FK506-binding proteins 12.6 (FKBP12.6) and 12 (FKBP12) tightly associate with the cardiac ryanodine receptor (RyR2). Studies suggest that dissociation of FKBP12.6 from mutant forms of RyR2 contributes to store overload-induced Ca2+ release (SOICR) and Ca2+-triggered arrhythmias. However, these findings are controversial. Previous studies focused on the effect of FKBP12.6 on the initiation of SOICR and did not explore changes in the termination of Ca2+ release. Less is known about FKBP12. We aimed to determine the effect of FKBP12.6 and FKBP12 on the termination of SOICR. Using single-cell imaging, in cells expressing wild-type RyR2, we found that FKBP12.6 and FKBP12 significantly increase the termination threshold of SOICR without changing the activation threshold of SOICR. This effect, dependent on the association of each FKBP with RyR2, reduced the magnitude of Ca2+ release but had no effect on the propensity for SOICR. In contrast, neither FKBP12.6 nor FKBP12 was able to regulate an arrhythmogenic variant of RyR2, despite a conserved protein interaction. Our results suggest that both FKBP12.6 and FKBP12 play critical roles in regulating RyR2 function by facilitating the termination of SOICR. The inability of FKBPs to mediate a similar effect on the mutant RyR2 represents a novel mechanism by which mutations within RyR2 lead to arrhythmia.
The cardiac ryanodine receptor (RyR2) is responsible for releasing Ca2+ from the sarcoplasmic reticulum (SR) during excitation–contraction coupling. Physiologically, Ca2+ release occurs in response to Ca2+ entry into the cell in a process known as Ca2+-induced Ca2+ release (CICR) . Pathologically, Ca2+ release occurs in the absence of Ca2+ entry and is termed spontaneous Ca2+ release, or due to its dependence on SR Ca2+ store content, store overload-induced Ca2+ release (SOICR) . Although the molecular mechanisms underlying both the physiological and pathological activation of RyR2 and Ca2+ release are fairly well defined, the molecular mechanisms underlying Ca2+ release termination remain less well understood.
Cellular studies emphasize that Ca2+ release terminates at a constant threshold (∼60% of the SR Ca2+ level) [3–6], suggesting that the termination of SR Ca2+ release is governed by SR Ca2+-dependent RyR2 gating. The recent breakthrough in identifying the molecular identity of an intrinsic luminal sensor within RyR2 reinforces the luminal Ca2+-dependent regulation of RyR2 . As with the activation of SR Ca2+ release, termination of SR Ca2+ release is subject to change under pathophysiological conditions. Domeier et al.  have shown that the termination threshold of SR Ca2+ release is reduced in heart failure (HF), and attribute this to sensitization of RyR2. RyR2 mutations causing cardiomyopathies can lead to delayed Ca2+ release termination . These observations indicate that RyR2 itself plays a critical role in modulating the termination of SR Ca2+ release. In addition to disease, dysfunctional calmodulin, a RyR2-associated regulatory protein has also been shown to have a potent effect on Ca2+ release termination . This suggests that SR Ca2+ release termination not only is governed by the SR Ca2+ content and intrinsic properties of RyR2, but can also be modulated by extrinsic factors such as RyR2-associated proteins.
FK506-binding protein 12.6 (FKBP12.6) is one of the most prominent RyR2-associated proteins. Dissociation of FKBP12.6 from RyR2, due to phosphorylation and catecholaminergic polymorphic ventricular tachycardia (CPVT)-associated RyR2 mutations, has been shown to lead to a leaky RyR2 channel, manifested as increased RyR2 open probability (Po), and to contribute to the development of HF and arrhythmias [11,12]. Additionally, FKBP12.6-null mice are more susceptible to arrhythmogenic triggers [12,13]. In contrast, overexpression of FKBP12.6 not only improves cardiac contractility , but also protects mice from arrhythmias . These findings have led to the proposal that restoration of FKBP12.6 binding to RyR2 may be a potential therapeutic target for HF and arrhythmias [15,16].
However, the importance of FKBP12.6 in modulating RyR2 function is not unanimously accepted. Others have suggested that loss of FKBP12.6 does not alter RyR2 single channel activation  and that the extent of endogenous association between FKBP12.6 and RyR2 is small . These findings raise questions about whether an interaction between RyR2 and FKBP12.6 is critical in regulating RyR2 and whether loss of FKBP12.6 from the RyR2 macromolecular complex leads to aberrant Ca2+ release and arrhythmias as previously suggested. A possible reason for the contrasting findings is that most of the molecular and cellular studies, including our own, have only monitored the activation of RyR2. These studies found no change in RyR2 activation in the presence or absence of FKBP12.6, leading to the conclusion that FKBP12.6 does not regulate RyR2. To date, no study has been designed to determine whether FKBP12.6 alters the termination of SR Ca2+ release, an emerging mechanism of regulating RyR2 function and arrhythmias.
FKBP12 is another member of the FKBP family, sharing a very similar crystal structure and ∼85% amino acid sequence homology with FKBP12.6 . The expression of FKBP12 in cardiac muscle is higher than FKBP12.6 [20,21], with a recent study estimating that the protein concentration of FKBP12 is 10-fold greater than FKBP12.6 in the murine heart . FKBP12 deficiency leads to embryonic and neonatal lethality due to severe dilated cardiomyopathy and ventricular septal defects , whereas FKBP12.6-null mice are viable and the heart is structurally normal . These findings suggest that FKBP12 is indispensable in cardiac development and at least as important as FKBP12.6 in regulating Ca2+ release. Indeed, at the single channel level, loss of FKBP12 has been shown to activate RyR2 by locking it into a sub-conductance state . However, as with FKBP12.6, the role of FKBP12 is not universally accepted as a previous study shows that FKBP12 activates RyR2 by increasing the RyR2 Po . As with FKBP12.6, nothing is known about the effect of FKBP12 on termination of Ca2+ release via RyR2. Importantly, to our knowledge, there are no studies designed to examine whether FKBP12 plays a role in CPVT. Therefore, more research is required to elucidate the molecular functions of both FKBP12.6 and FKBP12.
In the present study, we aimed to investigate the role of FKBP12.6 and FKBP12 in the termination of Ca2+ release. We found that both FKBP12.6 and FKBP12 mediate early termination of SOICR without altering when SOICR occurs, leading to a reduced magnitude of Ca2+ release. Interestingly, a commonly studied CPVT-associated RyR2 mutation (R4496C) abolished the effect of FKBPs on Ca2+ release termination and magnitude of release, despite a conserved interaction between mutant RyR2 and both FKBPs. Our findings provide a novel mechanism by which FKBPs regulate the function of RyR2, independent of changes in the occurrence of SOICR. By doing so, they help to reconcile a field which has been divided over the role of FKBPs for more than a decade.
MATERIALS AND METHODS
Single-cell Ca2+ imaging (cytosolic Ca2+)
Measurements of cytosolic Ca2+ were conducted after loading with the acetoxymethyl ester (AM) form of the Ca2+ indicator fluo-4 (Invitrogen) as described previously . Cells were superfused continuously with KRH (Krebs–Ringer–Hepes) buffer containing 125 mM NaCl, 5 mM KCl, 25 mM Hepes, 6 mM glucose and 1.2 mM MgCl2, adjusted to pH 7.4 with NaOH, containing various concentrations of CaCl2 (0.1–1.0 mM) at room temperature. At the end of experiments, 20 mM caffeine was applied to deplete intracellular Ca2+ store. Fluo-4 was excited at 470 nm every 2 s with an exposure time of 100 ms. Fluorescence of fluo-4 (>515 nm) was detected by a CoolSNAP HQ2 charge-coupled device (CCD) camera. The cytosolic Ca2+ fluorescence was represented by F/F0, where F was the Ca2+ fluorescence intensity at any time and F0 was the Ca2+ fluorescence intensity recorded in 0 Ca2+ KRH solution.
Single-cell Ca2+ imaging (luminal Ca2+)
Ca2+ release in human embryonic kidney (HEK)293 cells expressing wild-type (wt) or mutant RyR2 (R4496C) with and without FKBP12.6 or FKBP12 were measured using single-cell Ca2+ imaging and the Ca2+-sensitive FRET (fluorescence resonance energy transfer)-based cameleon protein D1ER as previously described . Briefly, stable inducible HEK293 cells expressing RyR2 (wt or mutant) were co-transfected with D1ER and FKBP12.6 or FKBP12 cDNA using Ca2+ phosphate precipitation (all clones in pcDNA3). Control cells (those not expressing FKBP) were transfected with an empty pcDNA3 vector. Expression of RyR2 was induced with tetracycline 24 h after transfection, and all experiments were conducted 40–42 h after transfection. This inducible cell model ensures a consistent level of expression between RyR2 wt and mutants . The cells were superfused continuously at room temperature with KRH. The extracellular Ca2+ concentration ([Ca2+]o) was adjusted as required for different experiments (1 or 2 mM). Tetracaine (2 mM) and caffeine (20 mM) were used to block RyR2 and deplete Ca2+ store respectively. To dissociate FKBP from RyR2, some assays were performed in the presence of 2 μM FK506 as previously described . Fluorescent images of HEK293 cells were acquired every 2 s with an exposure time of 100 ms and excitation at 436 nm (20 nm bandwidth). The emissions of YFP and CFP were captured simultaneously at 535 nm and 480 nm respectively. The amount of FRET was determined from the ratio of the emissions at 535 nm and 480 nm. The activation threshold of SOICR (FSOICR) was calculated from the mean of the FRET signal immediately preceding each SOICR event. The termination threshold of SOICR (Ftermi) was calculated from the mean nadir of each SOICR event. The maximum (Fmax) and minimum (Fmin) store levels were calculated from the mean signal of the plateau before the application of caffeine and the mean of the plateaued FRET level following the application of caffeine respectively.
Preparation of cell lysates
The preparation of cell lysates was performed as described previously . HEK293 cells were harvested with 2 mM EDTA in PBS solution by centrifugation at 3000 g for 5 min 18 h after induction of RyR2 expression. The supernatant was discarded and the cell pellet was resuspended in lysis buffer containing 137 mM NaCl, 25 mM Tris/Hepes, 1% CHAPS, 0.5% soyabean phosphatidylcholine, 1 mM benzamide, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 2 μg/ml pepstatin A, 0.5 mM PMSF and 2.5 mM DTT, pH 7.4. The mixture was incubated on ice for 1 h. Cell lysates were obtained by centrifugation twice at 16000 g for 30 min at 4°C to remove insoluble materials. Cell lysate was divided into aliquots and stored at −80°C until further use.
To create FKBP cell lysates, HEK293 cells were transiently transfected with FKBP cDNAs (in pcDNA3) prior to lysis.
RyR2–FKBP co-immunoprecipitation assays and immunoblotting
The association of RyR2 and FKBP was determined as previously described . A saturating amount of HEK293 cell lysate containing RyR2 wt was added to 30 μl of Protein G–Sepharose (Invitrogen) prebound with 0.5 μg of anti-RyR antibody (34C, Abcam) at 4°C overnight. FKBP12.6 or FKBP12 protein lysate was added until binding of FKBP to RyR2 became saturated. Subsequently, immunoprecipitation was repeated with RyR2 (wt or R4496C) and 0, 10, 30, 50, 70, 90 and 100% of the FKBP input required to achieve saturation of RyR2. The proteins bound to the Sepharose beads were solubilized by the addition 30 μl of 2× Laemmli's sample buffer plus 5% (v/v) 2-mercaptoethanol and boiled for 5 min. The samples were then separated by SDS/PAGE (6% for RyR2 or 10% for FKBP). The SDS/PAGE-resolved proteins were transferred on to nitrocellulose membranes at 100 V at 4°C for 100 min. The membranes containing the transferred protein were blocked with PBS containing 0.5% Tween 20 (PBS-T) and 5% (w/v) dried non-fat skimmed milk powder for 1 h. The blocked membranes were then incubated with anti-RyR (34C) (1:3000) or anti-FKBP (1:5000 dilution) (Pierce) at 4°C overnight and washed three times for 5 min each in PBS-T. The membranes were then incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (1:40000 dilution) for 1 h. The RyR2 or FKBP protein was detected by enhanced chemiluminescence and the blot images were analysed using myImageAnalysis Software 2.0 (Thermo Scientific).
All results are presented as means±S.E.M. For statistical analysis Student's unpaired t test and one-way ANOVA were applied. Differences were considered statistically significant if P<0.05. All data analysis and plotting were performed using GraphPad Prism 6.
FKBPs do not change the propensity for SOICR in HEK293 cells expressing either RyR2 wt or RyR2 R4496C
To determine whether the FKBP proteins have any effect on the occurrence of SOICR we measured the propensity for SOICR in cells expressing RyR2 wt or the CPVT mutant R4496C, in the presence and absence of FKBP12.6 and FKBP12. To avoid the possible complications of endogenous FKBP expression, this assay was performed in HEK293 cells. HEK293 cells stably expressing RyR2 have been shown to be a very good model for studying SOICR as they display SOICR events comparable with those of cardiac myocytes [3,7]. Importantly, they do not express FKBP12.6 and have only very low levels of FKBP12 (Supplementary Figure S1). HEK293 cells expressing RyR2 alone or co-expressing RyR2 and FKBP12.6 or FKBP12 were loaded with fluo-4 AM. SOICR was induced by increasing the external [Ca2+] and is seen by transient increases in fluo-4 fluorescence. Figure 1(A) shows representative fluo-4 traces recorded in RyR2 wt alone (black), RyR2 wt +FKBP12.6 (red) and RyR2 wt +FKBP12 (blue). As can be seen in Figure 1(B) the presence of either FKBP had no effect on the propensity for SOICR. In order to determine whether RyR2 CPVT mutations affected the occurrence of SOICR in the presence and absence of FKBP12.6 we performed the same assay in cells expressing an extensively characterized CPVT RyR2 mutation (R4496C). Consistent with our previous findings we found the RyR2 CPVT mutation led to an increase in SOICR compared with RyR2 wt . Similar to cells expressing RyR2 wt, the presence of FKBP12.6 or FKBP12 had no effect on the propensity for SOICR in cells expressing the RyR2 R4496C (Figures 1C and 1D).
Co-expression of FKBP12.6 or FKBP12 has no effect on the propensity for SOICR
Effect of FKBP12.6 on activation and termination of SOICR
As FKBP12.6 has been reported to reduce Ca2+ leak [14,26] but had no effect on the occurrence of SOICR, we hypothesized that FKBP12.6 would reduce the magnitude of leak by increasing the termination threshold of SOICR. SOICR was induced by increasing the external [Ca2+]o which led to an increase in luminal [Ca2+], observed as the initial increase in FRET signal in Figures 2(A)–2(C). When store Ca2+ reached a certain level SOICR events occurred (activation threshold of SOICR, FSOICR), as shown by the downward deflections. These SOICR events did not result in the complete depletion of the store Ca2+, but rather terminated at a given level (termination threshold of SOICR, Ftermi) that is consistent with observations in cardiomyocytes . A common RyR2 inhibitor, tetracaine, was applied to prevent SOICR and allow the store to fill to its maximal level (Fmax), followed by caffeine to deplete the Ca2+ store (Fmin). The activation and termination thresholds along with the amount of released Ca2+ per SOICR event were then calculated relative to the Ca2+ store size. In cells expressing RyR2 wt alone we found that SOICR occurred when the store Ca2+ reached 90.2±0.5% of its maximal capacity (Figure 2F). The addition of FKBP12.6 had no effect on the activation threshold of SOICR (90.9±0.6%, not significant). SOICR terminated when the store Ca2+ level fell to 53.8±1.0% of its maximal capacity in cells expressing RyR2 wt alone. Interestingly, the termination threshold of SOICR was significantly increased to 57.8±1.1% (P<0.01) in cells expressing FKBP12.6 (Figure 2D). The elevation in the termination threshold in the absence of any change in the activation threshold led to a significant reduction in the magnitude of Ca2+ release during each SOICR event (37.0±1.0% of total store in RyR2 wt alone group compared with 33.1±1.1% in +FKBP12.6 group, P<0.01) (Figure 2E). When normalized to the magnitude of release in cells expressing RyR2 wt alone this represents a 10.5% decrease in the amount released due to the expression of FKBP12.6. The total Ca2+ store size (difference between Fmax and Fmin, Figure 2G) was unaffected by the presence of FKBP12.6. The change in magnitude of release was confirmed by analysing the amplitude of Ca2+ release in cells loaded with fluo-4, which showed a 21% reduction due to FKBP12.6 (2.14±0.09 in RyR2 wt alone group compared with 1.70±0.08% in +FKBP12.6 group, P<0.0001) (Supplementary Figure S2). To confirm that the effect of FKBP12.6 was due to an interaction with RyR2, we treated cells expressing FKBP12.6 with 2 μM FK506 (Figure 2C). We have previously shown that this dissociates FKBP12.6 from RyR2 . In cells expressing FKBP12.6 the presence of FK506 resulted in a significant decrease in the termination threshold (from 57.8±1.1% to 50.9±1.6%, P<0.01) and a significant increase in the magnitude of SOICR (from 33.1±1.1% to 39.2±1.4%, P<0.01) compared with untreated cells expressing FKBP12.6. In doing so the presence of FK506 returned the termination threshold and magnitude of release to levels indistinguishable from cells without FKBP12.6 (Figures 2D and 2E).
FKBP12.6 increases the termination threshold of SOICR and reduces the magnitude of Ca2+ release in cells expressing RyR2 wt
The CPVT mutation R4496C abrogates the effect of FKBP12.6 on the termination threshold of SOICR
It is suggested that CPVT mutant variants of RyR2 experience greater levels of SOICR due to the partial dissociation of FKBP12.6 from RyR2 , thus we next investigated whether the effect of FKBP12.6 on the termination threshold of SOICR is preserved in cells expressing a RyR2 CPVT mutant. Figures 3(A) and 3(B) show a typical response of stable inducible HEK293 cells expressing the CPVT mutation R4496C to the experimental protocol described above. As shown in Figure 3(C) we found that, in the absence of FKBP12.6, the termination threshold of SOICR was 47.0±1.1%, reduced from 53.8±1.0% in RyR2 wt cells, illustrating the direct impact of RyR2 CPVT mutations on SOICR. Interestingly, in contrast with the effect of FKBP12.6 on RyR2 wt-expressing cells, FKBP12.6 had no influence on the termination threshold in cells expressing RyR2 R4496C (45.9±1.2% in R4496C+FKBP12.6 group). Consequently, the fractional release of Ca2+ per SOICR event was also unaffected by FKBP12.6 in these cells (Figure 3D). Figure 3(E) summarizes the activation threshold of SOICR in cells expressing RyR2 R4496C without or with FKBP12.6. Consistent with our previous findings, the activation threshold of SOICR was reduced by the CPVT mutant compared with RyR2 wt, and FKBP12.6 had no effect on this threshold (Figure 3E). FKBP12.6 expression did not alter the total Ca2+ store size in cells expressing the R4496C mutant (Figure 3F).
FKBP12.6 has no effect on SOICR in cells expressing an RyR2 CPVT mutant
Effect of FKBP12 on activation and termination of SOICR
In contrast with the extensive effort spent unravelling the role of FKBP12.6 in regulating RyR2 function, only a limited number of studies on the effect of FKBP12 have been reported. To understand the regulatory function of FKBP12 on RyR2, the same luminal Ca2+ imaging assays were applied to cells expressing FKBP12. Figures 4(A)–4(C) show representative FRET signals from HEK293 cells expressing RyR2 wt with and without FKBP12 and with FKBP12 and FK506. The presence of FKBP12 elevated the termination threshold of SOICR (Figure 4D, 53.0±1.0% in RyR2 wt alone group compared with 56.1±0.7% in +FKBP12 group, P<0.05), and attenuated the magnitude of Ca2+ release (Figure 4E, from 38.1±0.8% in RyR2 wt alone group to 35.8±0.6% in +FKBP12 group) in a similar manner (although to a lesser extent) to FKBP12.6. When normalized to the magnitude of release in cells expressing RyR2 wt alone, this represents a 6% decrease in the amount released due to the expression of FKBP12. The activation threshold of SOICR and the total Ca2+ store size were unaltered by FKBP12 (Figures 4F and 4G). The FKBP12-dependent increase in the termination threshold, lack of effect on the threshold for release and subsequent reduction in the magnitude of Ca2+ release mirrors the findings in cells expressing RyR2 wt and FKBP12.6. The change in magnitude of release was confirmed by analysing the amplitude of Ca2+ release in cells loaded with fluo-4, which showed a 13% reduction due to FKBP12 (2.14±0.09 in RyR2 wt alone group compared with 1.86±0.08% in +FKBP12 group, P<0.01) (Supplementary Figure S2). As with FKBP12.6 we also used FK506 to confirm that the effect of FKBP12 was due to an interaction with RyR2. The presence of FK506 in cells expressing FKBP12 resulted in a significant decrease in the termination threshold (from 56.1±07% to 51.5±1.3%, P<0.01) and a significant increase in the magnitude of SOICR (from 35.8±0.8% to 39.2±1.1%, P<0.01) compared with untreated cells expressing FKBP12. As with cells expressing FKBP12.6, the presence of FK506 returned the termination threshold and magnitude of release of cells with FKBP12 to the same levels as cells without FKBP12 (Figures 2D and 2E).
FKBP12 increases the termination threshold of SOICR and reduces the magnitude of Ca2+ release in cells expressing RyR2 wt
The RyR2 mutation R4496C abolishes the effect of FKBP12 on the termination and magnitude of SOICR
Given the similar effect of both FKBPs on RyR2 wt, we hypothesized that the presence of CPVT mutations within RyR2 would ablate the effect of FKBP12 on SOICR. As expected, the increase in termination of SOICR mediated by FKBP12 was completely eliminated in cells expressing the RyR2 CPVT mutant R4496C (47.0±1.1% in RyR2 alone group compared with 48.2±1.0% in +FKBP12 group, not significant) (Figures 5A–5C). Similar to FKBP12.6, the presence of FKBP12 had no effect on the activation threshold of SOICR (Figure 5E), therefore the magnitude of Ca2+ release was also unchanged (Figure 5D). FKBP12 had no effect on the total Ca2+ store size (Figure 5F).
FKBP12 has no effect on SOICR in cells expressing an RyR2 CPVT mutant
The loss of the effect of FKBPs on RyR2 CPVT mutants is not due to a reduction in FKBP binding to RyR2
Dissociation of FKBP12.6 from RyR2 CPVT mutants has been touted as a mechanism of the observed increase in RyR2 Ca2+ leak , and Figures 2 and 4 show that dissociation of FKBPs from RyR2 (via FK506) can prevent the effect of FKBPs on the termination of SOICR. Therefore, we next determined whether the mechanism by which the RyR2 CPVT mutation R4496C abolished the effect of FKBPs was due to an altered interaction between the RyR2 CPVT mutant and each FKBP. To assess the interaction between RyR2 and each FKBP, we performed co-immunoprecipitation assays with various quantities of FKBP, as previously described . The input of FKBP12.6 or FKBP12 required to saturate RyR2 binding was determined and set at 100%, then various fractional amounts of FKBP12.6 or FKBP12 were incubated with RyR2 wt or R4496C cell lysate in the presence of Protein G–Sepharose beads pre-bound to an anti-RyR2 antibody. Figure 6 shows that there was comparable binding of FKBP12.6 to RyR2 wt and R4496C over a wide range of FKBP12.6 levels (P =0.8). Similarly, the binding of FKBP12 to R4496C was commensurate with RyR2 wt (P =0.3). These findings demonstrate that RyR2 CPVT mutations do not alter the binding characteristics of either FKBP with RyR2 and suggest the lack of effect on the termination of SOICR observed is not due to the complete dissociation of either FKBP from RyR2.
The CPVT mutation R4496C does not alter the interaction of RyR2 with either FKBP12.6 or FKBP12
In the present study, we determined the effect of FKBP12.6 and FKBP12 on the activation and termination of SOICR. We demonstrated that expression of FKBP12.6 or FKBP12 increased the termination threshold of SOICR without altering the activation threshold of SOICR, leading to a reduced amount of Ca2+ release per SOICR event. The RyR2 CPVT mutation R4496C abolished the inhibitory effect of both FKBPs despite a conserved interaction between the RyR2 CPVT mutant and both FKBPs.
Termination of Ca2+ release
Under physiological conditions, RyR2 is activated by a process known as CICR. Given the positive-feedback nature of CICR, a concerted effort has been made to understand how Ca2+ release terminates. Although termination of Ca2+ release is still relatively poorly understood, several hypotheses have been proposed [28–30]. It is believed that SR Ca2+ release via RyR2 is highly dependent on the SR Ca2+ concentration. Since SR Ca2+ should reduce during release, the abrupt decline of local SR Ca2+ could interrupt the release process and lead to the closing of RyR2 [31,32]. Recent mathematical modelling work extends this proposal to show that, as SR Ca2+ decreases, the Ca2+ flux through RyR2 decreases to a level where CICR becomes unlikely and SR Ca2+ release terminates, a process termed induction decay . However, a large number of studies in myocytes and heterologous HEK293 cell models have demonstrated that in both triggered SR Ca2+ release (CICR) and diastolic spontaneous Ca2+ release (SOICR), Ca2+ release terminates at a constant level of SR Ca2+ (∼60% of SR Ca2+) [3–6,9,33,34]. This constant termination threshold of systolic Ca2+ release and diastolic Ca2+ release suggests a similar mechanism of termination that is not only dependent on SR Ca2+, but also regulated by other factors, such as the sensitivity of RyR2 to SR Ca2+ or RyR2-associated proteins. Indeed, it has been shown that dilated cardiomyopathy related RyR2 mutations cause aberrant SOICR termination . Moreover, dysfunctional calmodulin also leads to delayed SOICR termination and consequently enhances Ca2+ release . These intriguing findings support the hypothesis that termination of SR Ca2+ release is a complex process and is modulated by several factors. Using a heterologous HEK293 expression system, shown to display Ca2+ release and termination characteristics indistinguishable from myocytes [9,10,35], we show that the termination of SOICR is regulated by FKBP12.6 and FKBP12, both of which significantly enhance termination and subsequently reduce the magnitude of release. We confirmed that this regulation was via a direct interaction of FKBPs with RyR2. The application of FK506, a compound known to dissociate FKBPs from RyR2 , completely abolished the changes in the termination and magnitude of SOICR mediated by both FKBPs. Moreover, we found that the RyR2 CPVT mutation R4496C results in a significant reduction in the termination threshold strengthening the link between SR Ca2+ release termination and disease.
Role of FKBPs in RyR2 mutation-linked CPVT
One of the proposed mechanisms of CPVT is the dissociation of FKBP12.6 from RyR2, leading to leaky RyR2 channel, enhanced abnormal Ca2+ release and subsequent triggered arrhythmias . However, a number of studies have demonstrated that many CPVT mutations have no effect on the interaction between RyR2 and FKBP12.6 [2,3,35–37]. As we found the effect of FKBPs on the termination and magnitude of SOICR was dependent on an interaction between RyR2 and FKBP, we explored whether a well-studied CPVT mutant resulted in the dissociation of FKBP. Consistent with previous findings, we found the interaction between RyR2 R4496C and FKBP12.6 remained intact. However, despite a conserved interaction, FKBP12.6 (or FKBP12) was unable to increase the termination threshold or reduce the magnitude of Ca2+ release through the RyR2 CPVT mutant channel. These findings may explain why R4496C CPVT mice still undergo stress-induced arrhythmias despite a conserved interaction of FKBP12.6 and RyR2 . Although our data do not provide a mechanism to explain why, despite a conserved interaction, FKBPs are unable to regulate the termination of RyR2 CPVT mutants it could be due to the loss of allosteric gating of the channel normally conferred by FKBP .
Magnitude of Ca2+ release and its role in arrhythmias
It is well established that the propagation of arrhythmogenic Ca2+ waves is critically determined by the inter-RyR2 CICR activity. Only if the magnitude of the local cytosolic Ca2+ signal generated by Ca2+ release (SOICR) is large enough to diffuse through the cytosol to neighbouring RyR2s can a local Ca2+ release become cell-wide Ca2+ waves, in a so-called ‘fire–diffuse–fire’ manner [39,40]. Logically, the SR Ca2+ load determines the magnitude of each local Ca2+ release, because it sets the SR Ca2+ driving force. In addition to driving force Ca2+ uptake also affects the genesis of Ca2+ waves, delayed after depolarizations (DADs) and arrhythmias. Recently, we generated a phospholamban (PLN, a SR Ca2+ ATPase inhibitor)- knockout RyR2-R4496C mutant mouse model and found that by breaking up cell-wide Ca2+ waves into Ca2+ sparks or mini-waves, deletion of PLN protected RyR2-R4496C mutant mice from stress-induced ventricular tachycardia . These data suggest that enhancing Ca2+ uptake which indirectly reduces the persistence of cytosolic Ca2+ release events is able to suppress Ca2+-triggered arrhythmias. These previous findings demonstrate that the amount and duration of Ca2+ released per event, and its ability to propagate, are crucial determinants of whether spontaneous Ca2+ release is arrhythmogenic. Therefore, a reduction in the magnitude of a Ca2+ release event protects against Ca2+-triggered arrhythmias. Indeed, flecainide, a sodium channel blocker, has been found to prevent CPVT in both mice and humans  due to the inhibition of arrhythmogenic Ca2+ waves by reducing Ca2+ spark mass, which reflects the magnitude of single Ca2+-release events . The key finding of the present study is that FKBP12.6 elevates the termination threshold of SOICR, leading to a 10.5% reduction in magnitude of Ca2+ release per SOICR event. This is consistent with the view that overexpression of FKBP12.6 protects against cardiac arrhythmias by reducing RyR2-mediated diastolic abnormal Ca2+ release and spark magnitude [14,26]. Additionally, we found that FKBP12 mirrored the effect of FKBP12.6 on SOICR (reduction in magnitude of 6%) which is consistent with previous studies showing that FKBP12 reduces the amplitude of Ca2+ sparks  and Ca2+ waves  in permeabilized cardiac myocytes. Although the present study does not define the molecular mechanisms underlying the elevation in termination threshold of SOICR conferred by FKBPs, our findings provide novel insights into the effect of FKBPs on SOICR, and SOICR-induced cardiac arrhythmias.
FKBP12.6 or FKBP12
Given that both FKBP12.6 and FKBP12 are expressed in the heart, and share a similar effect on RyR2 function, a question arises as to which isoform is the most important for RyR2 regulation? A potential answer to this question may be provided by a study by Guo et al.  who elegantly define the association of FKBP12.6 and FKBP12 with RyR2 in isolated ventricular myocytes. Their data suggest that only 20% of RyR2s are associated with FKBP12.6. However, the cardiac concentration of FKBP12 is more than 10-fold greater than FKBP12.6, at a level stoichiometric with the concentration of RyR2 monomer. Therefore, although FKBP12 has a lower affinity and more modest effect on the magnitude of Ca2+ release, its increased expression suggests that it may work in tandem with FKBP12.6 to regulate RyR2 activity in the heart. In our study we used an over-expression system in HEK293 cells. Although this model has been validated as a faithful model of myocytes to isolate the effects of individual proteins on the properties of SOICR [10,24], the overexpression of both FKBP isoforms would better mimic the in vivo state of FKBP12 than FKBP12.6. Overexpression will result in RyR2 being saturated with both FKBP isoforms leading to stoichiometric binding, therefore the greater abundance of FKBP12.6 in our model may overestimate its impact in vivo. Notwithstanding, our data suggest that future studies examining the role of FKBPs in the heart should consider both isoforms of FKBP and their relative contribution to RyR2 regulation. The influence of FKBP12 on RyR2 function may also explain the inconsistent phenotype of FKBP12.6-knockout mice as it may compensate for the lack of FKBP12.6 under certain experimental conditions.
Combined, our data offer a novel mechanism by which FKBPs regulate RyR2 activity, and extend the recent discovery that changes in the termination threshold of SOICR are linked to arrhythmias. Importantly, our data help to reconcile a rift in the RyR2 field surrounding the role of FKBPs. Our conclusions support those studies that show that FKBPs have no effect on the frequency of spontaneous Ca2+ release (SOICR). However, by defining an alternative mechanism by which FKBPs alter RyR2 activity via modification of Ca2+ release termination, our data also support those studies that show FKBPs are important regulators of RyR2 and that this critical regulation is lost in CPVT-inducing RyR2 mutants.
Joe Zhang and Peter Jones designed the research; Joe Zhang, Helen Waddell, Ella Wu, Jhanvi Dholakia, Chidinma Okolo and Janet McLay performed the research; Joe Zhang, Helen Waddell, Ella Wu, Jhanvi Dholakia, Chidinma Okolo, Janet McLay and Peter Jones analysed the data and Joe Zhang, Helen Waddell and Peter Jones wrote the paper.
This work was supported by the Marsden Fund administered by the Royal Society of New Zealand [grant number 10-UOO-205 (to P.P.J.)]; and the Lottery Health Research Board [grant number AP353117 (to P.P.J.)].
Ca2+-induced Ca2+ release
catecholaminergic polymorphic ventricular tachycardia
delayed after depolarization
human embryonic kidney
cardiac ryanodine receptor
store overload-induced Ca2+ release