Insulin Receptor-mediated p62dok Tyrosine Phosphorylation at Residues 362 and 398 Plays Distinct Roles for Binding GTPase-activating Protein and Nck and Is Essential for Inhibiting Insulin-stimulated Activation of Ras and Akt*

A GTPase-activating protein (GAP)-associated 60-kDa protein has been found to undergo rapid tyrosine phosphorylation in response to insulin stimulation. However, whether this protein is a direct in vivo substrate for the insulin receptor (IR) tyrosine kinase and whether the tyrosine phosphorylation plays a role in insulin signaling remain to be established. Here we show that the insulin-stimulated tyrosine phosphorylation of the GAP-associated protein, now identified as p62dok, is inhibited by Grb10, an adaptor protein that binds directly to the kinase domain of the IR, both in vitro and in cells. Replacing Tyr362 and Tyr398 with phenylalanine greatly decreased the IR-catalyzed p62dok tyrosine phosphorylation in vitro, suggesting that these two residues are the major IR-mediated phosphorylation sites. However, mutations at Tyr362 and Tyr398 only partially blocked insulin-stimulated p62dok tyrosine phosphorylation in cells, indicating that p62dok is also a target for other cellular tyrosine kinase(s) in addition to the IR. Replacing Tyr362 with phenylalanine abolished the interaction between p62dok and Nck. Mutations at Tyr362/398 of p62dok disrupted the interaction between p62dokand GAP and decreased the inhibitory effect of p62dok on the insulin-stimulated activation of Ras and Akt, but not mitogen-activated protein kinase. Furthermore, the inhibitory effect of p62dok on Akt phosphorylation could be blocked by coexpression of a constitutively active Ras. Taken together, our findings indicate that p62dok is a direct substrate for the IR tyrosine kinase and that phosphorylation at Tyr362 and Tyr398 plays an essential role for p62dok to interact with its effectors and negatively regulate the insulin signaling pathway.

A GTPase-activating protein (GAP)-associated 60-kDa protein has been found to undergo rapid tyrosine phosphorylation in response to insulin stimulation. However, whether this protein is a direct in vivo substrate for the insulin receptor (IR) tyrosine kinase and whether the tyrosine phosphorylation plays a role in insulin signaling remain to be established. Here we show that the insulin-stimulated tyrosine phosphorylation of the GAP-associated protein, now identified as p62 dok , is inhibited by Grb10, an adaptor protein that binds directly to the kinase domain of the IR, both in vitro and in cells. Replacing Tyr 362 and Tyr 398 with phenylalanine greatly decreased the IR-catalyzed p62 dok tyrosine phosphorylation in vitro, suggesting that these two residues are the major IR-mediated phosphorylation sites. However, mutations at Tyr 362 and Tyr 398 only partially blocked insulin-stimulated p62 dok tyrosine phosphorylation in cells, indicating that p62 dok is also a target for other cellular tyrosine kinase(s) in addition to the IR. Replacing Tyr 362 with phenylalanine abolished the interaction between p62 dok and Nck. Mutations at Tyr 362/398 of p62 dok disrupted the interaction between p62 dok and GAP and decreased the inhibitory effect of p62 dok on the insulin-stimulated activation of Ras and Akt, but not mitogen-activated protein kinase. Furthermore, the inhibitory effect of p62 dok on Akt phosphorylation could be blocked by coexpression of a constitutively active Ras. Taken together, our findings indicate that p62 dok is a direct substrate for the IR tyrosine kinase and that phosphorylation at Tyr 362 and Tyr 398 plays an essential role for p62 dok to interact with its effectors and negatively regulate the insulin signaling pathway.
Insulin regulates a variety of biological activities by binding to its receptor on the cell membrane. The binding of insulin to its receptor leads to receptor tyrosine kinase activation and tyrosine phosphorylation of its cellular substrates. Several insulin receptor (IR) 1 substrates have been identified, including the IR substrate (IRS) proteins (1,2), Shc (3)(4)(5), and a 60-kDa protein that has been shown to interact with the p21 ras GTPase-activating protein (GAP) upon growth factor stimulation (6 -8).
The cDNA encoding the 60-kDa GAP-associated protein has been cloned recently, and the protein has been named p62 dok (62-kDa protein downstream of tyrosine kinase) or Dok-1 (9,10). Two p62 dok -like proteins have also been identified recently which include p56 dok or Dok-2 (also known as Dok-R or FRIP) (11)(12)(13) and Dok-L or Dok-3 (14). p62 dok contains an aminoterminal pleckstrin homology domain potentially involved in phospholipid interaction and membrane targeting, a central putative phosphotyrosine binding domain for interacting with tyrosine-phosphorylated proteins, and several growth factorstimulated tyrosine phosphorylation sites at its carboxyl terminus (9). Mutation of five tyrosine residues (tyrosine residues 296, 315, 362, 398, and 409) in p62 dok abolished its ability to bind to GAP in vitro (15), suggesting that tyrosine phosphorylation is necessary for the p62 dok /GAP interaction. Overexpression of p62 dok has been shown to inhibit Ras activity in human embryonic kidney 293 cells and B cell antigen receptor-mediated c-fos promoter activation in an immature B cell line (16), suggesting that p62 dok may play a negative role in Ras signaling. It has been shown that insulin stimulates tyrosine phosphorylation of p62 dok (1); however, the sites of phosphorylation and the physiological roles of p62 dok in insulin signal transduction remain largely unknown. It is of interest to note that the overall structure of p62 dok is similar to that of the IRS proteins (1). In addition, overexpression of p62 dok in Chinese hamster ovary cells expressing the IR (CHO/IR cells) has been shown to stimulate Ras membrane localization (17). However, although these findings suggest that p62 dok may play a role in insulin signaling, overexpression of p62 dok has been shown to have little effect, if any, on insulin-stimulated mitogen-activated protein kinase activity (17).
In addition to tyrosine phosphorylation of the IR cellular substrates, binding of insulin to its receptor also leads to the autophosphorylation of the IR on at least six tyrosine residues, including Tyr 1158 , Tyr 1162 , and Tyr 1163 in the activation domain. This phosphorylation of the receptor leads to the generation of docking sites for binding of adaptor proteins such as Shc and Grb2, which allows constitutive association with the guanine nucleotide exchange factor SOS and subsequent activation of Ras. Through an interaction with Raf-1, Ras activates the MAP kinase kinases MEK1 and MEK2, and in turn the extracellular signal-regulated kinases ERK1 and ERK2 (2). In recent years, it has become clear that Ras also controls the activity of other downstream effectors in addition to Raf-1. Ras has been shown to interact with and stimulate the activity of the catalytic p110 subunit of different isoforms of phosphatidylinositol 3-kinase (PI 3-kinase) (18). In addition, overexpression of the dominant negative N17 Ras mutant inhibits insulin, platelet-derived growth factor, and epidermal growth factorstimulated Akt activation (19). These findings suggest that Ras is an upstream activator of the PI 3-kinase/Akt pathway.
Grb10 is a pleckstrin homology and Src homology 2 (SH2) domain-containing protein that binds to the autophosphorylated IR (20). We and others have shown that Grb10 binds to autophosphorylated tyrosine residues in the activation loop of the IR (21,22). In addition, a human Grb10 isoform with a deletion in the pleckstrin homology domain inhibits insulinstimulated tyrosine phosphorylation of the 60-kDa GAP-associated protein when overexpressed in CHO/IR cells (23). However, the identity of this GAP-associated protein and the mechanism by which Grb10 inhibited its tyrosine phosphorylation remained unknown. Very recently, it has been shown that Grb10 inhibits the catalytic activity of the IR in vitro (24). These findings suggest that the binding of Grb10 to the activation domain of the IR blocks the ability of the receptor to phosphorylate its exogenous substrates.
In the present study, we investigated the insulin-stimulated tyrosine phosphorylation of the 60-kDa GAP-associated protein. We provide evidence that this 60-kDa GAP-associated protein is in fact p62 dok . In addition, we show that p62 dok is a direct substrate of the IR with phosphorylation occurring mainly at tyrosine residues 362 and 398. Replacing tyrosine residues of p62 dok at 362 and 398 with phenylalanine inhibited the interaction between p62 dok and GAP. We also found that overexpression of wild-type p62 dok in CHO/IR cells inhibited insulin-stimulated activation of Ras, MAP kinase, and Akt. Furthermore, inhibition of Ras and Akt could be rescued by mutating 362 and 398 of p62 dok . Taken together, our findings suggest that the IR-mediated tyrosine phosphorylation of p62 dok plays important and specific roles in regulating the insulin-mediated MAP kinase and PI 3-kinase pathways downstream of Ras.
Reagents-The cDNA encoding human p62 dok was a generous gift from Dr. Bruce Stillman (Cold Spring Harbor, NJ) and was described previously (10). The cDNAs encoding hemagglutinin (HA)-tagged wildtype and constitutively active p21 ras , Myc-tagged MAP kinase, and the GST-Raf-RBD were generous gifts from Dr. Kun-Liang Guan (University of Michigan) and Dr. Jun-Lin Guan (Cornell University), respectively. cDNA encoding Myc and FLAG-tagged Akt-1 was described previously (25). Monoclonal anti-GAP and anti-Myc antibodies were from Santa Cruz Biotechnology, Inc. Monoclonal anti-HA antibody was from BABCO. The anti-phosphotyrosine (RC-20) and monoclonal anti-Nck antibodies were from Transduction Laboratories. Polyclonal anti-Nck was from PharMigen. Phospho-specific polyclonal antibodies to Akt-Thr 308 , Akt-Ser 473 , and MAP kinase were from New England Biolabs. Monoclonal antibody to the FLAG tag (M2) was from Sigma.
Secondary antibodies conjugated to alkaline phosphatase and horseradish peroxidase were from Promega.
Expression and Purification of Proteins in Bacterial Cells-BL21(DE3) or DH5␣ cells containing plasmids encoding His-tagged full-length or various truncated mutants of p62 dok , His-Akt, GST-Grb10-C, or GST-Raf-RBD were grown in LB medium containing ampicillin. Expression of proteins was induced by the addition of 1 mM isopropyl-␤-D-thiogalactoside for 3.5 h at 30°C. Cells were harvested by centrifugation at 5,000 ϫ g for 10 min and lysed in Buffer A for 30 min at 4°C. The solution was sonicated and clarified by centrifugation at 12,000 ϫ g for 15 min. The proteins were purified by affinity chromatography using Ni 2ϩ -NTA-agarose (Qiagen) beads or glutathione-Sepharose beads (Sigma).
Cell Culture, Transfection, Immunoprecipitation, and Western Blot-CHO/IR and human embryonic kidney 293 cells were used in all experiments. Stable CHO cell lines overexpressing kinase-defective (IR KD ) and the tyrosine phosphorylation site mutant (IR Y1162F/Y1163F ) of the IR were generous gifts of Dr. Richard A. Roth (28). Transfections were performed using LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's protocol. Cells were lysed in Buffer B. Cell lysates were centrifuged at 12,000 ϫ g for 10 min at 4°C, and the clarified supernatant was used for either immunoprecipitation or Western blot analysis. Proteins in cell lysates were immunoprecipitated by incubation with the primary antibody conjugated to either protein G-or protein A-Sepharose beads for 6 -18 h at 4°C. The immunoprecipitates were washed three times with ice-cold Buffer C. For immunoblot analysis, cell lysates or immunoprecipitates were separated by SDS-PAGE using 10 -15% polyacrylamide gels. After electrophoresis, proteins were transferred onto nitrocellulose membranes, and bound proteins were detected by blotting with primary antibody followed by horseradish peroxidase-or alkaline phosphatase-conjugated secondary antibodies.
Purification and Activation of IR Tyrosine Kinase in Vitro-Wildtype or mutants of the IR were purified by incubation of cell lysates with wheat germ agglutinin-agarose (E. . Laboratories) at 4°C overnight. After extensive washing with Buffer C, the IR attached to the beads was activated in vitro by incubation at room temperature for 30 min with kinase buffer (Buffer D). Activated IR was eluted using 0.3 M N-acetylglucosamine (Sigma) and used immediately in kinase assays.
In Vitro Phosphorylation of p62 dok by the IR-His-tagged p62 dok bound to Ni 2ϩ -NTA-agarose beads was washed twice with ice-cold Buffer C and once with kinase reaction buffer (Buffer E). Kinase assays were carried out by incubation of p62 dok with Buffer E plus 2 Ci of [␥-32 P]ATP for 30 min at 25°C. The reaction was stopped by washing twice with ice-cold Buffer C followed by the addition of SDS-sample buffer and heating at 95°C for 3 min. The proteins were separated by SDS-PAGE using 10% (w/v) polyacrylamide gels, and tyrosine phosphorylation of p62 dok was detected by autoradiography. The protein level of p62 dok was determined by Coomassie staining.
Phosphopeptide Mapping of p62 dok -Wild-type or mutants of p62 dok were phosphorylated by purified and activated IR in the presence of [␥-32 P]ATP as described above. The proteins were separated by SDS-PAGE and transferred onto a nitrocellulose membrane. The 32 P-labeled bands corresponding to p62 dok were excised. The protein samples were trypsin treated, desalted, and separated by two-dimensional thin layer chromatography as described previously (29).
GST-Raf-RBD Pull-down Studies-HA-tagged Ras was coexpressed in CHO/IR or human embryonic kidney 293 cells together with wildtype or mutant p62 dok . Cells were left untreated or treated with insulin and lysed. Cell lysates were incubated at 4°C for 4 h with bacterially expressed GST-Raf-RBD fusion proteins bound to glutathione-Sepharose beads. After extensive washing with ice-cold Buffer C, bound protein was eluted by heating in SDS-sample buffer for 3 min at 95°C. The activated Ras pulled down by the GST-Raf-RBD protein was separated by SDS-PAGE using a 15% polyacrylamide gel, transferred onto nitrocellulose membrane, and detected by Western blot using antibody to the HA tag.
Akt Activity Assay-FLAG-tagged Akt was overexpressed alone or with wild-type or mutant p62 dok in CHO/IR cells. Cells were serum starved, treated with insulin or not treated, lysed, and Akt immunoprecipitated using antibody to the tag. Bound Akt was washed twice with ice-cold Buffer C and twice with Buffer F. Akt was then incubated in Buffer F containing 5 M cold ATP, 30 M peptide substrate (GRPRTSSFAEG), and 5 Ci of [␥-32 P]ATP/sample for 30 min at 30°C. Akt activity was determined by scintillation counting.

RESULTS
To study the insulin-induced tyrosine phosphorylation of p62 dok , we transiently transfected CHO/IR cells with plasmid encoding Myc-tagged p62 dok and stimulated the cells with insulin. Insulin-stimulated tyrosine phosphorylation of p62 dok could be detected as early as 15 s after treatment (Fig. 1A, upper panel). Similar results were also observed for endogenous p62 dok in CHO/IR cells (data not shown). The rapid tyrosine phosphorylation of p62 dok in response to insulin stimulation suggested that the protein might be a direct substrate for the IR kinase in cells. To test this idea further, we examined the in vitro phosphorylation of p62 dok in the presence of wild-type and two mutants of the IR (IR KD and IR Y1162F/Y1163F ) (28). We found that p62 dok was readily phosphorylated in vitro by the wild-type IR, but not by the kinase-defective IR KD or tyrosine phosphorylation site mutant IR Y1162F/Y1163F (Fig. 1B, upper panel, lanes 1-3). Under the same conditions, no tyrosine phosphorylation of the control protein Akt (Fig. 1B, upper panel, lanes 4 -6) or Grb10 (30) was observed by activated IR in vitro.
To provide additional evidence that p62 dok is a direct substrate for the IR, we examined the effect of Grb10 on the tyrosine phosphorylation of p62 dok by the IR in vitro. Because Grb10 binds to the kinase domain of the IR (21,22) and inhibits the tyrosine phosphorylation of endogenous p62 dok (23), we postulated that Grb10 might prevent p62 dok tyrosine phosphorylation by blocking the kinase activity of the IR. To test this idea, we expressed the carboxyl terminus of Grb10 (residues 369 -548 of hGrb10␣ (23)) in bacterial cells as a GST fusion protein and examined the effect of the recombinant protein on the tyrosine phosphorylation of p62 dok by the IR tyrosine kinase in vitro. The carboxyl terminus of Grb10 (Grb10-C) has previously been shown to mediate the binding of the protein to the tyrosine-phosphorylated IR (21, 23, 31, 32). As shown in Fig. 2A, incubation of p62 dok with activated IR in vitro resulted in substantial tyrosine phosphorylation of p62 dok (upper panel, A, time course of insulin-stimulated p62 dok tyrosine phosphorylation in CHO/IR cells. CHO/IR cells transiently expressing Myc-tagged p62 dok were serum starved for 5 h at 37°C and then treated with 10 nM insulin for the indicated times. Cells were lysed, and the Myc-tagged p62 dok was immunoprecipitated with antibody to the tag. Bound proteins were separated by SDS-PAGE, blotted to a nitrocellulose membrane, and the tyrosine phosphorylation of p62 dok was determined by Western blot (WB) with antibody to phosphotyrosine (upper panel). The same membrane was stripped and reblotted with antibody to the Myc tag (lower panel). Similar results were obtained in three independent experiments. B, in vitro phosphorylation of p62 dok by purified IR. His-tagged p62 dok (lanes 1-3) and Akt (lanes 4 -6) were expressed in bacterial cells, purified by Ni 2ϩ -NTA-agarose beads, and phosphorylated using purified wild-type (WT, lanes 3 and 6), kinase-defective (KD, lanes 1 and 4), or a tyrosine phosphorylation mutant (Y/F, lanes 2 and 5) IR. Tyrosine phosphorylation of p62 dok was determined by autoradiography (upper panels), and the protein levels of p62 dok and Akt were determined by Coomassie Blue staining (lower panels). lane 3). The tyrosine phosphorylation of p62 dok was inhibited competitively with increasing concentrations of the carboxyl terminus of Grb10 ( Fig. 2A, upper panel, lanes 4 -7). This inhibition was specific, and GST protein alone had no effect on p62 dok tyrosine phosphorylation with similar assay conditions ( Fig. 2A, upper panel, lane 1).
To verify that p62 dok is a direct substrate for the IR in vivo, we tested p62 dok tyrosine phosphorylation in CHO/IR cells in the presence of increasing amounts of Grb10. Treatment of cells with insulin led to substantial p62 dok tyrosine phosphorylation (Fig. 2B, top panel, lane 1). The insulin-stimulated tyrosine phosphorylation was inhibited competitively by increasing the expression of Grb10 (Fig. 2B, top panel, lanes 2-5). This inhibition was most likely the result of a direct binding of Grb10 to the IR kinase domain because little inhibition was observed in cells overexpressing the SH2 domain mutant of Grb10 (Fig. 2B, top panel, lane 5 versus lane 6), which has previously been shown to be unable to bind to the tyrosinephosphorylated IR in CHO/IR cells (30).
Having found p62 dok as a direct substrate for the IR, we next attempted to identify the IR-catalyzed tyrosine phosphorylation site(s) of p62 dok . To do this, we first tried to identify the regions on p62 dok which contained potential IR-mediated tyrosine phosphorylation sites. Full-length and several truncation mutants of p62 dok were expressed in Escherichia coli as Histagged proteins (Fig. 3A). The bacterially expressed p62 dok proteins were purified using Ni 2ϩ -NTA-agarose beads, and their phosphorylation was determined in the presence of purified, activated IR and [␥-32 P]ATP. In agreement with our earlier findings (Figs. 1B and 2A), full-length p62 dok was phosphorylated readily by the IR tyrosine kinase (Fig. 3B, upper panel,  lane 1). Deletion of the carboxyl terminus of p62 dok (residues 372-481, p62 dok ⌬C1) led to a decrease in the IR-catalyzed tyrosine phosphorylation of the protein (Fig. 3B, upper panel,  lane 2 versus lane 1). Further deletions greatly decreased or completely abolished p62 dok tyrosine phosphorylation (Fig. 3B, upper panel, lanes 3-5 (longer exposure of the film revealed a small but notable phosphorylation of p62 dok ⌬C2 (Fig. 3C, ⌬C2 and data not shown)). These findings suggested that the major IR-mediated tyrosine phosphorylation site(s) were located between residues 359 and 475 of p62 dok . To test this further, the in vitro phosphorylated wild-type and two truncation fragments (⌬C1 and ⌬C2) of p62 dok (Fig. 3A) were analyzed by phosphopeptide mapping. Two major phosphopeptides, a and b, were observed for wild-type p62 dok (Fig. 3C, WT). Deletion of residues 372-475 of p62 dok led to the loss of phosphopeptide b (Fig. 3C, ⌬C1), suggesting that one of the major phosphorylation sites lay between amino acid residues 372 and 475 of p62 dok . Phosphopeptide mapping of p62 dok ⌬C1 also revealed the presence of a new phosphopeptide (Fig. 3C, ⌬C1, phosphopeptide c). This phosphopeptide had a lower phosphorylation stoichiometry and was also observed with the other truncation and a point mutant of p62 dok (Fig. 3C, ⌬C2 and Y362F). The identity of this phosphopeptide is unknown but may have resulted from compensatory phosphorylation caused by a conformational change induced by truncating or mutating the protein. Further deletion of the carboxyl terminus of p62 dok (p62 dok ⌬C2) resulted in the loss of both phosphopeptides a and b (Fig. 3C, ⌬C2), indicating that the other major IR-catalyzed tyrosine phosphorylation site(s) was located between residues 359 and 371.
Amino acid sequence analysis of p62 dok revealed the presence of a single tyrosine residue (Tyr 362 ) between residues 359 and 371. To identify whether tyrosine 362 was an IR-catalyzed phosphorylation site, we carried out site-directed mutagenesis and changed this residue to phenylalanine. p62 dok/Y362F was expressed in bacterial cells as a His-tagged protein, purified by Ni 2ϩ -NTA-agarose beads, and phosphorylated in vitro by purified IR tyrosine kinase in the presence of [␥-32 P]ATP. We found that a mutation at Tyr 362 significantly decreased IR-catalyzed p62 dok tyrosine phosphorylation (Fig. 4A, upper panel). Phosphopeptide analysis revealed that mutation of Tyr 362 to phenylalanine resulted in a loss of phosphopeptide a (Fig. 3C,  Y362F). These findings confirmed Tyr 362 as one of the major IR-catalyzed tyrosine phosphorylation sites in p62 dok .
The finding that mutation at Tyr 362 decreased, but did not completely abolish, the IR-catalyzed p62 dok tyrosine phosphorylation suggested the presence of additional tyrosine phosphorylation site(s). To identify the site(s), we replaced every tyrosine residue between amino acids 359 and 475 of p62 dok with phenylalanine and examined the in vitro phosphorylation of these mutants of p62 dok by affinity-purified IR. We found that a mutation at Tyr 377 , Tyr 402 , Tyr 409 , or Tyr 449 had little effect on IR-catalyzed p62 dok tyrosine phosphorylation in vitro (Fig.  4A). On the other hand, a mutation at Tyr 398 almost completely blocked the in vitro phosphorylation of p62 dok by the IR (Fig.   FIG. 3. Identification of potential IR-catalyzed phosphorylation sites on p62 dok . A, diagrams of full-length and various truncations of p62 dok . B, phosphorylation of p62 dok proteins by the IR in vitro. Bacterially expressed, His-tagged full-length and truncated p62 dok proteins were affinity purified with Ni 2ϩ -NTA-agarose beads and phosphorylated by activated IR in the presence of [␥-32 P]ATP as described under "Experimental Procedures." The IR-mediated phosphorylation of p62 dok was visualized by autoradiography (upper panel). The protein levels of p62 dok were determined by Coomassie Blue staining (lower panel). C, two-dimensional phosphopeptide mapping of in vitro phosphorylated p62 dok . In vitro phosphorylated p62 dok protein was separated by SDS-PAGE and transferred onto a nitrocellulose membrane. Radioactive p62 dok protein bands were excised from the membrane, digested with trypsin, separated by thin layer electrophoresis and liquid chromatography, and phosphorylated peptides were visualized by autoradiography. WT, wild-type. 4A). To test further whether Tyr 398 of p62 dok is an IR-stimulated tyrosine phosphorylation site, a double mutant of p62 dok in which Tyr 362 and Tyr 398 were changed to phenylalanine was generated. In vitro phosphorylation studies revealed that mutations at these two sites led to a marked decrease of p62 dok phosphorylation by the IR in vitro (data not shown). However, detectable phosphorylation of this mutant was still observed, suggesting that additional IR-mediated tyrosine phosphorylation site(s) on p62 dok may exist or that a mutation at one or both of these sites generates a compensatory phosphorylation site(s), such as that observed in the phosphopeptide mapping studies (Fig. 3C, phosphopeptide c).
To test whether Tyr 362 and Tyr 398 of p62 dok are insulinstimulated phosphorylation sites in intact cells, we transiently expressed wild-type and mutants of p62 dok in CHO/IR cells. The Myc-tagged p62 dok proteins were immunoprecipitated with antibody to the tag, separated by SDS-PAGE, and the tyrosine phosphorylation of these proteins was determined by Western blot analysis using antibody to phosphotyrosine. Replacement of tyrosine residues at position 377, 402, 409, or 449 with phenylalanine had no significant effect on the insulin-stimulated tyrosine phosphorylation of p62 dok (data not shown). Conversely, mutations of Tyr 362 and Tyr 398 resulted in a notable decrease in insulin-stimulated tyrosine phosphorylation in cells (Fig. 4B, upper panel, lanes 5 and 7 versus lane 3). Consistent with the findings of Noguchi et al. (17), who showed that mutation at Tyr 362 had little effect on insulin-stimulated tyrosine phosphorylation of p62 dok in cells, we found that mutation of both Tyr 362 and Tyr 398 led to only a partial decrease in insulin-stimulated tyrosine phosphorylation of p62 dok (Fig. 4B,  upper panel, lane 9 versus lane 3), suggesting that in vivo, p62 dok was phosphorylated at several additional sites and that the protein is a target for other cellular tyrosine kinases in addition to the IR.
Mutations of five tyrosine residues in p62 dok , including Tyr 362 and Tyr 398 , have previously been shown to decrease p210 bcr-abl kinase-mediated p62 dok tyrosine phosphorylation and its association with GAP in vitro (15). To identify the potential role of Tyr 362/398 phosphorylation in insulin signaling, we overexpressed wild-type and tyrosine phosphorylation site mutants of p62 dok in CHO/IR cells. Cells were then treated with insulin or untreated, and the interaction between p62 dok and its associated proteins was examined by coimmunoprecipitation studies. As shown in Fig. 5A, insulin stimulation resulted in a significant increase in the association of p62 dok with endogenous GAP (upper panel, lane 2 versus lane 1). A partial decrease in p62 dok /GAP association was observed when either tyrosine residue 362 or 398 was individually mutated to phenylalanine (Fig. 5A, upper panel, lanes 4 or 6 versus lane 2). However, replacing both Tyr 362 and Tyr 398 with phenylalanine resulted in a nearly complete abolishment of the interaction between GAP and p62 dok (Fig. 5A, upper panel, lane 8 versus  lane 2). Taken together, these findings indicate that Tyr 362 and Tyr 398 play a key role in the insulin-stimulated interaction between p62 dok and GAP.
Tyrosine phosphorylation of p62 dok at 362 has recently been suggested to be essential for p62 dok to bind the SH2 and SH3 domain-containing adaptor protein Nck (17). To test whether the mutation of p62 dok at Tyr 362/398 affected the interaction between p62 dok and Nck in cells, coimmunoprecipitation studies were carried out. We found that insulin treatment led to association of p62 dok with endogenous Nck (Fig. 5B, top panel,  lane 2 versus lane 1). Whereas mutation of Tyr 398 had little effect on p62 dok /Nck interaction (Fig. 5B, top panel, lane 6  versus lane 2), replacement of Tyr 362 almost completely blocked the interaction between p62 dok and Nck (Fig. 5B, top panel,  lane 4 versus lane 2). Taken together, our findings suggest that  The interaction of p62 dok with GAP has recently been suggested to play a regulatory role in Ras signaling (15,16). Because we found that mutation of Tyr 362/398 affected p62 dok / GAP association (Fig. 5A), we wanted to test whether mutations at these sites had any effect on insulin-stimulated Ras activation. CHO/IR cells transiently expressing HA-tagged Ras protein together with Myc-tagged wild-type or mutant p62 dok were stimulated with or without insulin, and the activation of Ras was determined by Raf binding assays. The RBD of Raf-1 binds only to activated (GTP-loaded), but not inactivated (GDPloaded) Ras (33). As shown in Fig. 6A, treatment of cells with insulin led to an increase in coprecipitation of HA-tagged Ras with the GST-Raf -RBD (top panel, lane 3 versus lane 2). The insulin-stimulated Ras/Raf-RBD interaction was inhibited greatly in cells coexpressing wild-type p62 dok (Fig. 6A, top  panel, lane 5 versus lane 3) but not with those coexpressing p62 dokY362F/Y398F (Fig. 6A, top panel, lane 7 versus lane 3). Similar results were obtained in human embryonic kidney 293 cells transiently expressing Ras and wild-type or mutant p62 dok (data not shown). Quantification of Ras bound to GST-Raf-RBD indicated that wild-type, but not the tyrosine phosphorylation site mutant of p62 dok , inhibited insulin-stimulated Ras activation by ϳ3-fold (Fig. 6B). Taken together with the findings that mutations at Tyr 362 and Tyr 398 inhibited p62 dok / GAP association (Fig. 5A), our data suggest that tyrosine phosphorylation of p62 dok at Tyr 362 and Tyr 398 plays an important role in regulating Ras activity.
To test whether inhibition of Ras affects downstream signaling, we examined whether overexpression of p62 dok had any effect on insulin-stimulated MAP kinase activation. CHO/IR cells were transfected with Myc-tagged MAP kinase alone or together with either wild-type or the Y362F/Y398F mutant of p62 dok , and the insulin-stimulated MAP kinase phosphorylation was determined by Western blot using a phospho-specific antibody to MAP kinase. As expected, insulin treatment led to a marked increase in MAP kinase phosphorylation (Fig. 7A,  lane 2 versus lane 1). Overexpression of p62 dok significantly decreased insulin-stimulated MAP kinase phosphorylation (Fig. 7A, lane 4 versus lane 2). A similar inhibition of insulinstimulated MAP kinase phosphorylation was detected in cells expressing the Y362F/Y398F mutant of p62 dok (Fig. 7A, 1 and 2) or together with either p62 dok (lanes 3 and 4) or p62 dok Y362F/Y398F (lanes 5 and 6) were serum starved for 2 h, treated with 10 nM insulin for 5 min (ϩ) or not (Ϫ), and lysed. Cell lysates were analyzed by Western blot (WB) analysis using a phospho-specific antibody to MAP kinase (ERK1/2) (top panel). Expression of MAP kinase and p62 dok was determined by Western blot using antibody to the Myc tag (middle and bottom panels). Results are representative of three experiments with similar results. B, p62 dok inhibits insulin-stimulated Akt phosphorylation at Thr 308 and Ser 473 . CHO/IR cells transiently expressing Myc-tagged Akt and wildtype or mutant p62 dok were serum starved for 2 h, treated with 10 nM insulin for 5 min (ϩ) or not (Ϫ), and lysed. Cell lysates were analyzed by Western blot analysis using phospho-specific antibodies to Akt Thr-308 and Akt Ser-473 (top and middle panels, respectively). Expression of Akt and p62 dok was determined by Western blot using antibody to the tag (bottom panel). Results are representative of three experiments with similar results. C, p62 dok inhibits insulin-stimulated Akt activity. FLAG-tagged Akt was expressed alone or coexpressed with wild-type or mutant p62 dok in CHO/IR cells. 24 h after transfection, cells were starved for 2 h, left untreated or treated with 10 nM insulin for 5 min, and lysed. Akt was immunoprecipitated using antibody to the tag, and the activity of the bound protein was determined as described under "Experimental Procedures." Data are the means Ϯ S.E. from three independent experiments (* t test p Ͻ 0.002). WT, wild-type. versus lane 2). These results suggest that p62 dok can negatively regulate insulin-mediated MAP kinase activation in a GAP binding-independent manner. Because Ras has also been shown to activate PI 3-kinase and its downstream kinase Akt (19,34), we tested whether inhibition of Ras activation by p62 dok had any effect on Akt phosphorylation at Thr 308 and Ser 473 , both of which are essential for full activation of the enzyme (35). CHO/IR cells were transiently transfected with plasmids encoding Myc-tagged Akt alone or together with Myc-tagged wild-type or the Y362F/ Y398F mutant of p62 dok . Cells were untreated or treated with insulin, lysed, and the insulin-stimulated phosphorylation of Akt at Thr 308 and Ser 473 was examined by Western blot using phospho-specific antibodies. As shown in Fig. 7B Overexpression of wild-type but not the p62 dok Y362F/Y398F mutant also inhibited insulin-stimulated Akt kinase activity significantly (Fig. 7C). These findings suggest that phosphorylation of p62 dok at tyrosine 362/398 is essential for p62 dok to inhibit insulin-stimulated phosphorylation and activation of Akt.
To characterize the mechanism by which p62 dok inhibits insulin-stimulated Akt activation, we tested whether overexpression of wild-type or constitutively active Ras could block the inhibitory effect of p62 dok . CHO/IR cells were transfected with Myc-Akt alone or together with wild-type p62 dok and either wild-type or constitutively active Ras Val-12 . As expected, coexpression of wild-type p62 dok inhibited insulin-stimulated phosphorylation of Akt (Fig. 8, top panel, lane 4 versus lane 2). Coexpressing wild-type Ras had little effect on the inhibition of insulin-stimulated Akt phosphorylation by p62 dok (Fig. 8, top  panel, lane 6 versus lane 2). On the other hand, the inhibition of insulin-stimulated Akt phosphorylation by p62 dok was greatly rescued by coexpression of constitutively active Ras Val-12 (Fig. 8,  top panel, lane 8 versus lane 2). We also found that overexpression of p62 dok did not inhibit insulin-stimulated IRS-1 tyrosine phosphorylation (data not shown). Taken together, these findings suggest that Akt is a downstream effector of Ras and that inhibition of insulin-stimulated Akt phosphorylation by p62 dok is most likely caused by inhibition of Ras activation rather than inhibition of IRS-1 tyrosine phosphorylation. DISCUSSION The pleckstrin homology and phosphotyrosine binding domain-containing protein p62 dok contains multiple potential ty-rosine phosphorylation sites at its carboxyl terminus and undergoes rapid tyrosine phosphorylation in cells in response to stimulation with growth factors such as insulin and insulinlike growth factor I. p62 dok has been shown to be a substrate for nonreceptor tyrosine kinases such as Src (17,36), Lyn (37,38), Tec (38), and p210 bcr-abl (15); however, whether p62 dok is a direct in vivo substrate for receptor tyrosine kinases remains to be established. p62 dok undergoes insulin-stimulated tyrosine phosphorylation in CHO/IR cells (Ref. 17 and Fig. 4B). In this report, we have demonstrated that p62 dok is a direct substrate for the IR tyrosine kinase. First, insulin treatment led to very rapid tyrosine phosphorylation of p62 dok in cells (Fig. 1A), suggesting that the phosphorylation event occurred in close proximity to the IR. In addition, p62 dok was phosphorylated by purified IR in vitro (Fig. 1B). Furthermore, the in vitro IR-catalyzed and the in vivo insulin-stimulated p62 dok tyrosine phosphorylation were inhibited competitively by Grb10, an adaptor protein that binds to the kinase domain of the IR (21, 22) (Fig. 2). Our results using exogenously expressed p62 dok have confirmed previous findings that the 60-kDa GAP-associated protein could be phosphorylated directly by the IR (8).
Mutations at five tyrosine residues on p62 dok , including Tyr 296 , Tyr 315 , Tyr 362 , Tyr 398 , and Tyr 409 , have previously been shown to decrease the p210 bcr-abl kinase-catalyzed p62 dok tyrosine phosphorylation and its association with GAP in vitro (15). However, whether these tyrosine residues undergo insulinstimulated phosphorylation and whether all five sites are necessary for the interaction between p62 dok and GAP remain to be established. Our results show that mutations of Tyr 362 and Tyr 398 greatly reduced the insulin-stimulated tyrosine phosphorylation of p62 dok in vitro (Fig. 4A), suggesting that these residues are the major tyrosine phosphorylation sites phosphorylated directly by the IR. Phosphorylation at Tyr 362 and Tyr 398 appears to be critical for p62 dok binding to GAP because mutations at these sites almost completely blocked the interaction of p62 dok with GAP in vivo (Fig. 5A). However, mutations at these two sites only partially decreased the insulin-stimulated tyrosine phosphorylation of p62 dok in intact cells (Fig. 4B), indicating that additional tyrosine phosphorylation sites exist on p62 dok . These findings also suggest that in addition to the IR, other insulin-responsive tyrosine kinases may be involved in the phosphorylation of p62 dok in cells. Our studies also showed that mutation at tyrosine residue 362, but not 398, abolished the interaction between p62 dok with Nck (Fig. 5B). These findings indicate that tyrosine phosphorylation at distinct sites may provide specificity for p62 dok to interact with different signaling molecules in cells.
One of the downstream targets of Ras is PI 3-kinase (39,40), and activation of PI 3-kinase by Ras has been shown to promote survival through Akt in several cell systems (41). In agreement with these findings, we have shown that overexpression of p62 dok inhibits insulin-stimulated activation of Ras and Akt. The inhibition of Akt phosphorylation and activation by p62 dok are most likely the result of its association with GAP and inactivation of Ras, rather than inhibition of insulin-stimulated IRS-1 tyrosine phosphorylation. First, p62 dok interacts with GAP in a tyrosine phosphorylation-dependent manner, and the formation of the p62 dok ⅐GAP complex inhibits Ras activation (Fig. 6). In addition, mutations at tyrosine residues 362 and 398 of p62 dok inhibited its ability to associate with GAP (Fig. 5A) and reversed its ability to inhibit insulin-stimulated Ras and Akt activation (Figs. 6 and 7B). Furthermore, overexpression of constitutively active Ras was sufficient to block the inhibitory effect of p62 dok on Akt phosphorylation (Fig. 8). Finally, overexpression of p62 dok had no significant FIG. 8. p62 dok inhibits insulin-stimulated Akt phosphorylation through a Ras-dependent mechanism. CHO/IR cells were cotransfected with Myc-Akt and empty vector or wild-type p62 dok along with HA-tagged wild-type or constitutively active Ras. Cells were serumstarved for 2 h, treated with or without 10 nM insulin for 5 min, and lysed. Cell lysates were analyzed directly using phospho-specific antibodies to Akt Thr-308 (top panel). Expression of proteins was determined by Western blot (WB) using antibody to the Myc (Akt and p62 dok ) (middle panel) and HA (Ras) (bottom panel) tags. The data are representative of three experiments with similar results. effect on insulin-stimulated IRS-1 tyrosine phosphorylation in cells (data not shown), indicating that the inhibition of insulinstimulated Akt phosphorylation was IRS-1-independent. Our results are consistent with a recent finding that tyrosine phosphorylation of p62 dok is essential for inhibition of the activity of Ras (15,16). The finding that the inhibitory effect of p62 dok on insulin-stimulated Akt phosphorylation could be rescued by coexpression of constitutively active Ras Val-12 is in agreement with recent findings that overexpression of p62 dok inhibited Src tyrosine kinase-mediated cellular transformation but had no effect on Ras Val-12 -induced focus formation in NIH3T3 cells (42). Taken together, our results suggest that insulin-stimulated Akt activation may be mediated by both the IRS-1 and Ras signaling pathways. In addition, p62 dok acts upstream of Ras to block Ras-induced Akt phosphorylation and activation.
In addition to inhibiting insulin-stimulated Akt activation, we also found that overexpression of wild-type p62 dok inhibited insulin-stimulated phosphorylation of MAP kinase (Fig. 7A). This finding is consistent with many studies showing that Ras is an upstream effector of MAP kinase in the defined and conserved MAP kinase signaling module (43). However, our result is contrary to a recent finding that stably expressed p62 dok in CHO/IR cells had no effect on MAP kinase phosphorylation (17). Because activation of MAP kinase is essential for cell proliferation and growth (43), one possible explanation for this discrepancy may be that stable overexpression of p62 dok led to a selection of clones that had escaped the negative regulation of p62 dok . Unexpectedly, our results also showed that the p62 dok Y362F/Y398F mutant, which was unable to bind and inhibit Ras (Figs. 5A and 6), was still able to inhibit insulinstimulated MAP kinase phosphorylation (Fig. 7A). One possibility for this may be that, in addition to inhibition of Ras, p62 dok can inhibit an effector downstream of Ras in the MAP kinase signaling pathway. Another potential explanation may be that p62 dok inhibits a Ras-independent pathway necessary for insulin-stimulated MAP kinase phosphorylation. Further studies will be needed to test these possibilities.
In summary, the present work provides evidence that p62 dok is a direct substrate for the IR. In addition, we have presented the first evidence that phosphorylation of p62 dok by the IR at Tyr 362/398 is essential for p62 dok to regulate insulin-stimulated downstream events such as phosphorylation and activation of Ras and Akt negatively. This negative regulation may provide an important mechanism to keep insulin signaling in check so that uncontrolled amplification of the signal is prevented.