Differentiation of F9 Teratocarcinoma Stem Cells to Primitive Endoderm Is Regulated by the Giα2/Gsα Axis via Phospholipase C and Not Adenylylcyclase

Morphogen-induced decline in Giα triggers F9 teratocarcinoma stem cells to progress to primitive endoderm via activation of protein kinase C and mitogen-activated protein kinase (Gao, P., and Malbon, C. C. (1996) J. Biol. Chem. 271, 9002-9008). Constitutive expression of Giα2 blocks, whereas expression of Gsα provokes, progression to primitive endoderm, permitting identification of the effectors of the response-utilizing chimera created between Giα2 and Gsα. N-terminal substitution of Gsα with Giα2 sequence to create chimera Giα2 (1-54)/Gsα produced a chimera that activated adenylylcyclase but abolished progression to primitive endoderm and activation of phospholipase C. C-terminal substitution of Gsα with Giα2 sequence to Gsα/Giα2 (320-355) enhanced the ability of Gsα to promote progression. The Q205L-activated mutant of Giα2 suppresses, whereas the G225T-activated mutant of Gsα strongly activates phospholipase C and progression in these cells. The N-terminal region of Gsα (residues 62-86) appears to act as a dominant switch for the Gsα- (activation) versus Giα2- (suppression) mediated control of phospholipase C and progression to primitive endoderm.

Embryonal carcinoma cells mimic the early embryo and have proven to be a useful model for study of development. These stem cells can be induced to differentiate in vitro into cell types that resemble those found at various stages of early mouse development (1). Progression of F9 teratocarcinoma embryonic stem cells can be divided into two separate events, the production of primitive endodermal and of parietal endodermal cells. When F9 cells are exposed to retinoic acid (RA) 1 alone, cells become flat and show typical endodermal morphology, which is the functional equivalent of primitive endoderm (PE) of normal embryogenesis, and express the protease tissue plasminogen activator (tPA, a PE marker) as well as components of the basal lamina, such as type IV collagen and laminin B1 (2). Elevation of intracellular cyclic AMP levels in F9 cells treated in combination with RA promote primitive endodermal cells to a parietal endoderm-like phenotype in early mouse embryogenesis (1).
Many complex biological processes such as oncogenesis, differentiation, and early neonatal mouse development have been shown to be regulated by heterotrimeric G-proteins, such as G s␣ and G i␣2 . In the F9 mouse teratocarcinoma model of early mouse development, G i␣2 levels decline precipitously in response to the morphogen RA as the cells commit to PE (3). Constitutive expression of RNA antisense to G i␣2 , but not to G i␣1 or G i␣3 , induces progression to PE in the absence of RA (4). Overexpression of G i␣2 , a G-protein that antagonizes G s␣ with respect to one known effector adenylylcyclase, blocks stem cells from progression to PE even in the presence of RA (4). Expression of G s␣ or its constitutively active mutant form induced PE in the absence of RA (5). Furthermore, the RA-induced decline in G i␣2 triggers F9 teratocarcinoma stem cells to progress to PE via activation of protein kinase C and mitogen-activated protein (MAP) kinase (6). These observations suggest that G i␣2 suppresses the commitment of stem cells to primitive endoderm, while G s␣ expression somehow eliminates the suppression and/or induces differentiation itself.
In an effort to identify the effector for the G i␣2 /G s␣ axis with respect to regulating development of PE from stem cells, we expressed chimera with N-terminal substitutions of G i␣2 in the G s␣ molecule. Using stably transfected clones of F9 cells expressing elevated levels of the chimeric G-proteins, we probed the downstream events to the level of MAP kinase and progression to PE. The results identify a central role of phospholipase C in the regulation by G i␣2 /G s␣ , via protein kinase C and MAP kinase activation. Additionally, substitution of region -(55-64) of G i␣2 for G s␣ and the C-terminal 36 amino acids of G i␣2 for the corresponding 38 C-terminal residues of G s␣ promotes progression to PE in the absence of the morphogen and identifies these regions as important determinants for regulation of phospholipase C.

EXPERIMENTAL PROCEDURES
Stable Expression of Chimeras in F9 Stem Cells-Wild-type and stably-transfected clones of F9 mouse teratocarcinoma stem cells were cultured on 0.1% gelatin-coated 10-cm Falcon Petri dishes in Dulbecco's modified Eagle's medium supplemented with 15% of fetal calf serum (Hyclone Laboratories). The expression vector pCW1 harboring cDNA constructs encoding either wild-type and constitutively activated mutants of G i␣2 and G s␣ or chimera of G i␣2 /G s␣ under the control of the SV40 early promoter were used to stably transfect F9 stem cells by calcium phosphate precipitation. Clones were selected (500 g/ml) and maintained (100 g/ml) in the presence of the active form of gentamicin analogue, antibiotic G418 sulfate (Life Technologies, Inc.).
Immunoblotting-Aliquots of crude membrane fractions (5 g) from each subclone were subjected to SDS-polyacrylamide gel electrophoresis (PAGE), the separated proteins were transferred to nitrocellulose, and the blots were stained with antibodies specific for either G s␣ (CM129) or to G i␣2 (CM112). The immune complexes were made visible by staining with a second antibody (goat anti-rabbit IgG) coupled to calf alkaline phosphatase.
Cyclic AMP Accumulation-Cyclic AMP accumulation was deter-* 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. mined by the competitive protein binding assay (7). Cell aliquots (10 5 ) were washed and incubated in Krebs-Ringer phosphate buffer (2 ϫ 10 6 cells/ml) with the indicated agents for 15 min at 37°C. The reaction was terminated by the addition of ethanol maintained at Ϫ70°C. Aliquots of cells were assayed in triplicate in either the absence or the presence of the ␤-adrenergic agonist (Ϫ)isoproterenol (100 M) or the diterpene activator of adenylcyclase, forskolin (50 M).
Determination of PE-The activity of tissue plasminogen activator (tPA) is the hallmark of the PE phenotype. Stem cells induced to PE produce and secrete tPA as well as assume a characteristic morphology, i.e. extended spindle shape with defined foci of growth (1). For production of PE, all-trans-RA (100 nM, Eastman Kodak) was added for 4 days to induce progression to PE (8). For tPA determinations, the culture medium of cells was assayed using the amidolytic assay (9). One unit of tPA is arbitrarily defined as that amount of tPA produced at a reaction rate of 10 Ϫ7 ⌬A 405 min Ϫ2 (change in the optical absorbance at 405 nm divided by the square of the time, in min). For phase-contrast microscopy, the cells were cultured for 4 days and then fixed with 3% (w/v) paraformaldehyde, viewed, and photographed by phase-contrast microscopy using a Zeiss Axiophot system.
[ 3 H]Inositol 1,4,5-Trisphosphate (IP 3 ) Mass Assay-Cells (10 6 ) were cultured in 6-well plates for 4 days in the presence and absence of RA (100 nM). Cells were harvested and subjected to treatment with perchloric acid. After neutralization, the mass of IP 3 was determined by a competitive protein binding assay of [ 3 H]IP 3 , using the IP 3 receptor present in rabbit cerebellar membranes (10). Binding assays contained 10 6 cells, 2.5 nM [ 3 H]IP 3 , and a 100-l aliquot of binding protein in 50 mM Tris-HCl, 1 mM EDTA, pH 8.3. Reaction tubes were incubated for 10 min at 4°C, followed by centrifugation at 10,000 ϫ g for 5 min. The resultant pellets were washed, suspended in 4 ml of scintillation fluid, and counted by liquid scintillation spectrometry.
Quantitation of 1,2-Diacylglycerol (DAG) Mass-Cells plated in 6-well plates (10 6 ) were grown to confluence for 4 days in the absence or presence of RA. Total cellular lipids were extracted into chloroform. DAG content was measured using the DAG kinase assay (11) employing thin-layer chromatography to separate radioactivity incorporated into phosphatidic acid generated from the reaction. The DAG mass was calculated from a standard curve with authentic sn-1,2-diacylglycerol.
Protein Kinase C Assay-Cells were grown to confluence in 10-cm dishes for 4 days, released by treatment with phosphate-buffered saline containing 1 mM EDTA, and homogenized on ice in buffer A (20 mM Tris, pH 7.5, 0.5 mM EDTA, 0.5 mM EGTA, 10 mM ␤-mercaptoethanol, 0.5% Triton X-100, and 25 g/ml each aprotinin and leupeptin). Homogenates were incubated on ice for 30 min, and supernatants were collected after microcentrifugation. Protein kinase C was partially purified by chromatography on a DEAE-cellulose column (0.2 g of DE52, Whatman) equilibrated with buffer B (20 mM Tris, pH 7.5, 0.5 mM EDTA, 0.5 mM EGTA, 10 mM ␤-mercaptoethanol) (12). The cell extract was applied to the column, which was then washed with buffer B. The protein kinase C was eluted in buffer C (buffer B plus 0.2 M NaCl). Protein kinase C activity was assayed using a commercially prepared assay (Life Technologies, Inc.) that utilizes a 50-l reaction mixture containing 20 mM Tris, pH 7.5, 20 mM MgCl 2 , 1 mM CaCl 2 , 20 M [␥-32 P]ATP, 50 M acylated myelin basic protein substrate-(4 -14) and activator phorbol myristic acid (10 mM) in combination with phosphatidylserine (0.28 mg/ml) or the presence versus absence of the pseudosubstrate inhibitor peptide, protein kinase C-(19 -36) (20 M) (13). Reactions were initiated by the addition of 5 l of the DEAE-purified homogenate and incubated at 30°C for 5 min. Twenty l of the reaction sample were spotted onto a P-81 phosphocellulose filter, which was then washed with 1% (v/v) phosphoric acid. Protein kinase C activity is expressed as picomoles of 32 P transferred to the myelin basic protein peptide substrate (4 -14)/ min/10 6 minus those transferred to the myelin basic protein in the presence of the pseudosubstrate inhibitor peptide, protein kinase C- (19 -36).
Resolution and Determination of MAP Kinase Activity-MAP kinase was assayed as described previously (14). Cells plated in 10-cm plates and grown for 4 days were harvested (5 ϫ 10 7 cells) and lysed (70 mM ␤-glycerophosphate, pH 7.2, 100 M sodium vanadate, 2 mM MgCl 2 , 1 mM EGTA, 0.5% Triton X-100, 5 g/ml leupeptin, 2 g/ml aprotinin, and 1 mM dithiothreitol). Soluble extracts (1 ml) were loaded onto a Mono-Q FPLC column (Pharmacia Biotech, Inc.) and eluted with a 28-ml linear 0 -350 mM NaCl gradient developed in 1 ml fractions. Twenty l aliquots of each fraction were mixed with 20 l of 50 mM ␤-glycerophosphate, pH 7.2, 100 M sodium vanadate, 20 mM MgCl 2 , 200 M [␥-32 P]ATP, 10 M protein kinase A inhibitory fragment (6 -22), 1 mM EGTA, and 400 M EGF receptor peptide EGFR -(662-681) for 15 min at 30°C. The reactions were terminated by adding 10 l of 25% trichloro-acetic acid. Forty-five l of the reaction mix were spotted onto P-81 phosphocellulose paper, which was washed with 75 mM phosphoric acid and counted by liquid scintillation spectrometry. Column fractions (9 volumes) were precipitated with 1 volume of 72% trichloroacetic acid (w/v), 0.15% sodium deoxycholate for 2 h at 4°C. The precipitate was washed twice with 1 ml of ice-cold acetone, dried, resuspended in 40 l of SDS-PAGE sample buffer, and boiled. Proteins from the column fractions were separated by 10% SDS-PAGE and then identified by immunoblotting with a monoclonal anti-MAP kinase antibody (Life Technologies, Inc.).

RESULTS AND DISCUSSION
Stable Expression of Chimeric G i␣2 /G s␣ Subunits-The cDNAs encoding the chimeras (Fig. 1) were inserted in the pCW1 expression vector driven by the SV40 early promoter and harboring the neomycin resistance gene. N-terminal substitutions of G s␣ by analogous sequences of G i␣2 were employed (15), as was the substitution of the C-terminal 38 residues of G s␣ with the corresponding C-terminal 36 residues of G i␣2 (16). Similarly, wild-type and the constitutively active G225T mutant of G s␣ and wild-type and the constitutively active Q205L mutant of G i␣2 were expressed also. Clones selected in the presence of the gentamicin analogue G418 displayed elevated expression of the chimera (Fig. 2). Immunoblots of chimeras with G s␣ C-terminal sequences were stained by a specific antibody raised against the decapeptide C terminus of G s␣ (CM112; Fig. 2, top panel). Chimera with G i␣2 sequences expressed in the C terminus were stained with antibody specific to the decapeptide C terminus of G i␣2 (CM129; Fig. 2, bottom panel). Expression of each of the chimeras ranged from 1-to 2-fold over endogenous levels of the G-protein subunits expressed by F9 clones stably transfected with the empty vector pCW1 alone.
Activation of Adenylylcyclase-The ability of the chimeras to activate adenylylcyclase was explored by measuring cyclic AMP accumulation of the clones in response to stimulation either by the ␤-adrenergic agonist isoproterenol or by the diterpene forskolin (Fig. 3). Confirming earlier studies on adenylylcyclase in other cells (15,16), expression of G s␣ and G s␣ G225T significantly enhanced cyclic AMP accumulation. Substitution of G s␣ sequence with G i␣2 (1-7) , G i␣2 (1-64) , G i␣2  , and G i␣2 (1-212) sequences attenuated the enhanced response observed with expression of either G s␣ or G s␣ G225T. Expression of G i␣2 (1-54)/s displayed levels of cyclic AMP accumulation, in contrast, similar to the G s␣ G225T mutant. Expression of either G i␣2 or Q205L G i␣2 inhibited by Ͼ50% cyclic AMP accumulation in response to either the ␤-adrenergic agonist isoproterenol or to the diterpene forskolin. The patterns for cyclic AMP accumulation by clones expressing the various subunits and chimeras were similar for the basal, isoproterenol-stimulated (100 M), and forskolin-stimulated (50 M) conditions.These data demonstrate that, when stably expressed in the context of F9 teratocarcinoma cells, G s␣ stimulates, G i␣2 inhibits, and substitution of G s␣ with increasing N-terminal regions of G i␣2 attenuates activation of adenylylcyclase. Thus, the G-proteins expressed have retained the ability to recognize receptor, bind and exchange GTP, and to activate one known effector, adenylylcyclase, as previously noted (15,16).
Progression of Stem Cells to PE-Secretion of tPA is a hallmark for stem cell differentiation to PE (1). Using the sensitive amidolytic assay (9), we measured the ability of clones expressing G s␣ , G s␣ G225T, and chimeras of G s␣ with increasing substitution of G i␣2 to progress to PE, both in the absence and presence of the morphogen RA (Fig. 4). Expression of either G s␣ G225T or wild-type G s␣ promoted progression to PE in the absence of morphogen, as evidenced by expression of the tPA marker. In combination with RA, expression of G s␣ markedly potentiated (3-fold), whereas expression of G i␣2 suppressed (Ͼ50%), the progression to PE heralded by tPA production. Substitution of G s␣ N-terminal sequence with increasing amounts of G i␣2 sequence attenuated the effect on progression. The markedly elevated response to G s␣ expression was reduced to that of the empty-vector control by substitution in the N terminus with as little as seven residues of G i␣2 and reduced progressively below the control by expression of G i␣2 (1-54)/s , G i␣2 (1-122)/s , and G i␣2 (1-212)/s chimeras. The G i␣2 (1-54)/s construct, which displayed the highest activation of adenylylcyclase (Fig. 3), strongly inhibited stem cell progression, as measured by tPA production.
Further analysis of the N and C termini of G s␣ by substitution with G i␣2 sequences identified two regions of the molecule critical to progression of stem cells to PE in either the absence (Fig. 5) or the presence (not shown) of RA. Although expression of G i␣2 (1-54)/s strongly activates adenylylcyclase (Fig. 3) and inhibits progression (Fig. 4), expression of G i␣2 (1-64)/s strongly promoted progression of stem cells to PE with respect to expression of G s␣ although both chimeras were equivalent with respect to activation of adenylylcyclase. Substitution of G s␣ C-terminal domain with that of G i␣2 , G s␣/I (320 -355) , enhanced the progression to PE in the absence of morphogen for G s␣ and the G i␣2 (1-64)/s , but not for either the G i␣2 (1-54)/s or G i␣2 (1-122)/s chimera (not shown).

FIG. 3. Stable transfection of F9 stem cells with G i␣2 /G s␣ chimeras and effects on cyclic AMP accumulation.
For cyclic AMP accumulation, cells were incubated in the absence or presence of isoproterenol (Iso, 100 M) or forskolin (Fsk, 50 M) for 15 min at 37°C in media supplemented with adenosine deaminase (0.1 g/ml). Cyclic AMP accumulation by the cells was assayed in extracts by the competitive protein binding assay (7). Data are the mean values Ϯ S.E. from five independent experiments. *, denotes p Յ 0.05 for difference from vector control values; **, denotes p Յ 0.01 for difference from vector control values.
Signaling via G-protein Chimeras-The inability of agents that elevate intracellular cyclic AMP accumulation, like cholera toxin, pertussis toxin, and forskolin, as well as the addition of dibutyryl cyclic AMP itself to induce differentiation of F9 stem cells (1)(2)(3)(4)(5) are consistent with the lack of correlation observed between activation of adenylylcyclase and promotion of PE observed herein. The differential effects of expression of G i␣2 (1-54)/s on cyclic AMP accumulation (strongly stimulated) versus tPA secretion (strongly inhibited) highlight this point.
The recent demonstration that suppression of G i␣2 leads to increased phospholipase C activity in a variety of cells (17), prompted us to evaluate the PLC response by measuring IP 3 and DAG generation (Fig. 6). Expression of G s␣ and G s␣ G225T increased IP 3 accumulation in the absence of morphogen. Substituting G i␣2 N-terminal sequence from 7 to 212 residues for G s␣ attenuated IP 3 accumulation. This trend was broken, however, by clones expressing the G i␣2 (1-64)/s chimera, which displayed nearly a 4-fold increase in IP 3 accumulation over that observed in stem cells transfected with the empty vector alone (Fig. 6, left panel). These stimulated levels of IP 3 generated in clones promoted to PE through expression of the G i␣2 (1-64)/s chimera are in excess of those in F9 cells promoted to PE by RA itself. Expression of the G s␣/I-(38) chimera substituted in the N-terminus with G i␣2 also promoted markedly enhanced PLC activity. The PLC activity of the clones expressing G s␣ , G225T G s␣, G i␣2 (1-64)/s , and G s␣i (320 -355) chimeras remained elevated even when the clones were induced with RA (Fig. 6, center  panel).
DAG levels, explored in the absence of morphogen, confirmed the observations derived from IP 3 mass measurements. Although equivalent with respect to adenylylcyclase activation (Fig. 3), the G i␣2 (1-54)/s and G i␣2 (1-64)/s chimeras can be differentiated by the ability of the latter to increase IP 3 accumulation, to increase DAG accumulation (Fig. 6, right panel), and to promote conversion of stem cells to PE in the absence of morphogen (Fig. 5).
Role of Protein Kinase C-Activation of PLC activity, evidenced by elevated IP 3 and DAG levels, implicates protein kinase C in the signaling by the chimera. Analysis of protein kinase C activity in the absence of morphogen revealed a pattern of stimulation of protein kinase C (Fig. 7, left panel) not unlike that of IP 3 and DAG accumulation (Fig. 6). Expression of G s␣ G225T and G s␣ itself markedly activated protein kinase C, while substitution of G i␣2 sequence in the N terminus of G s␣ attenuated the elevated activity. The G i␣2 (1-64)/s chimera, in contrast to the other members of the array of chimeras with increasing G i␣2 substitution, displayed activation of protein kinase C similar to that of G225 G s␣ , the constitutively active mutant of G s␣ . Expression of the G s␣i (320 -355) C-terminal substituted chimera also strongly activated protein kinase C. The rank order and relative scale for protein kinase C activation FIG. 4. Analysis of progression to PE (tPA production) by F9 cells stably expressing G s␣ chimeras with increasing N-terminal substitutions of G i␣2 . F9 stem cells stably transfected with empty vector (pCW1) and pCW1 harboring a cDNA encoding G s␣ , G225TG s␣ , and G s␣ chimeras with increasing N-terminal substitutions of G i␣2 were cultured for 4 days either in the absence (ϪMorphogen) or in the presence (ϩMorphogen) of RA (10 Ϫ7 M) and refed 17 h prior to sampling of tPA secretion into the conditioned media. The activity of tPA, a marker protein for the differentiated phenotype, was measured by the amidolytic assay (9). The tPA production was normalized to the cell number. A unit of tPA activity is arbitrarily defined as 10 Ϫ7 ⌬A 405 min Ϫ2 . Data are the mean values Ϯ S.E. from five independent experiments. By the Student's t test, significance of the difference from vector control values yields p Յ 0.05 (denoted with *) or p Յ 0.01 (denoted with **).
FIG. 5. Analysis of progression to PE (tPA production) by F9 cells stably expressing G s␣ chimeras with C-and N-terminal substitutions of G i␣2 . F9 stem cells stably transfected with empty vector (pCW1) and pCW1 harboring a cDNA encoding G s␣ and G s␣ chimeras with C-and N-terminal substitutions of G i␣2 were cultured for 4 days either in the absence (ϪMorphogen) or in the presence (ϩMorphogen) of RA (10 Ϫ7 M) and refed 17 h prior to sampling of tPA secretion into the conditioned media. The activity of tPA, a marker protein for the differentiated phenotype, was measured by the amidolytic assay (9), as described in the legend to Fig. 4. The tPA production was normalized to the cell number. The values of 10 Ϫ7 ⌬A 405 min Ϫ2 in RA-treated cells were compared with that of untreated wild-type F9 stem cells to determine the fold-increase in the tPA activity. Data are the mean values Ϯ S.E. from five independent experiments. *, denotes p Յ 0.05 for difference from vector control values; **, denotes p Յ 0.01 for difference from vector control values. and progression to PE, measured by tPA secretion (Fig. 7, right  panel), is remarkably similar.
MAP Kinase Activation-The MAP kinase regulatory network is an important conduit for signaling from the cell surface to the nucleus. We explored MAP kinase activity only in clones expressing chimeras that induce progression to PE in the absence of retinoic acid (Fig. 8). The peak values of MAP kinase activity were resolved on Mono-Q FPLC separations and assayed with the EGF receptor peptide. In close agreement with the effects of expression of the G s␣ and G i␣2 (1-64)/s chimeras uniquely on protein kinase C and differentiation, MAP kinase was strongly activated. Expression of the constitutively active form of G i␣2 Q205L, providing a useful control, inhibits MAP kinase activation, much like it inhibited PLC activation, and the activation of protein kinase C and progression to PE. Immunoblotting of fractions from Mono-Q FPLC reveal multiple, phosphorylated forms of MAP kinase in extracts from the cells stably expressing the G i␣2 (1-64)/s chimera that progress to PE in the absence of RA as compared with the MAP kinase forms in the F9 stem cells (Fig. 8B). These changes in MAP kinase activity and profiles on FPLC separations were noted previously in a comparison of extracts prepared from F9 stem cells with cells promoted to PE by treatment with RA (6). The similarities between the rank orders for protein kinase C activation (Fig. 7, left panel), MAP kinase activity (Fig. 8), and progression to PE (Fig. 7, right panel) implicate both protein kinase C and MAP kinase in the regulation of stem cell conversion to PE.
Control of Stem Cell Progression via the G s␣ /G i␣2 Axis-The progression of F9 stem cells to PE is induced by the morphogen RA. Early gene products include G s␣ (18), ERA1/Hox1.6 (19), and RAR␣,␤ (20,21), which are synthesized within the first day of induction by RA. G i␣2 expression falls precipitously (3), early in RA-induced progression. Mimicking the decline in G i␣2 via expression of antisense RNA promotes progression to PE in the absence of RA, while constitutive expression of G i␣2 blocks the ability of RA to induce progression (4). Likewise, enhanced expression of either wild-type G s␣ or the G225T-activated mutant provokes progression in the absence of RA, mimicking the FIG. 6. Analysis of phospholipase C activity (IP 3 and DAG accumulation) in F9 cells stably expressing G s␣ , G i␣2 , activated mutants, and chimeras with C-and N-terminal substitutions of G i␣2 . To measure IP 3 accumulation, cells (10 6 ) were cultured in 6-well plates for 4 days either in the absence (ϪMorphogen) or presence (ϩMorphogen) of RA and then terminated by the addition of perchloric acid. After neutralization, the mass of IP 3 was determined by a competitive protein binding assay of [ 3 H]IP 3 using rabbit cerebellar membranes (10). For measurement of DAG levels, cells (10 6 ) were grown to confluence for 4 days and then extracted into chloroform. DAG was measured using a DAG kinase assay (11), employing thin-layer chromatography to separate radioactivity incorporated into phosphatidic acid generated from the reaction. Data are the mean values Ϯ S.E. from three independent experiments. *, denotes p Յ 0.05 for difference from vector control values; **, denotes p Յ 0.01 for difference from vector control values.

FIG. 7. Analysis of PKC activity of F9 cells stably expressing G s␣ chimeras with increasing N-terminal substitutions of G i␣2 .
Protein kinase C activity was measured in DEAE-purified cell homogenate (12) in the presence of [␥-32 P]ATP, acylated myelin basic protein substrate peptide substrate-(4 -14) (13), using a commercially-available protein kinase C assay (Life Technologies, Inc.). Protein kinase C activity is expressed as picomoles of 32 P transferred to the myelin basic protein peptide substrate/min/10 6 cells. In all cases, protein kinase C activity is defined as that activity sensitive to inhibition by the pseudosubstrate inhibitor peptide-(19 -36) of protein kinase C. The tPA production was measured as noted in the legend to Fig. 4. Data are the mean values Ϯ S.E. from three independent experiments. *, denotes p Յ 0.05 for difference from vector control values; **, denotes p Յ 0.01 for difference from vector control values. rise in G s␣ induced early by RA (18). G i␣2 has been shown to tonically inhibit PLC activity, and elimination of G i␣2 de-represses basal PLC as well as potentiates hormonal stimulation of PLC (17). Recently, we have shown that the decline in G i␣2 triggers activation of PLC, protein kinase C, and MAP kinase as stem cells are induced to PE (6). The G s␣ /G i␣2 axis is shown in the present work to control stem cell progression via this pathway. Using activated mutants of these G-protein ␣-subunits and chimeras with various N-and C-terminal substitutions of G i␣2 in G s␣ revealed several important insights. Comparison of the effects of expressing the activated mutants and chimeras on progression with those on cyclic AMP accumulation clearly eliminates adenylylcyclase as the effector for progression, although it is the most prominent effector controlled by the G s␣ /G i␣2 axis.
Based upon the concordance of the data for the activation of downstream elements by the activated subunits and chimeras, the target for regulation of progression appears to be PLC. Activation of PLC and protein kinase C by G s␣ chimeras with N-terminal sequence of G i␣2 demonstrated high sensitivity to even the smallest substitution, G i␣2 (1-7) /G s␣ (Figs. 6, 7). Substitution of G i␣2 sequence at the C terminus, in contrast, enhanced the activation of PLC, protein kinase C, and MAP kinase, and progression to PE. Unexpectedly, substitutions in the N-terminal region of G s␣ residues-(62-86) (corresponding to G i␣2 sequence-(55-64)) were found to have a profound effect on PLC regulation, suggesting that this region of the G s␣ molecule, when aligned with the crystal structure of other heterotrimeric G-protein ␣-subunits (22,23), is involved in regulation of PLC (Fig. 9, highlighted in blue). This region of the molecule is the site of the insertion of the intervening peptide sequence GGEEDPQAARSNSDG that distinguishes the long and short forms of G s␣ , a product of differential splicing of mRNA (24). Although speculation, this intervening sequence may act as an inhibitor of PLC activation in the long form of G s␣ , providing an explanation for the profound increase in PLC activation occurring in the G i␣2 (1-64)/s chimera that lacks this short region (14 -16). The -(62-86) region of G s␣ is also topographically distinct from the region (Fig. 9, highlighted in yellow) implicated in the control of adenylylcyclase (25), as well as from the regions on the aligned sequence of G q␣ (␣-helix 3) recently shown to be involved in PLC activation, as deduced by alanine scanning mutagenesis (26). More detailed analysis of this interesting region of G s␣ and its influence on PLC activation will be required. Our results do suggest that signaling via PLC, regulated via the G s␣ /G i␣2 axis, is critical to induction of stem cell progression to PE, and presumably mammalian embryogenesis.
FIG. 8. Analysis of MAP kinase activity of F9 cells stably expressing G s␣ chimeras with C-and N-terminal substitutions of G i␣2 . Panel A, MAP kinase (MAPK) activity was measured after separation on Mono-Q FPLC. Cells (5 ϫ 10 7 ) cultured for 4 days were harvested and lysed. Soluble extracts (1 ml) were fractionated onto a Mono-Q FPLC column with a 28-ml linear 0 -350 mM NaCl gradient, collecting 1-ml fractions. The enzyme activity was quantified by measurement of the phosphorylation of EGFR -(662-681) peptide, which contains the consensus MAP kinase phosphorylation site, PXTP. The data represent the peak of MAP kinase activity of the p42 form. Panel B, immunoblot analysis of Mono-Q FPLC fractions obtained from F9 stem cells (Stem) and clones expressing the G i␣2 (1-64)/ G s␣ chimera that progress to PE in the absence of RA. Column fractions were precipitated with trichloroacetic acid/sodium deoxycholate, and subjected to SDS-PAGE on 10% gels. The blots were probed with a monoclonal antibody to MAP2 kinase. The blots reveal the multiple phosphorylated forms of MAP kinase for the cells expressing the G i␣2 (1-64)/ G s␣ chimera that progress to PE in the absence of RA compared with the forms observed in the F9 stem cells. FIG. 9. Mapping of region -(62-86) of G s␣ onto the deduced crystal structure of G i␣1 and identification of regions regulating adenylylcyclase (yellow) and phospholipase C (blue). The landmarks for G s␣ were projected upon the deduced crystal structure of G i␣1 in which the space-filling model is initially red, with the region participating in the control of stem cell progression via phospholipase C colored blue and the region controlling adenylylcyclase colored yellow.