Antiapoptotic Activity of Akt Is Down-regulated by Ca2+ in Myocardiac H9c2 Cells

Cell survival signaling of the Akt/protein kinase B pathway was influenced by a change in the cytoplasmic free calcium concentration ([Ca2+]i) for over 2 h via the regulation of a Ser/Thr phosphatase, protein phosphatase 2Ac (PP2Ac), in rat myocardiac H9c2 cells. Akt was down-regulated when [Ca2+]i was elevated by thapsigargin, an inhibitor of the endoplasmic reticulum Ca2+-ATPase, but was up-regulated when it was suppressed by 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetra(acetoxymethyl)ester (BAPTA-AM), a cell permeable Ca2+ chelator. The inactivation of Akt was well correlated with the susceptibility to oxidant-induced apoptosis in H9c2 cells. To investigate the mechanism of the Ca2+-dependent regulation of Akt via the regulation of PP2A, we examined the transcriptional regulation of PP2Acα in H9c2 cells with Ca2+ modulators. Transcription of the PP2Acα gene was increased by thapsigargin but decreased by BAPTA-AM. The promoter activity was examined and the cAMP response element (CRE) was found responsible for the Ca2+-dependent regulation of PP2Acα. Furthermore, phosphorylation of CRE-binding protein increased with thapsigargin but decreased with BAPTA-AM. A long term change of [Ca2+]i regulates PP2Acα gene transcription via CRE, resulting in a change in the activation status of Akt leading to an altered susceptibility to apoptosis.

Calcium (Ca 2ϩ ) plays a signaling role in many important cellular functions, such as fertilization, embryonic pattern formation, differentiation, proliferation, contraction, secretion, and metabolism (1). The versatility of the Ca 2ϩ -signaling mechanism in terms of speed, amplitude and spatio-temporal patterning enables elevations of Ca 2ϩ to regulate many processes of cell activity. Ca 2ϩ exhibits cross-talk between a variety of signaling pathways (1). Ca 2ϩ affects the protein kinase A pathway by regulating the metabolism of cAMP. It also activates nitric-oxide (NO) synthase to generate NO, which in turn activates the cGMP pathway through the activation of guanylyl cyclase. The Ras/mitogen-activated protein kinase (MAPK) 1 and Ca 2ϩ /calmodulin/calmodulin kinase (CaMK) pathways are also controlled by Ca 2ϩ (1,2).
On the other hand, a cellular Ca 2ϩ overload or the perturbation of intracellular Ca 2ϩ compartmentalization can cause cytotoxicity and trigger apoptosis or necrosis (3,4). Under such circumstances, various Ca 2ϩ -dependent signaling cascades with kinases and phosphatases directly or indirectly influence cellular signaling. Protein kinase C family has been proposed to play an important role in the Ca 2ϩ -mediated signaling of apoptosis (5). Calcineurin/PP2B, a Ca 2ϩ -dependent Ser/Thr phosphatase (6), also appears to be involved in apoptosis (7). Together, these findings show that Ca 2ϩ has a pivotal role in the regulatory mechanism of signaling pathways in cell survival and death, although the precise mechanism of Ca 2ϩ -dependent cross-talk has not been fully clarified.
Akt/protein kinase B is a pleckstrin homology domain-containing Ser/Thr kinase (8,9,10). Akt is presently recognized as a cell survival or an antiapoptotic cellular signaling mediator. Akt is activated through a growth factor receptor-mediated activation of the phosphatidylinositol 3-kinase (PI3K) pathway (10). With growth factor signals, Akt is recruited to the plasma membrane and is activated through phosphorylation at Ser-473 and Thr-308 by phosphatidylinositol 3-phosphate-dependent protein kinase-1 (PDK1) or integrin-linked kinase (9,10). Akt can phosphorylate Bad, caspase-9, and forkhead-related transcription factors, leading to their inactivation and to enhanced cell survival (8,9,10). Inhibitor of nuclear factor B (IB) kinase is also phosphorylated by Akt leading to an upregulation of its activity and resulting in a promotion of the nuclear factor B (NFB)-mediated inhibition of apoptosis (8,11,12) Akt has been found to be involved in cell death following the withdrawal of extracellular signaling factors, oxidative and osmotic stress, irradiation, treatment with drugs and ischemic stress (10). However, in spite that a variety of cellular stressors influence cells through Ca 2ϩ signaling, the number of studies on Akt signaling and Ca 2ϩ is limited. As for Akt, Ca 2ϩ -dependent activation was reported in several studies (13,14). On the other hand, there was a report that the activation of Akt is independent of Ca 2ϩ (15). In contrast, we found that Akt was suppressed by an elevation of [Ca 2ϩ ] i in myocardiac H9c2 cells overexpressing the calreticulin gene (16). In the cells overexpressing calreticulin, protein phosphatase 2A (PP2A) was upregulated by Ca 2ϩ to decrease the phosphorylation level of Akt, and the inactivated status of Akt was well correlated with the susceptibility to apoptosis in H9c2 cells under conditions for differentiation induced by retinoic acid. Collectively, these results suggest that the Ca 2ϩ -dependent regulatory mechanism of Akt signaling may be important to a variety of apoptotic signaling mechanisms, although how has not been fully clarified.
In the present study, to investigate the mechanism of the Ca 2ϩ -dependent regulation of Akt signaling, we examined the influence of a change of [Ca 2ϩ ] i on susceptibility to oxidative stress-induced cell injury and on the Akt signaling pathway in myocardiac H9c2 cells. We show that the Ca 2ϩ -dependent regulation of PP2Ac␣ gene transcription is controlled through the cAMP responsive element (CRE), resulting in a change in the activation status of Akt leading to an altered susceptibility to apoptosis.

MATERIALS AND METHODS
Antibodies and Reagents-Antibodies against Akt, phospho-Akt (Ser-473), and phospho-Akt (Thr-308), CRE-binding protein (CREB), and phospho-CREB (Ser-133) were purchased from Cell Signaling Technology (Beverly, MA). The antibody against Sp1 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The reagents used in the study were all of high grade and from Sigma or Wako Pure Chemicals (Osaka, Japan).
Cell Culture-H9c2 cells from embryonic rat heart (16,17) were obtained from American Type Culture Collection (CRL-1446). H9c2 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum in a humidified atmosphere of 95% air and 5% CO 2 at 37°C. Before reaching confluence, the cells were split, and plated at low density in culture medium containing 10% fetal bovine serum.
Measurement of Cytoplasmic Free Ca 2ϩ -The cytoplasmic free Ca 2ϩ concentration, [Ca 2ϩ ] i , was measured using Fura-2-AM essentially as described previously (16). Briefly, cultured cells on glass coverslips were loaded with 5 M Fura-2-AM (Dojindo, Kumamoto, Japan) for 20 min in Earle's balanced salt solution (EBSS) in the presence of 0.01% pluronic acid F-127. After four washes with EBSS, the cover glass was positioned in a quartz cuvette containing 3.5 ml of fresh EBSS at a 45°angle to both excitation and emission light paths. The fura-2 fluorescence was determined at 37°C using a spectrofluorophotometer operating at an emission wavelength of 505 nm with an excitation wavelength of 340 and 380 nm. The maximal signal (R max ) was obtained by adding ionomycin at a final concentration of 4 M. Then the minimal signal (R min ) was obtained by adding EGTA at a final concentration of 7.5 mM, followed by Tris-free base to a final concentration of 30 mM, to increase the pH to 8.3. R is the ratio (F1/F2) of the fluorescence of Ex 340 nm, Em 505 nm (F1) to that of Ex 380 nm, Em 505 nm (F2). The actual Ca 2ϩ concentration was calculated as K d X (R Ϫ R min )/(R max Ϫ R) with the K d equal to 224 nM (18).
Lactate Dehydrogenase (LDH) Release Assay-After 4 h of treatment with 5 M thapsigargin or 10 M BAPTA-AM or not, cells were incubated with 75 M hydrogen peroxide (H 2 O 2 ) for 0 -120 min. The LDH activity was assayed by using a MTX LDH kit (Kyokuto Seiyaku, Tokyo, Japan) according to the manufacturer's instructions. Briefly, 50 l of supernatant was transferred to a 96-well plate, then 50 l of coloring reagent was added and incubated for 45 min at room temperature. After 100 l of stop solution was added, absorbance was measured at 560 nm with a microplate reader. The LDH release was shown as a rate of LDH released in the medium to total cellular LDH.
TUNEL Assay-Apoptosis was detected flow cytometrically by the terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) method (19) using an ApopTag Plus Fluorescein in situ Apoptosis Detection kit (Serologicals, Norcross, GA). Briefly, cells were harvested and fixed in 70% ethanol, treated with terminal deoxynucle-otidyl transferase for 1 h, and then with fluorescein isothiocyanate (FITC)-conjugated antidigoxigenin for 1 h at room temperature, and washed with phosphate-buffered saline (pH 7.0) (PBS) containing 0.1% Triton X-100. The fluorescence intensity was measured at 530 nm using a flow cytometer (BD Biosciences, San Jose, CA).
Northern Blot Analysis-The full-length rat PP1␣ catalytic subunit and PP2A catalytic ␣ cDNAs were generously provided by Dr. Kunimi Kikuchi (Hokkaido University, Japan) (20,21). A PstI-SmaI fragment of 600 bp and EcoRI-PvuII fragment of 680 bp were prepared from the cDNAs of PP1␣c and PP2Ac␣, respectively, and used as cDNA probes. The probes were labeled with 32 P using a Random Primer Labeling kit (Takara Biomedicals, Shiga, Japan). The isolation of cytoplasmic RNA and Northern blotting were essentially performed as described before (16). Isolated RNAs (10 g) were electrophoresed on a 1% agarose gel containing 0.6 M formaldehyde, transferred to a nylon membrane, and then hybridized with 32 P-labeled probes. Autoradiographed membranes were analyzed using a BAS5000 bioimage analyzer (Fuji Photo Film).
Immunoblot Analysis-Cultured cells were harvested and lysed for 20 min at 4°C in lysis buffer (20 mM Tris-HCl, pH 7.5, 130 mM NaCl, 1% Nonidet P-40, and 10% glycerol including protease inhibitors (20 M amidinophenylmethanesulfonyl fluoride, 50 M pepstatin, and 50 M leupeptin)). The supernatants obtained on centrifugation at 8,000 ϫ g for 10 min were used in subsequent experiments. Protein samples were subjected to 10% SDS-PAGE under reducing conditions and then transferred to nitrocellulose membrane as described (22). The membrane was blocked with 5% skim milk or 3% bovine serum albumin in Trisbuffered saline (pH 7.5) containing 0.05% Tween 20. The blots were coupled with the peroxidase-conjugated secondary antibodies, washed, and then developed using the ECL chemiluminescence delection kit (Amersham Biosciences) according to the manufacturer's instructions.
Akt Activity Assay-Akt acitivity was assayed using an Akt assay kit (Cell Signaling Technology) according to the manufacturer's protocol. Briefly, Akt was immunoprecipitated from cell lysates using the anti-Akt antibody, and then the immunoprecipitates were incubated at 30°C for 30 min in an assay mixture containing an Akt substrate, GSK-3␣/␤ fusion protein. Phosphorylated proteins were separated by 12.5% SDS-PAGE and then transferred to nitrocellulose membrane to detect phosphorylated GSK-3␣/␤ using an anti-phosphorylated GSK-3␣/␤ (Ser-21/9) antibody.
Protein Phosphatase Assay-Protein Ser/Thr phosphatase activity was assayed photometrically using Ser/Thr phosphatase assay kit 1 (Upstate Biotechnology, Lake Placid, NY), according to the manufacturer's directions. The activity was assayed in the presence or absence of 10 nM okadaic acid, and the okadaic acid-sensitive activity was estimated as PP2A-specific activity. The phosphopeptide (RK(pT)IRR) was used as a phosphatase substrate. Protein concentrations were determined using a BCA assay kit (Pierce).
Generation of Luciferase Reporter Constructs-A ϳ1.6-kb fragment of rat PP2Ac␣ gene promoter (Ϫ1350 to ϩ258) (23) was amplified with rat genome by PCR using Pfu turbo DNA polymerase (Stratagene). The primers used are as follows; a forward primer (5Ј-GATCTCAGGTACTT-TCTTCCGGAACACTAG-3Ј) and a reverse primer (5Ј-GTCCAGCTCCT-TGGTGAACAACTTC-3Ј). The PCR product was subcloned into pUC18 to obtain pUC18-pro-PP2Ac. The nucleotide sequence was confirmed by sequencing with an ALFexpress II system (Amersham Biosciences). pUC18-pro-PP2Ac was digested with HindIII, and the resulting fragment containing the promoter region from Ϫ1209 to ϩ258 was inserted into the HindIII site of the reporter vector pGL3-Basic (Stratagene) to give pGL3-pro-PP2Ac. To generate deleted mutants of the luciferase reporter construct, pGL3-pro-PP2Ac was digested with SacI and XhoI, and deletion mutants were made using a deletion kit for kilo sequence (Takara Biomedicals).
Luciferase Activity Assay-Each vector was transfected into H9c2 cells by using Lipofectamine2000 (Invitrogen) according to the manufacturer's instructions. After 24 h of transfection, cells were treated with thapsigargin (5 M) or BAPTA-AM (10 M), or left untreated for the periods indicated in the text. Then luciferase activity was assayed with cellular extracts by using a dual-luciferase reporter assay system (Promega).
Indirect Immunofluorescence Microscopy-After treatment with 5 M thapsigargin or 10 M BAPTA-AM for 2 h, cells incubated on Lab-Tek chamber slides (Nunc) were fixed with 3% paraformaldehyde for 20 min at room temperature and washed three times with PBS. Cells were permeabilized in 1% Triton X-100 in PBS for 10 min and washed three times with PBS. They were blocked with 1% bovine serum albumin in PBS for 30 min at room temperature, washed three times with PBS, and then incubated with antiphospho-CREB (Ser-133) overnight at 4°C. Cells were washed with PBS four times and incubated with FITC-labeled anti-rabbit IgG antibody for 30 min in a dark room. The immunoreactive signals were visualized by indirect immunofluorescence microscopy.  Fig. 1A, with thapsigargin, [Ca 2ϩ ] i shows a transient increase within the first 10 min. It then decreases to the basal level till 30 min, but again increases and retains elevated until 120 min, before gradually decreasing to the initial level. In contrast, with BAPTA-AM, [Ca 2ϩ ] i shows a continuous lowering during the treatment. To investigate the influence of the long term change of [Ca 2ϩ ] i on susceptibility to oxidative stress-induced cell injury, cells were treated with Ca 2ϩ modulators (i.e. thapsigargin and BAPTA-AM) for 4 h then exposed to 75 M hydrogen peroxide (H 2 O 2 ) for 0 -120 min, and cell damage was examined at predetermined times using the LDH release assay as described in the methods. In Fig. 1B, the release of LDH by H 2 O 2 was observed only at 120 min in the medium of untreated cells. However, the release by H 2 O 2 was initially observed at 60 min and increased at 120 min in the cells treated with thapsigargin. In contrast, BAPTA-AM treatment completely suppressed the release of LDH by H 2 O 2 throughout the 120 min. Next, to investigate whether apoptosis is involved in the mechanism of thapsigargin-induced cell damage, cells were treated with Ca 2ϩ modulators for 4 h then exposed to H 2 O 2 (75 M) for 2 h, and apoptosis was examined. Fig. 1C shows that TUNEL-positive fluorescence intensity was increased slightly by H 2 O 2 (upper), but was significantly enhanced after the pretreatment with thapsigargin (middle). In contrast, no change was observed in the fluorescent intensity of the BAPTA-AM-treated cells with or without H 2 O 2 (lower). Collectively, these results indicate that the susceptibility to apoptosis was enhanced with a long term elevation of [ ing. Previously, we found that the Akt signaling pathway is responsible for the cytoprotective mechanism in H9c2 cells under conditions of stress such as serum starvation with retinoic acid (16), and oxidative stress with hydrogen peroxide (27). Although elevations in Ca 2ϩ act as a signal, a prolonged increase in the concentration of Ca 2ϩ can be lethal (1). Moreover, cell signaling molecules including transcription factors are activated differentially by the amplitude and duration of the response to Ca 2ϩ (28). In the present study, H9c2 cells were treated with Ca 2ϩ modulators such as thapsigargin (5 M) and BAPTA-AM (10 M) for over 2 h to induce a long term change of [Ca 2ϩ ] i . After the treatments, the phosphorylation levels of Ser-473 and Thr-308 of Akt were examined. The phosphorylation of Thr-308 located in the kinase catalytic domain of Akt is necessary for the activation, and the phosphorylation of Ser-473 located in the regulatory domain of Akt supports the activation. As shown in Fig. 2, A (right) and B, the treatment with BAPTA-AM increased the phosphorylation of Akt both at Ser-473 and Thr-308, and the phosphorylation level increased to a maximum at 2 h, and was sustained thereafter till 4 h. In contrast, the phosphorylation of Akt decreased in a time-dependent manner with thapsigargin ( Fig. 2, A (left) and B). Next we examined whether Ca 2ϩ modulators also have an effect on the kinase activity of Akt. Fig. 2C shows that Akt activity is suppressed after the treatment with thapsigargin, but increased with BAPTA-AM. These results were consistent with the change in the phosphorylation status of Akt on treatment with Ca 2ϩ modulators such as thapsigargin and BAPTA-AM. Together, Akt signaling was suppressed by the long term elevation of [Ca 2ϩ ] i with thapsigargin, but was enhanced by the long term lowering of [Ca 2ϩ ] i with BAPTA-AM. This suggests a Ca 2ϩ -dependent regulation of Akt signaling in H9c2 cells. Furthermore, the Ca 2ϩ -induced suppression of Akt signaling was compatible with the enhanced susceptibility to apoptosis in H9c2 cells treated with H 2 O 2 (Fig. 1, B and C).

Elevation of [Ca 2ϩ ] i Gradually Accelerates Cell Damage and Apoptosis in H9c2 Cells
The Expression of PP2Ac␣ Is Transcriptionally Regulated by Ca 2ϩ Modulators-3-Phosphoinositide-dependent protein kinase (PDK1) is known to be responsible for phosphorylating Akt at Thr-308, and is activated by both phosphatidylinositol (3,4,5)-trisphosphate and phosphatidylinositol (3,4,)-bisphosphate, products of PI3K (9,10). To investigate whether the upstream kinases are involved in the regulation of Akt by the change of [Ca 2ϩ ] i , the activities for PI3K and PDK1 were measured in the cells treated with Ca 2ϩ modulators. However, neither activities showed any significant change even if the cells were treated with thapsigargin or BAPTA-AM for 0 -4 h (data not shown). Therefore, we focused on Ser/Thr protein phosphatases that could dephosphorylate and inactivate Akt to regulate the Akt signaling pathway (7). To investigate whether [Ca 2ϩ ] i levels affect the expression of protein Ser/Thr phosphatases, the cells were treated with 5 M thapsigargin or 10 M BAPTA-AM for 0 -4 h, and transcriptional levels were estimated by Northern blot analysis for protein phosphatase 2A catalytic subunit ␣ (PP2Ac␣) and protein phosphatase 1␣ catalytic subunit (PP1␣c). In Fig. 3A, the level of PP2Ac␣ mRNA was increased by thapsigargin but decreased by BAPTA-AM. In contrast, the mRNA level of PP1␣c was not significantly changed by the Ca 2ϩ modulators. In the immunoblot analysis, the protein level of PP2Ac␣ was increased by thapsigargin but decreased by BAPTA-AM (Fig. 3B). However, the protein level of PP1␣c was not influenced by thapsigargin or BAPTA-AM. The protein level of calcineurin/PP2B was not influenced by thapsigargin or BAPTA-AM either (data not shown). These results were consistent with results of the change of transcriptional levels for the phosphatases in the cells treated with each Ca 2ϩ modulator. The enzymatic activity of PP2A was also assayed in the cells treated with thapsigargin or BAPTA-AM for 0 -4 h. As shown in Fig. 3C, the activity of PP2A increased with thapsigargin by ϳ2-fold compared with that of untreated cells. In contrast, the activity was slightly suppressed after 2 h treatment with BAPTA-AM. Collectively, these results indicate that PP2Ac␣ expression is transcriptionally regulated by the long term change of [Ca 2ϩ ] i to control the phosphorylation status of target molecules including Akt.
Inhibition of PP2A Activity by Okadaic Acid Enhanced the Phosphorylation of Akt-Okadaic acid, a polyether toxin from the marine black sponge Halichondria okadai is a highly selective inhibitor of PP2A (29). To establish a link between Akt and PP2A, the influence of okadaic acid on Akt signaling was investigated with cells treated with okadaic acid. The cells were treated with okadaic acid (100 nM) for 0 -30 min. The phosphorylation level of Akt was estimated by immunoblot analysis using specific antibodies. Akt kinase and PP2A activities were measured as described above. As shown in Fig. 4A, okadaic acid reduced the activity of PP2A by ϳ30%, and in-creased the phosphorylation level of Akt (both Thr-308 and Ser-473) and Akt kinase activity. Next, we examined whether the decrease in phosphorylation of Akt caused by thapsigargin could be reversed by okadaic acid. The cells were preincubated with thapsigargin (5 M) for 4 h, and treated with or without okadaic acid (100 nM) for another 30 min, then the phosphorylation level of Akt was examined as described above. As shown in Fig. 4B, okadaic acid suppressed PP2A activity, and this reversed the decrease in phosphorylation of Akt with thapsigargin. Okadaic acid enhanced the phosphorylation of Akt at concentrations higher than 50 nM (data not shown). These results indicate that the function of PP2A can regulate the phosphorylation status of Akt in H9c2 cells.
Inhibition of PP2A Activity by Okadaic Acid Suppressed Apoptotic Cell Damage in H9c2 Cells Treated with Thapsigargin and H 2 O 2 -To establish a link between PP2A and Ca 2ϩ -dependent enhancement of apoptosis, the influence of okadaic acid on apoptosis was investigated using cells treated with okadaic acid. In Fig. 5A, cells were treated with or without anti-phospho-Akt (Thr-308), and anti-Akt antibodies. B, quantitative data for the phosphorylation status of Akt shown in A. The band intensity was estimated densitometrically, and the phosphorylation rate is expressed as the relative intensity of the phosphorylated Akt (Akt-P)/Akt. Each value represents the mean Ϯ S.D. of four independent experiments. C, Akt kinase activity was assayed as described under "Materials and Methods" using GSK-3␣/␤ as a substrate. Phosphorylated GSK-3␣/␤ (Ser-21/9), the enzymatic products of Akt, was detected by immunoblot analysis using specific antibody. The data represent three independent experiments. A, transcriptional levels of PP2Ac␣ and PP1␣c were evaluated by Northern blot analysis as described under "Materials and Methods." An EcoRI-PvuII fragment of 680 bp and a PstI-SmaI fragment of 600 bp were prepared from cDNAs for PP2Ac␣ and PP1␣c, respectively, and were used as cDNA probes after labeling with [␣-32 P]dCTP. B, protein levels for PP2Ac␣ and PP1c␣ were estimated by immunoblot analysis using specific antibodies. The data represent three independent experiments. C, protein phosphatase activity was assayed photometrically as described under "Materials and Methods." Each value represents the mean Ϯ S.D. of four independent experiments. thapsigargin (5 M) for 4 h then treated with okadaic acid (200 nM) for 30 min. Then cells were exposed to H 2 O 2 (75 M) for 0 -2 h, and cell damage was examined at predetermined times using LDH release assay. The results showed that treatment with okadaic acid suppressed the Ca 2ϩ -dependent enhancement of cell damage in cells treated with thapsigargin and H 2 O 2 . Fig.  5B shows a dose-dependent effect of okadaic acid on the Ca 2ϩdependent enhancement of cell damage. Cells were treated with thapsigargin (5 M) for 4 h then treated with different concentrations of okadaic acid (0 -500 nM) for 30 min. Thereafter, cells were exposed to H 2 O 2 (75 M) for 2 h, and cell damage was examined using LDH release assay. Okadaic acid showed maximal cytoprotective effects at a limited concentration range around 100 -200 nM. The protective effect of okadaic acid was rather diminished at concentrations higher than 500 nM (data not shown). Fig. 5C shows the effect of okadaic acid on Ca 2ϩdependent enhancement of apoptosis. Cells were treated with thapsigargin (5 M) for 4 h then treated with or without okadaic acid (100 nM) for 30 min. Then cells were exposed to H 2 O 2 (75 M) for 2 h, and apoptosis was examined by the TUNEL method. After thapsigargin treatment without okadaic acid, TUNEL-positive fluorescence intensity was significantly increased by H 2 O 2 (Figs. 1C and 5C, Tg(ϩ)OA(Ϫ)H 2 O 2 (ϩ)). In contrast, okadaic acid reduced the TUNEL-positive fluores-cence intensity even in the cells treated with thapsigargin and H 2 O 2 (Fig. 5C, Tg(ϩ)OA(ϩ)H 2 O 2 (ϩ)) compared with that of cells treated with thapsigargin and H 2 O 2 without okadaic acid. Okadaic acid did not solely affect the fluorescence intensity in cells with thapsigargin without H 2 O 2 (Fig. 5C, Tg(ϩ)OA(Ϫ) and Tg(ϩ)OA(ϩ)). However, it is noteworthy that okadaic acid shows an antiapoptotic effect at a limited concentration range around 100 nM. At concentrations above 500 nM, the antiapoptotic effect of okadaic acid was diminished or rather it showed enhancement of apoptosis in the cells treated with thapsigargin and H 2 O 2 (data not shown). This may be explained by the dualistic effect of inhibiting the effects of PP1 and PP2A on both apoptosis and cell proliferation in cells exposed to okadaic acid or microcystin-LR (30). Together, these results indicate that okadaic acid, a specific inhibitor of PP2A, inhibits apoptotic cell damage in H9c2 cells treated with thapsigargin and H 2 O 2 , and strongly suggests that PP2A up-regulates the Ca 2ϩdependent enhancement of apoptosis by dephosphorylating Akt to inhibit cell survival signaling. Characterization of PP2Ac␣ Gene Promoter-To investigate the mechanism of the transcriptional regulation of PP2Ac␣ expression, we used the 1.6-kb genomic fragment containing the promoter region of PP2Ac␣ inserted into a luciferase vector, pGL3 Basic. The transcriptional initiation site (nucleotide ϩ1) was denoted in accordance with the report of Kitagawa et al. (23) (Fig. 6, arrows). The promoter region contains neither a TATA box nor a consensus CAAT sequence. The promoter region contains various transcription factor binding sites such as, CRE at position Ϫ26, GC box for Sp1 (31) at positions Ϫ155 and Ϫ10, and binding sites for receptor-related orphan receptor ␣ (ROR␣) (32) at positions Ϫ778 and Ϫ553.
Both CRE and Sp1 Contribute to the Basal Expression of the PP2Ac␣ Gene-In the report by Kitagawa et al. (23), a fragment of 118 bp (Ϫ162 to Ϫ44) was defined as the essential region for the gene expression of PP2Ac␣. As shown in Fig. 6, the deletion fragments of the gene promoter were made from the HindIIIdigested fragment (Ϫ1209 to ϩ258, full-length) (Construct 1), and subcloned into a pGL3 Basic vector. Then, the activity of luciferase was assayed with the cells transfected with each deletion mutant vector. The activity was fully maintained in the 537-bp fragment (Ϫ279 to ϩ258) (Construct 2), but it decreased to ϳ65% of the full activity in the 403-bp fragment (Ϫ145 to ϩ258) (Construct 3) containing the CRE site. In the upstream fragment of 1064-bp (Ϫ1209 to Ϫ145) (Construct 6) containing the GC box, the activity decreased to ϳ40% of the full activity. However, the activity was almost lost in the upstream fragment of 1044 bp (Ϫ1209 to Ϫ165) (Construct 5) and the downstream fragment of 259 bp (Ϫ1 to ϩ258) (Construct 4). The results indicated that the full promoter activity was located in the sequence between Ϫ279 and Ϫ1. To examine whether the CRE and GC box contribute to the full promoter activity, disabled mutants were generated for the consensus sequences of CRE at Ϫ26 and GC box at Ϫ155 in the luciferase vector containing the 537-bp Construct 2. In Construct 2-Mt/G containing a CRE but no GC box, the activity decreased to ϳ70% of the full activity, and the level was similar to that of Construct 3. In Construct 2-Mt/G&C, the activity was significantly suppressed to less than 5% of full activity by both mutations. Furthermore, by the mutation with CRE, the activity was completely lost in Construct 3-Mt/C. Taken together, these results indicate that both the CRE and GC box contribute to the basal expression of the PP2Ac␣ gene in H9c2 cells.
Thapsigargin Enhanced Protein-CRE Complex Formation-To examine if a DNA-binding protein like Sp1 or CREB could interact with the PP2Ac␣ promoter, EMSA was performed with nuclear extracts from the cells treated with 5 M thapsigargin or 10 M BAPTA-AM using 32 P-labeled oligonucleotides designed for the GC box at Ϫ155 and CRE at Ϫ26. In the case of the CRE, a major band appeared but it disappeared in the presence of an excess of unlabeled probe (not shown) or 32 P-labeled probe with the disabled-mutant for CRE (Fig. 7, A  and C). The intensity of shifted band for CRE significantly increased with the nuclear extracts treated with thapsigargin for 1-2 h (Fig. 7A, left). In contrast, the intensity decreased with the nuclear extracts treated with BAPTA-AM for 1-2 h (Fig. 7A, right). In the case of the GC box, a major band appeared but it too disappeared in the presence of an excess of unlabeled probe (not shown) or 32 P-labeled probe with the disabled mutant for GC box (Fig. 7, B and C). However, the band intensity was not influenced by the treatment with thapsigargin or BAPTA-AM (Fig. 7). In the case of the GC box-like site at Ϫ10, no major band was observed in EMSA, suggesting that the site is nonfunctional (data not shown). In the case of the ROR␣ site at Ϫ553, a gel shift was observed but was not changed by the treatment with Ca 2ϩ modulators such as thapsigargin and BAPTA-AM (data not shown). Furthermore, there was no gel-shift observed with the probe for the ROR␣ site at Ϫ778 in EMSA, suggesting that the site is nonfunctional (data not shown). Together, these results indicate that the DNA binding activity involves both the CRE at Ϫ26 and GC box at Ϫ155, but it is specifically influenced by Ca 2ϩ modulators, such as thapsigargin and BAPTA-AM, especially in the CRE at Ϫ26.
The Gene Promoter Activity of PP2Ac␣ Is Regulated by Ca 2ϩ Modulators via CRE-To confirm the CRE-dependent regulation of the gene expression of PP2Ac␣ by Ca 2ϩ modulators, the gene promoter activity was examined by assaying the luciferase activity as describe above. The cells were transfected with luciferase vector construct 3 (Ϫ145 to ϩ258) (Fig. 6), which contains a CRE but no GC box. After 24 h of transfection, the cells were treated with 5 M thapsigargin or 10 M BAPTA-AM for predetermined periods, and the cell lysates were prepared and subjected to the assay for luciferase activity. As shown in Fig. 8A, the activity increased with thapsigargin to ϳ130% of the initial activity. The increase was not observed in the case of the disabled mutant for CRE with thapsigargin (data not shown). In contrast, with BAPTA-AM, the gene promoter activity decreased to ϳ85% of the initial level after 3 h of treat-ment, but returned to the initial level after 4 h (Fig. 8B). In the disabled mutant of CRE (Construct 3-Mut/C in Fig. 6), the promoter activity was lost and no activity was observed even on treatment with thapsigargin or BAPTA-AM (data not shown). Collectively, the results indicate that the gene expression of PP2Ac␣ is transcriptionally regulated by the change of intracellular Ca 2ϩ via CRE, and are consistent with the results of the EMSA for CREB binding (Fig. 6).
The Phosphorylation and Intracellular Localization of CREB Is Regulated by Ca 2ϩ Modulators in H9c2 Cells-CREB is a pivotal transcription factor for the regulation of cellular survival, and its activation is mediated by phosphorylation at a specific residue, Ser-133 (33). To investigate the effect of Ca 2ϩ modulators on the phosphorylation status of CREB, the level of CREB phosphorylated at Ser-133 was examined by immunoblot analysis using specific antibodies in the cells treated with thapsigargin or BAPTA-AM. As shown in Fig. 9A, the phosphorylation level of CREB increased on treatment with thapsigar-FIG. 7. Ca 2؉ modulators influence DNA binding to CRE in EMSA. H9c2 cells were incubated with thapsigargin (5 M) or BAPTA-AM (10 M) for the periods indicated, then the nuclear extracts were prepared as described under "Materials and Methods." 32 P-labeled oligonucleotides specific to CRE (Ϫ26) (A) and GC box (Ϫ155) (B) of the PP2Ac␣ gene promoter were prepared and incubated with each nuclear extract, and then subjected to a 6% nondenatured PAGE. In line F, nuclear extract is free. In line M, 32 P-labeled mutant oligonucleotides were used for the original ones. In line ϩAb, specific antibodies were added to the reaction mixture during the binding reaction for the supershift assay. C, quantitative data for the DNA binding to CRE and GC box shown in A and B. The intensity of gel-shift bands (arrows) was estimated densitometrically. Each value represents the mean of three experiments, and the S.D. was always within 10% of the mean. gin, but decreased with BAPTA-AM, indicating that the phosphorylation status is regulated by the change of [Ca 2ϩ ] i . In Fig.  9B, the intracellular localization of phosphorylated CREB (Ser-133) was investigated by indirect immunofluorescence microscopy in cells treated with or without the Ca 2ϩ modulators. The fluorescent signals for phosphorylated CREB increased especially in the nucleus of the cells treated with thapsigargin, but it was significantly decreased by BAPTA-AM compared with untreated cells. These results are consistent with the results that the gene of PP2Ac␣ is regulated via CRE in a Ca 2ϩ -dependent manner.

DISCUSSION
Akt is a Ser/Thr protein kinase with antiapoptotic and oncogenic activities (8 -10). With regard to the Ca 2ϩ -dependent regulation of Akt, Conus et al. (15) reported that the activation of Akt was independent of Ca 2ϩ in mouse fibroblasts treated with thapsigargin. However, Yano et al. (13) identified a Ca 2ϩtriggered signaling cascade in which CaMK kinase activates Akt in a PI 3-kinase-independent manner. Then Huber et al. (14) found that Akt was rapidly activated by treatment with thapsigargin through the activation of PI 3-kinase. In the present study, we found that Akt was suppressed by a long term elevation of [Ca 2ϩ ] i induced by thapsigargin, but was enhanced by a long term lowering of [Ca 2ϩ ] i caused by BAPTA-AM in H9c2 cells. The inactivation of Akt is highly correlated with susceptibility to apoptosis. Recently, Luo et al. (34) reported that inactivation of Akt is a causal mediator of cell death, and this is consistent with our present results. Although the underlying mechanism for these differences in the Ca 2ϩ -dependent regulation of Akt was not clear, we showed that transcriptional regulation of PP2Ac␣ was important to control the activation status of Akt in H9c2 cells under conditions where there is a long term change in [Ca 2ϩ ] i levels.
PP2A is a multifunctional protein Ser/Thr phosphatase that regulates a variety of signaling pathways in eukaryotic cells (35,36,37). The core structure is a dimer, consisting of a 36-kDa catalytic subunit (PP2Ac␣, ␤) and a 65-kDa constant regulatory (structural) subunit (PR65/A␣, ␤). A third, variable regulatory subunit (B, PR55/B␣, ␤, ␥, ␦; BЈ, PR56/61␣, ␤, ␥, ␦, ⑀; BЉ, PR48/58/72/130; Bٞ, PR99/110) can associate with this core enzyme. There are various reports of PP2A as a positive regulator of apoptosis (38), although a specific subunit of PP2A containing BЈ/PR61 is reported to be inhibitory for apoptosis in Drosophila (39,40). Bad is a pro-apoptotic member of the Bcl-2 family, whose function is highly regulated by reversible phosphorylation (41). PP2A was responsible for the dephosphorylation of Bad (42), and dephosphorylated Bad bound antiapoptotic Bcl-2 members at the mitochondrial membrane leading to apoptotic cell death (43). PP2A was also found to co-localize at the mitochondrial membrane with Bcl-2, and the proapoptotic sphingolipid ceramide has been shown to activate the PP2A involved (44,45). In anti-Fas-induced apoptosis, activation of caspase-3 caused cleavage of the regulatory A␣ subunit of PP2A, and this in turn increased PP2A activity (46). On the other hand, Liu et al. (47) reported that 4-hydroxynonenal induced dephosphorylation of Akt through activation of PP2A in a caspase-dependent apoptosis of Jurkat cells. In the study, the authors described that PP2A was activated by an altered intracellular localization of tyrosine-dephosphorylated PP2A, but not by the caspase-dependent cleavage of the regulatory A␣ subunit of PP2A. Furthermore, C2-ceramide induced dephosphorylation of both GSK3␤ and Akt by activating PP2A, resulting in apoptosis in rat cerebellar granule cells, and the apoptosis was blocked with lithium by inhibiting PP2A activity (48). Together, these findings indicate that PP2A plays a critical role in the positive regulation of apoptosis by dephosphorylating various apoptotic regulators including Akt, but the molecular mechanism for the activation of PP2A in apoptosis is not clearly understood.
PP2A is considered a phosphatase responsible for the dephosphorylation and inactivation of Akt (47,48,49,50,51,52,53). Previously, we also showed that Akt was dephosphorylated by PP2A in H9c2 cells (16), and PP2A interacted transiently with Akt in H9c2 cells under oxidative stress with H 2 O 2 (27). In the present study, the treatment with okadaic acid decreased PP2A activity to ϳ70% of the untreated control value, and it reversed the thapsigargin-dependent suppression of the phosphorylation of Akt in H9c2 cells (Fig. 4). Furthermore, treatment with okadaic acid could suppress the thapsigargin-induced enhancement of apoptosis in cells exposed to H 2 O 2 , although the effective okadaic acid concentration was limited to a range around 100 -200 nM (Fig. 5). These findings strongly suggest that PP2A plays an up-regulating role in the thapsigargin-induced enhancement of apoptosis by inhibiting Akt signaling. Collectively, the finding that up-regulation of PP2Ac␣ gene expression led to an increase of PP2A activity is consistent with the enhanced dephosphorylation and inactivation of Akt in H9c2 cells following the long term elevation of [Ca 2ϩ ] i due to thapsigargin exposure (Fig. 2). Although the expression of PP2Ac is tightly controlled by an autoregulatory translational mechanism (54), there are reports describing changes in PP2Ac levels, for instance, during all-trans retinoic acid-induced differentiation of HL-60 cells (55,56), during adipocyte differentiation induced by peroxisome proliferator-activated receptor-␥ (57), during stimulation by colony-stimulating factor in macrophages (58), and during a response to the disruption of cellular attachment in mouse C3 10T1/2 cells (59). We also observed transcriptional activation of PP2Ac␣ in H9c2 cells transfected with the expression vector for calreticulin, a molecular chaperone in the endoplasmic reticulum (16). However, the underlying mechanism for these differences in the regulation of PP2A expression was not clarified.
To investigate the molecular mechanism behind the Ca 2ϩdependent transcriptional regulation of PP2Ac expression in H9c2 cells, the promoter function of the PP2Ac␣ gene was characterized using a luciferase-based reporter assay in cells treated with Ca 2ϩ modulators such as thapsigargin and BAPTA-AM. The results showed that the expression of PP2Ac␣ was transcriptionally regulated by the change of [Ca 2ϩ ] i . The PP2Ac gene has been isolated and characterized in humans (60) and rats (23). In both species, the promoter region of PP2Ac␣ has a high GC content and does not contain either a TATA box or a CAAT box, which suggests that the gene is a typical housekeeping gene. Among various transcription factors including CREB, Sp1, and ROR␣ within the promoter region, CREB was revealed to be responsible for Ca 2ϩ -dependent regulation of the PP2Ac␣ gene in H9c2 cells treated with thapsigargin and BAPTA-AM. CREB is a bZIP transcription factor that forms homo-or heterodimers with itself or other members of the CREB family including ATF1 and CREM, and is a pivotal transcription factor that regulates cell proliferation, differentiation, and survival in a variety of cell types in vertebrates (33). The CREB dimers interact with a specific DNA sequence having the consensus motif TGACGTCA in the regulatory region of CREB target genes. CREB is inactive as a transcription factor until a cell is exposed to any extracellular stimuli that trigger its phosphorylation at a specific site, Ser-133, within its kinase-inducible domain (33). CREB was originally identified as a target of the cAMP signaling pathway, and regulated in response to diverse signals, including peptide hormones, growth factors, and Ca 2ϩ . CREB is known to be activated at high [Ca 2ϩ ] i and this is consistent with our findings that the level of CREB phosphorylated at Ser-133 increased with thapsigargin but decreased with BAPTA-AM in the nucleus of H9c2 cells. Though minute changes in [Ca 2ϩ ] i are quickly transformed into changes in the activity of several kinases including cAMP-dependent kinase, protein kinase C, MAPKs, Ca 2ϩ /CaMK and CaMK kinase, it is not clear whether these kinases are influenced in the case of long term change of [Ca 2ϩ ] i in H9c2 cells treated with Ca 2ϩ modulators. Among these kinases, CaMKII and CaMKIV were reported to be able to phosphorylate CREB directly (61,62). In addition, Rsk protein kinase phosphorylates CREB at Ser-133 through activation of the Ras/MAPK signaling pathway by Ca 2ϩ (63,64). Although CaMKIV mediates the early phase in the phosphorylation of Ser-133 in membrane-depolarized neurons, the MAPK pathway is responsible for prolonging the phosphorylation (65). In the present study, an increase in both the phosphorylation of CREB at Ser-133 and activity of CREB to bind the CRE site was observed after 2 h treatment with thapsigargin, suggesting a late activation of CREB caused by the long term elevation of [Ca 2ϩ ] i . The treatment with BAPTA-AM influenced the phosphorylation of CREB and suppressed the binding of CREB to the CRE after 1 h, and this also suggests a late inactivation of CREB on the long term suppression of [Ca 2ϩ ] i . Taken together, the long term change of [Ca 2ϩ ] i may regulate CREB function through the MAPK pathway rather than CaMK pathway in H9c2 cells treated with Ca 2ϩ modulators.
Thapsigargin causes an increase of [Ca 2ϩ ] i , and is also known as an inducer of endoplasmic reticulum (ER) stress (unfolding protein stress in the ER) (66). Recently, Song et al. (67) reported that thapsigargin-induced ER stress induced de-phosphorylation of both GSK3␤ and Akt by activating PP2A, resulting in apoptosis in neuroblastoma cells, and this was consistent with our results. In the study, the authors described that dephosphorylation of GSK3␤ by activated PP2A was critical for the activation of caspase-3 in ER stress-induced apoptosis, but the mechanism for the PP2A activation by ER stress was not clarified. In ER stress, a transcriptional up-regulation is seen in the ER stress responsive genes that code a variety of ER proteins related to molecular chaperone functions, such as BiP/Grp78, Grp94, Grp58/ERp57, ERp72, and calreticulin (66). In mammalian cells, the 19-nucleotide motif CCAAT-N9-CCACG was identified as an ER responsive element (ERSE) of various ER chaperone genes, and was recognized by the human bZIP transcriptional factor ATF6 for ER stress response (68). However, we could not find a consensus sequence within the 1.6-kbp promoter region of the PP2Ac␣ gene, and this suggests that the PP2Ac␣ gene is not a direct target for the ER stress response in thapsigargin-treated cells.
In this study, CREB was linked to the enhanced susceptibility to apoptosis through the induction of the PP2Ac gene in cells exposed to the long term elevation of [Ca 2ϩ ] i . To further investigate whether CREB is specifically responsible for the upregulation of apoptosis in cells treated with thapsigargin and H 2 O 2 , CREB expression was suppressed by the short interfering RNA (siRNA) for the CREB gene. Using mammalian CREB siRNA expression plasmid (pKD-CREB-v2, Upstate Biotechnology), the CREB expression level was suppressed in H9c2 cells to ϳ30% of non-transfected cells. Using the transfected cells, thapsigargin-dependent enhancement of cell damage and apoptosis were examined in cells treated with thapsigargin (5 M) and H 2 O 2 (75 M). However, despite the suppression of CREB protein, cell damage and apoptosis were not inhibited in cells treated with thapsigargin and H 2 O 2 , but rather were more enhanced (data not shown). This indicates that CREB is not specifically responsible for the mechanism enhancing apoptosis in cells showing long term elevation of [Ca 2ϩ ] i . However, this may be reasonable because CREB is well known as a transcription factor for cell survival and antiapoptotic genes such as Bcl-2 (69), and the suppression of CREB may firstly decrease the expression of such cell survival genes resulting in enhanced susceptibility to apoptosis. The effect of the suppressed expression of CREB on cell survival was also consistent with previous findings using dominant-negative CREB polypeptides (69). Although the suppressed expression of CREB itself did not specifically reduce the thapsigargin-dependent enhancement of apoptotic cell damage in H9c2 cells, it still may be possible that CREB works for apoptosis on some negative feedback-like loop of cell survival signaling in cells demonstrating long term elevation of [Ca 2ϩ ] i .
In conclusion, we found that the Akt kinase pathway was regulated by a long term change of [Ca 2ϩ ] i through transcriptional regulation of PP2Ac␣. With an elevation of [Ca 2ϩ ] i induced by thapsigargin, PP2Ac␣ gene expression is up-regulated through the activation of CRE at a late phase of the response. As a consequence, Akt is dephosphorylated and inactivated by PP2A, and this leads to an increase in susceptibility to apoptosis under the conditions with thapsigargin. Although the activation of CRE has been considered to function as a cell survival signaling, Ca 2ϩ -induced late activation of CRE leads to an enhancement of apoptotic signaling, and this suggests some feedback mechanism of CRE-mediating cell survival signaling. Luo et al. (34) reported that NMDA-induced deactivation of Akt was causative of neural cell death, and this suggests some Ca 2ϩ -dependent mechanism is involved in the inactivation of Akt. Although the Ca 2ϩ -dependent regulation of cell survival and death has been extensively studied (1, 3, 4), the present findings may indicate another novel regulatory pathway of Akt through PP2A in the cell survival signaling controlled by calcium homeostasis.