Autonomous activation of CaMKII exacerbates diastolic calcium leak during beta-adrenergic stimulation in cardiomyocytes of Metabolic Syndrome rats

Abstract

Autonomous Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) activation induces abnormal diastolic Ca²⁺ leak, which leads to triggered arrhythmias in a wide range of cardiovascular diseases, including diabetic cardiomyopathy. In hyperglycemia, Ca²⁺ handling alterations can be aggravated under stress conditions via the β-adrenergic signaling pathway, which also involves CaMKII activation. However, little is known about intracellular Ca²⁺ handling disturbances under β-adrenergic stimulation in cardiomyocytes of the prediabetic metabolic syndrome (MetS) model with obesity, and the participation of CaMKII in these alterations.

MetS was induced in male Wistar rats by administering 30% sucrose in drinking water for 16 weeks. Fluo 3-loaded MetS cardiomyocytes exhibited augmented diastolic Ca²⁺ leak (in the form of spontaneous Ca²⁺ waves) under basal conditions and that Ca²⁺ leakage was exacerbated by isoproterenol (ISO, 100 nM). At the molecular level, [³H]-ryanodine binding and basal phosphorylation of cardiac ryanodine receptor (RyR2) at Ser2814, a CaMKII site, were increased in heart homogenates of MetS rats with no changes in RyR2 expression. These alterations were not further augmented by Isoproterenol. SERCA pump activity was augmented 48 % in MetS hearts before β-adrenergic stimuli, which is associated to augmented PLN phosphorylation at T17, a target of CaMKII. In MetS hearts. CaMKII auto-phosphorylation (T287) was increased by 80%. The augmented diastolic Ca²⁺ leak was prevented by CaMKII inhibition with AIP. In conclusion, CaMKII autonomous activation in cardiomyocytes of MetS rats with central obesity significantly contributes to abnormal diastolic Ca²⁺ leak, increasing the propensity for β-adrenergic receptor-driven lethal arrhythmias.

Graphical Abstract

1. Introduction

Unhealthy contemporary lifestyles and dietary habits characterized by excessive consumption of carbohydrates and insufficient physical activity are associated with an alarming increase in the prevalence of obesity and cardiovascular diseases (CVDs) worldwide; the latter has become a global public health issue [1]. An early, adverse consequence of these unhealthy practices is the emergence of metabolic pathologies, which include Metabolic Syndrome (MetS). MetS represents a key risk factor for the development of Type II Diabetes Mellitus (DM2) and CVDs because the syndrome brings with it a series of such biochemical and physiological characteristics as obesity, high blood pressure, hypertriglyceridemia, low high-density lipoprotein cholesterol (HDL-c) levels, and insulin resistance with no alterations in fasting glucose levels [2–4]. The development of central obesity is considered an essential characteristic of MetS because it is an early step in the etiological cascade of the syndrome, linking waist circumference with CVD [5]. Thus, MetS is considered a pre-diabetic state and a key risk factor for heart disease [3]. Indeed, the prevalence of CVD events in subjects diagnosed with MetS is 31.9 %, but only 10.8 % in healthy subjects [6]. However, the molecular mechanisms underlying the higher incidence of CVD events in MetS patients, either at resting or under stress conditions (for example, in the presence of β-adrenergic receptor stimulation), are not fully elucidated.

Disturbances in the activity of cardiac Ca²⁺ handling proteins have been reported in insulin resistance and diabetic cardiomyopathy—this is particularly true of proteins involved in excitation-contraction coupling (ECC) [7,8]. In the heart, ECC begins with Ca²⁺ influx via voltage-dependent L-type Ca²⁺ channels (VDCCs). This Ca²⁺ entry activates the intracellular Ca²⁺ channel/ryanodine receptor (RyR2) by Ca²⁺-induced Ca²⁺ release, prompting a transient increase in the intracellular Ca²⁺ concentration ([Ca²⁺]_i), which is followed by cell contraction. Relaxation takes place due to a Ca²⁺ clearance process that primarily involves (1) Ca²⁺ recapture into the sarcoplasmic reticulum (SR) Ca²⁺ stores through the action of the SR Ca²⁺ ATPase (SERCA pump) and (2) Ca²⁺ extrusion through the sarcolemmal Na⁺/Ca²⁺ exchanger (NCX) [9]. The β-adrenergic response, induced by stress or strenuous physical activity, is the main regulatory pathway of the ECC mechanism [10,11]. Consequently, several ECC proteins—including VDCCs, RyRs, and phospholamban (PLN)—are phosphorylated by cAMP-dependent protein kinase (PKA), and Ca²⁺/calmodulin-dependent kinase type II (CaMKII), which increases their activity.

CaMKII has been proposed as a key kinase contributing to the deleterious effects of chronic β-adrenergic receptor activation in cardiac pathologies, mainly by exacerbating RyR2-mediated diastolic Ca²⁺ leak [12,13]. Studies in experimental models of insulin resistance and fructose fed-induced pre-diabetic cardiomyopathy have unveiled the participation of hyperglycemia and ROS in inducing abnormal CaMKII activation, the latter has been associated with alterations in cardiomyocyte Ca²⁺ handling and with the appearance of cardiac arrhythmic events [14–16]. Hyperglycemia leads to CaMKII glycosylation, increasing RyR2-mediated Ca²⁺ leak, and reducing SR Ca²⁺ load in cardiac cells [14]. The duration of CaMKII activation relies on the frequency of Ca²⁺ release events. Prolongation of CaMKII activation is related to auto-phosphorylation of the kinase at T287: this prevents CaMKII auto-inhibition, which causes the kinase to remain active even when [Ca²⁺]_i declines, constituting a mechanism for CaMKII autonomous activation [13,17].

Remarkably, the role of CaMKII in Ca²⁺ handling alterations of non-hyperglycemic prediabetic cardiomyocytes from an experimental model of MetS with obesity and under β-adrenergic stress has not been previously studied. Therefore, we investigated (1) the effects of β-adrenergic-mediated CaMKII activation in Ca²⁺ handling alterations of MetS cardiomyocytes and (2) the underlying CaMKII-related molecular mechanisms.

2. Materials and methods

2.1. Sucrose-induced Metabolic Syndrome model

Recently weaned male Wistar rats were divided into two groups and maintained in sterile conditions with a 12 h dark-light cycle of 12 h and controlled temperature of 22 ± 2 °C. The control group (CTL) received plain water, while the Metabolic Syndrome group (MetS) received 30% sucrose (refined commercial sugar) via drinking water; both groups were fed with commercial rat chow ad libitum (PicoLab Rodent Diet 20, LabDiet St. Louis, MO, USA). The animals were used for experimentation after 16 weeks of treatment.

2.2. Measurement of body and serum parameters

After fasting overnight, the rats were treated with 1000 U/kg heparin (Pisa, Guadalajara, México) and anesthetized with sodium pentobarbital (Pisa, Guadalajara, México), as previously described [18]. Measurements of serum glucose, triglycerides (TG), and HDL-Cholesterol (HDL-C) were obtained from whole blood samples using glucose and lipid panel strips (PTS lipid panel and PTS glucose strips) with the CardioCheck PA analyzer (PTS Diagnostics, Indianapolis, IN, USA). Other parameters such as body weight, tibia length, heart weight, left ventricle weight, visceral fat weight, and blood pressure were also determined.

2.3. Ventricular myocyte isolation

Rat ventricular myocytes were obtained using a previously reported enzymatic perfusion method [19]. Briefly, hearts were rapidly excised from anesthetized animals and placed in ice-cold, oxygenated Tyrode solution containing (in mM: NaCl 130, KCl 5.4, MgCl₂ 0.5, NaH₂PO₄ 0.4, glucose 22, HEPES 25; pH 7.4 with NaOH). Aortas were cannulated above the aortic valve and perfused with Tyrode solution supplemented with 0.23 mM EGTA for 5 min at 37 °C. Enzyme solution containing 1 mg/mL of collagenase Type II (Worthington Biochemical Corporation, Lakewood, NJ, USA) in Tyrode solution supplemented with 0.1 mM CaCl₂ was perfused until aortic valves were digested, as exhibited by increased efflux of perfusate. Hearts were then transferred to a tube containing enzyme solution supplemented with 10 mg/mL of BSA and gently shaken for 3 min at 37 °C to disperse myocytes. The resulting cell suspension was centrifuged for 3 min at 170 × g. The cell sedimentation was repeated two times, while CaCl₂ concentration on Tyrode solution was gradually increased (in mM: 0.25 and 0.5). Finally, the cell pellet was suspended in storage Tyrode solution supplemented with 0.8 mM CaCl₂.

2.4. Confocal Ca²⁺ Imaging

Isolated ventricular myocytes were loaded with the fluorescent Ca²⁺ indicator Fluo-3 AM (final concentration, 5 μM. Thermo Scientific Inc., Waltham, MA USA) for 25 min at room temperature. Fluo 3-loaded cardiomyocytes were placed in a field stimulation chamber with platinum electrodes (Warner Instruments LLC, Hamden CT, USA) attached to a perfusion system with valve controller (Model VC-6M, Warner Instruments LLC, Hamden CT, USA) and filled with recording Tyrode solution (in mM: NaCl 130, CaCl₂ 1, MgCl₂ 2.0, KCl 5.4, glucose 11, HEPES 10, pH 7.4 with NaOH). To activate the β-adrenergic response, Isoproterenol (ISO) was added to the recording Tyrode solution (final concentration 100 nM). Autocamptide-2 related inhibitory peptide (AIP, Cat N. SCP0001, Sigma Aldrich) was used in its myristoylated form for better membrane permeability. Cells were preincubated with AIP (10 μM, 1 h) in storage Tyrode solution before Ca²⁺ signaling recordings. Ca²⁺ sparks and Ca²⁺ waves were recorded in quiescent cardiomyocytes using a confocal microscope (Zeiss LSM700, Carl Zeiss Microscopy GmbH, Jena, Germany) in line-scan mode (1.5 ms/line). For this procedure, Fluo 3-loaded cardiomyocytes were paced at 0.5 Hz, electrical stimulation was stopped and confocal recordings (10 images per cell) were performed in quiescent cells after 5 min, in a lapse time of 20 min before washing the recording chamber and adding new Fluo 3-loaded cells. Ca²⁺ spark and Ca²⁺ wave frequencies were calculated as the ratio of the total number of events per cell divided by the recording time (events/s). Parameters such as amplitude (F/F₀), full width at half maximum (FWHM, in μm), full duration at half maximum (FDHM, in ms), time-to-peak (ms), and decay time constant (in ms) were measured using IDL version 5.5 software (Research Systems Inc., Boulder CO, USA) for image analysis, running a specialized home-made protocol written by Ana Maria Gómez (Inserm UMR-S 1180, LabEx LERMIT, Châtenay-Malabry, France). Ca²⁺ spark-mediated Ca²⁺ leak was calculated by multiplying spark frequency*amplitude*duration*width (as described by Biesmans et al. [20]). To determine SR Ca²⁺ load, cells were paced for at least 5 min at 0.5 Hz using an electric field stimulator (set a 70 V with a pulse duration of 20 ms) to restore steady-state SR Ca²⁺ load, after which a 10 mM caffeine stimulus was applied through a perfusion system with valve controller (Model VC-6M, Warner Instruments LLC, Hamden CT, USA). The confocal images were analyzed using Origin 8.1 software (Origin Lab Corporation, Northampton MA, USA) and SR Ca²⁺ load was measured by the amplitude of the caffeine-evoked Ca²⁺ transient.

2.5. Determination of basal [Ca²⁺]_i

Diastolic [Ca²⁺] was determined using an inverted microscope (Nikon Diaphot) coupled to a PTI microfluorometry system (RatioMaster™, PTI Technologies Inc. Oxnard CA, USA). Isolated ventricular cardiomyocytes were incubated with Fura-2 AM for 25 min at room temperature and washed. Fura 2-loaded cells were placed in a field stimulation chamber with platinum electrodes; the chamber was filled with recording Tyrode solution supplemented with CaCl₂ 1.8 mM. The cardiomyocytes were electrically stimulated at a frequency of 0.5 Hz before Fura-2 fluorescence signal was recorded. Diastolic [Ca²⁺]i was calculated using the Grynkiewicz equation [21]. The values of Rmax (7.01) and Rmin (0.22) were obtained in the presence of 2.5 mM of Ca²⁺ and 2 mM of EGTA, respectively; the Kd used was 282 nM.

2.6. Preparation of left ventricle homogenates

Left ventricle homogenates were prepared as previously described [18] with some modifications. Excised rat hearts were placed in ice-cold Tyrode solution. Aortas were cannulated, and hearts were placed in a Langendorff perfusion system. Hearts were retroperfused with warm (37 °C) oxygenated Tyrode solution supplemented with EGTA 0.23 mM for 5 min to stop beating and clean the remaining blood. Hearts were then retroperfused with Tyrode solution containing CaCl₂ 1.8 mM to restore extracellular Ca²⁺ levels. The β-adrenergic response was stimulated with ISO 1 μM for an additional 5 min. At the end of the perfusion protocol, left ventricles (LV) were isolated and frozen in liquid N₂ before pulverization. The tissue powder was homogenized in a Potter-Elvehjem glass homogenizer at 400 rpm for 45 s with lysis buffer (in mM: HEPES 20 and sucrose 0.3, pH 7.2. with KOH) supplemented with protein inhibitors: leupeptin 12 μM, PMSF 100 μM, benzamidine 500 μM, and aprotinin 0.153 μM, and with phosphatase cocktail inhibitor 3 (MilliporeSigma Corp. St. Louis MO, USA). LV homogenates were centrifuged at 2000 g for 10 min, and supernatants aliquoted and frozen at −80°C. Protein concentrations were determined using the Lowry method.

2.7. [³H]-ryanodine binding assay

To determine the maximum number of active RyRs, [³H]-ryanodine binding assays were performed using LV homogenates as previously reported [19,22], with some modifications. Briefly, 50 μg of protein per tube were incubated with 7 nM [³H]-ryanodine (PerkinElmer, Waltham MA, USA) for 90 min at 37°C in an incubation medium containing (in mM): 200 KCl, 20 HEPES (pH 7.2 with KOH), and CaCl₂ necessary to set free Ca²⁺ at 10 μM, as calculated by MaxChelator (http://maxchelator.stanford.edu/index.html). The samples were filtered onto glass fiber filters (Whatman GF-B, GE Healthcare, Chicago IL, USA) and washed three times with distilled water in an automatic collector (Brandel M24-R). The filters were placed in scintillation vials with 5 mL of liquid scintillation mixture, and the retained radioactivity was measured in a Beckman LS-6500 counter. The specific [³H]-ryanodine binding was defined as the difference between the binding in the absence (total binding) and the presence (nonspecific binding) of 20 μM unlabeled ryanodine (MP Biomedicals LLC, Santa Ana CA, USA).

2.8. ATPase activity assay

Inorganic phosphate (Pi) production due to ATP cleavage by thapsigargin-sensitive ATPases was determined as previously described [19], following the protocol reported by Bartolommei et al., [23] with some modifications. LV homogenates (100 μg of protein per tube) were incubated in buffer solution with ~10 μM free [Ca²⁺] (in mM: KCl 80, MgCl₂ 3, CaCl₂ 0.2, sodium azide 5, K-EGTA 0.2, MOPS 25, and A23187 0.002 at pH 7.0 adjusted with TRIS) at 37 °C for 5 min. The enzymatic reaction was initiated by adding 1mM ATP. Aliquots of 0.1 ml were taken after 5, 10, 15 and 30 min and were immediately added to 0.9 ml of Pi coloring solution (in mM: H₂SO₄ 125, ascorbic acid 10, ammonium-heptamolybdate tetrahydrate 0.5 and potassium antimony (III) tartrate hydrate 0.04). Absorbance was measured after 10 min at 850 nm (Beckman Spectrophotometer DU800). Control experiments were performed in the presence of 100 nM thapsigargin. The difference between total and thapsigargin-insensitive ATPase activities was considered SERCA pump hydrolytic activity.

2.9. SDS-PAGE and Western Blots

Gradient polyacrylamide gels (4–12%) were loaded with the indicated amount of protein from LV homogenates in Laemmli buffer. The SDS-PAGE resolved proteins were transferred to PVDF membranes that were subsequently blocked for 1 h at room temperature with low fat dry milk 5% in PBS-T or TBS-T buffer when indicated; PVDF membranes were then incubated with corresponding primary antibodies against: RyR2 (ab2827, Abcam, Cambridge MA, USA), pS2808-RyR2 (A010–30, Badrilla, Leeds UK), pS2814-RyR2 (A010–31, Badrilla, Leeds UK), SERCA2 (ab2861, Abcam, Cambridge MA, USA), PLN (A010–14, Badrilla, Leeds UK), pS16-PLN (A010–12, Badrilla, Leeds UK), pT17-PLN (A010–13, Badrilla Leeds UK), CaMKII (4436s, Cell Signaling, Danvers MA, USA), pT287-CaMKII (12716s, Cell Signaling. Danvers MA, USA); and GAPDH (Cat# AM4300, Ambion®, Thermo Fisher Scientific, Waltham, MA, USA) for 2 h at room temperature. After washing, the PVDF membranes were incubated with corresponding secondary conjugated-HRP antibody (goat anti-mouse IgG Cat. # 401215 and goat anti-rabbit IgG Cat. # 401315, MilliporeSigma Corp. St. Louis MO, USA) for 1 h at room temperature. Proteins were visualized by chemiluminescent reaction (SuperSignal® West Pico Chemiluminescent Substrate, Thermo Fisher Scientific, Waltham MA, USA) and the relative amount of protein was determined by densitometric analysis using ImageJ software (v. 1.50i NIH, USA).

2.10. AIP treatment of cardiomyocytes

Isolated ventricular myocytes were plated in 6-well plates with 1 mL of Tyrode solution supplemented with CaCl₂ 1 mM. After 10 min at 37 °C, cells were treated with AIP 10 μM for 1 h at 37 °C. The β-adrenergic response was induced with ISO 100 nM for an additional time of 10 min. At the end of the stimulation protocol, cells were collected and transferred to Eppendorf tubes and centrifuged at 3,000 rpm for 1 min. The supernatants were discarded, and pellets were immediately frozen in liquid N₂ and stored at −80 °C until use.

Cell pellets were resuspended in 100 μL of Laemmli buffer 1X, vortexed and sonicated for 40 s at 40 kHz. Then, the samples were heated at 99 °C for 5 min and centrifuged at 14,500 rpm for 5 min, supernatants were aliquoted and stored at −20 °C for future use in Western Blots. Protein was quantified using the RC-DC protein assay kit (BioRad) following the manufacturer’s instructions.

2.11. Statistical data analysis

Data are presented as the mean ± standard error of the mean (M ± S.E.M). The number of animals for each experiment is designated by “N”, while the number of cells or independent experiments is indicated as “n”. The Shapiro-Wilk normality distribution test was used to determine the Gaussian distribution of the values. When sets of data failed normality distribution test, non-parametric tests were used to evaluate statistical significance, either Mann-Whitney rank sum test or Kruskal-Wallis One-Way ANOVA on ranks followed by Dunn’s post-hoc test when appropriate. For sets of data with a normal Gaussian distribution, statistical significance was evaluated by two-tailed Student’s t-test or One-Way ANOVA followed by Tukey post-hoc test when appropriate. A P value of ≤ 0.05 was considered statistically significant. All analyses were performed using Origin Pro v.8 software (Origin Lab Corporation, Northampton MA, USA) or Sigma Plot v11.0 (Systat Software Inc., San Jose, CA, USA).

3. Results

3.1. Characteristics of Metabolic Syndrome rats

Administration of 30 % sucrose in the drinking water of male Wistar rats for 16 weeks caused many common features of prediabetic Metabolic Syndrome, including a significant rise in body weight (32 %) and a 3.1-fold increase in abdominal fat accumulation compared to the control group (Table 1). Although the heart-weight-to-body-weight ratio was decreased in MetS rats, features such as left ventricle weight and left ventricle-to-heart weight ratio of the rats showed no alterations, ruling out any cardiac remodeling produced by the sucrose diet (Table 1). In MetS rats, dyslipidemia was evidenced by a 2-fold increase in total triglyceride levels and a reduction in HDL-cholesterol levels (15 %), with no alteration in serum fasting glucose levels compared with control animals (Table 1). There was a 3.4-fold increase in the TG-to-HDL-C ratio in MetS rats, which is indicative of higher cardiometabolic risk in the insulin-resistant condition.[24].

Table 1.

General characteristics of control and Metabolic Syndrome animals.

	EXPERIMENTAL GROUP
Parameters	CTL	MetS
Body Weight (g)	398.1 ± 10.0	524.8 ± 35.0**
Abdominal Fat (g)	2.7 ± 0.2	11.2 ± 0.8**
Heart Weight (g)	1.22 ± 0.04	1.28 ± 0.04
HW/BW ratio x 100	0.30 ± 0.01	0.25 ± 0.03**
LV Weight (g)	0.81 ± 0.02	0.85 ± 0.03
LV/HW ratio	0.67 ± 0.02	0.67 ± 0.02
Tibia Length (cm)	4.2 ± 0.2	4.4 ± 0.4
Triglycerides (mg/dL)	94.3 ± 15.0	269.4 ± 18.0**
HDL-C (mg/dL)	37.2 ± 2.0	31.7 ± 5.0*
Glucose (mg/dL)	68.6 ± 6.6	78.5 ± 4.1
TG/HDL-C ratio	2.6 ± 0.4	8.8 ± 0.9**

Values are mean ± S.E.M of N=11 control rats (CTL), and N=10 metabolic syndrome (MetS) rats. BW, body weight; HDL-C, high-density lipoprotein cholesterol; HW, heart weight; LV, left ventricle; TG, triglycerides.

^*P ≤ 0.05 and

^**P ≤ 0.001 vs. CTL animals.

3.2. Abnormal diastolic Ca²⁺ leak in MetS cardiomyocytes is exacerbated after β-adrenergic stimulation

The primary outcomes of cardiometabolic risk in MetS patients are myocardial infarction, coronary and peripheral arterial diseases, congestive heart failure, arrhythmia, and stroke [7]. Previous studies have documented a direct association between abnormal diastolic Ca²⁺ leak in cardiomyocytes and Ca²⁺-driven cardiac arrhythmias [22,25]. The evidence of lethal arrhythmogenic activity in MetS rats (Supplementary Figure 1), also previously showed in sucrose-fed animals [26–28], prompted us to examine whether intracellular Ca²⁺ dynamics were compromised in this metabolic condition. A method for studying potential alterations of RyR2 activity in situ is by analyzing diastolic Ca²⁺ release events, such as spontaneous Ca²⁺ sparks and Ca²⁺ waves. Figures 1A and 1B show representative bi-dimensional images of these Ca²⁺ release events recorded with a confocal microscope in isolated Fluo 3-loaded cardiomyocytes. Recordings were taken under quiescent conditions and in the presence (or in the absence) of the β-adrenergic receptor agonist, isoproterenol (ISO). Under basal conditions (no ISO added), Ca²⁺ wave frequency was significantly higher in MetS cells compared with CTL cells (Figure 1C and Figure 1D), evidencing abnormal diastolic Ca²⁺ leak in the form of spontaneous Ca²⁺ waves in MetS cardiomyocytes.

Figure 1. Exacerbated diastolic Ca²⁺ leak induced by the β-adrenergic stimulus in isolated MetS cardiomyocytes.

Representative pseudo-colored confocal images of Ca²⁺ spark (A) and Ca²⁺ waves (B) recorded in quiescent Fluo-3 loaded cardiomyocytes from CTL and MetS rats in the presence of isoproterenol (ISO, 100 nM) or in its absence. C. Bar graph of average Ca²⁺ spark frequency. D. Ca²⁺ wave frequency (events/s). E. Caffeine-evoked Ca²⁺ transient amplitude (ΔF/F₀) of Fluo-3 loaded cardiomyocytes from CTL and MetS rats, perfused with recording solution in the presence of isoproterenol (ISO, 100 nM) or in its absence. F. Bar graph of diastolic [Ca²⁺]_i determined in cardiomyocytes loaded with Fura-2 of CTL and MetS rats before and after ISO administration. All values are presented as mean ± SE. Ca²⁺ sparks and Ca²⁺ waves data are obtained from n=30 cells from N= 6 rats for each experimental condition. Caffeine-evoked Ca²⁺ transient data are obtained from n_CTL= 16 cells, n_CTL+ISO= 11 cells, n_MetS= 13 cells and n_MetS+ISO= 9 cells, from N = 4 rats for each experimental condition. [Ca²⁺]_i data are obtained from n_CTL= 9 cells, n_CTL+ISO= 8 cells, n_MetS= 10 cells and n_MetS+ISO= 8 cells, from N = 3 rats for each experimental condition. *P ≤ 0.05 with respect to the CTL group, † P ≤ 0.05 with respect to the CTL+ISO group and # P ≤ 0.05 with respect to the MetS group.

As expected, ISO treatment increased Ca²⁺ spark frequency in cardiomyocytes of both experimental groups. Interestingly, the augmentation in Ca²⁺ spark frequency was larger in MetS cardiomyocytes (34 %) than in the CTL cells (20 %) (Figure 1C). However, ISO treatment significantly increased the Ca²⁺ spark-mediated Ca²⁺ leak (38 % respect to basal condition) in MetS cells only (Table 2). It is important to note that the Ca²⁺ spark amplitude remained unchanged in all the conditions, however, Ca²⁺ spark duration and width are decreased in MetS cell with and without ISO (Table 2). Histogram distributions of Ca²⁺ spark properties showed that the proportion of events with reduced duration and size is augmented in MetS cells may be due to the propensity of RyR clusters to be recruited to generate Ca²⁺ waves (Supplementary Figure 2 and Supplementary Table 1). Further analysis showed that β-adrenergic stimulus exacerbates already abnormal Ca²⁺ release in the form of spontaneous Ca²⁺ waves in MetS cardiomyocytes (Figure 1D), which attenuate changes in Ca²⁺ sparks properties (Supplementary Table 1). For cardiomyocytes of CTL group, we found that these modifications in diastolic Ca²⁺ dynamics are related to a significant increase in the SR Ca²⁺ load. The SR Ca²⁺ load was not modified in the MetS condition (Figure 1E, Supplementary Figure 3). Basal [Ca²⁺] did not change in any of the experimental groups (Figure 1F). These data suggest that in MetS cardiomyocytes, mainly Ca²⁺ wave-mediated Ca²⁺ leak under basal conditions could be exacerbated by the β-adrenergic stimulus and may underlie cardiac arrhythmogenic activity.

Table 2.

Characteristics of Ca²⁺ spark recorded in control (CTL) and metabolic syndrome (MetS) cardiomyocytes in the absence and presence of Isoproterenol (ISO).

	Experimental condition
Ca2+ spark properties	CTL	CTL+ISO	MetS	MetS+ISO
Amplitude (F/F0)	2.01 ± 0.04	2.11 ± 0.03	2.01 ± 0.03	2.01 ± 0.03
FDHM (ms)	60.3 ± 2.09	52.8 ± 1.34*	55.7 ± 1.80*	52.8 ± 1.33*
FWHM (μm)	3.14 ± 0.05	3.01 ± 0.04	2.92 ± 0.05*	2.90 ± 0.04*
RT (ms)	17.6 ± 0.79	17.3 ± 0.66	15.6 ± 0.65	17.2 ± 0.60
TTP (ms)	32.6 ± 2.15	29.2 ± 1.50	27.8 ± 1.53	30.7 ± 1.39
TAU (ms)	90.7 ± 8.55	73.7 ± 5.80	85.6 ± 8.1	71.7 ± 5.88
Spark-mediated Ca2+ leak	709.5 ± 52.8	715. 2 ± 38.6	600.2 ±32.9	828.2 ±37.8**,#

Values are mean ± S.E.M of n=30 cells from N= 6 rats for each experimental condition.

FDHM, full duration at half maximum, FWHM, full width at half maximum; RT, rising time; TTP, time-to-peak; Ca²⁺ leak, calculated as the product of Ca²⁺ spark frequency*Ca²⁺ spark amplitude*FWHM* FDHM. Ca²⁺ spark data are obtained from n_CTL= 339 events, n_CTL+ISO= 471 events, n_MetS= 421 events and n_MetS+ISO= 600 events.

^*P ≤ 0.05 and

^**P ≤ 0.001 vs. CTL cells,

^#P ≤ 0.05 vs. MetS cells.

3.3. Increased basal activity of RyR2s in MetS hearts is not further augmented by the β-adrenergic stimulus.

A possible explanation for the elevated incidence of diastolic Ca²⁺ release events in the MetS cardiomyocytes despite a similar SR Ca²⁺ load is the fact that activating the β-adrenergic signaling pathway increases RyR2 activity in situ [29]. To explore this possibility, we performed [³H]ryanodine binding assays in heart homogenates of MetS and CTL rats. Our results showed that under basal conditions (no ISO added) the maximum amount of active RyR2 at optimal [Ca²⁺] (5 μM) was higher (47 %) in MetS preparations with respect to controls; this was not further increased by ISO treatment (Figure 2A). These results suggest that the abnormal Ca²⁺ leak in MetS may be due to exacerbated basal activity of RyR2s. Next, we explored the molecular mechanisms underlying these changes. Because increases in the expression level or the phosphorylation status of RyR2s may account for the augmentation in RyR2 activity [30], we measured total RyR2 protein expression and RyR2 phosphorylation status by Western Blot. Figure 2B shows representative immunoblots of phosphorylated and total RyR2s from heart homogenates, both untreated and treated with ISO. There was no significant difference in the total amount of RyR2s under all experimental conditions (Figure 2C); this effectively rules out alterations in the expression of RyR2 as an explanation for the augmented basal RyR2 activity observed in MetS hearts.

Figure 2. Increased activity of Ryanodine Receptors in MetS associated with its phosphorylation levels.

A. Normalized [³H]-Ryanodine binding Bmax determined at pCa 5, pH 7.2 in CTL and MetS heart homogenates in the presence (1 μM) or absence of isoproterenol (ISO). B. Representative Western Blot images of total and phosphorylated RyR2 at S2808, S2814 and GAPDH, obtained from heart homogenates from CTL and MetS rats in the presence (1 μM) or absence of ISO. C. Total ryanodine receptor (RyR) expression normalized against corresponding GAPDH from N = 5 rats for each experimental condition. D. Phosphorylation levels of RyR at S2808 normalized against corresponding total RyR from N = 6 rats for each experimental condition. E. Phosphorylation levels of RyR at S2814 normalized against corresponding total RyR from N= 5 rats for each experimental condition. Data are expressed as mean ± SE. *P ≤ 0.05 with respect to CTL group and † P ≤ 0.05 with respect to the CTL + ISO group.

Conversely, the state of RyR2 phosphorylation at different residues can modify its activity [29]; therefore, we evaluated the phosphorylation levels of S2808 and S2814, two RyR2 sites identified as targets of PKA/CaMKII and CaMKII kinases, respectively. Basal phosphorylation levels of S2808 were substantially elevated—even in the absence of β-adrenergic stimulus; regardless, ISO treatment increased this parameter only for the CTL group (Figure 2D). Our data agree with previous work reporting that the phosphosite S2808 is at least ~50% constitutively phosphorylated [29,31]. Notably, MetS heart preparations under resting conditions had 59 % higher phosphorylation of S2814 than CTL heart homogenates; this augmented phosphorylation was not further increased by ISO treatment (Figure 2E).

3.4. Increased SERCA pump activity in MetS hearts under resting conditions compensates for augmented diastolic Ca²⁺ leak.

The abnormal diastolic Ca²⁺ leak observed in MetS cardiomyocytes under resting conditions may be associated with augmented basal RyR2 activity. However, caffeine-induced Ca²⁺ transients showed that SR Ca²⁺ load was not modified in MetS cardiomyocytes; this observation prompted us to evaluate whether perturbation of the SR Ca²⁺ recapture mechanism could explain the increase in Ca²⁺ availability for RyR2 release. The assessment of SERCA pump hydrolytic activity revealed that its Ca²⁺ recapture function was significantly increased in the hearts of MetS rats under basal conditions; this was not further increased with the β-adrenergic challenge. Conversely, ISO treatment did increase ATPase hydrolytic activity in CTL preparations (Figure 3A and 3B). The abnormal SERCA pump ATPase activity observed in MetS heart preparations was not due to changes in SERCA2 pump expression (Figure 3C and 3D).

Figure 3. SERCA pump ATPase activity is increased in MetS hearts under resting conditions.

A. SERCA pump ATPase hydrolytic activity evaluated by enzymatic assay in heart homogenates of CTL and MetS rats in the presence (1 μM) or absence of isoproterenol (ISO) B. Bar graph of SERCA2 pump reaction rate obtained from enzymatic assay. C. SERCA2 protein expression normalized against corresponding GAPDH and representative Western blot. Data are expressed as mean ± SE, N= 5 rats for each experimental condition. *P ≤ 0.05 and **P ≤ 0.001 respect to the CTL group.

3.5. CaMKII-induced phospholamban phosphorylation underlies augmented SERCA pump activity in MetS

Considering that phospholamban (PLN) is the main regulatory protein of SERCA pump activity [32], we determined the total protein expression levels of PLN. We also investigated PLN phosphorylation status at two different sites: S16 and T17, PKA-dependent and CaMKII-dependent phosphorylation sites, respectively. Figure 4A shows representative PLN immunoblots of phosphorylated and total PLN from heart homogenates, treated (or not) with ISO. Our results show that under resting conditions, PLN was not phosphorylated at S16 in either experimental group; activation of the β-adrenergic response significantly increased S16-PLN phosphorylation (Figure 4B). Notably, T17 site was found to be phosphorylated under basal conditions in MetS heart preparations; further, β-adrenergic stimulus significantly increased T17-PLN phosphorylation only in CTL preparations (Figure 4C). Total PLN expression and PLN/SERCA pump ratio were similar for both experimental groups (PLN/SERCA pump, normalized data: 1.0 ± 0.13 in n = 5 control samples vs. 0.80 ± 0.13 in n=5 MetS heart preparations, P < 0.3154).

Figure 4. Augmented T17-PLN phosphorylation in MetS.

A. Representative Western Blot images of total SERCA2, PLN, and phosphorylated PLN at S16 and T17, obtained from heart homogenates from CTL and MetS rats in the presence (1 μM) or absence of isoproterenol (ISO). B. Phosphorylation levels of PLN at S16 normalized against corresponding total PLN. C. Phosphorylation levels of PLN at T17 normalized against corresponding total PLN. Data are expressed as mean ± SE, N= 5 animals for each experimental condition. *P ≤ 0.05 with respect to the CTL group † P ≤ 0.05 with respect to the CTL + ISO group, and # P ≤ 0.05 with respect to the MetS group.

3.6. CaMKII is constitutively active in MetS hearts

The phosphorylation pattern of S2814-RyR2 and T17-PLN provided important insight into the participation of CaMKII in these post-translational modifications [33]; thus, we determined CaMKII expression and CaMKII auto-phosphorylation levels at the T287 site using Western Blot analysis (Figure 5). We primarily evaluated this post-translational modification because of its participation in the autonomous activation mechanism of CaMKII [13]. Figure 5A shows representative CaMKII immunoblots of pT287-CaMKII and total CaMKII from heart homogenates, untreated or treated with ISO. Our results demonstrate that under basal conditions, CaMKII auto-phosphorylation levels were significantly augmented (80%) in MetS heart homogenates compared to CTL samples (Figure 5C). ISO treatment did not increase the phosphorylation of T287 in CaMKII in either experimental group (Figure 5C). Moreover, CaMKII protein expression levels were unchanged in MetS homogenates (Figure 5B). Thus, CaMKII is constitutively active in MetS hearts: this autonomous activation may play a key role in the abnormal activity of RyR2s and SERCA pump under resting conditions.

Figure 5. Autonomous activation of CaMKII is increased in MetS hearts and is unchanged by ISO.

A. Representative Western Blot images of total and phosphorylated CaMKII at T287 and GAPDH, obtained from heart homogenates from CTL and MetS rats in the presence (1 μM) or absence of isoproterenol (ISO). B. CaMKII protein expression normalized against corresponding GAPDH C. Phosphorylation levels of CaMKII at T287 normalized against corresponding total CaMKII. Data are expressed as mean ± SE, N= 5 animals for each experimental condition. *P ≤ 0.05 with respect to the CTL group.

3.7. CaMKII inhibitor AIP reduces diastolic Ca²⁺ leak in MetS cardiomyocytes

Although the augmented auto-phosphorylation levels of T287-CaMKII suggested a role for CaMKII in the abnormal diastolic Ca²⁺ leak observed in MetS cardiomyocytes, further evidence was required. To better understand CaMKII role in cardiovascular pathologies, KN-93 has been used as its inhibitor of choice; however, this methoxy-benzene-sulfonamide mainly blocks Ca-CaM-mediated CaMKII activation not its autonomous activation [34]. Thus, we used autocamptide-2-related inhibitory peptide (AIP) to inhibit CaMKII. We measured properties of Ca²⁺ sparks and Ca²⁺ waves in cardiomyocytes pre-incubated with AIP (10 μM, 1 h) (Figure 6A and 6B). In the presence of AIP, Ca²⁺ spark frequency augmented in the presence of the β-adrenergic stimulus (Figure 6C). Importantly, the abnormal diastolic Ca²⁺ leak in the form of Ca²⁺ waves in MetS cardiomyocytes, observed before and after the β-adrenergic stimulus, was now prevented.

Figure 6. Incubation with AIP prevents exacerbated diastolic Ca²⁺ leak induced by the β-adrenergic stimulus in isolated MetS cardiomyocytes.

Representative pseudo-colored confocal images of Ca²⁺ sparks (A) and Ca²⁺ waves (B) recorded in quiescent Fluo-3 loaded cardiomyocytes from CTL and MetS rats pre-incubated with AIP 10 μM, in the presence or absence of isoproterenol (ISO, 100 nM). Bar graphs of Ca²⁺ spark frequencies (C, in events/s * 100 μm) and Ca²⁺ wave frequencies (D, events/s) for the different experimental conditions. E. Caffeine-evoked Ca²⁺ transient amplitude (ΔF/F₀) of Fluo 3-loaded cardiomyocytes from CTL and MetS rats pretreated with AIP (10 μM), and perfused with recording solution in the presence of isoproterenol (ISO, 100 nM) or in its absence. F. Bar graph of basal fluorescence (F₀) determined in Fluo 3-loaded, AIP-treated cardiomyocytes of CTL and MetS rats before and after ISO administration. All values are presented as mean ± SE. Ca²⁺ sparks, Ca²⁺ waves and F₀ data are obtained from n=25 cells from N= 6 rats for each experimental condition. Caffeine-evoked Ca²⁺ transient data are obtained from n_CTL= 9 cells, n_CTL+ISO= 6 cells, n_MetS= 9 cells and n_MetS+ISO= 9 cells, from N = 3 rats for each experimental condition. *P ≤ 0.05 respect to the CTL group, † P ≤ 0.05 respect to the CTL + ISO group and # P ≤ 0.05 respect to the MetS group.

Interestingly, although pre-incubation with AIP increased Ca²⁺ wave frequency in the CTL group, it inhibited the Ca²⁺ wave release in MetS cardiomyocytes (Figure 6D); further, preincubation with AIP prevented the increase in Ca²⁺ waves previously observed after β-adrenergic stimulus (Figure 1D).

In addition, Ca²⁺ spark amplitude and Ca²⁺ spark size were significantly diminished while Ca²⁺ spark time-to-peak was augmented by AIP in MetS cells, either treated or not with ISO (Table 3); suggesting a reduction of RyR2 activity. Thus, AIP treatment significantly decreased the Ca²⁺ spark-mediated Ca²⁺ leak in MetS cells regardless of ISO challenge. Histogram distributions of Ca²⁺ spark properties in AIP-treated cells (Supplementary Figure 4 and Supplementary Table 2) showed that the proportion of events with reduced amplitude and size is augmented in MetS cells. The AIP effect on the aforementioned Ca²⁺ spark properties remained even with ISO challenge; thus, the outcome was a significant decrease in Ca²⁺ spark-mediated Ca²⁺ leak. Additional Ca²⁺ spark properties were unaffected in MetS cells after exposure to AIP (Table 3). It is important to note that the differences observed in Ca²⁺ spark and Ca²⁺ wave incidence before and after ISO treatment were not associated with disturbances in SR-Ca²⁺ load or basal fluorescence (Figure 6E and 6F). The inhibition of abnormal diastolic Ca²⁺ leak in AIP-treated MetS cardiomyocytes is consistent with a significant reduction of RyR S2814 phosphorylation by CaMKII inhibition (Supplementary Figure 5) respect to basal MetS condition (pRyR S2814/RyR2 normalized data: 0.77 ± 0.12 in N = 5 AIP-treated cardiomyocyte samples vs. 1.66 ± 0.26 in N = 5 MetS heart preparations; P = 0.036).

Table 3.

Characteristics of Ca²⁺ spark recorded in control (CTL) and metabolic syndrome (MetS) cardiomyocytes exposed to AIP (10 μM, 1 h); and in the absence or presence of Isoproterenol (ISO).

	Experimental condition
Ca2+ spark properties	CTL	CTL+ISO	MetS	MetS+ISO
Amplitude (F/F0)	2.00 ± .006	2.05 ± 0.05	1.59 ± 0.04*,†	1.75 ± 0.05*,†
FDHM (ms)	37.7 ± 1.77	33.4 ± 1.44	39.7 ± 2.01	35.2 ± 1.79
FWHM (μm)	2.71 ± 0.08	2.55 ± 0.07	2.11 ± 0.08*,†	2.25 ± 0.08*,†
RT (ms)	17.3 ± 1.14	18.2 ± 01.09	20.6 ± 1.16	20.1 ± 1.30
TTP (ms)	19.7 ± 1.23	18.7 ± 1.08	21.8 ± 1.62*,†	23.1 ± 2.09*,†
TAU (ms)	65.2 ± 11.9	96.2 ± 17.1	107 ± 86.9	115 ± 32.2
Spark-mediated Ca2+ leak	203.8 ± 19.5	270.7 ± 33.1	144.6 ±16.0*,†	232.6 ±42.5*,†

Values are mean ± S.E.M of n=25 cells from N= 5 rats for each experimental condition.

FDHM, full duration at half maximum; FWHM, full width at half maximum; RT, rising time; TTP, time-to-peak; Ca²⁺ leak, calculated as the product of Ca spark frequency*Ca spark amplitude*FWHM* FDHM. Ca²⁺ spark data are obtained from n_CTL= 166 events, n_CTL+ISO= 194 events, n_MetS= 165 events and n_MetS+ISO= 186 events.

^*P ≤ 0.05 vs. CTL cells,

^†P ≤ 0.05 vs. CTL+ISO-treated cells and

^#P ≤ 0.05 vs. MetS cells.

4. Discussion

In this study, we examined the role of CaMKII in Ca²⁺ handling alterations of cardiomyocytes from MetS animals under the β-adrenergic response. Our results support the notion that CaMKII is autonomously activated in MetS hearts and is the underlying mechanism of the abnormal diastolic Ca²⁺ leak in the form of spontaneous Ca²⁺ waves observed in cardiomyocytes under resting conditions. Our results also demonstrate that β-adrenergic signaling pathway exacerbates the abnormal SR Ca²⁺ leak in MetS cardiomyocytes.

Our experimental MetS model was developed in rats by sucrose administration (30% in drinking water) for 16 weeks, which differs from our previous study in a MetS model at late stages [18]. In contrast with other sucrose-fed animal models [26,35], our rats displayed increased body weight, which was associated with significant abdominal fat accumulation (Table 1). The latter is a key component of our MetS model, particularly considering that obesity is a central factor of MetS pathophysiology in humans and crucial in predisposing to the development of CVD [36]. Similar to sucrose-fed rats [26], the MetS rats also developed dyslipidemia (Table 1), a metabolic condition independently related to stroke and/or myocardial infarction in MetS patients [37].

Clinical studies have also reported that MetS patients have a well-established hyper-adrenergic state that could account for CVD morbidity and mortality [7,38]. Previous reports in HTG and MetS rats reported arrhythmic events and impaired cardiac sinus rhythm recovery after in vivo or ex-vivo ischemia-reperfusion procedures [26–28]. Accordingly, MetS rats were more prompted to develop ventricular fibrillation after the administration of an arrhythmogenic cocktail (Supp. Fig 1) It is well documented that abnormal activation of CaMKII signaling pathways triggers cardiac arrhythmogenic activity in several congenital and acquired cardiac pathologies [14,39–41], and that this condition could be exacerbated under β-adrenergic stress. Indeed, resting MetS cardiomyocytes exhibited abnormal diastolic Ca²⁺ leak mainly in the form of Ca²⁺ waves (Figure 1). Interestingly, ISO effect on Ca²⁺ spark frequency was evident in both experimental groups. It is important to note that Ca²⁺ waves contribute to diastolic Ca²⁺ leak (Figure 1D); further, Ca²⁺ waves contribute more significantly to inward NCX current, thereby promoting arrhythmogenic activity [9]. Compelling evidence shows an association between Ca²⁺ wave formation and SR Ca²⁺ content threshold, which can be shifted toward spontaneous Ca²⁺ release due to changes in RyR2 properties [42] or increased SR Ca²⁺ content [43]. However, our data do not support the latter, because SR Ca²⁺ content, measured by the delta of the amplitude of the caffeine-evoked Ca²⁺ transients, remained unmodified in MetS cardiomyocytes even in the presence of ISO (Figure 1E), as it has been demonstrated in other CVD models [22]. Ca²⁺ spark frequency and morphometric properties are determined by SR Ca²⁺ load, cytoplasmic [Ca²⁺]_i, factors that control the opening and closure of the RyR2 channel; for instance, single current amplitude, RyR Ca²⁺ sensitivity, regulatory proteins, posttranslational modifications, and point mutations, among others [22,44–47]. Also, cytosolic and intra SR Ca²⁺ diffusion and clearance mechanisms are participating [44,47]. In normal cardiomyocytes and under very controlled conditions, β-adrenergic stimulus increases the potential for the RyRs to release a given amount of Ca²⁺ abruptly, as Ca²⁺ sparks [45], augmenting their frequency as it was observed in CTL+ISO and MetS+ISO cells. ISO did not modify Ca²⁺ spark amplitude in both experimental groups (Table 2), as previously reported in cardiomyocytes with similar SR Ca²⁺ load [44]. Other Ca²⁺ spark properties such as FWHM and FDHM that were expected to increase with ISO [45] in our conditions were reduced (FDHM in CTL+ISO cells) or remain similar (FWHM in both CTL+ISO and MetS+ISO cells) respect to control cells (Table 2). A possible explanation for our results in MetS cells, is that the augmented Ca²⁺ wave-mediated Ca²⁺ leak is preventing a modification in other Ca²⁺ spark characteristics.

Alterations in the properties of RyR2 are also known to trigger the occurrence of Ca²⁺ waves. Among the several post-translational modifications that the RyR2 undergoes, its phosphorylation via PKA and CaMKII is the primary mechanism for regulating its Ca²⁺ sensitivity and in situ activity. Thus, changes in RyR2 phosphorylation have been frequently associated with the occurrence of cardiac arrhythmic events [29,48]. In this regard, a study using a dog model of high-fat-induced MetS demonstrated a decrease in cardiac RyR2 activity and significantly increased phosphorylation levels at S2808 [49]. In contrast, a study with a murine MetS model (sucrose-fed) by Okatan et al., reported an increase in S2808-RyR2 phosphorylation associated with disturbances in cardiomyocyte Ca²⁺ handling, where RyR2 was more “leaky” during diastole. Further, the authors found that SERCA pump activity was depressed via a mechanism involving SERCA oxidation [35,50]. Despite these discrepancies, RyR2 phosphorylation clearly plays a crucial role in the cardiac MetS outcome. In our hands, we found only differences in S2808-RyR2 phosphorylation in the CTL condition after ISO treatment (Figure 2D), which agrees well with the considerable levels of S2808-RyR2 phosphorylation we observed under basal conditions [31,48].

In contrast, we found significant phosphorylation of the S2814-RyR2 site in MetS hearts under basal conditions (Figure 2E); this likely accounted for the increase in diastolic Ca²⁺ leak in MetS cardiomyocytes. Our results agree with previous reports associating this specific post-translational modification with RyR2- mediated Ca²⁺ release [12,51]. Likely, in a fructose-fed rat model, S2814-RyR2 basal phosphorylation levels were increased with a clear association with abnormal diastolic Ca²⁺ leak [15]; however, no evidence of ISO effects on S2814-RyR2 phosphorylation in this metabolic model was provided. Interestingly, Bers et al., [48,52] found that ISO treatment only increased S2814-RyR phosphorylation in rat cardiomyocytes after electrical stimulation, which could help to explain why β-adrenergic stimuli did not further increase S2814-RyR2 phosphorylation levels in either MetS or CTL hearts (Figure 2E).

Some studies using various prediabetic and diabetic models reported loss of SERCA pump function [18,35,53]. In contrast, SERCA pump activity was reportedly upregulated in a fructose-induced MetS model at early stages, which was interpreted as an initial compensatory mechanism in the heart [54]. Considering that we found abnormal diastolic Ca²⁺ leak without alterations of SR Ca²⁺ content in our experimental MetS model, we hypothesized that the SERCA pump activity would be upregulated. Indeed, our data showed that SERCA pump activity was increased in MetS hearts (Figure 3A and 3B). Intriguingly, the SERCA activity of MetS hearts was quite similar to the SERCA ATPase activity determined in ISO-stimulated hearts (in both CTL and MetS conditions), though PLN phosphorylation levels were alike (Figure 4). However, it has been previously demonstrated that either PKA- or ISO-induced stimulation of SERCA ATPase activity is clearly more effective at low free Ca²⁺ concentrations (~ 0.7 μM) that at optimal free Ca²⁺ (≥ 10 μM) in cardiac SR vesicles [55,56]. Because in the present work, the experimental protocol to measure the SERCA ATPase activity was optimized to obtain the maximal hydrolytic activity (no substrate restriction, ~10 μM free Ca²⁺ concentration; and in the presence of a Ca²⁺ ionophore to dissipate the calcium gradient), we assume that any increase in PLN phosphorylation will rise SERCA ATPase activity to the maximum level, thus no gradual changes in ATPase activity would be observed. In the future, a better approach to determine the fine regulation of PLN phosphorylation on SERCA pump hydrolytic activity could be accomplished by reducing the free Ca²⁺ concentration to suboptimal levels (≤ 1 μM) in the ATPase assay.

As mentioned previously, CaMKII has several Ca²⁺ handling targets, among them PLN, which constitutes an important regulator of the SERCA pump and therefore a key component of cardiomyocyte relaxation mechanisms. Importantly, we found that PLN-Thr17 phosphorylation was augmented in MetS (Figure 4C). This finding agrees with similar results obtained in fructose-fed rats [15]; and is in disagreement with a report in sucrose-fed MetS model, where PLN phosphorylation status did not explain SERCA pump dysfunction [35].

Given that both S2814-RyR2 and T17-PLN are sites exclusively phosphorylated by CaMKII [13], the finding of higher levels of phosphorylation of these sites clearly indicates that CaMKII plays a key role in the abnormal diastolic Ca²⁺ leak and arrhythmogenic activity in MetS. We validated the contribution of CaMKII by evaluating its auto-phosphorylation state (Figure 5C). Therefore, MetS upregulates this post-translational modification, which results in the long-term potentiation of CaMKII activity—no other stimuli are required. Our results are in line with those reporting increased CaMKII auto-phosphorylation levels after a high-carbohydrate diet [15,57]. Is important to note that autonomous activation of CaMKII by T287 auto-phosphorylation is a mechanism promoted by the frequency of the Ca²⁺ release events that prevails until the dephosphorylation of the site by PP1 or PP2a phosphatases [13]. Hence, Western blot results showed that only in the MetS condition this autonomous activation mechanism is prevalent, however this does not mean that CaMKII is not activated in other conditions.

Importantly, inhibition with the peptide AIP served as further evidence of CaMKII participation in the increased Ca²⁺ leak (Figure 6) induced by S2814 RyR phosphorylation (Supplementary Figure 5) in the MetS condition. Despite the fact that the inhibition of abnormal diastolic Ca²⁺ leak in AIP-treated MetS cardiomyocytes was not associated with disturbances in SR-Ca²⁺ load or diastolic (F₀) fluorescence (Figure 6), further investigation will be required to determine in-depth whether AIP treatment alters diastolic [Ca²⁺].We are aware that this MetS model entails a variety of pathophysiological implications that must be approached in the future. For instance, it is well known that oxidative stress is a consequence of obesity and dyslipidemia that is associated with the cardiometabolic complications of these conditions [58]. Given that obesity and dyslipidemia are main components of our MetS model, it would be interesting to evaluate the oxidative status in the MetS cardiomyocytes, particularly of those Ca²⁺ handling related proteins susceptible to oxidative post-translational modifications (Ox-PTMs) such as the RyR and CaMKII. Oxidative stress can exacerbate SR-Ca²⁺ leak directly by enhancing RyR activity and indirectly by contributing to CaMKII autonomous activation mechanism [40,59]. Thus, it would be interesting to unveil the association of obesity-induced oxidative stress in the intracellular Ca²⁺ disturbances observed in our MetS model. However, as far as we know, T287-CaMKII autophosphorylation is primarily driven by high Ca² dynamics, ruling out the participation of ROS in the autonomous activation of CaMKII in our experimental MetS model. This statement is supported by experiments in a sucrose-fed model where oxidized levels of CaMKII were not changed despite the increased ROS production [57].

Lately, abnormal CaMKII activation has been associated with diabetic cardiomyopathy and late-stage metabolic syndrome [14,18,60]. In this matter, it is worthy to mention a previous report from our group in a late-stage MetS model (30% sucrose in drinking water for 24 weeks); in which an impairment in RyR and SERCA pump activation due to decreased S2814-RyR and T17-PLN phosphorylation, respectively, was observed, in a similar way as the results in AIP-treated MetS cells, indicating a down-regulation in CaMKII activation[18]. Taken together, these results suggest a pivotal role of CaMKII in the progression of prediabetic cardiomyopathy. At early stages of MetS, when heart energy demand is still not compromised, CaMKII is upregulated; while at late stages of MetS, with marked diminution of ATP, phosphocreatine content; and creatine kinase activity [61], CaMKII is downregulated [18].

5. Conclusion

Our results support the notion that CaMKII autonomous activation in MetS hearts significantly contributes to exacerbating diastolic Ca²⁺ leak (mainly in the form of Ca²⁺ waves) after β-adrenergic stimulation. Main findings of this work are: 1) abnormal Ca²⁺ wave-driven diastolic Ca²⁺ leak in MetS cardiomyocytes is exacerbated after β-adrenergic stimulation; 2) RyR2 and PLN phosphorylation at CaMKII target sites underlie augmented diastolic Ca²⁺ leak in MetS cardiomyocytes; 3) auto-phosphorylation of T287-CaMKII accounts for CaMKII autonomous activation in MetS hearts; 4) conversely, pharmacological AIP-induced CaMKII inhibition reduces diastolic Ca²⁺ leak in myocytes from MetS hearts. Finally, exacerbated CaMKII activation in the heart could account for detrimental β-adrenergic responses at early stages of MetS.

HIGHLIGHTS.

• MetS cardiomyocytes show increased diastolic Ca²⁺ leak under basal conditions.

• Diastolic Ca²⁺ leak of Metabolic Syndrome (MetS) cardiomyocytes is exacerbated under β-adrenergic stimulation.

• Enhanced RyR and PLN basal phosphorylation by CaMKII underlies abnormal Ca²⁺ leak associated with MetS condition.

• Autonomous activation of CaMKII is involved in Ca²⁺ disturbances as evidenced by AIP reversion.

Abbreviations

AIP

Autocamptide-2-related inhibitory peptide

CaMKII

Ca²⁺/calmodulin-dependent protein kinase II

CICR

Calcium-induced calcium release

CTL

Control

CVD

Cardiovascular disease

EADs

Early after depolarizations

ECC

Excitation-contraction coupling

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

HDL

High-density lipoprotein

ISO

Isoproterenol

LV

Left ventricle

MetS

Metabolic Syndrome

NCX

Na⁺/Ca²⁺ exchanger

PKA

Protein Kinase A

PLN

Phospholamban

RyR

Ryanodine receptor

SERCA

Sarco/Endoplasmic reticulum Ca²⁺ ATPase

SR

Sarcoplasmic reticulum