Cardiovascular disease is the leading cause of death in the diabetic population. However, molecular mechanisms underlying diabetic cardiomyopathy remain unclear. We analyzed Ca2+-induced Ca2+ release and excitation-contraction coupling in db/db obese type 2 diabetic mice and their control littermates. Echocardiography showed a systolic dysfunction in db/db mice. Two-photon microscopy identified intracellular calcium concentration ([Ca2+]i) transient decrease in cardiomyocytes within the whole heart, which was also found in isolated myocytes by confocal microscopy. Global [Ca2+]i transients are constituted of individual Ca2+ sparks. Ca2+ sparks in db/db cardiomyocytes were less frequent than in +/+ myocytes, partly because of a depression in sarcoplasmic reticulum Ca2+ load but also because of a reduced expression of ryanodine receptor Ca2+ channels (RyRs), revealed by [3H]ryanodine binding assay. Ca2+ efflux through Na+/Ca2+ exchanger was increased in db/db myocytes. Calcium current, ICa, triggers sarcoplasmic reticulum Ca2+ release and is also involved in sarcoplasmic reticulum Ca2+ refilling. Macroscopic ICa was reduced in db/db cells, but single Ca2+ channel activity was similar, suggesting that diabetic myocytes express fewer functional Ca2+ channels, which was confirmed by Western blots. These results demonstrate that db/db mice show depressed cardiac function, at least in part, because of a general reduction in the membrane permeability to Ca2+. As less Ca2+ enters the cell through ICa, less Ca2+ is released through RyRs.

Diabetes prevalence is rising in Western countries and affects ∼143 million patients worldwide, thus approximately five times surpassing estimations made a decade ago (1). About 90% of diabetic patients exhibit non–insulin-dependent (or insulin-resistant) type 2 diabetes, and at least 80% of type 2 diabetic patients in Western countries are obese. These patients usually exhibit severe glucose intolerance compared with lean type 2 diabetic patients (2). Cardiovascular disease is the leading cause of death in the diabetic population (35). Although cardiovascular problems could be the result of associated comorbidities, a defect in cardiac performance has been detected in diabetic patients in the absence of vascular injury, supporting the existence of a specific diabetic cardiomyopathy. Reductions in diastolic compliance, contractility, and rate of myocardial relaxation characterize diabetic cardiomyopathy (68). However, cellular and molecular bases of cardiac dysfunction are largely unknown in type 2 diabetes.

Because Ca2+ release and uptake are at the root of contraction and relaxation, alteration of Ca2+ handling has been suspected to drive the progression of functional abnormalities in diseased hearts. However, data regarding Ca2+ handling in type 2 diabetes are limited and ambiguous. For example, a rat model of insulin-resistant obese type 2 diabetes (cp/cp) shows enhanced Ca2+ uptake into sarcoplasmic reticulum, but sarcoplasmic reticulum Ca2+ ATPase (SERCA) levels and sarcoplasmic reticulum Ca2+ load are unaltered (9). By contrast, other well-known models of type 2 diabetes such as Otsuka Long Evans Tokushima Fatty rats (10) and db/db mice (11) present defects in both function and expression of SERCA. Streptozotocin injection in neonatal rats induces adult non–insulin-dependent diabetes, and the activities of plasmalemmal Ca2+ ATPase (12,13) and Na+/Ca2+ exchanger (NCX) (13) are both decreased in this model. Thus, a distinct Ca2+ transporter malfunction has been identified in different diabetic models, but an integrated mechanism of altered Ca2+ homeostasis is lacking.

We characterized Ca2+-induced Ca2+ release and excitation-contraction coupling in hearts of db/db mice, an obesity-linked mouse model of type 2 diabetes. Like humans with diabetic cardiomyopathy, db/db mice display increased left ventricular end-diastolic pressure, decreased left ventricular systolic pressure at high preload, and lowered cardiac output (14). We found impaired intracellular calcium concentration ([Ca2+]i) cycling that may explain these functional changes: a reduction in l-type calcium current ICa underlies a reduction in [Ca2+]i transient and contraction, both of which are exacerbated by a decrease in ryanodine receptor Ca2+ channel (RyR) density and sarcoplasmic reticulum Ca2+ load.

M-mode echocardiography.

Fifteen-week-old male and female C57BL/KsJ-db (db/db) mice and their control littermates (+/+) were used. All experiments were carried out according to the ethical principles laid down by the French Ministry of Agriculture and European Union Council Directives for the care of laboratory animals. Transthoracic echocardiography was performed in anesthetized animals (35 mg/kg sodium pentobarbital) using an ultrasound machine (Esaote “Mylab”, Florence, Italy) and a 10-MHz phase-array transducer. Parameters were determined using the leading-edge method of the American Society of Echocardiography. End-diastole was determined at the maximal left ventricular diastolic dimension, and end-systole was taken at the peak of posterior wall motion. Systolic function was calculated as fractional shortening: % fractional shortening = (left ventricular end-diastolic dimension − left ventricular end-systolic dimension)/left ventricular end-diastolic dimension × 100. Heart rate was determined from measurements of cardiac cycle length on spectral Doppler tracings of the aortic outflow. All data are the average of at least three separate scans, each scan representing the average of three consecutive beats.

Two-photon microscopy in whole hearts.

At the time of the experiments, mice were weighed and anesthetized with sodium pentobarbital (100 mg/kg i.p.). The heart was quickly removed and weighed. The aorta was cannulated, and the heart was perfused with a HEPES-based Tyrode solution and used for two-photon microscopy. First, the heart was perfused ∼15 min with a solution similar to that used for cell isolation (15) but supplemented with 1.8 mmol/l CaCl2 and 50 mmol/l butanedione monoxime (16) to prevent contraction. Then, the same solution containing the fluorescence Ca2+ dye Rhod-2 AM (10 μmol/l; Tefflabs) was perfused for 20 min. After this loading time, the same solution free of Rhod-2 and of butanedione monoxime and supplemented with 50 μmol/l cytochalasin D to block contractions was perfused throughout the experiment. After 20–30 min, the heart was placed on the stage of an upright microscope (Axioskop 2FSmot Meta LSM510, adapted for two-photon microscopy, ceramic objective IR40×, numerical aperture 0.8; Zeiss) under continuous perfusion. The hearts were electrically stimulated through a bipolar platinum electrode at 2 Hz. Ca2+ images were obtained with illumination supplied by a mode-locked Ti:Saphire laser (Mira 900; Coherent) tuned to a center wavelength of 820 nm. Emission was collected at >560 nm wavelength. [Ca2+]i transients were recorded in the line-scan mode at 3 ms/line. To visualize cell membranes, the heart was loaded with the voltage-sensitive dye di-4 ANEPPS in a single bolus (16), and three-dimensional images were recorded with the two-photon microscope.

Isolated cells.

Ventricular myocytes were isolated from +/+ and db/db mice as previously described (15). Whole-cell patch clamp (Axopatch 200A; Axon Instruments) was used to monitor l-type Ca2+ currents (ICa). The patch pipette contained 50 μmol/l Fluo-3 pentapotassium salt to simultaneously record the associated [Ca2+]i transient. Voltage protocol and solutions were as previously described (17). [Ca2+]i transients were recorded with the confocal microscope (MetaZeiss LSM510, objective water immersion ×60, numerical aperture 1.2) in line-scan mode (1.5 ms/line). The 488-nm line of the argon laser was used to illuminate, and emission was collected at >505 nm. The line of scan was selected parallel to the longitudinal cell axis to measure cell shortening.

Sarcoplasmic reticulum Ca2+ load was estimated by rapid caffeine application to isolated cells. Myocytes were previously depolarized to 0 mV (150-ms steps) from a holding potential of −80 mV for 1 min, then 10 mmol/l caffeine was added, and the associated fluorescence image was recorded by confocal microscopy. In some cells, the associated NCX current was recorded by patch clamp and integrated to estimate the amount of Ca2+ released by the sarcoplasmic reticulum (18). Ca2+ sparks were recorded in saponin-permeabilized myocytes as previously described (19) using an internal solution containing 50 nmol/l free Ca2+. Image analyses were done using homemade routines in IDL (interactive data language) (RSI) (20). All experiments performed on isolated hearts or myocytes were carried out at room temperature (21–23°C).

Single-channel recording.

Cardiomyocytes were placed in a depolarizing solution containing 120 mmol/l potassium glutamate, 25 mmol/l KCl, 2 mmol/l MgCl2, 10 mmol/l HEPES, 2 mmol/l EGTA, 1 mmol/l CaCl2, 1 mmol/l Na2ATP, and 10 mmol/l glucose (pH 7.4 with KOH). Borosilicate pipettes (5–12 MΩ) were coated with Sylgard and filled with a solution containing 70 mmol/l BaCl2, 90 mmol/l sucrose, and 10 mmol/l HEPES (pH 7.4 with TEA-OH). Ba2+ currents carried by single l-type Ca2+ channels were recorded using the cell-attached configuration of the patch-clamp technique as previously described (21), applying voltage steps to +20 mV from a holding potential of −100 mV, at least 180 sweeps per experiment. Data were sampled at 10 KHz and filtered at 2 KHz using an Axopatch 200A amplifier and analyzed with pClamp6 software (Axon Instruments). In case of multichannel patches, all data were corrected by the number of channels as described previously (22).

[3H]ryanodine binding assay.

Hearts were homogenized in a buffer containing 30 mmol/l KH2PO4, 20 mmol/l NaF, 300 mmol/l sucrose, 0.012 mmol/l leupeptin, 0.1 mmol/l phenylmethylsulfonyl fluoride, and 0.01 mmol/l benzamidine (pH 7 with NaOH). Insoluble material was removed by centrifugation at 1,000 rpm for 1 min (4°C). Protein quantification was done with the Bradford assay using BioRad reagents. The [3H]ryanodine binding assay was performed as described previously (23) using 20 μg heart protein homogenate/assay tube in a buffer containing 1 mol/l NaCl, 50 μmol/l CaCl2, and 20 mmol/l Na-PIPES, pH 7.2. Bmax, the maximal number of receptor sites, and Kd, the affinity constant of the receptor-ligand interaction, were obtained by fitting data points with the formula B = Bmax × [3H]ryanodine/(Kd + [3H]ryanodine), where B is the specific binding. Nonspecific binding was determined in the presence of 20 μmol/l ryanodine.

Western blot.

Ventricular protein homogenates were made as described previously (21). Ventricles were rapidly frozen in liquid nitrogen and then placed in Tris solution (50 mmol/l, pH 7.4) containing 1 mmol/l phenylmethylsulfonyl fluoride and chopped using an Ultra-Turrax disperser. Suspension was centrifuged for 15 min (3,500 × g). Supernatant was then centrifuged for 60 min (55,000 × g). The pellet was dissolved in 50 mmol/l Tris solution (pH 6.8) containing 2% (wt/vol) SDS, 20% glycerol, and 5 mmol/l dithiothreitol. After shaking for 30 min, probes were frozen at −80°C until subjecting to SDS-PAGE on 8% running gel and 5% stacking gel. For protein analysis, equal amounts of 15 μg per probe were used. SDS-PAGE was blotted to polyvinylidene fluoride membranes (Bio-Rad) by using a semidry Western-blot system. Cav1.2 (α1C) proteins were detected by a specific antibody (C-1603; dilution, 1:600; Sigma-Aldrich) using alkaline-phosphatase reaction. Protein bands were analyzed densitometrically.

Data analysis.

Male and female animals were used in an attempt to identify sex differences in the diabetic cardiomyopathy (4). However, because no difference was found, both male and female data were pooled together. Data are presented as means ± SE, and t test was used to compare data obtained in +/+ and db/db mice. P < 0.05 was considered statistically significant.

db/db mice hearts are functionally defective but not hypertrophied.

Echocardiography confirmed functional impairment of db/db hearts (Fig. 1A). Decreased shortening fraction in db/db mice indicated contractile dysfunction (Fig. 1B), whereas left ventricular wall thickness was similar in +/+ and db/db mice (Table 1). Heart weight was similar for both groups (Fig. 1C). To assess hypertrophy of cardiomyocytes, we measured cell dimensions in hearts loaded with the voltage-sensitive dye di-4 ANEPPS, which labels the external cell membrane, and three-dimensional images were recorded. Deconvolution was applied with Huygens (Bitplane) to correct for thickness distortion (24). A deconvoluted two-photon image of cell membranes is shown in Fig. 1D. In agreement with unchanged cell capacitance (Table 1), calculated cell volume was not altered (Fig. 1E).

Cellular [Ca2+]i transients recorded in whole hearts.

To determine whether altered cellular Ca2+ signaling underlies this cardiac dysfunction, we recorded [Ca2+]i transients in whole hearts. Freshly explanted hearts were loaded with Rhod-2 (Fig. 1F) and electrically evoked [Ca2+]i transients in individual myocytes were visualized with two-photon microscopy in the line-scan mode. Figure 1G shows representative images of a +/+ cell (top), of a db/db cell (middle), and the corresponding calculated [Ca2+]i transients (bottom). The [Ca2+]i transients recorded in db/db cells were smaller in amplitude and decayed more slowly. Averaged data (Fig. 1H) indicate a significant reduction of [Ca2+]i transient amplitude in db/db myocytes compared with +/+ myocytes. The duration of [Ca2+]i transient at half-maximum amplitude, D50, was significantly longer (P < 0.01) in db/db myocytes (176.9 ± 5.4 ms) than in +/+ cells (104.1 ± 2.4 ms). Thus, db/db hearts have depressed contractile properties and decreased [Ca2+]i transients but with no apparent remodeling of the heart. To characterize impairment of [Ca2+]i transients in more detail, we analyzed the function and expression of l-type Ca2+ channels and RyRs and measured sarcoplasmic reticulum Ca2+ load, thus examining critical steps in cardiac Ca2+ cycling more comprehensively.

l-type Ca2+ channels and excitation-contraction coupling.

Ca2+ release from the sarcoplasmic reticulum is graded by ICa, the “trigger” current carried by l-type Ca2+ channels. We thus recorded whole-cell ICa in single myocytes. ICa was significantly decreased in db/db myocytes compared with controls (+/+) (Fig. 2A and D). To determine the ICa-sarcoplasmic reticulum Ca2+ release coupling efficiency, we simultaneously recorded ICa and its triggered [Ca2+]i transient by confocal microscopy. Consistent with reduced [Ca2+]i transients observed in single cells within the whole hearts, [Ca2+]i transients were decreased and slowed in isolated db/db cells as well. Figure 2C is a plot of peak [Ca2+]i transient (F/F0, with F, the fluorescence signal, and F0, the diastolic fluorescence) as a function of voltage step in db/db and +/+ cells. The decay time of the [Ca2+]i transient, calculated by fitting the decay phase of the fluorescence signal to a mono-exponential function, was prolonged in db/db myocytes (decay time constant at 0 mV, 160.1 ± 8.1 ms in 38 +/+ cells vs. 192.7 ± 12.4 ms in 16 db/db myocytes; P < 0.05). The reduction of sarcoplasmic reticulum Ca2+ release in db/db myocytes obviously paralleled the decrease of whole-cell ICa. Accordingly, cell contraction, which steeply follows the magnitude of the [Ca2+]i transient (25), was significantly reduced in db/db cells at most potentials (Fig. 2B). Cell shortening, as shown in Fig. 2A (lower traces), was also slowed in diabetic myocytes. Maximum derivative over time (dL/dt) at 0 mV (in % cell length/ms) was 0.33 ± 0.03 in +/+ cells vs. 0.21 ± 0.03 in db/db cells (P < 0.05). To estimate excitation-contraction coupling gain, we normalized Ca2+ release (amplitude of [Ca2+]i transient) by total Ca2+ entry (integral of ICa) and plotted it as a function of voltage (Fig. 2E). db/db and +/+ myocytes showed similar excitation-contraction coupling gain, indicating that the efficacy of ICa to trigger Ca2+ release remains unchanged in db/db myocytes.

The decrease of macroscopic ICa could be due to altered voltage dependence, impaired single-channel gating, or decreased expression of l-type Ca2+ channels. We first analyzed steady-state activation and inactivation of ICa. Figure 3A shows that steady-state inactivation curves in +/+ (solid line and open symbols) and db/db (dotted line and filled symbols) were not different regarding half-inactivation potential (V1/2: −33.1 ± 1.5 mV for +/+, n = 17 vs. −31.6 ± 1.1 mV for db/db, n = 11; P > 0.05) or slope factor (5.53 ± 0.31 for +/+, n = 17 vs. 5.44 ± 0.54 for db/db, n = 11; P > 0.05). Steady-state activation was slightly shifted toward more depolarized potentials in db/db myocytes (dotted line) compared with +/+ myocytes (solid line) (−17.0 ± 0.7 mV in +/+ cells, n = 43 vs. −13.8 ± 0.8 mV, n = 20; P < 0.01) without alteration of the slope factor (4.95 ± 0.29 in +/+ cells vs. 4.81 ± 1.19 in db/db).

Reduction of ICa may reflect altered single-channel properties and/or decreased number of functional channels in the membrane. Recording of single l-type Ca2+ channels (Fig. 3B) revealed no difference regarding fraction of active sweeps (46 ± 8 vs. 44 ± 3% in six +/+ vs. five db/db patches, respectively) and mean open time (0.61 ± 0.05 vs. 0.58 ± 0.02 ms in +/+ vs. db/db, respectively), open probability in active sweeps (+/+, 14.3 ± 4.2%; db/db, 21.3 ± 4.5%), and peak ensemble average currents (Fig. 3B, bottom panels) (−77 ± 23 fA in +/+ vs. −82 ± 17fA in db/db) obtained with db/db (n = 5 patches) or +/+ (n = 6 patches) myocytes. Thus, the activity of single l-type Ca2+ channels was unchanged. These findings suggest that the 31.2% decrease in macroscopic ICa may reflect lower expression of channel proteins. To test this possibility, we carried out Western-blot analysis of Cav1.2, the pore-forming α1C subunit of l-type Ca2+ channels (Fig. 3C). In agreement with the decrease in whole-cell ICa, the expression of α1C protein was reduced by ∼38% in db/db ventricles (Fig. 3C).

RyR activity and abundance.

The opening of a RyR cluster produces local, rapid, and brief elevations in [Ca2+]i termed Ca2+ sparks (26). Ca2+ sparks occur spontaneously during diastole and are activated during systole by ICa. To avoid the compounding effect of ICa, we analyzed Ca2+ sparks in saponin-permeabilized myocytes. In the absence of external membrane control, the excitation-contraction coupling machinery is reduced to Ca2+ release by RyR and Ca2+ uptake by SERCA2a, thus allowing direct assessment of these parameters. Moreover, permeabilized cells allow easy clamping of internal [Ca2+], so that Ca2+ uptake and release in +/+ and db/db cells may be determined under identical conditions. Figure 4A shows confocal line-scan images in a +/+ (top) and a db/db (bottom) myocyte, which exhibited fewer Ca2+ sparks. The averaged Ca2+ spark frequency recorded in db/db myocytes was reduced in diabetic myocytes (Fig. 4B). An analysis of RyR expression by [3H]ryanodine binding to cardiac homogenates (Fig. 4C) revealed that Bmax was decreased in db/db hearts (Fig. 4D) (0.10 ± 0.01 pmol/mg protein) compared with +/+ hearts (0.16 ± 0.02 pmol/mg protein), whereas Kd, the RyR-[3H]ryanodine affinity constant, was similar (3.7 ± 1.1 and 4.4 ± 0.8 nmol/l in +/+ and db/db hearts, respectively). Thus, a reduction in RyR expression levels may contribute, at least in part, to the decrease in Ca2+ spark occurrence observed in permeabilized cells.

Sarcoplasmic reticulum Ca2+ load.

Sarcoplasmic reticulum Ca2+ load is a major determinant of Ca2+ release and, by extension, a critical factor in Ca2+ spark frequency and [Ca2+]i transient amplitude. It is plausible, therefore, that sarcoplasmic reticulum Ca2+ load contributes to the decrease in both amplitude of [Ca2+]i transient (Figs. 1H and 2C) and Ca2+ spark frequency (Fig. 4B) observed in db/db hearts, independently or in combination with the decreased expression of l-type Ca2+ channels and RyRs. We thus measured sarcoplasmic reticulum Ca2+ content after caffeine application in conditions used to record Ca2+ sparks and [Ca2+]i transients, i.e., in permeabilized and intact myocytes, respectively. Two examples of line-scan images recorded during caffeine application in permeabilized myocytes and the average data are displayed in Fig. 4E. Permeabilized db/db myocytes have lower sarcoplasmic reticulum Ca2+ content, suggesting a decreased activity of SERCA pump. We also estimated sarcoplasmic reticulum Ca2+ load by rapid caffeine application to nonpermeabilized myocytes. Figure 5A shows examples of line-scan images recorded in a +/+ (top) and a db/db (bottom) myocyte. The caffeine-evoked [Ca2+]i transient was smaller in db/db myocyte, as demonstrated by the averaged data in Fig. 5B. The caffeine-evoked [Ca2+]i transient activates the forward-mode of the NCX, eliciting an inward current as Ca2+ is extruded. Hence, the integral of the inward NCX current is a reliable index of the sarcoplasmic reticulum Ca2+ load (18). Figure 5C shows examples of NCX currents (left) and their running integrals (right) in a +/+ (top) and a db/db (bottom) cell. Figure 5D shows the averaged data of the integral of the caffeine-evoked NCX current and confirms the reduction of sarcoplasmic reticulum Ca2+ load in db/db myocytes. Thus, a reduction in sarcoplasmic reticulum Ca2+ load may contribute, in combination with decreased expression of RyRs and l-type Ca2+ channels, to the decrease in Ca2+ spark frequency and [Ca2+]i transient amplitude and may be a major determinant of the contractile dysfunction of diabetic hearts.

The diminution in sarcoplasmic reticulum Ca2+ load may reflect decreased Ca2+ entry through ICa and slowing in sarcoplasmic reticulum Ca2+ uptake but also increased efflux through the NCX. In many models of heart failure, the NCX current (INCX) is increased (27). We measured peak INCX (pA) activated by the increase in [Ca2+]i (F/F0) at a membrane potential of −80 mV and found that INCX is increased in db/db myocytes (in pA/[F/F0], 27.6 ± 2.6 in 20 +/+ cells vs. 47.9 ± 7.2 in 20 db/db cells, P < 0.05). In heart failure after myocardial infarction, we found that although the INCX was increased, there was a concomitant increase in cell volume that cancels out the relative increase of INCX, thus unfaltering the capacity of the cell to extrude Ca2+ (28). We therefore normalized INCX by [Ca2+]i and by cell volume. We found that INCX (pA) normalized by [Ca2+]i (F/F0) and by cell volume (pL) was 1.0 ± 0.1 in +/+ myocytes (n = 20) and 1.6 ± 0.2 in db/db (n = 20), P < 0.05. So db/db myocytes show an increased [Ca2+]i efflux for a given [Ca2+]i.

Cardiovascular alterations constitute the most frequent cause of death in diabetic patients (35). At present, an integrated scheme of excitation-contraction coupling alterations in type 2 diabetic hearts is incomplete. Our study of db/db mice, which experiences obesity-associated type 2 diabetes, shows that the contractile dysfunction linked with the cardiomyopathy can be explained, at least in part, by disturbed Ca2+ cycling, i.e., reduced Ca2+ influx via l-type Ca2+ channels, lowered Ca2+ release from sarcoplasmic reticulum, slowed Ca2+ reuptake, and increased Ca2+ efflux.

db/db mice exhibit cardiac failure that is evident as early as 12 weeks of age (29). Here, we used 15-week-old db/db mice and confirmed the depression of systolic function by M-mode echocardiography (Fig. 1A and B). Systolic dysfunction is accompanied by a dramatic decrease of cellular [Ca2+]i transients recorded in whole hearts (Fig. 1G) and isolated field-stimulated (11) and patch-clamped myocytes (Fig. 2A). This depression of Ca2+ release could underlie the systolic dysfunction of db/db hearts, although other factors (e.g., contractile myofibrils and elastic properties of the heart) may contribute.

The simultaneous recording of cell shortening, [Ca2+]i transient and ICa in isolated db/db myocytes revealed that cell contraction is decreased without an apparent decrease in excitation-contraction coupling efficiency (Fig. 2E). The cellular contractile dysfunction is consistent with the echocardiographic measurements presented in Fig. 1. Furthermore, the amplitude of the [Ca2+]i transient recorded in isolated cells was reduced (Fig. 2C), which is consistent with data obtained by two-photon microscopy in whole hearts (Fig. 1). Interestingly, the trigger ICa was also reduced in db/db myocytes (Fig. 2D), thereby imposing a limitation on the amount of Ca2+ ions entering into diabetic myocytes. This limitation is likely exacerbated by the shift of ICa activation curve toward more depolarized potentials, which results in a reduction of the “window current.” Given the fact that the activity of single l-type Ca2+ channels is not modified and that the reduction in the density of pore-forming subunit α1c is similar to the reduction in whole-cell ICa, it is likely that the reduction in ICa is the result of fewer l-type Ca2+ channels in db/db mice hearts. A decrease in [Ca2+]i transient has been found in other models of cardiac pathology and contractile dysfunction (rev. in 27), but in most cases, cardiac hypertrophy and normal ICa density accompany the depressed [Ca2+]i transient. In the case of type 2 diabetic mice, a systematic reduction of ICa (Fig. 2D) may underlie the decreased sarcoplasmic reticulum Ca2+ release and contraction either directly, by triggering lower Ca2+ release, or indirectly, by lowering the sarcoplasmic reticulum Ca2+ load. Under a reduced ICa, a new equilibrium may be reached in diabetic hearts where Ca2+ entry, release, and uptake are reduced. The reduction in sarcoplasmic reticulum Ca2+ load may also be due to a depression in SERCA2a activity. In fact, we found the [Ca2+]i transient to decay slower in db/db myocytes. In mice, decline of the [Ca2+]i transient is mainly (∼90%) due to SERCA2a activity, with relatively minor contribution (∼9%) by the NCX (30). Thus slowing of the [Ca2+]i transient in db/db is most likely caused by decreased SERCA activity. In support of this notion, others have shown a large increase in the expression of the SERCA inhibiting protein phospholamban in db/db hearts (11). Moreover, the increased Ca2+ efflux through the NCX can contribute to decreasing the sarcoplasmic reticulum Ca2+ load.

RyR activity can be determined by examining Ca2+ sparks. We found that db/db cardiac myocytes exhibit reduced rate of Ca2+ spark frequency (Fig. 4A and B). This reduction cannot be explained by changes in diastolic [Ca2+]i (11) or ICa (Fig. 3) because Ca2+ sparks were recorded in permeabilized myocytes, where internal [Ca2+] is fixed and ICa is absent. However, permeabilization may wash out soluble components with capacity to modulate RyRs (Na+, Mg2+, sorcin...) that may be altered in db/db myocytes. The decreased Ca2+ spark frequency may instead be an indication of depressed expression of RyRs (Fig. 4C and D) and/or the sarcoplasmic reticulum Ca2+ load (Fig. 4E). The decrease in RyR abundance is consistent with data obtained in type 1 diabetic hearts (3134), although in type 2 diabetes, no change in RyR expression was found in two previous studies (9,11). A decrease in the frequency of spontaneous Ca2+ sparks could have a beneficial effect by preventing Ca2+ leakage from the sarcoplasmic reticulum and enhancing the sarcoplasmic reticulum Ca2+ load (18), analogous to effects seen with FKBP12.6 overexpression (35). But because sarcoplasmic reticulum Ca2+ load was decreased in db/db myocytes (Fig. 4E), this suggests that defects concomitant to reduced RyR expression contribute to altered Ca2+ cycling in db/db mice hearts.

This study has been performed in type 2 diabetic mice with obesity, which recapitulate human pathology. However, even if animal models are a valuable resource to elucidate molecular mechanisms of human pathology, there are clear limitations to translate results from bench to bedside. For example, the heart rate in mice is higher than in humans, so the relative contribution of SERCA2 over NCX is notably different (30).

In conclusion, we have identified impaired Ca2+- induced Ca2+ release in obesity-linked type 2 diabetes that underlies depressed cardiac function. A general reduction in the membrane permeability to Ca2+ was identified. Because less Ca2+ enters the cell through ICa, less Ca2+ is released through RyRs. This is due to a reduction in both sarcolemmal Ca2+ channels and sarcoplasmic reticulum Ca2+ release channels expression.

FIG. 1.

db/db mice show cardiac dysfunction associated with decreased [Ca2+]i transients but lack signs of cardiac hypertrophy. A: M-mode echocardiographic images obtained in +/+ (left) and db/db mouse (right). B: Fractional shortening in +/+ (n = 3) and db/db (n = 3) mice. C: +/+ and db/db heart weight. D: Deconvoluted two-photon image in a mouse heart loaded with di-4 ANEPPS. E: Cell volumes in +/+ (n = 40) and db/db mice (n = 66). F: Two-photon image of mouse heart loaded with Rhod-2 AM. G: Examples of two-photon line-scan images recorded in hearts loaded with Rhod-2 AM in cell from +/+ (top) and db/db heart (middle). Corresponding [Ca2+]i transients (expressed as F/F0) are superimposed at the bottom. Black trace, +/+ cell. Blue trace, db/db cell. Hearts were paced at 2 Hz. H: Averaged values of peak F/F0 in +/+ (113 cells from five hearts) and db/db (77 cells from four hearts). *P < 0.0001.

FIG. 1.

db/db mice show cardiac dysfunction associated with decreased [Ca2+]i transients but lack signs of cardiac hypertrophy. A: M-mode echocardiographic images obtained in +/+ (left) and db/db mouse (right). B: Fractional shortening in +/+ (n = 3) and db/db (n = 3) mice. C: +/+ and db/db heart weight. D: Deconvoluted two-photon image in a mouse heart loaded with di-4 ANEPPS. E: Cell volumes in +/+ (n = 40) and db/db mice (n = 66). F: Two-photon image of mouse heart loaded with Rhod-2 AM. G: Examples of two-photon line-scan images recorded in hearts loaded with Rhod-2 AM in cell from +/+ (top) and db/db heart (middle). Corresponding [Ca2+]i transients (expressed as F/F0) are superimposed at the bottom. Black trace, +/+ cell. Blue trace, db/db cell. Hearts were paced at 2 Hz. H: Averaged values of peak F/F0 in +/+ (113 cells from five hearts) and db/db (77 cells from four hearts). *P < 0.0001.

FIG. 2.

db/db mice exhibit depressed [Ca2+]i transients due to ICa decrease. A: Example recordings taken from +/+ cell (left) and db/db cell (right). From top to bottom: voltage protocol, ICa, line-scan confocal image, fluorescence trace, and shortening trace (CL%, percentage of cell length). B: Cell shortening as a function of voltage in +/+ (open circle and solid line, n = 20) and db/db cells (blue circle and dashed line, n = 11). C: Peak [Ca2+]i transient (F/F0, see Fig. 1) plotted against voltage in +/+ (n = 44) and db/db (n = 20) myocytes. Symbols as in B. D: ICa density plotted against voltage in +/+ and db/db cells. Symbols and number of samples as in C. E: Plot of excitation-contraction coupling gain calculated for each voltage dividing the F/F0 by the integral of ICa (pC/pF). Open circles, +/+ cells; blue circles, db/db cells. *P < 0.05; **P < 0.005.

FIG. 2.

db/db mice exhibit depressed [Ca2+]i transients due to ICa decrease. A: Example recordings taken from +/+ cell (left) and db/db cell (right). From top to bottom: voltage protocol, ICa, line-scan confocal image, fluorescence trace, and shortening trace (CL%, percentage of cell length). B: Cell shortening as a function of voltage in +/+ (open circle and solid line, n = 20) and db/db cells (blue circle and dashed line, n = 11). C: Peak [Ca2+]i transient (F/F0, see Fig. 1) plotted against voltage in +/+ (n = 44) and db/db (n = 20) myocytes. Symbols as in B. D: ICa density plotted against voltage in +/+ and db/db cells. Symbols and number of samples as in C. E: Plot of excitation-contraction coupling gain calculated for each voltage dividing the F/F0 by the integral of ICa (pC/pF). Open circles, +/+ cells; blue circles, db/db cells. *P < 0.05; **P < 0.005.

FIG. 3.

Biophysical properties of the calcium current. A: Steady-state activation and inactivation curves obtained from −60 to +20 mV for +/+ (○ and solid line, n = 6) and db/db (• and dashed line, n = 5). B: Consecutive traces obtained from single-channel recordings within +/+ (left) and db/db (right). Traces were obtained during 150-ms depolarization steps from −100 to +20mV. Bottom traces are ensemble currents obtained in each experiment. C: Top: Representative Western blot of +/+ and db/db samples from ventricular tissue homogenates; bottom: statistical analysis of db/db (n = 5) and +/+ hearts (n = 6) (*P < 0.05).

FIG. 3.

Biophysical properties of the calcium current. A: Steady-state activation and inactivation curves obtained from −60 to +20 mV for +/+ (○ and solid line, n = 6) and db/db (• and dashed line, n = 5). B: Consecutive traces obtained from single-channel recordings within +/+ (left) and db/db (right). Traces were obtained during 150-ms depolarization steps from −100 to +20mV. Bottom traces are ensemble currents obtained in each experiment. C: Top: Representative Western blot of +/+ and db/db samples from ventricular tissue homogenates; bottom: statistical analysis of db/db (n = 5) and +/+ hearts (n = 6) (*P < 0.05).

FIG. 4.

Ca2+ spark occurrence is decreased in db/db hearts, along with a reduction in RyR expression. A: Line-scan images showing spontaneous Ca2+ sparks recorded in +/+ and db/db permeabilized myocytes. B: Ca2+ sparks per second per 100 μm recorded in 10 +/+ and in 14 db/db myocytes. C: Specific [3H]ryanodine binding measured in +/+ (open circles) and db/db (blue circles) hearts. Lines represent fitting of the points with the equation B = Bmax × [3H]ryanodine/(Kd + [3H]ryanodine) with B as the specific binding. D: Averaged Bmax calculated for six +/+ and six db/db hearts. E: Left, line-scan images recorded in +/+ (top) and db/db (bottom) permeabilized myocytes during 20 mmol/l caffeine application. Right, averaged peak fluorescence ratio (measured as F/F0 as in Fig. 1) obtained during caffeine application in +/+ (n = 20) and db/db myocytes (n = 20). *P < 0.05; **P < 0.01.

FIG. 4.

Ca2+ spark occurrence is decreased in db/db hearts, along with a reduction in RyR expression. A: Line-scan images showing spontaneous Ca2+ sparks recorded in +/+ and db/db permeabilized myocytes. B: Ca2+ sparks per second per 100 μm recorded in 10 +/+ and in 14 db/db myocytes. C: Specific [3H]ryanodine binding measured in +/+ (open circles) and db/db (blue circles) hearts. Lines represent fitting of the points with the equation B = Bmax × [3H]ryanodine/(Kd + [3H]ryanodine) with B as the specific binding. D: Averaged Bmax calculated for six +/+ and six db/db hearts. E: Left, line-scan images recorded in +/+ (top) and db/db (bottom) permeabilized myocytes during 20 mmol/l caffeine application. Right, averaged peak fluorescence ratio (measured as F/F0 as in Fig. 1) obtained during caffeine application in +/+ (n = 20) and db/db myocytes (n = 20). *P < 0.05; **P < 0.01.

FIG. 5.

Sarcoplasmic reticulum Ca2+ load is decreased in db/db cardiomyocytes. A: Caffeine-evoked [Ca2+]i transients in a +/+ and db/db myocyte. B: Averaged peak fluorescence ratio (F/F0 as in Fig. 1) in +/+ (n = 35) and db/db (n = 32) myocytes. C: INCX evoked by caffeine (left) and its running integral (right) recorded in a +/+ (top) and a db/db (bottom) myocyte. D: Averaged integral of the INCX recorded during caffeine application in +/+ (n = 6) and db/db (n = 6) myocytes. *P < 0.01; **P < 0.001.

FIG. 5.

Sarcoplasmic reticulum Ca2+ load is decreased in db/db cardiomyocytes. A: Caffeine-evoked [Ca2+]i transients in a +/+ and db/db myocyte. B: Averaged peak fluorescence ratio (F/F0 as in Fig. 1) in +/+ (n = 35) and db/db (n = 32) myocytes. C: INCX evoked by caffeine (left) and its running integral (right) recorded in a +/+ (top) and a db/db (bottom) myocyte. D: Averaged integral of the INCX recorded during caffeine application in +/+ (n = 6) and db/db (n = 6) myocytes. *P < 0.01; **P < 0.001.

TABLE 1

Characteristics of +/+ and db/db mice

Body weight (g)Heart weight (mg)Cm (pF)LVEDD (mm)LVESD (mm)FS (%)SWT (mm)PWT (mm)HR (bpm)
+/+ 22.3 ± 0.8 (17) 171 ± 7 (17) 153 ± 7 (62) 3.1 ± 0.1 (3) 1.2 ± 0.1 (3) 60.6 ± 1.9 (3) 0.7 ± 0.7 (3) 0.7 ± 0.1 (3) 241 ± 18 (3) 
db/db 49.3 ± 1.1 (30)* 184 ± 7 (29) 169 ± 7 (53) 3.5 ± 0.1 (3) 2.3 ± 0.2 (3)* 33.7 ± 5.6 (3)* 0.7 ± 0.1 (3) 0.7 ± 0.1 (3) 238 ± 11 (3) 
Body weight (g)Heart weight (mg)Cm (pF)LVEDD (mm)LVESD (mm)FS (%)SWT (mm)PWT (mm)HR (bpm)
+/+ 22.3 ± 0.8 (17) 171 ± 7 (17) 153 ± 7 (62) 3.1 ± 0.1 (3) 1.2 ± 0.1 (3) 60.6 ± 1.9 (3) 0.7 ± 0.7 (3) 0.7 ± 0.1 (3) 241 ± 18 (3) 
db/db 49.3 ± 1.1 (30)* 184 ± 7 (29) 169 ± 7 (53) 3.5 ± 0.1 (3) 2.3 ± 0.2 (3)* 33.7 ± 5.6 (3)* 0.7 ± 0.1 (3) 0.7 ± 0.1 (3) 238 ± 11 (3) 

Data are means ± SE (n). Cm, membrane capacitance; FS, fractional shortening; HR, heart rate; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; PWT, posterior wall thickness; SWT, septal wall thickness.

*

P < 0.001;

P < 0.05.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

H.H.V. has received National Institutes of Health Grant HL55438. L.P. has received a travel grant from Deutscher Akademischer Austausch Dienst. S.R. and A.M.G. are Centre National de la Recherche Scientifique scientists. The project was funded by Institute National de la Santé et la Recherche Médicale, Université de Montpellier 1, Servier Laboratories, and Programme Nationale de Recherche.

We thank Nancy Benkuski, Georgina Gurrola, and Patrice Bideaux for excellent technical assistance. We also thank Prof. M. Dauzat for advice with echocardiography.

1.
Taegtmeyer H, McNulty P, Young ME: Adaptation and maladaptation of the heart in diabetes: Part I. General concepts.
Circulation
105
:
1727
–1733,
2002
2.
Schaffer SW: Cardiomyopathy associated with noninsulin-dependent diabetes.
Mol Cell Biochem
107
:
1
–20,
1991
3.
Bell DSH: Heart failure: the frequent, forgotten, and often fatal complication of diabetes.
Diabetes Care
26
:
2433
–2441,
2003
4.
Bauters C, Lamblin N, Mc Fadden EP, Van Belle E, Millaire A, De Groote P: Influence of diabetes mellitus on heart failure risk and outcome.
Cardiovasc Diabetol
2
:
1
,
2003
5.
Laakso M: Hyperglycemia and cardiovascular disease in type 2 diabetes.
Diabetes
48
:
937
–942,
1999
6.
Regan TJ, Khan MI, Jesrani MU, Oldewurtel HA, Ettinger PO: Alterations of myocardial function and metabolism in chronic diabetes mellitus.
Recent Adv Stud Cardiac Struct Metab
3
:
169
–178,
1973
7.
Penpargkul S, Schaible T, Yipintsoi T, Scheuer J: The effect of diabetes on performance and metabolism of rat hearts.
Circ Res
47
:
911
–921,
1980
8.
Fein F, Kornstein L, Strobeck J, Capasso J, Sonnenblick E: Altered myocardial mechanics in diabetic rats.
Circ Res
47
:
922
–933,
1980
9.
Misra T, Gilchrist JSC, Russell JC, Pierce GN: Cardiac myofibrillar and sarcoplasmic reticulum function are not depressed in insulin-resistant JCR:LA-cp rats.
Am J Physiol Heart Circ Physiol
276
:
H1811
–H1817,
1999
10.
Abe T, Ohga Y, Tabayashi N, Kobayashi S, Sakata S, Misawa H, Tsuji T, Kohzuki H, Suga H, Taniguchi S, Takaki M: Left ventricular diastolic dysfunction in type 2 diabetes mellitus model rats.
Am J Physiol Heart Circ Physiol
282
:
H138
–H148,
2002
11.
Belke DD, Swanson EA, Dillmann WH: Decreased sarcoplasmic reticulum activity and contractility in diabetic db/db mouse heart.
Diabetes
53
:
3201
–3208,
2004
12.
Schaffer SW, Allo S, Punna S, White T: Defective response to cAMP-dependent protein kinase in non-insulin-dependent diabetic heart.
Am J Physiol
261
:
E369
–E376,
1991
13.
Allo SN, Lincoln TM, Wilson GL, Green FJ, Watanabe AM, Schaffer SW: Non-insulin-dependent diabetes-induced defects in cardiac cellular calcium regulation.
Am J Physiol
260
:
C1165
–C1171,
1991
14.
Belke DD, Larsen TS, Gibbs EM, Severson DL: Altered metabolism causes cardiac dysfunction in perfused hearts from diabetic (db/db) mice.
Am J Physiol Endocrinol Metab
279
:
E1104
–E1113,
2000
15.
Bénitah J-P, Perrier E, Gómez AM, Vassort G: Effects of aldosterone on transient outward K+ current density in rat ventricular myocytes.
J Physiol (Lond)
537
:
151
–126,
2001
16.
Rubart M, Wang E, Dunn KW, Field LJ: Two-photon molecular excitation imaging of Ca2+ transients in Langendorff-perfused mouse hearts.
Am J Physiol Cell Physiol
284
:
C1654
–C1668,
2003
17.
Gómez AM, Kerfant BG, Vassort G: Microtubule disruption modulates Ca(2+) signaling in rat cardiac myocytes.
Circ Res
86
:
30
–36,
2000
18.
Overend CL, Eisner DA, O’Neill SC: The effect of tetracaine on spontaneous Ca2+ release and sarcoplasmic reticulum calcium content in rat ventricular myocytes.
J Physiol (Lond)
502
:
471
–479,
1997
19.
Farrell EF, Antaramián A, Rueda A, Gómez AM, Valdivia HH: Sorcin inhibits calcium release and modulates excitation-contraction coupling in the heart.
J Biol Chem
278
:
34660
–34666,
2003
20.
Heredia Mdel P, Delgado C, Pereira L, Perrier R, Richard S, Vassort G, Benitah JP, Gomez AM: Neuropeptide Y rapidly enhances [Ca(2+)](i) transients and Ca(2+) sparks in adult rat ventricular myocytes through Y(1) receptor and PLC activation.
J Mol Cell Cardiol
38
:
205
–212,
2005
21.
Foerster K, Groner F, Matthes J, Koch WJ, Birnbaumer L, Herzig S: Cardioprotection specific for the G protein Gi2 in chronic adrenergic signaling through beta 2-adrenoceptors.
Proc Natl Acad Sci U S A
100
:
14475
–14480,
2003
22.
Schroder F, Handrock R, Beuckelmann DJ, Hirt S, Hullin R, Priebe L, Schwinger RH, Weil J, Herzig S: Increased availability and open probability of single l-type calcium channels from failing compared with nonfailing human ventricle.
Circulation
98
:
969
–976,
1998
23.
Lokuta AJ, Meyers MB, Sander PR, Fishman GI, Valdivia HH: Modulation of cardiac ryanodine receptors by sorcin.
J Biol Chem
272
:
25333
–25338,
1997
24.
Perrier E, Kerfant BG, Lalevee N, Bideaux P, Rossier MF, Richard S, Gómez AM, Bénitah JP: Mineralocorticoid receptor antagonism prevents the electrical remodeling that precedes cellular hypertrophy after myocardial infarction.
Circulation
110
:
776
–783,
2004
25.
Solaro RJ, Rarick HM: Troponin and tropomyosin: proteins that switch on and tune in the activity of cardiac myofilaments.
Circ Res
83
:
471
–480,
1998
26.
Cheng H, Lederer WJ, Cannell MB: Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle.
Science
262
:
740
–744,
1993
27.
Bénitah JP, Kerfant BG, Vassort G, Richard S, Gómez AM: Altered communication between l-type calcium channels and ryanodine receptors in heart failure.
Front Biosci
7
:
e263
–e275,
2002
28.
Gómez AM, Schwaller B, Porzig H, Vassort G, Niggli E, Egger M: Increased exchange current but normal Ca2+ transport via Na+-Ca2+ exchange during cardiac hypertrophy after myocardial infarction.
Circ Res
91
:
323
–330,
2002
29.
Semeniuk LM, Kryski AJ, Severson DL: Echocardiographic assessment of cardiac function in diabetic db/db and transgenic db/db-hGLUT4 mice.
Am J Physiol Heart Circ Physiol
283
:
H976
–H982,
2002
30.
Bers DM:
Excitation-Contraction Coupling and Cardiac Contractile Force.
Dordrecht, the Netherlands, Kluwer Academic Publishers,
2001
31.
Yu Z, Tibbits GF, McNeill JH: Cellular functions of diabetic cardiomyocytes: contractility, rapid-cooling contracture, and ryanodine binding.
Am J Physiol
266
:
H2082
–H2089,
1994
32.
Netticadan T, Temsah RM, Kent A, Elimban V, Dhalla NS: Depressed levels of Ca2+-cycling proteins may underlie sarcoplasmic reticulum dysfunction in the diabetic heart.
Diabetes
50
:
2133
–2138,
2001
33.
Bidasee KR, Nallani K, Henry B, Dincer UD, Besch HR Jr: Chronic diabetes alters function and expression of ryanodine receptor calcium-release channels in rat hearts.
Mol Cell Biochem
249
:
113
–123,
2003
34.
Guner S, Arioglu E, Tay A, Tasdelen A, Aslamaci S, Bidasee KR, Dincer UD: Diabetes decreases mRNA levels of calcium-release channels in human atrial appendage.
Mol Cell Biochem
263
:
143
–150,
2004
35.
Gómez AM, Schuster I, Fauconnier J, Prestle J, Hasenfuss G, Richard S: FKBP12.6 overexpression decreases Ca2+ spark amplitude but enhances [Ca2+]i transient in rat cardiac myocytes.
Am J Physiol Heart Circ Physiol
287
:
H1987
–H1993,
2004