Insulin-like Growth Factor I-mediated Degradation of Insulin Receptor Substrate-1 Is Inhibited by Epidermal Growth Factor in Prostate Epithelial Cells*

We have sought to determine whether insulin-like growth factor I (IGF-I) regulates the levels of insulin receptor substrate-1 (IRS-1) in prostate epithelial cells. Exposure of prostate epithelial cells to IGF-I in the absence of other growth factors leads to a reduction in IRS-1 levels. Ubiquitin content of IRS-1 is increased in the presence of IGF-I, and inhibitors of the proteasome prevented the reduction of IRS-1 levels seen following IGF-I exposure. These results imply that IRS-1 is targeted to the proteasome upon exposure to IGF-I. The addition of epidermal growth factor (EGF) maintained IRS-1 levels even in the presence of IGF-I and inhibits IGF-I-dependent ubiquitination of IRS-1. Thus, these two growth factors, IGF-I and EGF, had antagonistic effects on IRS-1 protein levels in prostate epithelial cells. This regulation of IRS-1 reveals a novel level of cross-talk between the IGF-I and EGF signal pathways, which may have implications in tumors that harbor activating mutations in the EGF receptor.

Insulin and the insulin-like growth factors I and II (IGF-I and IGF-II) 1 are a family of structurally related polypeptide growth factors. Insulin is primarily involved in metabolic responses, specifically glucose metabolism (1,2), whereas IGF-I and IGF-II are mediators of cell proliferation and regulate body size during both prenatal and postnatal development (3)(4)(5). The biologic responses of both IGF-I and IGF-II are mediated by the insulin-like growth factor type 1 receptor (IGF-IR), which shows 80% homology to the insulin receptor (IR) (6). The IGF-IR is a heterotetrameric transmembrane glycoprotein composed of two extracellular ␣ subunits and two transmembrane ␤ subunits. The ␣ and ␤ subunits of the IGF-IR are synthesized as a single polypeptide that undergoes proteolytic cleavage at amino acid 710. The ␣ and ␤ subunits are linked in the mature IGF-IR by disulfide bonds (7). The ␤ subunits of the IGF-IR possess intrinsic tyrosine kinase activity that is activated upon ligand binding to the ␣ subunits. This tyrosine kinase activity is essential for the biologic activity of the IGF-IR, and mutation of the conserved ATP binding site at lysine 1003 abolishes kinase activity and mitogenic response to IGF-I. The activated IGF-IR associates with several intracellular targets in addition to the ␤ subunit itself. These intracellular targets include the proteins Shc, Grb 2, and the insulin receptor substrate 1 (IRS-1) (8).
The IRS-1 protein is part of a family of related proteins that includes at least four members, IRS-1, IRS-2 IRS-3, and IRS-4 (8 -13). The IRS proteins contain a conserved, amino-terminal pleckstrin homology domain adjacent to a phosphotyrosine binding domain. The phosphotyrosine binding domain of IRS-1 interacts with phosphorylated NPEY motifs including tyrosine 950 in the ␤ subunit of the IGF-IR, a site that is phosphorylated following ligand binding (14). IRS-1 contains 18 potential tyrosine phosphorylation sites that may be phosphorylated following activation of the IGF-IR. These phosphotyrosine motifs serve as binding sites for the Src homology 2 domain containing proteins such as Nck, Grb2, Grb 10, Shc, and the p85 regulatory subunit of PI 3-kinase (15). These protein-protein interactions subsequently activate a series of protein-serine kinase cascades characterized by the activation of MAP kinases and PI 3-kinase that ultimately affect cellular processes such as protein synthesis, glycogen metabolism, and transcription of specific gene targets.
Signals generated by IRS-1 phosphorylation are regulated through a variety of mechanisms. For example, serine phosphorylation of IRS-1 decreases tyrosine phosphorylation of IRS-1 (16) and association of IRS-1 with 14 -3-3 decreases IRS-1-associated PI 3-kinase activity (17,18). Treatment of cells with okadaic acid results in an increase in serine/threonine phosphorylation of IRS-1 concomitant with decreased tyrosine phosphorylation and decreased association with the insulin receptor (16,19). At least one specific residue, serine 612, is phosphorylated by both protein kinase C (20) and the extracellular signal-regulated kinase (21). Serine phosphorylation of IRS-1 is dependent upon multiple intracellular kinases. As mentioned above, both protein kinase C and the MAP kinase cascade phosphorylate IRS-1. In addition, PI 3-kinase-dependent signals including AKT-1/PKB (protein kinase B) and PI 3-kinase itself also phosphorylate IRS-1 (22)(23)(24)(25).
Phosphorylation of IRS-1 is not the only mechanism of regulation for this protein, the levels of IRS-1 are also subject to regulation. For example, chronic insulin treatment leads to decreased IRS-1 levels in 3T3 L1 adipocytes (26) and attenuation of insulin response (26 -28). In this system, IRS-1 is ubiquitinated and degraded via the proteasome pathway (29). IGF-I also down-regulates IRS-1 in a proteasome-dependent manner in human breast carcinoma MCF-7 cells. We have examined the possibility that IRS-1 levels are regulated by exposure to IGF-I. Using prostate epithelial cells, we find that IGF-I treatment targets IRS-1 for degradation via the proteasome. Interestingly, the addition of epidermal growth factor (EGF) prevents the IGF-I-mediated degradation of IRS-1. Thus, we have identified a previously unidentified mechanism of IRS-1 regulation in response to IGF-I in epithelial cells, and our results indicate that cross-talk between the IGF-I R and the epidermal growth factor receptor serve to regulate IRS-1 protein levels.

EXPERIMENTAL PROCEDURES
Materials-CPTX 1532 and NPTX 1532 cells (30) were cultured in Keratinocyte-SFM (Life Technologies, Inc.) with supplements supplied by the vendor (bovine pituitary extract and EGF) and additionally supplemented with 10% fetal bovine serum (Life Technologies, Inc.). IGF and EGF were purchased from Intergen (Purchase, NY). RPMI and methionine-free RPMI were purchased from Life Technologies, Inc. Cycloheximide, PD98059, LY294002, chloroquine, lactocystin, and MG132 were purchased from Biomol (Plymouth Meeting, PA). The anti-IRS-1 antibody was from Upstate Biotechnology (Lake Placid, NY). The anti-␤-actin antibody was purchased from Sigma. The anti-ubiquitin antibody was from Santa Cruz Biotechnology (Santa Cruz, CA).
Cell Culture-CPTX 1532 and NPTX 1532 cells were maintained in keratinocyte medium with supplements and passaged every 3-4 days using a seeding density of 1 ϫ 10 4 cells/cm 2 . All experiments followed the same protocol to transfer the cells from the growth factor-rich, undefined Keratinocyte medium into a defined medium containing only IGF-I and/or EGF. The cultures were grown to 70% confluence and washed three times with phosphate-buffered saline (PBS). Serum-free RPMI medium was then added to the cultures with the indicated concentrations of IGF-I or EGF. The cultures were viable in the RPMI medium for a period of 2-3 days, but this medium did not support sustained proliferation. For the purposes of these experiments, the cells were incubated in the serum-free RPMI medium for up to 24 h. During this period, the cells are completely viable with no loss of cell number, and the cultures will proliferate normally if placed into complete keratinocyte medium.
Western Blot and Immunoprecipitation-CPTX cell extracts were prepared by scraping cells into a lysis solution consisting of 50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.25% sodium deoxycholate, 160 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, leupeptin, and pepstatin, 1 mM Na 3 VO 4 , and 1 mM NaF. For Western analysis, equal amounts of protein (assayed using a protein assay based upon the Bradford method (31) purchased from Bio-Rad) were resolved by SDSpolyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto nitrocellulose. The IRS-1 was analyzed using a standard Western blot protocol. For immunoprecipitation, cell extracts were prepared as described above. The lysates were incubated with IRS-1 antibody, collected with protein A-Sepharose beads, washed with PBS, resolved by SDS-PAGE, and transferred onto membranes. IRS-1-conjugated ubiquitin was analyzed using standard Western blot procedures.
Antibodies were removed from nitrocellulose filters by incubating the filters in a solution of 100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7, at 50°C for 30 min with gentle agitation. The membranes were then washed twice with Tris-buffered saline containing 0.5% Triton X-100 prior to Western blot analysis.
Ubiquitination of IRS-1-Ubiquitinated IRS-1 was visualized by transfecting CPTX cells with hemagglutinin (HA)-tagged ubiquitin cDNA in an expression construct (BSSK) provided by Dr. Dirk Bohmann (49). Cells were transfected using LipofectAMINE 2000 for 4 h in serum-free medium, a protocol that provided a transfection efficiency of ϳ5% (determined using a ␤-galactosidase reporter construct). 24 h following transfection, cultures were washed with PBS three times and placed into RPMI medium containing IGF-I (40 ng/ml), EGF (40 ng/ml), and MG132 (10 M) in various combinations as described in the individual experiments. Total cellular proteins were isolated after a 3-h exposure to individual growth factors, and the IRS-1 protein was immunoprecipitated as described above. Immunoprecipitated proteins were separated by SDS-PAGE and ubiquitinated IRS-1 was visualized using an anti-HA monoclonal antibody 12CA5 (Roche Molecular Biochemicals). The IRS-1 protein was visualized by Western blot analysis on the same blot following removal of the anti-HA antibody.
Metabolic Labeling of IRS-1-CPTX 1532 cells were seeded in tissue culture flasks in complete keratinocyte SFM as described above. For measurements of IRS-1 degradation, cells were labeled with 1 Ci/ml [ 35 S]methionine for 4 h. At this time, the cells were washed three times with PBS and placed into RPMI medium containing either EGF (40 ng/ml) or IGF-I (40 ng/ml). Total cellular proteins were isolated at time intervals, and equal amounts of protein were used to immunoprecipitate the IRS-1 protein using an anti-IRS-1 polyclonal antibody (Upstate Biotechnology Inc., Saranac Lake, NY). Immunoprecipitated proteins were separated by SDS-PAGE and transferred to a nitrocellulose filter. 35 S-labeled IRS-1 was visualized by autoradiography. Measurements of the synthesis of IRS-1 involved placing CPTX cells into RPMI medium containing either EGF or IGF-I for 4 h. At this time, [ 32 S]methionine (1 Ci/ml) was added for 1 h. Total cellular proteins were then isolated, and the IRS-1 protein was immunoprecipitated, separated by SDS-PAGE, and visualized by autoradiography following transfer to nitrocellulose filters. The IRS-1 protein was positively identified on the autoradiograms by comparison with a Western blot of the same filter.
The relative amount of 35 S-labeled IRS-1 was quantitated by densitometry of the autoradiograms using a Bio-Rad Flour S system equipped with the Quantity One software, which is designed to provide quantitative measurements of densitometric scans using background subtraction algorithms.

RESULTS
We examined whether IGF-I exposure alters IRS-1 levels in a prostate epithelial cell line, CPTX 1532, which is derived from a primary carcinoma of the prostate (30). When these cells are placed into serum-free RPMI medium, the levels of the IRS-1 protein are relatively constant, whereas the addition of IGF-I to the medium causes a decline in IRS-1 protein levels (Fig. 1). The levels of IRS-1 are dramatically reduced following exposure to IGF-I for 24 h (Fig. 1). The decline in IRS-1 protein levels in response to IGF-I is dose-dependent with a maximal decrease in IRS-1 at 40 -60 ng/ml of IGF-I (Fig. 2). This range of IGF-I is similar to the optimal concentration required for proliferation, which is 20 -40 ng/ml (32). Interestingly, there is CPTX 1532 cells were exposed to increasing concentrations of IGF-I in serum-free RPMI medium. As described in Fig. 1, the cells were first grown to 70% confluence and then washed with PBS. Subsequently, the cells were exposed to serum-free RPMI 1640 medium containing the indicated concentrations of IGF-I. IRS-1 levels were determined by Western blot analysis of total cellular protein following 16 h. As a control for protein loading, the levels of ␤-actin were measured on the same blot following removal of the anti-IRS-1 antibody (see "Experimental Procedures"). The experiment shown is representative of three independent experiments, which all gave similar results. an increase in IRS-1 protein in cultures that are exposed to low levels (2 ng/ml) of IGF-I. This observation was consistent in several experiments but has not been investigated further at this time. A second prostate epithelial cell line derived from normal prostate epithelial tissue, NPTX 1532 (30), also responded to IGF-I treatment with a decline in IRS-1 protein levels (Fig. 1).
The decrease in IRS-1 is specific to IGF-I, because treatment of CPTX 1532 cells with EGF over the same time period did not cause a decrease in IRS-1 levels (Fig. 1). In CPTX 1532 cells the addition of EGF or IGF-I alone will support modest proliferation (10 -50% increase in cell number) for 24 -48 h, and the addition of EGF plus IGF-I enhances proliferation (100 -200% increase in cell number). Because EGF and IGF-I cooperate to stimulate proliferation of CPTX 1532 cells, it was of interest to determine whether the addition of EGF would mitigate the reduction in IRS-1 levels observed following IGF-I treatment.
The addition of EGF simultaneously with IGF-I prevents the decrease in IRS-1 levels (Fig. 3), and this effect is dependent upon the relative ratio of IGF-I to EGF. The simultaneous addition of EGF and IGF-I prevented the IGF-I-mediated decline in IRS-1 protein levels even when EGF was present at a 3-fold lower molar ratio (Fig. 3). The addition of increasing amounts of EGF lead to a progressive increase in the level of IRS-1 in the face of an amount of IGF-I that normally induced a reduction in IRS-1 protein levels. Thus, treatment of these prostate epithelial cell lines with either EGF or IGF-I had opposing effects on the levels of IRS-1, a critical component of the IGF-IR signaling complex.
It has been suggested that IRS-1 degradation may be mediated by the proteasome (29,50). A specific inhibitor of the proteasome, lactocystin, has been described (34). The addition of lactocystin to CPTX 1532 cells simultaneously with IGF-I prevented the IGF-I-mediated decline in IRS-1 protein levels (Fig. 4), implying that the proteasome may be involved in this process. A second inhibitor of the proteasome, MG132, also prevented the IGF-I-mediated decline in IRS-1 levels (not shown). The addition of a drug that inhibits receptor-mediated endocytosis, chloroquine (35), had no effect on IGF-I-mediated IRS-1 decline. Similarly, the addition of PD98059, which inhibits the MAP kinase pathway (36), has no effect on IGF-I-mediated decreases in IRS-1 protein levels (Fig. 4). However, inhibition of the PI 3-kinase enzyme using LY294002 (37) (Fig. 4) prevented the IGF-I-mediated decline in IRS-1 levels. These results are consistent with those obtained following insulin treatment of adipocytes. In that system, inhibition of PI 3-kinase prevented the insulin-mediated decline in IRS-1 levels (38).

FIG. 3. EGF prevents IGF-I-mediated decline in IRS-1 protein levels.
CPTX 1532 cells were grown to 70% confluence as described under "Experimental Procedures." The cells were washed with PBS and exposed to serum-free RPMI 1640 medium containing EGF and IGF-I in varying concentrations. The ratio of EGF:IGF-I was altered to determine the effects on IRS-1 protein levels. The concentration of IGF-I used in this experiment was 40 ng/ml, whereas the concentration of EGF was varied to achieve the indicated ratio of the two growth factors. CPTX 1532 cells were exposed to the indicated EGF:IGF-I ratio for 16 h, and total cellular proteins were isolated and subjected to Western blot analysis for the IRS-1 protein and ␤ actin, as a control for protein content. Similar results have been obtained in two independent experiments.  (51). Following transfection, the cells were exposed to the following conditions: RPMI medium without additions (Serum Free), RPMI containing IGF-I at 40 ng/ml (IGF-I), RPMI containing IGF-I at 40 ng/ml, and MG 132 at 10 M (IGF-I ϩ MG132), RPMI containing IGF-I at 40 ng/ml and EGF at 40 ng/ml (IGF-I ϩ EGF), and RPMI containing IGF-I at 40 ng/ml and EGF at 40 ng/ml in the presence of MG132 at 10 M (IGF-I ϩ EGF ϩ MG 132). The cells were exposed to these conditions for 3 h, and the IRS-1 protein was immunoprecipitated and subjected to SDS-PAGE and Western blot analysis. In A, a Western blot using an anti-HA antibody is shown, which recognizes the transfected HA-ubiquitin and thus will detect the ubiquitinated form of IRS-1. The anti-HA antibody was removed, and the same blot was reprobed with an anti-IRS-1 antibody in B. Higher molecular weight forms of IRS-1 can be seen in longer exposures of the blot shown in B. Similar results have been obtained using antibodies against the endogenous ubiquitin.
The fact that inhibition of the proteasome prevented the IGF-I-mediated decline in IRS-1 protein levels suggests that IGF-I stimulation targets IRS-1 to the proteasome. A common mechanism for the targeting of proteins to the proteasome is the covalent attachment of ubiquitin (50). Because ubiquitinated proteins are labile and anti-ubiquitin antibodies are of low affinity, we used an experimental protocol designed to enhance our ability to visualize IRS-1-associated ubiquitin. First, the time period of the exposure to growth factors was relatively short (3 h compared to 16 -24 h used in Figs. 1 and 2). Second, parallel cultures were treated with inhibitors of the proteasome such as lactocystin or MG132, and finally, the CPTX 1532 cells were transiently transfected with an HAtagged ubiquitin cDNA in a mammalian expression vector (51) to allow the use of a high affinity anti-HA monoclonal antibody (clone 12CA5). Following immunoprecipitation with an anti-IRS-1 antibody, the ubiquitin content of IRS-1 was examined by Western blot using the anti-HA antibody (Fig. 5). The ubiquitin content of the IRS-1 protein increased when CPTX 1532 cells were exposed to IGF-I, and the addition of an inhibitor of the proteasome, MG132, caused a dramatic increase in IRS-1associated ubiquitin. The addition of fetal bovine serum or EGF does not cause any increase in the ubiquitin content of IRS-1, and the addition of equimolar amounts of EGF decreased the ubiquitin content of IRS-1. This effect is most evident in samples treated with IGF-I, MG 132, and EGF (Fig. 5). These results indicate that increased ubiquitin content of the IRS-1 protein is specific for IGF-I treatment, and the presence of EGF prevented this IGF-I-mediated ubiquitination.
To determine whether the rate of IRS-1 synthesis is affected by either IGF-I or EGF, CPTX 1532 cells were exposed to IGF-I or EGF alone for 4 h and then were pulse-labeled with [ 35 S]methionine for 1 h, and the amount of labeled IRS-1 was examined. This provides a quantitative measure of the rate of IRS-1 synthesis in the presence of IGF-I or EGF. The rate of IRS-1 synthesis was similar whether the cells were exposed to IGF-I or to EGF (Fig. 6A). Thus, the rate of IRS-1 synthesis does not seem to be influenced by either IGF-I or EGF. To directly measure the rate of degradation of IRS-1 in the presence of IGF-I or EGF, CPTX 1532 cells were again labeled with [ 35 S]methionine. In this case, the cells were exposed to the [ 35 S]methionine in the presence of normal growth medium (keratinocytes SFM supplemented with 10% fetal bovine serum). Following the labeling period, the cells were exposed to either IGF-1 alone (40 ng/ml) or EGF alone (40 ng/ml), and the labeled IRS-1 was examined. Over a 14-h period the level of labeled IRS-1 decreased in the presence of IGF-I, whereas the level of labeled IRS-1 was relatively constant in the presence of EGF (Fig. 6B). Thus, IRS-1 degradation proceeds more rapidly when CPTX 1532 cells are exposed to IGF-I than when the cells are exposed to EGF. DISCUSSION This study identifies a novel level of regulation for the IRS-1 protein in response to IGF-I stimulation in prostate epithelial cells. The results presented here indicate that IRS-1 levels fluctuate following exposure to either IGF-I or EGF. It seems that, similar to adipocytes exposed to insulin, IGF-I negatively regulates IRS-1 levels in prostate epithelial cells. Surprisingly this effect is counterbalanced by the effects of EGF, which decreases and serves to maintain IRS-1 levels in the presence of IGF-I. This is the first demonstration that an extracellular factor (EGF) can positively influence IRS-1 levels.
The fact that EGF is able to modulate the levels of a signaling molecule critical to IGF-I-dependent responses implies a novel level of cross-talk between the EGF and IGF-I systems. These two growth factors cooperate to stimulate progression of fibroblast cells through G 1 and into S phase (39). In BALB 3T3 cells, there is a temporal requirement for the addition of EGF and IGF-I. EGF must be present prior to IGF-I treatment. If IGF-I is added prior to EGF, and subsequently removed, BALB 3T3 cells do not proceed into S phase (39 -43).
This information, combined with the results presented here, suggests that one mechanism underlying this cooperation between EGF and IGF-I in prostate epithelial cells is the maintenance of IRS-1 levels. Addition of EGF antagonizes the IGF-I-dependent, proteasome-mediated degradation of IRS-1 by preventing ubiquitination of IRS-1. Because IRS-1 phosphorylation regulates the activity of several critical intracellular enzymes activated by the IGF-IR, a change in the level of IRS-1 may have a profound effect on the ability of the cell to respond to mitogenic stimulation by IGF-I. In this regard, recent results Immunoprecipitated IRS-1 was separated by SDS-PAGE and detected by autoradiography. The relative abundance of the labeled IRS-1 was measured by densitometry performed on the autoradiogram (A). Shown below the autoradiogram is a quantitative result derived from three independent measurements of IRS-1 in the presence of either IGF-I or EGF. B, degradation of IRS-1 in the presence of either IGF-I or EGF was examined by labeling CPTX cells with [ 35 S]methionine in complete growth medium for 4 h. At this time, the cells were exposed to IGF-I (40 ng/ml) of EGF (40 ng/ml). The IRS-1 protein was immunoprecipitated at the indicated times (2, 4, 8, and 14 h), separated by SDS-PAGE, and detected by autoradiography. The relative abundance of labeled IRS-1 was determined by densitometry performed on the autoradiogram. The graph shown below the autoradiogram contains a quantitative analysis of the relative abundance of IRS-1 at each time point corrected for background in each lane. using 32 D cells indicate that the levels of IRS-1 dictate the degree of mitogenic response to IGF-I (44). We should note that the effect of IGF-I on IRS-1 seems to be restricted to epithelial cells as the levels of IRS-1 were not altered by IGF-I treatment of WI-38 fibroblasts (data not shown). This implies that the molecular mechanisms responsible for the cooperation of growth factors in terms of mitogenic responses may vary in different cell types.
In terms of cellular responses, IGF-I can stimulate proliferation or differentiation in several experimental settings. For example differentiation of muscle cell cultures is dependent upon IGF-I (45,46), yet IGF-I is also mitogenic for muscle cells (47). Similarly, differentiation of 32 D cells in response to IGF-I is dependent upon IRS-1 (44). Thus, modulation of IRS-1 levels may play a role in the type of response generated by IGF-I in prostate epithelial cells. In this light it is interesting that low levels of IGF-I actually increase IRS-1 levels in the prostate epithelial cells used in this study (see Fig. 2). Although the present study has not examined this effect in detail, the response was consistent in several experiments.
Degradation of IRS-1 via the proteasome has been reported to occur in adipocytes (29) and in human breast carcinoma cells (49). Thus, a proteasome-mediated degradation of IRS-1 may have relevance in a number of cell types in addition to the prostate epithelial cells reported here. However, a modulating effect by a physiologically relevant factor such as EGF has not been reported in any other cell type, and it remains to be seen if this is a general phenomenon.
The inhibitor studies presented here indicate that neither receptor endocytosis (inhibited by addition of chloroquine) nor inhibition of the MAP kinase pathway (see effects of PD98059 in Fig. 4) prevents the IGF-I-mediated decline in IRS-1 levels. It seems that ubiquitination of IRS-1 may occur prior to receptor endocytosis. Consistent with this, it has been reported that IRS-1 phosphorylation by the insulin receptor is not linked to receptor endocytosis, whereas Shc phosphorylation is linked to receptor endocytosis (48). Interestingly, the activation of PI 3-kinase is required for the down-regulation of IRS-1 in response to IGF-I (see effect of LY294002 in Fig. 4). Because PI 3-kinase activation is also required for insulin-mediated down-regulation of IRS-1 (38), the signal transduction pathways, which trigger IRS-1 decreases in response to IGF-I and insulin, may be similar.
In summary, our results indicate that the targeted degradation of IRS-1 via the proteasome occurs in response to IGF-I in prostate epithelial cells. In addition, EGF, a physiologically relevant growth factor known to act synergistically with IGF-I, prevents the degradation of IRS-1 and may enhance the mitogenic response to IGF-I. This information provides insight into a novel level of interaction between the EGF and IGF-I receptor signaling pathways that may be of particular importance in tumors harboring activating mutations in the EGF receptor.