Dissociation of cAMP-stimulated mitogenesis from activation of the mitogen-activated protein kinase cascade in Swiss 3T3 cells.

Elevation of intracellular cAMP by forskolin, 8-bromoadenosine 3′:5′-cyclic monophosphate, and prostaglandin E1, in synergy with insulin, stimulated DNA synthesis in quiescent Swiss 3T3 cells to the same level achieved by platelet-derived growth factor (PDGF) or bombesin. Both forskolin and 8-bromoadenosine 3′:5′-cyclic monophosphate stimulated a significant increase in cell number which, in the presence of insulin, reached the same levels achieved with PDGF. Treatment with either PDGF or bombesin caused a marked and persistent stimulation of p42MAPK and p44MAPK. In striking contrast, no activation was seen with mitogenic combinations of cAMP as shown by three different assays. Swiss 3T3 cells stably transfected with a constitutively activated Gsα subunit were 100-fold more sensitive to the mitogenic effects of forskolin but in this distinct cellular model forskolin did not activate p42MAPK. Swiss 3T3 cells stably transfected with interfering mutants of MEK-1 showed a 60% decrease in PDGF-stimulated p42 MAPK activation, but there was no inhibition of the mitogenic effect of forskolin in these cells. Furthermore, the upstream kinases MEK-1/MEK-2 and p74raf−1 were not activated by mitogenic combinations of cAMP while PDGF caused marked stimulation of their activity. Treatment of 3T3 cells with forskolin attenuated PDGF-stimulated p74raf−1 and p42MAPK activation but enhanced the mitogenic effects of this agent. Mitogenic combinations of cAMP strongly stimulated the phosphorylation and activation of p70s6k an effect that was inhibited by rapamycin. This agent markedly inhibited cAMP-stimulated DNA synthesis suggesting a critical role for p70s6k in cAMP mitogenic signaling. These results demonstrate that cAMP-induced mitogenesis can be dissociated from activation of the mitogen-activated protein kinase cascade and that this is not an obligatory point of convergence in mitogenic signaling in Swiss 3T3 cells.

Elevation of intracellular cAMP by forskolin, 8-bromoadenosine 3:5-cyclic monophosphate, and prostaglandin E 1 , in synergy with insulin, stimulated DNA synthesis in quiescent Swiss 3T3 cells to the same level achieved by platelet-derived growth factor (PDGF) or bombesin. Both forskolin and 8-bromoadenosine 3:5cyclic monophosphate stimulated a significant increase in cell number which, in the presence of insulin, reached the same levels achieved with PDGF. Treatment with either PDGF or bombesin caused a marked and persistent stimulation of p42 MAPK and p44 MAPK . In striking contrast, no activation was seen with mitogenic combinations of cAMP as shown by three different assays. Swiss 3T3 cells stably transfected with a constitutively activated Gs␣ subunit were 100-fold more sensitive to the mitogenic effects of forskolin but in this distinct cellular model forskolin did not activate p42 MAPK . Swiss 3T3 cells stably transfected with interfering mutants of MEK-1 showed a 60% decrease in PDGF-stimulated p42 MAPK activation, but there was no inhibition of the mitogenic effect of forskolin in these cells. Furthermore, the upstream kinases MEK-1/MEK-2 and p74 raf-1 were not activated by mitogenic combinations of cAMP while PDGF caused marked stimulation of their activity. Treatment of 3T3 cells with forskolin attenuated PDGFstimulated p74 raf-1 and p42 MAPK activation but enhanced the mitogenic effects of this agent. Mitogenic combinations of cAMP strongly stimulated the phosphorylation and activation of p70 s6k an effect that was inhibited by rapamycin. This agent markedly inhibited cAMP-stimulated DNA synthesis suggesting a critical role for p70 s6k in cAMP mitogenic signaling. These results demonstrate that cAMP-induced mitogenesis can be dissociated from activation of the mitogen-activated protein kinase cascade and that this is not an obligatory point of convergence in mitogenic signaling in Swiss 3T3 cells.
The mitogen-activated protein (MAP) 1 kinases (ERKs) are a family of highly conserved serine/threonine kinases that are activated in response to a wide range of extracellular signals including growth factors, hormones, and neuropeptides (1)(2)(3). The two best characterized isoforms p42 MAPK (ERK-2) and p44 MAPK (ERK-1) (4) can be activated through both tyrosine kinase receptors or G-protein-linked receptors (1)(2)(3). Once activated, p42 MAPK and p44 MAPK phosphorylate an array of cellular proteins including protein kinases such as p90 rsk (5), transcription factors (6 -8), and proteins involved in the regulation of cell growth (9). MAP kinases are themselves activated by phosphorylation on specific threonine and tyrosine residues by the dual-specificity MAP kinase kinase (or MEK) of which at least two isoforms have been identified in mammalian cells (10 -12). This kinase is itself regulated by upstream kinases including the Raf family (13) and MEK kinase (14). Studies with dominant-negative and activating mutants have provided evidence that this pathway can lead to the stimulation of DNA synthesis (15,16). However, it is unclear whether the activation of p42 MAPK and p44 MAPK is a point of convergence in the action of all signals that promote DNA synthesis (17)(18)(19).
The cAMP-protein kinase pathway links a number of extracellular signals to a range of cell functions including cell proliferation (20). Considerable evidence indicates that an increase in intracellular cAMP can act as a mitogenic signal for Swiss 3T3 cells (21). Agents that promote cAMP production and accumulation, such as forskolin and PGE 1 , as well as permeable cAMP analogues, stimulate DNA synthesis in 3T3 cells acting synergistically with insulin and other factors (21)(22)(23)(24). Cells expressing a mutated cAMP-protein kinase regulatory subunit show markedly reduced cAMP-protein kinase activation and mitogenesis in response to agents that elevate intracellular cAMP (25). Conversely, Swiss 3T3 cells expressing a constitutively active Gs␣ subunit are highly sensitive to the mitogenic effects of cAMP elevating agents (26). Increases in cAMP also lead to early intracellular events associated with cell proliferation including an increase in the expression of the proto-oncogene c-myc. (27). However, the exact relationship between the mitogenic effect of cAMP and activation of the MAP kinase cascade is as yet not defined.
Here we report that the mitogenic effects of cAMP are not associated with detectable activation of p42 MAPK and p44 MAPK , MEK-1/-2, or of p74 raf-1 . Interfering mutants of MEK-1 stably transfected into Swiss 3T3 cells significantly inhibited PDGFstimulated p42 MAPK activation but did not inhibit cAMP-induced mitogenesis. Further dissociation of mitogenesis from the MAPK cascade is demonstrated by the finding that elevating intracellular cAMP inhibits PDGF-stimulated p74 raf-1 and p42 MAPK activation but enhances PDGF-stimulated DNA synthesis. Mitogenic combinations of cAMP strongly stimulated the phosphorylation and activation of p70 s6k , an effect that was * 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  Cell Proliferation-Confluent, quiescent Swiss 3T3 cells were treated with various factors added directly to the culture medium. After 72 h, the cultures were washed twice with 0.02% EDTA in phosphate-buffered saline and the cells removed with 0.25% trypsin in Tris-buffered saline and cell number of the resulting suspension determined using a Coulter counter.
In-gel Kinase Assay-Quiescent cells in 33-mm dishes were washed twice with DMEM, treated with various factors in 2 ml of DMEM for designated time periods as indicated, and then lysed in SDS-PAGE sample buffer. Lysates were then subjected to a kinase assay in SDSpolyacrylamide gels using a modification of the method described by Kameshita and Fujisawa (28). Briefly, samples were subjected to SDS-PAGE in 10% polyacrylamide minigels containing 0.5 mg/ml MBP. After electrophoresis, SDS was removed from the gels by three 20-min washes with 20% (v/v) propan-2-ol in 50 mM Tris-HCl (pH 8.0) followed by three 20-min washes with 5 mM ␤-mercaptoethanol in 50 mM Tris-HCl (pH 8.0). Proteins were denatured by two 30-min washes with 6 M guanidine hydrochloride in 50 mM Tris-HCl (pH 8.0) and then renatured by incubation at 4°C in 5 changes of 50 mM Tris-HCl containing 0.04% (v/v) Tween-40 and 5 mM ␤-mercaptoethanol over 12-18 h. After preincubation of the gels at room temperature for 1 h in 40 mM HEPES (pH 8.0), 2 mM dithiothreitol, 10 mM MgCl 2 in gel phosphorylation of MBP was performed in 40 mM HEPES (pH 8.0), 0.5 mM EGTA, 10 mM MgCl 2 , 2 M cAMP-dependent protein kinase inhibitor peptide, 50 M ATP, 2.5 Ci/ml of [␥-32 P]ATP, for 1 h at room temperature. After extensive washing in 5% trichloroacetic acid (w/v) with 1% (w/v) sodium pyrophosphate, the gels were dried and autoradiographed. The same amount of the cell lysates used in this assay were subjected to SDS-PAGE and Western blotting for p42 MAPK to confirm equal loading of this protein in the in-gel kinase assay.
Shift Assay for MAP Kinase Activation-Activation of p42 MAPK and p44 MAPK was determined by the appearance of slower migrating forms in gel electrophoresis due to phosphorylation of specific threonine and tyrosine residues (29). Lysates from quiescent cells in 33-mm dishes prepared as above were subjected to SDS-PAGE and transferred to Immobilon membranes. Membranes were blocked using 5% nonfat dried milk in phosphate-buffered saline. Rabbit polyclonal antibodies raised against COOH-terminal peptides (EETARFQPGYRS for p42 MAPK and IFQETARFQPGAPE for p44 MAPK ) were used at 1/1000 dilution, and 125 I-protein A was used to visualize immunoreactive bands.
Immune Complex Assay for p42 MAPK Activation-Quiescent cells were treated with factors as above and lysed at 4°C in 1 ml of a solution containing 10 mM Tris-HCl, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 100 M Na 3 VO 4 , 1% Triton X-100, and 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin (pH 7.6, lysis buffer). Lysates were clarified by centrifugation at 15,000 ϫ g for 10 min at 4°C. Immunoprecipitation was performed using the polyclonal anti-p42 MAPK antibody as above incubating the samples on a rotating wheel for 2 h. Washed protein A-agarose beads (50 l 1:1 slurry) were added for the second hour. Immune complexes were collected by centrifugation and washed twice in lysis buffer and twice in kinase buffer (15 mM Tris-HCl, 15 mM MgCl 2 ). The kinase reaction was performed by resuspending the pellet in 25 l of kinase assay mixture containing kinase buffer with 0.5 mM EGTA, 1 mg/ml MBP-peptide (APRTPGGRR), 50 M ATP, 50 Ci/ml of [␥-32 P]ATP, 2 M cAMP-dependent protein kinase inhibitor peptide, and 100 nM microcystine LR. Incubations were performed under linear assay conditions at 30°C and, following centrifugation for 10 s, terminated by spotting 25 l of the supernatant onto Whatman P81 chromatography paper. Filters were washed four times for 5 min in 0.5% orthophosphoric acid, immersed in acetone, and dried before scintillation counting. The average radioactivity of two blank samples containing no immune complex was subtracted from the result of each sample. Results are expressed as cpm/1.5 ϫ 10 6 cells. The specific activity of [␥-32 P]ATP used was 900-1200 cpm/pmol.
Transfection of Interfering Mutants of MEK-1-MEK-1 mutants with alanine substitutions at serine 217 or serine 221 were used as interfering mutants to block MAP kinase activation in vivo (15). Wild type and MEK-1 mutants in the retroviral expression vector pBABEpuro (30) were transfected into Swiss 3T3 cells by co-culturing with the retrovirus producer cell line GPϩE pretreated with mitomycin C to render cells non-viable (30). Resistant clones were selected using puromycin (5 g/ml). Expression of MEK-1 was determined by Western blotting with a specific anti-MEK-1 monoclonal antibody. All clones selected for subsequent experiments had comparable levels of expression of exogenous MEK-1. Results shown are for one each of the wild type (WT), alanine 217 (Ala 217 ) and alanine 221 (Ala 221 ) mutant expressors but are typical of those obtained with two independent clones for each transfectant.
p74 raf-1 Kinase Assay-p74 raf-1 activity was assayed using a modification of the method described by Alessi et al. (31) using GST fusion proteins of MEK and p42 MAPK as substrates. Overnight cultures of Escherichia coli strain BL21 DE3 transformed with GST-p42 MAPK and GST-MEK expression vectors (pGEX-2T) were diluted 1 in 10 and grown for 1 h. GST-p42 MAPK was induced with 1 mM isopropyl-1-thio-␤-D-galactopyranoside for 4 h at 37°C and GST-MEK induced with 30 M isopropyl-1-thio-␤-D-galactopyranoside at 27°C overnight. The cells were then pelleted, freeze-thawed, and lysed. Cell debris was removed by centrifugation, and GST fusion proteins were purified by adding 0.5 ml of glutathione-Sepharose beads and rotating for 30 min at 4°C. The GST-p42 MAPK was cleaved from the GST in thrombin buffer while the GST-MEK was eluted from the glutathione-Sepharose beads with 5 mM glutathione in 50 mM Tris (pH 8). Both preparations were then dialyzed and concentrated. The purity of each preparation was checked by subjecting the proteins to SDS-PAGE and staining the gels with Coomassie Blue. For the GST-p42 MAPK typical yields were 10 g/ml of culture, with a purity of Ͼ95%. For the GST-MEK typical yields were 5 g/ml of culture with a purity of Ͼ95%. For the kinase assay quiescent cells were treated as indicated and lysed in lysis buffer as above. p74 raf-1 immunoprecipitation was performed with an affinity purified rabbit polyclonal antibody for 2 h with protein A-agarose added for the second hour. Immune complexes were collected by centrifugation and then washed three times in lysis buffer with no phenylmethylsulfonyl fluoride and once with buffer A (50 mM Tris-HCl (pH 7.5), 0.1 mM EGTA, 0.5 mM Na 3 VO 4 , and 0.1% ␤-mercaptoethanol). Pellets were then resuspended in 30 l of MEK/p42 MAPK buffer (30 mM Tris-HCl (pH 7.5), 0.1 mM EGTA, 0.1% ␤-mercaptoethanol, 6.5 g/ml GST-MEK, 100 g/ml GST-p42 MAPK , 0.03% Brij-35, 10 mM Mg-ATP, and 20 mM N-octyl-␤-Dglucopyranoside) and incubated at 30°C for 30 min. The reaction was then terminated by diluting the supernatant in 40 l of buffer A with 1 mg/ml of BSA and after mixing 10 l of the supernatant was removed to a fresh tube. p42 MAPK activation was then measured using the MBP peptide phosphorylation assay essentially as above.
MEK Assay-Activation of MEK-1 and MEK-2 was assayed using a modification of the above technique. Lysates from treated cells were incubated with a rabbit polyclonal antibody which recognized both MEK-1 and MEK-2. Immune complexes were incubated with GST-p42 MAPK as above, and the resultant activation of this enzyme was measured as before.
p70 s6k Immune Complex Kinase Assay-Quiescent cells were treated with various factors as above and lysed at 4°C for 20 min in 1 ml of a solution containing 50 mM Tris-HCl, 5 mM EDTA, 100 mM NaCl, 40 mM ␤-glycerophosphate, 50 mM NaF, 1 mM Na 3 VO 4 , 1% Triton X-100, and 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin (pH 7.6, lysis buffer). Lysates were clarified by centrifugation at 15,000 ϫ g for 10 min at 4°C. Immunoprecipitation was performed using a rabbit polyclonal anti-p70 s6k antibody (directed against a synthetic peptide corresponding to amino acids 2-30 of human p70 s6k ) for 4 h with with protein A-agarose added for the last hour. The immune complexes were washed three times in lysis buffer and once with S6 kinase buffer (20 mM HEPES (pH 7.4), 10 mM MgCl 2 , 1 mM dithithreitol, and 10 mM ␤-glycerophosphate). The kinase reaction was performed by resuspending the pellet in 25 l of kinase assay mixture containing kinase buffer with 0.2 mM S6 peptide (RRRLSSLRA), 20 M ATP, 5 Ci/ml of [␥-32 P]ATP, 2 M cAMP-dependent protein kinase inhibitor peptide and 100 nM microcystine LR. Incubations were performed under linear assay conditions at 30°C for 20 min and, following centrifugation for 10 s, terminated by spotting 25 l of the supernatant onto Whatman P81 chromatography paper. Filters were washed four times for 5 min in 0.5% orthophosphoric acid, immersed in acetone, and dried before scintillation counting. The average radioactivity of two blank samples containing no immune complex was subtracted from the result of each sample.
p70 s6k Mobility Shift Assay-Activation of p70 s6k was determined by the appearance of slower migrating forms in SDS-PAGE as a result of phosphorylation on several clustered serine and threonine residues (32). Immunoblot analysis on cell lysates was performed using a rabbit polyclonal antibody which recognized both ␣I and ␣II isoforms of p70/ 85 s6k (33).
Materials-Forskolin, 8-BrcA, IBMX, PGE 1 , bombesin, puromycin, myelin basic protein, and insulin were obtained from Sigma. Protein A-agarose conjugate was obtained from Boehringer Mannheim, Germany. cAMP-dependent protein kinase inhibitor peptide was obtained from Bachem (U.K.) Ltd., Saffron Walden, United Kingdom. Rapamycin was obtained from Calbiochem-Novabiochem (U.K.) Ltd., Nottingham, U.K. The anti-p44 MAPK , anti-p70 s6k , and anti-p74 raf-1 affinity purified rabbit polyclonal antibodies were obtained from Santa Cruz Biotechnology Ltd. Anti-MEK-1 monoclonal antibody and anti-MEK-1/MEK-2 polyclonal antibody was obtained from Affiniti Research Products Ltd, Nottingham, U.K. The N terminally directed anti-p70 s6k rabbit polyclonal antibody was obtained from Upstate Biotechnology Inc. The anti-p42 MAPK antibody was a kind gift from Dr J. Van Lint, Katholieke Universiteit Leuven, Belgium. GST-MEK and GST-p42 MAPK expression vectors and MEK-1 mutants in pBABEpuro were kind gifts from Professor C.  Fig. 1A, the addition of PDGF, which acts through a receptor with intrinsic tyrosine kinase activity or bombesin, which acts via a G-protein-linked receptor, induced striking stimulation of [ 3 H]thymidine incorporation in these cells, the effect of bombesin being further enhanced by the presence of insulin. Forskolin, a direct activator of adenylyl cyclase that increases intracellular cAMP in Swiss 3T3 cells and 8-BrcA, a cellpermeable cAMP analogue, also induced DNA synthesis in combination with insulin. As shown in Fig. 1A both these agents stimulated DNA synthesis to a level comparable to that achieved with PDGF or bombesin confirming that an increase in cAMP is a potent mitogenic signal for Swiss 3T3 cells.
To assess whether treatment of Swiss 3T3 cells with cAMPelevating agents resulted in cell proliferation, cultures of 3T3 cells in conditioned medium were treated for 72 h with the various factors and cell number determined. As shown in Table  I addition of either forskolin or 8-BrcA alone both caused a statistically significant increase in cell number compared to control cultures. In the presence of insulin, the stimulatory effect of forskolin was comparable to that achieved with PDGF.
To examine the effects of these mitogenic factors upon the activity of p42 MAPK and p44 MAPK , cell lysates from parallel cultures were analyzed by Western blotting using specific polyclonal antibodies to these proteins. Activation was determined by the appearance of slower migrating forms which results from the phosphorylation of specific threonine and tyrosine residues in these kinases (29). Treatment of cells with PDGF or bombesin, either alone or in combination with insulin, stimu-lated p42 MAPK as judged by the mobility shift (Fig. 1B, upper  panel). In contrast, no slower migrating form was seen with the mitogenic combinations of either forskolin or 8-BrcA with insulin or with insulin alone. Similar findings were obtained with p44 MAPK (data not shown).
The activation of p42 MAPK and p44 MAPK in response to mi-  14). B, upper panel, parallel, confluent, and quiescent cultures of Swiss 3T3 cells were washed and treated for 5 min at 37°C with factors as above, lysed in SDS sample buffer, and analyzed by Western blotting with anti-p42 MAPK polyclonal antibody. The positions of non-phosphorylated p42 MAPK and the slower migrating phosphorylated form pp42 MAPK are indicated. B, lower panel, confluent and quiescent cultures of Swiss 3T3 cells were washed and treated for 5 min at 37°C with factors as above and lysed in sample buffer. Samples were analyzed using in situ phosphorylation of MBP by renaturable kinases following SDS-PAGE as described under "Experimental Procedures." The positions of p42 MAPK and p44 MAPK are indicated. C, confluent and quiescent cultures of Swiss 3T3 cells were washed and treated for 5 min at 37°C with factors as above, lysed in lysis buffer, immunoprecipitated with anti-p42 MAPK polyclonal antibody, and the immune complexes analyzed in an immune complex kinase assay using MBP peptide as a substrate (see "Experimental Procedures"). Results are expressed as cpm/1.5 ϫ 10 6 cells, and the data are shown as the mean Ϯ S.E. for four independent experiments each performed in duplicate. The specific activity of [␥-32 P]ATP used was 900-1200 cpm/pmol. togens was further examined using in situ phosphorylation of MBP by renaturable kinases following SDS-PAGE. Fig. 1B, lower panel, shows that bombesin and PDGF stimulated MBP kinases of molecular masses 42 and 44 kDa corresponding to p42 MAPK and p44 MAPK . No activation of these MBP kinases was seen in response to the mitogenic combinations of either forskolin or 8-BrcA with insulin or with insulin alone. No additional MBP kinases of higher or lower molecular weight were activated by increasing intracellular cAMP (data not shown). This has been further corroborated by immune complex kinase assays using anti-p42 MAPK immunoprecipitates from cells treated with each of the mitogens. Fig. 1C shows that while PDGF and bombesin, either alone or in combination with insulin, stimulated p42 MAPK activity, the cAMP elevating agents in combination with insulin did not induce a significant increase in p42 MAPK activation. The results depicted in Fig. 1 suggest a striking dissociation of p42 MAPK and p44 MAPK activation (as assessed by three distinct assay techniques) from the mitogenic effects of cAMP, a finding that is in marked contrast with the activation stimulated by PDGF and bombesin.
cAMP Does Not Activate p42 MAPK and p44 MAPK Over Short and Long Time Courses-The demonstration that, unlike PDGF and bombesin, mitogenic combinations of cAMP elevating agents did not activate p42 MAPK and p44 MAPK after 5 min of stimulation led us to examine the possibility that they might stimulate these kinases at earlier or later times. Fig. 2A, upper panel, shows that in the gel mobility shift assay bombesin stimulated p42 MAPK activation as early as 2 min and that persistent activation was seen with incubations as long as 4 h. In contrast, forskolin and insulin did not stimulate a shift over a time range from 2 min to 12 h (Fig. 2B, lower panel). Similar results were obtained in the in-gel MBP-kinase assay which demonstrated similar kinetics for p42 MAPK and p44 MAPK activation in response to bombesin but an absence of activation of these or other MBP kinases by the combination of forskolin and insulin (data not shown).
To further assess the kinetics of p42 MAPK activation, immune complex kinase assays were performed over a time course of up to 120 min. As shown in Fig. 2B, bombesin stimulated a persistent activation in p42 MAPK activity beyond 120 min while forskolin and insulin did not induce a significant increase in activity over the same time period compared to untreated control cells. PGE 1 Is a Potent Mitogen but Does Not Activate p42 MAPK -To investigate the possibility that a receptor-mediated increase in intracellular cAMP, as distinct from that triggered by forskolin or 8-BrcA, might activate p42 MAPK , Swiss 3T3 cells were treated with PGE 1 at a dose of 50 ng/ml together with IBMX and insulin. This combination stimulates intracel-lular cAMP accumulation but does not promote Ca 2ϩ mobilization in these cells (24). 2 Fig. 3A shows that PGE 1 stimulated DNA synthesis to a level comparable to that seen with bombesin. However, in the gel mobility shift assay, no activation of p42 MAPK was seen over an extensive time course (Fig. 3B), and in immune complex kinase assays no stimulation of this enzyme was detected after 5-min incubations with mitogenic combinations of PGE 1 whereas bombesin again stimulated p42 MAPK activity (Fig. 3C).
Effect of Forskolin on Swiss 3T3 Cells Transfected with a Constitutively Activated Gs␣ Subunit-To further explore the dissociation between cAMP-induced mitogenesis and p42 MAPK activation an additional cellular model was employed. Mutational replacement of glutamine-227 with a leucine residue in the GTP-binding domain of the ␣ subunit of Gs (Q227L ␣ s ) reduces its ability to hydrolyze GTP and causes constitutive activation of the mutant protein (34). Swiss 3T3 cells expressing Q227L ␣ s have an increased sensitivity to the effects of forskolin upon intracellular cAMP accumulation and mitogenesis (26). Fig. 4A shows that half-maximal DNA synthesis was achieved with 10 M of forskolin (without IBMX) in wild type Swiss 3T3 cells whereas in Swiss 3T3 cells expressing Q227L ␣ s this level of stimulation of DNA synthesis was obtained with 2 D. J. Withers, S. R. Bloom, and E. Rozengurt, unpublished results.   (Fig. 4B) and immune complex kinase assays (Fig. 4C). In contrast bombesin-stimulated p42 MAPK activation could be detected using either assay. Thus, the dissociation of mitogenesis from MAP kinase activation could also be documented in Swiss 3T3 cells expressing Q227L ␣ s which are 100-fold more sensitive to forskolin than wild type cells.
Interfering Mutants of MEK-1 Stably Transfected into Swiss 3T3 Cells Significantly Attenuate PDGF-stimulated p42 MAPK Activation but Do Not Inhibit cAMP-induced Mitogenesis-Expression of interfering MEK-1 mutants with alanine substitutions at serine 217 or serine 221 have been shown to block MAP kinase activation in vivo (15). If, as indicated by the preceding results, cAMP induces DNA synthesis through a MAP kinaseindependent pathway in Swiss 3T3 cells, expression of interfering mutants of MEK-1 should not prevent [ 3 H]thymidine incorporation induced by forskolin and insulin in these cells. To test this prediction we isolated clones which overexpressed wild-type MEK-1 and the Ala 217 and Ala 221 mutants to comparable levels as determined by Western blotting (Fig. 5A). To verify that the level of expression of these mutants interfered with p42 MAPK activation in Swiss 3T3 cells, we performed immune complex kinase assays using lysates of transfected cells treated with or without PDGF. p42 MAPK activation in cells overexpressing wild-type MEK-1 was similar to that achieved with PDGF in untransfected cells. However, both the Ala 217 and Ala 221 mutants markedly inhibited PDGF-stimulated p42 MAPK activation by 50 and 60%, respectively (Fig. 5B). The striking finding shown in Fig. 5C was that the ability of forskolin to stimulate DNA synthesis was not impaired in cells expressing the interfering mutants. These results provide con-vincing evidence that cAMP-induced mitogenesis does not require the activation of MAP kinase.
Mitogenic Combinations of cAMP Do Not Activate MEK-1 and MEK-2 or p74 raf-1 -The finding of a striking dissociation between the mitogenic effects of cAMP and the activation p42 MAPK and p44 MAPK compared to other mitogens such PDGF and bombesin led us to explore the effect of cAMP upon upstream components of the MAP kinase cascade. This was achieved using highly sensitive assays in which the activation of immunoprecipitated kinases from stimulated cells was measured by their ability to activate, in vitro, GST fusion proteins of downstream kinases. Forskolin did not stimulate MEK-1 and MEK-2 activity whereas PDGF and bombesin both caused marked activation of these kinases (Fig. 6A). Next we examined the kinetics of p74 raf-1 activation in response to PDGF and forskolin with insulin. As shown in Fig. 6B, PDGF caused a marked stimulation of p74 raf-1 activity peaking at 5 min. In contrast, p74 raf-1 activity in forskolin-treated cells did not differ from that of control cells.
As cAMP is a mitogen in Swiss 3T3 cells, rather than growth inhibitory as it is in Rat-1 cells, we examined the effect of forskolin upon PDGF stimulated p74 raf-1 and p42 MAPK activation. Preincubation with forskolin abolished PDGF-stimulated p42 MAPK activation as shown in the mobility shift assay (Fig.  7A). This effect appeared to be mediated at the level of p74 raf-1 as PDGF-stimulated activation of this kinase was markedly inhibited by cAMP (Fig. 7B). However, PDGF-stimulated DNA synthesis in the presence of cAMP elevating agents was significantly enhanced (p Ͻ0.001) (Fig. 7C). In other experiments, we found that forskolin augmented [ 3 H]thymidine incorporation induced by PDGF at 2.5 and 5 ng/ml by 178 and 205%, respectively (data not shown). These findings again illustrate dissociation of activation of the MAP kinase cascade from cAMPinduced mitogenesis.
Mitogenic Combinations of cAMP Induce Phosphorylation and Activation of p70 s6k and Rapamycin Inhibits cAMP-stimulated Mitogenesis-The 90-kDa ribosomal S6 kinase (p90 rsk ) is phosphorylated and activated by p42 MAPK and p44 MAPK and thus lies on the MAP kinase cascade (5). Indeed, we could not demonstrate any activation of this enzyme by cAMP providing further corroborative evidence that the MAP kinase pathway was not activated (data not shown). A distinct S6 kinase family, p70/85kDa s6k , has been shown to be activated by a number of growth factors (42). Selective inhibition of the phosphorylation and activation of this kinase by the immunosuppressant rapamycin has suggested that p70 s6k plays a role in serum-stimulated DNA synthesis in Swiss 3T3 cells (33). 10 ng/ml PDGF, or no addition (Ϫ), lysed in lysis buffer, immunoprecipitated with anti-MEK-1/MEK-2 polyclonal antibody, and the immune complexes analyzed in a two-step immune complex kinase assay as described under "Experimental Procedures." Results are expressed as percent of maximum PDGF-stimulated activation (19000 -25000 cpm/1.5 ϫ 10 6 cells at 5 min), and the data are shown as the mean Ϯ S.E. for two independent experiments each performed in triplicate. The specific activity of [␥-32 P]ATP used was 900-1200 cpm/pmol. B, confluent and quiescent cultures of Swiss 3T3 cells were washed and treated with either 10 ng/ml PDGF (q), 10 M forskolin and 50 M IBMX with 1 g/ml of insulin ( ), or with control medium (serum-free DMEM) (f) for the times indicated, lysed in lysis buffer, immunoprecipitated with anti-p74 raf-1 polyclonal antibody, and the immune complexes analyzed in a two-step immune complex kinase assay as described in "Experimental Procedures." Results are the means of duplicates and are expressed as percent of maximum PDGF-stimulated activation (5000 -6000 cpm/1.5 ϫ 10 6 cells at 5 min) and are representative of three independent experiments. The specific activity of [␥-32 P]ATP used was 900-1200 cpm/pmol. Table II, treatment of Swiss 3T3 cells with either forskolin or insulin stimulated an increase in p70 s6k activity measured in an immune complex kinase assay by 2.5and 6-fold, respectively. The mitogenic combination of forskolin and insulin induced a further stimulation of p70 s6k activity. This stimulatory effect was completely inhibited by rapamycin at 20 ng/ml (Table II). Fig. 8, upper panel, shows in a mobility shift assay that forskolin in combination with insulin stimulated the phosphorylation of p70 s6k which accompanies the activation of this kinase. The mobility shift induced by these agents was comparable to that induced by PDGF which is presented for comparison (Fig. 8). In parallel experiments both forskolin and 8-BrcA alone also induced shift of p70 s6k (results not shown). The phosphorylation of p70 s6k induced by the mitogenic combination of forskolin with insulin was strongly inhibited by rapamycin. The striking finding shown in Fig. 8, lower panel, is that rapamycin markedly inhibited cAMP-stimulated DNA synthesis. The results presented in Fig. 8 thus identify one of the pathways utilized by cAMP and insulin to induce DNA synthesis that is distinct from the MAP kinase cascade.

DISCUSSION
The reinitiation of DNA synthesis in G 0 -arrested cells can be induced by multiple signaling pathways that act in a combinatorial and synergistic fashion (21,43,44). In the present study we utilized PDGF, bombesin, and cAMP in combination with insulin to promote maximum levels of DNA synthesis in Swiss 3T3 cells. Although these mitogenic signals must converge prior to DNA replication, the precise point of convergence in G 1 remains unknown. The redundancy in signaling pathways prior to the point of convergence implies that some events may be sufficient to stimulate DNA synthesis but not neccesarily obligatory for the action of all mitogens. p42 MAPK and p44 MAPK are activated by a number of mitogenic signaling pathways linked to both tyrosine kinase and G-protein-linked receptors (1)(2)(3). These include the Raf and MEK kinase pathways which transduce the signals from such effectors as p21 ras and PKC. p42 MAPK and p44 MAPK have been reported to phosphorylate an array of proteins involved in the mitogenic response including p90 rsk (5), the proto-oncogene products c-myc (7) and c-jun (8), and the transcription factor TCF 62 (6). Sustained activation of p42 MAPK and p44 MAPK , therefore, may be an obligatory step in the action of all mitogens leading to DNA synthesis as, in fact, has been proposed (45). However, several recent observations suggest that activation of p42 MAPK or p44 MAPK may not be an obligatory point of convergence in the action of all mitogenic signals. Interleukin-4 stimulates proliferation of two cell lines of T-lymphocyte and myeloid origin but not does activate MAP   kinase (18). Additionally, thyroid-stimulating hormone stimulates mitogenesis in thyrocytes but does not stimulate tyrosine phosphorylation of MAP kinase (17, but see also 46). In Swiss 3T3 cells, activin, a member of the transforming growth factor ␤ family of cytokines, is mitogenic but appears not to stimulate MAP kinase activation (19). These findings suggest but do not prove that p42 MAPK and p44 MAPK activation is not obligatory for DNA synthesis because the sensitivity of the assays used has been questioned (47) and none of the studies examined the effects of interfering mutants that block MAP kinase activation in vivo.
The results presented in this study demonstrate that the potent mitogenic effects of cAMP are not mediated via the MAP kinase cascade in Swiss 3T3 cells. Agents that elevate intracellular cAMP by distinct mechanisms are able to stimulate DNA synthesis to a level comparable to that seen with PDGF and bombesin. Additionally, cAMP-elevating agents either alone or in combination with insulin induce a significant increase in cell number demonstrating that these agents also stimulate progression through later stages of the cell cycle. However, using three separate assays we have shown that cAMP fails to detectably activate either p42 MAPK or p44 MAPK . This is in marked contrast to the effects of PDGF and bombesin which gave prolonged activation of these kinases. Interfering mutants of MEK-1 stably transfected into Swiss 3T3 also provide further convincing evidence that MAP kinase activation is not involved in cAMP-induced mitogenesis. In these cells PDGF-stimulated MAP kinase activation was significantly attenuated but the mitogenic effect of forskolin was uninhibited. In line with this conclusion, cAMP does not cause a significant induction of c-fos (27), the expression of which is regulated by the MAP kinase cascade through phosphorylation of TCF 62 (48). Swiss 3T3 cells stably transfected with a constitutively activated Gs␣ subunit are highly sensitive to the mitogenic effects of forskolin, and yet in this distinct cellular model again we did not detect p42 MAPK activation in response to this cAMP elevating agent. Furthermore, we have shown that cAMP fails to stimulate the upstream kinases in the cascade namely p74 raf-1 (see below) and MEK-1/MEK-2. In contrast, under identical experimental conditions we have confirmed that PDGF and bombesin potently stimulate p42 MAPK and p44 MAPK together with the upstream kinase MEK-1/-2 in Swiss 3T3 cells. Thus in this study we have demonstrated that cAMP does not utilize proteins at three levels of the MAP kinase cascade dissociating its mitogenic effects from a persistent activation of this pathway.
The serine/threonine protein kinase p74 raf-1 has been shown to play a central role in the mitogenic response of cells to growth factors and many oncogenes (49). It associates with activated Ras and stimulates the downstream elements of the MAP kinase cascade (50). However, our results also provide evidence that p74 raf-1 activation is not an obligatory step in cAMP mitogenic signal transduction. Mitogenic combinations of cAMP do not induce a significant increase in p74 raf-1 activity. Indeed, an increase in cAMP strikingly inhibited the activation of p74 raf-1 and p42 MAPK by PDGF but significantly increased the mitogenic effect of PDGF. Our results therefore demonstrate, for the first time, that cAMP can stimulate cell proliferation and inhibit the MAP kinase cascade in the same cell type providing further evidence for a dissociation between mitogenesis and activation of this pathway.
It is now clear that cAMP mitogenic signaling is not confined to Swiss 3T3 cells. Agents which elevate intracellular cAMP are mitogenic for number of cell types including mammary, keratinocyte, and kidney epithelial cells (51). Additionally, growth hormone-releasing factor, which stimulates cAMP ac-cumulation, is mitogenic for the rat anterior pituitary somatotroph (51). Further insight into the importance of the cAMP pathway in cell proliferation has come with the identification of constitutively activated Gs␣ subunits in a variety of tumors (52). These mutations, which result in a persistently elevated intracellular cAMP, are potentially oncogenic and demonstrate that cAMP may play a role in cellular transformation. Interestingly, our results with Swiss 3T3 cells stably transfected with a constitutively activated Gs␣ subunit indicate that cAMP can initiate a mitogenic response that is not mediated via the MAP kinase cascade in these cells.
Many growth factors activate a parallel but distinct signaling pathway leading to the phosphorylation and activation of p70 s6k which rapidly phosphorylates the S6 protein of 40 S ribosomal subunit (43). Inhibition of the activation of this enzyme with the immunosuppressant rapamycin or with neutralizing antibodies demonstrates that p70 s6k plays a role in serum-stimulated mitogenesis (33,53). Interestingly, it has been recently reported that p70 s6k phosphorylates CREM, which is also a target for cAMP-protein kinase, in response to mitogenic factors leading to a strong increase in transcriptional activation (54). This identifies a point of convergence between the p70 s6k and cAMP-protein kinase pathways that is distinct from the nuclear targets of the MAP kinase cascade. Our results show that cAMP in combination with insulin stimulates the activity and phosphorylation of p70 s6k as demonstrated by immune complex kinase and mobility shift assays. Rapamycin completely prevented the increase in activity and phosphorylation induced by cAMP. Crucially, we show here for the first time that rapamycin markedly attenuated the mitogenic effect of cAMP in combination with insulin. Thus our results identify p70 s6k , as opposed to MAP kinase, MEK, and Raf-1, as an important element in cAMP-induced mitogenic signaling pathways.
In conclusion, our results demonstrate that cAMP, a potent mitogen for Swiss 3T3 cells, does not induce a significant and persistent activation of p42 MAPK and p44 MAPK and fails to stimulate the upstream kinases of the cascade namely p74 raf-1 and MEK1/2. These findings, dissociating the mitogenic effects of cAMP from activation of the MAP kinase cascade, support the hypothesis that this kinase cascade is one of the parallel pathways that can lead to DNA synthesis rather than an obligatory point of convergence in mitogenic signaling.