THE JOURNAL OF Vol. 267, No. 15, of May 25, pp. 10588 ... · THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 hy The American Society for Biochemistry and Molecular Biology, Inc. Vol. 267,

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 hy The American Society for Biochemistry and Molecular Biology, Inc. Vol. 267, No. 15, Issue of May 25, pp. 10588-10695 1992

Printed in Ij.S.A.

Tyrosine-phosphorylated Epidermal Growth Factor Receptor and Cellular p130 Provide High Affinity Binding Substrates to Analyze Crk-Phosphotyrosine-dependent Interactions in Vitro*

(Received for publication, December 29, 1991)

Raymond B. BirgeS, J. Eduardo Fajardot, Bruce J. Mayerll, and Hidesaburo Hanafusall From the Department of Molecular Oncology, The Rockefeller University, New York, New York 10021

The genome of CTlO avian sarcoma virus encodes a 47-kDa fusion protein that consists of viral gag sequences fused to a cell-derived sequence containing SH2 and SH3 domains (v-crk). Genetic and biochemical evidence suggests that v-Crk can induce transformation of chicken embryo fibroblasts by influencing the activity of cellular proteins involved in growth regulation. In this report, we have developed an in vitro microtiter assay to study the binding of bacterially expressed glutathione S-transferase-fusion proteins of v-Crk and its cellular homolog, c-Crk, to the phosphorylated epidermal growth factor receptor (EGFR). Competitive binding data are presented that compare the abilities of heterologous glutathione S- transferase-fusion proteins containing GAPSHZ[N], AblSH2, SrcSH2, and PLC-rSHZ[N] sequences to inhibit Crk binding. Results indicate that both full-length Crk and GAPSH2[N] bind the phosphorylated EGFR with high affinity and can quantitatively compete the binding of each other by competitive enzyme-linked immunosorbent assay. Binding of full-length Crk or the isolated SH2 domains of GAP or Ab1 resulted in a significant protection of phosphorylated EGFR against dephosphorylation by cellular phosphatase activity, but did not appear to stimulate the intrinsic tyrosine kinase activity of the EGFR. To extend these findings to p130, the major phosphotyrosine-containing protein in CT10-transformed cells, we utilized a nitrocellulose filter binding assay. Results demonstrate high affinity binding of Crk toward denatured p130 and, as is the case for phosphorylated EGFR, Crk binding can par- tially protect p130 from phosphatase activity. How- ever, no apparent competition of Crk binding was noted with heterologous SH2-containing proteins including GAPSHZ[N], suggesting a possible specificity of Crk-pl3O binding. These data are consistent with a direct role of SH2 in the modulation of cellular phosphotyrosine status in vivo.

* This work was supported in part by United States Public Health Service Grant CA44356-03 from the National Institutes of Health and Grant 2517 from the Council for Tobacco Research. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

4 Supported by National Institutes of Health Postdoctoral Train- ing Grant AI-07233 and American Cancer Society Fellowship PF- 3653.

Fund Fellowship DRG-1049. Supported by Damon Runyon-Walter Winchell Cancer Research

1 Supported by United States Public Health Service Grant CA51462 and National Research Service Award 08875 from the National Cancer Institute.

11 To whom correspondence and reprint requests should be ad- dressed.

Binding of polypeptide growth factors to their respective receptor molecules results in an activation of the receptor tyrosine kinase activity, which triggers a complex array of enzymatic and biological events, ultimately leading to cell proliferation (1). Several cytosolic enzymes that regulate intracellular signal transduction, including phospholipase C-7 (PLC-7)’ (2, 3), Ras GTPase-activating protein (GAP) (4, 5), and the p85 subunit of phosphatidylinositol-3-kinase (6, 7), can bind to and co-immunoprecipitate with activated receptor-type tyrosine kinases. These interactions are dependent upon the tyrosine phosphorylation of the receptor and appear to be mediated by a conserved noncatalytic src homology 2 (SH2) domain present in the effector molecules (8, 9). While the general phenomena of SH2-phosphotyrosine associations have been clearly established, it is likely that cell-specific factors along with different SH2 phosphotyrosine binding affinities ultimately generate discrimination among unique signaling events (10).

The importance of the SH2 domain as a regulatory com- ponent in intracellular signaling is exemplified further by the Crk oncoprotein encoded by the avian sarcoma virus CTlO (11, 12). The product of CTlO virus consists of viral gag sequences fused to cellular sequences containing SH2 and SH3 domains. In this paper, we refer to gag-v-crk as v-crk. Although v-Crk lacks a tyrosine kinase domain, cells transformed by CTlO virus have increased tyrosine phosphorylation of several cellular proteins (12, 13). Furthermore, v-Crk was found to form stable associations with all the major tyrosine-phosphorylated proteins from these cells (14, 15). Both the transformation by v-Crk and the binding of v-Crk to phosphotyrosine-containing proteins depend upon an intact SH2 domain (15, 16), suggesting a direct role of SH2 in the modulation of intracellular tyrosine phosphorylations.

In the present study, we have examined the binding of v- Crk and its cellular homolog c-Crk to both the phosphorylated EGFR and p130, the major tyrosine-phosphorylated protein from CT10-transformed chicken embryo fibroblasts (CEF). Data are presented that compare the binding properties of glutathione S-transferase (GST)-Crk fusion proteins with that of heterologous GST-fusion proteins containing SH2 sequences from N-terminal GAP, Abl, Src, and N-terminal PLC-7. Moreover, we show that SH2 binding results in the protection of tyrosine-phosphorylated proteins from dephosphorylation by cellular phosphatases. These data are consist-

’ The abbreviations used are: PLC-7, phospholipase C-7; GAP, GTPase-activating protein; SH2, Src homology 2; CEF, chicken embryo fibroblasts; GST, glutathione S-transferase; EGFR, epidermal growth factor receptor; SDS, sodium dodecyl sulfate; PAGE, poly- acrylamide gel electrophoresis; PMSF, phenylmethylsulfonyl fluo- ride; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; ELISA, enzyme-linked immunosorbent assay; BSA, bovine serum albumin; DTT, dithiothreitol; mAb, monoclonal antibody.

10588

Crk-Phosphotyrosine Interactions 10589

ent with a regulatory role of SH2 sequences in the modulation of intracellular phosphotyrosine status.

MATERIALS AND METHODS

Cells and Viruses-CEF were cultured and maintained as described previously (17). To obtain v-Crk-transformedcells, CEF were infected by avian sarcoma virus CTlO in the presence of DEAE-dextran and passed every 3 days. In the controls, the CEF were infected with UR2AV helper virus alone (17). Cells were analyzed for transformation by morphological criteria and lysed about 10 days postinfection. For EGFR studies, A431 human carcinoma cells were utilized (18). Cells were grown to about 80% confluency in Dulbecco's minimum essential media supplemented with 10% fetal calf serum and then placed in Dulbecco's minimum essential media containing 0.5% fetal calf serum for 24 h prior to cell lysis.

Protein Analysis-Cell lysis, immunoprecipitations, in vitro kinase reactions, immunoblotting, and electrophoresis were performed as described previously (13,14). Briefly, CEFs were lysed in RIPA buffer, pH 7.5, containing 10 mM Tris-HC1, 150 mM NaCl, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, 1 mM Na3V0,, 0.1 mM Na2MoO4, 1 mM PMSF, and 100 units/ml of aprotinin, except where indicated. A431 cells were lysed in ice-cold HNTG buffer (50 mM Hepes, 1% Triton X-100, 150 mM NaCl, 10% glycerol) containing 1 mM Na3V04, 1 mM PMSF, 10 mM EDTA, and 100 mM NaF (18). To immunoprecipitate EGFR, a monoclonal antibody directed against the extracellular domain (Amersham: RPN.513) was incubated (1:20) with precleared lysate for 3 h, and immune complexes were collected by the addition of Protein A-Sepharose beads (Pharmacia LKB Biotechnology Inc.). Immunoprecipitates were washed five times in HNTG buffer (0.1% Triton X-100) lacking inhibitors and autophosphorylated by the addition of kinase buffer (HNTG containing 10 m M MnC12 and 100 p~ ATP) for 30 min at room temperature. For isotopic labelingexperiments, 50 pCi of [y-3ZP]ATP (3000 mCi/mmol) was added to the kinase buffer.

Construction of GST-Crk Fusion-expressing Plasmids-The SfiI- AlwNI c-crk fragment which contains the entire c-crk coding region starting at the initiator ATG, or the BamHI-Alw NI gag-crk fragment, beginning at codon 52 in the gag region of v-crk, was gel-purified, blunt-ended with T4 DNA polymerase, and ligated to EcoRI-linear- ized, blunt-ended pGEX-3X DNA (Fig. 1). After transformation of the DNA into Escherichia coli, random colonies were picked and analyzed by restriction digestion and DNA sequencing. Cells harbor- ing the appropriate expression vectors were then grown at 37 "C in LB media containing 100 pg/ml ampicillin and induced at an ODW of 1.0 with 0.1 mM isopropyl-1-thio-0-D-galactopyranoside for an additional 2 h. Cells were harvested by centrifugation, and the bacterial pellets were lysed in RIPA buffer containing 1 mM PMSF and 5 pg/ml leupeptin. After clarification of bacterial lysates (10,000 X g for 10 minutes), GST-fusion proteins were incubated with glutathione-Sepharose beads (Pharmacia), washed with Tris-buffered saline

GST

SH2 sH3 SH3 GST-c-Crk

GAG GST-v-Crk I

sH2 SH3 .:&>:kf@&$ v/////- d ,..,, .,~, *.....

GST-GAPSHP[N] 7-

GST-AblSH2 7-

GST-SrcSH2

GST-PLCSHZ[N]

FIG. 1. Constructs utilized in this study. All constructs were expressed as GST-fusion proteins and purified by GST-Sepharose beads. The structures and locations of SH2 are indicated in the figure.

containing 0.05% Tween 20 (TBS-T), and affinity-purified with 20 mM reduced to glutathione (GSH). Free GSH was removed prior to binding studies by Sephadex G-25 gel filtration resin.

Cloning of GAPLN], Abl, Src, and PLC-y[NI Sequences by Poly- merase Chain Reaction-The SH2 domain of c-ab1 was cloned down- stream of glutathione S-transferase in the vector pGEX-2T and expressed as a fusion protein in E. coli as described (19). The SH2 domain of c-src and the N-terminal SH2 domains of GAP and PLC- y were cloned by reverse transcription and polymerase chain reaction from NIH-3T3 fibroblast RNA. Polymerase chain reaction primers were synthesized with 5'-BamHI and 3'-EcoRI sites, and polymerase chain reaction products were digested with the two enzymes and cloned into EcoRI-BamHI double-digested pGEX-2T (Fig. 1). The GAP SH2 fragment corresponded to amino acids 175-279 of human GAP (20). The PLC-y SH2 corresponded to amino acids 548-649 of bovine PLC-y (21) and the src SH2 comprised amino acids 155-256 of murine neural src (amino acids 149-250 of the fibroblast form) (22). All constructions were verified by DNA sequencing.

Bacterially expressed fusion proteins were grown and purified essentially as described (23). Briefly, bacteria were grown to late log phase, induced for 2 h with 0.1 mM isopropyl-1-thio-P-D-galactopy-

taining 1% Triton X-100, 100 mM NazEDTA, and 1 mM PMSF. The ranoside, and lysed by sonication in phosphate-buffered saline con-

lysates were clarified by centrifugation, and fusion proteins were purified by binding to glutathione-agarose beads (Molecular Probes, Inc.). Proteins were eluted from the beads with 20 mM glutathione, and buffer was exchanged by gel filtration to phosphate-buffered saline containing 10% glycerol. Proteins were quantified by Coo- massie Blue dye binding (Bio-Rad) relative to y-globulin protein standards. All fusion proteins were analyzed by Western blotting and dot-blotting with affinity-purified anti-GST antibody (kindly pro- vided by R. Van Etten, Whitehead Institute, Boston, MA) to demonstrate similar reactivity of the recombinants (data not shown).

Characterization of SH2-dependent Binding to EGFR by Solid Phase Horseradish Peroxidase-ELISA-Association of SH2-containing polypeptides to the phosphorylated EGFR was assayed by ELISA. Wells of polystyrene microtiter plates (Nunc) were coated overnight at 4 "C with 45 pl of a 1 pg/ml solution (in carbonate buffer, pH 9.6) of monoclonal anti-EGFR antibody. Each well was blocked with BSA (10 pg/ml) in carbonate buffer for 30 min and then rinsed five times with HNTG buffer followed by the addition of 50 pl of cell lysate (initially diluted to 15 pg/ml from the original lysate) from serum- starved A431 cells. After 3 h at 4 "C, the plates were washed five times with HNTG buffer, and bound receptor molecules were autophosphorylated by the addition of 50 pl of HNTG buffer containing 100 p M ATP and 10 mM MnC12. After 30 min, the plates were again washed five times with TBS-T. To analyze the binding of the SH2- containing proteins, stock solutions (in TBS-T) of the proteins were diluted serially and applied to separate wells in duplicate. Binding was for 1 h at 37 "C unless otherwise indicated. Depending upon the experimental design, bound polypeptides were detected by either an affinity-purified anti-GST antibody or, in the case of Crk proteins, with polyclonal anti-Crk antibody (13). Following further washing, bound antibody was detected after the addition of peroxidase-conjugated goat anti-rabbit IgG (1:400 in TBS-T) (Organon Teknika) for 1 h at 37 "C. In some instances, purified GST-fusion proteins were biotinylated with sulfosuccimidyl 6-(biotinamid0)hexanote (Pierce) as described by the manufacturer and dialyzed against TBS-T prior to use. Biotinylated proteins were detected using peroxidase-conju- gatedstreptavidin (1:lOOO) (Pharmacia). After a series of final washes, peroxidase activity was monitored at 405 nm on an Artek plate reader by the addition of peroxidase substrate, 2,2'-azino-di-(3-ethylbenz- thiazoline sulfonic acid) in citrate buffer, pH 4.0, containing 0.12% hydrogen peroxide. Relative peroxidase activity was obtained by subtracting the readings at time 1 from time 2. For competitive ELISA, the microtiter plates were prepared exactly as described except that, after receptor phosphorylation, the wells were incubated with a premixed solution of ligands and their potential inhibitor molecules for 1 h at 37 "C. Following extensive washes, the bound ligand was detected and analyzed in a microtiter reader as described above.

Effect of SH2 Binding on Phosphatase Susceptibility and Intrinsic Kinase Actiuity of EGFR-Lysates from unstimulated A431 cells were incubated with anti-EGFR antibody for 3 h, and immune complexes were recovered with Protein A-Sepharose beads. After autophosphorylation of the immobilized EGFR with [y3'P]ATP and MnClZ, the immunoprecipitates were washed extensively with HNTG buffer and incubated with the SH2-containing proteins for 1 h at room temper-

10590 Crk-Phsphtyrosine Interactions

ature to achieve stable interactions. After several washes to remove nonbound SH2 protein, the immunoprecipitates were incubated for 1 h at 37 "C with 50 pl of CEF lysate (without phosphatase inhibitors) containing 1 mM DTT and 1 mM EDTA to favor phosphatase activity (24). The complexes were again washed and boiled in Laemmli's electrophoresis buffer for analysis by SDS-PAGE and autoradiogra-

For the in oitro kinase reactions, both the extent of receptor autophosphorylation and the tyrosine phosphorylation of an exogenous synthetic Src peptide (Arg-Arg-Leu-Ile-Glu-Asp-Ala-Gly-Tyr- Ala-Ala-Glu: Sigma A-4666) were analyzed. The effect of SH2 binding on receptor kinase activity was monitored by either of two strategies. In the first, the SH2 proteins were mixed directly with the kinase buffer and then added to the EGFR immobilized on Protein A- Sepharose beads. In the second method, immobilized receptors were phosphorylated in the presence of cold ATP for 30 min to provide preformed SH2 binding sites. Subsequently, the SH2 proteins were incubated for 1 h to achieve stable associations. After several washes with HNTG buffer to remove nonbound proteins, a kinase mixture containing [y3'P]ATP and the exogenous peptide substrate was added for an additional 30 min at room temperature. For the synthetic Src peptide phosphorylation, 10 pg was added to the kinase buffer, and, after the 30-min incubation, the beads were collected by centrifugation and the supernatant was precipitated on ice for 1 h with an equal volume of 20% trichloroacetic acid (25). Following centrifugation for 5 min to remove acid-insoluble proteins, the supernatant was spotted onto a Whatman P81 phosphocellulose disc (26) and washed five times with 15% acetic acid prior to quantitation by Cerenkov counting (27). For receptor phosphorylation, the beads were washed five times with ice-cold HNTG buffer containing 10 mM cold ATP and quantified by liquid scintillation counting or by SDS-PAGE/ autoradiography/densitometry.

Characterization of Crk Binding top130 by a Filter Binding Assay- 50 pg of RIPA-soluble protein from control or CT10-transformed CEFs was electrophoretically resolved by 8.5% SDS-PAGE and elec- trobIotted to nitrocellulose membranes (0.45 pm) for 16 h at 70 V in buffer containing 25 mM Tris (pH 8.3), 190 mM glycine, and 20% methanol (28). After transfer, the .region between the 97- and 200- kDa prestained molecular size markers was excised, placed in a 35- mm tissue culture dish, and blocked with TBS containing 1% BSA for 1 h. The membranes were then incubated with 500 pl of diluted Crk proteins in TBS-BSA for 1 h at room temperature with gentle shaking. In separate experiments, GST-Crk proteins were mixed with various concentrations of the heterologous GST-SH2 fusion proteins to examine the specificity of binding. Following three 10-min washes with TBS-T, the membrane strips were incubated further with anti- Crk antibody (1:lOOO) followed by '"I-conjugated Protein A (Amer- sham). Crk binding was visualized by autoradiography on Kodak XAR-5 film for 24 h at -70 "C with an intensifying screen.

To examine the effect of Crk binding on the susceptibility of p130 toward exogenous tyrosine phosphatase activity (PTPases), the membrane strips were incubated with or without Crk proteins as described above and then further incubated with 500 p1 of CEF lysate (approximately 1.0 mg of total cellular protein) in RIPA buffer containing 1 mM DTT and 1 mM EDTA, The membranes were washed three times in TBS-T and then incubated with TBS containing 1% SDS at 65 "C for 10 min to strip off the bound Crk proteins. Following several washes in TBS-T, the membrane strips were probed with an affinity- purified anti-phosphotyrosine (PY) antibody (1:lOOO) (29) and detected with '251-conjugated Protein A as above.

phy.

RESULTS

Analysis of Crk Binding to EGFR by Horseradish Peroxi- dase-ELISA-When purified GST-c-Crk or GST-v-Crk was incubated with EGFR immunoprecipitated from serum- starved A431 cells, both proteins bound to the receptor in a phosphorylation-dependent manner (Fig. 2). The binding was also concentration-dependent with both proteins exhibiting a biphasic titration profile in the solid phase binding assay as a function of protein dilution. In the linear region, both GST- c-Crk and GST-v-Crk bound similarly with half-maximal binding occurring at approximately lo-' M, while a t higher concentrations, the binding began to plateau but did not readily saturate. No receptor binding was observed with the isolated GST alone (data not shown). Based upon the results

c 1

0.8

0.6

0.4

0.2

0

1

150 75 37 18 9.0 4.5 2.3 1.1 0.6 0.3 0.15

[Crk] (x lO-'M)

FIG. 2. Binding of GST-c-Crk and GST-v-Crk to EGFR: effect of receptor phosphorylation. Wells of microtiter plates were coated with 45 ng of anti-EGFR monoclonal antibody. Wells were then incubated with 50 pl of lysate (15 pg of cellular protein/ ml) from unstimulated A431 cells as described under "Materials and Methods." After washing, the immobilized immune complexes were incubated in the presence (open symbols) or absence (closed symbols) of kinase buffer containing 10 mM MnC12 and 100 p~ ATP. GST-c- Crk (0) or GST-v-Crk (A), initially prepared as a 20 pg/ml stock solution in TBS-T, was serially diluted from 1:l to 1512, and 50 pl of each diluent was added to separate wells in duplicate. After 1 h at 37 "C, Crk binding was detected with anti-Crk antibody (1:lOOO) followed by peroxidase-conjugated goat anti-rabbit IgG.

from Fig. 2, it appeared that GST-c-Crk and GST-v-Crk bound phosphorylated EGFR with similar affinities. However, to examine the relative binding affinities of GST-c-Crk and GST-v-Crk to the EGFR in more detail, competitive ELISAs were performed (Fig. 3). Since the cellular derived sequences of v-Crk and c-Crk are very similar (Fig. l), it was first necessary to biotinylate each molecule in order to discriminate between the two in a competitive binding assay. Competitive binding studies between unlabeled and biotinylated molecules confirmed that biotinylation did not interfere with the SH2- mediated binding activity of the GST-fusion proteins (data not shown). Shown in Fig, 3 are the competition of GST-c- Crk-biotin (GST-c-Crk-B) by GST-v-Crk (panet A ) and the competition of GST-v-Crk-B by GST-c-Crk (panel B) . Each protein was able to effectively compete the binding of its cognate protein with similar Ki values (approximately M in both cases). To obtain 50% binding inhibition, about 5 times the amount of competitor protein was required. Taken together, these data indicate that GST-c-Crk and GST-v-Crk bind the phosphorylated EGFR with approximately equal affinity.

Competition of GST-Crk by Heterohgous SH2-containing Proteins-In an attempt to gain insights into the specificity of Crk binding, we investigated the ability of other known SH2 sequences to compete the binding of Crk to the phosphorylated EGFR. Fig. 4 shows the competition of GST-C- Crk-B binding toward the EGFR by GST, GST-GAPSH2[N], GST-AblSH2, GST-SrcSH2, and GST-PLC-rSH2[N]. In this experiment, the concentration of GST-c-Crk-B was held constant while the concentration of all competitor proteins was initially 500 times greater than GST-c-Crk-B and then serially diluted. GST-Crk proteins and competitors were premixed and incubated for 10 min prior to application to the immobilized EGFR in the microtiter plate. As indicated in Fig. 4A, each of the competitor proteins tested demonstrated a differential inhibitory effect on the binding of GST-c-Crk-


A A

400 2W 100 50 25 12.5 8.2 3.1 1.5 0.70

[v-Crk] (x 10%)

B

l o o 7

870 335 187 84 42 21 10.5 5.2 2.8 1.3

[c-Crk] (x 1O''M)

FIG. 3. Binding affinities of GST-c-Crk and GST-v-Crk toward the phosphorylated EGFR using competitive horseradish peroxidase-ELISA. Wells of microtiter plates were prepared as described in Fig. 2. In panel A, a 50 pg/ml solution of GST- v-Crk was serially diluted from 1:l to 1512 in TBS-T. 25 pl of each diluent was mixed with 25 p1 of a 2 pg/ml solution of GST-c-Crk-B and mixed for 10 min prior to addition to the wells. In pane l E , 25 pl of a 80 pg/ml solution of GST-c-Crk solution was serially diluted; each diluent was mixed with 25 pl of a 2 pg/ml solution of GST-V- Crk-B for 10 min prior to addition to the wells. Reactions were for 1 h at 37 "C. Competition of biotinylated proteins was determined after the addition of peroxidase-conjugated streptavidin. Data are expressed as percent inhibition of peroxidase activity compared to wells in which no inhibitor was added. All data represent the average of duplicate wells. Ki was obtained from the molar concentration of inhibitor necessary to achieve 50% inhibition of Crk binding. * indicates the position in which equivalent amounts of proteins were present.

B. Thus, GST-GAPSH2[N] was clearly the most effective followed by GST-AblSH2 which was more effective than GST-SrcSH2. No inhibition of GST-c-Crk-B binding was observed with GST alone or with GST-PLC-ySH2[N] even at the highest concentrations tested. At high concentrations of inhibitor, the inhibition of GST-c-Crk-B by GST- GAPSHB[N] was greater than 90% and the Kt was approximately lo" M. At 50% inhibition, the amount of GST- GAPSHZ[N] was about 4 times the amount of GST-c-Crk-B. A similar qualitative and quantitative pattern of GST-v-Crk inhibition was observed with these heterologous SH2 proteins (data not shown). In additional experiments, we biotinylated GST-GAPSHB[N] to test whether GST-c-Crk could also inhibit GST-GAPSHS[N] binding. Quantitative inhibition of binding was observed (about 90% at high GST-c-Crk levels) with similar protein stoichiometry as with the inhibition of GST-c-Crk by GST-GAPSHB[N] (data not shown). We also tested the ability of phosphoamino acids to compete GST-c- Crk binding. Neither phosphoserine nor phosphothreonine produced detectable inhibition at concentrations 2,500 times the amount of GST-c-Crk. In contrast, a small but consistent

6250 1560 390 48 12 3.0

[GST-SHP] (x lo%)

BlOO ~ ' " " " " ' " " " ' " ' ~ ~

.- I 6 60 I

6250 1560 390 48 12

[GST-SHP] (x lO.'M)

FIG. 4. Inhibition of GST-c-Crk binding to phosphorylated EGFR by GST-SH2-containing polypeptides. The inhibition of GST-c-Crk-B binding by GST (A), GST-GAPSHB[N] (A), GST- AblSH2 (O), GST-SrcSH2 (O), or GST-PLC-ySH2[N] (0) was determined by a competition ELISA using horseradish peroxidase- conjugated streptavidin. All inhibitor concentrations were initially 500 pg/ml and serially diluted from 1:l and 1:1024 in TBS-T. 25 pl of each diluent was mixed with 25 pl of a 1.0 pg/ml solution of GST- c-Crk-biotin for 10 min prior to addition to the wells. Competition reactions were for 1 h at 37 "C. In panel E , GST-c-Crk-biotin was incubated with the phosphorylated EGFR for 1 h at 37 "C and then competed with serially diluted GST-GAPSHB[N] or GST-AblSH2 for an additional 1 h. Data are the average of duplicate wells at 37 "C and are expressed as in Fig. 3. * indicates the position in which equivalent amounts of proteins were present.

inhibition of GST-c-Crk binding was observed with phosphotyrosine (30% inhibition at lo-* M phosphotyrosine) (data not shown).

Equilibrium Analysis of GST-c-Crk Binding Inhibition by GST-GAPSH2[N] and GST-AblSH2"In the experiments above, the competition ELISAs were performed by adding simultaneously an equilibrated GST-Crk plus competitor mixture (Fig. 4A). However, in a comparison of the effect of GST-c-Crk prebinding followed by the addition of competitors, the inhibition of GST-c-Crk by GST-GAPSHB[N] or GST-AblSH2 was completely abolished (Fig. 4B). These data suggest that the binding of GST-Crk protein was stable under these assay conditions.

Binding of Heterologous SH2-containing Proteins to Phos- phorylated EGFR-When the binding of serially diluted GST- fusion proteins to phosphorylated EGFR was analyzed directly with an affinity-purified anti-GST antibody, each SH2 protein bound with a different affinity (Fig. 5A) . Thus, GST- GAPSH2[N] had the highest affinity, followed by GST- AblSH2, followed by GST-SrcSH2 which was considerably lower and did not begin to plateau even at the highest concentrations tested. The binding of GST-PLC-rSH2[N] was only barely detectable at these concentrations although it did appear to be greater than that of GST alone (see also Fig.

10592 Crk-Phosphotyrosine Interactions

I I I I I I I I ,

30 60 120 240

Time (minutes)

FIG. 5. Binding of GST-SH2 polypeptides to phosphorylated EGFR. Microtiter plates were prepared as in Fig. 2. Binding of GST (A), GST-GAPSHB[N] (A), GST-AblSH2 (O), GST-SrcSH2 (O), or GST-PLC-rSH2[N] (0) was assayed by noncompetitive ELISA using affinity-purified anti-GST antibody followed by horseradish peroxidase-conjugated goat anti-rabbit. In panel A, all GST- SH2 fusion proteins were initially 500 pg/ml and serially diluted from 1:l to 1:8196 in TBS-T. 25 pl of each diluent was added for 1 h at 37 "C. Data are the average of duplicate wells. Panel B, proteins were diluted to 200 pg/ml and added to the wells for the indicated times.

5B). Only background binding was observed for all GST-SH2 fusion proteins when the receptors were not autophosphorylated (data not shown). It is apparent from a comparison of Figs. 4 and 5 that a strong correlation exists between the extent of SH2 binding to the EGFR and the ability to compete GST-c-Crk. Interestingly, GST-GAPSHB[N] binding consist- ently resulted in a higher relative peroxidase activity than GST-AblSH2 at higher concentrations. This may result if GST-GAPSHB[N] binds to a greater number of the potential phosphorylation sites on the EGFR (30). To test this hypoth- esis, we are currently evaluating the inhibition of SH2 binding by synthetic phosphopeptides corresponding to the known phosphorylation sites on the EGFR.

To address the possibility that the differences in binding affinities were due to differences in the initial rates of binding, we performed a kinetic study (Fig. 5B). At all time points up to 4 h, each of the GST-SH2 proteins bound with ranking similar to that observed in Fig. 5A. Since binding of all proteins produced similar curve profiles over the times indicated, it is unlikely that the length of incubation influenced the relative affinity measurements obtained in this study.

Effect of GST-c-Crk, GST-u-Crk, GST-GAP, and GST-Ab1 on PTPase Susceptibility and Intrinsic Tyrosine Kinase Actiu- ity of the EGFR-Due to the stable nature of the SH2- phosphotyrosine interactions, we were interested in determin- ing whether such binding affected the phosphorylation status or intrinsic kinase activity of the EGFR. For the phosphatase (PTPase) susceptibility studies, the EGFR was immobilized

on Protein A-Sepharose beads and autophosphorylated in the presence of [y-32P]ATP. Subsequently, the beads were divided into replicate samples and incubated with 10 pg of either GST alone, GST-c-Crk, GST-v-Crk, GST-GAPSHB[N], or GST- AblSH2. These concentrations corresponded to maximal binding assayed by ELISA. After removal of nonbound proteins, the beads were incubated with CEF lysate as a source of exogenous cellular PTPase activity. Binding of all SH2- containing proteins produced a significant protection against dephosphorylation by PTPase (Fig. 6 A ) . Thus, GST-c-Crk and GST-v-Crk protected about 7-fold compared to GST alone, and GST-AblSH2 and GST-GAPSH2[N] protected about 10- and 20-fold, respectively. A quantitatively similar result was obtained with GST-c-Crk and GST-v-Crk when phosphorylated EGFR was treated with 10 ng of the catalytic subunit of purified leukocyte common antigen related phosphatase (data not shown). The lack of complete protection may result in part from the fact that we did not observe complete saturation of binding after 1 h (Fig. 5B). To demonstrate a correlation between the extent of SH2 binding and the extent of protection from PTPase, we made 5-fold serial dilutions of GST-GAPSHB[N] and GST-AblSH2 (these proteins were most effective on PTPase protection above). As

1 2 3 4 5 6

B GAPSH2[N] I I , ablSH2 I . .

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3

FIG. 6. Effect of GST-SH2 binding on the susceptibility of phosphorylated EGFR toward cellular phosphatase activity. 200 p1 of 1.0 mg/ml cell lysate from unstimulated A431 cells was immunoprecipitated with 10 pl of anti-EGFR antibody and autophosphorylated with [Y-~'P]ATP. After washing, immobilized receptors were divided into 20 replicate aliquots and incubated with 10 pg of either GST (2), GST-GAPSHB[N] (3 ) , GST-AblSH2 (4), GST-c- Crk (5), or GST-v-Crk (6) for 1 h to achieve stable interactions (A). After washing, the beads were incubated with 50 pl of CEF lysate containing 1 mM DTT and 1 mM EDTA for 60 min at room temperature. Receptor phosphorylation was analyzed by 8.5% SDS-PAGE followed by autoradiography. Data were obtained from laser densi- tometric scanning of the films and is expressed as the percent of initial radioactivity for which no lysate was added (Fig. 7A, I ) . All data are from a single exposure. In Fig. 7B, 50 pl of 5-fold serial dilutions of a 100 pg/ml solution of GST-GAPSHB[N] or GST- AblSH2 was incubated with the beads as above. Lane I , no lysate. Lanes 2-6, serially diluted GST-GAPSH2[N]. Lanes 8-12, serially diluted GST-Abl. Lanes 7 and 13, without GST-SH2 binding. Data are shown as an autoradiogram of the bound radioactivity from an 8.5% acQlamide gel.


indicated in Fig. 6B, the extent of phosphorylation after PTPase was directly proportional to the extent of SH2 binding.

To examine the effect on tyrosine kinase activity, excess GST-c-Crk or GST-v-Crk was added to the immobilized EGFR Protein A beads for the duration of the kinase experiment. These concentrations produced no statistically significant effect on the tyrosine kinase activity of the receptor toward an exogenous Src peptide or on the extent of receptor autophosphorylation (+ S.E. of three independent experiments) (data not shown). We also did not detect any effect on the EGFR kinase activity toward the exogenous peptide when the receptors were phosphorylated initially with cold ATP to provide preformed phosphotyrosine binding sites and then incubated with Crk proteins. Finally, a kinase experiment was also conducted from an experiment similar to the one described in Fig. 6B. Again, no detectable differences in the kinase activity was noted after the binding of GST- GAPSHB[N] or GST-AblSH2. Thus, the intrinsic kinase activity of the EGFR appears to be independent of the binding of SH2 molecules to the phosphotyrosine residues.

Binding of GST-c-Crk and GST-u-Crk to Phosphorylated p130 from CTlO-transformed Cells-In order to extend our observations from the phosphorylated EGFR to the major tyrosine-phosphorylated protein in v-Crk-transformed cells (11-14), we utilized a nitrocellulose filter binding assay (19) in which lysates from control or CT10-transformed CEF were electrophoretically resolved, transferred to nitrocellulose membranes, and incubated with purified GST-c-Crk or GST- v-Crk proteins. Shown in Fig. 7 is an anti-Crk immunoblot of nitrocellulose strips probed with 50 nM GST-Crk fusion proteins. Both GST-c-Crk and GST-v-Crk bound directly to a 130-kDa protein in the CTlO lanes but not in the control lanes (Fig. 8, lanes 3-6 of the anti-Crk panel). Likewise, no 130-kDa band was noted if the membrane strips were not incubated with Crk proteins (Fig. 7, lanes 1 and 2 of the anti- Crk panel). In some instances, other minor bands in the CTlO lanes of approximately 150 and 120 kDa were also apparent in the CTlO lysate lanes but p130 was always the most intense band with this experimental system. Also shown in Fig. 7 is

A anti-PY B anti-Crk

. - "pp130

. -

"

"

1 2 3 4 1 2 3 4 5 6

FIG. 7. Binding of GST-c-Crk and GST-v-Crk to p130. In panel B, 50 pg of RIPA-soluble cellular protein from control UV2AV (lanes I , 3, and 5) or CT10-transformed CEFs (lanes 2 ,4 , and 6 ) was electrophoretically resolved by 8.5% acrylamide and electroblotted to nitrocellulose membranes (panel B ) . After blocking with 1% BSA, the strips were incubated with 50 nM GST-c-Crk (lanes 3 and 4 ) or GST-v-Crk (lanes 5 and 6 ) for 1 h. In lanes 1 and 2, no Crk proteins were preincubated. Crk proteins were detected with anti-Crk antibody followed by "'I-conjugated protein A. In panel A, lysate from control (lanes I and 3 ) or CT10-transformed CEFs (lanes 2 and 4 ) were analyzed directly by Western blotting with anti-phosphotyrosine antibody (lanes 1 and 2 ) or immunoprecipitated with anti-Gag mAb3C2 antibody (lanes 3 and 4 ) followed by Western blotting with anti-phosphotyrosine antibody.

[GST-c-Crk] 50 5 0.5 0.05

GAPSH2[N] ablSH2 srcSH2 PLCSHqV

FIG. 8. Characterization of GST-c-Crk binding to p130: effect of heterologous GST-SH2-containing polypeptides. Ly- sates from control (lanes 2) or CT10-transformed CEFs (lanes I ) were electrophoresed exactly as described in Fig. 7. In panel A, GST- c-Crk was initially 50 nM and then serially diluted from 1:l to 1:lOOO to yield the final nanomolar concentrations indicated. In panel R, 100-fold molar excess of the indicated GST-SH2 sequences were premixed with 50 nM GST-c-Crk and applied to the membrane strips. All data are anti-phosphotyrosine immunoblots and was obtained from a single exposure.

an anti-phosphotyrosine (anti-PY) immunoblot of total control and CTlO lysate (lanes 1 and 2 ) and of cellular phosphotyrosine-containing proteins immunoprecipitated with an anti-Gag monoclonal antibody (mAb3C2), to precipitate p47c0*crk and its associated proteins (lanes 3 and 4 ) . Consistent with our previous observations (13, 14), a clear correlation was noted between the patterns of phosphotyrosine proteins induced in Crk-transformed cells and the proteins to which Crk co-immunoprecipitates with.

To assess the relative binding affinity of Crk protein to p130, 10-fold serial dilutions of GST-c-Crk were incubated with replicate membrane strips. Under these conditions, GST- Crk fusion proteins bound to p130 with high affinity, estimated to be in the range of lo-" to 10"" M from the titration of Crk on the strips (Fig. 8, panel A ) . A similar titration was obtained with GST-v-Crk (data not shown). To examine whether the SH2-containing proteins that could compete GST-c-Crk binding to the phosphorylated EGFR could also compete GST-c-Crk binding to p130, up to 100-fold molar excesses of the SH2-containing proteins were premixed with GST-c-Crk and then incubated with the membranes strips. In contrast to the complete inhibition of GST-c-Crk binding to phosphorylated EGFR by GST-GAPSH2[N], these proteins produced no apparent competition of GST-c-Crk binding to the p130 tyrosine-phosphorylated protein from CT10- transformed cells. Furthermore, no detectable inhibition of GST-c-Crk was noted with excess GST-AblSH2, GST- SrcSH2, and GST-PLC-rSH2[N] (Fig. 8, panel B ) .

Effect of GST-c-Crk Binding on the PTPase Susceptibility of pl30"To examine the effects of GST-c-Crk binding to p130 on its susceptibility to dephosphorylation by PTPase, the nitrocellulose strips were incubated with CEF lysate in the presence or absence of bound GST-c-Crk. After 1 h, the strips were washed and probed with affinity-purified anti- phosphotyrosine antibodies (Fig. 9). Consistent with the results from the phosphorylated EGFR, the 130-kDa protein that directly bound GST-c-Crk retained a significantly higher percentage of phosphotyrosine than the untreated membrane strip (compare lane 4 versus lane 6 in panel E?). These data are consistent with a partial protection of phosphotyrosine- containing substrates after Crk association.

10594 Crk-Phosphotyrosine Interactions A + Crk B Tota l f Crk - Crk

ii 1 2 1 2 3 4 5 6

FIG. 9. Effect of GST-c-Crk binding on the susceptibility of phosphorylated p130 toward cellular phosphatase activity. Reactions are as described in Fig. 8. In panel A, the extent of GST- c-Crk binding is indicated. All lanes in panel B represent anti-PY immunoblots. After preincubation in the presence or absence of GST- c-Crk for 1 h, the membrane strips were washed and then treated with CEF lysate (without PTPase inhibitors) containing 1 mM DTT and 1 mM EDTA for 1 h at 37 "C. Lanes 1 and 2 are total lysate; Lanes 3 and 4, plus GST-c-Crk. Lunes 5 and 6, minus GST-c-Crk. Lanes 1, 3, and 5 are lysates from control and lanes 2, 4, and 6 are lysates from CT10-transformed cells. All strips were treated with SDS prior to anti-PY antibody and are from the same exposure.

DISCUSSION

Protein-protein interactions between SH2-containing and phosphotyrosine-containing proteins have been proposed to play an important role in intracellular signal transduction pathways (8, 9, 14). Furthermore, studies with the v-Crk oncoprotein indicate that SH2-containing sequences can be transforming when inappropriately regulated (10-15). The identification of physiologically relevant interactions of Crk proteins will undoubtedly aid in our understanding of the role of SH2 domains in regulation of cell growth. To gain insights into the mechanisms of Crk transformation, we have utilized the phosphorylated EGFR as a model substrate to character- ize the parameters of Crk binding and to compare this binding with that of other SH2 sequences. We extend this information to the interaction of Crk with p130, the major phosphotyrosine-containing protein that appears in CT10-transformed cells.

Using a microtiter binding assay, we have demonstrated that GST-fusion proteins containing Crk sequences and the N-terminal SH2 domain of GAP compete for similar phosphorylation sites on the EGFR in vitro. Thus, in competition binding assays, GST-GAPSHB[N] sequences and GST-Crk quantitatively competed (greater than 90%) the binding of each other with similar protein stoichiometry. Moreover, both GST-Crk and GST-GAPSH2[N] produced similar biphasic titration patterns in a solid phase binding assay, with an estimated 50% of maximal binding in the linear region occurring at approximately a lo-' M concentration of each protein. However, it is important to note that the binding of GST-Crk or GST-GAPSHB[N] did not completely saturate at high protein concentrations. Since the autophosphorylated EGFR contains at least four tyrosine phosphorylation sites (30), it is possible that these SH2-phosphotyrosine interactions represent a composite of individual binding affinities which may or may not be identical. The fact that we did not see complete saturation may indicate a low affinity binding site, perhaps analogous to that seen with receptor-ligand interactions (31). We are currently examining the binding of these proteins by Scatchard analysis to address this issue.

Despite the relatively high binding affinity of Crk to the phosphorylated EGFR in vitro, we have not detected the EGFR in Crk immunoprecipitates from CT10-transformed CEF. This may be due to the fact that, in vivo, such interactions are unstable or simply that the association is below the limit of detection in CEF. Consistent with this latter possibility, EGF stimulation of serum-starved A431 cells overexpressing v-Crk results in v-Crk association with the phosphorylated EGFR as demonstrated by co-immunoprecipita-

tion? Moran et al. (32) also demonstrated stzble association of TrpE-v-CrkSH2 with phosphorylated EGFR in vitro and, consistent with results presented here, found that the CrkSH2 domain and the N-terminal SH2 domain of GAP bound a similar set of phosphotyrosine-containing proteins from lysates of EGF-stimulated Rat-1 cells. Although stable association of intact GAP with the EGFR has not been demonstrated, GAP clearly becomes tyrosine-phosphorylated by EGF stimulation in vivo (33), and it is tempting to speculate that the biological activity of Crk might involve the competition of GAP binding sites intracellularly. However, the inability of GST-GAPSH2[N] to inhibit Crk binding to p130 demonstrates that each protein may also have unique cellular targets. It will be interesting to examine whether Crk can alter the phosphorylation status of GAP or compete the ability of GAP to complex with p62 and p190 in vivo (33). Similarly, it will be important to determine whether p21" becomes activated in Crk-overexpressing or transformed cells.

No differences in binding affinity of GST-c-Crk or GST-V- Crk toward the phosphorylated EGFR was noted in these in vitro studies. This observation is based upon three experimental findings. First, each protein exhibited a similar titration curve in the solid phase ELISA with anti-Crk antibodies. Second, each inhibited its cognate protein with similar affinity and stoichiometry and third, both were inhibited with similar characteristics by heterologous SH2-containing sequences. Moreover, both proteins bound equally well to denatured p130 on nitrocellulose strips. Both v-Crk and c-Crk have identical SH2 domains. However, v-Crk has viral-derived Gag sequences while c-Crk has an extra SH3 domain (Fig. 1). There- fore, it appears that, at least in vitro, the SH2 domain is the primary determinant in directing stable interactions with phosphotyrosine-containing proteins. This data are, however, curious in light of the different transforming activities of c- crk and v-crk in vivo. Cells overexpressing c-Crk, in contrast to cells overexpressing v-Crk, are not transformed in standard colony-forming assays nor are they tumorigenic when injected into syngeneic animals? At present, we cannot discount the possibility that subtle differences in the protein structures, specific post-translational modifications, or restricted subcel- lular locations contribute to these differences in vivo.

In light of the increased tyrosine phosphorylation observed in v-Crk-transformed cells, the information obtained from these in vitro studies may lend important insights into the biochemical phenotype of CT10-transformed cells. In this respect, the binding of SH2-containing proteins to phosphorylated EGFR or p130 appeared to be stable. Thus, under our assay conditions, inhibition of GST-c-Crk by GST- GAPSHP[N] was abolished by prebinding of GST-c-Crk to the EGFR. In addition, binding of SH2 proteins produced a significant protection of phosphorylated substrates from the action of cellular PTPase. Taken together with the fact that p47pw-crk co-immunoprecipitates with tyrosine-phosphorylated proteins (including p130) in vivo (13, 14 and Fig. 7), it is highly possible that binding and protection from PTPase contributes to the elevation of protein phosphotyrosine in Crk-transformed cells. Initially, these substrates would have to be phosphorylated by a cellular kinase(s) and the importance of Crk interactions with such kinase(s) is still uncertain. In this study, we did not see a direct activation of the EGFR by the binding of Crk or other heterologous SH2-containing polypeptides. Tyrosine kinase activity can be shown to immunoprecipitate with Crk in vivo (11,12), possibly indicating

J. E. Fajardo, H. Hanafusa, B. Margolis, and J. Schlessenger,

C. Reichman and H. Hanafusa, unpublished results. unpublished results.


that Crk may act in tram as a subunit of a tyrosine kinase, in a fashion analogous to the p85 subunit for phosphatidylinositol-3-kinase (34,35).

The observation that proteins with SH2 sequences can confer resistence to PTPase activity points to the importance of proper cellular control of phosphotyrosine-SH2 disassembly in uiuo. One potential mechanism could occur if phosphorylations of the SH2 protein resulted in a decreased affinity. Consistent with this interpretation are findings that tyrosine-phosphorylated PLC--, has a lower affinity for the PDGR than the nonphosphorylated form (36). In the case of v-Crk, tyrosine phosphorylations induced by EGFR in uitro does not appear to decrease the binding affinity of p47K4p-crk or GST-Crk fusion proteins toward the phosphorylated EGFR4 although we have not mapped the phosphorylation sites on Crk, Moreover, tyrosine phosphorylation accounts for only a minor fraction of the total phosphorylation of p47Kw-crk in vivo (13). The relationship between Crk phosphorylation and stable association with phosphorylated cellular proteins in uiuo is currently under investigation. Another less likely possibility is that SH2-phosphotyrosine interactions are targeted for proteolysis. Finally, it is possible that as yet undefined regulators of disassembly are required in uiuo. In this respect, it is intriguing that a tyrosine-specific phosphatase with SH2 homology has recently been identified and cloned (37).

Despite an absolute requirement for tyrosine phosphorylation in SH2 binding, only weak inhibition of Crk binding was observed with free phosphotyrosine (only 30% inhibition at 20 mM phosphotyrosine assayed by competition ELISA). These data are consistent with a model proposed by Mayer et al. (19) in which the phosphotyrosine moiety may direct an initial low affinity interaction with one or more conserved residues followed by specific interactions manifested by non- conserved sequences in the SH2. Consistent with this latter notion, several different mutations within the SH2 of Crk could abolish the ability of Crk to associate with phosphotyrosine-containing proteins (16). In this study, each of the GST-SH2-containing fusion proteins that bound to the phosphorylated EGFR in uitro displayed a different relative affinity. This clearly argues that different SH2 affinities toward phosphorylated substrates may be important in generating specificity.

We have utilized methodology to approximate the binding affinity of SH2-containing proteins to the phosphorylated EGFR. In addition, by competitive ELISA, we have been able to identify and quantify potential inhibitors that can cross- react with Crk. Not only can these approaches be utilized to compare the specificity of different SH2-containing proteins toward given phosphoproteins, but it should also be possible to map phosphorylation sites by competition with synthetic or enzymatically cleaved phosphopeptides. Furthermore, it should be possible to quantitatively define positive and neg- ative biological criteria that influence the extent of SH2 binding to phosphotyrosine-containing proteins. It is antici-

J. E. Fajardo, R. B. Birge, and H. Hanafusa, unpublished results.

pated that such assays will help establish models for the specific interactions of SH2-phosphotyrosine-containing proteins in vivo.

Acknowledgments-We thank Beatrice Knudson and David Stern- berg for helpful discussions on this manuscript. We are grateful to Charles Reichman for providing c-crk DNA and Haruo Saito for the purified leukocyte common antigen related phosphatase.

REFERENCES 1. Ullrich, A,, and Schlessinger, J. (1990) Cell 6 1 , 203-212 2. Kumjian, D. A., Wahl, M. I., Rhee, S. G., and Daniel, T. 0. (1989) Proc.

3. Meisenhelder, J., Suh, P. G., Rhee, S. G., and Hunter, T. (1989) CeU 67, Natl. Acad. SCI. U. S. A. 86,8232-8236

4. Kaplan, D. R., Morrison, D. K., Wong, G., McCormick, F., and Williams,

5. Kazlauskas, A., Ellis, C., Pawson, T., and Cooper, J. A. (1990) Science 247 ,

1109-1122

L. T. (1990) Cell 6 1 , 125-133

1 ~ 7 8 - 1 ~ 8 1 6. Kazlauskas, A. C., and Cooper, J. (1990) EMEO J. 9,3279-3286 7. Coughlin, S. R., Escobedo, J. A., and Williams, L. T. (1989) Science 2 4 3 ,

" ~ ~ . "

8. Pawson, T. (1988) Oncogene 3,491-495 9. Koch, C. A., Anderson, D., Moran, M. F., Ellis, C., and Pawson, T. (1991)

10. Cantley, L. C., Au er, K. R., Carpenter, C., Duckworth, B., Graziani, A.,

11. Mayer, B. J., Hamaguchi, M., and Hanafusa, H. (1988) Nature 3 3 2 , 272-

1191-1194

Scrence 262,666-674

Kapeller, R., anfsoltoff, S. (1991) Cell 64,281-302

076 12. M&&, B. J., Hamaguchi, M., and Hanafusa, H. (1988) Cold Spring Harbor

13. Mayer, B. J., and Hanafusa, H. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, Symp. Quant. Eiol. 63,907-914

9 I X Q - 0 C A 9

14. Matsuda, M., Mayer, B. J., Fukui, Y., and Hanafusa, H. (1990) Science

15. Matsuda, M., Mayer, B. J., and Hanafusa, H. (1991) Mol. Cell. Biol. 11 ,

16. Mayer, B. J., and Hanafusa H. (1990) J. Virol. 64,3581-3589 17. Hanafusa, H. (1969) Proc. h t l . Acad. sci. U. S. A. 63,318-325 18. Mar olis, B Rhee S. G Felder S. Mervic, M., Lyall, R., Levitzki, A.,

Ugrich, A:, Zilbe'rstein:' A., and' Slhlessinger, J. (1989) Cell 67, 1101- 11n7

"U" 1"s-

248,1537-1539

1607-1613

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29. 30.

31.

32.

33.

M a p , B. J., Jackson P. K., Etten, R. A. V., and Baltimore, D. (1992) Mol.

Trahey, M., Won, G., Halenbeck, R., Ruhinfeld, B., Martin, G. A., Ladner, M., Long, C. M. Crosier W. J. Watt, K., Koths, K., and McCormick, F. (1988) Science 242,169k17od

Stahl, M. L., Ferenz, C. R., Kelleher, K. L., Kriz, R. W., and Knopf, J. L. (1988) Nature 332,269-272

Martinez, R., Prevot, B. M., Bernards, A., and Baltimore, D. (1987) Science 237,411-415

Smith, D. B., and Johnson, K. S. (1988) Gene (Amst . ) 67,31-40 Tonks, N. K., Diltz, C. D., and Fischer, E. H. (1988) J. E d . Chem. 2 6 3 ,

A &" . ell Ezol. 1 2 , 609-618

6731-6737 Pi a, A , Wickremasinghe, R. G., Taheri, M. R., Yaxley, J. C., and Hoff-

Glass, D. B., Masaracchia, R. A., Feramisco, J. R., and Kemp, B. E. (1978)

Casnellie, J. E., Harrison, M. L., Pike, L. J., Hellstrom, K. E., and Krebs,

Towbin, H., Staehelin T., and Gordon, J. (1979) Proc. Natl. Acad. Sci.

Wang, J. Y. J. (1985) Mol. Cell. Eiol. 6, 3640-3643 Margolis, B. L., Lax, I., Kris, R., Dombalagian, M., Honegger, A. M., Howk,

R., Givol, D., Ullrich, A,, and Schlessinger, J. (1989) J. Eiol. Chem. 264,

Hempstead, B. L., Zanca, D. M., Kaplan, D. R., Parada, L. F., and Chao,

Moran, M. F., Koch, C. A., Anderson, D., Ellis, C:, En land, L., Martin, G. M. V. (1991) Nature 360,678-683

Ellis, C., Moran, M., McCormick, F., and Pawson, T. (1989) Nature 343 , S., and Pawson, T. (1990) Proc. Natl. Acad. SCI. U. 8. A. 87,8622-8626 ~ T - ~ Q I

. ~~

%rand, A. V. (1985) Exp. Cell Res. 169,103-112

Anal. Biochem. 87,566-575

E. G. (1982) Proc. Natl. Acad. Sci. U. S. A . 79,282-286

U. S. A. 76 ,43504 i54

10667-10671

34. Escobedo, J. A., Navankasattusas, S., Kavanaugh, W. M., Milfay, D., Fried,

35. Otsu, M., Hiles, I., Gout I., Fry, M. J., Larrea, F. R., Panayotou, G., V. A., and Williams, L. T. (1991) Cell 66,75-82

Thom son, A., Dhand, R., Hsuan, J., Tott , N Smith, A. D., Morgan S. J., tourtneidge, S. A., Parker, P. J., and &a&ield, M. D. (1991) Celi

36. Margolis, B., Li, N., Koch, A., Mohammadi, M., Hurwitz, D. R., Zilberstein, 66,91-104

A,, Ullrich, A,, Pawson, T., and Schlessmger, J. (1990) EMEO J. 9 , 4380-4375

Y l I -"A

37. Sheu, S. H., Bastieu, L., Posuer, B. L., and Chretien, P. (1991) Nature -. . . . - . - 362,736-739

Documents

THE JOURNAL OF Vol. 267, No. 15, of May 25, pp. 10588 ... · THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 hy The American Society for Biochemistry and Molecular Biology, Inc. Vol. 267,