α₁-Adrenoceptor stimulation potentiates L-type Ca²⁺ current through Ca²⁺/calmodulin-dependent PK II (CaMKII) activation in rat ventricular myocytes

Results

Effect of 10 μM Phenylephrine on I_Ca,L Measured by Using the Perforated Patch–Clamp Technique. Fig. 1 shows a representative result of the effect of 10 μM phenylephrine on I_Ca,L. The effect of phenylephrine on the Ca²⁺ responsiveness of the contractile element in rat ventricular myocytes is almost saturated at this concentration (5). As reported (8, 9), we observed a transient decrease followed by a sustained increase of I_Ca,L amplitude. As shown in Fig. 1 A,I_Ca,L transiently decreased for up to 2 min after the application of phenylephrine, and then it gradually increased and reached another steady-state level at ≈15 min after this application. The amplitudes of I_Ca,L at 15 and 20 min were not statistically different (114.8 ± 14.5% of control and 117.9 ± 13.5% of control, n = 12, P = 1.00), and thus, the effect of 10 μM phenylephrine on I_Ca,L potentiation reached a steady state at 15 min. The amplitude of I_Ca,L returned to the control level at 10 min after removal of phenylephrine (101.4 ± 6.2% of control, n = 5, P = 0.678). The initial transient decrease in I_Ca,L at 2 min (91.4 ± 6.0% of control, n = 12, P < 0.05) and the later increase at 15 min (114.8 ± 14.5% of control, n = 12, P < 0.001) (Fig. 1C) were small but significant. The increase in I_Ca,L occurred without changes in the shape of the current-voltage relationship (Fig. 1B). These negative and positive effects of phenylephrine on I_Ca,L were both completely blocked by the α₁-adrenoceptor antagonist prazosin (1 μM) (n = 12, data not shown). In the absence of phenylephrine, the amplitude of I_Ca,L was stable for up to 15 min (Fig. 1C).

We explored the effect of intracellular Ca²⁺ buffering on I_Ca,L in response to 10 μM phenylephrine (Fig. 1D). After establishing perforated patch, a cell was incubated with 20 μM BAPTA–acetoxymethyl ester, the cell permeable form of BAPTA, for 10 min and we confirmed that the cell did not contract during the pulse to evoke I_Ca,L (16). In this condition, phenylephrine did not show either a significant transient decrease in I_Ca,L at 2 min (97.2 ± 4.4% of control, n = 10, P = 1.00) or a sustained increase at 15 min (101.3 ± 17.1% of control, n = 10, P = 1.00). Thus, intracellular Ca²⁺ is a key factor for the potentiation in I_Ca,L induced by phenylephrine. The different responses of I_Ca,L to phenylephrine recorded by conventional patch and perforated patch can be explained by the intracellular Ca²⁺ concentration.

The Role of CaMKII in the Regulation of I_Ca,L by α₁ARS. Recent studies have reported that CaMKII is involved in various Ca²⁺-dependent effects both under physiological and pathophysiological conditions in mammalian ventricular myocytes (10). Thus, we investigated the role of CaMKII in the regulation of I_Ca,L by α₁ARS by using KN-93, a synthetic CaMKII inhibitor. At a concentration of 0.5 μM, KN-93 selectively inhibits CaMKII without affecting other PKs (21). A 10-min exposure to 0.5 μM KN-93 significantly decreased I_Ca,L from 9.12 ± 3.27 to 5.50 ± 3.39 pA/pF, corresponding to 39.1 ± 25.6% decrease (n = 12, P < 0.001) without changing the shape of the current–voltage relationship (n = 10) (see also ref. 22). In the presence of KN-93, 10 μM phenylephrine produced only a sustained decrease of I_Ca,L (Fig. 2A). At 15 min after application of 10 μM phenylephrine, the amplitude of I_Ca,L significantly decreased to 66.7 ± 22.0% of the value before application of phenylephrine in the presence of KN-93 (n = 12, P < 0.001) without changing the shape of the current-voltage relationship (n = 10). In contrast, 0.5 μM KN-92, an inactive KN-93 analogue, did not show significant effects on the biphasic change of I_Ca,L caused by α₁ARS (Fig. 2 A). We also investigated the effect of 10 μM phenylephrine in the presence of another CaMKII inhibitor, autocamtide-2 inhibitory peptide, a membrane-permeable and a highly specific peptide type inhibitor of CaMKII (23). When we used 10 μM autocamtide-2 inhibitory peptide, 10 μM phenylephrine produced only a sustained decrease of I_Ca,L without potentiation as in the presence of KN-93 (n = 5, data not shown). Thus, CaMKII inhibition abolished the potentiation of I_Ca,L during α₁ARS.

The Role of PKC in the Regulation of I_Ca,L During α₁ARS. PKC may be involved in the α₁-adrenoceptor-mediated modulation of cardiac K⁺ channels, intracellular alkalinization and myofibrillar responsiveness to Ca²⁺ (2). Therefore, we investigated the role of PKC in the regulation of I_Ca,L by α₁ARS by using chelerythrine as a PKC inhibitor. We found that 10 μM chelerythrine selectively inhibits PKC without affecting other PKs (24) and effectively blocks the effect of phenylephrine on the Ca²⁺ responsiveness of the contractile element (5). Exposure to 10 μM chelerythrine for 10 min significantly decreased I_Ca,L to a new steady state from 7.42 ± 1.61 to 3.78 ± 1.23 pA/pF, corresponding to a 49.7 ± 13.2% decrease (n = 9, P < 0.001) without changing the shape of the current–voltage relationship. In the presence of chelerythrine, 10 μM phenylephrine produced only a sustained decrease of I_Ca,L as observed in the presence of CaMKII inhibitors (Fig. 2B). At 15 min after application of 10 μM phenylephrine, the amplitude of I_Ca,L decreased to 42.3 ± 10.9% of the value before application of phenylephrine in the presence of chelerythrine (n = 9, P < 0.001) (Fig. 2B) without changing the shape of the current–voltage relationship (n = 9).

Activation of CaMKII During α₁ARS. Our electrophysiological experiments suggest that potentiation of I_Ca,L in response to α₁ARS is mediated by both CaMKII and PKC. Although PKC has been suggested to be involved in the α₁-adrenoceptor signal-transduction pathway (2), there are no reports that CaMKII is activated by α₁ARS in rat ventricular myocytes. The amount of CaMKII is low compared with other PKs; consequently, it is difficult to determine CaMKII activity directly in cardiac muscle (25). To overcome this difficulty, we used an antibody against the autophosphorylation site of CaMKII, Thr-286, and an enhanced chemiluminescence system to measure CaMKII activity in whole-cell lysates in response to α₁ARS (12, 13). Active CaMKII increased significantly at concentrations of ≥1 μM phenylephrine (Fig. 3A Bottom), although total CaMKII was not altered (Fig. 3A Top; data not shown). The increase of active CaMKII level induced by 10 μM phenylephrine was completely blocked by the α₁-adrenoceptor antagonist prazosin (1 μM), showing that this effect was mediated by α₁-adrenoceptor (Fig. 3B). The selective CaMKII inhibitor 0.5 μM KN-93 that we used in the perforated patch experiments partially blocked the basal CaMKII activity in the absence of phenylephrine and also completely blocked the increase in the active CaMKII level by 10 μM phenylephrine (Fig. 3C).

To determine whether PKC could be a regulator of CaMKII activation during α₁ARS, the effect of phenylephrine in the presence of chelerythrine on CaMKII activity was determined. Chelerythrine (10 μM) did not produce significant changes in the CaMKII activation before phenylephrine (86.6 ± 25.1% of control, n = 7, P = 0.39). However, the increase in active CaMKII by 10 μM phenylephrine was abolished by chelerythrine (102.2 ± 24.3% of control, n = 7, P = 1.00), as in the case of KN-93, indicating that PKC is involved in the activation of CaMKII during α₁ARS (Fig. 3D).

Immunocytochemical Localization of Active CaMKII by α₁ARS. It has been reported that CaMKII may directly phosphorylate the L-type Ca²⁺ channel and regulate the positive feedback system that facilitates I_Ca,L under physiological conditions (17, 26). Our biochemical experiments showed that CaMKII is significantly activated by α₁ARS in adult rat isolated ventricular myocytes. To investigate whether CaMKII could directly regulate the L-type Ca²⁺ channel during α₁ARS, we determined the localization of active CaMKII in isolated ventricular myocytes by using immunofluorescence microscopy, as shown in Fig. 4. We used WGA-FITC, a marker of sarcolemma including T-tubules (27), and anti-active CaMKII antibody to establish the intracellular localization of active CaMKII (Fig. 4 C and D). Before α₁ARS, active CaMKII was detectable at the plasmalemma (Fig. 4 A and E) as reported (17). After α₁ARS (100 μM phenylephrine) active CaMKII was still present at the plasmalemma (Fig. 4B), but it was also clearly visible along transverse bands that coincide with the location of WGA-FITC. This result suggests that a higher level of active CaMKII was localized at or near the T-tubules after α₁ARS than in the resting state (Fig. 4F).

To confirm the subcellular localization of active CaMKII before and after α₁ARS, immunoelectron microscopy was used. The number of gold particles over a standard area (SA) at the plasmalemma and at the level of the Z lines where T-tubules are located was counted in sections from four representative cells before and after α₁ARS. Before α₁ARS, the frequency of gold particles was very low both beneath the plasmalemma (0.10 ± 0.31 particles per SA, 0.39 ± 1.20 in normalized density, n = 30 areas) and at T-tubules (0.09 ± 0.29 particles per SA, 0.36 ± 1.15 in normalized density, n = 108 areas) (Fig. 5D). After application of phenylephrine (100 μM), as shown in Fig. 5 A and B, the frequency of gold particles (indicating active CaMKII) near T-tubules was increased (1.21 ± 0.92 particles per SA, 3.95 ± 3.00 in normalized density, n = 125 areas), showing ≈10-fold enrichment, consistent with the result of immunofluorescence image analysis (compare Figs. 4F and 5E). In contrast, the particle density beneath the plasmalemma remained low (0.49 ± 0.80 particles per SA, 1.60 ± 2.61 in normalized density, n = 53, P = 0.27, compared with that before α₁ARS) (Fig. 5C). Thus, immunogold labeling showed ≈2.5-fold enrichment of active CaMKII in T-tubules compared with the plasmalemma after α₁ARS (Fig. 5E). The particle density in the myofibrils areas was not significantly different before (0.25 ± 0.48 particles per SA, n = 110) and after α₁ARS (0.30 ± 0.52 particles per SA, n = 183).

There seems to be a discrepancy between the results obtained by using light and electron microscopy regarding the location of active CaMKII in the resting cells. The light-microscope images did not show active CaMKII at T-tubules, whereas electron-microscopic images detected a very low level at those sites. The confocal sections are relatively thick compared with the thickness of the T-tubules so that a small portion of each section is occupied by T-tubule membrane and a weak signal in this membrane is not detected. In contrast, the plasmalemma runs from one end of the section to the other, so that the signal from this membrane is detectable even if weak. In the thin sections for electron microscopy, plasmalemma and T-tubule membrane are more equally represented.

Discussion

In this study, we explored the involvement of CaMKII in the signal transduction pathway between α₁ARS and the potentiation of I_Ca,L in rat ventricular myocytes, and we show direct evidence indicating that CaMKII is activated by α₁ARS and has an important effect on I_Ca,L.

Our electrophysiological experiments using the perforated patch technique confirm the reported (8, 9) potentiation of I_Ca,L during α₁ARS in rat ventricular myocytes. When I_Ca,L is measured by using the whole-cell patch–clamp technique, phenylephrine does not affect I_Ca,L at any concentrations (4, 7), and we can mimic this effect by buffering the intracellular Ca²⁺ with BAPTA–acetoxymethyl ester. Thus, intracellular Ca²⁺ concentration is essential for the potentiation of I_Ca,L during α₁ARS. This result is consistent with the report that basal intracellular Ca²⁺ is sufficient for the initial CaMKII activation under physiological condition (28).

We chose 10 μM phenylephrine in our electrophysiological experiments because the effect of phenylephrine on I_Ca,L potentiation is saturated at this concentration (see Fig. 6, which is published as supporting information on the PNAS web site). α₁ARS has two opposite effects on I_Ca,L. (i) Higher concentration of phenylephrine (≥10 μM) cause a biphasic response: an initial brief (≈2 min) depression, or negative phase, followed by a potentiation or positive phase; and (ii) lower concentration of phenylephrine (≥1 μM) cause a monophasic positive effect (Fig. 6). The positive effect (potentiation) depends on CaMKII activation and PKC activation because CaMKII inhibition or PKC inhibition abolished the potentiation of I_Ca,L (Fig. 2 and Fig. 7, which is published as supporting information on the PNAS web site). There is less information about the negative effect of α₁ARS on I_Ca,L. In these experiments, we focused on the potentiation of I_Ca,L during α₁ARS (positive effect).

Our electrophysiological experiments using CaMKII inhibitors demonstrated the important role of CaMKII in the potentiation of I_Ca,L (positive phase) during α₁ARS in cardiac myocytes. Western immunoblot analysis confirmed an increase in active CaMKII (see also ref. 13) in parallel with the potentiation of I_Ca,L. Also, we showed that PKC, which is activated by the Gq-phospholipase C-diacylglycerol pathway (2), is involved in the activation of CaMKII during α₁ARS, based on the similar effects of CaMKII and PKC inhibition on I_Ca,L in the presence of phenylephrine. It has been reported that, in resting cardiac myocytes, there is significant activation of CaMKII but active CaMKII can be lost when the intracellular Ca²⁺ concentration is lowered to very low levels by removal of extracellular Ca²⁺ (17). Basal activity of CaMKII is determined mainly by the resting intracellular Ca²⁺ level (and not by the PKC activity). However, after α₁ARS, PKC activity has an important role in the additional and sustained activation of CaMKII. PKC can directly phosphorylate the autophosphorylation site of CaMKII in vitro, thus directly increasing CaMKII activity (29). This report strongly supports our hypothesis that there is a physiological linkage between PKC and CaMKII during α₁ARS.

Immunolabeling at the ultrastructural level demonstrates that the level of active CaMKII is very low along the plasmalemma and T-tubules in the resting cells but increases significantly along the T-tubules, where L-type Ca²⁺ channels are mostly present (11), after α₁ARS. The correspondence between the distributions of active CaMKII and L-type Ca²⁺ channels strongly supports our view that CaMKII directly phosphorylates the L-type Ca²⁺ channels and potentiates I_Ca,L in response to α₁ARS. Immunofluorescence microscopy confirms the increase in active CaMKII at the T-tubules from an undetectable to a detectable level with stimulation.

It has been reported (10) that CaMKII modulates I_Ca,L under physiological conditions. Several groups have demonstrated that Ca²⁺-dependent I_Ca,L facilitation is mediated by CaMKII-dependent phosphorylation of L-type Ca²⁺ channel, and this mechanism is considered to be related to the positive staircase of I_Ca,L induced by repeated depolarization from physiological holding potential (17, 26, 30). However, the molecular mechanisms of how I_Ca,L is potentiated by the activation of CaMKII have not been elucidated.

In summary, these experiments show that (i) α₁ARS potentiates I_Ca,L by PKC and CaMKII activation and (ii) activated CaMKII is highly localized close to the T-tubules after α₁ARS.