Abstract
Experiments using heterologous overexpression indicate that regulator of G-protein signaling (RGS) proteins play important roles in Gβγ-mediated ion channel modulation. However, the roles subserved by endogenous RGS proteins have not been extensively examined because tools for functionally inhibiting natively expressed RGS proteins are lacking. To address this void, we used a strategy in which GαoA was rendered insensitive to pertussis toxin (PTX) and RGS proteins by site-directed mutagenesis. Either PTX-insensitive (PTX-i) or both PTX- and RGS-insensitive (PTX/RGS-i) mutants of GαoA were expressed along with Gβ1 and Gγ2 subunits in rat sympathetic neurons. After overnight treatment with PTX to suppress natively expressed Gα subunits, voltage-dependent Ca2+ current inhibition by norepinephrine (NE) (10 μm) was reconstituted in neurons expressing either PTX-i or PTX/RGS-i GαoA. When compared with neurons expressing PTX-i GαoA, the steady-state concentration–response relationships for NE-induced Ca2+ current inhibition were shifted to lower concentrations in neurons expressing PTX/RGS-i GαoA. In addition to an increase in agonist potency, the expression of PTX/RGS-i GαoA dramatically retarded the current recovery after agonist removal. Interestingly, the alteration in current recovery was accompanied by a slowing in the onset of current inhibition. Together, our data suggest that endogenous RGS proteins contribute to membrane-delimited Ca2+ channel modulation by regulating agonist potency and kinetics of G-protein-mediated signaling in neuronal cells.
- calcium channel
- G-protein
- Gα
- Gβγ
- intranuclear injection
- RGS protein
- sympathetic neuron
- voltage-dependent inhibition
Biochemical studies indicate that members of the regulator of G-protein signaling (RGS) family of proteins serve as GTPase-activating proteins (GAPs) for Gαi and Gαq but not for Gαs (for review, see Dohlman and Thorner, 1997;Koelle, 1997; Berman and Gilman, 1998). Recent evidence implicates RGS proteins as key components in Gβγ-mediated modulation (Wickman and Clapham, 1995; Ikeda and Dunlap, 1999) of inwardly rectifying K+ (GIRK) and N-type Ca2+ channels. The potential roles of RGS proteins in Gβγ-mediated ion channel modulation were deduced from the discrepancy between the fast kinetics of modulation in vivo and the relatively slow intrinsic GTPase activities of Gαo/i in vitro. For example, after agonist removal, muscarinic (M2) receptor-activated GIRK currents in atrial myocytes deactivate ∼40-fold faster than the GTP hydrolysis rate of pertussis toxin (PTX)-sensitive Gα subunits in vitro (Breitwieser and Szabo, 1988; Kurachi, 1995; Shui et al., 1995). Logically, these discrepancies might be reconciled by invoking the ability of natively expressed RGS proteins to stimulate the intrinsic GTPase of Gα subunits thereby accelerating termination of Gβγ-mediated signaling by favoring reformation of the heterotrimeric state. In support of this notion, heterologous expression of RGS proteins has been shown to accelerate the activation and deactivation kinetics of GIRK currents inXenopus oocytes (Doupnik et al., 1997; Saitoh et al., 1997, 1999; Kovoor et al., 2000). In addition, heterologous expression of RGS proteins have been shown to alter the magnitude and kinetics of N-type Ca2+ channel inhibition (Jeong and Ikeda, 1998; Melliti et al., 1999). To date, however, the ability of natively expressed RGS proteins to influence Gβγ-mediated ion channel modulation in neurons has not been established primarily because few tools are available for inhibiting endogenous RGS proteins.
In the present study, we addressed this void in our knowledge by developing a strategy based on a combination of previously identified point mutations in Gα subunits. First, GαoAwas rendered insensitive to the actions of PTX by mutating the C-terminal cysteine in which PTX-mediated ADP ribosylation occurs (Milligan, 1988). In this way, natively expressed Gα subunits could be uncoupled from receptors by PTX treatment, thereby isolating the action of the mutated Gα subunit (Taussig et al., 1992; Hunt et al., 1994; Senogles 1994; Kozasa et al., 1996; Jeong and Ikeda, 2000). Second, point mutations were introduced into the switch I and II regions of GαoA that have been identified previously to prevent RGS binding and GAP activity without altering GDP release and basal GTP hydrolysis in other Gα subunits (Lan et al., 1998; Natochin and Artemyev, 1998a) (but see DiBello et al., 1998). Using this strategy, we examined alterations in the magnitude and kinetics of voltage-dependent (VD) N-type Ca2+ current inhibition reconstituted in PTX-treated rat sympathetic neurons expressing the GαoA mutants. Our results indicate that endogenous RGS proteins in sympathetic neurons play an integral role in controlling the magnitude and speed of neurotransmitter-induced N-type Ca2+ channel modulation.
Parts of this paper have been published previously in abstract form (Jeong and Ikeda, 1999a).
MATERIALS AND METHODS
Preparation of sympathetic neurons. Superior cervical ganglion (SCG) neurons were enzymatically dissociated as described previously (Ikeda, 1997). Briefly, adult (200–350 gm) male Wistar rats were decapitated using a laboratory guillotine. The SCG were dissected free of the carotid bifurcation and placed in cold (4°C) HBSS. The ganglia were desheathed, cut into small pieces, and transferred to the oxygenated Earle's balance salt solution (EBSS), pH 7.4, containing 0.6 mg/ml collagenase type D (Boehringer Mannheim, Indianapolis, IN), 0.4 mg/ml trypsin (TRL type; Worthington, Lakewood, NJ), and 0.1 mg/ml DNase Type I (Sigma, St. Louis, MO) in a 25 cm2 tissue culture flask. The EBSS was modified by adding 3.6 gm/l glucose and 10 mmHEPES. After incubation for 60 min in a water bath shaker at 35°C, neurons were dispersed by vigorous shaking of the flask. After centrifugation twice for 6 min at 50 × g, the neurons were resuspended in Minimum Essential Medium (Mediatech, Inc., Herndon, VA) containing 10% fetal calf serum (Atlanta Biologicals, Atlanta, GA) and 1% glutamine/penicillin–streptomycin solution (Life Technologies, Grand Island, NY). Neurons were then plated onto polystyrene culture dishes (35 mm), coated with poly-l-lysine, and maintained in a humidified atmosphere of 95% air–5% CO2 at 37°C. As appropriate, neurons were incubated overnight (16–20 hr) with 500 ng/ml PTX (List Biologic, Campbell, CA).
Construction and expression of RGS-insensitive mutants of Gα subunits. Previously, we have generated a PTX-insensitive GαoA (PTX-i GαoA) by introducing a cysteine (C) to glycine (G) mutation in the residue −4 from the C terminus (C351G) (Jeong and Ikeda, 2000). Using PTX-i GαoA as a template, additional point mutations (DiBello et al., 1998; Lan et al., 1998; Natochin and Artemyev, 1998a) were introduced to render GαoA both PTX- and RGS-insensitive (PTX/RGS-i GαoA). The following mutations were introduced into GαoA(C351G) using the GeneEditor site-directed mutagenesis kit (Promega, Madison, WI) per instructions of the manufacturer: G184S, S207D, and G184S/S207D. The following primers were used in constructing the mutations: G184S, 5′-GTCAAAACAACTTCCATCGTAGAAACCCAC-3′; and S207D, 5′-TCGGGGGCCAGCGAGATGAACGCAAGAAGTG-3′. The sequence of each construct was verified with an automated DNA sequencer (ABI 310; Applied Biosystems, Foster City, CA). Either PTX-i GαoA or PTX/RGS-i GαoAsubunits were coexpressed with Gβ1 and Gγ2 subunits (denoted Gβ1γ2, hereafter) by intranuclear microinjection as described previously (Ikeda, 1996,1997). All G-protein subunits were expressed from the cytomegalovirus promoter-driven vector, pCI (Promega). Neurons were used within 24 hr after intranuclear injection of vectors. Injected neurons were identified by fluorescence from coexpressed jellyfish green fluorescent protein (EGFP; Clontech, Palo Alto, CA).
Electrophysiology. Ca2+ channel currents were recorded using the whole-cell variant of the patch-clamp technique (Hamill et al., 1981) as described previously (Ikeda, 1991;Ikeda et al., 1995). To isolate Ca2+currents, patch electrodes were filled with a solution containing (in mm); 120N-methyl-d-glucamine methanesulfonate (MS), 20 tetraethylammonium (TEA)-MS, 20 HCl, 11 EGTA, 1 CaCl2, 10 HEPES, 4 MgATP, 0.3 Na2GTP, and 14 creatine phosphate, pH 7.2 (297 mOsm/kg H2O). The external recording solution contained (in mm): 145 TEA-MS, 10 HEPES, 10 CaCl2, 15 glucose, and 0.0003 tetrodotoxin, pH 7.4 (318 mOsm/kg H2O). All experiments were performed at room temperature (21–24°C). Norepinephrine (NE) was applied to single neurons via a gravity-fed fused silica capillary tube connected to an array of seven polyethylene tubes. The outlet of the perfusion system was located within 100 μm of the cell. Drug application was started by switching the control external solution to a drug solution. Complete solution exchange occurred within <1 sec.
Data analysis. The concentration–response curves were plotted as values normalized to the inhibition produced at the maximal [NE] (30 μm). The concentration–response curves were fit to the following Hill equation: I = 1/[1 + (k/[NE])n], whereI, k, [NE], and n are normalized inhibition, a constant, NE concentration, and Hill factor, respectively. The EC50 values were calculated using the following relationship: EC50 =k−n. The “on–off” times of the NE-induced Ca2+ current inhibition were approximated by fitting a polynomial function to the time courses and interpolating the appropriate values (i.e.,t0.5 andtrise). All curve fitting was performed with the IGOR PRO data analysis package (WaveMetrics, Lake Oswego, OR). Data are presented as means ± SEM. Student'st test (unpaired) or ANOVA followed by a post hoc Dunnett's test, as appropriate, were applied to the data to determine statistical significance. p < 0.05 was considered significant.
RESULTS
Reconstitution of NE-induced Ca2+ current inhibition after expression of PTX- and RGS-insensitive GαoA subunits
As a first step toward elucidating the potential roles of endogenous RGS proteins in Ca2+ channel modulation, we tested whether heterologous expression of PTX/RGS-i GαoA subunits reconstituted NE-induced VD Ca2+ current inhibition in SCG neurons after uncoupling of endogenous Go/i-proteins by PTX treatment. GαoA was selected to be rendered RGS-insensitive because Go has been implicated as the primary signaling element coupling α2-adrenergic receptors (α2-ARs) to N-type Ca2+ channels (Caulfield et al., 1994) in sympathetic neurons.
Figure 1 illustrates the effects of PTX/RGS-i GαoA subunits on NE-induced Ca2+ current inhibition when expressed in SCG neurons. Ca2+ currents were evoked from a holding potential of −80 mV with a double-pulse protocol consisting of two identical test pulses to +10 mV separated by a large depolarizing conditioning pulse to +80 mV (Fig.1A). In uninjected control neurons, NE (10 μm) inhibited Ca2+currents by 63 ± 2% (n = 30) (Figs.1A, 2). NE-induced Ca2+ current inhibition displayed the hallmarks of VD inhibition (Ikeda and Dunlap, 1999), i.e., slowed activation kinetics in the prepulse and an enhanced postpulse amplitude (Figs. 1A, 2). Prepulse facilitation ratio (PFR), defined as the ratio of the postpulse to prepulse current amplitude, increased from 1.20 ± 0.02 to 2.40 ± 0.10 (n = 30) after NE application. This form of N-type Ca2+ channel modulation has been shown to be mediated by the Gβγ subunit (Herlitze et al., 1996; Ikeda, 1996).
Overnight treatment of the neurons with PTX (500 ng/ml) greatly attenuated NE-induced Ca2+ current inhibition (8 ± 1%, n = 25) (Figs.1B, 2), confirming the involvement of a PTX-sensitive G-protein(s) in the signaling pathway (Hille, 1994). To reconstitute NE-induced Ca2+ current inhibition in PTX-treated neurons, either PTX-i GαoA or PTX/RGS-i GαoA subunits were coexpressed with Gβ1γ2 in SCG neurons using an intranuclear microinjection technique. As shown previously (Jeong and Ikeda, 2000), “balanced” expression of GαoA mutants and Gβ1γ2 was critical for successful reconstitution on the time scale used in these experiments (<24 hr after injection of cDNA). Briefly, if expression of Gα mutants greatly exceeded that of Gβγ, basal facilitation and NE-induced Ca2+ current inhibition were ablated. This effect likely reflects sequestration of free Gβγ subunits (both endogenous and exogenous) by the GDP-bound Gα (Ikeda, 1996; Jeong and Ikeda 1999b). In contrast, if expression of Gβγ exceeded that of Gα, a significant tonic inhibition (as indicated by increased basal facilitation) was observed. This result presumably reflects interaction of “free” or excess Gβγ with N-type Ca2+ channels (Herlitze et al., 1996;Ikeda, 1996; Garcı́a et al., 1998; Ruiz-Velasco and Ikeda, 2000). Thus, we have defined a basal PFR range of 1.02–1.50 as indicating adequate balance between PTX-i GαoA and Gβγ subunits (Jeong and Ikeda, 2000). In general, we applied the same criteria for PTX/RGS-i GαoA mutants. In a few neurons expressing GαoA(G184S:C351G), however, reconstitution of Ca2+ current inhibition occurred with basal PFR of <1. NE (10 μm) inhibited Ca2+ currents by 52 ± 3% (n = 9) when GαoA(C351G) was coexpressed with Gβ1γ2in PTX-treated neurons. The characteristics of the Ca2+ current inhibition reconstituted by the PTX-i GαoA were basically similar to those of uninjected controls neurons in terms of voltage dependence of the inhibition (PFR increased from 1.15 ± 0.05 to 1.90 ± 0.09 after NE application) and relatively rapid on–off kinetics of NE action. These results are consistent with previous findings using the GαoA(C351G) mutant (Jeong and Ikeda, 2000). Similarly, coexpression of the PTX/RGS-i GαoAsubunits GαoA(G184S:C351G) or GαoA(S207D:C351G), along with Gβ1γ2, resulted in successful reconstitution of NE-induced Ca2+ current inhibition (Fig. 1). These results demonstrate that introduction of the additional point mutations did not affect the ability of GαoA(C351G) to form heterotrimers or couple receptors to Ca2+ channels. On average, NE (10 μm) inhibited Ca2+currents by 69 ± 1 (n = 6) and 70 ± 2% (n = 8) in neurons expressing GαoA(G184S:C351G) and GαoA(S207D:C351G), respectively (Fig. 2). It should be noted that these magnitudes of inhibition were significantly larger than those of uninjected or PTX-i GαoA-expressing neurons (p < 0.05). The voltage dependence of the reconstituted Ca2+ current inhibition remained intact as evidenced by the kinetic slowing of current activation and increase in PFR after NE application [1.01 ± 0.02 to 2.23 ± 0.06 for GαoA(G184S:C351G); 1.19 ± 0.05 to 2.26 ± 0.07 for GαoA(S207D:C351G)] (Fig. 2). However, the most striking feature produced by expression of the PTX/RGS-i GαoA subunits was a sluggish recovery from inhibition after agonist removal (Fig.1D,E). Interestingly, the onset of agonist action also appeared slower when compared with control neurons (see Fig. 4). Coexpression of RGS8 (1 ng/μl DNA) with GαoA(G184S:C351G) in a limited number of neurons (n = 2) resulted in qualitatively similar results to those obtained in neurons expressing only GαoA(S207D:C351G) (data not shown). In particular, the recovery from inhibition after agonist removal remained very slow, indicating an inability of RGS8 to reverse this effect. In addition to GαoA(G184S:C351G) and GαoA(S207D:C351G), we generated a construct that contained both the G184S and S207D mutations. When expressed along with Gβ1γ2, GαoA(G184S:S207D:C351G) reconstituted VD Ca2+ current inhibition (62 ± 2%; PFR, 2.37 ± 0.07, n = 7) (Figs.1F, 2). The time course of current recovery after agonist removal was dramatically retarded in neurons expressing this mutant. It is not clear whether this effect reflects an intrinsically greater “resistance” to the effects of RGS proteins or a secondary effect of the combined mutations. Together, these data demonstrate that expression of the PTX/RGS-i GαoA subunits (together with Gβ1γ2) reconstituted functional coupling of α2-ARs to N-type Ca2+ channels in PTX-treated neurons. In addition, the data are consistent with biochemical data showing that the G184S and S207D mutations confer resistance to the GAP effects of exogenous RGS proteins. Thus, additional studies were undertaken to quantify the alterations in Ca2+ channel inhibition and time course.
Expression of PTX/RGS-i GαoA subunits produces a leftward shift in concentration–response curves
As noted above, Ca2+ current inhibitions produced by a single concentration of NE (10 μm) were modestly enhanced in neurons expressing PTX/RGS-i GαoA subunits when compared with neurons expressing PTX-i GαoA. These results are consistent with previous studies demonstrating that increasing levels of RGS proteins attenuate agonist-induced Ca2+ current inhibition in SCG neurons (Jeong and Ikeda, 1998) and HEK293 cells (Melliti et al., 1999). In the latter study, rightward shifts in steady-state concentration–response curves were observed after heterologous expression of RGS proteins. Therefore, it was predicted that expression of PTX/RGS-i GαoA subunits would shift concentration–response curves in the opposite direction, i.e., leftward, if the Gα subunit mutations interfered with RGS protein interactions. To test this possibility, we measured Ca2+ current inhibition over a wide range of NE concentrations (0.1–30 μm). Figure3A illustrates representative time courses of the current inhibition in neurons expressing GαoA(C351G) and GαoA(G184S:C351G) after sequential application of increasing concentrations of NE. Figure 3B summarized the steady-state concentration–response curves for neurons expressing PTX-i GαoA (control) or PTX/RGS-i GαoA. When compared with uninjected neurons (non-PTX-treated), NE-induced Ca2+ current inhibition in neurons expressing PTX-i GαoA, i.e., GαoA(C351G), was approximately twofold less potent but equally efficacious (data not shown). In neurons expressing PTX/RGS-i GαoA subunits, the steady-state concentration–response curves were shifted leftward with a rank order of potency (EC50) as follows: GαoA(S207D:C351G) (0.28 μm, n = 5) > GαoA(G184S:C351G) (0.41 μm, n = 6) > GαoA(G184S:S207D:C351G) (0.50 μm, n = 4) > GαoA(C351G) (2.35 μm,n = 4). The maximal inhibitions produced by NE (30 μm) in neurons expressing GαoA(C351G), GαoA(G184S:C351G), GαoA(S207D:C351G), and GαoA(G184S:S207D:C351G) were 52 ± 6, 67 ± 1, 63 ± 3, and 55 ± 2%, respectively. In addition, the Hill factor for the concentration–response curves ranged between 1.5 and 1.8. Together, these data indicate that endogenous RGS proteins influence the agonist potency, and to a minor extent, efficacy, in sympathetic neurons.
Expression of PTX/RGS-i GαoA subunits alters the kinetics of NE-induced Ca2+ current inhibition
It has been well established that RGS proteins, acting as GAPs, accelerate Gα-catalyzed GTP hydrolysis in vitro (for review, see Dohlman and Thorner, 1997; Koelle, 1997; Berman and Gilman, 1998). These biochemical results implicate RGS proteins in controlling the kinetics of G-protein signaling. In recent electrophysiological studies, heterologously expressed RGS proteins have been shown to accelerate recovery of Ca2+ channels from the inhibition mediated by Gz (Jeong and Ikeda, 1998) and Go/i (Melliti et al., 1999), and deactivation of GIRK channels (Doupnik et al., 1997; Saitoh et al., 1997; Kovoor et al. 2000). As described in Figure 1, the recovery of Ca2+ channels from inhibition was slowed, consistent with a disruption of endogenous RGS actions on expressed PTX/RGS-i GαoA subunits. Therefore, additional experiments were undertaken to quantify these effects. Figure4A illustrates time courses of NE-induced Ca2+ current inhibition onset and recovery in neurons expressing PTX-i GαoA (control) or PTX/RGS-i GαoA subunits. For these experiments, the Ca2+ currents were evoked every 2 or 5 sec by single 20 msec test pulses to +10 mV from a holding potential −80 mV. In neurons expressing PTX-i GαoA, NE (10 μm) produced a rapid onset of inhibition and current recovery after agonist removal similar to that observed in uninjected neurons (Fig. 4C) (Zhou et al., 1997). Conversely, the current recovery was dramatically retarded in neurons expressing PTX/RGS-i GαoA (Figs.4A,B). To compare the time course of current recovery, the half recovery time (t0.5) was determined for the relaxation phase after agonist removal (Fig. 4B). A time constant was not calculated because the recovery time course under these circumstances was clearly nonexponential. The meant0.5 for GαoA(C351G) (control), GαoA(G184S:C351G), GαoA(S207D:C351G), and GαoA (G184S:S207D:C351G) -expressing neurons was 9 ± 1 (n = 19), 59 ± 5 (n = 16), 47 ± 5 (n = 14), and 130 ± 13 sec (n = 10), respectively (Fig.4C, top). Interestingly, decreases in the off rate seemed to be accompanied by decreases in the on rate. Indeed, the onset of the steady-state current inhibition was significantly retarded in neurons expressing PTX/RGS-i GαoA when compared with neurons expressing PTX-i GαoA(Fig. 4A). Because the rate of current inhibition onset was relatively rapid in relationship to the frequency of test pulses (therefore limiting temporal resolution), the onset of the NE-induced Ca2+ current inhibition was estimated as a 10–90% rise time (trise) instead oft0.5. As summarized in Figure4C, mean trise for GαoA(C351G), GαoA(G184S:C351G), GαoA(S207D:C351G), and GαoA(G184S:S207D:C351G) was 3.4 ± 0.1, 15.0 ± 3.0, 8.0 ± 1.7, and 11.8 ± 1.3 sec, respectively. It should be noted that the slowing of current inhibition onset was not proportional to the slowing of current recovery. Together, these data suggest that endogenous RGS proteins play an important role in controlling the lifetime of Ca2+ current inhibition after receptor activation in neurons.
DISCUSSION
Increasing evidence suggests that RGS proteins play an integral role in G-protein signaling. To date, however, information regarding the functional roles of RGS proteins is limited primarily to in vitro biochemical assays and heterologous overexpression experiments. Thus, the role played by natively expressed RGS proteins in G-protein signaling remains unclear. A reason for this void in our knowledge is the lack of experimental tools for uncoupling endogenous RGS proteins from G-protein signaling pathways. Thus, the development of the better tools represents a major challenge for understanding the physiological roles subserved by RGS proteins. In this regard, several potential strategies are apparent. First, genetically ablating specific RGS proteins, either acutely (e.g., antisense techniques) or stably (e.g., knock-out mice), might lend insight into endogenous RGS protein functions. However, >20 mammalian RGS proteins have been identified so far, and it is likely that even single cells contain multiple different RGS protein subtypes (Gold et al., 1997; Kardestuncer et al., 1998). Moreover, of the RGS proteins studied to date, only limited Gα specificity (i.e., Gi/o and Gq families) was observed in regard to GAP activity (for review, see Dohlman and Thorner, 1997; Koelle, 1997;Berman and Gilman, 1998). Thus, it seems likely that redundancy of RGS protein actions might confound studies in which single RGS proteins are eliminated. In one study, however, dialysis of neurons with an antibody directed at the RGS4 C terminus appeared to produce a specific result (Diverse-Pierlussi et al., 1999), indicating that this approach may be a fruitful one. A second strategy is to produce a dominant negative mutant RGS protein. Unfortunately, the means of creating such a mutant are not apparent, because mutations that alter GAP activity also disrupt Gα binding (Srinivasa et al., 1998). Finally, it might be possible to construct a mutant Gα subunit that sequesters RGS proteins. It has been shown that RGS4 interacts with a GTPase-deficient mutant of Gαi1, Gαi1(R178C) (Berman et al., 1996b). In preliminary experiments, however, expression of GαoA(R178C) resulted in apparent sequestration of Gβγ (Ikeda, 1996; Jeong and Ikeda, 1999b), presumably arising from a population of GDP-bound GαoA(R178C) (S.-W. Jeong and S. R. Ikeda, unpublished data).
The strategy used in this study originated from the seminal findings of several groups. First, a founding member of the RGS family, SST2, has been shown to desensitize Gβγ-mediated pheromone signaling via direct interaction with GPA1, a Gα subunit in Saccharomyces cerevisiae (Dohlman et al., 1996). Biochemical and functional observations demonstrated that a GPA1 mutant (gpa1sst) phenotypically mimicked the loss of SST2, suggesting that the mutant was insensitive to RGS GAP activity (DiBello et al., 1998). Subsequent sequencing of the Gα mutant revealed that Gly302, a conserved amino acid in the Gα family, was mutated to Ser. Homologous mutations in mammalian Gα subunits (Gαo/i and Gαq) resulted in an equivalent phenotype, thus demonstrating the generality of the substitution (DiBello et al., 1998; Lan et al., 1998). Second, mutation of Ser202 to Asp in transducin was shown to abolish interaction with a human retinal RGS isoform, hRGSr, in vitro (Natochin and Artemyev, 1998a,b; Posner et al., 1999). Mutation of Ser202 was arrived at by comparing the sequences of RGS-sensitive (i.e., Gαi/o and Gαq/11) and -insensitive Gα subunits (i.e., Gαs and Gα12) (Berman et al., 1996a; Huang et al., 1997). A problem with using RGS-insensitive Gα mutants for investigating the functional roles of endogenous RGS proteins in neurons is suppressing the activity of endogenous Gα subunits. To this end, we adopted the method used by several laboratories (Taussig et al., 1992; Hunt et al., 1994;Senogles 1994; Kozasa et al., 1996) and subsequently studied in detail by Milligan and colleagues (Wise et al., 1997; Bahia et al., 1998) in which Gα subunits are rendered PTX-insensitive by mutating a Cys residue in the C terminus. We have shown previously that heterologously expressed G-protein heterotrimers (Gαβγs) containing PTX-i Gα subunits functionally replace (as determined from VD Ca2+ channel inhibition) natively expressed G-proteins (Go/i) after PTX treatment of SCG neurons (Jeong and Ikeda, 2000). Thus, the strategy we used combined these two sets of mutations.
Success in using this method relied on reconstituting VD Ca2+ channel inhibition in PTX-treated neurons expressing PTX/RGS-i GαoA subunits. Because the crystal structure of Gαi1-GDP-AlF4− indicates that the mutated Gly and Ser residues are located near Gβγ contact sites within the switch I and II regions, respectively (Lambright et al., 1996), there was concern that heterotrimer formation might be disrupted by the mutations. Preliminary experiments, however, indicated that overexpression of PTX/RGS-i GαoA (without coexpression of Gβγ) attenuated NE-induced Ca2+ current inhibition in a manner consistent with the sequestration of Gβγ subunits (data not shown) (Ikeda, 1996; Jeong and Ikeda, 1999b), thus providing evidence forin situ heterotrimer formation. Consistent with this observation, successful reconstitution of VD Ca2+ channel inhibition was observed for all PTX/RGS-i GαoA subunits (Fig. 1) whenever a functional stoichiometric match (Jeong and Ikeda, 2000) between PTX/RGS-i GαoA and Gβ1γ2 subunits was achieved.
Major features of the NE-induced Ca2+current inhibition reconstituted in neurons expressing PTX/RGS-i GαoA subunits were (1) a leftward shift in the NE concentration–response relationships, (2) a dramatically retarded recovery time course after agonist removal, and (3) an increase in the time to reach steady-state current inhibition after agonist application. Some of these results can be rationalized by the failure of endogenous RGS proteins to interact with PTX/RGS-i GαoA subunits (in the transition conformation) and, consequently, the resulting loss of RGS-mediated GTPase acceleration. In this situation, G-protein heterotrimer reformation is limited by the relatively slow intrinsic GTPase activities of GαoA subunits (Higashijima et al., 1987; Lan et al., 1998). Consequently, the concentration of free Gβγs around the Ca2+ channel would remain relatively high for a prolonged time after agonist removal, thus retarding recovery of channels from inhibition. Moreover, because steady-state concentrations of Gα-GTP (and hence free Gβγ) should be determined by the relative rates of Gα-GTP formation and destruction, the increase in agonist potency and modest increase in efficacy can be accommodated by this interpretation as well. These results are consistent with several previous studies in which RGS actions were increased by heterologously overexpressing RGS proteins. For example, expression of RGS4 in COS-7 cells produced a rightward shift in the concentration–response relationship for activation of mitogen-activated protein kinase (Yan et al., 1997). In HEK293 cells,Melliti et al. (1999) showed that expression of RGS3T, RGS3, and RGS8 altered the magnitude and kinetics of N-type Ca2+ channel inhibition, as well as produced a rightward shift of the steady-state concentration–response curve. In SCG neurons, Jeong and Ikeda (1998) have shown that overexpression of RGS4 and RGS10 accelerate the recovery of NE-induced inhibition of Ca2+ channels reconstituted with Gαz. In that study, overexpression RGS10 also attenuated NE-induced Ca2+ current inhibition in control SCG neurons, suggesting a rightward shift in the concentration–response curve. Conversely, RGS proteins appear to accelerate deactivation of GIRK currents without an apparent shift in the concentration–response relationship (Doupnik et al., 1997; Saitoh et al., 1997). Although the reason for this difference remains unclear, it possibly suggests that, in addition to GAP functions, RGS proteins subserve other roles, such as increasing Gβγ availability (Bünemann and Hosey, 1998) or effecting receptors/G-proteins coupling (Zeng et al., 1998; Xu et al., 1999).
The slowing of the on rate of steady-state Ca2+ current inhibition seen in this study is more difficult to explain within in the framework of established RGS functions. Multiple steps, such as agonist binding to receptor, receptor/G-protein interaction, GDP/GTP exchange rate, and Gβγ interaction with the channel, could contribute to this rate. A possibility that should be considered is that the Gα mutations used in our studies have additional actions, such as altering the receptor-mediated GDP/GTP exchange rate. However, in vitroexperiments indicate that basal GDP/GTP exchange rates are not significantly altered by these Gα mutations (Lan et al., 1998). Moreover, the current results are consistent with previous studies in which overexpression of RGS proteins exerted opposite effects, i.e., acceleration of the onset of steady-state Ca2+ current inhibition (Jeong and Ikeda, 1998; Melliti et al. 1999) and GIRK activation (Doupnik et al., 1997;Saitoh et al., 1997). Two possibilities might explain this phenomenon. First, RGS proteins may directly or indirectly stimulate GDP/GTP exchange via unknown mechanisms, although at present there is little evidence to support this view (Dohlman and Thorner, 1997; Koelle, 1997;Berman and Gilman, 1998). Second, RGS proteins may affect the receptor/G-protein coupling independently of the GAP activity associated with the RGS core domain (Zeng et al., 1998; Xu et al., 1999), thereby altering the rate of Ca2+channel inhibition. In this regard, it should be mentioned that the term RGS “insensitivity,” in the current context, refers solely to the GAP activity of RGS proteins and does not preclude effects that may be conferred by domains lying outside of the RGS core domain. In fact, preliminary data indicate that the kinetics of GIRK channel activation/deactivation after expression of PTX/RGS-i GαoA subunits are altered by noncore RGS domains (Jeong and Ikeda, unpublished data).
Previously, Zhou et al. (1997) have shown that N-type Ca2+ channels in SCG neurons recovered from NE-induced inhibition after agonist removal with a rate constant (between 0.09 and 0.18 sec−1) much higher than that reported for the intrinsic GTP hydrolysis rate (∼0.03 sec−1) of Gαodetermined in vitro (Higashijima et al., 1987). Our results indicate that the actions of endogenous RGS proteins might account for this discrepancy. Thus, in addition to receptors, G-proteins and N-type Ca2+ channels, RGS proteins must be added to the membrane-delimited pathway describing VD channel modulation. How RGS proteins contribute to the more integrative functions of N-type Ca2+ channels, such as presynaptic inhibition resulting from neurotransmitter release, remains to be determined. It is possible that the strategy used here, i.e., the combined use of PTX- and RGS-insensitive Gα subunits, will be useful in defining the functional roles of the endogenous RGS proteins in different systems.
Footnotes
This study was supported by National Institutes of Health Grant GM 56180 (S.R.I.). We thank M. King for excellent technical assistance and Dr. M. I. Simon (California Institute of Technology, Pasadena, CA) for G-protein plasmids.
Correspondence should be addressed to Dr. Stephen R. Ikeda, Laboratory of Molecular Physiology, Guthrie Research Institute, One Guthrie Square, Sayre, PA 18840. E-mail: sikeda{at}inet.guthrie.org.
Dr. Jeong's present address: Department of Physiology, Yonsei University, Wonju College of Medicine, Wonju, Korea.