BH3 mimetic-elicited Ca2+ signals in pancreatic acinar cells are dependent on Bax and can be reduced by Ca2+-like peptides

BH3 mimetics are small-molecule inhibitors of B-cell lymphoma-2 (Bcl-2) and Bcl-xL, which disrupt the heterodimerisation of anti- and pro-apoptotic Bcl-2 family members sensitising cells to apoptotic death. These compounds have been developed as anti-cancer agents to counteract increased levels of Bcl-2 proteins often present in cancer cells. Application of a chemotherapeutic drug supported with a BH3 mimetic has the potential to overcome drug resistance in cancers overexpressing anti-apoptotic Bcl-2 proteins and thus increase the success rate of the treatment. We have previously shown that the BH3 mimetics, BH3I-2′ and HA14-1, induce Ca2+ release from intracellular stores followed by a sustained elevation of the cytosolic Ca2+ concentration. Here we demonstrate that loss of Bax, but not Bcl-2 or Bak, inhibits this sustained Ca2+ elevation. What is more, in the absence of Bax, thapsigargin-elicited responses were decreased; and in two-photon-permeabilised bax−/− cells, Ca2+ loss from the ER was reduced compared to WT cells. The Ca2+-like peptides, CALP-1 and CALP-3, which activate EF hand motifs of Ca2+-binding proteins, significantly reduced excessive Ca2+ signals and necrosis caused by two BH3 mimetics: BH3I-2′ and gossypol. In the presence of CALP-1, cell death was shifted from necrotic towards apoptotic, whereas CALP-3 increased the proportion of live cells. Importantly, neither of the CALPs markedly affected physiological Ca2+ signals elicited by ACh, or cholecystokinin. In conclusion, the reduction in passive ER Ca2+ leak in bax−/− cells as well as the fact that BH3 mimetics trigger substantial Ca2+ signals by liberating Bax, indicate that Bax may regulate Ca2+ leak channels in the ER. This study also demonstrates proof-of-principle that pre-activation of EF hand Ca2+-binding sites by CALPs can be used to ameliorate excessive Ca2+ signals caused by BH3 mimetics and shift necrotic death towards apoptosis.

Disrupted regulation of apoptosis is the hallmark of carcinogenesis allowing accumulation of further genetic mutations and acquisition of metastatic properties. 1 Cancer cells often express increased levels of anti-apoptotic Bcl-2 (B-cell lymphoma-2) proteins, which provides additional protection against cell death signals 2,3 correlating with chemotherapy resistance and poor prognosis for patients. 4 As anti-cancer therapies often target the mitochondrial apoptotic pathway regulated by the Bcl-2 family, overexpression of anti-apoptotic proteins in cancer presents one of the leading challenges to overcome for effective treatment. 5 Mechanism of apoptosis induction in cancer is predominantly altered upstream of Bax and Bak. 6 Therefore pharmacological suppression of anti-apoptotic Bcl-2 members leading to activation of Bax and Bak should, in principle, be capable of recovering the programmed cell death. 7,8 BH3 mimetics are small-molecule synthetic inhibitors of Bcl-2 and Bcl-xL, specifically developed as anti-cancer agents. 7 They mimic activated BH3-only proteins by disrupting the heterodimerisation of anti-and pro-apoptotic Bcl-2 family members, and thus sensitising cells to apoptosis. HA14-1 was the first BH3 mimetic obtained by molecular modelling that had the ability to displace Bax from Bcl-2, followed by the induction of cell death. 7 Later, a family of seven members of BH3 mimetics, called BH3 inhibitors (BH3Is), was developed and shown to displace Bak peptide from Bcl-xL, trigger apoptosis, cytochrome c release and caspase activation. 9 A natural BH3 mimetic gossypol, isolated from the cotton plant (Gossypium), is also capable of inhibiting Bcl-2, Bcl-xL and Mcl-1. 10,11 Simultaneous application of a chemotherapeutic agent and a BH3 mimetic can potentially overcome drug resistance in cancers overexpressing anti-apoptotic Bcl-2 proteins and thus increase the treatment success rate. BH3I-2′ and HA14-1 were shown to sensitise leukaemic cells in vitro to TRAILinduced apoptosis. 12 Gossypol (AT-101) was found to increase radiation efficacy in head and neck cancer cell lines in vitro; 13 and recently underwent phase II clinical trials in combination with docetaxel 14 and with androgen deprivation therapy. 15 A particularly interesting approach is to develop chemically tailored BH3 mimetics that interact specifically with the antiapoptotic proteins overexpressed in a target cancer type. 8 As exciting as the prospect of Bcl-2 inhibition might seem, accumulating evidence shows that BH3 mimetics often affect intracellular Ca 2+ homoeostasis. Previously we have reported that HA14-1 and BH3I-2′ deplete the ER Ca 2+ store causing a sustained elevated cytosolic Ca 2+ concentration in pancreatic acinar cells (PACs). 16 Others have demonstrated that HA14-1 causes Ca 2+ deregulation in platelets, HeLa and HEK-293T cells. 17 This immediately triggers questions about safe and specific use of BH3 mimetics, especially given the crucial role of Ca 2+ in the regulation of a wide variety of intracellular process including muscle contraction, 18 enzyme secretion, 19 fertilisation, 20,21 cell proliferation 22 and death. 23 This is particularly important in the pancreas, as physiological Ca 2+ oscillations control enzyme secretion in PACs, whereas abnormal Ca 2+ signals are the hallmark of the initial stages of a severe necrotising disease of the pancreasacute pancreatitis. [24][25][26] This study aims to assess the effects of BH3 mimetics on Ca 2+ handling in relation to activation of proapoptotic Bax in PACs. Also, we draw conclusions about the involvement of Bax in intracellular Ca 2+ signalling and propose means to reduce the excessive Ca 2+ signals enabling modulation of the cell death mechanisms.

Results
Ca 2+ responses to BH3 mimetics in pancreatic acinar cells are dependent on Bax. BH3I-2′ and HA14-1 were shown to induce a slow Ca 2+ release from the intracellular stores. 16 Pharmacological inhibition of inositol 1,4,5-triphosphate receptors (IP 3 Rs) and ryanodine receptors (RyRs) led to a decreased Ca 2+ release from the ER of permeabilised PACs but did not completely block it. This indicates that IP 3 Rs and RyRs were not the primary source of BH3 mimeticelicited Ca 2+ release but merely amplified it. 16 Here PACs, isolated from wild-type (WT) mice as well as animals with a loss-of-function mutation in one of the following genes: bcl-2, bak or bax were treated with 5 μM BH3I-2′ (Figure 1a) or 30 μM HA14-1 (Figure 1b) in the absence of extracellular Ca 2+ . In WT, bcl-2 −/− and bak −/− cells, these BH3 mimetics caused sustained elevations of the cytosolic Ca 2+ concentration ([Ca 2+ ] i ), whereas in bax −/− cells the Ca 2+ signal generation was largely abolished (Figures 1a and b). Quantitative comparison of the responses to BH3I-2′ ( Figure 1c) and HA14-1 ( Figure 1d) confirmed that loss of Bax, but not Bcl-2 or Bak, dramatically inhibited the BH3 mimetic-induced elevation of the [Ca 2+ ] i (Po0.001). The responses to BH3I-2′ or HA14-1 in bax −/− cells, if present at all, were mainly short-lasting Ca 2+ transients, whereas sustained elevation was mostly inhibited (Figures 1e and f). Further, measurements of Ca 2+ release from the ER in twophoton-permeabilised PACs revealed that BH3I-2′ not only appears to release less Ca 2+ from the ER of bax −/− compared to WT cells (Po0.01; Figures 2a and b) but also that the apparent rate of release was slower than in WT cells. Consequently, τ 1/2 for bax −/− (226.5 ± 25.9 s) cells was higher than for WT cells (145.1 ± 19.3 s; Po0.05; Figure 2c), which translates into longer time needed for the fluorescence of the Ca 2+ indicator to decrease by half the difference between baseline and final values. Images of a PAC doublet before and after laser permeabilisation are depicted in Figure 2d. Increasing the ER store loading with CDN1163, a postulated allosteric activator of SERCA2b (sarco/endoplasmic reticulum Ca 2+ -ATPase, isoform 2b), 27 did not significantly affect the responses to BH3I-2′ in bax −/− cells (Figure 2e). Finally, BH3I-2′ increased apoptosis in WT cells by~20.6 ± 3.8% and necrosis by 26.1 ± 2.7% over control levels; these effects were completely abolished by intracellular Ca 2+ chelation with BAPTA (Po0.05; Figure 2f; bars and corresponding images). In bax −/− cells, BH3I-2′ did not affect apoptosis and only slightly increased necrosis (by 8.8 ± 3.8%); the latter was completely inhibited by BAPTA (Po0.05; Figure 2f). The effects of BH3I-2′ on cell death were substantially less pronounced in bax −/− cells compared to WT cells (Po0.05 for both apoptosis and necrosis; Figure 2f).
Loss of bax reduces Ca 2+ release from the ER. The effects of Bax on intracellular Ca 2+ homoeostasis were investigated in PACs that had their ER stores depleted with the SERCA blocker thapsigargin under simultaneous inhibition of IP 3 Rs (by 20 mM caffeine) and RyRs (by 10 μM ruthenium red). 16,28 Average traces ( Figure 3a) and average areas of responses (Figure 3a, inset) demonstrate that [Ca 2+ ] i elevation caused by the ER depletion was significantly smaller in bax −/− cells than in WT cells (Po0.001). Inhibition of Ca 2+ extrusion by 1 mM La 3+ preserved the difference in size of the thapsigargin-induced Ca 2+ release between WT and bax −/− cells (Po0.01; Figure 3b) indicating that this difference was not due to enhanced cytosolic Ca 2+ extrusion in bax −/− cells. Further, in two-photon-permeabilised PACs, thapsigargin also released less Ca 2+ from the ER of bax −/− than from WT cells CALPs inhibit Ca 2+ entry in pancreatic acinar cells. Calcium-like peptides (CALPs) are short peptides designed by inversion of the hydrophobic pattern of EF hand Ca 2+ -binding sites, 29 which generates molecules of a complementary surface contour, capable of interacting with the sequences of interest. 30,31 Cell permeable CALP-1 and CALP-3 can functionally mimic increased [Ca 2+ ] i by modulating the activity of calmodulin, Ca 2+ channels and pumps. 32 Emptying the ER Ca 2+ store with thapsigargin leads to opening of store-operated Ca 2+ entry (SOCE) channels in the plasma membrane and influx of Ca 2+ into the cytosol. 33 (Figures 4a-c). In the absence of CALPs, the first and second application of 10 mM Ca 2+ induced similar [Ca 2+ ] i elevations ( Figure 4a). Addition of low concentrations (10 μM) of CALP-1 (Figure 4b) or CALP-3 ( Figure 4c) before the second application of 10 mM Ca 2+ substantially inhibited Ca 2+ influx, as shown by markedly reduced amplitudes (Po0.001; Figure 4d).
CALPs reduce excessive cytosolic Ca 2+ signals and necrosis elicited by BH3I-2′. Excessive cytosolic Ca 2+ signals can trigger necrosis, associated with release of intracellular content followed by inflammation. 34 Necrosis is especially dangerous for the pancreas, where released activated digestive enzymes cause a serious threat for the integrity of the tissue. 35,36 Therefore apoptosis, executed by cells in a controlled manner, is a more favourable pathway for killing cells. Here, CALPs were applied in an attempt to reduce elevations in [Ca 2+ ] i and necrosis caused by BH3 mimetics. Preincubation of PACs either with 100 μM CALP-1 or CALP-3 led to a decrease in the average amplitude of the Ca 2+ signals occurring in response to BH3I-2′ (Po0.001 and Po0.05, respectively) compared to the control (average traces: Figure 5a; quantitative analysis: Figure 5b). This was due to a shift in the pattern of cytosolic Ca 2+ signals elicited by 5 μM BH3I-2′ (Figure 5c). The consequences of the CALP-mediated reductions in the cytosolic Ca 2+ signal generation are reflected by the cell death pattern (Figures 5d and e). In the untreated control the majority of cells were alive (low levels of apoptosis and  CALPs reduce excessive cytosolic Ca 2+ responses and necrosis elicited by gossypol. It has been reported that gossypol is also capable of mobilising intracellular Ca 2+ . 37,38 Similarly to BH3I-2′ and HA14-1, in the absence of extracellular Ca 2+ , bax −/− cells did not respond to gossypol (Figure 6a, black trace). In WT cells in the presence of external Ca 2+ , 20 μM gossypol alone caused an increase in [Ca 2+ ] i reaching a sustained elevated plateau (Figure 6a). Preincubation either with 100 μM CALP-1 or with CALP-3 led to a decrease in the magnitude of the Ca 2+ responses to gossypol (Figures 6a and b; Po0.001); CALP-1 markedly decreased the proportion of cells that responded at all (Figure 6c). The effects of CALP-1 and CALP-3 on gossypolinduced cell death (Figures 6d and e) were very similar to those for BH3I-2′ (Figure 5d). In the presence of CALP-1, necrosis was markedly decreased (Po0.05) and the proportion of apoptotic cells was increased (Po0.05). CALP-3 reduced gossypol-induced necrosis (Po0.05) and increased the fraction of live cells (Po0.05).
CALPs do not markedly affect physiological Ca 2+ signals in acinar cells. The ability of CALPs to reduce excessive pathological Ca 2+ signal generation is potentially very useful but given the crucial role of Ca 2+ in the regulation of a wide variety of intracellular processes, it was important to test whether CALPs also affect physiological Ca 2+ signalling. Nanomolar doses of ACh trigger enzyme secretion in PACs, an event controlled by Ca 2+ signals. 24 Here, 50 nM ACh induced repetitive [Ca 2+ ] i oscillations (Figure 7a). Preincubation with either 100 μM CALP-1 (Figure 7b The effects of CALPs on cell death are not limited to BH3mediated necrosis. As the primary mechanism by which CALPs exert their effects is the pre-activation of EF hand Ca 2+ -binding sites on a wide variety of intracellular targets 29 and inhibition of Ca 2+ entry (Figures 4a-d), it is unlikely that CALP-mediated reduction in excessive Ca 2+ signals and necrosis is limited only to the effects induced by BH3 mimetics. Menadione was previously shown to induce Ca 2+ -dependent cell death in PACs. 40 Here, 5 μM menadione was combined with a high physiological concentration of ACh (100 nM) in order to induce Ca 2+ -dependent cell death of similar proportions of apoptosis and necrosis (25.1 ± 5.0% and 20.7 ± 1.7%, respectively; Figure 7h; sample images in Figure 7i) to those caused by BH3I-2′ (Figure 5d) or gossypol (Figure 6d). In this protocol 100 μM CALP-1 decreased cell necrosis to control levels (Po0.05) while increasing slightly the proportions of both live and apoptotic cells; whereas 100 μM CALP-3 substantially reduced necrosis (Po0.05), leading to an increase in the number of live cells (Po0.05). The apoptotic fraction was essentially unaffected by CALP-3.

Discussion
It is now commonly accepted that Bcl-2 proteins are distributed in different cell compartments, not only at the outer mitochondrial membrane, but also in the cytosol, at the nuclear envelope and the ER. [41][42][43] An increasing number of reports indicates that Bcl-2 itself may either directly or indirectly modulate intracellular Ca 2+ fluxes such as: (1) Ca 2+ release from the ER by binding to IP 3 Rs or RyRs; [44][45][46] (2) Ca 2+ reuptake into the ER by SERCA; 47 (3) cellular Ca 2+ extrusion by PMCA (plasma membrane Ca 2+ ATPase); 48 (4) and mitochondrial Ca 2+ load. 49 We have previously demonstrated that the BH3 mimetics BH3I-2′ and HA14-1 induce Ca 2+ release from the intracellular stores in mouse PACs and in the rat pancreatic cancer cell line AR42J. 16 This Ca 2+ release was not completely blocked by inhibition of IP 3 Rs and RyRs indicating involvement of other Ca 2+ channels. We have also shown that BH3I-2′ and HA14-1 displace Bax from Bcl-2 and Bcl-xL. 16 Here we provide new insights into the mechanism of this phenomenon by demonstrating that loss of Bax, but not Bak or Bcl-2, substantially inhibited Ca 2+ responses to BH3I-2′ (Figure 1a), HA14-1 ( Figure 1b) and gossypol (Figure 6a) in the absence of extracellular Ca 2+ . HA14-1 and gossypol were suggested to have off-target effects, 50,51 and recently a novel Bcl-2 inhibitor, ABT-199, was found not to affect intracellular Ca 2+ signalling. 52 In our experiments not only the responses to HA14-1 and BH3I-2′ were markedly inhibited in bax −/− PACs (Figures 1a and b) but also BH3I-2′-elicited apoptosis was completely abolished (Figure 2f), indicating that these effects were dependent on Bax. Low levels of necrosis in bax −/− cells may indeed suggest a Bax-independent killing component in the mechanism of BH3I-2′ action. Importantly, BH3I-2′induced cell death in PACs appears to be mediated by Ca 2+ as it was blocked by chelation of intracellular Ca 2+ with BAPTA ( Figure 2f).
Our results appear to link Bax with the process of passive Ca 2+ leak from the ER. The leak, unmasked by thapsigargin, developed more slowly in bax −/− cells compared to WT cells (Figures 3c and e). Currently there is an on-going dispute about the nature of the channels mediating this process. As Bax can localise to the ER membranes 53 and given that its structure resembles that of pore-forming bacterial toxins, 54,55 Bax itself might be able to form a channel permeable to ions and thus contribute to the passive Ca 2+ leak. Two out of seven α-helices of Bax were suggested to function as pore-forming domains, but possibly more than one monomer of Bax would need to form a functional ion-permeable pore. 56,57 Alternatively, as opposed to acting as a channel itself, Bax could directly or indirectly regulate the ER Ca 2+ leak mediated by other channels. Consequently, it could be speculated that application of a BH3 mimetic increases the unbound fraction of Bax, which activates Ca 2+ channels in the ER, thus potentiating the leak (Figure 8). The leak may be further amplified by IP 3 Rs and RyRs.
The reduced cytosolic Ca 2+ release in response to thapsigargin treatment in bax −/− cells could, in principle, be explained by a diminished resting ER Ca 2+ content. However, as ACh-induced oscillations were essentially unaffected in bax −/− cells (Figures 3h-j) 63 Further, our results show that the BH3 mimetics, BH3I-2′ and gossypol not only induce apoptosis in PACs but also cause substantial levels of necrosis (Figures 5d and 6d). Cell death in PACs was a downstream effect of large cytosolic Ca 2+ signals triggered by application of BH3 mimetics, as (1) Ca 2+ chelation by BAPTA completely abolished both apoptosis and necrosis induced by BH3I-2′ and (2) reduction of Ca 2+ signals elicited by CALP-1 or CALP-3 (Figures 5a and 6a) led to a significant decrease in BH3I-2′-as well as gossypol-induced necrosis (Figures 5d and 6d). CALP-1 was shown to be particularly promising, as it effectively shifted the cell death mode from necrosis to apoptosis. In contrast, CALP-3 substantially reduced BH3 mimetic-induced necrosis while increasing the fraction of live cells. Both CALPs were capable of inhibiting Ca 2+ entry (Figures 4a-d), even though they differ in structure. CALP-1 has been designed to interact with EF hand Ca 2+ -binding sites of troponin C, whereas CALP-3 is complementary to the EF hand motif of calmodulin. 32 It is therefore likely that the two CALPs might preferentially interact with different intracellular targets, for example, Ca 2+ channels regulated by Ca 2+ versus calmodulin, abundant in PACs. 64 This could explain different effects on BH3 mimetic-induced cell death. Previously we demonstrated that CALP-3 efficiently inhibited Ca 2+ responses induced by ethanol in PACs and we attributed this effect to activation of calmodulin. 65 The exact assessment of binding affinities of CALPs to intracellular targets exceeds the scope of this work. It is clear, however, that the effects of CALPs on cell death are not limited to BH3 mimetics, as necrosis triggered by ACh and menadione was also inhibited in the presence of CALPs (Figure 7h).
Considering the vast spectrum of targets affected by CALPs, it was important to test whether application of these compounds would affect not only pathological Ca 2+ elevations but also physiological Ca 2+ signals. In PACs, nanomolar doses of ACh and picomolar concentrations of CCK are known to trigger Ca 2+ oscillations, which regulate pancreatic enzyme secretion. 24 In our experiments neither CALP-1 nor CALP-3 markedly affected responses to ACh (Figures 7a-c), whereas in the presence of CALP-3, CCK-elicited Ca 2+ oscillations were slightly reduced but not BH3 mimetic-elicited Ca 2+ signals are dependent on Bax PE Ferdek et al completely inhibited (Figures 7d-g). The previously demonstrated inhibition of Ca 2+ influx into PACs by the CRAC channel inhibitor GSK-7975A also affected only pathological Ca 2+ elevations but not the oscillations induced by physiological concentrations of ACh or CCK. 66 Although it can be triggered by sustained elevations of [Ca 2+ ] i , necrosis generally lacks intracellular regulatory mechanisms. This severely limits possible therapeutic approaches in diseases where necrosis plays a major role, such as acute pancreatitis. Inhibition of SOCE has already been demonstrated to decrease cell death in vitro 66 and in vivo. 67 This study provides another proof-of-principle demonstration that even non-specific attenuation of intracellular Ca 2+ fluxes can dampen pathophysiological Ca 2+ responses and thus result in necrosis inhibition (CALP-3) or a significant shift in cell death towards apoptosis (CALP-1).
Our results indicate that application of CALPs, especially CALP-1, could improve the outcome of BH3 mimetic-based therapies. Although the pharmacokinetic properties of CALPs have not yet been studied in detail, it is likely that CALPs share characteristics of other short peptides, such as limited cell permeability and metabolic instability. 68 However, Isolation of pancreatic acinar cells. PAC isolation and most of the experimental work was carried out in NaHEPES buffer. Unless otherwise stated, NaHEPES was supplemented with 1 mM Ca 2+ . Freshly isolated pancreas was washed twice in NaHEPES, injected with collagenase (200 µ/ml, in NaHEPES) and subsequently incubated at 37°C for 15 min in the collagenase solution to allow digestion of the tissue. After incubation, the pancreas was broken down by pipetting, suspended in NaHEPES, spun (1 min, 0.2 × g), resuspended in NaHEPES and spun again. Finally, isolated PACs were suspended in NaHEPES and loaded with a Ca 2+ sensitive dye as described below.
Cytosolic Ca 2+ measurements. Isolated PACs were loaded at room temperature with one of the Ca 2+ indicators: 5 μM Fluo-4 AM for 30 min or 10 μM Fura-2 AM for 1 h. After the incubation the cells were resuspended in fresh NaHEPES and used for experiments at room temperature in a flow chamber perfused with NaHEPES-based extracellular solution. Experiments with Fluo-4 AM were performed using the Leica confocal microscope TCS SPE (Leica Microsystems, Milton Keynes, UK): × 63 oil objective, excitation 488 nm, emission 500-600 nm. Static images were taken at 512 × 512 pixel resolution and series of images were recorded at 256 × 256 pixel resolution, two consecutive frames were averaged. Fluorescence signals were plotted as F/F 0 , where F 0 was an averaged signal from the first ten baseline images. Experiments with Fura-2 AM were performed using the Nikon Diaphot 200 imaging system (Nikon, Kingston, UK): excitation at 365 and 385 nm, emission at 510 nm. The signals were plotted as 365/385 nm ratio or the ratio normalized to baseline values (R/R 0 ).
ER Ca 2+ measurements. PACs were loaded with 5 μM Fluo-5N AM in NaHEPES for 45 min at 37°C. After the incubation, the cells were resuspended in fresh NaHEPES and used for experiments at room temperature in a flow chamber perfused with KHEPES-based intracellular solution. The cells were permeabilised with two-photon laser beam (720-750 nm), 70 which was applied at a small area of the plasma membrane of a PAC. The experiments were performed using the Leica two-photon confocal microscope TCS SP5 with the same settings as for Fluo-4 AM (see above).
Cell death assay. Cell death assay was performed using Annexin V-FITC Apoptosis Detection Kit (Sigma-Aldrich) according to a modified manufacturer′s protocol. PACs were isolated as described above and divided equally into four experimental groups. Two samples were pretreated with 100 μM CALP-1 or 100 μM CALP-3 for 15 min at room temperature; the remaining two were incubated without pretreatment. Then cell death was induced for 30 min in both samples containing CALPs and in one untreated sample. Depending on the experiment, either one of BH3 mimetics was used (5 μM BH3I-2′ or 5 μM gossypol) or 5 μM menadione together with 100 nM ACh. Analogical experiments were performed with 15 min pretreatment with BAPTA instead of CALPs. The control samples were left untreated. Fifteen minutes before the end of the incubation, annexin V-FITC and propidium iodide were added to all samples.  Figure 8 Schematic illustration of proposed effects of BH3 mimetics and CALPs on intracellular Ca 2+ homoeostasis. BH3 mimetics (such as BH3I-2′ or gossypol) disrupt the interaction between anti-apoptotic Bcl-2 and pro-apoptotic Bax. Unbound Bax may either form channels in the ER membranes or interact with the ER leak channel triggering Ca 2+ release from the ER stores. Ca 2+ release is amplified by IP 3 Rs and RyRs. These receptors can be inhibited by caffeine and ruthenium red (RR), respectively. CALPs affect multiple targets via interaction with their EF hand motifs, for example, inhibit SOCE and activate calmodulin (CaM) for detection of necrotic cells (excitation: 535 nm, emission: 585-705 nm). Multiple pictures (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) per treatment group were taken; live, apoptotic and necrotic cells were counted in each treatment group.
Statistical analysis. For quantitative analysis of Ca 2+ responses, areas under individual traces were calculated according to the formula: P F=F 0 À F 0 ð Þ Dt, where F is the recorded fluorescence (or ratio for Fura-2), F 0 is the baseline fluorescence (or ratio) and Δttime interval. Obtained values were then averaged and presented as bar charts with S.E.M. The Student′s t-test was applied for statistical comparison. The significance threshold was set at 0.05 and the range was indicated by asterisks (*Po0.05, **Po0.01, ***Po0.001). Where applicable, N indicates the number of individual experiments, whereas nindividual cells.
For cell death assays, three independent experiments were performed for each treatment group; average values and S.E.M. were calculated and the results presented as bar charts. Statistical analysis was performed using the non-parametric Mann-Whitney U-test with the significance threshold set at 0.05.