Ca²⁺ Channel β₃ Subunit Enhances Voltage-Dependent Relief of G-Protein Inhibition Induced by Muscarinic Receptor Activation and G_βγ

Abstract

The Ca²⁺ channel β subunit has been shown to reduce the magnitude of G-protein inhibition of Ca²⁺channels. However, neither the specificity of this action to different forms of G-protein inhibition nor the mechanism underlying this reduction in response is known. We have reported previously that coexpression of the Ca²⁺ channel β₃subunit causes M₂ muscarinic receptor-mediated inhibition of α_1B Ca²⁺ currents to become more voltage-dependent. We report here that the β₃ subunit increases the rate of relief of inhibition produced by a depolarizing prepulse and also shifts the voltage dependency of this relief to more hyperpolarized voltages; these effects are likely to be responsible for the reduction of inhibitory response of α_1B channels to G-protein-mediated inhibition seen after coexpression of the Ca²⁺ channel β₃ subunit. Additionally, the β₃ subunit alters the rate and voltage dependency of relief of the inhibition produced by coexpressed G_β1γ1, in a manner similar to the changes it produces in relief of M₂ receptor-induced inhibition. We conclude that the Ca²⁺ channel β₃ subunit reduces the magnitude of G-protein inhibition of α_1BCa²⁺ channels by enhancing the rate of dissociation of the G-protein βγ subunit from the Ca²⁺channel α_1B subunit.

G-protein-mediated inhibition of voltage-gated Ca²⁺ channels provides an important mechanism for regulating synaptic strength (Holz et al., 1986; Wheeler et al., 1994; Dittman and Regehr, 1996; Takahashi et al., 1996). Although many types of Ca²⁺ channels can undergo this class of inhibition, N-type Ca²⁺ current is the most frequently studied target of this modulation (Schultz et al., 1990; Anwyl, 1991; Dolphin, 1991; Hille, 1994). Various members of the seven membrane-spanning family of receptors, after binding neurotransmitter, transduce their signal via activation of a variety of heterotrimeric G-proteins. The activated G-proteins then either directly interact with the channel to cause inhibition, in a process known as membrane-delimited inhibition (Bean, 1989; Brown and Birnbaumer, 1990), or subsequently activate a second messenger cascade that ultimately acts on the channel to cause inhibition. N-type Ca²⁺ channels are inhibited via both a membrane-delimited pathway (Schultz et al., 1990; Anwyl, 1991; Dolphin, 1991; Hille, 1994) and a pathway requiring diffusible intracellular second messengers (Beech et al., 1992; Shapiro et al., 1994a).

Membrane-delimited G-protein inhibition encompasses both voltage-dependent and voltage-independent inhibition. Voltage-dependent inhibition exhibits two main characteristics in voltage-clamp studies: (1) slowed activation kinetics and (2) diminished inhibition at more depolarized voltages (Marchetti et al., 1986; Wanke et al., 1987; Bean, 1989; Kasai and Aosaki, 1989). The diminished inhibition at more depolarized voltages gives rise to a third characteristic of voltage-dependent inhibition, prepulse current facilitation (Elmslie et al., 1990; Ikeda, 1991; Lopez and Brown, 1991). Strongly depolarizing voltages are thought to cause a temporary dissociation of the G-protein from the Ca²⁺ channel (Bean, 1989; Lopez and Brown, 1991; Golard and Siegelbaum, 1993); thus, a current elicited during this period of G-protein dissociation will be facilitated compared with current elicited by the same test voltage step without a depolarizing prepulse.

Voltage-independent inhibition is characterized by equivalent current inhibition at all voltages, with no change in current kinetics during the inhibition. Frequently, voltage-independent G-protein inhibition requires intracellular signaling cascades and thus is not membrane-delimited (Beech et al., 1991, 1992; Bernheim et al., 1991;Shapiro et al., 1994a). However, there are instances of membrane-delimited voltage-independent inhibition (Shapiro and Hille, 1993; Diverse-Pierluissi et al., 1995; Wollmuth et al., 1995).

Voltage-dependent inhibition of N-type Ca²⁺ currents in rat superior cervical ganglion (SCG) sympathetic neurons (Herlitze et al., 1996; Ikeda, 1996), as well as α_1ACa²⁺ channel currents expressed in tsA-201 cells (Herlitze et al., 1996), is mediated by the G-protein βγ subunit. G_βγ, however, seems not to be responsible for voltage-dependent inhibition of N-type currents in embryonic chick dorsal root sympathetic ganglion neurons (Diverse-Pierluissi et al., 1995). The G-protein βγ subunit is capable of binding to at least two regions of the intracellular loop between transmembrane regions I and II of α_1A and α_1BCa²⁺ channels (De Waard et al., 1997; Zamponi et al., 1997); the same intracellular loop contains the binding region for the Ca²⁺ channel β subunit (Pragnell et al., 1994). Mutations that reduce in vitro G-protein βγ subunit binding to this region of the Ca²⁺ channel also block some characteristics of voltage-dependent G-protein inhibition of α_1A current (De Waard et al., 1997), although similar mutations do not affect somatostatin-induced inhibition of α_1B currents (Zhang et al., 1996). The critical amino acids within α_1A responsible for G_βγ binding are not the same as those critical for Ca²⁺ channel β subunit binding (De Waard et al., 1997), suggesting that direct competition for a binding site on the α₁ subunit is unlikely.

The Ca²⁺ channel β subunit reduces the magnitude of G-protein inhibition of both α_1A and α_1BCa²⁺ channels expressed in Xenopusoocytes (Roche et al., 1995), as well as Ca²⁺currents in rat dorsal root ganglion neurons (Campbell et al., 1995). Speculation on the mechanism underlying this reduction in sensitivity to G-protein inhibition includes: (1) direct competition between the Ca²⁺ channel β₃ subunit and the G-protein for the same site on the Ca²⁺ channel α₁ subunit (McAllister-Williams and Kelly, 1995; Roche et al., 1995; Bourinet et al., 1996; Clapham, 1996), (2) steric blockade of the G-protein binding site (Roche et al., 1995; Bourinet et al., 1996), and (3) a Ca²⁺ channel β subunit-induced increase in the GTPase activity of the G-protein (Campbell et al., 1995). Examination of M₂ muscarinic receptor-induced inhibition of α_1B currents in Xenopus oocytes revealed that not only is the magnitude of the G-protein inhibition reduced after coexpression of the Ca²⁺ channel β₃ subunit but the portion of inhibited current that is voltage-dependent is increased as well (Roche and Treistman, 1998). Here, we examine possible mechanisms that underlie the increase in voltage-dependence and discuss whether this mechanism can explain the reduction in the magnitude of M₂ receptor-induced inhibition of α_1B currents after coexpression of the Ca²⁺ channel β₃ subunit. We address these questions using α_1B Ca²⁺channels coexpressed with muscarinic M₂ receptors inXenopus oocytes. The coexpressed M₂ receptor couples to the endogenous pertussis toxin-sensitive G-proteins of theXenopus oocyte (Lechleiter et al., 1991). We also coexpress G-protein α and βγ subunits individually to determine the G-protein subunit mediating inhibition of both α_1B and α_1Bβ₃ Ca²⁺ channel currents and to assess the influence of the Ca²⁺channel β₃ subunit on the direct actions of these G-protein subunits on α_1B Ca²⁺channels.

MATERIALS AND METHODS

Expression plasmids and oocyte preparation. Capped RNA transcripts encoding full-length α_1A(XbaI-linearized/SP6 RNA polymerase; gift of Dr. Y. Mori, University of Cincinnati Medical Center), α_1B(SalI/SP6; gift of Dr. Y. Fujita, Kyoto University), and β₃ (NotI/T7; gift of Dr. Edward Perez-Reyes, Loyola University Medical Center) calcium channel subunits as well as the muscarinic M₂ receptor (EcoRI/BglII; gift of Dr. Wolfgang Sadee, University of California San Francisco) and G-protein αi₂(gift of Dr. Randall Reed, HHMI, Baltimore, MD) and β₁γ₁ (gift of Drs. Melvin Simon and Anna Aragay, California Institute of Technology, Pasadena, CA) subunits were synthesized using the mMESSAGE mMACHINE in vitrotranscription kit (Ambion, Austin, Texas). Xenopus laevisstage V–VI oocytes were removed and treated with collagenase (Sigma type IV; Sigma, St. Louis, MO) to remove the follicular layer. The oocytes were then injected with cRNA encoding α_1B in combination with M₂ in a ratio of 2:1 or in combination with both M₂ and β₃ (2:2:1). The concentration of all individual RNAs before injection was 0.1 μg/μl, with the exception of the G-protein α and βγ subunit RNA that was 0.5 μg/μl, and 20–60 nl of RNA mixed at the above ratios was injected. The oocytes were maintained in culture at 18°C for at least 2 d in ND-96 solution (96 mm NaCl, 2 mm KCl, 1.8 mm CaCl₂, and 5 mm HEPES, pH 7.5) supplemented with 2.5 mmsodium pyruvate and 2 mg/ml gentamycin.

Electrophysiological recording and experimental treatments.Two-electrode voltage-clamp currents were recorded using a Dagan CA-1 amplifier. Oocytes were clamped at a holding potential of −80 mV, and various electrophysiological protocols were used, as noted. Currents were filtered at 1 or 10 kHz, and a p/4 leak subtraction technique was used. Inhibition of current amplitude was determined by measurements of the peak current attained at any point during the 250 msec test pulse. Analysis was done off-line, using pClamp software version 6.0.2 (Axon Instruments, Foster City, CA). Electrodes contained 3 m KCl and had resistances of 0.5–2 MΩ. Oocytes were placed in a 1 ml chamber and perfused at a rate of 0.5 ml/min. All recordings were made at room temperature using bath solutions containing (in mm): Ba(OH)₂, 10; NaOH, 50; CsOH, 2; TEA-OH, 20; N-methyl-d-glucamine, 20; and HEPES, 5, titrated to pH 7.5 with methanesulfonic acid. In all experiments, 20 nl of a 100 mm stock solution of K₃-1,2-bis(aminophenoxy)ethane-N,N,N′,N′-tetra-acetic acid (BAPTA) (Sigma) was injected at least 2 hr before the experiment. The final concentration of BAPTA inside the oocyte was estimated to be between 2 and 5 mm, assuming an oocyte volume of 1 μl. For experiments using N-ethylmaleimide (NEM) (Aldrich, Milwaukee, WI), the NEM was dissolved in the external solution to a final concentration of 200 μm and was applied to the oocyte for 2 min. Acetylcholine (ACh) (Sigma) was stored as a 10 mm stock solution in water and dissolved in the recording medium to a final concentration of 50 μm.

RESULTS

Ca²⁺ channel β₃ subunit modulates voltage dependence of M₂-mediated inhibition

A protocol designed to remove tonic G-protein inhibition of α_1B Ca²⁺ channels allowed study of the isolated muscarinic M₂ receptor-induced G-protein inhibition of these channels. Briefly, we exposed the oocyte to 50 μm ACh. Immediately after removal of the ACh, there is a large rebound of current amplitude, resulting from temporary loss of tonic G-protein inhibition (Roche and Treistman, 1998). During the period in which tonic inhibition is abolished, ascertained by the loss of prepulse facilitation, the current remains sensitive to muscarinic receptor-induced inhibition. Loss of tonic inhibition occurred, in most cases, after a single 1 min application of ACh; on occasion, multiple ACh applications were required to remove tonic inhibition completely. Using this protocol, we have demonstrated that expression of the Ca²⁺ channel β₃ subunit reduced the magnitude of muscarinic M₂ receptor-induced inhibition (Fig.1A,B). However, the reduction in magnitude of inhibition was voltage-dependent, with substantial reductions of G-protein inhibition at voltages more positive than 0 mV, and no effect on calcium current inhibition during voltage steps to −10 or 0 mV (Fig. 1C). In addition to the reduced inhibition, a depolarizing prepulse during muscarinic inhibition elicits greater relief of G-protein inhibition after coexpression of the Ca²⁺ channel β₃ subunit (Fig. 1D).

Fig. 1.

The Ca²⁺ channel β₃ subunit modifies the voltage dependence of muscarinic-induced G-protein inhibition. A,B, Representative records of α_1B and α_1Bβ₃ Ca²⁺ currents for the control situation (Con), as well as after the application of 50 μm ACh and before (+ACh,−PP) and after (+ACh,+PP) a depolarizing prepulse to +100 mV for 75 msec. Oocytes were held at −80 mV and stepped to a test potential of +10 mV for 250 msec. The M₂ receptor is coupling to G-proteins that are endogenous to the oocyte.C, Inhibition of current amplitude at various test potentials for both the α_1B (open) and α_1Bβ₃ (filled) Ca²⁺ currents. D, Relief of M₂ receptor-induced inhibition by a depolarizing prepulse to +100 mV for 75 msec. The prepulse was given 20 msec before the test pulse. Facilitation was measured as the percentage of inhibited current that was reversed by the prepulse voltage protocol.

The Ca²⁺ channel β₃ subunit increases the rate of voltage-dependent relief of G-protein inhibition of α_1B currents

Voltage-dependent relief of G-protein inhibition of N-type currents is thought to result from temporary dissociation of the G-protein from the Ca²⁺ channel (Lopez and Brown, 1991; Golard and Siegelbaum, 1993). Thus, the heightened relief of the inhibited current by depolarizing voltages after coexpression of the Ca²⁺ channel β₃ subunit suggests that the rate of G-protein dissociation has changed. An increase in the G-protein dissociation rate could explain the reduced inhibition of current by M₂ receptor activation when the Ca²⁺ channel β₃ subunit is coexpressed, because the inhibition would be more easily reversed by moderate voltages, such as those in the range normally used to activate the Ca²⁺ channel.

This model was tested by increasing the duration or voltage of the prepulse incrementally and determining the rate of current facilitation of α_1B Ca²⁺ currents both with and without Ca²⁺ channel β₃ subunit associated with the α_1B channel. The Ca²⁺ channel β₃ subunit dramatically decreased the duration of the prepulse necessary for maximal facilitation from ∼160 msec to <20 msec; a single exponential fit to the data revealed a decrease in the time constant of relief by a voltage step to +100 mV from 67 msec in the absence of β₃auxiliary subunit to 6.9 msec after coexpression of the Ca²⁺ channel β₃ subunit (Fig.2A,B).

Fig. 2.

Modulation of time and voltage dependency of prepulse facilitation by the Ca²⁺ channel β₃ subunit. A, B, Facilitation of current amplitude after G-protein inhibition induced by application of acetylcholine. The prepulse was to +75 mV for varying periods of time, as indicated. The test potential was +10 mV. The data were fit with a single exponential, revealing time constants of 67 msec for the α_1B currents and 6.9 msec for the α_1Bβ₃ currents. C,D, Facilitation of current amplitude with a 75 msec prepulse of varying voltage, as indicated. The test potential was +10 mV. These data were fit with a Boltzmann curve, revealing V_1/2 values of 52 mV for the α_1B currents and 39 mV for the α_1Bβ₃ currents.

We also tested for changes in the voltage dependence of prepulse facilitation. This protocol was similar to the duration protocol used previously except, in this case, the voltage of the prepulse step was increased incrementally, while maintaining a fixed prepulse duration. The data were fitted with Boltzmann curves, revealing a V_1/2 for current facilitation of 52 mV for the α_1B current and 39 mV after coexpression of the Ca²⁺ channel β₃ subunit (Fig.2C,D). Thus, the rate of reversal of G-protein inhibition, as well as the voltage that is necessary to reverse the G-protein inhibition of the Ca²⁺ channel, has decreased after coexpression of the Ca²⁺ channel β₃ subunit.

G-protein βγ subunit mediates inhibition of α_1B currents

There is evidence that the voltage-dependent form of G-protein inhibition is mediated by the G-protein βγ subunit for N-type currents in rat SCG neurons (Herlitze et al., 1996; Ikeda, 1996). However, it is also clear that this is not the case in chick dorsal root ganglion neurons, in which the βγ subunit mediates a voltage-independent form of inhibition (Diverse-Pierluissi et al., 1995). We coexpressed subunits of a heterotrimeric G-protein with α_1B and α_1Bβ₃Ca²⁺ channels to determine the G-protein subunit mediating voltage-dependent inhibition in our system, first examining the tonic inhibition of Ca²⁺ currents produced by exogenously expressed G-proteins. Although α_1B currents display a large degree of tonic G-protein-mediated inhibition from G-proteins endogenous to the oocyte, activation of a coexpressed M₂ receptor results in both a further decrease in current amplitude and a slowing of activation kinetics (Roche and Treistman, 1998), suggesting that we should be able to detect any further G-protein inhibition induced by coexpression of a G-protein subunit. We first examine the results of G_βγ coexpression and then the influence of the G_α subunit.

Coexpression of the G-protein βγ subunit slowed current activation kinetics in comparison with current in oocytes in which no exogenous βγ subunits were expressed (Fig.3A), similar to the slowing of activation kinetics seen after muscarinic receptor-induced inhibition. Coexpression of the G-protein βγ subunit complex significantly increased the time necessary to reach peak current levels from 31.5 ± 2.2 to 70.0 ± 24.0 msec (p ≤ 0.05, Student’s t test) (Fig.3A,E). Facilitation of current amplitude by depolarizing prepulses dramatically increased after coexpression of the G-protein βγ subunit. Figure 3B shows the currents elicited both before (−PP) and after (+PP) a depolarizing prepulse for oocytes expressing the G-protein βγ subunit (+Gβγ). The mean facilitation of current amplitude was significantly increased from 108 ± 11 to 183 ± 10% after coexpression of the G-protein βγ subunit (p ≤ 0.05, Student’s t test) (Fig.3B,C). Slowed activation kinetics and increased prepulse facilitation are both consistent with increased voltage-dependent G-protein inhibition.

Fig. 3.

Effects of coexpressed G-protein α and βγ subunits on α_1B Ca²⁺ currents.A, Representative currents elicited by a voltage step from a holding potential of −80 mV to a test voltage of +10 mV. Control currents are labeled Con, whereas the currents elicited after application of 50 μm ACh are labeled +ACh. B, Representative currents elicited using the voltage protocol illustrated in oocytes coexpressing the G-protein βγ (G_βγ) or α (Gαi₂) subunits. Currents elicited without the prepulse are labeled −PP, whereas the currents elicited after a voltage step to +100 mV are labeled +PP. C, Summary of mean voltage-dependent facilitation before and after G-protein subunit coexpression.D, Summary of the inhibition of peak current amplitude by the initial application of ACh (Initial) and during the rebound phase induced after previous ACh application (Reb), as well as after coexpression of the G-protein βγ (+Gβγ) and α (+Gαi₂) subunits. E, Elapsed time from the beginning of the voltage step to the peak amplitude of the elicited current before (black) and after (gray) application of 50 μmacetylcholine. This was done for α_1B alone (α_1B), as well as after the coexpression of G-protein βγ (+Gβγ) or α (+Gαi₂). Sample size indicated above bars.

We next examined the effect of overexpression of the βγ subunit on M₂-mediated inhibition. The magnitude of inhibition of current amplitude after activation of the M₂ receptor was reduced after coexpression of the G-protein βγ subunit, from a value of 51 ± 3% inhibition for oocytes that were tonically inhibited but expressed no exogenous G-protein subunits (data not shown) to a value of 37 ± 8% inhibition after coexpression of the G-protein βγ subunit (Fig. 3D). This partial occlusion of the M₂-mediated inhibition is consistent with a common pathway for M₂- and exogenous βγ subunit-mediated inhibition. Further support for this conclusion is provided by examination of another measure of G-protein inhibition, the slowing of I_Ba activation kinetics measured as the time-to-peak current. The effects of M₂ activation and exogenous G_βγ were nonadditive, with similar values for maximal slowing obtained by M₂ receptor activation in the absence and presence of coexpressed G_βγ (Fig.3E). These data suggest a common pathway, consistent with voltage-dependent G-protein inhibition of α_1BCa²⁺ currents mediated by the G-protein βγ subunit.

G-protein α subunit blocks tonic G-protein inhibition

If G_βγ mediates the voltage-dependent inhibition of α_1B currents, we might expect that coexpression of the G-protein α subunit would block G-protein inhibition by acting as a “sink” for free βγ subunit. Such an effect of exogenous G-protein α subunit on G-protein βγ signaling has been suggested previously (Ikeda, 1996). Coexpression of G_α resulted in a significant decrease in the amount of tonic inhibition. We have shown previously that application of the alkylating agent NEM causes a potentiation of current amplitude (Roche et al., 1995), resulting from uncoupling of the basally active G-protein population. Coexpression of the G-protein α subunit should also eliminate potentiation of current amplitude by NEM, if the exogenous G-protein α subunit has blocked the tonic G-protein pathway. This is, indeed, the case. Potentiation of current amplitude by application of NEM to oocytes expressing α_1B currents and no exogenous G-protein subunits was 225 ± 25%, whereas the potentiation was reduced to 29 ± 9% after coexpression of the G-protein α subunit (data not shown). These data are consistent with the assumption that the G-protein α subunit acts as a sink for the tonically active βγ subunit, thus blocking the inhibition mediated by the G-protein βγ subunit. The G-protein α subunit did not, however, buffer M₂receptor-induced inhibition (79 ± 1.8% inhibition for control vs 86 ± 2.4% inhibition after coexpression of Gαi₂) (Fig. 3A,D).

Figure 3B shows representative currents elicited before and after a depolarizing prepulse to +100 mV, demonstrating the loss of prepulse facilitation after coexpression of G_α. Facilitation of current amplitude was reduced from 108 ± 11% facilitation for oocytes that expressed no exogenous G-protein subunits to −23 ± 10% facilitation after coexpression of exogenous G-protein α subunit (Fig. 3C). This loss of prepulse current facilitation is another indicator of the loss of voltage-dependent G-protein inhibition, supporting the conclusion that G_βγ mediates voltage-dependent inhibition of α_1B Ca²⁺ current.

Influence of Ca²⁺ channel β₃subunit on G-protein βγ subunit-mediated inhibition

The Ca²⁺ channel β subunit has been shown to significantly modify G-protein modulation of Ca²⁺channels, and we examined its influence on G_βγ-induced inhibition. The expression of exogenous G-protein βγ subunit was also effective in mediating voltage-dependent inhibition of α_1Bβ₃ Ca²⁺ currents. Similar to our results for the α_1B currents, the activation kinetics of the α_1Bβ₃ currents was significantly slowed by coexpression of the G-protein βγ subunit. Figure 4Ashows representative currents in the presence of exogenous G-protein subunits, demonstrating the slowed activation kinetics of the α_1Bβ₃ currents after coexpression of the G-protein βγ subunit. In addition, the G-protein βγ subunit also occludes the M₂ receptor-mediated inhibition (Fig.4A,C), again suggesting that βγ-induced inhibition is acting via the same mechanism as M₂-induced inhibition.

Fig. 4.

Effects of G-protein α and βγ subunit coexpression on α_1Bβ₃Ca²⁺ currents. A, Currents elicited by a voltage step from a holding potential of −80 mV to a test voltage of +10 mV. Control currents are labeled Con, whereas the currents elicited after application of 50 μm ACh are labeled +ACh. B, Representative currents elicited both before (−PP) and after (+PP) a depolarizing prepulse to +100 mV with and without coexpression of G-protein βγ (+Gβγ) and α (+Gαi₂) subunits.C, Summary of the inhibition of peak current amplitude by the initial application of ACh for α_1Bβ₃alone and after coexpression of the G-protein α (+Gαi₂) and βγ (+Gβγ) subunits. D, Summary of prepulse facilitation of current amplitude before and after coexpression of G-protein subunits. Sample size indicated abovebars.

There was very little facilitation of α_1Bβ₃current amplitude by depolarizing prepulses in the absence of exogenous G-protein subunits (Fig. 4D). However, after we coexpressed the G-protein βγ subunit, the facilitation of current amplitude by depolarizing prepulses was significantly increased. Figure4B shows representative α_1Bβ₃ currents, elicited before and after a depolarizing prepulse to +100 mV, in the absence and presence of coexpressed G_βγ. Prepulse facilitation using this protocol increased from 4 ± 4% when no exogenous G-protein subunits were expressed to 57 ± 8% after coexpression of the G-protein βγ subunit (Fig. 4D), indicative of a substantial increase in the amount of voltage-dependent inhibition.

Influence of Ca²⁺ channel β₃subunit on the ability of G-protein α subunit to block tonic G-protein inhibition

As with the α_1B current, the G-protein α subunit caused a small but significant increase in the magnitude of M₂-mediated inhibition of α_1Bβ₃current (Fig. 4C), from 55 ± 2.3% inhibition in oocytes that expressed no exogenous G-proteins to 68 ± 2% inhibition in oocytes that expressed exogenous G_αsubunit. Although the G-protein α subunit did not reduce the magnitude of M₂-induced inhibition, the G-protein α subunit did block a small tonic inhibition, evidenced by a decrease in the small amount of facilitation that was seen in the control (Fig.4D).

The Ca²⁺ channel β₃ subunit modulates the voltage sensitivity of G-protein βγ subunit-induced inhibition

A model for membrane-delimited voltage-dependent inhibition in which the G-protein βγ subunit binds directly to the α_1B Ca²⁺ channel has recently received experimental support (De Waard et al., 1997; Zamponi et al., 1997). Modulation of the inhibition mediated by exogenous G_βγby coexpression of the Ca²⁺ channel β₃subunit, therefore, would likely result from changes in the effectiveness of the interaction of G_βγ with the Ca²⁺ channel. We examined the influence of the Ca²⁺ channel β₃ subunit on the rate of voltage-dependent relief of G-protein βγ subunit-mediated inhibition. Figure 5 demonstrates that the Ca²⁺ channel β₃ subunit also dramatically increases the rate of relief of the inhibition produced by the coexpressed G_βγ subunit. A single exponential fit to the data revealed a shift in the rate at which the G-protein βγ-induced inhibition is reversed by depolarizing prepulses, from a time constant of 58 msec for α_1B alone to 6 msec after coexpression of the Ca²⁺ channel β₃subunit (Fig. 5A,B). This was similar to the increase in the rate of current facilitation produced by the Ca²⁺ channel β₃ subunit for M₂ receptor-induced inhibition of α_1B and α_1Bβ₃ currents (67 and 7 msec, respectively).

Fig. 5.

Effects of G-protein βγ subunit coexpression on rate and voltage dependence of prepulse facilitation.A, B, Exponential fit of prepulse facilitation by 100 mV prepulse of varying duration in the presence of the G-protein βγ subunit coexpressed with α_1B or α_1Bβ₃. C, D, Boltzmann fit of facilitation of α_1B current amplitude by a 75 msec prepulse to varying voltages in the presence of the G-protein βγ subunit coexpressed with α_1B or α_1Bβ₃.

Figure 5, C and D, shows also shows the voltage dependence of the relief of G-protein βγ subunit-induced current inhibition. A Boltzmann fit of the data revealed an ∼10 mV leftward shift in voltage sensitivity, similar to the shift in voltage-dependent relief of M₂ receptor-induced inhibition. Thus, the Ca²⁺ channel β subunit increases the rate and decreases the voltage necessary for facilitation of G_βγ-inhibited Ca²⁺ currents.

DISCUSSION

Our data demonstrate that the rate of reversal of M₂-mediated inhibition by depolarizing prepulses dramatically increases and the voltage necessary for reversal decreases after coexpression of the Ca²⁺ channel β₃ subunit. We have also confirmed that the G-protein βγ subunit mediates the inhibition of N-type currents and have extended this observation to include both α_1B and α_1Bβ₃ Ca²⁺ currents. In addition, we have demonstrated that the Ca²⁺ channel β₃ subunit increases the rate and decreases the voltage necessary for voltage-dependent reversal of G_βγ-induced inhibition. This results in voltage-dependent relief of inhibition at the moderate voltages used to activate the channel during voltage-clamp experiments and likely explains the reduction in the magnitude of G-protein inhibition of α_1B current after coexpression of the Ca²⁺ channel β₃ subunit. Although the leftward shift in the voltage dependence of facilitation is most likely the result of more rapid unbinding of the G-protein in the presence of the Ca²⁺ channel β₃subunit, caution should be used when interpreting this shift, because the Ca²⁺ channel β₃ subunit also causes a 10 mV leftward shift of peak current in theI–V relation (Roche and Treistman, 1998). Because reversal of G-protein inhibition is thought to result from a conformational change in the channel, produced by voltage, the apparent steeper voltage dependence of activation produced by the Ca²⁺ channel β₃ subunit may contribute to the leftward shift in voltage dependence of facilitation.

Recent findings suggest that some characteristics of voltage-dependent inhibition are a result of the G-protein βγ subunit binding to its consensus site next to the Ca²⁺ channel β subunit binding site (De Waard et al., 1997). The critical amino acids responsible for binding of the Ca²⁺ channel β subunit are not critical for G-protein βγ subunit binding and vice versa. The close proximity of the two sites, however, make modification of the G-protein βγ binding site by the bound Ca²⁺ channel β₃ subunit a likely possibility. However, it should be noted that some groups have reported that the G_βγ consensus binding sequence on the I–II loop of the Ca²⁺ channel is not responsible for mediating the effects of G-proteins (Zhang et al., 1996; Qin et al., 1997). Adjacent proximity of the G_βγ and calcium channel β₃ binding sites within the protein is not a requirement for the model we are proposing. The most reasonable interpretation, combining the information from mutagenesis studies and the results presented here, is that the bound β₃ subunit enhances the G_βγ dissociation rate and thus reduces the magnitude of G-protein inhibition of α_1BCa²⁺ channels.

It is interesting that coexpression of the G-protein α subunit eliminates tonic G-protein inhibition but not M₂-mediated inhibition of α_1B Ca²⁺ channel current. The differential effect of G_α might be explained by a variety of mechanisms. One possibility is that the endogenous free βγ subunit, a portion of which is responsible for tonic inhibition, exists at levels that saturate the exogenous free α subunit, so that the expressed G_α subunit cannot buffer the additional βγ subunit liberated by activation of the muscarinic receptor. In support of this mechanism, we find that M₂-mediated inhibition is only partially blocked by NEM, an agent that uncouples the G-protein α subunit from receptor activation (Jakobs et al., 1982; Nakajima et al., 1990), when no exogenous G-protein subunits are present. However, after coexpression of the NEM-sensitive Gαi (Shapiro et al., 1994b), the M₂-mediated inhibition is almost entirely blocked by NEM (data not shown). This result is predicted by a model in which exogenous G_α subunits form inactive heterotrimers with the tonically active endogenous G_βγ subunits. These newly formed heterotrimers are then activated after binding of an agonist to the M₂ receptor, liberating the G_βγ subunit, and overwhelming the buffering capacity of the coexpressed G_α subunits.

Regulation of responsiveness to G-proteins at the level of the ultimate target may be a widely used mechanism, enabling a channel or other protein to regulate its sensitivity to modulation while maintaining its basal properties. This mechanism may be necessary in situations in which a modulatory signal is greatly amplified or when the signal has a large number of ultimate targets. In such situations, downregulation of the receptor itself may have unwanted consequences or may be ineffective because of amplification of the signal.

Functional Ca²⁺ channels may exist in the absence of a component auxiliary β subunit (De Waard and Campbell, 1995). Additionally, a recent report (Qin et al., 1997) suggests that a second calcium channel β subunit binding site is located on the C terminal of α_1A, α_1B, and α_1E calcium channels and that this site is responsible for the antagonism of G-protein inhibition of these channels by the calcium channel β subunit. This site is distinct from that believed to be responsible for high expression and insertion of channels. Thus, it is possible that differential occupancy of this second site by the channel β subunit could serve a regulatory function, consistent with our observations with cloned channels. The increased voltage sensitivity of the inhibition observed after coexpression of the Ca²⁺ channel β₃ subunit may play an important role in the regulation of transmitter release in response to high-frequency or long-duration action potentials (Brody et al., 1997), in which depolarization of the presynaptic terminal would reach levels sufficient to relieve G-protein inhibition of Ca²⁺channels controlling release.