Advertisement

Direct Association of Grb2 with the p85 Subunit of Phosphatidylinositol 3-Kinase(∗)

  • Jing Wang
    Affiliations
    Division of Cell and Molecular Biology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

    Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
    Search for articles by this author
  • Kurt R. Auger
    Affiliations
    Division of Cell and Molecular Biology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

    Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
    Search for articles by this author
  • Lesley Jarvis
    Affiliations
    Division of Cell and Molecular Biology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

    Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
    Search for articles by this author
  • Yang Shi
    Affiliations
    Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
    Search for articles by this author
  • Thomas M. Roberts
    Correspondence
    To whom correspondence should be addressed: Dept. of Pathology, Harvard Medical School, Dana-Farber Cancer Inst., M 857, 44 Binney Street, Boston, MA 02115. Tel.: 617-632-3049; Fax: 617-632-4770
    Affiliations
    Division of Cell and Molecular Biology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

    Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
    Search for articles by this author
  • Author Footnotes
    ∗ This work was supported by National Institutes of Health Grant CA30002 and a Sandoz Fellowship (to J. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:May 26, 1995DOI:https://doi.org/10.1074/jbc.270.21.12774
      Phosphatidylinositol 3-kinase (PI 3-kinase) has been shown to play a key role in growth factor signaling pathways, although its signaling mechanism has not been fully elucidated. Using the yeast interaction trap system, we have identified Grb2 as a PI 3-kinase interacting protein. Our experiments demonstrate that p85, the regulatory subunit of PI 3-kinase, interacts with Grb2 in vivo, and this interaction is independent of growth factor stimulation. The direct association between Grb2 and p85 was reconstituted in vitro with glutathione S-transferase fusion proteins. Domain analyses and peptide competition indicate that the association is mediated by the SH3 domains of Grb2 and the proline-rich motifs of p85 and that only one SH3 domain is required for minimal binding. The interaction does not displace the catalytic subunit of PI 3-kinase but is exclusive of Sos. Signaling through PI 3-kinase, therefore, may involve the ubiquitous adapter Grb2, which serves as a convergence point for multiple pathways.

      INTRODUCTION

      The discovery of phosphatidylinositol 3-kinase (PI 3-kinase)1(
      The abbreviations used are: PI 3-kinase
      phosphatidylinositol 3-kinase
      bcr
      break point region
      GST
      glutathione S-transferase
      PDGF
      platelet-derived growth factor
      PBS
      phosphate-buffered saline
      PAGE
      polyacrylamide gel electrophoresis
      SH
      Src homology domain
      x-gal
      5-bromo-4-chloro-3-indoyl β-D-galactoside.
      ) and its ubiquitous presence in species from yeast to human has unveiled a new pathway of phosphatidylinositol (PI) metabolism and intracellular signaling (
      • Whitman M.
      • Downes C.P.
      • Keeler M.
      • Keller T.
      • Cantley L.
      ,
      • Auger K.R.
      • Carpenter C.L.
      • Cantley L.C.
      • Varticovski L.
      ,
      • Parker P.J.
      • Waterfield M.D.
      ). Mammalian PI 3-kinase exists as a heterodimer of an 85-kDa (p85) regulatory subunit and a 110-kDa (p110) catalytic subunit (
      • Carpenter C.L.
      • Duckworth B.C.
      • Auger K.R.
      • Cohen B.
      • Schaffhausen B.S.
      • Cantley L.C.
      ,
      • Otsu M.
      • Hiles I.
      • Gout I.
      • Fry M.J.
      • Ruiz-Larrea F.
      • Panayotou G.
      • Thompson A.
      • Dhand R.
      • Hsuan J.
      • Totty N.
      • Smith A.D.
      • Morgan S.J.
      • Courtneidge S.A.
      • Parker P.J.
      • Waterfield M.D.
      ,
      • Hiles I.D.
      • Otsu M.
      • Volinia S.
      • Fry M.J.
      • Gout I.
      • Dhand R.
      • Panayotou G.
      • Ruiz-Larrea F.
      • Thompson A.
      • Totty N.F.
      • Hsuan J.J.
      • Courtneidge S.A.
      • Parker P.J.
      • Waterfield M.D.
      ). The structure of the p85 protein is comprised of an SH3 domain, a break point region (bcr) homology domain flanked by two proline-rich sequences (
      • Kapeller R.
      • Prasad K.V.
      • Janssen O.
      • Hou W.
      • Schaffhausen B.S.
      • Rudd C.E.
      • Cantley L.C.
      ), and two SH2 domains flanking a p110-binding region (
      • Otsu M.
      • Hiles I.
      • Gout I.
      • Fry M.J.
      • Ruiz-Larrea F.
      • Panayotou G.
      • Thompson A.
      • Dhand R.
      • Hsuan J.
      • Totty N.
      • Smith A.D.
      • Morgan S.J.
      • Courtneidge S.A.
      • Parker P.J.
      • Waterfield M.D.
      ,
      • Hiles I.D.
      • Otsu M.
      • Volinia S.
      • Fry M.J.
      • Gout I.
      • Dhand R.
      • Panayotou G.
      • Ruiz-Larrea F.
      • Thompson A.
      • Totty N.F.
      • Hsuan J.J.
      • Courtneidge S.A.
      • Parker P.J.
      • Waterfield M.D.
      ). Initially studied in pp60v-src immunoprecipitates from transformed cells (
      • Sugimoto Y.
      • Whitman M.
      • Cantley L.C.
      • Erikson R.L.
      ), PI 3-kinase phosphorylates PI at the D-3 position of the inositol ring (
      • Whitman M.
      • Downes C.P.
      • Keeler M.
      • Keller T.
      • Cantley L.
      ). PI 3-kinase was also found to co-immunoprecipitate with middle T antigen (
      • Whitman M.
      • Kaplan D.R.
      • Schaffhausen B.
      • Cantley L.
      • Roberts T.M.
      ,
      • Kaplan D.R.
      • Whitman M.
      • Schaffhausen B.
      • Raptis L.
      • Garcea R.L.
      • Pallas D.
      • Roberts T.M.
      • Cantley L.
      ) and a variety of protein-tyrosine kinase receptors, including PDGF receptor (
      • Kaplan D.R.
      • Whitman M.
      • Schaffhausen B.
      • Pallas D.C.
      • White M.
      • Cantley L.
      • Roberts T.M.
      ,
      • Coughlin S.R.
      • Escobedo J.A.
      • Williams L.T.
      ), colony-stimulating factor 1 receptor (
      • Varticovski L.
      • Druker B.
      • Morrison D.
      • Cantley L.
      • Roberts T.
      ), and insulin receptor (
      • Ruderman N.B.
      • Kapeller R.
      • White M.F.
      • Cantley L.C.
      ). Moreover, two of the products of PI 3-kinase, PI-3,4-P2 and PI-3,4,5-P3, are elevated in cells activated by growth factors (
      • Auger K.R.
      • Serunian L.A.
      • Soltoff S.P.
      • Libby P.
      • Cantley L.C.
      ) or transformed by oncogenes (
      • Serunian L.A.
      • Auger K.R.
      • Roberts T.M.
      • Cantley L.C.
      ,
      • Fukui Y.
      • Saltiel A.R.
      • Hanafusa H.
      ).
      Three lines of evidence indicate that PI 3-kinase plays an important role in growth regulation and transformation. The first line of evidence comes from genetic analyses of the binding sites for PI 3-kinase on oncogenes and receptors. For instance, all mutants of polyoma virus middle T antigen, which either fail to associate with PI 3-kinase or are unable to elevate the levels of PI 3-kinase products in vivo, result in a transformation-defective phenotype (
      • Kaplan D.R.
      • Whitman M.
      • Schaffhausen B.
      • Raptis L.
      • Garcea R.L.
      • Pallas D.
      • Roberts T.M.
      • Cantley L.
      ,
      • Courtneidge S.A.
      • Heber A.
      ,
      • Pallas D.C.
      • Cherington V.
      • Morgan W.
      • DeAnda J.
      • Kaplan D.
      • Schaffhausen B.
      • Roberts T.M.
      ,
      • Ling L.E.
      • Druker B.J.
      • Cantley L.C.
      • Roberts T.M.
      ). Similarly, point mutations in the PI 3-kinase binding sites of the PDGF receptor impair the receptor's ability to initiate DNA synthesis (
      • Fantl W.J.
      • Escobedo J.A.
      • Martin G.A.
      • Turck C.W.
      • del Rosario M.
      • McCormick F.
      • Williams L.T.
      ,
      • Kazlauskas A.
      • Kashishian A.
      • Cooper J.A.
      • Valius M.
      ,
      • Valius M.
      • Kazlauskas A.
      ). Second, recent work by Roche et al.(
      • Roche S.
      • Koegl M.
      • Courtneidge S.A.
      ) has shown that microinjection of antibodies specific for the p110 subunit of the PI 3-kinase into quiescent fibroblasts inhibited PDGF-induced DNA synthesis. Finally, studies indicate that inhibition of PI 3-kinase activity by wortmannin, the specific PI 3-kinase inhibitor, results in blockage of serum-induced cell proliferation (
      • Vemuri G.S.
      • Rittenhouse S.E.
      ).
      The precise mechanisms by which PI 3-kinase regulates cell growth and transformation are not well defined. Evidence exists to link PI 3-kinase to the regulatory cascade that controls pp70S6k, but the actual mechanism is unknown (
      • Chung J.
      • Grammer T.C.
      • Lemon K.P.
      • Kazlauskas A.
      • Blenis J.
      ,
      • Monfar M.
      • Lemon K.P.
      • Grammer T.C.
      • Cheatham L.
      • Chung J.
      • Vlahos C.J.
      • Blenis J.
      ). The recent revelation that Ras (p21ras) interacts directly with the catalytic subunit of PI 3-kinase (p110 subunit) raises the possibility that PI 3-kinase may serve as an effector of Ras (
      • Rodriguez V.P.
      • Warne P.H.
      • Dhand R.
      • Vanhaesebroeck B.
      • Gout I.
      • Fry M.J.
      • Waterfield M.D.
      • Downward J.
      ), but again the details of the process are yet to be elucidated. Finally, other studies indicate a link between PI 3-kinase and actin polymerization and depolymerization in vivo(
      • Eberle M.
      • Traynor-Kaplan A.
      • Sklar L.A.
      • Norgauer J.
      ,
      • Severinsson L.
      • Ek B.
      • Mellstrom K.
      • Claesson W.L.
      • Heldin C.H.
      ) and more recently between PI 3-kinase and the small G proteins Rac and Rho, suggesting that the enzyme may also play a role in regulating cell shape (
      • Zhang J.
      • King W.G.
      • Dillon S.
      • Hall A.
      • Feig L.
      • Rittenhouse S.E.
      ,
      • Zheng Y.
      • Bagrodia S.
      • Cerione R.A.
      ).
      Additional information concerning the proteins that interact with PI 3-kinase is needed to define the biochemical interactions of PI 3-kinase that are involved in cell proliferation and cell shape. To this end, we constructed a bait with full-length human p85 and used it to search for p85 interactors in the yeast interaction trap system (
      • Zervos A.S.
      • Gyuris J.
      • Brent R.
      ). Here we report that an interactor for the 85-kDa subunit of PI 3-kinase is Grb2. Comprised entirely of SH2 and SH3 domains, Grb2 is a small adapter protein that binds phosphotyrosine motifs on activated receptors via the SH2 domain while the two flanking SH3 domains are used to bind the guanine nucleotide exchange protein Sos. In this manner, Grb2 links the activation of receptors to the GTP loading of p21ras(
      • Lowenstein E.J.
      • Daly R.J.
      • Batzer A.G.
      • Li W.
      • Margolis B.
      • Lammers R.
      • Ullrich A.
      • Skolnik E.Y.
      • Bar-Sagi D.
      • Schlessinger J.
      ). We have also confirmed that the interaction between Grb2 and PI 3-kinase also occurs in mammalian cells. Genetic and biochemical analyses indicate that the SH3 domains of Grb2 interact with the proline-rich motifs of p85. This interaction does not displace the catalytic subunit (p110) of PI 3-kinase but is exclusive of Sos.

      EXPERIMENTAL PROCEDURES

      Antibodies

      The mouse anti-p85 antibody and the mouse anti-Grb2 antibody used for immunoblotting were from Transduction Labs (Lexington, KY). The rabbit anti-p85 antibody used for immunoblotting and immunoprecipitation was made against glutathione S-transferase (GST) fusion protein with full-length p85.2(
      K. R. Auger, L. Jarvis, and T. M. Roberts, unpublished results.
      ) The rabbit anti-Grb2 antibody for immunoprecipitation was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The rabbit anti-mouse Sos1 was from Upstate Biotechnology Inc. (Lake Placid, NY). The peroxidase-conjugated goat anti-mouse IgG (Fcγ-specific) was from Jackson ImmunoResearch Labs, Inc. (West Grove, PA), and the alkaline phosphatase-conjugated goat anti-rabbit IgG and goat anti-mouse IgG were from Promega (New York, NY). The rabbit anti-LexA antibody was kindly provided by Russ Finley (Massachusetts General Hospital, Boston, MA).

      Library Screening through Yeast Two-hybrid System

      Manipulations of Escherichia coli and yeast were performed essentially as described (
      • Ausubel F.M.
      • Brent R.
      • Kingston R.
      • Moore D.
      • Seidman J.J.
      • Smith J.
      • Struhl K.
      ). E. coli K-12 strain KC8 pyrF::Tn5,hsdR,leuB600,trpC9830,lacD74,strA,galK,hisB436 was used for the rescue of yeast plasmids as described (
      • Zervos A.S.
      • Gyuris J.
      • Brent R.
      ). EGY48 MATa trp1,ura3,his3,LEU2::pLexAop6-LEU2 was used as a host for all interaction experiments (
      • Zervos A.S.
      • Gyuris J.
      • Brent R.
      ). Human PI 3-kinase p85 cDNA (
      • Skolnik E.Y.
      • Margolis B.
      • Mohammadi M.
      • Lowenstein E.
      • Fischer R.
      • Drepps A.
      • Ullrich A.
      • Schlessinger J.
      ), a gift from Dr. Lewis Cantley (Harvard Medical School), was linearized with BamHI, ligated to a BamHI to EcoRI linker, and then excised by EcoRI digestion. The EcoRI fragment containing the complete p85 cDNA was inserted into pEG202 at the EcoRI site, and the generated plasmid pEG-h85 was used as the bait. The oligo-primed HeLa cDNA yeast expression library was a generous gift from Dr. Russ Finley (Massachusetts General Hospital) and was screened essentially as described (
      • Zervos A.S.
      • Gyuris J.
      • Brent R.
      ).

      X-Gal Filter Assay on Yeast Colonies

      Yeast colonies freshly grown on HisUraTrp plates (glucose or galactose as energy source) were lifted onto nitrocellulose membranes and lysed by submerging in liquid nitrogen for 30 s to 1 min. The membranes were then placed gently on Whatman filter paper saturated with 3 ml of Z buffer (100 mM sodium phosphate, pH 7.0, 10 mM KCl, 1 mM MgSO4, 40 mM β-mercaptoethanol) containing 1 mg/ml X-gal, and the color of colonies was recorded through the course of 30 min.

      Cell Culture and Preparation of Cell Lysates

      NIH 3T3, Balb/c 3T3, and bovine kidney cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% calf serum. Cells from a confluent 100-mm culture dish were rinsed once in phosphate-buffered saline (PBS) and then lysed for 15 min in 1 ml of Nonidet P-40 lysis buffer (135 mM NaCl, 20 mM HEPES, 5 mM EDTA, 1% Nonidet P-40, 10% glycerol, 500 μM sodium orthovanadate) containing protease inhibitors (aprotinin (1 μg/ml), pepstatin (1 μg/ml), and leupeptin (0.75 μg/ml)). Lysates were scraped and cleared at 13,000 × g. Clarified supernatants were used fresh or stored at −80°C.

      PDGF Stimulation and Immunoprecipitation

      Cells 1-2 days postconfluence were switched to Dulbecco's modified Eagle's medium containing 0.2% calf serum for an overnight incubation. PDGF was added to a final concentration of 30 ng/ml for 15 min. The cells were then washed in PBS, and lysates were prepared as described above. Lysates were incubated with rocking at 4°C for 2 h with appropriate antibodies and then for 30 min with protein A-Sepharose (Pharmacia Biotech Inc.). Immune complexes were washed four times with PBS, 1% Nonidet P-40, 1 mM EDTA and two times with TNE (10 mM Tris, pH 8.0, 100 mM NaCl, 1 mM EDTA). They were resuspended in 40 μl of 1 × electrophoresis sample buffer, boiled for 5 min, and analyzed by SDS-PAGE.

      Immunoblot Analysis

      Samples were resolved by SDS-PAGE, transferred onto nitrocellulose filters (Schleicher & Schuell), and incubated with primary antibodies at the concentrations recommended by the manufacturer for 1-2 h at room temperature. Immunoblots were subsequently washed and incubated with corresponding secondary antibodies for 1 h at room temperature, washed three times in PBS-T (phosphate-buffered saline, 0.2% Tween 20), and washed one time in attophos buffer (50 mM Tris, pH 9.5, 100 mM NaCl, 0.1 mM MgCl2). Filters were then developed for 10 min in attophos buffer containing attophos substrate at 1:20 dilution and subjected to analysis by Fluorimager (Molecular Dynamics, Inc., Sunnyvale, CA). The data were transferred electronically to a Macintosh computer and printed with a Fujix Pictography 3000 (Fuji Photo Film USA, Inc., Elmsford, NY).

      Preparation of GST Fusion Proteins and Thrombin Cleavage

      A single colony of HB101 transformed with the plasmid of interest was grown overnight in 50 ml of LB containing 50 μg/ml ampicillin. Harvesting and purification of the fusion proteins by affinity to glutathione-Sepharose (Pharmacia) was carried out essentially as described by Smith and Johnson (
      • Smith D.B.
      • Johnson K.S.
      ).
      An aliquot of GST-beads was washed in PBS once and resuspended in 100 μl of PBS or 20 mM HEPES (pH 7.4). Human thrombin was added at a concentration of 0.2-0.4 units/μg protein and the reaction was incubated at room temperature for 15 min. PMSF was added to a final concentration of 1 mM and the mixture was incubated for an additional 5 min. The supernatant was separated from the beads in a microcentrifuge and stored at −80°C or used immediately.

      PI 3-kinase Assay

      PI 3-kinase assays were performed essentially as described (
      • Auger K.R.
      • Serunian L.A.
      • Soltoff S.P.
      • Libby P.
      • Cantley L.C.
      ). Briefly, immunoprecipitates were prepared as described above. Following the final wash, sonicated lipid substrates were added to the supernatants at a final concentration of 0.2 mg/ml, and the reaction was initiated by the addition of 5 mM MgCl2 and 100 μM [γ-32P]ATP at 10 μCi/reaction in 20 mM HEPES (pH 7.2). The reaction was incubated at room temperature for 5 min and stopped by extraction with 75 μl of 1 M HCl and 180 μl of methanol:chloroform (1:1). The organic phase was collected and stored at −80°C or immediately analyzed by thin layer chromatography (TLC) in an n-propyl alcohol, 2 M acetic acid solvent system (65:35).

      Protein Kinase Assay

      Immunoprecipitates were prepared as described above. Following the final wash, 45 μl of ATP mix (20 mM HEPES, 10 mM MnCl2, 5 μM ATP, 20 μCi of [γ-32P]ATP) was added to the beads, and the mixtures were incubated at room temperature for 20 min. The reaction was stopped by adding 40 μl of sample buffer, boiled for 5 min and analyzed by SDS-PAGE.

      RESULTS

      Association in Yeast Two-hybrid System

      We screened for p85 interactors from a HeLa cell cDNA library by the interaction trap technique developed by Zervos et al.(
      • Zervos A.S.
      • Gyuris J.
      • Brent R.
      ). Fig. 1A shows that the yeast host cells EGY48, carrying the pEG-h85 bait plasmid, correctly express the full-length human p85 fused to the LexA-binding domain (lane 3). We plated 1.7 × 106 primary library transformants onto Leu galactose selection plates. Of the colonies that grew, those which gave unambiguous, galactose-dependent blue color on X-gal filter assay and galactose-dependent growth on Leu selection plates were rescued for sequence analysis. Of twenty clones sequenced, the majority represent previously uncharacterized proteins. However, one of these clones, p1-74, was identical to the cDNA sequence of the adapter protein Grb2. As shown in Fig. 1B, p1-74 interacts specifically with the full-length p85 bait but not with the control baits of human p110 or the bcr domain of human p85. The same pattern was observed with the X-gal filter assay (Table I). As indicated in Fig. 2, p1-74 contains the C-terminal SH3 domain and the majority of the SH2 domain of Grb2.
      Figure thumbnail gr1
      Figure 1:Expression and interaction of human p85 with Grb2 in yeast. A, yeast cells EGY48 carrying different plasmids were lysed, and 100 μg of total protein was resolved by SDS-PAGE. Proteins were transferred to a nitrocellulose membrane and blotted with anti-LexA polyclonal antibody. Lane 1, EGY48 cells alone; lane 2, EGY48 carrying pEG202 and pSH18-34; lane 3, EGY48 carrying pEG-h85 and pSH18-34. B, colonies containing different combinations of baits and preys were streaked onto both glucose- and galactose-based HisUraTrpLeu plates and incubated at 30°C for 2 days. 1, 3, 5, 7, 9, and 11 represent human p85 bait with a panel of different plasmids as preys, including pJG4-5, negative control 1 and 2, p1-74, and positive control 1 and 2, respectively; 2, 4, 6, 8, 10, and 12 represent p110 bait with a panel of preys, including pJG4-5, negative control 1 and 2, p1-74, and positive control 1 and 2, respectively; 13, 14, 15, and 16 represent h85bcr bait with negative control 1 and 2, p1-74, and positive control 1, respectively.
      Table I:Human p85 specifically interacts with Grb2 in yeast
      Figure thumbnail gr2
      Figure 2:Complete nucleotide and peptide sequence of human Grb2 (
      • Lowenstein E.J.
      • Daly R.J.
      • Batzer A.G.
      • Li W.
      • Margolis B.
      • Lammers R.
      • Ullrich A.
      • Skolnik E.Y.
      • Bar-Sagi D.
      • Schlessinger J.
      ). The SH2 and SH3 domains are underlined in boldface. The sequence underlined in lightface represents clone p1-74.

      Direct Association of p85 to Grb2 in Vitro Mediated by SH3 and Proline-rich Domain

      To determine whether p85 and Grb2 can interact directly, we examined complex formation in vitro with bacterially produced p85 and Grb2. We found that soluble p85 released from GST-p85 fusion protein by thrombin cleavage was able to bind to GST-Grb2 immobilized on glutathione-Sepharose beads but not to the GST-2T control protein (Fig. 3A). Similarly, thrombin-cleaved Grb2 forms a complex with GST-p85 immobilized on beads but not with GST-2T (Fig. 3B).
      Figure thumbnail gr3
      Figure 3:Direct association of GST-85 and GST-Grb2 in vitro. A, soluble p85 was prepared by thrombin cleavage as described under “Experimental Procedures.” Approximately 5 μg of soluble p85 was incubated with GST-2T (lane 1) or GST-Grb2 (lane 2) beads for 30 min at 4°C. The complexes were washed and separated by SDS-PAGE and immunoblotted with anti-p85 monoclonal antibody. B, soluble Grb2 was prepared by thrombin cleavage, and approximately 5 μg of soluble Grb2 was incubated with GST-2T (lane 1) or GST-Grb2 beads for 30 min at 4°C. The complexes were washed and analyzed by SDS-PAGE and immunoblotted with anti-Grb2 antibody.
      To examine the molecular basis for the association, we prepared GST-fusion proteins that contained the N-terminal SH3, C-terminal SH3, and SH2 domain of Grb2 and the bcr, bcr with the flanking proline-rich sequences (bcrP), and SH3 domain of p85. When different fragments of p85 were examined for their ability to bind soluble Grb2 thrombin-cleaved from GST-Grb2 beads, only the bcrP domain was able to bind Grb2 (Fig. 4A), suggesting that the proline-rich sequences are involved in the association. We synthesized the two proline-rich peptides corresponding to the two proline-rich motifs in p85 (amino acids 83-100 and 313-329, respectively) (
      • Kapeller R.
      • Prasad K.V.
      • Janssen O.
      • Hou W.
      • Schaffhausen B.S.
      • Rudd C.E.
      • Cantley L.C.
      ). The effects of these peptides on the binding of the bcrP fragment to Grb2 were investigated (Fig. 4B). At 60 × molar excess, either peptide partially inhibited the binding of bcrP to Grb2. At 400 × molar excess, they blocked the binding to near completion. An additive effect was observed when both peptides were used in combination.
      Figure thumbnail gr4
      Figure 4:Domain analysis for p85 and Grb2 association. A, soluble Grb2 generated by thrombin cleavage was incubated with full-length (GST-p85, lane 1), bcr region (GST-bcr, lane 2), bcrP region (GST-bcrP, lane 3), and the SH3 domain (GST-SH3, lane 4) of p85. The complexes were separated by SDS-PAGE and immunoblotted with anti-Grb2 antibody. B, soluble Grb2 generated by thrombin cleavage was incubated with GST-bcrP beads alone (lane 1), or bcrP beads plus 400 × molar excess (lanes 2-4) of N-terminal (lane 2), C-terminal (lane 3), N- and C-terminal proline-rich peptides (lane 4), or bcrP beads plus 60 × molar excess (lanes 5-7) of N-terminal (lane 5), C-terminal (lane 6), N- and C-terminal proline-rich peptides (lane 7). The complexes were resolved by SDS-PAGE and immunoblotted with anti-Grb2 antibody. C, soluble p85 was incubated with GST-2T (lane 1), GST-Grb2 (lane 2), the N-terminal SH3 domain (lane 3), the SH2 domain (lane 4), the C-terminal SH3 domain (lane 5), or both N- and C-terminal SH3 domains of Grb2 (lane 6). The complexes were resolved by SDS-PAGE and immunoblotted with anti-p85 monoclonal antibody. D, soluble p85 generated by thrombin cleavage was incubated with GST-2T (lane 1), GST-SH2 of Grb2 (lane 2), GST-Grb2 (lane 3), or GST-W36K/W193K (K36/193) (lane 4). The complexes were analyzed by SDS-PAGE and immunoblotted with anti-p85 monoclonal antibody.
      Next, we examined the three domains of Grb2 for their ability to bind p85. Each domain of Grb2 was expressed as a GST fusion protein. As shown in Fig. 4, C and D, both N- and C-terminal SH3 domain but not the SH2 domain of Grb2 independently bound to p85. A double mutant of Grb2 (W36K/W193K), which contains point mutations in each SH3 domain, lost its ability to bind p85. Taken together, this evidence indicates that the association of p85 and Grb2 is mediated by the proline-rich sequences of p85 and the SH3 domains of Grb2.

      Association of p85 and Grb2 in Vivo

      To test whether complex formation between p85 and Grb2 also occurs in mammalian cells, we examined whether these two proteins co-immunoprecipitate. Cell lysates were prepared from one to two day postconfluent NIH 3T3 cells and subjected to immunoprecipitation with anti-p85, anti-Grb2, and control antibody. As shown in Fig. 5A, p85 can be precipitated with anti-Grb2 but not the control antibody. Similarly, Grb2 can be precipitated with anti-p85 antibody but not the preimmune serum (Fig. 5B). Identical results were obtained from Balb/c 3T3 and Bovine kidney cells (results not shown). These results provide the first evidence for an in vivo association between the PI 3-kinase p85 subunit and the adapter protein Grb2 in unstimulated mammalian cells.
      Figure thumbnail gr5
      Figure 5:Association of p85 with Grb2 in quiescent NIH 3T3 cells. A, immunoprecipitates were prepared from confluent NIH 3T3 cells using control serum (lane 1) or anti-Grb2 antibody (lane 2). The samples were washed four times with 1% Nonidet P-40, PBS, 1 mM EDTA and two times with 10 mM Tris, pH 8.0, 100 mM NaCl, 1 mM EDTA and were subjected to SDS-PAGE. Immunoblotting was performed with anti-p85 antibody. B, immunoprecipitates were prepared from confluent NIH 3T3 cells using preimmune serum (lane 1) or anti-p85 antibody. The samples were separated by SDS-PAGE and immunoblotted with anti-Grb2 antibody. C, immunoprecipitates were prepared from fresh lysates of confluent NIH 3T3 cells using anti-p85 antibody (lane 1), p85 preimmune serum (lane 2), or anti-Grb2 antibody (lane 3). Protein kinase assay was performed as described under “Experimental Procedures,” and the samples were boiled in the presence of 0.5% SDS and 10 mM β-mercaptoethanol for 5 min. The supernatants were diluted with lysis buffer and immunoprecipitated with anti-p85 antibody.
      Recent findings demonstrated that PI 3-kinase is a dual specificity kinase, phosphorylating both lipids and protein serine residues (
      • Carpenter C.L.
      • Auger K.R.
      • Duckworth B.C.
      • Hou W.M.
      • Schaffhausen B.
      • Cantley L.C.
      ,
      • Dhand R.
      • Hiles I.
      • Panayotou G.
      • Roche S.
      • Fry M.J.
      • Gout I.
      • Totty N.F.
      • Truong O.
      • Vicendo P.
      • Yonezawa K.
      • Kasuga M.
      • Courtneidge S.A.
      • Waterfield M.D.
      ). This enabled us to confirm the above interaction with a more sensitive method. Immunoprecipitates using anti-p85 antibody, p85 preimmune serum, and anti-Grb2 antibody were prepared from fresh NIH 3T3 cell lysates and were subjected to Mn2+-dependent protein kinase assay. The reactions were boiled, diluted, and subjected to an additional immunoprecipitation with anti-p85 antibody. As shown in Fig. 5C, both anti-p85 and anti-Grb2 antibodies but not the control serum precipitated p85 from the lysates.

      The Interaction Is Independent of PDGF Stimulation

      Since both Grb2 and p85 can be found in PDGF receptor immunoprecipitates, it is possible that PDGF receptor might be involved in bringing the two proteins together. The fact that the association occurs in quiescent cells suggest a receptor-independent interaction. Our in vitro data also support a direct binding mechanism. In an additional experiment to test this hypothesis, the effect of PDGF stimulation on the association was investigated. As shown in Fig. 6A, while PDGF stimulation of NIH 3T3 cells significantly increased the level of anti-phosphotyrosine immunoprecipitated p85, it did not appear to increase the amount of p85 brought down by anti-Grb2 antibody. The amount of Grb2 in anti-phosphotyrosine immunoprecipitates also increased significantly after PDGF treatment, whereas it remained essentially constant in anti-p85 immunoprecipitates (Fig. 6B). Taken together, these results indicate that PDGF receptor is not involved in mediating the interaction.
      Figure thumbnail gr6
      Figure 6:Association of p85 and Grb2 is independent of PDGF stimulation. A, confluent NIH 3T3 cells 1-2 days postconfluence were switched to 10 ml of Dulbecco's modified Eagle's medium containing 0.2% calf serum overnight. The cells were then incubated for 15 min in the presence or absence of PDGF at a final concentration of 30 ng/ml. Immunoprecipitates were prepared from the lysates using anti-Grb2 antibody (lanes 1 and 2) or 4G10 anti-phosphotyrosine antibody (lanes 3 and 4). Immunoblotting was performed with anti-p85 antibody. B, confluent NIH 3T3 cells were stimulated with PDGF as described in panel A. Immunoprecipitates were prepared from the lysates using anti-p85 antibody (lanes 1 and 2) or 4G10 antibody (lanes 3 and 4). Immunoblotting was performed with anti-Grb2 antibody.

      The Catalytic Subunit of PI 3-kinase Remains Bound to p85 after Its Association with Grb2

      The nature of the p85-Grb2 complex was examined further by testing whether p110 was present. Fig. 7A shows the result of PI 3-kinase assay performed on anti-Grb2 immunoprecipitates. A significant amount of PI 3-kinase activity was precipitated by both anti-p85 and anti-Grb2 antibodies (lanes 1 and 3), compared with the preimmune serum, which had background levels of phosphatidylinositol 3-phosphate and phosphatidylinositol 3,4,5-trisphosphate (lane 2). A variation of the experiment is shown in Fig. 7B where different GST-fusion proteins were examined for their ability to bring down PI-3 kinase activity from cell lysates. Compared with the vector control and the SH2 domain of Grb2, which had background level of PI 3-kinase activity, the full-length Grb2 fusion was able to precipitate a significant amount of activity. The double mutant W36K/W193K was comparable with the vector control. These results indicate that the catalytic subunit p110 is present in the same complex with p85 and Grb2.
      Figure thumbnail gr7
      Figure 7:The catalytic subunit p110 of PI 3-kinase co-exists in the complex of p85 and Grb2. A, immunoprecipitates were prepared from confluent NIH 3T3 cells using anti-p85, preimmune serum, or anti-Grb2 antibody (lanes 1, 2, 3) and washed four times with 1% Nonidet P-40, PBS, 1 mM EDTA and two times with 10 mM Tris, pH 8.0, 100 mM NaCl, 1 mM EDTA. The beads were subjected to PI 3-kinase assay as described under “Experimental Procedures” and analyzed on a TLC plate. B, GST-fusion constructs, each containing 10 μg of protein, were incubated with fresh lysates prepared from confluent NIH 3T3 cells for 1 h at 4°C. The beads were washed as in panel A and subjected to PI 3-kinase analysis. Lane 1, GST-2T; lane 2, GST-Grb-SH2; lane 3, GST-W36K/W193K (K36/193); lane 4, GST-Grb2. PIP and PIP3, phosphatidylinositol 3-phosphate and phosphatidylinositol 3,4,5-trisphosphate, respectively.

      p85 and Sos Do Not Co-exist in a Single Complex with Grb2

      We also attempted to test whether Sos and p85 could co-exist in the complex with Grb2. We were unable to detect Sos in anti-p85 immunoprecipitates. We next performed experiments using GST fusion proteins. Fig. 8 shows that only GST-Grb2 (lanes 1, 2, and 3), but not GST-bcrP (lanes 4, 5, and 6) or the vector control (lanes 7, 8, and 9), binds Sos from NIH 3T3 cells. Furthermore, thrombin-cleaved soluble bcrP was not able to displace Sos from binding Grb2 (lane 1). Similarly, while soluble Grb2 was brought down by GST-bcrP (lane 4), there was no detectable Sos in the precipitates, indicating that Sos could not displace bcrP from binding Grb2.
      Figure thumbnail gr8
      Figure 8:p85 and Sos do not co-exist in a single complex with Grb2. Lysates prepared from confluent NIH 3T3 cells were incubated for 1 h at 4°C with different combinations of GST fusion proteins on beads or in thrombin-released soluble form. GST-Grb2 beads were mixed with soluble bcrP (lane 1), 2T control (lane 2), or by themselves (lane 3); GST-bcrP beads were mixed with soluble Grb2 (lane 4), 2T control (lane 5), or by themselves (lane 6); GST-2T beads were mixed with soluble bcrP (lane 7), Grb2 (lane 8), or by themselves (lane 9). The complexes were resolved on SDS-PAGE and immunoblotted with anti-Sos antibody (upper panel) and stripped and reblotted with anti-Grb2 antibody (lower panel).

      DISCUSSION

      We have presented several lines of evidence that the p85 subunit of PI 3-kinase interacts directly with the adapter protein Grb2 both in vitro and in vivo. The association was first revealed in the yeast interaction trap system. The association between p85 and Grb2 was also observed in quiescent NIH 3T3 cells and Balb/c 3T3 cells on reciprocal immunoprecipitation using anti-p85 and anti-Grb2 antibodies, independent of PDGF stimulation. Therefore, at least most of the interaction we observed was not mediated by the activated receptor. The interaction appears to be direct since it can be reconstituted in vitro using purified bacterially produced proteins. Domain analyses demonstrate that the association is mediated by the SH3 domains of Grb2 and the proline-rich motifs flanking the bcr homology region of p85.
      The most parsimonious model to explain our results places both subunits of PI 3-kinase in a position equivalent to Sos in Grb2 signaling. Although the interaction of Grb2 is with the 85-kDa subunit, PI 3-kinase assays performed on anti-Grb2 immunoprecipitates indicate that p110 remains bound to p85 in the complex. While only a small fraction of PI 3-kinase binds Grb2 and vice versa, the interaction presents intriguing possibilities for linking PI 3-kinase to receptors that lack the ability to bind PI 3-kinase directly. For instance, recent work by Welham et al.(
      • Welham M.J.
      • Dechert U.
      • Leslie K.B.
      • Jirik F.
      • Schrader J.W.
      ) has suggested that SHPTP2 may play an important role in integrating signals from interleukin-3 and GM-CSF receptors to PI 3-kinase. Our data may provide a missing link between SHPTP2 and PI 3-kinase. Previous studies have shown that Grb2 associates with tyrosine-phosphorylated SHPTP2 via its SH2 domain (
      • Li W.
      • Nishimura R.
      • Kashishian A.
      • Batzer A.G.
      • Kim W.J.
      • Cooper J.A.
      • Schlessinger J.
      ,
      • Bennett A.M.
      • Tang T.L.
      • Sugimoto S.
      • Walsh C.T.
      • Neel B.G.
      ). Since Grb2 also interacts with PI 3-kinase p85 subunit through its SH3 domains, it serves as an excellent candidate connecting the two pathways. Analyses are currently under way to explore this possibility. In addition to SHPTP2, Shc (
      • Rozakis A.M.
      • McGlade J.
      • Mbamalu G.
      • Pelicci G.
      • Daly R.
      • Li W.
      • Batzer A.
      • Thomas S.
      • Brugge J.
      • Pelicci P.G.
      • Schlessinger J.
      • Pawson T.
      ,
      • Skolnik E.Y.
      • Batzer A.
      • Li N.
      • Lee C.H.
      • Lowenstein E.
      • Mohammadi M.
      • Margolis B.
      • Schlessinger J.
      ) and IRS-1 (
      • Myers M.J.
      • Wang L.M.
      • Sun X.J.
      • Zhang Y.
      • Yenush L.
      • Schlessinger J.
      • Pierce J.H.
      • White M.F.
      ) have been shown to interact with the SH2 domain of Grb2, as have non-receptor tyrosine kinases, including Abl (
      • Ren C.L.
      • Morio T.
      • Fu S.M.
      • Geha R.S.
      ), Bcr-Abl (
      • Pendergast A.M.
      • Quilliam L.A.
      • Cripe L.D.
      • Bassing C.H.
      • Dai Z.
      • Li N.
      • Batzer A.
      • Rabun K.M.
      • Der C.J.
      • Schlessinger J.
      • Gishizky M.L.
      ), and FAK (
      • Schlaepfer D.D.
      • Hanks S.K.
      • Hunter T.
      • van der Geer P.
      ). Alternatively, the interaction between Grb2 and PI 3-kinase might use p85 to present Grb2 to activated receptors that otherwise lack access to the adapter protein. Thus, the Grb2-PI 3-kinase interaction has the potential to play a role in a variety of signaling processes.
      The SH3 domain-mediated interaction provides a novel element for linking PI 3-kinase to molecules that otherwise lack access to the enzyme. The fact that only one SH3 domain of Grb2 is required for minimal binding in vitro raises the interesting possibility that Grb2 could use its free SH3 domain to bind other factors. However, our data suggest that Sos is not bound in the same complex with PI 3-kinase. The possibility remains that another molecule could occupy the second SH3 domain and thus would increase the signaling potential of the Grb2/PI 3-kinase system. Further work will need to be devoted to study the functionality of the interaction and to testing how the interaction results in specific physiological changes for the cell.

      Acknowledgments

      We thank Tiffany L. Holcombe for technical assistance and Dr. David R. Kaplan for helpful discussion during preparation of the manuscript.

      REFERENCES

        • Whitman M.
        • Downes C.P.
        • Keeler M.
        • Keller T.
        • Cantley L.
        Nature. 1988; 332: 644-646
        • Auger K.R.
        • Carpenter C.L.
        • Cantley L.C.
        • Varticovski L.
        J. Biol. Chem. 1989; 264: 20181-20184
        • Parker P.J.
        • Waterfield M.D.
        Cell Growth Differ. 1992; 3: 747-752
        • Carpenter C.L.
        • Duckworth B.C.
        • Auger K.R.
        • Cohen B.
        • Schaffhausen B.S.
        • Cantley L.C.
        J. Biol. Chem. 1990; 265: 19704-19711
        • Otsu M.
        • Hiles I.
        • Gout I.
        • Fry M.J.
        • Ruiz-Larrea F.
        • Panayotou G.
        • Thompson A.
        • Dhand R.
        • Hsuan J.
        • Totty N.
        • Smith A.D.
        • Morgan S.J.
        • Courtneidge S.A.
        • Parker P.J.
        • Waterfield M.D.
        Cell. 1991; 65: 91-104
        • Hiles I.D.
        • Otsu M.
        • Volinia S.
        • Fry M.J.
        • Gout I.
        • Dhand R.
        • Panayotou G.
        • Ruiz-Larrea F.
        • Thompson A.
        • Totty N.F.
        • Hsuan J.J.
        • Courtneidge S.A.
        • Parker P.J.
        • Waterfield M.D.
        Cell. 1992; 70: 419-429
        • Kapeller R.
        • Prasad K.V.
        • Janssen O.
        • Hou W.
        • Schaffhausen B.S.
        • Rudd C.E.
        • Cantley L.C.
        J. Biol. Chem. 1994; 269: 1927-1933
        • Sugimoto Y.
        • Whitman M.
        • Cantley L.C.
        • Erikson R.L.
        Proc. Natl. Acad. Sci. U. S. A. 1984; 81: 2117-2121
        • Whitman M.
        • Kaplan D.R.
        • Schaffhausen B.
        • Cantley L.
        • Roberts T.M.
        Nature. 1985; 315: 239-242
        • Kaplan D.R.
        • Whitman M.
        • Schaffhausen B.
        • Raptis L.
        • Garcea R.L.
        • Pallas D.
        • Roberts T.M.
        • Cantley L.
        Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 3624-3628
        • Kaplan D.R.
        • Whitman M.
        • Schaffhausen B.
        • Pallas D.C.
        • White M.
        • Cantley L.
        • Roberts T.M.
        Cell. 1987; 50: 1021-1029
        • Coughlin S.R.
        • Escobedo J.A.
        • Williams L.T.
        Science. 1989; 243: 1191-1194
        • Varticovski L.
        • Druker B.
        • Morrison D.
        • Cantley L.
        • Roberts T.
        Nature. 1989; 342: 699-702
        • Ruderman N.B.
        • Kapeller R.
        • White M.F.
        • Cantley L.C.
        Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 1411-1415
        • Auger K.R.
        • Serunian L.A.
        • Soltoff S.P.
        • Libby P.
        • Cantley L.C.
        Cell. 1989; 57: 167-175
        • Serunian L.A.
        • Auger K.R.
        • Roberts T.M.
        • Cantley L.C.
        J. Virol. 1990; 64: 4718-4725
        • Fukui Y.
        • Saltiel A.R.
        • Hanafusa H.
        Oncogene. 1991; 6: 407-411
        • Courtneidge S.A.
        • Heber A.
        Cell. 1987; 50: 1031-1037
        • Pallas D.C.
        • Cherington V.
        • Morgan W.
        • DeAnda J.
        • Kaplan D.
        • Schaffhausen B.
        • Roberts T.M.
        J. Virol. 1988; 62: 3934-3940
        • Ling L.E.
        • Druker B.J.
        • Cantley L.C.
        • Roberts T.M.
        J. Virol. 1992; 66: 1702-1708
        • Fantl W.J.
        • Escobedo J.A.
        • Martin G.A.
        • Turck C.W.
        • del Rosario M.
        • McCormick F.
        • Williams L.T.
        Cell. 1992; 69: 413-423
        • Kazlauskas A.
        • Kashishian A.
        • Cooper J.A.
        • Valius M.
        Mol. Cell. Biol. 1992; 12: 2534-2544
        • Valius M.
        • Kazlauskas A.
        Cell. 1993; 73: 321-334
        • Roche S.
        • Koegl M.
        • Courtneidge S.A.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 9185-9189
        • Vemuri G.S.
        • Rittenhouse S.E.
        Biochem. Biophys. Res. Commun. 1994; 202: 1619-1623
        • Chung J.
        • Grammer T.C.
        • Lemon K.P.
        • Kazlauskas A.
        • Blenis J.
        Nature. 1994; 370: 71-75
        • Monfar M.
        • Lemon K.P.
        • Grammer T.C.
        • Cheatham L.
        • Chung J.
        • Vlahos C.J.
        • Blenis J.
        Mol. Cell. Biol. 1995; 15: 326-337
        • Rodriguez V.P.
        • Warne P.H.
        • Dhand R.
        • Vanhaesebroeck B.
        • Gout I.
        • Fry M.J.
        • Waterfield M.D.
        • Downward J.
        Nature. 1994; 370: 527-532
        • Eberle M.
        • Traynor-Kaplan A.
        • Sklar L.A.
        • Norgauer J.
        J. Biol. Chem. 1990; 265: 16725-16728
        • Severinsson L.
        • Ek B.
        • Mellstrom K.
        • Claesson W.L.
        • Heldin C.H.
        Mol. Cell. Biol. 1990; 10: 801-809
        • Zhang J.
        • King W.G.
        • Dillon S.
        • Hall A.
        • Feig L.
        • Rittenhouse S.E.
        J. Biol. Chem. 1993; 268: 22251-22254
        • Zheng Y.
        • Bagrodia S.
        • Cerione R.A.
        J. Biol. Chem. 1994; 269: 18727-18730
        • Zervos A.S.
        • Gyuris J.
        • Brent R.
        Cell. 1993; 72: 223-232
        • Lowenstein E.J.
        • Daly R.J.
        • Batzer A.G.
        • Li W.
        • Margolis B.
        • Lammers R.
        • Ullrich A.
        • Skolnik E.Y.
        • Bar-Sagi D.
        • Schlessinger J.
        Cell. 1992; 70: 431-442
        • Ausubel F.M.
        • Brent R.
        • Kingston R.
        • Moore D.
        • Seidman J.J.
        • Smith J.
        • Struhl K.
        Current Protocols in Molecular Biology. John Wiley & Sons, Inc., Boston, MA1993: 1.0.3-1.8.6 (13.7.1-13.13.7)
        • Skolnik E.Y.
        • Margolis B.
        • Mohammadi M.
        • Lowenstein E.
        • Fischer R.
        • Drepps A.
        • Ullrich A.
        • Schlessinger J.
        Cell. 1991; 65: 83-90
        • Smith D.B.
        • Johnson K.S.
        Gene (Amst.). 1988; 67: 31-40
        • Carpenter C.L.
        • Auger K.R.
        • Duckworth B.C.
        • Hou W.M.
        • Schaffhausen B.
        • Cantley L.C.
        Mol. Cell. Biol. 1993; 13: 1657-1665
        • Dhand R.
        • Hiles I.
        • Panayotou G.
        • Roche S.
        • Fry M.J.
        • Gout I.
        • Totty N.F.
        • Truong O.
        • Vicendo P.
        • Yonezawa K.
        • Kasuga M.
        • Courtneidge S.A.
        • Waterfield M.D.
        EMBO J. 1994; 13: 522-533
        • Welham M.J.
        • Dechert U.
        • Leslie K.B.
        • Jirik F.
        • Schrader J.W.
        J. Biol. Chem. 1994; 269: 23764-23768
        • Li W.
        • Nishimura R.
        • Kashishian A.
        • Batzer A.G.
        • Kim W.J.
        • Cooper J.A.
        • Schlessinger J.
        Mol. Cell. Biol. 1994; 14: 509-517
        • Bennett A.M.
        • Tang T.L.
        • Sugimoto S.
        • Walsh C.T.
        • Neel B.G.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7335-7339
        • Rozakis A.M.
        • McGlade J.
        • Mbamalu G.
        • Pelicci G.
        • Daly R.
        • Li W.
        • Batzer A.
        • Thomas S.
        • Brugge J.
        • Pelicci P.G.
        • Schlessinger J.
        • Pawson T.
        Nature. 1992; 360: 689-692
        • Skolnik E.Y.
        • Batzer A.
        • Li N.
        • Lee C.H.
        • Lowenstein E.
        • Mohammadi M.
        • Margolis B.
        • Schlessinger J.
        Science. 1993; 260: 1953-1955
        • Myers M.J.
        • Wang L.M.
        • Sun X.J.
        • Zhang Y.
        • Yenush L.
        • Schlessinger J.
        • Pierce J.H.
        • White M.F.
        Mol. Cell. Biol. 1994; 14: 3577-3587
        • Ren C.L.
        • Morio T.
        • Fu S.M.
        • Geha R.S.
        J. Exp. Med. 1994; 179: 673-680
        • Pendergast A.M.
        • Quilliam L.A.
        • Cripe L.D.
        • Bassing C.H.
        • Dai Z.
        • Li N.
        • Batzer A.
        • Rabun K.M.
        • Der C.J.
        • Schlessinger J.
        • Gishizky M.L.
        Cell. 1993; 75: 175-185
        • Schlaepfer D.D.
        • Hanks S.K.
        • Hunter T.
        • van der Geer P.
        Nature. 1994; 372: 786-791