Blockade of Epidermal Growth Factor- or Heregulin-Dependent ErbB2 Activation with the Anti-ErbB2 Monoclonal Antibody 2C4 Has Divergent Downstream Signaling and Growth Effects

Reagents.

All chemicals and reagents were purchased from Sigma (St. Louis, MO) unless noted otherwise. Gab1 and p85 PI3K antibody were purchased from Upstate Biotechnology (Lake Placid, NY). Total and phospho-specific Akt antibodies were purchased from New England Biolabs (Beverly, MA). ErbB1 for immunoprecipitation (Ab12) and ErbB3 (Ab5) antibodies were from Neomarkers (Monroe, CA). Gab2, phospho-Erk1/2, total Erk1/2, and ErbB2 antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). ErbB1 for immunoblot, cbl, and RC-20 anti-phospho-tyrosine were purchased from Transduction Laboratories (Lexington, KY). Horseradish peroxidase-linked antirabbit and antimouse antibodies and molecular weight markers were from Amersham Pharmacia Biotech (Piscataway, NJ). Protein A-agarose was from Life Technologies, Inc. (Bethesda, MD). Acrylamide was from Fisher Biotech (Hampton, NH). GEO cells have been described previously (32) . 2C4 and Herceptin were obtained from Genentech, Inc. (South San Francisco, CA) and have been described previously (33, 34, 35) . A humanized variant of 2C4 (recombinant human mAb 2C4) was used in all studies described in this report (36) .

Cell Stimulation and Lysis.

GEO cells were regularly cultured in supplemental McCoy’s medium containing insulin, EGF, and transferrin (SF media). Before signaling experiments, cells were starved for 2–4 days in supplemental McCoy’s medium with no growth factors (SM media). For treatment with 2C4, cells were preincubated for 30 min in SM media with 20 μg/ml antibody, and then 50 ng/ml growth factor was added for the indicated times. Cells were washed twice in ice-cold PBS and lysed with 500 μl per 10-cm plate of 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 mixture tablet (Boehringer Mannheim, Mannheim, Germany) per 10 ml of lysis buffer]. Protein concentration of the supernatants was determined by the copper- bicinchoninic acid method with a Pierce Laboratory kit (Rockford, IL).

Immunoprecipitations.

All steps were performed with gentle rotation at 4°C. Because of the presence of 2C4 antibody in some treatment groups, lysates used in immunoprecipitation were not precleared with protein A-agarose. However, no effect from the presence of 2C4 was observed on the immunoprecipitation of any proteins. Protein from TNESV lysates (250 μg) was incubated with 2.5 μg of the indicated antibody overnight. Twenty-five μl of protein A-agarose or protein G-agarose was then added for 4 h, followed by three washes with TNESV buffer. Beads were resuspended in 2× Laemmli loading buffer with 2.5% β-mercaptoethanol, boiled, and separated by 8% SDS-PAGE.

Immunoblotting.

After SDS-PAGE proteins were transferred overnight to nitrocellulose membranes (Amersham Pharmacia Biotech, Arlington Heights, IL), the membranes were blocked in 5% nonfat dry milk in Tris-buffered saline-Tween 20 [TBST (0.15 m NaCl, 0.01 m Tris HCl (pH 7.4), and 0.05% 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 TBST for 1 h, and washed thoroughly; chemiluminescence was then performed as described below. For phospho-Akt immunoblot, membranes were incubated overnight in TBST at 1000:1. All other blots were incubated with a 1 μg/ml dilution of the indicated antibody in blocking buffer overnight at 4°C. After washing, blots were incubated with a 2000:1 dilution of horseradish peroxidase-linked secondary antibody in blocking buffer for 1 h, followed by further washing. Enhanced chemiluminescence was performed according to the manufacturer’s instructions (Amersham Pharmacia Biotech).

PI3K Assay.

After growth factor treatment, the cell monolayers were washed with PBS and lysed, and protein concentration was determined as described previously (37) . Lysates (250 μg of protein) were incubated with 2.5 μg of primary antibody for 30 min, followed by further incubation with protein A-agarose or protein G-agarose for 2 h. Immune complexes were washed twice with each buffer PBS and 1% NP40, 100 mm Tris and 5 mm LiCl, and 10 mm Tris, 150 mm NaCl, and 5 mm EDTA; all buffers included 10 mm sodium orthovanadate. After the last wash was removed, PI3K assays were performed as described previously (37) . Briefly, samples were resuspended in 50 μl of PI3K buffer [20 mm Tris (pH 7.5), 100 mm NaCl, and 0.5 mm EGTA], and 10 μg of phosphatidylinositol were added. After 10 min at room temperature, 10 μCi of [³²P]ATP and MgCl₂ to a final concentration of 20 μm was added. After 10 min at room temperature, lipids were extracted, first with 150 μl of CHCl₃:methanol:HCl (10:20:0.2) and 100 μl of pure CHCl₃. The second extraction used 80 μl of methanol:1N HCl (1:1). Samples were spotted on 1% potassium oxalate-treated TLC plates (Analtech, Newark, DE) and developed in CHCl₃:methanol:NH₄OH:H₂O (129:114:15:21). On an autoradiograph of the plate, the highest migrating spots on the TLC plate, representing phosphatidylinositol phosphate, were quantitated by densitometry using AlphaImager software (Alpha Innotech, San Leandro, CA).

Cell Monolayer Growth.

GEO cells were plated in 24-well plates at 30,000 cells/well in SF medium and allowed to adhere overnight. The next day, cells were washed with PBS and pretreated for 30 min in SM medium with or without 20 μg/ml 2C4, and then the indicated growth factors were added at 50 ng/ml. After 5 days, growth was analyzed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay (MTT) as described previously (38) . Sixty μl of 5 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reagent in PBS were added to each well, and then, after a 2-h incubation, wells were aspirated, and 0.5 ml of Me₂SO was added. Absorbance was measured at 545 nm.

Soft Agar Assay.

Experiments were performed essentially as described previously (39) . Briefly, 25,000 cells resuspended in 0.4% low melting point agarose were plated on a 0.8% agarose underlayer in a 6-well plate in triplicate. Both layers contained treatments in SM medium as indicated in the Fig. 6 ⇓ legend. Colonies were allowed to form for 8 days and then stained with 1 ml of p-iodonitrotetrazolium violet overnight, and images were captured on AlphaImager. Images of anchorage-independent growth at ×40 magnification were captured with a Nikon microscope fitted with an Olympus MagnaFire digital camera using Image-Pro Plus software from Media Cybernetics (Silver Spring, MD).

Monoclonal Antibody 2C4 Can Inhibit Heregulin- and EGF-Stimulated Phosphorylation of ErbB2.

GEO cells express ErbB1, ErbB2, and ErbB3 but not ErbB4 (data not shown; Refs. 5 , 30 , and 31 ). To test the responsiveness of GEO cells to heregulin and EGF, we stimulated quiescent cell monolayers with these growth factors and then immunoprecipitated lysates with ErbB2 antibody and performed anti-phospho-tyrosine immunoblots. We found a low basal level of ErbB2 phosphorylation in quiescent GEO cells (Fig. 1A ⇓ , Lane 1). After treatment with EGF or heregulin for 10 min, there was a large increase in phospho-tyrosine on ErbB2 (Fig. 1A ⇓ , Lanes 2 and 5, respectively). A 30-min pretreatment with the anti-ErbB2 mAb 2C4 (20 μg/ml) strongly inhibited ErbB2 phosphorylation after similar EGF or heregulin treatment (Fig. 1A ⇓ , Lanes 4 and 7). Consistent with others’ reports (35 , 40) , pretreatment with the ErbB2 mAb Herceptin did not reduce phosphorylation of ErbB2 after growth factor treatment, ruling out a nonspecific mAb effect or a vehicle effect on ErbB2 phosphorylation (Fig. 1A ⇓ , Lanes 3 and 6). The concentration of Herceptin and 2C4 used in these studies has been shown to be saturating but nontoxic in numerous previous studies (35 , 36 , 41, 42, 43, 44) . Anti-ErbB2 immunoblot of the same membrane showed that ErbB2 levels were not significantly changed by growth factor or mAb treatment (Fig. 1A ⇓ , bottom panel).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

2C4 blocks epidermal growth factor- and heregulin-stimulated phosphorylation of ErbB2. Growth factor-starved GEO monolayers were pretreated with or without 20 μg/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 μg 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.

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 blocked only the phosphorylation of ErbB2. ErbB2 and ErbB3 were activated after heregulin treatment, and 2C4 blocked the activation of both receptors. We next investigated the downstream consequences of these results.

Activation of Erk and Akt after 2C4 Treatment and Growth Factor Stimulation.

The MAPKs are well-characterized downstream targets of ErbB receptors. We determined the activation state of MAPK Erk1/2 after 2C4 treatment and growth factor stimulation in GEO cells. Erk1/2 phosphorylation was increased after a 10-min EGF treatment (Fig. 2A ⇓ , top panel, Lane 2), and this activation was not 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 α. 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).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Epidermal growth factor and heregulin activation of mitogen-activated protein kinase, Akt, and p85 signaling after 2C4 treatment. Growth factor-starved GEO monolayers were pretreated in growth factor-free media with or without monoclonal antibody 2C4 (20 μg/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 μg of protein were separated by SDS-PAGE and immunoblotted for phospho-extracellular signal-regulated kinase 1/2 (A, top panels) followed by total extracellular signal-regulated kinase 1/2 (A, bottom panels) and phospho-Akt (B, top panels) followed by total Akt (B, bottom panels) as described in “Materials and Methods.” C, 250 μg of protein were immunoprecipitated with anti-p85 phosphatidylinositol 3′-kinase antibody, separated by SDS-PAGE, and blotted for anti-phospho-tyrosine as described in “Materials and Methods.” Molecular weight markers are indicated to the right, and migration positions of phosphoproteins coimmunoprecipitating with p85 are indicated to the left. Data shown are representative of at least two independent experiments.

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 results, 2C4 blocked this effect (Fig. 2B ⇓ , top panel, Lanes 4–6). Anti-Akt immunoblots on the same membranes showed that total levels of Akt were not significantly changed by any of the treatment groups (Fig. 2B ⇓ , bottom panels). These data show that 2C4 could block heregulin-induced signaling to both MAPK and Akt, consistent with the notion that ErbB3-ErbB2 heterodimers are essential for all aspects of heregulin signaling. In EGF signaling, however, disruption of the ErbB2 component with 2C4 only inhibited signaling to Akt, not the MAPKs Erk1/2. These results indicated that Erk1/2 activation was the consequence of ErbB1 signaling, whereas Akt activation was due at least in part to ErbB2 signaling, thus raising the question of how PI3K was activated by EGF.

Coimmunoprecipitation of Distinct Phospho-Tyrosine-Containing Proteins after EGF and Heregulin Treatment.

Growth factor-induced activation of the p85 subunit of PI3K is initiated when, via its SH2 domain, p85 binds phosphorylated tyrosine residues of receptors or adaptor molecules (10) . To find phospho-tyrosine-containing molecules bound to p85 in GEO cells, we stimulated cells as described in 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 M_r 180,000 (Lane 1). These two bands were present in all treatment groups. After heregulin stimulation, the major phosphoprotein that coimmunoprecipitated with p85 was approximately M_r 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 M_r 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 M_r 190,000 range, but we did observe an increase in M_r 120,000 and M_r 100,000 phosphoproteins associated with p85 (Fig. 2C ⇓ , Lane 2). The intensity of the M_r 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 association of p85 with ErbB3 (12) or the adaptor molecules Gab1 (14, 15, 16) and cbl (13) , which migrate at the approximate molecular weight of pp120 of Fig. 2C ⇓ , and Gab2 (17) , which migrates at the approximate molecular weight of pp100 of Fig. 2C ⇓ . To determine which of these were activated in GEO cells, we stimulated cells as described in the Fig. 1 ⇓ legend, immunoprecipitated them with specific antibodies, and then immunoblotted 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 (M_r 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. 3 ⇓ , Lane 1 versus Lane 3). These findings correlate with the coimmunoprecipitation of a major M_r 190,000 phosphoprotein and a minor M_r 120,000 phosphoprotein with p85 antibody observed in Fig. 2C ⇓ . Further immunoblotting of the membrane with p85 antibody showed that p85 was associated with Gab1 and Gab2, but not cbl, after EGF treatment and that p85 was associated with ErbB3 after heregulin treatment (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Phosphorylation of Grb2-associated binder (Gab) 1/Gab2/cbl after epidermal growth factor treatment and ErbB3/Gab1 after heregulin treatment. Quiescent GEO monolayers were stimulated for 10 min with 50 ng/ml growth factor as indicated above the lanes. Cells were lysed, and then 250 μg of protein were immunoprecipitated with the indicated antibody, separated by SDS-PAGE, and blotted for anti-phospho-tyrosine as described in “Materials and Methods.” Molecular weight markers are indicated to the right, and migration positions of specific proteins are indicated to the left. Data shown are representative of three independent experiments.

To further verify the pathways containing PI3K activity after growth factor treatment in GEO cells, we next assessed the levels of PI3K activity associated with each of the signaling molecules identified in Fig. 3 ⇓ . Consistent with the phosphorylation data in Fig. 3 ⇓ , we found that EGF treatment increased the PI3K activity associated Gab1 (Fig. 4 ⇓ , Lane 1 versus 2) and Gab2 (Fig. 4 ⇓ , Lane 4 versus 5), but not ErbB3 (Fig. 4 ⇓ , Lanes 7 and 8). Heregulin increased PI3K activity associated with ErbB3 (Fig. 4 ⇓ , Lane 7 versus Lane 9) and Gab1 (Fig. 4 ⇓ , Lane 1 versus Lane 3), which is consistent with the heregulin-induced phosphorylation of ErbB3 and Gab1 observed in Fig. 3 ⇓ . There was minimal basal activity and no increase in PI3K activity associated with cbl after EGF or heregulin treatment (data not shown). These data provide evidence that EGF activates PI3K via the Gab1 and Gab2 adaptor molecules, but not through ErbB3 or cbl, and that heregulin activates PI3K via ErbB3 and, to a lesser extent, Gab1.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Phosphatidylinositol 3′-kinase activity is associated with Grb2-associated binder (Gab) 1/Gab2 after epidermal growth factor treatment and ErbB3/Gab1 after 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 μg of protein were immunoprecipitated with the indicated antibody, and then a phosphatidylinositol 3′-kinase assay was performed as described in “Materials and Methods.” Relative positions of phosphatidylinositol phosphate (PIP) and the origin (Ori) are indicated to the left. Densitometric analysis of the data is shown graphically. Data shown are representative of at least two independent experiments.

2C4 Inhibits EGF and Heregulin Stimulated PI3K Activity.

We next examined the effects of 2C4 on heregulin- and EGF-stimulated PI3K activity. We found that the activity of PI3K associated with ErbB3 after heregulin treatment for different times (Fig. 5A ⇓ , Lanes 2, 4, and 6) could be blocked by 2C4 (Fig. 5A ⇓ , Lanes 3, 5, and 7). In cells stimulated by EGF for different times, PI3K activity associated with Gab1 (Fig. 5B ⇓ , Lanes 2, 4, and 6) or Gab 2 (Fig. 5C ⇓ , Lanes 2, 4, and 6) was partially inhibited by 2C4 (Fig. 5B ⇓ , Lanes 3, 5, and 7 for Gab1; Fig. 5C ⇓ , Lanes 3, 5, and 7 for Gab2). 2C4 was less effective at inhibiting PI3K activity associated with Gab1 and Gab2 at 1 min of EGF treatment (Fig. 5, B and C ⇓ , Lane 3), whereas the greatest degree of inhibition was at 60 min (Fig. 5, B and C ⇓ , Lane 7).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

2C4 inhibits heregulin- and epidermal growth factor-stimulated phosphatidylinositol 3′-kinase activity. Quiescent GEO monolayers were pretreated in growth factor-free media with or without monoclonal antibody 2C4 (20 μg/ml) for 30 min and then stimulated with 50 ng/ml growth factor as indicated above the lanes. Cells were lysed; 250 μg of protein were immunoprecipitated with ErbB3 (A), Grb2-associated binder (Gab) 1 (B), or Gab2 (C) antibody; and then phosphatidylinositol 3′-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.

2C4 Inhibits Heregulin-Stimulated Cell Proliferation and EGF- and Heregulin-Stimulated Anchorage-Independent Growth.

After resolving how 2C4 alters the components of heregulin and EGF signal transduction, we next investigated the effects of 2C4 treatment on cell proliferation and anchorage-independent growth, an established marker of malignant potential. In cell monolayer growth over 5 days, we found that exogenous heregulin added to growth factor-free media (Fig. 6A ⇓ , Hrg) stimulated approximately a 3–4-fold increase in cell number over the serum-free media alone (C-SM), whereas EGF stimulated a 2-fold increase. Normal growth medium (SF), which contains a high concentration of insulin, also stimulated a 3–4-fold increase in cell number. 2C4 effectively blocked heregulin-induced proliferation but had only a minimal but reproducible effect on EGF-stimulated growth and no effect on insulin-stimulated (SF) cell growth (Fig. 6A) ⇓ .

• Download figure
• Open in new tab
• Download powerpoint

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 μg/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 μg/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 “Materials 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 μg/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.

We next assessed the anchorage-independent growth of GEO cells by assaying colony formation in soft agar. We found that cells plated in the absence of growth factors formed very few colonies, whereas cells plated in normal growth medium readily formed numerous colonies (data not shown). Cells plated in medium containing heregulin also readily formed colonies, and 2C4 completely blocked this effect (Fig. 6, B and C ⇓ , left panels). EGF also stimulated the formation of colonies, and 2C4 treatment reduced the number and size of colonies (Fig. 6, B and C ⇓ , right panels).