Activation of p38 Mitogen-Activated Protein Kinase Is Required for Death Receptor–Independent Caspase-8 Activation and Cell Death in Response to Sphingosine Free

Sphingolipid metabolites, such as ceramide, sphingosine, and sphingosine 1-phosphate (S1P), are key regulators of diverse cellular processes, including apoptosis, cell proliferation, differentiation, and inflammation (1). Considerable attention has been focused on ceramide as a potential endogenous mediator of apoptosis in response to cytokines, antigens, anticancer drugs, or environmental stress (2, 3). Indeed, these diverse exogenous stimuli rapidly enhance intracellular levels of ceramide through sphingomyelinase-mediated hydrolysis of sphingomyelin. Ceramide is further metabolized by ceramidase to generate sphingosine during the early stages of apoptosis. Moreover, sphingosine itself is capable of triggering apoptosis when added exogenously to a variety of leukemic cells or solid cancer cell lines (4). These findings have led to the proposal that sphingosine plays a key role in apoptosis signaling.

Accumulating evidence suggests a role of mitogen-activated protein kinase (MAPK) in sphingosine-induced apoptosis. For instance, sphingosine treatment results in complete inhibition of extracellular signal–regulated kinase (ERK) activity in leukemic and solid cancer cells (5-7), indicating that suppression of this pathway is required for sphingosine-induced apoptosis. Moreover, sphingosine is suggested to function as a mediator of stress responses, leading to the activation of c-jun NH₂-terminal kinase (JNK) or p38 MAPK (8-11). However, the molecular mechanism by which apoptotic cell death occurs in response to sphingosine has been widely explored but not precisely deciphered.

A close relationship exists between the ERK pathway and the p38 MAPK or JNK pathway in a variety of eukaryotic cells (8-11). The concomitant inactivation of survival signals may be necessary for JNK- and p38 MAPK–mediated cell death (11). Interestingly, deprivation of neurotrophic factors or UV irradiation not only activates stress kinase cascades but also leads to dramatic inhibition of the ERK pathway (12). In addition, ERK is required for survival signaling in response to cellular growth factors (13). Furthermore, overexpression of ERK in NIH 3T3 cells largely impairs the UV-induced apoptotic response (14). Thus, it seems that the cell fate (i.e., death or survival) is decided by a critical balance between the ERK pathway and the JNK or p38 MAPK pathway. Regulation of MAPKs involves dynamic interplay between kinases and phosphatases. ERKs are activated by phosphorylation of both conserved threonine and tyrosine residues and inactivated on dephosphorylation by tyrosine and serine-threonine phosphatases (15-20). Serine-threonine protein phosphatase 2A (PP2A) dephosphorylates MAPK/ERK kinase (MEK) and ERK family kinases in vitro (19). Moreover, inhibition of PP2A leads to the activation of MEK and ERK (21, 22). Recent studies show that ERK activity is down-regulated by serine-threonine phosphatase during p38 MAPK activation induced by arsenite in NIH 3T3 fibroblasts (23). Moreover, p38 MAPK regulates activation of the JNK pathway in human neutrophils via PP2A (24).

In the present study, we investigate the signaling pathway that triggers apoptotic cell death in response to sphingosine. Activation of p38 MAPK contributes to death receptor–independent caspase-8 activation, and suppression of ERK activity exerts positive effects on p38 MAPK signaling. We additionally present evidence that PP2A activity is necessary for sphingosine-induced suppression of ERK activity. The molecular signaling pathways leading to sphingosine-induced apoptotic cell death may be applied to improve chemotherapeutic approaches for human malignancies.

To investigate the kinetics of apoptotic cell death induced by sphingosine, we treated Jurkat T cells with different doses of sphingosine for various times and analyzed apoptotic cell death by Hoechst 33258 staining. Figure 1A depicts a dose- and time-dependent increase in the proportion of apoptotic cells after sphingosine treatment. Treatment of cells with sphingosine also caused activation of caspase-8 and caspase-3 (Fig. 1B). Requirement of caspase activity for sphingosine-induced apoptosis was additionally examined using a broad-spectrum caspase inhibitor, z-VAD-fmk, and a caspase-8 inhibitor, z-IETD-fmk. Treatment with the caspase inhibitors prevented sphingosine-induced apoptosis (Fig. 1C), supporting the theory that the cell death occurs in a caspase-8–dependent manner. We next examined the changes in intracellular levels of ceramide and S1P after treatment with sphingosine. Intracellular level of ceramide was significantly increased following sphingosine treatment, but the level of S1P was subtly increased (Fig. 1D). To rule out the effects of ceramide and S1P on apoptotic cell death after sphingosine treatment, we pretreated cells with fumonisin B1, an inhibitor of ceramide synthase, or small interfering RNA (siRNA) for sphingosine kinase. As expected, fumonisin B1 and si-sphingosine kinase completely inhibited induction of ceramide and S1P, respectively, in response to sphingosine (Fig. 1E). However, fumonisin B1 or si-sphingosine kinase failed to change sphingosine-induced apoptotic cell death (Fig. 1F). Moreover, pretreatment with fumonisin B1 and si-sphingosine kinase did not affect caspase-8 and caspase-3 activations in response to sphingosine treatment (Fig. 1G and H).

Several studies suggest that caspase-8 activation is dependent on its oligomerization, which is stimulated by association with the adaptor molecule, Fas-associated death domain (FADD), via the death effector domains of the two molecules. This step requires either direct or indirect interactions of FADD with surface receptors possessing death domains, such as Fas or tumor necrosis factor receptor (25, 26). To determine the involvement of the death receptor in sphingosine-induced apoptosis, we immunoprecipitated FADD, followed by immunoblotting for caspase-8. As shown in Fig. 2A, caspase-8 activation and interactions between FADD and caspase-8 were detected following treatment with Ch11, a soluble Fas ligand. However, whereas the addition of sphingosine led to caspase-8 activation, no interactions between caspase-8 and FADD were detected after sphingosine treatment. To further confirm these results, we examined caspase-8 activation in FADD-deficient Jurkat T cells in response to sphingosine. Despite FADD deficiency, sphingosine effectively induced caspase-8 activation (Fig. 2B) and apoptotic cell death (Fig. 2C). In addition, no changes in death receptor levels were observed after sphingosine treatment (data not shown). Our results establish that sphingosine-induced caspase-8 activation occurs in a death receptor–independent manner.

MAPKs play a pivotal role in regulating apoptosis in response to various stimuli (5-11). To investigate whether the MAPK signaling pathway is involved in sphingosine-induced cell death, we initially examined the MAPK activities and changes in protein levels following sphingosine treatment. As shown in Fig. 3A, the addition of sphingosine led to a dramatic increase in phosphorylated p38 MAPK. Increased p38 MAPK activity was initially detectable at 1 hour and gradually increased until 6 hours. However, sphingosine induced marked suppression of phosphorylated ERK1/2. Phosphorylated ERK started to diminish within 1 hour of sphingosine treatment and was sustained for 6 hours. The phosphorylated JNK level after sphingosine treatment was constant over time-course studies. In addition, pretreatments with fumonisin B1 and si-sphingosine kinase did not affect sphingosine-induced suppression of phosphorylated ERK1/2 and increase in phosphorylated p38 MAPK (Fig. 3B and C). These results suggest that sphingosine triggers rapid activation of p38 MAPK and suppression of ERK activity in Jurkat T cells.

A number of studies show that p38 MAPK acts at an early step, before caspase activation during the apoptotic process induced by specific stimuli (27, 28). In contrast, under certain circumstances, p38 MAPK activation is dependent on caspase activation (29). Therefore, we examined the effects of p38 MAPK inhibition on caspase-8 activation in response to sphingosine. Retroviral expression of dominant-negative p38 MAPK (MFG-Flag-DN-p38, T180A/Y182F) or pretreatment with PD169316 effectively blocked sphingosine-induced caspase-8 activation (Fig. 3D) and apoptotic cell death (Fig. 3E). However, ERK activity was not affected by p38 MAPK inhibition. Moreover, pretreatment with PD169316 clearly attenuated sphingosine-induced caspase-8 activation in FADD-deficient Jurkat cells (Supplementary Fig. S1).

To determine whether suppression of ERK activity is required for sphingosine-induced apoptosis, we transfected cells with constitutive active forms of MEK1 and examined apoptotic cell death. Overexpression of constitutively active forms of MEK1 (CA-MEK1 and S217E/S221E) attenuated sphingosine-induced decrease of phosphorylated ERK and activation of caspase-8 (Fig. 4A). Moreover, CA-MEK1 markedly suppressed apoptotic cell death in response to sphingosine treatment (Fig. 4B). Interestingly, overexpression of CA-MEK1 effectively attenuated sphingosine-induced p38 MAPK activation. To further clarify the relationship between ERK activation and sphingosine-induced cell death, we used U0126, a highly selective inhibitor of MEK. Simultaneous treatment with U0126 and sphingosine of cells overexpressing CA-MEK1 effectively reversed the inhibitory effect of MEK1 on sphingosine-induced activation of p38 MAPK and caspase-8 (Fig. 4C) and apoptotic cell death (Fig. 4D). These results suggest that during sphingosine-induced apoptosis, suppression of the MEK/ERK pathway is required for activation of p38 MAPK and subsequent caspase-8 activation.

The MEK/ERK pathway is regulated by the potent serine/threonine protein phosphatase PP2A (19). We examined whether PP2A is involved in suppression of the ERK pathway in response to sphingosine. Spectrophotometric analysis revealed a rapid increase in PP2A activity that reached a peak after 1 hour and declined to the starting level after 6 hours in the presence of sphingosine (Fig. 5A). However, the PP2A level was not altered in the presence of sphingosine over the time course examined (Fig. 5B). siRNA targeting of PP2A effectively attenuated sphingosine-induced increase in PP2A activity (Fig. 5C) and suppression of MEK and ERK activities (Fig. 5D). Moreover, si-PP2A significantly suppressed sphingosine-induced p38 MAPK activation, subsequent caspase-8 activation (Fig. 5D), and apoptotic cell death (Fig. 5E). These findings suggest that serine/threonine protein phosphatase, PP2A, is involved in sphingosine-induced apoptotic cell death through suppression of the MEK/ERK signaling pathway.

Despite extensive characterization, the precise molecular mechanism of anticancer activity of sphingolipid metabolites remains to be elucidated. In this investigation, we analyzed the basis for linkage between signaling pathways and apoptotic cascades during sphingosine-induced cell death. Our data show that sphingosine induces apoptotic cell death via p38 MAPK–mediated, death receptor–independent caspase-8 activation, and that suppression of the MEK/ERK pathway by PP2A is required for p38 MAPK activation in response to sphingosine.

p38 MAPK and/or JNK is positively associated with apoptosis induction in response to various cellular stimuli (8-11). Moreover, the p38 MAPK pathway plays an important role in sphingosine-induced apoptosis. Sphingosine treatment induced rapid phosphorylation/activation of p38 MAPK. This rapid activation suggests that p38 MAPK plays a key role in the early events of sphingosine-induced apoptosis. Consistent with this theory, inhibition of p38 MAPK completely attenuated sphingosine-induced caspase-8 activation and apoptotic cell death. Under specific conditions, caspases regulate the activation of p38 MAPK as in Fas-treated Jurkat T cells (30). However, activation of p38 MAPK by sphingosine was insensitive to the broad-spectrum caspase inhibitor z-VAD-fmk (data not shown), suggesting that this step occurs upstream of the caspase cascade during sphingosine-induced apoptosis.

Accumulating evidence suggests a role of ERK in sphingosine-induced apoptosis. For instance, sphingosine treatment results in complete inhibition of ERK activity in leukemic and solid cancer cells (5-7), indicating that suppression of this pathway is required for sphingosine-induced apoptosis. However, involvement of the apoptotic target machinery of the ERK pathway in sphingosine-induced cell death has not been established to date. We provide evidence in this study that suppression of the MEK/ERK pathway is required for p38 MAPK activation, caspase-8 activation, and apoptotic cell death in response to sphingosine. A relationship between the ERK pathway and the p38 MAPK or JNK pathway has been reported in a variety of eukaryotic cells (8-11). It seems that the ability of a cell to die or survive is decided by a critical balance between the ERK pathway and the p38 MAPK or JNK pathway. Deprivation of neurotrophic factors or UV irradiation not only activates stress kinase cascades but also leads to dramatic inhibition of the ERK pathway (13). Furthermore, ERK expression in NIH 3T3 cells impairs the majority of the UV-induced apoptotic response (14). The prosurvival function of ERK was confirmed in a recent report showing that inhibition of ERK signaling leads to increased sensitivity to cisplatin (cis-diammine-dichloroplatinum) in ovarian cancer cells (31). Conversely, activation of the MEK/ERK signaling pathway in activated Jurkat T cells suppresses tumor necrosis factor-related apoptosis-inducing ligand–mediated apoptosis in a similar manner to Fas-mediated apoptosis (30). These studies support the general view that activation of the ERK pathway delivers a survival signal that counteracts the proapoptotic effects associated with p38 MAPK or JNK activation.

Regulation of MAPKs involves a dynamic interplay between kinases and phosphatases. ERKs are activated by phosphorylation of both conserved threonine and tyrosine residues and inactivated on dephosphorylation by specific phosphatases (15-20). Serine-threonine PP2A can dephosphorylate MEK- and ERK-family kinases in vitro (21, 22). In this study, we also found that inhibition of PP2A clearly restored sphingosine-induced suppression of MEK and ERK. Moreover, inhibition of PP2A effectively reduced p38 MAPK, caspase-8 activation, and apoptotic cell death in response to sphingosine. From these results, we suggest that activation of PP2A induces p38 MAPK activation–mediated, caspase-8–dependent apoptotic cell death induced by sphingosine via suppression of the MEK/ERK pathway.

Our results conclusively indicate that activation p38 MAPK is required for death receptor–independent caspase-8 activation, and that suppression of the MEK/ERK pathway by PP2A is associated with p38 MAPK activation. An improved understanding of the mechanisms involved in sphingosine metabolite–induced apoptosis may ultimately provide novel strategies of intervention of specific signaling pathways to favorably alter therapeutic efficacy in the treatment of human malignancies.

The method of constructing the MFG retroviral vector by replacing the GFP sequence of MFG.GFP.IRES.puro was used to construct MFG-Flag-DN-p38 MAPK (T180A/Y182F). The MFG.GFP.IRES.puro itself was used as a negative control throughout the experiment. The retroviral plasmids were introduced into T293 retrovirus packaging cell line by transient transfection with Lipofectamine 2000 (Invitrogen). After 72 h, the supernatants were harvested and used for retroviral infection. The virus titers, measured in NIH 3T3 cell line by puromycin-resistant colony formation, were between 10⁵/mL and 5 × 10⁵/mL. Target cells were infected with 1 mL of MFG-Flag-DN-p38 supernatants in the presence of 8 μg/mL polybrene (Sigma Chemical Co.). Constitutively active forms of MEK1 (S217E/S221E) were purchased from Cell Biolabs. The plasmids were introduced into target cells by transient transfection with Lipofectamine 2000 (Invitrogen).

Jurkat, human T-cell lymphoma (type II), and FADD-deficient Jurkat T cells were obtained from American Type Culture Collection. Cells were grown in RPMI 1640 supplemented with 10% fetal bovine serum (Life Technologies, Inc.), penicillin, and streptomycin at 37°C in a humidified incubator with 5% CO₂.

Sphingosine was purchased from Sigma. Polyclonal antibodies specific for caspase-8, phospho-JNK, JNK, p38 MAPK, and PP2A and monoclonal anti–phospho-ERK1/2 and anti-ERK antibodies were from Santa Cruz. Polyclonal anti–β-actin antibody was obtained from Sigma. Polyclonal antibodies for cleaved caspase-3, MEK1, and phospho-p38 MAPK were from Cell Signaling Technology. The broad-spectrum caspase inhibitor z-VAD-fmk, caspase-8 inhibitor z-IETD-fmk, MEK-specific inhibitor PD98059, and p38 MAPK–specific inhibitor PD169316 were acquired from Calbiochem. Constitutively active forms of MEK1 (S217E/S221E) were purchased from Cell Biolabs.

Hoechst 33258 staining was done as described previously (32). Briefly, cells were fixed with 4% paraformaldehyde for 30 min at room temperature and washed once with PBS. Hoechst 33258 (50 ng/mL) was added to fixed cells, which were incubated for 30 min at room temperature and then washed with PBS. Cells were mounted and examined by fluorescence microscopy. Apoptotic cells were identified by condensation and fragmentation of their nuclei. The percentage of apoptotic cells was calculated from the ratio of apoptotic cells to total cells counted. A minimum of 500 cells were counted for each treatment.

The cells were washed twice with PBS. Reverse transcription-PCR was done. In brief, total RNA of the cells was isolated from cells by Trizol (Invitrogen). cDNA was obtained by reverse transcription of 2 μg of total RNA, and amplification of the respective cDNA region was conducted by PCR. PCR primers used were, for sphingosine kinase, CTCTGGTGGTCATGTCTGGA (sense) and CAGGTGTCTTGGAACCCACT (antisense). β-Actin cDNA was used as an internal standard. The PCR consisted of 30 cycles (94°C for 1 min, 58°C for 1 min, and 72°C for 1 min). PCR products were analyzed by electrophoresis on 1% agarose gels.

PP2A was immunoprecipitated and its activity determined using a nonradioactive kit (Upstate Biotechnologies). Jurkat cell lysates were prepared in 20 mmol/L imidazole-HCl, 2 mmol/L EDTA, 2 mmol/L EGTA (pH 7.0), with aprotinin, benzamidine, and 4-(2-aminoethyl)benzenesulfonylfluoride (Sigma-Aldrich). Total protein was immunoprecipitated with an anti-PP2A catalytic subunit antibody (Upstate) and protein A-Sepharose beads (Sigma-Aldrich). Equivalent immunoprecipitation of PP2A from all samples was confirmed by Western blot analysis. Immunoprecipitated PP2A was examined for activity in a 10-min reaction at 37°C in which phosphopeptide (K-R-pT-I-R-R) dephosphorylation was assayed spectrophotometrically at 650 nm using malachite green. PP2A activity was determined for all samples relative to control or a phosphate standard curve.

The lipids from cell pellets added to internal standard C₁₇ sphingosine–based ceramide (C₁₇ ceramide) were extracted with chloroform/methanol (1:2, v/v), and the extract residue was dissolved in methanol, spotted on TLC silica gel plates (Merck), and developed in chloroform/methanol/29% NH₄OH (40:10:1, v/v) to half of the plate length. After drying, the plate was rechromatographed in heptane/diisopropylether/acetic acid (60:40:3, v/v/v). Lipids were visualized by dipping into 10% sulfuric acid and drying. The band of ceramides corresponding to the band of standard C₁₇ ceramide was scraped off and eluted with methanol. Ceramidase was used to release sphingosine from ceramide by deacylation. The released sphingosine was derivatized with o-phthalaldehyde reagent. High-performance liquid chromatography (HPLC) analysis was done using a Shimadzu model LC-10AT pump and SIL-10AXL autoinjector. Nova-Pak C18 column was equilibrated with a mobile phase (92% methanol, 0.1% triethylamine) at a flow rate of 1 mL/min. The fluorescence detector (RF-10AXL) was set at an excitation wavelength of 340 nm and an emission wavelength of 455 nm.

Concentrations of sphingosine and S1P in cell pellets were determined by HPLC following a previously described procedure with modification (33). Cell pellets were lysed with 0.2 N NaOH for the determination of protein content. For the determination of sphingosine content, chloroform/methanol (1:2, v/v) and C₁₇ So as internal standard were added to cell lysates and the sphingolipids were extracted and analyzed by HPLC. For the determination of S1P content, chloroform/methanol-1 mol/L NaCl (1:1, v/v), 3N-NaOH, and C₁₇ S1P as internal standard were added to cell lysates and incubated for 5 h. The aqueous fraction of S1P extract was mixed thoroughly with reaction buffer [200 mmol/L Tris-HCl (pH 7.4), 75 mmol/L MgCl₂ in 2 mol/L glycine buffer, pH 9.0] and 50 units of alkaline phosphatase and then incubated at 37°C for 1 h. This procedure completely converted S1P to sphingosine by the action of alkaline phosphatase. The released sphingosine was analyzed by HPLC.

Data were analyzed using one-way ANOVA. Bonferroni's posttest multiple comparison procedure was used to determine statistical significance. Data are represented as mean ± SD.

No potential conflicts of interest were disclosed.