Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Inhibition of a Gi-activated Potassium Channel (GIRK1/4) by the Gq-coupled m1 Muscarinic Acetylcholine Receptor

2001; Elsevier BV; Volume: 276; Issue: 8 Linguagem: Inglês

10.1074/jbc.m008213200

ISSN

1083-351X

Autores

Jennifer J. Hill, Ernest G. Peralta,

Tópico(s)

Neuroscience and Neuropharmacology Research

Resumo

The G protein-coupled inwardly rectifying K+ channel, GIRK1/GIRK4, can be activated by receptors coupled to the Gαi subunit. An opposing role for Gαq receptor signaling in GIRK regulation has only recently begun to be established. We have studied the effects of m1 muscarinic acetylcholine receptor (mAChR) stimulation, which is known to mobilize calcium and activate protein kinase C (PKC) by a Gαq-dependent mechanism, on whole cell GIRK1/4 currents in Xenopus oocytes. We found that stimulation of the m1 mAChR suppresses both basal and dopamine 2 receptor-activated GIRK 1/4 currents. Overexpression of Gβγ subunits attenuates this effect, suggesting that increased binding of Gβγ to the GIRK channel can effectively compete with the Gq-mediated inhibitory signal. This Gq signal requires the use of second messenger molecules; pharmacology implicates a role for PKC and Ca2+ responses as m1 mAChR-mediated inhibition of GIRK channels is mimicked by PMA and Ca2+ionophore A23187. We have analyzed a series of mutant and chimeric channels suggesting that the GIRK4 subunit is capable of responding to Gq signals and that the resulting current inhibition does not occur via phosphorylation of a canonical PKC site on the channel itself. The G protein-coupled inwardly rectifying K+ channel, GIRK1/GIRK4, can be activated by receptors coupled to the Gαi subunit. An opposing role for Gαq receptor signaling in GIRK regulation has only recently begun to be established. We have studied the effects of m1 muscarinic acetylcholine receptor (mAChR) stimulation, which is known to mobilize calcium and activate protein kinase C (PKC) by a Gαq-dependent mechanism, on whole cell GIRK1/4 currents in Xenopus oocytes. We found that stimulation of the m1 mAChR suppresses both basal and dopamine 2 receptor-activated GIRK 1/4 currents. Overexpression of Gβγ subunits attenuates this effect, suggesting that increased binding of Gβγ to the GIRK channel can effectively compete with the Gq-mediated inhibitory signal. This Gq signal requires the use of second messenger molecules; pharmacology implicates a role for PKC and Ca2+ responses as m1 mAChR-mediated inhibition of GIRK channels is mimicked by PMA and Ca2+ionophore A23187. We have analyzed a series of mutant and chimeric channels suggesting that the GIRK4 subunit is capable of responding to Gq signals and that the resulting current inhibition does not occur via phosphorylation of a canonical PKC site on the channel itself. 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid G protein-coupled receptor G protein-coupled inwardly rectifying K+ channel protein kinase C muscarinic acetylcholine receptor phorbol 12-myristate 13-acetate dopamine 2 receptor pertussis toxin cAMP-dependent kinase Guanine nucleotide-binding protein coupled receptors (GPCRs)1 play an important role in many physiological processes, including the regulation of excitability in cardiac and neuronal cells (1Iismaa T.P. Shine J. Curr. Opin. Cell. Biol. 1992; 4: 195-202Crossref PubMed Scopus (34) Google Scholar). The downstream signaling pathways activated by GPCR agonists are dependent on the class of heterotrimeric guanine nucleotide-binding (G) protein that is coupled to the cytoplasmic domains of the receptor. For example, GPCRs such as the dopamine 2 receptor (d2R) are coupled to the PTX-sensitive Gi class of G proteins and thus mediate the inhibition of adenylyl cyclase and a lowering of cellular cAMP levels. In contrast, Gq-coupled receptors, such as the m1 muscarinic acetylcholine receptor (mAChR), activate phospholipase Cβ, resulting in the activation of protein kinase C (PKC) and an increase in cytoplasmic Ca2+ levels. In many neurons, treatment with somatostatin enhances the conductance of an inwardly rectifying K+ current, resulting in the inhibition of neuron excitability (2Mihara S. North R.A. Surprenant A. J. Physiol. (Lond.). 1987; 390: 335-355Crossref Scopus (112) Google Scholar, 3Inoue M. Nakajima S. Nakajima Y. J. Physiol. (Lond.). 1988; 407: 177-198Crossref Scopus (129) Google Scholar). This effect is mediated by a PTX-sensitive G protein (3Inoue M. Nakajima S. Nakajima Y. J. Physiol. (Lond.). 1988; 407: 177-198Crossref Scopus (129) Google Scholar). In contrast, treatment with another neurotransmitter, substance P, excites neurons by suppressing an inwardly rectifying current through a PTX-insensitive G protein pathway (4Stanfield P.R. Nakajima Y. Yamaguchi K. Nature. 1985; 315: 498-501Crossref PubMed Scopus (187) Google Scholar, 5Nakajima Y. Nakajima S. Inoue M. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 3643-3647Crossref PubMed Scopus (91) Google Scholar). Velimirovic et al. (6Velimirovic B.M. Koyano K. Nakajima S. Nakajima Y. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1590-1594Crossref PubMed Scopus (76) Google Scholar) presented evidence that the same inwardly rectifying channel is the recipient of these competing G protein signals in locus coeruleus neurons, supporting the idea that channels can be dually modulated by different classes of GPCRs. Similar effects of GPCRs on ion channels have been documented in defined expression systems using cloned channels and receptors. In cardiac pacemaker cells, the m2 mAChR, which is coupled to the PTX sensitive Gαi protein, is associated with the activation of the G protein-coupled inwardly rectifying K+ channel (GIRK). GIRK channels are characterized by a low basal level of activity that is dramatically increased upon stimulation of Gi-coupled receptors. The activation of these channels results in a decrease in cellular excitability by hyperpolarizing the cell, thus lengthening the time between action potentials. The GIRK protein family contains five homologous members (GIRK1–5), all of which contain two transmembrane domains flanking a hydrophobic pore region, with both the N and C termini located in the cytoplasm (7Dascal N. Cell Signal. 1997; 9: 551-573Crossref PubMed Scopus (269) Google Scholar). GIRK channels function in the membrane as tetramers consisting of two GIRK1 subunits and two subunits of one of the other GIRK family members (GIRK2–5) (8Chan K.W. Langan M.N. Sui J.L. Kozak J.A. Pabon A. Ladias J.A. Logothetis D.E. J. Gen. Physiol. 1996; 107: 381-397Crossref PubMed Scopus (67) Google Scholar, 9Kofuji P. Davidson N. Lester H.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6542-6546Crossref PubMed Scopus (272) Google Scholar, 10Silverman S.K. Lester H.A. Dougherty D.A. J. Biol. Chem. 1996; 271: 30524-30528Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar). In cardiac cells, heteromultimeric GIRK1/GIRK4 channels predominate (11Krapivinsky G. Gordon E.A. Wickman K. Velimirovic B. Krapivinsky L. Clapham D.E. Nature. 1995; 374: 135-141Crossref PubMed Scopus (763) Google Scholar). The signaling pathway between the activating Gi-coupled receptor and GIRK channels is membrane-delimited and dependent on the direct binding of Gβγ subunits to the cytoplasmic domains of both GIRK1 and GIRK4 (12Soejima M. Noma A. Pflugers Arch. Eur. J. Physiol. 1984; 400: 424-431Crossref PubMed Scopus (293) Google Scholar, 13Logothetis D.E. Kurachi Y. Galper J. Neer E.J. Clapham D.E. Nature. 1987; 325: 321-326Crossref PubMed Scopus (879) Google Scholar, 14Wickman K.D. Iniguez-Lluhi J.A. Davenport P.A. Taussig R. Krapivinsky G.B. Linder M.E. Gilman A.G. Clapham D.E. Nature. 1994; 368: 255-257Crossref PubMed Scopus (383) Google Scholar, 15Kunkel M.T. Peralta E.G. Cell. 1995; 83: 443-449Abstract Full Text PDF PubMed Scopus (134) Google Scholar, 16Krapivinsky G. Krapivinsky L. Wickman K. Clapham D.E. J. Biol. Chem. 1995; 270: 29059-29062Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar). Recently, it has been suggested that Gβγ binding to GIRK channels may stimulate channel activity by increasing the affinity of the cytoplasmic regions of GIRK for phosphatidylinositol 4,5-bisphosphate (17Huang C.L. Feng S. Hilgemann D.W. Nature. 1998; 391: 803-806Crossref PubMed Scopus (771) Google Scholar), an association that appears to be required for the activity of all inwardly rectifying potassium channels (17Huang C.L. Feng S. Hilgemann D.W. Nature. 1998; 391: 803-806Crossref PubMed Scopus (771) Google Scholar). Other groups have recently reported that Gq signaling can inhibit GIRK channels (18Sharon D. Vorobiov D. Dascal N. J. Gen. Physiol. 1997; 109: 477-490Crossref PubMed Scopus (137) Google Scholar, 19Stevens E.B. Shah B.S. Pinnock R.D. Lee K. Mol. Pharmacol. 1999; 55: 1020-1027Crossref PubMed Scopus (45) Google Scholar, 20Rogalski S.L. Cyr C. Chavkin C. J. Neurochem. 1999; 72: 1409-1416Crossref PubMed Scopus (19) Google Scholar). In this study, we corroborate and further their findings by investigating the effect of m1 mAChR stimulation on GIRK1/GIRK4 whole cell currents in a Xenopusoocyte expression system. We found that stimulation of the Gq-coupled m1 mAChR led to the suppression of GIRK1/GIRK4 currents and that this effect could be mimicked by a potent PKC activator, PMA, and a Ca2+ ionophore. Analysis of chimeric and mutant channels revealed that the GIRK4 subunit is sufficient for responding to these Gq signals but not via direct phosphorylation of a canonical PKC site on the channel itself. The following clones were used in these studies: rat GIRK1, rat GIRK4, rat IRK2, human m1 mAChR, human d2R (long form), rat Gαi2, murine Gβ1, and murine Gγ2. Chimeric GIRK1/IRK2 channels were generated as described previously (15Kunkel M.T. Peralta E.G. Cell. 1995; 83: 443-449Abstract Full Text PDF PubMed Scopus (134) Google Scholar). Mutant channels were generated using the QuikChange site-directed mutagenesis kit (Stratagene) protocol. RNA synthesis and oocyte preparation were performed as described previously (21Kunkel M.T. Peralta E.G. EMBO J. 1993; 12: 3809-3815Crossref PubMed Scopus (56) Google Scholar). Defolliculated oocytes were injected with 50 nl of a cRNA mixture containing the appropriate combination of transcripts in these approximate amounts: 5 ng of GIRK1, 5 ng of GIRK4, 5 ng of d2R, 1 ng of m1 mAChR, and 2 ng of IRK2. In Fig. 1 B, basal GIRK currents were increased by the injection of 2 ng of Gβ1 and 1 ng of Gγ2. A larger amount of Gβγ RNA was injected for Fig. 2 (∼8 ng of Gβ1, 3 ng of Gγ2). Oocytes were recorded 3–5 days later by two-electrode voltage clamp in 90 mmK+ saline (90 mm KCl, 3 mmMgCl2, 5 mm HEPES, pH 7.4) as described previously (15Kunkel M.T. Peralta E.G. Cell. 1995; 83: 443-449Abstract Full Text PDF PubMed Scopus (134) Google Scholar). Currents from a minimum of five oocytes were recorded for each treatment.Figure 2Overexpressed Gβγ attenuates m1 mAChR-mediated inhibition of GIRK1/4. Xenopus oocytes were injected with GIRK1/4 and m1 mAChR cRNA and then injected a second time with dopamine receptor and either Gβ1γ2 RNA alone or Gαi2 and Gβ1γ2 cRNA (A). The percentage of suppression over time as the result of carbachol addition for each of these batches of oocytes was determined (B). Overexpression of Gβ1γ2 subunits dramatically inhibits suppression of basal GIRK1/4 currents upon stimulation of the m1 mAChR by carbachol (B, open circles). This effect is rescued by the additional expression of Gαi2(B, closed circles). C, the maximal Ca2+-activated Cl− current at 0 mV elicited by m1 mAChR stimulation in each batch of oocytes. D, basal current amplitude at −80 mV of GIRK1/4 currents in each batch of oocytes.View Large Image Figure ViewerDownload Hi-res image Download (PPT) To minimize the contribution of endogenous currents, only oocytes with >1μA of whole cell current at −80 mV were used. Endogenous oocyte current amplitudes at −80 mV are generally less than 0.2 μA. Oocytes were held at 0 mV and pulsed to −80 mV for 500 ms at 10-s intervals. Currents were also monitored at 0 and +40 mV. Data from multiple oocytes were normalized by computing the percent suppression: % suppression = −((X − A)/A) × 100, where Xis the current amplitude at a given time point and A is the current amplitude immediately before carbachol addition and after d2R activation (if applicable). Cell-attached patch recordings (see Fig. 1 D) were performed with 140 mm K+ saline (140 mm KCl, 10 mm HEPES, pH 7.4, 1 mm MgCl2) in both the bath and the pipette. Quinpirole (100 μm) was added to the pipette saline to activate GIRK channels. Glass pipettes (Corning 8161) had tip resistances between 0.8 and 4 MΩ. Cell-attached patches containing multiple channels were held at 0mV. Every 5 s, the voltage was pulsed to −60 mV for 2700 ms, followed by a 460-ms pulse to +60 mV. Carbachol (1 μm) was added to the bath after 100 s and allowed to mix by diffusion. Data were analyzed using Fetchan (Axon Instruments) by calculating the percent total channel activity in a 100-s window beginning 100 s after carbachol (or control saline) addition relative to an identical window before carbachol (or saline) addition. To activate the d2R receptor, 10 μm (−)-quinpirole hydrochloride (Research Biochemicals International) was added to the recording solution. Similarly, the m1 mAChR was activated by addition of 100 nm to 10 μm carbachol. The concentration of carbachol used in each oocyte batch was chosen by normalizing the endogenous Ca2+-activated chloride transient current evoked by m1 mAChR stimulation in oocytes (22Dascal N. Landau E.M. Lass Y. J. Physiol. (Lond.). 1984; 352: 551-574Crossref Scopus (112) Google Scholar). A concentration of carbachol that evoked a visible but short lived (<30 s) transient current was used. Oocytes treated with calphostin C, the tyrosine kinase inhibitor mixture (4 μm tyrphostin A-48, 4 μmtyrphostin A-25, 1 μm lavendustin A, and 15 μm genistein), GF109203X, staurosporine, and H-89 were soaked for 20–30 min prior to recording in 90 mmK+ recording saline containing 1:1000 dilutions of both 1000× drug (in Me2SO) and 30% pluronic F-127 (in Me2SO). Pluronic F-127 alone had no effect on whole cell GIRK1/GIRK4 currents (data not shown). PMA and Ca2+ionophore were similarly dissolved and added to the recording solution in the final concentrations given in the text. 4K+-1,2-bis(0-aminophenoxy)ethane- N,N,N′,N′-tetraacetic acid (BAPTA) was dissolved in 100 mm KCl to a final concentration of 2 mm. A 50-nl bolus of this solution was injected into oocytes 20–30 min before recording. Control oocytes were similarly injected with 100 mm KCl. Similarly, KT5823 and the inhibitory PKA peptide were dissolved in Ca2+-free OR-2 (96 mm NaCl, 2 mm KCl, 1 mmMgCl2, and 5 mm HEPES, pH 7.5) to final concentrations of 20 μm and 200 nm, respectively. A 50-nl bolus was injected 20–30 min before recording and compared with oocytes injected with Ca2+-free OR-2 alone. To test the effect of m1 mAChR stimulation on GIRK currents, Xenopus oocytes expressing GIRK1, GIRK4, d2R, and m1 mAChR were placed under two-electrode voltage clamp. GIRK currents were first activated by the addition of 10 μm quinpirole. Fifty seconds later, the m1 mAChR was stimulated by 100 nmto 5 μm carbachol. Carbachol elicited a short lived endogenous oocyte current that strongly inactivated at negative potentials (Ref. 22Dascal N. Landau E.M. Lass Y. J. Physiol. (Lond.). 1984; 352: 551-574Crossref Scopus (112) Google Scholar and data not shown). After this transient current had subsided, we noted a carbachol-dependent inhibition in GIRK current amplitude (Fig.1 A, left panel). The right panel of Fig. 1 A shows the average percentage of suppression of activated GIRK currents over time (n = 12). Interestingly, carbachol-treated GIRK currents decreased at a rate that was three times faster than oocytes that were not treated with carbachol. This effect was dependent on signaling through the m1 mAChR because oocytes that were not injected with cRNA encoding the m1 mAChR failed to exhibit a carbachol-induced decrease in activated GIRK whole cell currents. (n = 3, data not shown). These results suggest that m1 mAChR, like the metabotropic glutamate receptor 1a, bombesin 1, and endothelin 1 receptors (18Sharon D. Vorobiov D. Dascal N. J. Gen. Physiol. 1997; 109: 477-490Crossref PubMed Scopus (137) Google Scholar, 19Stevens E.B. Shah B.S. Pinnock R.D. Lee K. Mol. Pharmacol. 1999; 55: 1020-1027Crossref PubMed Scopus (45) Google Scholar, 20Rogalski S.L. Cyr C. Chavkin C. J. Neurochem. 1999; 72: 1409-1416Crossref PubMed Scopus (19) Google Scholar), can inhibit GIRK activity. To eliminate the possibility that the m1 mAChR acts by directly inhibiting the d2R receptor itself, we next examined the effect of m1 mAChR stimulation on basal, nonreceptor-activated, GIRK1/GIRK4 currents. Oocytes injected with cRNA encoding GIRK1/4, m1 mAChR, Gβ1, and Gγ2 were screened for a high level of basal GIRK current. As previously reported by Kovoor et al. (23Kovoor A. Henry D.J. Chavkin C. J. Biol. Chem. 1995; 270: 589-595Abstract Full Text Full Text PDF PubMed Scopus (108) Google Scholar), we found that the basal current amplitude of GIRK1/GIRK4 decreased slowly in the absence of any agonists after placing the oocytes into high potassium saline. However, when carbachol (0.5–1.0 μm) was added to the bath, the basal GIRK current amplitude was significantly inhibited (Fig. 1 B) with similar kinetics to those seen in the inhibition of d2R-activated GIRK currents. This finding suggests that m1 mAChR-mediated inhibition of GIRK1/4 occurs downstream of the activating receptor. The inwardly rectifying potassium channel, IRK2 (Kir2.2), is 45% homologous to GIRK1 but does not respond to Gi-coupled receptor signaling. We found that the large basal current characteristic of IRK2 was also insensitive to Gq-coupled receptor signaling; carbachol stimulation of the m1 mAChR had no effect on IRK2 current amplitude (Fig. 1 C). Similarly, the small steady-state endogenous oocyte currents seen at −80 mV were also insensitive to m1 mAChR signaling (n = 5; data not shown). To determine whether second messenger molecules are necessary for m1 mAChR-mediated inhibition of GIRK1/4, we recorded GIRK currents in a cell-attached patch configuration and tested for their sensitivity to carbachol added to the bath. The left panel of Fig.1 D shows the openings of single GIRK channels in aXenopus oocyte patch. We found that carbachol added to the bath inhibited channels located within the pipette, implying the involvement of second messenger molecules (Fig. 1 D). This is in contrast to GIRK channel activation, which occurs by a membrane-delimited pathway and thus requires addition of agonist to the pipette; agonist addition to the bath does not activate GIRK currents (12Soejima M. Noma A. Pflugers Arch. Eur. J. Physiol. 1984; 400: 424-431Crossref PubMed Scopus (293) Google Scholar). Many groups have reported that overexpression of Gβγ subunits results in an increase in basal GIRK1/4 current, presumably because of increased binding of Gβγ to the cytoplasmic regions of the channel and subsequent channel activation (8Chan K.W. Langan M.N. Sui J.L. Kozak J.A. Pabon A. Ladias J.A. Logothetis D.E. J. Gen. Physiol. 1996; 107: 381-397Crossref PubMed Scopus (67) Google Scholar, 24Reuveny E. Slesinger P.A. Inglese J. Morales J.M. Iniguez-Lluhi J.A. Lefkowitz R.J. Bourne H.R. Jan Y.N. Jan L.Y. Nature. 1994; 370: 143-146Crossref PubMed Scopus (423) Google Scholar, 25Velimirovic B.M. Gordon E.A. Lim N.F. Navarro B. Clapham D.E. FEBS Lett. 1996; 379: 31-37Crossref PubMed Scopus (87) Google Scholar). However, heterotrimeric G proteins (Gαβγ) do not activate GIRK. The previous experiments described here utilized only G proteins that are endogenous to the oocyte (although small amounts of Gβ1 and Gγ2 cRNA were injected in Fig. 1 B). Interestingly, we found that GIRK currents in oocytes that were injected with Gβ1 and Gγ2cRNA were less responsive to Gq signaling than oocytes that also expressed the Gαi2 subunits, suggesting that free Gβγ subunits are capable of reversing or preventing Gqinhibition (Fig. 2, A andB). Notably, the slower agonist-independent GIRK suppression was similarly affected by the overexpression of G protein subunits (Fig. 2 B). Basal GIRK1/4 current was significantly decreased by the overexpression of Gαi2, suggesting that Gαi2 protein was expressed and functional (Fig.2 D). Furthermore, the amplitude of the transient Ca2+ activated Cl− current was comparable in the oocytes expressing Gβγ and Gαi2βγ, indicating similar levels of m1 mAChR stimulation under these two conditions (Fig.2 C). These experiments show that high levels of free Gβγ subunits have an inhibitory effect on Gq receptor signaling to GIRK1/4. Gq-coupled receptors, such as the m1 mAChR, have been linked to many downstream effector pathways including activation of phospholipase C, PKC, and tyrosine kinases (26Peralta E.G. Ashkenazi A. Winslow J.W. Ramachandran J. Capon D.J. Nature. 1988; 334: 434-437Crossref PubMed Scopus (548) Google Scholar, 27Ashkenazi A. Peralta E.G. Winslow J.W. Ramachandran J. Capon D.J. Trends Pharmacol. Sci. 1989; 10 (suppl.): 16-22Google Scholar, 28Huang X.Y. Morielli A.D. Peralta E.G. Cell. 1993; 75: 1145-1156Abstract Full Text PDF PubMed Scopus (243) Google Scholar). In an effort to determine whether these pathways are involved in the second messenger-dependent m1 mAChR-mediated suppression of GIRK1/4 currents, oocytes injected with GIRK1, GIRK4, d2R, and m1 mAChR mRNA were treated with various pharmacological agents. PKC is a ubiquitous kinase known to be activated by Gq-coupled receptors (29Newton A.C. Curr. Opin. Cell. Biol. 1997; 9: 161-167Crossref PubMed Scopus (851) Google Scholar). Interestingly, 10 μm phorbol 12-myristate 13-acetate (PMA), a potent PKC activator, mimicked the effect of m1 mAChR signaling on both activated (Fig. 3 A) and basal (Fig.3 B) GIRK1/4 currents. The kinetics of GIRK inhibition by PMA were very similar to the receptor-mediated suppression (compare with Fig. 1). PMA treatment also inhibited IRK2 currents; however, this inhibition took considerably longer to develop and occurred at a slower rate than GIRK1/4 inhibition (Fig. 3 C), suggesting that PMA acts on these two channels by different mechanisms. Treatment with a broad specificity serine/threonine kinase inhibitor, staurosporine (3 μm), slowed both m1 mAChR-mediated and agonist-independent channel suppression (Fig. 3 D). H-89 (50 μm), another broad specificity kinase inhibitor, also slowed m1 mAChR-mediated inhibition of activated currents (data not shown). However, we were unable to further illustrate the involvement of PKC in GIRK1/4 inhibition through the use of specific PKC inhibitors. The cell-permeable PKC inhibitor, calphostin C (up to 1 μm), did not affect m1 mAChR-mediated suppression of GIRK1/4 (Fig. 3 E). Another PKC inhibitor, GF109203X (1 μm), also failed to affect receptor-mediated inhibition of GIRK currents. At higher concentrations (15 μm), GF109203X blocked and disrupted GIRK1/4 currents (data not shown). Because both H-89 and staurosporine inhibit a broad range of kinases including PKC, PKA, and cGMP-dependent kinase at the concentrations used here, it is difficult to further define the role of a specific kinase in GIRK suppression. Additional experiments utilizing PKA and cGMP-dependent kinase inhibitors were unable to identify a specific role for these kinases. Both a protein kinase A inhibitory peptide (10 nm, Calbiochem) and KT5823 (1 μm, Calbiochem) failed to significantly affect m1 mAChR-mediated inhibition of GIRK channels (data not shown). In addition to the activation of PKC and other kinases, Gq-coupled receptors also strongly increase intracellular Ca2+ levels in the cell. To explore the possibility that Ca2+ signaling may mediate the Gq effect on GIRK channels, we modulated intracellular Ca2+ levels in oocytes with various pharmacological agents. Treatment with the Ca2+ ionophore, A23187, inhibited GIRK1/4 currents in a manner indistinguishable from m1 mAChR signaling or treatment with PMA (Fig. 3 F). However, inhibition of Ca2+ signaling by injection of the calcium chelating agent, BAPTA, did not significantly affect m1 mAChR signaling to GIRK1/4, despite the complete elimination of the transient Ca2+-activated Cl− current (Fig. 3 G). Tyrosine kinases have been implicated in the suppression of a voltage-gated potassium channel by m1 mAChR signaling (28Huang X.Y. Morielli A.D. Peralta E.G. Cell. 1993; 75: 1145-1156Abstract Full Text PDF PubMed Scopus (243) Google Scholar). However, we have found no evidence for tyrosine kinase phosphorylation in the regulation of GIRK channels by m1 mAChR. Treatment of oocytes with a mixture of tyrosine kinase inhibitors did not affect the carbachol-induced inhibition of GIRK1/4 currents (data not shown). Because pharmacology implicated a role for protein kinases in m1 mAChR-mediated suppression of GIRK1/4 currents, we searched for potential phosphorylation sites on the GIRK proteins that may mediate this effect. A ProSite data base search revealed 10 putative phosphorylation sites for both PKA and PKC in GIRK1 and three PKA sites and six PKC sites in GIRK4 (30Hofmann K. Bucher P. Falquet L. Bairoch A. Nucleic Acids Res. 1999; 27: 215-219Crossref PubMed Scopus (1006) Google Scholar). Interestingly, the extended C terminus of GIRK1 contains a particularly high density of phosphorylation sites. To narrow our search, we were interested in identifying the domains of the GIRK1 and GIRK4 proteins that mediate the inhibitory effect of Gq signaling on GIRK currents. A series of chimeric and mutant ion channels derived from GIRK1 and IRK2 were generated previously by Kunkel and Peralta (15Kunkel M.T. Peralta E.G. Cell. 1995; 83: 443-449Abstract Full Text PDF PubMed Scopus (134) Google Scholar) (Fig.4 A). We coexpressed these chimeras with the m1 mAChR and tested whether they were responsive to m1 mAChR signaling. The results from these experiments are summarized in Fig. 4 A. The d2R-activated current of both the GR2/GIRK4 and the ΔN1/GIRK4 mutant channels were suppressed upon stimulation of the m1 mAChR with carbachol. Taken together, these results suggest that the serine/threonine-rich distal N-terminal (amino acids 1–24) and C-terminal (amino acids 357–501) regions of GIRK1 are not required for GIRK inhibition by m1 mAChR. Thus, it appears that the numerous potential phosphorylation sites in these regions are not necessary for channel responsiveness to Gq signaling. In contrast, the large basal (nonreceptor activated) currents of the GR3 and GR7.1 chimeras were insensitive to Gq signaling, suggesting that amino acids 1–84 and 290–501 of GIRK1 are not capable of conferring Gq sensitivity onto IRK2. Interestingly, the chimeras that require coinjection of cRNA encoding the GIRK4 subunit to produce visible whole cell currents were suppressed upon m1 mAChR stimulation. This correlation suggested that the GIRK4 subunit itself may be responsive to Gq signals. To test this hypothesis, we took advantage of a mutation in the pore region of GIRK4, a serine to phenylalanine substitution at amino acid 143, that allows wild-type and S143F mutant GIRK4 subunits to form functional homomeric channels (31Chan K.W. Sui J.L. Vivaudou M. Logothetis D.E. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 14193-14198Crossref PubMed Scopus (101) Google Scholar, 32Vivaudou M. Chan K.W. Sui J.L. Jan L.Y. Reuveny E. Logothetis D.E. J. Biol. Chem. 1997; 272: 31553-31560Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). We found that GIRK4/GIRK4(S143F) homomeric channels are indeed inhibited by stimulation of the m1 mAChR (Fig. 4 B) in a similar manner to GIRK1/4 heteromeric channels. Thus, the GIRK4 subunit is sufficient to respond to Gq signals. Our pharmacological analysis implicates serine/threonine kinases in the regulation of GIRK channels by m1 mAChR signaling. Because GIRK channels contain numerous putative PKA and PKC phosphorylation sites, we wondered whether phosphorylation of GIRK1 or GIRK4 is required for GIRK current inhibition by m1 mAChR signaling. To address this question, we generated site-specific mutations to eliminate potential phosphorylation sites and tested these mutants for their ability to be suppressed by m1 mAChR. In particular, we took advantage of the GIRK4(S143F) pore mutation that allows GIRK4 to form homomeric channels and used site-directed mutagenesis to eliminate the canonical PKC phosphorylation sites found in this protein. GIRK4 contains six putative PKC sites ((S/T)X(R/K)), five of which are predicted to be accessible to intracellular kinases. The sixth PKC site is located immediately adjacent to the pore helix (H5 region), N-terminal to the second transmembrane domain (Fig.5 A). GIRK4 homomeric channels in which all five intracellular PKC sites have been mutated to alanine, designated GIRK4(φPKC), did not produce current in oocytes coinjected with the dopamine 2 receptor. However, the subsequent injection of Gβ1 and Gγ2 cRNAs produced whole cell GIRK currents of greater than 1 μA from these mutant channels. Interestingly, GIRK4(φPKC) channels are inhibited by Gqstimulation (Fig. 5 B) to a similar extent as wild-type GIRK4 homomeric channels. This result suggests that direct PKC phosphorylation of the channel is not necessary for Gq-mediated inhibition. In this paper, we have shown that a cloned inward rectifier, GIRK1/GIRK4, is functionally coupled to the m1 mAChR in aXenopus oocyte expression system. Thus, the regulation of GIRK1/4 conductance is mediated in opposing manners by different classes of GPCRs; it responds to an activation signal from Gi-coupled receptors such as d2R and an inhibitory signal from the Gq-coupled m1 mACh receptor. Other studies have demonstrated that stimulation of the m1 mAChR increases excitability in some cells by suppressing a voltage-gated K+ channel, Kv1.2 (28Huang X.Y. Morielli A.D. Peralta E.G. Cell. 1993; 75: 1145-1156Abstract Full Text PDF PubMed Scopus (243) Google Scholar). Our findings suggest that GIRK1/4 may also contribute to these m1 mAChR-mediated stimulatory signals by lowering the membrane permeability to K+, thus allowing the cell to reach firing threshold more quickly. It is interesting to speculate that the inhibition of GIRK currents by Gq-coupled receptors may be partially responsible for determining the specificity of GIRK activation. Although all G proteins release Gβγ subunits upon activation, only Gi-coupled receptors are capable of activating GIRK. Interestingly, GIRK channels do not seem to prefer particular combinations of β and γ subunits, suggesting that only the identity of the Gα subunit is important for determining specificity (14Wickman K.D. Iniguez-Lluhi J.A. Davenport P.A. Taussig R. Krapivinsky G.B. Linder M.E. Gilman A.G. Clapham D.E. Nature. 1994; 368: 255-257Crossref PubMed Scopus (383) Google Scholar). Gi-coupled receptors often have inhibitory physiological effects and are not generally thought to activate many cellular kinases. In contrast, other G proteins such as Gq and Gs stimulate many kinases, including PKC and PKA. Perhaps activation of these kinases inhibits GIRK channels by lowering their affinity for Gβγ, thus protecting them from subsequent activation. Other groups have also shown that GIRK channels can be suppressed by Gq-coupled receptors, including the metabotropic glutamate receptor 1a, the endothelin A receptor, and the bombesin 1 receptor (18Sharon D. Vorobiov D. Dascal N. J. Gen. Physiol. 1997; 109: 477-490Crossref PubMed Scopus (137) Google Scholar, 19Stevens E.B. Shah B.S. Pinnock R.D. Lee K. Mol. Pharmacol. 1999; 55: 1020-1027Crossref PubMed Scopus (45) Google Scholar, 20Rogalski S.L. Cyr C. Chavkin C. J. Neurochem. 1999; 72: 1409-1416Crossref PubMed Scopus (19) Google Scholar). Thus, it is likely that all Gq-coupled receptors are capable of inhibiting GIRK currents, just as GIRK currents are activated by any Gi-coupled receptor. However, the mechanism of Gq-mediated suppression of GIRK remains unclear. Activating protein kinase C (via PMA) or increasing cytosolic Ca2+ levels with A23187 mimicked the suppression of GIRK currents elicited by stimulation of the m1 mAChR with carbachol. Surprisingly, in our experiments, specific PKC targeting drugs that inactivate typical PKC subtypes failed to prevent or slow m1 mAChR-mediated suppression of GIRK1/4. However, broad specificity kinase inhibitors slowed both m1 mAChR-mediated and agonist-independent inhibition of GIRK1/4 currents, suggesting that serine/threonine kinases are likely to be involved in GIRK inhibition by Gq-coupled receptors. There are many different classes of PKC enzymes, all of which prefer similar phosphorylation site sequences (33Nishikawa K. Toker A. Johannes F.J. Songyang Z. Cantley L.C. J. Biol. Chem. 1997; 272: 952-960Abstract Full Text Full Text PDF PubMed Scopus (496) Google Scholar). It is possible that an atypical PKC family member is responsible for mediating the m1 mAChR effect on GIRK channels. In fact, Sharonet al. (18Sharon D. Vorobiov D. Dascal N. J. Gen. Physiol. 1997; 109: 477-490Crossref PubMed Scopus (137) Google Scholar) suggested PKC-μ, an atypical PKC, as a likely candidate. Stevens et al. (19Stevens E.B. Shah B.S. Pinnock R.D. Lee K. Mol. Pharmacol. 1999; 55: 1020-1027Crossref PubMed Scopus (45) Google Scholar) also found pharmacological evidence for the involvement of PKC. These pharmacological data led many of these groups to speculate that PKC may directly phosphorylate the channel itself resulting in a subsequent current inhibition. This mechanism is known to inhibit other channels, including an IRK family member, Kir2.3 (34Zhu G. Qu Z. Cui N. Jiang C. J. Biol. Chem. 1999; 274: 11643-11646Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). However, our results suggest that this mechanism is unlikely to explain Gq-mediated inhibition of GIRK currents because a mutant homomeric GIRK4 channel that does not contain accessible canonical PKC sites continues to respond to Gq signaling. Because stimulation of m1 mAChR can inhibit both basal and receptor-activated GIRK currents, m1 mAChR appears to regulate GIRK channel activity at a level downstream of the activating dopamine 2 receptor. GIRK activation occurs by a membrane-delimited pathway involving direct Gβγ binding to the channel (12Soejima M. Noma A. Pflugers Arch. Eur. J. Physiol. 1984; 400: 424-431Crossref PubMed Scopus (293) Google Scholar, 13Logothetis D.E. Kurachi Y. Galper J. Neer E.J. Clapham D.E. Nature. 1987; 325: 321-326Crossref PubMed Scopus (879) Google Scholar, 15Kunkel M.T. Peralta E.G. Cell. 1995; 83: 443-449Abstract Full Text PDF PubMed Scopus (134) Google Scholar, 16Krapivinsky G. Krapivinsky L. Wickman K. Clapham D.E. J. Biol. Chem. 1995; 270: 29059-29062Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar, 35Huang C.L. Slesinger P.A. Casey P.J. Jan Y.N. Jan L.Y. Neuron. 1995; 15: 1133-1143Abstract Full Text PDF PubMed Scopus (287) Google Scholar). Thus, it seems likely that m1 mAChR signaling targets either the activating heterotrimeric G protein or the GIRK channel itself. The exact nature of basal GIRK activity remains unclear, and it is a source of argument in the field whether GIRK channels display G protein-independent activity (see Ref. 7Dascal N. Cell Signal. 1997; 9: 551-573Crossref PubMed Scopus (269) Google Scholar). In oocytes, the basal current does not depend on the amount of expressed receptor, and ∼80% of the basal GIRK current is sensitive to PTX toxin, which targets the Gi heterotrimeric G protein (8Chan K.W. Langan M.N. Sui J.L. Kozak J.A. Pabon A. Ladias J.A. Logothetis D.E. J. Gen. Physiol. 1996; 107: 381-397Crossref PubMed Scopus (67) Google Scholar, 36Kubo Y. Reuveny E. Slesinger P.A. Jan Y.N. Jan L.Y. Nature. 1993; 364: 802-806Crossref PubMed Scopus (547) Google Scholar, 37Dascal N. Lim N.F. Schreibmayer W. Wang W. Davidson N. Lester H.A. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6596-6600Crossref PubMed Scopus (83) Google Scholar). Additionally, coexpression of βARK-PH domain or other Gβγ binding molecules can reduce basal GIRK activity (14Wickman K.D. Iniguez-Lluhi J.A. Davenport P.A. Taussig R. Krapivinsky G.B. Linder M.E. Gilman A.G. Clapham D.E. Nature. 1994; 368: 255-257Crossref PubMed Scopus (383) Google Scholar, 24Reuveny E. Slesinger P.A. Inglese J. Morales J.M. Iniguez-Lluhi J.A. Lefkowitz R.J. Bourne H.R. Jan Y.N. Jan L.Y. Nature. 1994; 370: 143-146Crossref PubMed Scopus (423) Google Scholar, 38Nair L.A. Inglese J. Stoffel R. Koch W.J. Lefkowitz R.J. Kwatra M.M. Grant A.O. Circ. Res. 1995; 76: 832-838Crossref PubMed Scopus (36) Google Scholar, 39Yamada M. Ho Y.K. Lee R.H. Kontanill K. Takahashill K. Katadall T. Kurachi Y. Biochem. Biophys. Res. Commun. 1994; 200: 1484-1490Crossref PubMed Scopus (36) Google Scholar). These findings suggest that a significant fraction of the basal GIRK current results from basal G protein activity. Thus, it remains possible that Gq signaling inhibits GIRK currents by inactivating the Gi protein rather than directly targeting the channel itself. I thank Maya Kunkel for providing the GIRK1/IRK2 chimera clones and assistance in the early phases of this project. I also thank Dr. N. Gautam for generously providing the Gβ1 and Gγ2 cDNA clones, the scientist who provided the d2R clone, and Guido Guidotti and Joel Bard for valuable suggestions about this manuscript.

Referência(s)