Transcriptional regulation of intermolecular Ca²⁺ signaling in hibernating ground squirrel cardiomyocytes: The myocardin–junctophilin axis

Significance

The efficiency of excitation–contraction coupling in heart cells depends partially on the interaction between junctophilin-2 in the sarcoplasmic reticulum and caveolin-3 on T-tubules. Decreased junctophilin-2 has been linked to heart failure. Here, we found that the transcription of junctophilin-2 and caveolin-3 is controlled concurrently by the serum response factor and myocardin. During the entree into hibernation in ground squirrels, the up-regulation of myocardin boosts junctophilin-2/caveolin-3 expression and enhances the efficiency of excitation–contraction coupling. This finding not only reveals an adaptive mechanism underlying the blood-pumping power in hibernation, but also offers a strategy to correct deficient Ca²⁺ signaling in heart diseases.

Abstract

The contraction of heart cells is controlled by the intermolecular signaling between L-type Ca²⁺ channels (LCCs) and ryanodine receptors (RyRs), and the nanodistance between them depends on the interaction between junctophilin-2 (JPH2) in the sarcoplasmic reticulum (SR) and caveolin-3 (CAV3) in the transversal tubule (TT). In heart failure, decreased expression of JPH2 compromises LCC–RyR communication leading to deficient blood-pumping power. In the present study, we found that JPH2 and CAV3 transcription was concurrently regulated by serum response factor (SRF) and myocardin. In cardiomyocytes from torpid ground squirrels, compared with those from euthermic counterparts, myocardin expression was up-regulated, which boosted both JPH2 and CAV3 expression. Transmission electron microscopic imaging showed that the physical coupling between TTs and SRs was tightened during hibernation and after myocardin overexpression. Confocal Ca²⁺ imaging under the whole-cell patch clamp condition revealed that these changes enhanced the efficiency of LCC–RyR intermolecular signaling and fully compensated the adaptive down-regulation of LCCs, maintaining the power of heart contraction while avoiding the risk of calcium overload during hibernation. Our finding not only revealed an essential molecular mechanism underlying the survival of hibernating mammals, but also demonstrated a “reverse model of heart failure” at the molecular level, suggesting a strategy for treating heart diseases.

Most eukaryotic cells generate and amplify Ca²⁺ signals via Ca²⁺-induced Ca²⁺ release (CICR) (1, 2). In the heart, the ryanodine receptor (RyR) Ca²⁺ release from the sarcoplasmic reticulum (SR) induced by the Ca²⁺ influx through L-type Ca²⁺ channels (LCCs) in the cell membrane/T-tubules (TTs) controls the tone of excitation–contraction (E-C) coupling and the vigor of blood pumping (3, 4). The trans-SR protein junctophilin-2 (JPH2) (5), by interacting with caveolin-3 (CAV3) on cell membrane and TTs (6), plays an important role in determining the LCC–RyR signaling efficiency (7–10). Pathological down-regulation of JPH2 leads to compromised Ca²⁺ signaling in severe cardiac diseases such as heart failure (7, 8, 11, 12). Identifying new strategies for rescuing the structure and function of Ca²⁺ signaling in heart diseases has been a challenge. Recently, JPH2 has been found to be translocated into the nucleus following its cleavage, acting as a transcription factor to regulate a variety of gene expression (13). Although posttranscriptional (14–16) and posttranslational (17) regulation of JPH2 have been reported, molecular mechanisms that regulate the transcription of JPH2 per se are yet unknown.

Hibernation is an adaptation of some small animals and birds to survive the cold winter. The body temperature of hibernating mammals, normally maintained at or near 37 °C in the euthermic state, is down-regulated to the environmental level (18–21) possibly through certain hypothalamic neuronal circuits (22–24). Sustained hypothermia usually disrupts intracellular homeostasis of Ca²⁺ in nonhibernating mammals (25–29). Intracellular Ca²⁺ overload stimulates the feedback loop with reactive oxygen species production and destroys membranes, organelles, and DNA, eventually leading to structural and functional disruption of cells and tissues (25, 29, 30). Ca²⁺ overload in heart cells also promotes cardiomyopathy, ventricular fibrillation, and sudden death (27, 31–33). In contrast, mammalian hibernators down-regulate their body temperature without pathological consequence (18, 20, 31). Previous studies have shown that the Ca²⁺ influx through LCCs (I_Ca) is adaptively decreased during hibernation (34, 35). A reduction in LCC I_Ca lowers the risks of intracellular Ca²⁺ overload and arrhythmogenesis and is thus believed to protect the function of hearts under conditions of low body temperature (27). Intuitively, a reduction in I_Ca means that the trigger signal for CICR is decreased and would be expected to compromise the RyR Ca²⁺ release and E-C coupling. However, we observed that hibernating ground squirrels usually exhibit stronger myocardial contractions than wakeful animals (SI Appendix, Fig. S1). We therefore hypothesized that a specific Ca²⁺ signaling mechanism should counterbalance the reduced Ca²⁺ trigger from I_Ca.

In testing this hypothesis, we found that TT–SR junctions were structurally optimized by increased JPH2 and CAV3 expression, which improved LCC–RyR signaling efficiency during hibernation. In seeking the underlying molecular mechanism, we found that JPH2 and CAV3 were transcriptionally regulated concurrently by serum response factor (SRF) and myocardin. The myocardin-mediated concerted up-regulation of JPH2/CAV3 not only identifies a key adaptive mechanism of hibernation, but also has important implications in understanding and treating related heart diseases.

Results

LCC–RyR Intermolecular Signaling Efficiency Is Enhanced during Hibernation.

We combined whole-cell patch clamp with confocal imaging to record the intracellular Ca²⁺ transients triggered by I_Ca during depolarizations from −70 to 0 mV in isolated ventricular cardiomyocytes from euthermic and torpid ground squirrels (Fig. 1A). Although the I_Ca density (SI Appendix, Fig. S2A) and LCC expression (SI Appendix, Fig. S3) were significantly lower in cardiomyocytes from the torpid group than that in cells from the euthermic group, the Ca²⁺ transient (Fig. 1B) and the cell contraction (Fig. 1C) triggered by the decreased I_Ca were comparable with, or even higher than, those in the euthermic group, rendering a near 100% increase in CICR gain (Fig. 1D) and E-C coupling efficiency (Fig. 1E).

Fig. 1.

Increased LCC–RyR Ca²⁺ signaling efficiency during hibernation. (A) Representative whole-cell I_Ca recordings (Upper) and confocal imaging of Ca²⁺ transients (Lower) in cardiomyocytes of euthermic and torpid ground squirrels. (B) Ca²⁺ transient amplitude. (C) Fractional shortening (FS) measured by cell edge-detection. (D) The gain of CICR was calculated as the amplitude of Ca²⁺ transient per unit I_Ca. (E) E-C efficiency calculated as FS per unit I_Ca. (F) Representative recordings of a 1-s near-threshold depolarization from −50 to −40 mV under whole-cell patch clamp conditions that activated Ca²⁺ sparks in a probabilistic manner. (G) Frequency of triggered Ca²⁺ sparks (F_spark). (H) Fidelity with which LCCs activate RyRs at −40 mV, defined as F_spark per unit I_Ca (F_spark/I_Ca). Data in B–H were from ≥25 cells (≥6 animals) per group. (I) Representative loose-patch Ca²⁺ image showing the coupling latency between an LCC sparklet and an RyR spark. (J) Distributions (bars) and exponential fits (curves) of the LCC–RyR intermolecular coupling latency. (K) Rate constants of LCC–RyR coupling. J and K measured ≥270 sparks (≥8 animals) per group. *P < 0.05 and **P < 0.01 vs. the euthermic group by Student’s t test.

We performed two different experiments to test whether the change of E-C coupling during hibernation is associated with enhanced LCC–RyR intermolecular signaling. In the first we applied a near-threshold depolarization from −50 to −40 mV to activate a few LCCs, which in turn triggered RyR Ca²⁺ sparks in a probabilistic manner (Fig. 1F). The frequency of the Ca²⁺ sparks (F_spark) in the torpid group was significantly higher than that in the euthermic group (Fig. 1G). Because the I_Ca at −40 mV was lower in the torpid group, determined by the Boltzmann fitting of I_Ca-voltage curves (SI Appendix, Fig. S2, insert), the fidelity for LCCs to activate RyRs, defined as F_spark per unit I_Ca (F_spark/I_Ca), increased by 150% during hibernation (Fig. 1H).

In the second experiment, we activated single LCCs and triggered Ca²⁺ sparks from RyRs (Fig. 1I) using the loose-patch confocal imaging technique (36). We measured the LCC–RyR signaling kinetics by gauging the latency from the onset of an LCC sparklet to the takeoff of a triggered spark (Fig. 1J). Exponential fitting of the LCC–RyR coupling latency distributions showed that the rate constant of LCC–RyR signaling increased from 235 ± 14 s⁻¹ in the euthermic group to 340 ± 21 s⁻¹ in the torpid group (Fig. 1K), indicating that RyRs became more responsive to Ca²⁺ triggers during hibernation.

In cardiomyocytes, RyR-residing SR meets the LCC-residing cell membrane/TTs with an ∼15-nm junctional cleft, forming an LCC-RyR signaling structure known as a couplon (37). Ultrastructural quantification (Fig. 2A and SI Appendix, Fig. S4) using a reported procedure (8) showed that the volume density (Fig. 2B) and surface area (Fig. 2C) of TTs coupled to the SR were significantly increased by 57.2% and 39.2%, respectively, in the torpid group. The apparent curvilinear length of individual TT–SR couplons (Fig. 2D) was increased by 25% (Fig. 2E), and the TT–SR distance (SI Appendix, Fig. S5) was decreased from 17.76 ± 0.09 nm in the euthermic state to 16.15 ± 0.11 nm in the torpid state (Fig. 2F). Given that the cytosolic domain of an RyR occupies ∼14 nm (38) in the TT–SR space (SI Appendix, Fig. S6), the 1.6-nm reduction of the TT–SR distance means that the LCC and RyR molecules were >40% closer during hibernation. Previous model simulations have demonstrated that increased couplon area (8) and decreased LCC–RyR distance (39) enhance LCC–RyR signaling efficiency.

Fig. 2.

Enhanced TT–SR membrane coupling during hibernation. (A) Representative transmission electron microscopy (TEM) images of cardiomyocytes. Yellow arrows denote TT–SR junctions. (B) Stereological analysis of volume density per unit volume of total TTs, TTs coupled with the SR, bald TTs, and junctional SRs (JSRs). (C) Stereological analysis of surface area. B and C measured ≥130 images (≥5 animals) per group. (D) Representative TEM images from torpid ground squirrels showing a TT–SR junction. (E) Apparent curvilinear length of parallel TT and SR membranes. (F) TT–SR distance was determined by fast Fourier transformation (see SI Appendix, Fig. S11 for the measurement). Data in E and F were from ≥200 junctions (≥5 animals) per group. (G) mRNA expression levels for JPH2 and CAV3. (H) Protein expression levels determined using JPH2 (sc-51313, Santa Cruz Biotechnology) and CAV3 (ab2912, Abcam) antibodies. n = 5 per group for G and H. *P < 0.05 and **P < 0.01 vs. the euthermic group by Student’s t test.

Increased Expression of JPH2 and CAV3 during Hibernation.

TT–SR junction and LCC–RyR coupling in cardiomyocytes are maintained by the interaction between JPH2 in the SR and CAV3 in the TT (6). We therefore assessed the levels of JPH2 and CAV3 by quantitative (q) reverse transcription (RT)-polymerase chain reaction (PCR) (Fig. 2G) and immunoblot analysis (Fig. 2H). In contrast to reduced JPH2 in heart failure (6, 12, 40), JPH2 expression was significantly up-regulated in torpid squirrels. Interestingly, both messenger RNA (mRNA) (Fig. 2G) and protein expression (Fig. 2H) of CAV3 were also significantly up-regulated during hibernation. Previous studies have shown that overexpression of JPH2 protected E-C coupling in models of heart failure (9). Therefore, the up-regulation of JPH2 and CAV3, by tightening the TT–SR junction, constitutes an important mechanism for the enhanced cardiac E-C coupling during hibernation.

Besides JPH2 and CAV3, the sarcoplasmic/endoplasmic reticulum Ca²⁺–ATPase (SERCA2) and RyR were also up-regulated in torpid squirrels (SI Appendix, Fig. S7). Other E-C coupling proteins, including sodium/calcium exchanger (NCX1), calsequestrin-2 (a Ca²⁺ storage protein in the SR), and junctin (a protein that coordinates the binding between calsequestrin2 and RyR) were unchanged during hibernation. The increase of RyR expression was marginal, less than the increased proportion of junctional SR (Fig. 2 B and C). Although the up-regulation of SERCA2 in hibernation (27, 35) could explain the accelerated SR Ca²⁺ uptake observed in torpid squirrels (SI Appendix, Fig. S8 A and B), the steady-state SR Ca²⁺ load, indexed by the caffeine-induced Ca²⁺ transient amplitude, did not significantly differ between euthermic and torpid squirrels (SI Appendix, Fig. S8 C and D). JPH2 appeared to be the major candidate factor underlying enhanced efficiency of intermolecular Ca²⁺ signaling during hibernation.

Mechanisms of JPH2 Transcription.

Although JPH2 is a key determinant of the CICR efficiency, its transcriptional regulation is unknown. To this end, we compared the JPH2 gene sequences among ground squirrels, rats, mice, and humans and identified a highly conserved region ∼1 kb upstream of the translation start codon. We cloned this region or its truncations into luciferase reporter plasmids and found that the promoter activity declined dramatically when the sequence length was cut from −109 to −67 bp (Fig. 3A), where a conserved CArG-like element was identified (Fig. 3B). Mutations in this element strongly suppressed the JPH2 promoter activity in neonatal cardiomyocytes (Fig. 3C), confirming that the CArG-like element was a key cis-element for JPH2 transcription. SRF, together with its cofactor myocardin, has been reported to bind the CArG and CArG-like elements in heart tissues (41). Electrophoretic mobility shift assays (EMSA; Fig. 3D) and chromatin immunoprecipitation (ChIP)–qPCR assays (Fig. 3E) confirmed that SRF specifically bound to the CArG-like element in the JPH2 promoter. Interestingly, cotransfection of SRF with CArG-like element vectors did not increase JPH2 promoter activity (Fig. 3F), but cotransfection of myocardin with CArG-like element vectors increased JPH2 promoter activity by fourfold (Fig. 3G). Because mutations of the CArG-like element eliminated the effect of myocardin (Fig. 3G), these results suggested that JPH2 transcription was regulated by the SRF–myocardin complex, with myocardin acting as a limiting factor.

Fig. 3.

Mechanisms for transcriptional regulation of JPH2. (A) The promoter activity of rat JPH2 promoter truncations was determined via luciferase assay. (B) The CArG-like element of JPH2 promoter was conserved among rats, mice, humans, and S. dauricus. (C) The promoter activity of wild and mutant types was determined via the luciferase assay. (D) EMSA showed that nuclear extracts from SRF-overexpressing cells shifted the mobility of biotin-labeled DNA probes containing the CArG-like element of JPH2 promoter. This effect was attenuated by an addition of a 50-fold excess of unlabeled wild-type CArG-like element DNA probes but not by mutated CArG-like element DNA probes. (E) ChIP–qPCR assay showed that SRF was specifically enriched in the CArG-like region. (F) The effect of SRF on the JPH2 promoter was examined via the luciferase assay. (G) The effect of myocardin on the JPH2 promoter was examined via the luciferase assay. n ≥ 3 per group. *P < 0.05 and **P < 0.01 by Student’s t test.

Myocardin Regulates Concerted Transcription of CAV3 and JPH2.

Interestingly, when we examined the 5′ upstream of the CAV3 translation start site, we also found a CArG-like element in the CAV3 promoter, which was highly conserved among ground squirrels, rats, mice, and humans (Fig. 4A). ChIP–qPCR assays showed that SRF accumulated specifically in this region (Fig. 4B). Additionally, mutations of this CArG-like element prevented myocardin from regulating CAV3 promoter activity (Fig. 4C), indicating that CAV3 transcription was also controlled by the SRF–myocardin complex.

Fig. 4.

Myocardin regulates concerted transcription of CAV3 and JPH2. (A) The conserved CArG-like element found in the CAV3 promoter. (B) Enrichment of SRF in the rat CAV3 promoter was examined via the ChIP–qPCR assay. (C) The effect of myocardin on the CAV3 promoter was examined via luciferase assay. (D) Relative mRNA expression levels of myocardin and SRF were determined by qPCR and compared between euthermic and torpid groups. (E) Western blot analyses of JPH2 and CAV3 expression in the neonatal rat cardiomyocytes overexpressing RFP or rat myocardin. n ≥ 3 per group. *P < 0.05 and **P < 0.01 by Student’s t test.

Because CAV3 interacts with JPH2 in forming the LCC–RyR signaling couplon (6), our above findings suggested that JPH2 and CAV3 are transcriptionally regulated concurrently. In accord with this concept and the increased JPH2/CAV3 expression during hibernation, we found that the expression of myocardin was significantly increased in torpid squirrels (Fig. 4D and SI Appendix, Fig. S9). Notably, the expression of SRF was unchanged (Fig. 4D), agreeing with the result that myocardin, but not SRF, limited the transcription. We then overexpressed myocardin in neonatal rat cardiomyocytes and observed that both JPH2 and CAV3 expressions are indeed increased (Fig. 4E).

Myocardin Overexpression Reproduces the Hibernation Changes of LCC–RyR Signaling.

In order to test whether increased myocardin is able to reproduce the enhanced structure and function of Ca²⁺ signaling apparatus, we adjusted adenoviral infection of myocardin in cardiomyocytes of adult euthermic squirrels such that the JPH2 and CAV3 expression was up-regulated (Fig. 5A) to similar levels as those seen in torpid squirrel cells (Fig. 2G). Ultrastructural analysis of electron microscopic images (Fig. 5B and SI Appendix, Fig. S10) showed that both the volume density (Fig. 5C) and the surface area (Fig. 5D) of TTs coupled to the SR were significantly increased by myocardin overexpression. We also observed that the length of individual TT–SR junctions was prolonged (Fig. 5E) and the TT–SR distance was shortened (Fig. 5F) in the myocardin overexpression group. Of note, comparison between Figs. 5 C–E and 2 B–E showed that TT–SR couplons were degraded during cell culture, and myocardin expression partially rescued these changes. Taken together, these results indicated that the myocardin-mediated JPH2/CAV3 up-regulation indeed tightened the nanospace between LCC-residing TTs and RyR-residing SRs.

Fig. 5.

Myocardin overexpression reproduced the enhanced LCC–RyR signaling. (A) Relative mRNA expression levels of JPH2 and CAV3 in S. dauricus cardiomyocytes infected with adenoviral (Ad-) RFP or Ad-myocardin. (B) Representative TEM images of S. dauricus cardiomyocytes infected with Ad-RFP or Ad-myocardin. The yellow arrows denote TTs. (C) Stereological analysis of volume density per unit volume of TTs coupled with the SR and JSRs. (D) Stereological analysis of surface area. C and D measured ≥130 images (≥5 animals) per group. (E) Curvilinear length of couplons. (F) TT–SR distance. Data in E and F were from ≥200 images per group. *P < 0.05 and **P < 0.01 by Student’s t test. (G) Representative whole-cell I_Ca recordings (Upper) and confocal imaging of Ca²⁺ transients (Lower) in myocardin overexpressed cardiomyocytes of euthermic ground squirrels. (H) Whole-cell I_Ca compared between the Ad-RFP and Ad-myocardin groups. (I) Ca²⁺ transient amplitude. (J) CICR gain calculated as the Ca²⁺ transient amplitude per unit I_Ca density. Data in H–J were from ≥10 cells per group. *P < 0.05 and **P < 0.01 by two-way ANOVA with repeated measures.

The structural changes of TT–SR junctions are expected to improve the performance of LCC–RyR Ca²⁺ signaling. We therefore characterized the effect of myocardin overexpression on CICR efficiency by whole-cell patch clamp and Ca²⁺ imaging (Fig. 5G). Although myocardin overexpression did not alter I_Ca density (Fig. 5H), it increased the Ca²⁺ transient amplitude by 49% (Fig. 5I). Therefore, the gain of CICR, calculated as the Ca²⁺ transient amplitude per unit I_Ca, was significantly increased (Fig. 5J). These results demonstrated that myocardin overexpression, via concerted up-regulation of JPH2 and CAV3, recapitulated the enhanced E-C coupling function observed in hibernating ground squirrels.

Discussion

In the present study, we systemically studied the molecular mechanisms that up-regulate the efficiency of Ca²⁺ signaling in cardiomyocytes of hibernating ground squirrels. We discovered that increased expression of JPH2 and CAV3 in hibernation tightened TT–SR junctions and facilitated LCC–RyR communication. Molecular analyses revealed that SRF and myocardin controlled the transcription of JPH2 and CAV3 concurrently. Myocardin overexpression in cardiomyocytes to mimic the increased expression of myocardin during hibernation reproduced the up-regulations of JPH2/CAV3 expression and LCC–RyR signaling efficiency observed in hibernating ground squirrels. These findings identified myocardin as a main switch in tuning the tone of cardiac Ca²⁺ signaling and E-C coupling (SI Appendix, Fig. S11) and suggested a concept of concerted transcription, which may be helpful for maintaining the stoichiometry of interacting partners. We and other laboratories have shown that JPH2 function is also regulated by posttranscriptional (14, 16) and posttranslational (17) mechanisms. These mechanisms comprise a diversified molecular toolkit to tune the efficiency of intermolecular Ca²⁺ signaling. Hibernating ground squirrel is so far the only model utilizing this molecular toolkit to reinforce the efficiency of E-C coupling.

Most mammals maintain body temperature within a narrow range around 37 °C at which the thermodynamic biochemical processes have been systemically optimized during evolution for coordinated and efficient operation; consequently, the homeostasis of endotherms is sensitive to temperature changes (25, 42). Intracellular Ca²⁺ overload occurring under hypothermia and other stress conditions is a common threat for all living cells (25–29). Ca²⁺ overload in heart cells leads to severe arrhythmia or ventricular fibrillation, as well as heart arrest (27, 31–33). To safely reduce their body temperatures and survive sustained deep hypothermia, hibernating mammals must avoid intracellular Ca²⁺ overload while maintaining vigorous Ca²⁺ transients to drive heart contraction. The remodeling of Ca²⁺ signaling during hibernation solves this question by at least three interdependent mechanisms. First, the decrease in the LCC current (34, 35) curtails the duration of the action potential and lowers the risks of Ca²⁺ overload and arrhythmogenesis. Second, enhanced LCC–RyR signaling efficiency, as revealed in the present study, compensates for the reduced LCC trigger, ensuring a full-sized Ca²⁺ transient. Third, the up-regulation of SERCA2 activity (27, 35) supports a quick relaxation and suffices the releasable store of SR Ca²⁺. These evolutionary adaptions of Ca²⁺ signaling perfectly solve the intuitively contradictory question to keep diastolic Ca²⁺ low and systolic Ca²⁺ high and optimize the balance between Ca²⁺ homeostasis and heart contraction. Given the importance of heart function, the above remodeling of Ca²⁺ signaling of hibernating animals represents a fundamental mechanism for their survival during the winter season.

In contrast to the “gain-of-function” remodeling of Ca²⁺ signaling in hibernation, the remodeling of Ca²⁺ signaling in hypertrophied and failing heart cells proceeds in the “loss-of-function” direction (11, 43, 44). During the pathological development of hypertrophy and heart failure, the structural basis and functional components of Ca²⁺ signaling appear to rewind toward the neonatal pattern. First, the TT network is disrupted, at least partially due to reduced JPH2 expression (8, 40, 45). Second, RyRs become “orphaned” (i.e., lose LCC control) due to TT retraction (43). Third, SR Ca²⁺ storage becomes insufficient due to decreased SERCA2 and increased NCX1 activities (46). As a result, there is a progressive decline of LCC–RyR signaling efficiency eventually leading to end-stage heart failure. Although enormous efforts have been made to study the transition from hypertrophy to heart failure, whether this pathological process can be retarded or reversed is still not known.

The gain-of-function remodeling of Ca²⁺ signaling during hibernation provides therapeutic insights for heart failure. This is not only because the regulation patterns are opposite in hibernation and in heart failure, but also because hibernation is a reversible process. Hibernating mammals usually enter hibernation in early winter, and return to euthermic state in the spring. With the euthermia–torpor cycle, intermolecular Ca²⁺ signaling between LCCs and RyRs switches between high-gain and low-gain modes to meet distinct physiological demands. Based on our present study, the expression level of myocardin is at least one of the key events in switching the Ca²⁺ signaling between these two modes. It will therefore be intriguing to explore whether the myocardin–SRF machinery can be used to reverse the E-C uncoupling in failing heart cells. Further dissection of the complex JPH2/CAV3 regulation under related pathological conditions will likely lead to novel strategies for treating defective E-C coupling in heart diseases.

Materials and Methods

Animals.

All animal experiments were approved by the Peking University Institutional Committee for Animal Care and Use. Ground squirrels (Spermophilus dauricus) were sampled in Hebei Province, China. The euthermic and torpid states of S. dauricus were compared in winter to avoid the involvement of possible seasonal changes. At least one hibernation bout was observed as a test of their capability of hibernation before sample collections. Torpid ground squirrels were reared at ∼5 °C in a constant dark room. Euthermic ground squirrels kept euthermic for at least 3–4 d in a warm room.

Cardiomyocyte Isolation.

Ventricular cardiomyocytes were enzymatically isolated as previously described (8, 45). Briefly, the hearts were rapidly removed under anesthesia and perfused at 37 °C for 5 min with a Ca²⁺-free Tyrode’s solution containing (in mM) 135 NaCl, 4 KCl, 1 MgCl₂, 1.2 NaH₂PO₄, 11.9 NaHCO₃, 10 glucose, and 10 taurine saturated with 95% O₂/5% CO₂. Collagenase Type II (0.7 mg/mL) (Worthington Biochemical) was then added to the perfusate. After 15–20 min, the ventricle was cut into small pieces. Finally, the myocytes were centrifuged and resuspended in Tyrode’s solution containing 0.5 mM Ca²⁺.

Cell Culture and Adenovirus Infection.

Isolated adult heart cells were resuspended in prebalanced M199 medium and plated at a density of 5 × 10⁵ cells/dish on laminin-coated glass coverslips. Isolated neonatal cardiomyocytes were preplated in a 10-cm dish for 2 h to remove fibroblasts and then cultured in Dulbecco’s modified Eagle medium (Corning) with 10% fetal bovine serum (HyClone Laboratories) at a density of 1 × 10⁶ cells/dish. Targeted gene sequences were cloned and inserted into pAd/CMV/V5–DEST gateway vector (V493-20, Invitrogen), and the adenovirus was packaged using a ViraPower adenoviral expression kit (K493000, Invitrogen) in HEK293A cells. Cultured cardiomyocytes were infected with adenovirus for ∼12 h with the titer carefully controlled. Biochemical and physiological experiments were performed 48 h after infection.

Whole-Cell Patch-Clamp Recording and Ca²⁺ Imaging.

I_Ca and Ca²⁺ transient recordings were performed as previously described (8). Isolated cardiomyocytes were bathed in Tyrode’s solution with 1 mM CaCl₂ and 0.02 mM tetrodotoxin and 4 mM 4-aminopyridine. The pipette solution contained (in mM) 110 CsCl, 6 MgCl₂, 5 Na₂ATP, 15 TEA-Cl, 0.2 Fluo-4 potassium, and 10 Hepes, pH 7.2 adjusted with CsOH. Line-scan Ca²⁺ images were acquired using confocal microscope (Carl Zeiss). The Ca²⁺ concentration was indexed by the fluorescence normalized to the resting level (R = F/F₀). The change of cell length was derived from edge-detection of the fluorescence.

Transmission electron microscopy.

Isolated ventricular cardiomyocytes were prepared for transmission electron microscopy (TEM) as previously described (8). Ultrathin sections were prepared at a thickness of ∼80 μm along the longitudinal axis of cardiomyocytes and stained with uranyl acetate and lead citrate. Images were randomly captured under an FEI Tecnai G² 20 electron microscope and measured as previously described (8).

Western Blot Analysis.

Ventricular tissues were homogenized in standard lysis buffer. Proteins were electrophoresed on 10% SDS–polyacrylamide gels and transferred to polyvinylidene fluoride membranes. After blocking in 5% skim milk, the membranes were incubated overnight at 4 °C with the respective primary antibodies. After washed three times, the membranes were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody for 1 h.

Quantitative RT-PCR.

Total RNA was extracted from ventricular tissues using TRIzol reagent (Invitrogen). The first-strand complementary DNA (cDNA) was synthesized using SuperScript III reverse transcriptase (Invitrogen). Quantitative RT-PCR was performed in an Mx3000p real-time PCR detection system (Stratagene) with Brilliant II SYBR Green (Stratagene) to label the amplified DNA (47). Detailed primer information was listed in SI Appendix, Table S2. The reference gene sequences of S. dauricus have been uploaded to ftp.ensembl.org (48).

5′ Rapid Amplification of cDNA Ends.

cDNA was synthesized using SuperScript III reverse transcriptase (Invitrogen) and was used in the reaction with TdT (New England Biolabs) to add a poly(A) tail in the presence of ATP. The tailed cDNA was used as the template in the first round of PCR using PrimSTAR (Takara) (49). A second round of PCR was performed with the Taq master mix (TSingKe) using the first-round PCR product as the template. The PCR product was purified and cloned into a pClone007 simple vector (TSingKe) for sequencing.

EMSA.

EMSAs were performed using the LightShift Chemiluminescent EMSA Kit (ThermoFisher Scientific). Nuclear proteins were extracted from HEK293T cells overexpressing SRF using NE–PER nuclear and cytoplasmic extraction reagents (ThermoFisher Scientific). A total of 1 nM biotin-labeled CArG-like element oligonucleotides of the JPH2 promoter (5′-CGA​CGC​GTG​GCCT​TGT​TAG​GGCG​CTA​GCT​AG -3′) and 1 μL nuclear extract were incubated in 17 μL mixed-buffer containing 1 × binding buffer (25 μM Hepes, 16 mM KCl, 50 mM NaCl), 8% Glycerol, 50 ng poly dIdC, and 2 μM ZnCl₂ for 20 min at room temperature. Nuclear extract was first incubated with an anti-SRF antibody (sc-25290, Santa Cruz Biotechnology) for 15 min before incubation with biotin-labeled oligonucleotides added to assess the supershift. Excessive wild-type and mutant CArG-like element oligonucleotides (5′-CGA​CGC​GTG​GCAG​CGT​CGT​GCGC​TAG​CTA​G-3′) without biotin labeling were added to the mix to test the binding specificity. The samples were separated on 5% PAGE (SDS-free) in 0.25 × tris borate-ethylenediaminetetraacetic acid (TBE) buffer, transferred to a nylon membrane in 0.5 × TBE buffer. The membrane was then immediately ultravioletradiation (UV) cross-linked at 120 mJ/cm² in a UV-light cross-linking instrument. Detection and analysis were performed according to the protocol of LightShift Chemiluminescent EMSA Kit (ThermoFisher Scientific).

ChIP–qPCR Assay.

Chromatin immunoprecipitation assays were performed using a ChIP IT high-sensitivity kit (Active Motif). Neonatal rat cardiomyocytes were fixed with 1% formaldehyde, and chromatin was extracted and sonicated using a BioRuptor (Diagnode). Sonicated chromatin was precipitated using 8 μg IgG (sc-2025, Santa Cruz Biotechnology) or SRF antibody (sc-25290, Santa Cruz Biotechnology) bound to protein G Sepharose. After washing and elution, chromatin was then reverse cross-linked, followed by the purification of genomic DNA. The JPH2 and CAV3 promoter region flanking binding sites were served as target regions, and the distal regions ∼1 and 2 kb from the transcription start site were served as control regions. The enrichment of target and control regions was quantified via qPCR using EVA green mix (Abm) in both the precipitated and input samples.

Reporter Plasmids and Luciferase Assays.

Upstream sequences of JPH2 and CAV3 of different lengths were amplified from the rat genomic extract and were cloned into pGL-3 plasmids (Promega). The mutations in the CArG-like element were generated by overlap-PCR and also cloned into pGL-3 plasmids. The coding sequences of SRF and myocardin were amplified and cloned into pcDNA 3.1 plasmid. For the transfection of neonatal rat cardiomyocytes, 500 ng of the constructed pGL-3 plasmid together with 500 ng pcDNA 3.1 plasmids were delivered using Lipofectamine 2000 (Invitrogen). A total of 200 ng of the PhRL–TK plasmid (Promega) was cotransfected with the constructed plasmids as a control. After 48 h transfection, the cells were washed with 500 μL phosphate buffered saline and lysed with 100 μL passive lysis buffer. Then, 15 μL of Luciferase Assay Reagent II (Promega) was added, and the luciferase activity was measured using a Varioskan Flash spectral scanning multimode reader (ThermoFisher Scientific). Then, 15 μL Stop & Glo Reagent (Promega) was added, and the Renilla luciferase activity was measured as a control. Detailed primer information was listed in SI Appendix, Table S1.

Data Statistics.

Data are expressed as means ± SEM. Statistical analysis was performed using Student’s t test, two-way ANOVA or Mann–Whitney rank-sum test. P value less than 0.05 was considered statistically significant.

Data Availability

All study data are included in the article and/or SI Appendix.