Báo cáo khoa học: Identification of the structural determinant responsible for the phosphorylation of G-protein activated potassium channel 1 by cAMP-dependent protein kinase pdf

Identification of the structural determinant responsible for the phosphorylation of G-protein activated potassium channel 1 by cAMP-dependent protein kinase Carmen Mu ¨ llner, Bibiane Steinecker, Astrid Gorischek and Wolfgang Schreibmayer Department of Biophysics, Center for Physiological Medicine, Medical University of Graz, Austria Introduction G-protein activated inwardly rectifying K + channels (GIRKs) link membrane potential to the presence of extracellular signalling molecules via G-protein cou- pled receptors and pertussis toxin sensitive, heterotri- meric, G-proteins in a membrane delimited manner [1,2]. Their role in brain function, as well as in the regulation of the heartbeat, is now well established and, increasingly, their importance in other organs is being acknowledged [3–7]. Generally, the G-protein b ⁄ c dimer is considered to be the primary opener of GIRKs, although the G-protein a-subunit is also involved in gating [8,9]. Besides this regulation via het- erotrimeric G-proteins, a huge body of work demon- strates that GIRK channel activity and trafficking is regulated by various signalling molecules, comprising nucleotides, PIP 2 , protein kinases and protein phos- phatases [10–16]. Most remarkably, the concerted action of protein phosphatase 2A (PP2A) [17] and cAMP-dependent protein kinase A (PKA) [18] provides an ‘on ⁄ off’ switch for G-protein activation. PKA-dependent activation of I K+ACh itself was observed in rat atrial cardiomyocytes [18,19] and the Keywords GIRK; IK+Ach; PKA Correspondence W. Schreibmayer, Department of Biophysics, Center for Physiological Medicine, Medical University of Graz, Harrachgasse 21 ⁄ 4, A-8010 Graz, Austria Tel: +43 316 380 4155 Fax: +43 316 380 9660 E-mail: wolfgang.schreibmayer@ medunigraz.at (Received 17 June 2009, revised 17 August 2009, accepted 24 August 2009) doi:10.1111/j.1742-4658.2009.07325.x Besides being activated by G-protein b ⁄ c subunits, G-protein activated potassium channels (GIRKs) are regulated by cAMP-dependent protein kinase. Back-phosphorylation experiments have revealed that the GIRK1 subunit is phosphorylated in vivo upon protein kinase A activation in Xeno- pus oocytes, whereas phosphorylation was eliminated when protein kinase A was blocked. In vitro phosphorylation experiments using truncated ver- sions of GIRK1 revealed that the structural determinant is located within the distant, unique cytosolic C-terminus of GIRK1. Serine 385, serine 401 and threonine 407 were identified to be responsible for the incorporation of radioactive 32 P into the protein. Furthermore, the functional effects of cAMP injections into oocytes on currents produced by GIRK1 homo- oligomers were significantly reduced when these three amino acids were mutated. The data obtained in the present study provide information about the structural determinants that are responsible for protein kinase A phosphorylation and the regulation of GIRK channels. Structured digital abstract l MINT-7260296, MINT-7260317, MINT-7260333, MINT-7260347, MINT-7260361, MINT- 7260270: PKA-cs (uniprotkb:P00517) phosphorylates (MI:0217) Girk1 (uniprotkb:P63251)by protein kinase assay ( MI:0424) Abbreviations GIRK, G-protein activated potassium channel; GST, glutathione S-transferase; PhB, phosphorylation buffer; PKA, protein kinase A; PKA-cs, catalytic subunit of PKA; PP2A, protein phosphatase 2A. 6218 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS stimulatory effect of PKA on GIRK1 ⁄ GIRK4 hete- rooligomeric channels had been attributed to a marked increase in the affinity of the phosphorylated channel protein to G b ⁄ c [20]. A signalling complex comprising (amongst other signalling molecules) GIRK1, GIRK4, G b ⁄ c , PKA, PP2A and protein phosphatase 1 was identified to exist in rat atrial membranes in situ [21], supporting the physiological relevance of this regula- tion. Several heterooligomeric combinations, including not only GIRK1 ⁄ GIRK4, but also the homooligomeric GIRK1 subunit alone, have been identified to be under PKA regulation [18]. In addition, the GIRK1 protein was also demonstrated to be a direct target for PKA- catalysed phosphorylation [17,22,23]. Despite several efforts undertaken to identify the responsible structural determinant on the GIRK1 subunit [17,23], the exact location still remains unknown. The present study aimed to obtain information about PKA phosphoryla- tion of GIRK1 in vivo and the structural motif(s) on the GIRK1 subunit serving as PKA substrate(s), as well as to assess its possible role in G-protein activa- tion of the channel. Results Back-phosphorylation of GIRK1 Previous studies had shown that the GIRK1 subunit, but not the GIRK4 subunit, isolated from bovine atrium, represented a prominent target for several S ⁄ T protein kinases in vitro, including cAMP-activated PKA [17]. To investigate whether GIRK1 represents a target for PKA also in vivo, back-phosphorylation experiments were performed. Using an antibody direc- ted against the entire C-terminus, GIRK1 was immu- nopreciptated from oocytes, expressing rat GIRK1 and subsequently submitted to back-phosphorylation using the catalytic subunit of PKA (PKA-cs) and [ 32 P]ATP[cP]. Autoradiograms of subsequent SDS gels revealed GIRK1 migrating in two bands, indicating a glycosylated and a nonglycosylated form, as reported previously [24]. Interestingly, prominent in vitro back- phosphorylation of immunoprecipitated GIRK1 was observed when RpCAMPS, a PKA inhibitor, was injected into the oocytes prior to immunoprecipitation, comparable to the control oocytes. This indicates that the heterologously expressed GIRK1 subunit was a prominent target for PKA in vitro, after the in vivo treatment of the oocytes with a PKA inhibitor and also in untreated oocytes. On the other hand, the in vitro phosphorylation signal was markedly dimin- ished when SpCAMPS, a PKA activator, was injected, indicating that the relevant PKA site(s) had been phosporylated already in the oocytes before the immu- nopreciptation (Fig. 1). Clearly, this indicates that GIRK1 is reversibly phosphorylated by native PKA in vivo in the oocytes and that the extent of basal PKA phosphorylation is low. Structural determinant responsible for PKA phosphorylation in vitro To identify the structural determinants that are respon- sible for phosphorylation of GIRK1 by PKA, fusion proteins comprising truncated forms of the cytosolic parts of GIRK1 and the glutathione S-transferase (GST) were generated and isolated from bacterial cultures. Recombinant proteins were submitted to PKA-cs-catalysed phosphorylation in vitro, using [ 32 P]ATP[cP] as a co-substrate. Whereas the entire C-terminus (amino acids 183–501) was found to be a prominent target for PKA phosphorylation in vitro, the N-terminus (amino acids 1–84) was only weakly phosphorylated. Further truncation of the C-terminus into two parts, a proximal one (G1 pC-T, amino acids 183–363) and a distal one (G1 dC-T, amino acids 365– 501), was performed to localize in more detail the phosphorylation sites within the cytosolic C-terminal part. The proximal part that is implicated in G bc bind- ing and activation [25] was not found to be prone to PKA phosphorylation, whereas the distal part was extensively involved (Fig. 2A). Interestingly, this distal part is unique for GIRK1 among the Kir3.x isoforms. Further truncations of the C-terminus revealed a 49 amino acid stretch (amino acids 362–411) that was the most prominent target for phosphorylation in vitro amongst the peptides tested (Fig. 2B). The incorpo- rated radioactivity relative to the amount of protein Fig. 1. Back-phosphorylation of GIRK1, expressed in Xenopus oocytes. Autoradiogram of SDS gel derived from immunoprecipi- tated GIRK1 showing the incorporation of radioactive 32 P into GIRK1. Before cell lysis and immunoprecipitation, SpCAMPS, RpCAMPS or nothing (control) was injected into the oocytes the oocytes. ), native oocytes; +, oocytes injected with RNA encoding GIRK1 and GIRK4. C. Mu ¨ llner et al. PKA phosphorylation of GIRK1 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS 6219 Fig. 2. In vitro phosphorylation of recombinant, truncated, GIRK1 by the catalytic subunit of PKA. In the lower panels, the mean values and standard error of mean values of relative specific radioactivity of the different constructs are plotted. The number of experiments is given in parenthesis above each bar. The mean value differs statistically significant at the P < 0.01 (**) and P < 0.001 (***) levels compared to GST alone. (A) Upper: autoradiogram of the different cytosolic regions of GIRK1. G1 pC-T, proximal C-terminus; G1 N-T, entire N-terminus; G1 dC-T, distal C-terminus. PKA-cs was present (+) or absent ()) from the reaction mixture. Lower: statistics of relative specific radioactivity (radioactivity incorporated ⁄ amount of protein) in the different cytosolic regions of the GIRK1. G1 C-T, C-terminus. (B) Upper: Autoradiogram of GIRK1 C-terminal fragments. Lower: statistics of relative specific radioactivity in the different regions of the GIRK1 C-terminus. (C) Upper: autoradiogram of the entire GIRK1 C-terminus (wild-type and mutated). Lower: statistics of relative specific radioactivity incorporation into the entire GIRK1 C-terminus (wild-type, single mutations, triple mutation and S385 + last 100 amino acids deleted). (D) Protein alignment of the four different GIRK isoforms from rat (for GIRK3, the human sequence is shown). Transmembrane regions (TM1, TM2) and pore helix (P) are marked. The N-terminal part that was used for G1 N-T, the region that was used for G1 pC-T and region that was used for G1 dC-T are marked in different shades of gray. S385, S401 and T407 are marked in black (bold and underlined). Arrows indicate the peptides tested for in vitro phosphorylation corresponding to Fig. 2B. PKA phosphorylation of GIRK1 C. Mu ¨ llner et al. 6220 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS increases in the order G1 C-T > G1 dC-T > G1 362)410 (Fig. 2A, B) and is inversely related to the molecular mass of the constructs, indicating the highest enrichment of phosphorylation sites in G1 362)410 . Three canonical PKA phosphorylation sites are located within this region of GIRK1, namely serine 385, serine 401 and threonine 407 [26]. Single mutations of these S ⁄ Ts to cysteine (an amino acid with physicochemical properties similar to serine and ⁄ or threonine but nonphosphoryla- ble) in the corresponding peptide (G1 362)411 ) signifi- cantly reduced the amount of incorporated radioactive 32 P. The most effective result was obtained by simulta- neous mutation of all three S ⁄ Ts (subsequently denoted 3*), resulting in an almost complete absence of PKA-cs- catalysed incorporation of 32 P (data not shown). A simi- lar pattern was observed for the entire C-terminus of GIRK1, when the same mutations were introduced alone or in combination (Fig. 2C). Mutation of S385C in combination with a deletion of the last 100 amino acids (DC 100 ) was slightly more effective than the 3* combination, indicating that an additional, but weak determinant may be located distal to amino acid 411. Functional aspects of mutation of S385, S401 and T407 The effects of PKA-catalysed phosphorylation on rat atrial I K+ACh as well as on basal and agonist induced GIRK1 ⁄ GIRK4 (and also homooligomeric GIRK1 F137S ) currents had been described previously [18]. To assess the role of the S ⁄ Ts in the regulation of GIRK1 via PKA in a manner that is unbiased by the eventual contributions of the other subunits in a het- erooligomer, the corresponding mutations were intro- duced into GIRK1 F137S , a mutant capable of forming functional, homooligomeric, channels in Xenopus oo- cytes [27,28]. The effects of cAMP injections on basal currents recorded from wild-type GIRK1⁄ GIRK4 het- erooligomers and GIRK1 F137S homooligomers were comparable in size, with the cAMP effect amounting to 0.31 ± 0.03 (mean ± SEM) in GIRK1 ⁄ GIRK4 heterooligomers and 0.35 ± 0.04 in GIRK1 F137S homooligomers (Fig. 3). Because cAMP injections had been shown to be effective on basal as well as on agonist-induced currents [18], cAMP injections were only occasionally performed during agonist application (data not shown) and the systematic analysis in this study was restricted to cAMP injections in the absence an agonist. The effects observed in the single amino acid mutant channels were generally reduced, with the reduction being statistically significant only for the S385C mutation (GIRK1 F137SS385C : 0.21 ± 0.04; GIRK1 F137SS401C : 0.26 ± 0.05; GIRK1 F137ST407C : Fig. 3. Effect of cAMP injections on homooligomeric GIRK1 wild- type and mutated channels. (A) Effect of cAMP injection on basal current of homooligomeric GIRK1 F137SWT channels. (B) As in (A), but with currents originating from the triple-mutated GIRK1 F137S*** protein. (C) Statistics of the effects of cAMP on basal currents of heterooligomeric GIRK1 ⁄ GIRK4, homooligomeric GIRK1 and homooligomeric, mutated GIRK1 (the cAMP effect was assessed as I cAMP ⁄ I HK of a given oocyte). Data are the mean ± SEM. The number of experiments is given in parenthesis above each bar. The mean value differs statistically significantly at the P < 0.05 (*) and P < 0.01 (**) levels compared to GIRK1 F137SWT . C. Mu ¨ llner et al. PKA phosphorylation of GIRK1 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS 6221 0.25 ± 0.05). The 3* mutant channel exerted a 50.8% reduction of the cAMP effect compared to GIRK1 F137SWT . This reduction was statistically signifi- cant at P < 0.005 (the cAMP effect for GIRK1 F137S3* in absolute numbers was 0.17 ± 0.05). It must be noted, however, that the remaining effects observed in all the mutations tested were still statistically significant com- pared to control values (= no injection). This indicates that direct PKA phosphorylation of the GIRK1 protein strongly contributes to the effects of cAMP injection, but that other, indirect, mediators of PKA action may also exist. Discussion The results obtained in the present study demonstrate that the GIRK1 subunit serves as PKA substrate both in vitro and in vivo. This is in line with observations obtained in another study demonstrating in vitro PKA phosphorylation of the GIRK1 subunit after immuno- precipitation of GIRK from bovine atrial plasma membranes [17]. In the present study, we report for the first time that the entire GIRK1 subunit serves as a substrate for PKA-catalysed phosphorylation in vivo, when coexpressed with GIRK4 in Xenopus oocytes. Furthermore, this phosphorylation was regulated by cytosolic injections of PKA activators and inhibitors, suggesting it to be the basis for the functional effects of cAMP injections on GIRK currents. PKA-induced phosphorylation of the recombinant entire GIRK1 carboxy terminus in vitro had been reported by us previously [22] and was recently confirmed by Lopes et al. [23]. Previously, attempts were made to identify the struc- tural determinant that is responsible for PKA phos- phorylation of GIRK1. In a detailed investigation, Medina et al. [17] coexpressed GIRK1 with GIRK4 in HEK-293 cells. Truncation of GIRK1 after amino acid 373 resulted in a complete loss of spontaneous in vivo phosphorous incorporation into the protein, whereas truncation to amino acid 419 had no effect. However, when all seven S ⁄ Ts located in this 46 amino acid stretch, including S385, S401 and T407 described in the present study, were mutated to alanines, incorpora- tion of radioactive phosphorus was still observed to a considerable extent. In this experiment, however, Med- ina et al. [17] had measured total 32 P incorporation in cell culture rather than phosphorylation directly cataly- sed by PKA (as performed in the present study). We conclude that the failure of these seven mutations to abolish protein phosphorylation was a result of other protein kinases masking the PKA-catalysed part. Indeed, protein kinases other than PKA, including tyrosine kinases, have been demonstrated to directly phosphorylate GIRK1 [12]. The observation that PP2A was unable to dephosphorylate the constitutively phosphorylated GIRK1 protein to a considerable extent but the 373–418 amino acid region was essential for functional regulation by PP2A [17] further fosters the hypothesis that GIRK1 serves as a substrate to a manifold of protein kinases, whereas S385, S401 and T407 are essential for specific PKA phosphorylation and likely also for the dephosphorylation by PP2A. Indeed, there is a broad overlap in substrate specificity between PKA and PP2A and both enzymes are known to colocalize in cellular microdomains, as well as in atrial cardiomyocytes, together with GIRK1 [21,29]. In another study, S221 and S315 were identified to be involved in the inhibitory action of H89, a PKA inhib- itor, on GIRK1 ⁄ GIRK4 currents in Xenopus oocytes [23]. In the present study, we were unable to observe PKA-catalysed incorporation of phosphate into the peptides comprising these residues. Hence, we conclude that both amino acids are indirectly involved in the PKA regulation of the GIRK1 subunit but do not serve directly as a substrate for PKA itself. Recently, using a different experimental approach employing mass spectroscopic methods, Rusinova et al. [30] has identified S385 as a prominent and specific target for PKA-catalysed phosphorylation in vitro, greatly sup- porting the result obtained in the present study. Amongst the three determinants identified by us, S385 had the greatest impact on PKA phosphorylation in vitro, being almost as effective in abolishing the phosphorylation of GIRK1 C-T as the triple mutation. This is in excellent agreement with theoretical predic- tions because: (i) arginines both at positions )2 and )3 (viewed from S385) exist; (ii) a hydrophobic valine is located at position +1; and (iii) a serine represents a stronger determinant than threonine does [31]. In com- parison, S401 and T407 have only a single specificity determinant in their surroundings, a lysine at )3 (S401) and at )2 (T407), with lysine representing a weaker determinant than arginine. Accordingly, their contribution to PKA-catalysed in vitro phosphoryla- tion is substantial, but smaller, compared to that of S385. On the other hand S ⁄ T protein phosphatases, especially PP2A, display a striking preference for phos- phothreonyl residues over phosphoseryl residues and hence T407 may represent an excellent target for this enzyme [31]. Activation of heterooligomeric GIRK1 ⁄ GIRK4 and homooligomeric GIRK1 F137S by PKA is well estab- lished [18,23]. The data obtained in the present study show that S385, S401 and T407 contribute substan- tially to this functional effect via the GIRK1 subunit PKA phosphorylation of GIRK1 C. Mu ¨ llner et al. 6222 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS but PKA activation was still observed to some extent, demonstrating that other, indirect, effects also contrib- ute. Furthermore, Medina et al. [17] were unable to completely eliminate dephosphorylation-mediated effects on heterooligomeric GIRK1 ⁄ GIRK4 channels by mutating seven S ⁄ Ts in GIRK1, including the struc- tural determinants identified in the present study. We suggest that also in this case the contribution of GIRK1 was masked by the indirect actions of PKA. For example, such an indirect mediator of PKA action was identified recently as RGS10 in rat atrial cells [32]. Another possibility explaining why we were unable to completely eliminate the effects of cAMP by mutating S385, S401 and T407 may be that isoforms other than GIRK1 may be also targets for protein phosphoryla- tion in the protein complex: for example, GIRK2 had been shown to exert dramatic PKA effects upon cAMP injection [18]. In our specific case, residual, end- ogeneous, GIRK5 subunits that were resistant to anti- sense oligonucleotide treatment may have contributed to the remaining PKA effect. The distal C-T of GIRK1, which is unique amongst GIRK isoforms, has given rise to various proposals about its peculiar func- tion in the past. In the present study, we identified three S⁄ Ts within this region that play an important role in direct phosphorylation by PKA and in mediat- ing PKA actions on the functional, homooligomeric complex. The next objective in this field will be to assess the possible contribution of GIRK1 isoforms to PKA-mediated effects on heterooligomeric channels, with the aim of understanding in more detail this important regulation concerning G-protein activation of K + channels. Experimental procedures Antibodies, reagents and solutions GIRK1-Ab: Anti-Kir3.1 (APC-005; Alomone Labs, Ltd, Jerusalem, Israel); Protein A Sepharose (CL-4B beads; Pharmacia LKB Biotechnology AB, Uppsala, Sweden); SpCAMPS, RpCAMPS (A166, A165, respectively; Sigma- Aldrich, St Louis, MO, USA); [ 32 P]ATP[cP] (25001748; GE Healthcare Europe GmbH, Vienna, Austria); PKA-cs (1529 307; Boehringer Ingelheim GmbH, Ingelheim, Germany; 400 mU per 80 mL). All other reagents used were of reagent grade throughout if not stated otherwise. Phosphor- ylation buffer (PhB): 25 mmolÆL )1 HEPES ⁄ Na; pH 7.4; 5 mmolÆL )1 MgCl 2 ; 5 mmolÆL )1 EGTA; 0.05% Tween-20. Homogenization buffer: 100 mmolÆL )1 sodium phosphate buffer; pH 5.8; 10 mmolÆL )1 EDTA; 5 mmolÆL )1 a-glycero- phosphate; 5 mmolÆL )1 BSA; 0.5 mmolÆL )1 vanadate; 50 mmolÆL )1 KF; 20% Triton-X-100. Seven mililiters of buffer were supplemented with one tablet of complete mini (Roche, Basel, Switzerland). 4· SDS-loading buffer: 400 mmolÆL )1 Tris ⁄ Cl, pH 6.8, 20% sucrose, 4% SDS, 20% mercaptoethnole; Comassie staining solution: 40 mgÆL )1 Co- massie blue, 500 mLÆL )1 methanol, 100 mLÆL )1 acetic acid; Destain I: 500 mLÆL )1 methanol, 100 mLÆL )1 acetic acid; Destain II: 50 mLÆL )1 methanol, 70 mLÆL )1 acetic acid. ND96: 96 mmolÆL )1 NaCl, 2 mmolÆL )1 KCl, 1 mmolÆL )1 MgCl 2 , 1 mmolÆL )1 CaCl 2 , 5 mmolÆL )1 Hepes, buffered with NaOH to pH 7.4; NDE: same as ND96, but CaCl 2 was 1.8 mmolÆL )1 and 2.5 mmolÆL )1 pyruvate and 0.1% antibiot- ics (G-1397; ·1000 stock from Sigma-Aldrich) were added; HK: 96 mmolÆL )1 KCl, 2 mmolÆL )1 NaCl, 1 mmolÆL )1 MgCl 2 , 1 mmolÆL )1 CaCl 2 , 5 mmolÆL )1 Hepes buffered with KOH to pH 7.4; Glutathione buffer: 120 mmolÆL )1 NaCl; 0.05% Tween-20; 100 mmolÆL )1 Tris; 15 mmolÆL )1 glutathione pH 8.0. Immunoprecipitation For the immunoprecipitation experiments, Xenopus laevis oocytes were injected with cRNAs coding for GIRK1 and GIRK4 (5 ng per oocyte for each RNA) and the KHA2 antisense oligonucleotide as described previously [33] (25 ng per oocyte). After incubation of cells for 6 days at 19 °C, the oocytes were checked for expression with the two elec- trode voltage clamp technique as described below and the cells were injected with either SpcAMPS or RpcAMPS (2.5 mmolÆL )1 , 20 nL per oocyte). Incubation with PKA activator ⁄ inhibitor was allowed to take place for approxi- mately 30 min; thereafter, oocytes were homogenized in homogenization buffer by pipetting up and down. Immuno- precipitation of GIRK1 channels from 15 oocytes solubi- lized in 100 lL of homogenization buffer was initiated by adding 4 lL of non-immune IgG and incubating for 1 h at room temperature to prevent unspecific binding. Twenty microliters of 10% Protein A Sepharose per reaction were added for precipitation of unspecific antibody complexes, whereas 4 lL of GIRK1-Ab were added to the supernatant. The immunoprecipitation reaction was incubated over night under constant agitation at 4 °C. Antibody complexes were precipitated by another addition of 10% Protein A Sepha- rose and incubation for 1 h at 4 °C. Back-phosphorylation In vitro back-phosphorylation was performed after the immunoprecipitate was washed twice with ice-cold phos- phorylation buffer. Then, 1 lL[ 32 P]ATP[cP] and 1 lLof PKA-cs were added to the pellet for 5 min at 30 °C. Subse- quently, the reaction was put on ice, washed twice with phosphorylation buffer and supplemented with 40 lLof SDS-loading buffer. The denatured proteins were loaded on a 10% SDS gel [34] and run for 1 h at 150 V. Afterwards, the gel was stained with Comassie staining solution, C. Mu ¨ llner et al. PKA phosphorylation of GIRK1 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS 6223 destained and dried on a slab gel dryer before exposure to X-ray film. Genetic engineering Plasmid vectors were grown in bacteria, purified and linear- ized using standard procedures [35]. Plasmids with inserts encoding m 2 R [18] and GIRK1 F137S [28] have been described previously. Plasmids for the production of recom- binant protein were constructed by a PCR, where forward and reverse primers encoded the desired parts of the C- and N-termini of GIRK1. These sequences were each preceded or followed, respectively, by restriction recognition sequences appropriate for cloning in frame with GST in the bacterial expression vector pGEX-4T-1. Isolated PCR prod- ucts were digested with the appropriate enzymes, ligated into pGEX-4T-1 and the sequences were verified by conventional sequencing. Mutants of GIRK1 F137S and truncated GIRK1 ⁄ GST fusion constructs for protein pro- duction were produced by PCR using homologous primers containing the appropriate mutation in addition to a silent mutation creating an additional restriction site to facilitate identification of the mutants. Before bacterial transforma- tion, template DNA was digested with DpnI as described previously [35]. Recombinant protein purification Constructs were transfected into BL-21(RIL) competent cells (Stratagene, La Jolla, CA, USA), the corresponding proteins were overexpressed and purified as described previ- ously [25]. Protein was quantitated by the method of Brad- ford [36], diluted to a concentration of 1 lgÆlL )1 , and aliquots were shock frozen in liquid N 2 and stored at )70 °C until use. In vitro phosphorylation One microgram of the appropriate protein was incubated in 300 lL of PhB containing 18.5 kBq [ 32 P]ATP[cP] and 0.4 lL of PKA-cs for 30 min at room temperature (agitated by a Labquake laboratory shaker; Cole-Parmer Instrument Company, Vernon Hills, IL, USA). Next, 30 lL of gluta- thione Sepharose 4B beads (washed and suspended in PhB) were added and incubation ⁄ agitation continued for another 30 min. Samples were centrifuged (1 min; maximum g; picofuge, Stratagene) and the supernatant discarded care- fully. Beads were washed three times in 1 mL of PhB by resuspension and centrifugation. Finally, the protein was eluted by adding 30 lL of glutathione buffer to the beads and incubating for 10 min. Thirty-two microliters of super- natant were removed, combined with 10 lLof4· SDS loading buffer and run on a 12% SDS gel [34]. Gels were stained with Comassie blue, dried and scanned. Subse- quently, autoradiograms were performed using the Storm Phosphorimager (GE Healthcare Europe GmbH, Vienna, Austria). Incorporation of radioactive 32 P into the protein was quantitated and normalized to the total amount of pro- tein as detected by the Comassie stain (= relative specific radioactivity). Xenopus laevis oocyte expression Oocytes were prepared as described previously [37]. Approxi- mately 24 h afterwards, they were injected with the KHA2 antisense oligonucleotide (25 ng per oocyte) together with the appropriate RNA (amounts in pg per oocyte): m 2 R: 1500; GIRK1 F137S : 37.5; GIRK1 F137SS385C : 37.5; GIRK1 F137SS401C : 37.5; GIRK1 F137ST407C : 37.5; GIRK1 F137SS385CS401CT407C : 150. Oocytes were kept in NDE at 19 °C for 3–5 days after injection before electrophysiological experiments were performed. Electrophysiology Oocytes were placed in a recording chamber, allowing superfusion with either ND96 or HK (with and without 10 )5 molÆL )1 acetylcholine) at 21 °C and currents were recorded via the two electrode voltage clamp technique using agarose cushion electrodes [38] and the Geneclamp 500 amplifier (Axon Instruments, Foster City, CA, USA). Membrane potential was kept at )80 mV and the medium was changed from ND96 to HK, HK+ acetylcholine, HK and back to ND96. Cytosolic injection of cAMP and cAMP analogs (30–60 pmol per oocyte) was performed as described previously [18]. The current increase, following cAMP injection, was normalized to the basal current of the given oocyte (where the basal current is defined as the current induced by a change of the extracellular medium from ND96 to HK, designated as I HK in Fig. 3). Current traces were low pass filtered at 10 Hz and digitized using the Digidata 1322A interface (Axon Instruments) con- nected to computer running pclamp 9.2 software (Axon Instruments). Statistical analysis Given experimental groups were tested for statistical signifi- cant differences using Student’s t-test and sigmaplot 9.0 (Systat Software Inc., Chicago, IL, USA). Acknowledgements We thank Dr D. Logothetis (Virginia Commonwealth University, Richmond, VA, USA) for kindly providing the clone encoding G1 F137S and Dr T. DeVaney (Medical University of Graz, Graz, Austria) for correcting the English language. Support provided by PKA phosphorylation of GIRK1 C. Mu ¨ llner et al. 6224 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS the Austrian Research Foundation (SFB708) and the Research Foundation of the Austrian National Bank (OENB12575) is gratefully acknowledged. References 1 Dascal N (2001) Ion-channel regulation by G proteins. Trends Endocrinol Metab 12, 391–398. 2 Dascal N (1997) Signalling via the G protein-activated K+ channels. Cell Signal 9, 551–573. 3 Smith PA, Sellers LA & Humphrey PPA (2001) Somatostatin activates two types of inwardly rectifying K+ channels in MIN-6 cells. J Physiol (Lond) 532, 127–142. 4 Shankar H, Kahner BN, Prabhakar J, Lakhani P, Kim S & Kunapuli SP (2006) G-protein-gated inwardly rectifying potassium channels regulate ADP-induced cPLA(2) activity in platelets through Src family kinases. Blood 108, 3027–3034. 5 Perry CA, Pravetoni M, Teske JA, Aguado C, Erickson DJ, Medrano JF, Lujan R, Kotz CM & Wickman K (2008) Predisposition to late-onset obesity in GIRK4 knockout mice. Proc Natl Acad Sci USA 105, 8148– 8153. 6 Iwanir S & Reuveny E (2008) Adrenaline-induced hyperpolarization of mouse pancreatic islet cells is mediated by G protein-gated inwardly rectifying potassium (GIRK) channels. Pflugers Arch 456, 1097– 1108. 7 Stringer BK, Cooper AG & Shepard SB (2001) Overex- pression of the G-protein inwardly rectifying potassium channel 1 (GIRK1) in primary breast carcinomas corre- lates with axillary lymph node metastasis. Cancer Res 61, 582–588. 8 Schreibmayer W, Dessauer CW, Vorobiov D, Gilman AG, Lester HA, Davidson N & Dascal N (1996) Inhibi- tion of an inwardly rectifying K+ channel by G-protein alpha-subunits. Nature 380, 624–627. 9 Rubinstein M, Peleg S, Berlin S, Brass D & Dascal N (2007) G alpha(i3) primes the G protein-activated K+ channels for activation by coexpressed G beta gamma in intact Xenopus oocytes. J Physiol (Lond) 581, 17–32. 10 Ruppersberg JP (2000) Intracellular regulation of inward rectifier K+ channels. Pflugers Arch 441, 1–11. 11 Mao JZ, Wang XR, Chen FX, Wang RP, Rojas A, Shi Y, Piao H & Jiang C (2004) Molecular basis for the inhibition of G protein-coupled inward rectifier K+ channels by protein kinase C. Proc Natl Acad Sci USA 101, 1087–1092. 12 Mora SI & Escobar LI (2005) Phosphorylation of a tyrosine at the N-terminus regulates the surface expres- sion of GIRK5 homomultimers. FEBS Lett 579, 3019– 3023. 13 Ippolito DL, Xu M, Bruchas MR, Wickman K & Chavkin C (2005) Tyrosine phosphorylation of K(ir)3.1 in spinal cord is induced by acute inflammation, chronic neuropathic pain, and behavioral stress. J Biol Chem 280, 41683–41693. 14 Chung HJ, Qian X, Ehlers M, Jan YN & Jan LY (2009) Neuronal activity regulates phosphorylation- dependent surface delivery of G protein-activated inwardly rectifying potassium channels. Proc Natl Acad Sci USA 106, 629–634. 15 Huang CL, Feng SY & Hilgemann DW (1998) Direct activation of inward rectifier potassium channels by PIP2 and its stabilization by G beta gamma. Nature 391, 803–806. 16 Zhang HL, He C, Yan XX, Mirshahi T & Logothetis DE (1999) Activation of inwardly rectifying K+ chan- nels by distinct PtdIns(4,5)P-2 interactions. Nat Cell Biol 1, 183–188. 17 Medina I, Krapivinsky G, Arnold S, Kovoor P, Krapivinsky L & Clapham DE (2000) A switch mechanism for G ss gamma activation of I-KACh. J Biol Chem 275, 29709–29716. 18 Mullner C, Vorobiov D, Bera AK, Uezono Y, Yakubovich D, Frohnwieser-Steinecker B, Dascal N & Schreibmayer W (2000) Heterologous facilitation of G protein-activated K+ channels by beta-adrenergic stimulation via cAMP-dependent protein kinase. J Gen Physiol 115, 547–557. 19 Kim D (1990) Beta-adrenergic regulation of the muscarinic-gated K+ channel via cyclic Amp-dependent protein-kinase in atrial cells. Circ Res 67, 1292–1298. 20 Mullner C, Yakubovich D, Dessauer CW, Platzer D & Schreibmayer W (2003) Single channel analysis of the regulation of GIRK1 ⁄ GIRK4 channels by protein phosphorylation. Biophys J 84, 1399–1409. 21 Nikolov EN & Ivanova-Nikolova TT (2004) Coordina- tion of membrane excitability through a GIRK1 signal- ing complex in the atria. J Biol Chem 279, 23630– 23636. 22 Muellner C, Steinecker-Frohnwieser B & Schreibmayer WF (2002) Structural determinant responsible for the cAMP dependent phosphorylation of the GIRK1 sub- unit. Biophysical Society Meeting Abstracts Biophys J Supplement, 82, 2890-Pos. 23 Lopes CMB, Remon JI, Matavel A, Sui JL, Keselman I, Medei E, Shen YM, Rosenhouse-Dantsker A, Rohacs T & Logothetis DE (2007) Protein kinase A modulates PLC-dependent regulation and PIP2-sensitivity of K+ channels. Channels 1, 124–134. 24 Krapivinsky G, Gordon EA, Wickman K, Velimirovic B, Krapivinsky L & Clapham DE (1995) The G-pro- tein-gated atrial K+ channel I-Kach is a Heteromulti- mer of 2 inwardly rectifying K+-channel proteins. Nature 374, 135–141. 25 Ivanina T, Rishal I, Varon D, Mullner C, Frohnwieser- Steinecke B, Schreibmayer W, Dessauer CW & Dascal N (2003) Mapping the G beta gamma-binding sites in C. Mu ¨ llner et al. PKA phosphorylation of GIRK1 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS 6225 GIRK1 and GIRK2 subunits of the G protein-activated K+ channel. J Biol Chem 278, 29174–29183. 26 Kennelly PJ & Krebs EG (1991) Consensus sequences as substrate-specificity determinants for protein-kinases and protein phosphatases. J Biol Chem 266, 15555– 15558. 27 Chan KW, Sui JL, Vivaudou M & Logothetis DE (1996) Control of channel activity through a unique amino acid residue of a G protein-gated inwardly recti- fying K+ channel subunit. Proc Natl Acad Sci USA 93, 14193–14198. 28 Vivaudou M, Chan KW, Sui JL, Jan LY, Reuveny E & Logothetis DE (1997) Probing the G-protein regulation of GIRK1 and GIRK4, the two subunits of the K-ACh channel, using functional homomeric mutants. J Biol Chem 272, 31553–31560. 29 McConnachie G, Langeberg LK & Scott JD (2006) AKAP signaling complexes: getting to the heart of the matter. Trends Mol Med 12, 317–323. 30 Rusinova R, Shen Y-MA, Dolios G, Padovan J, Yang H, Kirchberger M, Wang R & Logothetis DE (2009) Mass spectrometric analysis reveals a functionally important PKA phosphorylation site in a Kir3 channel subunit. Pflugers Arch-Eur J Physiol 458, 303–314. 31 Pinna LA & Ruzzene M (1996) How do protein kinases recognize their substrates? Biochim Biophys Acta 1314, 191–225. 32 Bender K, Nasrollahzadeh P, Timpert M, Liu B, Pott L & Kienitz MC (2008) A role for RGS10 in beta-adren- ergic modulation of G-protein-activated K+ (GIRK) channel current in rat atrial myocytes. J Physiol (Lond) 586, 2049–2060. 33 Hedin KE, Lim NF & Clapham DE (1996) Cloning of a Xenopus laevis inwardly rectifying K+ channel subunit that permits GIRK1 expression of I-KACh currents in oocytes. Neuron 16, 423–429. 34 Laemmli UK (1970) Cleavage of structural proteins during assembly of head of bacteriophage-T4. Nature 227, 680–685. 35 Sambrook J & Russel DW (2001) Molecular Cloning, 3 edn. CSH, Cold Spring Harbour. 36 Bradford MM (1976) Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein-dye binding. Anal Biochem 72, 248–254. 37 Dascal N & Lotan I (1992) Expression of exogenous ion channels and neurotransmitter receptors in RNA- injected Xenopus oocytes.InMethods in Neurobiology (Longstaff ARP, ed), pp. 205–225. Humana Press, Totowa, NJ. 38 Schreibmayer W, Lester HA & Dascal N (1994) Voltage clamping of Xenopus laevis oocytes utilizing agarose- cushion electrodes. Pflugers Arch 426, 453–458. PKA phosphorylation of GIRK1 C. Mu ¨ llner et al. 6226 FEBS Journal 276 (2009) 6218–6226 ª 2009 The Authors Journal compilation ª 2009 FEBS . Identification of the structural determinant responsible for the phosphorylation of G -protein activated potassium channel 1 by cAMP-dependent protein kinase Carmen. revised 17 August 2009, accepted 24 August 2009) doi :10 .11 11/ j .17 42-4658.2009.07325.x Besides being activated by G -protein b ⁄ c subunits, G -protein activated potassium