Cross-talk between G-protein and Protein Kinase C Modulation of N-type Calcium Channels Is Dependent on the G-protein β Subunit Isoform*

The modulation of N-type calcium current by protein kinases and G-proteins is a factor in the fine tuning of neurotransmitter release. We have previously shown that phosphorylation of threonine 422 in the α1B calcium channel domain I-II linker region resulted in a dramatic reduction in somatostatin receptor-mediated G-protein inhibition of the channels and that the I-II linker consequently serves as an integration center for cross-talk between protein kinase C (PKC) and G-proteins (Hamid, J., Nelson, D., Spaetgens, R., Dubel, S. J., Snutch, T. P., and Zamponi, G. W. (1999) J. Biol. Chem. 274, 6195–6202). Here we show that opioid receptor-mediated inhibition of N-type channels is affected to a lesser extent compared with that seen with somatostatin receptors, hinting at the possibility that PKC/G-protein cross-talk might be dependent on the G-protein subtype. To address this issue, we have examined the effects of four different types of G-protein β subunits on both wild type and mutant α1Bcalcium channels in which residue 422 has been replaced by glutamate to mimic PKC-dependent phosphorylation and on channels that have been directly phosphorylated by protein kinase C. Our data show that phosphorylation or mutation of residue 422 antagonizes the effect of Gβ1 on channel activity, whereas Gβ2, Gβ3, and Gβ4 are not affected. Our data therefore suggest that the observed cross-talk between G-proteins and protein kinase C modulation of N-type channels is a selective feature of the Gβ1 subunit.

The modulation of calcium channel activity by activation of intracellular messenger pathways is a key mechanism for fine tuning calcium entry into neurons. For example, the activation of protein kinase C has been shown to mediate an up-regulation of N-type calcium currents in intact neurons (1,2) and in transient expression systems (3,4). In contrast, the direct 1:1 binding of G protein ␤␥ subunits to the domain I-II linker region of N-type, P/Q-type, and possibly R-type calcium channels results in a depression of current activity (5-8) (reviewed in Refs. 9 and 10), which can be reversed by strong membrane depolarization (10 -12). Different types of calcium channels are modulated by G-proteins to different extents, such that N-type channels are typically inhibited more effectively than P/Q-type channels (13)(14)(15)(16). There is also increasing evidence that the degree of inhibition is dependent on the G-protein ␤ subunit species (16 -18). Finally, it has been shown that protein kinase C-dependent phosphorylation of the N-type calcium channel ␣ 1 subunit antagonizes receptor-mediated G-protein inhibition of the channel (1,2,12,19). This phenomenon (termed PKC 1 /Gprotein cross-talk) appears to be mediated by a single threonine residue (Thr-422) in the ␣ 1B domain I-II linker region (4). For somatostatin receptor-induced G-protein inhibition of N-type calcium channels, mutation of Thr-422 to glutamic acid mimics the antagonistic effect of protein kinase C on G protein inhibition, whereas a switch to alanine precludes the occurrence of PKC/G-protein cross-talk (4).
Here we have examined the dependence of PKC/G-protein cross-talk on the nature of the G protein ␤ subunit isoform. Using transient expression of either wild type or mutant (T422E) N-type calcium channels in combination with various G-protein ␤ subunit isoforms (G␤ 1 , G␤ 2 , G␤ 3 , and G␤ 4 ), we show that the effect of only the G␤ 1 isoform is reduced in the "permanently phosphorylated" mutant N-type channel. PKC/ G-protein cross-talk thus appears to be a selective feature of the G␤ 1 subunit isoform. In view of the notion that different types of G-protein-coupled receptors may combine with specific subsets of G␣␤␥ combinations (20), this may suggest that the extent of cross-talk occurring in intact neurons could be dependent on the type of neurotransmitter involved, thus providing a mechanism for the fine tuning of calcium homeostasis.

EXPERIMENTAL PROCEDURES
Calcium Channel and G-protein cDNAs-The calcium channel cDNA constructs (␣ 1B , ␣ 1B (T422E), ␤ 1b , ␣ 2 -␦) were the same as those discussed previously in Hamid et al. (4). Wild type constructs were kindly donated by Dr. T. P. Snutch. The cloning of the various G␤-subunits is described * This work was supported by an operating grant from the Canadian Institutes of Health Research (CIHR) (to G. W. Z.) and through a Scholarship Award from the EJLB Foundation (to G. W. Z.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Recipient of a studentship award from the Heart and Stroke Foundation of Canada.
¶ Recipient of a postdoctoral fellowship award from the Savoy Foundation.
Transient Transfection into tsa-201 Cells-Human embryonic kidney tsa-201 cells were grown in standard Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 0.5 mg/ml penicillin/ streptomycin. Cells were grown at 37°C to 85% confluency, split with trypsin-EDTA, and plated on glass coverslips at 5-10% confluency about 12 h before transfection. Immediately before transfection, the medium was replaced with fresh Dulbecco's modified Eagle's medium, and a standard calcium phosphate protocol was used to transiently transfect the cells with cDNA constructs encoding for mutant and wild type calcium channel ␣ 1 , ␤ 1b , and ␣ 2 -␦ subunits and green fluorescent protein as an expression marker and as appropriate with G-protein subunits or the -opioid receptor. After 12 h, cells were washed with fresh medium, allowed to recover for 12 h, and then incubated at 28°C in 5% CO 2 for 1-3 days prior to recording.
Electrophysiology-The exact whole cell recording procedures have been described in detail previously (16). The external recording solution was comprised of 20 mM BaCl 2 , 1 mM MgCl 2 , 10 mM HEPES, 40 mM tetraethylammonium chloride, 10 mM glucose, and 65 mM CsCl (pH 7.2), recording pipettes were filled with 108 mM cesium methanesulfonate, 4 mM MgCl 2 , 9 mM EGTA, 9 mM HEPES (pH 7.2) and showed typical resistances of 2 to 4 megaohms. For experiments involving opioid receptor activation, the intracellular solution was supplemented with 40 M GTP. DAMGO (purchased from RBI) was dissolved in water at a stock concentration of 2 mM, diluted into the recording solution at a final concentration of 1 M, and perfused directly into the vicinity of the cells via a gravity-driven microperfusion system. Typically, cells were held at Ϫ100 mV and currents were elicited upon depolarizations to various test potentials.
Tonic G-protein inhibition of the channels was assessed by application of strong depolarizing prepulses (to ϩ150 mV) followed by a test depolarization to ϩ20 mV (see Refs. 8 and 16). The time course of development of prepulse relief was determined by varying the duration of the prepulse (⌬t1) while leaving the duration between the prepulse and the test pulse constant at 5 ms (see Fig. 2). The reinhibition kinetics were determined by applying a 50-ms prepulse, followed by a test pulse spaced from the prepulse at variable intervals (⌬t2). The total degree of prepulse relief was obtained by extrapolating the exponential decay of the prepulse effect back to time t2 ϭ 0 (see Ref. 16), thus allowing us to assess the degree of prepulse relief without contamination from reinhibition. For the experiment in Fig. 2, the data were obtained at a fixed interval of 4 ms between the prepulse and the test pulse.
For experiments involving PKC-dependent phosphorylation of the channels, the phorbol ester PMA was dissolved in Me 2 SO at a stock concentration of 2 mM, diluted into the recording solution at a final concentration of 30 nM, and perfused directly onto the cell with a gravity-driven microperfusion system.
Western Blot Analysis-Western blots on cell lysates generated from transfected or sham-transfected tsa-201 cells were carried out as described in detail by Jarvis et al. (21). Antibodies to G protein ␤ subunits were purchased from Santa Cruz Laboratories (anti-G␤ 2 and anti-G␤ 3 ) and from Transduction Laboratories (anti-G␤ 1/4 ). Immunoblots were subjected to chemiluminescence analysis using ECL plus (Amersham Pharmacia Biotech) and detected on film.
Data Analysis and Statistics-Data were analyzed using Clampfit software. Preparation of figures and statistical analysis was performed via SigmaPlot (Jandel Scientific). Error bars are standard errors, p values reflect Student's t tests.

RESULTS AND DISCUSSION
We have previously shown that activation of protein kinase C in tsa-201 cells reduces a somatostatin receptor-induced Gprotein inhibition of N-type calcium channels (4). A point mutation in the domain I-II linker region of the channel (T422E) mimicked the effects of PKC activation, whereas a substitution to alanine precluded the effects of PKC on G-protein inhibition. We concluded that PKC/G-protein cross-talk occurs via phosphorylation of threonine 422 (4). To assess whether this effect was selective for somatostatin receptors, we coexpressed -opioid receptors with wild type or mutant (T422E) N-type (␣ 1B , ␤ 1b , ␣ 2 -␦) calcium channels and applied 1 M DAMGO to trigger G protein inhibition of the channels via the receptor. As seen in Fig. 1, wild type channels underwent a robust, reversible inhibition by 56 Ϯ 4% in response to opioid receptor activation. In contrast, the T422E mutant exhibited a reduced degree of G-protein inhibition (32 Ϯ 3%), in accord with our assertion that a negative charge (which permanently mimics phosphorylation) in position 422 reduces G protein efficacy. A comparison with our previous work with somatostatin receptors (4), however, shows that the somatostatin receptor-mediated inhibition of N-type calcium channels was significantly more strongly affected than the -opioid response (20 Ϯ 3% versus 32 Ϯ 3%, p Ͻ 0.05), whereas the wild type channels were similarly inhibited by both pathways (DAMGO, 56 Ϯ 4%; somatostatin, 53 Ϯ 5%, p Ͼ 0.05). Although the latter observation indicates that both receptor types are equally effectively coupled to wild type N-type channels in tsa-201 cells, the observation with the T422E mutant suggests that there are nonetheless differences in the way the two receptor types couple to the channels. It is possible that the two receptor types couple to the FIG. 1. Cross-talk between G-protein modulation and PKC-dependent phosphorylation of N-type calcium channels. A, schematic representation of the N-type calcium channel ␣ 1 subunit. The ␣ 1 subunit consists of four homologous domains that are linked by large intracellular loops (30). The cytoplasmic linker between domains I and II is one of the prime targets for binding of G␤␥ subunits (12,22). Residue Thr-422 forms the core of a protein kinase C consensus site (see amino acid sequence) which underlies G-protein PKC cross-talk (4). B, time course of -opioid receptor-mediated G-protein inhibition of wildtype and mutant (T422E) N-type calcium channels. Cells were held at Ϫ100 mV and depolarized to ϩ20 mV. Note that the effects are fully reversible. C, current records illustrating the effects of T422E mutation on -opioid receptor-mediated G-protein inhibition of N-type calcium channels. The solid line indicates the base line. D, comparison of the effects of T422E mutation on -opioid receptor-and somatostatin receptor-mediated inhibition of N-type calcium channels. The somatostatin data were taken from our previous work (4). Note that a mutation that permanently mimics phosphorylation of residue 422 affects the somatostatin receptor-mediated inhibition significantly more strongly than that induced by -opioid receptor activation. channels via different G-protein heterotrimer compositions, which may be differentially affected by the presence of the T422E mutation.
The key G-protein species involved in direct inhibition of N-type calcium channels is the G␤ subunit (5,6,16,18). To date, five different types of G␤ subunits have been identified in mammalian brain (17), and we have recently shown that Ntype channels expressed in tsa-201 cells are most effectively inhibited by G␤ 1 and G␤ 3 , whereas G␤ 4 and G␤ 2 mediate a somewhat smaller inhibition, and G␤ 5 is ineffective and is thus not further considered here (16).
To test the possibility that PKC/G-protein cross-talk might preferentially affect a subset of the G␤ subunit isoforms, we coexpressed wild type channels with one of four different G␤ subtypes (ϩG␥ 2 ) and used a strong depolarizing prepulse to compare the resulting tonic G-protein inhibition before and after phosphorylation of the channel by protein kinase C (elicited by 2-min application of 30 nM PMA). As shown in Fig. 2, in cells expressing G␤ 1 , the degree of prepulse relief became reduced from 2.4 Ϯ 0.3 to 1.4 Ϯ 0.1 (n ϭ 7, p Ͻ 0.02, paired t test) following PMA treatment, whereas inhibition by the three other G-protein ␤ subunit subtypes was not significantly al-tered (Fig. 2B). Thus, these data suggest that PKC/G-protein cross-talk selectively affects G␤ 1 -mediated responses.
Although N-type calcium channels expressed in tsa-201 cells show only negligible background G-protein inhibition in the absence of exogenous G␤␥ (16), it is true that these cells contain endogenous G protein ␤␥ subunits that are presumably complexed as ␣␤␥ heterotrimers. To assess the likelihood of contamination of our results by endogenously present G␤ proteins, we carried out Western blot analysis of control cells and cells transfected with either one of the four G␤ subunits. As seen in Fig. 2C, exogenous expression of each of the four subtypes tested resulted in a substantial increase in G␤ levels compared with those found in control cells, confirming that the exogenously expressed subunits are the predominant G-protein species in transfected cells.
To obtain an indication of the change in affinity of the channels for the G-proteins, which occurs after phosphorylation, we utilized dynamic prepulse protocols to determine the time constants of recovery from inhibition (by varying prepulse duration) and reinhibition after the prepulse (by varying the duration between the prepulse and the test pulse, see Fig. 3A). However, after prolonged application of PMA (Ͼ3 min), we noted that current levels began to run down, which would interfere with the accuracy of these types of experiments. Hence, we chose to utilize the T422E mutant for our kinetic measurements. Fig. 3B shows that the T422E mutant behaved like the phosphorylated wild type channel with regard to the G␤ subtype dependence of cross-talk, such that the mutation induced a selective decrease in the degree of prepulse relief seen in the presence of G␤ 1 (wild type, 2.8 Ϯ 0.3; T422E, 1.9 Ϯ 0.1; p Ͻ 0.05) whereas the facilitation ratios of the three remaining G-protein ␤ subunit isoforms did not differ significantly from those obtained with the wild type channels (note, however, that the data shown in Figs. 2B and 3B were obtained by slightly different protocols, and hence, absolute values for the degree of prepulse relief are not directly comparable). As seen in Fig. 3C, the T422E mutation selectively reduced the time constant of recovery from G␤ 1 inhibition (wild type, 11.0 Ϯ 0.6 ms; T422E, 8.0 Ϯ 0.6 ms; p Ͻ 0.05), indicating that one of the consequences of the T422E mutation is a slight destabilization of the G␤ 1channel complex. In addition, the time course of G␤ 1 reinhibition after the prepulse (Fig. 3D) was slowed 3-fold as a result of the T422E mutation (wild type, 18.5 Ϯ 2.4 ms; T422E, 51.7 Ϯ 4.7 ms; p Ͻ 0.05). Interestingly, a small increase in the time course of reinhibition was also observed with G␤ 2 (WT, 46.2 Ϯ 6.3 ms; T422E, 27.8 Ϯ 4.1 ms; p Ͻ 0.05), indicating that the T422E mutation may exert subtle effects on this subunit. Overall, however, the data shown in Figs. 2 and 3 indicate that a mutagenically phosphorylated threonine residue in position 422 affects predominantly the inhibition of N-type calcium channels by G␤ 1 subunits.
To date, little is known about the association of individual types of seven helix transmembrane receptors with specific subsets of G-protein subunits. It has been shown, however, that antisense depletion of G␤ 1 subunits abolishes somatostatin receptor signaling in rat pituitary GH 3 cells (22). It is thus likely that somatostatin receptors exclusively couple to G␤ 1 subunits, which can account nicely for our observation that the T422E mutation dramatically reduced the G-protein inhibition induced by both overexpression of G␤ 1 and upon activation of somatostatin receptors. In contrast, it is possible that the -opioid receptor couples to N-type calcium channels through more than one type of G-protein ␤ subunit. Whereas the observation that the opioid response was reduced in the T422E mutant would suggest that at least part of the opioid signaling FIG. 2. A, current records illustrating the effects of PKC-dependent phosphorylation of N-type calcium channels on the degree of prepulse relief seen with either G␤ 1 or G␤ 3 . The traces were arbitrarily scaled to overlap at the peak current levels obtained after the prepulse, and hence no legends for current levels are provided. Note that the degree of prepulse facilitation obtained with G␤ 1 becomes diminished following a 2-min application of 30 nM PMA. B, effect of 30 nM PMA on the degree of prepulse relief obtained with four types of G-protein ␤ subunits. In this experiment, a 50-ms prepulse to ϩ150 mV was applied, followed by a 4-ms repolarization to Ϫ100 mV and a test pulse to ϩ20 mV. Note that PMA selectively affects G␤ 1 -mediated responses. Error bars are standard errors, the asterisk denotes significance at the 0.05 level (paired t test). C, Western blot analysis of cell lysate obtained from either shamtransfected tsa-201 cells or from tsa-201 cells transfected with one of four G-protein ␤ subunits. The exogenous expression of either of the four G␤ subunits results in a substantial increase in G␤ levels despite the notion that our transfection efficiency is typically around 25%. The blots for G␤ 1 and G␤ 4 were probed with an antibody that recognizes both of these G-protein subtypes.
is mediated by G␤ 1 , our data showing that the opioid and somatostatin responses were not equally affected by the T422E substitution (Fig. 1D), however, support a mechanism in which -opioid receptors may couple to a mixed population of Gprotein ␤ subunit isoforms. This scenario would account for the intermediate response we observed.
The data shown in Fig. 3, C and D, provide some insights into the molecular mechanisms by which cross-talk may occur. It is now widely accepted that G␤␥ physically interacts with the N-type calcium channel domain I-II linker region (8,23,24), thus phosphorylation events or amino acid substitutions occurring in this region may affect the binding interactions between the channel and the G␤␥ dimers. Consistent with such an effect, the T422E substitution resulted in a significant decrease in the time constant for development of facilitation during the prepulse and an increase in the time constant for reinhibition after the prepulse. In view of evidence that G proteins must physically dissociate from the channel during the prepulse (8), these changes in kinetics likely reflect a change in the Gprotein association and dissociation kinetics, and thus an over-all reduction in the affinity of the channel for G␤ 1 , which may also account for the reduction in the degree of prepulse relief seen with the T422E mutant. Our experiments do not permit us to provide an absolute value for the PKC-induced changes in the equilibrium dissociation constant between the G proteins and the channels, because the kinetics for recovery from inhibition during the prepulse (ϩ150 mV) and reinhibition after the prepulse (i.e. repolarization to Ϫ100 mV) were obtained at different voltages. Nonetheless, based on the 3-fold decrease in reinhibition kinetics and the 1.4-fold speeding of the recovery time constant, we estimate that the presence of the T422E mutation may perhaps result in an ϳ5-fold change in G␤ 1 affinity.
An important aspect to consider is control over the G protein expression levels in our experiments. Our Western blot analysis shows that each of the four G␤ subtypes expressed well in tsa-201 cells, indicating that the exogenously expressed Gprotein subunits are much more abundant than endogenously present G␤, although it is difficult to predict relative expression levels among the four G␤ subtypes from Western blots as antibody sensitivity may vary. In the experiments shown in Fig. 2B, the inhibition of nonphosphorylated and phosphorylated channels by each G␤ subtype was studies in the same cell, and hence, at a constant G␤ concentration. In the experiments shown in Fig. 3B, the effects of each G␤ subtype on wild type and mutant channels needed was assessed in different cells. However, for each given G␤ subtype, wild type and mutant channels were studied under identical conditions, thus attributing any changes in channel inhibition to residue 422 rather than G␤ levels. Both types of experiments resulted in essentially the same result, namely that only G␤ 1 -mediated responses were affected by phosphorylation/mutation of the channel.
Although in vivo evidence supporting our conclusions is still lacking, it is tempting to speculate about the implications of observations for neurotransmission: calcium influx through Ntype and P/Q-type calcium channels is essential for the fast release of neurotransmitter (25). Regulation of presynaptic calcium channel activity by cytoplasmic messenger molecules is thus an essential means for the precise control of neurotransmission. The activation of opioid receptors, for example, depresses synaptic activity (13,26), and tonic G-protein inhibition at presynaptic nerve termini contributes to paired pulse facilitation (27)(28)(29). We have previously presented evidence that integration of protein kinase C and G␤␥ pathways by the N-type calcium channel ␣ 1 subunit can produce multiple levels of current activity (4). If our results are extrapolated to an in vivo situation, the notion that the extent of this integration appears to depend on the type of G␤ subunit isoform present may provide additional avenues by which neurotransmitter receptors may regulate calcium homeostasis in the presynapse. Together with the notion that different types of G-protein ␤ subunits do not only affect N-type channel activity to different degrees (16 -18) but also show pronounced differences in their abilities to inhibit P/Q-type calcium channels (16), the unique activation of specific G␤␥ combinations by different types of seven-helix transmembrane receptors could provide a highly complex regulatory mechanism for the fine tuning of presynaptic calcium levels and, thus, neurotransmitter release. FIG. 3. Dependence of PKC/G-protein cross-talk on the G␤ subunit isoform. The wild type data were taken from our previous work (16). A, prepulse paradigms used to determine the kinetics of removal of G-protein inhibition during the prepulse (i.e. recovery, left panel), and those of reinhibition following the prepulse (PP) (right panel). B, degree of G-protein inhibition of wild type and T422E mutant N-type calcium channels in form of facilitation ratios and as a function of G-protein ␤ subunit isoform. The facilitation ratios (I ϩ PP)/I(ϪPP) were obtained as described under "Experimental Procedures" and in Arnot et al. (16), which relies on back extrapolation of the data to t 2 ϭ 0 ms (hence the difference in absolute facilitation levels compared with Fig. 2B). Note that the T422E mutation selectively affects G␤ 1 -mediated inhibition. C, the T422E mutation selectively increases the rate of development of facilitation for the G␤ 1 subtype. D, T422E reduces the rate of G-protein reinhibition for G␤ 1 and G␤ 2 but does not affect the reinhibition rates observed with G␤ 3 and G␤ 4 .