Troglitazone Inhibits Voltage-Dependent Calcium Currents in Guinea Pig Cardiac Myocytes

Myocardial contractile dysfunction is a major complication of diabetes, known as diabetic cardiomyopathy.¹²³⁴ The subcellular mechanisms responsible for cardiomyopathy are unknown. However, several cellular defects, including depressions in sarcoplasmic reticular Ca²⁺ uptake,⁵ Na⁺-K⁺ pump,⁶ sarcolemmal Ca²⁺ pump, Na⁺-Ca²⁺ exchanger activities,⁷ and the alteration of mitochondrial functions,⁸ have been suggested to be contributors to the development of this disease. The net result of these changes in Ca²⁺ homeostasis causes an intracellular Ca²⁺ overload, thereby resulting in cellular damage and, ultimately, diabetic cardiomyopathy. Moreover, diabetes prolongs the action potential duration^9101112 and increases the number of myocardial voltage-dependent Ca²⁺ channels,¹³¹⁴ something that may also play a role in causing diabetic cardiomyopathy. In fact, it has been reported that in chronically diabetic rats, elevated tissue Ca²⁺ levels are present¹⁵ and treatment with verapamil or diltiazem, a voltage-dependent L-type Ca²⁺ channel blocker, lessens cardiac dysfunction.¹⁶¹⁷¹⁸ Thus, it is likely that excess Ca²⁺ influx through the voltage-dependent Ca²⁺ channels contributes to induce intracellular Ca²⁺ overload and consequently diabetic cardiomyopathy.

Troglitazone, a novel member of the insulin-sensitizing thiazolidinediones, has been widely used to treat patients with non–insulin-dependent diabetes mellitus and other insulin- resistant diseases. Treatment with troglitazone reduced hyperglycemia, plasma triglycerides, and blood pressure.^19202122 Recent studies show that troglitazone attenuates high- glucose–induced abnormalities in relaxation and intracellular calcium in rat ventricular myocytes²³ and may improve cardiac function in diabetic patients.²⁴ Until now, the mechanisms underlying the beneficial effects of troglitazone on hearts have not been clearly established, but several articles have shown that troglitazone inhibits the voltage-dependent L-type Ca²⁺ currents (I_Ca,L) in vascular smooth muscle cells.²⁵²⁶

Therefore, the purpose of the present study was to clarify the effects of troglitazone on the voltage-dependent Ca²⁺ currents (T-type [I_Ca,T] and L-type) in cardiac myocytes. We have also made comparisons with the classic Ca²⁺ antagonists verapamil and nifedipine.

Methods

Cell Preparation

Single atrial myocytes were obtained from guinea pig hearts by an enzymatic dissociation procedure described elsewhere.²⁷ Briefly, the animals were anesthetized with sodium pentobarbital, and their hearts were rapidly removed and retrogradely Langendorff-perfused at 35°C to 37°C with an oxygenated Tyrode’s solution. The hearts were then perfused with Ca²⁺-free Tyrode’s solution for ≈10 minutes and subsequently with the same solution containing collagenase (0.04% wt/vol type 1, Sigma Chemical Co) for 17 to 20 minutes. The digested hearts were stored in a high-K⁺/low-Cl⁻ solution²⁷ at 5°C for later experimentation. The atria were then removed, and cells were obtained by gentle mechanical agitation. This procedure consistently yielded an acceptable number of quiescent and relaxed atrial cells.

Solution and Drugs

The composition of the normal Tyrode’s solution was (in mmol/L) NaCl 136.5, KCl 5.4, CaCl₂ 1.8, MgCl₂ 0.53, glucose 5.5, and HEPES-NaOH buffer 5, pH 7.4. The Ca²⁺-free Tyrode’s solution was identical to normal Tyrode’s solution except that CaCl₂ was omitted. To record voltage-dependent Ca²⁺ currents, K⁺ currents were eliminated by the internal Cs and external Ba (5 mmol/L), and Ca²⁺-activated currents were blocked by 10 mmol/L EGTA and 2 mmol/L BAPTA in the internal solution. The composition of the internal solution was (in mmol/L) CsCl 140, EGTA 10, BAPTA 2, Na₂-ATP 3, GTP (sodium salt, Sigma) 0.1, MgCl₂ 1, and HEPES-CsOH buffer 5, pH 7.3. In the experiments in which the cells were held at −80 mV, the bath was perfused with the following solution (in mmol/L) to block the voltage-dependent Na⁺ current: tetraethylammonium chloride (TEA-Cl) 140, BaCl₂ 5, MgCl₂ 0.53, glucose 5.5, tetrodotoxin (TTX) 0.01, and HEPES-CsOH buffer 5, pH 7.4. Troglitazone was obtained from Sankyo Co Ltd. Troglitazone was dissolved in DMSO to give a stock solution of 1 to 30 mmol/L, and the final concentration of DMSO applied to the bathing solution was 0.1%. Nifedipine and verapamil were dissolved in ethanol to give a stock solution of 10 mmol/L. In several experiments, cAMP was added to the pipette solution. (±)Isoproterenol, cAMP, verapamil, and nifedipine were purchased from Sigma.

Recording Technique and Data Analysis

Membrane currents were recorded with patch electrodes in a whole-cell clamp configuration²⁷²⁸ and a patch-clamp amplifier (EPC-7, List Electronics). The heat-polished patch electrodes had a tip resistance of 3 to 6 MΩ. The membrane currents were monitored with a high-gain storage oscilloscope (COS 5020-ST, Kikusui Electronics). At the start of each experiment, the series resistance was compensated. The data were stored on video cassettes with a PCM converter system (RP-880, NF electronic circuit design). Later, the data were reproduced, low-pass–filtered at 2 kHz (−3 dB) with a Bessel filter (FV-665, NF, 48-dB/octave slope attenuation), sampled at 5 kHz, and analyzed off-line on a computer with p-Clamp software (Axon Instruments). In general, we used a holding potential of −40 mV at a frequency of 0.2 Hz to inactivate the voltage-dependent Na⁺ current. In experiments to evaluate the contribution of I_Ca,T or voltage-dependence of the drug, a holding potential of −80 mV was used in combination with the high-TEA solution containing Ba²⁺ 5 mmol/L in place of Ca²⁺ (see Methods). Statistical results are expressed as mean±SD. Student’s t tests were performed, with a value of P<0.05 considered significant.

The first data were usually taken after the current amplitude of Ca²⁺ currents had been stabilized (2 to 3 minutes after the rupture of the membrane). After that, we could investigate the effects of drugs on the voltage-dependent Ca²⁺ currents for ≈15 to 20 minutes. In experiments with cAMP, data were taken immediately after the rupture of the membrane. To measure the amplitude of the voltage-dependent Ca²⁺ currents, we subtracted from the peak amplitude of Ca²⁺ currents in the original trace to the current level in the presence of Cd²⁺ (1 mmol/L). In preliminary experiments, we confirmed that 0.1% DMSO did not affect the voltage-dependent Ca²⁺ currents significantly. Furthermore, to exclude the effects of DMSO, 0.1% DMSO was always added to the bathing solution. The steady-state inactivation parameters of the voltage-dependent Ca²⁺ currents were analyzed with double-pulse protocols. Conditioning voltage pulses (3 seconds in duration) for various membrane potentials between −70 and +0 mV were applied from a holding potential of −80 mV. Ten milliseconds after the end of each conditioning pulse, a test pulse of +10 mV (0.2 seconds in duration) was applied to elicit Ca²⁺ currents. The ratio between the amplitude of the Ca²⁺ currents with conditioning pulse and that without conditioning pulse was plotted for the membrane potential of each conditioning pulse. The interval between sets of double pulses was 20 seconds.

Results

Effects of Troglitazone on Voltage-Dependent I_Ca,L

The effects of troglitazone on the voltage-dependent I_Ca,L were examined in single atrial myocytes (Figure 1). The membrane potential was held at −40 mV, and command voltage pulses (320 ms in duration) to +0 mV were applied at 0.2 Hz. In control cells, a transient inward current was elicited during each voltage pulse (Figure 1A, a). The inward current was blocked by nifedipine 1 μmol/L (see Figures 9 and 10), verapamil 1 μmol/L (Figure 9), and Cd²⁺ 1 mmol/L (Figure 1A, f), indicating that it consisted of I_Ca,L. Troglitazone 10 μmol/L reduced the amplitude of I_Ca,L from −500 to −140 pA in this cell within 2 minutes (Figure 1A, b). After washout, I_Ca,L gradually returned to near control level. The time courses of changes in the peak I_Ca,L measured from the zero current level are shown in Figure 1B. Troglitazone 30 μmol/L almost completely blocked I_Ca,L (Figure 1A, d). After washout, the depressed current gradually increased, but the inhibitory effects of troglitazone 30 μmol/L were not easily reversible. Figure 2 shows the effects of troglitazone on current-voltage relationships of I_Ca,L. The cells were held at −40 mV, and command voltage steps to various membrane potentials were applied at 0.2 Hz. The current-voltage relationships of I_Ca,L measured at the peak inward current are shown in Figure 2B. Troglitazone reduced the current amplitude of I_Ca,L at any command voltage without affecting the voltage dependence of I_Ca,L. On average, troglitazone 5 μmol/L decreased peak I_Ca,L at +0 mV by 42±6% (n=5). Troglitazone 30 μmol/L almost completely blocked I_Ca,L. The reversal potential for I_Ca,L was not altered significantly by troglitazone. These results suggest that troglitazone inhibited I_Ca,L in cardiac myocytes. The effects of various concentrations of troglitazone on the amplitude of I_Ca,L are shown in Figure 3. The cells were held at −40 mV, and the command pulses to +0 mV were applied at 0.2 Hz. Troglitazone at concentrations >1 μmol/L decreased I_Ca,L, and the half-maximal inhibitory concentration (IC₅₀) of troglitazone on I_Ca,L was 6.3 μmol/L.

Figures 4 and 5 illustrate the effects of troglitazone on isoproterenol- and cAMP-enhanced I_Ca,L. Isoproterenol 1 μmol/L increased the amplitude of I_Ca,L (Figure 4, b). Immediately after application of isoproterenol, there was a rapid small increase in I_Ca,L, probably reflected by direct activation of the GTP-binding proteins (G_s),²⁹ and then a large increase in I_Ca,L was observed. The additional application of troglitazone 30 μmol/L completely abolished I_Ca,L (Figure 4, c). Moreover, when cAMP was applied through the patch pipette, I_Ca,L increased from −370 to −1080 pA in this cell (Figure 5A and 5B). Troglitazone 30 μmol/L also abolished I_Ca,L (Figure 5A, c and 5B, c). Figure 5D shows the current-voltage relationships of the peak I_Ca,L in the presence of cAMP (Figure 5C, a) and with the additional application of troglitazone 30 μmol/L (Figure 5C, b). Troglitazone 30 μmol/L decreased I_Ca,L at all command potentials. These results suggest that troglitazone antagonizes the effects of isoproterenol on I_Ca,L independently of β-adrenergic receptors.

Effects of Troglitazone on the Voltage-Dependent I_Ca,L and I_Ca,T

The existence of 2 distinct Ca²⁺ currents has been shown for cardiac myocytes in several kinds of mammalian hearts.^30313233I_Ca,T activates at low voltages and inactivates quickly; I_Ca,L activates at high voltages and inactivates slowly. In addition, the T-type Ca²⁺ channel is about equally permeable to Ca²⁺ and Ba²⁺ ions and has the same inactivation kinetics in Ba²⁺ as in Ca²⁺; the L-type Ca²⁺ channel is more permeable to Ba²⁺ and has a dramatically slower inactivation time in Ba²⁺ than Ca²⁺. To clarify whether both types of Ca²⁺ currents can be identified in guinea pig atrial myocytes, we carried out tests under the conditions in which extracellular Na⁺ ions were replaced by impermeable TEA⁺ ions, and 5 mmol/L BaCl₂ was added in place of Ca²⁺. Sodium removal induced cell contracture, but under our conditions with EGTA 10 mmol/L and BAPTA 2 mmol/L in the patch pipette, the cell attached to the patch electrode survives, probably owing to the diffusion of EGTA and BAPTA into the cytosol. The cells were held at −40 or −80 mV (Figure 6A), and command voltage steps (320 ms in duration) were applied to various membrane potentials. The current-voltage relationships of the peak inward current are shown in Figure 6C. At a holding potential of −40 mV, the inward current was elicited at positive potentials to −30 mV (Figure 6A, right). A small fraction of current was inactivated at the command pulses to −20 and −10 mV. Conversely, when the cell was held at −80 mV, the transient inward current was recorded at a command potential of −30 mV and was overlapped on the noninactivated component at a command potential of −20 mV (Figure 6A, left). The current traces subtracted from the current of a holding potential of −80 mV to that of a holding potential of −40 mV at command potentials of −30, −20, and +0 mV are shown in Figure 6B. The transient inward current rapidly inactivated within 50 ms and could be discriminated from the sustained component. Cd²⁺ 1 mmol/L abolished both types of inward current, but nifedipine 1 μmol/L (data not shown) failed to inhibit the transient component. These findings suggest that both types of Ca²⁺ currents exist in guinea pig atrial myocytes. The fast inward current consisted of I_Ca,T, and the slow component consisted primarily of I_Ca,L. Figure 7 shows the effects of troglitazone on both types of Ca²⁺ currents. I_Ca,T and I_Ca,L were elicited at a command voltage to −30 and +10 mV from a holding potential of −80 mV, respectively. Troglitazone 10 μmol/L inhibited both types of Ca²⁺ currents (Figure 7A and 7B) but inhibited I_Ca,L more effectively than I_Ca,T (Figure 7B).

Effects of Troglitazone on the Kinetic Parameters of the Voltage-Dependent I_Ca,L

Figure 8 shows the effects of troglitazone on the inactivation time courses of I_Ca,L. Under conditions in which the cell was perfused with normal Tyrode’s solution, the inactivation time courses of I_Ca,L were well fitted by the sum of 2 exponentials (Figure 8A and Table) as previously described.³⁴ Troglitazone 10 μmol/L did not affect the time courses of inactivation of I_Ca,L significantly (Figure 8 and Table). The differences between the values of τ₁ and τ₂ in the control and those in the presence of troglitazone were not statistically significant.

The use-dependent block of troglitazone was also examined and compared with the classic Ca²⁺ antagonists verapamil and nifedipine as shown in Figure 9. The changes in the amplitude of I_Ca,L elicited by successively applied command pulses were measured in the absence or presence of each drug with a test depolarizing pulse to +0 mV from a holding potential of −40 mV at 0.2 Hz. The amplitude of I_Ca,L recorded by the last pulse of a train stimulation before application of the agents (Figure 9A through D) was normalized to 1.0. In control conditions (Figure 9A), the amplitude of I_Ca,L elicited by the first command pulse (b) was not inhibited and remained nearly constant during the successive repetitive pulses (c). The small decrease of the current (8±3% of the first pulse, n=5) during 1-minute application of repetitive stimulation was thought to be induced simply by Ca²⁺ channel rundown. Verapamil (1 μmol/L, Figure 9B) produced very little inhibition of Ca²⁺ current in the absence of test pulses (b), but blockade increased with repeated depolarizations (c). Conversely, in studies with the same pulse protocol, nifedipine blockade was different (Figure 9C). The first current after the quiescent period gave a good estimate of the final level of blockade (b and c). Figure 9D shows the use-dependent effects of troglitazone 10 μmol/L. As in the case of nifedipine, the inward current elicited by the first command pulse after a 2-minute quiescent period was consistently inhibited by 63±10% (n=5, P<0.01), which was different from that recorded with verapamil. With the repetitive stimulations, the inward current decreased slightly, by 13±7% (n=5) from the first pulse during repetitive stimulation (Figure 9D, b), but could not discriminate the simple rundown of the Ca²⁺ channel. These results suggest that troglitazone inhibited I_Ca,L mainly in a use-independent manner.

The influence of the holding potential on the inhibitory effects of nifedipine and troglitazone was compared as shown in Figure 10. In these experiments, the command voltage steps (320 ms in duration) to +10 mV from a holding potential of −40 or −80 mV, where the voltage-dependent Ca²⁺ currents consisted primarily of I_Ca,L, were applied. The peak amplitude of I_Ca,L in the absence of drugs was normalized to 100. The percentage inhibition induced by nifedipine and troglitazone is shown in Figure 10B. Nifedipine 1 μmol/L reduced the amplitude of I_Ca,L by 96±4% at a holding potential of −40 mV, whereas it inhibited it by only 34±8% at a holding potential of −80 mV. Conversely, troglitazone 10 μmol/L inhibited I_Ca,L by ≈70% at each membrane potential. Furthermore, the effects of troglitazone and nifedipine on the voltage-dependent availability of L-type Ca²⁺ channels were examined by means of double-pulse protocols (Figure 11). The test pulse to +10 mV from a holding potential of −80 mV was preceded by a 3-second conditioning pulse to various membrane potentials. The relationships between membrane potentials and the f_∞ value in the absence and presence of the drug were fitted by the following Boltzmann equation using the least-squares method: f_∞(V)=f_∞max/{1+exp[(V−a)/b]}, where f_∞max is the maximal value of f_∞ (in control conditions, the value of f_∞max=1), V is membrane potential in mV, a is membrane potential at 1/2 f_∞max, and b is slope factor. In the absence of the drug, f_∞max=1, a=−21.2 mV, and b=5.13 mV. In the presence of nifedipine 1 μmol/L, f_∞max=0.68, a=−37.0 mV, and b=6.6 mV (Figure 11A). Thus, nifedipine decreased the maximal Ca²⁺ channel availability (0.69±0.05 of the control, n=5), with a significant shift of the curve toward the negative (−18±4 mV, n=5). Conversely, in the absence of troglitazone 10 μmol/L, f_∞max=1, a=−21.3 mV, and b=5.33 mV. In the presence of troglitazone, f_∞max=0.39, a=−22.3 mV, and b=6.0 mV (Figure 11B). Thus, troglitazone reduced the maximal Ca²⁺ channel availability (0.36±0.1 of control, n=5) but did not show any significant shift of the voltage-dependent inactivation curve (−23.4±3.4 mV in the control versus −25.9±5.0 mV in the presence of troglitazone, n=5, P=NS).

Discussion

We have demonstrated here that troglitazone had inhibitory effects on voltage-dependent Ca²⁺ currents in single atrial myocytes from the guinea pig. The inward Ba²⁺ current in place of Ca²⁺ could easily be divided into 2 components with distinct physiological properties, as described in mammalian cardiac myocytes.^30313233 One component had characteristics identical to the I_Ca,L, including slow inactivation (Figure 6) and sensitivity to dihydropyridine (Figures 9 and 10) and isoproterenol (Figure 4). The second component, I_Ca,T, had a fast inactivation (Figures 6 and 7) and small amplitude (≈−100 pA) even in the presence of Ba²⁺ and insensitivity to dihydropyridine (nifedipine 1 μmol/L) and isoproterenol (data not shown). We found that troglitazone inhibited both types of voltage-dependent Ca²⁺ currents in atrial myocytes, although it inhibited I_Ca,L more effectively than I_Ca,T. Thus, the effects of troglitazone on Ca²⁺ currents might not be restricted to L-type Ca²⁺ channels, in comparison with the classic Ca²⁺ channel antagonists nifedipine and verapamil, because nifedipine and verapamil 1 μmol/L did not inhibit I_Ca,T significantly (data not shown). Troglitazone 1 μmol/L reduced I_Ca,L by 10% to 20% within 2 minutes of application, and 10 μmol/L troglitazone reduced it by 60% to 80%. Troglitazone 30 μmol/L almost completely abolished I_Ca,L, and the IC₅₀ value was estimated at 6.3 μmol/L. The inhibitory potency of troglitazone on I_Ca,L was less than that of nifedipine and verapamil. However, because the therapeutic plasma concentration of troglitazone was estimated to be 0.6 to 2.7 μmol/L,³⁵ these concentrations are nearly the same as those required for the inhibition of I_Ca,L in this study. Thus, troglitazone may affect cardiac function by inhibiting the channel. The direct evidence showing that troglitazone inhibits I_Ca,L has been shown in vascular smooth muscle cells.²⁵²⁶ The IC₅₀ of troglitazone on I_Ca,L of vascular smooth muscle cells was ≈3 μmol/L,²⁶ which was relatively lower than that in the present study. However, we conclude that troglitazone inhibited the L-type Ca²⁺ channels in cardiac myocytes as well as vascular smooth muscle cells in therapeutic concentrations.

It has been reported that voltage-dependent L-type Ca²⁺ channel blockers such as verapamil and diltiazem prevent the development of diabetic cardiomyopathy.¹⁶¹⁷¹⁸ These cardioprotective effects of Ca²⁺-blocking drugs have also been reported in Syrian cardiomyopathic hamsters³⁶ and in patients with hypertrophic cardiomyopathy.³⁷ Therefore, the mode of action of troglitazone on I_Ca,L was compared with that of the classic Ca²⁺ antagonists verapamil and nifedipine. As shown in Figures 10 and 11, troglitazone reduced I_Ca,L but did not cause a significant shift in the steady-state inactivation curve. Conversely, nifedipine, a dihydropyridine Ca²⁺ antagonist, which has a high affinity for the inactivated state of the channel but much less affinity for other states (eg, closed, open), showed strong voltage-dependent effects and caused a distinct negative shift of the steady-state inactivation curve. Thus, it is unlikely that troglitazone inhibits I_Ca,L by preferentially binding the inactivated states of the channels. Also, troglitazone did not exhibit significant use-dependent characteristics, which was different from verapamil (Figure 9), as previously described.³⁸ Potencies of the use-dependent inhibition might be closely related to the ionization constants of the drug as shown by Sanguinetti and Kass.³⁹ According to this model, charged forms of the drug can reach their receptors inside the channel by a hydrophilic pathway available only when the channel gates are open and hence are characterized by a significant use-dependent block. In contrast, an uncharged form of the drug easily reaches its receptors via a hydrophobic region of the membrane without channel opening and thereby does not show significant use-dependent effects. Verapamil (pK_a=8.7) is almost entirely in the charged form at pH 7.4, whereas troglitazone (pK_a=6.1)⁴⁰ exists almost entirely in the neutral form at the same pH. Thus, opening of the channels may not be necessary for troglitazone to affect I_Ca,L, as shown in Figure 9. Furthermore, the time courses of Ca²⁺ current decay were little affected by troglitazone. From these observations, troglitazone did not appear to inhibit the Ca²⁺ channels by binding to activated Ca²⁺ channels. Thus, the mechanisms by which troglitazone affects the voltage-dependent Ca²⁺ channels are unknown at present, but troglitazone may interact with L-type Ca²⁺ channels in a manner distinct from the classic Ca²⁺ antagonists.⁴¹

The present study indicates that troglitazone inhibited the voltage-dependent Ca²⁺ currents (I_Ca,L and I_Ca,T) in cardiac myocytes in therapeutic concentrations. Under normal circumstances, the current through the T-type Ca²⁺ channel is unlikely to be very important in atrial and ventricular myocytes, because in a well-polarized cell, such as atrial and ventricular cells, Na current is much larger and activates in a similar voltage range. Also, because L-type Ca²⁺ channels inactivate more slowly, they are likely to be more important than T-type channels. However, T-type Ca²⁺ channels may contribute to the generation of pacemaker activities in pacemaker cells⁴² and may make hypertrophied ventricular myocytes more prone to spontaneous action potentials and increase the likelihood of arrhythmia in partially depolarized hypertrophied myocardium.⁴³ Troglitazone may affect the electrical activities under these conditions by inhibiting I_Ca,T. Conversely, troglitazone inhibited I_Ca,L more effectively than I_Ca,T. The inhibitory effects of troglitazone on I_Ca,L did not show significant voltage- and use-dependent properties as observed in classic Ca²⁺ antagonists.³⁸ From these unique actions, troglitazone may inhibit cardiac Ca²⁺ channels in a similar way in well-polarized cells as well as in partially depolarized cells. Also, it may antagonize the effects of isoproterenol on I_Ca,L. Several studies have shown that in diabetic animals, the duration of action potential in cardiac myocytes is markedly longer, whereas the resting membrane potential is not altered.^9101112 In addition, an augmented number of Ca²⁺ antagonist receptor binding sites and an increase of voltage-dependent L-type Ca²⁺ channels have been reported in diabetic hearts.¹³¹⁴ The increased influx of Ca²⁺ through the voltage-dependent Ca²⁺ channels may cause Ca²⁺ overload, which appears to be linked to the cardiac pathology in diabetic cardiomyopathy.¹⁶¹⁷¹⁸ The present study shows that troglitazone inhibits voltage-dependent Ca²⁺ currents (I_Ca,T and I_Ca,L) and then antagonizes the effects of isoproterenol in cardiac myocytes, which may play a role in preventing diabetes-induced intracellular Ca²⁺ overload and then myocardial changes. In fact, recent studies have shown that troglitazone attenuates high-glucose–induced abnormalities in relaxation and intracellular calcium in rat ventricular myocytes²³ and improves cardiac function in diabetes mellitus.²⁴ From these observations, troglitazone may be a unique agent for diabetic cardiomyopathy, but further studies are needed to clarify this possibility in diabetic patients.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes