Unique epitope-antibody interactions in the intrinsically disordered proteoglycan-like domain of human carbonic anhydrase IX defined by high-resolution NMR combined with yeast surface display.

Carbonic anhydrase (CA)-IX is an extracellular enzyme that is essential in the adaptation of tumor cells to their increasingly more hypoxic and acidic microenvironment. Within the family of carbonic anhydrases, CA-IX is unique in that it is the only CA with an N-terminal intrinsically disordered region (IDR) containing a proteoglycan (PG)-like domain. This PG-like IDR has been described to be instrumental in CA-IX's enzyme activity, as well as tumor cell motility and invasion. We have characterized the antibody-epitope interactions of two novel and unique antibodies (11H9 and 12H8) that are specific for the human CA-IX's IDR. Binding interactions of these antibodies to the intact IDR were studied by surface plasmon resonance and high-resolution nuclear magnetic resonance (NMR) spectroscopy, while the specific epitopes were determined by both NMR and yeast surface display (YSD). Our data show that 12H8 binds to the N-terminus of CA-IX, while 11H9 has a high affinity for an epitope located in the central region of the IDR containing three GEEDLP repeats in a manner that is different from the previously described M75 antibody. Titration NMR spectroscopy using CA-IX's entire IDR in addition identified a secondary epitope of 11H9 at the beginning of the PG-like domain that remains exposed and available for further binding events after the engagement at its primary epitope at the center of the PG-like domain. Transverse relaxation optimized NMR spectroscopy of 11H9-F(Ab) in complex with the CA-IX IDR outlines structural rigidification of a linear epitope, while the rest of the IDR remains largely unstructured upon complex formation. This study illustrates how high-resolution NMR and YSD are used as complementary tools for a comprehensive characterization of antibody-epitope interactions involving intrinsically unstructured antigen domains with highly repetitive sequences.

Affinity-controlled capture and release of engineered monoclonal antibodies by macroporous dextran hydrogels using coiled-coil interactions.

Long-term delivery is a successful strategy used to reduce the adverse effects of monoclonal antibody (mAb)-based treatments. Macroporous hydrogels and affinity-based strategies have shown promising results in sustained and localized delivery of the mAbs. Among the potential tools for affinity-based delivery systems, the de novo designed Ecoil and Kcoil peptides are engineered to form a high-affinity, heterodimeric coiled-coil complex under physiological conditions. In this study, we created a set of trastuzumab molecules tagged with various Ecoil peptides and evaluated their manufacturability and characteristics. Our data show that addition of an Ecoil tag at the C-termini of the antibody chains (light chains, heavy chains, or both) does not hinder the production of chimeric trastuzumab in CHO cells or affect antibody binding to its antigen. We also evaluated the influence of the number, length, and position of the Ecoil tags on the capture and release of Ecoil-tagged trastuzumab from macroporous dextran hydrogels functionalized with Kcoil peptide (the Ecoil peptide-binding partner). Notably, our data show that antibodies are released from the macroporous hydrogels in a biphasic manner; the first phase corresponding to the rapid release of residual, unbound trastuzumab from the macropores, followed by the affinity-controlled, slow-rate release of antibodies from the Kcoil-functionalized macropore surface.

Tuning the immune response: sulfated archaeal glycolipid archaeosomes as an effective vaccine adjuvant for induction of humoral and cell-mediated immunity towards the SARS-CoV-2 Omicron variant of concern.

Liposomes composed of sulfated lactosyl archaeol (SLA) have been shown to be a safe and effective vaccine adjuvant with a multitude of antigens in preclinical studies. In particular, SLA-adjuvanted SARS-CoV-2 subunit vaccines based on trimeric spike protein antigens were shown to be immunogenic and efficacious in mice and hamsters. With the continued emergence of SARS-CoV-2 variants, we sought to evaluate next-generation vaccine formulations with an updated antigenic identity. This was of particular interest for the widespread Omicron variant, given the abundance of mutations and structural changes observed within its spike protein compared to other variants. An updated version of our resistin-trimerized SmT1 corresponding to the B.1.1.529 variant was successfully generated in our Chinese Hamster Ovary (CHO) cell-based antigen production platform and characterized, revealing some differences in protein profile and ACE2 binding affinity as compared to reference strain-based SmT1. We next evaluated this Omicron-based spike antigen for its immunogenicity and ability to generate robust antigen-specific immune responses when paired with SLA liposomes or AddaS03 (a mimetic of the AS03 oil-in-water emulsion adjuvant system found in commercialized SARS-CoV-2 protein vaccines). Immunization of mice with vaccine formulations containing this updated antigen with either adjuvant stimulated neutralizing antibody responses favouring Omicron over the reference strain. Cell-mediated responses, which play an important role in the neutralization of intracellular infections, were induced to a much higher degree with the SLA adjuvant relative to the AddaS03-adjuvanted formulations. As such, updated vaccines that are better capable of targeting towards SARS-CoV-2 variants can be generated through an optimized combination of antigen and adjuvant components.

Immunogenicity of SARS-CoV-2 spike antigens derived from Beta & Delta variants of concern.

Using our strongly immunogenic SmT1 SARS-CoV-2 spike antigen platform, we developed antigens based on the Beta & Delta variants of concern (VOC). These antigens elicited higher neutralizing antibody activity to the corresponding variant than comparable vaccine formulations based on the original reference strain, while a multivalent vaccine generated cross-neutralizing activity in all three variants. This suggests that while current vaccines may be effective at reducing severe disease to existing VOC, variant-specific antigens, whether in a mono- or multivalent vaccine, may be required to induce optimal immune responses and reduce infection against arising variants.

Production of afucosylated antibodies in CHO cells by coexpression of an anti-FUT8 intrabody.

Some effector functions prompted by immunoglobulin G (IgG) antibodies, such as antibody-dependent cell-mediated cytotoxicity (ADCC), strongly depend on the N-glycans linked to asparagine 297 of the Fc region of the protein. A single α-(1,6)-fucosyltransferase (FUT8) is responsible for catalyzing the addition of an α-1,6-linked fucose residue to the first GlcNAc residue of the N-linked glycans. Antibodies missing this core fucose show a significantly enhanced ADCC and increased antitumor activity, which could help reduce therapeutic dose requirement, potentially translating into reduced safety concerns and manufacturing costs. Several approaches have been developed to modify glycans and improve the biological functions of antibodies. Here, we demonstrate that expression of a membrane-associated anti-FUT8 intrabody engineered to reside in the endoplasmic reticulum and Golgi apparatus can efficiently reduce FUT8 activity and therefore the core-fucosylation of the Fc N-glycan of an antibody. IgG1-producing CHO cells expressing the intrabody secrete antibodies with reduced core fucosylation as demonstrated by lectin blot analysis and UPLC-HILIC glycan analysis. Cells engineered to inhibit directly and specifically alpha-(1,6)-fucosyltransferase activity allows for the production of g/L levels of IgGs with strongly enhanced ADCC effector function, for which the level of fucosylation can be selected. The quick and efficient method described here should have broad practical applicability for the development of next-generation therapeutic antibodies with enhanced effector functions.

Isolation and characterization of monoclonal antibodies against human carbonic anhydrase-IX.

The architectural complexity and heterogeneity of the tumor microenvironment (TME) remains a substantial obstacle in the successful treatment of cancer. Hypoxia, caused by insufficient oxygen supply, and acidosis, resulting from the expulsion of acidic metabolites, are prominent features of the TME. To mitigate the consequences of the hostile TME, cancer cells metabolically rewire themselves and express a series of specific transporters and enzymes instrumental to this adaptation. One of these proteins is carbonic anhydrase (CA)IX, a zinc-containing extracellular membrane bound enzyme that has been shown to play a critical role in the maintenance of a neutral intracellular pH (pHi), allowing tumor cells to survive and thrive in these harsh conditions. Although CAIX has been considered a promising cancer target, only two antibody-based therapeutics have been clinically tested so far. To fill this gap, we generated a series of novel monoclonal antibodies (mAbs) that specifically recognize the extracellular domain (ECD) of human CAIX. Here we describe the biophysical and functional properties of a set of antibodies against the CAIX ECD domain and their applicability as: 1) suitable for development as an antibody-drug-conjugate, 2) an inhibitor of CAIX enzyme activity, or 3) an imaging/detection antibody. The results presented here demonstrate the potential of these specific hCAIX mAbs for further development as novel cancer therapeutic and/or diagnostic tools.

Hydrogen-deuterium exchange mass spectrometry reveals three unique binding responses of mAbs directed to the catalytic domain of hCAIX.

Human carbonic anhydrase (hCAIX), an extracellular enzyme that catalyzes the reversible hydration of CO2, is often overexpressed in solid tumors. This enzyme is instrumental in maintaining the survival of cancer cells in a hypoxic and acidic tumor microenvironment. Absent in most normal tissues, hCAIX is a promising therapeutic target for detection and treatment of solid tumors. Screening of a library of anti-hCAIX monoclonal antibodies (mAbs) previously identified three therapeutic candidates (mAb c2C7, m4A2 and m9B6) with distinct biophysical and functional characteristics. Selective binding to the catalytic domain was confirmed by yeast surface display and isothermal calorimetry, and deeper insight into the dynamic binding profiles of these mAbs upon binding were highlighted by bottom-up hydrogen-deuterium exchange mass spectrometry (HDX-MS). Here, a conformational and allosterically silent epitope was identified for the antibody-drug conjugate candidate c2C7. Unique binding profiles are described for both inhibitory antibodies, m4A2 and m9B6. M4A2 reduces the ability of the enzyme to hydrate CO2 by steric gating at the entrance of the catalytic cavity. Conversely, m9B6 disrupts the secondary structure that is necessary for substrate binding and hydration. The synergy of these two inhibitory mechanisms is demonstrated in in vitro activity assays and HDX-MS. Finally, the ability of m4A2 to modulate extracellular pH and intracellular metabolism is reported. By highlighting three unique modes by which hCAIX can be targeted, this study demonstrates both the utility of HDX-MS as an important tool in the characterization of anti-cancer biotherapeutics, and the underlying value of CAIX as a therapeutic target.

Engineering and therapeutic application of single-chain bivalent TGF-ß family traps.

Deregulation of TGF-ß superfamily signaling is a causative factor in many diseases. Here we describe a protein engineering strategy for the generation of single-chain bivalent receptor traps for TGF-ß superfamily ligands. Traps were assembled using the intrinsically disordered regions flanking the structured binding domain of each receptor as "native linkers" between two binding domains. This yields traps that are approximately threefold smaller than antibodies and consists entirely of native receptor sequences. Two TGF-ß type II receptor-based, single-chain traps were designed, termed (TßRII)2 and (TßRIIb)2, that have native linker lengths of 35 and 60 amino acids, respectively. Both single-chain traps exhibit a 100 to 1,000 fold higher in vitro ligand binding and neutralization activity compared with the monovalent ectodomain (TßRII-ED), and a similar or slightly better potency than pan-TGF-ß-neutralizing antibody 1D11 or an Fc-fused receptor trap (TßRII-Fc). Despite its short in vivo half-life (<1 hour), which is primarily due to kidney clearance, daily injections of the (TßRII)2 trap reduced the growth of 4T1 tumors in BALB/c mice by 50%, an efficacy that is comparable with 1D11 (dosed thrice weekly). In addition, (TßRII)2 treatment of mice with established 4T1 tumors (100 mm(3)) significantly inhibited further tumor growth, whereas the 1D11 antibody did not. Overall, our results indicate that our rationally designed bivalent, single-chain traps have promising therapeutic potential.

Molecular imaging of breast tumors using a near-infrared fluorescently labeled clusterin binding peptide.

Several reports have shown that secreted clusterin (sCLU) plays multiple roles in tumor development and metastasis. Here, we report on a 12-mer sCLU binding peptide (designated P3378) that was identified by screening a phage-display peptide library against purified human sCLU. Differential resonance perturbation nuclear magnetic resonance using P3378 and a scrambled control peptide (designated P3378R) confirmed the P3378-sCLU interaction and demonstrated that it was sequence specific. P3378 and P3378R peptides were conjugated to an Alexa680 near infrared fluorophore (NIRF) and assessed for their tumor homing abilities in in vivo time-domain fluorescence optical imaging experiments using living 4T1 tumor bearing BALB/c mice. When injected in separate animals, both peptides accumulated at the tumor site, however the NIRF-labeled P3378 peptide was retained for a significant longer period of time than the P3378R peptide. Similar observations were made after simultaneously injecting the same tumor-bearing animal with a peptide mixture of P3378 DyLight (DL)680 and the P3378R-DL800. Coinjection of P3378-DL680 with excess unlabeled P3378 blocked tumor accumulation of fluorescent signal while excess P3378R control peptide did not confirming the sequence specificity of the tumor accumulation. Finally, ex vivo fluorescence microscopy of these tumors confirmed the presence of P3378-DL680 in the tumor and its colocalization with CLU. These results confirm the tumor targeting specificity of the P3378 CLU-binding peptide and suggest its usefulness for the in vivo monitoring of solid tumors secreting detectable levels of CLU.

Identification of vascular breast tumor markers by laser capture microdissection and label-free LC-MS.

Blood vessels in tumors frequently show abnormal characteristics, such as tortuous morphology or leakiness, but very little is known about protein expression in tumor vessels. In this study, we have used laser capture microdissection (LCM) to isolate microvessels from clinical samples of invasive ductal carcinoma (IDC), the most common form of malignant breast cancer, and from patient-matched adjacent nonmalignant tissue. This approach eliminates many of the problems associated with the heterogeneity of clinical tumor tissues by controlling for differences in protein expression between both individual patients and different cell types. Proteins from the microvessels were trypsinized and the resulting peptides were quantified by a label-free nanoLC-MS method. A total of 86 proteins were identified that are overexpressed in tumor vessels relative to vessels isolated from the adjacent nonmalignant tissue. These proteins include well-known breast tumor markers such as Periostin and Tenascin C but also proteins with lesser-known or emerging roles in breast cancer and tumor angiogenesis (i.e., Serpin H1, Clic-1, and Transgelin 2). We also identified 40 proteins that were relatively under-expressed in IDC tumor vessels, including several components of the basement membrane whose lower expression could be responsible for weakening tumor vessels. Lastly, we show that a subset of 29 proteins, derived from our list of differentially expressed proteins, is able to predict survival in three publicly available clinical breast cancer microarray data sets, which suggests that this subset of proteins likely plays a functional role in cancer progression and outcome.

Identification of high-quality cancer prognostic markers and metastasis network modules.

Cancer patients are often overtreated because of a failure to identify low-risk cancer patients. Thus far, no algorithm has been able to successfully generate cancer prognostic gene signatures with high accuracy and robustness in order to identify these patients. In this paper, we developed an algorithm that identifies prognostic markers using tumour gene microarrays focusing on metastasis-driving gene expression signals. Application of the algorithm to breast cancer samples identified prognostic gene signature sets for both estrogen receptor (ER) negative (-) and positive (+) subtypes. A combinatorial use of the signatures allowed the stratification of patients into low-, intermediate- and high-risk groups in both the training set and in eight independent testing sets containing 1,375 samples. The predictive accuracy for the low-risk group reached 87-100%. Integrative network analysis identified modules in which each module contained the genes of a signature and their direct interacting partners that are cancer driver-mutating genes. These modules are recurrent in many breast tumours and contribute to metastasis.

A multi-strategy approach to informative gene identification from gene expression data.

An unsupervised multi-strategy approach has been developed to identify informative genes from high throughput genomic data. Several statistical methods have been used in the field to identify differentially expressed genes. Since different methods generate different lists of genes, it is very challenging to determine the most reliable gene list and the appropriate method. This paper presents a multi-strategy method, in which a combination of several data analysis techniques are applied to a given dataset and a confidence measure is established to select genes from the gene lists generated by these techniques to form the core of our final selection. The remainder of the genes that form the peripheral region are subject to exclusion or inclusion into the final selection. This paper demonstrates this methodology through its application to an in-house cancer genomics dataset and a public dataset. The results indicate that our method provides more reliable list of genes, which are validated using biological knowledge, biological experiments, and literature search. We further evaluated our multi-strategy method by consolidating two pairs of independent datasets, each pair is for the same disease, but generated by different labs using different platforms. The results showed that our method has produced far better results.

Glycoproteomic analysis of two mouse mammary cell lines during transforming growth factor (TGF)-beta induced epithelial to mesenchymal transition.

BACKGROUND: TGF-beta acts as an antiproliferative factor in normal epithelial cells and at early stages of oncogenesis. However, later in tumor development TGF-beta can become tumor promoting through mechanisms including the induction of epithelial-to-mesenchymal transition (EMT), a process that is thought to contribute to tumor progression, invasion and metastasis. To identify EMT-related breast cancer therapeutic targets and biomarkers, we have used two proteomic approaches to find proteins that change in abundance upon the induction of EMT by TGF-beta in two mouse mammary epithelial cell lines, NMuMG and BRI-JM01. RESULTS: Preliminary experiments based on two-dimensional electrophoresis of a hydrophobic cell fraction identified only 5 differentially expressed proteins from BRI-JM01 cells. Since 3 of these proteins were glycoproteins, we next used the lectin, wheat germ agglutinin (WGA), to enrich for glycoproteins, followed by relative quantification of tryptic peptides using a label-free LC-MS based method. Using these approaches, we identified several proteins that are modulated during the EMT process, including cell adhesion molecules (several members of the Integrin family, Fibronectin, Activated leukocyte cell adhesion molecule, and Neural cell adhesion molecule 1) and regulators of cellular signaling (Tumor-associated calcium signal transducer 2, Basigin). CONCLUSION: Interestingly, despite the fact that TGF-beta induces similar EMT phenotypes in NMuMG and BRI-JM01 cells, the proteomic results for the two cell lines showed only minimal overlap. These differences likely result in part from the conservative cut-off values used to define differentially-expressed proteins in these experiments. Alternatively, it is possible that the two cell lines may use different mechanisms to achieve an EMT transition.

Escherichia coli expression and refolding of E/K-coil-tagged EGF generates fully bioactive EGF for diverse applications.

Heterodimerizing peptides, such as the de novo designed E5/K5 peptide pair, have several applications including as tags for protein purification or immobilization. Recently, we demonstrated that E5-tagged epidermal growth factor (EGF), when bound to a K4 expressing adenovirus, promotes retargeting of the adenovirus to EGFR expressing target cells. In this study, we present the Escherichia coli expression, refolding and purification of human EGF fused with the E5-coil (E5-coil-EGF) or with the K5-coil (K5-coil-EGF). EGF receptor phosphorylation and cell proliferation assays demonstrated that the biological activity of the coil-tagged EGF versions was comparable to that of non-tagged EGF. Additionally, analysis of the binding of E5/K5-coil-EGF to cell surface EGFR or to soluble EGFR ectodomain, as measured by cell-based binding competition assays and by SPR-based biosensor experiments, indicated that the coil-tagged EGF versions bound to EGFR with affinities similar to that of non-tagged EGF. Finally, we show that E-coil-tagged EGF, but not non-tagged EGF, can retarget a K-coil containing adenovirus to EGF receptor expressing glioblastoma tumor cells. Overall these results indicate that E. coli expression offers a practical platform for the reproducible production of fully biologically active E5/K5-coil-tagged EGF, and support applications of heterodimerizing coil-tagged ligands, e.g. the targeting of viruses or other entities such as nanoparticles to tumor cells, or growth factor immobilization on cell culture scaffolds for tissue engineering.

Investigation of three new mouse mammary tumor cell lines as models for transforming growth factor (TGF)-beta and Neu pathway signaling studies: identification of a novel model for TGF-beta-induced epithelial-to-mesenchymal transition.

INTRODUCTION: This report describes the isolation and characterization of three new murine mammary epithelial cell lines derived from mammary tumors from MMTV (mouse mammary tumor virus)/activated Neu + TbetaRII-AS (transforming growth factor [TGF]-beta type II receptor antisense RNA) bigenic mice (BRI-JM01 and BRI-JM05 cell lines) and MMTV/activated Neu transgenic mice (BRI-JM04 cell line). METHODS: The BRI-JM01, BRI-JM04, and BRI-JM05 cell lines were analyzed for transgene expression, their general growth characteristics, and their sensitivities to several growth factors from the epidermal growth factor (EGF) and TGF-beta families (recombinant human EGF, heregulin-beta1 and TGF-beta1). The BRI-JM01 cells were observed to undergo a striking morphologic change in response to TGF-beta1, and they were therefore further investigated for their ability to undergo a TGF-beta-induced epithelial-to-mesenchymal transition (EMT) using motility assays and immunofluorescence microscopy. RESULTS: We found that two of the three cell lines (BRI-JM04 and BRI-JM05) express the Neu transgene, whereas, unexpectedly, both of the cell lines that were established from MMTV/activated Neu + TbetaRII-AS bigenic tumors (BRI-JM01 and BRI-JM05) do not express the TbetaRII-AS transgene. The cuboidal BRI-JM01 cells exhibit a short doubling time and are able to form confluent monolayers. The BRI-JM04 and BRI-JM05 cell lines are morphologically much less uniform, grow at a much slower rate, and do not form confluent monolayers. Only the BRI-JM05 cells can form colonies in soft agar. In contrast, all three cell lines form colonies in Matrigel, although the BRI-JM04 and BRI-JM05 cell lines do so more efficiently than the BRI-JM01 cell line. All three cell lines express the cell surface marker E-cadherin, confirming their epithelial character. Proliferation assays showed that the three cell lines respond differently to recombinant human EGF and heregulin-beta1, and that all are growth inhibited by TGF-beta1, but that only the BRI-JM01 cell line undergoes an EMT and exhibits increased motility upon TGF-beta1 treatment. CONCLUSION: We suggest that the BRI-JM04 and BRI-JM05 cell lines can be used to investigate Neu oncogene driven mammary tumorigenesis, whereas the BRI-JM01 cell line will be useful for studying TGF-beta1-induced EMT.

Expression of TGF-beta type II receptor antisense RNA impairs TGF-beta signaling in vitro and promotes mammary gland differentiation in vivo.

In order to clarify the role of TGF-beta in mammary development and tumorigenesis, we investigated the efficacy of full- or partial-length TbetaRII antisense RNA specifically to reduce TbetaRII levels in both in vitro and in vivo model systems. Here we show that the expression of TbetaRII antisense RNA in vitro reduced TbetaRII cell surface expression and inhibited the antiproliferative and transcriptional responses to exogenous TGF-beta. Expression of full-length TbetaRII antisense RNA in a transgenic mouse model under control of the mouse mammary tumor virus promotor resulted in precocious lobuloalveolar development of the mammary gland, a phenotype that resembles that of early pregnancy. These data demonstrate that TbetaRII plays a critical role in maintaining the nondifferentiated character of virgin mammary gland epithelium.

ErbB2/Neu-induced, cyclin D1-dependent transformation is accelerated in p27-haploinsufficient mammary epithelial cells but impaired in p27-null cells.

ErbB2/Neu destabilizes the cyclin-dependent kinase (Cdk) inhibitor p27 and increases expression of cyclin D1. Therefore, we studied the roles of p27 and cyclin D1 in ErbB2-mediated mammary epithelial cell transformation. Overexpression of ErbB2 or cyclin D1 in p27(+/-) primary murine mammary epithelial cells resulted in increased proliferation, cyclin D1 nuclear localization, and colony formation in soft agar compared to those in p27(+/+) cells. In contrast, ErbB2- or cyclin D1-overexpressing p27(-/-) cells displayed reduced proliferation, anchorage-independent growth, Cdk4 activity, cyclin D1 expression, and cyclin D1 nuclear localization compared to wild-type cells. A cyclin D1 mutation in its nuclear export sequence (T286A) partially rescued nuclear localization of cyclin D1 in p27(-/-) cells but did not increase proliferation or Cdk4 kinase activity. Overexpression of E2F1, however, increased proliferation to the same degree in p27(+/+), p27(+/-), and p27(-/-) cells. Mammary glands from MMTV (mouse mammary tumor virus)-neu/p27(+/-) mice exhibited alveolar hyperplasia, enhanced proliferation, decreased apoptosis, and accelerated tumor formation compared to MMTV-neu/p27(+/+) glands. However, MMTV-neu/p27(-/-) glands showed decreased proliferation, cyclin D1 expression, and Cdk4 activity, as well as markedly prolonged tumor latency, compared to MMTV-neu/p27(+/+) glands. These results suggest that p27(+/-) mammary epithelium may be more susceptible to oncogene-induced tumorigenesis, whereas p27-null glands, due to severely impaired cyclin D1/Cdk4 function, are more resistant to transformation.

Epidermal growth factor contains both positive and negative determinants for interaction with ErbB-2/ErbB-3 heterodimers.

Epidermal growth factor (EGF) and transforming growth factor (TGF)-alpha are potent activators of the ErbB-1 receptor, but, unlike TGF-alpha, EGF is also a weak activator of ErbB-2/ErbB-3 heterodimers. To understand the specificity of EGF-like growth factors for binding to distinct ErbB members, we used EGF/TGF-alpha chimeras to examine the requirements for ErbB-2/ErbB-3 activation. Here we show that in contrast to these two wild-type ligands, distinct EGF/TGF-alpha chimeras are potent activators of ErbB-2/ErbB-3 heterodimers. On the basis of differences in the potency of these various chimeras, specific residues in the linear N-terminal region and the so-called B-loop of these ligands were identified to be involved in interaction with ErbB-2/ErbB-3. A chimera consisting of human EGF sequences with the linear N-terminal region of human TGF-alpha was found to be almost as potent as the natural ligand neuregulin (NRG)-1beta in activating 32D cells expressing ErbB-2/ErbB-3 and human breast cancer cells. Binding studies revealed that this chimera, designated T1E, has high affinity for ErbB-2/ErbB-3 heterodimers, but not for ErbB-3 alone. Subsequent exchange studies revealed that introduction of both His2 and Phe3 into the linear N-terminal region was already sufficient to make EGF a potent activator of ErbB-2/ErbB-3 heterodimers, indicating that these two amino acids contribute positively to this receptor binding. Analysis of the B-loop revealed that Leu26 in EGF facilitates interaction with ErbB-2/ErbB-3 heterodimers, while the equivalent Glu residue in TGF-alpha impairs binding. Since all EGF/TGF-alpha chimeras tested have maintained high binding affinity for ErbB-1, it is concluded that the diversity of the ErbB signaling network is determined by specific amino acids that facilitate binding to one receptor member, in addition to residues that impede binding to other ErbB family members.