Phosphatidylinositol-3 kinase is necessary for 12-O-tetradecanoylphorbol-13-acetate-induced cell transformation and activated protein 1 activation.

Phorbol esters, which activate isoforms of protein kinase C, are general activators of the transcription factor activated protein 1 (AP-1). The pathway involved in this signal transduction is not very clear. Currently, little is known about whether phosphatidylinositol-3 (PI-3) kinase plays any role in phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced signal transduction. We demonstrate here that TPA not only has markedly synergistic effects on insulin-induced PI-3 kinase activity, but it also can induce PI-3 kinase activity and the PI-3 phosphates by itself. We also found that insulin, a PI-3 kinase activator, enhanced TPA-induced AP-1 trans-activation and transformation in JB6 promotion-sensitive cells. Furthermore, wortmannin and LY294002, two PI-3 kinase inhibitors, markedly decreased AP-1 activity induced by insulin, TPA, or TPA and insulin and inhibited JB6 promotion-sensitive cell transformation induced by TPA or TPA and insulin. Most importantly, constitutive overexpression of the dominant negative PI-3 kinase P85 mutants completely blocked insulin- or TPA-induced AP-1 trans-activation and TPA-induced cell transformation. All evidence from present studies suggests that PI-3 kinase acts as a mediator in TPA-induced AP-1 activation and transformation in JB6 cells.

12-O-Tetradecanoylphorbol-13-acetate (TPA) is not only a potent tumor promoter in mouse skin (14,15), but it also induces a wide range of other biological effects in cultured cells (16). Protein kinase C (PKC) is well known as a TPA receptor and a phospholipid-dependent kinase involved in basic cellular functions, including regulation of cell growth, differentiation, and gene expression (17,18). PKC isozymes include more than 11 different enzymes (␣, ␤I, ␤II, ␥, ␦, ⑀, , , , , and ). Previous studies indicated that the TPA-induced activation of AP-1, NFB, and other transcription factors in the nucleus is mediated by the Ras-Raf-1/MAP kinase cascade (18 -20). The role of PI-3 kinase in the TPA-induced signal transduction pathway, however, is not clear, even though some reports indicated that PI-3 kinase and p21 ras modulate each other's signals (2,4,21). For example, overexpression of activated p21 ras in PC12 cells increases PI-3 kinase activity and stimulates the accumulation of 3Ј-phosphorylated inositol lipids in the cells (21). GTP-bound p21 ras also directly binds and activates PI-3 kinase in vitro (21). Expression of activated PI-3 kinase in NIH 3T3 cells apparently potentiates p21 ras -dependent signaling events (2). Furthermore, PI-3 kinase products may play an important role in extensive cross-talk among multiple signaling pathways and regulation of cell function (1,12,13). Since both phorbol ester-and insulin-induced activation of the Ras-Raf/MAP kinase pathway and a combination of insulin and phorbol ester resulted in a synergistic activation of this pathway (22), we inquired whether TPA can activate PI-3 kinase and whether activation of PI-3 kinase is required for TPAinduced signal transduction and cell transformation. In the present study we used several approaches, which included a PI-3 kinase activator, two pharmacological PI-3 kinase inhibitors, and a dominant negative PI-3 kinase inhibitor, to study the role of PI-3 kinase in TPA-induced AP-1 activation and cell transformation in the well characterized mouse epidermal JB6 P ϩ (tumor promotion-sensitive) cells.

MATERIALS AND METHODS
Plasmids and Reagents-The AP-1 luciferase reporter plasmid (Ϫ73/ ϩ63 collagenase-luciferase) and cytomegalovirus-neo marker vector plasmid were constructed as previously reported (23); the bovine PI-3 kinase P85 subunit mutant plasmid (⌬P85) and vector plasmid SR␣ were as described by Hare et al. (24); agarose conjugated with monoclonal antiphosphotyrosine antibody Py20 was from Santa Cruz Biotechnology; fetal bovine serum (FBS) was from Life Technologies, Inc.; LipofectAMINE was from Life Technologies, Inc.; Eagle's minimal essential medium (MEM) and wortmannin were from Calbiochem; LY294002 was from BioMol; TPA was from Calbiochem; insulin was from Sigma; luciferase assay substrate was from Promega; and [␥-32 P]ATP was from DuPont NEN.
Generation of Stable Cotransfectants with AP-1 Reporter and Dominant Negative PI-3 Kinase Mutant-JB6 P ϩ cells, C1 41, were cultured in a six-well plate until they reached 85-90% confluence. We used 2 g of AP-1 luciferase reporter plasmid and 0.3 g of cytomegalovirus-neo vector with 6 g of a dominant negative mutant of PI-3 kinase P85 subunit plasmid ⌬P85 or vector SR␣ control plasmid DNA and 15 l of LipofectAMINE reagent to transfect each well in the absence of serum. After 10 -12 h, the medium was replaced by 5% FBS and MEM. Approximately 30 -36 h after the beginning of the transfection, the cells were trypsinized, and the cell suspension was seeded into 75-ml culture flasks and cultured for 24 -28 days with G418 selection (300 g/ml). Stable transfectants were screened by assay of the luciferase activity and Western blotting with rabbit polyclonal IgG against human PI-3 kinase P85␣. Stable transfected cells, ⌬P85 mass1, ⌬P85 mass2, and AP-1 mass1, were cultured in G418-free MEM for at least two passages before each experiment.
Assay for AP-1 Activity-Confluent monolayers of JB6 C1 41-19, ⌬P85 mass1, ⌬P85 mass2, or AP-1 mass1 cells were trypsinized, and 5 ϫ 10 3 viable cells suspended in 100 l 5% FBS and MEM were added into each well of a 96-well plate. Plates were incubated at 37°C in a humidified atmosphere of 5% CO 2 . Twelve to 24 h later, cells were starved by culturing them in 0.1% FBS and MEM for 12 h prior to exposure to TPA or insulin. The cells were exposed to TPA, insulin, and TPA and insulin for AP-1 induction with or without different concentrations of wortmannin or LY294002 for 24 h. The cells were extracted with lysis buffer, and luciferase activity was measured using a luminometer (Monolight 2010). The results are expressed as the relative AP-1 activity or relative luciferase units (26).
PI-3 Kinase Assay-PI-3 kinase activity was assayed according to the method of Endemann et al. (27). In brief, JB6 cells, C1 41, ⌬P85 mass1, and ⌬P85 mass 2 or AP-1 mass1, were cultured in monolayers in 100-mm plates. Then cells were incubated in 0.1% FBS and MEM for 24 h and in serum-free MEM for 3-4 h at 37°C, respectively. TPA (10 ng/ml) with or without wortmannin was added, and after 20 min at 37°C, 2.5 g/ml insulin was added. After 10 min at 37°C, the cells were washed once with ice-cold phosphate-buffered saline and lysed in 400 l of lysis buffer/plate (20 mM Tris, pH 8, 137 nM NaCl, 1 mM Mgcl 2 , 10% glycerol, 1% Nonidet P-40, 1 nM dithiothreitol, 0.4 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride). The lysates were centrifuged, and the supernatants were incubated overnight at 4°C with 40 l of agarose beads previously conjugated with the monoclonal antiphosphotyrosine antibody Py20. The beads were washed twice with each of the following buffers: 1) phosphate-buffered saline with 1% Nonidet P-40, 1 mM dithiothreitol; 2) 0.1 M Tris, pH 7.6, 0.5 M LiCl, 1 mM dithiothreitol; and 3) 10 mM Tris, pH 7.6, 0.1 M NaCl, 1 mM dithiothreitol. The beads were incubated for 5 min on ice in 20 l of buffer 3, and then 20 l of 0.5 mg/ml phosphatidylinositol (previously sonicated in 50 mM Hepes, pH 7.6, 1 mM EGTA, 1 mM NaH 2 PO 4 ) was added. After 5 min at room temperature, 10 l of the reaction buffer was added (50 mM MgCl 2 , 100 mM Hepes, pH 7.6, 250 M ATP containing 5 Ci of [␥-32 P]ATP), and the beads were incubated for an additional 5 min. The reactions were stopped by the addition of 15 l of 4 N HCl and 130 l of chloroform:methanol (1:1). After vortexing for 30 s, 30 l from the phospholipid-containing chloroform phase was spotted onto thin-layer chromatography plates coated with Silica Gel H containing 1.3% potassium oxalate and 2 mM EDTA applied in H 2 O:methanol (3:2). Plates were heated at 110°C for at least 3 h before use. The plates were placed in tanks containing chloroform:methanol:NH 4 OH:H 2 O (600:470:20:113) for 40 -50 min until the solvent reached the top of the plates. The plates were dried at room temperature and autoradiographed. The PI-3 phosphate fractions were scraped off and assayed by scintillation counting (27).
Intracellular PI-3,4-P 2 Assay-JB6 C1 41 cells were cultured in monolayers in 100-mm dishes. The cells were washed three times with phosphate-free MEM and starved in the same medium for 24 h. Before stimulation, the cell medium was changed to fresh phosphate-free MEM for 3-4 h, and then cells were incubated in the same medium with 0.5 mCi/ml [ 32 P]orthophosphate (ICN) for 3 h at 37°C. TPA (10 ng/ml) was or was not added, and after 20 min at 37°C, 2.5 g/ml insulin was or was not added. The cells were washed twice with ice-cold phosphatebuffered saline 10 min later. Then the cells were extracted by addition of 2 ml of methanol, 0.8 ml of chloroform, 0.1 ml of crude brain phosphoinositide (5 mg/ml in CHCl 3 ) and 0.1 ml of [ 3 H]PI-4,5-bisphosphate (10,000 dpm) in CHCl 3 . Lipids were deacylated by alkaline hydrolysis (28 -31). The deacylated products were separated by high-performance liquid chromatography on a Partisil SAX ion exchange column (Alltech) essentially as described by Scheid and Duronio (30). After a 10-min wash with deionized water, the column was eluted with a 20-min linear gradient from 0 -0.25 M ammonium phosphate, pH 3.8, followed by a linear gradient to 1 M ammonium phosphate in 50 min, with a flow rate of 1 ml/min. Fractions of 0.5 ml were collected, and radioactivity was determined by scintillation counting after addition of 3 ml of Ecolume (ICN). About 10 g of ATP was added to each sample to check the reproducibility of retention times, and the eluate was monitored at 254 nm.
Anchorage-independent Transformation Assay-Inhibition by wortmannin or LY294002 of TPA or TPA and insulin-induced cell transformation was investigated in JB6 P ϩ cells. Cells (1 ϫ 10 4 /ml) were exposed to TPA or TPA and insulin with or without wortmannin or LY294002 in 1 ml of 0.33% BME agar containing 10% FBS over 3.5 ml of 0.5% BME agar medium containing 10% FBS. The cultures were maintained in a 37°C, 5% CO 2 incubator for 14 -21 days, and the cell colonies were scored by the methods described by Colburn et al. (32). The effect of wortmannin or LY294002 on transformation of JB6 cells was presented as a percentage inhibition of cell transformation.

TPA Induces PI-3 Kinase Activity and Enhances Insulininduced PI-3 Kinase Activity-
To determine whether PI-3 kinase plays any role in TPA-induced signal transduction, we first investigated whether TPA can induce PI-3 kinase activity. We analyzed PI-3 kinase in JB6 P ϩ cells stimulated by TPA. As shown in Fig. 1, TPA not only had a synergistic effect on insulin-induced PI-3 kinase activity, but it also induced PI-3 kinase activity by itself. Induction or enhancement of PI-3 kinase activity by TPA could be blocked by wortmannin (Fig. 1). To test whether the PI-3 phosphates were increased in the cells stimulated by TPA, we also measured the products of PI-3 kinase in TPA-treated JB6 cells by HPLC. The results show that PI-3,4-P 2 in JB6 cells was increased after cells were stimulated by insulin or TPA (Fig. 2). PI-3-phosphate and PI-3,4,5triphosphate could not be conclusively identified.
Inhibition of TPA-induced AP-1 Activity in JB6 P ϩ Cells by Wortmannin-Since the above results show that TPA could stimulate PI-3 kinase activity and enhance insulin-induced PI-3 kinase activity, we hypothesized that PI-3 kinase may be involved in TPA-induced AP-1 trans-activation. To test this idea, we incubated JB6 P ϩ cells with wortmannin, a specific inhibitor of PI-3 kinase, for 30 min prior to TPA stimulation. The results showed that a nontoxic concentration of wortman-FIG. 1. TPA induces PI-3 kinase activity and enhances insulininduced PI-3 kinase activity. JB6 C1 41 cells were treated with medium or TPA (10 ng/ml) with or without wortmannin for 20 min at 37°C in 5% CO 2 . Cells were then exposed to insulin (2.5 g/ml) for another 10 min. The cells were harvested, and the PI-3 kinase was immunoprecipitated with monoclonal antibody PY20. The immunoprecipitates were incubated with PI and [␥-32 P]ATP. The phospholipids were separated by TLC. The TLC plates were dried at room temperature and autoradiographed. PI-3-P, PI-3-phosphate. nin (100 nM) inhibited AP-1 activity induced by insulin, TPA, or TPA and insulin by 92.0, 50.0, or 57.3%, respectively (Fig. 3).
Inhibition of TPA-induced AP-1 Activity in JB6 P ϩ Cells by LY294002-Wortmannin inhibits PI-3 kinase by covalently binding to P110 of the enzyme. Unlike wortmannin, LY294002 inhibits PI-3 kinase by competing with ATP for its substrate binding site (33). We therefore used this other PI-3 kinase inhibitor, LY294002, to determine the role of the PI-3 kinase in AP-1 trans-activation induced by insulin or TPA. The results showed that LY294002 inhibited AP-1 activity induced by insulin or TPA in a dose-dependent manner (Fig. 4).
Insulin Induces AP-1 Trans-activation and Enhances TPAinduced AP-1 Activity-Previous studies demonstrated that insulin or TPA induced partial activation of the extracellular signal-regulated protein kinase, whereas a combination of insulin and TPA resulted in a synergistic activation of the extracellular signal-regulated protein kinase in Rat-1 HIR cells (22). To investigate whether insulin could induce AP-1 or promote TPA-induced AP-1 activity in JB6 cells, we exposed JB6 P ϩ cells to insulin, TPA, or TPA and insulin. The results showed that insulin could markedly induce AP-1 activity and increase TPA-induced AP-1 activity in JB6 P ϩ cells in a 0.1% FBS medium culture system (Fig. 5A). The induction of AP-1 activity and the increase in TPA-induced AP-1 activity by insulin occur in a dose-dependent manner (Fig. 5B). At a serum concentration higher than 0.25%, the increase of TPA-induced AP-1 activity by insulin could not be observed (Fig. 6).
Overexpression of the Dominant Negative PI-3 Kinase P85␣ Blocks Insulin-or TPA-induced AP-1 Activity-Since significant inhibition of insulin-or TPA-induced AP-1 activity was achieved by using both PI-3 kinase inhibitors, wortmannin and LY294002, PI-3 kinase may play a critical role in TPA-or insulin-induced AP-1 activation. The dominant negative mutant of PI-3 kinase, ⌬P85, has been shown to be a specific inhibitor of PI-3 kinase in Chinese hamster ovary cells (24). To specifically block PI-3 kinase and to test the role of PI-3 kinase in TPA-induced AP-1 activation in JB6 cells, the dominant negative mutant of the PI-3 kinase regulatory subunit P85 plasmid and AP-1 reporter was cotransfected into JB6 cells by using a LipofectAMINE kit. Three stable mass cultures, two (⌬P85 mass1 and ⌬P85 mass2) from cotransfection AP-1 reporter and SR␣ ⌬P85 plasmids and one (AP-1 mass1) from cotransfection with AP-1 reporter and vector SR␣, were established by G418 selection (34). To determine whether overex-pression of the dominant negative PI-3 kinase protein blocks TPA-or TPA-and insulin-induced PI-3 kinase activity, we also tested the PI-3 kinase activity in ⌬P85 mass1 and mass2 cells induced by insulin, TPA, or TPA and insulin. The results show that the PI-3 kinase activity induced by insulin, TPA, or TPA and insulin was completely blocked by expression of the dominant negative PI-3 kinase protein (Fig. 7). Furthermore, overexpression of the dominant negative PI-3 kinase protein in ⌬P85 mass1 and ⌬P85 mass2 cells blocked insulin-or TPAstimulated AP-1 activity completely compared with that in AP-1 mass1 cells in all the sample points of the time and dose courses studied (Figs. 8 and 9).
Enhancement by Insulin of TPA-induced JB6 P ϩ Cell Trans- Cells were then exposed to insulin (2.5 g/ml), TPA (10 ng/ml), or insulin (2.5 g/ml) and TPA (10 ng/ml), respectively. After a 24-h culture at 37°C in 5% CO 2 , the AP-1 activity was measured and is presented as described under "Materials and Methods." Cells were then exposed to inducer as indicated. After 24 h at 37°C in 5% CO 2 , the AP-1 activity was measured and is presented as described under "Materials and Methods." formation-Since our previous results and other studies demonstrated that induced AP-1 activity is important and required for cell transformation, we tested whether insulin could induce transformation or promote TPA-induced transformation. The results showed that insulin could not induce JB6 P ϩ cell transformation alone; however, it markedly increased the TPA-induced JB6 P ϩ cell transformation rate (Fig. 10).
Inhibition by Wortmannin or LY294002 of JB6 P ϩ Cell Transformation Induced by TPA and TPA plus Insulin-As shown in Fig. 11, wortmannin and LY294002 inhibited not only TPA-plus insulin-induced JB6 P ϩ cell transformation but also TPA-induced JB6 P ϩ cell transformation. This inhibition is in a similar dose range as that observed for the inhibition of PI-3 kinase activity and the inhibition of AP-1 trans-activation.
Blocking of TPA-induced JB6 Cell Transformation by Overexpression of ⌬P85 Protein-We further explored whether the transfectants expressing the dominant negative PI-3 kinase mutant could repress TPA-induced transformation. Fig. 12 summarizes the results of these studies. The TPA-induced transformation in two stable mass transfectants, ⌬P85 mass1 and ⌬P85 mass2, were almost totally blocked, whereas the AP-1 mass1 cells showed a high frequency of the transformation rate with exposure to TPA. DISCUSSION Our present studies demonstrate that PI-3 kinase is a crucial mediator of TPA-induced cell transformation and AP-1 transactivation in JB6 cells. TPA alone could induce PI-3 kinase activity and increase the level of PI-3,4-P 2 in JB6 cells. More interestingly, TPA and insulin synergistically induced PI-3 kinase activity. Insulin, a strong PI-3 kinase activator, enhanced TPA-induced AP-1 activation and cell transformation. Furthermore, wortmannin and LY294002, two different kinds of PI-3 kinase inhibitors, which inhibit PI-3 kinase by interfering with P110 and P85, respectively, inhibit TPA-induced AP-1 activation as well as cell transformation. More convincingly, TPA-and TPA-and insulin-induced AP-1 activation and TPAinduced cell transformation could be blocked completely by overexpression of the dominant negative PI-3 kinase P85 mutants in all time courses and dose responses studied. In contrast, we found previously that wortmannin or LY294002, as well as overexpression of ⌬P85, had no significant inhibitory effect on JB6 cell proliferation and UV-induced AP-1 activity (34). These observations suggest that the inhibition of AP-1 activation and cell transformation occurred through blocking PI-3 kinase activity but not because of any cytotoxic effect.
Activation of PKC requires both association with the membrane and a number of activators and cofactors, the requirements for which differ for each isozyme (35). Thus, PKCs are grouped into three major classes: conventional PKC isoforms, FIG. 5. Enhancement of TPA-induced AP-1 activity in JB6 P ؉ cells by insulin. JB6 C1 41-19 cells were or were not exposed to TPA (10 ng/ml) with or without 2.5 g/ml insulin (A) or different concentrations of insulin (B) at 37°C in 5% CO 2 for 24 h. The AP-1 activity was measured by luciferase activity assay as described under "Materials and Methods." The results are presented as the relative luciferase units (RLU) or the luciferase activity relative to control cells.
FIG. 6. Influence of serum concentration on enhancement of TPA-induced AP-1 activity by insulin. JB6 C1 41-19 cells were seeded into each well of 96-well plates. The plates were incubated at 37°C in 5% CO 2 for 12-14 h. Cells were then starved by culturing cells in different concentrations of FBS and MEM for 12 h prior to exposure to stimulators. The cells were or were not exposed to TPA (10 ng/ml) with or without insulin (2.5 g/ml) in different concentrations of FBS and MEMat 37°C in 5% CO 2 for 24 h. The AP-1 activity was determined by luciferase activity assay as described under "Materials and Methods." The results are presented as relative luciferase units (RLU). such as ␣, ␤I, ␤II, and ␥; novel PKCs, including ␦, ⑀, , and ; and atypical PKCs, represented by the and isozymes of PKC (35,36). Activation of atypical PKCs could be carried out by either the PI-3 kinase pathway or the ceramide pathway (37,38). Conventional PKCs are activated by diacylglycerol in a Ca ϩ -dependent manner. In contrast, activation of novel PKCs is Ca ϩ -independent (35,36). In addition to the natural activator, conventional PKCs and novel PKCs are also activated with high specificity by TPA (17). For this reason, TPA is often used in the study of the mechanisms of conventional PKC and novel PKC activation and their function. Most of the previous studies have addressed the regulation of phosphorylation of the insulin receptor by PKC in PKC isozyme-transfected cells (39,40). Overexpression of PKC isozymes ␣, ␤I, ␥, and ⑀ did not affect expression of the insulin receptor or insulin-stimulated tyrosine phosphorylation of the receptor. However, in response to phorbol esters, cells expressing PKC ␣, ␥, and ␤I but not ⑀ exhibited 3-4-fold higher levels of insulin receptor (IR) phosphorylation. This TPA-stimulated IR phosphorylation inhibits the activation of the insulin receptor kinase and the insulininduced PI-3 kinase activity as well as the tyrosine phosphorylation of Shc (39,40), but this inhibition is not observed in the cells containing only the endogenous levels of PKC (40). In the present study, we demonstrated that TPA induces a low level of PI-3 kinase activity and has significant synergistic effects with insulin on activation of PI-3 kinase in mouse epidermal JB6 cells. The reason for the difference among data from different cells may be due to different levels of endogenous PKC in the various cell types studied as well as differences in ratios of various PKC isozymes present in different cells. The ratio of PKC:IR in different cells may be another reason for these differences.
Several studies suggested that the PI-3 kinase and its products PI-3,4-P 2 and PI-3,4,5-triphosphate are important regulators of cell proliferation and c-fos gene expression (1-3). The introduction of the NH 2 -terminal Src homology 2 domain of the P85 subunit of PI-3 kinase into cells abrogates the insulin-or FIG. 7. Inhibition of PI-3 kinase activity by overexpression of dominant negative PI-3 kinase mutant. Cells (⌬P85 mass1, ⌬P85 mass2, or AP-1 mass1) were or were not treated with TPA (10 ng/ml) for 20 min at 37°C in 5% CO 2 . Cells were then exposed to insulin (2.5 g/ml) for another 10 min. The cells were harvested, and PI-3 kinase activity was measured as described under "Materials and Methods." FIG. 8. Inhibition of insulin-, TPA-, or TPA-and insulin-induced AP-1 activity by overexpression of dominant negative PI-3 kinase mutant. Cells (⌬P85 mass1, ⌬P85 mass2, or AP-1 mass1) were or were not exposed to insulin (2.5 g/ml), TPA (10 ng/ml), or insulin (2.5 g/ml) and TPA (10 ng/ml), respectively. After a 24-h culture at 37°C in 5% CO 2 , the AP-1 activity was measured and presented as described under "Materials and Methods." For the dose-response study, cells (⌬P85 mass2 or AP-1 mass1) were treated with the indicated doses of insulin (A) or TPA (B) for 24 h. C, for the time course study, ⌬P85 mass2 (open symbols) or AP-1 mass1 (closed symbols) cells were or were not exposed to 2.5 g/ml insulin (circles), 10 ng/ml TPA (triangles), or 2.5 g/ml insulin and 10 ng/ml TPA (diamonds). The AP-1 activity was measured and presented as described under "Materials and Methods." IGF-1-stimulated DNA synthesis and prevents insulin stimulation of c-Fos protein expression (2,3). The microinjection of a dominant-negative p21 ras mutant or anti-Ras antibody inhibited insulin-induced DNA synthesis (3). A constitutively activated mutant P110 induced transcription from the Fos promoter; coexpression of dominant negative Ras blocked this response (2). Other studies have shown that PI-3,4-P 2 and PI-3,4,5-triphosphate are elevated in cells transformed by vabl, v-src, and polyoma middle T, and decreased levels of these lipids correlate with impaired cell transformation by mutated forms of these oncogenes (1,41,42). It has been reported that the presence of insulin-like growth factor I receptor (IGF-IR) is an obligatory requirement for the establishment and maintenance of the tumor phenotype (43)(44)(45)(46). Cells derived from mouse embryos with a targeted disruption of the IGF-IR gene (R Ϫ cells) cannot be transformed by SV40 T antigen or by an activated and overexpressed Ha-ras, even by a combination of both, all of which transform very efficiently the corresponding wild type cells or other 3T3-like cells (47). If a plasmid expressing a wild type human IGF-I receptor cDNA is stably transfected into R Ϫ cells, the cells can be transformed by SV40 T antigen. This indicates that the defect in transformability is specifically due to the lack of IGF-IR (46). Substantial evidence has been reported that PI-3 kinase is a critical component of signaling pathways used by the cell surface receptors for a variety of mammalian growth factors or other stimulators (1,10,12,48), especially IR and IGF-IR. Recently, it was reported that insulin could activate the Ras-Raf/MAP kinase pathway by interacting and activating its receptors (7). Dhand et al. (7) suggested that activation of this Ras/MAP kinase pathway is critical for the effect of insulin on mitogenesis and c-fos expression. Others found that neither insulin nor phorbol ester regulation of phosphoenolpyruvate carboxykinase gene expression requires activation of the Ras/MAP kinase pathway, but PI-3 kinase is required in this event (49). In contrast, Sakaue et al. (50) demonstrated that neither the Ras/MAP kinase cascade nor PI-3 kinase may be required for insulin-stimulated glycogen synthase activation in Chinese hamster ovary cell lines. Evidence from different groups using different models has confirmed the crucial importance of AP-1 activity in transformation and carcinogenesis (23,26,51,53). Our previous results have shown that AP-1 activation is required for tumor promotion in the JB6 cell model (23,26,52,53). High basal levels of AP-1 activity appear to be important for the maintenance of tumor phenotypes in the transformed cell line RT101 (52,53). Since our primary results (Fig. 1) showed that TPA can induce PI-3 kinase activity and increase the level of PI-3,4-P 2 (Fig. 2), as well as have a markedly synergistic effect with insulin on PI-3 kinase activation, a critical question is whether PI-3 ki-FIG. 10. Enhancement of insulin on TPA-induced JB6 P ؉ cell transformation. 1 ϫ 10 4 JB6 C1 41 cells were or were not exposed simultaneously to insulin (2.5 g/ml), TPA (10 ng/ml), or TPA (10 ng/ml) and insulin in 0.33% BME agar containing 10% FBS over 0.5% BME agar medium containing 10% fetal calf serum. Cell colonies were scored after 14 days of incubation at 37°C in 5% CO 2 .
FIG. 11. Inhibition of JB6 P ؉ cell transformation induced by TPA or TPA and insulin by wortmannin or LY294002. 1 ϫ 10 4 JB6 C1 41 cells were or were not exposed simultaneously to TPA (10 ng/ml) or TPA (10 ng/ml) and insulin (2.5 g/ml) with or without different concentrations of wortmannin or LY294002 in 0.33% BME agar containing 10% fetal calf serum over 0.5% BME agar containing 10% FBS. Cell colonies were scored after 14 days of incubation at 37°C in 5% CO 2 . The inhibition of cell transformation induced by TPA or TPA and insulin is expressed as described under "Materials and Methods." FIG. 12. Inhibition of JB6 P ؉ cell transformation induced by TPA by overexpression of dominant negative PI-3 kinase mutant. 1 ϫ 10 4 ⌬P85 mass1, ⌬P85 mass2, or AP-1 mass1 cells were or were not exposed to TPA (10 ng/ml) in 0.33% BME agar containing 10% FBS over 0.5% BME agar containing 10% FBS. Cell colonies were scored after 21 days of incubation at 37°C in 5% CO 2 . The inhibition of cell transformation induced by TPA is expressed as described under "Materials and Methods." nase plays any role in TPA-induced AP-1 activation and cell transformation. To test this hypothesis, we used several approaches. First, we treated cells with insulin, a very effective PI-3 kinase activator. The results showed that activation of PI-3 kinase by insulin resulted in a marked increase in TPAinduced AP-1 activity. Furthermore, we used two kinds of pharmacological PI-3 kinase inhibitors to block PI-3 kinase. The first inhibitor used in our study was a fungal metabolite, wortmannin, which covalently binds to the catalytic subunit P110 of PI-3 kinase and irreversibly inhibits the enzymatic activity at nanomolar concentrations (54). The second PI-3 kinase inhibitor used in this study was LY294002. Unlike wortmannin, LY294002 reversibly inhibits PI-3 kinase by competing with ATP for its substrate binding site (33). These two inhibitors markedly inhibited AP-1 activation and cell transformation induced by TPA or TPA and insulin in a dose-dependent manner. Finally, we used a dominant negative mutant of PI-3 kinase, ⌬P85, in our study. This dominant negative mutant has been shown to specifically block PI-3 kinase activity and its mediated cell function in intact cells (24,50). The stable introduction of a dominant negative mutant of the PI-3 kinase P85 subunit (⌬P85) into JB6 cells was shown to block PI-3 kinase activity by insulin, TPA, or TPA and insulin and also to completely block TPA-or TPA-and insulin-induced AP-1 activity and cell transformation. All results from our experiments indicate that PI-3 kinase is necessary for TPAstimulated AP-1 activation and cell transformation in JB6 cells. Recent reports suggest that PI-3 kinase and p21 ras modulate each other's signals (2,21,24). In JB6 cells, TPA alone induced a low level of PI-3 kinase activity. This TPA-induced low level of PI-3 kinase activity appears to be required for AP-1 activation and cell transformation. There are several models that may explain these data. One interpretation is that the cross-talk between PI-3 kinase and p21 ras is important for the TPA-induced Ras-Raf/MAP kinase cascade leading to the AP-1 activation and cell transformation. Another interpretation is that there are some growth factors (such as IGF-I) that exist in the serum used in cell transformation and have a synergistic effect with TPA on induction of PI-3 kinase activity. The last possibility is supported by data from Fig. 5, in that the enhancement of TPA-induced AP-1 activity by insulin was only observed at a low concentration of serum.
The signaling pathways induced by insulin have been the subject of intense research (3,22,24,49). Insulin is able to bind to both IR and IGF-IR, but affinities of these two receptors for insulin are different. The affinity of IR for insulin is at least 100 times higher than that of IGF-IR (11). Insulin binding to these two receptors results in receptor-mediated tyrosine phosphorylation of IRS-1 and Shc. These molecules then function as high affinity binding sites for the P85 subunit of PI-3 kinase, and this interaction subsequently results in the activation of PI-3 kinase. In the present studies, we used 2.5 g/ml insulin as the optimal concentration of insulin for AP-1 activation and costimulation of cell transformation. In this concentration, insulin may bind to both IR or IGF-IR. To whatever receptor insulin binds, after activation of these receptors and subsequent activation of IRS-1, PI-3 kinase is the downstream target of IRS-1. Therefore, the data from our study, in which insulin enhanced the TPA-induced AP-1 activation and cell transformation, still support the concept that PI-3 kinase is required in TPA-induced AP-1 activation and cell transformation in JB6 cells.
In conclusion, we have used several approaches to study the role of the PI-3 kinase in TPA-induced AP-1 activation and cell transformation in JB6 cells. TPA could induce PI-3 kinase, and this induction effect was synergistically enhanced by insulin.
The two pharmacological inhibitors (wortmannin and LY294002) or the biological inhibitor (⌬P85, a dominant negative mutant of the PI-3 kinase P85 subunit of PI-3 kinase) markedly blocked TPA-induced AP-1 activation and cell transformation. Specific blockage of the events required for cell transformation with few side effects on normal growth might be a promising target for cancer prevention and treatment. In fact, inhibiting induced PI-3 kinase and AP-1 activity by a pharmacological inhibitor (wortmannin) or a dominant negative mutant of PI-3 kinase (⌬P85) does not seem to have inhibitory effects on cell growth in JB6 cells. Further investigation of this topic may elucidate the precise mechanisms underlying the role of PI-3 kinase in phorbol ester-induced signal transduction and may thus provide a novel target for the prevention of carcinogenesis.