Activation of Gsalpha by the epidermal growth factor receptor involves phosphorylation.

Previous studies from our laboratory have shown that epidermal growth factor (EGF) stimulates cAMP accumulation in the heart via a process involving Gsalpha and the EGF receptor (EGFR) protein tyrosine kinase activity (Nair, B. G., Parikh, B., Milligan, G., and Patel, T. B. (1990) J. Biol. Chem. 265, 21317-21322; Nair, B. G., and Patel, T. B. (1993) Biochem. Pharmacol. 46, 1239-1245). Therefore, studies were performed to investigate the hypothesis that the EGFR protein tyrosine kinase phosphorylates Gsalpha and activates this protein. Employing purified EGFR and Gsalpha, we have demonstrated that the EGFR kinase phosphorylates Gsalpha in a time-dependent manner with a stoichiometry of 2 mol of phosphate incorporated/mol of Gsalpha. As determined by phosphoamino acid analysis, the phosphorylation of Gsalpha by the EGFR kinase was exclusively on tyrosine residues. Interestingly, GDP and guanosine 5'-3-O-(thio)triphosphate (GTPgammaS) inhibited the phosphorylation of Gsalpha without altering EGFR autophosphorylation. However, G protein betagamma subunits protected against GDP- and GTPgammaS-mediated inhibition of phosphorylation of Gsalpha. In functional studies, phospho-Gsalpha demonstrated a greater GTPase activity and also a greater capacity to bind GTPgammaS as compared to the nonphosphorylated Gsalpha. Moreover, the phospho-Gsalpha augmented adenylyl cyclase activity in S49 cyc- cell membranes to a greater extent than its nonphosphorylated counterpart. Therefore, we conclude that phosphorylation of Gsalpha on tyrosine residues by the EGFR kinase activates this G protein and increases its ability to stimulate adenylyl cyclase.

Previous studies from our laboratory have shown that epidermal growth factor (EGF) stimulates cAMP accumulation in the heart via a process involving G s␣ and the EGF receptor (EGFR) protein tyrosine kinase activity ( Pharmacol. 46, 1239 -1245). Therefore, studies were performed to investigate the hypothesis that the EGFR protein tyrosine kinase phosphorylates G s␣ and activates this protein. Employing purified EGFR and G s␣ , we have demonstrated that the EGFR kinase phosphorylates G s␣ in a time-dependent manner with a stoichiometry of 2 mol of phosphate incorporated/mol of G s␣ . As determined by phosphoamino acid analysis, the phosphorylation of G s␣ by the EGFR kinase was exclusively on tyrosine residues. Interestingly, GDP and guanosine 5-3-O-(thio)triphosphate (GTP␥S) inhibited the phosphorylation of G s␣ without altering EGFR autophosphorylation. However, G protein ␤␥ subunits protected against GDP-and GTP␥Smediated inhibition of phosphorylation of G s␣ . In functional studies, phospho-G s␣ demonstrated a greater GTPase activity and also a greater capacity to bind GTP␥S as compared to the nonphosphorylated G s␣ . Moreover, the phospho-G s␣ augmented adenylyl cyclase activity in S49 cyc ؊ cell membranes to a greater extent than its nonphosphorylated counterpart. Therefore, we conclude that phosphorylation of G s␣ on tyrosine residues by the EGFR kinase activates this G protein and increases its ability to stimulate adenylyl cyclase.
Epidermal growth factor (EGF) 1 exerts a variety of biological actions ranging from increased DNA synthesis, hyperplasia, and increased glucose and fatty acid metabolism, to alterations in muscular function (see Ref. 1 for review). These pleiotropic actions of EGF are mediated via the activation of several second messenger systems. For instance, following binding of EGF to its receptors, the intrinsic protein tyrosine kinase activity of the EGF receptor is increased, resulting in autophosphoryla-tion of the EGF receptor as well as of other cellular proteins (reviewed in Refs. 1 and 2). The autophosphorylation of the EGF receptor serves to recruit proteins containing the Src homology 2 (SH2) domains such as phospholipase C␥ (3), the subsequent phosphorylation of which increases phosphatidylinositol metabolism, and the generation of the second messengers inositol 1,4,5-trisphosphate and diacylglycerol (3). Likewise, recruitment of the SH2-containing proteins such as Grb2 and other adaptor proteins to the phosphotyrosine-containing domains on the EGF receptor also leads to the activation of serine/threonine phosphorylation cascades such as the mitogen-activated protein kinase cascade (4). In addition, EGF has also been documented to modulate the cAMP second messenger system. Studies from our laboratory have shown that EGF increases contractility, beating rate, and cAMP accumulation in the heart (5) by stimulating adenylyl cyclase via a process involving G s␣ (6,7). Moreover, the protein tyrosine kinase activity of the EGF receptor is important for EGF-mediated stimulation of cardiac adenylyl cyclase (8). One implication of this latter finding is that one, or more, of the signaling elements involved in stimulation of adenylyl cyclase by the activated EGF receptor is(are) phosphorylated. Therefore, we have proposed the hypothesis that EGF phosphorylates G s␣ and activates this G protein to stimulate adenylyl cyclase activity. This hypothesis incorporates the requirement for G s␣ and the EGF receptor protein tyrosine kinase, the two components which our previous studies have determined to be required for EGF-elicited stimulation of the effector, adenylyl cyclase (7,8).
Among the ␣ subunits of heterotrimeric G proteins, G i␣2 has been demonstrated to be phosphorylated and inactivated by protein kinase C (9 -11). The studies of Hausdorff et al. (12) have shown that in vitro, pp60 c-src phosphorylates G s␣ and that in reconstitution experiments with the ␤-adrenergic receptor, ␤-adrenoreceptor agonists such as isoproterenol augment GTP␥S binding to phosphorylated G s␣ to a greater extent than to nonphosphorylated G s␣ . The tyrosine kinase pp60 c-src phosphorylates G s␣ on tyrosine residues 37 and 377 (13) with a stoichiometry between 0.4 and 0.9 mol of phosphate/mol of G s␣ (12). To date, phosphorylation of G s␣ by tyrosine kinases other than pp60 c-src , and in particular receptor protein tyrosine kinases, has not been reported. Therefore, to address our aforementioned hypothesis we have investigated whether or not the EGF receptor protein tyrosine kinase phosphorylates G s␣ and modulates its activity. Our data demonstrate that the activated EGF receptor (EGFR) stoichiometrically phosphorylates G s␣ and activates this protein as monitored by its ability to bind GTP␥S and activate adenylyl cyclase. To our knowledge, this is the first demonstration of phosphorylation of G s␣ by the EGF receptor with a concomitant increase in activity of this protein.

MATERIALS AND METHODS
Purification of G s␣ and Bovine Brain ␤␥ Subunits-The BL21(DE3) strain of Escherichia coli transformed with the plasmid pQE-60, containing cDNA encoding the 45-kDa form of bovine G s␣ , was obtained * This work was supported in part by National Institutes of Health Grant HL 48308 and a grant-in-aid from the American Heart Association, National Center. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
from Alfred Gilman (University of Texas Southwestern Medical Center). Expression of G s␣ was induced with isopropyl ␤-D-thiogalactopyranoside and the protein was purified essentially as described by Graziano et al. (14). Bovine brain ␤␥ subunits of heterotrimeric G proteins were purified to homogeneity as described by Mumby et al. (15) and Neer et al. (16). Heterotrimeric G s was reconstituted by mixing active G s␣ and ␤␥ subunits (active G s␣ : ␤␥ ratio ϭ 1:1) and incubating for 30 min at 4°C; amount of active G s␣ was determined from maximal GTP␥ 35 S binding (described below).
Phosphoamino Acid Analysis of Phospho-G s␣ -Following phosphorylation of G s␣ and SDS-PAGE as described above, the proteins were electrophoretically transferred to PVDF membranes (Bio-Rad). Following autoradiography to locate the phosphoproteins on the PVDF membrane, the portions of the PVDF containing the proteins of interest were excised and subjected to phosphoamino acid analysis exactly as described by Martensen (18). Briefly, PVDF was rinsed with 2 ϫ 1.0 ml water and hydrolyzed for 60 min at 110°C in 6 N HCl, followed by evaporation to dryness in a vacuum desiccator (Speedvac, Savant). Hydrolyzate was washed with 1.0 ml of H 2 O and 1.0 ml of 0.1 N formic acid (HCOOH) prior to application on to a column of Dowex AG 50W-X2 (H ϩ ), 100 -200 mesh (Bio-Rad) in 0.2 ml of 0.1 N HCOOH. Phosphoamino acids were eluted with 2 ϫ 0.2 ml of 0.1 N HCOOH and 1 ϫ 0.6 ml of 0.1 N HCOOH and separated by thin layer electrophoresis on cellulose acetate-coated plates (Merck, Darmstadt, Germany) in pyridine/glacial acetic acid/H 2 O (10:100:1890), pH 3.5 at 750 V for 90 min. Phosphoamino acids were located by autoradiography and identified by comigration with phosphoamino acid standards visualized with ninhydrin.
Detection of Phosphorylated G s␣ by Western Blotting-After incubation with and without EGFR or ATP, G s␣ was separated from EGFR by SDS-PAGE (12% gel) and electrophoretically transferred to nitrocellulose. Membrane was blocked in 10% newborn calf serum in phosphatebuffered saline and exposed to polyclonal anti-phosphotyrosine antibody (Zymed Inc.; 1:1,000 dilution). After washing, the membranes were exposed to goat anti-rabbit IgG conjugated to horseradish peroxidase (Bio-Rad; 1:20,000 dilution). The Western blot was developed with the Amersham Corp. ECL system.
GTPase Activity Measurements-GTPase activity was measured as described by Brandt et al. (19) and modified by Okamoto et al. (20). Briefly, 500 ng of G s␣ (monomeric or heterotrimeric) were phosphorylated in the presence of 3 M GDP. GTPase activity was then monitored in medium containing the following at final concentration: 25 mM Hepes-NaOH, pH 8.0, 110 M EDTA, 200 M MgSO 4 , 1 mM dithiothreitol, and 100 nM [␥-32 P]GTP (total volume, 1 ml). Aliquots (100 l) were withdrawn at various times and transferred to tubes containing ice-cold 5% (w/v) Norit A in 50 mM NaH 2 PO 4 . Following centrifugation, the 32 P content of the supernatants was measured by scintillation counting. Adenylyl Cyclase Assays with S49 cyc Ϫ Cell Membranes-The ability of heterotrimeric nonphosphorylated G s and phospho-G s to activate adenylyl cyclase activity in mouse lymphoma S49 cyc Ϫ cell membranes was measured as described by Sun et al. (22). Briefly, heterotrimeric G s was phosphorylated in the presence of 10 M GDP; nonphosphorylated G s was similarly treated in the absence of EGFR. Both forms of G s were then preincubated for 60 min at 25°C, with GTP␥S (100 nM) in the GTP␥S binding buffer described above except that the Mg 2ϩ concentration was 500 M. The phosphorylated and nonphosphorylated G s (1.1 pmol) were then reconstituted with 10 g of S49 cyc Ϫ membrane protein for 5 min at 0°C. Adenylyl cyclase reactions were performed in quadruplicate as described by Sun et al. (22).

RESULTS AND DISCUSSION
Previous data from this laboratory have demonstrated that EGF stimulates adenylyl cyclase activity in cardiac membranes via activation of G s␣ (6,7) and that the EGFR protein tyrosine kinase activity is necessary for this effect (8). Therefore, in order to determine whether or not the EGFR protein tyrosine kinase phosphorylates G s␣ and to evaluate the functional significance of such a phosphorylation, experiments were performed with the purified EGFR and pure G s␣ . Initially, the ability of the EGFR kinase to phosphorylate G s␣ as a function of time and the dependence of this phosphorylation on the presence of EGFR and ATP were determined. As illustrated by the data in Fig. 1, the purified EGF receptor increased phosphorylation of G s␣ in a time-dependent manner (Fig. 1A); maximal phosphorylation of the G s␣ was obtained 60 min after initiation of the phosphorylation reaction at 25°C. Moreover, as demonstrated by the data in Fig. 1B, the phosphorylation of G s␣ was dependent upon the presence of the EGFR and ATP. In studies similar to those depicted in Fig. 1, the stoichiometry of phosphorylation of G s␣ by EGFR was determined to be 1.73 Ϯ 0.3 (n ϭ 4) mol of P i incorporated/mol of G s␣ . As a control, the stoichiometry of autophosphorylation of the EGF receptor was monitored in parallel and found to be 4.5 Ϯ 0.5 (n ϭ 5) sites phosphorylated per mol of the EGFR. While the latter data FIG. 1. Phosphorylation of G s␣ by the EGFR protein tyrosine kinase. Panel A, time course for phosphorylation of G s␣ by purified EGF receptor. Purified, recombinant G s␣ was incubated in the phosphorylation mixture described under "Materials and Methods" in the presence of [␥-32 P]ATP. Aliquots (10 l) were withdrawn at various time intervals and an equal volume of 2 ϫ Laemmli sample buffer was added. EGFR and G s␣ were separated by SDS-PAGE (12% acrylamide). Phosphoproteins were visualized by autoradiography. Panel B, ATP-and EGF receptor-dependent phosphorylation of G s␣ . Purified, recombinant G s␣ was phosphorylated as described above in the presence and absence of ATP and EGF receptor. Phosphoproteins were separated by SDS-PAGE and visualized by autoradiography.
confirm the published stoichiometry of the EGFR autophosphorylation (23), the former results indicate that there are two sites on the G s␣ that are phosphorylated by the EGFR.
To determine whether the phosphorylation of G s␣ by the EGF receptor was on tyrosine residues or involved phosphorylation of serine and threonine residues also, two approaches were pursued. First, G s␣ was phosphorylated by the EGFR in the presence of unlabeled ATP, and following separation of proteins by SDS-PAGE and transfer onto nitrocellulose, Western analyses with anti-phosphotyrosine antibody were performed. The data depicted in Fig. 2A demonstrate that both the EGFR and G s␣ are phosphorylated on tyrosine residues, and phosphorylation of G s␣ is dependent on the presence of ATP and EGFR (see also Fig. 1B). In additional experiments, phosphoamino acid analyses of 32 P-labeled G s␣ and autophosphorylated EGFR (positive control) demonstrated that the phosphorylation of both the EGFR and G s␣ was exclusively on tyrosine residues (Fig. 2B).
In the studies involving pp60 c-src -mediated phosphorylation of G s␣ , the GDP-bound form of G s␣ was reported to be a better substrate for the kinase (12). Therefore, we investigated the effects of GDP on phosphorylation of G s␣ by the EGFR kinase. In these experiments, G s␣ was phosphorylated by the EGFR kinase in the presence of different GDP concentrations. Following separation of the proteins by SDS-PAGE and autoradiography, the protein bands corresponding to EGFR and G s␣ were excised from the gels and counted for 32 P content. The data in Fig. 3A demonstrate that, while the autophosphorylation of the EGF receptor was not altered by GDP, phosphorylation of G s␣ was inhibited in a concentration-dependent manner such that, at GDP concentrations in excess of 3 M, the phosphorylation of G s␣ was decreased by 50%. Similarly, GTP␥S also inhibited by 50% the phosphorylation of G s␣ without altering the autophosphorylation of the EGFR (Fig. 3A). Thus, in the presence of guanine nucleotides, the stoichiometry of G s␣ phosphorylation by the EGFR kinase was decreased from 2 mol P i /mol of G s␣ to 1 mol of P i /mol of G s␣ (0.7 Ϯ 0.3 mol of P i /mol G s␣ (n ϭ 3)). Interestingly, in the absence of GDP or GTP␥S, the addition of G protein ␤␥ subunits did not alter the phosphorylation of G s␣ by the EGFR kinase (Fig. 3B, cf. lanes 1 and 3). However, in the presence of GDP (10 M), ␤␥ subunits protected against the GDP-mediated inhibition of phosphorylation of G s␣ (Fig. 3B, cf.  lanes 2 and 4), and therefore, the stoichiometry of G s␣ phos-phorylation by the EGFR kinase in the presence of GDP and ␤␥ subunits was 2 mol of P i /mol of G s␣ . Similar protection against inhibition of phosphorylation of G s␣ by GTP␥S was also seen in the presence of ␤␥ subunits (not shown). Notably, the ␤␥ subunits were not phosphorylated by the EGFR kinase (Fig. 3B). The finding that GDP inhibited phosphorylation of G s␣ (Fig.  3A) is in stark contrast to the effects of this nucleotide on phosphorylation of G s␣ by pp60 c-src (12) but similar to the effects of GDP on phosphorylation of the small molecular weight G protein by EGFR reported by Hart et al. (24). The opposite effects of GDP on phosphorylation of G s␣ by EGFR and pp60 c-src suggest that the sites on G s␣ phosphorylated by these two protein tyrosine kinases may be different. To determine the specificity of phosphorylation of G s␣ by EGFR kinase, we also investigated the ability of the EGFR kinase to phosphorylate G i␣2 . As compared with phosphorylation of G s␣ , the EGFR kinase phosphorylated G i␣2 with a very low stoichiometry (0.4 pmol P i /mole of G i␣2 ) (data not shown). This finding is consistent with the report of Hart et al. (24) that EGFR kinase phosphorylated G i␣ and G o␣ with a very low stoichiometry. Hence, it would appear that the EGFR kinase selectively phosphorylates G s␣ .
In vivo, under basal conditions, prior to activation of the EGFR, the G s␣ would be predominantly in the GDP bound heterotrimeric form. Therefore, to study the functional consequences of G s␣ phosphorylation, this situation was recreated in in vitro studies by phosphorylating the G s␣ by the EGFR kinase as a heterotrimer (i.e. presence of ␤␥ and GDP) a condition under which G s␣ is phosphorylated on two sites (see Fig. 3B). The data in Fig. 4A demonstrate that the GTPase activity of the phospho-G s␣ was 2-fold greater than that of its nonphosphorylated counterpart; in these experiments, controls with ␤␥ subunits plus EGFR alone were found to have negligible activity (not shown). This increase in GTPase activity of the phosphorylated G s␣ also suggests that the EGFR kinase phosphorylates G s␣ on tyrosine residues which may be different to those phosphorylated by the pp60 c-src (13) since phosphorylation of G s␣ by the latter kinase does not alter its GTPase activity (12). As a second approach to assess functional significance, experiments were also performed to determine if phosphorylation of G s␣ by EGFR kinase altered the ability of the protein to bind GTP␥S. As demonstrated by the data in Fig. 4B, phospho-G s␣ bound GTP␥S at a greater rate and to a larger extent as compared to the nonphosphorylated G s␣ ; GTP␥S binding to ␤␥ subunits plus EGFR (control) was negligible (not shown). These data (Figs. 4, A and B) strongly suggested that phosphorylation of G s␣ in its heterotrimeric form (i.e. in presence of GDP and ␤␥ subunits) markedly activated this G protein. Therefore, to determine whether or not the phosphorylated G s␣ was indeed more effective as an activator of adenylyl cyclase, experiments were performed employing S49 cyc Ϫ cell membranes which do not contain any endogenous G s␣ (25). The data from these experiments demonstrated that upon reconstitution of equal amounts of nonphosphorylated and phosphorylated G s␣ with cyc Ϫ membranes, the adenylyl cyclase activity was 2-fold greater in the presence of the phospho-G s␣ than its non-phosphorylated counterpart (Fig. 4C). In controls performed with ␤␥ subunits with the EGFR (control), adenylyl cyclase activity was very low and comparable to the activity in the absence of any additions (cf. Fig. 4C and Ref. 22).
Since GDP and GTP␥S inhibit phosphorylation of G s␣ similarly (Fig. 3), and because the structures of GDP-and GTPbound forms of G ␣ subunits are different; see e.g. with transducin ␣ subunit (G t␣ ) (26) and G i␣ (27,28), our data (Fig. 3A) would suggest that the tyrosine residue(s) whose phosphorylation is(are) altered by the guanine nucleotides is(are) not located in the regions of the molecule that change conformation upon exchange of GDP for GTP. Indeed the crystal structure data of the GDP-and GTP-bound forms of G t␣ and G i␣ (26 -28) indicate that the GTP-bound form of G ␣ subunits are different from the GDP-bound form in essentially three regions (switch I, switch II, and switch III) none of which contains tyrosine residues (26). However, on G s␣ , tyrosine residues (Tyr-176 and -239) are located proximal to switch I and switch III regions and it is plausible that phosphorylation of one or both of these residues by the EGF receptor increases the rate of GTP␥S binding that is observed (Fig. 4B). Likewise since tyrosine residues (Tyr-325, -344, and -346) are also located in the proximity of adenylyl cyclase interacting regions on G s␣ (see e.g. Ref. 26), it is tempting to speculate that phosphorylation of one of these residues increases the interaction of G s␣ with adenylyl cyclase, thereby augmenting activity as observed in Fig. 4C. Interestingly, phosphorylation of the heterotrimeric form of G s by the EGF receptor was not altered by either GDP (Fig. 3B) or GTP␥S (not shown) suggesting that the tyrosine residues on G s␣ that are phoshorylated reside in region(s) whose conformation is stabilized by the ␤␥ subunits. Additional studies which will identify the sites on G s␣ that are phosphorylated by the EGF receptor will provide more mechanistic information in the light of the crystal structure of G t␣ and G i␣ . Presently, from the data presented in Fig. 3, we can conclude that guanine nucleotide binding (either GDP or GTP) inhibits the phosphorylation of G s␣ by the EGF receptor and that ␤␥ subunits protect against such inhibition of phosphorylation.
Several laboratories have reported the phosphorylation of ␣ subunits of G proteins by different kinases as a potential regulatory event in signal transduction. However, functional significance of the phosphorylation has been demonstrated in only a few of the studies. Thus it is clear that phosphorylation of G i␣2 by protein kinase C decreases the activity of this G protein (9 -11). On the other hand, the functional consequence of phosphorylation of G z␣ by protein kinase C (29) remains to be elucidated. Similarly, although the ␣ subunits of transducin (30), G o , and G i (31) have been shown to be phosphorylated by the insulin receptor protein tyrosine kinase, the functional consequences of these phosphorylations remain unknown. It should be noted that G s␣ is not phosphorylated to any significant extent by either protein kinase C (29) or the insulin receptor protein tyrosine kinase (31). In this respect, our findings for the first time, demonstrate the phosphorylation of G s␣ by a receptor protein tyrosine kinase and provide information concerning the alteration in G s␣ function due to these phosphorylations. Most importantly, the data presented here provide a tenable mechanism for EGF-mediated stimulation of adenylyl cyclase activity. Notably, however, this may not be the only mechanism involved in EGF-elicited stimulation of adenylyl cyclase activity but may represent one of two different, but not mutually exclusive, manners by which EGF may augment adenylyl cyclase activity. Hence, recently we have shown that a juxtamembrane 13-amino acid sequence in the EGFR can activate G s and thereby stimulate adenylyl cyclase activity (22). This finding is consistent with the hypothesis that upon binding to its receptors, EGF stimulates autophosphorylation of the receptors which results in a change in the conformation of the cytosolic domain of the EGF receptor from a compact to an extended form (32) and thereby, allows the juxtamembrane FIG. 3. GDP and GTP␥S inhibit the phosphorylation by EGFR kinase of G s␣ in its monomeric but not heterotrimeric form. Panel A, inhibition of monomeric G s␣ phosphorylation in the presence of varying concentrations of GDP (circles) or GTP␥S (squares). Purified G s␣ was phosphorylated by the EGFR kinase in the absence and presence of guanine nucleotides as described under "Materials and Methods." After separation of proteins by SDS-PAGE and location of the G s␣ and EGFR by autoradiography, the two protein bands were excised and counted for 32 P content. Data are presented as percent of radioactivity in the EGFR (open symbols) and G s␣ (filled symbols) in the absence of guanine nucleotides. Panel B, GDP does not inhibit phosphorylation of the G s␣ in its heterotrimeric form. Purified G s␣ mixed with G protein ␤␥ subunits (G s␣ :␤␥ ϭ 1:1) were phosphorylated in the presence and absence of GDP (10 M). Following separation of proteins, the gels were dried and subjected to autoradiography. Lane 1, G s␣ ϩ EGFR; lane 2, G s␣ ϩ EGFR ϩ GDP (10 M); lane 3, G s␣ ␤␥ ϩ EGFR; lane 4, G s␣ ␤␥ ϩ EGFR ϩ GDP (10 M). The asterisk indicates the migration of ␤ subunit as determined by Coomassie staining of the gel. region to interact with and stimulate G s . However, in addition to this mode of activation of G s , simultaneous phosphorylation of G s␣ by the EGF receptor protein tyrosine kinase could, in a mutually reinforcing manner, amplify the signal from EGFR to G s and ultimately to adenylyl cyclase, since the phospho-G s␣ is a better activator of this effector (Fig. 4C). This latter possibility, i.e. amplification of signaling by combinatorial effects of phosphorylation and interaction of the cytosolic, juxtamembrane, region of the EGF receptor with G s and its implications on adenylyl cyclase activity remains to be experimentally tested.
In conclusion, we have presented experimental evidence to demonstrate that EGF receptor protein tyrosine kinase can phosphorylate G s␣ on tyrosine residues and stimulate the functional activity of this G protein. To our knowledge, this is the first demonstration that a receptor protein tyrosine kinase can phosphorylate and activate G s . Moreover, these data provide mechanistic insights into EGF-elicited stimulation of adenylyl cyclase activity. Presently, the identity of the tyrosine residues on G s␣ that are phosphorylated by EGFR kinase remain to be elucidated. However, stoichiometry analyses indicate that at least 2 tyrosine residues are phosphorylated. Whether or not the phosphorylation of one of these 2 tyrosine residues alters the GTPase and GTP␥S binding activities preferentially or phosphorylation of both tyrosines is required to observe the functional changes is not known. These questions and the possibility that there may be a hierarchy in the phosphorylation of the two tyrosines on G s␣ by the EGFR kinase forms the subject of our future investigations.