Protein phosphatase 1 and an opposing protein kinase regulate steady-state L-type Ca²⁺ current in mouse cardiac myocytes

Abstract

Studies have suggested that integration of kinase and phosphatase activities maintains the steady-state L-type Ca²⁺ current in ventricular myocytes, a balance disrupted in failing hearts. As we have recently reported that the PP1/PP2A inhibitor calyculin A evokes pronounced increases in L-type I_Ca, the goal of this study was to identify the counteracting kinase and phosphatase that determine ‘basal’I_Ca in isolated mouse ventricular myocytes. Whole-cell voltage-clamp studies, with filling solutions containing 10 mm EGTA, revealed that calyculin A (100 nm) increased I_Ca at test potentials between −42 and +49mV (44% at 0mV) from a holding potential of −80mV. It also shifted the V_0.5 (membrane potential at half-maximal) of both activation (from −17 to −25mV) and steady-state inactivation (from −32 to −37mV) in the hyperpolarizing direction. The broad-spectrum protein kinase inhibitor, staurosporine (300 nm), was without effect on I_Ca when added after calyculin A. However, by itself, staurosporine decreased I_Ca throughout the voltage range examined (50% at 0mV) and blocked the response to calyculin A, indicating that the phosphatase inhibitor's effects depend upon an opposing kinase activity. The PKA inhibitors Rp-cAMPs (100 μm in the pipette) and H89 (1 μm) failed to reduce basal I_Ca or to block the calyculin A-evoked increase in I_Ca. Likewise, calyculin A was still active with 10 mm intracellular BAPTA or when Ba²⁺ was used as the charge carrier. These data eliminate roles for protein kinase A (PKA) and calmodulin-dependent protein kinase II (CaMKII) as counteracting kinases. However, the protein kinase C (PKC) inhibitors Ro 31-8220 (1 μm) and Gö 6976 (200 nm) decreased steady-state I_Ca and blunted the effect of calyculin A. PP2A is not involved in this regulation as intracellular applications of 10–100 nm okadaic acid or 500 nm fostriecin failed to increase I_Ca. However, PP1 is important, as dialysis with 2 μm okadaic acid or 500 nm inhibitor-2 mimicked the increases in I_Ca seen with calyculin A. These in situ studies identify constitutive activity of PP1 and the counteracting activity of certain isoforms of PKC, in pathways distinct from receptor-mediated signalling cascades, as regulatory components that determine the steady-state level of cardiac L-type I_Ca.

Excitation–contraction coupling in cardiac cells is initiated by openings of L-type Ca²⁺ channels, which increases the local [Ca²⁺]_i that activates nearby sarcoplasmic reticulum Ca²⁺ release channels, or ryanodine receptors (Fabiato, 1983). The activation of a large number of ryanodine receptors gives rise to the [Ca²⁺]_i transient that activates contraction (Cheng et al. 1993). Excitation–contraction coupling is regulated by receptor-activated signalling pathways that frequently alter the phosphorylation state of Ca²⁺ handling proteins (Tada & Katz, 1982; Hartzell, 1988). It has been established that the action of kinases is integrated with protein phosphatase activity in these signalling pathways. For example, during the course of the cAMP–PKA cascade the serine/threonine phosphatase PP1 is inhibited (Ahmad et al. 1989; Gupta et al. 1996), augmenting the cAMP signal, while the Ca²⁺-dependent phosphatase calcineurin is activated, perhaps limiting the action of PKA (Santana et al. 2002). The functional importance of phosphatases is underscored by observations from immunoprecipitation studies revealing that another phosphatase, PP2A, is bound to the RyR (Marx et al. 2001) and the α1-subunit of the L-type Ca²⁺ channel (Davare et al. 2000).

Several lines of evidence have established that even in the absence of humoral stimulation there is a dynamic balance between kinase and phosphatase activity that controls ventricular myocyte function. Our laboratory has shown that intracellular dialysis of rat ventricular myocytes with exogenous PP2A decreases the [Ca²⁺]_i transient and steady-state L-type I_Ca (duBell et al. 1996). Application of PP1/PP2A inhibitors such as okadaic acid can activate L-type Ca²⁺ channels both in intact myocytes (Hescheler et al. 1988; Neumann et al. 1993; Hirayama & Hartzell, 1997) and in cell-attached patches (Ono & Fozzard, 1993; Wiechen et al. 1995). Recently, we have shown that the phosphatase inhibitor calyculin A increases contractility in isolated mouse ventricular myocytes by increasing L-type Ca²⁺ channel activity (duBell et al. 2002). These studies reveal that there is an ‘I_Ca reserve’ that is maintained by this dynamic phosphorylation–dephosphorylation balance. The functional importance of this balance is seen in ventricular myocytes from failing human heart in which a blunted response to β-adrenergic agonists is accompanied by a loss in phosphatase-sensitive L-type Ca²⁺ channel reserve (Schröder et al. 1998; Chen et al. 2002).

The major goal of the present study was to identify the specific signalling enzymes that determine the set point, or basal level, of L-type I_Ca. This kinase–phosphatase balance was examined in situ in murine ventricular myocytes, a model system that is increasingly important due to the many transgenic mouse heart failure models that are available. The main findings are that PP1 is an important phosphatase in the steady-state regulation of L-type I_Ca and that the opposing kinase in this setting is neither PKA nor CaMKII, but rather one or more isoforms of PKC. The significance of these new results is seen in recent studies of both transgenic mouse lines and human myocardium that implicate PP1 in heart failure (Neumann et al. 1997; Carr et al. 2002).

Methods

Cardiac myocyte preparation

Ventricular myocytes were isolated from the hearts of adult male CD-1 mice. After deep anaesthesia was induced with 30mgkg⁻¹ (i.p.) sodium pentobarbital (Abbott Laboratories), the heart was removed and washed. The ascending aorta was rapidly cannulated for Langendorff perfusion. Following a brief period of perfusion with enzyme-free buffer, the heart was perfused with buffer containing Type 2 collagenase (Worthington Biochemical Corp.) and protease (Fraction XIV, Sigma Chemical Corp.) as previously described (duBell et al. 2000). Following the isolation procedure, the cells were maintained at room temperature in Hepes-buffered Dulbecco's modified Eagle's medium supplemented with heat-inactivated fetal calf serum (10%) and insulin (1 unit ml⁻¹). All experiments were carried out within 8 h of cell isolation

Voltage clamp and I_Ca

L-type Ca²⁺ currents were recorded at 35°C from single, isolated ventricular myocytes using low resistance (<1.5MΩ) patch-type microelectrodes in the whole-cell mode. All experiments were conducted on a Nikon Diaphot 300 inverted microscope using an Axopatch 200A voltage-clamp amplifier with a CV202A headstage (Axon Instruments). Cells were placed in the experimental chamber and superfused with a solution containing 140mm NaCl, 5mm KCl, 10mm Hepes, 1.0mm MgCl₂, 0.33mm NaH₂PO₄ and 10mmd-glucose at pH 7.4 (NaOH). When Ca²⁺ currents were evoked from a holding potential of −40mV (Figs 3–7), the cells were superfused with a similar solution with the exception that 5.0mm CsCl replaced the KCl and 1.8mm CaCl₂, 5mm 4-aminopyridine (4-AP), 10mm tetraethylammonium chloride (TEA) and 10μm tetrodotoxin (TTX) were added. When Ca²⁺ currents were evoked from a holding potential of −80mV (Figs 1 and 2), the extracellular solution was Na⁺ and K⁺ free and consisted of 150mm TEACl, 5mm CsCl, 10mm Hepes, 1.0mm MgCl₂, 10mm dextrose, 1.8mm CaCl₂, 5mm 4-AP and 10μm TTX. For all experiments, the pipette filling solution consisted of 120mm CsCl, 15mm TEACl, 5mm MgATP, 10mm Hepes, 1.0mm MgCl₂ and 10mm of either EGTA or BAPTA (pH to 7.2 with CsOH). The use of Cs⁺, TEACl and 4-AP assured block of currents through K⁺ channels and TTX was present to block currents through Na⁺ channels.

Figure 3. Effects of calyculin A and staurosporine on I_Ca–V relationship.

A, top, L-type Ca²⁺ currents recorded from the same cell in control buffer (left), after 5 min in 100nm calyculin A (centre) and after 5 min in 100nm calyculin A and 300nm staurosporine (right). The holding potential was −70mV. After a 1000 ms voltage ramp to −40mV, the cell was held at −40mV for 330 ms and I_Ca was elicited by a 300 ms test pulse to a potential between −35 and +60mV. There was a 5 s interval between each test pulse. A, bottom, superimposed plots of the mean (±s.e.m., n= 4, 2) voltage dependence of I_Ca in control (▪), calyculin A (♦) and calyculin A–staurosporine (▿). B, top, L-type Ca²⁺ currents recorded from the same cell in control (left), after 5 min in 300nm staurosporine (centre) and after 5 min in 300nm staurosporine and 100nm calyculin A (right). B, bottom, superimposed plots of the mean (±s.e.m., n= 5, 2) voltage dependence of I_Ca in control (▪), staurosporine (•) and staurosporine–calyculin A (▵).

Figure 7. Effects of intracellular okadaic acid and inhibitor–2 on the I_Ca–V relationship.

A, mean (± s.e.m.) I_Ca–V plots from cells dialysed with okadaic acid (OA) in concentrations of 10nm (▪ n= 5, 2), 100nm (□; n= 4, 2) and 2μm▴ n= 6, 2). B, average I_Ca–V plots from control cells (▪; n= 9, 4) and cells dialysed with 500nm inhibitor-2 (□; n= 11, 6). C, summary histogram of the effects of calyculin A, okadaic acid, fostriecin and inhibitor-2 at 0mV. The first two bars represent the average I_Ca before and after calyculin A from all of the experiments in which calyculin A was added in the absence of any inhibitors (n= 19, 8; data from Figs 3A, 4A and 6A and B).

Figure 1. The effect of calyculin A on the voltage dependence and steady-state inactivation of I_Ca.

A, L-type Ca²⁺ currents from the same cell before (left) and after (right) 5 min exposure to 100nm calyculin A (Caly A). Internal and external solution were Na⁺ and K⁺ free. Test depolarizations (300 ms) were given to potentials between −70 and +63mV from a holding potential of −80mV. The interval between test pulses was 5 s. B, plots of the mean (±s.e.m., n= 7, 3) voltage dependence of I_Ca before (▪) and after (▴) 5 min of exposure to 100nm calyculin A. C, L-type Ca²⁺ currents from the same cell before (left) and after (right) 5 min of exposure to 100nm calyculin A. Cells were held for 1200 ms at potentials between −90 and −6mV following a 1000 ms ramp from −70mV. I_Ca was elicited by a 300 ms test pulse to 0mV, followed by repolarization to −70mV, with an 8 s interval between test pulses (see inset for schematic). D, plots of mean (±s.e.m., n= 8, 3) I_Ca density versus holding potential before (▪) and after (▴) 5 min of 100nm calyculin A.

Figure 2. Effects of calyculin A on the voltage dependencies of activation and steady-state inactivation.

A, normalized plots of the mean (±s.e.m., n= 7, 3) voltage dependence of I_Ca before (▪) and after (▴) 5 min of exposure to 100nm calyculin A. The current at each test potential was normalized to the maximum current from the experiment and the average values were plotted versus test potential. The plots are from the data used for Fig. 1A and B). Normalized plots of the voltage dependencies of activation (open symbols) and steady-state inactivation (filled symbols) from cells before (▪, □) and after (▴, ▵) 100nm calyculin A (n= 8, 3). Voltage dependence of activation was determined from the I_Ca–V relationships shown in Fig. 1A. The conductance (g) at each potential was calculated as g=I_Ca/(E_m−E_rev), where E_m is the test potential and E_rev the reversal potential, as determined from the I_Ca–V plot. The conductance was then normalized to the maximal conductance obtained during the experiment. Voltage dependence of steady-state inactivation was determined from data in Fig. 1B. The current at each holding potential was normalized to the largest current recorded during the experiment. The average values for each were plotted versus holding potential.

The Ca²⁺ current recordings were low-pass filtered at 2 kHz by the voltage-clamp amplifier, digitized at 5 kHz (Digidata 1200, Axon Instruments) and stored on a computer for subsequent analysis. The current amplitudes were normalized to cell capacitance to account for differences in cell size and are reported as current densities (pA pF⁻¹). Cell capacitance was calculated by integrating the uncompensated capacity transients elicited by 10 ms hyperpolarizing pulses from −70 to −80mV.

Calyculin A and staurosporine (Alexis) were prepared as 1mm stocks in DMSO. H89 and Ro 31-8220 (Biomol) were prepared as 10mm stocks in DMSO. Gö 6976 (Biomol) was prepared as a 2mm stock in DMSO. These stocks were added to the extracellular buffers to bring each agent to the indicated concentration. The control solutions included the same concentration of DMSO as the experimental solutions. The total extracellular concentration of DMSO never exceeded 0.04%. Okadaic acid (Alexis) was prepared as a 500μm stock in DMSO and added to the filling solution. The final concentration of DMSO in the filling solution for each concentration of okadaic acid was 0.4%. Fostriecin (Alexis), 8-bromo cAMP (Calbiochem) and inhibitor-2 (Upstate Biotechnology) were prepared as aqueous stock solutions and added to the filling solution to the indicated concentrations.

Statistics

The experimental values are expressed as the means ± s.e.m. The sample size for each experiment is indicated in the figure legend or text, followed by the number of hearts from which those cells were taken. Statistical significance (P < 0.05) was assessed using either the paired or unpaired Student's t test, as appropriate.

Results

We have recently reported that the PP1/PP2A inhibitor calyculin A evokes a rapid and pronounced increase in the magnitudes of the mouse ventricular [Ca²⁺]_i transient and twitch contraction, responses that result principally from an increase in L-type I_Ca (duBell et al. 2002). In the present study, we extend these findings to characterize, in situ, the mechanisms of the dynamic control of the steady-state murine ventricular L-type Ca²⁺ current in the absence of humoral signalling.

Effects of calyculin A on the biophysical properties of the L-type Ca²⁺ current

Initial studies were performed to characterize the effects of calyculin A on the biophysical properties of macroscopic L-type I_Ca. In contrast to our previous study, which was carried out under conditions in which we could examine excitation−contraction coupling (duBell et al. 2002), the present experiments were performed with pipette filling solution containing 10mm EGTA to minimize the effects of sarcoplasmic reticulum Ca²⁺ release on I_Ca. In addition, the initial experiments (Figs 1 and 2) were performed in the absence of internal and external Na⁺ and K⁺ to increase the voltage range over which I_Ca could be examined and to minimize overlapping inward and outward currents. It was not certain that calyculin A would increase I_Ca in the absence of intracellular Ca²⁺ transients, as we had previously seen in their presence (duBell et al. 2002). However, Fig. 1 shows that calyculin A did, in fact, increase I_Ca under these conditions. The magnitude of the current was significantly greater in calyculin A at all potentials between −42 and +49mV (Fig. 1B). For example, the current at 0mV increased from −12.0 ± 0.80 to −17.2 ± 1.2 pA pF⁻¹(P= 0.001). The increases at −7 and −14mV were even greater because, in addition to increasing the amplitude of the current, calyculin A also shifted the peak of the I_Ca–voltage relationship in the hyperpolarizing direction, from 0mV to −14mV (Fig. 1B). This observation is emphasized in Fig. 2A, which shows the control and calyculin A I_Ca–voltage relationships normalized to the peak I_Ca observed in each experiment. Calyculin A exposure also resulted in an apparent shift in the voltage dependence of activation of I_Ca, with the threshold for activation of I_Ca shifting from −35 to −42 mV (control, 0.09 ± 0.06 pA pF⁻¹; calyculin A, −0.25 ± 0.04 pA pF⁻¹). However, calyculin A did not affect the kinetics of I_Ca decay. At 0mV, the decay of the control current was well fitted by two exponentials with time constants (τ) of 22 ± 1 and 72 ± 2 ms. After calyculin A, the fast and slow time constants were 24 ± 1 ms (P= 0.12) and 82 ± 4 ms (P= 0.07), respectively. While the amplitude of the slow component remained unchanged following calyculin A (−4.77 ± 0.4 pA pF⁻¹ in control versus−4.58 ± 0.6 with calyculin A, P= 0.60), the amplitude of the fast component was increased, from −7.7 ± 0.7 to −13.5 ± 0.8 pA pF⁻¹(P= 0.0005).

The effects of calyculin A on the voltage dependence of steady-state inactivation were also examined (Fig. 1C and D). At holding potentials between −90 and −48mV, the current was significantly greater after calyculin A (−60mV; control, 13.1 ± 1.0 pA pF⁻¹; calyculin A, 18.6 ± 1.2 pA pF⁻¹, P= 0.0003, Fig. 1D). However, at potentials between −27 and −13mV, the current was significantly less after calyculin A (−60mV; control, 3.61 ± 0.44 pA pF⁻¹; calyculin A, 2.72 ± 0.50 pA pF⁻¹, P= 0.001, Fig. 1D). These data suggest that in addition to its other effects, calyculin A shifted the voltage dependence of steady-state inactivation in the hyperpolarizing direction.

The data presented in Fig. 1 were used to construct normalized plots of both the voltage dependence of peak conductance (activation) and the voltage dependence of steady-state inactivation (Fig. 2B, see legend for details). Consistent with the hyperpolarizing shift in both the threshold for activation and the peak of the I_Ca−V relationship, calyculin A also produced a hyperpolarizing shift in the voltage dependence of activation, with the voltage at which activation was half-maximal (V_0.5) changing from −17mV in control to −25mV after calyculin A. As suggested by the data presented in Fig. 1D, the inactivation V_0.5 is also shifted in the hyperpolarizing direction, from a potential of −32mV in control to −37 mV in calyculin A. Thus, these data reveal that a potent inhibitor of PP1/PP2A evokes fundamental changes in the macroscopic properties of murine ventricular whole-cell L-type Ca²⁺ current in the steady state.

Sequential addition of staurosporine and calyculin A

The large and rapid responses seen following application of calyculin A (Fig. 1) suggest that an active dynamic balance between phosphorylation (stimulation of I_Ca) and dephosphorylation (inhibition of I_Ca) determines the magnitude of the L-type Ca²⁺ current in the absence of any neurohumoral stimulation. In this scheme, the increase in I_Ca in response to calyculin A would result from kinase activity that becomes predominant following the inhibition of the counteracting phosphatase. Thus the notion that the action of calyculin A is dependent on kinase activity was tested through consecutive additions of calyculin A and the broad-spectrum kinase inhibitor staurosporine.

These experiments began with the recording of an I_Ca–voltage relationship under control conditions. The cells were then exposed to either calyculin A or staurosporine alone for 5 min. Following the recording of an I_Ca–voltage relationship in this condition, the cells were treated with both calyculin A and staurosporine for an additional 5 min and another I_Ca–voltage relationship was recorded. These and all subsequent experiments were conducted in Na⁺-containing solutions and the I_Ca–voltage relationships were determined from a holding potential of −40mV to inactivate I_Na (see Methods and legend for Fig. 3). Control experiments (not shown) demonstrated that the effects of both calyculin A and staurosporine were complete within 5 min. In addition, control experiments in the absence of any inhibitors demonstrated that there was minimal I_Ca rundown over the same time period. For example, the magnitude of the current at 0mV decreased by only about 5%, from −12.7 ± 1.7 to −12.0 ± 1.7 pA pF⁻¹(n= 6) and the decreases at other test potentials were comparably small. Figure 3A shows the results obtained when calyculin A was added first. After 5 min in calyculin A, there was a statistically significant increase in the magnitude of I_Ca at each test potential between −28 and +56mV. For example, I_Ca increased by 33% from −13.9 ± 2.9 to −18.5 ± 2.9 pA pF⁻¹ at 0mV (P= 0.0005). In addition, calyculin A altered the voltage dependence of I_Ca, shifting the peak of the I_Ca–V relationship from 0 to −7mV (Fig. 3A). Importantly, subsequent addition of staurosporine (300nm) had no effect on the magnitude of I_Ca at any potential (−17.9 ± 3.3 pA pF⁻¹ at 0mV, P= 0.20). This latter experiment indicated that the kinase inhibitor alone did not display any significant non-specific effects on I_Ca.

In contrast, the results obtained when staurosporine was added first are quite different (Fig. 3B). Initial application of staurosporine had a marked effect on the magnitude of I_Ca, decreasing the current at all potentials between −21 and +56mV. The current at 0mV decreased by approximately 50%, from −11.8 ± 0.9 to −6.05 ± 0.42 pA pF⁻¹, after 5 min in staurosporine (P= 0.0004). Importantly, when calyculin A was added after staurosporine there was no increase in I_Ca at any potential, nor was there any shift in the voltage dependence of the current (Fig. 3B). These data support a model in which the amplitude of the steady-state L-type Ca²⁺ current is controlled by a balance between kinase and phosphatase activities, with the effects of calyculin A on I_Ca dependent on the stimulatory effects of the kinase and the effects of staurosporine dependent on the inhibitory effects of the phosphatase.

Calyculin A and protein kinase A inhibition

The effects of calyculin A on the macroscopic biophysical properties of the L-type Ca²⁺ current shown in Figs 1 and 2 are quite similar to those seen following stimulation of the β-adrenergic receptor–PKA cascade (Piacentino et al. 2000; Chen et al. 2002). Thus, it was intriguing to speculate that the increase in I_Ca evoked by calyculin A resulted from a shift in the kinase–phosphatase balance toward the activity of PKA. This hypothesis was further supported by the observation that 1μm isoprenaline (isoproterenol) was without effect on I_Ca after 5 min exposure to 100nm calyculin A (Fig. 4A).

Figure 4. The effect of PKA inhibition on the response to calyculin A.

A, I_Ca–V plots from control (▪), after 5 min in 100nm calyculin A (□) and after 5 min in 1μm isoprenaline and 100nm calyculin A (▴, n= 4, 2). B, I_Ca–V plots from control cells before (▪) and after (□) 100nm isoprenaline (n= 4, 2) and from cells dialysed with 100μm Rp-cAMPs before (▴) and after (▵) isoprenaline (n= 5, 3). C, I_Ca–V plots from control cells (▪; n= 14, 7), cells dialysed with 100μm Rp-cAMPs (□; n= 10, 4) and a subset of the cells dialysed with 100μm Rp-cAMPs that were subsequently superfused with 100nm calyculin A (▴; n= 5, 2). D, I_Ca–V plots in control (▪), after 5 min in 1μm H89 (▴) and after 5 min in H89 and 100nm calyculin A (▵) (n= 4, 3).

Accordingly, the notion that PKA is involved in the steady-state regulation of murine ventricular L-type Ca²⁺ current was tested by using inhibitors of PKA. Rp-cAMPs, a competitive antagonist to cAMP, was used to block the activity of PKA (Rothermel & Parker-Botelho, 1988). Control experiments revealed that 100μm Rp-cAMPs in the pipette filling solution completely blocked the effects of 100nm isoprenaline on the I_Ca–voltage relationship (Fig. 4B). For example, the control value of I_Ca at 0mV increased from −7.03 ± 0.9 to −16.6 ± 2.0 pA pF⁻¹ in isoprenaline. However, when 100μm Rp-cAMPs was included in the filling solution, isoprenaline failed to increase the amplitude of I_Ca at any potential (0mV: −7.51 ± 0.8 pA pF⁻¹ basal; −7.20 ± 0.8 pA pF⁻¹ in the presence of isoprenaline). Unlike staurosporine (Fig. 3B), the cAMP-antagonist had no effect by itself on I_Ca density at any test potential (Fig. 4C). A comparison between 14 cells with and 10 cells without Rp-cAMPs revealed I_Ca densities of −9.2 ± 0.7 and 10.3 ± 1.2 pA pF⁻¹, respectively, at 0mV (P= 0.40, unpaired t test) and a similar lack of effect at other test potentials (Fig. 4C). Furthermore, the stimulatory action of calyculin A was insensitive to this PKA-inhibitory dose of Rp-cAMPs (Fig. 4C). A subset of the cells dialysed with 100μm Rp-cAMPs were exposed to calyculin A (see legend to Fig. 4). Under these conditions, calyculin A still increased I_Ca at all potentials between −21 and +56mV. For example, I_Ca at 0mV was increased by 60%, from −13.1 ± 1.4 to −20.9 pA pF⁻¹(P= 0.006).

It is important to note that Rp-cAMPs blocks PKA activation by inhibiting the cAMP-evoked dissociation of the regulatory subunit RII from the enzyme complex (Rothermel & Parker-Botelho, 1988). Although not likely, it is possible that a constitutively active, RII-free form of PKA could underlie the response to calyculin A. Accordingly, complementary experiments were performed with the PKA inhibitor H89, an established direct inhibitor, of the PKA catalytic subunit in ventricular myocytes (Yuan & Bers, 1994, 1995; You et al. 1997). As shown in Fig. 4D, cellular responses to H89 were nearly identical to those of Rp-cAMPs. For example, the average current densities at 0mV in control and H89 were −12.9 ± 1.7 and −12.0 ± 1.9 pA pF⁻¹, respectively (P= 0.27). Further, I_Ca was significantly increased by calyculin A at all test potentials between −2 and +56mV in H89-treated cells. At 0mV, the current was increased by 47%, from −9.49 ± 1.4 to −17.9 ± 2.6 pA pF⁻¹(P= 0.036). Taken together, these results eliminate PKA as the candidate kinase in the tonic regulation of I_Ca.

Calyculin A and CaMKII inhibition

Another kinase whose activity could be unmasked by calyculin A is the Ca²⁺- and calmodulin-dependent protein kinase, CaMKII. Although the experiments in the present study were carried out under conditions of high Ca²⁺-buffering, it has been shown that 10mm EGTA does not buffer [Ca²⁺]_i sufficiently rapidly to prevent local increases in [Ca²⁺]_i and the subsequent activation of CaMKII (Naraghi & Neher, 1997; You et al. 1997). Thus, experiments were carried out using filling solution that contained 10mm BAPTA. BAPTA is a more rapid buffer of Ca²⁺ (Naraghi & Neher, 1997; You et al. 1997) that has been shown to block CaMKII-dependent processes in ventricular myocytes (Xiao et al. 1994; Vinogradova et al. 2000). However, it is clear from the results presented in Fig. 5A that the calyculin A-evoked increases in I_Ca were completely insensitive to intracellular BAPTA (compare with Fig. 3). There was a significant increase in I_Ca at all potentials between −21 and +49mV. The calyculin A-evoked increase at −7mV was from −13.3 ± 1.9 to −24.4 ± 1.8 pA pF⁻¹(P= 0.003), comparable to the increase evoked by calyculin A in EGTA-treated cells (Fig. 3), from a control value of −13.6 ± 1.9 to −20.4 ± 2.8 pA pF⁻¹ in calyculin A. To address the possibility that inclusion of 10mm BAPTA still provided insufficient buffering to prevent local Ca²⁺-dependent activation of CaMKII proximate to the channel, a series of complementary experiments was performed in which 1.8mm Ba²⁺ was substituted for Ca²⁺ as the charge carrier. While Ba²⁺ is widely used as the permeant ion in studies of the L-type Ca²⁺ channel (Ono & Fozzard, 1993; Xiao et al. 1994; Wiechen et al. 1995; Wu et al. 1999; Dzhura et al. 2000), this ion cannot activate CaMKII (Chao et al. 1984). As shown in Fig. 5B, L-type Ca²⁺ channel-mediated Ba²⁺ currents were still increased by calyculin A and the peak of the I_Ca–V relationship was shifted in the hyperpolarizing direction. Taken together, these data indicate that CaMKII activity is not responsible for the activation of I_Ca observed following phosphatase inhibition with calyculin A.

Figure 5. Effects of minimizing CaMKII activity on the response to calyculin A.

A, top, currents recorded before (left) and after (right) exposure to 100 nm calyculin A using pipette filling solution containing 10 mm BAPTA. A, bottom, mean (±s.e.m., n= 5, 2) I_Ca–V relationships before (▪) and after (□) 100nm calyculin A. B, top, representative currents elicited by steps to 0mV with 1.8mm Ba²⁺, instead of Ca²⁺, as the charge carrier. B, bottom, mean (±s.e.m., n= 4, 1) current–voltage relationships from these experiments before (▪) and after (□) 100nm calyculin A.

Calyculin A and PKC inhibition

Several studies have implicated PKC in receptor-mediated regulation of the L-type Ca²⁺ channel (for review see Kamp & Hell, 2000). Accordingly, it was a priority to assess the role of PKC in the action of calyculin A on I_Ca. The pan-specific PKC inhibitor Ro 31-8220 was used for this purpose. As shown in Fig. 6A, Ro 31-8220 had no effect on calyculin A-stimulated I_Ca. However, when Ro 31-8220 was applied before calyculin A, the results were quite different. Ro 31-8220 decreased the amplitude of I_Ca at each potential between −14 and +49mV. For example, the current density at 0mV decreased from −12.6 ± 1.2 to −8.5 ± 0.50 pA pF⁻¹ (P= 0.006, see Fig. 6A). Following Ro 31-8220, calyculin A significantly increased I_Ca at all potentials between −28 and +49mV. However, the response was significantly attenuated, as I_Ca increased only to the initial control values (Fig. 6A). These results suggest the possibility that PKC is involved in the determination of the basal level of I_Ca in these cells. This idea was tested further in studies with another inhibitor, Gö 6976. The effects of Gö 6976 were similar to those of Ro 31-8220 (Fig. 6B). Gö 6976 decreased the current only when added before calyculin A, and prior exposure to this inhibitor blunted the response of I_Ca to calyculin A. For example, the average current at 0mV when calyculin A was added after Gö 6976 was −11.6 ± 1.2 pA pF⁻¹, identical to the control value of −11.9 ± 0.78 pA pF⁻¹. It is important to note that, of the specific kinase inhibitors utilized in the present study, only the PKC inhibitors decreased the amplitude of basal I_Ca and attenuated the calyculin A-induced activation of I_Ca.

Figure 6. The effects of PKC inhibition with Ro 31-8220 and Gö 6976 on the response to calyculin A.

A, left, superimposed plots of the voltage dependence of I_Ca in control (▪), after 5 min in 100nm calyculin A (○) and after 5 min in calyculin A and 1μm Ro 31-8220 (•, n= 6, 2). A, right, similar panel with the reverse order of application (n= 8, 4): Control (▪), 1μm Ro 31-8220 (□) and Ro 31-8220 and 100nm calyculin A (•). B, left, voltage dependence of I_Ca in control (▪), after 5 min in 100nm calyculin A (○) and after 5 min in calyculin A and 200nm Gö 6976 (•, n= 5, 2). B, right, similar panel with the reverse order of application (n= 9, 2): Control (▪); 200nm Gö 6976 (□) and Gö 6976 and 100nm calyculin A (•).

Potential non-specific effects of PKC inhibitors are a concern. However, since neither Gö 6976 nor Ro 31-8220 had any effect on I_Ca in calyculin A-treated cells, direct effects on the channel are unlikely. In addition, β-adrenergic stimulation has been reported to result in activation of PKC (Yabe et al. 1998) so it is possible that the effects of the PKC inhibitors may result from an interaction with the PKA pathway. Thus, a series of control experiments were carried out to examine this possibility by comparing the ability of isoprenaline to increase I_Ca in the absence and presence of 1μm Ro 31-8220. In control cells, there was little change over 5 min in I_Ca at any potential. The values at 0mV were −12.9 ± 1.3 and −11.4 ± 0.7 pA pF⁻¹ after 5 min (n= 3, 3; P= 0.12). After 5 min in 100nm isoprenaline, I_Ca at 0mV reached −21.8 ± 2.5 pA pF⁻¹. As was seen in Fig. 6A, 5 min exposure to 1μm Ro 31-8220 decreased I_Ca at all potentials. In this sample (n= 6, 5), I_Ca at 0mV decreased from −12.5 ± 0.7 to −9.2 ± 0.6 pA pF⁻¹(P= 0.001). Following Ro 31-8220, isoprenaline increased I_Ca at all potentials, with the average at 0mV reaching −19.4 ± 1.9 pA pF⁻¹. The amplitude of I_Ca after isoprenaline in the presence of Ro 31-8220 was statistically indistinguishable from its amplitude after isoprenaline in control cells (P= 0.46 at 0mV, unpaired t test). These observations support the view that the effects of PKC inhibitors in the present study result only from PKC inhibition. Thus, along with the results shown in Fig. 6, they are consistent with the view that at least one, and possibly more, isoforms of PKC are involved in the tonic regulation of I_Ca.

The effects of specific PP1 and PP2A inhibitors

An important goal of this study was to identify the specific phosphatase that controls steady-state I_Ca activity in situ. We have previously demonstrated that calyculin A blocks both PP1 and PP2A under the conditions used in the present study (duBell et al. 2002). Okadaic acid also inhibits both PP1 and PP2A but it has a dose–inhibition profile that allows for a distinction between the activities of these two enzymes (Bialojan & Takai, 1988). Accordingly, okadaic acid was dialysed into cells as a component of the pipette filling solution at concentrations of 10nm, 100nm and 2μm (Fig. 7A). Note that the I_Ca–V plots for both 10 and 100nm okadaic acid superimpose and that the peak of each is at 0mV. The current densities at 0mV were −13.6 ± 2.06 and −13.8 ± 0.9 pA pF⁻¹ with 10 and 100nm okadaic acid, respectively (P= 0.95, unpaired t test). Further, these values are very close to the established control values. This is emphasized in Fig. 7C, a histogram showing the average effects on I_Ca at 0mV of the phosphatase inhibitors used in the present study. When the pipette solution contained 2μm okadaic acid, however, there was a marked increase in I_Ca density. For example, the average current density at 0mV was −21.2 ± 1.9 pA pF⁻¹ (P= 0.015, unpaired t test, compared to 10nm okadaic acid). In addition, the increase in I_Ca was accompanied by the characteristic shift in the peak of the I_Ca–V relationship from 0 to −7mV (Fig. 7A). These data suggest that PP1, rather than PP2A, is important in the tonic regulation of I_Ca.

This hypothesis was further explored in more precise experiments using fostriecin and inhibitor-2, specific inhibitors of PP2A and PP1, respectively. When fostriecin (Walsh et al. 1997) was included in the pipette at 500nm, there was no change in either the amplitude of I_Ca (Fig. 7C and 0mV, control, −11.6 ± 0.85 pA pF⁻¹(n= 6, 2); fostriecin, −11.9 ± 1.8 pA pF⁻¹(n= 5, 2), P= 0.85, unpaired t test) or on the peak of the I_Ca–V relationship (data not shown). However, as shown in Fig. 7B, a calyculin A-like effect was indeed observed in the cells dialysed with 500nm inhibitor-2, a specific protein inhibitor of PP1 (Silberman et al. 1984). Note that in the presence of inhibitor-2, the current density was significantly increased at all potentials between −21 and +14mV (P < 0.05, unpaired t tests) and that the characteristic leftward shift, from 0 to −7mV, was observed in the peak of the I_Ca–V relationship. Thus, the data presented in Fig. 7 reveal that PP1, and not PP2A, is an important regulator of the steady-state L-type Ca²⁺ current in intact cardiac myocytes.

Discussion

Since L-type Ca²⁺ channels play a central role in Ca²⁺ homeostasis in cardiac cells, neurohumoral regulation of I_Ca through kinases and phosphatases has been the subject of study in several laboratories (Hartzell, 1988; Trautwein & Hescheler, 1990; Kamp & Hell, 2000; Herzig & Neumann, 2000). Studies of signal transduction pathways have revealed that active modulation of Ca²⁺ influx through I_Ca results from a coordinated interplay in the activities of kinases and phosphatases. For example, in the well-known β-adrenergic receptor–PKA cascade, inhibitor-1 is a downstream PKA target, and is activated to attenuate PP1 activity (Ahmad et al. 1989; Gupta et al. 1996). Recently, Santana et al. (2002) provided evidence that the phosphatase calcineurin opposes the action of PKA in mouse ventricular myocytes, providing a Ca²⁺-dependent mechanism for limiting the action of β-agonists. However it is not clear at present how these two distinct phosphatase pathways integrate to regulate I_Ca during the PKA cascade.

Other data demonstrate a regulatory model in which I_Ca is precisely modulated by a dynamic kinase–phosphatase balance even in the absence of humoral stimulation. For example, application of serine/threonine phosphatase inhibitors alone is sufficient to increase whole-cell I_Ca in frog (Frace & Hartzell, 1993), guinea-pig (Hescheler et al. 1987, 1988; Neumann et al. 1993, 1994; Hirayama & Hartzell, 1997), and more recently mouse ventricular myocytes (duBell et al. 2002). In a complementary study we have shown that addition of exogenous protein phosphatase 2A decreased I_Ca in rat ventricular myocytes (duBell et al. 1996). While the molecular details of this dynamic signalling system remain to be defined, the critical importance of this constitutive kinase–phosphatase balance in human heart disease was recently illuminated by studies on myocytes from failing human hearts. Defective Ca²⁺ homeostasis in cells from failing hearts results in part from a disruption of kinase–phosphatase control of I_Ca, as phosphatase inhibitors fail to increase I_Ca in the normal manner (Schröder et al. 1998; Chen et al. 2002). These studies suggest that a loss of ‘I_Ca reserve’, whereby the ability of I_Ca to increase in response to physiological demand is severely limited, may be a central feature of human heart failure. Thus, it is important to identify the molecular details of the kinase–phosphatase balance responsible for the tonic regulation of I_Ca in cardiac myocytes.

The goal of the present study was to identify signalling enzymes that control ‘basal’ cardiac L-type Ca²⁺ channel activity in situ. We examined this issue in mouse ventricular myocytes because of the growing importance of murine models in the study of heart disease. Further, we recently reported that increased twitch contraction in such cells after exposure to the PP1/PP2A inhibitor calyculin A was entirely attributable to increased L-type I_Ca (duBell et al. 2002). A finding reported here is that PP1 is an important modulator of steady-state I_Ca. Another conclusion is that the opposing kinase, rather than PKA or CaMKII, is probably one or more members of the protein kinase C family. Importantly, these studies provide new insights into the local control of I_Ca and illuminate the distinction between the pathways that underlie the tonic regulation of I_Ca and those that modulate I_Ca in response to receptor-activated signalling pathways.

Throughout the studies reported here there is a consistent, unifying pattern to the results that allowed an in situ analysis of the dynamic integration of phosphatase–kinase signalling. First, basal, steady-state, Ca²⁺ current levels were in the region of −12 pA pF⁻¹ under the conditions used (see Figs 3–7). Second, serine/threonine phosphatase inhibitors, when effective, acted only to increase basal I_Ca to the −18 to −21 pA pF⁻¹ range (see Figs 3–7). Third, protein kinase inhibitors, when effective, displayed two clear, reproducible effects. When applied first they reduced the basal, steady-state Ca²⁺ current levels to the 6–8 pA pF⁻¹ range (Figs 3B and 6A and B) and they markedly limited the calyculin A-evoked increases to the 11–12 pA pF⁻¹ range (Figs 3B and 6A and B) compared to control cells treated with calyculin A alone (21 pA pF⁻¹). Other experiments argued against a non-specific effect of the protein kinase inhibitors, as these agents had no effect on I_Ca in cells that were pretreated with calyculin A (Figs 3A and 6A and B). Thus, such striking reciprocity of action of kinase and phosphatase inhibitors on I_Ca allowed us to analyse the steady-state balance between kinases and phosphatases in situ.

PP1 versus PP2A

The use of broad spectrum serine/threonine phosphatase inhibitors has revealed the importance of dephosphorylation mechanisms in the control of L-type Ca²⁺ channels in normal and failing heart (Hescheler et al. 1987, 1988; Neumann et al. 1993, 1994; duBell et al. 1996; Hirayama & Hartzell, 1997; Schröder et al. 1998; Chen et al. 2002). Accordingly, an important goal in this study was to identify the phosphatase that contributes to the ‘set point’ or reserve of steady-state I_Ca. A variety of previous studies led us to suspect that PP2A was that phosphatase. First, we have shown that intracellular addition of exogenous PP2A decreases basal L-type I_Ca (duBell et al. 1996). Further, PP2A is bound constitutively to the α_1c subunit of the L-type Ca²⁺ channel. In brain, this phosphatase is part of the L-type Ca²⁺ channel macromolecular signalling complex and reverses the effects of PKA (Davare et al. 2000, 2001). Finally, we have previously shown that calyculin A is a powerful PP2A inhibitor under the conditions used here (duBell et al. 2002). However, our results (Fig. 7) show that low concentrations of okadaic acid (10 and 100nm), conditions that inhibit PP2A and not PP1 (Bialojan & Takai, 1988), and the selective PP2A inhibitor fostriecin (Walsh et al. 1997) are without effect on I_Ca. Rather, several results presented here reveal that PP1 is important in this regulation. First, okadaic acid increased I_Ca only at concentrations within the range that inhibits PP1 (Bialojan & Takai, 1988). In addition, more precise studies showed that inhibitor-2, a selective protein inhibitor of PP1 (Silberman et al. 1984), also markedly increased I_Ca. These results provide evidence that PP1 is important in controlling steady-state L-type Ca²⁺ channel activity. They are consistent with cell-attached patch-clamp studies of L-type Ca²⁺ channels conducting Ba²⁺ currents which suggest that PP-1 inhibition results in increased ion flux through single L-type Ca²⁺ channels (Ono & Fozzard, 1993; Wiechen et al. 1995). Taken together, our results indicate the importance of PP1, and little or no role for PP2A, in determining the level of I_Ca in the absence of receptor-mediated signalling.

There is an apparent inconsistency with our previous studies which showed that PP2A, and not PP1, inhibited basal I_Ca (duBell et al. 1996). However in this earlier study the role of phosphatases was studied by adding excess exogenous enzymes in the pipette filling solution. In contrast, the present experiments critically examine the role of endogenous phosphatases in situ, including locally targeted forms.

PKA

Several lines of evidence strongly implicated PKA as the counter-acting kinase that regulated I_Ca. First, it is well established that PKA stimulates ventricular I_Ca (Hartzell, 1988; Trautwein & Hescheler, 1990; Kamp & Hell, 2000) and recent work has shown that PKA is part of the L-type Ca²⁺ channel complex (Davare et al. 1999, 2001) through interactions with AKAPs (Gao et al. 1997; Davare et al. 1999). Further, the electrophysiological analysis reported here strongly implicated this kinase in mouse ventricle. In particular, calyculin A produced a hyperpolarizing shift in the voltage dependence of I_Ca, as well as the activation and steady-state inactivation curves; effects similar to those reported by others for β-adrenergic stimulation (Piacentino et al. 2000; Chen et al. 2002). Also, isoprenaline was without effect when added following calyculin A (Fig. 4A).

Thus, studies that critically examined the importance of PKA were warranted. Accordingly, inhibitors were used that would either prevent the cAMP-induced activation of PKA (Rp-cAMPs, Fig. 4B and C) or directly inhibit its activity (H89, Fig. 4D). Our experiments demonstrated that both of these agents failed to alter either the amplitude of steady-state I_Ca or the calyculin A-induced increase of I_Ca. A recent report by Santana et al. (2002) has suggested that PKA regulates basal I_Ca in mouse ventricular myocytes. However, as this was deduced primarily from an H89-dependent decrease in contraction, it is not clear if H89 decreased I_Ca. In contrast, others report that PKA inhibition has no effect on mammalian ventricular I_Ca. For example, Hirayama & Hartzell (1997) showed that intracellular dialysis with 2mm Rp-cAMPs failed to decrease guinea-pig ventricular I_Ca, and more recently Chen et al. (2002) reported that H89 has no effect on I_Ca in human ventricular myocytes. Thus, the results presented here combined with those of others (Hirayama & Hartzell, 1997; Chen et al. 2002) support the view that constitutive PKA is not important in maintaining steady-state mammalian ventricular I_Ca. However, the case may be different in atrial cells, as Wang & Lipsius (1995) showed that basal I_Ca in feline atrial cells was decreased by both Rp-cAMPs and H89, implicating the importance of cAMP-PKA cascade in these cells in the absence of external signalling.

CaMKII

CaMKII was another protein kinase that required consideration. Blitzer et al. (1998) have shown in brain that inhibition of PP1 is linked to the activation of CaMKII. In heart, CaMKII not only colocalizes with L-type Ca²⁺ channels (Wu et al. 1999; Vinogradova et al. 2000) but activation of the kinase by flash photolysis of caged Ca²⁺ increases I_Ca as well (Anderson et al. 1994). Further, stimulation-dependent facilitation of I_Ca has been attributed to CaMKII activity (Yuan & Bers, 1994; Xiao et al. 1994; Dzhura et al. 2000).

Our initial strategy was to assess the importance of CaMKII using the rapid Ca²⁺ chelator BAPTA. While EGTA can effectively buffer global [Ca²⁺]_i, it is not a sufficiently rapid buffer to prevent accumulation of local Ca²⁺ at the mouth of the channel during Ca²⁺ influx (Naraghi & Neher, 1997; You et al. 1997). In contrast, BAPTA, with its rapid Ca²⁺ binding kinetics, is a more useful buffer for preventing an increase in local [Ca²⁺]_i (Naraghi & Neher, 1997; You et al. 1997). Thus, CaMKII-dependent facilitation of rat ventricular I_Ca is blocked by 10mm intracellular BAPTA but not by 10mm EGTA (Xiao et al. 1994) and BAPTA blocks the CaMKII-mediated effects on I_Ca inactivation and reactivation that promote pacemaker activity in sinoatrial-nodal cells (Vinogradova et al. 2000). In a second, independent approach, Ba²⁺ was used as the charge carrier. While Ba²⁺ can freely permeate through the L-type Ca²⁺ channel, it cannot activate calmodulin and CaMKII (Chao et al. 1984). Like BAPTA, it has been shown to prevent CaMKII-dependent regulation of the L-type Ca²⁺ current (Xiao et al. 1994; Dzhura et al. 2000). Nonetheless, neither intracellular BAPTA nor Ba²⁺ blocked the increase in current through the L-type Ca²⁺ channels in response to calyculin A. While the increase in I_Ca elicited by calyculin A in the presence of BAPTA is comparable to that seen with EGTA (compare Fig. 5A to Figs 3A and 6A and B), it appears as if the effect of calyculin A on Ba²⁺ currents is reduced. However, this is likely to be the result of the reduced ion-dependent inactivation that characterizes Ba²⁺ currents through L-type Ca²⁺ channels (Dzhura et al. 2000) and the fact that current amplitude was calculated as the difference between the peak inward current and the current remaining at the end of the voltage-clamp pulse. Thus, although CaMKII has been linked to Ca²⁺ channel regulation in several settings, the combined results of the BAPTA and Ba²⁺ experiments exclude a role for CaMKII in the regulation of the steady-state level of I_Ca.

PKC

The role for PKC in regulating cardiac I_Ca has been the subject of much investigation, with different effects of PKC activation reported for different species and for different PKC-activating agonists (Kamp & Hell, 2000). It is likely that some of these differences arise from the large number of different isoforms of PKC that exist within cardiac cells as well as from the method of activation of PKC. In the present study, we considered the possibility of a constitutively active PKC, presumably localized near the L-type Ca²⁺ channel, which is part of the kinase–phosphatase balance that we have observed. It is interesting to note that Tseng & Boyden (1991) found that activation of PKC increased I_Ca and left-shifted both activation and steady-state inactivation in canine ventricular myocytes and Purkinje cells, similar to the effect that we observe when background kinase activity is unmasked by inhibition of PP1 with calyculin A (Fig. 3) or inhibitor-2 (Fig. 7). A role for PKC in the steady-state regulation of mouse ventricular L-type Ca current was further supported by our results with two distinct PKC inhibitors, Ro 31-8220 and Gö 6976. Both of these inhibitors were effective in blocking the action of calyculin A on I_Ca (Fig. 6). Although non-specific effects of PKC inhibitors on I_Ca have been reported (Hartzell & Rinderknecht, 1996), neither Ro 31-8220 nor Gö 6976 had any effect on I_Ca when added after calyculin A, arguing against any such non-specific actions of these compounds. In addition, Ro 31-8220 did not modulate the effects of β-adrenergic stimulation, further arguing against any non-specific interactions.

Thus, the results reported in the present study illuminate an I_Ca-activating pathway for PKC in heart. While these results might seem to be in contrast to recent reports identifying an inhibitory role for PKC on L-type current (McHugh et al. 2000; Hu et al. 2000), there are many different forms of PKC and it is likely that distinct isozymes of PKC act at different sites and have opposing regulatory effects on the L-type Ca²⁺ channel. For example, phorbol esters inhibit I_Ca in TSA-201 cells expressing the recombinant rabbit cardiac a_1c subunit (McHugh et al. 2000), yet these agents increase the current when the same protein is expressed in Xenopus oocytes (Shistik et al. 1998). Recent studies link PKCε activation with inhibition of I_Ca in heart cells (Hu et al. 2000; Mochly Rosen et al. 2000) and Thr27 and Thr31 of the a_1c subunit may be the targets of PKC in this pathway (McHugh et al. 2000). However, there are also many reports that reveal a PKC-dependent pathway that increases L-type I_Ca in heart (Tseng & Boyden, 1991; Kelso et al. 1996; Woo & Lee, 1999; Kurata et al. 1999; He et al. 2000). It is important to note that because our experimental protocols do not involve the direct activation of PKC, the results of these experiments cannot be interpreted to either agree or disagree with the reports cited above. Rather, our present results support the view that one or more isoforms of PKC exerts tonic, positive control of ventricular L-type I_Ca. However, the identification of the underlying PKC species and its molecular substrate(s) requires further study.

In summary, the results of the present study support the view that PP1 and an isoform of PKC counteract one another to determine the set-point of the mouse ventricular L-type Ca²⁺ current. It will be important to determine how this system, which maintains a functional reserve of I_Ca, is altered in failing heart. A thorough understanding of this regulatory pathway, which is distinct from receptor-activated signalling, may lead to new clinical options for the enhancement of contractility.