Blockade of Epidermal Growth Factor

[CANCER RESEARCH 64, 2601–2609, April 1, 2004] Blockade of Epidermal Growth Factor- or Heregulin-Dependent ErbB2 Activation with the Anti-ErbB2 Monoclonal Antibody 2C4 Has Divergent Downstream Signaling and Growth Effects

James G. Jackson,1 Patricia St. Clair,1 Mark X. Sliwkowski,2 and Michael G. Brattain1 1Department of Surgery, University of Texas Health Science Center, San Antonio, Texas, and 2Genentech, Inc., South San Francisco, California

ABSTRACT by initiating a cascade involving SOS, ras, raf, and MAP/ERK kinase (9). Due to heterodimerization and a variety of stimulating ligands, the Another major downstream effector of oncogenic signals, phos- ErbB receptor system is both diverse and flexible, which proves particu- phatidylinositol 3Ј-kinase [PI3K (10)], can be activated by several larly advantageous to the aberrant signaling of cancer cells. However, specific mechanisms of how a particular receptor contributes to generat- different mechanisms in ErbB1 and ErbB2 signaling. ErbB1 and ing the flexibility that leads to aberrant growth regulation have not been ErbB2 only weakly bind the p85 subunit of PI3K in their COOH well described. We compared the utilization of ErbB2 in response to termini (11); therefore, they recruit help in transmitting a PI3K signal. epidermal growth factor (EGF) and heregulin stimulation in colon carci- ErbB1 can form heterodimers with ErbB3, which contains several noma cells. Anti-ErbB2 monoclonal antibody 2C4 blocked heregulin- binding motifs in its COOH terminus for p85 binding (12). ErbB1 can stimulated phosphorylation of ErbB2 and ErbB3; activation of mitogen- also recruit the binding of adaptor molecules such as cbl (13), Grb2- activated protein kinase (MAPK), phosphatidylinositol 3؅-kinase (PI3K), associated binder (Gab) 1 (14–16), and Gab2 (17), each of which also and Akt; proliferation; and anchorage-independent growth. 2C4 blocked contains high-affinity p85 binding sites and can thereby promote an EGF-mediated phosphorylation of ErbB2 and inhibited PI3K/Akt and increase in PI3K activity (18). Phospholipid products of PI3K then anchorage-independent growth but did not affect ErbB1 or MAPK. Im- munoprecipitations showed that ErbB3 and Grb2-associated binder (Gab) initiate the activation of molecules such as phosphoinositide-depend- 1 were phosphorylated and associated with PI3K activity after heregulin ent protein kinase 1 (PDK1) and Akt (19). Akt has been shown to treatment and that Gab1 and Gab2, but not ErbB3, were phosphorylated transmit antiapoptotic signals by phosphorylating molecules such as and associated with PI3K activity after EGF treatment. These data show Bad (20) and FKHR/FOXO1 (21). ErbB2 has a weak affinity for cbl that monoclonal antibody 2C4 inhibited all aspects of heregulin signaling (22), and its affinity for Gab1/Gab2 is unknown. as well as anchorage-independent and monolayer growth. Furthermore, Further demonstrating complexity in the family, ErbB3 (or Her3) we identify ErbB2 as a critical component of EGF signaling to the Gab1/ does not possess an active kinase and is therefore dependent on ErbB2 Gab2-PI3K-Akt pathway and anchorage-independent growth, but EGF to form a functional receptor for its ligands (23), the heregulins (also stimulation of MAPK and monolayer growth can occur efficiently without called neuregulins or neu differentiation factors). ErbB4 (or Her4) can the contribution of ErbB2. signal in homodimers but also dimerizes preferentially with ErbB2 (24). ErbB3 signals to the MAPK pathway in a very similar fashion to INTRODUCTION ErbB1 (25). However, ErbB3 contains multiple p85 consensus binding motifs in its COOH terminus and can signal without using any of The ErbB family of tyrosine kinase receptors (ErbB1–4) plays an the adaptor molecules (cbl, Gab1, or Gab2) used by ErbB1 for PI3K important role in the progression of numerous types of malignancies activation (12). (1, 2). The complexity of available ligands, receptors, and oligomer- The ability of ErbB family receptors to undergo heterodimerization is ization provide ErbB receptor signaling with a great deal of flexibility an important basis for the system’s diversity and flexibility (26). ErbB2 and diversity, enabling cells to respond to environmental stimuli. is the preferred partner of all ErbB family members, and heterodimer- Given the extensive capabilities of this signaling system, it is not ization is preferred over homodimerization. Comparison of homodimer- surprising that the aberrant regulation of ErbB family signaling is ization and heterodimerization has typically involved overexpression frequently associated with several histological types of cancer in a approaches in cell lines null or nearly devoid of background ErbB manner that is not dependent on receptor overexpression (3–5). receptors to reduce the complexity of the ensuing receptor oligomeriza- These diverse signals originate when ErbB family receptors are tion and subsequent downstream signaling (27–29). Specific signaling trans-phosphorylated at specific tyrosine residues in their COOH mechanisms that generate diversity in those cancer cells that exhibit termini after ligand binding. This phosphorylation can occur in the multiple ErbB family receptors and provide the aberrant signaling that context of homodimers, such as ErbB1-ErbB1, or heterodimers, such sustains specific malignant properties have not been elucidated within a as ErbB1-ErbB2. Various signaling molecules and adaptor proteins completely native receptor system context. are then recruited to these docking sites, and downstream signals are Here, we use GEO colon cancer cells to examine how signaling transmitted (2). For example, activation of the mitogen-activated diversity and flexibility result from ErbB2 heterodimerization with protein kinase (MAPK) pathway can occur when the adaptor molecule either ErbB1 or ErbB3 to sustain an important hallmark of malignant Grb2 directly binds ErbB receptors via an SH2 domain (6, 7), or when cells: the ability of anchorage-independent growth. GEO cells express Shc, which itself has Grb2 binding sites, binds ErbB receptors via a ErbB1, ErbB2, and ErbB3 (5, 30, 31). They also express endogenous phospho-tyrosine binding domain (8). Grb2 can then activate MAPKs ligands that activate ErbB1, as well as heregulin, which activates ErbB2 through heterodimerization with ErbB3 (5, 30). We show that Received 10/2/03; revised 12/9/03; accepted 2/5/04. exogenous epidermal growth factor (EGF) or heregulin conferred the Grant support: NIH Grants CA54807 and CA34432 (M. G. Brattain). ability of anchorage-independent growth in GEO cells, which was, in The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with both cases, dependent on ErbB2 heterodimerization and activation. 18 U.S.C. Section 1734 solely to indicate this fact. We have characterized the diverse signaling mechanisms elicited by Requests for reprints: Michael G. Brattain, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, New York 14263 (present address). Phone: (716) 845-3044; Fax: ErbB2 that are dependent on its binding partner. Specifically, by using (716) 845-4437; E-mail: [email protected]. an ErbB2-blocking antibody to inhibit heterodimerization of ErbB1 2601

and ErbB2 in response to EGF stimulation, we show a bifurcation of washing, blots were incubated with a 2000:1 dilution of horseradish peroxi- signal transduction from ErbB1 versus that from ErbB2, both in terms dase-linked secondary antibody in blocking buffer for 1 h, followed by further of the pathways activated and function. The anti-ErbB2 monoclonal washing. Enhanced chemiluminescence was performed according to the man- antibody (mAb) 2C4 blocked EGF-induced phosphorylation of ErbB2 ufacturer’s instructions (Amersham Pharmacia Biotech). and inhibited PI3K activity associated with both Gab1 and Gab2 and PI3K Assay. After growth factor treatment, the cell monolayers were washed with PBS and lysed, and protein concentration was determined as activation of the kinase Akt. Inhibition of these pathways was suffi- described previously (37). Lysates (250 ␮g of protein) were incubated with 2.5 cient to reduce EGF-induced colony formation in soft agar. 2C4 had ␮g of primary antibody for 30 min, followed by further incubation with protein no effect on ErbB1 phosphorylation or the EGF-induced activation of A-agarose or protein G-agarose for 2 h. Immune complexes were washed twice the MAPK extracellular signal-regulated kinase (Erk) 1/2 and had with each buffer PBS and 1% NP40, 100 mM Tris and 5 mM LiCl, and 10 mM only a minimal effect on monolayer growth. These data identify Gab1 Tris, 150 mM NaCl, and 5 mM EDTA; all buffers included 10 mM sodium and Gab2 as important signaling intermediates and ErbB2 as a nec- orthovanadate. After the last wash was removed, PI3K assays were performed essary component in EGF-induced activation of the PI3K pathway, as described previously (37). Briefly, samples were resuspended in 50 ␮lof but not the MAPK pathway. To our knowledge, this is the first PI3K buffer [20 mM Tris (pH 7.5), 100 mM NaCl, and 0.5 mM EGTA], and 10 demonstration of specific pathway activation as well as subsequent ␮g of phosphatidylinositol were added. After 10 min at room temperature, 10 ␮Ci of [32P]ATP and MgCl to a final concentration of 20 ␮M was added. After function for endogenous heterodimers of ErbB1 and ErbB2. 2 10 min at room temperature, lipids were extracted, first with 150 ␮lof ␮ CHCl3:methanol:HCl (10:20:0.2) and 100 l of pure CHCl3. The second ␮ MATERIALS AND METHODS extraction used 80 l of methanol:1N HCl (1:1). Samples were spotted on 1% potassium oxalate-treated TLC plates (Analtech, Newark, DE) and developed

Reagents. All chemicals and reagents were purchased from Sigma (St. in CHCl3:methanol:NH4OH:H2O (129:114:15:21). On an autoradiograph of Louis, MO) unless noted otherwise. Gab1 and p85 PI3K antibody were the plate, the highest migrating spots on the TLC plate, representing phos- purchased from Upstate Biotechnology (Lake Placid, NY). Total and phospho- phatidylinositol phosphate, were quantitated by densitometry using AlphaIm- specific Akt antibodies were purchased from New England Biolabs (Beverly, ager software (Alpha Innotech, San Leandro, CA). MA). ErbB1 for immunoprecipitation (Ab12) and ErbB3 (Ab5) antibodies Cell Monolayer Growth. GEO cells were plated in 24-well plates at were from Neomarkers (Monroe, CA). Gab2, phospho-Erk1/2, total Erk1/2, 30,000 cells/well in SF medium and allowed to adhere overnight. The next day, and ErbB2 antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). cells were washed with PBS and pretreated for 30 min in SM medium with or ErbB1 for immunoblot, cbl, and RC-20 anti-phospho-tyrosine were purchased without 20 ␮g/ml 2C4, and then the indicated growth factors were added at 50 from Transduction Laboratories (Lexington, KY). Horseradish peroxidase- ng/ml. After 5 days, growth was analyzed using the 3-(4,5-dimethylthiazol-2- linked antirabbit and antimouse antibodies and molecular weight markers were yl)-2,5-diphenyltetrazolium bromide assay (MTT) as described previously from Amersham Pharmacia Biotech (Piscataway, NJ). Protein A-agarose was (38). Sixty ␮l of 5 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo- from Life Technologies, Inc. (Bethesda, MD). Acrylamide was from Fisher lium bromide reagent in PBS were added to each well, and then, after a 2-h Biotech (Hampton, NH). GEO cells have been described previously (32). 2C4 incubation, wells were aspirated, and 0.5 ml of Me2SO was added. Absorbance and Herceptin were obtained from Genentech, Inc. (South San Francisco, CA) was measured at 545 nm. and have been described previously (33–35). A humanized variant of 2C4 Soft Agar Assay. Experiments were performed essentially as described (recombinant human mAb 2C4) was used in all studies described in this report previously (39). Briefly, 25,000 cells resuspended in 0.4% low melting point (36). agarose were plated on a 0.8% agarose underlayer in a 6-well plate in triplicate. Cell Stimulation and Lysis. GEO cells were regularly cultured in supple- Both layers contained treatments in SM medium as indicated in the Fig. 6 mental McCoy’s medium containing insulin, EGF, and transferrin (SF media). legend. Colonies were allowed to form for 8 days and then stained with 1 ml Before signaling experiments, cells were starved for 2–4 days in supplemental of p-iodonitrotetrazolium violet overnight, and images were captured on Al- McCoy’s medium with no growth factors (SM media). For treatment with 2C4, phaImager. Images of anchorage-independent growth at ϫ40 magnification cells were preincubated for 30 min in SM media with 20 ␮g/ml antibody, and were captured with a Nikon microscope fitted with an Olympus MagnaFire then 50 ng/ml growth factor was added for the indicated times. Cells were digital camera using Image-Pro Plus software from Media Cybernetics (Silver washed twice in ice-cold PBS and lysed with 500 ␮l per 10-cm plate of Spring, MD). TNESV buffer [50 mM Tris (pH 7.4), 1% NP40, 2 mM EDTA, 100 mM NaCl, 0.1% SDS, 10 mM sodium orthovanadate, and 1 complete protease inhibitor RESULTS mixture tablet (Boehringer Mannheim, Mannheim, Germany) per 10 ml of lysis buffer]. Protein concentration of the supernatants was determined by the Monoclonal Antibody 2C4 Can Inhibit Heregulin- and EGF- copper- bicinchoninic acid method with a Pierce Laboratory kit (Rockford, IL). Stimulated Phosphorylation of ErbB2. GEO cells express ErbB1, Immunoprecipitations. All steps were performed with gentle rotation at ErbB2, and ErbB3 but not ErbB4 (data not shown; Refs. 5, 30, and 4°C. Because of the presence of 2C4 antibody in some treatment groups, 31). To test the responsiveness of GEO cells to heregulin and EGF, we lysates used in immunoprecipitation were not precleared with protein A- stimulated quiescent cell monolayers with these growth factors and agarose. However, no effect from the presence of 2C4 was observed on the immunoprecipitation of any proteins. Protein from TNESV lysates (250 ␮g) then immunoprecipitated lysates with ErbB2 antibody and performed was incubated with 2.5 ␮g of the indicated antibody overnight. Twenty-five ␮l anti-phospho-tyrosine immunoblots. We found a low basal level of of protein A-agarose or protein G-agarose was then added for 4 h, followed by ErbB2 phosphorylation in quiescent GEO cells (Fig. 1A, Lane 1). three washes with TNESV buffer. Beads were resuspended in 2ϫ Laemmli After treatment with EGF or heregulin for 10 min, there was a large loading buffer with 2.5% ␤-mercaptoethanol, boiled, and separated by 8% increase in phospho-tyrosine on ErbB2 (Fig. 1A, Lanes 2 and 5, SDS-PAGE. respectively). A 30-min pretreatment with the anti-ErbB2 mAb 2C4 Immunoblotting. After SDS-PAGE proteins were transferred overnight to (20 ␮g/ml) strongly inhibited ErbB2 phosphorylation after similar nitrocellulose membranes (Amersham Pharmacia Biotech, Arlington Heights, EGF or heregulin treatment (Fig. 1A, Lanes 4 and 7). Consistent with IL), the membranes were blocked in 5% nonfat dry milk in Tris-buffered others’ reports (35, 40), pretreatment with the ErbB2 mAb Herceptin saline-Tween 20 [TBST (0.15 M NaCl, 0.01 M Tris HCl (pH 7.4), and 0.05% did not reduce phosphorylation of ErbB2 after growth factor treat- Tween 20)]. For anti-phospho-tyrosine blotting with RC-20, membranes were washed five times with TBST after blocking, incubated in a 5000:1 dilution in ment, ruling out a nonspecific mAb effect or a vehicle effect on ErbB2 TBST for 1 h, and washed thoroughly; chemiluminescence was then performed phosphorylation (Fig. 1A, Lanes 3 and 6). The concentration of as described below. For phospho-Akt immunoblot, membranes were incubated Herceptin and 2C4 used in these studies has been shown to be overnight in TBST at 1000:1. All other blots were incubated with a 1 ␮g/ml saturating but nontoxic in numerous previous studies (35, 36, 41–44). dilution of the indicated antibody in blocking buffer overnight at 4°C. After Anti-ErbB2 immunoblot of the same membrane showed that ErbB2 2602

inhibited by pretreatment with 2C4 at 10 min (Fig. 2A, top panels, Lane 3) or at other time points up to 60 min (data not shown). These data are in contrast to those observed in MCF-7 breast cancer cells (36), where 2C4 was able to block MAPK activation by the ErbB1 ligand transforming growth factor ␣. Heregulin activation of Erk1/2 was effectively blocked by 2C4 (Fig. 2A, top panels, Lanes 4–6). Anti-Erk1/2 immunoblots on the same membranes show that total levels of Erk1/2 were unchanged in all treatment groups (Fig. 2A, bottom panels). A second well-characterized downstream signaling molecule of the ErbB family is Akt, whose activation is initiated by products of PI3K. Akt was phosphorylated after EGF treatment (Fig. 2B, top panel, Lane 2), and, in contrast to observations in MAPK signaling, Akt phosphorylation was inhibited by 2C4 (Fig. 2B, top panel, Lane 3). Heregulin treatment resulted in Akt activation, and, consistent with MAPK

Fig. 1. 2C4 blocks epidermal growth factor- and heregulin-stimulated phosphorylation of ErbB2. Growth factor-starved GEO monolayers were pretreated with or without 20 ␮g/ml of the monoclonal antibodies Herceptin (Hrc) or 2C4 for 30 min in growth factor-free media and then stimulated for 10 min with 50 ng/ml growth factor as indicated above the lanes. Cells were lysed, and then 250 ␮g of protein were immunoprecipitated with the indicated antibody, separated by SDS-PAGE, and blotted for anti-phospho- tyrosine (top panels) followed by total ErbB antibody (bottom panels) as described in “Materials and Methods.” Molecular weight markers are indicated to the right of each panel. Data shown are representative of at least two independent experiments. levels were not significantly changed by growth factor or mAb treatment (Fig. 1A, bottom panel). We next sought to determine the fates of the other ErbB family members after 2C4 treatment and stimulation. Fig. 1B shows that ErbB1 was phosphorylated after EGF treatment (Lane 2) but not heregulin treatment (Lane 4), and this phosphorylation was unaffected by treatment with 2C4 (Lane 3). Likewise, ErbB3 was phosphorylated after heregulin treatment (Fig. 1B, Lane 9); however, this activation could be blocked by 2C4 pretreatment (Fig. 1B, Lane 10). EGF did not stimulate activation of ErbB3 (Fig. 1B, Lane 7). Immunoblots on the same membranes with ErbB1 or ErbB3 antibodies showed that total levels of these receptors were not changed by growth factor or 2C4 treatment (Fig. 1B, bottom panels). As observed in Fig. 1A for ErbB2, Herceptin had no effect on ErbB3 phosphorylation after heregulin treatment (data not shown), nor did Herceptin change the tyrosine phosphorylation status of any proteins after EGF treatment in a total phospho-tyrosine blot, including those in the molecular weight range of the ErbB receptor family (data not shown). The presence of Herceptin in lysates of Herceptin-treated cells did not allow for ErbB1 immunoprecipitation/Western blot because ErbB2 was also immunoprecipitated and detected. This was not observed for 2C4. Thus, GEO cells activated ErbB1-ErbB2 after EGF treatment, and 2C4 Fig. 2. Epidermal growth factor and heregulin activation of mitogen-activated protein blocked only the phosphorylation of ErbB2. ErbB2 and ErbB3 were kinase, Akt, and p85 signaling after 2C4 treatment. Growth factor-starved GEO mono- activated after heregulin treatment, and 2C4 blocked the activation of layers were pretreated in growth factor-free media with or without monoclonal antibody both receptors. We next investigated the downstream consequences of 2C4 (20 ␮g/ml) for 30 min and then stimulated for 10 min with 50 ng/ml growth factor as indicated above the lanes. A and B, cells were lysed, and 50 ␮g of protein were these results. separated by SDS-PAGE and immunoblotted for phospho-extracellular signal-regulated Activation of Erk and Akt after 2C4 Treatment and Growth kinase 1/2 (A, top panels) followed by total extracellular signal-regulated kinase 1/2 (A, Factor Stimulation. The MAPKs are well-characterized downstream bottom panels) and phospho-Akt (B, top panels) followed by total Akt (B, bottom panels) as described in “Materials and Methods.” C, 250 ␮g of protein were immunoprecipitated targets of ErbB receptors. We determined the activation state of with anti-p85 phosphatidylinositol 3Ј-kinase antibody, separated by SDS-PAGE, and MAPK Erk1/2 after 2C4 treatment and growth factor stimulation in blotted for anti-phospho-tyrosine as described in “Materials and Methods.” Molecular weight markers are indicated to the right, and migration positions of phosphoproteins GEO cells. Erk1/2 phosphorylation was increased after a 10-min EGF coimmunoprecipitating with p85 are indicated to the left. Data shown are representative treatment (Fig. 2A, top panel, Lane 2), and this activation was not of at least two independent experiments. 2603

munoblotted for anti-phospho-tyrosine. Fig. 3 shows that EGF induced phosphorylation of Gab1 (Lane 2), Gab2 (Lane 5), and cbl (Lane 8) but not ErbB3 (Lane 11). Gab1 and cbl each have molecular weights of approximately 120,000, consistent with the pp120 observed in Fig. 2C.

The molecular weight of Gab2 (Mr 97,000) and its less intense degree of phosphorylation are both consistent with the pp100 observed in Fig. 2C. After heregulin treatment, ErbB3 antibody immunoprecipitated a strong phospho-tyrosine-containing double band, likely ErbB3/ErbB2 (Fig. 3, Lane 12), Heregulin also stimulated some phosphorylation of Gab1 (Fig. Fig. 3. Phosphorylation of Grb2-associated binder (Gab) 1/Gab2/cbl after epidermal growth factor treatment and ErbB3/Gab1 after heregulin treatment. Quiescent GEO 3, Lane 1 versus Lane 3). These findings correlate with the coimmuno- monolayers were stimulated for 10 min with 50 ng/ml growth factor as indicated above the precipitation of a major Mr 190,000 phosphoprotein and a minor Mr lanes. Cells were lysed, and then 250 ␮g of protein were immunoprecipitated with the indicated antibody, separated by SDS-PAGE, and blotted for anti-phospho-tyrosine as 120,000 phosphoprotein with p85 antibody observed in Fig. 2C. Further described in “Materials and Methods.” Molecular weight markers are indicated to the immunoblotting of the membrane with p85 antibody showed that p85 right, and migration positions of specific proteins are indicated to the left. Data shown are was associated with Gab1 and Gab2, but not cbl, after EGF treatment and representative of three independent experiments. that p85 was associated with ErbB3 after heregulin treatment (data not shown). results, 2C4 blocked this effect (Fig. 2B, top panel, Lanes 4–6). To further verify the pathways containing PI3K activity after Anti-Akt immunoblots on the same membranes showed that total growth factor treatment in GEO cells, we next assessed the levels of levels of Akt were not significantly changed by any of the treatment PI3K activity associated with each of the signaling molecules identi- groups (Fig. 2B, bottom panels). These data show that 2C4 could fied in Fig. 3. Consistent with the phosphorylation data in Fig. 3, we block heregulin-induced signaling to both MAPK and Akt, consistent found that EGF treatment increased the PI3K activity associated Gab1 with the notion that ErbB3-ErbB2 heterodimers are essential for all (Fig. 4, Lane 1 versus 2) and Gab2 (Fig. 4, Lane 4 versus 5), but not aspects of heregulin signaling. In EGF signaling, however, disruption ErbB3 (Fig. 4, Lanes 7 and 8). Heregulin increased PI3K activity of the ErbB2 component with 2C4 only inhibited signaling to Akt, not associated with ErbB3 (Fig. 4, Lane 7 versus Lane 9) and Gab1 (Fig. the MAPKs Erk1/2. These results indicated that Erk1/2 activation was 4, Lane 1 versus Lane 3), which is consistent with the heregulin- the consequence of ErbB1 signaling, whereas Akt activation was due induced phosphorylation of ErbB3 and Gab1 observed in Fig. 3. There at least in part to ErbB2 signaling, thus raising the question of how was minimal basal activity and no increase in PI3K activity associated PI3K was activated by EGF. with cbl after EGF or heregulin treatment (data not shown). These Coimmunoprecipitation of Distinct Phospho-Tyrosine-Contain- data provide evidence that EGF activates PI3K via the Gab1 and Gab2 ing Proteins after EGF and Heregulin Treatment. Growth factor- adaptor molecules, but not through ErbB3 or cbl, and that heregulin induced activation of the p85 subunit of PI3K is initiated when, via its activates PI3K via ErbB3 and, to a lesser extent, Gab1. SH2 domain, p85 binds phosphorylated tyrosine residues of receptors 2C4 Inhibits EGF and Heregulin Stimulated PI3K Activity. We or adaptor molecules (10). To find phospho-tyrosine-containing mol- next examined the effects of 2C4 on heregulin- and EGF-stimulated ecules bound to p85 in GEO cells, we stimulated cells as described in PI3K activity. We found that the activity of PI3K associated with the Fig. 1 legend, immunoprecipitated lysates with p85 antibody, and then immunoblotted with anti-phospho-tyrosine antibody. Fig. 2C shows that in the absence of stimulation, p85 was associated with two phospho-tyrosine-containing molecules of approximately Mr 180,000 (Lane 1). These two bands were present in all treatment groups. After heregulin stimulation, the major phosphoprotein that coimmunopre- cipitated with p85 was approximately Mr 190,000, and this association was abrogated by 2C4 (Fig. 2C, Lanes 4 and 5). The molecular weight of this phosphoprotein is consistent with that of ErbB3, whose binding to p85 after heregulin stimulation has been well characterized (45,

46). Of note, there was an apparent slight increase in a Mr 120,000 phosphoprotein after heregulin treatment (Fig. 2C, Lane 4). After EGF stimulation, we saw no change in phospho-tyrosine content of proteins in the Mr 190,000 range, but we did observe an increase in Mr 120,000 and Mr 100,000 phosphoproteins associated with p85 (Fig. 2C, Lane 2). The intensity of the Mr 120,000 band appeared to be reduced by 2C4 treatment (Fig. 2C, Lane 3). To determine the probable identities of these phosphoproteins, we next performed specific immunoprecipitations. Identification of Gab1/Gab2 and ErbB3/Gab1 as Modulators of PI3K for EGF and Heregulin, Respectively. As stated above, it has been shown that the major phospho-tyrosine-containing protein associated with p85 after heregulin stimulation is ErbB3. EGF can induce Fig. 4. Phosphatidylinositol 3Ј-kinase activity is associated with Grb2-associated association of p85 with ErbB3 (12) or the adaptor molecules Gab1 binder (Gab) 1/Gab2 after epidermal growth factor treatment and ErbB3/Gab1 after (14–16) and cbl (13), which migrate at the approximate molecular weight heregulin treatment. Quiescent GEO monolayers were stimulated for 10 min with 50 ng/ml growth factor as indicated above the lanes. Cells were lysed, 250 ␮g of protein were of pp120 of Fig. 2C, and Gab2 (17), which migrates at the approximate immunoprecipitated with the indicated antibody, and then a phosphatidylinositol 3Ј-kinase molecular weight of pp100 of Fig. 2C. To determine which of these were assay was performed as described in “Materials and Methods.” Relative positions of phosphatidylinositol phosphate (PIP) and the origin (Ori) are indicated to the left. activated in GEO cells, we stimulated cells as described in the Fig. 1 Densitometric analysis of the data is shown graphically. Data shown are representative of legend, immunoprecipitated them with specific antibodies, and then im- at least two independent experiments. 2604

Fig. 5. 2C4 inhibits heregulin- and epidermal growth factor-stimulated phosphatidylinositol 3Ј-kinase activity. Quiescent GEO monolayers were pretreated in growth factor- free media with or without monoclonal antibody 2C4 (20 ␮g/ml) for 30 min and then stimulated with 50 ng/ml growth factor as indicated above the lanes. Cells were lysed; 250 ␮g of protein were immunoprecipitated with ErbB3 (A), Grb2-associated binder (Gab) 1 (B), or Gab2 (C) antibody; and then phosphatidylinositol 3Ј-kinase assays were performed as described in “Materials and Methods.” Densitometric analysis of the data is shown graphically. Relative positions of phosphatidylinositol phosphate (PIP) and the origin (Ori) are indicated to the left. Data shown are representative of at least two independent experiments.

ErbB3 after heregulin treatment for different times (Fig. 5A, Lanes 2, EGF treatment (Fig. 5, B and C, Lane 3), whereas the greatest degree 4, and 6) could be blocked by 2C4 (Fig. 5A, Lanes 3, 5, and 7). In cells of inhibition was at 60 min (Fig. 5, B and C, Lane 7). stimulated by EGF for different times, PI3K activity associated with 2C4 Inhibits Heregulin-Stimulated Cell Proliferation and EGF- Gab1 (Fig. 5B, Lanes 2, 4, and 6) or Gab 2 (Fig. 5C, Lanes 2, 4, and and Heregulin-Stimulated Anchorage-Independent Growth. Af- 6) was partially inhibited by 2C4 (Fig. 5B, Lanes 3, 5, and 7 for Gab1; ter resolving how 2C4 alters the components of heregulin and EGF Fig. 5C, Lanes 3, 5, and 7 for Gab2). 2C4 was less effective at signal transduction, we next investigated the effects of 2C4 treatment inhibiting PI3K activity associated with Gab1 and Gab2 at 1 min of on cell proliferation and anchorage-independent growth, an estab- 2605

Fig. 6. 2C4 inhibits monolayer and anchorage- independent growth in GEO cells. GEO cells were seeded at 30,000 cells/well in a 24-well plate and allowed to adhere overnight. A, the next day, wells were washed with PBS and then pretreated in growth factor-free media (SM) with or without 20 ␮g/ml monoclonal antibody 2C4 for 30 min and then stimulated with 50 ng/ml growth factor as indicated below the lanes (Hrg, heregulin; SF, 20 ␮g/ml insulin, 5 ng/ml epidermal growth factor, and 40 ng/ml transferrin). After 5 days, 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay was performed as described in “Mate- rials and Methods.” Error bars represent SE. Data shown are representative of at least three independent experiments. B and C, 25,000 cells resuspended in 0.4% low melting point agarose were plated on a 0.8% agarose under layer in 6-well plates in triplicate for each treatment group. As indicated, triplicate wells contained epidermal growth factor or heregulin at a final concentration of 50 ng/ml and 2C4 at a final concentration of 20 ␮g/ml in both the top and bottom layers. Colonies were allowed to form for 8 days and then stained with 1 ml of p-iodonitrotetrazolium violet overnight, and images were captured using AlphaImager. Results shown are representative of two independent experiments.

lished marker of malignant potential. In cell monolayer growth over 5 We next assessed the anchorage-independent growth of GEO days, we found that exogenous heregulin added to growth factor-free cells by assaying colony formation in soft agar. We found that media (Fig. 6A, Hrg) stimulated approximately a 3–4-fold increase in cells plated in the absence of growth factors formed very few cell number over the serum-free media alone (C-SM), whereas EGF colonies, whereas cells plated in normal growth medium readily stimulated a 2-fold increase. Normal growth medium (SF), which formed numerous colonies (data not shown). Cells plated in me- contains a high concentration of insulin, also stimulated a 3–4-fold dium containing heregulin also readily formed colonies, and 2C4 increase in cell number. 2C4 effectively blocked heregulin-induced completely blocked this effect (Fig. 6, B and C, left panels). EGF proliferation but had only a minimal but reproducible effect on EGF- also stimulated the formation of colonies, and 2C4 treatment stimulated growth and no effect on insulin-stimulated (SF) cell growth reduced the number and size of colonies (Fig. 6, B and C, right (Fig. 6A). panels). 2606

DISCUSSION responses to the same therapy, and care must be taken in extrapolating results shown here or in the study by Agus et al. (36) to other cancer When regulation of the ErbB receptor pathways is lost, increased types. The PI3K pathway was inhibited by 2C4 after EGF treatment cell proliferation and survival can result, leading to the malignant in GEO cells, and this finding led us to further investigate activation phenotype. This has been observed in numerous types of cancers (26), of PI3K in these cells. and designing therapies targeting points of lost regulation is an area of Activation of PI3K after EGF stimulation has been reported to increasing interest (47). A wide diversity of signaling events origi- occur by several different mechanisms. Whereas ErbB1 does not bind nates from the ErbB receptors due to a complexity generated by the p85 regulatory subunit of PI3K with high affinity, it can use numerous ligands as well as different receptor combinations (2). It is adaptor proteins such as cbl (13), Gab1 (14–16), and Gab2 (17), or it important to understand the specific contributions of individual re- can signal laterally to ErbB3 via ErbB2 (24). In this study, we found ceptors in these involved signaling complexes to increase our under- that EGF treatment resulted in phosphorylation of Gab1, Gab2, and standing of the development of malignancy. In this study, we used cbl; however, p85 and PI3K activity were associated only with Gab1 2C4, a mAb directed against the extracellular domain of ErbB2, which and Gab2. can prevent the heterodimerization of ErbB3 with ErbB2 (23, 40). Recent data show that receptor tyrosine kinase binding of Gab1/2 We first assessed the responsiveness of the human colon carcinoma can occur by several different mechanisms. Gab1 can bind directly to cell line GEO to exogenous heregulin and EGF, and then we deter- tyrosine-phosphorylated receptors (56) or through an intermediary mined which downstream signals of heregulin and EGF were dis- Grb2 or by both methods in a “double hook” scheme of binding (57). rupted by 2C4. GEO cells express autocrine transforming growth The latter mechanism is consistent with recent observations of ErbB1 ␣ factor and have been shown to be responsive to different ErbB1- binding of Gab1 (16). Furthermore, the ErbB1-Gab1 complex can also targeted therapies (48–50). However, we found that GEO cells were include SHP2 (17). In our results, we did not see any phospho- also very sensitive to the ErbB2/ErbB3 ligand heregulin, a pathway tyrosine-containing proteins in the molecular weight range of SHP2 or present in colon cancer (51–54). One would not expect that targeting Grb2 coimmunoprecipitate with Gab1/2 or p85 antibodies as others ErbB1 would inhibit heregulin-stimulated growth in these cells have observed (17, 58). We do not, however, rule out the possibility (heregulin treatment did not result in any detectable activation of that more complex aggregates involving Gab1/2 exist, but this possi- ErbB1). 2C4, however, completely abrogated heregulin-stimulated bility does not change our conclusion that EGF treatment activated activation of ErbB2-ErbB3, PI3K, Akt, and MAPK Erk1/2 in GEO Gab1 and Gab2 and that PI3K activity was associated with them. cell cultures. 2C4 also blocked heregulin stimulation of cell prolifer- The EGF-Gab1/2-PI3K pathway was somehow enhanced by ErbB2 ation and anchorage-independent growth of GEO cells. 2C4 effec- because 2C4 treatment inhibited PI3K activation without altering tively blocked the activation of ErbB2 after stimulation with EGF and ErbB1 or MAPK pathway activation. This raises the question of the inhibited the PI3K pathway and the anchorage-independent growth of mechanism by which 2C4 inhibited EGF activation of PI3K. One GEO cells. The effectiveness of 2C4 in blocking ErbB2 transphos- possibility is that Gab1/2 molecules bound directly or indirectly phorylation by both ErbB1 and ErbB3 is potentially important be- (through Grb2) to phospho-tyrosine sites on ErbB2, and abrogation of cause it has been shown that the level of ErbB2 activation has ErbB2 phosphorylation by 2C4 simply reduced the number of avail- prognostic significance (55). To help rule out the possibility that this able binding sites. Although Gab1 has not been shown previously to 2C4 effect was nonspecific or a vehicle effect, we used Herceptin, a bind any ErbB family receptors other than ErbB1, we found that Gab1 mAb directed against a different epitope of ErbB2 that is effective phosphorylation and associated PI3K activity were increased after only in settings where ErbB2 is overexpressed (33, 35). Herceptin did heregulin treatment (Figs. 5 and 6), suggesting that Gab1 can and does not block activation of ErbB2 after heregulin or EGF treatment (Fig. bind ErbB receptors other than ErbB1 because only ErbB2-ErbB3, 1A), consistent with the findings of others (35, 40). Furthermore, 2C4 and not ErbB1, was phosphorylated after heregulin treatment. In was unable to impair the growth of GEO cells in medium containing addition, ErbB2 does contain a YXNQ motif at 1139 in its COOH a high concentration of insulin (Fig. 6A) and did not alter EGF terminus, homologous to the Y1068 in ErbB1, a site that is known to activation of ErbB1. Taken together, these data suggest that 2C4, be responsible for Gab1 binding. A possible connection between Gab1 ErbB2 mAbs in general, or the vehicle did not produce a global and ErbB2 is further suggested by the fact that the knockout mice of decrease in phosphorylation or a nonspecific toxicity in the cells that each have seriously impaired cardiac development (59, 60). could not be overcome by an unrelated growth factor (insulin). Alternatively, a signaling complex of ErbB1-ErbB2 heterodimer, Treatment with 2C4 blocked EGF-stimulated ErbB2 phosphoryla- with Gab1/2 bound to ErbB1, could have been stabilized by the tion but did not diminish the activation of ErbB1. This observation is presence of ErbB2. It has been shown that EGF dissociates from an consistent with the notion that, in GEO cells, ErbB2 was dependent on ErbB2-ErbB1 heterodimer more slowly than an ErbB1 homodimer ErbB1 for trans-phosphorylation, but not vice versa. Also consistent and that ErbB1-ErbB2 heterodimers are recycled back to the cell with our data is the hypothesis that after EGF bound ErbB1, het- membrane more efficiently than ErbB1 homodimers (61, 62). In our erodimers with ErbB2 were formed, leading to full activation of cells, it is possible that after 2C4 interrupted the heterodimerization of downstream pathways, MAPK and PI3K. However, when 2C4 ErbB1 and ErbB2, the ErbB1-Gab signal was enervated. Indeed, we blocked ErbB2, ErbB1-ErbB1 homodimerization occurred, which, did find that at 1 min, EGF-induced PI3K activity was only slightly interestingly, could fully activate the MAPK pathway. In MCF-7 inhibited by 2C4, but at 60 min, the inhibition was near complete. breast cancer cells, Agus et al. (36) found that MAPK activation was Thus, in either of these scenarios, 2C4 impaired the ability of ErbB2 dependent on ErbB2 because 2C4 inhibited activation of MAPK after to augment an EGF signal to PI3K. Immunoblots for phospho-Akt on stimulation with the ErbB1 ligand transforming growth factor ␣. This the same samples did not reveal a difference in degree of inhibition of difference could be due to differences in breast and colon cancers or Akt activation at different time points; however, phospho-Akt anti- the individual cell lines used; however, it does further demonstrate body has proven in our hands to be much less sensitive at detecting heterogeneity in ErbB signaling across different cell types. This modest differences in activation compared with the radioactive enzy- example of diversity in response across cell cultures and/or cancer matic PI3K assay. types after 2C4 treatment serves as a reminder that seemingly similar Interestingly, stimulation of cell monolayer growth by exogenous tumors, i.e., those that are ErbB2 driven, may still exhibit divergent EGF was only minimally inhibited by 2C4, whereas a greater, al- 2607

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2004 American Association for Cancer Research. ROLE OF ErbB2 IN EGF AND HEREGULIN SIGNALING though not complete, reduction was observed in anchorage-indepen- 18. Hibi M, Hirano T. Gab-family adapter molecules in signal transduction of cytokine dent growth. These data suggest that ErbB1 homodimers are still and growth factor receptors, and T and B cell antigen receptors. Leuk Lymphoma 2000;37:299–307. capable of mitogenic signaling in cell monolayers in the absence of a 19. Downward J. Mechanisms and consequences of activation of protein kinase B/Akt. contribution from ErbB2; however, anchorage-independent growth, a Curr Opin Cell Biol 1998;10:262–7. more widely accepted marker for tumorigenicity (63), is more de- 20. Datta SR, Dudek H, Tao X, et al. Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 1997;91:231–41. pendent on ErbB2 and likely PI3K, the pathway inhibited by ErbB2 21. Jackson JG, Kreisberg JI, Koterba AP, Yee D, Brattain MG. Phosphorylation and blockade. nuclear exclusion of the forkhead transcription factor FKHR after epidermal growth We show here that heregulin treatment activated ErbB2/ErbB3, factor treatment in human breast cancer cells. Oncogene 2000;19:4574–81. 22. Levkowitz G, Klapper LN, Tzahar E, et al. Coupling of the c-Cbl protooncogene PI3K activity associated with ErbB3 and Gab1, Akt, and MAPK product to ErbB-1/EGF-receptor but not to other ErbB proteins. Oncogene 1996;12: Erk1/2. Furthermore, heregulin was a potent mitogen for colon car- 1117–25. cinoma cells. The anti-ErbB2 mAb 2C4 effectively blocked the 23. Sliwkowski MX, Schaefer G, Akita RW, et al. Coexpression of erbB2 and erbB3 proteins reconstitutes a high affinity receptor for heregulin. J Biol Chem 1994;269: heregulin-stimulated activation of these pathways and reduced mono- 14661–5. layer and anchorage-independent growth to basal levels. We show that 24. Graus-Porta D, Beerli RR, Daly JM, Hynes NE. ErbB-2, the preferred heterodimer- EGF treatment activated ErbB1/ErbB2, PI3K activity associated with ization partner of all ErbB receptors, is a mediator of lateral signaling. EMBO J 1997;16:1647–55. Gab1/Gab2, Akt, and Erk1/2. 2C4 inhibited only the activation of 25. Vijapurkar U, Cheng K, Koland JG. Mutation of a Shc binding site tyrosine residue ErbB2 and the PI3K pathway, but not ErbB1 or Erk1/2. However, this in ErbB3/HER3 blocks heregulin-dependent activation of mitogen-activated protein selective inhibition was sufficient to inhibit anchorage-independent kinase. J Biol Chem 1998;273:20996–1002. 26. Olayioye MA, Neve RM, Lane HA, Hynes NE. The ErbB signaling network: receptor growth. These data suggest an important role for ErbB2 in potentiat- heterodimerization in development and cancer. EMBO J 2000;19:3159–67. ing a mitogenic signal from EGF to PI3K but that ErbB1 activation 27. Di Fiore PP, Pierce JH, Kraus MH, et al. erbB-2 is a potent oncogene when alone is sufficient for EGF to activate MAPK Erk1/2. overexpressed in NIH/3T3 cells. Science (Wash. DC) 1987;237:178–82. 28. Olayioye MA, Graus-Porta D, Beerli RR, et al. ErbB-1 and ErbB-2 acquire distinct Others have demonstrated that heterodimerization leads to signal- signaling properties dependent upon their dimerization partner. Mol Cell Biol 1998; ing diversity as indicated by adaptor recruitment (26) or signaling 18:5042–51. outcome (64). However, to our knowledge, this is the first documen- 29. Riese DJ II, van Raaij TM, Plowman GD, Andrews GC, Stern DF. The cellular response to neuregulins is governed by complex interactions of the erbB receptor tation of how heterodimers contribute to signaling diversity with family. Mol Cell Biol 1995;15:5770–6. regard to the identification of a specific signaling pathway (PI3K/Akt) 30. Ziober BL, Willson JK, Hymphrey LE, Childress-Fields K, Brattain MG. Autocrine transforming growth factor-␣ is associated with progression of transformed properties and its ensuing outcome (anchorage-independent growth) as a func- in human colon cancer cells. J Biol Chem 1993;268:691–8. tion of heterodimerization. 31. Ciardiello F, Caputo R, Troiani T, et al. Antisense oligonucleotides targeting the epidermal growth factor receptor inhibit proliferation, induce apoptosis, and cooper- ate with cytotoxic drugs in human cancer cell lines. Int J Cancer 2001;93:172–8. REFERENCES 32. Mulder KM, Brattain MG. Effects of growth stimulatory factors on mitogenicity and c-myc expression in poorly differentiated and well differentiated human colon car- 1. Schlessinger J. Cell signaling by receptor tyrosine kinases. Cell 2000;103:211–25. cinoma cells. Mol Endocrinol 1989;3:1215–22. 2. Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol 33. Lewis GD, Figari I, Fendly B, et al. Differential responses of human tumor cell lines Cell Biol 2001;2:127–37. to anti-p185HER2 monoclonal antibodies. Cancer Immunol Immunother 1993;37: 3. Jiang D, Yang H, Willson JK, et al. Autocrine transforming growth factor ␣ provides 255–63. a growth advantage to malignant cells by facilitating re-entry into the cell cycle from 34. Fendly BM, Winget M, Hudziak RM, et al. Characterization of murine monoclonal suboptimal growth states. J Biol Chem 1998;273:31471–9. antibodies reactive to either the human epidermal growth factor receptor or HER2/neu 4. Sawhney RS, Zhou GH, Humphrey LE, et al. Differences in sensitivity of biological gene product. Cancer Res 1990;50:1550–8. functions mediated by epidermal growth factor receptor activation with respect to 35. Lewis GD, Lofgren JA, McMurtrey AE, et al. Growth regulation of human breast and endogenous and exogenous ligands. J Biol Chem 2002;277:75–86. ovarian tumor cells by heregulin: evidence for the requirement of ErbB2 as a critical 5. Venkateswarlu S, Dawson DM, St Clair P, et al. Autocrine heregulin generates component in mediating heregulin responsiveness. Cancer Res 1996;56:1457–65. growth factor independence and blocks apoptosis in colon cancer cells. Oncogene 36. Agus DB, Akita RW, Fox WD, et al. Targeting ligand-activated ErbB2 signaling 2002;21:78–86. inhibits breast and prostate tumor growth. Cancer Cell 2002;2:127–37. 6. Gale NW, Kaplan S, Lowenstein EJ, Schlessinger J, Bar-Sagi D. Grb2 mediates the 37. Jackson JG, White MF, Yee D. Insulin receptor substrate-1 is the predominant EGF-dependent activation of guanine nucleotide exchange on Ras. Nature (Lond.) signaling molecule activated by insulin-like growth factor-I, insulin, and interleukin-4 1993;363:88–92. in estrogen receptor-positive human breast cancer cells. J Biol Chem 1998;273:9994– 7. Li N, Batzer A, Daly R, et al. Guanine-nucleotide-releasing factor hSos1 binds to 10003. Grb2 and links receptor tyrosine kinases to Ras signalling. Nature (Lond.) 1993;363: 38. Twentyman PR, Luscombe M. A study of some variables in a tetrazolium dye (MTT) 85–8. based assay for cell growth and chemosensitivity. Br J Cancer 1987;56:279–85. 8. Basu T, Warne PH, Downward J. Role of Shc in the activation of Ras in response to 39. Brattain MG, Levine AE, Chakrabarty S, et al. Heterogeneity of human colon epidermal growth factor and nerve growth factor. Oncogene 1994;9:3483–91. carcinoma. Cancer Metastasis Rev 1984;3:177–91. 9. Wilkinson MG, Millar JB. Control of the eukaryotic cell cycle by MAP kinase 40. Schaefer G, Fitzpatrick VD, Sliwkowski MX. ␥-Heregulin: a novel heregulin isoform signaling pathways. FASEB J 2000;14:2147–57. that is an autocrine growth factor for the human breast cancer cell line, MDA-MB- 10. Roymans D, Slegers H. Phosphatidylinositol 3-kinases in tumor progression. Eur 175. Oncogene 1997;15:1385–94. J Biochem 2001;268:487–98. 41. Le XF, Claret FX, Lammayot A, et al. The role of cyclin-dependent kinase inhibitor 11. Hu P, Margolis B, Skolnik EY, et al. Interaction of phosphatidylinositol 3-kinase- p27Kip1 in anti-HER2 antibody-induced G cell cycle arrest and tumor growth associated p85 with epidermal growth factor and platelet-derived growth factor 1 inhibition. J Biol Chem 2003;278:23441–50. receptors. Mol Cell Biol 1992;12:981–90. 42. Yakes FM, Chinratanalab W, Ritter CA, et al. Herceptin-induced inhibition of 12. Kim HH, Sierke SL, Koland JG. Epidermal growth factor-dependent association of phosphatidylinositol-3 kinase and Akt is required for antibody-mediated effects on phosphatidylinositol 3-kinase with the erbB3 gene product. J Biol Chem 1994;269: 24747–55. p27, cyclin D1, and antitumor action. Cancer Res 2002;62:4132–41. 13. Soltoff SP, Cantley LC. p120cbl is a cytosolic adapter protein that associates with 43. Wu K, Wang C, D’Amico M, et al. Flavopiridol and trastuzumab synergistically phosphoinositide 3-kinase in response to epidermal growth factor in PC12 and other inhibit proliferation of breast cancer cells: association with selective cooperative cells. J Biol Chem 1996;271:563–7. inhibition of cyclin D1-dependent kinase and Akt signaling pathways. Mol Cancer 14. Holgado-Madruga M, Emlet DR, Moscatello DK, Godwin AK, Wong AJ. A Grb2- Ther 2002;1:695–706. associated docking protein in EGF- and insulin-receptor signalling. Nature (Lond.) 44. Jones JT, Akita RW, Sliwkowski MX. Binding specificities and affinities of egf 1996;379:560–4. domains for ErbB receptors. FEBS Lett 1999;447:227–31. 15. Laffargue M, Raynal P, Yart A, et al. An epidermal growth factor receptor/Gab1 45. Ram TG, Ethier SP. Phosphatidylinositol 3-kinase recruitment by p185erbB-2 and signaling pathway is required for activation of phosphoinositide 3-kinase by lyso- erbB-3 is potently induced by neu differentiation factor/heregulin during mitogenesis phosphatidic acid. J Biol Chem 1999;274:32835–41. and is constitutively elevated in growth factor-independent breast carcinoma cells 16. Rodrigues GA, Falasca M, Zhang Z, Ong SH, Schlessinger J. A novel positive with cerbB-2 gene amplification. Cell Growth Differ 1996;7:551–61. feedback loop mediated by the docking protein Gab1 and phosphatidylinositol 3-ki- 46. Ram TG, Hosick HL, Ethier SP. Heregulin-␤ is especially potent in activating nase in epidermal growth factor receptor signaling. Mol Cell Biol 2000;20:1448–59. phosphatidylinositol 3-kinase in nontransformed human mammary epithelial cells. 17. Kong M, Mounier C, Wu J, Posner BI. Epidermal growth factor-induced phosphati- J Cell Physiol 2000;183:301–13. dylinositol 3-kinase activation and DNA synthesis. Identification of Grb2-associated 47. Stern DF. Tyrosine kinase signalling in breast cancer: ErbB family receptor tyrosine binder 2 as the major mediator in rat hepatocytes. J Biol Chem 2000;275:36035–42. kinases. Breast Cancer Res 2000;2:176–83. 2608

48. Normanno N, Bianco C, Damiano V, et al. Growth inhibition of human colon and -independent recruitment of Gab1 to receptor tyrosine kinases. J Biol Chem carcinoma cells by combinations of anti-epidermal growth factor-related growth 2000;275:31536–45. factor antisense oligonucleotides. Clin Cancer Res 1996;2:601–9. 57. Bardelli A, Longati P, Gramaglia D, Stella MC, Comoglio PM. Gab1 coupling to the 49. Ciardiello F, Bianco R, Damiano V, et al. Antiangiogenic and antitumor activity of HGF/Met receptor multifunctional docking site requires binding of Grb2 and corre- anti-epidermal growth factor receptor C225 monoclonal antibody in combination with lates with the transforming potential. Oncogene 1997;15:3103–11. vascular endothelial growth factor antisense oligonucleotide in human GEO colon 58. Kameda H, Risinger JI, Han BB, et al. Identification of epidermal growth factor cancer cells. Clin Cancer Res 2000;6:3739–47. receptor-Grb2-associated binder-1-SHP-2 complex formation and its functional loss 50. Ciardiello F, Caputo R, Bianco R, et al. Antitumor effect and potentiation of cytotoxic during neoplastic cell progression. Cell Growth Differ 2001;12:307–18. drugs activity in human cancer cells by ZD-1839 (Iressa), an epidermal growth factor 59. Itoh M, Yoshida Y, Nishida K, et al. Role of Gab1 in heart, placenta, and skin receptor-selective tyrosine kinase inhibitor. Clin Cancer Res 2000;6:2053–63. development and growth factor- and cytokine-induced extracellular signal-regu- 51. Poller DN, Spendlove I, Baker C, et al. Production and characterization of a poly- lated kinase mitogen-activated protein kinase activation. Mol Cell Biol 2000;20: clonal antibody to the c-erbB-3 protein: examination of c-erbB-3 protein expression 3695–704. in adenocarcinomas. J Pathol 1992;168:275–80. 60. Lee KF, Simon H, Chen H, et al. Requirement for neuregulin receptor erbB2 in neural 52. Kapitanovic S, Radosevic S, Slade N, et al. Expression of erbB-3 protein in colorectal and cardiac development. Nature (Lond.) 1995;378:394–8. adenocarcinoma: correlation with poor survival. J Cancer Res Clin Oncol 2000;126: 205–11. 61. Worthylake R, Opresko LK, Wiley HS. ErbB-2 amplification inhibits down-regula- 53. Natali PG, Nicotra MR, Bigotti A, et al. Expression of the p185 encoded by HER2 tion and induces constitutive activation of both ErbB-2 and epidermal growth factor oncogene in normal and transformed human tissues. Int J Cancer 1990;45:457–61. receptors. J Biol Chem 1999;274:8865–74. 54. Ciardiello F, Kim N, Saeki T, et al. Differential expression of epidermal growth 62. Wada T, Qian XL, Greene MI. Intermolecular association of the p185neu protein and factor-related proteins in human colorectal tumors. Proc Natl Acad Sci USA 1991; EGF receptor modulates EGF receptor function. Cell 1990;61:1339–47. 88:7792–6. 63. Cifone MA, Fidler IJ. Correlation of patterns of anchorage-independent growth with 55. Thor AD, Liu S, Edgerton S, et al. Activation (tyrosine phosphorylation) of ErbB-2 in vivo behavior of cells from a murine fibrosarcoma. Proc Natl Acad Sci USA (HER-2/neu): a study of incidence and correlation with outcome in breast cancer. 1980;77:1039–43. J Clin Oncol 2000;18:3230–9. 64. Riese DJ, Kim ED, Elenius K, et al. The epidermal growth factor receptor couples 56. Lock LS, Royal I, Naujokas MA, Park M. Identification of an atypical Grb2 carboxyl- transforming growth factor-␣, heparin-binding epidermal growth factor-like factor, terminal SH3 domain binding site in Gab docking proteins reveals Grb2-dependent and amphiregulin to Neu, ErbB-3, and ErbB-4. J Biol Chem 1996;271:20047–52.

2609

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2004 American Association for Cancer Research. Blockade of Epidermal Growth Factor- or Heregulin-Dependent ErbB2 Activation with the Anti-ErbB2 Monoclonal Antibody 2C4 Has Divergent Downstream Signaling and Growth Effects

James G. Jackson, Patricia St. Clair, Mark X. Sliwkowski, et al.

Cancer Res 2004;64:2601-2609.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/7/2601

Cited articles This article cites 64 articles, 34 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/7/2601.full#ref-list-1

Citing articles This article has been cited by 24 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/7/2601.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/7/2601. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.