Akt as a mediator of cell death

Abstract

Protein kinase B/Akt possesses prosurvival and antiapoptotic activities and is involved in growth factor-mediated neuronal protection. In this study we establish Akt deactivation as a causal mediator of cell death. Akt deactivation occurs in multiple models of cell death including N-methyl-d-aspartate excitotoxicity, vascular stroke, and nitric oxide (NO)- and hydrogen peroxide (H₂O₂)-elicited death of HeLa, PC12, and Jurkat T cells. Akt deactivation characterizes both caspase-dependent and -independent cell death. Conditions rescuing cell death, such as treatment with poly(ADP-ribose) polymerase or NO synthase inhibitors and preconditioning with sublethal concentrations of N-methyl-d-aspartate, restore Akt activity. Infection of neurons with adenovirus expressing constitutively active Akt prevents excitotoxicity, whereas phosphatidylinositol 3-kinase inhibitors or infection with dominant negative Akt induce death of untreated neuronal cells.

Akt [protein kinase B (PKB)] is a serine/threonine protein kinase with oncogenic and antiapoptotic activities (1–3). Phosphatidylinositol 3-kinase (PI3-kinase) generates phosphatidylinositol (3,4,5)-trisphosphate [PtdIns(3,4,5)P3], which activates Akt (4–6). Akt, in turn, phosphorylates a variety of proteins, including several associated with cell death pathways such as BAD, Forkhead, CREB, MDM2, and NF-κB, leading to diminished apoptotic cell death (1, 6). Whether death-promoting stimuli lead to deactivation of Akt has not been established. In the present study we demonstrate deactivation of Akt in a variety of forms of cell death and establish a causal role for the deactivation process in N-methyl-d-aspartate (NMDA) neurotoxicity.

Materials and Methods

Neuronal Cell Cultures and Cytotoxicity. Primary cortical or hippocampal neuronal cultures were prepared as previously described (7). To induce excitotoxicity, the cells were prewashed with Tris-buffered control salt solution (CSS; 120 mM NaCl/5.4 mM KCl/1.8 mMCaCl₂/25 mM Tris·HCl, pH 7.4/15 mM glucose) and treated with CSS containing 300 μM NMDA for 5 min. Toxicity was assayed 20–24 h after NMDA exposure by microscopic examination with computer-assisted cell counting. Total and dead cells were determined by nuclei staining with 100 ng/ml 4′,6-diamidino-2-phenylindole (DAPI) and propidium iodide (PI) (10 μM), respectively. After a 10-min incubation, the cells were examined under a fluorescence microscope (Zeiss) with excitation at 360 nm. Cell death was determined as the ratio of dead to total cells and quantified by counting 1,000 cells. For staining of dead cells by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) assay, cells were fixed in 4% paraformaldehyde/PBS and then stained by using a TUNEL Assay Kit (Molecular Probes) following protocols provided by the manufacturer. The cell death inhibitory effect of various agents was examined essentially as described (7, 8). Western blotting was performed essentially as described (9).

Focal Cerebral Ischemia Model. C57BL/6 mice weighing 17–25 g were used for transient focal cerebral ischemia. After a midline neck incision, the left common carotid artery was isolated from the vagus nerve and ligated. The external carotid artery also was ligated, and the internal carotid artery was isolated carefully from the surrounding tissue. An 8-0 nylon filament (Ethicon, Somerville, NJ) was inserted into the common carotid artery through a small incision made in the proximity of the carotid bifurcation and advanced to the proximal part of the anterior cerebral artery to compromise the middle cerebral artery (MCA) flow. The filament was fixed in position by ligature. In sham-operated animals, the above procedures were performed except for the insertion of an intraluminal filament. For histological examinations, mice were perfused transcardially with heparinized PBS followed by 4% paraformaldehyde/PBS for tissue fixation. Brains were removed and postfixed in 4% paraformaldehyde/PBS at 4°C overnight. Coronal frozen sections (20 μm) were prepared on a cryostat and stored at –80°C until use. The frozen sections were thawed, washed three times in PBS, permeabilized with 0.1% Triton X-100/PBS at room temperature for 5 min, and then blocked in 5% skim milk/3% BSA/PBS for 60 min. Subsequently, they were incubated with primary antibodies (1:200) at 4°C overnight and with secondary antibodies at room temperature for 2 h, and immunoreactivity was visualized by the avidin–biotin complex (ABC) method.

Cell Lines and Cell Death Assays. HeLa cells, a human cervical carcinoma-derived cell line, were maintained in DMEM with 10% FBS, 2 mM l-glutamine, and 100 units of penicillin/streptomycin at 37°C with a 5% CO₂ atmosphere in a humidified incubator. PC12 cells, a rat pheochromocytoma cell line, were cultured on polylysine-coated plates in DMEM containing 10% FBS, 5% horse serum, and 2 mM l-glutamine. Cells were plated at 5 × 10⁶ cells per well in six-well plastic plates 1 day before the experiment, and cell death was induced by addition of NO donor S-nitroso-acetylpenicillamine (SNAP) or hydrogen peroxide (H₂O₂) (10). At indicated time points, cells were lysed with 0.5 ml of boiling protein loading buffer (Invitrogen), and 20-μl samples were analyzed by Western blotting. Jurkat T cells were maintained in RPMI medium 1640 supplemented with 2 mM l-glutamine and 10% FBS. In experiments that involved caspase-independent cell death, the broad-spectrum caspase inhibitor N-benzyloxycarbonyl-Val-Ala-Asp-f luoromethylketone (ZVAD.fmk) was added 2 h before UV-B treatment to eliminate caspase-dependent effects.

Generation of Recombinant Adenovirus and Infection of Neurons. Recombinant adenoviruses were generated by Cre-lox recombination using CRE8 cells that stably express Cre recombinase (11). The viruses were amplified and purified as previously described (12) and then aliquoted, frozen on dry ice, and stored at –80°C. The adenovirus expressing the constitutively active form of Akt (AdE-Akt c.a.) contained cDNA encoding hemagglutinin (HA)-tagged myristoylated wild-type Akt, whereas the one expressing the dominant negative form of Akt (AdE-Akt d.n.) contained cDNA encoding HA-tagged myristoylated mutant Akt (Ser-473/Ala and Thr-308/Ala double mutant). Infection usually was performed 48 h before the cell death assay. Primary neuronal cells were coinfected with AdVgRXR plus AdE-Akt c.a., AdE-Akt d.n., or AdE-GFP (9) in Neurobasal/B27 medium (Invitrogen) containing 2% FBS. Infection was carried out at 37°C for 24 h, and a multiplicity of infection (moi) value of 100 was used for all virus constructs. Protein expression was induced by addition of the ecdysone promoter ligand ponasterone A (3 μM), and further assays were performed 24 h after the induction.

Results

NMDA Excitotoxic Cell Death Selectively Deactivates Akt. High concentrations of glutamate induce neuronal cell death (excitotoxicity), a process that contributes to cell loss in brain ischemia, trauma, seizure, stroke, epilepsy, and hypoxia, as well as in certain neurodegenerative diseases (13, 14). Excitotoxicity triggers delayed cell death, which shares many features of apoptosis, such as cell body shrinkage, nuclear condensation, and DNA fragmentation (15). However, this type of cell death does not require the activation of caspases, which is the hallmark of classic apoptosis (8, 16). We examined activation of Akt, which was monitored by phospho-Akt levels, in hippocampal and cerebral cortical cultures subjected to NMDA-elicited neurotoxicity. Treatment of these hippocampal and cerebral cortical cultures elicited cell death as evidenced by morphologic alterations (Fig. 1 A and E, respectively), nuclear morphology, PI staining, and TUNEL assay (Fig. 1 B and C). In both hippocampal and cortical cultures (Fig. 1 D and F, respectively), phospho-Akt levels initially increased at 15 min after NMDA treatment, followed by a decline evident at 2 h, with negligible levels after 17 h. The initial increase of phospho-Akt might reflect physiologic NMDA receptor-mediated signal transduction. Glycogen synthase kinase (GSK)-3β is a substrate of activated Akt (17). Levels of phospho-GSK-3β declined in hippocampal cultures after NMDA treatment with a time course similar to that of the decline in levels of phospho-Akt. By contrast, total levels of Akt and GSK-3α, as well as GSK-3β, did not change. Although Akt can be a substrate of caspase-3 (18, 19), we detected only a very slight increase in cleaved Akt at late time points after NMDA treatment.

Fig. 1.

Akt/PKB kinase activity decreases during neuronal cell death. (A–D) Hippocampal neurons. (A) Phase-contrast photomicrograph of hippocampal neurons. Primary neuronal cultures (14 days) were treated with 300 μM NMDA for 5 min, and toxicity was assayed 20–24 h after NMDA exposure. (B) Cell viability was assessed by fluorescence microscopy after DAPI nuclear staining, PI staining, or the TUNEL assay. For the TUNEL assay, incorporated BrdUrd was detected by using Alexa Fluor 488-conjugated anti-BrdUrd monoclonal antibody. (C) Quantitative analysis of cell death. The percentage of cell death was determined at the indicated time points after NMDA treatment. At least three separate experiments were performed with a minimum of 1,000 neurons counted per data point. The results are the means of three independent experiments. Bars indicate mean ± SD. (D) Deactivation of Akt/PKB during cell death. Protein extracts collected at indicated time points were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. Phosphorylated Akt/PKB, the active form, was detected by using phospho-Akt (Ser-473) antibody. GSK-3 is one of the targets of Akt, and its phosphorylation was detected by using phospho-GSK-3β (Ser-9) antibody. (E and F) Cortical neurons. NMDA treatment (300 μM) causes cell death and Akt/PKB deactivation in primary cortical neuronal cultures. The experiments were performed by using the same procedures as in A and D, respectively. (G–I) Akt/PKB activity decreases during neuronal cell death in an in vivo mouse ischemia model. C57BL/6 mice were subjected to transient focal cerebral ischemia by middle cerebral artery occlusion (MCAO) for 1 h. Brain slices were prepared at the indicated time points. Amounts of total and phosphorylated Akt/PKB were detected by using anti-Akt and anti-phospho-Akt (Ser-473) antibodies, respectively. (G) Immunostaining of cortex. (H) Deactivation of Akt also was determined by Western blotting analysis. At the indicated time points after MCAO, brains were quickly removed from the skull and lysed. Aliquots containing 40 μg of protein were separated on gels and visualized as described above. (I) The decreases in phospho-Akt during NMDA-induced neuronal cell death are selective. Activation of PKCα/β 2, PKCδ, PKD, and p38 pathway was determined by Western blotting analysis using phospho-PKCα/β 2 (Thr-638), phospho-PKCδ (Ser-643), phospho-PKD (Ser-916), and phospho-p38 (Thr-180/Tyr-182) antibodies, respectively.

To ascertain whether alterations in Akt activation are relevant to glutamate-mediated neurotoxicity in intact organisms, we subjected mice to middle cerebral artery occlusion (MCAO), which elicits substantial cell death in the area of the occlusion via glutamate receptor-mediated events (13). In mice subjected to MCAO, we observed declines in phospho-Akt, but not total Akt, levels 24 and 48 h after the lesion, whether monitored by immunohistochemistry or Western blot analysis (Fig. 1 G and H).

The decreases in phospho-Akt are selective. We measured phospho-PKCα/β 2, phospho-PKCδ, phospho-protein kinase D (phospho-PKD), and phospho-p38 mitogen-activated protein kinase (MAPK) (Fig. 1I). In hippocampal cultures treated with NMDA, we detected no alteration in any of these phosphokinases 15 min to 28 h after NMDA treatment.

Akt Deactivation Characterizes Both Caspase-Dependent and -Independent Cell Death. Multiple subtypes of cell death have been delineated. The family of caspases is activated in certain forms of apoptotic cell death. Cleavage of pro-caspase-3 leads to its activation and is often used as an index of the role of caspases in cell death.

In NMDA-treated hippocampal cultures, we monitored levels of cleaved and uncleaved caspase-3 and caspase-7 (Fig. 2A). We detected no augmentation of cleavage, hence activation, of either caspase-3 or caspase-7. Poly(ADP-ribose) polymerase-1 (PARP) is a nuclear enzyme that is activated by DNA damage and participates in the DNA repair process. After massive cellular insults, excess activation of PARP depletes energy sources and is associated with necrotic cell death (20, 21). Caspase-dependent apoptotic cell death leads to cleavage of PARP by caspases, which presumably prevents energy depletion, preserving the energy required for apoptosis. We did not detect PARP cleavage after NMDA treatment. These findings are consistent with previous observations that caspase inhibitors do not block NMDA excitotoxic cell death (8, 16).

Fig. 2.

Akt/PKB deactivation occurs in both caspase-independent and -dependent cell death. (A) NMDA-induced neuronal cell death is caspase-independent. Primary hippocampal neurons were treated with 300 μM NMDA for 5 min. Protein extracts collected at the indicated time points were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. Active caspase-3 was detected by using anti-cleaved caspase-3 (Asp-175) antibody. Caspase-7 and PARP are targets of caspase-3, and their cleaved products were detected by using anti-cleaved caspase-7 (Asp-198) and anti-cleaved PARP (Asp-214) antibody, respectively. (B) Deactivation of Akt/PKB during caspase-independent and -dependent cell death in HeLa cells. Caspase-independent cell death was induced with 500 μM SNAP (NO donor), and caspase-dependent cell death was induced by 1 mM H₂O₂. Representative confocal images show cell death as determined by PI staining. Akt/PKB and caspase-3 activation was examined by Western blotting with the indicated antibodies. Caspase-3 protease was activated during H₂O₂-induced, but not NO-induced, cell death, as indicated by the appearance of the cleaved active form of caspase-3 (p17). (C) Deactivation of Akt/PKB during caspase-independent and -dependent cell death in PC12 cells. Caspase-independent cell death was induced with 500 μM SNAP, and caspase-dependent cell death was induced by 0.2 mM H₂O₂ (10). (D) Deactivation of Akt/PKB during caspase-independent and -dependent cell death in Jurkat T cells. Caspase-independent cell death was induced with 300 J/m² UV-B treatment (22), and caspase-dependent cell death was induced by 25 μM etoposide (23).

To examine the generality of deactivation of Akt in cell death and its relationship to caspases, we used HeLa (Fig. 2B), PC12 (Fig. 2C), and Jurkat T (Fig. 2D) cells. We treated HeLa and PC12 cells with the NO donor SNAP, which causes cell death independent of caspase, and with H₂O₂, which kills cells in a caspase-dependent fashion (10) (Fig. 2 B and C). Both SNAP and H₂O₂ elicited morphologic changes characteristic of cell death. SNAP led to deactivation of Akt with only minimal cleavage of Akt and no evident cleavage of pro-caspase-3. By contrast, H₂O₂ treatment of the two cell types resulted in substantial cleavage of both Akt and caspase-3. The decline of phospho-Akt after H₂O₂ treatment seems to be related to decreased levels of total Akt and does not seem to reflect a direct deactivation of the enzyme.

In Jurkat T cells, we used UV-B irradiation and etoposide to induce caspase-independent (22) and -dependent (23) cytotoxicity, respectively (Fig. 2D). UV-B treatment led to deactivation of Akt with no change in total Akt levels and no evident cleavage of Akt or pro-caspase-3. By contrast, etoposide elicited cleavage of both Akt and caspase, with a decline of phospho-Akt coinciding with diminished levels of total Akt.

In summary, in multiple cellular systems we observed a selective association of Akt deactivation with forms of cell death that are dependent or independent of caspases.

PI3-Kinase Inhibitors Deactivate Akt and Elicit Cell Death in Brain Cultures. PI3-kinase is upstream of Akt, because its product PtdIns(3,4,5)P3 stimulates Akt activation (4–6). Inhibition of PI3-kinase is associated with deactivation of Akt and cell death in several models (2, 5). To determine whether brain cultures used in our study are similarly influenced by PI3-kinase, we treated hippocampal cultures with the PI3-kinase inhibitors LY294002 and wortmannin (Fig. 3). Treatment with these drugs markedly deactivates Akt without altering total Akt levels (Fig. 3A). The two drugs produce evidence of cell death whether monitored by morphology (Fig. 3B), nuclear shape, PI staining, or TUNEL assay (Fig. 3 C and D).

Fig. 3.

Deactivation of Akt/PKB by PI3-kinase inhibitors causes neuronal cell death. (A) Deactivation of Akt/PKB by PI3-kinase inhibitors. Primary hippocampal neurons were treated with 20 μM LY294002 or 50 nM wortmannin for 30 h, and Akt/PKB activity was assessed by Western blotting analysis. Both inhibitors markedly decreased Akt/PKB activity. (B) PI3-kinase inhibitor-induced neuronal cell death was examined by phase-contrast photomicrography. (C) Cell viability was assessed by fluorescence microscopy after DAPI staining, PI staining, or the TUNEL assay. Alexa Fluor 488-conjugated anti-BrdUrd monoclonal antibody was used for the TUNEL assay. (D) Quantitative analysis of PI3-kinase inhibitor-induced cell death. Both PI staining and TUNEL assay were performed, and at least three separate experiments were carried out with a minimum of 1,000 neurons counted per data point. The results are the means of three independent experiments. Bars indicate mean ± SD.

Treatments That Diminish Cell Death Prevent Akt Deactivation. If Akt deactivation mediates cell death, then treatments that prevent cell death should inhibit the deactivation. NMDA neurotoxicity is associated with excess production of NO and pronounced activation of PARP (24, 25). Agents that inhibit NO synthase and PARP block NMDA neurotoxicity (7). We have confirmed that the PARP inhibitors 3,4-dihydro-5-[4-(1-piperidinyl)butoxyl]-1(2H)-isoquinolinone (DPQ) and PJ34 prevent NMDA toxicity, as do the NO synthase inhibitors l-nitroarginine and l-nitroarginine methylester (l-NAME) (Fig. 4A). The PARP and NO synthase inhibitors also prevent deactivation of Akt with no alterations in total Akt levels (Fig. 4B).

Fig. 4.

Drugs or conditions preventing neuronal cell death restore Akt/PKB kinase activity. (A) Both PARP and NO synthase inhibitors rescue neurons from excitotoxicity-induced cell death. Primary hippocampal neurons were challenged with 300 μM NMDA as described above. The NO synthase inhibitors N-Arg (100 μM) or l-NAME (500 μM) were included in the NMDA challenge solution. The PARP inhibitors DPQ (30 μM) or PJ34 (0.5 μM) were added to the Neurobasal/B27 medium after NMDA treatment was initiated. (B) Both PARP and NO synthase inhibitors restored Akt/PKB activity. Levels of total Akt and phospho-Akt were measured by Western blotting analysis. (C and D) Preconditioning is neuroprotective and restores Akt activity. To generate the preconditioning effect, neuronal cultures were pretreated with 20 μM NMDA for 10 min, and excitotoxicity was elicited with 300 μM NMDA 24 h later.

Whereas high concentrations of NMDA (300 μM) are neurotoxic to hippocampal cultures, preconditioning with low concentrations of NMDA (20 μM) prevents the neurotoxic actions of 300 μM NMDA subsequently administered (26). We have confirmed the cytoprotective effects of preconditioning with 20 μM NMDA in hippocampal cultures (Fig. 4C). This preconditioning diminishes Akt deactivation by NMDA (300 μM) without altering total Akt levels (Fig. 4D).

Down-Regulation of Endogenous Akt Augments, and Overexpression of Active Akt Prevents, NMDA Neurotoxicity. If Akt deactivation mediates NMDA neurotoxicity, then overexpression of active Akt should diminish such toxicity. We infected hippocampal cultures with an adenovirus containing an inducible promoter of a constitutively active form of Akt (Fig. 5 A and B). Infection with the constitutively active form of Akt prevented NMDA cell death whether monitored by cellular morphology, nuclear morphology, TUNEL assay, or PI staining (Fig. 5 C–E). These findings are consistent with the reported ability of Akt transfection to prevent cell death in a variety of model systems (1, 2).

Fig. 5.

Dominant negative Akt triggers neuronal cell death, and constitutively active Akt rescues cells from NMDA-induced cell death. (A) Schematic representation of the ecdysone (Ecd)-inducible adenovirus constructs. The receptor virus (AdVgRXR) expresses two hormone receptors (VgEcR and RXR) that can activate the Ecd promoter in the presence of ponasterone A. Akt/PKB viruses (AdE-Akt) contained Akt cDNA under the control of the Ecd promoter. The constitutively active form of Akt (Akt c.a.) contained the cDNA encoding HA-tagged myristoylated wild-type Akt, whereas the dominant negative form of Akt (Akt d.n.) contained the cDNA encoding HA-tagged myristoylated mutant Akt (Ser-473/Ala and Thr-308/Ala double mutant). Ecd, ecdysone; RSV, Rous sarcoma virus; VgEcR, modified ecdysone receptor; RXR, retinoid X receptor. (B) Ponasterone A induced expression of recombinant Akt. Primary neuronal cells were coinfected with AdVgRXR plus AdE-Akt c.a., AdE-Akt d.n (12), or AdE-GFP (9). A multiplicity of infection (moi) value of 100 was used for all virus constructs. Protein expression was induced by addition of ponasterone A (3 μM). Expression of recombinant proteins was examined 24 h after induction by Western blotting analysis using anti-HA antibody. (C) Primary hippocampal neurons were infected with or without the indicated adenovirus. Ponasterone A was added 24 h after infection, and cells were challenged with 300 μM NMDA after another 24 h. Akt c.a. and Akt d.n. were detected by immunocytochemistry using fluorescein anti-HA antibody conjugate. Cell viability was assessed 24 h after NMDA treatment by DAPI staining, PI staining, or the TUNEL assay. Rhodamine-conjugated anti-BrdUrd monoclonal antibody was used for the TUNEL assay. (D and E) Quantitative analysis of cell viability by TUNEL assay (D) and PI staining (E). The results are the means of three independent experiments. Bars indicate mean ± SD. *, P < 0.001 versus GFP-virus infected neurons by Student's t test.

To provide a direct test of the hypothesis that deactivation of Akt plays a causal role in NMDA-mediated cell death, we sought to down-regulate endogenous Akt, using a dominant negative construct. Infection of hippocampal cultures with dominant negative Akt substantially augmented cell death of untreated cultures or cultures treated only with the GFP vector as monitored by cellular morphology, nuclear staining, TUNEL assay, or PI staining (Fig. 5 C–E). Previous studies have established that cell death of hippocampal and cortical cultures by treatment with NMDA reflects endogenous NMDA glutamate neurotransmission acting through NO and is prevented by agents that block NMDA transmission (13) and by inhibitors of NO synthase (7). The dominant negative Akt construct does not cause additional cell death compared with GFP vector-treated cells, although it significantly augments cell death compared with untreated cultures. Evidently, the profound cytotoxicity elicited by NMDA is so great that it is difficult to produce further enhancement.

Discussion

In the present study we have established a causal and general role for deactivation of Akt in eliciting cell death. Akt is deactivated in multiple types of cell death, including NMDA neurotoxicity and treatment of HeLa, PC12, and Jurkat T cells with diverse stimuli including NO donors and UV-B irradiation. Treatments that prevent NMDA toxicity, such as preconditioning with NMDA or PARP or NO synthase inhibitors, also prevent the deactivation of Akt. Compelling evidence for a causal link between cell death and Akt activation comes from experiments in which down-regulation of endogenous Akt enhances basal cell death.

We found Akt deactivation selectively associated with both caspase-independent and -dependent cell death in multiple cellular systems. What mechanisms might account for the cytotoxic actions of Akt deactivation in a caspase-independent fashion? The cytoprotective influences of Akt reflect phosphorylation of various proteins involved in cell death processes, such as BAD, Forkhead, CREB, MDM2, and NF-κB. Our findings suggest that, under basal conditions, in the absence of cytotoxic cell stressors, Akt exerts a restraining effect on cytotoxic processes. Stressors deactivate Akt, terminating this protective effect.

What might be the upstream deactivators of Akt? The best established activator of Akt is PtdIns(3,4,5)P3, generated by PI3-kinase (4–6). Conceivably, deactivation of PI3-kinase is responsible for Akt's deactivation. The PtdIns(3,4,5)P3 level can also be regulated by the tumor suppressor PTEN, a phosphatidylinositol 3′-phosphatase that converts PtdIns(3,4,5)P3 to PtdIns(4,5)P2 (27, 28). Cell stress might activate PTEN, leading to down-regulation of Akt. Phosphorylation of Akt can be elicited by PI3-kinase-independent pathways, such as those mediated by dopamine (29) or β-adrenoreceptors (30), which could down-regulate Akt. Deactivation of Akt might also reflect activation of protein phosphatase-2A, which can dephosphorylate Akt directly (31, 32). Calcium is released in multiple forms of cell death (33, 34), and its entry through NMDA ion channels is the major cause of NMDA excitotoxicity (13, 14, 26). How an elevated intracellular calcium level deactivates Akt remains to be elucidated.