The phosphodiesterase inhibitor SQ 20006 selectively blocks mitogen activation of p70S6k and transition to S phase of the cell division cycle without affecting the steady state phosphorylation of eIF-4E.

In quiescent cells high levels of protein synthesis are required in order to re-enter the cell cycle upon stimulation. Initiation of polypeptide synthesis is the step most often subject to regulation, controlled in part by phosphorylation of 40 S ribosomal protein S6 and a number of initiation factors. The kinase responsible for S6 phosphorylation is p70S6k. We now show that the p70S6k pathway can be selectively blocked by the aminopurine analogue, SQ 20006. This agent is known to raise cAMP levels, resulting in activation of protein kinase A. We present evidence that the increase in cAMP is not responsible for the inhibitory effect observed. We also show that SQ 20006 can prevent the activation of p70S6k in a rapid and reversible manner. The compound does not exert its inhibitory activity on p70S6k but can inhibit in vitro two protein kinase C isozymes (α and γ). In a B lymphoblastoid cell line, treatment with SQ 20006 results in inhibition of protein synthesis at the initiation stage. In contrast, when tested directly upon the translational machinery in the reticulocyte lysate, inhibition is manifest at both the level of initiation and elongation. The role of protein kinase A in the modulation of p70S6k and the rate of translation is discussed.

In quiescent cells high levels of protein synthesis are required in order to re-enter the cell cycle upon stimulation. Initiation of polypeptide synthesis is the step most often subject to regulation, controlled in part by phosphorylation of 40 S ribosomal protein S6 and a number of initiation factors. The kinase responsible for S6 phosphorylation is p70 S6k . We now show that the p70 S6k pathway can be selectively blocked by the aminopurine analogue, SQ 20006. This agent is known to raise cAMP levels, resulting in activation of protein kinase A. We present evidence that the increase in cAMP is not responsible for the inhibitory effect observed. We also show that SQ 20006 can prevent the activation of p70 S6k in a rapid and reversible manner. The compound does not exert its inhibitory activity on p70 S6k but can inhibit in vitro two protein kinase C isozymes (␣ and ␥). In a B lymphoblastoid cell line, treatment with SQ 20006 results in inhibition of protein synthesis at the initiation stage. In contrast, when tested directly upon the translational machinery in the reticulocyte lysate, inhibition is manifest at both the level of initiation and elongation. The role of protein kinase A in the modulation of p70 S6k and the rate of translation is discussed.
Stimulation of cell growth and proliferation is initiated at the cell surface by specific ligand-receptor interactions. Such interactions lead to receptor dimerization, cross-phosphorylation (1) and recruitment of Src-homology 2-containing signal transducers, which dock at phosphorylated tyrosine residues (2). This in turn causes activation of signaling molecules by a variety of mechanisms, including tyrosine phosphorylation (3), conformational changes (4,5), and translocation to the plasma membrane (6,7). The signal is further propagated and amplified by cascades of cytosolic protein kinases (8,9), which ultimately activate transcription factors in order to initiate metabolic processes necessary for growth.
One obligatory step required for progression through the cell cycle is the activation and maintenance of high rates of protein synthesis (10). Control of translation plays an important role in cell proliferation (reviewed in Ref. 11), with physiological reg-ulation almost always exerted at the level of polypeptide chain initiation (12,13). This phase is regulated, in part, by the phosphorylation of initiation factors involved in binding mRNA to the 40 S ribosomal subunit (11)(12)(13)(14)(15)(16). The only protein in the ribosome that has been reported to undergo phosphorylation in vivo in response to a number of mitogens is the 40 S ribosomal protein S6 (17). This has been mapped to an area of the 40 S ribosomal subunit that is implicated in mRNA binding and is thought to reside near the tRNA acceptor site (17). S6 phosphorylation can be correlated with a selective translational up-regulation of a family of mRNAs encoding for proteins required for cell growth (18). The kinase believed to modulate the level of S6 phosphorylation is p70 S6k , which is itself activated by phosphorylation (19 -21, 23). 1 However, the signaling pathway responsible for inducing p70 S6k activation remains unidentified (24,25). Contrary to other mitogen-regulated kinases, p70 S6k activity remains high throughout G 1 , and microinjection of inhibitory antibodies at any time before S phase can block G 1 /S transition (26,27).
The cap structure present on the 5Ј end of mRNA facilitates its binding to the ribosome, a process mediated by three initiation factors (eIF-4A, -4B, and -4F) and ATP hydrolysis (11-13, 28, 29). eIF-4F is a cap binding protein complex composed of three subunits: eIF-4E, which specifically recognizes the cap structure (13); eIF-4A, an ATP-dependent, single strand RNAbinding protein with helicase activity (13,28); and eIF-4␥ (p220, eIF-4G), whose function is unknown but whose integrity is required for eIF-4F complex activity (11)(12)(13). It is believed that eIF-4F functions to unwind secondary structure in the mRNA 5Ј-untranslated region to facilitate binding to the 40 S ribosomal subunit (11,12,28,29). Consistent with its proposed regulatory role, eIF-4E exists in both phosphorylated and nonphosphorylated forms (11)(12)(13)15) and is believed to be the least abundant of the initiation factors (15,30). In response to the appropriate stimuli, increased levels of eIF-4E phosphorylation have been directly correlated with increased rates of translation in a variety of cell types (reviewed in Ref. 11). More recently, it has been proposed that in adipocytes the regulated phosphorylation of an eIF-4␣-associated protein (PHAS-I, 4E-BP1) plays a role in modulating the availability of eIF-4E to enter the initiation pathway (31,32). However, the role of phosphorylation of eIF-4E in this interaction is at present poorly defined, and moreover, it is not known whether such interactions occur in other cell types.
These observations have prompted us to investigate the role of p70 S6k in G 1 progression and in the phosphorylation of eIF-4E and its association with PHAS-I. Here we report on the properties of a p70 S6k -specific inhibitor, SQ 20006; we show that SQ 20006 prevents the mitogen-induced activation of p70 S6k and causes inhibition of protein synthesis initiation in Swiss 3T3 cells and Raji cells. SQ 20006 blocks entry of cells into S phase but does not affect the steady state phosphorylation of eIF-4E or its association with PHAS-I (4E-BP1).

EXPERIMENTAL PROCEDURES
Materials -1-Ethyl-4-hydrazino-1H-pyrazolo-(3-4-b)-pyridine-5carboxylic acid, ethyl ester, and hydrochloride (SQ 20006) as well as M1 and M5 polyclonal antisera to p70 S6k were kindly provided by Dr. G. Thomas, Basel, Switzerland. Unless specified, chemicals were from Merck or Calbiochem. Media for cell culture and fetal bovine serum were from Bio-Whittaker (Dulbecco's modified Eagle's medium) or from Life Technologies, Inc. (RPMI 1640). Recombinant baculovirus for protein kinase C isozymes were provided by Dr. S. Stabel (Cologne, Germany). Expression and partial purification of protein kinase Cs and determination of activities were carried out as described (33). The cAMP-dependent protein kinase was a gift of Dr. B. Hemmings (Basel, Switzerland). p34 cdc2 /cyclin B from starfish oocytes was obtained from Dr. L. Meijer (Roscoff, France) and assayed as in Ref. 34.
Cell Culture and Cell Cycle Analysis-Swiss 3T3 fibroblasts were seeded and maintained as described previously (35). Cells at approximately 60 -70% confluency were arrested in G 1 by serum deprivation for 24 h in Dulbecco's modified Eagle's medium containing 0.5% fetal bovine serum (serum starvation protocol) and restimulated to enter the cell cycle by adding Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (complete medium). Assessment of the resting state as well as determination of cell population in each phase of the cell cycle were carried out by labeling nuclei with propidium iodide (CycleTest kit, Becton-Dickinson). For this, 1 ϫ 10 4 cells were analyzed for DNA content in a FACScan cell analyzer (Becton-Dickinson) employing the LYSIS II program, version 1.1. To monitor cell cycle progression into S phase, Swiss 3T3 cells were seeded at 2.5 ϫ 10 3 cells/well in 200 l in a 96-well plate in complete medium and allowed to grow for 48 h. Cells were then arrested following the serum starvation protocol and restimulated to enter the cell cycle by the addition of complete medium. Finally, cells were labeled in a time course with 1 Ci of [ 3 H]thymidine (DuPont NEN)/well and after a 2-h pulse, cells were harvested with a PHD cell harvester (Cambridge Technology, Inc.). Raji cells (a B lymphoblastoid cell line, ATCC CCL 86) were maintained in mid-log phase in RPMI 1640, supplemented with 10% fetal calf serum, 2 mM glutamine, and 50 g/ml gentamycin. Raji cells (1 ϫ 10 5 cells in 200 l) were seeded into a 96-well plate and pulse-labeled for 4 h with 2.5 Ci of [ 3 H]thymidine at the times indicated. Cells were then harvested using a Skaton Combi cell harvester, and radioactivity incorporated into DNA was quantified by scintillation counting.
Protein Synthesis Measurement-Raji cells (6 ϫ 10 4 cells in 200 l) were removed to a 96-well plate, incubated with the indicated concentration of SQ 20006 for 20 h at 37°C, and then assayed for protein synthesis by pulse-labeling for 4 h with 5 Ci/ml [ 35 S]methionine (ICN) in complete RPMI 1640. Cells were then harvested, and incorporation of radiolabel into protein was estimated by liquid scintillation counting. Reticulocyte lysates were prepared and incubated under the conditions described in Ref. 36, and protein synthesis was estimated using Polysome Analysis-Raji cells (1 ϫ 10 7 cells in 10 ml) in mid-log phase were incubated with 0.56 mM SQ 20006 for 24 h at 37°C, cycloheximide was then added to a final concentration of 200 g/ml, and cells were immediately harvested by pouring over crushed ice made from phosphate-buffered saline (with cycloheximide). Cells were pelleted, washed in the same buffer, and resuspended in 180 l of Buffer A (200 mM Tris-HCl, pH 8.0, 2 mM MgCl 2 , 25 mM KCl, 200 g/ml cycloheximide). Cell lysis was carried out by the addition of 0.1% Triton X-100 and 0.5% sodium deoxycholate (v/v), vortexing, and the removal of cell debris by centrifugation in a microcentrifuge at 4°C. The resultant supernatants, adjusted to equal protein concentration, were layered onto 10 -50% (w/v) sucrose gradients (in Buffer A), and ribosomes were fractionated by centrifugation in an SW 50.1 rotor (Beckman) for 90 min at 42,000 rpm at 4°C. The gradients were fractionated using an ISCO density gradient fractionator connected to a UA-5 absorbance monitor.
Analysis of eIF-4E and PHAS-I-eIF-4E was isolated by m 7 GTP-Sepharose chromatography (37) and visualized by SDS-PAGE 2 and immunoblotting, as described in Ref. 38. PHAS-I was visualized by the use of polyclonal antiserum kindly provided by Dr. J. Lawrence, (Washington University School of Medicine, St. Louis, MO). The phosphorylation status of eIF-4E was analyzed as described in Ref. 39. Protein Measurements, Western blotting, Immunoprecipitations, and Protein Kinase Assay-Cell extracts were prepared at the indicated times by Dounce homogenization in ice-cold Buffer B (50 mM Tris-HCl, pH 7.5, 120 mM NaCl, 20 mM NaF, 1 mM EDTA, 6 mM EGTA, 15 mM sodium pyrophosphate, 30 mM p-nitrophenyl phosphate, 1 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, 1% Nonidet P-40). Extracts were centifuged at 18,000 ϫ g for 10 min at 4°C. Protein concentration was determined with a BCA protein assay kit (Pierce).
To analyze p70 S6k and ERK2, equal amounts of total protein (5 g) were resolved on SDS-PAGE as described (20,40), and proteins were transferred to polyvinylidene difluoride (Millipore). The resulting membrane was decorated with either anti-ERK2 polyclonal antibody (40) or with the p70 S6k M1 antibody (20) and revealed using the ECL system (Amersham Corp.). In order to assay protein kinase activity, equal amounts of total protein (50 g) were immunoprecipitated with the p70 S6k antibody M5 (20) or with an anti-ERK2 polyclonal antibody (40) and immobilized on protein A-Sepharose beads. Pellets were washed with 3 ϫ 1 ml of ice cold Buffer B followed by 1 ml of ice-cold Buffer C (50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 1 mM dithiothreitol, 0.1% Triton X-100). Finally, pellets were dried and resuspended in 20 l of ice cold Buffer C. Kinase activity was assayed in Buffer C in a final volume of 25 l containing 50 M [␥-32 P]ATP (Amersham Corp.) (specific activity, 3.2 Ci/nmol), and 50 M peptide substrate. The synthetic peptides employed were S6 230 -249 for assaying p70 S6k activity and p70 S6k 408 -427 for assaying ERK2 activity (19). Reactions were terminated by the addition of EDTA/adenosine to a final concentration of 20 mM and 1.5 mM, respectively. Incorporation of [ 32 P]phosphate onto the peptide substrate was determined by spotting 20-l aliquots on P81 paper (Whatman) as described (41). In the case of cyclin-dependent kinase 2, immunoprecipitation was carried out on 300 g of total protein in Buffer D (25 mM Tris-HCl, pH 7.5, 60 mM ␤-glycerophosphate, 15 mM MgCl 2 , 15 mM EGTA, 0.1 mM NaF, 15 mM p-nitrophenyl phosphate, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 0.1% Nonidet P-40), and immobilized on protein A-Sepharose beads. Pellets were washed with 3 ϫ 1 ml of ice-cold Buffer D and 1 ml of ice-cold Buffer E (50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 1 mM dithiothreitol). Pellets were processed for kinase activity as described for ERK2 with the exception that Buffer E was employed and [␥-32 P]ATP was at a final concentration of 5 M (specific activity, 50 Ci/nmol).

SQ 20006 Specifically Blocks p70 S6k Activation in Swiss 3T3
Cells-SQ 20006 is a methylxanthine analogue and a potent phosphodiesterase inhibitor (42). The drug was first employed in the characterization of S6 phosphorylation in order to mimic the effect of high concentration of cAMP (43). Although SQ 20006 caused an inhibition of S6 phosphorylation and protein synthesis initiation, it was shown that total changes in intracellular levels of cAMP were not involved in inhibiting either process. Evidence obtained in recent years suggests that p70 S6k is the enzyme responsible for S6 phosphorylation in vivo (17). This prompted us to examine whether p70 S6k might be the target of SQ 20006. In order to test this hypothesis, we examined p70 S6k activity in Swiss 3T3 fibroblasts. Cells were arrested in G 0 by serum starvation for 24 h and restimulated with 10% serum in the presence or absence of SQ 20006. Fig.  1A shows that following re-entry into G 1 p70 S6k activity increases progressively, reaching a maximum at 1 h post stimulation. Contrary to epidermal growth factor stimulation (44), serum addition induces a sustained kinase activation that persists until transition to S phase. Addition of SQ 20006 at the time of serum stimulation completely prevents the activation of p70 S6k (Fig. 1A). The lack of kinase activity in the p70 S6k immune complex is reflected by the absence of the character-istic band shift on SDS-PAGE that accompanies p70 S6k activation ( Fig. 1C and Ref. 20). In order to examine whether this inhibition was specific for p70 S6k , we analyzed the activation of ERK2, which lies on a parallel but distinct pathway from p70 S6k (45). As shown in Fig. 1, B and D, ERK2 activity is totally unaffected by the presence of SQ 20006 in the medium. This suggests that SQ 20006 is specifically preventing the activation of p70 S6k . Moreover, addition of SQ 20006 at any time in G 1 fully reversed p70 S6k activation (see below) without affecting the activation of ERK2 (data not shown).
SQ 20006 Does Not Directly Affect the Activity of p70 S6k but Does Inhibit the Activity of Protein Kinase C␣ and -␥ in Vitro-One possible explanation for the effects of SQ 20006 on the activation of p70 S6k is that it is acting as a competitive inhibitor for ATP (46). In order to test whether SQ 20006 is a direct inhibitor of protein kinases, the compound was tested in vitro on numerous kinases, as summarized in Table I. The data presented show that p70 S6k activity was unaffected by the presence of SQ 20006, even up to levels in excess of 1 mM. In agreement with the data presented in Fig. 1, ERK2 activity was similarly unaffected in these assays, as was the case with cAMP-dependent protein kinase (protein kinase A) and cyclindependent kinase 2/cyclin E, although p34 cdc2 cyclin B was more sensitive to the inhibitor. We also tested the effect of SQ 20006 on the protein kinase C family of molecules; while little or no effect was observed on the majority of isoforms, both protein kinase C␣ and protein kinase C␥ were inhibited by SQ 20006 in vitro, displaying an IC 50 of 270 M and 160 M, respectively. These data suggest that protein kinase C␣ and/or protein kinase C␥ may have a role in the activation of p70 S6k . Indeed, it has been shown that p70 S6k is activated in a biphasic manner following the addition of growth factors to cells, with part of the response regulated by a protein kinase C-dependent signaling pathway (44).
The Inhibition of p70 S6k Activity Induced by SQ 20006 Is Reversible-To further characterize the kinetics of p70 S6k in-hibition by SQ 20006, quiescent Swiss 3T3 fibroblasts were stimulated with serum for 1 h and then treated with SQ 20006; extracts were then prepared at different times following incubation, and the activation state of p70 S6k was monitored by SDS-PAGE and immunoblotting. The data presented in Fig. 2A shows that p70 S6k is partially inactivated within 10 min of exposure of the cells to SQ 20006 and fully inactivated within 30 min. Next, we examined whether SQ 20006 addition is toxic to cells, leading to general cell death. Swiss 3T3 fibroblasts rendered quiescent by serum deprivation were restimulated to grow in the presence of SQ 20006 for 1 h. Cells were then washed free of SQ 20006, fresh medium was added, and p70 S6k activation was examined by Western blot analysis. The data presented in Fig. 2B indicate that within 60 min, full activation of p70 S6k was observed, as judged by its retarded mobility on SDS-PAGE, comparable with that obtained following 1 h of  serum stimulation in the absence of SQ 20006 (Fig. 2B, lane 6). Therefore, these data suggest that SQ 20006 is not toxic to Swiss 3T3 cells over the times that these experiments have been performed.
The Effect of SQ 20006 on Cell Cycle Progression after Serum Starvation and Stimulation of Swiss 3T3 Cells-Previous studies from the Thomas group (43) have shown that SQ 20006 causes acute effects on the inhibition of polysome assembly and protein synthesis initiation when added to cells re-entering the cell cycle from a quiescent state. An essential step in this process is the phosphorylation of 40 S ribosomal protein S6 (17). Since SQ 20006 blocks p70 S6k activation (Fig. 1, A and C), this could account for this observation. Whether such early inhibition is then bypassed, allowing cells to progress to S phase, is not known. To address this possibility, we have examined the effect of SQ 20006 addition on G 1 progression in Swiss 3T3 fibroblasts. Quiescent cells were stimulated with serum in the absence or presence of SQ 20006 or aphidicolin (to arrest cells in G 1 ), and progression through the cell cycle was visualized by FACScan analysis. Fig. 3 shows that in the absence of inhibitor, cells progressed through the cell cycle, with DNA synthesis clearly observed at 24 h (panel E). This is prevented by the inclusion of either SQ 20006 or aphidicolin (Fig. 3, I and G, respectively) to the restimulated cells or the addition of SQ 20006 to an asynchronous cell population (panel H). These data suggest that SQ 20006 induces a G 1 block that is not bypassed by the cell.
In order to investigate in more detail the inhibition of S phase entry following SQ 20006 addition, we first considered the response of cells to increasing doses of the drug. Fig. 4 shows that the concentration of SQ 20006 required to inhibit S phase entry by 50% (IC 50 ) is 0.2 mM, with full block at concentrations of 0.5 mM and above (Table II). We have also compared the effect of SQ 20006 addition with that of a potent inhibitor of the p70 S6k activation pathway, the immunosuppressant, rapamycin. As shown in Table II, rapamycin decreases the extent of S phase transition to 48% in Swiss 3T3 cells, whereas SQ 20006 fully blocks DNA synthesis. One possible explanation for the inhibitory effect of SQ 20006 could be that it is due to inhibition of phosphodiesterase and a subsequent rise in the levels of cAMP. To test this, cells were incubated in the presence of 8-Br-cAMP. As shown in Table II, 1 mM 8-Br-cAMP did not block entry of cells into S phase whether added at the start of the incubation or at any time during the G 1 period (data not shown). To complement these studies, aliquots of cells treated as above for 1 h were also analyzed for the activation state of p70 S6k . While the addition of SQ 20006 (Fig. 2) or rapamycin (Fig. 5) to serum-stimulated cells prevented the activation of p70 S6k , neither 8-Br-cAMP nor the phosphodiesterase inhibitor, IBMX was effective at altering the activity of p70 S6k (Fig. 5).
Next, we addressed the question of whether p70 S6k and protein synthesis inhibition by SQ 20006 after the restriction point (47) are critical to S phase transition. For these studies, quiescent cells were induced to grow by the addition of serum, SQ 20006 was added at various times, and DNA synthesis was monitored at 14 h following stimulation. Fig. 6A shows that induction of DNA synthesis was very sensitive to the presence of SQ 20006 during the first 10 h following serum stimulation, with sensitivity decreased by 12 h. The addition of SQ 20006 to cells during S phase transition (i.e. 12-18 h) did not significantly affect the extent of [ 3 H]thymidine incorporation into DNA (data not shown). Accordingly, withdrawal of the drug early in G 1 allowed cells to regularly proceed to S phase, whereas later removal caused a delay in the transition to S ( Fig. 6B and Table III).
Since SQ 20006 yields an apparent G 1 /S block, we decided to test whether the drug might also function in a manner similar to known inhibitors of S phase entry. This is the case for hydroxyurea, an inhibitor of the DNA precursor pool synthesis. Cells were restimulated with serum for 15 h, in the absence or presence of either SQ 20006 or hydroxyurea, the drugs were removed by washing the cells, and DNA synthesis was measured at 2-h intervals during the following 20 h by pulse labeling with [ 3 H]thymidine, as described. The data presented in Table  III show that release of the cells from the hydroxyurea block was rapid, with cells progressing into S phase within 4 h. On the contrary, cells released from SQ 20006 block required 15 h to reach mid-S phase, independent of whether SQ 20006 was added at the start of the incubation (Table III) or during G 1 (data not shown). Taken together, the data above suggest that SQ 20006 is not directly involved in the inhibition of DNA replication but rather it blocks early in G 1 and in a reversible manner.
SQ 20006 Inhibits DNA and Protein Synthesis in the B Lymphoblastoid Cell Line, Raji-To complement the above studies with Swiss 3T3 cells, we have also examined the effect of SQ 20006 upon DNA and protein synthesis in the B lymphoblastoid cell line, Raji, maintained in the mid-log phase of cell growth. Raji cells were incubated in the absence or presence of different concentrations of SQ 20006 (Fig. 7A, left panel) or rapamycin (Fig. 7A, right panel) for 20 h, prior to measuring the rates of DNA or protein synthesis, as described under "Experimental Procedures." Panel A shows that, as with Swiss 3T3 cells, SQ 20006 inhibited DNA synthesis, with an IC 50 of 0.1-0.2 mM; inhibition of protein synthesis was also evident over this concentration range. However, the two structurally unrelated p70 S6k inhibitors, SQ 20006 and rapamycin, display a different potency with regard to inhibition of DNA synthesis. The former fully blocks it (Fig. 7A, left panel), whereas the latter displays only a weak effect (Fig. 7A, right panel). In Swiss 3T3 cells, rapamycin has been reported to slightly decrease the rate of initiation and block the preferential translation of a class of mRNAs (18,48). To analyze whether SQ 20006 was affecting the initiation phase of translation, extracts were prepared from control or SQ 20006-treated cells, and their ribosomes were analyzed by sucrose density centrifugation. Relative to the control cells, polysome profiles obtained from SQ 20006-treated Raji cells (Fig. 7B) show a clear decrease in heavy polysomes, which is suggestive of a decrease in the initiation rate.
Considering the differences in the magnitude of inhibition observed between SQ 20006 and rapamycin, we asked whether SQ 20006 might act on translational targets other than p70 S6k . One potential site of regulation is initiation factor 4E (eIF-4E), which is a limiting factor in the process of initiation and has been demonstrated to undergo phosphorylation in response to numerous growth factors and hormones in a variety of cells (11)(12)(13). Recently it has been suggested that, in addition to phosphorylation, interaction of eIF-4E with other proteins, such as PHAS-I (4E-BP1 (31,32,49)) may play a role in translational control. Therefore, we have looked at the association between eIF-4E and PHAS-I following treatment of cells with SQ 20006, by isolation of the former on m 7 GTP-Sepharose, as described under "Experimental Procedures." As shown in Fig. 7C, prolonged treatment of Raji cells with SQ 20006 did not affect the interaction of eIF-4E with its inhibitory partner PHAS-I. We have also examined the phosphorylation status of eIF-4E by vertical slab isoelectric focusing and immunoblotting. Fig. 7D shows that SQ 20006 has no effect on the steady state phosphorylation of eIF-4E. In order to test whether SQ 20006 has a direct inhibitory effect on the translational machinery, we have also employed the reticulocyte lysate translation system. This in vitro system is unique, as it maintains high levels of protein synthesis, which is largely independent of any requirement for S6 phosphorylation; indeed, p70 S6k activ-ity in the reticulocyte lysate is low relative to that found in mitogen-stimulated cells. 3 As shown in Fig. 8A, time course and dose-response experiments indicate that SQ 20006 displayed a significant inhibition of translation at concentrations comparable with those used above. Analysis of polysomes by sucrose density gradient centrifugation (Fig. 8B) shows that SQ 20006 induces a weak and incomplete disaggregation of ribosomes from polysomes and a rise in the content of free 80 S ribosome couples. This was also seen in the presence of the elongation inhibitor, emetine, suggesting that part of the inhibitory effect of SQ 20006 is due to activation of low levels of nuclease. However, this level of nuclease is insufficient to account for the large inhibition of translation shown in panel A (data not shown). An alternative conclusion is that SQ 20006 induces a weak inhibition at the level of initiation, with a dominant effect at the level of elongation, the latter possibly through activation of eEF-2 kinase via protein kinase A (50). As with the Raji cells, SQ 20006 had little or no effect on the association between eIF-4E and PHAS-I; neither did it affect the steady state phosphorylation of eIF-4E (Fig. 8, C and D). The inhibition of translation in the reticulocyte lysate appears to be an in vitro effect; treatment of intact reticulocytes (37)  association between eIF-4E and PHAS-I, or the steady state phosphorylation of eIF-4E (data not shown). This possibly reflects the lack of requirement of this translation system for S6 phosphorylation and activation of p70 S6k .

DISCUSSION
In many cell systems examined, stimulation of quiescent cells to re-enter the cell cycle with serum or growth factors causes an immediate drop in the intracellular concentration of cAMP (51). Concomitant with this event is the activation of kinase cascades, which mediate the mitogenic effect of growth factors (11,17), leading to evidence of a negative correlation between higher levels of cAMP and mitogenesis. This has been shown in studies that have addressed the role of the cAMP-de-pendent protein kinase (protein kinase A) during cell cycle progression in yeast (52), during meiotic maturation of Xenopus oocytes (53) and directly on intracellular signaling pathways (54). A number of hypotheses have been proposed to explain the mechanism by which raised levels of cAMP yield a G 1 block; these include inhibition of activation of the raf kinase (54) and increased levels of p27 K , which in turn inhibits cyclindependent kinase 4 activity (55).  Recent data obtained using a lymphoid cell line has indicated another target of protein kinase A to be the inactivation of p70 S6k (56). On the other hand, previous data on S6 phosphorylation in Swiss 3T3 fibroblasts have shown that changes in total intracellular levels of cAMP were not involved in the inhibition of either S6 phosphorylation or protein synthesis initiation (43). The data presented in Fig. 1 show that in Swiss 3T3 cells, SQ 20006 is a potent inhibitor of the pathway leading to p70 S6k activation, although not a direct inhibitor of the activated kinase itself in vitro (Table I). This effect is fairly specific, as SQ 20006 selectively blocks the p70 S6k pathway without affecting the activation of the ERK2 kinase cascade (Fig. 1, C and D) or directly affecting the activity of ERK2 in vitro (Table I). SQ 20006 is a potent inhibitor of phosphodiesterase activity (42), leading to enhanced levels of cAMP and activation of protein kinase A. Considering that it also inhibits p70 S6k activation, one might be tempted to draw a parallel between the two phenomena (56). However, this may not be the case in Swiss 3T3 cells, as addition of the membrane-permeable analogue 8-Br-cAMP to cells at the time of restimulation (Fig.  5) or at any following time (data not shown) did not affect p70 S6k activity. Accordingly, the addition of the phosphodiesterase inhibitor, IBMX, was also ineffective in this respect (Fig.   5). Therefore, it appears that the elevation of cellular levels of cAMP alone is insufficient to explain the prevention of the serum-induced activation of p70 S6k . In Swiss 3T3 cells, p70 S6k is activated in a biphasic manner in response to growth factor stimulation, with the second phase under the control of protein kinase C (44). When SQ 20006 was tested in vitro on several kinases, it appeared to be particularly effective in inhibiting protein kinase C␣ and protein kinase C␥ (Table I). These data suggest that protein kinase C␣ and/or protein kinase C␥ might function upstream of p70 S6k , or else a still, as yet unknown, p70 S6k kinase(s) might be target for inhibition by SQ 20006.
Contrary to other mitogen-induced protein kinases (such as ERK2; Fig. 1), p70 S6k activity remains high throughout G 1 , and microinjection of inhibitory antibodies at any time in G 1 can block the G 1 /S transition (26,27). Considering the low turnover rate of SQ 20006 and that it was not toxic to cells, we set out to examine whether prolonged inhibition of p70 S6k was involved in blocking transition to S phase. As shown by FACScan analysis, SQ 20006 addition to cells yields an apparent G 1 /S block, similar to that seen with aphidicolin ( Fig. 3 and Table II), which could be reversed by removing the drug (Table III). We conclude that the inhibition set by SQ 20006 cannot be bypassed and does not allow cells to progress to S phase. In Panel C, Raji cell extracts were prepared as in panel B, and eIF-4E was isolated as described under "Experimental Procedures." Proteins recovered from the affinity resin were subjected to SDS-PAGE, resolved proteins were transferred to polyvinylidene difluoride, and eIF-4E and PHAS-I were identified using specific antiserum. Panel D, samples prepared as in panel C to enrich for eIF-4E were subjected to one-dimensional vertical slab isoelectric focussing and immunoblot analysis with antiserum specific for eIF-4E, as described. The migration of the phosphorylated form of eIF-4E is indicated. similar experiments, the immunosuppressant rapamycin is known to cause only a delay in the transition to S phase despite full inhibition of p70 S6k activity (Table II and Refs. 20, 22, and 57). Furthermore, we observed that release from SQ 20006 did not allow a rapid initiation of DNA replication. Cells appeared to require at least 10 h to enter S phase (Table III), indicating that SQ 20006 is an early G 1 blocker. Treatment of cells with SQ 20006 at different times following restimulation with serum ( Fig. 6A) showed that the compound could set an effective block only before the restriction point (47). In agreement with published data, this implies that p70 S6k function is necessary throughout the G 1 phase of the cell cycle (26). Accordingly, withdrawal of the drug from the medium within the first 2 h of exposure allowed cells to proceed to S phase, whereas removal at times after 4 h did not prevent the inhibitory effect of SQ 20006 on the transition to S phase ( Fig. 6B and Table III). These data suggest that as a result of, or as a consequence of, the inactivation of p70 S6k , cells are returned to an early point in G 1 , possibly to enable the cells to resynthesize labile proteins necessary for G 1 progression. Indeed, immunoprecipitation of cyclin-dependent kinase 2 with anti-cyclin E antiserum indicated a dramatic drop in cyclin-dependent kinase 2 activity 10 h following serum stimulation in the presence of SQ 20006 (data not shown) without affecting the activity of this kinase in vitro (Table I).
Early studies with SQ 20006 in Swiss 3T3 cells showed that the addition of the drug decreased the mobilization of 80 S ribosomes into polysomes (43). This is indicative of a lesion at the level of polypeptide chain initiation. We now show that SQ 20006 can inhibit translation initiation in the B lymphoblastoid cell line, Raji, when maintained in the mid-log phase of growth (Fig. 7A, left panel). The effect of SQ 20006 on translation was more pronounced than that seen with rapamycin (Fig. 7A, right panel). Because both SQ 20006 and rapamycin inhibit the activation of p70 S6k , we have studied other aspects of protein synthesis initiation in an attempt to explain the different effects of these compounds on protein synthesis. In mammalian cells, there is evidence for the regulation of translation by phosphorylation of initiation factors and their associated proteins (PHAS-I, 4E-BP1) involved in binding mRNA to the 40 S ribosomal subunit (11-13, 16, 29, 31, 32, 49). Although S6 phosphorylation may play a role in the selective binding of mRNA species to ribosomes (17,18,48), the activity of initiation factors, such as the eIF-4F complex, may also influence the selection of mRNA from the cellular pool for translation (11)(12)(13). It is believed that the eIF-4F complex functions to unwind secondary structure in the mRNA 5Ј-untranslated region to facilitate binding of the 40 S ribosome. As with S6, in response to the appropriate stimuli, increased levels of eIF-4E phosphorylation have been directly correlated with increased rates of translation in a variety of cell types (11)(12)(13). The data presented in Figs. 7 and 8 show that SQ 20006 affected neither the association between eIF-4E and PHAS-I nor the phosphorylation status of eIF-4E in Raji cells or reticulocyte lysate, respectively. Together these data confirm the finding that eIF-4E lies on a signaling pathway distinct from that of p70 S6k (39) and suggest that protein kinase C␣ and/or protein kinase C␥ is not involved in the serum-stimulated phosphorylation of eIF-4E in vivo. At this time, the exact function of the coordinate phosphorylation of S6 and eIF-4E in recruiting mRNA to the ribosome is not understood, although each appears to be mediated by separate signaling pathways.