Coordinated Agonist Regulation of Receptor and G Protein Palmitoylation and Functional Rescue of Palmitoylation-deficient Mutants of the G Protein G11α following Fusion to the α1b-Adrenoreceptor

Transfection of either the α1b-adrenoreceptor or Gα11 into a fibroblast cell line derived from a Gαq/Gα11 double knockout mouse failed to produce elevation of intracellular [Ca2+] upon the addition of agonist. Co-expression of these two polypeptides, however, produced a significant stimulation. Co-transfection of the α1b-adrenoreceptor with the palmitoylation-resistant C9S,C10S Gα11 also failed to produce a signal, and much reduced and kinetically delayed signals were obtained using either C9S Gα11 or C10S Gα11. Expression of a fusion protein between the α1b-adrenoreceptor and Gα11 allowed [Ca2+] i elevation, and this was also true for a fusion protein between the α1b-adrenoreceptor and C9S,C10S Gα11, since this strategy ensures proximity of the two polypeptides at the cell membrane. For both fusion proteins, co-expression of transducin α, as a β·γ-sequestering agent, fully attenuated the Ca2+signal. Both of these fusion proteins and one in which an acylation-resistant form of the receptor was linked to wild type Gα11 were also targets for agonist-regulated [3H]palmitoylation and bound [35S]guanosine 5′-3-O-(thio)triphosphate (GTPγS) in an agonist concentration-dependent manner. The potency of agonist to stimulate [35S]GTPγS binding was unaffected by the palmitoylation potential of either receptor or G protein. These studies provide clear evidence for coordinated, agonist-mediated regulation of the post-translational acylation of both a receptor and partner G protein and demonstrate the capacity of such fusions to bind and then release β·γ complex upon agonist stimulation whether or not the G protein can be palmitoylated. They also demonstrate that Ca2+ signaling in EF88 cells by such fusion proteins is mediated via release of the G protein β·γ complex.

The phosphoinositidase C-linked G proteins G␣ q and G␣ 11 are widely co-expressed (1)(2)(3)(4)(5)(6). To gain insight into their function and into potential functions of G protein-coupled receptors (GPCRs) 1 in the absence of these G proteins, their genes have been inactivated in mice (6 -12). Although double G␣ q /G␣ 11 knockout mice are not viable, cells derived from such embryos have been extremely useful, particularly in the elucidation of function of the G 12 /G 13 class of ubiquitously expressed, pertussis-insensitive G proteins (11,12). Such cells have also been used to demonstrate that agonist-induced internalization of G␣ q /G␣ 11 -coupled GPCRs is not dependent upon the presence of these G proteins (13).
Post-translational palmitoylation close to the N terminus of the ␣ subunits of heterotrimeric G proteins appears to be central to their effective interaction with the plasma membrane and thus their capacity to transduce signals from GPCRs to effectors (14 -16). Since the thioester bond between the protein and the fatty acid is easily cleaved, there is the potential for dynamic regulation of G protein acylation. This has been best examined for the adenylyl cyclase stimulatory G protein G s , where activation mediated by GPCRs or cholera toxin has been shown to alter [ 3 H]palmitoylation of the G protein (17)(18)(19). G␣ 11 possesses two adjacent cysteine residues at positions 9 and 10, which have been established to be sites of acylation (20,21). Mutation of these two cysteine residues results in production of a soluble polypeptide. Few reports have provided evidence for regulation of the palmitoylation of G␣ q /G␣ 11 , although elevated incorporation of [ 3 H]palmitate into these G proteins by stimulation of the gonadotrophin-releasing hormone receptor has been reported (22) and by molecularly undefined receptors for 5-hydroxytryptamine in rat brain cortical membranes (23).
Production of fusion proteins between GPCRs and G protein ␣ subunits has become a popular means to explore many aspects of the detailed interactions between these protein classes (24,25). Because the N terminus of the G protein ␣ subunit is fused directly to the C terminus of the GPCR in such constructs, it is often unclear whether this might limit interaction with the G protein ␤⅐␥ complex. This uncertainty reflects that the N terminus of the ␣ subunit is an important contact interface for ␤⅐␥, although key amino acids for this interaction are thought to be located some 15 amino acids away from the site(s) of palmitoylation (26).
Herein, we use fusion proteins between the ␣ 1b -adrenoreceptor and forms of G␣ 11 to demonstrate that the fusion proteins are activated and regulate their palmitoylation status in re-sponse to agonist. By examining fusion constructs in which either the receptor or the G protein is resistant to palmitoylation, we also demonstrate that the acylation status of both polypeptide partners is dynamically regulated by agonist. Moreover, palmitoylation of G␣ 11 is not necessary for the binding and release of ␤⅐␥ complex and further transduction of the signal.

MATERIALS AND METHODS
A fibroblast cell line (EF88) derived from a combined G␣ q /G␣ 11 double knockout mouse (9 -10, 13) was the gift of Dr. M. I. Simon (California Institute of Technology, Pasadena, CA).
Construction of Fusion Proteins-Production and subcloning of wild type and palmitoylation-resistant ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins was performed in two separate stages. In the first step, the coding sequence of each form of G␣ 11 (20,21) was modified by PCR amplification using the amino-terminal primer 5Ј-GAGGACGGTACCACTCTG-GAGTCCATG-3Ј the initiating Met of G␣ 11 was removed, and both a KpnI restriction site (underlined) and a two-amino acid spacer (Gly-Asn) were introduced. Using the C-terminal primer 5Ј-TTGTGCGGC-CGCCGGTCACACCAGGTT-3, a NotI restriction site (underlined) was introduced downstream of the stop codon of G␣ 11 . The amplified fragments digested with KpnI and NotI were subcloned into similarly digested pcDNA3 expression vector (Invitrogen). To obtain the various ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins, the coding sequence of the wild type or C365A, C367G hamster ␣ 1b -adrenoreceptor was amplified by PCR. Using the amino-terminal primer 5Ј-GACGGTACCTCTA-AAATGAATCCCGAT-3Ј, a KpnI restriction site (underlined) was introduced upstream of the initiator Met. Using the carboxyl-terminal primer 5Ј-GTCCCTGGTACCAAAGTGCCCGGGTG-3Ј, a second KpnI restriction site (underlined) was introduced immediately upstream of the stop codon. Finally, the G␣ 11 constructs in pcDNA3 were digested with KpnI and ligated together with the PCR product of the ␣ 1b -adrenoreceptor amplification also digested with KpnI. The open reading frames thus produced represent the coding sequence of either ␣ 1badrenoreceptor-G␣ 11 , C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 , or ␣ 1badrenoreceptor-C9S,C10S G␣ 11 . Each was fully sequenced before its expression and analysis.
Transient Transfection of HEK293 Cells-HEK293 cells were maintained in DMEM supplemented with 0.292 g/liter L-glutamine and 10% (v/v) newborn calf serum at 37°C in a 5% CO 2 humidified atmosphere. Cells were grown to 60 -80% confluence before transient transfection in 60-mm dishes. Transfection was performed using LipofectAMINE reagent (Life Technologies) according to the manufacturer's instructions.
3 H Palmitoylation-Cells were labeled with 0.5 mCi/ml [9,10-3 H]palmitic acid in DMEM supplemented with 0.292 g/liter L-glutamine, 5% (v/v) dialyzed newborn calf serum, and 5 mM pyruvic acid at 37°C in a 5% CO 2 humidified atmosphere. After incubation for the appropriate time in the presence and absence of varying concentrations of phenylephrine, reactions were terminated by the addition of 200 l of 1% (w/v) SDS. Proteins were denatured by passage through a 25-gauge needle followed by 5-min incubation at 100°C. After chilling to 4°C, 800 l of Kahn solubilization buffer (1% (v/v) Triton X-100, 10 mM EDTA, 100 mM NaH 2 PO 4 , 10 mM NaF, 50 mM HEPES (pH 7.2)) was added, and the samples were precleared by incubation for 1 h at 4°C with 100 l of Pansorbin (Calbiochem). The precleared supernatants were then incubated for 16 h at 4°C with protein-A-Sepharose and 10 l of antiserum CQ (27,28). Immune complexes were isolated by centrifugation, washed three times with Kahn immunoprecipitation buffer (1% (v/v) Triton X-100, 100 mM NaCl, 100 mM NaF, 50 mM NaH 2 PO 4 , 50 mM HEPES (pH 7.2) plus 0.5% SDS), and eluted from the protein A-Sepharose by the addition of electrophoresis buffer containing 20 mM dithiothreitol and heating to 80°C for 3 min. Analysis was by SDSpolyacrylamide gel electrophoresis, using 10% (w/v) polyacrylamide resolving gels and by autoradiography. Reactions were incubated for 15 min at 30°C and were terminated by the addition of 0.5 ml of ice-cold buffer, containing 20 mM HEPES (pH 7.4), 3 mM MgCl 2 and 100 mM NaCl. The samples were centrifuged at 16,000 ϫ g for 15 min at 4°C, and the resulting pellets were resuspended in solubilization buffer (100 mM Tris, 200 mM NaCl, 1 mM EDTA, 1.25% Nonidet P-40) plus 0.2% sodium dodecylsulfate. Samples were precleared with Pansorbin (Calbiochem), followed by immunoprecipitation with CQ antiserum. Finally, the immunocomplexes were washed twice with solubilization buffer, and bound [ 35 S]GTP␥S was estimated by liquid-scintillation spectrometry.
[ 3 H]Prazosin Binding Studies-Binding assays were initiated by the addition of 3 g of cell membranes to an assay buffer (50 mM Tris-HCl, 100 mM NaCl, 3 mM MgCl 2 , pH 7.4) containing [ 3 H] prazosin (0.05-10 nM in saturation assays and 0.5 nM for competition assays) in the absence or presence of increasing concentrations of phenylephrine (200-l final volume). Nonspecific binding was determined in the presence of 100 M phentolamine. Reactions were incubated for 30 min at 30°C, and bound ligand was separated from free by vacuum filtration through GF/B filters. The filters were washed twice with assay buffer, and bound ligand was estimated by liquid scintillation spectrometry.
[Ca 2ϩ ] Imaging-A fibroblast cell line, (EF88), derived from the embryos of mice in which the ␣ subunits of both G q and G 11 had been knocked out by targeted gene disruption (9 -10, 13) were grown in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum and L-glutamine (1 mM) in a 95% air and 5% CO 2 atmosphere at 37°C. For transfection experiments, a portion of the cells harvested during trypsinization were plated onto glass coverslips (22-mm diameter, grade 0 thickness), and after a 24-h growth period, cells were transfected using LipofectAMINE (Life Technologies) according to the manufacturer's instructions. After 3 h, cells were washed twice with OPTI-MEM-1 and then cultured in DMEM growth medium for a further 24 h. In some experiments, after an initial 24-h transfection/growth period, the transfected cells were treated with pertussis toxin (25 ng/ml, 24 h).
Measurement of [Ca 2ϩ ] i -Transfected cells growing on coverslips were loaded with the Ca 2ϩ -sensitive dye Fura-2 by incubation (15-20 min, 37°C) in physiological control saline solution, 130 mM NaCl, 5 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 20 mM HEPES, 10 mM D-glucose, pH adjusted to 7.4 using NaOH) containing the dye's membrane-permeant acetoxymethylester form (1.0 M). A rise in [Ca 2ϩ ] i causes a corresponding rise in the Fura-2 fluorescence ratio recorded from cells loaded with this dye, which allows receptor-mediated changes in [Ca 2ϩ ] i to be monitored using standard, microspectrofluorimetric techniques (29). An Optoscan monochromator (Cairn Research, Faversham, Kent, UK) was used to alternate the excitation wavelength between 340 and 380 nm (band pass of 10 nm) and to control the excitation frequency. Fura-2 fluorescence emission at 510 nm was monitored either by a low noise COHU CCD camera or a photomultiplier tube with a bialkali photocathode. Images acquired with the CCD camera were stored and analyzed digitally under the control of Meta Fluor imaging software (Universal Imaging Corp., West Chester, PA).
Agonist-evoked [Ca 2ϩ ] i responses were quantified by peak height (i.e. difference between the base-line resting ratio level and that attained at the peak response). Responses were pooled and are expressed as the mean Ϯ S.E. of at least five experiments, with vertical lines (see Fig. 3) representing S.E. Statistical significance of any difference between means was determined using Student's t test. Differences in the magnitude of [Ca 2ϩ ] i responses evoked by phenylephrine in untreated and pertussis toxin-treated cells were evaluated by Student's paired t test.

RESULTS
Cells of a fibroblast-derived line (EF88) from the embryo of a combined G␣ q /G␣ 11 double knockout mouse were grown on glass coverslips. These were transiently transfected with either the hamster ␣ 1b -adrenoreceptor or the mouse G protein G␣ 11 . Co-transfection with enhanced green fluorescent protein allowed identification of positively transfected cells. Following loading of the cells with the Ca 2ϩ indicator Fura-2/AM (1 M, 15 min, 37°C), single cell Ca 2ϩ imaging was performed in the absence of external Ca 2ϩ . In both cases, the addition of the ␣ 1 -adrenoreceptor-selective agonist phenylephrine (10 M) failed to alter basal intracellular [Ca 2ϩ ] ([Ca 2ϩ ] i ) (Fig. 1, A and  B). However, co-expression of both the ␣ 1b -adrenoreceptor and G␣ 11 resulted in a robust and rapid elevation of [Ca 2ϩ ]i ( Fig. 1, We have previously demonstrated that G␣ 11 can be post-translationally acylated on both Cys 9 and Cys 10 (20) and that mutation of both of these amino acids to Ser prevents membrane association of the G protein (20,21). Co-transfection of EF88 cells with the ␣ 1b -adrenoreceptor and C9S,C10S G␣ 11 thus also failed to result in a phenylephrine-mediated elevation of [Ca 2ϩ ] i (Fig. 2). Equivalent studies with either C9S G␣ 11 or C10S G␣ 11 did produce an agonist-dependent rise in [Ca 2ϩ ] i (Fig. 2), but the magnitude of the response was substantially less than with wild type G␣ 11 and was kinetically much slower. In both of these regards, C10S G␣ 11 performed more poorly than C9S G␣ 11 (Fig. 2).
To potentially overcome these deficits, fusion proteins were constructed between the ␣ 1b -adrenoreceptor and forms of G␣ 11 . The G protein sequence was attached directly to the C-terminal tail of the receptor cDNA from which the stop codon was eliminated. This allows production of single open reading frames containing the features of both polypeptides (24,25). Expression in EF88 cells of the chimeric polypeptide containing the wild type sequences of both receptor and G protein resulted in an elevation of [Ca 2ϩ ] i upon the addition of phenylephrine (Fig.  3A), although the kinetics of the response were markedly slower than for the isolated but co-transfected receptor and G protein. Now the same was true when a fusion protein between the ␣ 1b -adrenoreceptor and C9S,C10S G␣ 11 was used (Fig. 3B). This was also the case when a C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion protein was expressed (data not shown). Elevation of [Ca 2ϩ ] i in response to activation of a GPCR can proceed from activation of members of the phosphoinositidase C family by either ␣ subunits of the G q /G 11 family or ␤⅐␥ complexes (Fig. 4A). To ascertain if the signal from the fusion proteins derived from the receptor-attached G protein ␣ subunit, EF88 cells were co-transfected with the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein and transducin ␣. Transducin ␣ is used regularly as a ␤⅐␥-sequestering agent, and in this situation the effect of phenylephrine was fully attenuated (Fig. 3A). The N-terminal region of G␣ subunits is an important binding interface for ␤⅐␥ (30 -31). However, transducin ␣ also fully attenuated the phenylephrine signal from the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion protein (Fig. 3B).
Elevation of [Ca 2ϩ ] i could also potentially arise from ␤⅐␥ complex released by interaction of the fusion proteins with members of the pertussis toxin-sensitive G i family (Fig. 4A), which, unlike G␣ q and G␣ 11 , are expressed in EF88 cells (Fig.  4B). Expression of receptors in heterologous systems can result in a reduction in specificity of G protein coupling (32). To eliminate this possibility, experiments were repeated following sustained treatment with pertussis toxin of EF88 cells that had been transfected to express the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein. This did not alter the phenylephrine-mediated elevation of [Ca 2ϩ ] i (Fig. 4C). The combined data of Figs. 3 and 4 demonstrate that the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein both binds endogenous ␤⅐␥ and is able to release it upon agonist occupancy and that the palmitoylation status and potential of G␣ 11 does not limit either its binding or release of ␤⅐␥ complex.
To directly explore palmitoylation of the fusion proteins and its regulation, ␣ 1b -adrenoreceptor-G␣ 11 was expressed transiently in HEK293 cells. Both these and mock-transfected cells were labeled with [ 3 H]palmitate for 2 h, and the samples were immunoprecipitated with an antiserum (CQ) that identifies the C-terminal 10 amino acids shared by G␣ q and G␣ 11 (27,28). Following SDS-polyacrylamide gel electrophoresis and autoradiography, a band of some 42 kDa was observed in both mocktransfected and positively transfected cells (Fig. 5). This corresponds to a mixture of G␣ q and G␣ 11 , which are co-expressed by HEK293 cells and not resolved by the gel conditions employed. A 100-kDa [ 3 H]palmitoylated polypeptide corresponding to the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein was also observed but only in the positively transfected cells (Fig. 5). When the time course of [ 3 H]palmitoylation of the fusion protein was monitored in the presence and absence of phenylephrine, the rate, but not the maximal extent, of [ 3 H]palmitoylation was markedly enhanced by the agonist (Fig. 6). This effect was specific for agonist, since the presence of phentolamine, an antagonist/ inverse agonist at the ␣ 1b -adrenoreceptor, did not alter the kinetics or extent of palmitoylation (data not shown). Endogenous G␣ q /G␣ 11 was also [ 3 H]palmitoylated in a time-dependent manner. However, there was no indication that this was regulated significantly by agonist occupation of the fusion protein (Fig. 6), consistent with the notion that the fusion protein does not interact to a significant extent with the endogenous G protein pool. Both the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion protein and the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion protein were also targets for [ 3 H]palmitoylation (Fig. 7,  A and B). Each of the three fusion proteins expressed equally well as measured by the specific binding of the high affinity ␣ 1 -adrenoreceptor ligand [ 3 H]prazosin (see below). Thus, the lower incorporation of [ 3 H]palmitate into ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 and C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 (Fig. 7, A and B) reflects the reduced number of sites for acylation in these fusion proteins, which contain the acylationresistant, mutated G protein and the acylation-resistant receptor, respectively. Furthermore, since the labeling of ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 must represent fatty acylation of the receptor, the acylation status of the ␣ 1b -adrenoreceptor is clearly regulated by the presence of agonist (Fig. 7, A and B). Equally, palmitoylation of C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 must represent labeling of the G protein. This was also regulated by agonist (Fig. 7, A and B) in a concentration-dependent manner with EC 50 ϭ 7.8 ϫ 10 Ϫ7 M (Fig. 7C).
To further explore the functionality of the ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins, the capacity of phenylephrine to stimulate binding of [ 35 S]GTP␥S was measured. Traditionally, it is difficult to observe significant elevation of [ 35 S]GTP␥S for receptors that couple to members of the G q /G 11 family (33). This  1-10) or mock-transfected (lanes 11 and 12) and exposed to [ 3 H]palmitate for 5 (lanes 1 and 2), 15 (lanes 3 and 4), 30 (lanes 5 and 6), 60 (lanes 7 and 8), or 120 min (lanes 9 -12) in the absence (Ϫ) or presence (ϩ) of phenylephrine (PE; 10 M). A, samples were immunoprecipitated and processed as in Fig. 5. B, the autoradiograph to A was scanned, and the intensity of labeling of the ␣ 1b -adrenoreceptor-G␣ 11  indeed was the case when experiments were performed on membranes expressing ␣ 1b -adrenoreceptor-G␣ 11 , since basal levels of [ 35 S]GTP␥S binding were high (data not shown). However, by immunoprecipitating the fusion protein from the membranes with antiserum CQ following the [ 35 S]GTP␥S binding assay, the background was reduced to the extent that a greater than 30-fold stimulation of [ 35 S]GTP␥S binding by phenylephrine was observed (Fig. 8A). Phenylephrine was without effect in mock-transfected cells (Fig. 8A), and the effect of phenylephrine on [ 35 S]GTP␥S binding was suppressed completely when experiments were performed in the presence of a high concentration of unlabeled GTP␥S (Fig. 8B). The ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion and the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion were also able to produce effective stimulation of [ 35 S]GTP␥S binding upon the addition of phenylephrine, and the potency of the agonist (EC 50 ϭ 1.2 ϫ 10 Ϫ6 M) was unaffected by the palmitoylation potential of either the receptor or the G protein (Fig. 8C). Prazosin is a high affinity antagonist/inverse agonist at the hamster ␣ 1b -adrenoreceptor (34). [ 3 H]Prazosin bound to each of the ␣ 1b -adrenoreceptor-G␣ 11 , C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 , and ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion proteins with high affinity (Fig. 9A). Phenylephrine was able to compete with [ 3 H]prazosin for binding to these constructs (Fig. 9B) with similar estimated affinity (K i of ␣ 1badrenoreceptor-G␣ 11 ϭ 7.2 ϫ 10 Ϫ6 M; K i of C365A,C367G ␣ 1badrenoreceptor-G␣ 11 ϭ 9.1 ϫ 10 Ϫ6 M; and K i of ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 ϭ 1.1 ϫ 10 Ϫ5 M).

DISCUSSION
Both GPCRs and G protein ␣ subunits are targets for posttranslational palmitoylation (14 -16). This can be a regulated process and agonist occupation of a GPCR may enhance the rate of turnover of [ 3 H]palmitate labeling in both the GPCR and the G protein activated by the GPCR (35, 36). There are, however, a variety of technical limitations inherent in studies that attempt to examine GPCR-mediated regulation of G protein palmitoylation. Significant among them are lability of the thioester link between the fatty acid and the protein, the length of time usually required to produce autoradiograms of appropriate intensity when using [ 3 H]palmitate as the label and the FIG. 7. Agonist-regulation of palmitoylation of the ␣ 1b -adrenoreceptor and G␣ 11 . A, the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein (lanes 1 and 2), the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion protein (lanes 3 and 4), or the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion protein (lanes 5 and 6) were expressed transiently in HEK293 cells. Experiments akin to those of Fig. 6 were then performed in the absence (lanes 1, 3, and 5) or presence (lanes 2, 4, and 6)  possibility that observed effects predominantly reflect alterations in the specific activity of the palmitoyl CoA pool (37,38). A further issue is that G proteins are often present in marked excess compared with a GPCR in cells (39,40). Thus, if only a small fraction of the total pool of a G protein is activated by a GPCR, it may be difficult to detect agonist-mediated regulation of palmitoylation of this fraction. One method to overcome the latter issue is to ensure that only G proteins activated by the GPCR are isolated and analyzed.
To do this for the pairing of the ␣ 1b -adrenoreceptor and G␣ 11 , a fusion protein was constructed in which the N terminus of G␣ 11 was linked to the C-terminal tail of the GPCR such that a single open reading frame was generated containing the sequences and functions of both partner proteins. Following expression of this fusion protein and labeling of cells with [ 3 H]palmitate in the presence or absence of the ␣ 1 -adrenoreceptor agonist phenylephrine, the fusion protein could be effectively immunoprecipitated and analyzed using an antiserum that identifies the C terminus of G␣ 11 . Since the receptor-G protein fusion protein has a predicted molecular mass of some 100 kDa, it is extremely well resolved following SDS-polyacrylamide gel electrophoresis from the endogenous G␣ q /G␣ 11 expressed by most cells. This approach allowed clear demonstration that agonist enhanced the kinetics of turnover of [ 3 H]palmitate of the fusion protein but not the maximal extent of labeling (Fig. 6). Although such results are consistent with reports on agonist effects on co-expressed ␤ 2 -adrenoreceptor and G s ␣ (17-19), they are very different from those reported for a ␤ 2 -adrenoreceptor-G s ␣ fusion protein (41), in which the palmitoylation of the fusion protein was greatly reduced in the presence of agonist ligands. The basis for this variation is unclear, particularly since Loisel et al. (41) have suggested that repalmitoylation of the depalmitoylated ␤ 2 -adrenoreceptor-G␣ s fusion protein is inhibited and that this might be related to the poor capacity of this construct to become desensitized. Interestingly, however, we have observed that agonist also greatly accelerates the kinetics of acylation of a 5-hydroxytryptamine 1A-receptor-G␣ o1 fusion protein, 2 consistent with observations of an enhanced rate of [ 3 H]palmitoylation of G␣ i when activated by a co-expressed 5-hydroxytryptamine 1A receptor (42).
When the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein was expressed in HEK293 cells, which express both G␣ 11 and G␣ q endogenously, these G proteins also became [ 3 H]palmitoylated in a time-dependent manner. However, this was not modified by agonist treatment, indicative that the endogenous G proteins were not accessed by the fusion protein to a significant extent. Further confirmation of functional activation of the fused G␣ 11 by the receptor was obtained in [ 35 S]GTP␥S binding studies. Such assays provide a direct monitor of receptor-mediated guanine nucleotide exchange on a G protein. It is traditionally difficult, however, to monitor such an effect of agonists for G q family G proteins because of the high background contributed by other G proteins (33). This was also the case in the current studies when such assays were performed on membrane preparations. However, immunoprecipitation of the fusion protein allowed stimulation to be observed easily. Although the low basal guanine nucleotide exchange rate of the G q family G proteins is often cited as a limitation in efforts to monitor agonist stimulation of [ 35 S]GTP␥S binding to them, in the immunoprecipitation approach employed herein, it is a distinct advantage, since there was very little radioactivity bound to the fusion protein in the absence of agonist. When equivalent studies were performed in EF88 cells transfected with the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein, immunoprecipitation also provided direct evidence that the agonist stimulation of [ 35 S]GTP␥S binding was on the G␣ 11 of the fusion protein (data not shown).
Phenylephrine stimulation of [ 35 S]GTP␥S binding was achieved with an EC 50 of 1.2 ϫ 10 Ϫ6 M (Fig. 8). This is some 6-fold more potent than the estimated K i for phenylephrine for the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein calculated from the ability of this agonist to compete with [ 3 H]prazosin for binding to the fusion protein (Fig. 9). Since the 1:1 stoichiometry of the protein partners of the fusion predicts that activation of the G protein would require occupancy of the receptor by agonist, the basis for this difference is currently uncertain. This may suggest that the GPCR is not restricted to stimulating only the G protein that is directly linked to the receptor, but further analysis will be required to explore this possibility. However, it should also be borne in mind that the conditions required for the two assays are not identical, and although we tried to perform the binding assays in buffer conditions similar to those FIG. 9. The affinity of phenylephrine binding to the ␣ 1b -adrenoreceptor is unaffected by the palmitoylation potential of G␣ 11 . A, the specific binding of a range of concentrations of [ 3 H]prazosin to the ␣ 1b -adrenoreceptor-wild type G␣ 11 (circles), the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 (squares), or the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 (triangles) fusion proteins expressed transiently in HEK293 cells was measured. Analysis of such saturation isotherms demonstrated that [ 3 H]prazosin bound with similar affinity to each. B, the binding of [ 3 H]prazosin (0.5 nM) to the ␣ 1b -adrenoreceptor-wild type G␣ 11 , the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 , or the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins was competed for by varying concentrations of phenylephrine. of the [ 35 S]GTP␥S binding assay, subtle variations may alter estimates of ligand affinity and potency in different assays. It was noticeable, however, that in both assays elimination of the acylation sites in either the receptor or G protein did not alter these parameters for phenylephrine (Figs. 8 and 9).
Direct evidence of the capacity of the wild type fusion protein to regulate downstream end points was obtained following its expression in EF88 cells. Since these cells are derived from a mouse embryo in which the genes for both G␣ 11 and G␣ q were inactivated, signal cannot proceed via endogenous forms of these G proteins. The addition of phenylephrine resulted in an elevation of intracellular [Ca 2ϩ ]. This was derived from intracellular Ca 2ϩ stores, since all such assays were performed in the absence of extracellular Ca 2ϩ . Co-expression of transducin ␣ with the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein fully blocked the phenylephrine-induced rise in [Ca 2ϩ ]. Since transducin ␣ is used routinely as a ␤⅐␥ complex-sequestering agent, these results demonstrate that signal was transduced via ␤⅐␥. Such results prove that the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein can both interact with and release ␤⅐␥ in response to agonist stimulation. Furthermore, since the co-expression of transducin ␣ also fully attenuated signal from the ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion protein, the lack of palmitoylation of the G protein ␣ subunit prevents neither binding nor agonistmediated release of ␤⅐␥ complex. Although previous studies have indicated that a fusion protein between the ␤ 2 -adrenoreceptor and G s ␣ can co-immunoprecipitate ␤⅐␥ complex (43), this is the first study that has convincingly shown agonist-mediated signaling produced by ␤⅐␥ derived from a GPCR-G protein fusion protein. In the single cell imaging studies, elevation of [Ca 2ϩ ] appeared to occur somewhat more rapidly for the fusion protein incorporating the wild type G protein compared with the one containing the palmitoylation-resistant form of G␣ 11 (Fig. 3). EF88 cells are difficult to transfect, and thus we used single cell Ca 2ϩ imaging in concert with co-expression of green fluorescent protein for these sets of studies because only a small fraction of the cells in any particular field were positively transfected. Thus, although we noted that overall levels of expression of the two forms of the fusion protein, monitored by binding of [ 3 H]prazosin, were similar following transfection of HEK293 cells for the [ 3 H]palmitoylation and [ 35 S]GTP␥S binding studies, we do not know the levels of expression of the constructs in the individual EF88 cells that were imaged. Since each copy of the construct can release one copy of the ␤⅐␥ complex, higher levels of the fusion proteins in a particular cell might be expected to result in more rapid kinetics of liberation of [Ca 2ϩ ] i . Overall, these time courses were not very different. It is also noteworthy that Ca 2ϩ elevation in cells expressing any of the forms of the fusion protein was significantly slower than following co-expression of the isolated receptor and G protein and, more importantly, that the kinetics of Ca 2ϩ elevation were considerably slower when the ␣ 1b -adrenoreceptor was co-expressed with either C9S G␣ 11 or C10S G␣ 11 compared with the wild type G protein (Fig. 2). Both of these mutants have been shown to partition between membrane and cytosol much less effectively than wild type G␣ 11 (20,21), and this thus restricts their relative membrane concentration compared with the receptor.
It is also of considerable interest that interactions with ␤⅐␥ are thought to be required for membrane targeting and palmitoylation of both G␣ q and G␣ s (26) and play key roles in the targeting of G␣ subunits to the correct membrane compartments with lipid acylation and then stabilizing membrane association (44,45). In the native state, there is a complex interplay between the palmitoylation of G protein ␣ subunits and interaction with the ␤⅐␥ complex. For example, mutants of G␣ q unable to bind ␤⅐␥ are largely soluble and as such are not palmitoylated (26), since this acylation appears to occur at the membrane. Tethering of the ␣ subunit to the membrane is potentially sufficient to allow palmitoylation, however, since the addition of an N-terminal myristoylation signal to the ␤/␥ interaction-defective mutants of G␣ q allowed their palmitoylation but did not restore ␤/␥ interaction (26). These two processes thus can clearly be resolved. Detailed review of this topic has recently been provided (46). In the current studies, we have used attachment to a receptor to direct trafficking of the palmitoylation-resistant form of G␣ 11 to the membrane and to demonstrate that ␤/␥ interaction does not require prior palmitoylation. This is consistent with the concept that ␤/␥ interaction can stabilize the localization of a nonacylated G protein at the membrane until it becomes palmitoylated, which then provides extra anchorage, and with older studies, which indicated that high level overexpression of ␤⅐␥ complex can assist with trafficking of acylation-resistant forms of G␣ to the membrane (see Ref. 46 for details). Previous studies on G␣ i1 and G␣ z , which like the other G i family G proteins are modified by both palmitoylation and myristoylation at their N terminus, have also shown that lack of palmitoylation does not impair GPCR-mediated release of ␤⅐␥ (47,48). However, palmitoylation of G␣ s has been reported to increase its affinity for ␤⅐␥ by some 5-fold (49). Others have used different strategies to tether a G protein at the membrane. These have included linking the G protein to a single transmembrane-spanning element of a GPCR (50). However, the current approach ensures proximity of the G protein to the GPCR and has been successfully applied previously for acylation-resistant forms of G␣ i1 (51).
The capacity of phenylephrine to stimulate the binding of [ 35 S]GTP␥S binding to fusion proteins in which the acylation sites in either the receptor or G protein were mutated indicates that Cys residues at positions 9 and 10, and thus potential palmitoylation, is not inherently required for information transfer from the ␣ 1b -adrenoreceptor to the G protein. Furthermore, since the binding affinity of phenylephrine was little affected by these mutations (Fig. 9), it is not surprising that the measured EC 50 for the agonist to stimulate [ 35 S]GTP␥S binding was not different between the various fusion proteins employed (Fig. 8C).
The vast majority of rhodopsin-like GPCRs have one or more Cys residues in the first 20 amino acids of the predicted Cterminal tail. In many cases, these have been shown directly to be sites for palmitoylation (36,(52)(53)(54)(55). Mutation of these residues has been reported to have effects ranging from alterations in G protein coupling (52)(53)(54) and receptor internalization/ membrane delivery (55,56) to effects on the phosphorylation of the GPCR by regulatory kinases (36,57) and interactions with ␤-arrestin-1 (58). The ␣ 1b -adrenoreceptor-C9S,C10S G␣ 11 fusion protein was also able to incorporate [ 3 H]palmitate. This therefore must represent acylation at Cys 365 and/or Cys 367 of the receptor. The kinetics of incorporation of [ 3 H]palmitate into this form of the fusion protein was also enhanced by treatment with phenylephrine, indicating that turnover of receptor palmitoylation is not a process that has to be integrated with equivalent regulation of the acylation status of the G protein (Fig. 7). Equally, agonist regulated palmitoylation of the C365A,C367G ␣ 1b -adrenoreceptor-G␣ 11 fusion protein and thus of the G protein activated by the receptor.
The current studies provide new insights into the coordinated agonist regulation of post-translational acylation of a GPCR and its associated G protein. Furthermore, they introduce a strategy to monitor the true extent of receptor-enhanced [ 35 S]GTP␥S binding to G q family G proteins. Finally, they also provide novel insights into the mechanism of [Ca 2ϩ ] signaling in cells lacking the ␣ subunits of the G proteins which are usually assumed to transduce these signals in cells.