Activation of Akt/Protein Kinase B by G Protein-coupled Receptors

The serine/threonine protein kinase Akt has recently been shown to be implicated in the pathway leading to cell survival in response to serum and growth factors in a variety of cellular systems. However, the existence of a biochemical route connecting this kinase to the large family of receptors that signal through heterotrimeric G proteins is yet to be explored. In this study, we set out to investigate whether GTP-binding protein (G protein)-coupled receptors (GPCRs) can stimulate Akt activity and survival pathways and, if so, to define the mechanism(s) whereby this class of cell surface receptors could regulate Akt function. Using ectopic expression of GPCRs in COS-7 cells as a model, we have observed that both m1 and m2 muscarinic acetylcholine receptors, representative of those GPCRs coupled to Gq and Giproteins, respectively, can readily activate an epitope-tagged form of Akt kinase and prevent UV-induced apoptosis. We have also found that the pathway connecting G proteins to Akt implicates signals emanating from Gαq, Gαi, and βγ dimers, but not from Gαs or Gα12, in each case acting through a pathway that involves a phosphatidylinositol-3-OH kinase activity. Moreover, our findings suggest a role for a novel βγ-sensitive complex, p101·phosphatidylinositol-3-OH kinase-γ, in the transduction of signals leading to Akt stimulation and cell survival by GPCRs and open new avenues for research on the function of the large family of G protein-linked receptors in the regulation of anti-apoptotic pathways.

The serine/threonine protein kinase Akt has recently been shown to be implicated in the pathway leading to cell survival in response to serum and growth factors in a variety of cellular systems. However, the existence of a biochemical route connecting this kinase to the large family of receptors that signal through heterotrimeric G proteins is yet to be explored. In this study, we set out to investigate whether GTP-binding protein (G protein)coupled receptors (GPCRs) can stimulate Akt activity and survival pathways and, if so, to define the mechanism(s) whereby this class of cell surface receptors could regulate Akt function. Using ectopic expression of GPCRs in COS-7 cells as a model, we have observed that both m1 and m2 muscarinic acetylcholine receptors, representative of those GPCRs coupled to G q and G i proteins, respectively, can readily activate an epitopetagged form of Akt kinase and prevent UV-induced apoptosis. We have also found that the pathway connecting G proteins to Akt implicates signals emanating from G␣ q , G␣ i , and ␤␥ dimers, but not from G␣ s or G␣ 12 , in each case acting through a pathway that involves a phosphatidylinositol-3-OH kinase activity. Moreover, our findings suggest a role for a novel ␤␥-sensitive complex, p101⅐phosphatidylinositol-3-OH kinase-␥, in the transduction of signals leading to Akt stimulation and cell survival by GPCRs and open new avenues for research on the function of the large family of G proteinlinked receptors in the regulation of anti-apoptotic pathways.
Receptors coupled to heterotrimeric GTP-binding proteins (G proteins) 1 are the largest group of integral membrane receptors involved in the transmission of signals from the extracellular environment to the cytoplasm (1). A wide range of external stimuli, including neurotransmitters, growth factors, hormones, light, odorants, and certain taste ligands, can activate specific members of this family and promote a conformational change that is transmitted to the cytoplasmic side of the receptor protein (1). This leads to a physical interaction between the receptor and the GDP-bound G protein heterotrimer which causes the dissociation of the guanine nucleotide and the incorporation of GTP in the G protein ␣ subunit, thus releasing the ␤␥ heterodimer (2). In turn, GTP-bound G protein ␣ subunits and ␤␥ complexes initiate, independently, a wide variety of intracellular signaling pathways (3).
Although the G protein-coupled receptor (GPCR) family is involved in many functions performed by fully differentiated cells, these receptors are also expressed in most proliferating cells, and they have been implicated in embryogenesis, tissue regeneration, and growth stimulation (4). The nature of the growth regulatory pathway(s) stimulated by GPCRs has just begun to be elucidated (5). In our laboratory, we have used the ectopic expression of human muscarinic receptors for acetylcholine (mAChRs) in NIH 3T3 cells as a model system for studying mitogenic signaling through G protein-linked receptors. In this biological setting, we have shown that certain mAChR subtypes can effectively transduce mitogenic signals and, when activated persistently, induce the transformed phenotype (6). Interestingly, two recent reports have demonstrated that the activation of endogenously expressed muscarinic receptors is per se sufficient to block apoptosis in neuronal cells (7,8). These results suggest that in addition to their role in cell growth, GPCRs might also activate yet to be defined survival pathways.
In this regard, the serine-threonine kinase Akt/PKB, which was first identified as the human homolog of a transforming oncogene (9), has been shown recently to control intracellular pathways preventing cell death in response to a variety of extracellular stimuli (10) and in a wide range of cellular systems (11,12). The mechanism of activation of Akt by tyrosine kinase growth factor receptors has been established recently (13). However, the regulation of this intriguing kinase by GPCRs is still unclear. Initial reports showed that lysophosphatidic acid, acting through G i -coupled receptors, was unable to stimulate Akt activity in NIH 3T3 cells (14). Furthermore, direct activation of PKC by phorbol esters also failed to activate Akt in this fibroblast cell line (14,15). On the other hand, recent reports have described Akt stimulation in response to GPCRs in rat epididymal fat cells (16) and human embryonic kidney 293-EBNA cells (17), albeit by a yet to be determined mechanism.
In this study, we set out to investigate whether Akt can be activated effectively by GPCRs using ectopically expressed receptors and different G protein subunits in COS-7 cells as a model system. We found that both G q -and G i -coupled GPCRs, m1 and m2 receptors, respectively, can readily activate an epitope-tagged form of Akt kinase. We also show that both signal-transducing molecules generated upon GPCR activation, ␤␥ complexes and ␣ subunits, can effectively promote Akt activation in a PI3K-dependent manner. Moreover, our findings suggest a role for the novel PI3K␥ and its associated regulatory subunit, p101, in the transduction of signals leading to Akt stimulation by GPCRs. We also present evidence that GPCRs can induce survival pathways through PI3K␥ acting on Akt.

EXPERIMENTAL PROCEDURES
Cell Lines and Transfection-COS-7 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Cells were transfected by the DEAE-dextran technique, adjusting the total amount of DNA to 1-4 g/plate with pcDNA3-␤-galactosidase DNA, when necessary (18).
Expression Plasmids-An epitope-tagged Akt (pCEFL-HA-Akt) as well as the membrane-targeted mutant (pCEFL-myr-HA-Akt) were generated by inserting the coding region of Akt and myr-Akt (kindly provided by Dr. P. N. Tsichlis) into the pCEFL expression vector. Expression plasmids for m1 and m2 mAChRs, for the ␣ and ␥ isoforms of the PI3K and their mutants, as well as expression plasmids for ␤ 1 , ␥ 2 , and ␥ 2 * subunits of heterotrimeric G proteins were reported previously (18 -20). Plasmids expressing the coding region of PI3K␥ and the p101 protein fused to the Glu-Glu tag were kindly provided by L. R. Stephens and are described elsewhere (21). Plasmids expressing the GTPasedeficient, constitutively activated forms of representative G protein ␣ subunits (G␣ 12 -Q227L, G␣ i2 -Q205L, G␣ q -Q209L, and G␣ s -R201C) have been described already (19,22,23). An expression plasmid for a chimeric molecule between the extracellular and transmembrane domain of CD8 fused to the carboxyl-terminal 222 amino acids of ␤ARK, which includes both the ␤␥-binding domain and the PH domain of ␤ARK (pcDNA-CD8-␤ARK), has also been described recently (22,24).
Apoptosis Assay-COS-7 cells were grown on coverslips and transfected with the indicated plasmids together with an expression vector for ␤-galactosidase as a marker for transfection, using the Lipo-fectAMINE-Plus reagent (Life Technologies, Inc.) following the manufacturer's instructions. Protection from UV-induced apoptosis was performed essentially as described (12). Briefly, 24 h after transfection, cells were serum starved overnight in Dulbecco's modified Eagle's medium containing 10 mM Hepes and subjected subsequently to UV irra-diation (120 mJ, UV-Stratalinker 1800, Stratagene). After the addition of fresh serum-free medium containing or not 1 mM carbachol, cells were maintained in the incubator, fixed 8 h later in 4% paraformaldehyde, and permeabilized with 0.01% Triton X-100. Transfected cells were identified by immunostaining for ␤-galactosidase expression with a mouse anti-␤-galactosidase antibody (Promega, 1:100) followed by a rabbit anti-mouse rhodamine-coupled secondary antibody (Sigma, 1:200). Fragmented DNA was then visualized by the terminal deoxynucleotidyltransferase-mediated dUTP-FITC nick end labeling (TUNEL) technique using a kit from Boehringer Manheim, following the manufacturer's instructions, except that the reaction was carried out at room temperature instead of at 37°C. The frequency of apoptosis was scored by counting several hundred rhodamine-stained (transfected) cells from at least 20 different fields/coverslip and examining them for FITC staining (TUNEL-positive) under UV light in an Axio-plan2 fluorescence microscope (Zeiss).

RESULTS
Stimulation of muscarinic GPCRs has been reported to induce protection from apoptosis both in cerebellar granule neurons (8) and in the pheochromocytoma cell line PC12 (7), and the Akt kinase has been implicated in survival pathways in many cell types, including PC12 and other neuroectodermalderived cells (11,13). Thus, we asked whether activation of endogenous muscarinc receptors would lead to Akt activation in PC12 cells. As shown in Fig. 1A, exposure of PC12 cells to the cholinergic agonist carbachol induced the rapid stimulation of the Akt phosphotransferase activity, to an extent similar to that observed in these cells in response to nerve growth factor acting on its cognate receptors.
To investigate further the molecular mechanism(s) whereby muscarinic receptors activate Akt and induce protection from cell death, we chose to use a reconstituted system, consisting of the coexpression of m1 and m2 muscarinic receptors together with an epitope-tagged Akt (HA-Akt) in COS-7 cells. Whereas m1 receptors are typical of those coupled through G proteins of the G q family to phospholipase C activation, m2 is known to couple through G i to a number of effector pathways, including the inhibition of adenylyl cyclases (25). In cells expressing either muscarinic receptor we observed that carbachol induced an increase in Akt activity, as judged by immune complex kinase reactions using histone 2B as a substrate. When mediated by m1 receptors, induction of histone 2B phosphorylation was evident as early as 1 min after the addition of agonist, showing an early peak at approximately 3 min after stimulation (Fig. 1B). Stimulation of m2 also caused a very rapid activation, which was evident after 1 min and reached a maximum around 15 min (Fig. 1B). Both m1-and m2-induced Akt kinase activity remained elevated for an extended period of time, decreasing to the basal activity as late as 2-3 h after treatment (data not shown).
Recent work has demonstrated that PI3K activity is required for Akt activation in the majority of the systems described to date (26). However, PI3K-independent pathways have also been described (17,27), including those mediating Akt activation by ␤ 3 -adrenergic GPCR in epididymal fat cells (16). In our experimental system, preincubation of cells with 50 nM wortmannin, a potent PI3K inhibitor, completely blocked both m1and m2-mediated stimulation of Akt (Fig. 1C). Thus both G qand G i -coupled receptors appear to activate Akt irrespective of their G protein coupling specificity, utilizing a PI3K-dependent pathway.
As an approach to investigate which G protein(s) mediates Akt activation, we used the expression of GTPase-deficient mutationally activated forms of G protein ␣ subunits, which can activate effector pathways by obviating the need for receptor stimulation (2). Thus, we coexpressed the epitope-tagged Akt together with GTPase-deficient mutants for G␣ s , G␣ i2 , G␣ q , and G␣ 12 , representing each of the four known ␣ subunits (3). As shown in Fig. 2A, the activated mutant of G␣ q was able to trigger Akt activation similar to that caused by Ras when used as control (15,28). Activated G␣ i2 also stimulated Akt, although less efficiently than G␣ q , and the activated mutants of G␣ s and G␣ 12 had no demonstrable effect under our experimental conditions. These data indicate that G␣ q , and to a lesser extent G␣ i , can mediate Akt activation by G protein-linked receptors.
When activated, GPCRs catalyze the replacement of GDP by GTP bound to the ␣ subunit and induce the dissociation of ␣-GTP from ␤␥ dimers. Although the ␣ subunits were thought to be the sole responsible molecules for coupling receptors to second messenger-generating systems, recent work has estab-lished a critical role for ␤␥ dimers in signal transduction (3,19). These data prompted us to explore whether ␤␥ dimers might also participate in signaling to Akt. We observed that when cotransfected, ␤ 1 ␥ 2 subunits induce a remarkable increase in the phosphorylating activity of the epitope-tagged Akt, although expression of the HA-Akt was similar for each transfected cell population (Fig. 2B). In contrast, Akt was activated poorly when coexpressed with ␤ 1 or ␥ 2 alone or when coexpressed with ␤ 1 and an altered form of ␥ 2 subunit, designated ␥ 2 *, which lacks the ␥-isoprenylation signal and therefore fails to associate to the plasma membrane (19) (Fig. 2B). Based upon these results, we conclude that membrane-bound forms of ␤␥ subunits of heterotrimeric G proteins can potently stimulate Akt activity in COS-7 cells. In view of these results, we next sought to explore the relative contributions of ␣ and ␤␥ proteins in Akt stimulation by mAChRs. To approach this question, we employed a chimeric molecule combining the extracellular and transmembrane domain of CD8 fused to the carboxyl-terminal domain of ␤ARK which includes the high affinity ␤␥ binding region of this kinase as described (22). This chimeric molecule targets the COOH-terminal part of ␤ARK to the plasma mem- cells were grown in 10-cm plates to 80% confluence and serum-starved overnight before being subjected to stimulation with carbachol (cch) (1 mM, 2 min) or nerve growth factor (NGF) (50 ng/ml, 10 min). Cell lysates were then immunoprecipitated with goat anti-Akt serum, and kinase reactions were performed as described under "Experimental Procedures." The graph represents the mean Ϯ S.E. of three independent experiments. A representative experiment is shown in the inset. H2B, histone 2B. Panel B, COS-7 cells were transfected with expression plasmids for m1 or m2 mAChRs (0.5 g/10-cm plate) together with a plasmid expressing an epitope-tagged Akt (pCEFL-HA-Akt, 0.5 g/ dish). Cultures were stimulated by the addition of carbachol (1 mM) for the indicated times. Subsequently cells were quickly lysed on ice, and Akt activity was determined in the HA-immunoprecipitates as described under "Experimental Procedures." The amount of kinase protein present in each immunoprecipitate was assessed by Western blot analysis with anti-HA antibody. Autoradiograms are from a representative experiment that was repeated three times with nearly identical results. Panel C, transfected cells were treated for 30 min before lysis with wortmannin (50 nM) or vehicle (dimethyl sulfoxide) for controls. Plates were then stimulated for 5 or 15 min, respectively, for m1 and m2, and processed as described in panel A. Data represent the mean Ϯ S.E. of three independent experiments, expressed as the fold increase in carbachol-stimulated Akt activity with respect to controls.

FIG. 2. Overexpression of ␣ or ␤␥ subunit of heterotrimeric G proteins results in stimulation of Akt activity. COS-7 cells were
transfected with pCEFL-HA-Akt (0.5 g/plate) together with pcDNA3-␤-galactosidase vector (control), an activated form of Ras (pCEFL-AU5-RasV12, 0.5 g/plate), or with expression vectors carrying cDNAs for the activated forms of the ␣ subunits of G s , G i2 , G q , (panel A) or expressing ␤ 1 and ␥ 2 G protein subunits (panel B) (0.5 g/plate in each case). As indicated, kinase reactions and Western blot analysis were performed in anti-HA immunoprecipitates from the corresponding lysates or in 50 l of the corresponding total lysates to confirm the expression of the different G proteins as described under "Experimental Procedures" (data not shown). The autoradiograms shown correspond to a representative experiment that was repeated three times. 32 P-Labeled products as well as specific bands detected by the anti-HA antibody are indicated with an arrow. Panel C, Akt activity was measured in cells cotransfected with HA-Akt (0.5 g/plate) and a plasmid carrying the COOH terminus of ␤ARK fused to the CD8 receptor (CD8-␤ARK, 2 g/plate) or the same amount of a vector expressing ␤-galactosidase as a control. Data represent the mean Ϯ S.E. of three to five independent experiments expressed as the percentage of activation with respect to control transfected cells.
brane where it is expected to bind and sequester free ␤␥ complexes when dissociated from G␣ subunits upon receptor stimulation, thus blocking ␤␥-dependent pathways (22). As shown in Fig. 2C, coexpression of CD8-␤ARK with the m2 mAChR nearly abolished the activation of Akt in response to carbachol, whereas m1-mediated Akt stimulation was only partially impaired by overexpression of this ␤␥-sequestering molecule. In contrast, Akt activation by other effectors such as Ras was unaffected by CD8-␤ARK. Taken together, these findings strongly suggest that signaling from m2 mAChR to Akt is mediated primarily by the ␤␥ subunits of heterotrimeric G proteins, whereas m1-mediated activation is achieved via G␤␥dependent and -independent pathways. m1 receptors and activated G␣ q efficiently stimulate phospholipase C␤ causing the hydrolysis of phosphoinositides (29). This results in the generation of two second messengers: inositol trisphosphate, which leads to the elevation of intracellular [Ca 2ϩ ], and diacylglycerol, which activates PKC (25). However, several lines of evidence suggested that PKC does not participate in signaling from m1 and G␣ q to Akt; direct activation of PKC by phorbol 12-myristate-13-acetate (100 ng/ml for 15 min) provoked no change in Akt activity, nor did specific inhibitors of PKC such as bisindolylmaleimide (10 M for 30 min) affect m1 or G q -induced Akt activation (data not shown). On the other hand, wortmannin treatment clearly blocked the stimulation of Akt by m1 (see above, Fig. 1C) and, similarly, prevented Akt activation by G␣ q -QL and ␤␥ coexpression (data not shown). These data implicated a function for PI3K rather than for PKC in the pathway connecting G protein-initiated signals to Akt. In this regard, whereas many cell surface receptors activate PI3K␣ and ␤ through the tyrosine phosphorylation of their p85 subunit, GPCRs were shown recently to activate a novel PI3K isoform, PI3K␥, which does not interact with p85 (21). To explore whether this novel PI3K isoform participates in signaling from G proteins to Akt, we cotransfected a dominant negative form PI3K␥ together with activated G␣ q , G␣ i2 , ␤␥, and activated Ras. As shown in Fig. 3, cotransfection of a dominant negative PI3K␥ did not affect Ras-induced stimulation of Akt, suggesting that in COS-7 cells Ras might act mainly via PI3K␣, as proposed previously (28). However, this PI3K␥ mutant nearly abolished Akt activation by each heterotrimeric G protein subunit, G␣ q -QL, G␣ i2 -QL, and ␤␥ complexes (Fig. 3), as well as by m1 and m2 receptors (data not shown), thus suggesting that a PI3K␥ or a PI3K␥-like kinase is required for both ␣and ␤␥-mediated pathways elicited by GPCRs.
To assess further the involvement of a PI3K activity downstream of GPCR in the signaling pathway to Akt, we investi-gated whether the overexpression of different isoforms of PI3K was per se sufficient to induce Akt activation. As observed in Fig. 4, expression of both ␣ and ␥ isoforms of PI3K triggered Akt activation poorly, although a constitutively active form of PI3K␥ (myr-PI3K␥) (19) revealed a greater ability to stimulate Akt than each of the wild type forms.
For PI3K␥, a novel noncatalytic subunit unrelated to p85 has been identified recently (21) and named p101. It has been shown recently that coexpression of this molecule with PI3K␥ enhances its basal activity and potentiates PI3K activation by ␤␥ dimers (21). We therefore asked whether the expression of this noncatalytic PI3K subunit p101 could potentiate the effect of PI3K␥ on Akt. As shown in Fig. 4, Akt activity increased dramatically when both proteins were expressed together. Interestingly, however, we observed that expression of p101 was able per se to induce an increase in Akt activity, to an extent comparable to that elicited by PI3K␥ alone (Fig. 4). These data suggest that p101 may activate an endogenous PI3K␥ or a PI3K␥-like protein thereby stimulating Akt phosphorylating activity.
To assess further the biological consequences of Akt activation by GPCRs, we set out to investigate whether m1 and m2 receptors could induce survival pathways in COS-7 cells. For that, we took advantage of the recent observation that transfected COS-7 cells undergo apoptosis upon UV irradiation (12) and that Akt activation can protect cells from death in this cellular setting (12). In preliminary experiments, the ED 50 for FIG. 3. Effect of overexpression of a kinase-deficient form of PI3K␥ on Akt activation by constitutively active G␣ subunits. pCEFL-HA-Akt (0.5 g/plate) was cotransfected together with expression plasmids for the constitutively active mutants of G q , G i , and Ras described in Fig. 2 (0.5 g/plate in each case) and a kinase-deficient form of PI3K␥ (pCEFL-PI3K␥ K799R, 0.5 g/plate) or the same amount of pCDNA3-␤-galactosidase as a control. Quantitation of Akt activity present in anti-HA immunoprecipitates was performed as in Fig. 1. We also cotransfected ␤ 1 and ␥ 2 subunits (0.5 g of each/plate) with PI3K␥ K799R or ␤-galactosidase as an additional control. The autoradiograms shown correspond to a representative experiment that was repeated three times. the apoptotic effect of UV was found to be approximately 120 mJ, a dose of UV which was utilized for the subsequent assays. As shown in Fig. 5, under control conditions less than 10% of the transfected cells displayed an apoptotic phenotype and were labeled by the TUNEL reaction. However, when irradiated by UV a fraction of the untransfected and transfected cells underwent apoptosis, the former visualized as TUNEL-positive cells (FITC-stained) and the latter as both anti-␤-galactosidase and TUNEL-positive cells (rhodamine-and FITC-stained, respectively), as depicted in Fig. 5B. Of interest, the addition of carbachol protected m1-and m2-transfected cells from UVinduced apoptosis, to an extent similar to that caused by transfection of an activated form of Akt, myr-Akt (Fig. 5A). In contrast, carbachol treatment produced no apparent consequences in mock-transfected or myr-Akt-transfected cells. Thus, in COS-7 cells m1 and m2 GPCRs can effectively activate survival mechanisms that are able to counteract apoptotic insults, most likely through Akt.
Based on our previous results, we wanted to study further the role of PI3K␥ in the survival pathway stimulated by GPCRs. Initial experiments exploring the effect of wortmannin on the protective activity of m1 and m2 were inconclusive, as treatment of serum-starved cells with wortmannin caused per se a noticeable apoptotic effect (data not shown). Accordingly, we next examined the effect of overexpression of the kinasedeficient mutant of PI3K␥ on the antiapoptotic activity elicited by m1 and m2 receptors. As shown in Fig. 6, expression of the kinase-deficient mutant of PI3K␥ prevented the protective effect of m1 and m2 stimulation without affecting the basal UV-induced apoptosis or the protection conferred by expression of an activated form of Akt, thus suggesting a role for PI3K␥ in the survival pathway elicited by GPCRs.

DISCUSSION
The multiplicity of intracellular signaling pathways activated by the large family of GPCRs has just begun to be appreciated. In particular, recent work has provided a wealth of evidence that this family of cell surface receptors can trigger biochemical routes communicating the membrane to the nucleus, thereby controlling key molecules involved in gene expression regulation (4,5). Furthermore, these receptors can affect normal and aberrant cell growth as well as regulate antiapoptotic pathways (7,8). In this regard, it has been shown that activation of muscarinic GPCRs in PC12 cells can induce cell survival (7). Here, we show that stimulation of PC12 cells with the cholinergic agonist carbachol can activate Akt to an extent comparable to that observed in response to nerve growth factor, a natural survival-promoting factor for these cells (30), thus suggesting that the Akt kinase might participate in the survival pathway(s) elicited by GPCRs. These cells, as many neuroectodermal derived cells, express a mixed population of muscarinic receptor subtypes (31), each exhibiting a distinct coupling selectivity, thus limiting the ability to characterize the pathway linking these receptors to Akt. However, when ectopically expressed in COS-7 cells both m1 and m2, G q -and G i -coupled receptors, respectively, were able to activate Akt efficiently, thus providing an experimental model where the activation of Akt by these GPCRs could be examined in a molecularly defined reconstituted system. In these cells, we found that whereas activated G␣ i2 can induce Akt activity poorly, both activated G␣ q and ␤␥ complexes were potent stimulators of this serine/threonine kinase. In line with this observation, a ␤␥-sequestrant, CD8-␤ARK, nearly abolished Akt stimulation in response to m2 activation but had a more lim-  6. A kinase-deficient PI3K␥ is able to prevent the protective effect of m1 and m2 receptors on UV-induced apoptosis in COS-7 cells. COS-7 cells were grown onto coverslips in six-well plates and transfected with 0.5 g/well pcDNA3-␤-galactosidase together with 0.5 g/well plasmids containing m1 or m2 receptors or a myristoylated form of Akt (pCEFL-myr-HA-Akt). Where indicated, 1 g of pCEFL-PI3K␥-K799R or empty vector as a control was included in the transfection mix. Cells were then irradiated with UV light and treated with carbachol, as indicated. The percentage of apoptotic cells was determined as described under "Experimental Procedures." Data represent the mean Ϯ S.E. from three independent experiments. ited effect on m1-induced Akt activation. Collectively, these data indicate that whereas G i -coupled receptors signal primarily to Akt through ␤␥ dimers, G q -coupled receptors utilize both ␤␥-dependent and -independent mechanisms, the latter likely acting through G␣ q .
Of interest, PKC did not appear to link G␣ q to Akt, but pharmacological and biochemical evidence supported a role for a PI3K downstream from GPCRs and heterotrimeric G protein subunits in the pathway leading to Akt, and we obtained data to suggest that the PI3K␥⅐p101 dimer is a likely candidate to communicate G proteins to Akt. This finding can help explain the seemingly contradictory results on Akt activation by GPCRs (14 -17; see above) because PI3K␥ is highly expressed in hematopoietic cells but poorly expressed in other tissues. In COS-7 cells, we can detect limited expression of PI3K␥ (not shown) (20) which is consistent with the observation that p101 expression alone is sufficient to activate Akt, albeit to a limited extent compared with that achieved upon coexpression of PI3K␥ and p101. Thus, in tissues and cell lines lacking PI3K␥, either GPCR would fail to activate Akt (14,15) or other PI3K isoforms, such as PI3K␤ (32), might link heterotrimeric G proteins to Akt.
To examine the biological consequences of activating GPCRs in COS-7 cells, we investigated the ability of m1 and m2 receptors to protect COS-7 cells from UV-induced apoptosis. In this reconstituted system, we found that both GPCRs were able to activate cell survival pathways effectively, most likely through Akt. Furthermore, we observed that the protective effect elicied by m1 and m2 was nearly abolished by expression of a dominantnegative mutant form of PI3K␥, suggesting that the GPCR-PI3K␥ pathway is biologically relevant in this cellular setting.
In summary, our findings raise the possibility of the existence of a novel pathway activating Akt and preventing apoptosis by cell surface receptors. This pathway involves extracellular ligands acting on GPCRs and the consequent activation of G i or G q proteins, depending on the coupling selectivity. In turn, these G proteins will release ␤␥ subunits, and ␤␥ dimers and activated G␣, particularly G␣ q , will then stimulate PI3K␥⅐p101 complexes or other yet to be identified PI3K isoforms, leading to Akt activation and promoting cell survival.