L-type voltage-gated Ca2+ channels (LVGCs) are functionally downregulated in arterial smooth muscle (SM) cells (ASMCs) of mice with SM-specific knockout of Na+/Ca2+ exchanger type-1 (NCX1SM−/−) (32). Here, using activators and inhibitors of protein kinase C (PKC), we explore the regulation of these channels by a PKC-dependent mechanism. In both wild-type (WT) and NCX1SM−/− myocytes, the PKC activator phorbol 12,13-dibutyrate (PDBu) increases LVGC conductance, decreases channel closing rate, and shifts the voltage dependence of channel opening to more negative potentials. Three different PKC inhibitors, bisindolylmaleimide, Ro-31-8220, and PKC 19-31, all decrease LVGC currents in WT myocytes and prevent the PDBu-induced increase in LVGC current. Dialysis of WT ASMCs with activated PKC increases LVGC current and decreases channel closing rate. These results demonstrate that PKC activates LVGCs in ASMCs. The phosphatase inhibitor calyculin A increases LVGC conductance by over 50%, indicating that the level of LVGC activation is a balance between phosphatase and PKC activities. PDBu causes a larger increase in LVGC conductance and a larger shift in voltage dependence in NCX1SM−/− myocytes than in WT myocytes. The inhibition of PKC with PKC 19-31 decreased LVGC conductance by 65% in WT myocytes but by only 37% in NCX1SM−/− myocytes. These results suggest that LVGCs are in a state of low PKC-induced phosphorylation in NCX1SM−/− myocytes. We conclude that in NCX1SM−/− myocytes, reduced Ca2+ entry via NCX1 lowers cytosolic [Ca2+], thereby reducing PKC activation that lowers LVGC activation.
- sodium/calcium exchanger type-1 knockdown
- calcium channels
l-type voltage-gated Ca2+ channels (LVGCs) play an important role in the maintenance of myogenic tone (MT) in arterial smooth muscle (SM) cells (ASMCs). During the myogenic response, an increase in intralumenal pressure in small resistance arteries depolarizes the ASMCs (8a; 14). The depolarization is believed to be caused by the opening of stretch-activated channels, such as transient receptor potential melastatin 4 channels (7) or transient receptor potential canonical 6 channels (27), or the closure of K+ channels (25). This depolarization opens LVGCs, and Ca2+ influx through these open channels causes a rise in the cytosolic Ca2+ concentration ([Ca2+]Cyt), contractile activation of ASMCs, and arterial constriction. The LVGCs are critical to the maintenance of MT (5, 10). In fact, when LVGCs are blocked by dihydropyridines, MT decreases by 80–90% (30). Na+/Ca2+ exchanger type-1 (NCX1)-mediated Ca2+ influx may be responsible for much of the remaining MT because the blockade of NCX reduces MT by about 15% (32).
In many SM preparations, protein kinase C (PKC) increases LVGC currents (2, 3, 18). Channel activation may occur as a result of the phosphorylation of residues on the NH2-terminus of the α-subunit, which relieves a tonic inhibitory control on channel activation (24). There are nine PKC genes; they code for at least 11 PKC isozymes that can be classified into three groups (17). Conventional PKCs (PKCα, PKCβI, PKCβII, and PKCγ) are Ca2+ dependent and are activated by diacylglycerol (DAG). Novel PKCs (PKCδ, PKCε, PKCη, PKCθ, and PKCσ) are regulated by DAG but are Ca2+ independent. Atypical PKCs (PKCλ and PKCζ) are Ca2+ and DAG independent. The predominant PKC isozymes expressed in vascular SM are PKCα and PKCβ (29). Recent evidence suggests that PKCα is the isoform necessary for Ca2+ sparklet activity in ASMCs (18).
The accompanying report (32) reveals that the activation of ASMC LVGCs was significantly reduced in SM-specific NCX1 knockout (NCX1SM−/−) mice. Similar results were previously described in cardiac-specific NCX1 knockout mice, where the reduced LVGC current is due to faster and more complete Ca2+-dependent inactivation (9, 20). In contrast, in ASMCs, the decreased LVGC conductance is not due to an increase in channel inactivation or to a decreased expression of the channels. In NCX1SM−/− mouse ASMCs, reduced Ca2+ entry via NCX1 lowers [Ca2+]Cyt and thereby reduces MT and blood pressure (BP) (32). Because the phosphorylation of LVGCs by PKC can activate these voltage-gated channels, it is possible that the decreased Ca2+-dependent phosphorylation of LVGCs in NCX1SM−/− ASMCs might explain the reduced LVGC conductance in myocytes from these animals.
To explore the role of PKC in the reduction of LVGC currents in NCX1SM−/− myocytes, we applied several PKC activators and inhibitors. Using either a known activator of PKC, phorbol 12,13-dibutyrate (PDBu), or activated PKC itself, we found that the activation of PKC dramatically increased LVGC currents. Moreover, this effect was much greater in NCX1SM−/− myocytes than in wild-type (WT) control myocytes. In addition, several PKC inhibitors decreased LVGC currents and prevented the ability of PDBu to enhance the currents. The ability of PKC inhibitors to decrease LVGC currents was greater in WT myocytes than in NCX1SM−/− myocytes. These results are consistent with the hypothesis that LVGCs in NCX1SM−/− myocytes are in a state of low activation by PKC.
All mouse protocols were approved by the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine.
Solutions, myocyte isolation, patch-clamp recording, and data analysis.
All methods are described in the accompanying report (32).
Reagents and sources were as follows: nifedipine, collagenase, elastase, and BSA (Sigma-Aldrich, St. Louis, MO); TTX, Bay K 8644, and activated PKC (Calbiochem, San Diego, CA); and bisindolylmaleimide (BIM 1), Ro-31-8220, calyculin A, and PKC 19-31 (Biomol, Enzo Life Sciences, Farmingdale, NY). The activated PKC contained a mixture of all the PKC isoforms found in the rat brain. Other reagents were reagent grade or the highest grade available. TTX was dissolved in deionized water; nifedipine was dissolved in DMSO.
PDBu increases LVGC current in WT mouse myocytes.
PKC has been reported to activate LVGCs in various SM preparations (2, 3). As a first step in this study, we tested whether LVGCs in mesenteric artery myocytes from WT mice can be activated by PKC. Tumor-promoting phorbol esters such as PDBu activate a variety of PKC isozymes (1). In isolated ASMCs, PDBu increases Ba2+ current (Fig. 1, A and B) as well as the maximum conductance of LVGCs (Fig. 1C). The maximum LVGC current, at +10 mV, was increased by 70% in the presence of 200 nM PDBu, and the maximum conductance was increased by 68%. Furthermore, the kinetics of decline of the LVGC tail currents were markedly slowed by PDBu (Fig. 1A), indicating a reduction in channel closing rate. The additional inward current that was induced by 200 nM PDBu was blocked by 1 μM nifedipine, indicating that the extra current flows through LVGCs (Fig. 1D).
PKC inhibition reduces LVGC current and prevents activation by PDBu.
If PDBu increases LVGC activity by activating PKC, as the preceding data suggest, it should be possible to block this effect of PDBu with a PKC inhibitor. A variety of inhibitors are available with different isozyme selectivities. We first tested a pair of drugs with relatively broad isozyme specificity: BIM 1 and Ro-31-8220. These agents, which probably block PKC activity by competitively inhibiting ATP binding (28), had similar effects on the LVGCs. In the absence of PDBu, Ro-31-8220 (1 μM) decreased the maximum LVGC current by 28% (Fig. 2A,a) and the maximum LVGC conductance by 26% (Fig. 2A,b). In addition, Ro-31-8220 prevented the increase in LVGC activity caused by PDBu (Fig. 2A, a and b). BIM 1 (1 μM) had very similar effects: it decreased the maximum LVGC current by 28% and conductance by 35%, and it prevented PDBu activation of LVGCs (Fig. 2B, a and b).
PKC pseudosubstrate peptide inhibitors such as PKC 19-31 are derived from the regulatory C1 domain of PKC, which inhibits enzyme activity (11). PKC 19-31 selectively inhibits PKCα and -β. Because this peptide is membrane impermeable, we applied the inhibitor by adding it to the pipette (intracellular) solution. After about 3 min of whole cell dialysis with 1 μM PKC 19-31 in the pipette, the LVGC current declined to about half of its maximum value (Fig. 3A). A subsequent addition of 200 nM PDBu resulted in only a small increase in LVGC current (Fig. 3A). In contrast, in the absence of intracellular PKC 19-31, PDBu induced a robust increase in LVGC current after 8 min of dialysis (Fig. 3B). Summarized data on the effects of PDBu on Ba2+ current in WT artery myocytes in the absence and presence of PKC 19-31 (right- and left-hand sets of bars, respectively) are shown in Fig. 3C. Thus, in the experiment of Fig. 3A, it is the PKC 19-31 and not the extended period of cell dialysis that prevented PDBu activation of LVGCs.
The fact that these three different PKC inhibitors, BIM 1, Ro-31-8220, and PKC 19-31, all prevent the PDBu-induced increase in LVGC current implies that PDBu increases LVGC activation by activating PKC. The activated PKC, in turn, phosphorylates and thereby activates LVGCs. Because of the selectivity of PKC 19-31 for PKCα and PKCβ, one of these two Ca2+-dependent PKC isozymes is most likely responsible for the PKC-mediated regulation of LVGCs in mesenteric ASMCs. PKCα is required for basal Ca2+ sparklet activity in ASMCs (18).
Intracellular dialysis with activated PKC increases LVGC current.
The aforementioned results demonstrate that the activation of PKC enhances LVGC activity in ASMCs. In that case, the direct addition of activated PKC to the cytoplasm should also increase LVGC conductance. To perform this experiment, we added activated PKC to the pipette solution to dialyze the cell with PKC during whole cell recording. Indeed, LVGC current increased significantly after several minutes of dialysis with PKC in the pipette (Fig. 4A). In addition, the kinetics of decline of the LVGC tail current were markedly slowed (Fig. 4B) and the voltage-dependence of channel activation was shifted to more negative voltages (Fig. 4C) as was the case following PDBu activation of PKC (Fig. 1). Thus LVGC currents can be increased either by activating PKC with PDBu or by dialyzing the cell interior with activated PKC.
PDBu eliminates the difference in LVGC activation between NCX1SM−/− and WT myocytes.
Available evidence implies that [Ca2+]Cyt is low in NCX1SM−/− myocytes (32). Therefore, a possible explanation for the reduced LVGC activation in these cells is that Ca2+-dependent PKC activity is lower, and fewer LVGCs are phosphorylated, than in WT myocytes. To test this hypothesis, we compared the effect of PDBu activation of PKC in WT and NCX1SM−/− myocytes. PDBu increased both the LVGC current and the maximum conductance in both WT and NCX1SM−/− myocytes, but it had a much greater effect in NCX1SM−/− myocytes (Fig. 5). In WT myocytes, the maximum conductance was increased by 68%, whereas in the NCX1SM−/− myocytes it was increased by 178%. In fact, in the presence of PDBu, there were no significant differences between either the current-voltage curves (Fig. 5A) or the conductance-voltage (G-V) curves (Fig. 5B) in WT and in NCX1SM−/− myocytes. These results are consistent with the idea that the LVGC conductance is smaller in NCX1SM−/− myocytes because fewer channels are phosphorylated by PKC. When PKC is activated by PDBu, the maximum LVGC conductance is the same in WT and NCX1SM−/− myocytes (Fig. 5B), presumably because the channels are maximally phosphorylated. These whole cell current data are consistent with the results of LVGC activation by Bay K 8644 (32) and further support the view that NCX1SM−/− myocytes express a normal number of LVGCs.
PDBu shifts LVGC activation to more negative voltages.
We have seen that the activation of LVGCs by PKC dramatically slows channel closing (Figs. 1A and 4). Consistent with this kinetic effect is the fact that channel activation appears to be shifted to more negative voltages by PDBu (Fig. 5B). To characterize LVGC voltage dependence, we fit a Boltzman distribution to the LVGC G-V curves (Fig. 6). In WT myocytes, PDBu shifted the Vm for half-maximal activation (V0.5) by about 13 mV, from +3.4 in control to −9.5 mV in PDBu (Fig. 6A). In contrast, in NCX1SM−/− myocytes, V0.5 was shifted by about 21 mV: from +9.8 in control to −11.1 mV in PDBu (Fig. 6B). These results are consistent with the idea that PKC phosphorylation of LVGCs activates the channels and shifts their activation to more negative voltages. In NCX1SM−/− myocytes, LVGCs activate at relatively positive voltages (V0.5 = +9.8 mV) because their phosphorylation state is low. After maximal activation of PKC with PDBu, more LVGCs become phosphorylated, and the G-V curves are shifted to a similar, more negative potential in both WT (V0.5 = −9.5 mV) and NCX1SM−/− (V0.5 = −11.1 mV) myocytes.
PKC 19-31 reduces LVGC current in both NCX1SM−/− and control myocytes.
Because PKC activation with PDBu had a larger effect in NCX1SM−/− than in WT myocytes (Fig. 5), we also compared the effect of PKC inhibitors on ASMCs from these two mouse genotypes. The inhibition of PKC with PKC 19-31 reduced LVGC conductance in WT myocytes by about 67%, but by only 37% in NCX1SM−/− myocytes (Fig. 7). Indeed, in the presence of PKC 19-31, both the current-voltage curves and the G-V curves of NCX1SM−/− and control myocytes were virtually superimposable. This result, too, suggests that LVGC activity was low in NCX1SM−/− myocytes because few of the channels were phosphorylated by PKC.
Calyculin A, phosphatase inhibitor, increases LVGC current.
The serine/threonine phosphatases PP1 and PP2A influence LVGC activity in cardiac and vascular SM cells (6, 18, 21). This raises the possibility that PKC and the phosphatases have opposing effects on LVGC activation. Indeed, the fact that PKC inhibition decreases LVGC activity in mesenteric artery myocytes (Figs. 2 and 3) implies that phosphatase activity is present in these cells. Therefore, we investigated the effect of the PP1/PP2A inhibitor calyculin A (12) on LVGCs. In WT myocytes, 20 nM calyculin A increased the maximum LVGC current and conductance by over 50% (Fig. 8). This is evidence that the prevailing level of activation of the LVGCs is the result of a balance between PKC and phosphatase activities.
In the preceding report (32), we noted that LVGC activity was markedly reduced in ASMCs from NCX1SM−/− mice. The LVGCs are also downregulated in cardiac-specific NCX1 knockout mice (9, 20), but the mechanism of the current decline is apparently different in ASMCs. In the heart, Ca2+-dependent inactivation of LVGCs is faster and more complete in the NCX1 knockout mouse myocytes than in control myocytes, resulting in reduced LVGC current. Also in the heart, the reduced Ca2+ entry through LVGCs compensates for the reduced ability of NCX1 to extrude Ca2+ and, as a result, cardiac contractions are nearly normal in the cardiac-specific NCX1 knockout animals. There is no evidence that increased inactivation is responsible for the reduced LVGC current in ASMCs.
PKC upregulates LVGCs in arterial SM.
The PKC agonist PDBu markedly increased LVGC activity in WT ASMCs (Fig. 1). The maximum current and conductance were both increased by about 70%. In addition, the LVGC closing rate was decreased (Fig. 1A) and the voltage dependence of channel activation was shifted to more negative membrane potentials (Fig. 6). Three different inhibitors of PKC, BIM 1, Ro-31-8220, and PKC 19-31, all decreased LVGC activation and prevented the PDBu-induced increase in LVGC activity, indicating that PDBu enhances channel activity by activating PKC. Several previous reports have suggested that PKC activates LVGCs in SM (2, 3, 8, 19, 26). The stimulation of LVGCs by PKC is often followed by inhibition, both in cardiac and SM myocytes (16, 23). We did not look for secondary inhibition of LVGCs because it was difficult to hold these myocytes in the whole cell configuration. Furthermore, the long-term decay of LVGC currents is caused, in part, by channel rundown, which may be independent of PKC.
Recently, a new mechanism of Ca2+ entry through L-type Ca2+ channels was described, in which single channels or clusters of channels open with high probability and create sites of sustained Ca2+ influx, called “persistent Ca2+ sparklet sites” (19). PKCα is required for basal persistent Ca2+ sparklet activity (18), and PKC activation increases the activity of Ca2+ sparklet sites (19). These findings suggested that PKC increases Ca2+ influx by causing additional channels or clusters of channels to operate in the “persistent” mode. These persistent channels gate differently than the average LVGC: the open probability of a persistent LVGC is as much as 1.8 × 107 times higher than the average LVGC at −70 mV (19). The results described in this report demonstrate that PKC activation has additional effects on the overall population of LVGCs. For example, the maximum LVGC conductance is increased (Figs. 1 and 5), which indicates that the number of voltage-activatable LVGCs is increased by PKC activation. In addition, the kinetic and steady-state properties of the channels are altered: LVGC closing rate is decreased (Figs. 1 and 4), and the activation voltage range is shifted to more negative voltages (Fig. 6). In atrial myocytes, the inhibition of PKC shifts LVGC voltage dependence to more positive potentials (15), which is consistent with our results. Thus PKC activation has multiple effects on LVGCs in vascular SM. It increases the activity of persistent Ca2+ channels that have a high open probability at negative membrane potentials, and it activates the average LVGC by increasing its open probability and shifting its voltage dependence.
PKC activation has a larger effect and PKC inhibition a smaller effect in NCX1SM−/− myocytes than in WT myocytes.
The small LVGC currents in NCX1SM−/− myocytes raised the possibility that the low BP in NCX1SM−/− mice and reduced MT in their myocytes (32) might be attributable primarily to the reduced Ca2+ entry through the LVGCs. This apparently is not the case, however. The reduction in LVGC currents appears to be a consequence of the reduction in [Ca2+]Cyt that occurs as a result of lower Ca2+ entry via NCX1 and not vice versa. The evidence presented in this article suggests that LVGC activation is reduced in NCX1SM−/− myocytes because of a reduced phosphorylation of LVGCs by conventional, Ca2+-dependent PKC. The evidence is as follows. The PKC agonist PDBu has a larger stimulatory effect on LVGCs in NCX1SM−/− myocytes than it does in WT cells. Since the LVGCs in NCX1SM−/− myocytes can be activated to a greater extent by PKC than can WT myocyte channels, they must be less phosphorylated than WT channels. Conversely, the inhibition of PKC by PKC 19-31 decreases LVGC conductance to a greater extent in WT myocytes than in NCX1SM−/− myocytes. This must mean that LVGCs in NCX1SM−/− myocytes are already in a state of low phosphorylation; therefore, a further inhibition of PKC has a relatively small effect.
The implication that the reduced PKC activation in pressurized NCX1SM−/− arteries is the result of a low [Ca2+]Cyt raises the question of the Ca2+ dependence of PKC (presumably the α-isoform; see results). The relationship between [Ca2+]Cyt and PKCα activation is not known for in vivo conditions; nevertheless, PKCα must be regulated by [Ca2+]Cyt within the dynamic physiological range (22), which is about 100–300 nM [Ca2+] (14, 31). Ca2+ binding to the C2 domain of PKCα activates the kinase and targets it to the plasma membrane. The Ca2+ affinity of the C2 domain of PKCα is highly sensitive to lipids, such as phosphatidylinositol 4,5-bisphosphate. For example, in vitro studies using a protein-to-membrane Förster resonance energy transfer assay showed that the Ca2+ concentration required for PKCα-C2 to dock to membranes decreased dramatically when the membrane lipids included 6% phosphatidylinositol 4,5-bisphosphate (4). In vivo, the Ca2+ affinity of PKCα may be increased to an even greater extent by the specific lipid microenvironment present at the inner surface of the plasma membrane. It seems reasonable, therefore, that a decrease in [Ca2+]Cyt of about 25–50 nM, which is the magnitude of change anticipated on the basis of the reduced myogenic reactivity of NCX1SM−/− arteries (32), could significantly reduce PKCα activity.
A comparison of LVGC voltage dependence under various conditions (Table 1) supports the idea that channel phosphorylation activates the channels and shifts their voltage dependence to more negative voltages. All of the following interventions, which should promote channel phosphorylation, activate LVGCs: activation of PKC with PDBu in both WT and NCX1SM−/− myocytes, dialysis of the cell interior with activated PKC, and inhibition of phosphatase with calyculin A. In all of these cases, the maximum tail-current amplitude was increased to about 40 pA/pF and the V0.5 was shifted to approximately −10 mV (range, −8.3 to −11.6 mV; Table 1). The PKC inhibitors all decrease the maximum tail-current amplitude, and the V0.5 values are more positive (range, −4.7 to +5.5 mV; Table 1).
These considerations lead us to conclude that the reduction of LVGC activation in NCX1SM−/− myocytes is not a direct consequence of reduced NCX1 expression and that it does not directly cause the reduced MT and BP in these animals. NCX1 normally operates in Ca2+ entry mode (13, 32). Consequently, NCX1 knockout should reduce [Ca2+]Cyt. The activation of conventional (Ca2+ dependent) PKC, presumably PKCα, therefore should be reduced. This, in turn, should reduce the phosphorylation of LVGCs and thereby decrease their activity. In other words, the reduced LVGC activity must be a consequence and not the primary cause of the low [Ca2+]Cyt. Nevertheless, the depressed LVGC activity in NCX1SM−/− myocytes is likely a compensation for the reduced NCX1 activity that helps to avoid Ca2+ overload. The low LVGC activity also helps to keep [Ca2+]Cyt low and therefore must help to maintain the low MT and low BP.
This work was supported by a National Scientist Development grant from the American Heart Association (to J. Zhang), an International Society of Hypertension-Pfizer award (to J. Zhang), and National Institutes of Health grants (to M. P. Blaustein, K. D. Philipson, and M. I. Kotlikoff).
No conflicts of interest are declared by the author(s).
We thank Fernando Santana for very helpful discussion, Stephen P. Kinsey and Michelle Izuka for technical support and transgenic mouse genotyping, and Katherine Frankel for assistance with the manuscript.
Present address of C. Ren: Neuroscience Program, Otolaryngology/Dept. of Surgery, University of Utah, Salt Lake City, UT.
- Copyright © 2010 the American Physiological Society