Protein carbonylation causes sarcoplasmic reticulum Ca2+ overload by increasing intracellular Na+ level in ventricular myocytes

Abstract

Diabetes is commonly associated with an elevated level of reactive carbonyl species due to alteration of glucose and fatty acid metabolism. These metabolic changes cause an abnormality in cardiac Ca²⁺ regulation that can lead to cardiomyopathies. In this study, we explored how the reactive α-dicarbonyl methylglyoxal (MGO) affects Ca²⁺ regulation in mouse ventricular myocytes. Analysis of intracellular Ca²⁺ dynamics revealed that MGO (200 μM) increases action potential (AP)-induced Ca²⁺ transients and sarcoplasmic reticulum (SR) Ca²⁺ load, with a limited effect on L-type Ca²⁺ channel-mediated Ca²⁺ transients and SERCA-mediated Ca²⁺ uptake. At the same time, MGO significantly slowed down cytosolic Ca²⁺ extrusion by Na⁺/Ca²⁺ exchanger (NCX). MGO also increased the frequency of Ca²⁺ waves during rest and these Ca²⁺ release events were abolished by an external solution with zero [Na⁺] and [Ca²⁺]. Adrenergic receptor activation with isoproterenol (10 nM) increased Ca²⁺ transients and SR Ca²⁺ load, but it also triggered spontaneous Ca²⁺ waves in 27% of studied cells. Pretreatment of myocytes with MGO increased the fraction of cells with Ca²⁺ waves during adrenergic receptor stimulation by 163%. Measurements of intracellular [Na⁺] revealed that MGO increases cytosolic [Na⁺] by 57% from the maximal effect produced by the Na⁺-K⁺ ATPase inhibitor ouabain (20 μM). This increase in cytosolic [Na⁺] was a result of activation of a tetrodotoxin-sensitive Na⁺ influx, but not an inhibition of Na⁺-K⁺ ATPase. An increase in cytosolic [Na⁺] after treating cells with ouabain produced similar effects on Ca²⁺ regulation as MGO. These results suggest that protein carbonylation can affect cardiac Ca²⁺ regulation by increasing cytosolic [Na⁺] via a tetrodotoxin-sensitive pathway. This, in turn, reduces Ca²⁺ extrusion by NCX, causing SR Ca²⁺ overload and spontaneous Ca²⁺ waves.

INTRODUCTION

Regular heart contraction critically depends on well-controlled intracellular Ca²⁺ ([Ca²⁺]_i) homeostasis. In adult ventricular myocytes, the majority of Ca²⁺ that triggers contraction during systole is released from the sarcoplasmic reticulum (SR) through the ryanodine receptor (RyR2) Ca²⁺ release channel. It occurs as a result of RyR2 activation by Ca²⁺ influx through the voltage-gated L-type Ca²⁺ channels (LTCC) during the action potential [6]. During diastole, the excessive cytosolic Ca²⁺ is removed by the sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA2a) and the Na⁺/Ca²⁺ exchanger (NCX) [6], so heart can relax and fill with blood. It is well established that defects in [Ca²⁺]_i regulation can lead to arrythmias and contractile dysfunctions in many cardiac pathologies [1, 2, 5, 7, 8, 12, 37]. In cardiomyocytes, several important sarcolemmal transport mechanisms, including NCX, are coupled with the inward Na⁺ gradient. The gradient is established and maintained by the Na⁺/K⁺ ATPase (NKA), also called Na⁺ pump [32]. As a result, an inhibition of the Na⁺ pump causes accumulation of intracellular Na⁺ ([Na⁺]_i) and, therefore, reduction of Ca²⁺ extrusion by NCX. Selective and moderate inhibition of NKA with glycosides has been traditionally used as a strategy to increase [Ca²⁺]_i and heart contraction in patients with chronic heart failure (HF) [18]. On the another hand, excessive [Na⁺]_i accumulation during myocardial infarction plays a critical role in cytosolic Ca²⁺ overload and myocardial injury [21]. Furthermore, several models of HF are associated with abnormal Na⁺ regulation and increased [Na⁺]_i [13, 22]. Thus, well-controlled intracellular Ca²⁺ and Na⁺ homeostasis are critically important for proper heart function.

More than 500 million people around the world live with type 1 and type II diabetes, the chronic conditions that affect the ability of the body to produce or sense insulin and, therefore, control of the blood glucose level. Diabetes is commonly associated with an elevated level of reactive carbonyl species such as α-dicarbonyl methylglyoxal (MGO) in serum and urine [11, 33]. Accumulation of MGO is mainly a result of hyperglycemia and increased oxidation of fatty acid. While a low level of reactive carbonyl species is necessary for some physiological processes such as cell growth, immune defense, and differentiation, an excessive concentration of MGO is a hallmark of several pathological conditions associated with neurodegeneration and diabetes [28]. At its high level, MGO can react with several amino acids (including proline, arginine, lysine, and threonine) to form glycated adducts that compromise the function of the altered proteins. If blood glucose levels are not properly controlled, patients with diabetes can develop cardiovascular diseases, including diabetic cardiomyopathy. These cardiomyopathies are usually associated with diastolic dysfunction as well as ventricular arrhythmias [17, 25, 19]. Carbonylation of several key Ca²⁺ transport systems have been suggested contributing to [Ca²⁺]_i dysregulation during diabetic cardiomyopathy. It has been shown that carbonylation of SERCA2a and RyR2 reduces SR Ca²⁺ uptake and desynchronize SR Ca²⁺ release, causing diastolic dysfunction in the type 1 diabetic rat model [29, 30]. NKA is also a suitable substrate for carbonylation in renal proximal tubules and this post translational modification reduces NKA function [35]. Despite its clinical significance, the effect of reactive carbonyl species on cardiac [Ca²⁺]_i and [Na⁺]_i regulation are still not fully understood.

The aim of this study was to determine mechanisms by which MGO affects intracellular Ca²⁺ and Na⁺ homeostasis in ventricular myocytes. By measuring intracellular [Ca²⁺] and [Na⁺] dynamics and NKA activity, we found that MGO affects intracellular Ca²⁺ regulation by increasing [Na⁺]_i, primarily due to activation of a tetrodotoxin (TTX)-sensitive Na⁺ inward pathway. This, in turn, reduces the NCX-mediated Ca²⁺ extrusion, causing SR Ca²⁺ overload and pro-arrhythmogenic Ca²⁺ waves.

MATERIALS and METHODS

Isolation of ventricular myocytes.

All animal experiments were carried out in accordance with the National Institutes of Health guide for the care and use of Laboratory animals [20]. Male and female C57Bl6/J mice, (11 of male and 8 of female animals, Jackson Laboratories) were housed according to approved IACUC guidelines. Mice aged between 2–3 months were anesthetized using Isoflurane (1%). Following thoracotomy, hearts were quickly excised, immersed in Ca²⁺ free buffer, mounted on a Langendorff apparatus, and retrogradely perfused with the zero Ca²⁺ Tyrode buffer (in mM: NaCl 140; KCl 4; MgCl2 1; glucose 10; Hepes 10; pH 7.4). (37°C) containing Liberase H (Roche), according to a procedure described previously [9, 14]. The left ventricle was excised from the digested heart, placed in stop buffer containing BSA 1 mg/mL, cut into several pieces (average size 1 mm) and gently triturated into single cells. Myocytes were pelleted by gravity and resuspended in the low-Ca²⁺ Tyrode buffer (in mM: NaCl 140; KCl 4; CaCl₂ 0.1; MgCl₂ 1; glucose 10; Hepes 10; pH 7.4). [Ca²⁺] was gradually adjusted to 1 mM. Isolated cardiomyocytes were stored at room temperature (20°C). All chemicals and reagents were purchased from Sigma-Aldrich (St Louis, USA).

Measurement of [Ca ²⁺] _i.

Changes in the cytosolic [Ca²⁺] ([Ca²⁺]_i) were measured with laser scanning confocal microscopy (Radiance 2000 MP, Bio-Rad, UK) equipped with a ×40 oil-immersion objective lens (N.A.=1.3) as described previously [38, 39]. To record [Ca²⁺]_i, we used the high-affinity Ca²⁺ indicator Fluo-4 (Molecular Probes/Invitrogen, Carlsbad, CA, USA). To load the cytosol with Fluo-4, ventricular myocytes were incubated at room temperature with 10 μM Fluo-4 AM for 15 min in Tyrode solution (in mM: NaCl 140; KCl 4; CaCl₂ 1; MgCl₂ 1; glucose 10; Hepes 10; pH 7.4), followed by a 20 min wash. Fluo-4 was excited with the 488 nm line of an argon laser and the emission signal collected at wavelengths above 515 nm. Changes in [Ca²⁺]_i were expressed as changes in F/F₀, where F₀ is the Fluo-4 signal at the resting condition.

Ca²⁺ spark measurements were conducted in permeabilized ventricular myocytes as described previously [38–40]. After the surface membrane permeabilization with saponin, Ca²⁺ sparks were studied in an internal solution composed of (in mM): K-aspartate 100; KCl 15; KH₂PO₄ 5; MgATP 5; EGTA 0.35; CaCl₂ 0.1; MgCl₂ 0.75; phosphocreatine 10; HEPES 10; Fluo-4 pentapotassium salt 0.04 mM; dextran (MW: 40,000) 8%, and pH 7.2 (KOH). Free [Ca²⁺] and [Mg²⁺] in this solution were 100 nM and 1 mM, respectively. Sparks were detected and analyzed using the SparkMaster algorithm [27].

Measurement of [Na ⁺] _i.

Changes in [Na⁺]_i were monitored using inverted fluorescence microscope (Nikon Ti2 Eclipse) equipped with air immersion 40x objective and Lumencore Spectra X excitation system. The fluorescent sodium indicator SBFI/AM (Molecular Probes/Invitrogen, Carlsbad, CA, USA) was used to trace the Na⁺-dependent fluorescent signal. To load the cytosol with SBFI, ventricular myocytes were incubated at room temperature with 20 μM SBFI/AM for 30 min in Tyrode solution, followed by a 20 min wash. The SBFI signal was detected using the stream acquisition tool in Metamorph software (Molecular Devices) with 100 ms excitation (excitation 325/25 nm, emission 530 nm). Emitted light was passed through a dichroic emission filter cube (DAPI/GFP/TagRFP Spectra X emission filter set) and detected with Photometrics Prime 95B 25 mm camera. Two-dimensional images were obtained at 15 s intervals. Changes of [Na⁺]_i are presented as background-subtracted normalized fluorescence (1−F/F₀). The experiments were conducted in Tyrode solution. All 2-D images and line scan measurements for [Ca²⁺]_SR were analyzed with ImageJ software (NIH, USA).

Preparation of purified NKA.

A purified sodium pump was prepared from the pig’s outer medulla as described previously [23]. The SDS-solubilized microsomal membranes were separated using differential ultracentrifugation and the final protein was aliquoted at a concentration of 1 mg/ml and stored at −80°C.

Measurements of NKA-specific ATPase activity.

NKA ATPase activity was determined by the Baginsky method as described previously [3]. The method detects the presence of inorganic phosphate in the sample by measuring absorbance at 710 nm. Specifically, 2 μl of purified NKA was mixed with 48 μl ATPase buffer (in mM: Na-ATP 3.3, NaCl 192, MgCl₂ 4, KCl 20; pH 7.2) and incubated for 6 minutes at room temperature. The inorganic phosphate was stained with 75 μl of staining solution (160 mM ascorbic acid, 3.7% (v/v) HCl, 3% (w/v) SDS, and 0.5% ammonium molybdate) and the reaction was stopped after 8 minutes by addition of 125 μl of stopping solution (0.9% (w/v) bismuth citrate, 0.9% (w/v) sodium citrate and 3.7% HCl). Immediately after stopping the reaction, the plate was read using the microplate reader Clario Star (BMG Labtech, Germany). The calibration line was determined using the set of KH₂PO₄ solutions with concentrations ranging between 0–37.5 nM. To estimate the NKA-specific ATPase activity, the signal in the presence of the NKA inhibitor ouabain (50 μM) was subtracted from the total ATPase activity.

Statistics.

Data are presented as mean ± standard error of the mean (SEM) of n measurements. Statistical comparisons between groups were performed using Student’s t test for unpaired data sets. Differences were considered statistically significant at P < 0.05. Statistical analysis and graphical representation of averaged data was carried out using the OriginPro7.5 software (OriginLab, USA).

RESULTS

Effect of protein carbonylation on Ca²⁺ regulation in mouse ventricular myocytes

To determine how protein carbonylation affects cardiac Ca²⁺ signaling, we studied the effect of the reactive α-dicarbonyl methylglyoxal (MGO) on action potential (AP)- and caffeine- induced Ca²⁺ transients in mouse ventricular myocytes (Fig. 1A). Myocytes were electrically stimulated at 0.5 Hz to evoke AP-induced Ca²⁺ transients. The amplitude of Ca²⁺ transient during caffeine (10 mM) application was analyzed to estimate changes in SR Ca²⁺ load. MGO (200 μM) increased AP-induced Ca²⁺ transients to 145 ± 11% (n = 12 cells; Fig. 1B) and SR Ca²⁺ load to 116 ± 6% (n = 12 cells; Fig. 1C) compared to 100% in control. As a result, MGO enhanced the fractional release (Ca²⁺ transient/Ca²⁺ load) by 26%. The maximal effect of MGO on AP-induced Ca²⁺ transient developed within 3–5 min.

Figure 1. Effect of MGO on Ca²⁺ regulation in intact mouse ventricular myocytes.

A, AP- and caffeine- induced Ca²⁺ transients in control conditions and during MGO (200 mM) application. The amplitude of Ca²⁺ transient during caffeine (10 mM) application was analyzed to assess changes in SR Ca²⁺ load. Duration of caffeine (Caff) applications is marked by white boxes. Effect of MGO on AP-induced Ca²⁺ transient amplitude (B), SR Ca²⁺ load (C), LTCC-mediated Ca²⁺ transient amplitude (D) and NCX-mediated [Ca²⁺]_i decay rate (E). The data were collected from 12 myocytes isolated from 4 animals. *P<0.05 vs Control.

The first AP-induced Ca²⁺ transient after complete depletion of [Ca²⁺]_SR with caffeine is mainly mediated by Ca²⁺ influx via L-type Ca²⁺ channels (LTCC) [10]. Analysis of the LTCC-mediated Ca²⁺ transient amplitude suggested that MGO does not significantly affect LTCC activity (Fig. 1D). MGO drastically slowed down the rate of [Ca²⁺]_i decay during the caffeine application (Fig. 1E), but not during AP. As the Ca²⁺ decay during the caffeine application is predominantly mediated by NCX-mediated Ca²⁺ extrusion [4], these results suggest that carbonylation reduces cytosolic Ca²⁺ extrusion by NCX. Consequently, more cytosolic Ca²⁺ could be pumped by SERCA2a into the SR, leading to an increase in SR Ca²⁺ load.

Next, we studied whether an increase in SR Ca²⁺ load caused by MGO increases the propensity of spontaneous SR Ca²⁺ release events. After 10 min electrical pacing (0.5 Hz) to stabilize SR Ca²⁺ load, Ca²⁺ waves were recorded during a following period of rest (Fig. 2A). MGO (200 μM) increased Ca²⁺ wave frequency from 2.7 ± 1.1 to 11.7 ± 3.7 waves/min (n = 8 cells) or more than 4 times (Fig. 2B). The Ca²⁺/Na⁺ free Tyrode solution abolished the spontaneous Ca²⁺ release events in the presence of MGO (Fig. 2B). These results suggest that the reduced Ca²⁺ extrusion by NCX and SR Ca²⁺ overload increase the propensity of spontaneous SR Ca²⁺ release events during protein carbonylation.

Figure 2. Effect of MGO on spontaneous Ca²⁺ waves during rest in intact ventricular myocytes.

A, AP-induced Ca²⁺ transients (left panel) and spontaneous Ca²⁺ waves during rest (right panel) in control conditions, during MGO (200 mM) application and during MGO application in the presence of the 0 Ca²⁺/0 Na⁺ Tyrode solution. B, effect of MGO on Ca²⁺ wave frequency during rest. The data were collected from 8 myocytes isolated from 3 animals. *P<0.05 vs Control. N/D – not detected.

To determine whether MGO directly affects the RyR-mediated Ca²⁺ leak, we studied the effect of MGO on Ca²⁺ sparks in permeabilized ventricular myocytes (Fig. 3A). Studying permeabilized cells enabled exclusion of sarcolemmal Ca²⁺ transporters from SR Ca²⁺ regulation. In the control conditions, Ca²⁺ sparks were detected at a stable frequency of 16.5 ± 0.4 sparks s⁻¹ (100 μm)⁻¹. MGO (200 μM) decreased Ca²⁺ spark frequency to 14.4 ± 0.6 sparks⋅s⁻¹ ⋅ (100 μm)⁻¹ or by 13% (n = 17 cells; Fig. 3B). The inhibitory effect on Ca²⁺ spark frequency developed within 3 min and remained stable during MGO application (15 min). The inhibitory effect of MGO on Ca²⁺ spark frequency was associated with a decrease of SR Ca²⁺ load by 14 ± 3% (n = 10 cells).

Figure 3. Effect of MGO on spontaneous Ca²⁺ sparks in permeabilized ventricular myocytes.

A, confocal line scan images of [Ca²⁺]_i with corresponding F/F₀ profiles of Ca²⁺ sparks in control conditions and during MGO (200 mM) application. F/F₀ profiles were obtained by averaging the fluorescence over the 1 mm wide regions indicated by the black arrows. B, Effect of MGO on Ca²⁺ spark frequency. The data were collected from 17 myocytes isolated from 4 animals. *P<0.05 vs Control.

Effect of protein carbonylation on Ca²⁺ regulation during adrenergic receptor activation

b-adrenergic receptor (b-AR) activation with isoproterenol (ISO; 0.1 μM) increased the AP-induced Ca²⁺ transient amplitude to 230 ± 25% (n = 11 cells) and SR Ca²⁺ load to 153 ± 9% compared to 100% in control (n = 12 cells; Figs. 4A, C and D). This effect is mainly mediated by an increase in SR Ca²⁺ uptake and sarcolemmal Ca²⁺ influx due to phospholamban and LTCC phosphorylation by protein kinase A (PKA) [6, 15]. We studied whether protein carbonylation affects the β-AR stimulation effects on Ca²⁺ signaling. These experiments revealed that in the presence of MGO (200 μM), ISO increased the AP-induced Ca²⁺ transient amplitude to 268 ± 27% (n = 11 cells; Fig. 4B) and SR Ca²⁺ load to 130 ± 7% (n = 11 cells; Fig. 4C) compared to 100% in control. The results between control and MGO groups were not statistically different, suggesting that the effect of adrenergic stimulation on Ca²⁺ regulation remained largely preserved during protein carbonylation. In ventricular myocytes, SR Ca²⁺ overload induced by adrenergic stress can lead to spontaneous SR Ca²⁺ release in the form of diastolic Ca²⁺ waves [7]. We found that in control conditions, 3 myocytes out of 11 studied or 27% exhibited diastolic Ca²⁺ waves during b-AR activation with ISO. Treatment of cells with MGO increased the fraction of myocytes with Ca²⁺ waves during adrenergic stimulation to 71% or more than 3 times (Fig. 4B and E).

Figure 4. Effect of MGO on Ca²⁺ signaling during adrenergic receptor activation.

A, F/F₀ profiles of changes in [Ca²⁺]_i during AP- and caffeine- induced Ca²⁺ transients in control conditions, during adrenergic receptor activation with ISO (0.1 mM). B, F/F₀ profiles of changes in AP- and caffeine- induced Ca²⁺ transients in the presence of MGO (200 mM) and during the consecutive application of ISO (0.1 mM). Ca²⁺ waves are marked with asterisk. Effect of ISO in control conditions and the presence of MGO on AP-induced Ca²⁺ transient amplitude (C), SR Ca²⁺ load (D) and percentage of myocytes with spontaneous Ca²⁺ waves (E). A total of 11 myocytes isolated from 3 animals were studied in these experiments. *P<0.05 vs ISO alone.

Effect of protein carbonylation on [Na⁺]_i and NKA-dependent ATPase activity

To assess whether the decreased NCX-mediated Ca²⁺ extrusion induced by MGO was a result of elevated intracellular [Na⁺] (Na⁺]_i), we measured changes in [Na⁺]_i in intact ventricular myocytes. We found that MGO (200 μM) increased [Na⁺]_i by 57% from the maximal effect produced by the selective Na⁺-K⁺ ATPase (NKA) inhibitor ouabain (20 μM). In contrast to ouabain which caused the maximal [Na⁺]_i elevation within 1 min of its application, the effect of MGO on [Na⁺]_i required more than 3 min to fully develop (Fig. 5A). These results suggest that protein carbonylation reduces the NCX-mediated Ca²⁺ extrusion by increasing cytosolic [Na⁺].

Figure 5. Effect of MGO and ouabain on intracellular [Na⁺] homeostasis and NKA-dependent ATPase activity.

A, Effect of MGO (200 mM; black circles) and ouabain (20 mM; empty circles) on [Na⁺]_i in intact ventricular myocytes. Effect of MGO (gray triangles) and ouabain (black triangles) on [Na⁺]_i in the presence of TTX (10 mM). The black arrow marks the application either MGO or ouabain. [Na⁺]_i was measured with SBFI. 15 myocytes were studied to measure the effect of MGO, and 10 myocytes were studied to measure the effect of ouabain. Myocytes were isolated from 3 animals. B, Effect of MGO on NKA-dependent ATPase activity. The pump activity was measured as an accumulation of inorganic phosphate sensitive to the NKA inhibitor ouabain.

Since NKA is the main mechanism that controls cytosolic [Na⁺], we tested whether carbonylation affects the NKA-dependent ATPase activity. NKA samples were preincubated with different MGO concentrations for 60 min at room temperature and the ouabain-specific ATPase activity was measured as accumulation of inorganic phosphate [3]. These experiments revealed that MGO does not affect the NKA-dependent ATPase activity (Fig. 5B). It has been shown that an accumulation of [Na⁺]_i in ventricular myocytes from failing heart was sensitive to the Na⁺ channel inhibitor TTX [13]. We found that TTX (10 μM) significantly prevented [Na⁺]_i accumulation induced by MGO. In the presence of TTX, MGO (200 μM) increased [Na⁺]_i only by 11% from the maximal effect produced by ouabain (20 μM). At the same time, TTX did not prevent an increase in [Na⁺]_i during the ouabain application (Fig. 5A).

Effect of elevated [Na⁺]_i on Ca²⁺ regulation

If the elevated [Na⁺]_i is a main factor that causes SR Ca²⁺ overload during protein carbonylation, then an increase of [Na⁺]_i alone should produce a similar effect on Ca²⁺ regulation as MGO. To selectively increase [Na⁺]_i, NKA pump was inhibited with ouabain. We found that ouabain (20 μM) increased an amplitude of AP-induced Ca²⁺ transients to 187 ± 10%, increased SR Ca²⁺ load to 147 ± 7% and slowed down NCX-mediated Ca²⁺ extrusion by 272 ± 27% (n = 12 cells; Fig. 6A). In the presence of ouabain, ISO (0.1 μM) increased the AP-induced Ca²⁺ transient amplitude to 223 ± 6% and increased SR Ca²⁺ load to 185 ± 8% (n = 10; Fig. 6B). Similar to the MGO effect, ouabain increased propensity of spontaneous Ca²⁺ waves in cells treated with ISO from 17 to 65% or more than 3 times.

Figure 6. Effect of ouabain on Ca²⁺ signaling in control conditions and during adrenergic receptor activation.

A, F/F₀ profiles of changes in [Ca²⁺]_i during AP- and caffeine- induced Ca²⁺ transients in control conditions and during NKA inhibition with ouabain (20 mM). B, F/F₀ profiles of changes in AP- and caffeine- induced Ca²⁺ transients in control conditions, the presence of ouabain (20 mM) and during the consecutive application of ISO (0.1 mM). Ca²⁺ waves are marked with asterisk. A total of 10 myocytes isolated from 2 animals were studied in these experiments.

DISCUSSION

Cardiomyopathy is one of the leading causes of morbidity and mortality in type 1 and type 2 diabetic patients. Among several factors, increased production of reactive carbonyl species is viewed as a critical step involved in the development of diabetic cardiomyopathy [33]. It has been shown that myocardium from patients with diabetic heart failure (HF) exhibits the increased level of carbonylation of the sarcomeric proteins as compared with myocardium from non-diabetic HF patients [26]. Several studies have shown that in hearts from the type 1 diabetic animal model, RyR2 and SERCA2a are characterized by highly glycated MGO-adducts [29, 30]. These modifications have been proposed altering the function of these Ca²⁺ handling proteins and, therefore, intracellular Ca²⁺ dynamics. It has been shown that in myocytes from type 1 diabetic hearts RyR2-mediated Ca²⁺ sparks were significantly increased and AP-evoked Ca²⁺ transients were largely dyssynchronous [31]. Moreover, SERCA2a-mediated Ca²⁺ uptake decreased in type 1 diabetic hearts and this effect was associated with reduced the affnity of SERCA2a for Ca²⁺ [24]. Moreover, it has been shown that cardiomyocytes contraction and Ca²⁺ signaling were depressed in the type 2 diabetic rat model together with reduced RyR2 expression level [16]. The type 2 diabetics hearts were also characterized by reduced SERCA2a function [34]. While carbonylation of RyR2 and SERCA2a might contribute to Ca²⁺ mishandling in the diabetic heart, the effect of reactive carbonyl species on intracellular Ca²⁺ regulation is still not fully understood. Moreover, an alteration of Ca²⁺ homeostasis in the diabetic heart is likely to be a result of complex changes of many intracellular processes, including cell energetics, gene regulation, protein expression and protein post-translational modifications (e.g., oxidation, phosphorylation, and carboxylation). To define specific molecular mechanisms by which carboxylation affects cardiac Ca²⁺ regulation, we characterized the effect of the reactive carbonyl byproduct MGO on Ca²⁺ dynamics in mouse ventricular myocytes.

Using confocal microscopy and in-cell Ca²⁺ imaging, we found that MGO increases the amplitude of Ca²⁺ transients triggered by the action potential. The effect was accompanied by an increase in SR Ca²⁺ load (Fig. 1). MGO also enhanced the fractional SR Ca²⁺ release and increased propensity of spontaneous Ca²⁺ waves during rest (Fig. 2). These results are in good agreement with the previously published work [30] that suggests that RyR2 carboxylation increases RyR2 activity. In permeabilized myocytes, however, MGO reduced spontaneous Ca²⁺ spark frequency, and this effect was associated with a decrease in SR Ca²⁺ load (Fig. 3). Since Ca²⁺ spark frequency highly depends on SR Ca²⁺ load [36], the observed decrease spark activity can be explained by activation of RyR2-mediated Ca²⁺ leak by MGO with following a depletion of SR Ca²⁺ load. However, this mechanism alone cannot explain the observed differences in the MGO effect on SR Ca²⁺ load. While MGO increased SR Ca load in intact myocytes, it decreased SR Ca load in permeabilized cells. Such discrepancy suggests that other cellular mechanisms of the MGO action have been disrupted after the sarcolemma permeabilization. Analysis of Ca²⁺ signaling in intact myocytes revealed that MGO significantly slows down the decay of Ca²⁺ transient during caffeine application (Fig. 1). As this decay is predominantly mediated by NCX-mediated Ca²⁺ extrusion [4], these findings suggest that protein carbonylation reduces cytosolic Ca²⁺ removal by NCX. Subsequently, more cytosolic Ca²⁺ could be pumped into the SR by SERCA2a, leading to an increase in SR Ca²⁺ load. We hypothesized that MGO might reduce the NCX-mediated Ca²⁺ extrusion via a Na⁺-dependent mechanism, similar to the positive inotropic effect produced by glycosides. Similar to the MGO effect on Ca²⁺ regulation, we observed that an increase of [Na⁺]_i after treating cells with a the selective NKA inhibitor ouabain slows down the NCX-mediated Ca²⁺ extrusion, increases the AP-triggered Ca²⁺ transients, and increases SR Ca²⁺ load (Fig. 6). Direct measurements of [Na⁺]_i revealed that MGO significantly increased [Na⁺]_i (Fig. 5). Since NKA is the main mechanism that controls cytosolic [Na⁺], we tested whether protein carbonylation directly inhibits NKA. However, analysis of the ouabain-specific ATPase activity revealed that MGO does not affect NKA function (Fig. 5). It has been shown that an accumulation of [Na⁺]_i in ventricular myocytes from failing heart was sensitive to the Na⁺ channel inhibitor TTX [13]. Here, we tested whether the MGO effect on [Na⁺]_i was also sensitive to TTX. Measurements of [Na⁺]_i revealed that TTX significantly prevented the [Na⁺]_i increase induced by MGO (Fig. 5). Thus, this study, for the first time, uncovered a new mechanism by which reactive carbonyl species can affect [Ca²⁺]_i in ventricular myocytes. This mechanism involves [Na⁺]_i accumulation due to activation of Na⁺ influx via a TTX-sensitive pathway.

Intuitively, protein carbonylation could be viewed as beneficial post-translational modification for Ca²⁺ regulation and, therefore, for heart function. However, as the inward Na⁺ gradient plays a critical role in many cellular processes, persistent cytosolic Na⁺ overload would have a negative impact on cardiac function. First, [Na⁺]_i accumulation would reduce the efficacy not only Ca²⁺ transport by NCX, but regulation of intracellular pH by the Na⁺-H⁺ exchanger and the Na⁺-driven HCO₃⁻ transport. It would also reduce the transport of glucose, pyruvate, and several amino acids, therefore, unbalancing cellular metabolism. Moreover, protein carboxylation with MGO significantly increases the propensity of spontaneous Ca²⁺ waves during adrenergic receptor activation (Fig. 4). It appears that SR Ca²⁺ overload due to the reduced NCX function and the increased RyR2 activity create an effective substrate for Ca²⁺⁻ dependent cardiac arrythmias during adrenergic stress. Additionally, these spontaneous Ca²⁺ waves would reduce the efficacy of myocardial relaxation, increasing ventricular stiffness during diastole. In fact, it has been reported that diabetic cardiomyopathies are commonly associated with ventricular arrhythmias as well as diastolic dysfunction [17, 19]. Therefore, a therapeutic approach aimed at limiting protein carbonylation could be beneficial for protecting cardiac function in diabetic patients.

Funding Statement

This work was supported by the National Institutes of Health Grants R01HL151990 (to A.V.Z.) and R01 HL092321 (to S.L.R.).

Availability of data and materials

The data supporting this article and other findings are available within the manuscript, figures, and from the corresponding authors upon request.