Decreased l-Type Ca2+ Current in Cardiac Myocytes of Type 1 Diabetic Akita Mice Due to Reduced Phosphatidylinositol 3-Kinase Signaling

  1. Zhongju Lu1,
  2. Ya-Ping Jiang2,
  3. Xin-Hua Xu3,
  4. Lisa M. Ballou2,
  5. Ira S. Cohen1 and
  6. Richard Z. Lin1,2,4
  1. 1Department of Physiology and Biophysics and the Institute of Molecular Cardiology, Stony Brook University, Stony Brook, New York
  2. 2Department of Medicine, Stony Brook University, Stony Brook, New York
  3. 3Department of Cardiac Surgery, 2nd Hospital Xiang Ya Medical School, Central South University, Changsha, China
  4. 4Department of Veterans Affairs Medical Center, Northport, New York
  1. Address correspondence and reprint requests to Richard Z. Lin, Department of Medicine, Division of Hematology and Oncology, Stony Brook University, Stony Brook, NY 11794-8151. E-mail: richard.lin{at}sunysb.edu

Abstract

OBJECTIVE—Contraction of cardiac myocytes is initiated by Ca2+ entry through the voltage-dependent l-type Ca2+ channel (LTCC). Previous studies have shown that phosphatidylinositol (PI) 3-kinase signaling modulates LTCC function. Because PI 3-kinases are key mediators of insulin action, we investigated whether LTCC function is affected in diabetic animals due to reduced PI 3-kinase signaling.

RESEARCH DESIGN AND METHODS—We used whole-cell patch clamping and biochemical assays to compare cardiac LTCC function and PI 3-kinase signaling in insulin-deficient diabetic mice heterozygous for the Ins2Akita mutation versus nondiabetic littermates.

RESULTS—Diabetic mice had a cardiac contractility defect, reduced PI 3-kinase signaling in the heart, and decreased l-type Ca2+ current (ICa,L) density in myocytes compared with control nondiabetic littermates. The lower ICa,L density in myocytes from diabetic mice is due at least in part to reduced cell surface expression of the LTCC. ICa,L density in myocytes from diabetic mice was increased to control levels by insulin treatment or intracellular infusion of PI 3,4,5-trisphosphate [PI(3,4,5)P3]. This stimulatory effect was blocked by taxol, suggesting that PI(3,4,5)P3 stimulates microtubule-dependent trafficking of the LTCC to the cell surface. The voltage dependence of steady-state activation and inactivation of ICa,L was also shifted to more positive potentials in myocytes from diabetic versus nondiabetic animals. PI(3,4,5)P3 infusion eliminated only the difference in voltage dependence of steady-state inactivation of ICa,L.

CONCLUSIONS—Decreased PI 3-kinase signaling in myocytes from type 1 diabetic mice leads to reduced Ca2+ entry through the LTCC, which might contribute to the negative effect of diabetes on cardiac contractility.

Cardiac complications are an important cause of morbidity and mortality in type 1 diabetic patients. This is partly due to the presence of hypertension and coronary artery disease, which are commonly associated with diabetes. Diabetes also increases the risk of developing cardiac dysfunction independently of these risk factors, supporting the existence of a distinct diabetic cardiomyopathy (1). A number of studies have shown that Ca2+ entry through the voltage-dependent l-type Ca2+ channel (LTCC) is reduced in cardiac myocytes from streptozotocin-induced diabetic rats and from obese db/db mice, a well-known model of type 2 diabetes (25). This inward Ca2+ current (ICa,L) is the critical initiator of the contractile cycle in cardiac myocytes, and inhibition of LTCC function would reduce Ca2+ entry and contractile force. However, the molecular mechanism that underlies the LTCC defect in diabetic myocytes is unclear.

Class I phosphatidylinositol (PI) 3-kinases preferentially phosphorylate PI 4,5-bisphosphate [PI(4,5)P2] to form phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P3] in vivo and exhibit substantial activation in response to stimulation with insulin or other hormones. Studies in neurons and vascular smooth muscle cells indicate that the LTCC is regulated by the PI 3-kinase signaling pathway in these cell types (6,7). Studies by our laboratory and other investigators indicate that PI 3-kinase signaling also modulates LTCC function in cardiac myocytes (810). Our group recently reported that inhibition of PI 3-kinase by Gαq led to a reduction in ICa,L density in cardiac myocytes (9,11). We also found that infusion of PI(3,4,5)P3 into Gαq transgenic myocytes fully restored the ICa,L density to wild-type levels (9).

LTCCs are composed of α1-, β-, and δ/α2-subunits and, in some forms, an additional γ-subunit. There are multiple isoforms of each subunit. The α1-subunit forms the pore of the channel, and the intracellular β-subunit regulates cellular localization. Cardiac LTCCs are mostly composed of α1C- and β2a-isoforms, although other β-variants are also present (12). In a heterologous expression system, PI 3-kinase signaling stimulated ICa,L density through Akt-mediated phosphorylation of β2a-subunits, leading to increased trafficking of the LTCC to the plasma membrane (13). The protein kinase Akt (also known as PKB) is an important downstream effector of PI 3-kinase that is activated upon binding to PI(3,4,5)P3.

These results led us to speculate that decreased PI 3-kinase signaling as a result of insulin deficiency in type 1 diabetes would result in reduced ICa,L density. To test this hypothesis, we used mice heterozygous for the Ins2Akita mutation. These animals start to develop hyperglycemia at 3–4 weeks of age, and the diabetic animals show a significant reduction in plasma insulin levels at 7 weeks (14). Islets from these mice are depleted of β-cells, and those remaining release very little mature insulin. It is thought that the dominant phenotype of the Ins2Akita mutation is due to accumulation of misfolded mutant proinsulin 2 protein in β-cells. This disrupts the secretory pathways that are needed for secretion of wild-type insulin (15). We show here that ICa,L density is reduced in cardiac myocytes isolated from the diabetic Akita mice compared with nondiabetic control mice, and acute insulin treatment or intracellular infusion of PI(3,4,5)P3 reverses this defect.

RESEARCH DESIGN AND METHODS

Phospho-Thr308 Akt antibody was purchased from Cell Signaling Technology (Danvers, MA). Akt1/2 antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal antibody against α1C was from Calbiochem (La Jolla, CA). Insulin, taxol, and antibody against actin were from Sigma. Phosphoinositides (all di-C8) were from Echelon Biosciences (Salt Lake City, UT). The glucometer was from Home Diagnostics (Ft. Lauderdale, FL).

M-mode echocardiography.

Heterozygous C57BL/6-Ins2Akita/J (Akita) mice were purchased from The Jackson Laboratories and bred to wild-type C57BL/6 mice in our institutional animal facility to generate Akita and wild-type littermates. All animal-related experimental protocols were approved by the Institutional Animal Care and Use Committee. Transthoracic echocardiography was performed on anesthetized animals (1–2% isoflurane) using a Vevo 770 ultrasound device (VisualSonics, Toronto, Canada) with a 30-MHz transducer. The Vevo Measurement and Calculations software package was used to generate left ventricular measurements and calculations derived from the tracings of the collected echocardiographic data. Parameters were determined using the leading-edge method of the American Society of Echocardiography. The following parameters were obtained using M-mode transthoracic views: left ventricular end-systolic diameter (LVESD), left ventricular end-diastolic diameter (LVEDD), interventricular septal wall thickness, and left ventricular posterior wall thickness. Contractile function was calculated as fractional shortening = (LVEDD − LVESD)/LVEDD × 100%.

Ventricular myocyte isolation and electrophysiology.

Mice were killed by intraperitoneal injection of 100 mg/kg body wt sodium pentobarbital, and ventricular myocytes were isolated as previously described (8). The preparation of phosphoinositides, the solutions used, and the method of whole-cell patch clamping of isolated cardiac myocytes were as previously described (8). Myocytes isolated from at least three different animals in each group were analyzed in each experiment.

Akt kinase assays and Western blots.

After the animals were killed, hearts were rapidly removed, rinsed in three changes of saline, and frozen in liquid nitrogen. Frozen hearts were subsequently powdered in a mortar and pestle cooled by liquid nitrogen, and aliquots were stored in liquid nitrogen. Frozen heart powder was homogenized and centrifuged to make tissue lysates (16). The protein concentration was measured by a Bradford-based assay (Bio-Rad Protein Assay). Akt activity was assayed in immunoprecipitates following a method described earlier using 0.5 mg tissue lysate protein (16). Western blot signals were visualized using horseradish peroxidase–linked secondary antibodies (Amersham Biosciences) and a chemiluminescence kit (PerkinElmer Life Sciences).

Cell surface biotinylation of the LTCC.

To assess cell surface expression of the LTCC, we used a protein biotinylation technique that is based on an assay used to assess dopamine transporter trafficking (17). Isolated myocytes were washed three times on ice with K-reversal Tyrode buffer (140 mmol/l KCl, 8 mmol/l KHCO3, 2 mmol/l MgCl2, and 0.4 mmol/l KH2PO4, pH 8.0). Equal numbers of suspended cells (as determined by protein assay on separate aliquots) were then incubated on ice for 2 h with 20 mmol/l EZ-Link Sulfo-NHS-LC-Biotin (Pierce Chemicals, Rockford, IL) to biotinylate cell surface proteins, including the LTCC α1C-subunit. After the reaction was quenched with 100 mmol/l glycine, cells were lysed with 1% Triton X-100 on ice. Equal amounts of extract protein were incubated with neutrAvidin agarose beads (Pierce Chemicals) to pull down biotinylated proteins. α1C in the biotinylated fraction was detected by Western blotting. The total amount of α1C in the mouse hearts was assessed by Western blot analysis of heart membranes prepared from frozen heart powder as previously described (8).

Data analysis.

Unless otherwise specified, data were analyzed by one-way ANOVA, and pairwise comparisons were made using Tukey's post hoc tests. A value of P < 0.05 was considered significant. The OriginPro 7.5 statistical software (OriginLab, Northhampton, MA) was used to perform the analyses.

RESULTS

Reduced cardiac contractility in diabetic mice.

In this study, we used Akita mice 3–4 months of age that were frankly diabetic (random tail vein blood glucose levels >500 mg/dl) paired with sex-matched control mice from the same litter that exhibited normal blood glucose levels (mean ± SD of 149 ± 20 mg/dl, n = 24). Echocardiography showed that most of the parameters examined were significantly altered in the diabetic animals (Table 1). In particular, fractional shortening was reduced by 45% in the diabetic compared with control animals (Table 1).

TABLE 1

Echocardiographic measurements in nondiabetic and diabetic mice

Reduced PI 3-kinase signaling in diabetic mouse hearts.

We have shown that insulin activates PI 3-kinase/Akt signaling in the heart of wild-type C57BL/6 mice (9), and others have shown that cardiac Akt activity is regulated by food intake (18). Because diabetic Akita mice produce little or no insulin, we expected that PI 3-kinase/Akt signaling in these animals would be lower than that in nondiabetic littermates, especially after feeding. Akt activity was significantly lower in heart lysates of fed diabetic mice compared with that in control animals (Fig. 1A). The average Akt activity in hearts from fasted diabetic mice also tended to be lower than that in the control samples, but the difference was not statistically significant. As expected, insulin injection strongly activated Akt to a similar degree in both groups of animals (Fig. 1A). Upon binding to PI(3,4,5)P3, Akt is phosphorylated on Thr308 and Ser473, which leads to activation of the enzyme. Consistent with results from the activity assays, we found that Akt in diabetic hearts was less phosphorylated than that in control hearts as assessed by Western blotting using an antibody that recognizes Akt phosphorylated on Thr308 (Fig. 1B). Total Akt protein in the heart was not significantly different between the two groups (Fig. 1B).

FIG. 1.

Akt activity in heart lysates. A: Animals were fasted overnight or fed ad libitum and killed in the morning. The insulin-injected animals were fasted overnight, injected intraperitoneally with insulin (1 unit/kg body wt), and killed 5 min later. Akt kinase activity was assayed in heart lysates. Values are means ± SE; n = 6 for each group. **Significantly different from nondiabetic ad libitum. B: Equal amounts of heart lysate protein from two ad libitum–fed nondiabetic and diabetic mice were analyzed by Western blotting using a phospho-Thr308 Akt antibody. The blot was stripped and reprobed with a general Akt antibody.

Reduced ICa,L density in diabetic myocytes.

We previously used three different approaches to show that inhibition of PI 3-kinase signaling leads to decreased ICa,L density in cardiac myocytes: 1) Gαq was activated in the hearts of transgenic animals to directly inhibit PI 3-kinase, 2) cardiac myocytes from wild-type mice were infused with the lipid phosphatase PTEN to reduce PI(3,4,5)P3 levels, and 3) wild-type myocytes were infused with a peptide that sequesters PI(3,4,5)P3 (9). Because the diabetic Akita mice also have decreased PI 3-kinase signaling, we investigated whether ICa,L density is decreased in cardiac myocytes isolated from these animals. ICa,L was measured by whole-cell patch clamping in myocytes isolated from diabetic and nondiabetic mice. The left panels of Fig. 2A show typical recordings of ICa,L activation, which was elicited by depolarizing voltage pulses from −50 to 50 mV in 10-mV increments (300-ms duration) from a holding potential of −50 mV. The membrane capacitance was measured by applying a voltage step from a holding potential of −50 mV, and the current amplitudes were normalized to cell capacitance to obtain current densities in picoamps (pA)/picofarads (pF). Membrane capacitance, which is proportional to surface area, was not significantly different between the two groups of myocytes (141.2 ± 6 pF [nondiabetic, n = 31] and 141.8 ± 9.5 pF [diabetic, n = 45]). The peak inward current densities from these records were used to construct current density–voltage (I-V) relationships that show a reduction in ICa,L density in diabetic myocytes compared with control cells at most of the voltages tested (Fig. 2B, left). The ratio of ICa,L density in diabetic to nondiabetic cells was not the same at all voltages. The ratios were smaller at voltages between −30 and 0 mV than those at voltages between 10 and 30 mV. This suggests that there is a difference in the voltage dependence of activation of the LTCC in the two cell types (see below). For statistical comparisons, the peak ICa,L density was measured at 10 mV after a single depolarizing step of 300-ms duration from a holding potential of −50 mV. The peak ICa,L density at 10 mV was reduced by 38% in the diabetic myocytes compared with that in nondiabetic controls (Fig. 3A).

FIG. 2.

ICa,L activation in myocytes infused with or without PI(3,4,5)P3. A: Representative recordings of ICa,L activation in myocytes from nondiabetic (top) or diabetic (bottom) mice infused with internal solution without (left panels) or with (right panels) 1 μmol/l PI(3,4,5)P3. B: I-V relationships in the absence (left) or presence (right) of 1 μmol/l PI(3,4,5)P3 were constructed by plotting the mean peak inward current density for each dataset at each voltage potential. n = 12 cells for all groups except nondiabetic plus PI(3,4,5)P3, for which n = 10.

FIG. 3.

Effect of PI(3,4,5)P3 infusion on ICa,L density in myocytes. A: Peak ICa,L densities at 10 mV were measured by whole-cell patch clamping in myocytes isolated from nondiabetic and diabetic mice. Internal solution with or without phosphoinositides (1 μmol/l) was infused through the patch pipette. The values are means ± SE. The number of cells examined in each group is indicated in parentheses. *Significantly different from nondiabetic in the absence of PI(3,4,5)P3. **Significantly different from diabetic in the absence of PI(3,4,5)P3. B: Time dependence of ICa,L activation in diabetic myocytes infused with internal solution without (left) or with (right) 1 μmol/l PI(3,4,5)P3. Values are average ICa,L densities normalized to the value of the first ICa,L recording, which is taken immediately after breaking into the whole-cell mode. Control, n = 8; PI(3,4,5)P3, n = 10.

PI(3,4,5)P3 increases ICa,L density in diabetic myocytes.

To determine whether decreased PI 3-kinase signaling in diabetic Akita hearts might be responsible for this difference in ICa,L density, we tested whether PI(3,4,5)P3 infusion can restore the ICa,L density to control levels. The right panels of Fig. 2A show recordings of ICa,L activation in nondiabetic and diabetic myocytes infused with 1 μmol/l PI(3,4,5)P3 through the patch pipette, and the right panel in Fig. 2B shows the corresponding I-V curves. There was an enhancement of ICa,L density by PI(3,4,5)P3 in the diabetic myocytes at most of the voltages tested (compare closed circles in Fig. 2B, right and left). As we previously reported (9), PI(3,4,5)P3 had little or no effect on ICa,L density in nondiabetic myocytes (compare open circles in Fig. 2B, right and left). In the presence of PI(3,4,5)P3, the current density in both cell types was nearly identical at voltages >0 mV, but at negative voltages, the current density tended to be lower in diabetic cells than that in nondiabetic cells (Fig. 2B, right). This remaining difference is due to changes in the steady-state activation of the Ca2+ channel (discussed later).

Although the peak ICa,L density at 10 mV in diabetic myocytes in the absence of PI(3,4,5)P3 was significantly lower than in nondiabetic cells, in the presence of PI(3,4,5)P3, ICa,L density in the diabetic myocytes increased to a value that was statistically indistinguishable from that observed in nondiabetic myocytes (Fig. 3A). By contrast, infusion with PI(3,4,5)P3 had no effect on ICa,L density in control myocytes (Fig. 3A). We performed additional experiments in diabetic myocytes using two other phosphoinositides to show that the stimulatory effect is specific to PI(3,4,5)P3. Infusion with PI(4,5)P2 or phosphatidylinositol 3,5-bisphosphate, which is generated from phosphatidylinositol 3-phosphate by a phosphatidylinositol 5-kinase, did not significantly affect the ICa,L density in diabetic myocytes (Fig. 3A). These results are consistent with the hypothesis that decreased ICa,L density in diabetic Akita myocytes is due to reduced PI 3-kinase activity.

To further characterize the action of PI(3,4,5)P3 in diabetic myocytes, we examined its time-dependent effect on ICa,L activation in single cells. In this protocol, myocytes were infused with or without 1 μmol/l PI(3,4,5)P3, and repeatedly depolarized every 12 s with voltage steps 300 ms in duration to 10 mV from a holding potential of −50 mV. The ICa,L densities were normalized to the value obtained from the first voltage step after opening of the patch and initiation of whole-cell recording. The typical “run-down” of ICa,L density was observed in diabetic myocytes when the patch pipette contained the control internal solution (Fig. 3B, left). After 240 s, the normalized ICa,L density in these cells decreased by ∼15%. In contrast, there was a “run-up” of ICa,L density in the diabetic myocytes in the presence of PI(3,4,5)P3 (Fig. 3B, right). ICa,L density then started to decrease after reaching a maximum after ∼130 s. These results indicate that ICa,L density in diabetic myocytes is restored to control levels in a relatively short period of time when PI 3-kinase signaling is artificially increased by intracellular infusion with PI(3,4,5)P3.

Insulin treatment of diabetic mice increases ICa,L density in myocytes.

To determine whether PI(3,4,5)P3 infusion recapitulates the physiological increase in PI(3,4,5)P3 levels that accompanies insulin stimulation, we next examined whether acute insulin injection of diabetic mice corrects the ICa,L density deficit in myocytes. Myocytes isolated from insulin-treated versus untreated diabetic animals showed a significant increase in peak ICa,L density at 10 mV (Fig. 4A). Interestingly, ICa,L density in myocytes isolated from nondiabetic mice was not changed by insulin treatment (Fig. 4A). I-V relationships for ICa,L activation in myocytes from insulin-treated diabetic versus nondiabetic mice were nearly indistinguishable (Fig. 4B, right).

FIG. 4.

Effect of insulin treatment of mice on ICa,L density in myocytes. Nondiabetic and diabetic mice were injected with insulin (10 units/kg body wt) intraperitoneally and killed 10 min later or left untreated as control. A: Myocytes were isolated, and peak ICa,L densities were measured at 10 mV by whole-cell patch clamping. *Significantly different from control nondiabetic. **Significantly different from control diabetic. The membrane current was evoked by a 300-ms depolarizing pulse to 10 mV from a holding potential of −50 mV. The values in parentheses show the number of cells analyzed. B: I-V relationships for myocytes from nondiabetic and diabetic mice without (left) or with (right) insulin treatment. n = 20 cells for all groups except control diabetic, for which n = 14.

The effect of PI(3,4,5)P3 on steady-state ICa,L activation and inactivation.

The decrease in ICa,L density in myocytes from diabetic mice could be due to any combination of altered voltage dependence, decreased cell surface expression of the LTCC, or altered single channel conductance. We examined voltage dependence by first analyzing the steady-state activation of ICa,L. Cells were exposed to depolarizing voltage steps from −50 to 10 mV (300-ms duration) from a holding potential of −50 mV, and the data were fitted by a Boltzmann two-state model to produce steady-state activation curves. Without PI(3,4,5)P3, the steady-state activation curve for diabetic myocytes was shifted toward more positive potentials compared with the curve for nondiabetic myocytes (Fig. 5A). The membrane potentials at which 50% of the channels are activated (Vh) were significantly different (−18.2 ± 0.3 [nondiabetic] and −12 ± 1.1 mV [diabetic]). The slope factors were not significantly different (3.9 ± 0.4 [nondiabetic] and 5.1 ± 0.5 mV [diabetic]). This difference in Vh probably explains why the ratio of ICa,L density in diabetic to nondiabetic cells is smaller at negative potentials than positive potentials in Fig. 2B (left). Infusion of 1 μmol/l PI(3,4,5)P3 caused a slight shift (not statistically significant) of the steady-state activation curves to more negative potentials in both nondiabetic and diabetic cells (Fig. 5B). More importantly, Vh values were still significantly different between the two groups in the presence of PI(3,4,5)P3 (−21 ± 1.2 [nondiabetic] and −15.4 ± 1.5 mV [diabetic]). The slope factors were not significantly different (4.6 ± 0.9 [nondiabetic] and 4.8 ± 0.5 mV [diabetic]). This difference in activation in the presence of PI(3,4,5)P3 probably explains the reduced peak ICa,L density observed in diabetic myocytes at negative potentials (Fig. 2B, right). This result also indicates that reduced PI 3-kinase signaling does not fully account for the alteration in voltage dependence of LTCC activation in diabetic myocytes. It is possible that this change is secondary to the hyperglycemia in Akita animals.

FIG. 5.

Steady-state activation and inactivation of ICa,L in myocytes. A and B: Steady-state activation curves of ICa,L for nondiabetic (n = 10) and diabetic (n = 12) cells infused with internal solution without (control) or with PI(3,4,5)P3. C and D: Steady-state inactivation curves of ICa,L for nondiabetic and diabetic cells infused with internal solution without (control) (n = 10 [nondiabetic] and 12 [diabetic]) or with (n = 7 [nondiabetic] and 12 [diabetic]) PI(3,4,5)P3. I/Imax is peak inward current at each voltage normalized to the maximum current. The inset shows the voltage protocol used to generate the steady-state inactivation curves.

To investigate the steady-state inactivation of ICa,L, conditioning pulses (2-s duration) were performed from −80 to 10 mV in 10-mV increments from a holding potential of −50 mV followed by a test step to 10 mV. The data were fitted by a Boltzmann two-state model to produce steady-state inactivation curves. In the absence of PI(3,4,5)P3, the membrane potentials at which 50% of the channels are inactivated (V1/2) were significantly different (−31.5 ± 0.6 [nondiabetic] and −22.8 ± 0.6 mV [diabetic]) (Fig. 5C). The slope factors were not significantly different (5.2 ± 0.5 [nondiabetic] and 5.1 ± 0.5 mV [diabetic]). In contrast, the inactivation curves for the two cell types were very similar in the presence of PI(3,4,5)P3 (Fig. 5D). V1/2 values were −31.6 ± 0.5 (nondiabetic) and −29.1 ± 0.4 mV (diabetic), and the slope factors were 5.3 ± 0.4 (nondiabetic) and 5.9 ± 0.3 mV (diabetic). The differences between these values were not statistically significant. Moreover, the difference between V1/2 values plus and minus PI(3,4,5)P3 in the diabetic myocytes was significant, whereas PI(3,4,5)P3 did not alter V1/2 in the nondiabetic myocytes. These results indicate that reduced PI 3-kinase signaling is responsible for the changes in voltage dependence of LTCC inactivation in diabetic myocytes.

We also examined the kinetics of inactivation of ICa,L at different voltages in the presence or absence of PI(3,4,5)P3. The average time constant (τ) of ICa,L inactivation at different test voltages was fitted by a single exponential. We did not detect statistically significant differences in τ between nondiabetic or diabetic myocytes with or without PI(3,4,5)P3 (data not shown). We then used a two-pulse protocol to investigate the kinetics of recovery from inactivation of ICa,L in both cell types. First, a depolarizing pulse 300 ms in duration was applied from the holding potential of −50 mV to the test potential of 10 mV to activate ICa,L. A variable interval was allowed for recovery, and then a second identical test pulse was applied. The cycle length was 8 s. The data were fitted with the equation: I/Imax = 1 − e(−t/τ) to calculate τ. The difference in τ between nondiabetic (330 ± 39 ms, n = 7) and diabetic (340 ± 36 ms, n = 5) myocytes was not statistically significant. These results, together with the steady-state activation and inactivation data, indicate that the reduction in peak ICa,L density at positive voltages in diabetic myocytes (Fig. 2B, left) is not due to altered gating.

Reduced cell surface expression of the LTCC in diabetic myocytes.

Another possible explanation for the lower ICa,L density in diabetic myocytes is that they express fewer LTCCs at the cell surface than nondiabetic cells. To test this hypothesis, we used a protein biotinylation technique to assess the level of LTCC α1C-subunits present on the cell surface. The amount of biotinylated α1C was markedly reduced in the diabetic versus nondiabetic myocytes (Fig. 6A), but the total level of α1C in the heart was not appreciably different between the two groups (Fig. 6B).

FIG. 6.

Reduced cell surface expression of the LTCC in Akita myocytes. A: Cell surface proteins of cardiac myocytes isolated from two ad libitum–fed nondiabetic and diabetic mice were biotinylated, followed by detection of α1C in the biotinylated fraction by Western blotting. B: Total α1C in membranes prepared from hearts of ad libitum–fed nondiabetic and diabetic mice (n = 3 in each group) was detected by immunoblotting. The blot was stripped and reprobed with an actin antibody.

PI(3,4,5)P3 and insulin stimulation of ICa,L is blocked by taxol treatment.

It was reported that activation of PI 3-kinase stimulates ICa,L density through Akt-mediated phosphorylation of β2a-subunits, leading to increased trafficking of exogenously expressed LTCC to the plasma membrane in COS7 cells (13). We hypothesized that PI(3,4,5)P3 infusion stimulates trafficking of the LTCC to the cell surface of diabetic cardiac myocytes, thus increasing ICa,L density. To test this hypothesis, we used taxol to inhibit microtubule-dependent LTCC trafficking. Taxol stabilizes microtubles, thus blocking the dynamic process of polymerization and depolymerization needed for trafficking of macromolecules within the cell. Interestingly, Gomez and colleagues (19,20) reported that microtubule disruption with colchicine but not taxol affects ICa,L in rat cardiac myocytes. We also found that colchicine, which causes microtubule depolymerization, increased ICa,L density and blocked isoproterenol stimulation of ICa,L in mouse myocytes (data not shown).

Diabetic myocytes were pretreated with 10 μmol/l taxol or vehicle for 1 h, and then the time-dependent effect of PI(3,4,5)P3 on ICa,L activation was examined. Taxol pretreatment completely blocked the PI(3,4,5)P3-stimulated “run-up” in ICa,L density (Fig. 7A, right; compare with Fig. 3B). Surprisingly, taxol also blocked the “run-down” in current density in the control cells (Fig. 7A, left). Taxol treatment did not affect the magnitude of ICa,L density at 10 mV but completely blocked the PI(3,4,5)P3-induced increases (Fig. 7B). In contrast, taxol did not affect the β-adrenergic receptor–mediated increase in ICa,L density in response to isoproterenol in diabetic myocytes (Fig. 7B). We also tested whether treatment with insulin in vitro corrects the ICa,L density deficit in the diabetic myocytes. Similar to the in vivo effect of insulin, we found that insulin incubation increased the peak ICa,L density by 61% (Fig. 7B). This response was also blocked in the presence of taxol (Fig. 7B). These results indicate that the LTCC is still functional in the presence of taxol and suggest that drug treatment blocks PI(3,4,5)P3-stimulated trafficking of the LTCC to the cell surface.

FIG. 7.

Taxol inhibits PI(3,4,5)P3 stimulation of ICa,L. A: Time dependence of ICa,L activation in diabetic myocytes infused with internal solution without (control) or with 1 μmol/l PI(3,4,5)P3. Cells were incubated with 10 μmol/l taxol at room temperature for 1 h before measuring ICa,L density as described for Fig. 3B. Values are average ± SE. ICa,L densities normalized to the value of the first ICa,L recording, which was taken immediately after breaking into the whole-cell mode. Control, n = 20; PI(3,4,5)P3, n = 8. B: Peak ICa,L densities at 10 mV were measured in diabetic myocytes as described in Fig. 3A. Cells were incubated with or without 10 μmol/l taxol for 1 h at room temperature before perfusion with 2 μmol/l isoproterenol (ISO) or incubation for 1 h with 1 μmol/l insulin in the external solution or infusion with 1 μmol/l PI(3,4,5)P3 in the internal solution. The values are means ± SE. The number of cells examined in each group is indicated in parentheses. *Significantly different from control. **Significantly different from insulin minus taxol.

DISCUSSION

Results from these studies support the hypothesis that PI 3-kinase signaling plays an important role in regulating the cardiac LTCC. We previously demonstrated that Gαq inhibition of PI 3-kinase in myocytes from transgenic mice decreased ICa,L density, which was reversed upon infusion of PI(3,4,5)P3 or PI 3-kinase proteins (9). We now show that myocytes from insulin-deficient diabetic mice exhibit a reduction in ICa,L density that is reversible upon intracellular infusion of PI(3,4,5)P3 and by injecting the animal or incubating the myocytes with insulin. These results suggest that the ICa,L defect is due to a lack of insulin-activated PI 3-kinase signaling and not to cell damage caused by chronic hyperglycemia. These results are consistent with studies showing that ICa,L density was decreased in cardiac myocytes from streptozotocin-induced insulin-deficient and obese insulin-resistant animals (25). We speculate that reduced ICa,L density in insulin-resistant db/db myocytes is also due to decreased PI 3-kinase signaling. Future studies using isoform-specific PI 3-kinase knockout myocytes will directly address the role of these lipid kinases in regulating the LTCC.

Our echographic analysis reveals a large decrease in cardiac contractility in diabetic Akita mice (Table 1). Using various techniques to evaluate cardiac contractile function, some investigators also found depressed contractility in murine models of diabetes, while others found only minimal or mildly decreased systolic function (5,2125). The wide variability in the magnitude of the reduction in fractional shortening or ejection fraction reported in these studies may be related to the duration of diabetes or age. Our study of the Akita mouse is limited by a small sample size. A more detailed study of cardiac function will be necessary in the future to completely characterize the nature of the contractile defect in the Akita mouse.

The decreased ICa,L density probably contributes to the contractility defect observed in diabetic Akita mice. Further studies are needed to address this hypothesis because diabetes has complex effects on Ca2+ homeostasis, as seen in db/db myocytes (5). Future studies using genetic mouse models that selectively ablate the insulin receptor or PI 3-kinase in cardiac myocytes will help clarify how this signaling pathway regulates the LTCC and its role in the development of contractile defects in diabetes. For example, it would be informative to determine whether ICa,L density is reduced in myocytes of cardiac-selective insulin receptor knockout mice, which exhibit a contractile defect but not diabetes (26). We recently generated cardiac-selective PI 3-kinase knockout mice and observed that myocytes isolated from these animals had decreased ICa,L density that was reversed by PI(3,4,5)P3 infusion (data not shown). By contrast, it was previously reported that transgenic mice expressing a dominant-negative PI 3-kinase in the heart had normal ICa,L density (10) and fractional shortening by echocardiography (27). It will be important to determine whether the reduction in ICa,L density is sufficient to cause a contractile defect in the PI 3-kinase knockout mouse or whether adaptations might preserve cardiac contractility.

This information may have important therapeutic considerations. Overexpression of Akt in adult rat hearts using an adenoviral gene transfer method enhanced myocyte contractility (28). Conceivably, strategies that activate the PI 3-kinase/Akt pathway can be developed to enhance the contractility of a failing heart. Interestingly, contractility of cardiac myocytes isolated from failing human myocardium was improved by treatment with IGF-I, and this effect was blocked by wortmannin, a PI 3-kinase inhibitor (29). Acute intravenous administration of IGF-I to patients with chronic heart failure improved cardiac index by 27% and stroke volume index by 21% (30). Studies have shown that infusion with supraphysiological concentrations of insulin and glucose to maintain euglycemia increases left ventricular ejection fraction in diabetic patients (31). Glucose-insulin-potassium infusion is being evaluated in clinical trials as a treatment for patients suffering from acute myocardial infarction or who are critically ill in an intensive care unit. Based on our results, we speculate that insulin treatment in these patients will stimulate LTCC function that may lead to enhanced cardiac performance.

It has been reported that insulin treatment of rat ventricular and human atrial cardiac myocytes enhances ICa,L (32,33). One group also showed that IGF-I treatment of mouse cardiac myocytes increased ICa,L density by ∼15% (10). We have not observed an increase in ICa,L in response to insulin treatment or infusion with PI(3,4,5)P3 in myocytes isolated from wild-type mice (this study) (9). This divergence could be due to technical differences in the patch-clamping procedure. It is also possible that mouse cardiac myocytes have a low intracellular reserve of the LTCC. Therefore, insulin treatment would stimulate a small or no net movement of the LTCC to the cell membrane in normal myocytes. In contrast, our data suggest that diabetic Akita myocytes accumulate intracellular LTCC that moves to the cell surface in response to insulin or PI(3,4,5)P3, thus increasing ICa,L density.

Intracellular infusion of PI(3,4,5)P3 had a rapid effect on ICa,L density (Fig. 3B, right). We also found that nondiabetic and diabetic myocytes contained similar amounts of LTCC protein, but the amount of the protein on the cell surface was reduced in diabetic myocytes (Fig. 6). Although we have not directly demonstrated that insulin treatment led to an increase in cell surface LTCC protein, the inhibitory effect of taxol treatment on the stimulatory effect of PI(3,4,5)P3 or insulin on ICa,L density (Fig. 7) is consistent with the hypothesis that impaired LTCC trafficking contributes to the reduction in ICa,L density in diabetic myocytes. Our results are also consistent with a model proposed by Viard et al. (13) in which PI(3,4,5)P3 activation of Akt results in phosphorylation of the LTCC β2a-subunit, which leads to increased trafficking of the channels to the cell surface. It was reported that the neuronal N-type Ca2+ channel rapidly internalized and reappeared on the cell surface after GABA receptor stimulation within 80 s (34). These investigators postulated that the channels are preassociated in arrestin-containing cellular complexes that allow rapid trafficking of the protein. Additional studies are needed to further elucidate how trafficking of the cardiac LTCC is regulated by PI 3-kinase signaling.

One observation of note was that taxol eliminated the “run-down” of the LTCC, which is normally observed after breaking the patch and achieving the whole-cell patch-clamp configuration with its subsequent dialysis of intracellular constituents (compare Fig. 3B with Fig. 7A). This result suggests the intriguing possibility that “run-down” might at least in part be the consequence of net channel trafficking out of the cell surface membrane after loss of relevant intracellular factors.

Our study indicates that the biophysical properties of the LTCC are also altered in the diabetic Akita myocytes and that reduced PI 3-kinase signaling is responsible for some but not all of the biophysical changes. The voltage dependence of channel activation and inactivation were both shifted toward more positive potentials in the diabetic myocytes, but PI(3,4,5)P3 normalized the defect in inactivation only (Fig. 5). The change in activation is particularly important because it suggests that a smaller fraction of the LTCCs present in the membrane are likely to open during an action potential. This will reduce calcium entry, and in the absence of other compensatory effects, reduce contraction. Additional studies are needed to determine how PI(3,4,5)P3 or a downstream effector such as Akt modulates the biophysical properties of the LTCC.

In conclusion, we have shown that reduced PI 3-kinase signaling in insulin-deficient mice leads to decreased Ca2+ entry through the LTCC in cardiac myocytes. The reduction in the magnitude of ICa,L density is secondary to a lower number of Ca2+ channels at the surface of diabetic myocytes. This negative effect likely contributes to the depressed cardiac contractility associated with this animal model of diabetes.

Acknowledgments

R.Z.L. has received grants from the American Heart Association, the Juvenile Diabetes Research Foundation, the Department of Veterans Affairs, and the National Institutes of Health (DK-62722). I.S.C. has received grant support from the National Institutes of Health (HL-70161, HL-28958, HL-85221, and HL-67101).

We greatly appreciate the expert technical assistance of Joan Zuckerman. We thank Dr. Irwin Kurland for his helpful suggestions regarding the Akita mice.

Footnotes

  • Published ahead of print at http://diabetes.diabetesjournals.org on 31 July 2007. DOI: 10.2337/db06-1629.

    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.

    • Received November 21, 2006.
    • Accepted July 26, 2007.

REFERENCES

| Table of Contents

This Article

  1. Diabetes vol. 56 no. 11 2780-2789
  1. All Versions of this Article:
    1. db06-1629v1
    2. 56/11/2780 most recent