Insulin-induced Activation of Phosphatidylinositol (PI) 3-Kinase INSULIN-INDUCED PHOSPHORYLATION OF INSULIN RECEPTORS AND INSULIN RECEPTOR SUBSTRATE-1 DISPLACES PHOSPHORYLATED PLATELET-DERIVED GROWTH FACTOR RECEPTORS FROM BINDING SITES ON PI 3-KINASE*

Phosphatidylinositol (PI) 3-kinase is an enzyme that functions in the signaling pathways downstream from multiple cell surface receptors. The p85 regulatory sub- unit of PI 3-kinase binds to phosphotyrosine residues of various phosphoproteins including the platelet-derived growth factor (PDGF) receptor, the insulin receptor, and insulin receptor substrate-1 (IRS-1). Using NIH-3T3 cells overexpressing the human insulin receptor, we demonstrate that the p85 regulatory subunit of PI 3-ki- nase binds to phosphorylated PDGF receptor in cells incubated in the absence of insulin. When insulin is added, p85 is released from phosphorylated PDGF receptors and binds to phosphorylated insulin receptors and insulin receptor substrate-1. Moreover, insulin-in-duced dissociation of PDGF receptors from binding sites on PI 3-kinase requires a functional insulin receptor and is not prevented by vanadate treatment. In contrast, insulin activation does not displace PDGF receptors from binding sites on Ras GTPase-activating protein. This competition for binding to PI 3-kinase provides a mechanism for cross-talk among signaling pathways ini- tiated by distinct peptide hormones and growth factors such as insulin and PDGF. cellular


Phosphatidylinositol (PI) 3-kinase is an enzyme that functions in the signaling pathways downstream from multiple cell surface receptors. The p85 regulatory subunit of PI 3-kinase binds to phosphotyrosine residues of various phosphoproteins including the platelet-derived growth factor (PDGF) receptor, the insulin receptor, and insulin receptor substrate-1 (IRS-1). Using NIH-3T3 cells overexpressing the human insulin receptor, we demonstrate that the p85 regulatory subunit of PI 3-kinase binds to phosphorylated PDGF receptor in cells incubated in the absence of insulin. When insulin is added, p85 is released from phosphorylated PDGF receptors and binds to phosphorylated insulin receptors and insulin receptor substrate-1. Moreover, insulin-induced dissociation of PDGF receptors from binding sites on PI 3-kinase requires a functional insulin receptor and is not prevented by vanadate treatment. In contrast, insulin activation does not displace PDGF receptors from binding sites on Ras GTPase-activating protein.
This competition for binding to PI 3-kinase provides a mechanism for cross-talk among signaling pathways initiated by distinct peptide hormones and growth factors such as insulin and PDGF.
Insulin resembles many peptide growth factors in that it can elicit multiple biological responses in target cells. These cellular responses include promotion of cellular growth, regulation of cell differentiation programs, and regulation of cellular metabolism. Despite the fact that individual growth factors may trigger distinct biological responses on the part of the target cell, it is remarkable that most of the receptors share in common an overlapping set of downstream signaling pathways, e.g. activation of phosphatidylinositol (PI) 1 3-kinase and activation of Ras and the protein kinase cascade involving Raf and MAP kinase. In this study, we have explored the consequences of the fact that two receptors both compete to activate the same PI 3-kinase molecules. When PDGF stimulates tyrosine phosphorylation of its receptor, this causes the p85 regulatory subunit of PI 3-kinase to bind to phosphorylated Tyr-Xaa-Xaa-Met motifs in the PDGF receptor (1,2). Similarly, when insulin binds to its receptor, this leads to phosphorylation of Tyr-Xaa-Xaa-Met motifs in the insulin receptor and insulin receptor substrate-1 (IRS-1) (3). Our data demonstrate that the binding of phosphorylated insulin receptors and/or IRS-1 to p85, competitively displaces p85 molecules from binding sites on constitutively phosphorylated PDGF receptors. This phenomenon of receptors competing to bind the same pool of downstream signaling molecules provides a novel mechanism for cross-talk among receptors for peptide growth factors. It is possible that this mechanism of cross-talk may play an important role in the regulation of cell growth, for example, the interactions among various peptide growth factors in regulating the proliferation of malignant cells in vivo.

Expression of Insulin Receptors by Transfection of cDNA in Cultured
Cells-NIH-3T3 cells were stably transfected with pBPV (Pharmacia Biotech Inc.) expression vector and cDNA encoding wild-type human insulin receptor (4) or ⌬43 insulin receptor cDNA, a truncated receptor lacking the 43 amino acid residues at the COOH terminus of the ␤-subunit (5), or Ile 1153 mutant insulin receptor (a kinase-deficient mutant insulin receptor in which Met 1153 is replaced by Ile) (6) as described previously (5). Expression of insulin receptors was assayed by measuring 125 I-insulin binding (7) and/or immunoblotting (8).
Immunoprecipitation-Confluent cells in 10-cm Petri dishes, grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, were incubated in the presence or the absence of 10 Ϫ7 M insulin or in the presence of 100 ng/ml of epidermal growth factor for 3 min at 37°C. (In some experiments, serum was omitted from the tissue culture medium.) When indicated in the figure legends, Na 3 VO 4 (100 M) was added 15 min prior to insulin stimulation. To test for the presence of an autocrine loop for PDGF, the cells were incubated with 100 g/ml of suramin 30 min prior to the stimulation with insulin in some experiments (9). The cells were quickly washed once with ice-cold phosphate-buffered saline followed by two washes with washing buffer (20 mM Tris-HCl, pH 7.5, 137 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 100 M Na 3 VO 4 ). Thereafter, the cells were solubilized in 500 l of washing buffer containing Nonidet P-40 (1%), glycerol (10%), phenylmethylsulfonyl fluoride (2 mM). After normalization for protein concentration, about one-third of the cell lysate was immunoprecipitated using either a polyclonal antibody directed against the PDGF receptor at a dilution of 1:100 (10); a polyclonal antibody directed against the p85 regulatory subunit of PI 3-kinase (Upstate Biotechnology Inc., Lake Placid, NY; catalog number, 06-195) at a dilution of 1:250; B-10, an anti-insulin receptor antibody directed against the ␣-subunit at a dilution of 1:100; or a polyclonal anti-GAP (GTPase-activating protein) antibody (Upstate Biotechnology Inc.; catalog number, 06-157) at a concentration of 6 g/ml. The immune complexes were precipitated using either protein A-agarose (Life Technologies, Inc.) in which nonspecific sites were saturated by washing with a buffer containing Tris-HCl (10 mM, pH 7.5) and albumin (10 mg/ml) or Ultra-Link Immobilized Protein A Plus (Pierce). In some experiments, in which sequential immunoprecipitations were performed, the supernatant was saved and immunoprecipitated again with either anti-PDGF receptor antibody or anti-PI3 kinase antibody at the same conditions as above. The immunoprecipitates were washed once with phosphate-buffered saline containing Nonidet P-40 (0.1%) and vanadate (100 M) and twice with a buffer containing Tris-HCl (10 mM, pH 7.5), NaCl (100 mM), EDTA (1 mM), and vanadate * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Immunoblotting-After immunoprecipitation, the complexes were boiled for 3 min in 40 l of Laemmli sample buffer containing dithiothreitol (80 mM). The samples were analyzed by SDS-polyacrylamide gel electrophoresis. Proteins were transferred from the gel to nitrocellulose sheets by electroblotting at 90 V for 1 h at 4°C in a solution containing Tris (25 mM), glycine (192 mM), and methanol (20%). The immunoblots were probed either with 4G10 monoclonal antiphosphotyrosine antibody (Upstate Biotechnology Inc.) at a concentration of 250 ng/ml or a monoclonal antibody directed against the p85 subunit of PI 3-kinase at a concentration of 1 g/ml (Upstate Biotechnology Inc.; catalog number, 05-217), a rabbit antibody directed against the human PDGF receptor (1:500), or a polyclonal anti-GAP antibody (Upstate Biotechnology Inc.) at a concentration of 2 g/ml; proteins were detected by enhanced chemiluminescence using either horseradish peroxidaselabeled anti-mouse ␥-globulin or anti-rabbit ␥-globulin (Amersham Corp.) as described elsewhere (5).

Insulin Inhibits Binding of PI 3-Kinase to Phosphorylated
PDGF Receptors-PI 3-kinase is an enzyme that functions in the signaling pathways downstream from multiple cell surface receptors. For example, the p85 regulatory subunit of PI 3-kinase binds to phosphotyrosine residues in the PDGF receptor (2), the insulin receptor (11)(12)(13), and IRS-1 (14). NIH-3T3 cells expressing recombinant human insulin receptors were incubated in the presence or absence of insulin (10 Ϫ7 M), and the extracts were immunoprecipitated with anti-p85 antibody ( Fig.  1, lanes 3 and 4). In cells incubated in the absence of insulin, the major phosphotyrosine-containing protein has M r Ϸ 190,000 (Fig. 1A, lane 3). Based upon the observation that it is immunoprecipitated by antibody to the PDGF receptor, this M r Ϸ 190,000 phophoprotein is identified as the PDGF receptor (Fig. 1A, lane 1). When cells were incubated with insulin, this led to the disappearance of the phosphorylated PDGF receptor from the immune complex immunoprecipitated by anti-p85 antibody (Fig. 1A, lane 4). However, when the extracts were immunoprecipitated by anti-PDGF receptor antibody, we observed that insulin did not alter the phosphotyrosine content of the PDGF receptor (Fig. 1A, lanes 1 and 2). Furthermore, as expected, insulin stimulation led to the appearance of two additional phosphotyrosine-containing proteins coimmunoprecipitated with p85. The M r Ϸ 95,000 band is the ␤-subunit of the insulin receptor. Previously, using the criterion of immunoprecipitation with antibody to IRS-1, we have demonstrated that the M r Ϸ 180,000 band in these cells contains phosphorylated IRS-1 (5,13). When the extracts were immunoprecipitated with anti-p85 antibody, insulin did not alter the cellular content of p85 (Fig. 1B, lanes 3 and 4). In extracts of cells incubated in the absence of insulin, antibody to the PDGF receptor co-immunoprecipitates p85 (Fig. 1B, lane 1), presumably due to the well known ability of phosphotyrosine residues in the PDGF receptor to bind to SH2 domains in the p85 subunit of PI 3-kinase (2). Interestingly, incubation of cells with insulin led to the disappearance of p85 from the immune complex immunoprecipitated by anti-PDGF receptor antibody (Fig. 1B, lanes 1  and 2). This suggests that insulin somehow led to a displacement of phosphorylated PDGF receptors from binding sites in SH2 domains of PI 3-kinase. This effect of insulin was rapid, being achieved within approximately 1 min (data not shown). In contrast, p85 was not co-immunoprecipitated with insulin receptors in extracts of cells in basal state (Fig. 1B, lane 5). However, incubation of the cells with insulin led to the appearance of p85 in immune complexes immunoprecipitated by antiinsulin receptor antibody (Fig. 1B, lane 6). This raises the possibility that phosphorylated Tyr-Xaa-Xaa-Met motifs in insulin receptors and IRS-1 bind to p85 and, as a result competitively inhibit the binding of phosphorylated PDGF receptors.
When extracts of cells are immunoprecipitated with anti-PDGF receptor antibody, two bands are detected by immunoblotting with the same antibody (Fig. 1C, lanes 1 and 2); the intensities of these bands were not altered when the cells were incubated with insulin. The upper band (M r Ϸ 190,000) comigrates with the phosphorylated PDGF receptor (Fig. 1A,  lanes 1-3). Although the identity of the lower band is not known with certainty, it is possible that it represents partially processed forms of the immature PDGF receptor. In extracts of unstimulated cells, the phosphorylated PDGF receptor was co-immunoprecipitated by anti-p85 antibody (Figs. 1C, lane 3,  and Fig. 2A, lane 3). However, treatment of the cells with insulin led to the disappearance of the PDGF receptor from anti-p85 immunoprecipitates (Figs. 1C, lane 4, and 2A, lane 4). These observations further support the conclusion that insulin leads to displacement of phosphorylated PDGF receptors from binding sites on p85. These experiments were extended by carrying out sequential immunoprecipitation studies with two antibodies, anti-p85 and anti-PDGF receptor (Fig. 2, lanes  5-8). To detect "unbound" PDGF receptors (i.e. unbound to p85), supernatants of anti-p85 immunoprecipitates were subjected to immunoprecipitation by anti-PDGF receptor antibody. Using this assay, we demonstrated that insulin markedly increased the quantity of "unbound" phosphorylated PDGF receptors ( Fig. 2A, lanes 5 and 6). We carried out similar studies in which supernatants of anti-PDGF receptor immunoprecipitates were subjected to immunoprecipitation by anti-p85 antibody. Prior to treatment of the cells with insulin, most of the p85 was bound to phosphorylated PDGF receptors (Fig. 2B,  lanes 3 and 7). However, insulin treatment markedly increased the quantity of p85 that was not bound to PDGF receptors (Fig.  2B, lane 8).
Requirement for Insulin Receptor Tyrosine Kinase Activity-When insulin binds to the insulin receptor, this leads to phosphorylation of multiple sites in the cytoplasmic domain of the ␤-subunit. One phosphorylation site (Tyr 1322 ), located near the COOH terminus, is embedded in a Tyr-His-Thr-Met sequence that is a binding site for the SH2 domains in the p85 subunit of PI 3-kinase (11)(12)(13). In addition, the insulin receptor phosphorylates IRS-1, a docking protein with multiple Tyr-Xaa-Xaa-Met motifs that also provide binding sites for the SH2 domains of p85 (14). Therefore, it seems likely that insulin-stimulated phosphorylation would provide phosphotyrosine residues with the ability to competitively inhibit phosphorylated PDGF receptors from binding to p85. Thus, we inquired whether receptor tyrosine kinase activity is required for the ability of the insulin receptor to mediate insulin-induced inhibition of PDGF receptor binding to p85. To address this question, we studied NIH-3T3 cells expressing Ile 1153 mutant insulin receptors that are deficient in receptor tyrosine kinase activity (6,15). As was observed in cells expressing wild-type insulin receptors, p85 bound to phosphorylated PDGF receptors as shown in co-immunoprecipitation studies in extracts of cells expressing Ile 1153 mutant insulin receptors incubated in the absence of insulin (data not shown). However, in contrast to what was observed with cells expressing wild-type insulin receptors, there was no significant dissociation of p85 from PDGF receptors when insulin was added to cells expressing Ile 1153 mutant insulin receptors. Similarly, in untransfected NIH-3T3 cells that express relatively few insulin receptors, insulin did not significantly affect the binding of p85 to phosphorylated PDGF receptors (data not shown).
Next, we inquired whether phosphorylation of the Tyr-His-Thr-Met sequence at the COOH terminus of the insulin receptor ␤-subunit is required to mediate insulin's ability to dissociate phosphorylated PDGF receptors from p85. To address this question, we studied cells overexpressing ⌬43 mutant insulin receptors (Fig. 3, lanes 1-4). The ⌬43 mutant is a truncated receptor that lacks 43 amino acid residues at the COOH terminus; these deleted amino acids include the phosphorylation site at Tyr 1322 . Previously, we have demonstrated that deletion of this 43-amino acid sequence from the carboxyl terminus of the insulin receptor abolishes binding of p85 to the receptor (13). This supports the conclusion that Tyr 1322 is the principal binding site for p85 on the phosphorylated insulin receptor. Cells expressing ⌬43 mutant receptors were incubated in the presence or absence of insulin. As in the experiments described above, extracts were immunoprecipitated with antibodies directed against either p85 or the PDGF receptor. Incubation of the cells with insulin led to a decrease in the quantity of p85 co-immunoprecipitated by anti-PDGF receptor antibody (Fig.  3B, lanes 1 and 3). Similarly, insulin led to a decrease in the quantity of phosphorylated PDGF receptor co-immunoprecipitated by anti-p85 antibody (Fig. 3A, lanes 2 and 4). These data demonstrate that phosphorylation of Tyr 1322 is not required for the insulin receptor to mediate the effect of insulin to displace phosphorylated PDGF receptors from binding sites on p85. Inasmuch as ⌬43 mutant receptors retain the ability to phosphorylate IRS-1 (5,16), it is likely that binding of phosphorylated IRS-1 to p85 is principally responsible for displacing PDGF receptors.
Specificity of Insulin's Effect upon the Binding of p85 to Phosphorylated PDGF Receptors-We inquired whether this effect of insulin to cause displacement of PDGF receptors from p85 was specific. To address this question, we carried out experiments in which we studied the interaction between phosphorylated insulin receptors and GAP. When cells were incubated with insulin, we observed only a slight decrease in the quantity of phosphorylated PDGF receptor that was co-immunoprecipitated by anti-GAP antibody (Fig. 4, lanes 5 and 6); this slight decrease was not reproducible in other experiments. As a control, we demonstrated that, even in extracts from cells incubated in the presence of insulin, PDGF receptors were co-immunoprecipitated by anti-GAP antibodies (Fig. 4, lanes 7  and 8). In contrast, incubation with insulin led to a near total displacement of phosphorylated PDGF receptors from anti-p85 immunoprecipitates (Fig. 4, lanes 3 and 4). Furthermore, we demonstrated that epidermal growth factor did not mimic the action of insulin to dissociate p85 from PDGF receptors in NIH-3T3 cells over-expressing insulin receptors (Fig. 3, lanes 5  and 6).
Phosphorylation of PDGF Receptors in the Basal State-

FIG. 3. COOH terminus of the insulin receptor is not required for insulin-induced inhibition of binding of PDGF receptor to p85. Confluent monolayers of stably transfected NIH-3T3 cells express-
ing ⌬43 human insulin receptor, a truncated receptor lacking the 43 amino acid residues from the COOH terminus of the ␤-subunit (lanes 1-4) were incubated in the absence (lanes 1 and 2) or the presence (lanes 3 and 4) of insulin. In the same experiment, NIH-3T3 cells expressing the wild-type human insulin receptor were incubated with epidermal growth factor (100 ng/ml) (lanes 5 and 6). After solubilization, the cell extracts were immunoprecipitated using either anti-PDGF receptor antibody (lanes 1, 3, and 5) or anti-p85 antibody (lanes 2, 4, and  6). Immunoprecipitates were analyzed by SDS-polyacrylamide (6.5%) gel electrophoresis followed by immunoblotting either with antiphosphotyrosine (panel A) or with anti-p85 antibody (panel B).
Under our experimental conditions, we observed easily detectable phosphorylation of PDGF receptors in the basal state. While phosphorylation appeared to be slightly higher when NIH-3T3 cells were incubated in the presence of serum (Fig. 5,  lanes 1-4), phosphorylated PDGF receptors were also detected in cells cultivated in the absence of serum (Fig. 5, lanes 9 -12). This suggests that basal phosphorylation of PDGF receptors did not absolutely require the presence of a factor present in fetal bovine serum. Furthermore, addition of suramin to the culture medium did not abolish basal phosphorylation of PDGF receptors (Fig. 5, lanes 5-8). This suggests that phosphorylation was not due to autocrine stimulation by endogenously secreted PDGF (9).
Nevertheless, PDGF receptors were not maximally phosphorylated in the basal state. For example, if vanadate is added to the culture medium, this led to a marked increase in the phosphotyrosine content of PDGF receptors (Fig. 5, lane 13). Under these conditions, insulin retained its ability to effect a total displacement of phosphorylated PDGF receptors from p85 (Fig. 5, lane 14). Moreover, the presence of vanadate plus insulin led to a marked increase in the phosphorylation of a band with slightly higher mobility than the PDGF receptor, presumably corresponding to IRS-1 (Fig. 5, lane 14). DISCUSSION The receptors for insulin and PDGF share a common signaling pathway involving PI 3-kinase. The p85 regulatory subunit of PI 3-kinase binds to phosphotyrosine residues in various phosphoproteins including the PDGF receptor, the insulin receptor, and IRS-1 (2,12,14,17). Using NIH-3T3 cells overexpressing the human insulin receptor, we demonstrate in the present study that insulin decreases the binding of the p85 the regulatory subunit of PI3-kinase to constitutively phosphorylated PDGF receptors.
Cross-talk between Receptors for PDGF and Insulin via PI 3-Kinase-In our experimental system, there is detectable phosphorylation of tyrosine residues in the PDGF receptor, even in the basal state. Furthermore, these constitutively phosphorylated PDGF receptors bind to p85. Incubation of the cells with insulin stimulates phosphorylation of the insulin receptor and IRS-1. Possibly as a consequence of the binding of p85 to phosphotyrosine residues in IRS-1, PDGF receptors are no longer found in association with p85. Insulin receptor tyrosine kinase activity is required for the ability of insulin to promote the dissociation of p85 from PDGF receptors (data not shown). However, the Tyr-Xaa-Xaa-Met motif (Tyr 1322 -His-Thr-Met) in the COOH terminus of the insulin receptor ␤-subunit is not required (Fig. 3). This latter observation is consistent with the hypothesis that it is IRS-1 rather than the insulin receptor that directly competes with PDGF receptors for binding of p85. Moreover, vanadate treatment does not prevent the release of PDGF receptor from p85 binding; this observation suggests that the dissociation of p85 is not due to dephosphorylation of PDGF receptors by insulin-activated phosphotyrosine phosphatases.
Specificity of Insulin Action-The ability of the p85 subunit of PI 3-kinase to bind to multiple phosphotyrosine-containing proteins is not unique among SH2-domain containing intracellular proteins. In fact, many such proteins function downstream in the signaling pathways initiated by multiple receptor tyrosine kinases (18 -22). For example, GAP binds to phosphotyrosine residues in the receptors for PDGF and epidermal growth factor (among other receptors) (20), but not to either the insulin receptor or IRS-1. Assuming our hypothesis is correct, one would predict that insulin-induced tyrosine phosphorylation would not lead to competitive displacement of GAP from binding sites on the PDGF receptor. Indeed, as predicted, we did not detect binding of GAP to either IRS-1 or the insulin receptor; moreover, insulin did not significantly alter the binding of GAP to the PDGF receptor.
Constitutive Phosphorylation of PDGF Receptors-Unlike many other published reports, there was a detectable level of tyrosine phosphorylation in PDGF receptors in our NIH-3T3 cells. Nevertheless, we confirmed that PDGF (both PDGF-A and PDGF-B) increased the phosphotyrosine content of PDGF receptors. However, the increment in phosphorylation due to PDGF was approximately 2-fold, considerably less than the magnitude of effect that is usually reported (10,23). We wondered whether the basal phosphorylation might be the result of an autocrine loop due to endogenous production of PDGF. However, the observation that suramin did not decrease receptor phosphorylation made this explanation unlikely. Based upon the report that the E5 protein of bovine papilloma virus increases the phosphorylation of the PDGF receptor (24,25), we also considered the possibility that the E5 protein might contribute to the increased phosphorylation of the PDGF receptor. However, we did not detect the presence of the E5 protein in cells transfected with the pBPV expression vector. Moreover, FIG. 4. Insulin stimulation slightly decreases PDGF receptor binding to GAP. Cell extracts of NIH-3T3 cells expressing the human insulin receptor were immunoprecipitated using either anti-PDGF receptor antibody (lanes 1 and 2), anti-p85 antibody (lanes 3 and 4), or anti-GAP antibody (lanes 5-8) as described under "Materials and Methods." Immunoprecipitates were analyzed by SDS-polyacrylamide gel electrophoresis followed by immunoblotting either with antiphosphotyrosine (lanes 1-6) or with anti-GAP antibody (lanes 7 and 8).
Significance of Receptor Cross-talk-Assuming that the effect of insulin is due to competition between IRS-1 and PDGF receptors for binding to p85, one would predict that the effect of insulin would be favored by the presence of large concentrations of phosphorylated IRS-1. Therefore, because overexpression of recombinant insulin receptors increases the cellular content of phosphorylated IRS-1 molecules (26), this is predicted to maximize the ability of insulin to promote the dissociation of p85 from PDGF receptors. Nevertheless, it seems likely that this phenomenon occurs to some extent in target cells that express physiological numbers of insulin receptors. What, then, is the physiological significance of this type of receptor cross-talk in which multiple phosphotyrosine-containing proteins compete for the binding of p85? While the answer to this question is not clear, several possibilities deserve consideration. First, multiple proteins containing phosphorylated Tyr-Xaa-Xaa-Met motifs share in common the ability to bind to p85, thereby activating PI 3-kinase. Nevertheless, it is possible that some phosphoproteins may exert large effects and other phosphoproteins may exert small effects upon PI 3-kinase activity. Second, different phosphoproteins may be located in different subcellular compartments. For example, the PDGF receptor is located at the plasma membrane while IRS-1 is a cytosolic protein. If 3-phosphorylated inositol derivatives interact with multiple proteins located in distinct subcellular compartments, the subcellular localization of PI 3-kinase may determine distinct signaling functions for the same messenger molecule. Third, PI 3-kinase may not function in isolation but as part of signaling complex. Thus, when PI 3-kinase binds to the PDGF receptor, it is located in close proximity to one set of signaling molecules; when PI 3-kinase binds to IRS-1, it is located in proximity to a different set of signaling molecules. It is possible that this sort of combinatorial specificity may determine which biological responses are initiated in the target cell.
In conclusion, many intracellular signaling molecules such as PI 3-kinase function in pathways downstream from receptors for multiple growth factors and peptide hormones. These overlapping signaling pathways create the possibility of many different types of interactions among the various pathways. In principle, these interactions may be additive, synergistic, or antagonistic. In this investigation, we have demonstrated the possibility of competitive interactions whereby insulin-stimulated tyrosine phosphorylation causes the p85 subunit of PI 3-kinase to dissociate from the PDGF receptor and rather to associate with IRS-1 and/or the insulin receptor. Future studies may provide important insights into the physiological sig-nificance of this type of receptor cross-talk.