Int. J. Cancer: 70, 208–213 (1997) r 1997 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer MULTIPLE Grb2–PROTEIN COMPLEXES IN HUMAN CANCER CELLS Lakshmi SASTRY*, Tin CAO and C. Richter KING Department of Biochemistry, Lombardi Cancer Center, Georgetown University Medical Center, Washington DC 20007, USA Grb2 is an SH2/SH3 domain-containing adaptor protein that links receptor tyrosine kinases to the ras signaling pathway. The Grb2-SH2 domain binds phosphotyrosine sequences on activated tyrosine kinases, and one target of the SH3 domains is the ras-nucleotide-exchange factor Sos1. We have examined Grb2-protein interactions in human cancer cells that over-express the receptor tyrosine kinase erbB2. Our results show that the 2 Grb2-SH3 domains complex with Sos1, dynamin and at least 4 other proteins (p228, p140, p55, p28) in these cells. The 2 Grb2-SH3 domains bind these proteins differently, with the N-terminal SH3 domain interacting preferentially with p228, Sos1, p140 and dynamin. The C-terminal SH3 domain has higher affinity toward p28. The Grb2-SH3 domain interactions appear to be similar in erbB2 over-expressing breast, ovarian and lung cancer cells. Also, the major tyrosine-phosphorylated proteins that associate with Grb2 in erbB2 over-expressing cancer cells appear to be erbB2 and Shc. The multiple Grb2-SH3 domain interactions in these cells may mediate novel cellular functions. Int. J. Cancer, 70:208–213, 1997. r 1997 Wiley-Liss, Inc. The adaptor protein Grb2 plays an important role in signaling by receptor tyrosine kinases. Grb2 is a ubiquitously expressed 25 kDa protein composed of a single SH2 domain flanked by aminoterminal and carboxy-terminal SH3 domains (NSH3 and CSH3, respectively) (Lowenstein et al., 1992). One function of Grb2 is to link receptor tyrosine kinases to the ras signaling pathway (Gale et al., 1993; Egan et al., 1993; Buday and Downward, 1993; Rozakis-Adcock et al., 1993). The Grb2-SH2 domain binds several tyrosine-phosphorylated receptor type molecules, including erbB2 (Janes et al., 1994), EGFR (Lowenstein et al., 1992) and PDGFR (Arvidsson et al., 1994), as well as other phosphotyrosinecontaining proteins like Shc (Rozakis-Adcock et al., 1992), insulin receptor substrate 1 (Skolnik et al., 1993) and focal adhesion kinase (Schlaepfer et al., 1994). Grb2 binds the ras-guanine nucleotide exchange factor Sos1 constitutively through its SH3 domains (Chardin et al., 1993; Li et al., 1993). The Grb2-SH3 domainbinding sequences have been localized to a proline-rich region in the C-terminus of Sos1 (Chardin et al., 1993; Li et al., 1993). Upon growth factor stimulation, the Grb2-Sos1 complex binds to activated receptors (Gale et al., 1993; Egan et al., 1993; Buday and Downward, 1993; Rozakis-Adcock et al., 1993; Lowenstein et al., 1992) or to an intermediate protein, Shc (Rozakis-Adcock et al., 1992). It is proposed that binding of the Grb2-Sos1 complex to phosphorylated receptors activates the ras pathway by bringing Sos1 to the proximity of the membrane-bound ras (McCormick, 1993). This is supported by the observation that targeting Sos1 to the plasma membrane is sufficient for the activation of the ras pathway (Quilliam et al., 1994). Alternatively, Grb2 may activate Sos1 by relieving the inhibition imposed by the C-terminus of Sos1 on its catalytic activity (Wang, W., et al., 1995). Grb2 may have other cellular functions as well. Micro-injection studies with anti-Grb2 antibodies in rat kidney cells suggest that Grb2 may play a role in signaling from receptor tyrosine kinases to the small GTP-binding protein Rac (Matuoka et al., 1993). Also, the Grb2-SH3 domains have been shown to bind several proteins besides Sos1, including dynamin (Miki et al., 1994), synapsin (McPherson et al., 1994), p85a (Wang, J., et al., 1995), Sos2 (Yang et al., 1995b), Cbl (Odai et al., 1995), Vav (Hanazono et al., 1995; Ye and Baltimore, 1994; Ramos-Morales et al., 1995) and dystroglycan (Yang et al., 1995a). The exact cellular functions of these Grb2/Grb2 SH3 domain-binding protein (GSBP) complexes are not known. However, Cbl is implicated in the signal transduction of erythropoietin and granulocyte macrophage colony-stimulating factor in hematopoietic cells (Odai et al., 1995), and the GTPase dynamin has been shown to be important in receptor endocytosis (van der Bliek et al., 1993). It is therefore likely that Grb2 participates in signaling pathways other than ras. Grb2 plays an important role in human cancers over-expressing the receptor tyrosine kinase p185erbB2. In breast cancer cells over-expressing the erbB2 tyrosine kinase, the ras signaling pathway is activated, and a higher amount of the Grb2–Sos1 complex is associated with erbB2 (Janes et al., 1994). Moreover, Grb2 is over-expressed in breast cancer cells and over-expression correlates with high expression of erbB2 (Daly et al., 1994). These findings suggest an important role for Grb2 in erbB2-induced malignancies. We are interested in investigating the signals carried by Grb2protein complexes in erbB2 over-expressing cancer cells. For this purpose, we examined Grb2-protein complexes in these cells by immunoprecipitation of the complexes followed by far-Western blot analysis using isolated Grb2-SH3 domains. Our results show that Grb2 can associate with Sos1, dynamin and at least 4 other proteins (p228, p140, p55, p28) through SH3 domain interactions. These proteins are bound differentially by the N-terminal and C-terminal SH3 domains. Also, the Grb2-SH3 domain interactions are similar in erbB2 over-expressing breast, ovarian and lung cancer cells and in erbB2-transformed and -untransformed fibroblasts. Our findings support the view that Grb2 may have multiple functions mediated by distinct SH3-binding proteins. MATERIAL AND METHODS Antibodies Polyclonal antibodies against Grb2 (C-23) and monoclonal antibodies (MAbs) against GST (SC-138) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-Shc (S14620), anti-Sos1 (S15520) and anti-dynamin (D25520) MAbs were obtained from Transduction Labs (Lexington, KY). Anti-erbB2 MAbs (OP15) were from Oncogene Science (Uniondale, NY). Expression and purification of GST-Grb2 fusion proteins GST fusion constructs of whole Grb2, N-terminal SH3 (aa 1–59) and C-terminal SH3 (aa 159–217) domains were expressed in a modified pGEX 2T-2 expression vector (Pharmacia, Uppsala, Sweden). For protein expression analysis, fresh transformants were grown in 3 ml LB with 100 µg/ml ampicillin at 37°C to an OD600 of 0.6–1.0. Expression was induced by adding IPTG to a final concentration of 1 mM for 1 hr; protein was analyzed on an 8-16% SDS-PAGE gel. GST-fusion proteins were purified using glutathione beads according to the manufacturer’s instructions (Glutathione Sepharose 4B Instructions, Pharmacia). Purified proteins were dialyzed against PBS and stored in aliquots at 270°C. Cell lines MDA-MB-453, SK-BR-3, MCF-7, SK-OV-3, Calu-3, NIH 3T3 and NIH 3T3/erbB2 cells were grown in IMEM supplemented with 10% FBS (Biofluids, Gaithersburg, MD) and 1 mM glutamine (GIBCO, Gaithersburg, MD) in 150 3 25 mm dishes. Twenty-four *Correspondence to: E508, Research Building, 3970 Reservoir Rd., NW, Washington, DC 20007, USA. Fax: (202) 687-7505. Received 19 June 1996; revised 4 October 1996. Grb2-BINDING PROTEINS 209 hours before harvesting, cells were washed twice with sterile PBS (GIBCO) and grown in IMEM supplemented with 10% FBS and 1 mM glutamine or IMEM supplemented with 1% FBS and 1 mM glutamine (for starving cells) at 37°C, in a 5% CO2 incubator. Immunoprecipitation, far-Western blotting and Western blotting For immunoprecipitation experiments, harvested cells were rinsed twice with cold PBS and lysed with the addition of 0.5 ml cold immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 50 mM Hepes [pH 7.5], 10% glycerol, 1.5 mM MgCl2, 1 mM EGTA, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 1 mM PMSF, 200 µM NaVO4 and 100 mM NaF) for 10 min at 4°C. Lysed cells were centrifuged at 4°C for 10 min at 4,850g, and the supernatant, the total native cell lysate, was collected. Total protein concentration of the cell lysate was determined by the BCA method (BioRad, Hercules, CA). For immunoprecipitations, 12.5 µg of rabbit polyclonal IgG against Grb2 or a control rabbit IgG were added to 5 mg of total native cell lysate and incubated with rotation at 4°C for 2 hr. Protein A-agarose (100 µl; Santa Cruz Biotechnologies) was then added and the mixture incubated with rotation at 4°C overnight. Protein A beads were recovered by centrifugation at 1,485g for 1 min. After 4 washes with cold immunoprecipitation buffer, 60 µl of 23 sample buffer (ISS, Natico, MA) were added to the agarose beads, boiled for 8 min and stored at 220°C. For far-Western blots, immunoprecipitated material was resolved on SDS-PAGE gels; blotted onto nitrocellulose filters; and analyzed using 0.5 µg/ml of GST, GST-NSH3 or GST-CSH3 proteins. Blots were blocked in 5% BSA in TBST (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% Tween 20) overnight and then in 3 mM reduced glutathione (Sigma, St. Louis, MO) in 5% BSA-TBST for 1 hr at room temperature (RT). Prior to incubation of the blots in GST, GST-NSH3 or GST-CSH3 proteins, proteins were preincubated in 3 mM reduced glutathione for 1 hr at RT. Following this, blots were incubated for 1 hr at RT with the appropriate fusion protein, washed 43 with TBST and incubated for 1 hr at RT with anti-GST MAbs in 5% BSA-TBST at 0.5 µg/ml. Blots were then washed with TBST and incubated with sheep anti-mouse horseradish peroxidase (Amersham, Arlington Heights, IL) at 1:5,000 dilution in 5% BSA-TBST for 1 hr at RT. After extensive washing in TBST, blots were visualized with enhanced chemiluminescence (ECL; Amersham). For direct Western blotting, Grb2 or control immunoprecipitates were electrophoresed, transferred to nitrocellulose membranes, probed with MAbs against Sos1 (1/250 dilution), dynamin (1/1,000 dilution), Shc (1/1,000 dilution) or erbB2 (1/1,000 dilution) and detected using sheep anti-mouse IgG-HRP conjugates and chemiluminescence. In co-precipitation experiments, Grb2 immunoprecipitates were probed with anti-Grb2 MAbs (1/1,000 dilution; Transduction Labs). RESULTS Grb2-protein complexes in erbB2 over-expressing human cancer cells To investigate Grb2-SH3 domain specificities in cancer cells, we first immunoprecipitated Grb2 under conditions where the SH3 domain complexes are intact and detected co-precipitating proteins by far-Western analysis using the GST fusion proteins of Grb2-SH3 domains (IP-far-Western). Proteins detected in this analysis most likely bind Grb2 through the SH3 domains. For this analysis, it was important to choose an antibody that specifically immunoprecipitated Grb2 and had minimal cross-reactivity with other SH2/SH3 domain-containing proteins. The specificity of the antibody used in these studies was determined by first immunoprecipitating Grb2 and blotting the precipitated material with the same immunoprecipitating antibody (Fig. 1). These results show that the antibody is specific for Grb2 and has little cross-reactivity with other cellular proteins under our experimental conditions. FIGURE 1 – Specificity of the anti-Grb2 antibody. Lysates from erbB2-over-expressing (MDA-MB-453, SK-BR-3) and non-overexpressing (MCF-7) breast cancer cells were prepared, and Grb2 was immunoprecipitated from each cell lysate (5 mg) using anti-Grb2 antibodies. Immunoprecipitated material was resolved on 8–16% SDS-PAGE gels and analyzed by Western blotting with anti-Grb2 antibodies. Anti-Grb2 antibodies used for immunoblotting were the same as those used for immunoprecipitation. Blots were detected with ECL. Arrows indicate Grb2, heavy chain (HC) and light chain (LC); m.w. markers are shown in kDa. The Grb2-SH3 domain complexes were first examined in human breast cancer cells over-expressing the erbB2 receptor tyrosine kinase (SK-BR-3, MDA-MB-453) and non-over-expressing cells (MCF-7) under serum-treated and serum-starved conditions. IP-farWestern analysis shows that the Grb2-SH3 domains bind 6 proteins (p228, p170, p140, p125, p55, p28) that co-precipitate with Grb2 in these cells (Fig. 2). This binding is specific for SH3 domains as GST shows no reactivity when used to probe the blots of Grb2 immunoprecipitates (Fig. 2, GST). The SH3 domains show minimal reactivity with proteins immunoprecipitated with control antibodies, indicating that far-Western blot analysis detects proteins associating with Grb2 specifically (Fig. 2, lane NS). The specificity for Grb2-binding proteins is further supported by the fact that far-Western blot analysis of Sos1 immunoprecipitates with Grb2-SH3 domains detects only Sos1 (data not shown). Interestingly, the Grb2-NSH3 domain shows higher affinity toward most proteins detected in the far-Western analysis, including p228, p170, p140 and p125. The CSH3 domain appears to have preferential interaction with p28. Both domains seem to interact with the 55 kDa protein. The Grb2-binding proteins detected in this assay are similar in erbB2-over-expressing (MDA-MB-453, SK-BR-3) and -non- 210 SASTRY ET AL. FIGURE 2 – Grb2-protein complexes in breast cancer cells. Lysates were prepared from serum-treated and serum-starved erbB2-over-expressing (MDA-MB-453, SK-BR-3) and non-overexpressing (MCF-7) breast cancer cells. Grb2-protein complexes were immunoprecipitated from each lysate (5 mg) with anti-Grb2 antibodies. Immunoprecipitation was also carried out with a control antibody from MDA-MB-453 cell lysates (NS). Immunoprecipitated material was resolved on 8–16% SDS-PAGE gels and analyzed by far-Western blotting with GST, GST-NSH3 or GST-CSH3 proteins. Grb2-SH3 domain-binding proteins (p228, p170, p140, p125, p55, p28) are shown by arrows. Sos1-c, the purified C-terminal region of Sos1, binds the Grb2-NSH3 domain (Sastry et al., 1995) and is included as a positive control for far-Western blotting; m.w. markers are shown in kDa. FIGURE 3 – Grb2-protein complexes in erbB2-over-expressing cancer cells. Lysates were prepared from serum-treated erbB2-over-expressing ovarian (SK-OV-3), lung (Calu-3) and breast (MDA-MB-453, SK-BR-3) cancer cells and erbB2-transformed (N/erbB2) and untransformed fibroblasts (NIH 3T3). Grb2-protein complexes were immunoprecipitated from each lysate (5 mg) with anti-Grb2 antibodies and analyzed by far-Western blotting with GST, GST-NSH3 or GST-CSH3 proteins. Grb2-SH3 domain-binding proteins (p228, p170, p140, p125, p55, p28) are shown by arrows; m.w. markers are shown in kDa. NS is the immunoprecipitation with a control antibody from MDA-MB-453 cell lysates, and Sos1-c is the positive control for far-Western blotting as described in Figure 2. overexpressing (MCF-7) cells (Fig. 2). The assay also shows that Grb2 binds similar m.w. proteins in starved and serum-treated cells, though there was an increase in the amount of Grb2-binding proteins in serum-treated MDA-MB-453 and MCF-7 cells (Fig. 2). No quantitative differences are seen in Grb2-binding proteins from starved or serum-treated SK-BR-3 cells. The quantitative increase in Grb2-binding proteins in MDA-MB-453 and MCF-7 cells is most likely due to a higher amount of Grb2 in these cells upon serum treatment (data not shown). The receptor tyrosine kinase erbB2 is also over-expressed in certain lung and ovarian cancer cell lines and primary tumors. To investigate if there are any tissue-specific GSBPs, we compared Grb2-associating proteins in erbB2 over-expressing breast (SKBR-3, MDA-MB 453), ovarian (SK-OV-3) and lung (Calu-3) cancer cells by IP-far-Western analysis (Fig. 3). This comparison shows that the major proteins associated with Grb2-SH3 domains (p228, p170, p140, p125, p55, p28) are similar in the breast, lung and ovarian cancer cells tested (Fig. 3). A minor band detected at approx. 160 kDa in Grb2 immunoprecipitates from all cells may be a degradation product as this is not seen reproducibly. A 46 kDa band seen in all cells does not appear to be specifically associated with Grb2 as this is present in the control antibody immunoprecipitates as well (Fig. 3, lane NS). Also, untransformed and erbB2transformed NIH 3T3 (N/erbB2) cells show similar Grb2-SH3 domain-associating proteins (Fig. 3). In these cells, a protein slightly smaller than 228 kDa is observed. It is likely that this is the p228 protein observed in human cancer cells, and the slightly smaller size may be due to a species difference. The identity of Grb2-binding proteins To examine whether any of the Grb2-SH3 domain-associating proteins in our study correspond to previously characterized Grb2-binding proteins, we carried out co-precipitation experiments. For these experiments, Grb2 immunoprecipitates from erbB2-over-expressing SK-BR-3 cells were probed with antibodies against known GSBPs. As shown in Figure 4, probing with the anti-Sos1 antibodies indicates that the 170 kDa band detected in the far-Western blot analysis is Sos1. The 125 kDa protein was identified as dynamin by similar co-precipitation experiments using anti-dynamin antibodies (Fig. 4). The identity of the 140 kDa Grb2-BINDING PROTEINS 211 MCF7 cells that do not over-express erbB2, the predominant phosphoproteins in Grb2 immunoprecipitates appear to be Shc and a 80 kDa protein that does not react with Shc antibodies (Fig. 5b). The 80 kDa protein also associates with Grb2 in MDA-MB-453 cells. DISCUSSION FIGURE 4 – Identity of Grb2-binding proteins. Grb2 was immunoprecipitated from lysates of serum-treated SK-BR-3 cells with anti-Grb2 antibodies (I) or control antibodies (NA), electrophoresed and blotted onto nitrocellulose filters. Blots were probed with anti-Grb2 (a), anti-Sos1 (b) or anti-dynamin (c) antibodies and detected with sheepanti-mouse-HRP conjugates using ECL. Grb2, Sos1 and dynamin are indicated by arrows. The 28 kDa protein detected in the far-Western studies has a different electrophoretic mobility than Grb2 and is unlikely to be Grb2 itself; m.w. markers are shown in kDa. protein is unknown, though a similar m.w. protein was found to complex with Grb2 in hematopoietic cells (Pelicci et al., 1995) and in synaptic fractions (McPherson et al., 1994). A 145 kDa protein capable of binding Grb2-SH3 domains has been identified as synaptojanin in brain tissue (McPherson et al., 1996). However, probing Grb2 immunoprecipitates with anti-synaptojanin antibodies showed no reactivity, indicating that the 140 kDa protein seen in our studies is unlikely to be synaptojanin (McPherson, personal communication). Previous studies indicate that Grb2 can associate with Cbl (Odai et al., 1995), Vav (Hanazono et al., 1995; Ramos-Morales et al., 1994; Ramos-Morales et al., 1995) and p85a (Wang, J., et al., 1995) constitutively in hematopoietic cells and in mouse fibroblasts, respectively. Antibodies against Cbl and p85a showed no reactivity with Grb2 IPs, though both were expressed in all of the cells that we tested (data not shown). Vav is expressed exclusively in cells of hematopoietic origin and is not expected to associate with Grb2 in the cells being examined. The identity of p228, p55 and p28 is currently not known, and they have previously not been shown to associate with Grb2-SH3 domains. The 28 kDa protein is unlikely to be Grb2 itself as it has a slightly different electrophoretic mobility than Grb2 and is recognized by Grb2-SH3 domains. Phosphotyrosine proteins associated with Grb2 in erbB2-over-expressing cancer cells To investigate the major tyrosine-phosphorylated proteins associated with Grb2 in erbB2-over-expressing cancer cells, we conducted direct Western blotting of Grb2 immunoprecipitates with anti-phosphotyrosine antibodies (Fig. 5a). This analysis shows that there are 3 major phosphoproteins (p200, p52, p46) associated with Grb2 in erbB2-over-expressing breast and ovarian cancer cells. Longer exposure shows the presence of these bands in erbB2-overexpressing lung cancer cells (Calu-3) as well. Immunoblotting with anti-erbB2 and anti-Shc antibodies identified the p200 and p52/p47 proteins as erbB2 and Shc, respectively (Fig. 5b). The slightly higher m.w. of the erbB2 protein in these experiments is due to the use of gradient gels. This study suggests that in erbB2-overexpressing cancer cells, Grb2 can associate with both erbB2 and Shc (Fig. 5b). This association is most likely through the SH2 domain, as previously described (Rozakis-Adcock et al., 1992; Daly et al., 1994). None of the GSBPs observed in our study were tyrosine-phosphorylated under these experimental conditions. In The major function of Grb2-SH3 domains is thought to be association with the ras nucleotide exchange factor Sos1, thus linking tyrosine kinase activation to the ras signaling pathway (Lowenstein et al., 1992; Gale et al., 1993; Egan et al., 1993; Buday and Downward, 1993; Rozakis-Adcock et al., 1993). However, Grb2 may have other cellular functions as it has been shown to bind a number of proteins, including dynamin, synapsin, Cbl and Vav. Our results show that Grb2 can associate with at least 4 other proteins (p228, p140, p55, p28) besides Sos1 and dynamin, through SH3 domain interactions (Fig. 6). These proteins coprecipitate with Grb2 and interact directly with the Grb2-SH3 domains, indicating the existence of at least 6 different pools of Grb2-protein complexes in cells, including Grb2–Sos1 and Grb2– dynamin complexes. Each of these Grb2 complexes may mediate a different function. Our results, therefore, strongly support the suggestion that Grb2 has other cellular functions in addition to linking receptor tyrosine kinases to the ras signaling pathway (McPherson et al., 1994; Miki et al., 1994; Odai et al., 1995). The Grb2–GSBP complexes detected in our study are constitutively formed as they are similar in serum-starved and serumtreated cells. This is expected of Grb2–Sos1 (Rozakis-Adcock et al., 1993; Egan et al., 1993; Buday and Downward, 1993) and Grb2–dynamin (Miki et al., 1994) interactions and is consistent with the fact that most reported Grb2–SH3 domain interactions are constitutive and are not altered in response to cell growth conditions (Miki et al., 1994; Odai et al., 1995; Ramos-Morales et al., 1994). In addition, near identical Grb2–GSBP complexes are seen in erbB2-transformed and untransformed NIH 3T3 cells, indicating that a powerful transforming signal does not alter the state of the complexes. Similar Grb2–GSBP complexes are seen in fibroblasts and in tumor epithelial cells derived from different tissues, suggesting that the complexes may have important cellular functions and that they are not unique to transformed cells. Our results show that the 2 Grb2-SH3 domains have different binding specificities. The NSH3 domain shows higher affinity toward most proteins seen in the far-Western blot analysis, including p228, Sos1, p140 and dynamin. The CSH3 domain shows minimal interaction with all of these proteins and preferentially interacts with p28. Only one protein, p55, appears to have similar affinity toward both SH3 domains. These results suggest that the NSH3 and CSH3 domains of Grb2 may have different functions. This is supported by studies on genetic mutations in these domains in the Caenorhabditis elegans Grb2 homologue (Sem-5), which lead to different phenotypes (Clark et al., 1992). We and others have previously shown that the Grb2–Sos1 interaction is primarily mediated by the NSH3 domain of Grb2 (Reif et al., 1994; Sastry et al., 1995). The CSH3 domain has minimal reactivity toward Sos1, and its cellular function is not completely clear. Dominant negative mutants of Grb2 lacking the NSH3 or CSH3 domain have different effects on the transformed phenotype caused by mutation-activated neu and on p185-mediated ras activation. The NSH3 domain deletion has a much stronger effect than the CSH3 deletion (Xie et al., 1995). These results are consistent with the fact that Grb2-Sos1 binding is mediated by the NSH3 domain in cells. The authors also suggest the existence of multiple routes to ras activation based on the observation that the Grb2 dominant negative mutants do not completely suppress ras activation. Our results identify proteins that may participate in such pathways. 212 SASTRY ET AL. FIGURE 5 – Phosphotyrosine proteins associated with Grb2 in erbB2-over-expressing cancer cells. Lysates were prepared from serum-treated erbB2-over-expressing ovarian (SK-OV-3), lung (Calu-3) and breast (MDA-MB-453, SK-BR-3) cancer cells and non-overexpressing breast cancer cells (MCF-7). Grb2 and Grb2-binding proteins were immunoprecipitated from each lysate (5 mg) with anti-Grb2 antibodies and analyzed by Western blotting with (a) anti-phosphotyrosine antibodies and (b) anti-erbB2 or anti-Shc antibodies. In (a), the major phosphotyrosine proteins in Grb2 immunoprecipitates are indicated by arrows; m.w. markers are shown in kDa. In (b), erbB2 (top panel) and the 2 isoforms of Shc (bottom panel) are indicated. Longer exposures clearly show the p46 isoform of Shc in MCF-7 cells. In (b), NS represents an immunoprecipitation with control antibodies from the 453 cell lysate. FIGURE 6 – Schematic representation of Grb2-protein complexes in erbB2-over-expressing human cancer cells. The Grb2-SH3 domainbinding proteins and the major SH2 domain-binding proteins in erbB2-over-expressing human cancer cells are shown. v-src-transformed or fibronectin-treated NIH 3T3 cells (Schlaepfer et al., 1994) and Gab1 (Holgado-Madruga et al., 1996) in insulin/EGF-stimulated A431 cells. However, under our experimental conditions, Grb2 seems to associate mostly with erbB2 and Shc in erbB2-over-expressing cancer cells. This strongly suggests that in these cells Grb2 is primarily involved in transmitting signals from erbB2 and Shc. Similarly, in non-erbB2-over-expressing breast cancer cells (MCF-7), Shc is a major up-stream target of Grb2. Phosphorylation of Shc in response to growth factors and subsequent recruitment of the Grb2–Sos1 complex is a major mechanism for the activation of ras signaling (Rozakis-Adcock et al., 1992). Our results add to the evidence for the importance of the association of Grb2 with erbB2 or Shc in cancer cells. Our findings reveal the existence of multiple Grb2-protein interactions in human cancer cells and substantiate the idea that Grb2 may have several cellular functions. The elucidation of novel Grb2 functions may require the identification and characterization of the Grb2-SH3-binding proteins described. ACKNOWLEDGEMENTS The major tyrosine-phosphorylated proteins associated with Grb2 in erbB2-over-expressing cancer cells appear to be erbB2 and Shc (Fig. 6). In these cells, none of the GSBPs are tyrosinephosphorylated, indicating that they are not substrates for tyrosine kinases. Grb2 has previously been reported to bind other tyrosinephosphorylated proteins, including focal adhesion kinase (FAK) in We thank Dr. P. McPherson (McGill University, Canada) for probing Grb2 immunoprecipitation blots with anti-synaptojanin antibodies. We also thank Drs. M. Lippman, D. Goldstein and A. Dent for critically reviewing the manuscript and Drs. K. Rosenberg and M. Hizaji for helpful suggestions and comments. This research was supported in part by a generous gift from Mr. William Baker. REFERENCES ARVIDSSON, A.K., RUPP, E., NANBERG, E., DOWNWARD, J., RONNSTRAND, L., WENNSTROM, S., SCHLESSINGER, J., HELDIN, C.H. and CLAESSON-WELSH, L., Tyr-716 in the platelet-derived growth factor beta-receptor kinase insert is involved in Grb2 binding and Ras activation. Mol. cell. Biol., 14, 6715–6726 (1994). BUDAY, L. and DOWNWARD, J., Epidermal growth factor regulates p21ras through the formation of a complex of receptor, Grb2 adapter protein, and Sos nucleotide exchange factor. Cell, 73, 611–620 (1993). CHARDIN, P., CAMONIS, J.H., GALE, N.W., VAN AELST, L., SCHLESSINGER, J., WIGLER, M.H. and BAR-SAGI, D., Human Sos1: a guanine nucleotide exchange factor for Ras that binds to Grb2. Science, 260, 1338–1343 (1993). CLARK, S.G., STERN, M.J. and HORVITZ, H.R., C. elegans cell signaling gene sem-5 encodes a protein with SH2 and SH3 domains. Nature (Lond.), 356, 340–344 (1992). DALY, R.J., BINDER, M.D. and SUTHERLAND, R.L., Overexpression of the Grb2-BINDING PROTEINS Grb2 gene in human breast cancer cell lines. Oncogene, 9, 2723–2727 (1994). EGAN, S.E., GIDDINGS, B.W., BROOKS, M.W., BUDAY, L., SIZELAND, A.M. and WEINBERG, R.A., Association of Sos Ras exchange protein with Grb2 is implicated in tyrosine kinase signal transduction and transformation [see comments]. Nature (Lond.), 363, 45–51 (1993). GALE, N.W., KAPLAN, S., LOWENSTEIN, E.J., SCHLESSINGER, J. and BARSAGI, D., Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on Ras [see comments]. Nature (Lond.), 363, 88–92 (1993). HANAZONO, Y., SASAKI, K., ODAI, H., MIMURA, T., MITANI, K., YAZAKI, Y. and HIRAI, H., Tyrosine phosphorylation of the proto-oncogene product Vav and its association with the adapter Grb2/Ash in a human leukemia cell line UT-7. Jpn. J. Cancer Res., 86, 336–341 (1995). HOLGADO-MADRUGA, M., EMLET, D.R., MOSCATELLO, D.K., GODWIN, A.K. and WONG, A.J., A Grb2 associated docking protein in EGF and insulin stimulated receptor signaling. Nature (Lond.), 379, 560–564 (1996). JANES, P.W., DALY, R.J., DE FAZIO, A. and SUTHERLAND, R.L., Activation of the Ras signalling pathway in human breast cancer cells overexpressing erbB-2. Oncogene, 9, 3601–3608 (1994). LI, N., BATZER, A., DALY, R., YAJNIK, V., SKOLNIK, E., CHARDIN, P., BAR-SAGI, D., MARGOLIS, B. and SCHLESSINGER, J., Guanine-nucleotidereleasing factor hSos1 binds to Grb2 and links receptor tyrosine kinases to Ras signalling [see comments]. Nature (Lond.), 363, 85–88 (1993). LOWENSTEIN, E.J., DALY, R.J., BATZER, A.G., LI, W., MARGOLIS, B., LAMMERS, R., ULLRICH, A., SKOLNIK, E.Y., BAR-SAGI, D. and SCHLESSINGER, J., The SH2 and SH3 domain-containing protein GRB2 links receptor tyrosine kinases to ras signalling. Cell, 70, 431–442 (1992). MATUOKA, K., SHIBASAKI, F., SHIBATA, M. and TAKENAWA, T., Ash/Grb-2, a SH2/SH3-containing protein, couples to signaling for mitogenesis and cytoskeletal reorganization by EGF and PDGF. EMBO J., 12, 3467–3473 (1993). MCCORMICK, F., Signal transduction. How receptors turn Ras on [news; comment]. Nature (Lond.), 363, 15–16 (1993). MCPHERSON, P.S., CZERNIK, A.J., CHILCOTE, T.J., ONOFRI, F., BENFENATI, F., GREENGARD, P., SCHLESSINGER, J. and DE CAMILLI, P., Interaction of Grb2 via its Src homology 3 domains with synaptic proteins including synapsin I. Proc. nat. Acad. Sci. (Wash.), 91, 6486–6490 (1994). MCPHERSON, P.S., GARCIA, E.P., SLEPNEV, V.I., DAVID, C., ZHANG, X., GRABS, D., SOSSIN, W., BAUERFEIND, R., NEMOTO, Y. and DE CAMILLI, P., A presynaptic inositol-5-phosphatase. Nature (Lond.), 379, 353 (1996). MIKI, H., MIURA, K., MATUOKA, K., NAKATA, T., HIROKAWA, N., ORITA, S., KAIBUCHI, K., TAKAI, Y. and TAKENAWA, T., Association of Ash/Grb-2 with dynamin through the Src homology 3 domain. J. biol. Chem., 269, 5489–5492 (1994). ODAI, H., SASAKI, K., IWAMATSU, A., HANAZONO, Y., TANAKA, T., MITANI, K., YAZAKI, Y. and HIRAI, H., The proto-oncogene product c-Cbl becomes tyrosine phosphorylated by stimulation with GM-CSF or Epo and constitutively binds to the SH3 domain of Grb2/Ash in human hematopoietic cells. J. biol. Chem., 270, 10800–10805 (1995). PELICCI, G., LANFRANCONE, L., SALCINI, A.E., ROMANO, A., MELE, S., GRAZIA BORRELLO, M., SEGATTO, O., DI FIORE, P.P. and PELICCI, P.G., Constitutive phosphorylation of Shc proteins in human tumors. Oncogene, 11, 899–907 (1995). QUILLIAM, L.A., HUFF, S.Y., RABUN, K.M., WEI, W., PARK, W., BROEK, D. and DER, C.J., Membrane-targeting potentiates guanine nucleotide ex- 213 change factor CDC25 and SOS1 activation of Ras transforming activity. Proc. nat. Acad. Sci. (Wash.), 91, 8512–8516 (1994). RAMOS-MORALES, F., DRUKER, B.J. and FISCHER, S., Vav binds to several SH2/SH3 containing proteins in activated lymphocytes. Oncogene, 9, 1917–1923 (1994). RAMOS-MORALES, F., ROMERO, F., SCHWEIGHOFFER, F., BISMUTH, G., CAMONIS, J., TORTOLERO, M. and FISCHER, S., The proline-rich region of Vav binds to Grb2 and Grb3-3. Oncogene, 11, 1665–1669 (1995). REIF, K., BUDAY, L., DOWNWARD, J. and CANTRELL, D.A., SH3 domains of the adapter molecule Grb2 complex with two proteins in T cells: the guanine nucleotide exchange protein Sos and a 75-kDa protein that is a substrate for T cell antigen receptor-activated tyrosine kinases. J. biol. Chem., 269, 14081–14087 (1994). ROZAKIS-ADCOCK, M., FERNLEY, R., WADE, J., PAWSON, T. and BOWTELL, D., The SH2 and SH3 domains of mammalian Grb2 couple the EGF receptor to the Ras activator mSos1 [see comments]. Nature (Lond.), 363, 83–85 (1993). ROZAKIS-ADCOCK, M., MCGLADE, J., MBAMALU, G., PELICCI, G., DALY, R., LI, W., BATZER, A., THOMAS, S., BRUGGE, J., PELICCI, P.G., SCHLESSINGER, J. and PAWSON, T., Association of the Shc and Grb2/Sem5 SH2-containing proteins is implicated in activation of the Ras pathway by tyrosine kinases. Nature (Lond.), 360, 689–692 (1992). SASTRY, L., LIN, W., WONG, W.T., DI FIORE, P.P., SCOPPA, C.A. and KING, C.R., Quantitative analysis of Grb2-Sos1 interaction: the N-terminal SH3 domain of Grb2 mediates affinity. Oncogene, 11, 1107–1112 (1995). SCHLAEPFER, D.D., HANKS, S.K., HUNTER, T. and VAN DER GEER, P., Integrin-mediated signal transduction linked to Ras pathway by GRB2 binding to focal adhesion kinase. Nature (Lond.), 372, 786–791 (1994). SKOLNIK, E.Y., LEE, C.H., BATZER, A., VICENTINI, L.M., ZHOU, M., DALY, R., MYERS, M.J., JR., BACKER, J.M., ULLRICH, A., WHITE, M.F. and SCHLESSINGER, J., The SH2/SH3 domain-containing protein Grb2 interacts with tyrosine-phosphorylated IRS1 and Shc: implications for insulin control of ras signalling. EMBO J., 12, 1929–1936 (1993). VAN DER BLIEK, A.M., REDELMEIER, T.E., DAMKE, H., TISDALE, E.J., MEYEROWITZ, E.M. and SCHMID, S.L., Mutations in human dynamin block an intermediate stage in coated vesicle formation. J. Cell Biol., 122, 553–563 (1993). WANG, J., AUGER, K.R., JARVIS, L., SHI, Y. and ROBERTS, T.M., Direct association of Grb2 with the p85 subunit of phosphatidylinositol 3-kinase. J. biol. Chem., 270, 12774–12780 (1995). WANG, W., FISHER, E.M., JIA, Q., DUNN, J.M., PORFIRI, E., DOWNWARD, J. and EGAN, S.E., The Grb2 binding domain of mSos1 is not required for downstream signal transduction. Nature (Genet.), 10, 294–300 (1995). XIE, Y., PENDERGAST, A.M. and HUNG, M.C., Dominant negative mutants of Grb2 induced reversal of the transformed phenotypes caused by the point mutation activated rat Her2/neu. J. biol. Chem., 270, 30717–30724 (1995). YANG, B., JUNG, D., MOTTO, D., MEYER, J., KORETZKY, G. and CAMPBELL, K.P., SH3 domain-mediated interaction of dystroglycan and Grb2. J. biol. Chem., 270, 11711–11714 (1995a). YANG, S.S., VAN AELST, L. and BAR-SAGI, D., Differential interactions of human Sos1 and Sos2 with Grb2. J. biol. Chem., 270, 18212–18215 (1995b). YE, Z.S. and BALTIMORE, D., Binding of Vav to Grb2 through dimerization of Src homology 3 domains. Proc. nat. Acad. Sci. (Wash.), 91, 12629– 12633 (1994).