Caspase-3-induced Truncation of Type 1 Inositol Trisphosphate Receptor Accelerates Apoptotic Cell Death and Induces Inositol Trisphosphate-independent Calcium Release during Apoptosis*

Inositol 1,4,5-trisphosphate receptor-deficient (IP3RKO) B-lymphocytes were used to investigate the functional relevance of type 1 inositol 1,4,5-trisphosphate receptor (IP3R1) and its cleavage by caspase-3 in apoptosis. We showed that inositol 1,4,5-trisphosphate receptor-deficient cells were largely resistant to apoptosis induced by both staurosporine (STS) and B-cell receptor (BCR) stimulation. Expression of either the wild-type IP3R1 or an N-terminal deletion mutant (Δ1-225) that lacks inositol 1,4,5-trisphosphate-induced Ca2+ release activity restored sensitivity to apoptosis and the consequent rise in free cytosolic Ca2+ concentration ([Ca2+]i). Expression of caspase-3-non-cleavable mutant receptor, however, dramatically slowed down the rate of apoptosis and prevented both Ca2+ overload and secondary necrosis. Conversely, expression of the “channel-only” domain of IP3R1, a fragment of the receptor generated by caspase-3 cleavage, strongly increased the propensity of the cells to undergo apoptosis. In agreement with these observations, caspase inhibitors impeded apoptosis and the associated rise in [Ca2+]i. Both the staurosporine- and B-cell receptor-induced apoptosis and increase in [Ca2+]i could be induced in nominally Ca2+-free and serum-free culture media, suggesting that the apoptosis-related rise in [Ca2+]i was primarily because of the release from internal stores rather than of influx through the plasma membrane. Altogether, our results suggest that IP3R1 plays a pivotal role in apoptosis and that the increase in [Ca2+]i during apoptosis is mainly the consequence of IP3R1 cleavage by caspase-3. These observations also indicate that expression of a functional IP3R1 per se is not enough to generate the significant levels of cytosolic Ca2+ needed for the rapid execution of apoptosis, but a prior activation of caspase-3 and the resulting truncation of the IP3R1 are required.

Apoptosis is a highly regulated and evolutionarily conserved form of cell death that plays an important role in normal embryonic development and maintenance of adult tissue homeostasis (1). Apoptotic cell death involves a characteristic sequence of morphological and biochemical features (2,3). Most, but not all, forms of apoptotic cell death processes are characterized by the activation of a family of aspartate-specific cysteine proteases called caspases that cleave a wide range of cellular proteins leading to the manifestation of the major phenotypes in apoptosis (4).
Early studies of glucocorticoid-induced cell death suggested that an increase in the cytosolic free Ca 2ϩ concentration ([Ca 2ϩ ] i ) was a key component of the apoptotic process (5). Various reports have since then established that a prolonged and up-regulated intracellular Ca 2ϩ signal is a general feature of apoptosis (6 -9). Apoptosis-related cleavage of a range of regulatory proteins and ion channels seems to be common to many apoptotic paradigms. During apoptosis, caspase-3, the main effector caspase, cleaves a wide array of cellular proteins including many that play significant roles in intracellular Ca 2ϩ regulation such as the Ca 2ϩ /calmodulin-dependent protein kinase IV (10), plasma membrane Ca 2ϩ ATPase (11,12), Na ϩ / Ca 2ϩ exchanger (13), and the ␤-subunit of the Na ϩ /K ϩ -ATPase (14). Caspase cleavage could bring about a gain or loss of function on the target proteins leading to aberrant intracellular Ca 2ϩ regulation that can directly influence the commitment of cells to apoptosis. Inositol 1,4,5-trisphosphate (IP 3 ) 1 receptors (IP 3 Rs) are ubiquitous intracellular Ca 2ϩ release channels, and their involvement in apoptosis has been demonstrated in different cell types. It was initially reported that the mRNA and protein levels of IP 3 R3 increase during apoptosis in lymphocytes, with no change in the IP 3 R1 level (15). Also, expression of an antisense cDNA construct of IP 3 R3 blocked the dexamethasoneinduced apoptosis and increase in [Ca 2ϩ ] i , whereas that of IP 3 R1 had no effect. Subsequently, however, it was reported that Jurkat cells deficient in IP 3 R1 were resistant to apoptosis induced by Fas, dexamethasone, and ␥-irradiation despite the presence of IP 3 R3 (16). IP 3 Rs are functionally redundant in chicken B-lymphocytes as apoptosis induced by B-cell receptor (BCR) stimulation was significantly inhibited only in cells deficient of all three receptors (17). The reason for these discrepancies is a matter of speculation, but together the reports indicate that Ca 2ϩ flux through IP 3 Rs plays a fundamental role in apoptotic cell death induced by various stimuli. A more direct involvement of IP 3 Rs in apoptosis was demonstrated by reports that identified IP 3 R1 as a substrate of caspase-3 during apoptosis (18,19). IP 3 R1, but not IP 3 R2 or IP 3 R3, contains a single DEVD-specific cleavage site for caspase-3 at amino acids 1888 -1891 (mouse sequence), and this sequence is conserved in rat and human proteins. Caspase-3-mediated degradation may actually explain a previous observation that the IP 3 R1 level shows a dramatic decrease following dexamethasone treatment of thymocytes and S49 cells (15). Cleavage by caspase-3 removes the cytoplasmic segment of IP 3 R1 comprising the IP 3 -binding domain and most of the regulatory domain. This cleavage obviously abolishes the IP 3 -induced Ca 2ϩ release (IICR) activity of the receptor and produces a "channel-only" domain that apparently remains constitutively open in transiently transfected COS-1 and HeLa cells (20). The significance of this cleavage either to the process of cell death or to the apoptosis-related increase in [Ca 2ϩ ] i is not yet clear.
Because IP 3 R1 is the most ubiquitous isoform of the IP 3 R family, its direct involvement in apoptosis as a caspase-3 substrate could have a far reaching physiological significance. Therefore, we set out to investigate the functional relevance of IP 3 R1 and its cleavage by caspase-3 to apoptosis induced by staurosporine (STS) and BCR stimulation using IP 3 R-deficient chicken B-lymphocytes (IP 3 R-KO) that stably express various mutants of the IP 3 R1. We report here that expression of either the wild-type IP 3 R1 or a mutant receptor that is no longer activated by IP 3 could render the IP 3 R-KO cells susceptible to apoptosis and support the consequent rise in [Ca 2ϩ ] i . Pretreatment of the cells with caspase inhibitors blocked the cleavage of IP 3 R1 and the rise in [Ca 2ϩ ] i as well as apoptosis. A functional receptor that was mutated at the caspase-3 cleavage site significantly slowed down the kinetics of apoptosis and prevented both Ca 2ϩ overload and secondary necrosis. In contrast, stable expression of the channelonly domain, corresponding to the C-terminal fragment generated by caspase-3, predisposed the cells to undergo apoptosis. Both STS-and BCR-induced apoptotic cell death and the associated rise in [Ca 2ϩ ] i could also be induced in nominally Ca 2ϩ -free culture medium, suggesting Ca 2ϩ release from internal stores as the primary cause of the [Ca 2ϩ ] i rise rather than a major influx from the extracellular medium. Our data also indicate that IP 3 R1 plays a pivotal role in apoptosis not necessarily through its IICR activity but mainly as a substrate of caspase-3. The perturbance of intracellular Ca 2ϩ homeostasis during the execution phase of apoptosis seems to be related to the cleavage of IP 3 R1 by caspase-3. Accordingly, the amplification of apoptotic signals and the rapid execution of apoptosis necessitate a prior activation of caspase-3 and the resulting truncation of the IP 3 R1. In addition, our results suggest that the specific pattern of changes in [Ca 2ϩ ] i during apoptosis in different cell types may be related to the relative distribution of IP 3 R1 among the different tissues.
Cells and Culture Conditions-The DT40 chicken B-lymphocytes lacking all three IP 3 Rs (IP 3 R-KO) were a kind gift from Dr. T. Kurosaki (Tokyo, Japan). IP 3 R-KO cells were maintained in RPMI 1640 medium containing 10% fetal calf serum, 1% chicken serum, 50 M 2-mercaptoethanol, 85 units/ml penicillin, 85 g/ml streptomycin, and 3.5 mM L-glutamine in a humidified incubator at 5% CO 2 and 37°C. These culture media and additives and Dulbecco's modified Eagle's medium with no added calcium (catalog no. 21068) were purchased from Invitrogen.
DNA Constructs and Transfection-Mouse cerebellum IP 3 R1 cDNA (a kind gift from Dr. K. Mikoshiba, Tokyo, Japan) in pcDNA-3.1(ϩ) vector was used as a template to generate different mutants of the receptor. Mutagenesis was carried out using the QuikChange XL sitedirected mutagenesis kit (Stratagene) according to the manufacturer's protocol. To construct the caspase-3-non-cleavable mutant of IP 3 R1 (IP 3 R1⌬casp), a fragment containing the region 2129 -6819 of the fullsize IP 3 R1 cDNA, flanked by two BamHI restriction sites, was first subcloned into the pBlueScript II SK(ϩ) vector. This construct was then used as template for the insertion of the mutations using 5Ј-GGGAA-ACAAAAAGAAAGATATCGAAGTGGCCAGGGATGCCCCGTC-3Ј as a forward primer and 5Ј-GAGGGGGCATCCCTGGCCACTTCGATATCT-TTCTTTTTGTTTCCC-3Ј as a reverse primer (underlined are the inserted mutations). These oligonucleotides replace the known caspase-3 cleavage site DEVD, encoded between nucleotides 2584 and 2594 with IEVA, thereby mutating the aspartic acids at positions 1888 and-1891 to isoleucine and alanine, respectively. The mutated IP 3 R1 region in the recombinant vector was then recovered with BamHI and religated into a BamHI-digested pcDNA3.1(ϩ)/IP 3 R1. The N-terminal deletion mutant of the IP 3 R1 lacking the first 225 amino acids ((⌬1-225)-IP 3 R1) was constructed as described earlier (22), and the deletion mutant (⌬1-1891)IP 3 R1 encoding the 95-kDa caspase-3-generated C-terminal region of the receptor was constructed by PCR amplification of the relevant region and subsequent substitution in the pcDNA3.1(ϩ)/IP 3 R1 plasmid. The sequences of the different constructs were confirmed by the automated fluorescent sequencing system (Amersham Biosciences). All constructs, including that of the wild-type receptor (WT-IP 3 R1), were transfected into IP 3 R-KO cells by electroporation using a Gene Pulser apparatus (Bio-Rad). Briefly, about 10 7 cells in 0.5 ml of serum-free medium were transferred to a 4-mm electroporation cuvette (Eurogentec, Seraing, Belgium) and pulsed at 550 V and 25 F in the presence of 100 g of plasmids. The electroporated cells were incubated in 30 ml of normal culture medium for 24 h before starting the selection with 1.5 mg of G418/ml to generate stable cell lines.
IP 3 R1 Subcellular Localization-The cells (5 ϫ 10 5 ) were attached to poly-L-lysine-coated, 2-well chambered slides (Nalge Nunc, Naperville, IL) for 3 h before fixation in 3% paraformaldehyde for 15 min at room temperature. The fixed cells were then permeabilized with 0.5% Triton X-100 in PBS for 5 min, and the nonspecific binding sites were blocked with 20% goat serum in PBS for 1 h at room temperature. The cells were then incubated for 1 h with the Rbt04 primary antibody in PBS containing 1.5% goat serum. Subsequently, the slides were washed three times with PBS and incubated with the Alexa Fluor 488 goat anti-rabbit secondary antibody (Molecular Probes) in PBS containing 5% goat serum. BODIPY-thapsigargin and MitoTracker dyes (Molecular Probes) were used to visualize endoplasmic reticulum (ER) and mitochondria, respectively. BODIPY-thapsigargin was added together with the secondary antibody. Mitochondria were visualized by incubating the cells with MitoTracker for 30 min at 37°C prior to fixation. Images were acquired using an LSM510 confocal laser-scanning microscope (Carl Zeiss, Germany) with a Plan-Neofluar ϫ100 numerical aperture 1.3 oil immersion objective. Excitation wavelengths of 488 nm for Alexa Fluor 488 and 543 nm for BODIPY-thapsigargin and MitoTracker were used. Emission fluorescence was monitored by a photomultiplier fitted with a BP filter of 505-530 nm for Alexa Fluor 488 and with an LP filter of 585 nm for BODIPY-thapsigargin and MitoTracker.
Preparation of Cell Lysates and Microsomes-Following treatments, cells were harvested and washed in ice-cold PBS before preparation of lysates as described previously (23). Total microsomes were prepared as described previously with minor modifications (21). Briefly, cells were harvested by centrifugation for 5 min at 400 ϫ g and washed twice with ice-cold PBS without Ca 2ϩ and Mg 2ϩ . Cell pellets were then resuspended in homogenization buffer (10 mM Tris/HCl, pH 7.4, 1 mM EGTA, 0.8 mM benzamidine, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 g/ml leupeptin, 0.5 g/ml aprotinin, 0.5 g/ml pepstatin A) and homogenized on ice with a probe sonicator (MSE Ltd., UK). Total microsomes were obtained by centrifugation for 25 min at 125,000 ϫ g. The membranous pellet was then resuspended in end medium (20 mM Tris/HCl, pH 7.4, 300 mM sucrose, 0.8 mM benzamidine, and 0.2 mM phenylmethylsulfonyl fluoride). Cell lysates and microsomal preparations were frozen in liquid nitrogen and stored at Ϫ80°C. Protein concentrations were determined using either the BCA protocol (Pierce) or the Lowry method (24) with bovine serum albumin as a standard.
Cleavage of IP 3 R1 by Recombinant Caspase-3-For in vitro assay of IP 3 R1 cleavage by caspase-3, microsomes (200 g) were incubated with purified active recombinant human caspase-3 (Pharmingen, BD Biosciences) at 37°C for 1 h. The samples were then analyzed by Western blot using the Rbt04 antibody against the C-terminal region of IP 3 R1. Briefly, the samples were subjected to electrophoresis on 3-8% Trisacetate SDS-polyacrylamide gradient gels (Invitrogen), transferred to a polyvinylidene difluoride membrane, and subsequently incubated with the primary antibody (1:3000 dilution) and alkaline phosphatase-conjugated secondary antibody (1:8000 dilution). The immunoreactive bands were developed using the enhanced chemifluorescence reagent from Amersham Biosciences and then detected by the Storm840 FluorImager equipped with the ImageQuant software (Amersham Biosciences).
Induction and Analysis of Apoptosis-Cells were seeded at a density of 0.5 ϫ 10 6 cells/ml (1.5 ml of medium) in 12-well plates before treatment with either STS or anti-chicken IgM and then incubated for a specified period of time before harvesting. In some experiments, the anti-chicken IgM antibody was supplemented with anti-mouse IgM to further cross-link the surface IgM. For analysis of apoptosis, the protocol included in the annexin V-FITC apoptosis detection kit from Pharmingen was used as provided. Cell death detection was performed on a Coulter Epics flow cytometer (Beckman-Coulter Inc., Miami, FL) using the standard emission filters for green (FL1) and red (FL3) fluorescence photomultipliers. The Expo32 MultiCOMP software from Coulter Corporation was used to analyze the data. Cells having a reduced overall volume and staining with annexin V-FITC while retaining the plasma membrane integrity (propidium iodide (PI)-negative) were regarded as apoptotic. Primary necrosis was identified as the loss of plasma membrane integrity without a clear reduction in cell volume and no annexin staining. Primary necrotic cells stain with PI only. Cells that stain with both annexin V-FITC and PI were considered as those undergoing secondary necrosis, subsequent to apoptotic cell death.
Caspase-3 Assay-Cells treated with STS or anti-chicken IgM antibody and incubated for the indicated time were harvested in ice-cold buffer containing 50 mM Tris/HCl, pH 7.6, 150 mM NaCl, 1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 g of each/ml leupeptin, aprotinin, and pepstatin A, 1 mM dithiothreitol, and 1% Triton X-100. Caspase-3 assay was then performed by the colorimetric CaspACE assay system (Promega) exactly as recommended by the manufacturer, using 50 -100 g of protein/assay. After incubation for 3 h at 37°C, A 405 readings were taken using a 96-well plate reader.

Analysis of [Ca 2ϩ ] i by Flow Cytometry-Measurement of [Ca 2ϩ
] i was performed essentially as described previously (25,26). Briefly, 10 6 cells were first treated with STS or anti-chicken IgM for a specific period of time and then loaded with either 3 M Fluo-3/AM or 6 M Fura red/AM (Molecular Probes) in 200 l of culture medium for 30 and 90 min, respectively, at 37°C. At the end of the incubation period, extra medium was added to make the final cell density up to about 10 6 cells/ml. 15,000 cells/sample were analyzed for the Ca 2ϩ -dependent increase in Fluo-3 fluorescence and the Ca 2ϩ -dependent decrease in the Fura red fluorescence emission by a Coulter Epics flow cytometer by exciting the cells at 488 nm.

Expression of Different IP 3 R1 Constructs in IP 3 R-deficient
Cells-The basic structure of IP 3 R1 comprises the N-terminal IP 3 -binding domain, the regulatory domain, and the C-terminal channel domain as shown schematically in Fig. 1A. Previous studies have shown that caspase-3 mediates the cleavage of mouse IP 3 R1 at Asp-1891 in cells undergoing apoptosis (18,19). In this study, we aimed to investigate the exact role of the IP 3 R1 channel activity and the significance of this cleavage in apoptotic cell death. For this purpose, we first generated a caspase-non-cleavable mutant (IP 3 R1⌬casp) by introducing the mutations D1888I and D1891A (Fig. 1B) using mouse IP 3 R1 as a template. IP 3 R1⌬casp possesses the same level of IICR activity as that of the WT-IP 3 R1 in stably expressing cell lines (data not shown). We also generated a deletion mutant lacking the first 225 amino acids ((⌬1-225)IP 3 R1). This segment immediately precedes the ligand-binding core (Fig. 1C) and has been designated as a suppressor domain because of the observation that its deletion significantly enhances the affinity of the receptor for IP 3 binding (27,28). However, we (22) and others (29) have observed that despite the higher level of IP 3 binding, (⌬1-225)IP 3 R1 lacks any detectable IICR activity. The use of this mutant will help clarify whether the role of IP 3 R1 in apoptosis requires its IICR activity. Finally, a mutant receptor that lacks amino acids 1-1891 ((⌬1-1891)IP 3 R1) and mainly comprising the channel domain (as generated by caspase-3 cleavage of IP 3 R1) was created by PCR amplification of the specific region (Fig. 1D). All constructs including the WT-IP 3 R1 were transfected by electroporation into chicken embryonic B-lymphocytes (DT40 cells) lacking all three IP 3 R isoforms (IP 3 R-KO cells), and permanent cell lines were then established by selection with G418. The level of expression of all constructs in cell lines was comparable, as determined by Western blot analysis of membrane preparations (data not shown). In addition, the subcellular localization of the constructs was determined using confocal laser-scanning microscopy by costaining of the IP 3 R1 (Rbt04 antibody and Alexa Fluor 488conjugated secondary antibody) and the ER (BODIPY-thapsigargin) or mitochondria (MitoTracker). Representative images from the cells expressing the (⌬1-1891)IP 3 R1 construct are shown in Fig. 1, E and F. The truncated IP 3 R1 was expressed exclusively in the perinuclear region and was strongly associated with the distribution of the ER. It was excluded from mitochondria and the plasma membrane. These results are in agreement with the subcellular localization of a similar construct in other cell types (20).
Cleavage of IP 3 R1 in Vitro and in Cells Undergoing Apoptosis-The first data demonstrating a possible involvement of IP 3 R1 in cell death were published a few years ago showing that cells deficient in IP 3 R1 were resistant to apoptosis (16). This study, together with reports that caspase-3 mediates the cleavage of IP 3 R1 (18,19), prompted us to examine the relevance of IP 3 R1 cleavage to the cell death processes and its effect on intracellular Ca 2ϩ distribution. First, to examine whether the cleavage process could occur in vitro as well as in stable cell lines, crude microsomal preparations from WT-IP 3 R1 and IP 3 R1⌬casp cells were incubated with recombinant human caspase-3. In addition, we also studied IP 3 R1 degradation in cells treated with 50 nM STS to induce apoptosis. The extent of receptor degradation was then analyzed by Western blotting. As shown in Fig. 2A, recombinant caspase-3 cleaved IP 3 R1 in vitro in a dose-dependent manner generating exactly the same 95-kDa fragment as in cells undergoing STS-induced apoptosis. No such cleavage product could be detected in microsomal preparations from IP 3 R1⌬casp cells. Pretreatment of WT-IP 3 R1 cells with 100 M z-VAD-fmk (a pan-caspase inhibitor) or z-DEVD-fmk (a specific inhibitor of caspase-3) completely inhibited the degradation of IP 3 R1 during STS-induced apoptosis (data not shown). These results confirm that caspase-3 was responsible for the cleavage of IP 3 R1 in cells undergoing apoptosis. The same 95-kDa fragment was also generated in (⌬1-225)IP 3 R1 cells induced to undergo apoptosis by either 50 nM STS or BCR cross-linking using 15 g/ml anti-chicken IgM (Fig. 2B). Again, no cleavage product was detected in IP 3 R1⌬casp cells. These results suggest that the process of IP 3 R1 cleavage by caspase-3 is probably common to a wide range of apoptosis-inducing agents. Importantly, the results from the (⌬1-225)IP 3 R1 cells indicate that the initial caspase-3 activation during both STS-and anti-chicken IgMinduced apoptosis does not require IICR activity.
Role of Caspase-3-mediated Cleavage of IP 3 R1 in the Process of Apoptotic Cell Death-Previous studies have shown that the degradation of IP 3 R1 by caspase-3 inhibits IICR activity in microsomal preparations from cerebellum (18) and in digitonin-permeabilized A7r5 cells. 2 However, the point at which Ca 2ϩ is involved in apoptosis and the possible contribution of the IP 3 R1 and its cleavage to the process of cell death are not yet thoroughly investigated. We addressed these points by inducing apoptosis by 50 nM STS or by BCR stimulation with 15 g/ml anti-chicken IgM in cells expressing the different constructs of IP 3 R1. The level and rate of STS-induced caspase-3 activation, as determined by using DEVD-p-nitroaniline substrate, were drastically reduced in IP 3 R-KO cells (Fig. 3A). The caspase-3 activity in WT-IP 3 R1, (⌬1-225)IP 3 R1, and (⌬1-1891)IP 3 R1 cells showed a steady time-dependent increase of up to about 4.5-fold higher than that in IP 3 R-KO cells at the

Elevation of [Ca 2ϩ ] i during Apoptosis Requires the Cleavage of IP 3 R1 by Caspase-3-We next examined whether the alterations in [Ca 2ϩ
] i that usually accompany apoptotic cell death are influenced by the cleavage status of the IP 3 R1 constructs. Following the treatment of the cells with 50 nM STS, the induction of apoptosis and changes in [Ca 2ϩ ] i were monitored in parallel by flow cytometry using annexin V-FITC and Ca 2ϩbinding dyes Fluo-3/AM and Fura red/AM. Cell populations known from preliminary experiments to be positive for PI staining, indicative of the loss of plasma membrane integrity, were excluded during analysis. The increase in [Ca 2ϩ ] i was observed as a shift in the peak in the cell number versus fluorescence intensity distribution. As depicted in Fig. 4, A and  B, the increase in [Ca 2ϩ ] i was indicated by a shift to the right in Fluo-3 fluorescence (higher intensity upon binding Ca 2ϩ ) (Fig. 4A) and a corresponding shift to the left in Fura red fluorescence (lower intensity upon binding Ca 2ϩ ) (Fig. 4B). These shifts were clearly observed during apoptosis in WT-IP 3 R1, (⌬1-225)IP 3 R1, and (⌬1-1891)IP 3 R1 cells, with no significant change in IP 3 R-KO and IP 3 R1⌬casp cells, consistent also with the low level of apoptosis in these cells (see Fig. 3B).
Predictably, the largest increase in the percentage of cells with elevated [Ca 2ϩ ] i was observed in cells expressing (⌬1-1891)IP 3 R1, the channel-only fragment, which also had a considerably higher percentage of apoptotic cells. The rise in the level of [Ca 2ϩ ] i showed a pattern of time-dependent increase, commencing at 6 -8 h after the addition of STS (data not shown), suggesting that it was the consequence of the induction of apoptosis and caspase-3 activation (see Fig. 3A). The results depicted in Fig. 5 confirm that the apoptosis-associated increase in [Ca 2ϩ ] i was indeed caused by the activity of caspases. Pretreatment of cells with 100 M z-VAD-fmk blocked the STSinduced increase in [Ca 2ϩ ] i . Similar results were obtained with the specific caspase-3 inhibitor z-DEVD-fmk (data not shown). Taken together, the results indicate that the increase in [Ca 2ϩ ] i measured here was the consequence rather than the primary cause of apoptosis as it requires caspase-mediated cleavage of IP 3 R1. These results also establish that the truncation of IP 3 R1 by caspase-3 was a crucial requirement for the rise in [Ca 2ϩ ] i in cells undergoing apoptosis. However, it should be noted that the cleavage of IP 3 R1 by caspase-3, although absolutely necessary, is not sufficient by itself to induce a spontaneous rise in

STS-induced Elevation of [Ca 2ϩ ] i in Nominally
Ca 2ϩ -free Medium-To determine whether the rise in [Ca 2ϩ ] i was because of the release of calcium from intracellular stores or because of an influx of extracellular calcium, changes in intracellular calcium were determined in cells treated with STS in serum-free medium with no added calcium. Incubation of the cells in this nominally Ca 2ϩ -free medium did not change the basal level of intracellular calcium in untreated control cells as compared with those cells cultured in normal medium. Treatment of the cells with STS induces a similar level of apoptotic cell death (data not shown) and rise in [Ca 2ϩ ] i (Fig. 6) whether or not extracellular calcium was present. These results indicate that the elevated calcium level in STS-treated cells is derived mainly from the intracellular stores. It is conceivable that an influx of extracellular Ca 2ϩ triggered by the depletion of intracellular stores could further contribute to the sustained increase in [Ca 2ϩ ] i . At a later stage of apoptosis, Ca 2ϩ influx may also increase as a result of the reduced plasma membrane integrity.
Excessive Intracellular Free Ca 2ϩ Enhances the Rate of Apoptosis Leading to a Secondary Necrosis-Cells were incubated with 50 nM STS for a prolonged period (24 h), and the percentage of apoptotic, necrotic, and secondary necrotic cells was determined by flow cytometry using the annexin V-FITC/PI apoptosis detection kit. Secondary necrosis can be regarded as a postapoptotic event that is observed particularly during apoptosis in cultured cells. Secondary necrotic cells readily stain both with annexin V-FITC and PI. Fig. 7 shows the percentage of cells that undergo secondary necrosis following treatment with 50 nM STS for 24 h. The proportion of cells in this phase of cell death ranged from about 40% in WT-IP 3 R1 cells to over 65% in (⌬1-1891)IP 3 R1 cells. The number of secondary necrotic cells in IP 3 R-KO and IP 3 R1⌬casp cells did not differ appreciably and was less than 15% on average. The results suggest that caspase cleavage of IP 3 R1 and the resulting rise in [Ca 2ϩ ] i augment the rate of apoptotic cell death to the point where it switches into necrosis because of Ca 2ϩ overload. The deficiency of IP 3 R1 or mutation of its caspase-3 cleavage site precludes an excessive increase in [Ca 2ϩ ] i thereby resulting in an exceedingly slow rate of apoptosis. DISCUSSION Changes in Ca 2ϩ homeostasis in general and elevation of [Ca 2ϩ ] i in particular are an integral part of the process of cell death in many systems. Several mechanisms have been proposed regarding the contribution of cytosolic Ca 2ϩ to apoptosis (13,30). Because all mammalian cell types express at least one of the three IP 3 Rs, it can be presumed that these receptors play a central role in the regulation of [Ca 2ϩ ] i during apoptosis. In this study, we showed that IP 3 Rs are required for STS-and BCR-induced apoptosis because IP 3 R-KO cells were mostly resistant to cell death induced by both stimuli. Expression of IP 3 R1 was sufficient to restore the susceptibility of the cells to apoptosis and a consequent rise in [Ca 2ϩ ] i . Surprisingly, expression of a deletion mutant that lacks IICR activity renders the cells even slightly more sensitive to apoptotic cell death than those expressing the wild-type receptor. One conclusion that could be drawn from these results is that the IICR activity of IP 3 R1 per se was not required to mediate either the cell death or the associated rise in [Ca 2ϩ ] i . Significantly, cells expressing a mutant IP 3 R1 resistant to caspase-3 cleavage underwent extremely slow apoptotic cell death. Conversely, expression of the channel-only domain of the IP 3 R1 strongly increased the propensity of the cells to undergo apoptosis. Our results indicate that IP 3 R1 plays an important role in apoptosis and that the disturbance in intracellular Ca 2ϩ homeostasis during apoptosis is mainly caused by IP 3 R1 cleavage by caspase-3. This essential role of caspase-3 in this process was substantiated by the observation that the expression of a caspase-non-cleavable mutant of IP 3 R1 or the treatment of cells with caspase inhibitors strongly impedes such a rise in [Ca 2ϩ ] i . We also observe that rapid execution of apoptosis does not require the expression of a functional IP 3 R1 per se but a prior activation of caspase-3 and the resulting cleavage of the receptor. The requirement for some level of caspase activity illustrates that the elevation in [Ca 2ϩ ] i was the consequence but not the cause of the initiation of STS-or BCR-induced apoptosis. As such, increased [Ca 2ϩ ] i may function as an additional stress that sustains and amplifies the apoptosis signals rather than serving as an obligatory messenger for the induction of death. Both apoptosis and the accompanying increase in [Ca 2ϩ ] i could be induced in nominally Ca 2ϩ -free culture medium, suggesting that the primary source of increased [Ca 2ϩ ] i was the intracellular store, in agreement with a previous report (16). The finding is also compatible with the subcellular localization of the IP 3 R1 in DT40 cells as well as other cell types (20), which was coincident with an ER marker. The strong dependence of a sustained [Ca 2ϩ ] i increase on caspase-3 truncation suggests that it is primarily caused by an increased leak from the ER, with conceivably a consequent activation of store-operated Ca 2ϩ influx. However, the cleavage of other regulators of intracellular Ca 2ϩ ,

FIG. 5. Caspase inhibitors block apoptosis-related increase in [Ca 2؉ ] i .
The indicated cells were preincubated with 100 M z-VAD-fmk for 1 h before treatment with 50 nM STS. After incubation for 12 h, cells were harvested for analysis of changes in Fluo-3 and Fura red fluorescence by flow cytometry exactly as described in Fig. 4. notably the plasma membrane Ca 2ϩ -ATPase (12) and the Na ϩ / Ca 2ϩ exchanger (13), can significantly contribute to the overall rise in [Ca 2ϩ ] i during apoptosis. In addition, we cannot exclude the possibility that an influx of extracellular Ca 2ϩ could occur in the late phase of apoptosis because of the emptying of intracellular stores and/or the deterioration of cell membrane integrity.
A new impetus to the study of IP 3 Rs emanated from reports implying IP 3 Rs in apoptosis and IP 3 R1 as a caspase-3 substrate. Structure-function relationship studies of IP 3 R1 have led to the proposal that the large regulatory domain of the receptor was necessary to maintain a closed state of the channel in resting cells (29). In principle, therefore, the cleavage and removal of this region should specifically abolish the IICR, but the effect on the general properties of the channel remains mostly unclear. A recent study has demonstrated that caspase cleavage of IP 3 R1 results in a constitutively leaky channel, resulting in an almost complete emptying of the stores in transiently overexpressing cells (20). The extent of leakiness of the channel and the level of this apparently unregulated release of Ca 2ϩ from the ER remain to be fully characterized. We reasonably expect that the stably expressed (⌬1-1891)IP 3 R1 behaves similarly to the endogenously generated channel-only domain as both have a comparable expression level and the same localization at the ER. However, it is clear that in our model system using stable expression, the assumed passive leak did not result in a considerably increased [Ca 2ϩ ] i in resting cells and was not sufficient by itself to trigger apoptotic cell death. Moreover, a similar level of thapsigargin-induced increase in [Ca 2ϩ ] i was observed in cells that stably express the truncated receptor as in those expressing the WT-IP 3 R1 (data not shown), suggesting that thapsigargin-sensitive stores of these cells are sufficiently filled. Nevertheless, after treatment with suboptimal levels of either STS or anti-chicken IgM, the FIG. 6. Intracellular Ca 2؉ increase in normal and in nominally Ca 2؉ -free medium. Apoptosis was induced using 50 nM STS in the indicated cells grown in normal medium or in those that were transferred to nominally Ca 2ϩ -free medium just before treatment. The level of [Ca 2ϩ ] i was determined using Fluo-3 and a flow cytometer as described above, 8 h after treatment. A representative result from at least three independent experiments is shown. cells expressing the channel-only domain undergo a swift apoptotic cell death with an enhanced rate of caspase-3 activation and phosphatidylserine exposure relative to those expressing the WT-IP 3 R1. Therefore, it seems that an enhanced response to cell death stimuli and a related rise in [Ca 2ϩ ] i require apoptosis-associated changes and/or activation of an essential cofactor during apoptosis as well as the caspase-3 cleavage of IP 3 R1.
The significance of an increased [Ca 2ϩ ] i , especially downstream of caspase activation, is not clear. As an executioner caspase, the activation of caspase-3 requires upstream stimuli in the form of death receptor stimulation and/or the release of mitochondrial cytochrome c, depending on the apoptosis-inducing agent (4). Thus, a detectable level of caspase-3 activity would eventually lead to apoptotic cell death albeit at a rate dictated by the intensity of the upstream signal. Most likely, the late cleavage of IP 3 R1 and the resulting buildup of cytosolic Ca 2ϩ were required to sustain and augment the apoptotic signals and the rate of cell death, thereby ensuring a speedy demise and removal of apoptotic cells. Recently, a study has shown that mitochondrial cytochrome c translocates to the ER early in apoptosis, selectively binds to the C-terminal tail of IP 3 Rs, and blocks the Ca 2ϩ -dependent inhibition of IP 3 R function, which results in an oscillatory [Ca 2ϩ ] i increase (31). This study envisages a universal role for cytochrome c as an agonist of all IP 3 Rs leading to increased Ca 2ϩ release from the ER, apparently irrespective of the apoptosis-inducing agent. In effect, all three IP 3 Rs would have identical roles as apoptosis signal amplifiers (through cytochrome c-induced Ca 2ϩ release activities), which, however, was not found in other notable reports (15,16). From our observation, the lack of elevated [Ca 2ϩ ] i in IP 3 R1⌬casp-expressing cells strongly suggests that the increased Ca 2ϩ release during apoptosis was mainly because of the cleavage of the receptor by caspase-3 rather than cytochrome c binding. However, it is possible that cytochrome c released from a limited population of mitochondria can bind to IP 3 Rs on adjacent ER to induce a local Ca 2ϩ release that may ultimately lead to a rise in mitochondrial and cytosolic Ca 2ϩ levels under certain conditions. It is also possible that regulation by cytochrome c may be more relevant for type 2 and 3 IP 3 Rs that do not have caspase cleavage sites. In any case, the observations that caspase-3 cleaves IP 3 R1 and that cytochrome c may bind to IP 3 Rs, in both cases to increase [Ca 2ϩ ] i , illustrate the existence of an elaborate and vital autoamplification loop, whereby the release of mitochondrial cytochrome c and/or, more likely, caspase-3 activation leads to enhanced ER Ca 2ϩ release at different time points during apoptosis, resulting in an enhanced cell death signal. Tombal et al. (32) have demonstrated that such a late rise in [Ca 2ϩ ] i was essential to complete the execution phase of cell death regardless of the apoptosisinducing agent. In this regard, it should be emphasized that the expression of a caspase-non-cleavable mutant of IP 3 R1 did not block the apoptotic cell death but merely delayed the process for a considerable period of time, probably until the cells attain the supramicromolar concentration of intracellular Ca 2ϩ that was suggested to be a prerequisite for the activation of the execution phase of apoptosis (32).
Conditions associated with mitochondrial Ca 2ϩ overload lead to rapid loss of mitochondrial function and cell death by necrosis (7,33,34) as a result of the consequent irreversible permeability transition pore opening and depolarization of the mitochondrial membrane, collapse of ATP production, and generation of reactive oxygen species (3,35). Some of these responses form a self-amplification loop as they further increase [Ca 2ϩ ] i by favoring the release of stored Ca 2ϩ from the ER and/or inhibiting extrusion of the ion from the cell. The switch from Ca 2ϩ overload-induced apoptotic signals to necrosis depends, in part, on the intensity of the death-inducing signal (36). Our results indicate that such secondary necrosis events can occur downstream of caspase activation and apoptotic cell death. Cells that express caspase-cleavable IP 3 R1 constructs and the channel-only domain invariably undergo an extensive necrosis after an initial robust apoptotic cell death, suggesting that the generation of the channel domain results in cytoplasmic and/or mitochondrial Ca 2ϩ overload when cells are challenged by apoptosis-inducing agents. The dependence of secondary necrosis on caspase activity may account for the previous observations where caspase inhibitors protected against cell death in ischemic and excitotoxic brain injury (34,37,38), which was also associated with Ca 2ϩ overload. Neuronal cells predominantly express IP 3 R1 (39,40), and dysregulation of Ca 2ϩ signaling is involved in neuronal cell death (41)(42)(43). Moreover, combined treatments with caspase inhibitors and Ca 2ϩ channel blockers synergistically protect against cerebral histotoxic hypoxia (37). Therefore, it is imperative to investigate the contribution of IP 3 Rs in stroke and neurodegenerative disorders.