Effective Information Transfer from the (cid:1) 1b -Adrenoceptor to G (cid:1) 11 Requires Both (cid:2) / (cid:3) Interactions and an Aromatic Group Four Amino Acids from the C Terminus of the G Protein*

Co-expression of the (cid:1) 1b -adrenoreceptor and G (cid:1) 11 in cells derived from a G (cid:1) q /G (cid:1) 11 knock-out mouse allows agonist-mediated elevation of intracellular Ca 2 (cid:4) levels that is trans-duced by (cid:2) / (cid:3) released from the G protein (cid:1) subunit. Mutation of Tyr 356 of G (cid:1) 11 to Phe, within a receptor contact domain, had little effect on function but this was reduced greatly by alteration to Ser and virtually eliminated by conversion to Asp. This pattern was replicated following incorporationofeachformofG (cid:1) 11 intofusionproteinswith the (cid:1) 1b -adrenoreceptor. Following a [ 35 S]guanosine 5 (cid:1) - O -(3-thiotriphosphate) (GTP (cid:3) S) binding assay, immunoprecipitation of the wild type (cid:1) 1b -adrenoreceptor-G (cid:1) 11 fusion protein indicated that the agonist phenylephrine stimulated guanine nucleotide exchange on G (cid:1) 11 more than 30-fold. Information transfer by agonist was controlled in residue 356 G (cid:1) 11 mutants with rank order Tyr > Phe > Trp > Ile

Information transfer between G protein-coupled receptors (GPCRs) 1 and G protein ␣ subunits involves the induced release of GDP from the nucleotide binding site of the G protein and its subsequent replacement by GTP (1). This is often studied by monitoring the binding of an analogue of GTP, [ 35 S]GTP␥S, that is resistant to the in built GTPase activity of the G protein ␣ subunit (2,3). Receptor-mediated activation of phospholipase C␤1 is mediated via members of the G q family of G proteins (4 -6). Resistance of these G proteins to ADP-ribosylation by bacterial toxins and their low rates of guanine nucleotide exchange made their initial purification and characterization an extremely difficult task (7)(8)(9). This second feature has also limited efforts to use the binding of [ 35 S]GTP␥S to monitor their activation. In contrast, the ease of use of this assay for members of the pertussis toxin-sensitive G i family of G proteins has resulted in its widespread application (2,3).
The pertussis toxin-sensitive G proteins have a conserved Cys located 4 amino acids from the C terminus that is the site of ADP-ribosylation by the toxin. As this modification prevents GPCR-mediated activation of these G proteins, it provided early compelling evidence of a key role for the extreme Cterminal region of G protein ␣ subunits in productive interactions with GPCRs (10). Recently, extensive mutagenesis at this site in both G␣ i1 (11) and G␣ i3 (12) has indicated that the Cys is not essential for information transfer and that the effectiveness of receptor-mediated activation is determined by the hydrophobicity of the residue at this position.
The widely expressed G q family G proteins, G␣ q and G␣ 11 , share an identical C-terminal decapeptide with Tyr 356 located 4 amino acids from the C terminus of G␣ 11 (4). Previous studies have suggested that this Tyr can become phosphorylated in response to GPCR activation and that this may be a key event in activation of the G protein (13). We have recently employed an immunoprecipitation strategy in concert with a [ 35 S]GTP␥S binding assay to monitor directly GPCR and agonist-induced guanine nucleotide exchange on G␣ 11 (14). Herein we extend this approach to analyze the effects on information transfer between the ␣ 1b -adrenoreceptor and forms of G␣ 11 in which Tyr 356 was altered to a range of amino acids. There is no inherent requirement for Tyr at this position and thus for its potential phosphorylation in the activation of the G protein. However, as Phe and Trp are the other amino acids that can substitute effectively a key role for a bulky aromatic group is evident.
Although the guanine nucleotide binding site and GTPase machinery are defined by the G protein ␣ subunit, the ␤/␥ complex plays a key role, with growing evidence for a direct role of the ␥ subunit in contacting the receptor to enhance guanine nucleotide exchange (15,16). By mutating a key ␤/␥ contact site in G␣ 11 we also provide support for this notion, because the modified ␣ subunit exchanges guanine nucleotide significantly less effectively in response to receptor activation that does the wild type. Overexpression of ␤/␥ with a fusion protein between the ␣ 1b -adrenoreceptor and the mutant form of G␣ 11 with reduced ␤/␥ binding affinity enhances the effectiveness of agonist-stimulated guanine nucleotide exchange.
G Protein Mutations and Construction of Fusion Proteins-Residue 356 of G␣ 11 (wild type is Tyr) was converted to a range of other amino acids. The Ile 25 3 Ala,Glu 26 3 Ala (IE) mutant was constructed based on the studies of Evanko et al. (25) on G␣ q . These modified forms of G␣ 11 were also incorporated into ␣ 1b -adrenoreceptor-G␣ 11 fusions proteins as described previously (14). 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 (Invitrogen) according to the manufacturer's instructions.
[Ca 2ϩ ] i Measurement; Cell Culture and Transfection-EF88 is a fibroblast cell line derived from the embryos of mice in which expression of the ␣ subunits of both G q and G 11 had been eliminated by targeted gene disruption (17)(18)(19). These 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. A portion of the cells harvested during trypsinization was plated on to glass coverslips, and after a 24-h growth period they were transfected using LipofectAMINE (Invitrogen) according to the manufacturers' instructions. After 3 h cells were washed twice with Opti-MEM I and then cultured in DMEM growth medium for a further 24 h. A total of 3 g of pCDNA3 containing the relevant cDNA species were used to transfect each coverslip.
[Ca 2ϩ ] i Imaging-Transfected cells were loaded with the Ca 2ϩ -sensitive dye Fura-2 by incubation (15-20 min, 37°C) under reduced light in DMEM growth medium containing the dye's membrane-permeant acetoxymethyl ester form (1.5 M). Loaded cells were illuminated with an ultra high point intensity 75-watt xenon arc lamp (Optosource, Cairn Research, Faversham, Kent, UK) and subsequently imaged using a Nikon Diaphot inverted microscope equipped with a Nikon 40ϫ oil immersion Fluor objective lens (NA ϭ 1.3) and a monochromator (Optoscan, Cairn Research), which was used to alternate the excitation wavelength between 340/380 nm and to control the excitation band pass (340 nm band pass ϭ 10 nm; 380 nm band pass ϭ 8 nm). Fura-2 fluorescence emission at 510 nm was monitored using a high resolution interline-transfer cooled digital CCD camera (Cool Snap-HQ, Roper Scientific/Photometrics, Tucson, AZ). MetaFluor imaging software (version 4.6.8, Universal Imaging Corp., Downing, PA) was used for control of the monochromator, CCD camera, and for processing of the cell image data. Sequential images (2 ϫ 2 binning) were collected every 2 s, exposure to excitation light was 100 ms/image, and all experiments were undertaken in the absence of extracellular Ca 2ϩ in saline solution comprising: 130 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 20 mM HEPES, 10 mM D-glucose, 0.01 mM EGTA, pH adjusted to 7.4 using NaOH.
[Ca 2ϩ ] i Image Analysis-Ratio images were presented in MetaFluor intensity-modulated display mode (26), which associates the color hue with the excitation ratio value and the intensity of each hue with the source image brightness. Briefly background subtracted images acquired at 340 and 380 nm excitation were first used for calculating the 340/380 nm ratio of each pixel. After determination of the upper and lower thresholds, the ratio value of each pixel was associated with one of the 24 hues from blue (low [Ca 2ϩ ] i ) to red (high [Ca 2ϩ ] i ). Pooled average intensity-modulated display ratio intensity values measured from single cells were expressed as the mean Ϯ S.E. mean of at least 10 cells with the vertical lines representing S.E. of mean.
[ 35  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 dodecyl sulfate. Samples were precleared with Pansorbin (Calbiochem), followed by immunoprecipitation with CQ antiserum (22,23) or an antiserum against the N-terminal section of the ␣ 1b -adrenoreceptor. Finally, the immunocomplexes were washed twice with solubilization buffer and bound [ 35 S]GTP␥S 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 (0.05-5 nM in saturation assays and 1.0 nM for competition assays), in the absence or presence of increasing concentrations of phenylephrine. Nonspecific binding was determined in the presence of 100 M phentolamine. Reactions were incubated for 30 min at 30°C and bound ligand 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.
Immunoprecipitation and Immunodetection Studies-For immunoprecipitations cells were washed once with ice-cold phosphate-buffered saline and immediately homogenized in a lysis medium, containing 50 mM Hepes (pH 7.4), 10 mM Na 4 P 2 O 7 , 100 mM NaF, 10 mM EDTA, 0.2 mM Na 3 VO 4 , 1% Triton X-100, and a protease inhibitor mixture (Complete, Roche Molecular Biochemicals). Cell lysates were centrifuged (15 min, 13,000 rpm) and the supernatants precleared for 1 h with nonspecific serum and protein A. Next, samples were incubated overnight with the appropriate antiserum. The immunocomplexes were then captured with protein A-agarose.
For immmunoblotting, cell lysates or immunoprecipitates were subjected to SDS-polyacrylamide gel electrophoresis. Proteins were transferred to polyvinylidene difluoride membranes and blocked for 2 h with 5% nonfat dried milk in 0.05% Tween 20/Tris buffered saline (TTBS). Then, the polyvinylidene difluoride membranes were probed overnight at 4°C with the appropriate antiserum and washed with TTBS. Membranes were incubated for 20 min with horseradish peroxidase conjugated to anti-rabbit IgG (1:20,000) (Amersham Biosciences). Finally, membranes were washed with TTBS and developed by enhanced chemiluminescence.

RESULTS
Co-expression of the ␣ 1b -adrenoreceptor and G␣ 11 is required to elevate intracellular [Ca 2ϩ ] in response to the ␣ 1 -adrenoreceptor agonist phenylephrine in cells (EF88) derived from a combined G␣ q and G␣ 11 knock-out mouse ( Fig. 1) (14). Because this effect is blocked by co-expression of ␣-transducin ( Fig. 1), which is an effective ␤/␥ sequestering agent, the response to agonist reflects activation of the G protein, the release of ␤/␥ complex from the ␣ subunit, and regulation of a ␤/␥-sensitive isoform of phospholipase C␤. Co-expression of the ␣ 1 -adrenoreceptor with Tyr 356 3 Asp G␣ 11 did not result in a significant elevation of intracellular Ca 2ϩ (Fig. 1). These results indicate an important role for this residue in the G protein. When the ␣ 1b -adrenoreceptor was co-expressed with Tyr 356 3 Phe G␣ 11 the response to phenylephrine was only slightly lower and more delayed kinetically than with wild type G␣ 11 (Fig. 1). Such results negate the suggestion that phosphorylation of Tyr 356 G␣ 11 is required for its activation (13). Tyr 356 3 Ser G␣ 11 was also capable of being activated by the ␣ 1b -adrenoreceptor, but response was substantially lower and kinetically substantially slower than with wild type or Tyr 356 3 Phe G␣ 11 ( Fig. 1).
We have recently demonstrated that a fusion protein in which the N terminus of wild type G␣ 11 is linked to the Cterminal tail of the ␣ 1b -adrenoreceptor from which the stop codon was eliminated also releases ␤/␥ complex and elevates Ca 2ϩ levels in EF88 cells upon agonist stimulation (14). Fusion proteins were thus also constructed between the ␣ 1b -adrenoreceptor and each of the Tyr 356 3 Phe G␣ 11 , Tyr 356 3 Ser G␣ 11 , and Tyr 356 3 Asp G␣ 11 G proteins. Each of these fusion proteins was expressed in EF88 cells and the ability of phenylephrine to elevate intracellular Ca 2ϩ measured. These studies recapitulated the co-expression data. Virtually no signal was generated from the fusion containing Tyr 356 3 Asp G␣ 11 , a poor signal was obtained from the Tyr 356 3 Ser G␣ 11 containing version, and similar and significantly greater signals were obtained from the fusion proteins containing the Tyr 356 3 Phe or wild type G␣ 11 (Fig. 2). Again, the signals obtained with the Phe and Ser mutants were delayed kinetically compared with those from the construct containing the wild type G protein (Fig. 2). As reported previously (14) co-expression of ␣-transducin fully attenuated the phenylephrine-mediated elevation of intracellular Ca 2ϩ produced by ␣ 1b -adrenoreceptor-G␣ 11 , and this was also the case for the fusion proteins containing both Tyr 356 3 Phe G␣ 11 and Tyr 356 3 Ser G␣ 11 (data not shown).
As EF88 cells are difficult to transfect efficiently, many of the subsequent studies were performed in HEK293 cells. The transfected ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins were detected in lysates of HEK293 cells as distinct doublets that probably reflect differential glycosylation of the receptor element of the polypeptides (Fig. 3A). Each of the ␣ 1b -adrenoreceptor-G␣ 11 fusion constructs containing the mutated forms of the G protein expressed as well as the fusion containing the wild type G protein as assessed by immunoblotting HEK293 cell lysates with an antiserum that identifies the extreme N terminus of the ␣ 1b -adrenoreceptor (Fig. 3A, top). This was apparently not true when the same lysates were immunoblotted with an antiserum (CQ) that identifies the C-terminal decapeptide of G␣ 11 . Only the construct containing the wild type G protein was identified effectively (Fig. 3A, bottom). Unlike EF88 cells, HEK293 cells express both G␣ q and G␣ 11 (14). Thus, although antiserum CQ was unable to identify the fusion proteins containing the G protein mutants, it did identify the G␣ q /G␣ 11 endogenously expressed in each sample (Fig.  3A, bottom). Furthermore, ligand binding studies monitoring the specific binding of the ␣ 1 -adrenoreceptor antagonist [ 3 H]prazosin confirmed that each fusion protein was expressed to similar levels (Fig. 3B). As the alteration in the G protein sequence was within the epitope used as antigen, it is thus likely that Tyr 356 is a immunodominant element for identification of G␣ 11 by antiserum CQ (see also later under "Results"). HEK293 membranes expressing 100 fmol of the ␣ 1b -adrenoreceptor-wild type G␣ 11 fusion protein, as measured by the specific binding of [ 3 H]prazosin, were subjected to a [ 35 S]GTP␥S binding assay followed by immunoprecipitation with the anti-G␣ q /G␣ 11 C-terminal antiserum. Very little [ 35 S]GTP␥S was present in the immunoprecipitate (Fig. 4). However, addition of phenylephrine during the assay increased the levels of [ 35 S]GTP␥S present in the immunoprecipitate in a concentration-dependent manner. A maximally effective concentration of phenylephrine increased the level of bound [ 35 S]GTP␥S nearly 30-fold with EC 50 for phenylephrine of 3.3 ϫ 10 Ϫ7 M (Fig. 4A). Equivalent results were obtained when the experiments were repeated but immunoprecipitation performed with the antibody directed toward the N terminus of the ␣ 1b -adrenoreceptor. The EC 50 for phenylephrine was now 6.0 ϫ 10 Ϫ7 M, and although a lower number of counts of [ 35 S]GTP␥S were present in these immunoprecipitates (Fig. 4B), this simply reflects the immunoprecipitation efficiencies achieved with the amounts of the two antibodies employed (data not shown).
Further fusion proteins were constructed between the ␣ 1badrenoreceptor and forms of G␣ 11 with each of the Trp, Ile, Ala, Gln, and Arg residues at position 356 of the G protein. These also were identified poorly by antiserum CQ but were expressed at similar levels as the other fusion proteins as monitored by saturation binding assays using [ 3 H]prazosin. Membrane amounts corresponding to 100 fmol of the ␣ 1badrenoreceptor-G␣ 11 fusion proteins containing Ser or Trp as residue 356 of the G protein were then used in [ 35 S]GTP␥S binding assays employing a range of concentrations of phenyl-ephrine. Samples were immunoprecipitated with the anti-␣ 1badrenoreceptor antibody and counted (Fig. 4B). Although the effectiveness of phenylephrine to stimulate nucleotide binding to the ␣ 1b -adrenoreceptor-G␣ 11 fusion proteins containing Trp 356 3 G␣ 11 and particularly Ser 356 3 G␣ 11 was substantially lower than for the wild type (Fig. 4B), this did not reflect a reduction in potency of the agonist. There was less than a 2-fold difference in the EC 50 for phenylephrine between the three constructs (Fig. 4B). Each of the nine fusion proteins were then expressed in HEK293 cells, and following saturation [ 3 H]prazosin binding assays to determine expression levels, membrane amounts containing 100 fmol of each construct were used to measure basal and phenylephrine-stimulated [ 35 S]GTP␥S binding. Samples were subsequently immunoprecipitated with the anti-␣ 1b -adrenoreceptor antibody and counted (Fig. 5A). A clear pattern emerged with the profile of [ 35 S]GTP␥S binding being that Tyr Ͼ Phe Ͼ Trp Ͼ Ile Ͼ Ala,Gln,Arg Ͼ Ser Ͼ Asp. Most effective guanine nucleotide exchange was obtained with amino acids containing an aromatic ring in the side chain. This did not reflect differences in the affinity of phenylephrine to bind to the individual constructs (Fig. 5, B and C). However, results were not the same when such [ 35 S]GTP␥S binding assays were subjected to immunoprecipitation with the anti-G␣ q /G␣ 11 antiserum CQ. Most of the constructs apparently functioned very poorly compared with the wild type sequence (data not shown). However, as noted earlier, this reflects that the various fusion proteins are not immunoprecipitated or immunodetected equally by this antiserum (see Fig. 3B).
As Ca 2ϩ elevation in EF88 cells is dependent upon ␤/␥ release, we generated a form of the ␣ 1b -adrenoreceptor-G␣ 11 fusion protein by mutation of Gly 208 of the G protein to Ala. This was anticipated to be unable to exchange guanine nucleotide and thus should not be able to release ␤/␥ in an agonist-dependent manner. We also mutated residues in G␣ 11 (Ile 25 3 Ala,Glu 26 3 Ala) in a region shown previously (25) to be a key ␤/␥ contact region for G␣ q . Following co-transfection of either the wild type ␣ 1b -adrenoreceptor-G␣ 11 fusion protein or ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 with a combination of the G protein ␤ 1 and ␥ 2 subunits cell lysates were immunoprecipitated with the anti-G␣ q /G␣ 11 antiserum. These samples were resolved by SDS-PAGE and immunoblotted to detect the ␤ 1 subunit (Fig. 6 top). This was effectively coimmunoprecipitated with the wild type ␣ 1b -adrenoreceptor-G␣ 11 fusion protein, but it was hardly detectable following immunoprecipitation of the ␣ 1b -adrenoreceptor-(Ile 25 3 Ala, Glu 26 3 Ala) G␣ 11 fusion protein (Fig. 6, top). This was not a reflection of poor expression of the ␤ 1 subunit with the ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 fusion protein as direct immunoblots of the cell lysates demonstrated equal levels of this subunit in the two samples (Fig. 6, middle). The ␣ 1b -adrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 fusion protein was also expressed as effectively as the fusion protein containing the wild type G protein (Fig. 6, bottom). Neither the ␣ 1b -adrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 nor the ␣ 1b -adrenoreceptor-(Gly 208 3 Ala) G␣ 11 fusion protein caused effective elevation of intracellular [Ca 2ϩ ] in response to phenylephrine following expression in EF88 cells (Fig. 7).
We have previously demonstrated that the ␣ 1b -adrenoreceptor-(Gly 208 3 Ala) G␣ 11 fusion protein is unable to bind [ 35 S]GTP␥S in response to phenylephrine (39). Following expression in HEK293 cells and [ 3 H]prazosin binding studies, 100 fmol of both wild type and the ␣ 1b -adrenoreceptor-(Ile 25 3 Ala,Glu 26  was used, as the C-terminal region of the G protein is not different between these two constructs, and as noted earlier, this antiserum is the more efficient in immunoprecipitation if the C-terminal of the G protein is wild type. Although ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 clearly bound [ 35 S]GTP␥S in response to a maximally effective concentration of phenylephrine (10 M), it was to a markedly lower level than the wild type fusion protein (Fig. 8A). To explore the ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 construct fur-ther, the time course of phenylephrine-stimulated [ 35 S]GTP␥S binding was compared with the ␣ 1b -adrenoreceptor-G␣ 11 fusion. Phenylephrine-stimulated [ 35 S]GTP␥S binding to the ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 fusion protein was substantially lower at all time points measured (Fig. 8B). Co-expression of ␣ 1b -adrenoreceptor-G␣ 11 with ␤ 1 and ␥ 2 subunits did not significantly increase phenylephrine-mediated binding of [ 35 S]GTP␥S to the wild type fusion protein (Fig. 8, A  and B). By contrast, co-expression of ␤ 1 and ␥ 2 with the ␣ 1badrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 fusion protein did increase agonist-stimulated [ 35 S]GTP␥S binding to this fusion protein, although this remained lower than for the construct containing the wild type G protein (Fig. 8, A and B). It was not possible to assess whether co-expression of ␤ 1 and ␥ 2 subunits with the ␣ 1b -adrenoreceptor-(Ile 25 3 Ala,Glu 26 3 Ala) G␣ 11 fusion protein would allow agonist-mediated elevation of Ca 2ϩ levels in EF88 cells. Expression of ␤ 1 /␥ 2 with or without the fusion protein resulted in spontaneous elevation of [Ca 2ϩ ] i (data not shown). This presumably reflects that significant amounts of the introduced ␤ 1 /␥ 2 did not become associated with G protein ␣ subunits and thus were active signal transducing complexes in the absence of receptor stimulation. Despite these differences ␣ 1b -adrenoreceptor-(Ile 25 3 Ala, Glu 26 3 Ala) G␣ 11 bound both phenylephrine (Fig. 8C) and [ 3 H]prazosin (Fig. 8D) with similar affinities as the construct containing the wild type G protein, and these characteristics were not altered significantly by co-expression of ␤ 1 /␥ 2 (Fig. 8,  C and D).  FIG. 7. Fusion proteins that fail to bind or release ␤ 1 /␥ 2 complex do not elevate [Ca 2؉ ] i effectively in EF88 cells. EF88 cells were transfected to express the wild type ␣ 1b -adrenoreceptor-G␣ 11 fusion protein (black) or variants of this (␣ 1b -adrenoreceptor-Ile 25 3 Ala,-Gln 26 3 Ala G␣ 11) (green) that fail to bind ␤ 1 /␥ 2 efficiently or (␣ 1badrenoreceptor-Gly 208 3 Ala G␣ 11 ) that fails to release it (red). The effects of phenylephrine (3 M) on [Ca 2ϩ ] i were then imaged as in Fig.  2. Data represent means Ϯ S.E. from analysis of 27 (wild type G␣ 11 ), 10 (Gly 208 3 Ala G␣ 11 ), and 19 (Ile 25 3 Ala,Gln 26 3 Ala G␣ 11 ) positively transfected cells.

DISCUSSION
The extreme C terminus of G protein ␣ subunits provides a key contact interface for GPCRs, and thus modification of this region is expected to alter either the effectiveness of information transfer or receptor selectivity (27,28). This knowledge has been applied to the production of chimeric G proteins in which alteration of as few as the last 3 amino acids of G␣ q to the sequence of G␣ i has been shown to alter the classes of receptors able to activate the G protein (29 -32). Such chimeric G proteins are widely used in drug discovery programs as they can allow a single, easy to measure, end-point assay to be used to infer activation of different classes of GPCRs (31,32). Detailed studies of the effectiveness of interactions between GPCRs and G proteins of the pertussis toxin-sensitive G i class have often taken advantage of the relatively high rate of both basal and agonist-stimulated guanine nucleotide exchange of these G proteins. By contrast, efforts to employ this approach for other G protein classes are generally limited by the poor signal to noise that is obtained. By immunoprecipitation of a fusion protein between the ␣ 1b -adrenoreceptor and G␣ 11 at termination of a [ 35 S]GTP␥S binding assay, we have shown that ago-nists produce a large increase in binding of this nucleotide and that in this format the extremely low basal nucleotide exchange on G␣ 11 is a distinct advantage (14).
Previous studies have suggested that the presence of the Tyr 4 amino acids from the C terminus of G␣ q and G␣ 11 is integral to activation and indeed that this amino acid may well be a target for tyrosine phosphorylation (13). Using either co-transfection of a receptor and mutants of this site in G␣ 11 or by constructing a series of fusion proteins between the ␣ 1b -adrenoreceptor and variants of G␣ 11 modified at this position, we now show that this is not the case. Although Tyr is indeed the most effective amino acid at that location in allowing activation of the G protein, either Phe or Trp are well tolerated (Fig. 5). As no other amino acid we tested was more than 40% as effective as Tyr, there is, however, a clear requirement for a bulky and aromatic structure at this position. Amino acids with small side chains, such as Ser, or either positive or negative charge were poorly tolerated (Fig. 5). In initial studies we were able to immunoprecipitate the wild type ␣ 1b -adrenoreceptor-G␣ 11 fusion protein with antibodies directed against either the Nterminal region of the receptor or the C-terminal decapeptide of the G protein. These produced equivalent data for the EC 50 of phenylephrine as agonist at the construct (Fig. 4). However, when we repeated such experiments using the range of fusion proteins incorporating alterations in Tyr 356 of the G protein, it was clear that the anti-G protein antibody was not equally effective in identifying each fusion protein and thus could not be used to compare the effectiveness of agonist-induced [ 35 S]GTP␥S binding. This antibody was generated against the G protein C-terminal peptide bound to a carrier protein via an N-terminal Cys. In such approaches bulky amino acids close to the C-terminal of the peptide sequence used as antigen are likely to be immunodominant and contribute significantly to the interaction between antibody and antigen. Indeed, we have noted previously that ADP-ribosylated G␣ i is a higher affinity target for an antibody against the C-terminal decapeptide of this G protein than the unmodified sequence even though the native sequence was the antigen (33). As the N terminus of each fusion protein used herein is identical, we used the antireceptor antibody for the immunoprecipitation studies that were designed to determine the effect of Tyr 356 mutation on signal transduction, even though the immunoprecipitation efficiency of this antiserum was lower than for the anti-G protein antiserum when this was assessed in parallel on the wild type fusion construct (Fig. 4). The anti-G␣ q /G␣ 11 antiserum used herein, and all those widely available from commercial sources, are directed against the C-terminal tail of these polypeptides. It would not thus have been possible to quantitate the effect of mutants in this region of G␣ 11 on information transfer from a receptor without constructing the receptor-G protein fusions used herein, because linking the G protein to the receptor was the best practical means to immunoprecipitate, and hence enrich, the mutant forms of G␣ 11 at the termination of [ 35 S]GTP␥S binding assays. Although immunoprecipitation of G␣ q /G␣ 11 with C-terminal anti-G protein antisera has been used previously to enrich [ 35 S]GTP␥S binding assays (34 -36), we show herein that this is only practical in receptor and G protein co-transfection experiments that do not analyze such mutants of the G proteins. The major limitation in the current studies was that many of the mutations we wished to study were within the epitope identified by the antiserum. However, mutations also frequently alter the expression levels of a polypeptide. The fusion protein approach also overcame this concern as we measured levels of expression of each construct in ligand binding assays and thus were able to add the same amount of each distinct construct to the [ 35 S]GTP␥S binding studies. G proteins are activated by the exchange of GTP for GDP (1). We thus also tested directly downstream signal transduction via these fusion proteins. EF88 cells are derived from a G␣ q / G␣ 11 double knock-out mouse and thus require expression of both an appropriate GPCR and a signal transduction-competent G protein to elevate intracellular Ca 2ϩ in response to agonist (14,18). We have previously shown that a fusion protein between the ␣ 1b -adrenoreceptor and wild type G␣ 11 is functional (14). The rank order of capacity of fusions with alterations at Tyr 356 of the G protein to elevate Ca 2ϩ in response to phenylephrine was the same as observed for the agonist-stimulated binding of [ 35 S]GTP␥S. We also noted that mutant forms of G␣ 11 that coupled less effectively to the receptor as measured in the [ 35 S]GTP␥S binding assays displayed slower kinetics of elevation of [Ca 2ϩ ] in both co-transfection and fusion protein expression studies (Figs. 1 and 2). This is likely a reflection of the reduced effectiveness of information transfer and thus activation of these mutants as they were expressed equally well as the wild type G protein (Fig. 3). A key point about agonist-mediated elevation of Ca 2ϩ in EF88 cells is that it is blocked by co-expression of ␣-transducin. As such, it is a measure of ␤/␥ release and function (14). It was thus anticipated that incorporation of a ␤/␥ release-deficient G protein into the fusion protein would result in a construct unable to signal. Mutation of a conserved Gly in the G3 nucleotide binding domain of G protein ␣ subunits to Ala is known to prevent ␤/␥ release from the ␣ subunit, although this alteration has little obvious effect on G protein structure (37,38). Mutations centered at Ile 25 and Glu 26 of G␣ q have previously been inferred to lack the capacity to bind ␤/␥ effectively (25). Fusion proteins between the ␣ 1b -adrenoreceptor and both wild type and Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 were co-expressed with ␤ 1 and ␥ 2 subunits, and following immunoprecipitation of the fusion proteins, co-immunoprecipitation of the ␤ 1 subunit was assessed. This polypeptide was shown to be associated with the wild type G protein containing fusion, but only small amounts of ␤ 1 were present along with the Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 -containing fusion (Fig. 6). Such results confirm the poor ability of Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 to interact with the ␤ 1 /␥ 2 complex. The Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 containing fusion was only able to elevate Ca 2ϩ very poorly in EF88 cells in response to phenylephrine, and even this was not observed in all transfected cells. Although this was also the case for the fusion containing Gly 208 3 Ala G␣ 11 , in this case it rather reflects that this construct was also unable to bind [ 35 S]GTP␥S (39) and thus adopt the conformation required to dissociate the G protein subunits.
As ␤/␥ complex is also key for agonist-mediated guanine nucleotide exchange on the ␣ subunit (16), we tested this quantitatively. Co-expression of the ␤ 1 /␥ 2 complex with the fusion protein containing the wild type protein did not result in higher levels of phenylephrine-stimulated [ 35 S]GTP␥S binding when the same number of fusion protein receptor binding sites were added to the assay. This is likely to reflect good interactions between the transfected fusion protein and endogenously expressed ␤/␥ complexes. Indeed, phenylephrine-mediated elevation of [Ca 2ϩ ] in EF88 cells (Fig. 2) could not occur via the wild type fusion protein without this interaction. However, although the fusion protein containing Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 bound significantly less [ 35 S]GTP␥S than the wild type, phenylephrine-stimulated binding of the nucleotide was increased to this fusion protein in the presence of co-expressed ␤ 1 /␥ 2 . This is likely to reflect that Ile 25 3 Ala,Glu 26 3 Ala G␣ 11 has a significantly lower affinity to bind ␤ 1 /␥ 2 than wild type G␣ 11 (and thus was poor in producing co-immunoprecipitation) rather than being entirely lacking in this regard.
These studies expand the recent use (14) of a receptor-G protein fusion protein to overcome the traditional difficulties in monitoring directly guanine nucleotide exchange on G q family G proteins to allow analysis of the importance of both ␤/␥ interactions and the role of Tyr 356 in effective information transfer between a receptor and G␣ 11 . This basic approach should be equally amenable to any sets of mutations in receptors and G proteins from these families.