 and ced-4 are required for the execution of apoptosis (6,7), whereas ced-9 protects cells from undergoing apoptosis (8). The structural and functional homolog of Ced-9 in mammals has been identified as Bcl-2, which belongs to a large family of molecules involved in either pro-apoptotic or anti-apoptotic signals (9,10). The mammalian counterpart of Ced-3 has been identified as caspase-3, a member of a large family of intracellular proteases that play central roles in apoptotic machinery (11). Ced-4 homolog has been identified and termed apoptosis protease-activating factor 1 (Apaf1), which is essential for the activation of caspase-9, another Ced-3 homolog (12,13).
Despite the diversity of pro-apoptotic signals, the activation of caspases is consistently implicated in apoptosis (14,15). Among the caspase family, the effector caspases represented by caspase-3 are frequently activated in response to various apoptosis inducers and essential for the execution and completion of apoptosis in many, but not all, cell types or death stimuli (16). Pathways leading to activation of the effector caspases have been identified to be either dependent or independent on the mitochondrial release of cytochrome c into the cytosol (17). The released cytochrome c, together with dATP, binds to Apaf-1, which in turn activates an initiator caspase, caspase-9 (13). Another initiator caspase, caspase-8, mainly mediates cytochrome c-independent death signal, which is initiated upon ligation of death receptors (18,19). These two pathways converge on the activation of the effector caspases followed by cleavage of a variety of apoptotic substrates.
Activation of the effector caspases appears to be tightly regulated by a divergent signaling mechanism including the members of Bcl-2 family; e.g., a pro-apoptotic Bcl-2 family member, Bid, is cleaved by caspase-8, and the truncated Bid induces the mitochondrial cytochrome c release (20,21). In addition to caspase and Bcl-2 families, a number of signaling molecules have been suggested to regulate apoptosis (22). Members of the mitogen-activated protein (MAP) 1 kinase family are involved in signal transduction of apoptosis as well as cell growth and differentiation (23,24). Two different MAP kinase cascades that converge on c-Jun N-terminal kinase (JNK; also known as SAPK, stress-activated protein kinase) and p38 MAP kinase are preferentially activated by cytotoxic stressors such as UV radiation, x-ray, heat shock and osmotic shock, and by proin-flammatory cytokines such as tumor necrosis factor and interleukin-1 (25)(26)(27). The possible roles of JNK pathway in apoptosis signaling have been demonstrated by knockout mouse studies. Mice lacking the JNK3 gene were reported to exhibit marked reduction in excitotoxicity-induced apoptosis of hippocampal neurons (28). JNK2 was shown to be required for apoptosis of immature thymocytes induced by anti-CD3 antibody but not for activation-induced cell death of mature T cells (29). Compound mutant mice lacking the JNK1 and JNK2 genes suggested that JNK1 and JNK2 regulate region-specific apoptosis during early brain development (30).
Apoptosis signal-regulating kinase 1 (ASK1) is a ubiquitously expressed MAP kinase kinase kinase (MAPKKK) that activates the SEK1-JNK and MKK3/MKK6-p38 signaling cascades (31). Overexpression of ASK1 in epithelial cells in a low serum condition-induced apoptosis, and in ovarian cancer cells, expression of a kinase-inactive mutant of ASK1 inhibited microtubule-interfering agent-induced apoptosis, suggesting that ASK1 plays a role in the mechanism of stress-induced apoptosis (31)(32)(33). Although moderate expression of a constitutively active form of ASK1 induced neuronal differentiation in naive PC12 cells (34), overexpression of ASK1 induced apoptosis in nerve growth factor-differentiated PC12 cells and primary rat sympathetic neurons (35). Moreover, dominant-negative ASK1 reduced the neuronal apoptosis induced by nerve growth factor withdrawal from these cells (35). ASK1 was activated upon treatment with tumor necrosis factor-␣ or agonistic anti-Fas antibody, and a kinase-inactive mutant of ASK1 reduced tumor necrosis factor-␣-and Fas-induced JNK activation and apoptosis, suggesting that ASK1 is a pivotal component in cytokineinduced apoptosis as well (31,36,37). However, it is still unknown how ASK1-induced apoptosis is executed, particularly in relation to caspase activities. Here we show that ASK1induced apoptosis is dependent on caspase activities and that the mitochondria-dependent apoptosis pathway is the main mechanism of execution of ASK1-induced apoptosis.
Antibodies and Reagents-A monoclonal antibody to the hemagglutinin (HA) tag (clone 3F10) was purchased from Roche Molecular Biochemicals. Monoclonal antibodies to caspase-3, cytochrome c, and ␣-tubulin were purchased from Transduction Laboratories, Pharmingen, and Harlan Sera-Lab, respectively. zVAD-fmk was purchased from Peptide Institute, Inc.
Infection of Recombinant Adenoviruses-Recombinant adenoviruses encoding HA-tagged ASK1 mutants and ␤-galactosidase were constructed as described (32). Cells were plated in assay plates and allowed to grow in DMEM supplemented with 10% fetal calf serum overnight. Mv1Lu cells and MEFs were washed twice with phosphate-buffered saline (PBS) and infected with recombinant adenoviruses in DMEM supplemented with 1% and 0.1% FBS, respectively. Nearly 100% infection of recombinant adenoviruses to Mv1Lu cells and MEFs can be achieved at a multiplicity of infection (m.o.i.) of 100 and 400, respectively, as determined by ␤-galactosidase staining (data not shown).
Determination of Cell Viability-Cell viability of Mv1Lu cells was determined by trypan blue exclusion assay. Briefly, cells were washed with PBS, trypsinized, and collected by centrifugation. The cell pellets were resuspended in 0.4% trypan blue in PBS, and the proportion of dead blue cells was determined. For MEFs, cell viability was determined using a cell counting kit-8 (Dojindo), in which 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium monosodium salt (WST-8) was used as a substrate. The relative number of surviving cells was determined in duplicate by estimating the value of uninfected cells as 100%.
Annexin V Staining-Cells were trypsinized, washed with PBS, and resuspended in a binding buffer containing 10 mM Hepes, pH 7.4, 140 mM NaCl, 2.5 mM CaCl 2 . After 15 min of incubation with annexin V-fluorescein isothiocyanate and propidium iodide (PI; Pharmingen) at room temperature, the fluorescence emitted by cells (10,000 cells/sample) was analyzed using a FACScan flow cytometer (Becton Dickinson).
DNA Fragmentation Analysis-Approximately 2 ϫ 10 6 cells were lysed with 200 l of a buffer containing 20 mM Tris-HCl, pH 7.5, 10 mM EDTA, and 0.5% Triton X-100. Cell extracts were clarified by centrifugation at 15,000 rpm for 5 min. The lysate was incubated with 0.2 mg/ml proteinase K at 42°C for 1 h. DNA was purified by standard phenol-chloroform extraction and ethanol precipitation. Dry DNA pellets were then resuspended in TE buffer (10 mM Tris-HCl, pH 7.5, and 1 mM EDTA) containing 0.2 mg/ml ribonuclease A. Samples were loaded on a 2% agarose gel and visualized by ethidium bromide staining.
Measurement of Caspase Activities-Approximately 4 ϫ 10 5 cells were used for measurement of caspase activities. Caspase-3-like activity was measured by a CPP32/caspase-3 fluorometric protease assay kit (MBL) following the manufacturer's instructions, in which a fluorogenic synthetic peptide DEVD-7-amino-4-trifluoromethyl coumarin (AFC) was used as a substrate. Caspase-8-and -9-like activities were measured using IETD-AFC and LEHD-AFC (Calbiochem), respectively. The fluorescence of the released AFC was measured with an excitation wavelength of 360 nm and an emission wavelength of 530 nm.
Detection of Cytochrome c Release-Approximately 1 ϫ 10 7 cells were trypsinized and collected by centrifugation. The cell pellets were washed with PBS and resuspended in 125 l of buffer containing 250 mM sucrose, 20 mM Hepes-KOH, pH 7.5, 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride. The cells were then homogenized with 30 strokes of a homogenizer (Wheaton). The mitochondria-enriched fraction was pelleted by centrifugation at 15,000 rpm for 15 min. The supernatant was subjected to Western blot analysis with a monoclonal antibody to cytochrome c.

Characterization of ASK1-induced Cell
Death-To investigate the mechanism regulating an execution process of ASK1induced apoptosis, we first characterized the death-inducing activity of ASK1 in Mv1Lu cells and primary MEF cells. Mv1Lu and MEFs were infected with recombinant adenoviruses encoding a hemagglutinin (HA) epitope-tagged constitutively active mutant of ASK1 (Ad-ASK1⌬N) or an HA-tagged catalytically inactive mutant of ASK1 (Ad-ASK1KM) (32). Expression of ASK1⌬N and ASK1KM was detectable as early as 8 h after infection by Western blot analysis (data not shown) and was clearly detected in a m.o.i.-dependent manner after 24 h in both cells (Fig. 1, lower panel, and data not shown). Forty-eight h after infection, ASK1-induced cell death became apparent in Mv1Lu as determined by a trypan blue exclusion assay (Fig. 1,  upper panel). Only ASK1⌬N but not ASK1KM induced a significant cell death, confirming that the death-inducing effect of ASK1 depends on its enzymatic activity. Although a longer time period was required (3 to 6 days culture after infection), adenovirus-mediated expression of ASK1⌬N but not ASK1KM strongly induced cell death in MEFs as well (data not shown).
To examine whether the ASK1-induced cell death is accompanied by apoptotic features, distribution of phosphatidyl-serine on the plasma membrane and fragmentation of chromosomal DNA were analyzed by annexin V staining and DNA laddering assay, respectively. Compared with uninfected or Ad-ASK1KM-infected cells, Ad-ASK1⌬N-infected cells displayed an increased annexin V-positive population in Mv1Lu ( Fig. 2A). Of note, annexin V-positive but PI-negative population was also increased in the ASK1⌬N-expressing cells, suggesting that ASK1-induced cell death follows an ordered apoptotic process. When the cytoplasmic small-fragmented DNA was extracted and analyzed by agarose gel electrophoresis, DNA fragmentation was observed in Ad-ASK1⌬N-infected MEFs but not in uninfected or Ad-ASK1KM-infected cells (Fig.  2B). These findings indicate that the expression of ASK1⌬N effectively induces apoptosis in Mv1Lu cells and MEFs and that these cells provide useful models to investigate the mechanism of ASK1-induced apoptosis.
Caspase Activities in ASK1-induced Apoptosis-Since members of the caspase family are crucial mediators of apoptosis (14 -16), we examined the requirement of caspase activities for ASK1-induced apoptosis using zVAD-fmk, a broad-spectrum caspase inhibitor. We confirmed that the addition of zVAD-fmk did not reduce the adenovirus-mediated expression of ASK1⌬N in Mv1Lu cells (Fig. 3, lower panel). When cell death was determined by trypan blue exclusion assay, ASK1-induced death was significantly inhibited in the presence of zVAD-fmk, suggesting that caspase activity is required for ASK1-induced apoptosis (Fig. 3, upper panel). Consistently, proteolytic processing of pro-caspase-3 was induced by the expression of ASK1⌬N but not ASK1KM in a m.o.i.-dependent manner, suggesting that caspase-3 is activated during the course of ASK1induced apoptosis (Fig. 4).
To investigate the involvement of initiator caspases in ASK1induced apoptosis, we next examined the activities of caspase-8-and -9-like proteases in comparison with caspase-3like activity induced by the expression of ASK1⌬N. Fluorogenic synthetic peptides, DEVD-AFC, IETD-AFC, and LEHD-AFC were utilized as substrates to assess the activities of caspase-3-, caspase-8-, and caspase-9-like proteases, respectively. Consistent with the ASK1⌬N-induced processing of pro-caspase-3 (Fig. 4), expression of ASK1⌬N but not ASK1KM strongly induced caspase-3-like activity 48 h after the infection in  Mv1Lu cells (Fig. 5A, left panel). At this time point, a substantial level of caspase-9-like activity was also induced by the expression of ASK1⌬N, whereas only a marginal activation of caspase-8-like proteases was detected (Fig. 5A, middle and right panels). To exclude a possibility that activation of caspase-8-like proteases occurred earlier, we examined the time course of activities of caspase-8-and caspase-3-like proteases (Fig. 5B). Activation of caspase-3-like proteases was first detected 28 h after the infection of ASK1⌬N in Mv1Lu cells. In contrast, no significant activation of caspase-8-like proteases was detected throughout the course of ASK1-induced apoptosis, suggesting that activation of caspase-8-like proteases is not involved in ASK1-induced apoptosis.
In contrast, ASK1-dependent activation of caspase-3-like proteases was completely absent in Casp9 Ϫ/Ϫ MEFs (Fig. 7A),   FIG. 4. Processing of pro-caspase-3 induced by the expression  of ASK1⌬N. MEFs were left uninfected or infected with the indicated m.o.i. of Ad-ASK1⌬N or Ad-ASK1KM. Four days after the infection, cells were lysed and immunoblotted with a monoclonal antibody to caspase-3. As a positive control for processing of pro-caspase-3, cells were exposed to 5 g/ml of anisomysin for 12 h (indicated as Aniso). suggesting that caspase-9 is prerequisite to ASK1-induced activation of caspase-3. On the other hand, Casp9 Ϫ/Ϫ MEFs were still sensitive to ASK1-induced cell death, although less sensitive than the wild type (Casp9 ϩ/ϩ ) MEFs (Fig. 7B). These results suggest that other caspases, which lack DEVDase activity, may be involved in the execution of apoptosis induced by ASK1 (see "Discussion").
Involvement of Cytochrome c Release in ASK1-induced Apoptosis-Since activation of caspase-9 requires the mitochondrial release of cytochrome c into the cytosol (13), we examined whether cytochrome c release occurs in ASK1-induced apoptosis. We isolated the cytosolic fractions from ASK1⌬N-or ASK1KM-expressing cells and detected the released cytochrome c from mitochondria by Western blot analysis. The expression of ASK1⌬N but not ASK1KM resulted in release of cytochrome c from mitochondria in Mv1Lu cells and MEFs ( Fig.  8A and data not shown), clearly demonstrating the involvement of cytochrome c release in ASK1-induced apoptosis. Importantly, the addition of zVAD-fmk did not inhibit the cytochrome c release from mitochondria, suggesting that the cytochrome c release is a trigger rather than a consequence of the caspase activation in ASK1-induced apoptosis. Consistent with this notion, ASK1⌬N-induced cytochrome c release was detected as early as 18 h after the infection of Ad-ASK1⌬N and increased thereafter (Fig. 8B). Since no apparent cell death was detected within 30 h in ASK1⌬N-expressing Mv1Lu cells (data not shown), cytochrome c release obviously preceded the onset of cell death and was likely to be one of the determinants of ASK1-induced apoptosis. Taken together, ASK1 appears to induce apoptosis mainly via the mitochondrial pathway followed by caspase-9 and -3 activation. DISCUSSION In the present study, we investigated the roles of mitochondria and caspases in the execution of ASK1-induced apoptosis. We found that caspase activities are required for ASK1-induced apoptosis at least in two different cell-types, epithelial cells (Mv1Lu) and fibroblasts (MEFs). In the presence of a broad-spectrum caspase inhibitor, zVAD-fmk, ASK1-induced apoptosis was almost suppressed (Fig. 3), suggesting that ASK1-induced apoptosis is largely dependent on caspase activities. Based on the requirement of caspase activities, we first presumed that ASK1 might kill cells through death receptorinduced direct activation of caspase-8, since activation of JNK has been reported to lead to de novo synthesis of pro-apoptotic proteins, e.g. Fas ligand (FasL), via c-Jun-dependent transcription (40,41). In fact, expression of an activated form of MEKK1, another MAP kinase kinase kinase (MAPKKK) in the JNK pathway, results in apoptosis of Jurkat T cells and PC12 cells in a FasL induction-dependent manner (40,41). However, ASK1 induced only a limited increase in the caspase-8-like activity (Fig. 5), and MEFs deficient in caspase-8 were still sensitive to ASK1-induced caspase-3 activation and apoptosis (Fig. 6). These results clearly demonstrate that caspase-8 is not a minimum requirement for ASK1-induced apoptosis at least in these cells.
In contrast, another initiator caspase, caspase-9, appears to be crucial for caspase-3 activation in ASK1 signaling. Caspase-9 was activated by the expression of ASK1⌬N (Fig. 5), and more convincingly, ASK1-induced caspase-3 activation was completely absent in Casp9 Ϫ/Ϫ MEFs (Fig. 7A). These results strongly suggest that caspase-9 is indispensable for ASK1-induced activation of caspase-3 and exclude a possibility that caspase-3 is directly activated downstream of the ASK1-JNK and/or -p38 axes. Unexpectedly, however, Casp9 Ϫ/Ϫ MEFs were not completely resistant to ASK1-induced apoptosis (Fig. 7B). An explanation for the discrepancy between the complete absence of ASK1-induced caspase-3-like activity and the limited resistance to apoptosis induced by ASK1 in Casp9 Ϫ/Ϫ MEFs is that other caspases, which lack DEVDase activity, may participate in the execution of ASK1-induced apoptosis or may be activated in compensation for caspase-9 deficiency (42). A candidate for such among effector caspases is caspase-6, since the optimal tetrapeptide sequence for caspase-6 is VEHD and considerably differs from DEVD, a preferential sequence for caspase-3 and -7 (43). Alternatively, incomplete inhibition of ASK1-induced apoptosis by zVAD-fmk (Fig. 3) may suggest the possible involvement of the caspase-independent mechanism in ASK1-induced apoptosis, including a recently identified caspase-independent effector of apoptosis, AIF (apoptosis inducing factor) (44). On the other hand, Casp9 Ϫ/Ϫ MEFs were in fact less sensitive to ASK1-induced apoptosis than Casp9 ϩ/ϩ MEFs (Fig. 7B). The observed difference in the sensitivity to ASK1-induced apoptosis is likely to be attributable to the lack of the increase in DEVDase activity in Casp9 Ϫ/Ϫ MEFs.
The mitochondrial release of cytochrome c into the cytosol is required for activation of capase-9 (13). Consistently, we observed cytochrome c release in ASK1-induced apoptosis (Fig.  8). Of note, the addition of zVAD-fmk, which effectively inhibited the cell death induced by ASK1 (Fig. 3), caused no significant reduction in cytochrome c release, indicating that caspase activities are not required for ASK1-induced release of cytochrome c. Moreover, ASK1-induced release of cytochrome c obviously preceded the onset of cell death. These results strongly suggest that cytochrome c release is not a consequence but a trigger of ASK1-induced caspase activation. Mitochondrion may thus be a direct target of death signals mediated via ASK1-JNK and/or -p38 axes.
The molecular target for the activated JNK or p38 to induce cytochrome c release from mitochondria remains to be elucidated. Candidates as such are the Bcl-2 family members, which play pivotal roles in decision of cell death and survival mainly at the level of mitochondria (9,10). Recently, Bcl-2 was reported to be phosphorylated in a normal physiologic process at G 2 /M as well as in response to microtubule-interfering agents via the ASK1-JNK1 pathway (45). The phosphorylated Bcl-2 is suggested to lose its anti-apoptotic activity, and therefore cells appear to be susceptible to death signals (45). It will be interesting to study whether other members in the Bcl-2 family may also be regulated by ASK1-JNK and/or -p38 signals.
In conclusion, the results presented here strongly suggest that the mitochondria-initiated activation of effector caspases is a main mechanism operating in the execution of ASK1induced apoptosis. However, further investigations will be needed to elucidate the issue of cell-type specificity in ASK1 signaling and the possible involvement of other signaling pathways mediating ASK1-induced death signals.