Introduction

The anti-apoptotic B-cell lymphoma 2 (Bcl-2)-family members, like Bcl-2 and Bcl-XL, are critically involved in maintaining mitochondrial integrity by scavenging and inhibiting pro-apoptotic Bcl-2-family members, such as Bax and Bak1. This interaction occurs via the hydrophobic cleft, formed by the Bcl-2 homology (BH) domain 1, 2, and 3, of anti-apoptotic Bcl-2 proteins and the BH3 domain of the pro-apoptotic family members. In addition to the hydrophobic cleft the last most N-terminal BH domain, the BH4 domain, is also critical for the anti-apoptotic properties of Bcl-22,3,4.

Besides neutralizing pro-apoptotic Bcl-2-family members, anti-apoptotic Bcl-2 proteins also emerged as critical modulators of intracellular Ca2+ signaling5,6,7. In particular, Bcl-2 is present at the membranes of the endoplasmic reticulum (ER), the main intracellular Ca2+-store8,9. At the ER, Bcl-2 directly inhibits the inositol 1,4,5-trisphosphate (IP3) receptor (IP3R)3,4, a ubiquitously expressed IP3-gated ER-located Ca2+ release channel10. Bcl-2 forms a protein complex with the IP3R by targeting the central modulatory domain of the channel. As such, Bcl-2 limits excessive Ca2+-release that may trigger apoptotic cell death. Importantly, the BH4 domain of Bcl-2 by itself is sufficient for inhibiting IP3R-mediated Ca2+ release, protecting cells against Ca2+-induced mitochondrial outer membrane permeabilisation (MOMP) and subsequent apoptosis3. Although, the BH4 domain of Bcl-2 and that of Bcl-XL are very similar in size, sequence and structure, the latter is unable to bind to and inhibit IP3R-mediated Ca2+ release2. The difference in IP3R-inhibitory properties between the BH4 domain of Bcl-2 and Bcl-XL could be largely attributed to a single amino acid change. Substituting Lys17 for an Asp residue in the BH4 domain of Bcl-2 abrogated its IP3R-inhibitory function, while changing Asp11 into a Lys residue in BH4 domain of Bcl-XL enhanced its IP3R-inhibitory function2.

In addition to the IP3R, we recently showed that anti-apoptotic Bcl-2 proteins also bind to and inhibit ryanodine receptors (RyR)7,11. RyRs form a second class of the ER-located Ca2+ release channels, mainly expressed in specialized tissues such as the heart, brain, muscle, but also in the pancreas12,13. In these tissues, RyR-mediated Ca2+ release is involved in muscle contraction, memory formation and secretion of digestive enzymes13,14. Endogenous activation of RyRs occurs via Ca2+-induced Ca2+ release or via NAADP and/or cyclic-ADP ribose production15. The interaction between RyRs and Bcl-2 is mediated by the BH4 domain of the latter. Binding of the BH4 domain of Bcl-2 to the RyR is sufficient for inhibiting RyR-mediated Ca2+ release7. In contrast to what was shown for the IP3R, the BH4 domain of Bcl-XL also binds to and inhibits RyRs11. In our previous work, we demonstrated that Bcl-2 and Bcl-XL (via their BH4 domains) could modulate pharmacological activation of RyRs by caffeine7,11. However, prior to the present study it was unclear whether RyR responses to physiological agonists can also be regulated by the BH4 domains.

Furthermore, excessive RyR-mediated Ca2+ release is linked to several pathologies and is a hallmark for the onset of several diseases of the brain, heart, muscle, and pancreas, such as Alzheimerā€™s disease, catecholaminergic polymorphic ventricular tachycardia, malignant hyperthermia, and acute pancreatitis (AP), respectively13,16. The BH4 domains of Bcl-2 proteins, particularly Bcl-XL, have been shown to have protective anti-apoptotic effects in several of these tissues17,18,19,20,21. However, a link between cell protection and RyR inhibition by these BH4 domains has not yet been reported. Therefore, it remains unknown whether the BH4 domains of Bcl-2 and Bcl-XL could be exploited as inhibitors of RyRs in diseases associated with excessive RyR activity in orderĀ to dampen disease burden.

To address this, we have chosen pancreatic acinar cells (PACs), in which physiological and pathophysiological intracellular Ca2+ signaling has been extensively studied22,23,24,25,26. Importantly, in PACs physiologically relevant agonists can be utilized to trigger IP3R or RyR-dependent Ca2+ releases, allowing to compare the effects of the BH4 domain peptides on both channels26. Exposing PACs to low nanomolar concentrations of acetylcholine (ACh) is known to trigger primarily IP3R-mediated Ca2+ oscillations via activation of phospholipase C leading to IP3 production. In contrast, cholecystokinin (CCK) mainly triggers RyR-mediated Ca2+ release via the production of NAADP and/or cyclic-ADP ribose26. Finally, bile acids, such as taurolithocholic acid 3-sulfate (TLC-S), induce IP3R and RyR-mediated Ca2+ overload22, leading to premature activation of digestive enzymes and subsequent necrosis, which is an initiating event for AP26. It has been demonstrated that noxious Ca2+ signals induced by TLC-S can be reduced by inhibition of IP3Rs or RyRs, using caffeine or dantrolene, respectively24,25. Importantly, these approaches not only protected PACs against necrosis but also ameliorated the severity of AP in mouse models. Although it is well known that intracellular Ca2+ signaling plays important roles in the development of AP, currently no effective treatment exists for this disease. Here, we test whether the BH4 domains of Bcl-2/Bcl-XL proteins can be utilized to suppress IP3R or RyR hyperactivity associated with the onset of AP.

In this study, by employing primary isolated mouse PACs, we show that peptides derived from the BH4 domains of Bcl-2 and Bcl-XL inhibit both physiological and pathophysiological RyR-mediated Ca2+ release, as well as protect PACs from TLC-S-induced necrosis. These findings show for the first time that the BH4 domains of Bcl-2 and Bcl-XL can be applied as innovative peptide tools to limit excessive RyR-mediated Ca2+ release associated with the pathology of AP. As such, BH4-domain-based molecules and mimetics may originate a novel group of therapeutics with the potential application in diseases associated with excessive Ca2+ release.

Results

The BH4 domains of Bcl-2 and Bcl-XL inhibit RyR-mediated Ca2+ release in PACs

In the first set of experiments, we aimed to assess the ability of the BH4 domains of Bcl-2 and Bcl-XL to inhibit IP3R or RyR-mediated Ca2+ release in isolated PACs. IP3R or RyR-mediated Ca2+ release was triggered by the physiological activators ACh or CCK, respectively. The sequences used to design the peptides corresponding to the BH4 domain of Bcl-2 and Bcl-XL with their predicted Ī±-helical properties is depicted in Fig.Ā 1a. In PACs, low nanomolar concentrations of ACh and low picomolar concentrations of CCK generate long lasting Ca2+ oscillations, which vary in terms of amplitude and frequency between the cells. Therefore, in single PAC, we compared and quantified the Ca2+ oscillations induced by each agonist before and after addition of the BH4 domain peptides or a control peptide. We first assessed the effects of the BH4 domains of Bcl-2 and Bcl-XL on CCK-induced-RyR-mediated Ca2+ release (representative traces shown in Fig.Ā 1bā€“d). Ca2+ oscillations were evoked by 5ā€‰pM CCK and recorded for 5ā€‰min. Then 50ā€‰ĀµM control peptide (Fig.Ā 1b), the BH4 domain of Bcl-2 (Fig.Ā 1c) or the BH4 domain of Bcl-XL (Fig.Ā 1d) was added and Ca2+ responses were measured for another 10ā€‰min, in the continuous presence of CCK. Quantitative analysis of the responses was performed by comparing the area under the curve (AUC). These values were further divided by the length of the recording before (5ā€‰min) and after (10ā€‰min) addition of the peptides, resulting in the response area normalized per unit of time (AUC/sec). These experiments show that the BH4 domains of Bcl-2 and Bcl-XL dampen RyR-mediated Ca2+ oscillations evoked by the physiological stimulus CCK (Fig.Ā 1e) and are in line with our previous findings in dissociated hippocampal neurons and RyR-overexpression models7,11. In addition, this also indicates that the BH4 domain peptides are taken up by the PACs and are capable of inhibiting physiological RyR-mediated Ca2+ release.

Fig. 1: The BH4 domains of Bcl-2 and Bcl-XL inhibit RyR-mediated Ca2+ release in PACs.
figure 1

a Linear representation of the Bcl-2/Bcl-XL protein together with the sequences used to design the BH4 domain peptides originating from Bcl-2 and Bcl-XL and the predicted secondary structure using the PSIPRED secondary structure predictor (http://bioinf.cs.ucl.ac.uk/psipred/). bā€“d Typical traces of the Fluo-4AM single-cell Ca2+ measurements performed in mouse PACs. At the start of each experiment cells were perfused with NaHEPES containing DMSO (vehicle). After establishing the baseline (200ā€‰s), CCK (5ā€‰pM) was added and RyR-mediated Ca2+ oscillations were measured for 5ā€‰min. Then, 50ā€‰ĀµM of either the control peptide (b), the BH4 domain of Bcl-2 (c), or the BH4 domain of Bcl-XL (d) were added for 10ā€‰min, in the continuous presence of 5ā€‰pM CCK. e Quantification of the experiments: The area under the curve per second (AUC/sec) of the CCK-induced Ca2+ release after peptide addition (10ā€‰min recording) was compared to the AUC/sec before peptide addition (5ā€‰min recording). Each data point represents the measurement of an individual cell. The average valuesā€‰Ā±ā€‰SEM are shown (P-value <0.0001). At least three independent experiments were performed per condition (Nā€‰ā‰„ā€‰3). For each condition at least 35 cells were recorded

The BH4 domain of Bcl-2 does not inhibit IP3R-induced Ca2+release in PACs

Next, we assessed the effects of the BH4 domain peptides on IP3R-mediated Ca2+ release. Analogical to CCK experiments single-cell Ca2+ measurements were performed using ACh (75ā€‰nM) in order to trigger IP3R-mediated Ca2+ oscillations (Fig.Ā 2). As expected, subsequent addition of the control peptide or the BH4 domain of Bcl-XL did not result in an inhibition of ACh-induced-IP3R-mediated Ca2+ release (Fig.Ā 2a, c, d). Surprisingly, also the BH4 domain of Bcl-2 was unable to inhibit IP3R-mediated Ca2+ release in PACs (Fig.Ā 2b, d). These results differ compared to what we showed previously in cultured or permeabilized cells where the BH4 domain of Bcl-2 inhibited IP3R-mediated Ca2+ release via its BH4 domain4,27,28,29.

Fig. 2: The BH4 domain of Bcl-2 does not inhibit IP3R-induced Ca2+ release in PACs.
figure 2

aā€“c Typical traces of the Fluo-4AM single-cell Ca2+ measurements performed in mouse PACs. At the start of each experiment cells were perfused with NaHEPES containing DMSO (vehicle). After establishing the baseline (200ā€‰s), ACh (75ā€‰nM) was added and IP3R-mediated oscillations were measured for 5ā€‰min. Then, 50ā€‰ĀµM of either the control peptide (a), the BH4 domain of Bcl-2 (b), or the BH4 domain of Bcl-XL (c) were added for 10ā€‰min, in the continuous presence of 75ā€‰nM ACh. d Quantification of the experiments: The area under the curve per second (AUC/sec) of the ACh-induced Ca2+ release after peptide addition (10ā€‰min recording) was compared to the AUC/sec before peptide addition (5ā€‰min recording). Each data point represents the measurement of an individual cell. The average valuesā€‰Ā±ā€‰SEM are shown (P-value 0.3467). At least three independent experiments were performed per condition (Nā€‰ā‰„ā€‰3). For each condition at least 14 cells were recorded

The BH4 domain of Bcl-2 inhibits IP3R-mediated Ca2+ release in Bcl-2 knock out PACs

The lack of inhibition of IP3R-mediated Ca2+ release upon application of the BH4 domain of Bcl-2 could be due to the fact that endogenous Bcl-2 in wild-type (WT) PACs is already associated with IP3Rs, preventing its regulation by exogenously added BH4 domain of Bcl-2. To address this, we performed single-cell Ca2+ measurements in PACs isolated from Bcl-2 knock out (KO) mice, devoid of endogenous Bcl-2.

In order to obtain repeatable, long lasting Ca2+ oscillations in Bcl-2 KO PACs, similar in frequency and amplitude to those evoked in WT PACs, high-nanomolar concentrations (200ā€‰nM) of ACh were required. Consistent with our previous findings, addition of 50ā€‰ĀµM control peptide or the BH4 domain of Bcl-XL (Fig.Ā 3a, c, d) did not alter IP3R-mediated Ca2+ releases. However, in Bcl-2 KO PACs, the BH4 domain of Bcl-2 does significantly inhibit ACh-induced-IP3R-mediated Ca2+ release (Fig.Ā 3b, d) compared to control peptide or the BH4 domain of Bcl-XL.

Fig. 3: The BH4 domain of Bcl-2 inhibits IP3R-mediated Ca2+ release in Bcl-2 KO PACs.
figure 3

aā€“c Typical traces of the Fluo-4AM single-cell Ca2+ measurements performed in Bcl-2 KO mouse PACs. At the start of each experiment cells were perfused with NaHEPES containing DMSO (vehicle, Fig.Ā 3a). After establishing the baseline (200ā€‰s), ACh (200ā€‰nM) was added and IP3R-mediated Ca2+ oscillations were measured for 5ā€‰min. Then, 50ā€‰ĀµM of either the control peptide (a), the BH4 domain of Bcl-2 (b) or the BH4 domain of Bcl-XL (c) were added for 10ā€‰min, in the continuous presence of 200ā€‰nM ACh. d Quantification of the experiments: The area under the curve per second (AUC/sec) of the ACh-induced Ca2+ release after peptide addition (10ā€‰min recording) was compared to the AUC/sec before peptide addition (5ā€‰min recording). Each data point represents the measurement of an individual cell. The average valuesā€‰Ā±ā€‰SEM are shown (P-value <0.0001). At least three independent experiments were performed per condition (Nā€‰ā‰„ā€‰3). For each conditionĀ at least 35 cells were recorded

The inhibitory properties of the BH4 domain peptides were also tested on CCK-evoked oscillations in Bcl-2 KO PACs (Fig.Ā 4aā€“d). In the Bcl-2 KO PACs, both the BH4 domains of Bcl-2 and Bcl-XL showed a similar response inhibition (Ā±50% AUC/sec) as seen in the WT PACs. Comparing WT to Bcl-2 KO PACs (Figs.Ā 1, 4) shows that the BH4 domains did not confer additional inhibition of CCK-induced Ca2+ release in the absence of Bcl-2. Taken together these functional experiments suggest that in WT PACs endogenous Bcl-2 mainly occupies/regulates IP3Rs and to a lesser extent RyRs. This potentially leaves more RyR channels available for binding to exogenously added BH4 domain of Bcl-2, resulting in efficient inhibition of RyRs but not IP3Rs in WT PACs. Of note, just like in the experiments utilizing ACh in Bcl-2 KO PACs (Fig.Ā 3), higher concentrations of CCK (10ā€‰pM) were needed to trigger RyR-mediated Ca2+ oscillations of comparable frequency and amplitude to those recorded in WT PACs (Fig.Ā 4).

Fig. 4: The BH4 domains of Bcl-2 and Bcl-XL inhibit RyR-mediated Ca2+ release in Bcl-2 KO PACs.
figure 4

aā€“c Typical traces of the Fluo-4AM single-cell Ca2+ measurements performed in Bcl-2 KO mouse PACs. At the start of each experiment cells were perfused with NaHEPES containing DMSO (vehicle). After establishing the baseline (200ā€‰s), CCK (10ā€‰pM) was added and RyR-mediated Ca2+ oscillations were measured for 5ā€‰min. Then, 50ā€‰ĀµM of either the control peptide (a), the BH4 domain of Bcl-2 (b), or the BH4 domain of Bcl-XL (c) were added for 10ā€‰min, in the continuous presence of 10ā€‰pM CCK. d Quantification of the experiments: The area under the curve per second (AUC/sec) of the CCK-induced Ca2+ release after peptide addition (10ā€‰min recording) was compared to the AUC/sec before peptide addition (5ā€‰min recording). Each data point represents the measurement of an individual cell. The average valuesā€‰Ā±ā€‰SEM are shown (P-value <0.0001). At least three independent experiments were performed per condition (Nā€‰ā‰„ā€‰3). For each condition at least 40 cells were recorded

The BH4 domains of Bcl-2 and Bcl-XL inhibit pathological TLC-S induced Ca2+ releases in PACs

Combining the above results from ACh and CCK-induced Ca2+ release, we can conclude that in WT PACs the inhibitory effects of the BH4 domains of Bcl-2 and Bcl-XL on Ca2+ release from the ER is due to the inhibition of RyR- but not IP3R-mediated Ca2+ release. This has prompted us to investigate whether the BH4 domains of Bcl-2 and Bcl-XL could also suppress pathological RyR-mediated Ca2+ releases.

The bile acid TLC-S has been described to mediate excessive, pathological intracellular Ca2+ release in PACs and both IP3Rs and RyRs have been proposed to be involved in this process22. Here we aimed to test whether the BH4 domains of Bcl-2 and Bcl-XL could also inhibit pathophysiological TLC-S-induced Ca2+ release (Fig.Ā 5). WT PACs were pre-treated for 5ā€‰min with either DMSO (vehicle) or 50ā€‰ĀµM of the peptides, then TLC-S (200ā€‰ĀµM) was added and Ca2+ responses were measured (Fig.Ā 5a). Detailed comparison of the response areas recorded for the individual cells (AUC) per unit of time (AUC/sec) revealed that both BH4 domains inhibited pathological TLC-S-induced Ca2+ responses compared to the control peptide and DMSO vehicle (Fig.Ā 5b). Both BH4 domains inhibit RyR but not IP3R-mediated Ca2+ release (Figs.Ā 1, 2) and similarly suppress TLC-S-induced Ca2+ release in WT PACs (Fig.Ā 5a, b). Collectively, these data indicate that the BH4 domains suppress pathophysiological Ca2+ signaling by acting as modulators of the excessive RyR activity.

Fig. 5: The BH4 domains of Bcl-2 and Bcl-XL inhibit pathological TLC-S-induced Ca2+ release and necrosis in PACs by inhibiting RyR activity.
figure 5

a Average traces of Fluo-4AM single-cell Ca2+ measurements performed in PACs. The cells were pre-treated for 5ā€‰min with either DMSO (vehicle; red) or 50ā€‰ĀµM of the peptides: control (blue), the BH4 domain of Bcl-2 (black), and the BH4 domain of Bcl-XL (pink). Then, TLC-S (200ā€‰ĀµM) was added and pathological intracellular Ca2+ release was measured for 5ā€‰min, in the continuous presence of either vehicle or the peptides. Average tracesā€‰Ā±ā€‰SEM of all performed experiments are shown. b Quantification of the experiments: Analysis of area under the curve per second (AUC/sec) of the TLC-S-induced Ca2+ releases, in the presence or absence of the BH4 domains or control peptide. Each data point represents the measurement of an individual cell, and the average valuesā€‰Ā±ā€‰SEM are shown (P-value <0.0001). At least three independent experiments were performed per condition (Nā€‰ā‰„ā€‰3). For each condition at least 35 cells were recorded. c Quantification of the necrosis assay: Isolated PACs were treated with DMSO (vehicle) or 50ā€‰ĀµM of the indicated peptides. 15ā€‰min later, TLC-S was added (200ā€‰ĀµM final concentration) to induce necrotic cell death. Propidium iodide staining (necrosis indicator) was assessed 2ā€‰h after TLC-S addition. The negative control was treated with DMSO (vehicle) only. In each experimental repeat (Nā€‰=ā€‰4) at least 15 images were taken per treatment group (nā€‰ā‰„ā€‰100 cells/treatment/experiment). The percentages of propidium iodide-positive necrotic cells were assessed for each experimental condition. Each data point represents an independent repeat of the experiment, and the average valuesā€‰Ā±ā€‰SEM are shown (P-values <0.001)

The BH4 domains of Bcl-2 and Bcl-XL inhibit TLC-S-induced necrosis in PACs by inhibiting excessive RyR activity

It has been well established that TLC-S-induced Ca2+ release in PACs leads to necrosis which is an underlying cause of AP23. In addition, pharmacological inhibition of either RyRs or IP3Rs, using dantrolene or caffeine, respectively, not only attenuated TLC-S-induced necrosis, but also reduced the severity of AP evoked in mouse models of this disease24,25.

Since the BH4 domains of Bcl-2 and Bcl-XL proved to be the effective inhibitors of TLC-S-induced Ca2+ release via inhibiting RyRs (Fig.Ā 5a, b), we aimed to determine whether these peptides could protect PACs against the cytotoxic effects of TLC-S. In the PACs, 50ā€‰ĀµM BH4 domain of Bcl-2 or Bcl-XL almost completely inhibited necrotic cell death induced by 2ā€‰h treatment with 200ā€‰ĀµM TLC-S, compared to the control peptide or the vehicle control (Fig.Ā 5c and S1). This demonstrates a proof-of-principle that the BH4 domains of Bcl-2 and Bcl-XL could be applied as a basis for developing therapeutic tools to decrease the necrotic burden of bile acids on PACs by preventing the excessive RyR-mediated Ca2+ release.

Discussion

The main conclusions of this paper are that (i) the BH4 domains of Bcl-2 and Bcl-XL suppress physiological RyR-mediated Ca2+ release in isolated PACs exposed to the hormone CCK; (ii) the Ca2+-signaling-inhibitory properties of these BH4 domains can be utilized to dampen pathophysiological RyR-mediated cytosolic Ca2+ overload associated with AP, protecting PACs against necrosis. Furthermore, these results indicate for the first time a potential for the therapeutic application of BH4 domains, or therapeutic tools derived from them, as suppressors of (excessive) RyR-mediated Ca2+ release in the treatment of AP.

In our previous work, we have already shown that Bcl-2, Bcl-XL and the BH4 domains of these proteins inhibit RyR-mediated Ca2+ release7,11. In those studies, RyR-overexpression models and dissociated hippocampal neurons were investigated and RyR activation was attained pharmacologically using caffeine. In this study, we provide evidence that the BH4 domains of Bcl-2 and Bcl-XL inhibit RyR activity triggered by the hormone CCK in a physiologically relevant concentration (5ā€‰pM) (Fig.Ā 1). Therefore, in PACs the BH4 domains of Bcl-2 and Bcl-XL may serve as the modulators of RyR-mediated Ca2+ signals.

The lack of inhibition of IP3R-mediated Ca2+ release in WT PACs by the BH4 domain of Bcl-2 was somewhat surprising (Fig.Ā 2). However, this might be due to several factors. First, in most of the previous studies, the peptides were either loaded into the cells via electroporation or were added to already permeabilized cells2,4,28. Here, the BH4 domain peptides were added directly to intact cells without any adjuvant approach. This, however, could limit the intracellular concentration of the peptides and prevent it from reaching the levels required for inhibiting IP3Rs. Another possible explanation of these findings may be related to the proteolytic activation of the digestive enzymes and degradation of the BH4 domain of Bcl-2 or Bcl-XL into smaller fragments, capable of inhibiting RyRs but not IP3Rs. Although our experiments were performed in the presence of an extracellular protease inhibitor (inhibiting trypsin and chymotrypsin), we cannot fully rule out the possibility of enzymatic degradation of the BH4 domains. Finally, we hypothesize that IP3Rs in the WT PACs may already be associated with endogenous Bcl-2 and thus application of the BH4 domain of Bcl-2 may not provide any additional inhibition. To investigate this, we performed control experiments using PACs isolated from Bcl-2 KO mice. Importantly, in the cells lacking endogenous Bcl-2, the exogenous BH4 domain of Bcl-2 was capable of inhibiting ACh-induced IP3R-mediated Ca2+ release (Fig.Ā 3). Also in the Bcl-2 KO PACs, no additional inhibition of CCK-induced-RyR-mediated Ca2+ release by either BH4 domain was detected compared to the WT PACs (Figs.Ā 1 and 4), suggesting that in WT PACs RyRs are not heavily regulated by endogenous Bcl-2. Collectively, these results support the hypothesis that in WT PACs endogenous Bcl-2 may be associated mainly with IP3Rs, limiting the potential of the exogenously added BH4 domain of Bcl-2 to inhibit IP3R-mediated Ca2+ release.

We also want to highlight that higher concentrations of ACh (200ā€‰nM) or CCK (10ā€‰pM) were needed in Bcl-2 KO PACs compared to wild-type PACs to evoke IP3R- or RyR-mediated Ca2+ oscillations, respectively (Figs.Ā 3 and 4). Since Bcl-2 is an inhibitor of both IP3Rs and RyRs, it is anticipated that Bcl-2 KO PACs should be more sensitive to ACh and CCK. It remains unclear whether in Bcl-2 KO PACs this phenomenon occurs due to the compensatory mechanisms or other adaptive processes during pancreatic development. It is possible that lack of Bcl-2 in PACs results in reduced expression of ACh and CCK receptors, declined levels of the proteins involved in generation of IP3, NAADP, cADP-ribose, or decreased levels of IP3Rs and RyRs. It is also important to note that, compared to the WT mice, the Bcl-2 KO mice suffer from growth retardation, facial malformation and development of polycystic kidney disease30. Also, the pancreata of these Bcl-2 KO mice are much smaller compared to the WT pancreata. Thus, this phenotypic difference may account for the apparent reduced responsiveness of Bcl-2 KO mouse PACs towards extracellular agonists.

Bile acids, such as TLC-S, are well known to induce pathological Ca2+ release in PACs, leading to extensive tissue necrosis and development of AP23. The involvement of both IP3R- and RyR-mediated Ca2+ release in this process has been previously demonstrated22. From our previous work, we already knew that the BH4 domain of Bcl-XL does not inhibit IP3R-mediated Ca2+ release2. In the present study, we confirmed that observation in PACs (Fig.Ā 2). We also demonstrated that the BH4 domain of Bcl-2 inhibits IP3R-mediated Ca2+ release in Bcl-2 KO but not in WT PACs (Figs.Ā 2, 3). In addition, we show that both the BH4 domains of Bcl-2 and Bcl-XL similarly inhibit pathophysiological Ca2+ overload induced by the bile acid TLC-S in WT PACs (Fig.Ā 5a, b). Combining these observations we conclude that, in WT PACs the inhibition of TLC-S-induced Ca2+ release conferred by the BH4 domain peptides is mainly due to the inhibition of RyRs and not IP3Rs.

Pharmacological inhibition of either IP3Rs or RyRs has been shown to reduce TLC-S-induced necrosis in PACs and the severity of AP24,25. Here we confirmed this by showing that the BH4 domains were able to inhibit TLC-S induced necrosis (Fig.Ā 5c). These findings underpin the therapeutic properties of Ca2+-signaling modulation in AP and add the BH4 domains of Bcl-2 and Bcl-XL to the arsenal of tools with the therapeutic potential to reduce AP burden by limiting the excessive RyR activity and cell necrosis.

Previous studies on the BH4 domains of the Bcl-2-family proteins have revealed their protective properties against a wide range of pathological stimuli17,18,19,20,21,31,32,33,34,35, particularly with respect to the regulation of mitochondrial integrity and thus protection from apoptosis17,18,19,20,21. However, whether these BH4 domains could also antagonize cell damage by inhibiting excessive intracellular Ca2+ release, especially aberrant RyR activity under pathological conditions, has never been addressed. Here we show for the first time that these BH4 domains may provide protection in AP, a disease characterized by Ca2+-induced necrosis, via suppressing excessive RyR activity (Fig.Ā 5).

Although, the exact length of the BH4 domains used in different studies may differ slightly, the amino acid stretch conferring the Ī±-helical properties in the BH4 domains (Fig.Ā 1a) is always present17,18,19,20,21,31,32,33,34,35. The Ī±-helical properties of the BH4-domain peptide of Bcl-2 were previously shown to be crucial for inhibiting both IP3R activity and suppressing apoptosis induction, rendering this an important feature for the biological activity of the BH4 domain of Bcl-228. In the full-length protein, the BH4 domain also contributes to the overall stability of the Bcl-2 proteins36. PACs are known to take up small peptides that retain their bio-activity37. Hence, in this work, unmodified BH4-domain peptides could be applied, as they were taken up by intact primary PACs and they retained their biological activity (Figs.Ā 1ā€“5). At this point, it remains unclear whether other cells besides PACs are able to accumulate these unmodified peptides and whether they can be applied in vivo. An often used strategy to improve the uptake of peptides by cells is to couple peptides to a cell-penetrating sequence, like TAT, the protein-transduction domain of the HIV-1 TAT protein38. Besides enhancing cell uptake, this tag also introduces several positive charges, improving peptide solubility and bio-availability. This approach has already been used for the BH4 domain of Bcl-XL. TAT-tagged BH4 domain of Bcl-XL retained its biological activity when injectedĀ intravenously or intraperitonealy in mice, protecting hearth cells from ischemia reperfusion-induced cell death17,21, rescuing astrocyte degeneration in amyotrophic lateral sclerosis39 and protecting neurons from apoptosis upon frataxin depletion by restoring proper Ca2+ homeostasis and dynamics40. Another option/modification for further investigating the therapeutic potential of these BH4 domains is the use of stapled BH4 domains41,42. Stapled BH4 domains are cyclic peptides with increased stability and cell permeability. Such tools may enable to explore the therapeutic potential of BH4-domains of Bcl-2/Bcl-XL proteins and derivatives in in vivo models for AP.

Taken together, we here report that the BH4-domains of Bcl-2-family members can be utilized as peptide tools with Ca2+-modulatory properties to reduce disease burden in disorders such as AP. This work reveals that the BH4 domain of Bcl-2 and Bcl-XL, or tools mimicking their functions, may prove useful for therapeutic modulation of the pathologies in which excessive intracellular Ca2+ release is a critical driver.

Materials and methods

Reagents, mouse strains, secondary structure prediction, and peptides

Unless otherwise specified all reagents were obtained from Sigma-Aldrich (Dorset, UK). Transgenic Bcl-2 KO mice (B6;129S2-Bcl-2) were purchased from The Jackson Laboratory. Homozygous Bcl-2 KO and the WT litter mates were used for the experiments. All regulated animal procedures were subject to approval by the Universityā€™s Animal Welfare and Ethical Review Body (AWERB), and covered by a Project Licence granted by the Home Office under the Animal (Scientific Procedures) Act, 1986. PSIPRED version 3.3 (http://bioinf.cs.ucl.ac.uk/psipred/) was utilized to predict the secondary structure of the BH4 domains of Bcl-2 and Bcl-XL.

The following peptides (purity ā‰„80%) were obtained from LifeTein, validated via mass spectrometric analysis, and dissolved in DMSO (30ā€‰mM stock solution).

Control peptide: WYEKQRSLHGIMYYVIEDRNTKGYR

BH4 domain Bcl-2: RTGYDNREIVMKYIHYKLSQRGYEW

BH4 domain Bcl-XL: MSQSNRELVVDFLSYKLSQKGYSW

PAC isolation

Mouse PACs were isolated using a modified protocol described in ref. 37. Briefly, the pancreas of WT or Bcl-2 KO mouse was dissected and washed twice in NaHEPES buffer (140ā€‰mM NaCl, 4.8ā€‰mM KCl, 1ā€‰M MgCl2, 10ā€‰mM HEPES, 10ā€‰mM glucose and 1ā€‰mM CaCl2; pH 7.2). 1ā€‰ml collagenase (25ā€‰U/ml) (C9263) was injected into the pancreas and then the tissue was incubated at 37ā€‰Ā°C in a water bath shaker. After this the pancreas was broken down by several cycles of pipetting in 1ā€‰ml NaHEPES buffer. The isolated PACs were spun down (200Ɨg, 1ā€‰min), washed with NaHEPES and then were suspended in fresh NaHEPES.

PAC loading and single-cell Ca2+ measurements

Isolated PACs were incubated in NaHEPES containing 4ā€‰ĀµM Fluo4-AM (Life Technologies, Loughborough, UK) for 30ā€‰min at room temperature. Then PACs were spun down and resuspended in NaHEPES. Single-cell Ca2+ measurements were performed as indicated in the figure descriptions, using a Scientifica (East Sussex, UK) imaging system connected to an Olympus (Cardiff, UK) IX71 microscope. In order to prevent enzymatic digestion of the peptides, the extracellular buffer was supplemented with 0.2ā€‰mg/ml trypsin-chymotrypsin inhibitor. The peptides and indicated stimuli were introduced to the cells using a syringe-driven perfusion system.

Necrosis assay

Isolated PACs were treated with either vehicle (DMSO) or the indicated peptides (50ā€‰ĀµM) for 15ā€‰min. Then cell death was induced by adding TLC-S (final concentration: 200ā€‰Ī¼M) to the PACs for 2ā€‰h at room temperature. Propidium iodide (2ā€‰Āµg/ml; Life Technologies) was present during the last 15ā€‰min of the TLC-S treatment. For the peptide-treated cells, 0.2ā€‰mg/ml trypsin-chymotrypsin inhibitor was added in order to protect the peptides from proteolytic damage. A TCS SPE confocal microscope or a multiphoton SP5 (both from Leica, Milton Keynes, UK) were used to image the propidium iodide staining and cell death was quantified by counting the propidium iodide-positive cells in relation to the control. In each experiment at least 100 cells were imaged and counted per treatment group.

Statistical analysis

For statistical analysis GraphPad prism 7 was used. Because of lack of normal distribution and unequal variance in several tested groups, the Kruskalā€“Wallis test with the Dunnā€™s multiple comparison post hoc test was performed for all Ca2+ measurements. One-way ANOVA with the Tukeyā€™s multiple comparison post hoc tests were performed for the necrosis assays. When using the One-way ANOVA, the data were normally distributed and variances were not significantly different between the tested groups. P-values for each statistical analysis are given in the figure legends. *, **, *** are used to indicate P-values <0.05, 0.01, and 0.001 between the different groups as indicated by the post hoc test. For each experimental setup, PACs were isolated and used independently from at least three different mice.