Insulin-like growth factor-I receptor internalization regulates signaling via the Shc/mitogen-activated protein kinase pathway, but not the insulin receptor substrate-1 pathway.

Insulin-like growth factor-I (IGF-I) receptors activate divergent signaling pathways by phosphorylating multiple cellular proteins, including insulin receptor substrate-1 (IRS-1) and the Shc proteins. Following hormone binding, IGF-I receptors cluster into clathrin-coated pits and are internalized via an endocytotic mechanism. This study investigates the relationship between IGF-I receptor internalization and signaling via IRS-1 and Shc. A mutation in the C terminus of the IGF-I receptor decreased both the rate of receptor internalization and IGF-I-stimulated Shc phosphorylation by more than 50%, but did not affect IRS-1 phosphorylation. Low temperature (15 degrees C) decreased IGF-I receptor internalization and completely inhibited Shc phosphorylation. Although receptor and IRS-1 phosphorylation were decreased in accordance with delayed binding kinetics at 15 degrees C, the ratio of IRS-1 to receptor phosphorylation was increased more than 2-fold. Dansylcadaverine decreased receptor internalization and Shc phosphorylation, but did not change receptor or IRS-1 phosphorylation. Consistent with these findings, dansylcadaverine inhibited IGF-I-stimulated Shc-Grb2 association, mitogen-activated protein kinase phosphorylation, and p90 ribosomal S6 kinase activation, but did not affect the association of phosphatidylinositide 3-kinase with IRS-1 or activation of p70 S6 kinase. These data support the concept that Shc/mitogen-activated protein kinase pathway activation requires IGF-I receptor internalization, whereas the IRS-1 pathway is activated by both cell surface and endosomal receptors.

Insulin-like growth factor-I (IGF-I) 1 initiates pleiotropic cellular growth and metabolic responses by binding to the type I insulin-like growth factor receptor (1). The activated intrinsic tyrosine kinase of the IGF-I receptor catalyzes receptor autophosphorylation and the tyrosine phosphorylation of early signaling intermediates, including insulin receptor substrate-1 (IRS-1) and the Shc proteins (2)(3)(4). IRS-1 contains multiple phosphotyrosine residues that serve as docking sites and activators of a number of signaling molecules, including phosphatidylinositide (PI) 3-kinase (5,6). Stimulation of PI 3-kinase leads to the activation of downstream mediators of IGF-I action, including p70 S6 kinase, a serine/threonine kinase that phosphorylates 40 S ribosomal protein S6 (7). The Shc proteins, which include several isoforms (46,52, and 66 kDa), contain specific tyrosine phosphorylation sites necessary for their association with Grb2 and the consequent activation of the GTPbinding protein Ras (8,9). Activation of Ras initiates a series of phosphorylation events that results in the dual phosphorylation of the mitogen-activated protein (MAP) kinases, ERK1 and ERK2 (10). These MAP kinases can phosphorylate a number of substrates, including p90 ribosomal S6 kinase (11), a protein kinase that can activate several transcription factors in the nucleus (12,13). Thus, tyrosine phosphorylation of IRS-1 and Shc activates divergent pathways that are ultimately important for cell metabolism, growth, and differentiation (14). However, little is known about cellular mechanisms that may regulate the relative intensity of signaling via these intermediates in response to IGF-I.
Previous studies have suggested a role for ligand-induced receptor trafficking in modulating the activation of intracellular signaling molecules (15,16). Based in part on direct evidence and on homologies with the insulin signaling pathway (17,18), it is believed that IGF-I binding initiates the migration of IGF-I receptors to clathrin-coated pits and the subsequent formation of early endosomes containing internalized but still active receptors. The ligand-receptor complex ultimately becomes dissociated and inactivated in the acidic environment of late endosomes, where ligands and receptors are sorted for degradation in lysosomes or recycling to the cell surface (17,18). Activated IGF-I receptors at the cell surface, in coated pits, and in early endosomes have the potential to initiate specific cellular responses. For other hormone-receptor complexes, endosomal compartments have been shown to participate in the initiation and/or continuation of intracellular signaling reactions (15). Hormone-stimulated tyrosine phosphorylation of the epidermal growth factor receptor begins at the cell surface but persists within the endosome compartment, such that there is a greater level of receptor phosphorylation in the endosome than in the plasma membrane (19). Generation of ceramide by tumor necrosis factor receptors in endosomes initiates the activation of NF-B, while ceramide synthesis at the plasma membrane directs the activation of serine/threonine protein kinases and phospholipase A 2 (20). Thus, hormone signaling can occur at multiple steps in receptor internalization pathways, and this may provide a mechanism for regulating the intensity and pattern of signals generated by activated receptors.
The relationship between ligand-induced IGF-I receptor internalization and the activation of specific cellular substrates that lead to signaling responses has not been studied. We and others have observed that the time courses of IGF-I stimulated IRS-1 and Shc phosphorylation are different (21,22). IRS-1 is maximally phosphorylated within 1 to 2 min of IGF-I stimulation, whereas Shc phosphorylation has been shown to reach a maximum only after 5 to 10 min (21,22). The time course of Shc but not IRS-1 phosphorylation correlates with the time required for ligand-induced receptor internalization (23). The objective of the current study was to determine whether IRS-1 and Shc phosphorylation results from their interaction with IGF-I receptors at distinct sites in the ligand-induced internalization pathway.

EXPERIMENTAL PROCEDURES
Materials-Recombinant human IGF-I was provided by Eli Lilly Inc. (Indianapolis, IN). Shc, Grb2, and pan-ERK antibodies were purchased from Transduction Laboratories (Lexington, KY). Phosphospecific MAP kinase antibody was from Promega Corp. (Madison, WI). The 3R S6 RSK substrate peptide and antibodies to the RSK-2 isoform of p90 ribosomal S6 kinase and the p85 regulatory subunit of PI 3-kinase were from Upstate Biotechnology, Inc. (Lake Placid, NY). Polyclonal phosphotyrosine and IGF-I receptor (C-terminal) antibodies were prepared as described previously (24). The glutathione S-transferase-S6 substrate peptide and antibodies to p70 ribosomal S6 kinase, IRS-1, and IRS-2 were provided by Dr. Morris White. The monoclonal IGF-I receptor antibody (␣IR-3) was provided by Dr. Steve Jacobs. Protein A-Sepharose was purchased from Pharmacia Inc. (Piscataway, NJ).
[␥-32 P]ATP was purchased from NEN Life Science Products Inc. (Boston, MA). 125 I-Protein A and 125 I-IGF-I were from ICN Biomedical (Costa Mesa, CA) and Amersham Life Science, Inc. (Arlington Heights, IL), respectively. All other reagents were purchased from Sigma.
CHO Cell Lines-Unless otherwise indicated, all studies were conducted with multiple clones of CHO cells stably transfected with the CAGϩ isoform of the human IGF-I receptor. These cells, which have been described previously (25), express approximately 6 ϫ 10 5 receptors/cell as assessed by ligand binding and Scatchard analysis. For studies on cells expressing mutated IGF-I receptors, a 961-base pair HindIII-BamHI fragment of the wild-type IGF-I receptor cDNA (base pairs 3195-4156) was subcloned into Bluescript (Stratagene, La Jolla, CA) and used as a template for site-directed mutagenesis as described by Kunkel (26). The oligonucleotide TCCGCGCCAGCTACGACGAG-AGACA (mutant nucleotide underlined) was used to generate a cDNA fragment encoding a Phe to Tyr substitution at position 1310 in the C-terminal portion of the IGF-I receptor ␤-subunit. The mutated IGF-I receptor fragment was fully sequenced to exclude the possibility of additional unwanted mutations and used to replace the analogous fragment in the previously described AP r M8 expression vector containing the full-length cDNA of the wild-type CAGϩ IGF-I receptor (25). Multiple stably transfected clones of CHO cells expressing approximately 6 ϫ 10 5 mutant receptors/cell were obtained by a co-transfection protocol using a neomycin resistance selection method as described previously (25). The isolation of CHO cells expressing similar numbers of kinase-inactive IGF-I receptors (Lys to Ala substitution at position 1003) has been described (25).
Ligand-induced Internalization-For comparative studies on wildtype and mutant IGF-I receptors, the rate of receptor internalization was determined as described previously (20). Cells were replica plated in 24-well dishes in Ham's F-12 medium supplemented with 10% fetal bovine serum. Subconfluent cell monolayers were washed twice with phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA), preincubated for 20 min at 37°C in assay buffer containing Ham's F-12 medium supplemented with 0.5% BSA and 50 mM Hepes (pH 7.4), and rinsed twice more with wash buffer. Subsequently, 40,000 cpm of 125 I-IGF-I (final IGF-I concentration 5 ϫ 10 Ϫ11 M) in assay buffer was added to each well, and individual plates were incubated at 37°C for timed periods of 2, 4, 6, 8, or 10 min. Following incubation, the medium was removed, and the plates were immersed in ice-cold PBS plus 0.1% BSA (pH 7.5) and further washed by immersing twice in ice-cold PBS plus 0.1% BSA at pH 2.75 (determination of internalized IGF-I) or pH 7.5 (determination of total cell-associated IGF-I). The washed cells were solubilized with 0.1 N NaOH in 0.1% SDS, and 125 I was quantified by ␥-counting. The internalization rate constant (K e ), representing the slope of the line relating internalized to surface-bound hormone at each time point, was determined as described previously (27). An integrated measure of surface-bound hormone was approximated by the trapezoidal rule using intervals of 2 min, and the slope of the line defining K e was determined by linear regression.
For the low temperature and inhibitor studies, the amount of IGF-I internalized was determined as described by Hsu and Olefsky (28). Cells were grown to subconfluency in 35-mm wells, serum depleted for 16 -18 h, washed twice with PBS and once with Krebs-Ringer phosphate-Hepes binding buffer (pH 7.5) at room temperature, and incubated in binding buffer containing 125 I-IGF-I (10 Ϫ8 M) for 10 min. The monolayer was then washed 3 times with ice-cold PBS, the surfacebound ligand was extracted with acidic binding buffer (pH 2.75) for 5 min at 4°C, and the cell-associated (internalized) radioactivity was determined by lysing the cells with 0.4 N NaOH. The percentage of ligand internalized was calculated by dividing the internalized counts/ min by surface-bound plus internalized counts/min.

IGF-I Stimulated Protein
Tyrosine Phosphorylation-To determine IGF-I stimulation of protein tyrosine phosphorylation in intact cells, subconfluent monolayers were incubated for 16 -18 h in serum-free Ham's F-12 medium containing 0.5% BSA and 25 mM Hepes (pH 7.4). IGF-I was subsequently added to the medium at a dose of 10 Ϫ8 M for the indicated period of time, the medium was removed, and the cells were washed twice with ice-cold buffer (137 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , and 0.1 mM Na 3 VO 4 in 20 mM Tris-HCl, pH 7.6) and lysed in the same buffer supplemented with 1% Nonidet P-40, 10% glycerol, 2 mM EDTA, 10 mM sodium pyrophosphate, 10 mM sodium fluoride, 2 mM Na 3 VO 4 , 2 mM phenylmethylsulfonyl fluoride, and 8 g/ml leupeptin). Cell lysates were centrifuged at 12,000 ϫ g for 5 min at 4°C, the supernatants were collected, and protein concentrations determined using a Bradford dye binding assay kit with BSA as standard (Bio-Rad). For studies on the effects of low temperature, cell monolayers were incubated overnight with serum-free medium, and this was replaced with fresh medium preadjusted to 15 or 37°C. After incubation for 30 min at the appropriate temperature, IGF-I was added, and cell lysates prepared as described above. For inhibitor studies, dansylcadaverine (500 M), chloroquine (200 M), or an appropriate diluent were added for 30 min prior to IGF-I stimulation.
To detect tyrosine-phosphorylated proteins, equal amounts of solubilized protein (1 mg) from cells treated under various conditions were incubated in lysis buffer with either ␣IR3 (1:200 dilution), IRS-1 (1:100 dilution), or Shc (1 g/ml) antibody at 4°C overnight. The antibody was then adsorbed to protein-A Sepharose beads for 2 h at 4°C, and the resulting immunocomplexes were washed three times by centrifugation and resuspension in buffer containing 10 mM Tris-HCl (pH 7.8), 150 mM NaCl, 1% Nonidet P-40, 1 mM Na 3 VO 4 , and 1 mM phenylmethylsulfonyl fluoride. Specific protein immunoprecipitates, plus additional cell extracts not treated with antibodies (200 g), were heated in Laemmli buffer with 100 mM DTT at 100°C for 5 min.
Proteins were separated by SDS-PAGE and electroblotted onto nitrocellulose membranes (Schleicher & Schuell, Inc., Keene, NH). The blots were blocked in 5% BSA, probed with phosphotyrosine antibody (2 g/ml), washed as described previously, and then incubated with 125 I-Protein A for 1 h at 25°C (29). Labeled protein bands were detected and quantitated using a PhosphorImaging system (Molecular Dynamics, Sunnyvale, CA). The identities of the phosphorylated bands on antiphosphotyrosine immunoblots corresponding to the ␤-subunit of the IGF-I receptor (105 kDa), IRS-1 (170 kDa), and Shc proteins (66, 52, and 46 kDa) were confirmed with antibodies specific for each of these proteins.
MAP Kinase Activation and Co-precipitation Studies-IGF-I-stimulated MAP kinase activation was measured by immunoblotting CHO cell lysates (prepared above) with a phosphospecific MAP kinase antibody (1:2000 dilution). After detection with 125 I-Protein A, these blots were stripped and re-probed with a monoclonal pan-ERK antibody (1:5000 dilution) as described by the manufacturer (Transduction Laboratories). Protein bands corresponding to ERK 1 and 2 were visualized by the enhanced chemiluminescence (ECL) method (Amersham).
For co-precipitation experiments, cell lysates were incubated with IRS-1, IRS-2 (1:100 dilution), or Shc antibody followed by SDS-PAGE and electroblotting as described above. Blots of IRS-1 and -2 immunoprecipitates were subsequently probed with either an antibody that recognizes the p85 regulatory subunit of PI-3 kinase (1:1000 dilution), or the respective immunoprecipitating antibody. To measure IGF-Istimulated Shc-Grb2 association, Shc immunoprecipitates were immunoblotted with a monoclonal Grb2 antibody (1:1000) as described by the manufacturer (Transduction Laboratories). The protein band corre-sponding to Grb2 was detected by the ECL method.
Kinase Assays-RSK-2 activity was measured by an immune complex kinase assay as described previously (30). For each treatment, 500 g of cell extract was incubated with polyclonal RSK-2 antibody for 2 h at 4°C and subsequently complexed to protein A-Sepharose beads for an additional 2 h. Immunocomplexes were washed 3 times with lysis buffer, 3 times with LiCl buffer, 3 times with RSK kinase buffer (30), and resuspended in kinase buffer containing 30 mM Tris (pH 7.4), 10 mM MgCl 2 , 0.1 mM EGTA, 1 mM DTT, 3R S6 peptide (RRLSSLRA), 40 M ATP, and 10 Ci of [␥-32 P]ATP. Reactions were performed at 30°C for 15 min with agitation, terminated, and spotted onto P81 phosphocellulose papers. The papers were washed several times with 1% phosphoric acid and measured for radioactivity by the Cerenkov method.
Statistical Analysis-Quantitative data are presented as mean Ϯ S.E. and analyzed using a statistical model based on a one-way or two-way classification analysis of variance. Tests of significance for all possible comparisons were determined by Bonferroni's t test (SYSTAT, version 5.0, Evanston, IL).

Time Course of IGF-I Stimulated IRS-1 and Shc
Phosphorylation-To determine the time course of IGF-I stimulated IRS-1 and Shc tyrosine phosphorylation, multiple clones of stably transfected CHO cells expressing approximately 6 ϫ 10 5 human wild-type IGF-I receptors/cell were stimulated with 10 Ϫ8 M IGF-I for 1, 2, 5, 10, or 20 min at 37°C. Cell lysates were prepared, resolved by SDS-PAGE, and the IGF-I receptor and IRS-1 were identified by immunoblotting with phosphotyrosine antibody. For quantitation of phosphorylation of the Shc proteins, the cell extracts were first subjected to immunoprecipitation with Shc antibody, and the resulting precipitates were analyzed by phosphotyrosine antibody immunoblotting. As shown in Fig. 1A, phosphorylation of the IGF-I receptor was rapid, reaching a maximum within 2 min. Insulin receptor tyrosine phosphorylation was unmeasurable in these cell extracts (data not shown), reflecting the specificity of IGF-I binding and the relatively low number of insulin receptors in the CHO cells. The tyrosine phosphorylation of IRS-1 also was very rapid, reaching maximal levels within 1 to 2 min after the addition of IGF-I (Fig. 1, A and B). At 10 min, there was a 20% decrease in IRS-1 phosphorylation and no further change through 20 min of incubation. In comparison to IRS-1, Shc tyrosine phosphorylation was more gradual, reaching a maximal and sustained level after 5 to 10 min of incubation with IGF-I (Fig. 1A). Quantitative data are shown for the 52-kDa Shc isoform in Fig. 1B, and a similar time course of IGF-I stimulated phosphorylation was also observed for the 46-and 66-kDa Shc isoforms.

Protein Tyrosine Phosphorylation in CHO Cells Transfected with Internalization-defective Mutant IGF-I Receptors-Previ-
ous studies in our laboratory have shown that mutation of the Phe residue to Tyr at position 1310 in the C terminus of the IGF-I receptor leads to a decreased rate of ligand-induced receptor internalization in transfected CHO cells (31). As shown in Fig. 2, the rate of ligand-induced internalization of Tyr-1310 mutant receptors is decreased by 50% in comparison with wildtype IGF-I receptors. The internalization rate of the Tyr-1310 mutant is only slightly greater than the rate of ligand-induced internalization of receptors that have absent kinase activity as a consequence of a Lys to Ala mutation at position 1003 in the ATP-binding site.
Although internalization of the Tyr-1310 mutant receptor was markedly inhibited, IGF-I-induced autophosphorylation of this receptor construct was preserved and, in fact, modestly increased in comparison with wild-type receptors (Fig. 3A, top  and bottom panels). Thus, the mutant receptor has an effective IGF-I activated tyrosine kinase. The increased receptor autophosphorylation cannot be explained by a difference in receptor content, which was similar to that in cells transfected with the wild-type receptor as assessed by immunoblotting (middle panel in Fig. 3A) or by Scatchard analysis of IGF-I binding data (not shown). The higher level of phosphorylation in the Tyr-1310 mutant may reflect, at least in part, phosphorylation of the additional C-terminal tyrosine residue. IRS-1 tyrosine phosphorylation was markedly stimulated by IGF-I in cells transfected with wild-type or Tyr-1310 receptors. There was no difference in IRS-1 content or in the level of IRS-1 phosphorylation catalyzed by these two receptor constructs (Fig. 3B). Thus, the marked decrease in ligand-induced internalization of the Tyr-1310 receptor does not interfere with its capacity to interact with IRS-1 in intact cells. This contrasts with the effects of IGF-I on Shc phosphorylation (Fig. 3C). Phosphorylation of all three Shc isoforms was stimulated by IGF-I in cells transfected with either receptor construct, but Shc phosphorylation was markedly decreased in cells expressing the internalization-defective Tyr-1310 receptor mutant. Quantitation of the extent of tyrosine phosphorylation of the dominant 52-kDa Shc isoform from multiple experiments (Fig. 3C, bottom panel) demonstrated a 65% decrease in the level of Shc phosphorylation 10 min after IGF-I stimulation of cells expressing the Tyr-1310 receptor as compared with the wild-type receptor. A similar or greater decrease in Shc phosphorylation was evident at multiple time points extending to 60 min after IGF-I stimulation (data not shown).
The protein level of the 52-kDa Shc isoform, as determined by direct immunoblotting, was also decreased in the Tyr-1310 cells (35% lower than in cells expressing wild-type receptors, p Ͻ 0.05), but the decrease in Shc protein content was less than the decrease in Shc phosphorylation in these cells. IGF-I stimulation of 52-kDa Shc phosphorylation was significantly decreased in the Tyr-1310 cells after correction for the change in Shc protein content (100 Ϯ 8.6% for wild-type versus 70 Ϯ 4.3% for Tyr-1310, p Ͻ 0.05). This observation was consistent for two different clones of wild-type Tyr-1310 cells that were studied. Therefore, the IGF-I-stimulated phosphorylation of Shc, but not IRS-1, was decreased in cells transfected with the internalization-defective mutant IGF-I receptor.

Effects of Low Temperature on Internalization and Protein
Tyrosine Phosphorylation-To further investigate the relationship between receptor-mediated endocytosis and protein tyrosine phosphorylation, IGF-I receptor internalization in CHO cells expressing the wild-type receptor was inhibited by reducing the cell incubation temperature from 37 to 15°C. Ligandinduced IGF-I receptor internalization after 10 min was decreased by 64% at 15°C compared with cells at 37°C (Fig. 4A). Low temperature also decreased IGF-I stimulated receptor autophosphorylation at early time points (0 -20 min) (Fig. 4B). This is consistent with delayed ligand binding kinetics at 15°C, which have previously been described (32). The time course of IGF-I stimulated IRS-1 tyrosine phosphorylation was delayed, and the level of IRS-1 phosphorylation observed at all time points between 0 and 20 min was reduced at 15°C versus 37°C (Fig. 5A). Since diminished IRS-1 phosphorylation might be explained by delayed IGF-I binding and receptor activation, the level of IRS-1 phosphorylation was expressed per unit of receptor autophosphorylation. When data from cells stimulated with IGF-I for periods ranging from 2 to 20 min were analyzed, a significant increase in the amount of phosphorylated IRS-1 per phosphorylated receptor was evident at 15°C compared with 37°C (p Ͻ 0.05 by ANOVA). As shown for the 20-min time point in Fig. 5B, the amount of phosphorylated IRS-1 per phosphorylated receptor was increased by more than 2-fold at 15°C. Thus, the inhibition of ligand-induced receptor internalization at 15°C is associated with an augmentation of IRS-1 phosphorylation relative to the number of autophosphorylated (active) IGF-I receptors.
In contrast with IRS-1, IGF-I stimulation of Shc phosphorylation was completely inhibited at 15°C (Fig. 6A). This inhibition of Shc phosphorylation at 15°C was evident at time points extending at least 60 min after the addition of IGF-I (data not shown). When Shc phosphorylation was expressed per unit of autophosphorylated IGF-I receptor, a marked decrease was apparent at 15°C (Fig. 6B). Thus, the inhibition of ligandinduced IGF-I receptor internalization by reduced temperature is associated with increased IRS-1 phosphorylation per autophosphorylated receptor and inhibited Shc phosphorylation.
Effects of Dansylcadaverine and Chloroquine on IGF-I Receptor Internalization and Protein Tyrosine Phosphoryla- tion-As an additional method of investigating the relationship between receptor endocytosis and signaling, IGF-I-stimulated protein tyrosine phosphorylation in CHO cells was determined after treatment with two chemical inhibitors of receptor trafficking, dansylcadaverine and chloroquine. Although the mechanism of action of dansylcadaverine has not been precisely defined, extensive evidence indicates that it inhibits receptor trafficking at a step proximal to the formation of endocytotic vesicles (33,34). Chloroquine interferes with receptor trafficking by blocking a later step in the acidification of endosomes, thereby preventing recycling of receptors to the cell surface (35). Consistent with these early and late sites of action in the receptor internalization/trafficking pathway, pretreatment of CHO cells expressing the wild-type IGF-I receptor with dansylcadaverine resulted in a 65% decrease in ligand-induced receptor internalization, whereas chloroquine did not significantly alter receptor internalization (Fig. 7).
Although not as effectively visualized in Fig. 8C, inhibition of phosphorylation of the 66-and 46-kDa isoforms of Shc by dansylcadaverine was also evident on blots exposed for a longer period of time. Consistent with its lack of effect on receptor internalization, chloroquine did not alter receptor, IRS-1, or Shc phosphorylation (data not shown).
Effects of Dansylcadaverine on the IRS-1 and Shc/MAP Kinase Signaling Pathways-To determine whether ligand-induced receptor internalization influences signaling events that are downstream from IRS-1, CHO cells expressing IGF-I receptors were incubated with the internalization inhibitor, dansylcadaverine. Following stimulation with IGF-I for 5 min, cell lysates were immunoprecipitated with IRS-1 antibody and immunoblotted with an antibody specific for the p85 regulatory subunit of PI3-kinase to measure p85 association with IRS-1. Fig. 9A demonstrates that dansylcadaverine has no effect on IGF-I-stimulated p85 association with IRS-1, consistent with the previous observation of unaltered IRS-1 tyrosine phosphorylation in dansylcadaverine-treated cells. Under these same conditions, IRS-2 tyrosine phosphorylation and p85 association with IRS-2 was not affected by dansylcadaverine (data not shown) indicating that the activation of this alternate IRS pathway also is not affected by changes in receptor internalization. As a signaling response downstream from IRS activation of PI 3-kinase, p70 S6 kinase was shown to be markedly stimulated by IGF-I, and this activation was not altered by pretreatment with dansylcadaverine (Fig. 9B).
In contrast to the lack of effects on IRS pathway signaling, inhibition of receptor internalization by dansylcadaverine decreased IGF-I activation of multiple signaling responses downstream from Shc. As shown in Fig. 10A, the IGF-I-stimulated association of Grb2 with Shc was decreased by approximately 35% in dansylcadaverine-treated cells (p Ͻ 0.05), which is similar in magnitude to the inhibition of IGF-I-stimulated Shc tyrosine phosphorylation by dansylcadaverine (Fig. 8C). This was associated with a decrease in IGF-I-stimulated phosphorylation of the 44-kDa ERK 1 isoform of MAP kinase (Fig. 10B), a less marked decrease in the 42-kDa ERK 2 isoform, and a significant decrease in IGF-I-stimulated p90 S6 kinase activity (Fig. 10C). DISCUSSION The comparison of the time courses of IGF-I-stimulated tyrosine phosphorylation of IRS-1 and Shc proteins in CHO fibroblasts expressing human IGF-I receptors confirms, in this cell system, the previously reported observation that Shc phosphorylation is delayed in comparison with IRS-1 phosphorylation. A similar slower time course of Shc compared with IRS-1 phosphorylation has also been observed in Rat1 fibroblasts stimulated with IGF-I (22) or insulin (9). The temporal pattern of Shc phosphorylation corresponds to the time course of IGF-I receptor internalization previously described in CHO cells (23), suggesting that receptor endocytosis may play a role in cell signaling via Shc phosphorylation. Therefore, we investigated the relationships between IRS-1 and Shc-dependent signaling and ligand-induced IGF-I receptor internalization.
IGF-I-induced receptor internalization was inhibited by mutation of the C-terminal region of the IGF-I receptor (Phe to Tyr substitution at amino acid position 1310), incubation of cells at low temperature (15°C), or treatment with the endocytosis . Cells were subsequently incubated in the presence of 125 I-IGF-I for 10 min. Surfacebound ligand was extracted with acidic binding buffer and cell-associated (internalized) radioactivity was determined by alkaline lysis as described under "Experimental Procedures." Data are expressed as the % of total cell associated radioactivity internalized after 10 min and represent mean Ϯ S.E. from four independent experiments. *, p Ͻ 0.05. inhibitor dansylcadaverine. With each of these different experimental approaches, there was a significant decrease in ligandinduced receptor internalization and a corresponding decrease in IGF-I stimulation of Shc tyrosine phosphorylation. Cells expressing the Tyr-1310 mutant receptor exhibited maximum phosphorylation of Shc 10 min after IGF-I stimulation and decreased Shc phosphorylation at multiple time points extending for at least 60 min after IGF-I treatment. Thus, decreased Shc tyrosine phosphorylation in cells expressing the Tyr-1310 receptor cannot be explained by delayed kinetics of Shc phosphorylation. Two independent Tyr-1310 cell clones had a similar 35% decrease in Shc protein content in comparison with two clones expressing the wild-type receptor. This suggests a possible relationship between the activation state and expres-sion of the Shc proteins, which we have not yet investigated. The decrease in Shc phosphorylation was greater than the decrease in Shc content, however, and IGF-I-stimulated Shc phosphorylation was significantly diminished in the Tyr-1310 cells even after correction for the lower level of Shc protein. The cellular content of the Shc proteins did not change in cells incubated at low temperature or with the inhibitor dansylcadaverine, further confirming a consistent relationship between decreased IGF-I receptor internalization and decreased IGF-I stimulation of Shc phosphorylation.
The step in the endocytotic pathway inhibited by the Tyr-1310 mutation is not known. However, there is evidence that low temperature and dansylcadaverine delay receptor internalization at steps proximal to the formation of early endosomes. It has been shown that incubating cells at a 16°C results in a marked decrease in the fraction of bound insulin that cannot dissociate from intact cells by low pH washing, indicative of a delay in early endosome formation via clathrin-coated pits (36). In the CHO cells investigated in this study, we have demonstrated a similar marked inhibition of ligand-induced IGF-I receptor internalization at 15°C. Previous studies have shown that dansylcadaverine appears to inhibit a step in endocytotic vesicle formation that involves the pinching off of vesicles from the plasma membrane (33,37). Based on these sites of inhibition, our findings are consistent with a model in which the tyrosine phosphorylation of Shc occurs predominantly by activated IGF-I receptors in endosomes, or at least by receptors in vesicles that have moved past the point in the endocytotic pathway inhibited by dansylcadaverine. Our findings further demonstrate that changes in Shc phosphorylation as a consequence of altered receptor internalization correlate with the intensity of IGF-I induced signaling via the MAP kinase pathway.
The endosome has previously been implicated as a major site of Shc tyrosine phosphorylation in response to epidermal growth factor stimulation in liver parenchyma (38), but the relative activity of receptors at different points in the ligandinduced internalization pathway in phosphorylating Shc or other substrates was not investigated in this earlier study. In a report published subsequent to the completion of our study, the inhibition of ligand-induced insulin receptor internalization in receptor overexpressing CHO cells by mutation or incubation at 4°C was shown to be associated with decreased insulinstimulated tyrosine phosphorylation of annexin II (39). It was hypothesized that annexin II is selectively phosphorylated by receptors that have moved from the microvillous to nonvillous regions of the plasma membrane (40), but not yet entered coated pits and endosomes. In accordance with our findings, it was briefly noted in this study that insulin-stimulated phosphorylation of the 64-kDa isoform of Shc was also inhibited at 4°C (39). Thus, the internalization of insulin receptors as well as IGF-I receptors appears to be required for phosphorylation and activation of the Shc proteins.
In contrast to Shc phosphorylation, we did not observe a decrease in IGF-I-stimulated IRS-1 tyrosine phosphorylation when receptor internalization was inhibited by mutation, low temperature, or dansylcadaverine, indicating that IRS-1 phosphorylation is not dependent on IGF-I receptor internalization. Furthermore, data from dansylcadaverine-treated cells demonstrate that IGF-I receptor internalization does not regulate the intensity of signaling via the IRS pathway, as measured by association of the p85 regulatory subunit of PI 3-kinase with IRS proteins or p70 S6 kinase activation. The observation that IGF-I receptor internalization is not required for IRS-1 phosphorylation is consistent with a previous study on insulin receptors demonstrating that insulin-stimulated IRS-1 tyrosine phosphorylation can be observed at 4°C in CHO fibroblasts and 3T3-L1 adipocytes (41). Thus, it appears that IRS-1 phosphorylation can be catalyzed by activated IGF-I or insulin receptors at the cell surface and/or in coated pits closely associated with the plasma membrane. When receptor internalization is inhibited by low temperature, the amount of tyrosinephosphorylated IRS-1 per activated (phosphorylated) IGF-I receptor in these structures increases, whereas Shc phosphorylation is completely inhibited. It is important to note that these findings of effective IRS-1 tyrosine phosphorylation despite decreased receptor internalization do not exclude an important role for activated endosomal or intracellular membrane-associated receptors in the phosphorylation of IRS-1. In vivo administration of colchicine and a peroxovanadium compound has been shown to exclusively activate endosomal insulin receptors in the liver, and this leads to tyrosine phospho-  ). B, cell lysates were directly resolved by SDS-PAGE and immunoblotted with MAP kinase antibodies. The top panel shows a representative immunoblot with phosphospecific MAP kinase antibody, and the middle panel is the same blot stripped and re-probed with antibody that recognizes MAP kinase independent of its phosphorylation state. The bar graph at the bottom was derived from the quantitation of the 44-kDa bands in six independent experiments with the phosphospecific antibody. C, p90 S6 kinase was immunoprecipitated from cell lysates, and kinase activity was measured based on phosphorylation of S6 peptide as described under "Experimental Procedures." The data were derived from six independent experiments. The results in all bar graphs are expressed as % of control (IGF-I stimulated without dansylcadaverine). *, p Ͻ 0.05 versus basal; **, p Ͻ 0.05 versus basal or stimulated control cells. rylation of IRS-1, demonstrating the signaling capacity of endosomal insulin receptors in signaling (42). Another study has provided evidence that IRS-1 can be phosphorylated by activated insulin receptors in intracellular membrane compartments in rat adipocytes after insulin stimulation (43).
In a recently published report (44), the Shc proteins were shown to be localized on rough endoplasmic reticulum membranes and redistributed to several regions in the cell, including clathrin-coated pits and endosomes, after tyrosine kinase receptor activation. Furthermore, the presence of distinct subcellular compartmentalized pools of Shc was demonstrated in PC-12 cells (45). Based on these data, plus the findings in the current study, it will be important to explore in future investigations the possibility that activation of receptors and their subsequent endocytosis may lead to the recruitment and tyrosine phosphorylation of the Shc proteins. Rapidly emerging evidence supports a role for compartmentation in targeting the actions of multiple cellular signaling intermediates, including PI 3-kinase (46), protein phosphatases (47), protein kinase C (48), cAMP-dependent protein kinase (49), and the MAP kinase cascade (47). The observation that different substrates or combinations of substrates are phosphorylated by IGF-I receptors at distinct cellular sites defined by the internalization pathway suggests an important role for the ligand-induced internalization mechanism in establishing the subcellular compartmentation of receptors. Factors that modify receptor internalization, either by altering the membrane microenvironment or directly interacting with receptors, could change the distribution of receptors in different cellular compartments and, thus, change the specific pattern of post-receptor responses initiated by receptor activation.
In conclusion, the present study demonstrates that IGF-I receptor internalization is required for cell signaling via the Shc/MAP kinase pathway, but not the IRS-1 pathway. Using three different approaches to modify receptor internalization in intact CHO fibroblasts, our data provide strong evidence for the role of IGF-I receptor endocytosis in modulating post-receptor signaling events. Cellular compartmentation established by ligand-induced receptor internalization may provide a mechanism for independently modulating the stimulation of specific signaling molecules and, thus, the activation of divergent pathways that lead to distinct biological responses.