Chimeric receptor analyses of the interactions of the ectodomains of ErbB-1 with epidermal growth factor and of those of ErbB-4 with neuregulin Jae-Hoon Kim 1, *, Kazuki Saito 1,2 and Shigeyuki Yokoyama 1,2,3 1 Yokoyama CytoLogic Project, ERATO, Japan Science and Technology Corporation, c/o Tsukuba Research Consortium, Tokodai, Tsukuba, Japan; 2 RIKEN Genomic Sciences Center, Suehiro-cho, Tsurumi, Yokohama, Japan; 3 Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Hongo, Bunkyo-ku, Tokyo, Japan A series of chimeric receptors was generated between the epidermal growth factor (EGF) receptor, ErbB-1, and its homologue, ErbB-4, to investigate the roles of the extracel- lular domains (I–IV) in the ligand specificities. As compared with ErbB-1 and the chimeras with both domains I and III of ErbB-1, the chimeras with only one of these domains exhibited reduced binding of 125 I-labeled EGF. Particularly, the contribution of domain III was appreciably larger than that of domain I of ErbB-1 in 125 I-labeled EGF binding. Nevertheless, the chimeras with domain III of ErbB-1 and domain I of ErbB-4 were prevented from binding to 125 I-labeled EGF competitively by the ErbB-4 ligand, neu- regulin (NRG). On the other hand, NRG did not compete with 125 I-labeled EGF for binding to the chimeras with the ErbB-1 domain I and the ErbB-4 domain III. Therefore, NRG binding to ErbB-4 depends much more on domain I than on domain III. With respect to autophosphorylation and subsequent ERK activation, EGF activated the chi- meras with either domain I or III of ErbB-1. In contrast, NRG activated the chimeras with the ErbB-4 domain I and the ErbB-1 domain III, but not those with the ErbB-1 domain I and the ErbB-4 domain III. Therefore, the relative contributions between domains I and III of ErbB-4 in the NRG signaling are different from those of ErbB-1 in the EGF signaling. Keywords: chimeric receptors; epidermal growth factor; ErbB family; ligand recognition; neuregulin. The ErbB-family tyrosine kinases play central roles in the proliferation, differentiation, and development of cells [1]. The family is composed of four members, including the ErbB-1/epidermal growth factor (EGF) receptor [2], ErbB- 2/neu [3], ErbB-3 [4], and ErbB-4 [5]. Each of these receptors has an extracellular region, a single transmembrane region, and a cytoplasmic sequence containing a tyrosine kinase domain and a C-terminal tail. The extracellular region has about 40% homology among the four family members and can be further divided into four domains (I–IV); the N-terminal domain I has sequence similarity to domain III, which is flanked by two cysteine-rich domains, II and IV. More than a dozen ligands have been found to interact with the ErbB-family receptors [6]. These ligands have a characteristic structure called the EGF-like motif, which is defined by three disulfide bridges [7–9], and can be classified into three major groups according to their receptor-binding specificities. The first group consists of EGF, transforming growth factor a, and amphiregulin, all of which bind directly to ErbB-1. The isoforms of neuregulin (NRG, also known as heregulin and neu differentiation factor) are members of the second group, and have specific affinity for ErbB-3 and ErbB-4. The third group is composed of the ligands that bind to both ErbB-1 and ErbB-4, such as betacellulin, heparin-binding EGF, and epiregulin. The binding of ligands to the extracellular region of receptors causes receptor dimerization and autophosphory- lation of the C-terminal tail [10]. The phosphorylated tyrosine residues in the tail serve as docking sites for the proteins that possess a src homology 2 (SH2) domain [11] or a phosphotyrosine-binding (PTB) domain [12]. All members of the ErbB family have docking sites for growth factor receptor-bound protein 2 (Grb2) and/or SH2-containing polypeptide (Shc), both of which have SH2 and/or PTB domains. Recruitment of adapter proteins by the phos- phorylated receptors can stimulate the Ras signaling pathway, leading to the activation of extracellular signal- regulated protein kinases (ERKs) [13]. The Ras/ERK pathway is one of the most important and well-studied pathways that transduce extracellular signals into the intranuclear activation of gene expression [14]. A number of studies have described the ligand interac- tions of the ErbB family, but most of them have dealt with only the EGF receptor (ErbB-1). Among the four extracel- lular domains of ErbB-1, domain III is considered to be the major binding site for EGF. First, cross-linking of Correspondence to K. Saito or S. Yokoyama, RIKEN Genomic Sciences Center, Suehiro-cho, Tsurumi, Yokohama 230-0045, Japan. E-mail: saito@gsc.riken.go.jp or E-mail: yokoyama@biochem.s.u-tokyo.ac.jp Abbreviations: CHO, Chinese hamster ovary; DMEM/F-12, Dulbecco’s modified Eagle’s medium/nutrient mixture F-12; EGF, epidermal growth factor; ERK, extracellular signal-regulated protein kinase; NRG, neuregulin; PTB domain, phosphotyrosine-binding domain; SH2 domain, src homology 2 domain. *Present address: Center for Cellular Switch Protein Structure, Korea Research Institute of Bioscience and Biotechnology, Yusong, Taejon, South Korea. (Received 26 November 2001, revised 7 March 2002, accepted 12 March 2002) Eur. J. Biochem. 269, 2323–2329 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02877.x 125 I-labeled EGF to ErbB-1 in A431 cells resulted in a single- labeled CNBr fragment (residues 294–543), which involves domain III [15]. In addition, some monoclonal antibodies that recognize epitopes in domain III competitively inhibited the EGF binding [16]. By replacing domain III with that of the human EGF receptor, the chicken EGF receptor gained higher affinity for mammalian EGF, similar to that of the entire human receptor [17,18]. The C-terminal part of EGF was found to be located near Lys456 of domain III in the ligand–receptor complex by a cross-linking experiment [19]. Among the ectodomain fragments, the one corresponding to domain III showed the highest affinity for EGF [20]. Nevertheless, it has been reported that domain I as well as domain III is involved in EGF binding, suggesting bivalent binding of EGF to the EGF receptor. A deletion in the N-terminal region of domain I impaired the EGF binding of ErbB-1 [21]. The N-terminus of EGF was linked to Tyr101 in domain I by using a covalent cross-linking reagent [22]. Chimeras between the chicken and human EGF receptors revealed that domain I contributes somewhat to the binding of EGF, in addition to the major contribution of domain III [18]. At present, the bivalent manner of EGF binding to the receptor, in which both domains I and III are utilized, is accepted within the mechanism of receptor dimerization [23]. Considering that the extracellular domains share high sequence homology among the four ErbB members, it is possible that the ligand binding by the ectodomains of other family members is similar to that of ErbB-1. However, with respect to the EGF-like motif of the ErbB-3/4 ligand, NRG, the determinant residues for the specific binding to ErbB-3/4 are somewhat different from those of EGF [24]. The determinants of NRG are clustered in a part that corres- ponds to the N-terminal part of EGF, while those of EGF are in two different parts: the central antiparallel b sheet and the surface including Tyr13, Leu15, Arg41, and Leu47 [25–28]. In the present study, we constructed a series of chimeric receptors between ErbB-1 and ErbB-4 (Fig. 1). These two members of the ErbB family have different specificities for the ligands, EGF and NRG, respectively, but are very similar in their other properties. Actually, both ErbB-1 and ErbB-4 have a ligand-promoted tyrosine kinase activity, whereas ErbB-2 does not have any authentic ligands and ErbB-3 is deficient in the kinase activity. Thus, we found that the relative contributions among the extracellular domains to the cognate-ligand binding are different between ErbB-1 and ErbB-4. EXPERIMENTAL PROCEDURES Materials Human EGF (recombinant) and NRG1-b1 (the EGF-like motif of neuregulin 1-b1, amino-acid residues 176–246, recombinant) were purchased from R&D Systems Inc. (Minneapolis, MN, USA). Murine 125 I-labeled EGF was from NEN Life Science Products, Inc. (Boston, MA, USA). Anti-(ErbB-1) Ig (sc-03), anti-(ErbB-4) Ig (sc-283), and anti- ERK2 Ig (sc-153), and a monoclonal antibody to phosphotyrosine (sc-508), were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). The anti- [phospho-p44/42 MAP kinase (Thr202/Tyr204)] Ig was obtained from New England Biolabs, Inc. (Beverly, MA, USA). Construction of expression plasmids for ErbB-1, ErbB-4, and chimeric receptors The mammalian expression plasmid for ErbB-1 was con- structed as described previously [29]. The full-length cDNA molecule encoding ErbB-4 was obtained from human brain Quick-Clone cDNA (Clontech) by the PCR, and was inserted into the AflII–Xba I sites of the mammalian expression plasmid pcDNA3.1/Zeo(+) (Invitrogen). A schematic diagram of the constructed chimeric receptors is shown in Fig. 1A. The constructs were carefully designed to maintain the disulfide bond connections within the domains [30]. The chimeric receptor 1111-4 was generated by replacing the full-length extracellular region of ErbB-4 (1– 639) with that of ErbB-1 ()24 to 614). The chimeric receptors 1114-4, 1144-4, and 1444-4 were engineered to Fig. 1. Schematic representation and expression of ErbB-1, ErbB-4, and chimeric receptors. (A) EC represents the extracellular region, which consists of domains I, II, III, and IV; TM, the transmembrane region; TK, the tyrosine kinase domain; and CT, the C-terminal tail. Con- structs 1111-1 and 4444-4 correspond to the human wild-type ErbB-1 and ErbB-4, respectively. (B) Whole-cell lysates of CHO cell clones were resolved by 7.5% SDS/PAGE and were transferred to a nitro- cellulose membrane. To confirm the expression of the receptors, the membrane was immunoblotted with an appropriate antibody that recognizes a region of the C-terminal tail of ErbB-1 or ErbB-4. The positions of the molecular mass markers (kDa) are shown on the left. 2324 J H. Kim et al. (Eur. J. Biochem. 269) Ó FEBS 2002 contain N-terminal portions of the extracellular region of ErbB-1 ()24 to 479, )24 to 311, and )24 to 163) and C-terminal portions of ErbB-4 (500–1308, 333–1308, and 187–1308), respectively. To create chimeric receptors that have the ErbB-1 cytoplasmic region in common (4444-1, 4441-1, 4411-1, and 4111-1), the entire extracellular region of ErbB-1 or portions thereof ()24 to 615, )24 to 477, )24 to 311, and )24 to 163) were replaced by the corresponding regions of ErbB-4 (1–639, 1–498, 1–332, and 1–186, respectively). All of the chimeric receptors mentioned above were constructed by the PCR with appropriate primers, and were confirmed by DNA sequence analysis. Cell lines and cell culture Chinese hamster ovary (CHO) cells, which lack endogenous ErbB-1 and ErbB-4, were grown in Dulbecco’s modified Eagle’s medium/nutrient mixture F-12 (DMEM/F-12) medium (Life Technologies, Inc) supplemented with 10% fetal bovine serum and antibiotics. The mammalian expression vectors, which were constructed to express thewild-typeorchimericreceptors,wereintroducedinto CHO cells by the LipofectAMINE method (Life Technologies Inc). Several transfectants were selected in complete medium containing Zeocin (0.2 mgÆmL )1 ). Comparable expression of the receptors was confirmed by immunoblotting of the cell lysates (Fig. 1B). As often observed for membrane proteins, the bands were rather broad in the blots, due to the heterogeneity of the attached carbohydrate chains. EGF binding assay Confluent cells in 12-well plates were incubated in duplicate with 10 ngÆmL )1125 I-labeled EGF in DMEM/F-12 medium containing 1 mgÆmL )1 BSA at 4 °C for 2 h. The free 125 I- labeled EGF was removed by washing three times with ice- cold NaCl/P i containing 1 mgÆmL )1 BSA. The cells were lysedin0.5mLof0.5 M NaOH, and the radioactivity was measured by a gamma counter. The extent of nonspecific binding was determined in the presence of a 200-fold excess of unlabeled EGF. Stimulation, lysate preparation, immunoprecipitation, and immunoblotting The transfected CHO cells were starved in serum-free medium containing 1 mgÆmL )1 BSA for 24 h, and then human EGF or NRG1-b1 was added to stimulate the cells for 5 min. The cells were washed with ice-cold NaCl/P i and lysed in a buffer containing 30 m M Tris/HCl, pH 7.4, 150 m M NaCl, 5 m M EDTA, 40 m M 2-glycerophosphate, 10% glycerol, 1% Triton X-100, 1 m M phenyl- methanesulfonyl fluoride, 1 m M sodium orthovanadate, 10 lgÆmL )1 aprotinin, and 10 lgÆmL )1 leupeptin. Cell debris was removed by microcentrifugation at 4 °Cfor 10 min, and the protein concentrations of the cell lysates were measured with a Protein Assay Kit (Bio-Rad). The lysates were resolved by SDS/PAGE, and then were transferred to a nitrocellulose membrane for immunoblot- ting with an appropriate antibody. Protein bands were visualized by the ECL system (Amersham Pharmacia Biotech). The ERK activation was determined by immunoblotting with an antibody that recognizes activated ERK specifically. RESULTS Ligand-binding abilities of the chimeric receptors The ligand-binding abilities of the chimeric receptors were measured (Fig. 2). After the incubation with 125 I-labeled EGF, the cells expressing ErbB-1 retained strong radioac- tivity as compared to the control cells, while those expres- sing ErbB-4 did not show any EGF binding (closed bars). First, by elucidating the relative contributions of the four extracellular domains of ErbB-1 to the specific binding of the cognate ligand, EGF, we verified the assays using the chimeras, because the contributions of the ErbB-1 ectodo- mains to the EGF binding had already been established by the study with the chicken and human chimeric EGF receptors [18]. Chimera 1111-4, in which the transmembrane and cytoplasmic regions of ErbB-1 (1111-1) were replaced by those of ErbB-4 (4444-4), showed an adequate affinity for 125 I-labeled EGF. As the difference in the affinity between 1111-1 and 1111-4 may arise from the expression level of each transfectant, 1111-4 has a similar affinity for 125 I-labeled EGF to that of 1111-1, as reported for a similar chimera between ErbB-1 and ErbB-2 [31]. The transmem- brane and cytoplasmic regions of ErbB-1 are not involved in the specific EGF binding. Similarly, 1114-4 showed nearly the same 125 I-labeled EGF binding as those of 1111-1 and 1111-4, indicating that the extracellular domain IV is not important for the EGF binding. Furthermore, the 125 I- labeled EGF bindings of 1111-1, 1111-4, and 1114-4 were reduced to the background level by the addition of an excess amount of unlabeled EGF (open bars), but not by NRG1- b1 (gray bars). Consequently, the three N-terminal do- mains, I–III, of ErbB-1 possess the binding sites specific for EGF. In contrast, the replacement of domain III and of domains III and II, resulting in 1144-4 and 1444-4, respectively, greatly decreased the 125 I-labeled EGF binding, indicating that domain III of ErbB-1 is primarily important Fig. 2. Binding of 125 I-labeled EGF to the wild-type ErbB-1, ErbB-4, and the chimeric receptors. Monolayers of CHO cell clones expressing the wild-type ErbB-1, ErbB-4, or the indicated chimeric receptor were incubated with 125 I-labeled EGF at 4 °C for 2 h. After the incubation, the cells were washed three times with ice-cold NaCl/P i containing 1mgÆmL )1 BSA and were solubilized with 0.5 M NaOH. Radioactivity retained on the cells was then measured by a gamma counter (closed bars). For competition assays, the cells were incubated with 125 I- labeled EGF in the presence of a 200-fold excess amount of unlabeled EGF (open bars) or NRG1-b1(grey bars). Ó FEBS 2002 Ligand specificities of ErbB-1/ErbB-4 chimeras (Eur. J. Biochem. 269) 2325 for the EGF binding. This agrees with previous reports that domain III of ErbB-1 is the major binding site for EGF, even in the bivalent manner of EGF binding. On the other hand, chimeras 4111-1 and 4411-1, generated by the replacement of domain I and of domains I and II, respectively, of ErbB-1 by the corresponding domain(s) of ErbB-4, retained significant 125 I-labeled EGF binding. However, the replacement of domain III and of domains III and IV, which resulted in 4441-1 and 4444-1, respect- ively, completely abolished the 125 I-labeled EGF binding. These results confirm that domain III of ErbB-1 plays an important role in the cognate–ligand interaction of ErbB-1. Nevertheless, the 125 I-labeled EGF bindings of 4111-1 and 4411-1 were appreciably smaller than those of 1111-1, 1111-4, and 1114-4, indicating that domain I of ErbB-1 participates somewhat in the EGF binding, in addition to domain III. Chimeras 1144-4 and 1444-4, which contain domain I but lack domain III of ErbB-1, showed weak, but detectable, 125 I-labeled EGF binding, because the bindings were obviously reduced by unlabeled EGF in the assays using the same transfectant cells. This also indicates that the ErbB-1 domain I is involved in EGF binding, but the contribution of domain I is smaller than that of domain III. This agrees with the previously reported conclusions, that EGF binding to the receptor depends mainly on domain III and less on domain I in the bivalent binding [18]. Therefore, binding assays using the chimeras between ErbB-1 and ErbB-4 properly evaluate the relative contributions of the extracellular domains to the cognate-ligand bindings. Thus, the relative contributions of the four extracellular domains of ErbB-4 to the specific binding for the cognate ligand, NRG1-b1, were then examined on the basis of the competition of unlabeled NRG1-b1with 125 I-labeled EGF for the chimera binding (grey bars in Fig. 2). Even though 4111-1 has the ErbB-1 domain III, which binds EGF most strongly among the ectodomains of ErbB-1, the 125 I-labeled EGF binding to 4111-1 was competitively reduced to the background level by the addition of unlabeled NRG1-b1, whereas those of 1114-4, 1111-4, and 1111-1 were not affected. This suggests that domain I of ErbB-4 greatly contributes to the cognate-NRG binding of the receptor. Similarly, the 125 I-labeled EGF binding to 4411-1, which also has the ErbB-4 domain I and the ErbB-1 domain III, was reduced by the addition of unlabeled NRG1-b1. In addition, unlabeled NRG1-b1 hardly reduced the 125 I- labeled EGF binding of either 1444-4 or 1144-4, even though these chimeras lack the ErbB-1 domain III, and therefore exhibit only weak 125 I-labeled EGF binding. These results confirm the conclusion that domain I contributes the most to the cognate-NRG binding of ErbB-4. Autophosphorylation of the chimeric receptors ErbB-1, ErbB-4, and the chimeric receptors were tested for autophosphorylation (Fig. 3). In response to the ligand binding to the ectodomains, the dimerized receptors phos- phorylate their own cytoplasmic C-terminal tails. After the cells expressing ErbB-1 were stimulated with EGF or NRG1-b1, ErbB-1 was immunoprecipitated with an anti- (ErbB-1) Ig, and the phosphorylation of the receptor was visualized by immunoblotting with an anti-phosphotyrosine Ig. ErbB-1 was phosphorylated in response to EGF, but not to NRG1-b1. Among the chimeric receptors constructed here, 1111-4, 1114-4, 1144-4, 1444-4, 4111-1, and 4411-1 were phosphorylated by stimulation with EGF. The extents of the autophosphorylation of 1111-1, 1111-4, 1114-4, 4111- 1, and 4411-1, which all have the ErbB-1 domain III, were stronger than those of 1144-4 and 1444-4 without the ErbB- 1 domain III. In the 125 I-labeled EGF binding assay (closed bars in Fig. 2), the former five receptors bound larger amounts of 125 I-labeled EGF than the latter two. The receptors with neither domain III nor domain I of ErbB-1, 4444-4, 4444-1, and 4441-1, showed negligible autophosph- orylation (Fig. 3) and no 125 I-labeled EGF binding (Fig. 2). Therefore, the contributions of domains III and I in the EGF-induced autophosphorylation of ErbB-1 were fully consistent with those in the EGF binding. Furthermore, the extent of the EGF-induced auto- phosphorylation of 1444-4 was slightly lower than that of 1144-4, while that of 4111-1 was higher than that of 4411-1 Fig. 3. Ligand-induced tyrosine autophosphorylation of ErbB-1, ErbB-4, and the chimeric receptors. CHO cell clones expressing ErbB-1, ErbB-4, or each of the chimeric receptors were serum-starved for 24 h in serum-free medium containing 1 mgÆmL )1 BSA. After the starvation, the cells were treated with the indicated ligands (20 ngÆmL )1 ) for 5 min or left untreated (–). Lysates were subjected to immunoprecipitation with anti-(ErbB-1) Ig or anti-(ErbB-4) Ig. Immunoprecipitates were resolved by 7.5% SDS/PAGE, transferred to a nitrocellulose membrane, and visualized by immunoblotting with an anti-phosphotyrosine Ig. Membranes were stripped and reprobed with the corresponding anti-ErbB Ig to control for protein loading. 2326 J H. Kim et al. (Eur. J. Biochem. 269) Ó FEBS 2002 (Fig. 3). This indicates that domain II may also contribute to the ligand interaction. In this context, 4411-1 bound less 125 I-labeled EGF than 4111-1 (closed bars in Fig. 2). In the case of transforming growth factor a, which belongs to the same ligand group for the ErbB members as EGF, domain II of ErbB-1 was found to be involved in the ligand interaction by insertion mutagenesis [32]. Although three- dimensional structures are not yet available for the ErbB members, a comparative model of ErbB-1, based on the structure of the type-1 insulin-like growth factor receptor [33], shows that the N-terminal part of domain II is involved in a lobe of domain I [34]. Domain II of ErbB-1 might participate in the ligand interaction as a part of ÔstructuralÕ domain I, or it may support the relative positions between the ligand-binding domains, I and III. Cells transfected with either 4414-4 or 1414-4 did not show any response to the ligands (data not shown). In chimeras with such complica- ted constructions, misfolding of the receptors might prevent domains I and III from assuming their correct positions. On the other hand, ErbB-4 was phosphorylated only by NRG1-b1, but hardly by EGF. The chimeras with domain I of ErbB-4, such as 4444-1, 4441-1, 4411-1, and 4111-1, showed sufficient autophosphorylation in response to NRG1-b1, while those without domain I of ErbB-4 did not exhibit any NRG1-b1-induced autophosphorylation. In contrast, domain III contributes much less than domain I to NRG binding by ErbB-4, considering the loaded amounts of the chimeras, shown in a control strip of Fig. 3, and thus the NRG1-b1-induced autophosphorylations of 4444-1 and 4441-1 were just slightly stronger than those of 4411-1 and 4111–1. Even if the binding of NRG1-b1toErbB-4is bivalent, like that of EGF to ErbB-1, domain I is predominant in the specific interaction of ErbB-4 with the cognate ligand, NRG1-b1. Activation of downstream ERK by the chimeric receptors To determine whether the receptor phosphorylation triggers the activation of downstream cellular pathways, the chimeric receptors were tested for ligand-induced ERK activation (Fig. 4). In control cells, the ERK activity was not affected by either EGF or NRG1-b1. The cells expressing ErbB-1 activated ERK when treated with EGF, but not with NRG1-b1. Chimeras 1111-4, 1114-4, 1144-4, 1444-4, 4111-1 and 4411-1, which were all auto- phosphorylated in response to EGF, induced obvious ERK activation upon stimulation with EGF. Although transient overexpression of the receptors in cells may cause non- physiological autophosphorylation, these observations show that the phosphorylation of the receptors certainly transmits the external EGF signal to the downstream enzymes. However, as compared with the 125 I-labeled EGF binding and autophosphorylation assays, it is more difficult to see the contribution of domain III of ErbB-1 to the ERK activation, for example, from the difference between 1114-4 and 1144-4. This may be due to the saturation of ERK activation resulting from the amplifying effect of the signaling cascade. DISCUSSION In the present study, by using chimeric receptors, the contributions of the extracellular domains of ErbB-1 and ErbB-4 to ligand-specific signaling were examined. In the case of EGF signaling by ErbB-1, the receptor has two major functional binding sites, domains I and III, as suggested by the bivalent binding of EGF. From many studies using individual mutations of ligand residues, two contact sites have been mapped on EGF for the interaction with the receptor: the central antiparallel b sheet and the surface including Tyr13, Leu15, Arg41, and Leu47 [25–28]. Affinity labeling between ErbB-1 and EGF, using a heterobifunctional reagent, showed that the N- and C-terminal parts of the ligand are cross-linked to domains I and III, respectively, of the receptor [19,22]. The b sheet of EGF may bind to domain I of ErbB-1, and the other surface of the ligand binds to domain III of the receptor. Recently, Gly441 of the ErbB-1 domain III was proposed to be involved in the binding site that recognizes Arg45 of human EGF [35]. In contrast, domain I is predominant in the NRG signaling by ErbB-4. In the case of NRG, although cross- linking experiments have not been applied to the complex with ErbB-4, several residues in the N- and C-terminal parts Fig. 4. Ligand-induced ERK phosphorylation in the cell clones expressing ErbB-1, ErbB-4, or the chimeric receptors. CHO cell clones expressing ErbB-1, ErbB-4, or each of the chimeric receptors were serum-starved for 24 h in serum-free medium containing 1 mgÆmL )1 BSA. After the starvation, the cells were treated with the indicated ligands (20 ngÆmL )1 ) for 5 min or left untreated (–). Whole-cell lysates were resolved by 10% SDS/PAGE and were immunoblotted with an antibody specific to the active, doubly phosphorylated form of ERK. Membranes were stripped and reprobed with an anti-ERK Ig to control for protein loading. Ó FEBS 2002 Ligand specificities of ErbB-1/ErbB-4 chimeras (Eur. J. Biochem. 269) 2327 of the EGF-like motif of the ligand have already been characterized as high and low affinity sites for the receptor, respectively [36]. Considering the results of this study, the N-terminal high affinity site of NRG1-b1 binds to the dominant ligand-binding site, domain I, of ErbB-4. Recently, domain I of ErbB-3 was also found to have a binding site for NRG1-b1 [37]. Thus, our results suggest that both EGF and NRG1-b1 have a similar orientation in the complex with the cognate receptors, which suggests a common mechanism for both homodimerization and heterodimerization of the ErbB- family receptors. Nevertheless, the interaction of the ligand with the receptor domain I may be somewhat different between EGF and NRG1-b1. Domain I of ErbB-1 binds the central b sheet of EGF, while that of ErbB-4 recognizes several N-terminal residues of NRG1- b1. Although the b sheet of the ligands is located adjacent to the N-terminus in the deduced three dimensional structures, domain I of the receptors may bind a different part of the cognate ligands, because an artificial ligand, ÔbiregulinÕ, which was made by the substitution of a partial NRG sequence for the N-terminus of EGF, bound with high affinity to both ErbB-1 and ErbB-4 [24]. When bound to the receptors, this artificial molecule uses the N-terminal and b sheet parts to bind with domains I of ErbB-4 and ErbB-1, respectively. In conclusion, we have demonstrated differences in the relative contributions of the domains of ErbB-1 to the EGF signaling and those of ErbB-4 to the NRG signaling. Analyses using chimeric receptors are very useful to elucidate the relative contributions among the domains of the receptors in living cells. REFERENCES 1. Walker, R.A. (1998) The erbB/HER type 1 tyrosine kinase receptor family. J. Pathol. 185, 234–235. 2. Ullrich, A., Coussens, L., Hayflick, J.S., Dull, T.J., Gray, A., Tam, A.W., Lee, J., Yarden, Y., Libermann, T.A., Schlessinger, J. et al. (1984) Human epidermal growth factor receptor cDNA sequence and aberrant expression in A431 epidermoid carcinoma cells. Nature 309, 418–425. 3. Yamamoto, T., Ikawa, S., Akiyama, T., Semba, K., Nomura, N., Miyajima, N., Saito, T. & Toyoshima, K. (1986) Similarity of protein encoded by the human c-erb-B-2 gene to epidermal growth factor receptor. Nature 319, 230–234. 4. Kraus, M.H., Issing, W., Miki, T., Popescu, N.C. & Aaronson, S.A. (1989) Isolation and characterization of ERBB3,athird member of the ERBB/epidermal growth factor receptor family: evidence for overexpression in a subset of human mammary tumors. Proc. Natl Acad. Sci. USA 86, 9193–9197. 5. Plowman, G.D., Green, J.M., Culouscou, J.M., Carlton, G.W., Rothwell, V.M. & Buckley, S. (1993) Heregulin induces tyrosine phosphorylation of HER4/p180 erbb4 . Nature 366, 473–475. 6. Riese, D.J. II & Stern, D.F. (1998) Specificity within the EGF family/ErbB receptor family signaling network. Bioessays 20, 41–48. 7. Hommel, U., Harvey, T.S., Driscoll, P.C. & Campbell, I.D. (1992) Human epidermal growth factor. High resolution solution struc- ture and comparison with human transforming growth factor a. J. Mol. Biol. 227, 271–282. 8. Kohda, D. & Inagaki, F. (1992) Three-dimensional nuclear magnetic resonance structures of mouse epidermal growth factor in acidic and physiological pH solutions. Biochemistry 31, 11928– 11939. 9. Montelione, G.T., Wu ¨ thrich, K., Burgess, A.W., Nice, E.C., Wagner,G.,Gibson,K.D.&Scheraga,H.A.(1992)Solution structure of epidermal growth factor determined by NMR spec- troscopy and refined by energy minimization with restraints. Biochemistry 31, 236–249. 10. Lemmon, M.A. & Schlessinger, J. (1994) Regulation of signal transduction and signal diversity by receptor oligomerization. Trends. Biochem. Sci. 19, 459–463. 11. Lowenstein, E.J., Daly, R.J., Batzer, A.G., Li, W., Margolis, B., Lammers, R., Ullrich, A., Skolnik, E.Y., Bar-Sagi, D. & Schlessinger, J. (1992) The SH2 and SH3 domain-containing protein GRB2 links receptor tyrosine kinases to ras signaling. Cell 70, 431–442. 12. Batzer, A.G., Blaikie, P., Nelson, K., Schlessinger, J. & Margolis, B. (1995) The phosphotyrosine interaction domain of Shc binds an LXNPXY motif on the epidermal growth factor receptor. Mol. Cell. Biol. 15, 4403–4409. 13. Alroy, I. & Yarden, Y. (1997) The ErbB signaling network in embryogenesis and oncogenesis: signal diversification through combinatorial ligand–receptor interactions. FEBS Lett. 410, 83–86. 14. Wood,K.W.,Sarnecki,C.,Roberts,T.M.&Blenis,J.(1992)ras mediates nerve growth factor receptor modulation of three signal- transducing protein kinases: MAP kinase, Raf-1, and RSK. Cell 68, 1041–1050. 15. Lax, I., Burgess, W.H., Bellot, F., Ullrich, A., Schlessinger, J. & Givol, D. (1988) Localization of a major receptor-binding domain for epidermal growth factor by affinity labeling. Mol. Cell. Biol. 8, 1831–1834. 16. Wu, D., Wang, L., Sato, G.H., West, K.A., Harris, W.R., Crabb, J.W. & Sato, J.D. (1989) Human epidermal growth factor (EGF) receptor sequence recognized by EGF competitive monoclonal antibodies. Evidence for the localization of the EGF-binding site. J. Biol. Chem. 264, 17469–17475. 17. Lax, I., Bellot, F., Howk, R., Ullrich, A., Givol, D. & Schlessinger, J. (1989) Functional analysis of the ligand binding site of EGF- receptor utilizing chimeric chicken/human receptor molecules. EMBO J. 8, 421–427. 18. Lax,I.,Fischer,R.,Ng,C.,Segre,J.,Ullrich,A.,Givol,D.& Schlessinger, J. (1991) Noncontiguous regions in the extracellular domain of EGF receptor define ligand-binding specificity. Cell Regul. 2, 337–345. 19. Summerfield, A.E., Hudnall, A.K., Lukas, T.J., Guyer, C.A. & Staros, J.V. (1996) Identification of residues of the epidermal growth factor receptor proximal to residue 45 of bound epidermal growth factor. J. Biol. Chem. 271, 19656–19659. 20. Kohda, D., Odaka, M., Lax, I., Kawasaki, H., Suzuki, K., Ullrich, A., Schlessinger, J. & Inagaki, F. (1993) A 40-kDa epidermal growth factor/transforming growth factor a-binding domain produced by limited proteolysis of the extracellular domain of the epidermal growth factor receptor. J. Biol. Chem. 268, 1976–1981. 21. Lax, I., Bellot, F., Honegger, A.M., Schmidt, A., Ullrich, A., Givol, D. & Schlessinger, J. (1990) Domain deletion in the extra- cellular portion of the EGF-receptor reduces ligand binding and impairs cell surface expression. Cell Regul. 1, 173–188. 22. Woltjer, R.L., Lukas, T.J. & Staros, J.V. (1992) Direct identifi- cation of residues of the epidermal growth factor receptor in close proximity to the amino terminus of bound epidermal growth factor. Proc.NatlAcad.Sci.USA89, 7801–7805. 23. Lemmon, M.A., Bu, Z., Ladbury, J.E., Zhou, M., Pinchasi, D., Lax, I., Engelman, D.M. & Schlessinger, J. (1997) Two EGF molecules contribute additively to stabilization of the EGFR dimer. EMBO J. 16, 281–294. 24. Barbacci, E.G., Guarino, B.C., Stroh, J.G., Singleton, D.H., Rosnack, K.J., Moyer, J.D. & Andrews, G.C. (1995) The struc- tural basis for the specificity of epidermal growth factor and heregulin binding. J. Biol. Chem. 270, 9585–9589. 2328 J H. Kim et al. (Eur. J. Biochem. 269) Ó FEBS 2002 25. Tadaki, D.K. & Niyogi, S.K. (1993) The functional importance of hydrophobicity of the tyrosine at position 13 of human growth factor in receptor binding. J. Biol. Chem. 268, 10114–10119. 26. Campion, S.R. & Niyogi, S.K. (1994) Interaction of epidermal growth factor with its receptor. Prog.NucleicAcidRes.Mol.Biol. 49, 353–383. 27. Nandagopal, K., Tadaki, D.K., Lamerdin, J.A., Serpersu, E.H. & Niyogi, S.K. (1996) The functional importance of Leu15 of human epidermal growth factor in receptor binding and activation. Pro- tein Eng. 9, 781–788. 28. Murray, M.B., Tadaki, D.K., Campion, S.R., Lamerdin, J.A., Serpersu, E.H., Bradrick, T.D. & Niyogi, S.K. (1998) Structure- function analysis of a conserved aromatic cluster in the N-terminal domain of human epidermal growth factor. Protein Eng. 11, 1041– 1050. 29. Sato, C., Kim, J H., Abe, Y., Saito, K., Yokoyama, S. & Kohda, D. (2000) Characterization of the N-oligosaccharides attached to the atypical Asn-X-Cys sequence of recombinant human epi- dermal growth factor receptor. J. Biochem. 127, 65–72. 30.Abe,Y.,Odaka,M.,Inagaki,F.,Lax,I.,Schlessinger,J.& Kohda, D. (1998) Disulfide bond structure of human epidermal growth factor receptor. J. Biol. Chem. 273, 11150–11157. 31. Lotti, L.V., Lanfrancone, L., Migliaccio, E., Zompetta, C., Pelicci, G., Salcini, A.E., Falini, B., Pelicci, P.G. & Torrisi, M.R. (1996) Shc proteins are located on endoplasmic reticulum membranes and are redistributed after tyrosine kinase receptor activation. Mol. Cell. Biol. 16, 1946–1954. 32. Harte, M.T. & Gentry, L.E. (1995) Mutations within subdomain II of the extracellular region of epidermal growth factor receptor selectively alter TGFa binding. Arch. Biochem. Biophys. 322, 378–389. 33. Garrett, T.P.J., McKern, N.M., Lou, M., Frenkel, M.J., Bentley, J.D., Lovrecz, G.O., Elleman, T.C., Cosgrove, L.J. & Ward, C.W. (1998) Crystal structure of the first three domains of the type-1 insulin-like growth factor receptor. Nature 394, 395–399. 34. Jorissen,R.N.,Epa,V.C.,Treutlein,H.R.,Garrett,T.P.J.,Ward, C.W. & Burgess, A.W. (2000) Characterization of a comparative model of the extracellular domain of the epidermal growth factor receptor. Protein Sci. 9, 310–324. 35. Elleman, T.C., Domagala, T., McKern, N.M., Nerrie, M., Lo ¨ nnqvist, B., Adams, T.E., Lewis, J., Lovrecz, G.O., Hoyne, P.A., Richards, K.M. et al. (2001) Identification of a determinant of epidermal growth factor receptor ligand-binding specificity using a truncated, high-affinity form of the ectodomain. Bio- chemistry 40, 8930–8939. 36.Tzahar,E.,Pinkas-Kramarski,R.,Moyer,D.J.,Klapper, L.N., Alroy, I., Levkowitz, G., Shelly, M., Henis, S., Eisenstein, M., Ratzkin, B.J. et al. (1997) Bivalence of EGF- like ligands drives the ErbB signaling network. EMBO J. 16, 4938–4950. 37. Singer, E., Landgraf, R., Horan, T., Slamon, D. & Eisenberg, D. (2001) Identification of a heregulin binding site in HER3 extra- cellular domain. J. Biol. Chem. 276, 44266–44274. Ó FEBS 2002 Ligand specificities of ErbB-1/ErbB-4 chimeras (Eur. J. Biochem. 269) 2329 . Chimeric receptor analyses of the interactions of the ectodomains of ErbB-1 with epidermal growth factor and of those of ErbB-4 with neuregulin Jae-Hoon. respectively, of the receptor [19,22]. The b sheet of EGF may bind to domain I of ErbB-1, and the other surface of the ligand binds to domain III of the receptor.