k JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 269, No. 14, Issue of April 8, pp. 10734-10738, 1994 Printed in U S A

Shc Is the Predominant Signaling Molecule Coupling Insulin Receptors to Activation of Guanine Nucleotide Releasing Factor and pBl"-GTP Formation*

(Received for publication, January 7, 1994, and in revised form, February 1, 1994) 'lbshiyasu Sasaoka, Boris DrazninS, J. Wayne LeitnerS, W. John Langlois, and Jerrold M. OlefskyQ From the Department of Medicine, Division of Endocrinology and Metabolism, University of California, Sun Diego, La Jolla, California 92093, the Veterans Administration Medical Center, Medical Research Service, Sun Diego, California 92161, and the $Medical Research Service and the Department of Medicine, Veterans Affairs Medical Center and the University of Colorado Health Sciences Center, Denver, Colorado 80220

Insulin stimulates tyrosine phosphorylation of insulin receptor substrate-1 (IRS-1) and Shc in Rat1 fibroblasts overexpressing wild type insulin receptors. We investi- gated the relative role of IRS-1 and Shc in insulin acti- vation of guanine nucleotide releasing factor (GNRF) and p21"GTP formation. The time course of insulin- stimulated tyrosine phosphorylation of IRS-1 was rapid, whereas Shc phosphorylation was relatively slow. Growth factor receptor bound protein-2 (Grb2) associ- ated with IRS-1 rapidly and gradually dissociated after 5 min, whereas Grb2 association with Shc was slower and reached a maximum at 10 min after insulin stimu- lation. Thus, the kinetics of Grb2 association with IRS-1 and Shc corresponded closely to the time course of ty- rosine phosphorylation of IRS-1 and Shc, respectively. Importantly, 3-13-fold more Grb2 was associated with Shc than with IRS-1. In addition, the kinetics of insulin- stimulated GNRF activity and p21"GTP formation corresponded more closely to the time course of Shc phosphorylation than to the kinetics of IRS-1 phospho- rylation. Furthermore, immunoprecipitation of Shc pro- teins from cell lysates of insulin-stimulated cells re- moved 67% of the GNRF activity, whereas precipitation of IRS-1 had a negligible effect on GNRF activity. Thus, although both IRS-1 and Shc associate with Grb2, the current results indicate that Shc plays a more important role than IRS-1 in insulin stimulation of GNRF activity and subsequent p21"GTP formation.

Insulin binding to the extracellular a-subunits activates the intrinsic tyrosine kinase activity of the cytoplasmic portion of the insulin receptor P-subunit (1). One early molecular event linking the receptor kinase to insulin's biologic actions is tyro- sine phosphorylation of IRS-1' (2). Current evidence suggests that IRS-1 acts as a multisite "docking" protein by binding to downstream signal-transducing molecules (3). The physical in-

* This work was supported in part by Grant DK33651 from the Na- tional Institute of Diabetes and Digestive and Kidney Diseases, Na- tional Institutes of Health, by the Veterans Administration Medical Research Service, by the Sankyo Diabetes Research Fund, by a research fellowship grant from the Medical Research Council of Canada, and by an American Diabetes Association Mentor-Based Fellowship Award. The costs of publication of this article were defrayed in part by the

"advertisement" in accordance with 18 U.S.C. Section 1734 solely to payment of page charges. This article must therefore be hereby marked

indicate this fact. 0 To whom correspondence should be addressed: Dept. of Medicine

(06731, University of California, San Diego, 9500 Gilman Dr., La Jolla,

Src homology; PI 3-kinase, phosphatidylinositol 3-kinase; GAP, GT- The abbreviations used are: IRS-1, insulin receptor substrate 1; SH,

Pase-activating protein; EGF, epidermal growth factor; PMSF, phenyl- methylsulfonyl fluoride; PAGE, polyacrylamide gel electrophoresis; GNRF, guanine nucleotide releasing factor.

CA 92093-0673. Tel.: 619-534-6651; Fax: 619-534-6653.

teraction is mediated by binding of phosphotyrosine IRS-1 mo- tifs with Src homology 2 (SH2) domains of cytoplasmic signal- ing proteins (4,5). Recently, Pelicci et al. (6) have characterized a cDNA clone that encodes a protein termed Shc. Shc, which is ubiquitously expressed, contains a single SH2 domain at the C terminus but lacks catalytic activity (6). Shc protein is also tyrosine-phosphorylated after insulin stimulation but does not associate stably with either the insulin receptor or IRS-1(7,8).

Current evidence indicates that p21" is a key molecule in the mitogenic signaling pathways initiated by receptor tyrosine kinases, including the insulin receptor (9-12). p21" becomes active as a signaling molecule when it is converted from the GDP- to GTP-bound form, and this process is controlled by Ras GTPase-activating protein (GAP) and GNRF activity (12). Ras GAP promotes the intrinsic GTPase activity of p21m (13), whereas GNRF mediates the dissociation of GDP from p21"" (14-16). Recent reports have provided evidence that insulin increases p21"-GTP primarily by stimulation of GNRF activ- ity, rather than by inhibition of GAP activity (17, 18). Other growth factors also appear to increase p21""-GTP levels through GNRF (19,20). A protein called son of sevenless (Sos) was identified as a GNRF in Drosophila (141, and a mammalian homologue to Sos (16), appears to be the GNRF mediating growth factor actions. The proline rich region of Sos binds to the SH3 domain of Grb2, which is an adapter protein composed of one SH2 domain and two SH3 domains (21), and preformed Grb2.Sos complexes exist within unstimulated cells (16, 22- 27). Through the Grb2 SH2 domain, Grb2.Sos complexes can bind to phosphorylated epidermal growth factor (EGF) and platelet-derived growth factor receptors, providing a mecha- nism whereby these growth factors can stimulate p2lm8-GTP formation (16, 22-25). However, Grb2.Sos complexes do not associate with the insulin receptor (8, 26). Activation of the insulin receptor leads to tyrosine phosphorylation of IRS-1 and Shc, both of which can then bind to Grb2.Sos complexes (7, 8, 26, 27). Thus, there are two downstream coupling molecules, IRS-1 and Shc, which can potentially independently connect the activated insulin receptor to Grb2.Sos and stimulation of p21"GTP formation. We have recently used single cell micro- injection studies to inhibit Shc and IRS-1, and we have shown that both molecules are necessary for insulin's overall growth stimulus (28).' It is possible that Shc and IRS-1 represent ad- ditive complementary pathways allowing insulin to couple more effectively to p21". On the other hand, one of these path- ways could be dominant with the other redundant. Therefore, the purpose of this study was to evaluate the relative roles of Shc and IRS-1 in mediating the insulin-induced increase in cellular p21'"-GTP.

Sasaoka, T., Rose, D. W., Jhun, B. H., Saltie1,A. R., Draznin, B., and Olefsky, J. M. (1994) J. Biol. Chern. 269, in press.

10734

Role of She in Insulin Action 10735

EXPERIMENTAL PROCEDURES Materials-Rat1 cells expressing wild type insulin receptors (HI&)

were maintained as previously described (29). Porcine insulin was kindly provided by Lilly. The p21" probe (v-Ha-Ras) was purchased from Oncor, Inc. (Gaithersburg, MD). [32P10rthophosphate (0.25 mCil ml) and L3H1GDP (32 Ci/mmol) were from DuPont NEN. Electrophoresis reagents were from Bio-Rad. Enhanced chemiluminescence reagents were from Amersham Corp. A monoclonal anti-phosphotyrosine anti- body (pY20) was from ICN. An anti-p21m" monoclonal antibody (Y13- 259) was from Oncogene Science, Inc. (Uniondale, NY). Apolyclonal and a monoclonal anti-Shc antibody, as well as a monoclonal anti-Grb2 antibody, were from Transduction Laboratories (Lexington, KY). A poly- clonal anti-IRS-1 antibody was kindly provided by Dr. Hiroshi Maegawa (Shiga University of Medical Science, Japan). All other routine reagents were purchased from Sigma.

Western Blotting Studies-Cells were starved for 24 h in serum-free Dulbecco's modified Eagle's medium. The cells were then treated with 17 I ~ M insulin a t 37 "C. After the indicated time, cells were lysed in a buffer containing 30 m Tris, 150 m NaCl, 10 m EDTA, 0.5% sodium deoxycholate, 1% Triton X-100, 1 m phenylmethylsulfonyl fluoride (PMSF), 10 pg/ml aprotinin, 10 pg/ml leupeptin, 1 m Na3V04, pH 7.4. The cell lysates were centrifuged, and supernatants were used for im- munoprecipitation with the indicated antibodies. Immunoprecipitated or supernatant proteins were separated by SDS-PAGE and transfemed onto Immobilon-P by electroblotting. After incubation with the specified antibody, enhanced chemiluminescence detection was performed ac- cording to the manufacturer's instructions (Amersham Corp.).

GNRF Activity in HIRc Cell LysatesSerum-starved cells were stimulated with 17 m insulin for the indicated time a t 37 "C. Cells were then lysed in a buffer containing 50 m Hepes, 150 m NaCl, 5 m MgCl,, 1 m PMSF, 1 m Na,VO,, 1% Triton X-100, 0.05% SDS, 10 pg/ml aprotinin, 10 pg/ml leupeptin, 500 w GTP, and 500 w GDP, pH 7.5. GNRF activity in the lysates was determined by measuring the dissociation of protein-bound c3H1GDP radioactivity. Purified v-Ha-Ras was incubated with c3H1GDP in a buffer containing 25 m Tris, 100 m NaCI, 1 m EDTA, 1 m dithiothreitol, 400 pg/ml bovine serum albu- min, pH 7.5, for 15 min a t 30 "C. The complex was stabilized by addition of the incubation buffer supplemented with 10.7 m MgCl,. The Ras.GDP complex was added to the cell lysates and incubated at 23 "C. At the indicated time intervals, aliquots were removed and filtered through 0.45-pm Millipore nitrocellulose filters. The amount of I3H1GDP radioactivity bound to p21"" was quantitated by scintillation counting. c3H1GDP binding to p21" was confirmed by immunoprecipi- tation with anti-Ras antibody (17).

Measurements of GTP- and GDP-bound p21"Serum-starved cells were incubated for 1 h in phosphate-free serum-free Dulbecco's modified Eagle's medium and labeled with [32Plorthophosphate for 4 h. Insulin (17 m) was added a t 37 "C. f i r the indicated time, cells were lysed in a buffer containing 50 m Hepes, 150 mM NaCl, 5 m MgCI,, 1 m PMSF, 1 m dithiothreitol, 1 m sodium vanadate, 1 m Na,PO,, 1% Triton X-100, 0.05% SDS, 10 pg/ml aprotinin, 10 pg/ml leupeptin, 500 w GTP, 500 GDP, pH 7.5. The lysates were centrifuged, and GTP- and GDP-bound p21" was immunoprecipitated with the monoclonal antibody Y13-259. Nucleotides were eluted by suspending immunopre- cipitates, and heating a t 85 "C for 3 min. Separation of GTP from GDP was then performed by thin layer chromatography (17).

RESULTS To investigate the kinetics of tyrosine phosphorylation of

Shc, the insulin receptor 0-subunit, and IRS-1, we studied the time course of tyrosine phosphorylation after insulin treatment in Rat1 fibroblasts overexpressing insulin receptors. As shown in Fig. lA, phosphorylation of the receptor P-subunit and IRS-1 was observed after 30 s of insulin stimulation, peaked by -1 min, and declined after 10 min. The bands corresponding to the insulin receptor 0-subunit and IRS-1 were confirmed by using specific antibodies (data not shown). To assess Shc phosphoryl- ation, cell lysates were immunoprecipitated with a polyclonal anti-Shc antibody, and the immunoprecipitates were immuno- blotted with a monoclonal anti-phosphotyrosine antibody. The time course of insulin-induced Shc phosphorylation was slower compared with the insulin receptor 0-subunit and IRS-1. Phos- phorylation was first observed at 1 min, reached a maximum a t 10 min, and then declined (Fig. 1B 1. These results are summa-

0 0.5 1 2.5 5 10 20 Tfme (mln)

66 4

28== "

0 0.5 1 2.5 5 10 20 Time (min)

C

Time (min)

FIG. 1. Time course of insulin-stimulated tyrosine phospho- rylation of insulin receptor p-subunit, IRS-1, and Shc. A, time course of insulin receptor P-subunit and IRS-1 phosphorylation. Serum- starved cells were treated with insulin for the indicated times. The

lyzed by immunoblotting with an anti-phosphotyrosine antibody. Mo- supernatants after anti8hc antibody immunoprecipitation were ana-

lecular mass ofinsulin receptor ksubunit (95 kDa) and IRS-l(l75 kDa) is shown by arrows. B, time course of Shc phosphorylation. The cells were treated as described above, and anti-Shc antibody immunoprecipi- tates were analyzed by anti-phosphotyrosine antibody. Molecular mass of Shc isoforms (46,52, and 66 kDa) is shown by arrows. C, percentage maximal tyrosine phosphorylation of insulin receptor P-subunit (A), IRS-1 (O), and Shc (0). Results are representative of three separate experiments.

rized in Fig. 1C. It is clear that tyrosine phosphorylation of the insulin receptor P-subunit and IRS-1 proceeds more rapidly compared with Shc.

It has been shown that both IRS-1 and Shc associate with Grb2 following insulin stimulation (8, 26,27), and we assessed the time course of Grb2 association with these two molecules. After insulin treatment, cell lysates were immunoprecipitated with antibodies against IRS-1 or Shc, and the precipitates were immunoblotted with a Grb2 antibody. The IRS-1-Grb2 associa- tion was rapid, with complexes detected after 30 s of insulin treatment, peaking a t 1 min, and declining thereafter (Fig. 2A). On the other hand, complex formation between Shc and Grb2 proceeded more slowly. Association was first apparent at 1 min, reached a maximum by 10 min, and declined thereafter (Fig. 2B). These results are summarized in Fig. 2C, and the pattern of these time courses were comparable to the kinetics of Shc uersus IRS-1 phosphorylation, as seen in Fig. 1.

To assess Grb2 association with IRS-1 or Shc more directly, we performed sequential immunoprecipitation studies using anti-IRS-1 and anti-Shc antibodies. After insulin stimulation, cell lysates were first immunoprecipitated with the anti-Shc antibody, and the remaining supernatants were re-immunopre- cipitated with anti-IRS-1 antibody. The anti-Shc and anti-

10736 A

Role of Shc in Insulin Action A

0 0.5 1 2.5 5 10 20 Time (min)

B

25- ”-

0 0.5 1 2.5 5 10 20 Time (mid

C

0 5 10 15 20 Time(m1n)

FIG. 2. Insulin-induced IRS-1 and Shc association with Grb2.A, IRS-1 association with Grb2. Serum-starved cells were treated with insulin for the indicated times. Cell lysates were immunoprecipitated with anti-IRS-1 antibody and analyzed by immunoblotting with Grb2 antibody. Molecular mass of Grb2 (25 kDa) is shown by an arrow. B, Shc association with Grb2. The cells were treated as described above, and anti-Shc antibody precipitates were analyzed by immunoblotting with anti-Grb2 antibody. C, Grb2 association with IRS-1 (0) or Shc (0) is summarized as a percentage of maximal association, and results are representative of three separate experiments.

IRS-1 immunoprecipitates and the final supernatants were then immunoblotted with anti-Grb2 antibody. As can be seen in Fig. 3A, there was minimal association of Grb2 with IRS-I or Shc in the basal state. After 1 min of insulin stimulation, a small amount of Grb2 associated with both IRS-1 and Shc, but even at this early time, the absolute amount of Shc.Grb2 com- plexes far exceeded the amount of IRS-1-Grb2. By 10 min, a large amount of Grb2 was associated with Shc, whereas Grb2 association with IRS-1 had already declined (Fig. 3A). Three separate experiments are summarized in Fig. 3B. Of total Grb2, 5.1 e 1.3 and 14.1 k 0.8% associated with IRS-1 or Shc, respectively, after 1 min of insulin treatment. At 10 min, 2.4 k

1.3 and 31.7 e 1.1% of the Grb2 was complexed with IRS-1 versus Shc, respectively. Reversing the order of the immuno- precipitation (first with anti-IRS-1 and second with anti-Shc) produced the same results (data not shown), thus, demonstrat- ing that Grb2 does not bind to IRS-1 and Shc simultaneously.

Preformed Grb2.Sos complexes exist within the cell (16,22- 271, and insulin stimulation leads to the association of these complexes with either IRS-1 or Shc (8, 26, 27). To further ex- plore which association is more important for insulin-stimu- lated GNRF activity, we measured GNRF activity in cell lysates before and after precipitation of the lysates with anti-IRS-1 or anti-Shc antibody. As can be seen in Table I, precipitation of Shc from insulin-stimulated cells removed 67% of the total GNRF activity, whereas precipitation of IRS-1 had a negligible effect on GNRF activity.

To confirm that our results were not related to the particular antibodies used in these studies, comparable experiments were conducted with a different set of antibodies, with identical re-

25+ - B

p1 ‘ 60 5 40

20

3 0

n ” 0 1 1 0

Time (mid

FIG. 3. Comparison of Grb2 association with IRS-1 or She. A, serum-starved cells were treated with insulin for the indicated times. Cell lysates were immunoprecipitated with anti-Shc antibody and the supernatants were re-immunoprecipitated with anti-IRS-1 antibody. The anti-Shc and anti-IRS-1 precipitates and the final supernatants were separated by SDS-PAGE and immunoblotted with anti-Grb2 an- tibody. Molecular mass of Grb2 (25 kDa) is shown by an arrow. B , summary of Grb2 association with IRS-1 and Shc. Results of Grb2 association with IRS-1 (open bars) or Shc (hatched bars), as well as remaining non-precipitated Grb2 (dotted bars), are summarized as a percentage of total Grb2. The results are the mean S.E. of three separate experiments. The anti-Shc antibody precipitated 90% of phos- phorylated Shc proteins, and the anti-IRS-1 antibody precipitated 70% of phosphorylated IRS-1. Given the large differences shown inpanel B, correcting for the lower efficiency of the anti-IRS-1 antibody would not alter the overall results.

TABLE I GNRF actiuvity in cell lysates afier immunoprecipitation with

anti-IRS-1 or anti-Shc antibody Serum-starved cells were stimulated with 17 m insulin for 10 min at

37 “C. GNRF activity in the cell lysates before and after immunopre- cipitation with antibodies was determined by measuring the loss of protein-bound f3H1GDP radioactivity as described under ‘Materials and Methods.” Results are expressed as the percentage of reduction from total GNRF activity by immunodepletion, and are shown as the mean = S.E. of four separate experiments.

Antibody Reduction of GNRF activity

9%

Control preimmune IgG 4.3 f 7.9 Anti-IRS-1 antibody 10.6 = 5.5 Anti-SHC antibody 67.1 = 12.6

sults. Thus, a monoclonal anti8hc antibody was used to per- form the Grb2 association and GNRF immunodepletion experi- ments, and the results were comparable to those obtained with the polyclonal antibody in Fig. 3 and Table I. In addition, we used an anti-phosphotyrosine antibody (pY20) to immunopre- cipitate post-anti-Shc antibody supernatants. The major phos- phoprotein species in these supernatants are the insulin recep- tor P-subunit and IRS-1 (Fig. LA). Immunoprecipitation with pY20 removes more than 90% of phosphorylated IRS-1 (data not shown), and, consequently, any Grb2-Sos associated with IRS-1 would be found in the pY20 precipitates. When the post

Role of She in Insulin Action 10737

A

I I

0 10 2o 30 Time [min)

FIG. 4. Time course of insulin stimulation of GNRF activity and p21"GTP formation. A, GNRF activity. The Ras.GDP complex was incubated with cell lysates obtained after insulin stimulation for the indicated time. GNRF activity was determined by measuring the loss of protein-bound radioactivity using the nitrocellulose filter binding assay as described under "Materials and Methods". The results were expressed as the mean 2 S.E. of six separate experiments. B , p2lros-GTP formation. p21"" was immunoprecipitated from the cell lysates with Y13-259 antibody. GTP and GDP were separated by thin layer chro- matography and analyzed by autoradiography. Quantitation of nucle- otides was determined by counting the labeled spots in a liquid scintil- lation counter. The results were expressed as GTP/(GTP + GDP) x 100% after subtraction of the percentage of GTP-bound p21"" in the absence of insulin. Results are the mean 2 S.E. of six separate experiments.

anti-Shc, pY20 immunoprecipitates were analyzed by SDS- PAGE and Western blotting with the anti-Grb2 antibody, neg- ligible Grb2 was observed (data not shown), whereas the ma- jority of the Grb2.Sos was contained in the anti-Shc precipitates (Fig. 3 and Table I).

We have previously shown that insulin stimulates GNRF activity (171, and this accounts for the increase in p2lraS-GTP formation. The time course of these effects is seen in Fig. 4. GNRF stimulation was not statistically significant until 2.5 min, and reached a maximum at 10 min of insulin treatment (Fig. 4A). The increase in p21'""-GTP was slightly delayed rela- tive to the time course of GNRF activity (t , , , -3 min for GNRF and -6 min for p2lrUs-GTP), consistent with the notion that GNRF mediated GDP exchange leads to the insulin-induced increase in p21""-GTP (Fig. 4B 1. These time courses correspond more closely to the kinetics of Shc phosphorylation and Shc- Grb2 association than to the time course of IRS-1 phosphory- lation.

DISCUSSION

The SH2 domain of Grb2 can directly bind t o a phos- photyrosine motif in the EGF receptor, forming EGF re- ceptor.Grb2.Sos complexes (22-251, In contrast, Grb2 does not associate directly with the insulin receptor (7,8). However, it can associate through its SH2 domain with IRS-1 and Shc (7, 8, 26, 271, two proteins that undergo tyrosine phosphorylation upon insulin stimulation (2, 3, 7, 8). Consequently, either IRS-1

or Shc, or both, could couple insulin receptors to Grb2.Sos. In this study, we sought to elucidate the roles of IRS-1 and Shc in linking the insulin signal to Grb2.Sos and ultimately to stimu- lation of p2lraS-GTP.

IRS-1 was rapidly phosphorylated after addition of insulin, consistent with its role as a direct substrate of the insulin receptor (2, 3). In contrast, Shc phosphorylation proceeded more slowly, with a lag time before appreciable phosphoryla- tion was detected, consistent with previous observations (7). Shc does not bind to the insulin receptors through SH2 domain interactions (7, 8). However, direct, transient, low affinity as- sociation with the receptor is possible. On the other hand, the lag phase in Shc phosphorylation raises the possibility that an intermediary tyrosine kinase mediates the phosphorylation of Shc. This is in contrast to the EGF signaling system, in which the time course of Shc phosphorylation is very rapid, peaking a t about 1 min (data not shown), and in which Shc binds directly to phosphorylated EGF receptors (22-25). In any event, the time course of insulin-stimulated Shc phosphorylation corre- sponds more closely to the kinetics of GNRF activity and p21""- GTP formation than does the time course of IRS-1 phosphory- lation, suggesting that Shc plays a more prominent role than IRS-1 in coupling the insulin receptor to ~ 2 1 " ~ activation.

We also found that immunoprecipitation of lysates from in- sulin-stimulated cells with anti-Shc antibody removed 67% of the cellular GNRF activity, whereas anti-IRS-1 antibody pre- cipitation had only a negligible effect to deplete GNRF. Since Sos mediates GNRF activity and it is constitutively bound to Grb2 (16,22-27), one would expect insulin-induced association of Grb2 with Shc and/or IRS-1 to parallel GNRF activity. This was indeed the case. Thus, the kinetics of Grb2 association with Shc closely paralleled the time course of insulin-stimulated Shc phosphorylation (Fig. 21, and much more of the cellular Grb2 formed complexes with Shc than with IRS-1. For these reasons, we conclude that Shc is the quantitatively dominant pathway linking the activated insulin receptor, through GrbB.Sos, to p21"" activation.

Since the initial identification of the Shc cDNA, increasing evidence has accrued indicating that Shc is an important com- ponent of growth signaling pathways (6,30-33). To specifically identify a functional role for Shc in insulin action, we previ- ously carried out single cell microinjection studies with anti- Shc antibody. Our results demonstrated that microinjection of anti-Shc antibody into HIRc cells led to an 80% inhibition of subsequent insulin-induced DNA synthesis.2 These findings, taken together with the current results, underscore the impor- tance of Shc in the insulin mitogenic signaling pathway and indicate that it is the major molecular connection between the insulin receptor and the Grb2~Sos-p2lra" signaling cascade.

Although it is clear that Shc is a key molecule in the insulin mitogenic signaling pathway, and that IRS-1 is not an up- stream regulator of Shc, substantial evidence exists indicating that IRS-1 is also a necessary component of insulin's growth promoting effects. For example, expression of antisense IRS-1 mRNA leads to inhibition of insulin-stimulated thymidine in- corporation into Chinese hamster ovary cells (34). In addition, expression of IRS-1 in the hematopoietic 32D cell line, which lacks endogenous IRS-1, allows these cells to respond to insulin with DNA synthesis (35). Lastly, microinjection of an IRS-1 antibody almost completely inhibits subsequent insulin stimu- lation of DNA synthesis in HIRc cells (28). Thus, it is evident that IRS-1 is also an important molecule in insulin's mitogenic signaling actions.

To reconcile the above observations, it is unlikely that all of the involved molecules line up in a linear pathway, and parallel pathways seem possible. For example, IRS-1 and Shc may sub- serve parallel colinear pathways, leading to activation of p21'"",

10738 Role of Shc in Insulin Action

with Shc having the major input. Even though the Shc pathway 7. Pronk, G. J., McGlade, J., Pelicci, G., Pawson, T., and BOS, J. L. (1993) J. B i d .

is both inputs each Other to generate 8. Skolnik, E. Y., Lee, C. -H,, Batzer, A,, Vicentini, L. M., Zhou, M., Daly, R., Chem. 268,57486753

the necessary stimulatory amount of p21"'"GTP. Alternatively, Myers, M. J., Jr., Backer, J. M., Ullrich, A., White, M. F., and Schlessinger, since IRS-l.Grb2.Sos complex formation occurs more quickly, J. (1993) EMBO J. 12, 1929-1936

is Dossible that the same Grb2.Sos molecules which initially (1990) Pmc. Natl. Acad. Sei. U. S. A. 87. 792G7929 to a much lesser than Shc.Grb2'Sos it 10. Satoh, T., Endo, M., Nakafuku, M., Akiyama, T., Yamamoto, T., and Kaziro, Y.

9. Mulcahy, L. S., Smith, M. R., and Stacey, D. W. (1985) Nature 313,241-243

bind to IRS-1 subsequently associate with Shc. This might lead to the appropriate translocation or redistribution of Grb2Sos to facilitate stimulation of membrane bound p21"". Finally, it seems possible that IRS-1 participates in mitogenic signaling through mechanisms largely independent of p21"" stimulation. Thus, IRS-1 serves as a multisite docking protein binding to the SH2 domain of the p85 regulatory subunit of phosphatidyli- nositol (PI) 3-kinase (361, and inhibition of PI 3-kinase action will abrogate insulin's mitogenic effects. Thus, Yamauchi et al. (37) have shown that overexpression of the p85 subunit in Chinese hamster ovary cells inhibits insulin's ability to stimu- late c-Fos transcription, and we have found that microinjection of a glutathione S-transferase fusion protein containing the N-terminal SH2 domain of the p85 subunit markedly inhibits insulin stimulation of DNA synthesis? Perhaps PI 3-kinase activation and p21""-GTP formation are both necessary for cell cycle progression. Kapeller et al. (38) have shown that PI 3-ki- nase is closely localized to cytoskeletal structures, and is also a component of various endocytic vesicles. Based on studies in yeast, an important role for PI 3-kinase was identified in ve- sicular trafficking and cytoskeletal integrity (39). We hypoth- esize that, through IRS-1, activation of PI 3-kinase provides important signals to cytoskeletal architecture andor vesicular trafficking, which normally complement Shc-mediated p21"" activation to allow efficient ceil cycle progression.

Acknowledgments-We thank Dr. Hiroshi Maegawa for the IRS-1 antibody and Dr. Tahir S. Pillay for advice on the Western blotting studies. We are grateful to Elizabeth Martinez for assistance in the preparation of this manuscript.

REFERENCES 1. Ullrich, A., Bell, J. R., Chen, E. Y., Herrera, R., Petruzzelli, L. M., Dull, T. J.,

Gray, A,, Coussens, L., Liao. Y.-C., Tsubokawa, M., Mason, A,, Seeburg, F? H.. Grunfeld, C., Rosen, 0. M., and Ramachandran, J. (1985) Nature 313, 756761

2. White, M. F., Maron, R., and Kahn, C. R. (1985) Nature 318, 183-186 3. Sun, X. J., Rothenberg, P., Kahn, C. R., Backer, J. M., Araki, E., Wilden, P. A.,

4. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A,, Kapel- Cahill, D. A,, Goldstein, B. J., and White, M. F.(1991) Nature 352, 73-77

5. Koch, C . A,, Anderson, D., Moran, M. F., Ellis, C., and Pawson, T. (1991) ler, R., and Soltoff, S. (1991) Cell 64,281-302

6. Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F., Forni, G., Science 252, 668-674

Nicoletti, I., Grignani, F., Pawson, T., and Pelicci, P. G. (1992) Cell 70, 93-104

B. H. Jhun, B. L. Seely, A. R. Saltiel, and J. M. Olefsky, unpublished data.

, ~ , ~~~ . ~~ ~

11. Burgering, B. M. Th., Medema, R. H., Maassen, J. A,, Wetering, M. L., Eh, A.

12. Satoh, T., Nakafuku, M., and Kaziro, Y. (1992) J. B i d . Chem. 267, 24149-

13. Trahey, M., and McCormick, F. (1987) Science 238,542445 14. Simon, M. A., Bowtell, D. D. L., Dodson, G. S., Laverty, T. R., and Rubin, G. M.

15. Shou, C., Farnsworth, C. L., Neel, B. G., and Feig, L. A. (1992) Nature 568,

16. Chardin, P., Camonis, J . H., Gale, N. W., Aelst, L. V., Schlessinger, J., Wigler,

17. Draznin, B., Chang, L., Leitner, J . W., Takata, Y., and Olefsky, J. M. (1993) J.

18. Medema, R. H., Vries-Smits, A. M. M., Zon, G. C. M., Maassen, J. A,, and Bos,

20. Li, B.-Q., Subleski, M., Shalloway, D., Kung, H.-F., and Kamata, T. (1993)Pmc. 19. Buday, L., and Downward, J. (1993) Mol. Cell. B i d . 13, 1903-1910

21. Lowenstein, E. J., Daly, R. J., Batzer, A. G., Li, W., Margolis, B., Lammers, R., Natl. Acad. Sci. U. S. A. 90, 8504-8508

Ullrich, A., Skolnik, E. Y., Bar-Sagi, D., and Schlessinger, J. (1992) Cell 70, 431-442

22. Rozakis-Adcock, M., Fernley, R., Wade, J., Pawson, T., and Bowtell, D (1993) Nature 363, 83-85

23. Egan, S. E., Giddings, B. W., Brooks, M. W., Buday, L., Sizeland, A. M., and Weinberg, R. A. (1993) Nature 363,4541

24. Buday, L., and Downward, J. (1993) Cell 73,611-620 25. Li, N., Batzer, A,, Daly, R., Yajnik, V., Skolnik, E., Chardin, P., Bar-Sagi, D.,

26. Skolnik, E. Y., Batzer, A., Li, N., Lee, C.-H., Lawenstein, E., Mohammadi, M., Margolis, B., and Schlessinger, J . (1993) Nature 363,8588

27. Baltensperger, K., Kozma, L. M., Chemiack,A. D., Klarlund, J. K., Chawla,A., Margolis, B., and Schlessinger, J. (1993) Science 260, 1953-1955

28. Rose, D. W., Saltiel, A. R., Majumdar, M., Decker, S. J.. and Olefsky, J. M. Banerjee, U., and Czech, M. P. (1993) Science 260, 195CL1952

29. McClain, D. A,, Maegawa, H., Lee, J., Dull, T. J., Ullrich, A., and Olefsky, J. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 797-801

30. Rozakis-Adcock, M., McGlade, J., Mbamalu, G., Pelicci, G., Daly, R., Li, W., (1987) J. Biol. Chem. 262, 14663-14671

Batzer, A., Thomas, S., Brugge, J., Pelicci, P. G., Sehlessinger, J., and Paw- son, T. (1992) Nature 360, 689-692

31. Ravichandran, K. S., Lee., K. K., Songyang, Z., Cantley, L. C., Burn, P., and Burakoff, S. J. (1993) Science 262,902-905

32. Obermeier,A., Lammers, R., Wiesmuller, K.-H., Jung, G., Schlessinger, J., and Ullrich, A. (1993) J. Biol. Chem. 268,22963-22966

33. Cutler, R. L., Liu, L., Damen, J. E., and Krystal, G. (1993) J. Biol. Chern. 268, 21463-21465

J., McCormick, F., and Bos, J. L. (1991) EMBO J. 10, 1103-1109

24152

(1991) Cell 67, 701-716

351-354

M. H., and &v-Sagi, D. (1993) Science 260,1338-1343

Bid . Chem. 268, 1999&20001

J. L. (1993) Mol. Cell. Biol. 13, 155-162

34.

35.

36.

37.

38.

39.

Waters, S. B., Yamauchi, K., and Pessin, J. E. (1993) J. B i d . Chem. 268,

Wang, L.-M., Myers, M. G., Jr., Sun, X . J . , Aaronson, S. A., White, M. F., and 22231-22234

Backer, J. M., Myers, M. G., Jr., Shoelson, S. E., Chin, D. J., Sun, X.-J., Pierce, J. H. (1993) Science 261, 1591-1594

Miralpeix, M., Hu, P., Margolis, B., Skolnik, E. Y., Schlessinger, J., and White, M. F. (1992) EMBO J. 11, 3469-3479

Yamauchi, K., Holt, K., and Pessin, J. E. (1993) J. Biol. Chern. 268, 14597- 14600

Kapeller, R., Chakrabarti, R., Cantley, L., Fay, F., and Corvera, S. (1993) Mol. Cell. Biol. 13, 6052-6063

Hiles, I. D., Otsu, M., Volinia, S., Fry, M. J., Gout, I., Dhand, R., Panayotou. G., Ruiz-Larrea, E , Thompson, A., Totty, N. F., Hsuan, J. J., Courtneidge, S. A,, Parker, P. J., and Waterfield, M. D. (1992) Cell 70, 419-429

~~~~~ ~ ~ . "