Regulation of L-type Ca²⁺ channels in rabbit portal vein by G protein α_s and βγ subunits

Abstract

1. The effect of purified G protein subunits α_s and βγ on L-type Ca²⁺ channels in vascular smooth muscle and the possible pathways involved were investigated using freshly isolated smooth muscle cells from rabbit portal vein and the whole-cell patch clamp technique.

2. Cells dialysed with either Gα_s or Gβγ exhibited significant increases in peak Ba²⁺ current (I_Ba) density (148 % and 131 %, respectively) compared with control cells. The combination of Gα_s and Gβγ further increased peak I_Ba density (181 %). Inactive Gα_s and Gβγ did not have any effect on Ca²⁺ channels.

3. The stimulatory effect of Gα_s on peak I_Ba was entirely abolished by the protein kinase A inhibitor Rp-8-Br-cAMPS, or the adenylyl cyclase inhibitor SQ 22536. On the other hand, the stimulatory response of Ca²⁺ channels to Gβγ was not affected by the protein kinase A inhibitors Rp-8-Br-cAMPS and KT 5720, or by the Ca²⁺-dependent protein kinase C inhibitor bisindolylmaleimide 1, but was completely blocked by the protein kinase C inhibitor calphostin C. Pretreatment of cells with phorbol 12-myristate 13-acetate for over 18 h prevented the stimulatory effect of Gβγ on peak I_Ba. In addition, acute application of phorbol 12,13-dibutyrate enhanced peak I_Ba density in control cells, which could be entirely blocked by calphostin C.

4. These data indicate that enhancement of Ba²⁺ currents by Gα_s and Gβγ can be attributed to increased activity of protein kinase A and protein kinase C, respectively. No direct membrane-delimited pathway for Ca²⁺ channel regulation by activated G_s proteins could be detected in vascular smooth muscle cells.

The exact mechanisms by which voltage-dependent Ca²⁺ channels (L-type) are modulated by β-adrenergic stimulation in vascular smooth muscle (VSM) remain controversial. For instance, Sperelakis and co-workers (Xiong et al. 1994a,b) observed in rabbit portal vein cells that 10 μM isoprenaline (isoproterenol) exerts a dual effect on L-type Ca²⁺ channels: a transient increase followed by a decrease in channel activity. They also reported (Xiong & Sperelakis, 1995) that the α_s subunit of G protein increased Ca²⁺ currents in rabbit portal vein smooth muscle cells and concluded that the stimulatory effect of β-adrenergic receptor activation may involve a direct membrane-delimited modulation of L-type Ca²⁺ channels by the activated G proteins whereas the inhibitory effect was attributed to G_s activation of adenylate cyclases and subsequent phosphorylation of the channel by protein kinase A (PKA). On the other hand, previous studies from our laboratory suggest that stimulation of the cAMP-PKA pathway causes enhancement of Ca²⁺ channel activity in rabbit portal vein smooth muscle cells, while higher levels of cAMP may lead to a cross-over activation of protein kinase G (PKG), which then leads to inhibition of Ca²⁺ channel activity (Ishikawa et al. 1993; Ruiz-Velasco et al. 1998). In addition, reports from other researchers also support a stimulatory effect of the cAMP-PKA pathway on L-type Ca²⁺ channels in VSM cells (e.g. Fukumitsu et al. 1990; Loirand et al. 1992; Tewari & Simard, 1994; Shi & Cox, 1995; Farrugia, 1997). However, there is presently little evidence available to support a direct modulation of VSM L-type Ca²⁺ channels by G_s proteins.

In their inactive state, G proteins are membrane-associated heterotrimers composed of α, β and γ subunits with GDP bound to the α subunits. Upon dissociation of α subunits from βγ dimers by exchange of GTP for GDP, both GTP-bound α subunits and βγ dimers are activated and interact with their effectors such as adenylyl cyclases and ion channels (Hepler & Gilman, 1992). Although it is well established that α subunits of G_s protein play an important role in the regulation of L-type Ca²⁺ channels, there is no direct evidence for modulation of L-type Ca²⁺ channels by βγ subunits of G proteins. Furthermore, the role of G protein subunits in the regulation of VSM L-type Ca²⁺ channels has not yet been examined in any detail. In the present study, we investigated the effects of purified α_s and βγ subunits of G proteins on L-type Ca²⁺ channels in isolated rabbit portal vein smooth muscle cells. In addition, we examined whether there is a direct membrane-delimited effect of these subunits, independent of intracellular messengers, on Ca²⁺ channels in vascular smooth muscle cells.

METHODS

Isolation of rabbit portal vein myocytes

Myocytes were isolated using the methods reported previously (Ruiz-Velasco et al. 1998) with modification. Male albino rabbits (1.5-2.0 kg) were killed with an intravenous overdose of sodium pentobarbital (50 mg kg⁻¹). The portal vein was rapidly removed and cleaned of connective tissue in ice-cold Krebs solution (mM): 125 NaCl, 4.2 KCl, 1.2 MgCl₂, 1.8 CaCl₂, 11 glucose, 1.2 K₂HPO₄, 23.8 NaHCO₃ and 11 Hepes; pH 7.4 with Trizma base. The portal vein was then cut into small segments (∼4 mm × 4 mm) and pre-incubated for 30 min in a shaking water-bath at 35°C in a dispersion solution (enzyme-free, mM): 90 NaCl, 1.2 MgCl₂, 1.2 K₂HPO₄, 20 glucose, 50 taurine and 5 Hepes; pH 7.1 with NaOH. Following pre-incubation, the segments were incubated in the dispersion solution containing 2 mg ml⁻¹ collagenase Type I (Sigma), 0.5 mg ml⁻¹ protease Type XXVII (Sigma) and 2 mg ml⁻¹ bovine serum albumin (BSA; Sigma) for 10-14 min at 35°C, and then rinsed 4 times with the enzyme-free dispersion solution. Smooth muscle cells were dispersed by gentle trituration of the segments with a wide-tipped fire-polished Pasteur pipette. The cell suspension was stored in the enzyme-free dispersion solution containing BSA (1 mg ml⁻¹) and Ca²⁺ (0.1 mM) at 4°C and used within 10 h. The animal use protocol was reviewed and approved by the Animal Care and Use Committee of the University of Nevada.

Electrophysiology

Ba²⁺ currents (I_Ba) in portal vein smooth muscle cells were measured using the whole-cell patch clamp. Previous studies from this laboratory have demonstrated that the inward Ba²⁺ currents measured from rabbit portal vein myocytes were completely blocked by 10 μM nicardipine, suggesting the presence of predominantly L-type Ca²⁺ channels in these cells (Ishikawa et al. 1993). A drop of cell suspension was added to a small recording chamber mounted on the stage of an inverted microscope (Nikon, Japan). The cells in the chamber were superfused by gravity at a constant rate (∼1-2 ml min⁻¹) and the complete exchange of the superfusate in the recording chamber required about 1 min. All the experiments were performed at room temperature (20-22°C). Inward currents were measured using an Axopatch-1D patch-clamp amplifier (Axon Instruments). Patch electrodes were made from borosilicate glass pulled with a Sutter P80-PC Flaming-Brown micropipette horizontal puller and fire-polished with an MF-83 Narishige microforge. Pipette resistance was 3-5 MΩ when filled with the pipette solution. After establishing the whole-cell configuration, cell membrane capacitance and series resistance were determined using a 20 mV hyperpolarizing pulse and were partially compensated. Inward current was elicited by stepping voltage from a holding potential of -70 mV to 0 mV at 30 s intervals. Voltage clamp protocols were applied to the cells using the data acquisition package pCLAMP 6 (Axon Instruments) and filtered at 2 kHz (-3 dB). Data analysis was performed using the pCLAMP 6 software package.

The bath solution used to record I_Ba in portal vein cells was composed of (mM): 117.5 NaCl, 10 tetraethylammonium chloride (TEACl), 5 BaCl₂, 0.5 MgCl₂, 5.5 glucose, 5 CsCl and 10 Hepes; pH 7.40 with NaOH. Both TEACl and CsCl are used to block K⁺ currents. The pipette solution consisted of (mM): 75 glutamic acid, 55 CsCl, 1 K₂HPO₄, 5 glucose, 5.7 MgSO₄, 5 ATP, 10 EGTA and 10 Hepes; pH 7.2 with CsOH. GTP was intentionally omitted from the pipette solution, unless otherwise stated, to avoid possible activation of endogenous G proteins and production of cGMP. The osmolality of the solutions (external and internal) was measured and maintained between 290 and 300 mosmol (kg H₂O)⁻¹.

Purification of G protein subunits

The α_s subunit of the G protein, Gα_s, was purified from E. coli as described in detail (Lee et al. 1994) and activated by incubation with 50 mM NaHepes (pH 8.0), 10 mM MgSO₄, 1 mM EDTA, 2 mM dithiothreitol (DTT) and 400 μM GTPγS at 30°C for 30 min. Free GTPγS was removed by gel filtration. After purification, Gα_s was kept at -70°C in a solution of composition (mM): 20 Hepes, 1 EDTA, 2 DTT and 5 MgSO₄ until use. The recombinant subunits β1γ2 and non-prenylated β1γ2 Cys⁶⁸ to Ser were purified from Sf9 cells (Kozasa & Gilman, 1995). These βγ subunits were stored at -70°C in a solution of composition (mM): 20 Hepes, 2 DTT, 50 NaCl, 11.4 3-((3-cholamidopropyl)-dimethylammonio)-1-propanesulphonate (CHAPS). The final concentration of CHAPS in the pipette solution during experiments was 20 μM, which alone did not have any effect on peak Ba²⁺ current.

Drugs

Isoprenaline (Iso), phorbol 12-myristate 13-acetate (PMA), phorbol 12,13-dibutyrate (PDBu) and all chemicals were purchased from Sigma. 8-Bromoadenosine-3′,5′-monophosphorothioate R_P-isomer (Rp-8-Br-cAMPS) was obtained from Biolog Life Science Institute (La Jolla, CA, USA). KT 5720, SQ 22536, calphostin C and bisindolylmaleimide 1 (BIM) were from Calbiochem (La Jolla, CA, USA). Drugs insoluble in water were first dissolved in dimethylsulphoxide (DMSO) and were then further diluted in the solution with the final concentration of DMSO less than 0.2 %. DMSO alone at a concentration of 0.2 % had no effect on I_Ba.

Data analysis

G protein subunits were included in the patch pipette and dialysed intracellularly. The effects of these compounds on I_Ba were assessed by applying repetitive voltage clamp steps to 0 mV from a holding potential of -70 mV. Time-dependent effects were compared with matched control cells using identical voltage clamp protocols. All experimental values are presented as means ±s.e.m., and n refers to the number of cells tested. Differences between the values from different groups were compared using both Student's paired and unpaired t tests, and two-way analysis of variance where appropriate. P values of less than 0.05 were considered significantly different.

RESULTS

Effect of Gα_s and Gβγ on peak I_Ba

To characterize the effect of G_s protein subunits on the L-type Ca²⁺ channel in smooth muscle cells from rabbit portal vein, Ba²⁺ currents were recorded from freshly isolated cells dialysed with pipette solution containing 50 nM of either active G protein subunits, Gα_s-GTPγS (Gα_s) and Gβ1γ2 (Gβγ), or relatively inactive subunits, Gα_s-GDPβS (Gα_si) and non-prenylated Gβ1γ2C68S (Gβγ_i). Another set of cells was dialysed with pipette solution containing no added G protein subunits, which served as controls. After establishment of the whole-cell configuration, I_Ba was elicited by stepping the voltage to 0 mV from a holding potential of -70 mV. In control cells, peak I_Ba reached a steady state at ∼5 min, but cells from other groups took a longer time (6-10 min) to reach a steady state (Fig. 1A). All the measurements of peak currents were determined when peak I_Ba reached a steady state (5-10 min after whole-cell configuration). As shown in Fig. 1B and C, peak I_Ba densities from cells dialysed with active Gα_s or Gβγ were significantly higher compared with that of control cells. The combination of Gα_s and Gβγ produced an even larger increase in peak current density. The percentage increase of peak I_Ba density from cells dialysed with Gα_s and Gβγ alone was 48 and 31 %, respectively, compared with control cells. In cells dialysed with both Gα_s and Gβγ, peak current density was 81 % greater than their time-matched control cells, indicating an additive effect of Gα_s and Gβγ on Ca²⁺ channel activity (Fig. 1B and C). The higher peak I_Ba density in cells dialysed with Gα_s and Gβγ was not due to possible differences in osmolarity of the pipette solution because dialysis of Gα_s-GDPβS (Gα_si) and Gβ1γ2C68S (Gβγ_i) had no significant effect on peak I_Ba density compared with control cells. In another set of experiments, when I_Ba was elicited from a holding potential of -40 mV, peak I_Ba density in cells dialysed with either Gα_s or Gβγ was still significantly higher than that in control cells (Fig. 1D).

Figure 1. Effects of Gα_s and Gβγ on I_Ba in rabbit portal vein myocytes.

A, representative recordings from a control cell (left), a cell dialysed with Gα_s (50 nM, centre), and a cell dialysed with Gβγ (50 nM, right) at 1 min (a), 5 min (b) and 15 min (c) after establishment of the whole-cell configuration. Currents were elicited by stepping the potential to 0 mV from a holding potential of -70 mV. B, time dependence of peak Ba²⁺ current density. Peak current density was calculated by measuring the current at 10 ms after the initiation of the command pulse and dividing by the cell capacitance. Each symbol represents one cell recording from each group. Gα_si and Gβγ_i represent Gα_s-GDPβS and Gβ1γ2C68S, respectively. C, averaged peak I_Ba density for cells from different groups with a holding potential of -70 mV. Bars represent values of mean ±s.e.m. for control cells (Con; n = 42), and cells dialysed with 50 nM of Gα_s (n = 38), Gα_si (n = 10), Gβγ (n = 45), Gβγ_i (n = 11), or a combination of Gα_s and Gβγ (n = 9). D, averaged peak I_Ba density measured with a holding potential of -40 mV. Bars represent values of mean ±s.e.m. for control cells (n = 7), and cells dialysed with 50 nM of either Gα_s (n = 7), or Gβγ (n = 7). * Significantly different from control value with P < 0.05.

The current-voltage (I-V) relationships from Gα_s- and Gβγ-treated cells exhibited similar patterns to those from the control cells, although the voltage dependence of the I-V relationship from cells dialysed with Gβγ was shifted to slightly more negative potentials. Figure 2A shows the representative recordings of one cell from each group with the test potentials ranging between -60 and +60 mV from a holding potential of -70 mV. Averaged peak current density was significantly higher at test pulse potentials between -20 and +30 mV for cells dialysed with Gα_s, and between -20 and +10 mV for cells dialysed with Gβγ (Fig. 2B). We also examined the effects of G protein subunits on steady-state inactivation of I_Ba, measured using a two-pulse protocol. The membrane potential of cells from different groups was held at -70 mV. A conditioning prepulse ranging from -60 to +40 mV in 10 mV increments was applied for 300 ms, followed by a test pulse to 0 mV for 200 ms. The two pulses were separated by an interpulse resting interval of 5 ms. A representative recording from a cell dialysed with Gα_s using the two-pulse protocol is shown in Fig. 2C. Increasing the potential of the conditioning prepulse reduced I_Ba elicited by the following test pulse in all three groups of cells (Fig. 2D). Relative availability of peak I_Ba (peak I_Ba/peak I_Ba,max) versus the prepulse potential was fitted using the Boltzmann equation: I/I_max=[1 + exp(V - V_0.5)/k]⁻¹, where V_0.5 is the voltage producing half-maximal inactivation, k is the slope of the curve, I is the peak I_Ba measured after different prepulses, and I_max is the peak I_Ba measured with a prepulse of -60 mV. Neither V_0.5 nor k of the curves in cells dialysed with Gα_s (-17.6 ± 0.6 mV, 7.6 ± 0.6) or with Gβγ (-19.3 ± 0.9 mV, 8.4 ± 0.8) were significantly different from those values in control cells (-18.4 ± 0.5 mV, 8.0 ± 0.5).

Figure 2. Voltage dependence of peak I_Ba.

A, sample traces of current recording from cells dialysed without G protein subunits (left), with 50 nM Gα_s (centre) or 50 nM Gβγ (right). The membrane potential was held at -70 mV and the command potentials were stepped between -60 and +60 mV with increments of 10 mV. Pulses were applied every 20 s for 200 ms. B, current density-voltage relationship for cells dialysed with Gα_s (n = 16), Gβγ (n = 17), or without any subunits (control, n = 15). C, a representative recording of I_Ba inactivation in a cell dialysed with 50 nM Gα_s. The conditioning prepulses ranged from -60 to +40 mV. The test pulses were stepped to 0 mV from the resting potential of -70 mV during the brief resting interval of 5 ms. D, relative peak I_Ba-prepulse potential relationship for cells dialysed with Gα_s (n = 8), Gβγ (n = 7), or without G protein subunits (n = 7). The relative peak I_Ba was determined by dividing the peak I_Ba measured following each prepulse by the peak I_Ba measured with a prepulse of -60 mV in the same cell. * Significantly different from control value (P < 0.05).

Effect of Rp-8-Br-cAMPS and SQ 22536 on Gα_s-stimulated peak Ba²⁺ currents

Modulation of L-type Ca²⁺ channels in vascular smooth muscle cells by β-adrenergic receptor activation is believed to involve the cAMP-PKA pathway through phosphorylation of channel subunits by protein kinase A (Ishikawa et al. 1993) and/or possible direct modulation of the channel activity by α subunits of G_s protein (Xiong et al. 1995). To evaluate if there is a direct membrane-delimited effect of Gα_s on Ca²⁺ channel activity, we first tested the effect of Rp-8-Br-cAMPS, a specific inhibitor of PKA, on the Gα_s-stimulated Ba²⁺ currents. Steady-state peak I_Ba in Gα_s cells was significantly higher than that of time-matched control cells in the absence of Rp-8-Br-cAMPS. Superfusion with Rp-8-Br-cAMPS alone did not have any effect on I_Ba in control cells, but the enhanced peak I_Ba density in Gα_s-dialysed cells was reversed by subsequent superfusion with Rp-8-Br-cAMPS (Fig. 3). The relatively slow rate of blockade and washout of Rp-8-Br-cAMPS reflects the rate of solution change of the perfusion system as well as the time required for the drug to cross the sarcolemma. In another set of experiments, cells were superfused constantly with Rp-8-Br-cAMPS before and throughout the current recording period. Under these conditions, peak I_Ba density in cells dialysed with Gα_s was not significantly different from that of control cells at any time throughout the 20 min recording period (n = 3 and 4 for control and Gα_s-dialysed cells, respectively, data not shown).

Figure 3. Effect of Rp-8-Br-cAMPS on Gα_s-stimulated I_Ba.

Currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. A, time course of current recordings from a control and a Gα_s-dialysed cell. Cells were first superfused with basal solution and then exposed to 30 μM Rp-8-Br-cAMPS during the period indicated by the horizontal bar. The individual traces shown in the upper inset of the figure were obtained before (1), during (2) and after (3) superfusion of Rp-8-Br-cAMPS. B, averaged data for control cells (n = 11) and cells dialysed with 50 nM Gα_s (n = 9) before and 5 min after exposure to Rp-8-Br-cAMPS. * Significantly different from control value; † significantly different from the value before exposure to Rp-8-Br-cAMPS in the same cells (P < 0.05).

The transmembrane signalling pathway involved in the stimulation of β-adrenergic receptors includes activation of G_s proteins, adenylyl cyclase, followed by cAMP activation of PKA. To further confirm the hypothesis that PKA is responsible for the stimulation of Ca²⁺ channels by Gα_s under our experimental conditions, we measured peak I_Ba in both Gα_s and control cells in the absence and presence of SQ 22536, a specific inhibitor of adenylyl cyclase. Addition of SQ 22536 entirely abolished the stimulatory effect of Gα_s on peak I_Ba, and this effect of SQ 22536 could be washed out (Fig. 4A). In contrast, SQ 22536 had no detectable effect on I_Ba in control cells. The mean value of peak I_Ba density in the Gα_s-dialysed group was not significantly different from that in control cells during the superfusion of SQ 22536, although peak I_Ba density in the same Gα_s group of cells was higher than control cells before superfusion with SQ 22536 (Fig. 4B). These data, along with the results from the Rp-8-Br-cAMPS experiments, suggest that the stimulation of Ca²⁺ channels in the portal vein smooth muscle cells by Gα_s is entirely mediated by the PKA pathway, and no evidence for a direct membrane-delimited effect of Gα_s was observed.

Figure 4. Effect of SQ 22536 on Gα_s-stimulated I_Ba.

Currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. A, time course of current recordings from a control and a Gα_s-dialysed cell. Cells were first superfused with basal solution and then exposed to 200 μM SQ 22536 during the period indicated by the bar. The individual traces shown in the upper inset of the figure were obtained before (1), during (2) and after (3) superfusion of SQ 22536. B, averaged data for control cells (n = 5) and cells dialysed with Gα_s (n = 10) before and 5 min after exposure to SQ 22536. * Significantly different from control value; † significantly different from the value before exposure to SQ 22536 in the same cells (P < 0.05).

Effect of Rp-8-Br-cAMPS and KT 5720 on Gβγ-stimulated peak Ba²⁺ currents

Experiments in Fig. 1 demonstrated a stimulatory effect of Gβγ on Ca²⁺ channel activity. To evaluate the intracellular pathway involved in the stimulatory effect of Gβγ, we first tested the effect of PKA inhibitors on Gβγ-stimulated I_Ba. When peak I_Ba from cells dialysed with or without 50 nM Gβγ reached a steady state, cells were superfused with a bathing solution containing either Rp-8-Br-cAMPS (30 μM) or KT 5720 (200 nM) (Kase et al. 1987). These drugs, at the concentration used in this set of experiments, have been shown to completely block the stimulatory response of I_Ba to either cAMP or the catalytic subunit of PKA (Ruiz-Velasco et al. 1998). Superfusion with Rp-8-Br-cAMPS or KT 5720 did not change the peak I_Ba density in control cells nor affect the stimulatory response of I_Ba to Gβγ. Peak I_Ba density in cells dialysed with Gβγ was significantly higher compared with control cells, in the absence or presence of either Rp-8-Br-cAMPS (Fig. 5A) or KT 5720 (Fig. 5B). These results suggest that PKA is not involved in the stimulation of Ca²⁺ channel activity in VSM by G protein βγ subunits.

Figure 5. Effects of Rp-8-Br-cAMPS and KT 5720 on Gβγ-stimulated I_Ba.

Cells were first superfused with basal solution and then exposed to either 30 μM Rp-8-Br-cAMPS or 200 nM KT 5720 and currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. A, averaged values of peak I_Ba density in control (n = 11) and Gβγ-dialysed (n = 7) cells before and after exposure to Rp-8-Br-cAMPS. B, averaged values of peak I_Ba density in control (n = 9) and Gβγ-dialysed (n = 8) cells before and after exposure to KT 5720. * Significantly different from control values under the same experimental conditions (P < 0.05).

Effect of bisindolylmaleimide 1 and calphostin C on Gβγ-stimulated Ba²⁺ currents

The inability of PKA inhibitors to reduce the stimulatory response of I_Ba to G protein βγ subunits suggests that Gβγ may activate Ca²⁺ channels through another protein kinase or through a direct membrane-delimited pathway. To answer this question, we tested the possible involvement of PKC in Gβγ stimulation of Ba²⁺ currents. In this set of experiments, cells were dialysed with or without Gβγ and Ba²⁺ currents were measured. When peak current reached a steady state, cells were then superfused with either bisindolylmaleimide 1 (BIM, 100-200 nM) or calphostin C (100 nM), both of which are PKC inhibitors. The difference between BIM and calphostin C is that the latter inhibits both Ca²⁺-dependent and Ca²⁺-independent isozymes of PKC while the former is more selective for Ca²⁺-dependent isoforms of PKC (Gordge & Ryves, 1994). Application of BIM did not affect I_Ba recorded in the presence of Gβγ. Cumulative data for Gβγ-dialysed cells superfused with 200 nM BIM are shown in Fig. 6A. In contrast to BIM, calphostin C at 100 nM concentration significantly decreased peak I_Ba density in cells dialysed with Gβγ but had no effect on currents in control cells. The steady-state inhibition of Gβγ-stimulated currents by calphostin C occurred at 5-10 min. Figure 6B shows that calphostin C selectively abolished the stimulatory effect of Gβγ on peak I_Ba density.

Figure 6. Effects of calphostin C and BIM on Gβγ-stimulated I_Ba.

Cells were first superfused with basal solution and then exposed to either 200 nM BIM or 200 nM calphostin C and currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. A, averaged values of peak I_Ba density in cells dialysed with 50 nM Gβγ (n = 7) before and after exposure to BIM. B, averaged values of peak I_Ba density in control (n = 5) and Gβγ-dialysed (n = 17) cells before and 10 min after exposure to calphostin C. * Significantly different from the control value under the same experimental conditions; † significantly different from the value under basal conditions in the same cells (P < 0.05).

Effect of phorbol esters on L-type Ca²⁺ channel activity

Results from the above set of experiments suggest that protein kinase C may be involved in the Ba²⁺ current response to Gβγ and may be responsible for the observed stimulation of Ca²⁺ channels under our experimental conditions. The possible involvement of PKC in the regulation of Ca²⁺ channels in portal vein smooth muscle cells was further tested by two sets of experiments using phorbol esters. First, cells were pretreated with phorbol 12-myristate 13-acetate (PMA, 100 nM) for >18 h. Long-term exposure of cells to PMA is a common method to down-regulate PKC activity (Roman et al. 1998). Pretreated cells were then dialysed with or without Gβγ (50 nM) and I_Ba recorded. As shown in Fig. 7, peak I_Ba in both control and Gβγ-dialysed cells reached a steady state within ∼5 min. Application of calphostin C had no effect on peak I_Ba in both groups of pretreated cells. Averaged peak I_Ba density in cells dialysed with Gβγ was not different from control in the presence or absence of calphostin C (Fig. 7B). In another set of experiments, cells were dialysed with a pipette solution without G protein subunits and Ba²⁺ currents were measured. When peak current reached a steady state, phorbol 12,13-dibutyrate (PDBu, 200 nM) was added to the superfusate. PDBu is a phorbol ester known to activate PKC (Gordge & Ryves, 1994). Application of PDBu consistently increased peak I_Ba in these cells. The stimulatory response to PDBu reached a steady state after 5-10 min superfusion and remained stable throughout the experiment for up to 30 min superfusion with PDBu (n = 4, data not shown). In addition, the stimulatory response of I_Ba to PDBu could be blocked by calphostin C. Figure 8A shows an example of a typical recording under these conditions. Figure 8B shows that the mean peak I_Ba density was increased 40 ± 13 % by PDBu but was not different from the value under basal control conditions when cells were superfused with both PDBu and 200 nM calphostin C.

Figure 7. Effect of long-term treatment with PMA on Gβγ-stimulated I_Ba.

Cells were first exposed to PMA (100 nM) for over 18 h before current measurement. The next day, cells were dialysed with or without Gβγ (50 nM) and currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. After peak I_Ba reached a steady state, cells were superfused with the bathing solution containing calphostin C (100 nM) for 10 min. A, representative recordings from a control and a Gβγ-dialysed cell. Note that application of calphostin C for a 10 min period did not change peak I_Ba in cells from either treatment. B, averaged values of peak I_Ba density in control (n = 4) and Gβγ-dialysed (n = 5) cells before and 10 min after exposure to calphostin C. There is no significant difference between the values in control and Gβγ-dialysed cells before or after exposure to calphostin C.

Figure 8. Effect of PDBu on I_Ba in the presence and absence of calphostin C.

Cells were dialysed with a pipette solution without added G protein subunits, superfused with basal solution, and currents were elicited by voltage steps to 0 mV from a holding potential of -70 mV every 30 s. When peak current reached the steady state, cells were superfused with bathing solution containing 200 nM PDBu and then with solution containing 200 nM PDBu plus 200 nM calphostin C. A, time course of current recording from a representative cell. The individual traces shown in the upper inset of the figure were obtained from the recording times indicated. B, averaged values of peak I_Ba density under different experimental conditions from the same cells (n = 9). * Significantly different from the value under basal conditions (P < 0.05).

DISCUSSION

The results from the present study suggest that in vascular smooth muscle: (1) both α_s subunits and βγ dimers of G proteins exert stimulatory effects on L-type Ca²⁺ channel activity; (2) the stimulatory effect of Gα_s could be completely blocked by the adenylyl cyclase and cAMP-dependent protein kinase inhibitors, SQ 22536 and Rp-8-Br-cAMPS; and (3) the stimulatory effect of Gβγ could be abolished by the protein kinase C inhibitor, calphostin C. Furthermore, these data failed to detect any direct membrane-delimited modulation of Ca²⁺ channels in vascular smooth muscle by either α_s or βγ forms of activated G protein subunits.

There is general consensus that α subunits of activated G_s protein play an important role in regulation of L-type Ca²⁺ channels in cardiac and smooth muscle cells during β-adrenergic stimulation (McDonald et al. 1994; Xiong & Sperelakis, 1995). However, using an anti-G protein β subunit antibody, Macrez et al. (1997) recently reported that a βγ dimer from G₁₃ might be responsible for the signal transduction pathway during angiotensin II-induced stimulation of L-type Ca²⁺ channels in rat portal vein myocytes. In the present study, both purified Gα_s and Gβγ stimulated the activity of L-type Ca²⁺ channels in smooth muscle cells freshly isolated from rabbit portal vein. The enhancement of I_Ba by Gα_s and Gβγ appears specific because dialysis with the GDPβS-bound form of Gα_s and non-prenylated βγ dimers had no effect. In addition, the effects of Gα_s and Gβγ on Ca²⁺ channels in vascular smooth muscle cells were additive. When cells were dialysed with combined Gα_s and Gβγ, peak I_Ba density was significantly higher than that in cells dialysed with either Gα_s or Gβγ alone. Thus, the results from our study suggest that both α subunits and βγ dimers of activated G_s proteins may play a role in the regulation of L-type Ca²⁺ channels in vascular smooth muscle cells during β-adrenergic stimulation. Although it is not clear whether the βγ combination used in this study (i.e. β1γ2) is coupled specifically with α_s in vascular smooth muscle, it appears that many different combinations of β and γ subunits (except β1γ1) have similar actions (Ueda et al. 1994; Dolphin, 1998). For example, Ueda et al. (1994) compared the ability of different combinations of recombinant βγ subunits to modulate types I and II adenylyl cyclase activities, stimulate phosphoinositide-specific phospholipase C_β, support pertussis toxin-catalysed ADP-ribosylation of rGα1_i and Gα_o, and inhibit steady-state GTP hydrolysis catalysed by Gα_s, Gα_o and myristoylated rGα2_i. The results from their study fail to discriminate between the many isoforms of βγ (except β1γ1). These authors further suggest that different combinations of βγ subunits may be functionally interchangeable among α subunits (Ueda et al. 1994).

Although it is well established that G_sα subunits play an important role in the β-adrenergic stimulation of L-type Ca²⁺ channels in the cardiovascular system, the signalling pathways underlying the modulation of L-type Ca²⁺ channels by activated G_s proteins remain controversial. In cardiac myocytes, it is generally agreed that the majority of Ca²⁺ channel stimulatory effects of β-adrenergic receptor activation is via activation of adenylyl cyclase and subsequent phosphorylation of the channel (McDonald et al. 1994). Early evidence also suggested there is a direct G protein activation of Ca²⁺ channels in the heart. First, Yatani et al. (1987) showed that Gα_s could restore activity to rundown patches containing L-type Ca²⁺ channel current. Second, a small, fast component of Ca²⁺ channel activation in response to β-adrenergic receptor stimulation was attributed to direct G protein effects due to its rapid kinetics (Brown, 1990). Third, when partially purified cardiac sarcolemmal vesicles were incorporated in bilayers, Gα_s potentiated the open probability of Ca²⁺ channels four- to sixfold (Imoto et al. 1988). Even with this evidence, the existence of a direct G protein modulation of cardiac Ca²⁺ channels and the physiological importance of such a pathway are still under debate (Hartzell & Fischmeister, 1992; Clapham, 1994). In smooth muscle cells, there are few reports available to support a direct modulation of L-type Ca²⁺ channels by G proteins. Xiong et al. (1994a) observed a dual effect of Iso on L-type Ca²⁺ channels in rabbit portal vein myocytes. In the presence of 10 μM Iso, Ca²⁺ current was initially increased and subsequently decreased. The stimulatory effect of Iso could be mimicked by the activated G protein α subunit. In addition, H-7, a non-specific inhibitor of protein kinases, blocked the inhibitory phase but not the initial stimulatory phase of the channel response to Iso. Their conclusions, based on these observations, were that direct G protein regulation of the channel was solely responsible for the stimulatory effects of Iso, whereas the inhibitory effects were mediated by adenylate cyclase, cAMP and PKA phosphorylation of the channel (Xiong et al. 1994a). However, a recent report from the same group (Liu et al. 1997) demonstrated that forskolin produced a stimulatory effect on Ca²⁺ channels in smooth muscle cells of rat mesenteric artery and this effect could be blocked by a PKA inhibitor, PKI. In addition, we demonstrated in a recent study (Ruiz-Velasco et al. 1998) that both 8-bromo cAMP and the catalytic subunit of PKA significantly increased peak Ba²⁺ currents, and their effects could be entirely blocked by specific PKA inhibitors. In the present study, peak I_Ba density in cells dialysed with Gα_s was significantly higher than that in cells dialysed with pipette solution containing inactive or no G protein subunits. The stimulatory effects of Gα_s were also entirely blocked by Rp-8-Br-cAMPS and SQ 22536, specific inhibitors of protein kinase A and adenylyl cyclase, respectively. These data suggest that the activated α subunits of G_s proteins elicit their stimulatory effect through the cAMP-PKA pathway, not through a direct G protein gating of the channel.

Modulation of L-type Ca²⁺ channels by G protein βγ subunits has received much less attention. βγ subunits of G proteins have been shown to modulate many effectors in different pathways (Clapham & Neer, 1997). For example, Gβγ inhibits N- and P/Q-type Ca²⁺ channels by direct binding of βγ dimers to the α1 interaction domain of these Ca²⁺ channels (De et al. 1997; Dolphin, 1998). βγ subunits also activate K⁺ channels in cardiac cells (Clapham & Neer, 1993). In the present study, the stimulatory effects of Gβγ could not be abolished by Rp-8-Br-cAMPS or KT 5720, which eliminates the possible involvement of the cAMP-PKA pathway. In contrast, calphostin C completely blocked the stimulatory effect of Gβγ on Ca²⁺ channels. Calphostin C also abolished the stimulatory response elicited by PDBu. In addition, downregulation of PKC with long-term pretreatment of PMA successfully prevented the stimulatory effect of Gβγ. These data indicate a possible involvement of protein kinase C. Although it is unclear exactly which isozymes of PKC might be activated by Gβγ under our experimental conditions, the inability of BIM to abolish the stimulatory effect of Gβγ suggests the possible involvement of Ca²⁺-independent PKC isoforms in our study since this compound has a higher affinity to block Ca²⁺-dependent PKC isozymes (Gordge & Ryves, 1994). In fact, under the experimental conditions in the present study, there was no Ca²⁺ added in either the pipette solution or the superfusate, and intracellular Ca²⁺ was buffered by 10 mM EGTA. In rabbit portal vein smooth muscle three isoforms of PKC, α, ε and ξ, were found to be present (Clement-Chomienne et al. 1996). PKCα is a Ca²⁺-dependent isoform whereas the other two are Ca²⁺ independent (Gordge & Ryves, 1994). However, PKCξ is insensitive to 1,2-diacylglycerol (DAG) as well as to phorbol esters and is not antagonized by calphostin C (Gordge & Ryves, 1994). Thus, PKCε appears to be the most likely isozyme responsible for the signalling transduction in Gβγ stimulation of L-type Ca²⁺ channels in rabbit portal vein smooth muscle. Furthermore, although not tested in this study, other additional effects of Gβγ on L-type Ca²⁺ channels might occur in the presence of Ca²⁺.

The possible involvement of PKC in the coupling of βγ subunits of G proteins with L-type Ca²⁺ channels is supported by observations from other research groups. Activation of PKC has been shown previously to stimulate L-type Ca²⁺ channels in vascular smooth muscle cells (McHugh & Beech, 1997). In addition, a recent report from Macrez et al. (1997) showed that intracellular dialysis of rat portal vein myocytes with an antibody to the G protein β subunit blocked the stimulatory effect of angiotensin II on L-type Ca²⁺ channels leading to the conclusion that the βγ subunit was involved. Although not directly tested, they proposed that βγ stimulation was due to activation of protein kinase C because both angiotensin II and PKC activators produced a similar change in the I-V relationship of Ca²⁺ channels. Other evidence indicates that βγ subunits of G proteins have no direct effect on either α1C L-type Ca²⁺ channel subunits (Dolphin, 1998), or types V and VI adenylyl cyclases (cardiac and smooth muscle types) (Tang & Gilman, 1992; Ishikawa & Homcy, 1997). There are several possible pathways by which Gβγ may stimulate PKC in cells, such as Gβγ-activated phospholipase C, D and A₂ (Cockcroft, 1992; Clapham & Neer, 1997). Further experiments are needed to define the possible link between Gβγ and PKC in vascular smooth muscle cells.

The results of the present study suggest a simple schematic model to explain the intracellular mechanisms underlying stimulation of L-type Ca²⁺ channels in vascular smooth muscle cells in response to β-adrenergic receptor binding (Fig. 9). Upon receptor binding with agonist, both the α subunit and βγ dimer of G_s protein are activated. The α subunit then activates adenylyl cyclase, which activates PKA via production of cAMP. PKA phosphorylation of the VSM α1C subunit may lead to an increase in the number of functional channels and increases in open probability due to changes in fast and slow gating behaviour, in a manner analogous to that described for cardiac α1C subunits (McDonald et al. 1994), since the two isoforms have similar complements of putative PKA phosphorylation sites in the carboxyl termini and greater than 90 % amino acid homology (Stea et al. 1995). In fact, cAMP has been shown to increase I_Ba in cells expressing cardiac (Gao et al. 1997) or VSM (Klockner et al. 1992) α1C co-expressed with a skeletal muscle β subunit of Ca²⁺ channels. In contrast to α subunits, βγ subunits activate PKC, possibly through stimulation of phospholipase C and/or phospholipase D, which also leads to channel phosphorylation. The proposed model contrasts markedly with earlier models put forward to explain the regulation of VSM L-type Ca²⁺ channels by protein kinases and G protein subunits (Xiong et al. 1994a; Xiong & Sperelakis, 1995; Liu et al. 1997). Our data failed to provide any evidence supporting the potential role of a membrane-delimited direct G protein pathway in the regulation of L-type Ca²⁺ channels in vascular smooth muscle cells involving either activated Gα_s or Gβγ.

Figure 9. Proposed signalling pathways underlying the modulation of vascular L-type Ca²⁺ channels by G_s protein subunits.

Note that both α and βγ subunits of G_s proteins are involved in the stimulation of Ca²⁺ channel activity during β-adrenergic receptor binding. The stimulatory effects of Gα_s and Gβγ on Ca²⁺ channels are via protein kinases A and C, respectively. In addition, higher levels of cAMP activate not only PKA but also PKG, which then leads to an inhibition of Ca²⁺ channels (Ruiz-Velasco et al. 1998). No evidence for a membrane-delimited direct G protein regulation of Ca²⁺ channels in vascular smooth muscle cells was observed.