Calpain-Dependent Protein Degradation Contributes to CASQ2-R33Q CPVT

Data Availability

The proteomics data generated in this study have been deposited in the jPOST repository under accession number JPST004689. Other data that support the findings of this study are available from the corresponding author on reasonable request. Descriptions of additional experimental procedures and materials are provided in the Supplemental Methods. Key experimental resources are listed in the Major Resources Table in the Supplemental Material.

Mice

The homozygous Casq2^R33Q/R33Q knock-in strain was generated by homologous recombination using genetic manipulation of embryonic stem cells.¹¹ Homozygous Casq2 null (Casq2^–/–) mice were also chosen because of their extensively expressed phenotypic manifestations of CPVT.¹³ C57BL/6NCrL mice from Charles River Laboratories were designated Casq2^WT/WT and used as WT controls. Experiments were performed in 2-month-old adult mice and in 1-day-old (P1) neonatal mice. Because no consistent sex-dependent differences have been described in patients with CPVT or in mouse models of CPVT associated with CASQ2 variants, pooled data from both sexes were used, as previously described in this model.¹² Neonatal mice were euthanized by decapitation, and adult mice were euthanized by cervical dislocation. Animals were maintained and studied according to the protocols approved by the local ethics committee of the Spanish National Center for Cardiovascular Research and the Regional Government of Madrid (PROEX 177/17 and PROEX 191.5/22). Power calculations were performed to determine the required number of animals based on previous data (P=0.05; power=80%).

For pharmacological treatment, Casq2^WT/WT and Casq2^R33Q/R33Q 2-month-old mice were randomized to different treatment groups using simple randomization. Drugs were administered by intraperitoneal injection. Flecainide (20 mg/kg; Focus Biomolecules) was administered every 12 hours for 3 consecutive days. Proteasome activity was inhibited with bortezomib (1 mg/kg; MedChemExpress) once daily for 3 consecutive days. For autophagy inhibition, mice were treated with a single dose of hydroxychloroquine sulphate (50 mg/kg; Sigma) for 24 hours. For calpain inhibition, MDL28170 (70 mg/kg; MedChemExpress) or calpeptin (20 mg/kg; Sigma-Aldrich CAS 117591-20-5) was administered once daily for 3 consecutive days. At the end of the treatment, mice were euthanized, and myocardial proteins were extracted.

For AAV administration, 7×10¹⁰ viral particles of AAV-TRDN and AAV-EGFP (enhanced GFP) were injected into the femoral vein of 2-month-old mice after simple randomization. In both AAV vectors, TRDN and EGFP were expressed under the control of the TnT (Troponin T) promoter. Vectors were packaged into an AAVmyo capsid. Mice were euthanized 4 weeks later, and myocardial proteins were extracted.

Neonatal Cardiomyocytes Isolation and Treatment

Cardiomyocytes were isolated from Casq2^WT/WT and Casq2^R33Q/R33Q newborn mice (1 day after birth, P1), as previously described by our group.¹⁴ Briefly, hearts were removed from the newborn mice on the day of birth. Sequential, partial digestions were performed with type 2 collagenase (LS004176, Worthington) until the hearts were completely digested, and the isolated cells were stored in fetal bovine serum. Cells were seeded in 10 cm² plates pretreated with 1% gelatin and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 20% 199 medium, 15% fetal bovine serum, 15% HEPES (4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid) and 2 mmol/L L-glutamine.

For pharmacological treatment, neonatal cardiomyocytes were isolated and seeded in a 24-well plate in 0.5 mL of culture medium supplemented as described above. The next day, cells were treated with different proteolytic inhibitors or the corresponding vehicle. Protein synthesis was blocked by treating neonatal cardiomyocytes with cycloheximide (Sigma) at 50 μM for 2, 4, and 8 hours. Flecainide (Focus Biomolecules) was used at 20 μM for 8 hours. For proteasome and ER-associated degradation (ERAD) inhibition, cells were exposed to MG132 (Sigma) and eeyarestatin (Sigma), respectively, at 10 μM for 8 hours. Autophagy was inhibited with hydroxychloroquine (Sigma) at 30 μM for 8 hours. To inhibit calpain, cells were treated with calpeptin (Abcam) at 10 μM for 8 hours. At the end of the treatment, cells were harvested, and proteins were extracted and analyzed as described below.

ECG Recording by Telemetry

Mice were anesthetized with 3% to 3.5% sevoflurane administered with 100% oxygen, and the telemetry device (Data Sciences International) was implanted subcutaneously in their back. Baseline ECGs were recorded for 5 minutes. Epinephrine (2 mg/kg; B. Brown) was injected intraperitoneally, and ECGs were recorded for a further 20 minutes. ECG data acquisition was performed using the Ponemah software (Data Sciences International). Both acquisition and analysis of ECG data were performed by a blinded operator.

Statistical Analysis

Normality was assessed using the D’Agostino-Pearson test. The vast majority of the subgroups successfully passed the normality threshold. For subgroups showing mild deviations from normality, an ordinary 2-way ANOVA was still used, considering the robust nature of a balanced factorial design against minor distribution violations. For quantitative values, the Mann-Whitney U test was used to compare means of 2 groups for a single time point. Two-way ANOVA with an appropriate posttest was used to compare multiple conditions with 2 variables, as indicated in the figure legends. Categorical variables were analyzed using contingency tables with the Fisher exact test. Changes were considered statistically significant if P<0.05. GraphPad Prism 8.0.1 and R version 4.5.2 were used to perform all the statistical analyses and graphs. Additional statistical methods for the proteomics analysis are provided in the Supplemental Material.

Changes in Protein Expression and Posttranslational Modification in Casq2^R33Q/R33Q Mice

To investigate the molecular mechanisms underlying the detrimental effects of the CASQ2-R33Q mutation, proteomics analysis was performed on adult Casq2^WT/WT and Casq2^R33Q/R33Q mouse hearts (Figure 1A through 1F). Mutant mice showed downregulation of several calcium-handling proteins in the SR, including CASQ2, TRDN, JCTN/ASPH (Aspartate Beta-Hydroxylase), and SRL (sarcalumenin), as well as increased expression of HRC (histidine-rich calcium-binding protein), which competes with CASQ2 to bind RyR2. Overall, the results showed a mild decrease in proteins related to myocardial contraction and the electron transport chain, together with an increase in proteins associated with inflammation, glycolysis, protein synthesis and oligomerization, and the proteasome complex (Figure 1A and 1C; Table S1). Specifically, Casq2^R33Q/R33Q mice showed an upregulation of proteins associated with protein degradation, including TOLLIP (toll-interacting protein), the ubiquitin-like protein ISG15 (interferon-stimulated gene 15), the lysosomal protein LAMP1 (lysosome-associated membrane glycoprotein 1), and PSMD2 (proteasome 26S subunit ubiquitin receptor non-ATPases 2) and PSMD11. Furthermore, mutant mice showed increased levels of ER chaperones required for protein folding and quality control, including GRP94 (glucose-regulated protein 94), GRP78, and CNX (calnexin), a Ca²⁺-binding protein, suggesting activation of ER stress (Figure 1A and 1C). Interestingly, no upregulation of ER stress markers was observed in CASQ2-knockout mice (Figure 1B and 1D). Furthermore, we found a poor correlation between protein expression changes in knockout and mutant mice, suggesting that the mechanisms leading to CPVT in different models are diverse (Figure 1E).

Because Casq2^R33Q/R33Q mice showed an enrichment in the protein phosphorylation category (Figure 1C), we quantified with SanXoT the differences in protein phosphorylation between WT and mutant animals. The results showed changes in the phosphorylation of some SR Ca²⁺ regulators, including an increase in HRC (Ser96) and a decrease in SERCA2a (sarcoplasmic/endoplasmic reticulum calcium ATPase 2a; Ser38), PLN (phospholamban; Ser16 [Serine 16] and Thr17), and the chaperones HSP90AB1 (Heat shock protein 90 kDa alpha, class B, member 1; Ser287) and CNX (Ser582; Figure 1F; Table S2). In addition, changes in the phosphorylation of sarcomeric proteins TTN and MYBPC3 (myosin-binding protein C3) were detected (Figure 1F).

TRDN Degradation Precedes CASQ2 Loss and Its Restoration Rescues CASQ2 Stability

To establish the mechanisms responsible for the downregulation of couplon proteins in Casq2^R33Q/R33Q mice, we first investigated their degradation rate. After cycloheximide treatment to inhibit protein translation, we found that CASQ2 half-life was significantly reduced in neonatal cardiomyocytes from Casq2^R33Q/R33Q mice compared with WT controls (Figure S1A and S1B). TRDN also showed a faster decrease after cycloheximide treatment in neonatal cardiomyocytes from mutant mice (Figure S1A and S1C). Importantly, TRDN degradation rate seemed to be faster than that of CASQ2, reaching an 84% decrease by 8 hours.

To determine whether the faster degradation of TRDN might contribute to CASQ2 destabilization, we reexpressed TRDN in mutant cardiomyocytes and assessed its impact on CASQ2 levels. First, we transfected neonatal cardiomyocytes from Casq2^R33Q/R33Q mice with modified RNA encoding TRDN or a GFP (green fluorescent protein) control. Western blot analysis showed that induction of TRDN expression led to increased CASQ2 protein levels in mutant cardiomyocytes (Figure S2A and S2B). To confirm these findings in vivo, we generated an AAV vector expressing TRDN under the control of the cardiac troponin promoter, packaged into the AAVmyo capsid. Systemic administration of AAV-TRDN to Casq2^R33Q/R33Q mice resulted in elevated myocardial TRDN levels, as expected, and a parallel increase in CASQ2 expression (Figure S2C and S2D). Together, these results suggest that loss of TRDN precedes and facilitates subsequent CASQ2 degradation in Casq2^R33Q/R33Q hearts.

Aberrant Calcium Handling Contributes to Couplon Protein Degradation in Casq2^R33Q/R33Q Mice

To confirm the changes in calcium regulators identified by proteomics, left ventricle samples from adult Casq2^R33Q/R33Q and Casq2^WT/WT mice were analyzed using Western blot. Samples from Casq2^R33Q/R33Q mice showed a significant increase in HRC expression and a decrease in P-PLN (PLN phosphorylation; Ser16/Thr17), with no changes in total PLN or SERCA2a levels (Figure 2A and 2B). Because reduced P-PLN results in attenuated SERCA2a activity,¹⁵ these changes may be associated with an increased Ca²⁺ concentration in the cytosol.

The antiarrhythmic drug flecainide is a RyR2 inhibitor that reduces spontaneous SR Ca²⁺ release into the cytosol during diastole.¹⁶ Flecainide has been shown to reduce the incidence of adverse arrhythmic events in a large cohort of patients with CPVT, especially among those who were refractory to β-blocker treatment.¹⁷ It also prevented CPVT in a mouse model bearing the R4496C mutation in RyR2 and in CASQ2 knockout mice.^18,19 To determine whether flecainide had any effect on couplon protein stability in Casq2^R33Q/R33Q mice, neonatal cardiomyocytes isolated from mutant mice were treated with flecainide for 8 hours. The results showed an increase in CASQ2 and TRDN protein levels in flecainide-treated cardiomyocytes compared with untreated cells (Figure 2C and 2D). To validate these results in vivo, 2-month-old Casq2^R33Q/R33Q and Casq2^WT/WT mice were injected intraperitoneally with flecainide or vehicle for 3 consecutive days. Western blot analysis showed an increase in CASQ2 and TRDN proteins in flecainide-treated Casq2^R33Q/R33Q mice compared with untreated Casq2^R33Q/R33Q mice, with no changes in WT mice. However, this beneficial effect was not evident with JCTN levels (Figure 2E and 2F). These results suggest that degradation of CASQ2 and TRDN in mutant mice is the result rather than the cause of Ca²⁺ mishandling, at least partially.

CASQ2 Degradation in Casq2^R33Q/R33Q Mice Depends on the ERAD Pathway

Because the proteomics analysis had shown an increase in ER stress proteins in mutant mice, we explored the potential role of this pathway in the degradation of couplon proteins in CPVT. We aimed to validate the identified changes using Western blot, and we investigated additional pathways associated with ER stress. As shown in Figure 3A and 3B, Casq2^R33Q/R33Q adult mice showed a significant increase in GRP94, GRP78, and CNX expression. In addition, the unfolded protein response–related transcription factors ATF6 (Activating transcription factor 6) and ATF4, and the spliced form of XBP1 (XBP1s, X-box binding protein 1, short), which is also associated with ER stress, were significantly upregulated in mutant mice (Figure 3A through 3C). To determine whether the ERAD pathway might be involved in the degradation of CASQ2 and TRDN, isolated neonatal cardiomyocytes were treated for 8 hours with the ERAD pathway inhibitor eeyarestatin. The results showed an increase in CASQ2 protein levels in Casq2^R33Q/R33Q neonatal cardiomyocytes treated with eeyarestatin, compared with untreated Casq2^R33Q/R33Q neonatal cardiomyocytes, whereas TRDN levels remained unchanged (Figure 3D and 3E). These results pointed toward distinct degradation pathways for the different couplon proteins.

CASQ2 But Not TRDN Is Degraded by the Ubiquitin-Proteasome System

Given the increase in proteasome-related proteins identified in the proteomics analysis, we explored whether the ubiquitin-proteasome system (UPS) was involved in the degradation of couplon proteins in Casq2^R33Q/R33Q mice. Neonatal cardiomyocytes isolated from Casq2^R33Q/R33Q and Casq2^WT/WT mice were treated with the proteasome inhibitor MG132 or vehicle. Results showed an increase in CASQ2 levels in Casq2^R33Q/R33Q treated cardiomyocytes compared with untreated cells, whereas TRDN remained unchanged (Figure 4A and 4B). To confirm these changes in vivo, Casq2^R33Q/R33Q and WT adult mice were injected intraperitoneally with the proteasome inhibitor bortezomib or vehicle once a day for 3 consecutive days. CASQ2 protein levels were increased in bortezomib-treated Casq2^R33Q/R33Q mice, while no changes were observed in TRDN or JCTN protein levels (Figure 4C and 4D), confirming the results obtained in neonatal cardiomyocytes. These results suggested that the UPS is necessary for CASQ2 degradation in CPVT.

CASQ2 Is Partially Degraded Through Autophagy in Casq2^R33Q/R33Q Mice

Previous studies have reported connections between the UPS and autophagy pathways. More specifically, when the ER stress response and the ERAD system are insufficient to restore ER homeostasis, additional degradation pathways are activated, including autophagy.^20,21 To determine whether autophagy was altered in this CPVT model, we first looked for changes in the global autophagic flux by measuring the amount of LC3II protein in neonatal cardiomyocytes from mutant and WT mice by Western blot. After treatment with the autophagosome-lysosome fusion inhibitor hydroxychloroquine, LC3II protein levels were significantly increased in cardiomyocytes from Casq2^R33Q/R33Q compared with those of Casq2^WT/WT mice (Figure 5A and 5B), suggesting increased autophagic flux. In addition, we found an increase in LAMP1 in protein extracts from Casq2^R33Q/R33Q mice (Figure 5C and 5D).

Interestingly, we found that CASQ2 protein levels increased when neonatal cardiomyocytes isolated from Casq2^R33Q/R33Q mice were treated with hydroxychloroquine compared with those of untreated Casq2^R33Q/R33Q mice, whereas there were no changes in TRDN protein levels (Figure 5E and 5F). To validate these changes in vivo, Casq2^R33Q/R33Q and Casq2^WT/WT mice were treated with chloroquine for 24 hours, and protein extracts were analyzed using Western blot. Autophagy flux inhibition with chloroquine prevented CASQ2 degradation in Casq2^R33Q/R33Q mouse hearts, whereas TRDN and JCTN remained unchanged (Figure 5G and 5H), suggesting that autophagy partially mediates CASQ2 degradation. No changes in protein levels were observed when WT mice were treated with chloroquine.

Calpain Is Activated in Casq2^R33Q/R33Q Mice and Mediates Couplon Protein Degradation

Alterations in calcium homeostasis and ER stress may activate calcium-dependent proteolytic enzymes, such as calpain.^22–25 Therefore, we hypothesized that calpain could play an important role in the degradation of couplon proteins. Western blot analysis showed a significant increase in CAPN1 (calpain 1) protein levels in Casq2^R33Q/R33Q mice and a decrease in the endogenous calpain inhibitor CAST, which regulates the activity of CAPN1 and CAPN2 (Figure 6A and 6B). In addition, Casq2^R33Q/R33Q mice showed a significant increase in the cleavage of α-fodrin (alpha-fodrin), a known calpain target that often serves as an indicator of calpain activity. A strong increase was observed in the cleaved (150 kDa) to full-length (240 kDa) α-fodrin ratio (Figure 6C and 6D), suggesting that calpain activity was increased in Casq2^R33Q/R33Q mice. To determine whether calpain inhibition could prevent couplon protein degradation, neonatal cardiomyocytes from Casq2^R33Q/R33Q and Casq2^WT/WT mice were treated with calpeptin for 8 hours. An increase in CASQ2 and TRDN levels was detected in Casq2^R33Q/R33Q neonatal cardiomyocytes treated with calpeptin compared with untreated controls (Figure 7A and 7B). To confirm these results in vivo, Casq2^R33Q/R33Q mice were injected intraperitoneally with the calpain inhibitor MDL28170 for 3 days. Western blot analysis of myocardial samples showed a decrease in α-fodrin cleavage in the treated mice, indicating that the inhibitor was effective (Figure 7C and 7D). Importantly, a significant increase in CASQ2 and TRDN protein levels was observed after calpain inhibition, accompanied by increased JCTN protein levels (Figure 7E and 7F). Furthermore, restoration of TRDN and CASQ2 expression significantly reduced the levels of ER stress–related proteins GRP78 and GRP94 (Figure 7G and 7H). To confirm the increase in TRDN and CASQ2 expression after calpain inhibition, we treated Casq2^R33Q/R33Q mice with a second calpain inhibitor in vivo. As shown in Figure S3A and S3B, calpeptin treatment resulted in a significant upregulation of TRDN and CASQ2, confirming the results obtained with MDL28170.

Finally, we explored whether TRDN could be a direct calpain substrate. Incubation of recombinant human TRDN with recombinant calpain in vitro resulted in strong cleavage of TRDN that was prevented by the calpain inhibitor calpeptin (Figure S4A and S4B). Similarly, recombinant calpain cleaved mouse TRDN in extracts from WT neonatal cardiomyocytes, and this processing was inhibited by calpeptin (Figure S4C and S4D). These results demonstrate that TRDN is a direct target of calpain, supporting the hypothesis that increased calpain activity contributes to couplon protein degradation in Casq2^R33Q/R33Q hearts.

Calpain Inhibition Improves the VT Phenotype in Casq2^R33Q/R33Q Mice

To investigate whether the increase in calpain activity could play a detrimental role in the development of CPVT, we first investigated whether calpain inhibition improved calcium handling in adult mutant mice. Cardiomyocytes isolated from adult Casq2^R33Q/R33Q mice showed a significantly higher propensity for developing triggered activity as compared with WT counterparts, particularly when exposed to isoproterenol, in line with our previous results with these mice²⁶ (Figure 8A through 8C). Mutant mice treated with flecainide were significantly less likely to develop triggered activity, as expected (Figure 8A through 8C; Table S3). Importantly, exposure of adult Casq2^R33Q/R33Q mice to the calpain inhibitor MDL28170 also reduced triggered activity (odds ratio, 0.2 [95% CI, 0.1–0.6]; Figure 8A through 8C; Table S3). Because flecainide and MDL28170 were administered to the animals before cardiomyocyte isolation and were absent during Ca²⁺ imaging, these findings indicate that the reduction in triggered activity persisted after drug removal, consistent with changes induced in vivo rather than an acute effect of the compounds during Ca²⁺ recording.

In our calcium imaging recordings, triggered events were consistently preceded by spontaneous diastolic Ca²⁺ waves. These spontaneous Ca²⁺ release events displayed a consistent latency measured from the prior diastole and were visually identifiable as discrete propagating mini-waves. We did not observe biphasic or dome-and-plateau Ca²⁺ release patterns, nor the irregular Ca²⁺ dynamics typically associated with early afterdepolarizations. Instead, the observed patterns are more consistent with delayed afterdepolarization–mediated triggered activity.

To determine whether this improvement translated into arrhythmia susceptibility, an ECG telemetry device was implanted into Casq2^R33Q/R33Q mice. As shown in Figure 8D and 8E, treatment with the calpain inhibitor MDL28170 for 3 days significantly reduced the incidence of ventricular arrhythmias after adrenergic stimulation (epinephrine, 2 mg/kg), indicating that calpain inhibition improves both cellular Ca²⁺ handling and susceptibility to ventricular arrhythmias in Casq2^R33Q/R33Q mice.

Acknowledgments

We thank the Center for Cardiovascular Research (CNIC) animal facility for mouse breeding, the Viral Vectors Unit for adeno-associated virus (AAV) production, and the Proteomics Unit for analyses.

Disclosures

None.

Supplemental Material

Supplemental Methods

Tables S1–S3

Figures S1–S4

Major Resources Table

Uncropped Western Blots

Data Set

ARRIVE Guidelines

References 12,18,44–52