Tumor Necrosis Factor Induces Apoptosis in Hepatoma Cells by Increasing Ca2+ Release from the Endoplasmic Reticulum and Suppressing Bcl-2 Expression*

Tumor necrosis factor (TNF) plays an import role in the control of apoptosis. The most well known apoptotic pathway regulated by TNF involves the TNFR1-associated death domain protein, Fas-associated death domain protein, and caspase-8. This study examines the mechanism of TNF-induced apoptosis in FaO rat hepatoma cells. TNF treatment significantly increased the percentage of apoptotic cells. TNF did not activate caspase-8 but activated caspase-3, -10, and -12. The effect of TNF on the expression of different members of the Bcl-2 family in these cells was studied. We observed no detectable changes in the steady-state levels of Bcl-XL, Bax, and Bid, although TNF suppresses Bcl-2 expression. Dantrolene suppressed the inhibitory effect of TNF on Bcl-2 expression. TNF induced release of Ca2+ from the endoplasmic reticulum (ER) that was blocked by dantrolene. Importantly, the expression of Bcl-2 blocked TNF-induced apoptosis and decreased TNF-induced Ca2+ release. These results suggest that TNF induces apoptosis by a mechanism that involves increasing Ca2+ release from the ER and suppression of Bcl-2 expression.

Tumor necrosis factor (TNF) 1 elicits a wide range of biological responses including inflammation, infection, injury, and apoptosis (1,2). TNF exerts its pleiotropic function through two distinct receptors, TNFR1 and TNFR2 (3). Both receptors contain several cysteine repeats in their extracellular domains, whereas their intracellular domain contains no significant homology. Both TNFR1 and TNFR2 can regulate TNF-mediated responses, but apoptosis is mainly induced through TNFR1, which has a cytoplasmic death domain that interacts with the adaptor protein TNFR1-associated death domain protein (TRADD) following ligand binding (4). TRADD serves as a platform to recruit at least three additional mediators, receptor-interacting protein 1, Fas-associated death domain protein (FADD) and TNF receptor-associated factor 2 (TRAF2) (3,(5)(6)(7)(8). Caspase-8 is next recruited to this protein complex, and the active caspase-8 initiates a caspase cascade, which results in apoptosis (9). TNFR2 lacks a death domain but synergistically enhances TNFR1-induced apoptosis. Recently, it has been shown in human T cells that TNFR2-mediated apoptosis requires RIP, which is induced during T-cell activation and promotes a change in TNFR2 signaling from the nuclear factor-B (NF-B) activation pathway to apoptosis (10).
The endoplasmic reticulum (ER) regulates protein synthesis, protein folding and trafficking, cellular responses to stress, and intracellular calcium (Ca 2ϩ ) levels (11). Intracellular Ca 2ϩ homeostasis is very important in maintaining the normal function of the cell. Alterations in intracellular Ca 2ϩ homeostasis are implicated in the control of apoptosis. Endogenous endonuclease activation proceeds via a Ca 2ϩ -dependent mechanism in thymocytes and certain other cell types exposed to a wide range of stimuli (12). Bcl-2 has been demonstrated to act as an ion channel in isolated lipid bilayers (13,14) and displays a complex distribution in cells (15). Bcl-2 is localized to the mitochondrial membrane as well as to the ER and the nuclear membrane (16). The co-localization of Bcl-2 with Ca 2ϩ pumps and channels on ER and the nuclear membrane has raised the possibility of a role for Bcl-2 in the maintenance of Ca 2ϩ homeostasis in these compartments. The exact mechanism by which Bcl-2 regulates ER Ca 2ϩ is not yet understood. Recent data suggest that Bcl-2 decreases the flux of Ca 2ϩ from a mobilizable pool located within the ER lumen, thereby abrogating Ca 2ϩ signaling of apoptosis (12,(17)(18)(19)(20)(21). Overexpression of Bcl-2 reduces the calcium concentration in the endoplasmic reticulum. However, overexpression of Bcl-2 has been shown to modulate the ER store of calcium by up-regulating calcium pump (SERCA) expression without affecting the release channel (IP3R) (22). Other studies have shown that Bcl-2 regulates mitochondrial Ca 2ϩ homeostasis (23,24) and prevents Ca 2ϩinduced cytochrome c release (25).
FaO rat hepatoma cells are very sensitive to various apoptotic stimuli, including TGF-␤ (26 -30). Cleavage of full-length BAD is rapidly induced during TGF-␤1-induced apoptosis, but steady-state levels of Bcl-2, Bcl-X L , and Bax are not changed (30). In this study, we have examined the effect of TNF on apoptosis in FaO cells. We showed that TNF treatment induced apoptosis and activated caspase-3, caspase-10, and caspase-12. TNF specifically suppressed Bcl-2 expression and induced a rise in cytoplasmic Ca 2ϩ concentration by releasing Ca 2ϩ from intracellular stores. Altogether the present data suggest that TNF induces apoptosis via mobilization of intracellular Ca 2ϩ and suppression of Bcl-2.

MATERIALS AND METHODS
Cell Culture-FaO rat hepatoma cells were maintained at 37°C in DMEM (Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum, penicillin, and streptomycin (100 g/ml)). Generation of FaO cells stably expressing Bcl-2 was previously described (28).
Reagents-Propidium iodide, SKF96365, TMB-8, and dantrolene, a blocker of intracellular calcium release from the sarcoplasmic reticulum that acts by binding to the ryanodine receptor, were purchased from Calbiochem. EGTA and thapsigargin were purchased from Sigma. Recombinant murine TNF was purchased from R & D Systems. All other chemicals were from standard sources and were molecular grade or higher. Caspase-8 inhibitor, z-Ile-Glu(OMe)-Thr-Asp(OMe)-CH 2 F (IETD-fmk), was purchased from Calbiochem and was reconstituted in Me 2 SO to a concentration of 50 mM. It was then added to medium at a concentration of 50 M 30 min prior to treatment with either TNF or TGF-␤1.
DNA Fragmentation Assay-FaO cells were treated with lysis buffer (10 mM Tris-Cl, pH 7.4, 10 mM NaCl, 10 mM EDTA, and 0.5% SDS, and 0.1 mg/ml proteinase K) and were incubated at 50°C for 2 h. The lysate was extracted with phenol, phenol/chloroform (1:1), and chloroform and precipitated with 2.5 volumes of ice-cold ethanol. The DNA was resuspended in Tris-EDTA buffer supplemented with 100 g/ml RNase A. DNA samples were electrophoretically separated on 2% agarose gel for 2 h at 50 V.
Deoxynucleotidyl Transferase-mediated dUTP Nick End Labeling (TUNEL) Assay-FaO cells were plated at 5 ϫ 10 4 cells/eight-well chamber slide (Nalge Nunc International, Rochester, NY) and incubated for 24 h. The cells were treated with TGF-␤1 (5 ng/ml) or TNF (20 ng/ml) for 12 h and fixed with 4% paraformaldehyde (pH 7.4) for 10 min. Apoptotic cells were assessed by measuring DNA fragmentation in a standard TUNEL assay according to the instructions with the kit (In Situ Cell Death Detection Kit, POD; Roche Molecular Biochemicals).
Flow Cytometry Analysis-For the flow cytometry assay, FaO and FaO-Bcl-2 cells were grown in six-well plates and incubated for 24 h at 37°C in the presence or absence of 20 ng/ml TNF-␣. Cells were harvested and washed twice with phosphate-buffered saline (pH 7.4). After A, cellular DNA fragmentation after TNF treatment (20 ng/ml). Genomic DNA was extracted from control, TGF-␤1 (5 ng/ml) (12 h), and TNF-treated cells at 6, 12, and 24 h. Extracted DNA was resolved on 1.5% agarose gel. M, size markers. B, the TUNEL procedure was carried out, and pictures were taken under a light microscope (magnification, ϫ200). C, expression of Bcl-2 family proteins during TNF-induced apoptosis in FaO rat hepatoma cells. The effect of the TNF on Bcl-X L , Bcl-2, Bax, and Bid protein levels in FaO cells were determined. Cells were incubated with 20 ng/ml TNF. Cell lysates were made, and equivalent amounts of cellular proteins were separated by SDS-15% PAGE, blotted, and then probed with the appropriate antibody. Blots shown are typical of at least three individual experiments.

FIG. 2. Activation of caspases after TNF and TGF-␤1 treatment.
A and B, samples were taken at the indicated time points and assayed for caspase-3 and caspase-8 cleavage by Western blotting using caspase-3-and caspase-8-specific antibodies. The full-length caspase protein and cleaved products (sizes indicated in kDa) are indicated. C, examination of the effect of caspase-8 inhibitor on TNFinduced apoptosis in FaO cells by TUNEL apoptotic assay. Using the nucleus-specific staining 4Ј,6-diamidino-2-phenylindole and TUNEL assay, an increase in the rate of nuclear fragmentation and apoptosis was observed in FaO cells following treatment with either TNF (20 ng/ml) or TGF-␤1 (5 ng/ml). Pretreatment with caspase-8 inhibitor, IETD-fmk, partially blocked TGF-␤1-mediated apoptosis but showed very little effect on TNF-mediated apoptosis. D, TUNEL-positive apoptotic cells at the respective incubation time were counted, and the percentage of apoptotic cells was graphed. Similar results were achieved in three separate experiments with comparable outcomes. fixing in 80% ethanol for 30 min, cells were washed twice and resuspended in phosphate-buffered saline (pH 7.4) containing 0.1% Triton X-100, 5 g/ml propidium iodide (PI) and 50 g/ml ribonuclease A for DNA staining. Cells were then analyzed by a FACScan cytometer (CELLQUEST program; Becton Dickinson). Red fluorescence due to PI staining of DNA was expressed on a logarithmic scale simultaneously with the forward scatter of the particles. Four thousand events were counted on the scatter gate. The number of apoptotic nuclei was expressed as a percentage of the total number of events.

FIG. 3. Activation of caspase-10 and caspase-12 after TNF and TGF-␤1 treatment.
A, immunoblot for caspase-10. FaO cells were treated with either TGF-␤1 (5 ng/ml) or TNF (20 ng/ml) for 24 h. B, samples were taken at the indicated time points after TNF treatment and assayed by Western blotting analysis using a caspase-12-specific antibody that recognizes full-length as well as the active form. ␤-Actin was used as a control for protein loading. C, immunoblot for caspase-12. FaO cells were treated with either TGF-␤1 (5 ng/ml) or TNF (20 ng/ ml) for 24 h.

FIG. 4. TNF-induced apoptosis is dependent on [Ca 2؉ ] i in FaO rat hepatoma cells.
A, EGTA (2 mM) or SKF96365 (1 M) were added 1 h before TNF application. TUNEL procedure was carried out, and pictures were taken under a light microscope (magnification, ϫ200). Representative pictures demonstrate the effect of either EGTA or SKF96365 on TNF-induced apoptosis. B, TUNEL-positive apoptotic cells at the respective incubation times were counted, and the percentage of apoptotic cells was graphed. Similar results were achieved in three separate experiments with comparable outcomes. C, dantrolene (50 M) and TMB-8 (20 M) were added 1 h before TNF treatment. The changes in cellular morphologies were observed by phase-contrast microscopy (magnification, ϫ200). DMSO, Me 2 SO.
[Ca 2ϩ ] i Measurements-Fura2 fluorescence in single cells was measured as described earlier (31) by using an SLM 8000/DMX 100 spectrofluorimeter attached to an inverted Nikon Diaphot microscope with a Fluor ϫ40 oil immersion objective. Images were acquired using an enhanced CCD camera (CCD-72; MTI) and Image-1 software (Universal Imaging Corp., Downingtown, PA). Analog plots of the fluorescence ratio (340/380) in single cells are shown.

TNF Induces Apoptosis in FaO Rat
Hepatoma Cells-Treatment of FaO rat hepatoma cells with TNF induced a 180 -200-bp internucleosomal DNA cleavage as early as 12 h after TNF treatment (Fig. 1A). We have previously shown that TGF-␤1 rapidly induces apoptosis in these cells (27,28). Therefore, we used TGF-␤1 treatment as a positive control for the apoptotic response. To show that the TNF-induced cell death is due to apoptosis, the TUNEL assay was also carried out (Fig.  1B). As indicated by the number of dark brown positive cells, there was a significant increase in the rate of apoptosis in a time-dependent manner following TNF treatment.
To characterize the mechanism of TNF-mediated apoptosis, the steady-state levels of several Bcl-2 family proteins were measured by Western blotting analysis. We detected no changes in expression of Bcl-X L , Bax, or Bid over the TNF treatment time course; nor were any cleaved products of these proteins detected (Fig. 1C). However, Bcl-2 expression decreased in a time-dependent manner (Fig. 1C).
Caspases in TNF-mediated Apoptosis-Since caspase-8 and caspase-3 have been implicated in TNF-induced apoptosis, we examined activation of caspase-8 and caspase-3 in TNF-induced apoptosis in FaO cells. TNF treatment of FaO cells induced a time-dependent processing of caspase-3 but not caspase-8 ( Fig. 2A). We also used TGF-␤1 as a control for the caspase activation. As expected, TGF-␤1 treatment caused activation of both caspase-3 and caspase-8 (Fig. 2B). To confirm this observation, we examined the effect of inhibitor of caspase-8 on TNF-mediated apoptosis. The addition of an inhibitor of caspase-8 (IETD-fmk) reduced the percentage of TUNEL-positive nuclei induced by TGF-␤1, but the presence of IETD-fmk showed very little effect on TNF-mediated apoptosis (Fig. 2, C and D). We next examined activation of caspase-10 in TNF-induced apoptosis in FaO cells. It is known that in some cases caspase-10 can functionally substitute for caspase-8 in death receptor signal transduction (32). The anti-caspase-10 antibody from Cell Signaling Technology does not detect the cleaved form of caspase-10. Treatment with TNF decreased the level of procaspase-10, whereas TGF-␤1 treatment showed no effect on the level of procaspase-10, suggesting that TNF may induce apoptosis through the activation of caspase-10 (Fig. 3A). Caspase-12, an ER-resident caspase, is specifically involved in apoptosis that results from ER stress (33)(34)(35)(36). Caspase-12 participates in ER stress-induced apoptosis that is blocked by benzyloxycarbonyl-Val-Ala-Asp-fluoromethyl ketone (z-VADfmk), a general caspase inhibitor (34). Alterations in Ca 2ϩ homeostasis and accumulation of unfolded proteins in the ER cause ER stress. Using a polyclonal serum directed against caspase-12, a single band corresponding to the p53 proform of caspase-12 was detected. After treatment with TNF, smaller fragments between 17 and 40 kDa became visible representing active caspase-12 (Fig. 3, B and C). However, cleavage of procaspase-12 was not detected during TGF-␤1-induced apoptosis (Fig. 3C). These results suggest that the TNF-induced cell death is apoptosis and that the actual apoptotic pathway involves activation of caspase-3 and caspase-12.
Effects of Ca 2ϩ on TNF-induced Apoptosis-Bcl-2 is known to regulate the flux of Ca 2ϩ across the ER membrane, thereby abrogating Ca 2ϩ signaling of apoptosis (20). To determine whether an increase in cytosolic Ca 2ϩ was required for the induction of apoptosis by TNF, the action of extracellular calcium chelators, EGTA and SKF96365, was studied. In a Ca 2ϩcontaining medium, TNF induced apoptosis in FaO cells. The low extracellular Ca 2ϩ caused by either EGTA or SKF96365 did not alter TNF-induced apoptosis (Fig. 4, A and B). Dantrolene, a drug used to treat malignant hyperthermia, inhibits Ca 2ϩ release from ER and has been shown to inhibit cell damage induced by thapsigargin-induced ER Ca 2ϩ depletion (37,38). To assess the effects of dantrolene and TMB-8, which blocks Ca 2ϩ release from ER, against TNF-induced apoptosis, we treated the FaO cells with TMB-8 and dantrolene. TNFinduced apoptosis was almost completely blocked by TMB-8 and dantrolene, whereas TGF-␤1-induced apoptosis was unaffected by these compounds (Fig. 4C).
Effects of TNF on Cellular Ca 2ϩ Homeostasis-To support the view that TNF-induced apoptosis is triggered by the release of Ca 2ϩ from intracellular Ca 2ϩ stores, we investigated whether TNF had a direct effect on intracellular Ca 2ϩ homeostasis. FaO cells were loaded with the Ca 2ϩ dye, and changes in [Ca 2ϩ ] i (expressed as 340/380-nm fluorescence ratio) were measured. Treatment with TNF caused a relatively fast elevation in [Ca 2ϩ ] i , which peaked within 5 min and then decreased to a lower sustained level. In a Ca 2ϩ -free medium, a smaller, transient increase was seen (Fig. 5, A, C, and H). However, TGF-␤1 has no effect on intracellular Ca 2ϩ homeostasis (Fig.  5G). This finding suggests that TNF induces both internal Ca 2ϩ release and Ca 2ϩ influx. The trace in Fig. 6B represents the internal Ca 2ϩ release component. The increases in cytosolic Ca 2ϩ elevation after TNF treatment in both Ca 2ϩ -containing and Ca 2ϩ -free conditions, were significantly reduced in cells pretreated with dantrolene (Fig. 5, B, D, and H). This suggests that dantrolene reduces TNF-induced internal Ca 2ϩ release. Dantrolene also inhibited internal Ca 2ϩ release induced by a thapsigargin, a highly specific inhibitor of the ER-associated Ca 2ϩ pump that is used to deplete internal Ca 2ϩ stores (Fig. 5,  E, F, and I).
We next determined the effect of Bcl-2 overexpression on TNF-induced calcium homeostasis. We previously generated FaO cell lines expressing Bcl-2 (28). These cell lines are individually isolated clones. In both Ca 2ϩ -containing and Ca 2ϩ -free medium, the increase of cytosolic Ca 2ϩ after the addition of TNF was significantly reduced in FaO/Bcl-2 cells compared with FaO cells (Fig. 6).
Dantrolene Increases Bcl-2 Protein Level in FaO Cells-The skeletal muscle relaxant dantrolene is an inhibitor of Ca 2ϩ release through rynodine receptor. The expression of rynodine receptor protein was confirmed in FaO cells by Western blot analysis using a specific antibody raised against the ryanodine receptor (Fig. 7A). We next examined the effect of dantrolene on Bcl-2 protein levels by Western blotting. As shown in Fig. 7, B and C, Bcl-2 protein levels increased significantly to 320% of the control after treatment with 25 M dantrolene for 24 h. TNF treatment markedly decreased Bcl-2 protein level, but treatment with dantrolene blocked suppression of Bcl-2 expression by TNF, suggesting that TNF suppresses Bcl-2 expression by regulating the flux of Ca 2ϩ .
Dantrolene Blocks TNF-induced Apoptosis-To investigate whether dantrolene also blocks TNF-induced apoptosis in FaO cells, we measured nuclear incorporation of propidium iodide by fluorescence-activated cell sorting analysis. Cells were incubated in the absence or presence of TNF. TGF-␤1 or thapsigargin was used as control. Dantrolene pretreatment robustly suppressed TNF-induced-or thapsigargin-induced population of apoptotic nuclei, whereas dantrolene did not block TGF-␤1induced apoptosis (Fig. 8). These results further suggest that TNF activates intracellular Ca 2ϩ release from ER, which, in turn, induces apoptosis.
Overexpression of Bcl-2 has been shown to repress apoptosis by regulating ER-associated Ca 2ϩ release. To investigate the protective effect of Bcl-2 on TNF-induced apoptosis, stable cell lines expressing Bcl-2 were generated (28) (Fig. 9A). As shown in Fig. 9A, Bcl-2 protein level was significantly increased. Most of the Bcl-2-overexpressing cell lines were completely resistant to TGF-␤1-induced apoptosis. We next examined the propidium iodide incorporation in the FaO cells overexpressing Bcl-2. A   FIG. 8. Intracellular Ca 2؉ release is responsible for TNF-induced apoptosis in FaO cells. Dantrolene (50 M) was added 1 h before TNF and thapsigargin application. A representative illustration is shown of PI incorporation measured in control and TNF-or thapsigargin-stimulated conditions by flow cytometry analysis. The number of apoptotic nuclei is expressed as a percentage of the total number of events. significant decrease in PI incorporation was seen after the addition of TNF in FaO/Bcl-2 cells compared with control cells (Fig. 9B).

Effects of Calcium Chelators and an Inhibitor of Intracellular Calcium Release on Activation of Caspase-3 and Caspase-12
Induced by TNF-Up to this point, our results suggest that ER-controlled calcium release is involved in TNF-induced apoptosis in FaO cells. The activation of caspase-12 is involved in a specific form of apoptosis in the ER unfolded protein response. TNF treatment increased cleaved active products of caspase-12. However, pretreatment with dantrolene or TMB-8 completely blocked activation of caspase-12 induced by TNF, whereas extracellular calcium chelators, EGTA or SKF96365, did not block the TNF-induced caspase-12 cleavage (Fig. 10). DISCUSSION Receptor-mediated apoptosis has been demonstrated for various growth factors, including TNF. The TNFs act via a large family of receptors expressed on the surface of the target cell, known as the TNF receptor superfamily (1-3). TNF is responsible for a diverse range of signaling events within cells, leading to either necrosis or apoptosis. TNF exerts many of its effects by binding to either a 55-kDa cell membrane receptor termed TNFR-1 or a 75-kDa cell membrane receptor termed TNFR-2. TNF induces apoptosis by more than one pathway. The most widely accepted pathway involves TRADD, FADD, and caspase-8 (39). Indeed, FADD-and caspase-8-deficient fibroblasts are resistant to TNFR1-induced apoptosis (40,41). Ligation of TNFR1 by TNF recruits the TRADD to the receptor's death domain. TRADD in turn recruits FADD, which recruits procaspase-8. TRADD also serves to recruit the serine/ threonine kinase RIP and the TNF receptor-associated factor 2 (TRAF2), which are implicated in activation of the NF-B and c-Jun N-terminal kinase/AP-1 pathways. This report shows that TNF also induces apoptosis through a caspase-8-independent mechanism in FaO rat hepatoma cells (Fig. 2). TNF treatment did not activate the cleavage of caspase-8; however, it induced activation of caspase-10. It has been suggested that in some cases caspase-10 can functionally substitute for caspase-8 in death receptor signal transduction (32). However, the role of caspase-10 is not clear in TNF-mediated apoptosis in FaO cells. TNF induces activation of caspase-12 in FaO cells. Caspase-12 is known to be essential for cell death induced by ER stress. Procaspase-12 is enriched in ER-containing microsomal fractions from the brain and is processed by the family of cytosolic calcium-dependent cysteine proteases, calpain (34). Procaspase-12 is cleaved, and the activated forms accumulate under ER stress conditions (35). Caspase-12-deficient mice are resistant to ER stress-induced apoptosis, but their cells are subject to apoptosis in response to other stimuli. Activation of caspase-12 can be directly stimulated by depletion of the ER calcium pool. Our results also support the concept that a rise in cytosolic free calcium concentration triggers caspase-12 activation and ER apoptotic signaling. Treatment with dantrolene, an inhibitor of calcium release from the ER, blocks the activation of caspase-12 and apoptotic signaling induced by TNF (Fig. 10).
The ER is a principal site for protein synthesis and folding and also serves as a cellular storage site for calcium (42,43). Agents that block protein folding or export, inhibitors of protein glycosylation, and agents that affect calcium uptake and re- lease from the ER can all lead to ER stress and ultimately cell death (44 -46). Several previous studies have implicated depletion of the ER calcium pool depletion in the initiation of apoptosis (19,20,47). TNF treatment also induces mobilization of Ca 2ϩ from intracellular stores in cultured sensory neurons (48). We confirmed that TNF treatment increases the release of Ca 2ϩ from the ER Ca 2ϩ pool and induces apoptosis in this study (Fig. 5).
Both the ER and mitochondria act as calcium stores controlling the capacitive calcium influx and cytoplasmic calcium homeostasis (49). It has been suggested that the ER Ca 2ϩ depletion caused by Bcl-2 overexpression is an integral part of the antiapoptotic program set by various agents (47). Ceramide induces a rise in cytoplasmic Ca 2ϩ concentration by releasing Ca 2ϩ from intracellular stores and activating the capacitative Ca 2ϩ entry pathway, resulting in prolonged mitochondrial Ca 2ϩ accumulation and alterations in organelle morphology (swelling and fragmentation) (50). Overexpression of Bcl-2 in WEHI7.2 cells blocks thapsigargin-induced cell death and thapsigargin-induced mobilization of ER Ca 2ϩ . The effect of Bcl-2 on the release of Ca 2ϩ from the ER may be cell type-dependent. Other laboratories have reported different effects of Bcl-2 on the intracellular calcium pool (19,20,22,23,47). Three groups have shown that Bcl-2 lowers the steady-state level of Ca 2ϩ within the ER and inhibits apoptosis by reducing Ca 2ϩ efflux across the ER membrane (19,20,23,47). In contrast, another group (22) showed that the overexpression of Bcl-2 increased luminal Ca 2ϩ concentrations in the ER by up-regulating calcium pump (SERCA) expression in breast epithelial cells. Although we cannot explain these contrasting findings at present, they are probably related to the use of different cellular model systems. It is conceivable that cellular context dictates precisely how Bcl-2 will influence intracellular Ca 2ϩ pools. Importantly, the connection between ER Ca 2ϩ pool emptying and apoptosis is not disputed. TNF treatment suppressed the protein level of Bcl-2 in FaO cells. This result is somewhat unexpected, given the evidence that Bcl-2 is capable of blocking Ca 2ϩ release from the ER, resulting in the protection of Ca 2ϩ -induced toxicity. Although the mechanism of suppression of Bcl-2 by TNF in FaO cells remains to be investigated, this study suggests another apoptotic pathway triggered by TNF. It will be interesting to find whether the residual apoptotic response observed upon TNF treatment of FADD-deficient cells (41) is mediated through the mobilization of Ca 2ϩ from the ER.
Interference of Ca 2ϩ release from the ER increases Bcl-2 mRNA and protein levels. Treatment with dantrolene, an inhibitor of Ca 2ϩ release from the ER, prevented cell death, and this protection by dantrolene was associated with a marked increase in the protein levels of Bcl-2 in GT1 hypothalamic neurosecretory cells (38). As the ER Ca 2ϩ level has a prominent role in regulating protein synthesis (51), the effect of dantrolene on ER Ca 2ϩ could be the underlying mechanism for Bcl-2 induction. In another study, treatment with lithium, which interferes with Ca 2ϩ release from the ER (52), protected cell death induced by glutamate, and this protection was associated with increases in Bcl-2 mRNA and protein levels in cultured cerebellar neurons (53). Our study also supports the concept that inhibition of Ca 2ϩ release from the ER induces Bcl-2 expression. Treatment of FaO cells with dantrolene inhibited TNF␣-induced apoptosis and increased the protein level of Bcl-2. Further investigation will be required to identify mechanisms contributing to the induction of the levels of Bcl-2 mRNA and protein by dantrolene. Regardless of the detailed mechanisms underlying the cytoprotective effects of dantrolene, the ability of dantrolene to induce Bcl-2 raises the possibility that this drug may be potentially useful in the treatment of some forms of neurodegenerative diseases.  (1 mM) were added 1 h before TNF treatment. Samples were taken at the indicated time points and assayed for caspase-3 cleavage by Western blotting using caspase-3-specific antibody. The full-length caspase protein and cleaved products (sizes indicated in kDa) are indicated. Con, control. B, cells were pretreated with either SK96365 or different doses of datrolene and then treated with TNF (20 ng/ml) for 24 h. Samples were taken at the indicated time points and assayed by Western blotting analysis using caspase-12-specific antibody that recognizes full-length protein as well as the active form. Cell lysates were made, and equivalent amounts of cellular proteins were separated by SDS-15% PAGE, blotted, and then probed with the appropriate antibody. Blots were shown are typical of at least three individual experiments.