The basic biology of HER2

Annals of Oncology 12 (Suppl. I): S3-S8, 2001. © 2001 Kluwer Academic Publishers. Primed in the Netherlands.

Symposium article The basic biology of HER2 I. Rubin & Y. Yarden Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, Israel

Summary

Introduction

The neu gene was originally identified in rat neuroectodermal tumors [1] and later its close human relative was isolated. This human gene encodes a protein that has a structure consistent with a growth factor receptor and has been designated HER2, due to its similarity to the human epidermal growth factor receptor [2, 3]. The HER2 receptor has an important role in normal cell growth and differentiation. However, amplification of the HER2 gene leads to overexpression of the receptor, which is linked to the development of many types of human cancers including breast, ovarian and those of the gastrointestinal tract [4, 5]. Breast cancer is a leading cause of cancer-related death in women. In many cases, the cancer is resistant to conventional drugs, resulting in treatment failure. The prognosis for patients with breast cancer is determined by well-recognized pathologic features associated with aggressiveness, histologic grade, tumor size and nodal involvement. Approximately 20%-30% of breast cancers overexpress the HER2 receptor, and it is now recognized that high levels of HER2-receptor expression are an indicator of poor prognosis in patients with breast cancer. Furthermore, overexpression of HER2 is associated with aggressive tumor growth and metastatic activity [4]. Slamon et al. [6] first observed that HER2gene amplification independently predicted overall (OS) and disease-free survival (DFS) and these findings have

Key words: breast cancer, HER2 (erbB-2, neu), network, signal transduction, tumor growth, tyrosine phosphorylation

been confirmed by subsequent studies [7]. These patients have poor response rates and short DFS and OS compared with women whose tumors do not overexpress this receptor. There have been attempts to use HER2 status as a predictive factor for response to hormonal and chemotherapy in breast cancer. Although the presence of estrogen receptors indicates response to hormonal therapy, HER2 overexpression is associated with poor response to tamoxifen, possibly because of different growth-signaling pathways. In contrast, HER2 overexpression may be associated with a better response to chemotherapy, particularly anthracyclines, than expected. Therefore, HER2 status is now a well-characterized indicator of aggressiveness and poor prognosis. Moreover, HER2 is being investigated as a target for new breast cancer treatments. Monoclonal antibodies directed against the HER2 receptor are currently the most promising treatments. It is important to understand the biology of the receptor in order to elucidate the mechanism of action of these newly-developed treatments. The HER family of receptors The HER/erbB signaling network has developed in parallel with the evolution of complex life forms. The ancestors of HER receptors may be seen in invertebrate organisms such as Caenorhabiditis elegans (C. elegans)

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

Human epidermal growth factor receptors (HER/erbB) constitute a family of four cell surface receptors involved in transmission of signals controlling normal cell growth and differentiation. A range of growth factors serve as ligands, but none is specific for the HER2 receptor. HER receptors exist as both monomers and dimers, either homo- or heterodimers. Ligand binding to HER1, HER3 or HER4 induces rapid receptor dimerization, with a marked preference for HER2 as a dimer partner. Moreover, HER2-containing heterodimers generate intracellular signals that are significantly stronger than signals emanating from other HER combinations. In normal cells, few HER2 molecules exist at the cell surface, so few heterodimers are formed and growth signals are relatively

weak and controllable. When HER2 is overexpressed multiple HER2 heterodimers are formed and cell signaling is stronger, resulting in enhanced responsiveness to growth factors and malignant growth. This explains why HER2 overexpression is an indicator of poor prognosis in breast tumors and may be predictive of response to treatment. HER2 is a highly specific and promising target for new breast cancer treatments. The recombinant human anti-HER2 monoclonal antibody (rhuMAb-HER2, trastuzumab, Herceptin) induces rapid removal of HER2 from the cell surface, thereby reducing its availability to heterodimers and reducing oncogenicity.

Extracellular domain (632 amino acids) Ligand-binding site

7

High affinity/ narrow specificity

Low affinity/ broad specificity

Intracellular domain (580 amino acids) Tyrosme kinase activity Cytoplasm

Figure I. Structure of the HER2 protein and its transmembrane topology.

NDF-p: erf>B-4>erbB-3 TGF-a: e/fcB-1

ert>B-2>ert>B-4=erbB-3>ert>B-1 ert>B-2>ert>B-1>erfjB-4>ert>B-3

Figure 2. Bivalence of EGF-like ligands. The model refers to neuregulin-1 and is based on results described in Tzahar et al. [19], (Reproduced by permission of Elsevier Science from Tzahar and Yarden [8].)

HER2 is a co-receptor for many different ligands and is often transactivated by EGF-like ligands, resulting in the formation of HER1-HER2 heterodimers [12, 13]. Neuregulins can induce the formation of HER2-HER3 and HER2-HER4 heterodimers [11, 14-16]. However, no known ligand can promote the HER2 homodimer (HER2-HER2), suggesting that no existing ligand binds directly to HER2. In support of this theory, analysis of representative pox virus growth factors, HER ligands of viral origin, revealed that none bind directly to HER2, but all are able to recruit HER2 into heterodimers [17]. This observation led us to propose that HER2 functions as a ligand-less receptor for many, if not all, HER ligands [18]. Inter-receptor interactions within the HER family are not random. Instead, there is a distinct hierarchy that prefers HER2 as an interaction partner [16]. An examination of the structure of one of the ligands helps Ligands of HER receptors to explain this preference. Neuregulin-1 (NDF) seems to An understanding of the action of ligands of HER be a bivalent molecule with two binding sites for HER receptors is crucial to understanding the role of the receptors: a high affinity/narrow specificity site (Nreceptors themselves. The HER receptors exist as mono- terminal) and a low affinity with broad specificity site mers. However, on ligand binding they form receptor (C-terminal) [19]. The high affinity site binds first to its dimers, which can either be homodimers with the same specific site (HER3 or HER4 receptor). Once the low receptor type (e.g., HER1-HER1) or heterodimers with affinity site is effectively 'immobilized' at the plasma different receptor types (e.g., HER1-HER2) [10], Dimer membrane, its operational affinity to potential partners formation is driven by the higher stability of the com- of the dimer is increased. Importantly, the receptor that plex formed between a ligand and two receptors com- preferentially binds to the immobilized low affinity arm of the ligand is HER2 (Figure 2). As a result of pared with the monomeric receptor. The HER1 receptor has many roles and is activated this heterodimerization, the HER2 receptor is able to by six ligands: epidermal growth factor (EGF), trans- participate in signal transduction in the absence of a forming growth factor a (TGFa), amphiregulin, hepa- specific ligand. This preferential binding with HER2 is rin-binding EGF-like growth factor, betacellulin and also enhanced by the overexpression of HER2 in human epiregulin. HER3 and HER4 receptors bind neuregulins cancer cells [8]. (NGFs), a family of structurally diverse peptides [10]. Heterodimers generate more potent signals than Signaling by all neuregulins is directed through mitogen- homodimers, and those containing HER2 have a particactivated protein kinase (MAPK) [11]. ularly high ligand binding and signaling potency com-

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

and Drosophila melanogaster. These organisms only have one HER receptor. Although the receptor in C. elegans has one ligand, the receptor in fruit flies {Drosophila melanogaster) has four ligands. In contrast, the mammalian HER receptor family comprises four closely-related growth factor receptors that show a high degree of homology with each other. The receptors are all located on the cell membrane and are found in a variety of tissues. They interact with a range of ligands, all sharing an EGF-like motif of 50-55 amino acids, including six highly conserved cysteine residues [8]. The HER2 receptor is encoded by the HER2 gene, a proto-oncogene mapped to chromosome 17q21 [2]. The HER2 receptor is a 1255 amino acid, 185kD transmembrane glycoprotein also designated as pl85 HER [2]. All four HER receptors comprise a cysteine-rich extracellular ligand binding site, a transmembrane lipophilic segment, and an intracellular domain with tyrosine kinase catalytic activity (Figure 1) [9]. A terminal carboxyl autophosphorylation segment is most likely to be responsible for translation of the activation signals initiated by extracellular ligand binding into physiologic action.

oo oo oo Proliferation index

3.2

5.0

3.1

o® o® 3.1

6.5

10.5

Figure 3 Relative potency of HER dimers. Note that H ER4-containing combinations were omitted. The mitogenic index of each HER dimer was calculated on the basis of experiments described in Pinkas-Kramarski et al. [12]. (Reproduced by permission of Oxford University Press from Pinkas-Kramarski et al. [12].)

ert>B-1

e/6B-1T3

erbB-3

EC

TK

CT

Signal

No signal

Figure 4. HER1-HER3 chimera is more potent than any other receptor combination. The presented erb-lT3 artificial fusion protein was designed and tested in animal cells. Its activity exceeded that of the HER2-HER3 combination. (Reproduced by permission of Oxford University Press from Waterman et al. [24].)

Negative regulation of HER receptors and its relevance to cancer immunotherapy

Of the 10 different HER homo- and heterodimer combinations, those containing HER2 are long-lived and pared with hetero- and homodimers without HER2 [20, transmit strong signals, and are thus associated with 21]. The increased potency of HER2-containing hetero- malignant growth (Figure 5) [5]. As mentioned earlier, dimers is attributable to two reasons: first, the hetero- overexpression of HER2 promotes the formation of dimers are characterized by a relatively slow rate of more HER2 heterodimers. By contrast, the non-HER2 ligand dissociation [10], and second, unlike HER1 whose combinations signal relatively weakly. This signaling rate of ligand-induced endocytosis is rapid, HER2 is a is normal, and these combinations are essential for slowly internalizing receptor [12, 22]. Thus, signaling by normal cell growth but will not lead to tumor growth HER2-containing receptor combinations is relatively (Figure 5) [5]. In mice carrying human tumor cells overexpressing prolonged and results in enhanced activation of signaling pathways such as the MAPK route. The most potent a rodent oncogenic mutant of HER2, the growth of heterodimer is HER2-HER3 (Figure 3). This combina- tumors was inhibited by monoclonal antibodies directed tion is the most mitogenic and HER3 is the predom- at the HER2 receptor [25]. Tumor regression can be inant partner of HER2 in carcinoma cells [10, 11]. This demonstrated also in a cell system that overexpresses demonstrates how heterodimerization can be used to the human HER2. When a number of different anticontrol cell activity. The HER 3 receptor has only a bodies to HER2 were examined, the rate of tumor weakly active kinase domain [23] and HER3 homo- regression depended on type of antibody; some antidimers are completely inactive [12]. However, HER3 bodies almost completely inhibited tumor growth, while has many binding sites for signaling proteins and it is others were only partial antagonists [26]. However, possible that an active kinase with so many binding sites enhancement of growth also occurred upon treatment would be harmful to the cell, so control is maintained by of cells with certain monoclonal antibodies. In these latter formation of the heterodimer requiring both a ligand for cases, labeling of tumor cells or the antibodies revealed HER3 and the HER2 partner. This hypothesis was that the tumor inhibitory antibodies were not processed tested using a chimeric receptor comprised of the kinase by the tumor cells. By contrast, tumor-inhibitory antidomain of HER1 fused to the signaling domain of HER3 bodies were rapidly internalized and metabolized by

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

TM JM

[24]. As predicted, this engineered HER1-HER3 chimera was more active than any other HER protein or HER combination (Figure 4). In conclusion, signal transduction by the HER family may be considered in terms of information processing by networks. The network is a layered organization of units. Thus, the HER network consists of an input layer (ligands, growth factors), information processing layer (receptors, SH2-proteins, transcription factors), and an output layer (cell growth, differentiation or migration). Thus, receptor dimerization is important as it allows a signaling network with an enormous potential for diversification of biologic messages instead of four receptors with only four linear pathways. In effect, HER2 may be considered as a master regulator of the network because it promotes and controls signaling.

EGF-like ligands (>30) erbB dimers (10 types)

SH2/PTB adaptors (e.g., Src, She) Enzyme cascades (e.g.MAPK.JNK)

Malignant growth Figure 5. The HER signaling network. The layered structure of the HER signaling network is presented. Numbers refer to the four HER proteins. Note that HER3 carries an inactive catalytic domain. (Reproduced with permission from Klapper et al. [5].)

cancer cells overexpressing HER2 [27]. These observations led us to suggest that tumor inhibition is due to the endocytic removal of HER2 from the cell surface [5]. Consistent with this model, combining two distinct antibodies to HER2 better inhibits tumors than each antibody alone, probably because the combination leads to accelerated degradation of HER2 inside tumor cells [28]. In conclusion, the anti-tumor effect of HER2specific antibodies is due to the ability of certain antibodies to accelerate internalization and degradation of the oncoprotein. The attribution of immunotherapy to enhanced endocytosis focused the interest on processes leading to HER2 degradation. Generally, HER receptors are continuously recycled back to the cell surface after endocytosis through endosomes. Alternatively, the receptor may be degraded after passing from endosome to lysosome. The recycling/degradation pathway is controlled through an adaptor molecule, c-Cbl, which is a major tyrosine phosphorylation substrate of HER [29]. c-Cbl is the normal form of this sorting molecule and is needed for the HER1 receptor to move from endosome to lysosome and thus be degraded [30]. Unlike HER1, which is strongly coupled to c-Cbl, HER2 only weakly interacts with the adaptor, but the two neuregulin receptors, HER3 and HER4, do not recruit c-Cbl [31]. The biochemical basis of c-Cbl's action has been recently resolved. The ring finger domain of c-Cbl, a motif detected in two oncogenic forms of c-Cbl, serves as a binding site for a ubiquitin processing enzyme (E2).

Thereby, when c-Cbl is recruited to a specific tyrosine phosphorylated residue of HER1, it enhances polyubiquitination of the receptor (Figure 6). Polyubiquitin tracts are sufficient for targeting HER1 to proteasomal and lysosomal degradation [32]. It should be noted that HER3 is continuously recycled [33] because is does not have an active kinase domain, nor does it carry a c-Cbl docking site. Our most recent results suggest that antitumor antibodies tag the HER2 receptor to degradation by recruiting c-Cbl and enhance ubiquitination of HER2.

Conclusions: Implications of HER2 biology for the treatment of breast cancer

In conclusion, the HER2 gene and HER2 receptor play a crucial role in the network of cell-signaling processes controlling normal growth and development. The HER family binds a range of ligands, and although HER2 does not bind a specific ligand, it is the preferred partner for receptor dimerization and this enables it to perform its role through growth factor binding at its partner receptor. The role of HER2 in cell growth can be summarized in simple terms. When HER2 is normally expressed, ligands that bind to the HER receptors form only a few HER2 heterodimers, and the responses to these growth factors are relatively weak, resulting in normal growth of cells. However, when HER2 is overexpressed as in cancer cells, many ligands originating primarily in the

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

Transcription factors teg., Egr-1. Sp-1)

Transfection

Cont.

c-Cbl

ing breast tumors that overexpress HER2 have been investigated, including tyrosine kinase inhibitors, antiEGF sense approaches, designed to downregulate expression of the HER2 gene, and immunization to actively boost anti-HER2 responses. In addition, selective targeting can be achieved using monoclonal antibodies directed against the extracellular domain of the HER2 protein. As HER2 is expressed only in low levels in normal tissue IB Ab: anti Ub [37], this permits a suitable therapeutic window to minimize damage to normal cells. A number of murine monoclonal antibodies directed against the HER2 receptor have been developed. The monoclonal antibodies inhibited tumor growth in laboratory model systems. The mechanism of action of their anti-tumor effects is not yet completely understood although, as suggested earlier in this paper, the monoclonal antibodies may direct HER2 towards a degradation pathway. However, the murine antibodies caused IB Ab: anti erfjB-1 the development of neutralizing human anti-mouse antibodies (HAMA) and did not progress beyond experimental stages. Since these experiments, the murine 4D5 monoclonal antibody has been humanized [38], resultIP Ab: anti ert>B-1 ing in the recombinant human anti-HER2 monoclonal Figure 6. c-Cbl controls HER degradation and recycling. Chinese hamster ovary cells were transfected with plasmids directing the antibody (rhuMAb-HER2, trastuzumab, Herceptin). expression of HER1. Some cells were transfected also with a c-Cbl This antibody is being introduced into clinical practice expression vector. Following 40 hours, cells were treated with EGF for patients with metastatic breast cancer. It is not (100 ng/ml) or left untreated for 10 minutes. Cell extracts were associated with the development of HAMA and was examined by immunoprecipitation (IP) and immunoblotting (IB) with effective in phase II and III clinical trials. In addition to the indicated antibodies. Closed arrowheads mark the location of the being active as a single agent, trastuzumab potentiates HERl/erfeB-1 protein band, and the open arrowhead marks the ubiqthe anti-tumor activity of paclitaxel and doxorubicin uitinated protein. [39] and cisplatin [40]. Therefore, monoclonal antibodies are the potential stroma or in the tumor cells themselves will recruit greatest advance in the treatment of tumors overHER2 into heterodimers. The heterodimers stay longer expressing HER2. Elucidation of the role of HER2 in at the cell surface, and their signaling is enhanced. cell growth should allow the mechanism of action of This results in potent stroma-to-epithelium signaling, monoclonal antibodies to be determined and should enhanced responsiveness to growth factors and, eventu- help optimize treatment of aggressive HER2-positive ally, malignant growth. tumors. In breast cancers, 92% of cases of HER2 overexpression result from HER2-gene amplification (generation of more than the normal two gene copies) [34]. This Acknowledgements leads to increased transcription of the HER2 gene and concomitant increased synthesis of HER2 protein. In Research in our laboratory is supported by the National cancer cells, HER2 protein levels may be 100 times those Institutes of Health and the Israel Basic Research Fund. in normal cells [35]. This is a consequence of gene amplification and/or transcriptional alterations. Amplification is also the most common cause of overexpres- Note sion in ovarian and gastric tumors [4]. The mechanism of gene amplification has not been determined. How- Dr Yarden has reported that he is a consultant for Novartis and has ever, the HER2 gene is not amplified in a few tumor received research support from Genentech, Inc. types (lung, mesenchymal, bladder and esophageal) overexpressing HER2 and, in these cases, overexpression may result from transcriptional or post-transcriptional References dysregulation [36]. 1. Shih C, Padhy LC, Murray M,Weinberg RA. Transforming genes As HER2 is frequently overexpressed in human tumors of carcinomas and neuroblastomas introduced into mouse fibroand often confers a more aggressive clinical course, blasts. Nature 1981; 290: 261-4. HER2 is being investigated as a target for cancer therapy. 2. Coussens L, Yang-Feng TL, Liao Y-C et al. Tyrosine kinase Its localization at the cell surface makes it an easy target receptor with extensive homology to EGF receptor shares chroto access. A wide range of therapeutic strategies targetmosomal location with neu oncogene. Science 1985; 230: 1132-9.

I

II

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

24.

25.

26.

27.

28.

29. 30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

pl8OErbB3 possesses an impaired tyrosine kinase activity. Proc Natl Acad Sci USA 1994; 91: 8132-6. Waterman H, Alroy I, Strano S et al. The carboxyl terminus of the kinase-defective neuregulin receptor ErbB-3 confers mitogenic superiority and dictates endocytic routing. EMBO J 1999; 18: 3348-58. Drebin JA, Link VC, Stern DF et al. Down-modulation of an oncogene protein product and reversion of the transformed phenotype by monoclonal antibodies. Cell 1985; 41: 697-706. Stancovski I, Hurwitz E, Leitner O et al. Mechanistic aspects of the opposing effects of monoclonal antibodies to the ERBB2 receptor on tumor growth. Proc Natl Acad Sci USA 1991; 88: 8691-5. Hurwitz E, Stancovski I, Sela M, Yarden Y. Suppression and promotion of tumor growth by monoclonal antibodies to ErbB-2 differentially correlate with cellular uptake. Proc Natl Acad Sci USA 1995; 92: 3353-7. Kasprzyk PG, Song SU, Di Fiore PP, King CR. Therapy of an animal model of human gastric cancer using a combination of anti-erbB-2 monoclonal antibodies. Cancer Res 1992; 52: 2771-6. Thien CBF, Langdon W. c-Cbl: A regulator of T-cell receptormediated signaling. Immunol Cell Biol 1998; 76: 473-82. Levkowitz G, Waterman H, Zamir L et al. c-Cbl/Sli-1 regulates endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev 1998; 12: 3663-74. Levkowitz G, Klapper LN, Tzahar E et al. Coupling of the c-Cbl protooncogene product to ErbB-1 /EGF-receptor but not to other ErbB proteins. Oncogene 1996; 12: 1117-25. Levkowitz G, Waterman H, Ettenberg SA et al. Ubiquitin ligase activity and tyrosine phosphorylation underlie suppression of growth factor signaling by c-Cbl /Sli-1. Mol Cell 1999; 4: 1029-40. Waterman H, Sabanai I, Geiger B, Yarden Y. Alternative mtracellular routing of ErbB receptors may determine signaling potency. J Biol Chem 1998; 273: 13819-27. Pauletti G, Godolphin W, Press MF, Slamon DJ. Detection and quantitation of HER-2/neu gene amplification in human breast cancer archival material using fluorescence in situ hybridization. Oncogene 1996; 13: 63-72. Venter DJ, Tuzi NL, Kumar S, Gullick WJ. Overexpression of the c-erbB-2 oncoprotein in human breast carcinomas: Immunohistological assessment correlates with gene amplification. Lancet 1987; ii: 69-72. Slamon D, Godolphin W, Jones LA et al. Studies of the HER-2/ neu protooncogene in human breast and ovarian cancer. Science 1989; 244. 707-12. De Potter CR, Schelfhout AM, Verbeeck P et al. Neu overexpression correlates with extent of disease in large-cell ductal in situ carcinoma of the breast. Hum Pathol 1995; 26: 601-6. Carter P, Presta L, Gorman CM et al. Humanization of the antipl85 H E R 2 antibody for human cancer therapy. Proc Natl Acad Sci USA 1992; 89: 4285-9. Baselga J, Norton L, Albanell J et al. Recombinant humanized anti-HER2 antibody (Herceptin™) enhances the antitumor activity of paclitaxel and doxorubicin against HER2/«e« overexpressing human breast cancer xenografts. Cancer Res 1998; 58: 2825-31. Pietras RJ, Fendly BM, Chazin VR et al. Antibody to HER-2/neu receptor blocks DNA repair after cisplatin in human breast and ovarian cancer cells. Oncogene 1994; 54: 1829-38.

Correspondence to • Y. Yarden, PhD Department of Biological Regulation The Weizmann Institute of Science Rehovot 76100 Israel E-mail: [email protected]

Downloaded from http://annonc.oxfordjournals.org/ at Serials Acq (Law) on May 6, 2012

3. King CR, Kraus MH, Aaronson SA. Amplification of a novel v-erAB-related gene in a human mammary carcinoma. Science 1985; 229: 974-6. 4. Hynes NE, Stern DF. The biology oferbB-2/neu/HER-2 and its role in cancer. Biochem Biophys Acta Rev Cancer 1994; 1198: 165-84. 5. Klapper LN, Kirschbaum MH, Sela M, Yarden Y. Biochemical and clinical implications of the ErbB/HER signaling network of growth factor receptors. In Klein G, Woude V (eds): Advances in Cancer Research. New York: Academic Press 2000; 25-79. 6. Slamon DJ, Clark GM, Wong SG et al. Human breast cancer: Correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987; 235: 177-82. 7. Ross JS, Fletcher JA. The HER-2/neu oncogene in breast cancer: Prognostic factor, predictive factor, and target for therapy. Stem Cells 1998; 16: 413-28. 8. Tzahar E, Yarden Y. The ErbB-2/HER2 oncogenic receptor of adenocarcinomas: From orphanhood to multiple stromal hgands. BBA Rev Cancer 1998; 1377: M25-37. 9. van de Geer P, Hunter T, Lindberg RA. Receptor protein-kinases and their signal transduction pathways. Annu Rev Cell Biol 1994; 10: 251-337. 10. Alroy I, Yarden Y. The ErbB signaling network in embryogenesis and oncogenesis: Signal diversification through combinatorial ligand-receptor interactions. FEBS Lett 1997; 410: 83-6. 11 Pinkas-Kramarski R, Shelly M, Guarino BC et al. ErbB tyrosine kinases and the two neuregulin families constitute a ligandreceptor network. Mol Cell Biol 1998; 18: 6090-101. 12 Pinkas-Kramarski R, Soussan L, Waterman H et al. Diversification of Neu differentiation factor and epidermal growth factor signaling by combinatorial receptor interactions. EMBO J 1996; 15: 2452-67. 13. Pinkas-Kramarski R, Eilam R, Alroy I et al. Differential expression of NDF/neuregulin receptors ErbB-3 and ErbB-4 and involvement in inhibition of neuronal differentiation. Oncogene 1997; 15: 2803-15. 14. Baly DL, Wirth CM, Allison DA et al. Development and characterization of a rhuMAb HER2 antibody assay for clinical evaluation of cytotoxic potency. Proc Am Assoc Cancer Res 1997; 38: 181A. 15. Burden S, Yarden Y. Neuregulins and their receptors: A versatile signaling module in organogenesis and oncogenesis. Neuron 1997; 18: 847-55. 16. Tzahar E, Waterman H, Chen X et al. A hierarchical network of interreceptor interactions determines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Mol Cell Biol 1996; 16: 5276-87. 17. Tzahar E, Guarino BC, Waterman H et al. Pathogenic poxviruses reveal viral strategies to exploit the ErbB signaling network. EMBOJ 1998; 17:5948-63. 18. Klapper LN, Glathe S, Vaisman N et al. The ErbB-2/HER2 oncoprotein of human carcinomas may function solely as a shared coreceptor for multiple stroma-denved growth factors. Proc Natl Acad Sci USA 1999; 96: 4995-5000. 19. Tzahar E, Pinkas-Kramarski R, Moyer JD et al. Bivalence of EGF-like ligands drives the ErbB signaling network. EMBO J 1997; 16: 4938-50. 20. Karunagaren D, Tzahar E, Beerli RR et al. ErbB-2 is a common auxiliary subunit of N D F and EGF receptors: Implications for breast cancer. EMBOJ 1996; 15: 254-64. 21. 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: 14661-5. 22. Baulida J, Kraus MH, Alimandi M et al. All ErbB receptors other than the epidermal growth factor receptor are endocytosis impaired. J Biol Chem 1996; 271: 5251-7. 23. Guy PM. Platko JV, Cantley LC et al. Insect cell-expressed