Novel EGFR-binding protein

Abstract

This invention relates to an antigen-binding protein that binds to the ligand-binding site of EGFR and clinically significant EGFR escape mutations. Specifically, the invention relates to an antigen-binding protein comprising a first antibody-variable domain and a second antibody-variable domain, wherein the variable domains are determined by biolayer interferometry (BLI). -5 It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more mutations among V441D, S464L, G465R, and S492R at a KD of less than molal concentration, and 10 are determined by biolayer interferometry (BLI). -5 We provide an antigen-binding protein that does not bind to human EGFR, possessing the mutation Q435P and / or the double mutation F436A / I462A at a KD exceeding molar concentration. Binding of the antigen-binding protein to human EGFR inhibits ligand-induced activation of human EGFR.

Description

[Technical Field]

[0001] This invention relates to an antigen-binding protein that binds to the ligand-binding site of the epidermal growth factor receptor (EGFR), and to clinically significant EGFR escape mutations. The antigen-binding protein can block EGFR activation and overcome acquired resistance observed in cancer patients treated with anti-EGFR antibodies. [Background technology]

[0002] EGFR (also known as HER1), a member of the ErbB (HER) receptor family, is a membrane-bound tyrosine kinase that is activated upon ligand binding, leading to conformational changes and the formation of homotype or heterotype receptor dimers with either EGFR or one of the other three members of the receptor family (HER2, HER3, HER4) (Roskoski, 2014, Pharmacol. Res. 79; 34-74). The extracellular domain of EGFR contains four domains. Domains I and III are involved in ligand binding, while domains II and IV contribute to receptor dimerization. Ligands for EGFR include EGF (epidermal growth factor), EPG (epigen), TGFα (transforming growth factor-α), ARG (amphiregulin), BTC (beta-cell phosphate), HB-EGF (heparin-binding epidermal growth factor), and EPR (epigulin). These ligand-activated receptor dimers phosphorylate tyrosine residues in the intracellular signaling domain of the receptor, which function as binding sites for adapter molecules, and subsequently activate downstream signaling pathways. These activated signaling pathways ultimately lead to increased proliferation, migration, and survival of cancer cells, contributing to a malignant phenotype (Cai et al., 2020, Front. Oncol.24: 1249).

[0003] Elevated EGFR expression levels and activation status have been found in lung (NSCLC), prostate, thoracic, colorectal, head and neck, esophageal and gastric, liver, glioblastoma, cervix, ovarian, bladder, kidney, and pancreatic cancers (Thomas & Weihua, 2019, Front. Oncol. 9: 800). This has led to the development of EGFR-targeted drugs, including antibodies that block receptor activation and small molecules that bind to and inhibit the kinase domain (tyrosine kinase inhibitors, TKIs) (Wieduwelt & Moasser, 2008, Cell. Mol. Life Sci. 65: 1566-1584, Nedergaard et al., 2012, Biodrugs 26: 83-99, Dokala & Thakur, 2017, Oncogene 36: 2337-2344). For both types, treatment molecules have been approved, and numerous further molecules have been investigated in preclinical and clinical studies, highlighting the potential of EGFR as a target in cancer therapy (Kwapiszewski et al., 2016, Targ. Oncol. 11: 739-752, Yamaoka et al., Molecular Sciences 2017).

[0004] Several anti-EGFR antibodies are approved for the treatment of CRC (colorectal cancer), HNSCC (head and neck squamous cell carcinoma), and NSCLC (non-small cell lung cancer). These include cetuximab, panitumumab, necitumumab, and nimotuzumab, i.e., human antibodies or humanized antibodies, all of which bind to domain III of EGFR. Cetuximab, panitumumab, and necitumumab bind to the same region of domain III, which is part of the ligand-binding site (Li et al., 2005, Cancer Cell 7: 301-311, Sickmier et al., 2016, PLoS ONE 11: e163366, Bagchi et al., 2018, Mol. Cancer Ther. 17: 521-531), but the three antibodies differ in the residues directly involved in antigen binding.

[0005] Cetuximab is approved for the treatment of patients with K-Ras wild-type, EGFR-expressing metastatic colorectal cancer and recurrent or metastatic squamous cell carcinoma of the head and neck. Panitumumab is approved for the treatment of metastatic colorectal cancer with RAS wild-type. Necitumumab is approved for the treatment of patients with advanced squamous non-small cell lung cancer. Nimotuzumab is approved for the treatment of patients with squamous cell carcinoma of the head and neck.

[0006] Further antibodies that have been studied in clinical trials or are currently in clinical development include tomzotuximab (CeuGEX), imugatuzumab (GA201), matsuzumab (EMD72000), saltumumab (huMax-EGFr), GC1118, JMT101, SCT200, ZZ06, and ABT-806, as well as antibody mixtures such as modotuximab + futuximab and MM-151.

[0007] Furthermore, anti-EGFR antibodies are used to generate bispecific antibodies targeting EGFR and other surface receptors, such as cMETs (amivantamab, MCLA-129, EMB-01, GB-263), HER3 (SI-B001), HGFR (CKD-702), and LGR5 (MCLA-158) for dual targeting and signal inhibition. Other examples include a bispecific antibody targeting EGFR and CD28 for targeted immune costimulation (REGN7075), and a bispecific antibody targeting 4-1BB for targeted checkpoint inhibition in cancer immunotherapy (HLX35). In cancer immunotherapy, EGFR antibodies are further used to generate bispecific antibodies for mobilizing T cells (T cell engagers; TCEs) or NK cells by binding to CD3 or CD16 using a second arm. Furthermore, EGFR-targeting antibody therapies include single or bispecific antibody-drug conjugates (including the approved cetuximab derivative (cetuximab sarotracan)), antibody fusion proteins, and CAR-T cells (Cai et al., 2020, Front. Oncol.24: 1249).

[0008] While EGFR antibody treatment improves overall patient outcomes, several pre-existing or acquired resistance mechanisms have been found to limit the benefits for individual patients. Pre-existing resistance includes mutations in downstream signaling molecules, such as those found in K-Ras, and patients with these mutations are excluded from therapy by diagnostic analysis of K-Ras mutation status. Acquired resistance includes increased expression of EGFR-binding ligands (e.g., TGFα) and compensatory expression of members of other receptors, such as the ErbB family (e.g., HER3) or other receptor families (e.g., c-Met).

[0009] In particular, mutations in the extracellular domain of EGFR, which are part of the epitopes of approved anti-EGFR antibodies such as cetuximab and panitumumab, are observed in approximately 15–20% of patients treated with approved EGFR antibodies and cause reduced binding of the therapeutic antibody. Commonly observed mutations after cetuximab treatment include V441D, S442R, G465R, and S492R within domain III (numbered based on the full-length EGFR sequence, SEQ ID NO: 54). Because the approved antibodies cetuximab, necitumumab, and panitumumab share a common epitope, some of these mutations (V441D, S442R, G465R) can also cause reduced or lost binding to necitumumab and panitumumab, potentially leading to resistance to treatment with these antibodies (Strickler et al., Cancer Discov., 2018, 8: 164-173, Cai et al., 2020, Front. Oncol.24: 1249).

[0010] Currently, there are no available antibodies that can broadly overcome the resistance mutations in the extracellular domain of EGFR mentioned above. [Overview of the Initiative] [Problems that the invention aims to solve]

[0011] Therefore, there is a need for fully human antagonistic anti-EGFR binding proteins, particularly antibodies that can bind to the ligand-binding site of EGFR and thus inhibit receptor activation, receptor signaling, and cell proliferation, and that also tolerate mutations in the extracellular domain of EGFR that would otherwise result in resistance to antibody treatment. Furthermore, such antigen-binding proteins are needed for use in the treatment of patients with EGFR-expressing tumors. [Means for solving the problem]

[0012] Selection from a human antibody phage display library identified novel human anti-EGFR antigen-binding proteins capable of binding to EGFR domain III, efficiently inhibiting EGFR activation, signaling, and cell proliferation, and binding to several clinically significant EGFR escape mutations, including V441D, S442R, S464L, G465R, and S492R, within the ligand-binding site of domain III. These antigen-binding proteins bind to specific epitopes containing amino acid residues G434, Q435, F436, and I462 and are cross-reactive with mouse EGFR. Using cell lines expressing either wild-type EGFR or mutants such as the clinically significant mutations G465R and S492R, these antigen-binding proteins exhibit superior inhibitory effects compared to cetuximab, necitumumab, panitumumab, GC1118, and matsuzumab. Therefore, the antigen-binding protein of the present invention makes it possible to treat cancers having wild-type EGFR and EGFR with clinically significant escape mutations.

[0013] According to a first aspect, the present invention relates to an antigen-binding protein comprising a first antibody-variable domain and a second antibody-variable domain, wherein the variable domain is a) determined by biolayer interferometry (BLI) -5 K below molar concentration Da) provides an antigen-binding protein that specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more of the mutations V441D, S464L, G465R, and S492R, and b) forms a binding site that does not specifically bind to human EGFR having the mutation Q435P and / or the double mutation F436A / I462A, thereby preventing ligand-induced activation of human EGFR upon binding of the antigen-binding protein to human EGFR.

[0014] According to one embodiment, the variable domain forms a binding site that specifically binds to domain III of human EGFR as defined by Sequence ID No. 61.

[0015] In further embodiments, the antigen-binding protein is selected from the group consisting of antibodies or their antigen-binding fragments, and chimeric antigen receptors (CARs). Preferably, the antigen-binding protein is selected from the group consisting of Fab, Fab', (Fab')2, Fv, disulfide-linked Fv, BsFv, dsFv, (dsFv)2, dsFv-dsFv', scFv, scFv dimer, single-chain domain antibody, diabody, ds-diabody, and bivalent domain antibody.

[0016] According to a preferred embodiment, the antibody is selected from the group consisting of chimeric antibodies, humanized antibodies, bispecific antibodies, and polyspecific antibodies. The antibody is preferably human or humanized.

[0017] In yet another embodiment, the first antibody variable domain comprises light chain complementarity determining regions (CDRL)1 (CDRL1), CDRL2, and CDRL3 of the variable light chain region of SEQ ID NO: 83, 84, or 85, and / or the second variable domain comprises CDRH1, CDRH2, and CDRH3 of the heavy chain variable region of SEQ ID NO: 74 or 75, where CDR is identified by Kabat, Chothia, IMGT, or contact definitions, where each CDR may contain one or two amino acid substitutions.

[0018] According to a preferred embodiment, the first antibody variable domain comprises CDRL1 of SEQ ID NO: 86, CDRL2 of SEQ ID NO: 87, and CDRL3 of SEQ ID NO: 88, and the second antibody variable domain comprises CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73, where each CDR may contain one or two amino acid substitutions.

[0019] According to another preferred embodiment, the first antibody variable domain comprises CDRL1 of SEQ ID NO: 76 or 77, CDRL2 of SEQ ID NO: 78 or 79, and CDRL3 of SEQ ID NO: 80, 81 or 82, and the second antibody variable domain comprises CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73, where each CDR may comprise one or two amino acid substitutions.

[0020] According to another embodiment, the first antibody variable domain comprises one or more of the framework (FR) sequence of SEQ ID NO: 83, 84, or 85, or variants of the FR sequence having at least 90% sequence identity with respect to the FR sequence of SEQ ID NO: 83, 84, or 85, and / or the second antibody variable domain comprises one or more of the FR sequence of SEQ ID NO: 74 or 75, or variants of the FR sequence having at least 90% sequence identity with respect to the FR sequence of SEQ ID NO: 74 or 75, where FR is identified by Kabat definition, Chothia definition, IMGT definition, or contact definition.

[0021] In yet another embodiment, the first variable domain comprises one of the amino acid sequences according to SEQ ID NO: 83, 84, or 85, and the second variable domain comprises one of the amino acid sequences according to SEQ ID NO: 74 or 75. In a preferred embodiment, the antigen-binding protein comprises one of the combinations: SEQ ID NO: 74 and SEQ ID NO: 83, SEQ ID NO: 75 and SEQ ID NO: 84, or SEQ ID NO: 74 and SEQ ID NO: 85, where each CDR may comprise one or two amino acid substitutions.

[0022] In a particularly preferred embodiment, the antigen-binding protein comprises one of the following sequence combinations: a) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of SEQ ID NOs. 76, 78, and 80; b) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of SEQ ID NOs. 77, 79, and 81; or c) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of SEQ ID NOs. 77, 79, and 82, where each CDR may contain one or two amino acid substitutions.

[0023] According to one embodiment, the antigen-binding protein competes for binding to the IgG1 antibody containing the VL of SEQ ID NO: 83, 84, or 85 and the VH of SEQ ID NO: 74 or 75, or binds to the same epitope as the IgG1 antibody containing the VL of SEQ ID NO: 83, 84, or 85 and the VH of SEQ ID NO: 74 or 75, where each CDR may contain one or two amino acid substitutions.

[0024] In a further embodiment, the present invention provides nucleic acids (one or more) encoding antigen-binding proteins according to the present invention.

[0025] In another aspect, the present invention provides a vector comprising nucleic acids (one or more) according to the present invention.

[0026] In a further embodiment, the present invention provides recombinant cells expressing antigen-binding proteins, nucleic acids (one or more), or vectors according to the present invention.

[0027] The present invention further provides a pharmaceutical composition comprising an antigen-binding protein, nucleic acid(s), vector, or recombinant cell according to the present invention, and a pharmaceutically acceptable excipient as appropriate.

[0028] In a further embodiment, the present invention provides antigen-binding proteins, nucleic acids (one or more), vectors, recombinant cells, or pharmaceutical compositions according to the present invention for use in pharmaceuticals.

[0029] In yet another aspect, the present invention provides antigen-binding proteins, nucleic acids (one or more), vectors, recombinant cells, or pharmaceutical compositions according to the present invention for use in the treatment of cancer.

[0030] According to a preferred embodiment, the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic head and neck squamous cell carcinoma.

[0031] In another aspect, the present invention provides a method for treating cancer, comprising administering to a patient in need of treatment a therapeutically effective amount of the antigen-binding protein, nucleic acid(s), vector, recombinant cells, or pharmaceutical composition of the present invention. The cancer is preferably selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic head and neck squamous cell carcinoma.

[0032] In another aspect, the present invention provides a method for treating cancer, comprising administering to a patient in need of treatment a therapeutically effective amount of an antigen-binding protein, nucleic acid(s), vector, recombinant cells, or pharmaceutical composition according to the present invention.

[0033] According to one embodiment, the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably from the group consisting of head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

[0034] Further aspects and embodiments are disclosed in the accompanying claims and can be derived from the following detailed description of the invention.

[0035] The contents of the drawings included in this specification are described below. In this context, please also refer to the detailed description of the invention above and / or below. [Brief explanation of the drawing]

[0036] [Figure 1] Figure 1 shows the inhibition of EGFR phosphorylation and downstream signaling in FaDu cells treated with 10 μg / mL anti-EGFR scFv-huFc, either without stimulation (a) or with stimulation with 50 ng / mL of EGF (b). (n=1) [Figure 2] Figure 2 shows SEC(a) and SDS-PAGE(b) analyses of IgG1 MKU011-A7(a, b) and Fab-His MKU011-A7(b) after transient expression in HEK293-6E and purification with Protein A or CaptureSelect™ CH1 XL. [Figure 3] Figure 3 shows the SEC (a, b, d, e) and SDS-PAGE (c, f) analyses of IgG1 YU250-A02 (a-c) and YU250-H10 (d-e) after transient expression and protein A purification in HEK293-6E. [Figure 4] Figure 4 shows the SEC (a, c) and SDS-PAGE (b, d) analyses of Fab-His YU250-A02 (a, b) and YU250-H10 (c, d) after transient expression in HEK293-6E and purification with CaptureSelect™ CH1 XL. [Figure 5] Figure 5 shows analyses of IgG YU250-H10 wild-type and mutant VL N126K binding to EGFR by SEC-HPLC (a, b), SDS-PAGE (c), and ELISA (d) (n=1 - data are shown as double repeat mean ± SD). [Figure 6] Figure 6 shows the analysis of monovalent binding of Fab-His (a) and divalent binding of IgG1 (b) to MKU011-A7 and YU250-H10 by EGFR-binding ELISA (n=3 - data are shown as mean ± SD). [Figure 7] Figure 7 shows the analysis of cetuximab, IgG MKU011-A7, IgG YU250-A02, and IgG YU250-H10 binding to EGFR by ELISA (a) and to EGFR-expressing cancer cells DiFi (b) and FaDu (c) by flow cytometry (n=1 - data are shown as double repeats, mean ± SD). [Figure 8] Figure 8 shows the ELISA analysis of binding of 10 nM IgG YU250-A02, IgG YU250-H10, and cetuximab to human (a) and cynomolgus monkey EGFR (b). ELISA determination of EC50 values ​​for IgG YU250-A02 (c) and IgG YU250-H10 (d) for both cynomolgus monkey and human EGFR (n=3 - data shown as mean ± SD). [Figure 9] Figure 9 shows the inhibition of EGFR phosphorylation and MAPK pathway signaling in DiFi(a) and FaDu(b) in response to anti-EGFR IgG MKU011-A7, YU250-A02, YU250-H10, and cetuximab. Analysis of proliferation in DiFi(c) and FaDu(d) treated with anti-EGFR IgG (n=3 - data presented as mean ± SD - normalized for cells without mAb - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01 / *** - p<0.001 / **** - p<0.0001). [Figure 10]Figure 10 shows the inhibition of EGFR dimerization by anti-EGFR IgG by a β-galactosidase complementation assay, analyzed with respect to the concentration range using 30 ng / mL EGF as a stimulus (a) or at a concentration of 10 nM (b) (n=3 - data are shown as mean ± SD - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01). [Figure 11] Figure 11 shows the analysis of the internal transfer of pH-rodo-labeled cetuximab and IgG YU250-H10 to FaDu cells at 37°C (a) and 4°C (b) (n=3 - data shown as mean ± SD - statistics for 48 hours at 37°C: t-test - p=0.05 / ns - not significant). [Figure 12] Figure 12 shows an ELISA analysis of competitive binding of cetuximab to scFv-huFc YU250-A02 and YU250-H10(a), and to anti-EGFR IgG to Fab-His YU250-A02(b) or YU250-H10(c) (n=3 - data presented as mean ± SD - normalized for binding of cetuximab (a) or anti-EGFR IgG without competitor (b, c) at 100 nM - statistics: one-way ANOVA with Tukey's post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01 / *** - p<0.001 / **** - p<0.0001). [Figure 13] Figure 13 shows the ELISA analysis of binding of anti-EGFR IgG YU250-A02, YU250-H10, cetuximab, GC1118, matsuzumab, nesitumumab, and panitumumab to EGFR wt(a), G434A(b), Q435P(c), F436A(d), I462A(e), and F436A / I462A(f) (n=3 - data are shown as mean ± SD). [Figure 14]Figure 14 shows the ELISA analysis of binding of anti-EGFR IgG YU250-A02, YU250-H10, cetuximab, and GC1118 to EGFR wt(a), F381A(b), H383A(c), and F381A / H383A(d), as well as the binding of anti-EGFR IgG YU250-A02, YU250-H10, cetuximab, and matsuzumab to EGFR S484A(e) and K487(f) (n=3 - data are shown as mean ± SD). [Figure 15] Figure 15 shows the ELISA analysis of binding of anti-EGFR IgG YU250-A02, YU250-H10, cetuximab, nesitumumab, and panitumumab to EGFR wt(a), V441D(b), V441G(c), S442R(d), S464L(e), G465R(f), K489E(g), and S492R(h) (n=3 - data are shown as mean ± SD). [Figure 16] Figure 16 shows an overview of epitope mapping resulting from the introduction of mutations into EGFR that lead to reduced binding for anti-EGFR IgG YU250-A02, YU250-H10, cetuximab, GC1118, matsuzumab, nesitumumab, and panitumumab (data are shown as the mean of n=3 for reduced binding normalized to wt[%] at a concentration of 10 nM). [Figure 17] Figure 17 shows the affinity (KD) of monovalent anti-EGFR Fab to the histidine-tagged extracellular domain of EGFR, as analyzed by biolayer interferometry. [Figure 18] Figure 18 shows the structure of human EGFR domain III, with the positions Q435P and F436A / I462A (cluster 1) that influence the binding of antibodies MKU011-A7, YU250-A02, or YU250-H10, and the positions V441D, S464L, G465R, and S492R (cluster 2) that influence the binding of cetuximab highlighted. A) Top view of domain III, B) Side view of domain III complexed with EGF. The two clusters are structurally separated but located at the EGF-EGFR-DIII interface. [Figure 19] Figure 19 shows the ELISA analysis of binding of 10nM IgG YU250-A02, IgG YU250-H10, cetuximab, GC1118, matsuzumab, necitumumab, and panitumab to human (a) and mouse EGFR (b). ELISA determination of EC50 values ​​for both mouse and human EGFR for IgG YU250-A02 (c) and IgG YU250-H10 (d) (n=3 - data shown as mean ± SD). [Figure 20] Figure 20 shows the analysis of anti-EGFR IgG binding to A431 cells (a), inhibition of A431 cell proliferation (b), and inhibition of EGFR phosphorylation and MAPK pathway signaling (c) (n=3 for binding and proliferation / n=1 for signaling inhibition - data are shown as mean ± SD - proliferation is normalized to cells without mAbs - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05). [Figure 21] Figure 21 shows the analysis of anti-EGFR IgG binding to DiFi (a), inhibition of DiFi-related proliferation (b), and inhibition of EGFR phosphorylation and MAPK pathway signaling (c) (n=3 for binding and proliferation / n=2 for signaling inhibition - data are shown as mean ± SD - proliferation is normalized to cells without mAbs - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01 / *** - p<0.001). [Figure 22] Figure 22 shows the analysis of anti-EGFR IgG binding to FaDu (a), inhibition of FaDu cell proliferation (b), and inhibition of EGFR phosphorylation and MAPK pathway signaling (c) (n=3 for binding and proliferation / n=1 for signaling inhibition - data are shown as mean ± SD - proliferation is normalized to cells without mAbs - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01). [Figure 23]Figure 23 shows the analysis of anti-EGFR IgG binding to LIM1215 (a), inhibition of proliferation related to LIM1215 (b), and inhibition of EGFR phosphorylation and MAPK pathway signaling (c) (n=3 for binding and proliferation / n=1 for signaling inhibition - data are shown as mean ± SD - proliferation is normalized to cells without mAbs - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01 / *** - p<0.001 / **** - p<0.0001). [Figure 24] Figure 24 shows the inhibition of proliferation of human cancer cell lines DiFi (a / c) and LIM1215 (b / d) stimulated with EGF (a / b) or HB-EGF (c / d) (n=3 - normalized to cells without mAb - statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01 / *** - p<0.001 / **** - p<0.0001). [Figure 25] Figure 25 shows EGFR expression in stable NIH-3T3 cell lines by Western blotting (a) and analysis of anti-EGFR IgG binding to cell lines NIH-3T3+EGFR wt (b), NIH-3T3+EGFR G465R (c), and NIH-3T3+EGFR S492R (d) analyzed by flow cytometry (data are shown as mean ± SD, n=3). [Figure 26] Figure 26 shows the suppression of EGFR phosphorylation and activation of the MAPK pathway by anti-EGFR IgG in stable mouse fibroblast lines NIH-3T3+EGFR wt(a / b), NIH-3T3+EGFR G465R(c), and NIH-3T3+EGFR S492R(d) under either no stimulation (a) or EGF stimulation (b-c) (n=1). [Figure 27]Figure 27 shows the binding of bispecific EGFR × CD3 antibodies to CD3 on Jurcut cells (a) and to EGFR on cancer cells FaDu (b) and LIM1215 (c), as analyzed by flow cytometry. Cytotoxic assays of FaDu (d) and LIM1215 (e) in co-culture with PBMCs isolated from one healthy donor (n=1 for cell binding, shown as mean ± SD of double repeats; n=3 for cytotoxicity, shown as mean ± SD normalized against untreated target cells). [Figure 28] Figure 28 shows the cell binding (a, c, e) and cytotoxicity (b, d, f) of bispecific EGFR×CD3 antibodies against FaDu (a, b), HCT-116 (c, d), and SW620 (e, f) (for cell binding, n=1 data is shown as mean ± SD of double repeats / for cytotoxicity, n=3 data is shown as mean ± SD of normalized cytotoxicity from three different donors against untreated target cells). [Figure 29] Figure 29 shows quality control of Fab-eIg cetuximab by SEC-HPLC (a). Cytotoxicity analysis of bispecific EGFR×CD3 antibodies on stable NIH-3T3 cell lines expressing EGFR wt, G465R, or S492R, comparing two anti-EGFR antibodies YU250-H10 and cetuximab (b) (n=3 - 3 different donors normalized to untreated target cells - data shown as mean ± SD). [Figure 30] Figure 30 shows the quality control of stable CT-26 cell lines by flow cytometry (a) and Western blotting (b) to quantify EGFR expression levels. The binding of a bispecific EGFR×CD3 antibody, which cross-reactive with mouse CD3, to stable CT-26+EGFR wt (c) and CD3+ spleen cells (d) was analyzed by flow cytometry (n=3(a / c) / n=4(d) - data shown as mean ± SD). [Figure 31]Figure 31 shows T cell activation by a bispecific EGFR × CD3 antibody substitute and T cell proliferation of CD4+(c) and CD8+(d) subtypes, analyzed for CD69+ T cells of CD4+(a) and CD8+(b) subtypes isolated from BALB / c (n=3 - data shown as mean ± SD). [Figure 32] Figure 32 shows the in vivo study of alternatives to bispecific EGFR×CD3 antibodies. The pharmacokinetics of eIg and Fab-eIg were analyzed in BALB / c (a), and the tumor volume of CT-26+EGFR wt was analyzed in BALB / c treated with various doses of eIg or Fab-eIg at 17 days (b) (3 mice per group for pharmacokinetics - data shown as mean ± SD / 6-7 mice per group for tumor models - data shown as mean ± 95% CI / statistics: one-way ANOVA with Tukey post-hoc test - p=0.05 / * - p<0.05 / ** - p<0.01). [Figure 33] Figure 33 shows a formal embodiment of the antigen-binding protein of the present invention, which includes heterodimerization domains. The term "BD" refers to a binding domain, where two binding domains together form an antigen-binding site, for example, BD1 and BD2 together form a first binding site for binding a first antigen, and BD3 and BD4 together form a second binding site for binding the same or a different antigen. [Figure 34] Figure 34 shows the analysis of the bivalent, bispecific antibody Dab-Fc (1+1 form) targeting EGFR and HER3. A. Schematic structure of the domain in the Dab-Fc antibody. B. SDS-PAGE analysis of Dab-Fc antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the Dab-Fc antibody. D. Binding of bivalent, bispecific Dab-Fc was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 35] Figure 35 shows the analysis of bivalent, bispecific antibody eIg (1+1 format) targeting EGFR and HER3. A. Schematic structure of the domain in the eIg antibody. B. SDS-PAGE analysis of eIg antibodies under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the eIg antibody. D. Binding of bivalent, bispecific eIg was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 36] Figure 36 shows the analysis of the tetravalent, bispecific antibody Fab-eIg (2+2 format) targeting EGFR and HER3. A Schematic structure of the domain in the Fab-eIg antibody. B SDS-PAGE analysis of Fab-eIg antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C Size exclusion chromatography of the Fab-eIg antibody. D Binding of tetravalent, bispecific Fab-eIg was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 37] Figure 37 shows the analysis of the quadrivalent, bispecific antibody Db-Ig (2+2 format) targeting EGFR and HER3. A. Schematic structure of the domain in the Db-Ig antibody. B. SDS-PAGE analysis of Db-Ig antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the Db-Ig antibody. D. Binding of quadrivalent, bispecific Db-Ig was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 38]Figure 38 shows the analysis of the trivalent, bispecific antibody Dab / Fab-Fc (2+1 format) targeting EGFR and HER3. A. Schematic structure of the domain in the Dab / Fab-Fc antibody. B. SDS-PAGE analysis of Dab / Fab-Fc antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the Dab / Fab-Fc antibody. D. Binding of trivalent, bispecific Dab / Fab-Fc was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 39] Figure 39 shows the analysis of the bivalent, bispecific antibody Fab / scFv-Fc (1+1 format) targeting EGFR and HER3. A Schematic structure of the domain in the Fab / scFv-Fc antibody. B SDS-PAGE analysis of Fab / scFv-Fc antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C Size exclusion chromatography of the Fab / scFv-Fc antibody. D Binding of bivalent, bispecific Fab / scFv-Fc was analyzed by ELISA using the extracellular domain of EGFR or HER3 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fc antibody. Optical density was measured at 450 nm. [Figure 40] Figure 40 shows the analysis of bivalent, bispecific antibody eIg (1+1 format) targeting EGFR and MET. A. Schematic structure of the domain in the eIg antibody. B. SDS-PAGE analysis of eIg antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the eIg antibody. D. Binding of bivalent, bispecific eIg was analyzed by ELISA using the extracellular domain of EGFR fused to the mouse Fc portion or MET fused to the human Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fab antibody. Optical density was measured at 450 nm. [Figure 41]Figure 41 shows the analysis of bivalent, bispecific antibody eIg (1+1 format) targeting EGFR and HER2. A. Schematic structure of the domain in the eIg antibody. B. SDS-PAGE analysis of eIg antibody under reduced (R) and unreduced (NR) conditions (12% PAA; Coomassie staining) (M: marker). C. Size exclusion chromatography of the eIg antibody. D. Binding of bivalent, bispecific eIg was analyzed by ELISA using the extracellular domain of EGFR or HER2 fused to the mouse Fc portion as the antigen. Binding proteins were detected with HRP-conjugated anti-human Fab antibody. Optical density was measured at 450 nm.

[0037] array The sequences referred to herein are listed in the attached sequence listing. Furthermore, particularly important sequences and their combinations are listed in the table below. The CDR sequences listed in the table below are defined according to Kabat.

[0038] [Table 1-1] [Table 1-2] [Table 1-3]

[0039] The table below shows the CDR sequences included in the variable heavy chain of preferred antibodies of the present invention, as defined using the Kabat definition, IMGT definition, Chothia definition, or Contact definition:

[0040] [Table 2]

[0041] The table below shows the CDR sequences contained in the variable light chains of preferred antibodies of the present invention as defined using the Kabat, IMGT, Chothia, or Contact definitions:

[0042] [Table 3] [Modes for carrying out the invention]

[0043] Before detailing the present invention below, it will be understood that the present invention is not limited to the specific methodologies, protocols, and reagents described herein, as they can be diverse. It will also be understood that the terms used herein are for the purpose of describing specific embodiments only and are not intended to limit the scope of the present invention, which is limited solely by the accompanying claims. Unless otherwise noted, the technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art.

[0044] Preferably, the terms used herein are defined as described in "A multilingual glossary of biotechnological terms: (IUPAC Recommendations)", Leuenberger, HGW, Nagel, B. and Klbl, H. eds. (1995), Helvetica Chimica Acta, CH-4010 Basel, Switzerland.

[0045] The implementation of this invention will, unless otherwise specified, utilize conventional methods of biochemistry, cell biology, immunology, and recombinant DNA techniques, which are described in the literature in those fields (e.g., Molecular Cloning: A Laboratory Manual, 2 ndEdition, J. Sambrook et al. eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor 1989).

[0046] Throughout this specification and the accompanying claims, unless contextually required, the phrase “includes” and its variations such as “includes (third person singular present)” and “contains” implies the inclusion of any member, integer or process, or group of members, integers or processes, described, but not the exclusion of any other member, integer or process, or group of members, integers or processes, except in some embodiments where such other members, integers or processes, or groups of members, integers or processes, are excluded; that is, the subject matter is understood to consist of the inclusion of any member, integer or process, or group of members, integers or processes described.

[0047] In the context describing the present invention (particularly in the context of the claims), “a,” “an,” “the,” and similar reference terms should be construed to encompass both singular and plural forms unless otherwise specified herein or unless it is clearly inconsistent with the context. The enumeration of value ranges herein is intended to function simply as a concise way of referring individually to each separate value that falls within the range. Unless otherwise specified herein, each individual value is incorporated herein as it is individually enumerated herein.

[0048] The elements of the present invention are described below. These elements are listed together with specific embodiments, but it should be understood that they can be combined in any manner and in any number to create further embodiments. The various examples and preferred embodiments described should not be construed as limiting the invention to only the embodiments specified. This description should be understood as supporting and encompassing embodiments that combine the specified embodiments with any number of disclosed and / or preferred elements. Furthermore, unless otherwise indicated in the context, any rearrangement and combination of all the elements described in this application should be considered as disclosed by the description of this application.

[0049] Below, definitions of some terms that are frequently used herein are provided. These terms have their respective defined meanings and preferred meanings in each instance of their use in the remainder of this specification.

[0050] As used herein and in the accompanying claims, the singular forms "a," "an," and "the" include the plural form unless the context clearly indicates otherwise.

[0051] When used in the context of this invention, the term "domain III of human EGFR" refers to the amino acid sequence shown in SEQ ID NO: 61, which corresponds to the amino acids at positions 311-480 of the amino acid sequence in SEQ ID NO: 69, and represents the amino acid sequences of extracellular domains I-IV of human EGFR. SEQ ID NO: 54 shows the amino acid sequence of full-length human EGFR including the signal peptide. SEQ ID NO: 54 includes domain III at positions 335-505. Definitions of each of the domains of human EGFR are further provided in Roskoski et al., 2014.

[0052] As used herein, the term "antigen-binding protein" refers to a molecule containing an antigen-binding site that immunospecifically binds an antigen. It also includes immunoglobulin-like proteins selected by techniques such as phage display that specifically bind to a target molecule or target epitope. The binding and / or specificity of an antigen-binding protein, such as an antibody or an immunologically functional fragment thereof, is determined by whether the antigen-binding protein substantially inhibits the binding of its ligand to its binding partner when an excess of the antigen-binding protein reduces the amount of its binding partner bound to its ligand by at least about 1-20%, 20-30%, 30-40%, 40-50%, 50-60%, 60-70%, 70-80%, 80-85%, 85-90%, 90-95%, 95-97%, 97-98%, 98-99% or more (e.g., as measured in an in vitro competitive binding assay). The neutralizing ability may be described in terms of the IC 50 or EC 50 value.

[0053] The "IC 50 " value refers to the half-maximal inhibitory concentration of a substance and is thus a measure of the effectiveness of a substance in inhibiting a particular biological or biochemical function. The value is usually expressed in molar concentration. The IC 50 of a drug can be determined in a functional antagonism assay by constructing a dose-response curve and examining the inhibitory effect of the test substance at various concentrations. Alternatively, a competitive binding assay may be performed to determine the IC 50 value. Usually, an inhibitory antigen-binding protein exhibits an IC 50 value of 50 nM to 1 pM, i.e., 50 nM, 15 nM, 10 nM, 1 nM, 900 pM, 800 pM, 700 pM, 600 pM, 500 pM, 400 pM, 300 pM, 200 pM, 100 pM, 50 pM, 30 pM, or 1 pM.

[0054] The "EC 50The EC value refers to the maximum effective concentration (FCP) of a substance and is therefore a measure of the concentration of the substance that induces an intermediate response between the baseline and the maximum value after a particular exposure time. It is commonly used as a measure of drug potency. Thus, the EC value of a stepwise dose-response curve 50 This represents the concentration of the substance at which 50% of its maximum effect is observed. EC of the counting dose-response curve 50 This represents the concentration of a compound in which 50% of the population responds after a specific exposure period. Typically, inhibitory antigen-binding proteins are found at concentrations of 50 nM to 1 pM, i.e., 50 nM, 15 nM, 10 nM, 1 nM, 900 pM, 800 pM, 700 pM, 600 pM, 500 pM, 400 pM, 300 pM, 200 pM, 100 pM, 50 pM, 390 pM, or 1 pM EC 50 Show the value.

[0055] The term "binding ability" in this invention preferably refers to specific binding. "Specific binding" means that a binding protein (e.g., an antibody) binds more strongly to a target, such as an epitope, to which it is specific, compared to binding to another target. D When binding to the first target, the binding protein binds more strongly to the first target than to the second target. Preferably, the dissociation constant (K) with respect to the target to which the binding protein specifically binds is D ) is the dissociation constant (K) for targets to which binding proteins do not specifically bind. D Less than 1 / 10 of ), preferably less than 1 / 20, more preferably less than 1 / 50, even more preferably 1 / 100, 1 / 200, 1 / 500 or 1 / 1000, 10 4 1 / 10 5 1 / 10 6 It is less than 1 / 1. Preferably, the term "specific binding" is 10 -5 M or less, 10 -6 M or less, 10 -7 M or less, 10 -8 M or less, 10 -9 M or less, or 10 -10M or less K D This refers to binding to the default target.

[0056] When used herein, "K D The term K₀ (measured in mol / L, sometimes abbreviated as "M") is intended to refer to the dissociation equilibrium constant of a particular interaction between a binding protein (e.g., an antibody or a fragment thereof) and a target molecule (e.g., an antigen or its epitope). The method for determining the binding affinity of a compound, i.e., the dissociation constant K₀, is also used. D Methods for determining are known to those skilled in the art and can be selected from, for example, the following methods known in the art: surface plasmon resonance (SPR) based techniques, biolayer interferometry (BLI), quartz crystal microbalancing (QCM), enzyme-linked immunosorbent assay (ELISA), flow cytometry, isothermal titration calorimetry (ITC), analytical ultracentrifugation, radioimmunoassay (RIA or IRMA), and enhanced chemiluminescence (ECL). In the context of this application, K D The values ​​are determined by biolayer interferometry (BLI). Preferably, biolayer interferometry is performed as described in Example 15.

[0057] The terms “binding site” or “binding domain,” as used herein, refer to an amino acid sequence that has the ability to specifically bind to an antigen, and may, for example, originate from an antibody or its antigen-binding fragment. Examples included in these terms include VL(V L ) domain, VH(V HExamples of variable domains include, but are not limited to, Fab fragments, which are monovalent fragments consisting of domains; Fv fragments, which consist of VL and VH domains of a single arm of an antibody; dAb fragments, which consist of a VH domain or a VL domain (Ward et al., (1989) Nature 341: 544-546); VHH, nanobodies, or variable domains of IgNARs; isolated complementarity-determining regions (CDRs); and combinations of two or more isolated CDRs that may be linked by synthetic peptide linkers. The term also refers to variable domains or regions of the TCRα and TCRβ chains, or the TCRγ and TCRδ chains.

[0058] In the context of this invention, the terms "complementarity-determining region" or "CDR" refer to a variable domain of immunoglobulin, for example, V H , V L , V α and V βThis refers to discontinuous antigen combination sites observed in [the following location]. CDR is referenced in Lefranc (2003). Developmental and Comparative Immunology 27: 55, Kabat et al., J. Biol. Chem. 252: 6609-6616 (1977), Kabat et al., US Dept. of Health and Human Services, "Sequences of proteins of immunological interest", 1991, Chothia et al., J. Mol. Biol. I96: 90I-917, 1987, and Contact annotation (MacCallum et al., J. Mol. Biol. 262: 732-745 (1996) for the Contact annotation), Abhinandan and Martin, Mol. Immunol. (2008), 45(14): 3832-9. for AbM annotation, and IMGT (Lefranc MP. Unique database numbering system for immunogenetic analysis; As described in Immunol Today (1997) 18: 509), the definition includes duplication or subsets of amino acid residues when compared to one another. Nevertheless, the application of any definition to refer to a CDR of an antibody or graft antibody, or their variants or fragments, is intended to be within the scope of terms as defined and used herein. Thus, a CDR may be identified, for example, by the Kabat definition, Chothia definition, IMGT definition, or contact definition. Preferably, the location of the CDR and framework region as defined herein is assigned according to Kabat or Chothia, in particular according to Kabat numbering. Thus, according to a particularly preferred embodiment of the present invention, V described herein H and V L The numbering follows Kabat.

[0059] Those skilled in the art will understand that, particularly in the sequence of a CDR, hypervariable and variable regions can be modified without losing their ability to bind to a target. For example, a CDR region may be identical to or highly homologous to a CDR disclosed herein. "Highly homologous" means that 1 to 3, preferably 1 to 2, or 1 substitution may occur in the CDR.

[0060] The term "antigen," as used herein, refers to a drug containing an epitope recognized by an antigen-binding domain. The term "antigen" includes proteins and peptides in particular. Antigens are preferably equivalent to or derived from naturally occurring antigens. Such naturally occurring antigens may include or be derived from allergens, viruses, bacteria, fungi, parasites, and other infectious substances and pathogens, or antigens may also be tumor antigens. Antigens may be equivalent to naturally occurring products, such as viral proteins or parts thereof, or tumor proteins.

[0061] Variable heavy chain domain (V) used in the context of the present invention H ) and variable light chain domain (V L The antibody is preferably derived from an antibody or immunoglobulin. Such an antibody or immunoglobulin may be a natural or conventional antibody having a "Y" shape and consisting of four polypeptide chains, two identical heavy chains and two identical light chains linked by disulfide bonds. Each chain consists of a structural domain. The two heavy chains are linked to each other by disulfide bonds, and each heavy chain is linked to a light chain by disulfide bonds. There are two types of light chains, lambda (λ) and kappa (κ), and five major heavy chain species (or isotypes) that determine the functional activity of the antibody molecule: IgM, IgD, IgG, IgA, and IgE. Each chain contains a distinct sequence domain. The light chain consists of two domains or regions, V L and steady domain (C L ) contains. The heavy chain has four domains, V H Domain and three constant domains (CH 1, C H 2 and C H 3. Collectively referred to as C H (referred to as) or, in the case of IgE and IgM, five domains, V H Domain and four constant domains (C H 1, C H 2, C H 3, C H 4) includes. V L and V H This determines the binding recognition and specificity to the antigen. Light (C L ) Chain and heavy (C H The constant region domain of the antibody chain confers important biological properties such as antibody chain binding, secretion, transplacental mobility, complement binding, and binding to the Fc receptor (FcR). The "arms" of a "Y"-shaped antibody contain sites that can bind to specific molecules, enabling the recognition of specific antigens. This region of the antibody is called the Fab (fragment, antigen-binding) region. It consists of one constant domain and one variable domain derived from the heavy chain and light chain, respectively. The base of the Y plays a role in regulating immune cell activity. This region is called the Fc (fragment, crystallizable) region and consists of two heavy chains that provide two or three constant domains depending on the type of antibody. The Fv fragment is the N-terminal portion of the Fab region of immunoglobulin and consists of one light chain and one variable portion of the heavy chain. Antibody specificity is attributable to the structural complementarity between the antibody combination site and the antigenic determinant. The antibody combination site consists mainly of hypervariable or CDR-derived residues. In some cases, non-hypervariable or framework (FR)-derived residues affect the entire domain structure and, therefore, the combination site. As mentioned above, CDRs refer to amino acid sequences that determine the binding affinity and specificity of the innate Fv region of the innate immunoglobulin binding site. The light and heavy chains of immunoglobulins each have three CDRs, called CDRL1, CDRL2, CDRL3 and CDRH1, CDRH2, CDRH3, respectively. Therefore, the antigen-binding site of a conventional antibody has six CDRs, including sets of CDRs derived from the heavy chain and light chain variable ("V") regions, respectively.

[0062] The antibodies and their antigen-binding fragments usable in the present invention may originate from any animal, including birds and mammals. Preferably, the antibodies or fragments originate from humans, chimpanzees, rodents (e.g., mice, rats, guinea pigs, or rabbits), chickens, turkeys, pigs, sheep, goats, camels, cattle, horses, donkeys, cats, or dogs. The antibodies are also particularly preferably of human or mouse origin. The antibodies of the present invention also include chimeric molecules in which a constant region of one type of antibody, preferably derived from humans, is combined with an antigen-binding site of another type, e.g., derived from mouse. Furthermore, the antibodies of the present invention include humanized molecules in which an antigen-binding site of an antibody derived from a non-human species (e.g., mouse) is combined with constant and framework regions of human origin.

[0063] As illustrated herein, antibodies can be obtained directly from a hybridoma expressing the antibody, or they can be cloned in a host cell (e.g., a CHO cell or lymphocyte) and recombinantly expressed. Further examples of host cells include microorganisms such as Escherichia coli (E. coli) and fungi such as yeast. Alternatively, further examples of host cells can be recombinantly produced in transgenic non-human animals or plants.

[0064] Suitable antibodies and their antigen-binding fragments in the context of the present invention include, but are not limited to, polyclonal, monoclonal, monovalent, bispecific, heteroconjugate, polyspecific, recombinant, heterogeneous, heterohybrid, chimeric, humanized (especially CDR graft), deimmunized, or human antibodies, Fab fragments, Fab' fragments, F(ab')2 fragments, fragments produced by Fab expression libraries, Fd, Fv, disulfide-linked Fv (dsFv), single-chain antibodies (e.g., scFv), diabodies or tetrabodies (Holliger P. et al. (1993) Proc. Natl. Acad. Sci. USA 90(14), 6444-6448), nanobodies (also known as single-domain antibodies), anti-idiotype (anti-Id) antibodies (e.g., anti-Id antibodies against the antibodies of the present invention), and any of the above epitope-binding fragments.

[0065] In the context of this invention, the term "humanized antibody" refers to an antibody that is entirely or partially of non-human origin and has been modified to avoid or minimize the immune response in humans, particularly by replacing certain amino acids in the framework regions of the heavy and light chains. The constant domains of a humanized antibody are primarily human heavy and light chain domains. Methods relating to the humanization of antibody sequences are known in the art (Almagro & Fransson (2008) Front Biosci. 13: 1619-1633). One commonly used method is the CDR grafting method, or antibody remodeling, which involves grafting the CDR sequence of a donor antibody, generally a mouse antibody, onto a framework scaffold of a human antibody of different specificity. In the CDR grafting method, reverse mutations may be introduced at selected locations in the CDR grafted antibody to preserve the binding specificity and affinity of the parent antibody, as this can reduce the binding specificity and affinity and therefore the biological activity of the CDR grafted non-human antibody. While amino acid residues that make up a CDR are usually left unchanged, in certain cases it may be desirable to modify individual CDR amino acid residues, for example, to remove glycosylation sites, deamide sites, isomerization sites, or undesirable cysteine ​​residues. N-linked glycosylation occurs by the attachment of oligosaccharide chains to asparagine residues in the tripeptide sequences Asn-X-Ser or Asn-X-Thr (wherein X can be any amino acid except Pro). Removal of N-glycosylation sites can be achieved by mutating either the Asn or Ser / Thr residue to a different residue, particularly by conservative substitution. Deamide of asparagine and glutamine residues can occur depending on factors such as pH and surface exposure. Asparagine residues are particularly susceptible to deamide, mainly when present in the Asn-Gly sequence, and to a lesser extent in other dipeptide sequences such as Asn-Ser. Therefore, if such a deamidation site, particularly Asn-Gly, is present in the CDR sequence, it may be desirable to remove the site by conservative substitution, which usually involves removing one of the residues involved.Substitutions in the CDR sequence that remove one of the involved residues are also intended to be included in the present invention.

[0066] The term "dimerizing domain," as used herein, refers to a domain capable of forming a dimer of two peptide or protein chains, wherein at least one dimerizing domain is present on the first chain and at least a second dimerizing domain is present on the second chain. The dimerizing domain comprises an Fc region, a heterodimerizing Fc region, and C H 1, C L , the second heavy chain constant domain of IgE and IgM (C H 2) (EHD2, MHD2), modified EHD2, IgG, IgD, IgA, or the last heavy chain constant domain of IgE (C H or C H 4) and their heterodimerization derivatives, as well as the group consisting of the constant domains C-alpha and C-beta of the T cell receptor, can be selected. Depending on each dimerization domain, the C-terminus and N-terminus of the dimerization domain may be diverse. If the dimerization domain is derived from a naturally occurring protein, such as an immunoglobulin, it is preferable that the dimerization domain is directly linked to the variable domain in the sense of the present invention, i.e., if there is a naturally occurring amino acid at its C-terminus or N-terminus, the dimerization domain is linked without a peptide linker. When used in the context of the present invention, the dimerization domains preferably form one or more covalent bonds between themselves, preferably one or two covalent bonds, and most preferably one covalent bond.

[0067] Where used herein, the term “eIg” refers to a heterodimerized domain containing a modified IgE heavy chain domain 2 (EHD2). An example of an eIg domain is described, for example, in International Publication No. 2021 / 058804, which is incorporated herein by reference. Where used herein, the terms “EHD2” and “EHD2 domain” refer to the second constant domain of the IgE heavy chain (Seifert et al., 2012, Protein Eng Des Sel.; 25: 603-12, and Seifert et al., 2014, Mol Cancer Ther.; 13: 101-11). Similarly, the term “modified EHD2 domain” refers to a portion of the EHD2 domain whose sequence has been modified compared to the original EHD2 sequence from which it is derived. Modifications include amino acid deletions, substitutions, and insertions. Modifications preferably involve one or more amino acid substitutions. Preferred substitutions include C14S and C102S, respectively, with respect to the wild-type human EHD2 core amino acid sequence, as shown in Sequence ID No. 91. As used herein, the term "hetEHD2" refers to a heterodimer of two modified EHD2 domains, i.e., a dimer containing first and second EHD2 domains that are distinct from each other at at least one amino acid position. Exemplary hetEHD2 domains to be used in the context of the present invention include or consist of amino acid sequences as shown in Sequence ID Nos. 94 and 95. It will be understood by those skilled in the art that a heterodimerizing domain requires a first domain and a second domain to which the first domain binds. To prevent homodimerization, the two domains are distinct from each other, allowing the first heterodimerizing domain to bind only to each of the distinct second heterodimerizing domains. Thus, according to a preferred embodiment of the present invention, the first heterodimerizing domain includes an amino acid sequence as described in Sequence ID No. 94, and each second heterodimerizing domain includes an amino acid sequence as described in Sequence ID No. 95. The eIg form described herein is preferably used for a bivalent bispecific antigen-binding protein.The construct having the eIg form is described in more detail in International Publication No. 2021 / 058804. Specific embodiments of the antigen-binding protein of the present invention using the hetEHD2 domain are shown in Figure 33.

[0068] Where referred to herein, the term “Fab-eIg” refers to a Fab directly or indirectly linked to an eIg as described herein. The Fab-eIg format is as described in International Publication No. 2021 / 058804, C H 1 / C L V instead H / V L This is based on applying a heterodimerization module eIg designed for dimerization. The Fab-eIg form is preferably used for trivalent bispecific antigen-binding proteins. For this purpose, the first antigen-binding moiety is formed by the first variable heavy chain C H 1. Also, the first variable light chain C L It is produced by fusing the first hetEHD2 with the second hetEHD2. The second antigen-binding moiety is produced, as described above, by fusing the second variable heavy chain with the first hetEHD2 and the second variable light chain with each of the second hetEHD2s. A specific embodiment of the antigen-binding protein of the present invention using the Fab-eIg format is shown in Figure 33D.

[0069] The term “heterodimizing Fc” refers to variants of Fc capable of forming heterodimers. Examples of heterodimerizing Fc include other variants of Fc described in the literature on knob-into-hole techniques and the generation of heterodimeric Fc (Krah et al., 2007, Ha et al., 2016, Mimoto et al., 2016, Brinkmann & Kontermann, 2017). The technique, also known as knob-into-hole or knobs-into-holes, preferably refers to both the CH3-CH3 interface mutations Y349C, T366S, L368A, and Y407V (holes) and S354C and T366W (knobs) for promoting heteromultimerization, described in particular U.S. Patents 5,731,168 and 8,216,805, both of which are incorporated herein by reference.

[0070] As used herein, the term “diabody” refers to a bivalent molecule capable of binding to either two antigens of the same type (monospecificity) or two antigens of different types (bispecificity). Diabodies are described, for example, by Holliger et al. (1993) Proc. Natl. Acad. Sci. USA 90(14), 6444-6448, the contents of which are incorporated herein by reference.

[0071] As used herein, the term “single-chain diabody (scDb)” refers to a derivative of a diabody in which four variable domains of one or two antibodies are linked by three linkers.

[0072] The term "competing" is used in the context of antigen-binding proteins (e.g., neutralizing antigen-binding proteins or neutralizing antibodies) competing for the same epitope, meaning competition between antigen-binding proteins when the antigen-binding protein being tested (e.g., an antibody or an immunologically functional fragment thereof) is determined by an assay that prevents or inhibits (e.g., reduces) the specific binding of a reference antigen-binding protein (e.g., a ligand, or a reference antibody) to a common antigen. Numerous types of competitive binding assays can be used to determine whether one antigen-binding protein competes with another: for example, solid-phase direct or indirect radioimmunoassays (RIAs), solid-phase direct or indirect enzymatic immunoassays (EIAs), sandwich competition assays (see, e.g., Stahli et al., 1983, Methods in Enzymology. 2: 242-253), solid-phase direct biotin-avidin EIA (see, e.g., Kirkland et al., 1986, J. Immunol. 137: 3614-3619), solid-phase direct labeling assays, solid-phase direct labeling sandwich assays (see, e.g., Harlow and Lane, 1988, Antibodies, A Laboratory Manual, Cold Spring Harbor Press), and solid-phase direct labeling RIA using I-125 labeling (see, e.g., Morel et al., 1988, Molec. Jmmunol. 25: (See 7-15), solid-phase direct biotin avidin EIA (see, e.g., Cheung et al., 1990, Virology 176: 546-552), and directly labeled RIA (Moldenhauer et al., 1990, Scand. J Immunol. 32: 77-82). Typically, such assays involve the use of purified antigen bound to a solid surface or cells possessing either of these, an unlabeled test antigen-binding protein, and a labeled reference antigen-binding protein. Competitive inhibition is measured by determining the amount of label bound to the solid surface or cells in the presence of the test antigen-binding protein. Typically, the test antigen-binding protein is present in excess.Antigen-binding proteins identified by competitive assays (competitive antigen-binding proteins) include antigen-binding proteins that bind to the same epitope as the reference antigen-binding protein, and antigen-binding proteins that bind to an adjacent epitope sufficiently proximal to the epitope to which the reference antigen-binding protein binds to cause steric hindrance. Typically, when competitive antigen-binding proteins are present in excess, they inhibit (reduce) the specific binding of the reference antigen-binding protein to the common antigen by at least 40-45%, 45-50%, 50-55%, 55-60%, 60-65%, 65-70%, 70-75%, or 75% or more. In some cases, binding is inhibited by at least 80-85%, 85-90%, 90-95%, 95-97%, or 97% or more.

[0073] The terms “nucleic acid” and “nucleic acid (plural)” and “nucleic acid molecule (plural)” are used herein as synonyms and are understood as single- or double-stranded oligomers or polymers of deoxyribonucleotides or ribonucleotide bases, or both. Nucleic acid monomers consist of a nucleic acid base, a 5-carbon sugar (such as, but not limited to, ribose or 2'-deoxyribose), and one to three phosphate groups. Typically, nucleic acids are formed by phosphate diester bonds between individual nucleotide monomers. In the context of this invention, the term nucleic acid includes, but is not limited to, ribonucleic acid (RNA) and deoxyribonucleic acid (DNA) molecules, as well as synthetic forms of nucleic acids containing other bonds (e.g., peptide nucleic acids, as described in Nielsen et al. (Science 254: 1497-1500, 1991)). Typically, nucleic acids are single- or double-stranded molecules composed of naturally occurring nucleotides. The single-stranded description of nucleic acids also (at least partially) defines the sequence of the complementary strand. Nucleic acids may be single-stranded or double-stranded, or may contain portions of both double-stranded and single-stranded sequences. Exemplary double-stranded nucleic acid molecules may have 3' or 5' overhangs and therefore do not need to be, nor are they assumed to be, entirely double-stranded over their entire length. Nucleic acids may be obtained by any method known in the art, including but not limited to biological, biochemical, or chemical synthesis methods, or methods of amplification and reverse transcription of RNA. The term nucleic acid includes chromosomes or chromosomal segments, vectors (e.g., expression vectors), expression cassettes, naked DNA or RNA polymers, primers, probes, cDNA, genomic DNA, recombinant DNA, cRNA, mRNA, tRNA, microRNA (miRNA), or small interfering RNA (siRNA). Nucleic acids may be, for example, single-stranded, double-stranded, or triple-stranded, and are not limited to any particular length. Unless otherwise noted, a particular nucleic acid sequence includes or encodes a complementary sequence in addition to any sequence specified.

[0074] The term "pharmaceutical composition," as used herein, refers to a substance and / or combination of substances used to identify, prevent, or treat a disease or tissue condition. Pharmaceutical compositions are formulated to be suitable for administration to a patient to prevent and / or treat a disease. A pharmaceutical composition refers to a combination of an activator and an inactive or active carrier that makes the composition suitable for therapeutic use. Depending on its chemical and physical properties, pharmaceutical compositions may be formulated for oral, parenteral, topical, inhalation, rectal, sublingual, transdermal, subcutaneous, or vaginal application routes. Pharmaceutical compositions include solids, semi-solids, liquids, and transdermal therapeutic systems (TTS). Solid compositions are selected from the group consisting of tablets, coatings, powders, granules, pellets, capsules, effervescent tablets, or transdermal therapeutic systems. Liquid compositions are also included, selected from the group consisting of solutions, syrups, infusions, extracts, solutions for intravenous application, solutions for infusions, or solutions of the carrier systems of the present invention. Semi-solid formulations that may be used in the context of the present invention include emulsions, suspensions, creams, lotions, gels, globules, buccal tablets, and suppositories.

[0075] The term "activator" refers to a substance in a pharmaceutical composition or formulation that is biologically active, i.e., provides pharmaceutically valuable. According to the present invention, a pharmaceutical composition comprises at least one or more binding proteins, nucleic acids, vectors, or recombinant cells according to the present invention as activators. A pharmaceutical composition may contain more than one activator that can act in conjunction with or independently of each other. Activators can be formulated, for example, in neutral or salt form. Examples of pharmaceutically acceptable salts include salts formed with free amino groups, such as those derived from hydrochloric acid, phosphoric acid, acetic acid, oxalic acid, tartaric acid, etc., and salts formed with free carboxyl groups, such as, but not limited to, those derived from sodium, potassium, ammonium, calcium, ferric hydroxide, isopropylamine, triethylamine, 2-ethylaminoethanol, histidine, procaine, etc.

[0076] The term "identity," also referred to as "amino acid sequence identity," in the context of this invention refers to a percentage of sequence identity, determined by comparing two optimally aligned sequences across a comparison window, where the portion of the sequence in the comparison window may include additions or deletions (i.e., gaps) when compared to a reference sequence for optimal alignment of the two sequences (which does not include insertions or deletions). The percentage is calculated by determining the number of positions in which identical nucleic acid bases or amino acid residues are found in both sequences to obtain the number of matching positions, dividing the number of matching positions by the total number of positions in the comparison window, and multiplying the result by 100 to obtain the percentage of sequence identity.

[0077] The term “identical” in the context of two or more polypeptide or nucleic acid sequences means that two or more sequences or subsequences are the same, i.e., contain the same sequence of amino acids or nucleic acids. Sequences are substantially identical to each other if, when compared and aligned for the maximum correspondence across a comparison window or specified region, as measured by one of the sequence comparison algorithms below, or by manual alignment and visual inspection, they have a specified percentage of identical amino acid residues (e.g., at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identity). These definitions also refer to complements of the sequences being compared. Thus, the term “at least 90% sequence identity” is used throughout this specification in reference to the comparison of polypeptide and polynucleotide sequences. This expression preferably refers to sequence identity of at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% with respect to each reference polypeptide or each reference polynucleotide.

[0078] In the context of this invention, comparing two sequences refers to the process in which one sequence acts as a reference sequence and a check sequence is compared to it. When using a sequence comparison algorithm, the check sequence and the reference sequence are input into a computer, the coordinates of subsequences are specified as needed, and the sequence algorithm program parameters are specified. Default program parameters are usually used, or alternative parameters may be specified. Subsequently, the sequence comparison algorithm calculates a percentage of sequence identity or similarity of the check sequence to the reference sequence based on the program parameters. If two sequences are compared and the reference sequence for which the sequence identity percentage should be calculated is not specified in the comparison, the sequence identity should be calculated using the longer of the two sequences being compared as the reference, unless otherwise specifically stated. If a reference sequence is provided, the sequence identity is determined based on the full length of the reference sequence, as indicated by its sequence number, unless otherwise specifically stated.

[0079] Methods for aligning sequences for comparison are well known in the art. Optimal alignment of sequences for comparison can be performed, for example, by Smith and Waterman's local phase-dynamic algorithm (Adv. Appl. Math. 2: 482, 1970), Needleman and Wunsch's homology alignment algorithm (J. Mol. Biol. 48: 443, 1970), Pearson and Lipman's similarity search method (Proc. Natl. Acad. Sci. USA 85: 2444, 1988), by computer implementations of these algorithms (e.g., the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Dr., Madison, Wisconsin, GAP, BESTFIT, FASTA, and TFASTA), or by manual alignment and visual inspection (e.g., see Ausubel et al., Current Protocols in Molecular Biology (1995 supplement)). Suitable algorithms for determining sequence identity and sequence similarity percentages are the BLAST and BLAST 2.0 algorithms, described in Altschul et al. (Nuc. Acids Res. 25: 3389-402, 1977) and Altschul et al. (J. Mol. Biol. 215: 403-10, 1990), respectively. Software for performing BLAST analysis is available through the National Center for Biotechnology Information (http: / / www.ncbi.nlm.nih.gov / ).

[0080] All methods described herein may be carried out in any suitable order, unless otherwise specified herein or unless it is clearly inconsistent with the context. Any examples or exemplary language provided herein (e.g., “e.g., “etc.”) are intended solely to better illustrate the invention and do not impose any limitation on the scope of the invention as otherwise claimed. The language herein should not be construed as indicating any unclaimed element essential to the carrying out of the invention.

[0081] Throughout this specification, several documents are referenced. All documents referenced herein (including all patents, patent applications, scientific publications, manufacturer specifications, instructions, etc.) are incorporated herein by reference, whether above or below. Nothing in this specification should be construed as acknowledging that the present invention does not have prior rights to such disclosure by prior art. Some of the documents referenced herein are characterized as “incorporated by reference.” In the event of any conflict between the definitions or teachings of such incorporated references and the definitions or teachings listed herein, the text of this specification shall prevail.

[0082] Various aspects of the present invention are described in more detail below. Each of these aspects may be combined with any other aspect(s) unless otherwise clearly indicated. In particular, any feature indicated as preferred or desirable may be combined with any other feature(s) indicated as preferred or desirable.

[0083] In a first aspect, the present invention provides an antigen-binding protein comprising a first antibody-variable domain and a second antibody-variable domain, wherein the variable domain is a) determined by biolayer interferometry (BLI) -5 K below molar concentration Da) It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more of the mutations V441D, S464L, G465R, and S492R, and b) forms a binding site that does not specifically bind to human EGFR having the mutation Q435P and / or the double mutation F436A / I462A determined by biolayer interferometry (BLI), where the binding of the antigen-binding protein to human EGFR prevents ligand-induced activation of human EGFR. The antigen-binding protein of the present invention is determined by biolayer interferometry (BLI) to form a binding site that does not specifically bind to human EGFR having the mutation Q435P and / or the double mutation F436A / I462A, where the binding of the antigen-binding protein to human EGFR prevents ligand-induced activation of human EGFR. -5 A concentration exceeding the molar concentration, preferably 10 -4 More preferably 10 -3 A concentration exceeding the molar concentration, more preferably 10 -2 K exceeding molar concentration D Therefore, it does not specifically bind to human EGFR having the mutation Q435P and / or the double mutation F436A / I462A. Preferably, the variable domain of the first antibody is V L Therefore, the variable domain of the second antibody is V H That is the case.

[0084] According to a preferred embodiment, the antigen-binding protein of the present invention is determined by biolayer interference (BLI) 10 -5The antigen-binding protein specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more of the mutations V441D, V441G, S442R, S464L, G465R, K489E, and S492R at a KD concentration of less than molal, but does not specifically bind to human EGFR having the mutation Q435P and / or the double mutation F436A / I462A. More specifically, the antigen-binding protein specifically binds to domain III (SEQ ID NO: 61) of human EGFR, and preferably, the antigen-binding protein of the present invention binds to one or more of the mutations V441D, S464L, G465R, and S492R, for example, V441D, S464L, G465R, or S492R alone, or any combination thereof, for example, V441D and S464L, or V441D and G465R, or V441D and G465R, or V441D The antigen-binding protein of the present invention specifically binds to variant domain III of human EGFR having one or more, preferably all four, mutations V441D, S464L, G465R, and S492R, or S465R and S492R, or V441D, S464L, G465R, and S492R, or S464L, G465R, and S492R, or all four of V441D, S464L, G465R, and S492R. In other words, the antigen-binding protein of the present invention recognizes EGFR variants having one or more, preferably all four, mutations V441D, S464L, G465R, and S492R, regardless of which (one or more) of these mutations are present in the EGFR variant.

[0085] In a particularly preferred embodiment, the antigen-binding protein of the present invention is determined by biolayer interference (BLI) as defined herein. -5 It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more mutations among V441D, S464L, G465R, and S492R at a KD of less than molal concentration, but does not specifically bind to human EGFR having mutations Q435P and double mutation F436A / I462A.

[0086] According to an alternative preferred embodiment, the antigen-binding protein of the present invention is determined by biolayer interference (BLI) as defined herein. -5 It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more mutations among V441D, S464L, G465R, and S492R at a KD of less than molal concentration, but does not specifically bind to human EGFR having mutation Q435P or double mutation F436A / I462A.

[0087] The amino acid sequence of the human epidermal growth factor receptor (EGFR) containing the signal peptide is as shown in SEQ ID NO: 54. The locations of mutations in human EGFR, particularly the mutations V441D, S464L, G465R, and S492R, as well as Q435P and F436A / I462A, which are referred to herein, refer to SEQ ID NO: 54.

[0088] Common ligands for EGFR are EGF, EPG, TGFα, AR, BTC, HB-EGF, and EPR (Roskoski, Pharmacological Research 2014: 79: 34-74). Inhibition of ligand-induced activation of EGFR can be determined by methods known in the art. One exemplary method is Western blotting to detect phosphorylation of EGFR, such as phosphorylation at position Y1068 and / or Y1173, using one or more phosphorus-specific antibodies (Roskoski et al. 2014, above).

[0089] The antigen-binding protein of the present invention preferably binds to a specific epitope in human EGFR, which includes amino acid residues G434, Q435, F436, and I462 with respect to SEQ ID NO: 54.

[0090] According to the present invention, the variable domain forms a binding site that specifically binds to domain III of human EGFR having one or more of the mutations V441D, S464L, G465R, and S492R (SEQ ID NO: 61). According to a preferred embodiment, the variable domain forms a binding site that recognizes domain III of human EGFR having one or more of the mutations V441D, S464L, G465R, and S492R as defined above (SEQ ID NO: 61). As used in this context, the term "specifically binds to domain III of human EGFR" means 10 as determined by biolayer interferometry (BLI). -5 K below molar concentration D Bonding at, preferably 10 -6 K below molar concentration D Bonding in, furnace 10 -7 K below molar concentration D This means binding at [location]. According to a preferred embodiment, the binding site is less than 1 μM K when determined by biolayer interferometry (BLI). D Preferably, K is 0.1nM to 999nM, more preferably 0.5nM to 850nM. DThe mutant specifically binds to domain III of human EGFR having one or more of the mutations V441D, S464L, G465R, and S492R. The specific sites referred to herein, in particular the mutations V441D, S464L, G465R, and S492R, refer to Sequence ID No. 54. Since the amino acid sequence of domain III of human EGFR as identified herein is contained within Sequence ID No. 54, the mutation sites referred to herein can also be identified in Sequence ID No. 61, which shows the amino acid sequence of domain III of human EGFR. Thus, the mutation V441D referred to herein corresponds to V107D in Sequence ID No. 61. Similarly, the mutation S464L referred to herein corresponds to S130L in Sequence ID No. 61, the mutation G465R referred to herein corresponds to G131R in Sequence ID No. 61, and the mutation S492R referred to herein corresponds to S158R in Sequence ID No. 61. Accordingly, according to one embodiment, the variable domain of the antigen-binding protein of the present invention forms a binding site that specifically binds to domain III of human EGFR having one or more mutations V107D, S130L, G131R, and S158R, as per SEQ ID NO: 61.

[0091] Accordingly, the other mutations in EGFR referred herein with respect to SEQ ID NO: 54, Q435P and F436A / I462A, are identical to the mutations Q101P and F102A / I128A with respect to SEQ ID NO: 61. Therefore, according to one embodiment, the antigen-binding protein of the present invention does not specifically bind to human EGFR having mutation Q101P or double mutation F102A / I128A with respect to SEQ ID NO: 61. Similarly, amino acids G434, Q435, F436 and I462, which are included by the specific epitopes to which the antigen-binding protein of the present invention preferably binds, refer to SEQ ID NO: 54 and correspond to amino acids G100, Q101, F102 and I128 of SEQ ID NO: 61, respectively.

[0092] By a simple comparison (e.g., alignment) between SEQ ID NO: 54 and SEQ ID NO: 61 or any other amino acid sequence, all amino acid positions referred to herein can be associated with such other amino acid sequences, such as the amino acid sequence shown in SEQ ID NO: 61.

[0093] According to the present invention, the antigen-binding protein is not limited to any particular form or shape. The antigen-binding protein may consist of only a single polypeptide chain containing first and second variable domains, for example in the form of scFv, or the antigen-binding protein may contain first and second variable domains on two separate polypeptides. The antigen-binding protein may further comprise identical or different polypeptide chains, for example, a total of three, four, five, or six polypeptide chains. These polypeptide chains of the antigen-binding protein of the present invention, or these polypeptide chains(s), may include further functionality linked directly or indirectly to the first and / or second variable domains, for example, via a peptide linker. Exemplary functionality is described in more detail below and may include: (i) One or more antigen-binding sites having the same or different specificity, (ii) homodimerized domain, (iii) Trimerizing domain or tetramerizing domain, (iv) CH1 or C L domain.

[0094] The antigen-binding proteins of the present invention are preferably selected from the group consisting of Fab, Fab', (Fab')2, Fv, disulfide-linked Fv, BsFv, dsFv, (dsFv)2, dsFv-dsFv', scFv, scFv dimers, single-chain (sc) domain antibodies, diabody (Db), scDb, dsDb, and bivalent domain antibodies, but are not limited thereto. They also include forms that retain the ability to specifically bind to domain III of human EGFR. Examples include Fab fragments, F(ab')2 fragments, Fd fragments, Fv fragments, and dAb fragments. The antigen-binding proteins of the present invention may also exist in the form of chimeric antigen receptors (CARs). The most preferred form of the antigen-binding proteins of the present invention is an antibody. According to a more preferred embodiment of the present invention, the antibody is selected from the group consisting of chimeric antibodies, bispecific antibodies, and polyspecific antibodies. The antibody is preferably human or humanized.

[0095] In certain embodiments of the present invention, the antigen-binding protein is monospecific, bispecific, or polyspecific. In certain embodiments, the bispecific or polyspecific antigen-binding protein specifically binds to at least one further second bicellular target. In preferred embodiments, the at least one second cellular target is selected from the group consisting of proteins expressed on the surface of immune cells, such as CD2, CD3, CD16, CD44, CD64, CD69, CD89, Mel14, Ly-6.2C, TCR complex, Vy9V52 TCR, and NKG2D, and is preferably CD3. Particularly preferred anti-CD3 antigen-binding proteins include the variable light and heavy chains (SEQ ID NOs. 96 and 97) of the anti-human CD3 antibody UCHT1(huU3). In a more preferred embodiment, the second cell target is selected from the group consisting of human epidermal growth factor receptor 2 (HER2), human epidermal growth factor receptor 3 (HER3), human epidermal growth factor receptor 4 (HER4), insulin-like growth factor 1 receptor (IGF-1R), hepatocyte growth factor receptor (HGFR, c-MET), and derivatives thereof. In a particularly preferred embodiment, the second cell target is HER2, HER3, or MET, more preferably HER3. In a particularly preferred embodiment, the bispecific or polyspecific binding protein according to the present invention specifically binds to domains III and CD3 of human EGFR, where specifically binding means 10 as determined by biolayer interferometry (BLI). -5 This means that the binding occurs at a molar concentration below a certain level.

[0096] In certain embodiments of the present invention, the antigen-binding protein is trivalent or tetravalent. In certain embodiments, the antigen-binding protein includes an effector domain to which the Fc receptor, neonatal Fc receptor (FcRn), or complement system binds. In certain embodiments, the Fc domain is a domain to which the Fc gamma receptor, in particular CD16, CD32, and / or CD64 binds. In certain embodiments, the Fc domain is a domain that activates the complement system, in particular by binding to C1q of the complement system.

[0097] In a preferred embodiment of the present invention, the antigen-binding protein is bivalent. Unless otherwise specified, the structure of the antigen-binding protein embodiments described herein is depicted from the left N-terminus to the right C-terminus. It is more preferred that the antigen-binding protein is bivalent and bispecific. In a further preferred embodiment, the bivalent and bispecific antigen-binding protein is a diabody. The bispecific diabody comprises two chains each containing a VH and a VL domain derived from different antigen-binding molecules. The two variable domains VH and VL are preferably linked by a short linker of 3 to 5 residues.

[0098] The diabody according to the present invention can be a two-chain diabody (Db) or a single-chain diabody (scDb). For the two-chain diabody, the two chains may have V H A-V L B and V H B-V L A or V L A-V H B and V L B-V H A, where A and B represent two different specificities. For the single-chain diabody, the first chain, V H A-V L B or V L A-V H B, and the second chain, V H B-V L A or V L B-V H A are covalently bonded to each other. Preferably, the first and second chains are linked via a peptide linker having a length of 10 to 15 amino acids. Preferably, the bispecific diabody is scDb. Preferably, the antigen-binding protein has a structure (V H A-V L B-V H B-V L A)scDb. In a particularly preferred embodiment, the antigen-binding protein comprises or consists of the amino acid sequence of SEQ ID NO: 19.

[0099] In a more preferred embodiment, the antigen-binding protein is one or more scFvs, preferably one or two scFvs bound to a bispecific Db or bispecific scDb, preferably bispecific scDb. Two or more scFvs may be bound in tandem. The scFv preferably contains the VH and VL domains of the same antigen-binding protein linked by a peptide linker of about 10 to 25 amino acids. The scFv is composed of VH and VL domains. H -V L or V L -V H It may have the following. Preferably, one or more scFv have the specificity of either a bispecificity Db or a bispecificity scDb, or both. Therefore, the scFv preferably has the specificity of a bispecificity V H AV L A or V L AV H Having A, or configuration V H BV L B or V L BV H B may be present. In a more preferred embodiment, one or more scFv may have a specificity different from that of the double specificity Db or the double specificity scDb. Therefore, one or more scFv may have a specificity different from that of the double specificity V H CV L C or V L CV H C, or V H DV L D or V L DV H It may have D, etc. In a preferred embodiment, the antigen-binding protein is a bispecific trivalent antigen-binding protein. Preferably, the antigen-binding protein has a composition (V H AV L BV H BV L A)scDb-(V H AV L A) It has scFv. In a particularly preferred embodiment, the antigen-binding protein contains or consists of the amino acid sequence of SEQ ID NO: 20.

[0100] In a more preferred embodiment, the antigen-binding protein is composed of two bispecific Db or bispecific scDb, preferably bispecific scDb, each bound to an Fc region, where the Fc region functions as a homodimerization domain. In a preferred embodiment, the antigen-binding protein is composed of (V H AV L BV H BV L A) It contains two parts, scDb-Fc. The two parts can be joined by a covalent or non-covalent bond.

[0101] In a further preferred embodiment of the bispecific antigen-binding protein, further V of the second specificity H Domain and V L The domains are attached to the light chain and heavy chain, respectively, where the Fc region of the heavy chain functions as a dimerization domain. Two V with different specificities H Domain and two Vs L The domains may be bound to the light and heavy chains in various combinations, resulting in different configurations. In a preferred embodiment of the antigen-binding protein, the light chain is composed of V H AV H BC L It has k, and the heavy chain is composed of V L AV L It has B-CH1-CH2-CH3. A certain configuration is V H Domain and V L Enables crossover pair formation of domains. In a preferred embodiment, the light chain is configured V H AV L BC L It has k, and the heavy chain is composed of V H BV L It has A-CH1-CH2-CH3. In a preferred embodiment, the light chain is V L AV L BC L It has a k configuration, and the heavy chain has a V configuration. H BV H It has A-CH1-CH2-CH3.

[0102] In each of the above examples, the letters "A," "B," "C," and "D" symbolize the antigenic specificity of the antigen-binding protein of the present invention. At least one of "A," "B," "C," and "D" in each antigen-binding protein of the present invention specifically binds to the conformational epitope formed by domain III of the human epidermal growth factor receptor EGFR, as defined by SEQ ID NO: 61. Other specificities may be the same or different. Preferred second and further specificities are outlined below.

[0103] Further examples of bispecific antibodies are described in Brinkmann U & Kontermann RE, MABS, 2017, 9(2), 182-212 and International Publication No. 2021 / 058804, both of which are specifically incorporated into this specification by reference.

[0104] Particularly preferred forms of the antigen-binding protein of the present invention include bivalent and monospecific, bivalent and bispecific, trivalent and bispecific, and tetravalent and bispecific. Particularly preferred forms of the antigen-binding protein of the present invention are shown in Figure 33.

[0105] In certain embodiments, the antigen-binding protein of the present invention comprises a multimerizing domain. Preferred examples include a dimerizing domain, a trimerizing domain, or a tetramerizing domain. When two protein chains bind to each other, each comprises a dimerizing domain capable of binding to at least one dimerizing domain in the other protein. Thus, when the antigen-binding protein comprises three protein chains, each comprises at least one trimerizing domain capable of interacting with each trimerizing domain of the other chain. In certain embodiments, the dimerizing domain is selected from the group consisting of heavy chain domain 2 (CH2) of IgM (MHD2) or IgE (EHD2), immunoglobulin Fc region, heavy chain domain 3 (CH3) of IgG or IgA, heavy chain domain 4 (CH4) of IgM or IgE, Fab, Fab2, leucine zipper motif, burnase-burster dimer, mini-antibodies, and ZIP mini-antibodies.

[0106] In certain embodiments, the trimerizing domain is selected from the group consisting of tenascin C (TNC), the trimerizing region of the C-terminal noncollagenous domain (NC1) of collagen XVIII, Fab3-like molecules, and TriBi minibodies.

[0107] In certain embodiments, the tetramerization domain is selected from the group consisting of the tetramerization domain of p53, the tetramerization domain of general regulatory protein 4 (GCN4), the tetramerization domain of VASP (vasodilation-stimulated phosphorylated protein), tandem diabodies, and di-diabodies.

[0108] In some embodiments, particularly when two protein chains with different antigen specificities are present, the use of heterodimerizing domains is preferred. Particularly preferred heterodimerizing domains are eIg and Fab-eIg, as described herein.

[0109] The present invention provides an antigen-binding molecule comprising a first antibody-variable domain and a second antibody-variable domain. The first antibody-variable domain preferably comprises the light chain complementarity-determining regions CDRL1, CDRL2, and CDRL3 of the variable light chain region of SEQ ID NO: 83, 84, or 85. The second antibody-variable domain preferably comprises the heavy chain complementarity-determining regions CDRH1, CDRH2, and CDRH3 of the heavy chain variable region of SEQ ID NO: 74 or 75, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, it is preferable that the substitutions are conservative amino acid substitutions.

[0110] Accordingly, according to a preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 83 and HCDR of SEQ ID NO: 74, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0111] In a more preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 84 and HCDR of SEQ ID NO: 74, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0112] In a more preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 85 and HCDR of SEQ ID NO: 74, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0113] In a more preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 83 and HCDR of SEQ ID NO: 75, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0114] In a more preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 84 and HCDR of SEQ ID NO: 75, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0115] In a more preferred embodiment, the antigen-binding molecule comprises LCDR of SEQ ID NO: 85 and HCDR of SEQ ID NO: 75, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0116] When referring to CDR, the sequence and position of amino acid residues are preferably identified by Kabat definition, Chothia definition, IMGT definition, or contact definition. According to a preferred embodiment, CDR is identified by Kabat definition.

[0117] The antigen-binding molecule according to the present invention may also be defined by the following complementarity-determining regions of the variable heavy and light chains.

[0118] In a preferred embodiment, the first antibody variable domain comprises CDRL1 of SEQ ID NO: 86, CDRL2 of SEQ ID NO: 87, and CDRL3 of SEQ ID NO: 88, and the second antibody variable domain comprises CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0119] In another preferred embodiment, the first antibody-variable domain comprises CDRL1 of SEQ ID NO: 76 or 77, CDRL2 of SEQ ID NO: 78 or 79, and CDRL3 of SEQ ID NO: 80, 81 or 82, and the second antibody-variable domain comprises CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73. Preferably, CDRL1 comprises one or two, preferably one substitution, and / or CDRL3 comprises one or two, preferably one substitution. Preferably, CDRH1 comprises one or two, preferably one substitution, and / or CDRH3 comprises one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0120] Therefore, according to a preferred embodiment, the antigen-binding molecule is the following combination of CDRL: (i) CDRL1 of sequence number 76, CDRL2 of sequence number 78, and CDRL3 of sequence number 80, (ii) CDRL1 of sequence number 76, CDRL2 of sequence number 78, and CDRL3 of sequence number 81, (iii) CDRL1 of sequence number 76, CDRL2 of sequence number 78, and CDRL3 of sequence number 82, (iv) CDRL1 of sequence number 76, CDRL2 of sequence number 79, and CDRL3 of sequence number 80, (v) CDRL1 of sequence number 76, CDRL2 of sequence number 79, and CDRL3 of sequence number 81, (vi) CDRL1 of sequence number 76, CDRL2 of sequence number 79, and CDRL3 of sequence number 82, (vii) CDRL1 of sequence number 77, CDRL2 of sequence number 78, and CDRL3 of sequence number 80, (viii) CDRL1 of sequence number 77, CDRL2 of sequence number 78, and CDRL3 of sequence number 81, (ix) CDRL1 of sequence number 77, CDRL2 of sequence number 78, and CDRL3 of sequence number 82, (x) CDRL1 of sequence number 77, CDRL2 of sequence number 79, and CDRL3 of sequence number 80, (xi) CDRL1 of sequence number 77, CDRL2 of sequence number 79, and CDRL3 of sequence number 81, (xii) CDRL1 of sequence number 77, CDRL2 of sequence number 79, and CDRL3 of sequence number 82 The antibody variable domains include CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73, in combination with an antibody variable domain containing one of the above, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one substitution, and / or CDRH3 contains one or two, preferably one substitution. In any case, the substitutions are preferably conservative amino acid substitutions.

[0121] According to another embodiment, the first antibody variable domain comprises one or more of the framework (FR) sequences of SEQ ID NO: 83, 84, or 85, or variants of FR sequences having at least 80%, at least 85%, at least 90%, or at least 95% sequence identity to the FR sequences of SEQ ID NO: 83, 84, or 85, and / or the second antibody variable domain comprises one or more of the FR sequences of SEQ ID NO: 74 or 75, or variants of FR sequences having at least 80%, at least 85%, at least 90%, or at least 95% sequence identity to the FR sequences of SEQ ID NO: 74 or 75, where FR is identified by Kabat definition, Chothia definition, IMGT definition, or contact definition. According to a preferred embodiment of the present invention, FR is identified by Kabat definition. The sequence identity of the FR sequences is preferably determined individually for each FR. This is because V H In this context, FR1, FR2, FR3, and FR4 each have at least 80%, at least 85%, at least 90%, or at least 95% sequence identity with the respective FR1, FR2, FR3, and FR4 sequences of sequence number 74 or 75. Similarly, V LIn this case, FR1, FR2, FR3, and FR4 each have at least 80%, at least 85%, at least 90%, or at least 95% sequence identity with the respective FR1, FR2, FR3, and FR4 sequences of sequence numbers 83, 84, or 85.

[0122] In yet another embodiment, the first variable domain comprises one of the amino acid sequences of SEQ ID NOs. 83, 84, or 85, or a variant thereof having at least 90%, preferably at least 92%, more preferably at least 95% or 98% sequence identity over the entire length, and maintaining the three CDLs as defined above, and the second variable domain comprises one of the amino acid sequences described in SEQ ID NOs. 74 or 75, or a variant having at least 90%, preferably at least 92%, more preferably at least 95% or 98% sequence identity over the entire length, and maintaining the three CDRHs as defined above. In a preferred embodiment, the antigen-binding protein comprises one of the following combinations: SEQ ID NOs. 74 and SEQ ID NOs. 84, SEQ ID NOs. 75 and SEQ ID NOs. 83, SEQ ID NOs. 75 and SEQ ID NOs. 85, SEQ ID NOs. 74 and SEQ ID NOs. 83, SEQ ID NOs. 75 and SEQ ID NOs. 84, or SEQ ID NOs. 74 and SEQ ID NOs. 85. Particularly preferred combinations of variable domains are SEQ ID NO: 74 and SEQ ID NO: 83 (A7), SEQ ID NO: 75 and SEQ ID NO: 84 (A02), or SEQ ID NO: 74 and SEQ ID NO: 85 (H10), or variants of each designated SEQ ID NO: preferably having at least 90%, preferably at least 92%, more preferably at least 95% or 98% sequence identity over the entire length and maintaining the respective CDRs as defined above. According to a particularly preferred embodiment of the present invention, the antigen-binding protein includes a combination of SEQ ID NO: 74 and SEQ ID NO: 83, or a combination of SEQ ID NO: 75 and SEQ ID NO: 84, or a combination of SEQ ID NO: 74 and SEQ ID NO: 85, or variants of each designated SEQ ID NO: preferably having at least 90%, preferably at least 92%, more preferably at least 95% or 98% sequence identity over the entire length and maintaining the respective CDRs as defined above.

[0123] In a particularly preferred embodiment, the antigen-binding protein comprises one of the following sequence combinations: a) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 (A7) of SEQ ID NOs. 76, 78, and 80; b) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 (A02) of SEQ ID NOs. 77, 79, and 81; or c) CDRH1, CDRH2, and CDRH3 of SEQ ID NOs. 71, 72, and 73, and CDRH1, CDRH2, and CDRH3 (H10) of SEQ ID NOs. 77, 79, and 82, where each CDR may contain one or two amino acid substitutions. Preferably, CDRL1 contains one or two, preferably one substitution, and / or CDRL3 contains one or two, preferably one substitution. Preferably, CDRH1 contains one or two, preferably one, substitutions, and / or CDRH3 contains one or two, preferably one, substitutions. In either case, the substitutions are preferably conservative amino acid substitutions.

[0124] The present invention also provides an antigen-binding protein, such as an antibody, that competes for binding to the antigen-binding protein of the present invention, which functions as a reference antigen-binding protein, preferably comprising the VL of SEQ ID NO: 83, 84, or 85 and the VH of SEQ ID NO: 74 or 75, or that binds to the same epitope as the antigen-binding protein of the present invention, which comprises the VL of SEQ ID NO: 83, 84, or 85 and the VH of SEQ ID NO: 74 or 75. Accordingly, according to one embodiment of the present invention, the antigen-binding protein competes for binding to the antigen-binding protein comprising the VL of SEQ ID NO: 83 and the VH of SEQ ID NO: 74, or binds to the same epitope as the antigen-binding protein comprising the VL of SEQ ID NO: 83 and the VH of SEQ ID NO: 74. According to a further embodiment of the present invention, the antigen-binding protein competes for binding to the antigen-binding protein comprising the VL of SEQ ID NO: 84 and the VH of SEQ ID NO: 74, or binds to the same epitope as the antigen-binding protein comprising the VL of SEQ ID NO: 84 and the VH of SEQ ID NO: 74. According to a further embodiment of the present invention, antigen-binding proteins compete for binding to antigen-binding proteins including VL of SEQ ID NO: 85 and VH of SEQ ID NO: 74, or bind to the same epitope as antigen-binding proteins including VL of SEQ ID NO: 85 and VH of SEQ ID NO: 74. According to a further embodiment of the present invention, antigen-binding proteins compete for binding to antigen-binding proteins including VL of SEQ ID NO: 83 and VH of SEQ ID NO: 75, or bind to the same epitope as antigen-binding proteins including VL of SEQ ID NO: 83 and VH of SEQ ID NO: 75. According to a further embodiment of the present invention, antigen-binding proteins compete for binding to antigen-binding proteins including VL of SEQ ID NO: 84 and VH of SEQ ID NO: 75, or bind to the same epitope as antigen-binding proteins including VL of SEQ ID NO: 84 and VH of SEQ ID NO: 75. According to a further embodiment of the present invention, antigen-binding proteins compete for binding to antigen-binding proteins including VL of SEQ ID NO: 85 and VH of SEQ ID NO: 75, or bind to the same epitope as antigen-binding proteins including VL of SEQ ID NO: 85 and VH of SEQ ID NO: 75.In each of the above embodiments, the antigen-binding protein of the present invention that functions as a reference is preferably in the scFv format. Appropriate competition assays are described in the examples. For example, when the antigen-binding protein of the present invention that functions as a reference is present at a concentration of 1 nM in a competition assay, the competing antibody has an IC of 100 nM or less, more preferably 50 nM, even more preferably 10 nM or less. 50 Preferably has.

[0125] According to a particularly preferred embodiment, the bispecific antigen-binding protein of the present invention is in the form of scDb, scDb-scFc, eIg or Fab-eIg.

[0126] According to an even more preferred embodiment, the bispecific antigen-binding protein of the present invention comprises the anti-EGFR antibody variable domain of the present invention and further the anti-CD3 antibody variable domain.

[0127] A particularly preferred embodiment of the present invention is a bispecific anti-EGFR and anti-CD antigen-binding protein comprising an amino acid sequence as described in SEQ ID NO: 19 (scDb), SEQ ID NO: 20 (scDb-scFv), combinations of SEQ ID NO: 9, 21, 23 and 24 (eIg), or combinations of SEQ ID NO: 9, 21, 22 and 24 (Fab-eIg).

[0128] An even more preferred embodiment of the present invention is a bispecific anti-EGFR antigen and anti-HER3 antigen-binding protein. Particularly preferred embodiments of such bispecific anti-EGFR and anti-HER3 antigen-binding proteins include: V H H10-V L 3-43-C H 1-Fc ホール And VH3-43-VLH10-CLk-Fc ノブ A bispecific and bivalent antibody comprising Light chain V L H10-C L λ (SEQ ID NO: 9), light chain V L 3-43-hetEHD2 (C247S, N275Q), heavy chain VH H10-C H 1-Fc ホール (Sequence ID 21), and heavy chain V H 3-43-hetEHD2(C337S)-Fc ノブ A bispecific and bivalent antibody containing Light Chain V L 3-43-hetEHD2(C247S, N275Q), light chain V L H10-C L λ (SEQ ID NO: 9), and heavy chain V H H10-C H 1-V H A bispecific and tetravalent antibody containing 3-43-hetEHD2(C337S)-Fc, Light Chain V H 3-43×V L H10-C L and heavy chain V H H10×V L hu3-43-C H 1-C H 2-C H A bispecific and tetravalent antibody containing 3, Light Chain V H 3-43×V L 3-43-C L λ, heavy chain V H 3-43×V L 3-43-C H 1-C H 2-C H 3 ホール , heavy chain V H H10-C H 1-C H 2-C H 3 ノブ (Sequence ID 123), and light chain V L H10-C L A bispecific and trivalent antibody containing λ (SEQ ID NO: 9), Light Chain V L H10-C L λ (SEQ ID NO: 9), heavy chain V H H10-C H 1-Fc ホール (Sequence ID 21) and heavy chain scFv3-43-Fc ノブ A bispecific and bivalent antibody containing These are some examples.

[0129] A more preferred embodiment of the present invention is a bispecific antigen-binding protein that binds EGFR and HER2 or MET. A particularly preferred embodiment of such a bispecific anti-EGFR and anti-HER2 or anti-MET antigen-binding protein is: Light Chain V L H10-C L λ (SEQ ID NO: 9), light chain V L 5D5-hetEHD2(C247S, N275Q)(Sequence ID 126), heavy chain V H H10-C H 1-Fc ノブ (Sequence ID 124) and heavy chain V H 5D5-hetEHD2(C337S)-Fc ホール (SEQ ID NO: 125) is a bispecific, bivalent anti-EGFR and anti-MET antibody. Light Chain V L H10-C L λ (SEQ ID NO: 9), light chain V L 4D5-hetEHD2(C247S, N275Q)(Sequence ID 138), heavy chain V H H10-C H 1-Fc ノブ (Sequence ID 124) and heavy chain V H 4D5-hetEHD2(C337S)-Fc ホール (SEQ ID NO: 127) is a bispecific, bivalent anti-EGFR and anti-HER2 antibody. These are some examples.

[0130] In view of the numerous examples presented in this application, those skilled in the art will readily recognize that the anti-EGFR antigen-binding proteins disclosed herein can be combined with other antigen-binding domains or proteins to form polyvalent and polyspecific antigen-binding proteins, such as bispecific antigen-binding proteins that are, for example, bivalent, trivalent, or tetravalent. However, those skilled in the art will readily recognize that the specificity and number of valencies are not particularly limited, and therefore the particular embodiments disclosed herein do not limit the invention in any way.

[0131] In a second embodiment, the present invention further provides nucleic acids (one or more) or one or more sets of nucleic acids (hereinafter referred to as nucleic acids (one or more)) encoding antigen-binding proteins according to the present invention. Nucleic acids can be degraded by endonucleases or exonucleases, particularly DNases and RNases, that may be found in cells. Therefore, it may be desirable to modify the nucleic acids of the present invention to stabilize them against degradation, thereby ensuring that high concentrations of nucleic acids are reliably maintained in cells over long periods. Typically, such stabilization can be achieved by introducing one or more internucleotide phosphorus groups, or by introducing one or more internucleotide nonphosphorus groups. Therefore, nucleic acids may consist of modifications to naturally occurring nucleotides and / or naturally occurring nucleotides, and / or modifications to the molecular backbone. Examples of modified internucleotide phosphate groups and / or non-phosphorus crosslinks in nucleic acids include, but are not limited to, methylphosphonates, phosphorothioates, phosphoramidates, phosphorodithioates, and / or phosphate esters. On the other hand, examples of non-phosphorus internucleotide analogs include, but are not limited to, siloxane crosslinks, carbonate crosslinks, carboxymethyl esters, acetamidoate crosslinks, and / or thioether crosslinks.Further examples of nucleotide modifications include, but are not limited to, phosphorylation of 5' or 3' nucleotides that enable ligation or prevention of exonuclease degradation / polymerase elongation, respectively; amino, thiol, alkyne, or biotinyl modifications for covalent and near-covalent bonds; fluorophores and quenchers; and modified bases such as deoxyinosine (dI), 5-bromodeoxyuridine (5-bromo-dU), deoxyuridine, 2-aminopurine, 2,6-diaminopurine, inverted dT, inverted dideoxy-T, dideoxycytidine (ddC), 5-methyldeoxycytidine (5-methyldC), locked nucleic acid (LNA), 5-nitroindole, iso-dC and -dG bases, 2'-O-methylRNA base, hydroxymethyldC, 5-hydroxybutyl (hydroxybutynl)-2'-deoxyuridine, 8-aza-7-deazaguanosine, and fluorine-modified bases. Therefore, nucleic acids can also be synthetic nucleic acids, including but not limited to polyamide or peptide nucleic acids (PNA), morpholino and locked nucleic acids (LNA), as well as glycol nucleic acids (GNA) and threose nucleic acids (TNA).

[0132] In a further aspect, the present invention provides a vector comprising the nucleic acid or nucleic acid set of the present invention. Vectors suitable for expressing nucleic acids in host cells are well known in the art.

[0133] The present invention also relates to a method for producing recombinant host cells expressing a binding molecule according to the present invention, the method comprising: (i) introducing a nucleic acid or nucleic acid set or vector as described above into a competent host cell in vitro or ex vivo; (ii) culturing the resulting recombinant host cell in vitro or ex vivo; and (iii) optionally selecting cells that express and / or secrete the binding molecule. Accordingly, the present invention further provides recombinant cells or host cells comprising nucleic acids, nucleic acids or vectors according to the present invention. The present invention further provides recombinant cells or host cells expressing antigen-binding proteins, nucleic acids or nucleic acid sets or vectors according to the present invention. Suitable cells in this context are Escherichia coli cells, insect cells, and mammalian cells. Examples of cells to be used as recombinant or host cells include, but are not limited to, prokaryotic cells (e.g., bacteria) and eukaryotic cells (e.g., yeast cells, mammalian cells, insect cells, plant cells, etc.). Specific examples include Escherichia coli, yeasts of the genera Kluyveromyces or Saccharomyces, mammalian cell lines (e.g., Vero cells, CHO cells, 3T3 cells, COS cells, etc.), and primary or established mammalian cell cultures (e.g., those produced from lymphoblasts, fibroblasts, germ cells, epithelial cells, nerve cells, adipocytes, etc.). Examples include mouse SP2 / 0-Ag14 cells (ATCC CRL1581), mouse P3X63-Ag8.653 cells (ATCC CRL1580), CHO cells lacking the dihydrofolate reductase gene (hereinafter referred to as the "DHFR gene") (Urlaub G et al; 1980), rat YB2 / 3HL.P2.G11.16Ag.20 cells (ATCC CRL1662, hereafter referred to as "YB2 / 0 cells"), and HEK293 cells. Preferred cells are HEK293-6E cells (Durocher et al., 2002, Nucl. Acids Res. 30(2)e9) or CHO cells.

[0134] In a further embodiment, the present invention provides a pharmaceutical composition comprising an antigen-binding molecule, nucleic acid(s), vector, or recombinant cell or host cell according to the present invention as an activator, and a pharmaceutically acceptable carrier and / or suitable excipient as appropriate. The pharmaceutical composition may further comprise one or more additional activators. The pharmaceutical composition is preferably selected from the group consisting of a solid, liquid, semi-solid, or transdermal therapeutic system.

[0135] In further embodiments, the present invention relates to antigen-binding molecules, nucleic acids (one or more), vectors, recombinant cells or host cells, or pharmaceutical compositions for use in pharmaceuticals.

[0136] The antigen-binding molecules, nucleic acids or nucleic acid sets, vectors, recombinant cells or host cells, or pharmaceutical compositions of the present invention are particularly suitable for use in the treatment of cancer. According to a preferred embodiment, the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

[0137] Cancer is preferably selected from the group consisting of lung cancer (NSCLC), prostate cancer, breast cancer, colon and rectal cancer, head and neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer (Thomas & Weihua, 2019, Front. Oncol. 9: 800).

[0138] In another embodiment, the present invention provides a method for treating cancer, the method comprising administering to a patient in need of treatment a therapeutically effective amount of the antigen-binding protein, nucleic acid(s), vector, recombinant cell or host cell, or pharmaceutical composition of the present invention, thereby treating the cancer. In a preferred embodiment, the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

[0139] According to the above, the present invention preferably relates to the following items:

[0140] Item 1. An antigen-binding protein comprising a first antibody variable domain and a second antibody variable domain, wherein the variable domains are a) 10 determined by biolayer interferometry (BLI) -5 It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more mutations among V441D, S464L, G465R, and S492R at a KD of less than molal concentration. b) Does not specifically bind to human EGFR having mutation Q435P and / or double mutation F436A / I462A. Antigen-binding proteins form binding sites, and their binding to human EGFR inhibits ligand-induced activation of human EGFR.

[0141] Item 2. An antigen-binding protein as described in Item 1, wherein the variable domain forms a binding site that specifically binds to domain III of EGFR (SEQ ID NO: 61).

[0142] Item 3. The antigen-binding protein according to item 1 or 2, which is selected from the group consisting of an antibody or an antigen-binding fragment thereof, and a chimeric antigen receptor (CAR), and preferably is selected from the group consisting of Fab, Fab’, (Fab’)2, Fv, disulfide-linked Fv, BsFv, dsFv, (dsFv)2, dsFv-dsFv’, scFv, scFv dimer, single-chain domain antibody, diabody, ds diabody, and bivalent domain antibody.

[0143] Item 4. The antigen-binding protein according to item 3, wherein the antibody is selected from the group consisting of a chimeric antibody, a humanized antibody, a bispecific antibody, and a multispecific antibody, and preferably the antibody is a human antibody.

[0144] Item 5. The first antibody variable domain contains the light chain complementarity-determining region quality (CDRL) 1 (CDRL1), CDRL2, and CDRL3 of the variable light chain region of SEQ ID NO: 83, 84, or 85, and / or the second variable domain contains CDRH1, CDRH2, and CDRH3 of the heavy chain variable region of SEQ ID NO: 74 or 75, the CDRs are identified by the Kabat definition, the Chothia definition, the IMGT definition, or the contact definition, and each CDR may contain one or two amino acid substitutions. The antigen-binding protein according to any one of items 1 to 4.

[0145] Item 6. (a) The first antibody variable domain contains CDRL1 of SEQ ID NO: 86, CDRL2 of SEQ ID NO: 87, and CDRL3 of SEQ ID NO: 88 and (b) The second antibody variable domain contains CDRH1 of SEQ ID NO: 71, CDRH2 of SEQ ID NO: 72, and CDRH3 of SEQ ID NO: 73 and each CDR may contain one or two amino acid substitutions. The antigen-binding protein according to any one of items 1 to 4.

[0146] Item 7. (a) The first antibody variable domain contains CDRL1 with sequence number 76 or 77, CDRL2 with sequence number 78 or 79, and CDRL3 with sequence numbers 80, 81, or 82 Includes, (b) The second antibody variable domain is CDRH1 of sequence number 71, CDRH2 of sequence number 72, and CDRH3, sequence number 73 An antigen-binding protein as described in any one of items 1 to 6, comprising, wherein each CDR may contain one or two amino acid substitutions.

[0147] Item 8. An antigen-binding protein according to any one of items 1 to 7, wherein the first antibody-variable domain comprises one or more of the framework (FR) sequence of SEQ ID NO: 83, 84, or 85, or variants of the FR sequence having at least 90% sequence identity to the FR sequence of SEQ ID NO: 83, 84, or 85, and / or the second antibody-variable domain comprises one or more of the FR sequence of SEQ ID NO: 74 or 75, or variants of the FR sequence having at least 90% sequence identity to the FR sequence of SEQ ID NO: 74 or 75, and the FR is identified by the Kabat definition, Chothia definition, IMGT definition, or contact definition, and the CDR may each contain one or two amino acid substitutions.

[0148] Item 9. The first variable domain contains one of the amino acid sequences according to SEQ ID NO: 83, 84, or 85, and the second variable domain contains one of the amino acid sequences according to SEQ ID NO: 74 or 75, preferably an antigen-binding protein in combination: a) Sequence IDs 74 and 83, b) Sequence IDs 75 and 84, or c) Sequence IDs 74 and 85 An antigen-binding protein as described in any one of items 1 to 8, comprising one of the following, wherein each CDR may contain one or two amino acid substitutions.

[0149] Item 10. The antigen-binding protein has the following sequence: a) CDRH1, CDRH2, and CDRH3 of sequence numbers 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of sequence numbers 76, 78, and 80, b) CDRH1, CDRH2, and CDRH3 of sequence numbers 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of sequence numbers 77, 79, and 81, or c) CDRH1, CDRH2, and CDRH3 of sequence numbers 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of sequence numbers 77, 79, and 82 An antigen-binding protein as described in any one of items 1 to 9, comprising, wherein each CDR may contain one or two amino acid substitutions.

[0150] Item 11. (i) V of sequence number 83, 84, or 85 L and V of sequence number 74 or 75 H It competes for binding to IgG1 antibodies, or (ii) V of sequence number 83, 84 or 85 L and V of sequence number 74 or 75 H The same epitope binding as IgG1 antibodies containing an antigen-binding protein listed in any one of items 1 through 10.

[0151] Item 12. Nucleic acids (one or more) encoding an antigen-binding protein according to any one of claims 1 to 11.

[0152] Item 13. A vector containing one or more nucleic acids as described in Item 12.

[0153] Item 14. Recombinant cells expressing an antigen-binding protein as described in any one of items 1 through 11, one or more nucleic acids as described in item 12, or a vector as described in item 13.

[0154] Item 15. A pharmaceutical composition comprising an antigen-binding protein as described in any one of Items 1 through 11, one or more nucleic acids as described in Item 12, a vector as described in Item 13, or recombinant cells as described in Item 14, and a pharmaceutically acceptable excipient.

[0155] Item 16. An antigen-binding protein as described in any one of Items 1 to 11, a nucleic acid(s) as described in Item 12, a vector as described in Item 13, a recombinant cell as described in Item 14, or a pharmaceutical composition as described in Item 15, for use in pharmaceuticals.

[0156] Item 17. An antigen-binding protein as described in any one of Items 1 through 11, a nucleic acid(s) as described in Item 12, a vector as described in Item 13, a recombinant cell as described in Item 14, or a pharmaceutical composition as described in Item 15, for use in the treatment of cancer.

[0157] Item 18. An antigen-binding protein, nucleic acid(s), vector, recombinant cell, or pharmaceutical composition as described in Item 17, wherein the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

[0158] Item 19. A method for treating cancer, comprising administering to a patient in need of treatment a therapeutically effective amount of an antigen-binding protein described in any one of Items 1 to 11, one or more nucleic acids described in Item 12, a vector described in Item 13, recombinant cells described in Item 14, or a pharmaceutical composition described in Item 15.

[0159] Item 20. The method according to Item 19, wherein the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, preferably head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck. [Examples]

[0160] [Example 1] Selection of EGFR antibodies by phage display For the selection of EGFR-specific human single-chain fragment variable (scFv) antibodies, the extracellular domain of EGFRvIII fused to human EGFR or human Fc was used as the antigen (huEGFR- / huEGFRvIII-huFc). Monoclonal scFv antibody selection (panning) was performed on 96-well microtiter plates (MTPs), and therefore on surface-immobilized EGFR- / EGFRvIII-huFc. Antibodies were selected from a library of human naive antibody genes packaged in Hyperphage, i.e., the HAL9 / 10 library (Kuegler et al., 2015), over three consecutive selection rounds.

[0161] In short, 2 μg / well of huEGFR-Fc or huEGFRvIII-huFc was immobilized overnight in 150 μL of PBS at 4°C in two separate wells of flat-bottomed MTP (High Binding Costar / Corning, New York, USA). Furthermore, four other wells were coated with IgG1-huFc protein under the same conditions, but with IgG1-huFc protein, for antibody phage library preclearance. Afterward, all coated wells were emptied and blocked at RT for 1 hour in 300 μL of 2% MPBS-T (1×PBS + 2% powdered milk in 0.1% Tween-20), followed by washing. Human naive antibody gene library input was added at a final concentration of 100 μL. 11The phage was mixed 1:1 in panning block solution (1% BSA, 1% MPBS-T) to a concentration of cfu / well. Removal of huFc-specific antibody phage particles was performed by incubation at RT for 1 hour in the preclearance wells. Subsequently, the preclearanced antibody phage library was incubated at RT for 2 hours in wells containing immobilized huEGFR-huFc or huEGFRvIII-huFc.

[0162] Furthermore, to eliminate huFc-specific antibody phages, 1 μg / well of huIgG1-Fc soluble protein was added for competition during library incubation on huEGFR-huFc or huEGFRvIII-huFc immobilized antigens. In various selection strategies, 1 μg / well of huEGFR-huFc or huEGFRvIII-huFc protein was used as competition instead of hIgG1-Fc alone to select EGFRvIII or EGFR-specific antibodies, respectively.

[0163] After completing the incubation, stringent plate bottom washing (1st and 2nd panning rounds: 20 times, 3rd panning round: 30 times) was performed to remove unbound antibody phages. From the remaining antigen-bound antibody phages, the corresponding phage particles were eluted with 150 μL of 10 μg / ml trypsin in PBS at 37°C for 30 minutes. The eluted phages were rescued by infection with E. coli OD600 0.5 at 37°C for 30 minutes, followed by incubation at 37°C and 650 rpm for 30 minutes. Subsequently, fresh selective medium was added (2×YT 1000 μL + 10×glucose / ampicillin 150 μL), and the infected E. coli were grown at 37°C and 650 rpm for 1 hour. To produce new antibody phage particles, M13K07 helper phage was added to the bacteria (1.5 × 10⁻¹² cfu / mL). Infection was repeated at 37°C for 30 minutes, followed by incubation at 37°C and 650 rpm for 30 minutes. The cells were then centrifuged at 3,220 × g for 10 minutes, and the culture medium was replaced with 2 × TY (tryptone-yeast) medium containing ampicillin and kanamycin for antibody-phage production (at 30°C and 650 rpm).

[0164] The following day, the bacterial culture was centrifuged (3,220 × g, 10 min) to separate the bacteria from the antibody-phage particles in the supernatant. The supernatant was collected and used for the second cycle of panning. Amplified antibody-phage particles (50 μL each) from each of the four selection strategies were separately mixed with the panning block to a final volume of 200 μL. This antibody-phage preparation was used as input for panning rounds 2 and 3 in place of the human naive antibody library. All other steps were carried out as described above, except that no eluted phages were amplified after the third panning round. Instead, E. coli infected with eluted phage particles were directly plated into 2 × YT-GA agar plates at dilutions 10E-4, 10E-5, and 10E-6 and incubated overnight at 37°C.

[0165] Antibody screening and sequence analysis A total of 92 single bacterial colonies, equivalent to as many antibody clones as possible, were collected from each of the four panning strategies and used to inoculate 150 μL of 2×YT-GA medium in a U-shaped MTP. The plates were incubated overnight at 37°C and 800 rpm.

[0166] Subsequently, 10 μL of 2×YT-GA medium was used to inoculate 180 μL of fresh U-shaped MTP plates. Wells H3 and H9 were left uninoculated (negative control), while wells H6 and H12 were inoculated with known antibody cultures (positive control). After incubation at 37°C and 800 rpm for 2 hours, the plates were centrifuged at 3,220×g for 10 minutes. The bacterial pellet was resuspended in 180 μL of fresh production medium supplemented with 50 μM IPTG to induce scFv expression, and incubated overnight at 30°C at 800 rpm. The following morning, the bacteria were pelletized again at 3,220×g for 10 minutes, and the supernatant was used for antigen binding analysis in ELISA.

[0167] For this purpose, 96-well ELISA plates were coated with 1 μg / mL of huEGFR-huFc, huEGFRvIII-huFc, or huIgG1-Fc in 100 μL / well. After washing and blocking the ELISA plates, soluble scFv from the crude bacterial supernatant mixed with MPBS-T was added and incubated at RT for 1 hour. Binding of scFv was detected using an anti-myc tag antibody (clone 9E10, produced by the University of Braunschweig) and an HRP conjugate secondary antibody (Sigma A0168). After quantifying and measuring the binding signal by TMB reaction, absorbance was read at reference wavelengths of 450 nm and 620 nm.

[0168] For the analysis, the signal-to-noise ratio (S / N) between the ELISA binding signals of the positive and negative antigens was calculated. Antibody clones exhibiting an S / N ratio greater than 3 and negligible binding to huIgG1-Fc were selected as potential lead candidates. A total of 94 clones met these selection criteria and were used to confirm the results obtained so far in the screening ELISA and subsequent sequence analysis. A total of 22 clones exhibited unique antibody sequences, and one was discarded due to the presence of an amber stop codon in the VH gene.

[0169] Antibody production and testing The remaining 21 scFv antibody genes were fused to the mouse IgG2a-Fc gene in mammalian cells to produce scFv-moFc fusion proteins, which were purified by protein A affinity chromatography for preliminary binding analysis on A549 cells in titration ELISA. In ELISA, no antibodies were found to be selective for EGFRvIII compared to EGFR, while four showed selectivity for the WT protein. Of the 21 antibodies tested, 15 showed specific binding to A549 cells.

[0170] [Example 2] Screening for neutralizing anti-EGFR antibodies Antibodies screened for scFV by phage display were cloned into expression vectors fused with the selected scFv to the human Fc portion. Plasmids were transiently transfected into HEK293 cells. Antibodies were purified from the supernatant by protein A affinity chromatography.

[0171] Inhibition of EGFR signaling was analyzed by Western blotting using the cancer cell lineage FaDu (Figure 1). Therefore, 200,000 cells per 6 wells were seeded in DMEM + 10% FCS + P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 μg / mL anti-EGFR scFv-huFc for 1 hour, followed by stimulation with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% Nu-PAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) or 0.05% (v / v) Tween-20 in PBS), and detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma-Aldrich-clone B5-1-2), and secondary HRP conjugate antibodies induced against mouse IgG (Sigma [A2554]) or rabbit IgG (Cell Signaling [CS#7074]).

[0172] In cells treated with anti-EGFR scFv-huFc MKU011-A7 and scFv-huFc MKU012-E11, a reduction in EGFR phosphorylation was observed compared to untreated controls, but the inhibition was not as efficient as that with cetuximab (Figure 1b). This inhibitory effect was stronger with MKU011-A7 than with MKU012-E11. Treatment of unstimulated cells with anti-EGFR scFv-huFc MKU011-A7 and MKU012-E11 did not induce EGFR phosphorylation or activation of the MAPK pathway, which was analyzed by ERK phosphorylation (Figure 1a). Any other tested anti-EGFR scFv-huFc did not inhibit EGFR phosphorylation.

[0173] In summary, anti-EGFR antibodies scFv-huFcMKU011-A7 and MKU012-E11 were identified as antagonistic antibodies that inhibit ligand-induced phosphorylation of EGFR in FaDu cells.

[0174] [Example 3] Affinity maturation of MKU011-A7 by light chain shuffling Library generation The VH gene of two EGFR-specific antibodies (MKU011-A7 (variable heavy and light chains according to SEQ ID NOs. 4 and 5) and MKU012-E11 (variable heavy and light chains according to SEQ ID NOs. 89 and 90)) was amplified by PCR and purified by gel extraction.

[0175] Next, the purified VH DNA was individually cloned into antibody-phage display vectors containing either a human kappa or lambda VL gene library isolated from human B cells. Insertion of the VH gene into the vector resulted in gene fusion between the VH gene, the peptide linker, and the VL gene, thus generating the scFv antibody gene.

[0176] Cloning products were pooled and desalted using an Amicon Ultra centrifugation column. Commercially available electrocompetent E. coli were then mixed with the desalted DNA, and the cells were transformed with phagemid DNA by electroporation. The transformed bacteria were harvested in SOC medium and propagated at 37°C and 650 rpm for 1 hour. Theoretical diversity of the scFv library was determined by serial dilution and colony counting.

[0177] Regarding the scFv lambda library, the total diversity is 3.0 x 10 8 The transformants were calculated. For the scFv Kappa library, the total diversity was 0.8 × 10⁻⁶. 8 A transformed product was obtained.

[0178] The remaining bacterial culture was used for antibody-phage production. For this purpose, the bacteria were grown again in selective culture medium until the OD600 reached 0.5. Subsequently, the bacteria were infected with the M13K07 helper phage. After another growth phase of the infected bacteria in selective medium (30 minutes at 37°C and 30 minutes at 37°C and 250 rpm), the culture medium was changed, and antibody-phage production was induced by overnight incubation at 25°C and 250 rpm.

[0179] Bacteria were separated from antibody-phage particles by centrifugation and purified by PEG precipitation. Briefly, a 20% (v / v) PEG / NaCl solution was added to the clarified culture supernatant and incubated on ice for 1 hour. After centrifugation at 4°C and 4,000×g for 1 hour, the supernatant was discarded and the antibody-phage pellet was resuspended in PBS. The resuspended antibody-phage was again centrifuged at 17,000×g for 10 minutes, and the soluble fraction was collected and stored at 4°C.

[0180] To determine the antibody-phage concentration, serial dilutions were first prepared and used to infect E. coli with an OD600 of 0.5. The infected cells were then cultured on selective medium, and the number of transformants was determined by colony counting.

[0181] For both scFv libraries, approximately 4 × 10 11 Antibody-phage concentrations in cfu / ml were calculated (approximately 3 × 10 in total). 11 cfu).

[0182] The display of scFv on the antibody-phage surface was confirmed by SDS-PAGE of 10E9 antibody-phage particles, followed by Western blotting and immunoblotting using anti-pIII antibody.

[0183] Antibody-phage selection To select VL-shuffled antibodies with improved binding properties, antibody-phage libraries were used for biopanning with antigen restriction and competition. Briefly, antibody-phage particles (approximately 4 × 10⁻¹⁰ cfu) were diluted in 2% BSA-PBST (approximately 1 mL) and incubated with magnetic Dynabeads® M-280 streptavidin (25 μL, ThermoScientific, #11206D) and mouse IgG (10 μg) at room temperature under rotation for 1 hour. The beads were separated from the supernatant by magnetic force, thereby separating the cross-reactive antibody-phages from the solution. After clarification of the library, biotinylated target antigens were added. As the target antigen, the extracellular domain (ECD) of the EGFR protein fused to mouse IgG2a Fc was used. In the first cycle of biopanning, the antigen at a final concentration of 20 nM was used and incubated with the clarified library at room temperature under rotation for 1 hour. Subsequently, 25 μl of magnetic Dynabeads® M-280 streptavidin was added, and the mixture was incubated at room temperature under rotation for 30 minutes. Antibody-phages bound to the biotinylated target antigen were separated from unbound antibody-phages in solution via Dynabeads using magnetic force. The supernatant was removed, and the beads were washed three times with 1 ml of 2% BSA-PBST. The beads were then incubated in 1 mL of 2% BSA-PBST for 15 minutes, followed by three repeated washes with 1 mL of PBST. Antibody-phages still bound to the antigen-bead complex were eluted by incubation with trypsin (10 μg / mL, 150 μL) at 37°C for 30 minutes.

[0184] The eluted antibody-phages were rescued by infecting OD600 0.5 E. coli (1 mL) at 37°C for 30 minutes, followed by incubation at 37°C and 500 rpm for 30 minutes. Then, fresh selective medium (approximately 4.5 ml) was added, and the infected E. coli were allowed to grow for 1 hour at 37°C and 500 rpm. To produce new antibody-phage particles, the M13K07 helper phage was added to the bacteria (MOI: approximately 1:20). Further infection was carried out at 37°C for 30 minutes, followed by incubation at 37°C and 500 rpm for 30 minutes.

[0185] Subsequently, the cells were centrifuged at 3,220 × g for 10 minutes, and the culture medium was replaced with 2 × TY medium containing ampicillin and kanamycin to induce antibody-phage production (overnight at 30°C and 500 rpm). The following day, the bacterial culture was centrifuged (3,220 × g, 10 minutes) to separate the bacteria from the antibody-phage particles in the supernatant. The supernatant was removed and used for a second cycle of biopanning under more stringent conditions (100 μL each).

[0186] In Cycle 2, two different selection strategies were implemented using antibody phages amplified from Cycle 1. In one strategy, the amount of biotinylated target antigen was reduced to 2 nM. In the other strategy, the amount of biotinylated target antigen was reduced to 0.2 nM. After incubation of the antibody-phage and biotinylated target antigen at room temperature under rotation for 1 hour, the unbiotinylated target antigen was added to the selection mixture as a soluble competitor (75 nM). After continuing incubation for a further 30 minutes, washing, elution, and rescue were performed as described above.

[0187] Antibody screening and sequence analysis Bacteria infected with antibody-phages eluted from the second selection round were transferred to selective medium and incubated at 37°C until single colonies were observed. Monoclonal colonies were collected and transferred to a 96-well plate containing selective culture medium (180 μL / well) and incubated overnight at 30°C and 300 rpm. The overnight culture (15 μL) was transferred to a new 96-well plate pre-filled with production medium (180 μL / well), and soluble scFv production was induced at 30°C and 300 rpm.

[0188] The binding characteristics of scFv-containing bacterial cultures were measured by ELISA. For this purpose, 384-well ELISA plates were coated with streptavidin (40 ng / well) and biotinylated target antigen (20 ng / well), streptavidin alone (40 ng / well), mouse IgG (20 ng / well), or 2% BSA solution. After washing and blocking the ELISA plates, scFv was added and incubated at room temperature for 1 hour. Binding of scFv was detected using a secondary antibody. The binding signal was quantified and measured by the TMB reaction, followed by absorbance readings at reference wavelengths of 450 nm and 620 nm.

[0189] For analysis, the signal-to-noise (S / N) ratio was calculated between the ELISA binding signals of the positive and negative antigens. Antibody clones were selected from the most stringent selection strategy (0.2 nM target antigen for positive selection), and those with very high ELISA S / N ratios (greater than 100) were selected as potential lead candidates. A total of 88 clones met these selection criteria and were used for DNA sequencing analysis. Thirteen clones revealed unique antibody sequences.

[0190] Antibody production and testing First, the 12 identified antibodies were produced as scFv-human Fc fusion proteins in mammalian cells and purified by protein A affinity chromatography. Following initial EGFR binding experiments using ELISA, two lead antibodies (YU250-A02, YU250-H10) derived from the parent antibody MKU011-A7 were identified, and subsequently, these were produced as human IgG1 antibodies.

[0191] [Example 4] Production of anti-EGFR IgG and Fab-His against MKU011-A7, YU250-A02, and YU250-H10 Fully human monospecific IgG1 (SEQ ID NOs. 4 and 10) and Fab-His (SEQ ID NOs. 4 and 5), containing the anti-EGFR antibody MKU011-A7 variable domain sequence, were cloned and expressed in suspension culture-adapted HEK293-6E cells. For this purpose, 1 × 10¹⁶ molecules were expressed in FreeStyle® F17 expression medium (Thermo Fisher, A1383501) supplemented with 4 mM GlutaMAX® (Thermo Fisher, 35050061) and 0.1% Kolliphor® P188 (Sigma, 15759-1KG). 6 HEK293-6E at a density of 1 / mL was transiently transfected using polyethyleneamine (PEI, 25kDa, linear - Polysciences). For 100mL of cells, 100μg of DNA and 200μg of PEI were mixed in 10mL of culture medium, incubated for 15 minutes, and then added to the cells. The cells were incubated at 37°C at 120rpm / 5%CO2 in a humidified incubator for 24 hours, after which 5mL of 20% (w / v) tryptone N1 (Organo Technie) dissolved in culture medium was added to 100mL of culture medium to initiate protein production. After a further 4 days of incubation at 37°C at 120rpm / 5%CO2 in a humidified incubator, the supernatant containing the protein was collected by centrifugation. The proteins were eluted using 100mM glycine (pH 3.5) as the elution buffer, with IgG eluted with Protein A, and Fab-His eluted with CaptureSelect® CH The proteins were purified by 1 XL (Thermo Fisher) affinity chromatography. The integrity of IgG after protein A purification was confirmed by size exclusion chromatography using a TSKgel SuperSW mAb HR column (Tosoh) (Figure 2a). SDS-PAGE analysis of the proteins showed single bands for Fab-His (50 kDa) and IgG (150 kDa) under non-reducing conditions. Under reducing conditions, the expected chains were confirmed for Fab-His (two 25 kDa chains) and IgG (a 50 kDa heavy chain and a 25 kDa light chain) (Figure 2b).

[0192] One molecule of fully human IgG (IgG YU250-A02 and YU250-H10) containing the YU250-A02 variable domain sequence (SEQ ID NOs. 6 and 7) or the YU250-H10 variable domain sequence (SEQ ID NOs. 9 and 10) was cloned and expressed in suspension-culture-adapted HEK293-6E cells as described above. The proteins were purified from the supernatant of transiently transfected cells by protein A affinity chromatography using 100 mM glycine (pH 3.5) as the elution buffer. Size exclusion chromatography after protein A purification using a TSKgel SuperSW mAb HR column (Tosoh) confirmed IgG YU250-A02 and YU250-H10 with purities of 94.4% and 93.1%, respectively (Figures 3a and 3d). No aggregates were detected after preparative SEC-FPLC (Figures 3b+e). SDS-PAGE analysis of FPLC-purified IgG YU250-A02 and YU250-H10 showed a single band (approximately 150 kDa) under non-reducing conditions, and two bands corresponding to the heavy chain (50 kDa) and light chain (25 kDa) under reducing conditions (Figures 3c and 3f).

[0193] Fully human Fab-His, containing the variable domain sequences of the anti-EGFR antibody YU250-A02 (SEQ ID NOs. 6 and 8) and the YU250-H10 (SEQ ID NOs. 5 and 9), was cloned and expressed in suspension culture-adapted HEK293-6E cells as described above (fullly human Fab-His containing the anti-EGFR antibody MKU011-A7 with SEQ ID NOs. 4 and 5). The protein was eluted using CaptureSelect® C with 100 mM glycine (pH 3.5) as the elution buffer. H Purified from the supernatant by 1 XL (Thermo Fisher) affinity chromatography. Size exclusion chromatography using a TSKgel SuperSW mAb HR column (Tosoh) was performed, followed by CaptureSelect® C H The integrity of Fab-His after 1 XL purification was confirmed (Figures 4a and 4c). SDS-PAGE analysis of the protein showed a single 50 kDa band under non-reducing conditions and two 25 kDa bands under reducing conditions (Figures 4b and 4d).

[0194] In summary, the anti-EGFR antibodies MKU011-A7, YU250-A02, and YU250-H10 are each compatible with Protein A or CaptureSelect® C, respectively. H IgG and Fab-His can be produced with a purity of over 93.1% after one-step purification using either XL (Thermo Fisher) affinity chromatography. Compared to their parent antibody MKU011-A7, V L Yields were increased for the shuffled mutants YU250-A02 and YU250-H10 (Table 1).

[0195] [Table 4]

[0196] [Example 5] V of YU250-H10 LDeletion of the glycosylation site in increases the productivity of IgG YU250-H10. IgG YU250-H10 V L By deleting the potential glycosylation site in the mutant V through site-directed mutagenesis, L N126K was generated. Parental antibody (wt) (SEQ ID NOs. 9 and 58) and mutant (V L N126K) (SEQ ID NOs. 57 and 58) was expressed in suspension-culture-adapted HEK293-6E cells as described above. The protein was purified from the supernatant of transiently transfected cells by protein A affinity chromatography using 100 mM glycine (pH 3.5) as elution buffer. The integrity of both proteins was confirmed by size exclusion chromatography after protein A purification using a TSKgel SuperSW mAb HR column (Tosoh) (Figures 5a and 5b). Deletion of potential glycosylation sites resulted in an approximately 50% increase in protein yield (16.8 mg / L / V compared to wild-type). L (26.3 mg / L for N126K). The integrity of both proteins was analyzed by SDS-PAGE (Figure 5c), and under non-reducing conditions, a band corresponding to approximately 150 kDa was shown. Under reducing conditions, both the heavy chain (50 kDa) and light chain (25 kDa) were observed. Due to the deletion of the glycosylation site, V LA smaller apparent molecular weight was observed for the N1216K mutant. Anti-EGFR IgG binding was analyzed by ELISA using immobilized EGFR-His fusion protein (SEQ ID NO: 53) containing the extracellular domains of human EGFR (aa25-645) (Figure 5d). The EGFR-His fusion protein was coated on a polystyrene microtiter plate at 3 μg / mL in PBS. The remaining region was blocked with PBS and 2% skim milk (MPBS). The plate was incubated with serial dilutions of the antibody in MPBS. Binding antibodies were detected with HRP-conjugated anti-huFc antibody (1:5,000) diluted in MPBS. Diluted 1% TMB and 0.06% H2O2 in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates. IgG YU250-H10 V L When the glycosylation site was deleted, no difference in binding to EGFR was observed (EC50 0.49±0.09 nM / V compared to the wild type). L For N126K, the value is 0.53 ± 0.05 nM.

[0197] [Example 6] The anti-EGFR antibody YU250-H10 shows increased binding to EGFR compared to MKU011-A7. Monovalent binding of Fab-His to anti-EGFR antibodies MKU011-A7 and YU250-H10 (sequence as defined in Example 4) (Figure 6a) and bivalent binding of IgG1 (Figure 6b) was analyzed by ELISA using immobilized EGFR-moIgG2a-Fc fusion protein (SEQ ID NO: 33) containing the extracellular domain of human EGFR (aa25-645). The EGFR-moIgG2a-Fc fusion protein was coated on polystyrene microtiter plates at 3 μg / mL in PBS. The remaining region was blocked with PBS and 2% skim milk (MPBS). The plates were incubated with a series of serial dilutions of the antibody in MPBS. The bound antibody was detected with HRP-conjugated anti-huFab antibody (1:20,000) diluted in MPBS. For detection, 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used (n=3).

[0198] Compared to the anti-EGFR antibody MKU011-A7, YU250-H10 showed a 4.7-fold increase in binding to monovalent Fab-His (EC2). 50 (2.79±1.19nM vs. 0.59±0.19nM). No difference was observed for divalent bonds (EC). 50 (0.32±0.08 nM vs. 0.27±0.10 nM) (Table 2).

[0199] [Table 5]

[0200] [Example 7] IgG YU250-A02 and YU250-H10 bind as efficiently as cetuximab to cells analyzed by EGFR in ELISA and flow cytometry. Anti-EGFR IgG binding was analyzed by ELISA using immobilized EGFR-His fusion protein (SEQ ID NO: 53) containing the extracellular domains of human EGFR (aa25-645) (Figure 7a). The EGFR-His fusion protein was coated on a polystyrene microtiter plate at 3 μg / mL in PBS. The remaining region was blocked with PBS and 2% skim milk (MPBS). The plate was incubated with serial dilutions of the antibody in MPBS. The bound antibody was detected with HRP-conjugated anti-huFab antibody (1:5,000) diluted in MPBS. 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates.

[0201] The binding of anti-EGFR IgG to cells was analyzed by flow cytometry for DiFi (Figure 7b) and FaDu (Figure 7c). Therefore, 100,000 cells were harvested and incubated for 1 hour in a U-bottom 96-well plate with serial dilutions of the antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (cell collection by centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA), the cells were incubated for 1 hour with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgGMKU011-A7, YU250-A02, and YU250-H10 (sequences as defined in Example 4) were found to have EC similar effects to cetuximab, as observed in ELISA against EGFR-His and in human cancer cell lines DiFi and FaDu, as analyzed by flow cytometry. 50 The data was joined by value (Table 3).

[0202] [Table 6]

[0203] [Example 8] The anti-EGFR antibodies YU250-A02 and YU250-H10 are cross-reactive against EGFR in cynomolgus monkeys. The binding of anti-EGFR IgG YU250-A02 and YU250-H10 (sequences as defined in Example 4) to cynomolgus monkey EGFR was analyzed at 10 nM compared to cetuximab (Figures 8a and 8b), and concentration-dependent binding of anti-EGFR IgG YU250-A02 and YU250-H10 was analyzed (Figures 8c and 8d). Binding was analyzed by ELISA using an immobilized EGFR-His fusion protein (Sino Biological[SIN-90285-C08H-50]) (SEQ ID NO: 52) containing the extracellular domain (aa1-645) of cynomolgus monkey EGFR. The fusion protein was coated on a polystyrene microtiter plate at 3 μg / mL in PBS. The remaining region was blocked with PBS and 2% skim milk (MPBS). Plates were incubated with 10 nM anti-EGFR IgG in MPBS (Figures 8a and 8b) or serial dilutions of anti-EGFR IgG (Figures 8c and d). Binding IgG was detected with HRP-conjugated anti-huFc antibody (1:5,000) diluted in MPBS. 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates. Anti-EGFR IgG YU250-A02 and YU250-H10 were cross-reactive to cynomolgus monkey EGFR and exhibited similar EC reactions to both human and cynomolgus monkey EGFR. 50 The values ​​are shown (Table 4).

[0204] [Table 7]

[0205] [Example 9] Affinity-mature IgG YU250-A02 and YU250-H10 are more potent than IgGMKU011-A7 in inhibiting EGFR signaling and cell proliferation. Inhibition of EGFR signaling was analyzed by Western blotting using cancer cell lines DiFi and FaDu (Figures 9a and 9b). Therefore, 200,000 cells per 6-well plate were seeded in RPMI + 10% FCS + P / S (DiFi) or DMEM + 10% FCS + P / S (FaDu), incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 1 hour, followed by stimulation with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% Nu-PAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) in PBS, 0.05% (v / v) Tween-20), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma-clone B5-1-2), as well as secondary HRP conjugate antibodies induced against mouse IgG (Sigma [A2554]) or rabbit IgG (Cell Signaling [CS#7074]).In both cell lines treated with the anti-EGFR IgG antibody MKU011-A7, YU250-A02 and YU250-H10 inhibited EGFR phosphorylation at position Y1068 compared to untreated cells. IgG YU250-A02 and YU250-H10 inhibited EGFR phosphorylation at position Y1068 more efficiently than MKU011-A7 for both EGF-stimulated DiFi and FaDu. This inhibition was comparable to that observed with cetuximab treatment.

[0206] Proliferation inhibition was analyzed for the cancer cell lineages DiFi and FaDu (Figures 9c and 9d). Therefore, 2,000 cells per 96-well plate were seeded in RPMI + 10% FCS + P / S (DiFi) or DMEM + 10% FCS + P / S (FaDu), incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 5 days until cell count quantification using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). Anti-EGFR IgG MKU011-A7, YU250-A02, and YU250-H10 (sequences as defined in Example 4) resulted in statistically significant inhibition of proliferation for serum-starved DiFi compared to untreated cells. Anti-EGFR IgG MKU011-A7, YU250-A02, and YU250-H10 inhibited serum-starved DiFi proliferation as efficiently as cetuximab. Anti-EGFR IgG MKU011-A7, YU250-A02, and YU250-H10 resulted in statistically significant inhibition of serum-starved FaDu proliferation compared to untreated cells. Anti-EGFR IgG YU250-A02 and YU250-H10 resulted in statistically significant and potent inhibition of serum-starved FaDu proliferation compared to cells treated with MKU011-A7.

[0207] In summary, IgG YU250-A02 and YU250-H10 showed comparable inhibitory effects to cetuximab in inhibiting EGFR phosphorylation and proliferation of FaDu and DiFi. These inhibitory effects were superior to those of IgGMKU011-A7 and were comparable to those of cetuximab.

[0208] [Example 10] Anti-EGFR IgG YU250-A02 and YU250-H10 inhibit EGFR dimerization as efficiently as cetuximab in β-galactosidase complementation assays. Stable mouse fibroblast cell lines were generated using a retroviral MSCV-based vector system, and EGFR dimerization via β-galactosidase complementation was studied. For this purpose, i) the extracellular and transmembrane domains (aa1-679) of human EGFR were fused to the α-peptide of β-galactosidase (SEQ ID NO: 59) using a linker (AAAGSGGGGS) and cloned into a pMSCV-LTR-puromycin vector, and ii) the extracellular and transmembrane domains (aa1-679) of human EGFR were fused to the α-acceptor (Δ1-31) of β-galactosidase using a linker (AAAGSGGGGS) and cloned into a pMSCV-LTR-hygromycin B vector (SEQ ID NO: 60). Transduced NIH-3T3 cells were selected using 1 μg / mL puromycin (Sigma [P8833-25MG]) and 200 μg / mL hygromycin B (Thermo Fisher [10687010]) in DMEM + 10% FCS + P / S medium. To monitor EGFR dimerization, 20,000 stable cells per 96-well plate were seeded in DMEM + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with serial dilutions (Figure 10a) or 10 nM (Figure 10b) IgG for 15 minutes, and then stimulated with 30 ng / mL EGF (Biotechne [236-EG-200]) for 1 hour. After washing the cells once with PBS, luminescence was measured according to the instructions for the Galacto-Light Plus® β-Galactosidase Reporter Gene Assay System (Thermo Fisher [T1011]) (TECAN Spark®). Anti-EGFR IgG YU250-A02 and YU250-H10 (sequences as described in Example 4) resulted in statistically significant inhibition of EGF-mediated EGFR dimerization in β-galactosidase-complementarily mediated cell lines, similar to cetuximab, while trastuzumab, included as a negative control, had no effect.

[0209] [Example 11] Anti-EGFR IgG YU250-H10 efficiently translocates to FaDu cells. Anti-EGFR IgG YU250-H10 (sequence as described in Example 4) and cetuximab were labeled with pHrodo® Red (succinimide ester) according to instructions (Thermo Fisher [P36600]) and antibody internal migration was analyzed. For this purpose, 20,000 FaDu cells per 96 wells were seeded in DMEM + 10% FCS + P / S, incubated for 24 hours, and then treated with 50 nM pHrodo® Red-labeled anti-EGFR IgG. Cells incubated at either 37°C (Figure 11a) or 4°C (Figure 11b) after 5 minutes, 1 hour, 2 hours, 6 hours, 24 hours, and 48 hours were harvested by trypsin treatment, resuspended in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA), and fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Labeled anti-EGFR IgG YU250-H10 was rapidly translocated to FaDu, similar to cetuximab, while only minimal antibody uptake was observed at 4°C.

[0210] [Example 12] The anti-EGFR antibodies YU250-A02 and YU250-H10 compete with cetuximab, matsuzumab, necitumumab, and panitumab for binding to EGFR. Competition for binding to EGFR with cetuximab was analyzed by ELISA using an immobilized EGFR-moIgG2a-Fc fusion protein (SEQ ID NO: 33) containing the extracellular domains of human EGFR (aa25-645). The EGFR-moIgG2a-Fc fusion protein was coated onto polystyrene microtiter plates at 3 μg / mL in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). The plates were incubated with serial dilutions of cetuximab in MPBS. After washing, 1 nM scFv-Fc YU250-A02 (SEQ ID NO: 92) or scFv-Fc YU250-H10 (SEQ ID NO: 93) added to a dilution of cetuximab in MPBS was incubated. After washing, conjugated scFv-huFc was detected against the variable domain using rabbit serum (Stork et al., JBC 2008, 12: 7804-7812-1:1,000 dilution), followed by incubation with HRP-conjugated anti-rabbit Fc antibody (Abcam-1:5,000) diluted in MPBS. Conjugated cetuximab was detected in separate wells by incubation with HRP-conjugated anti-human Fc antibody (1:5,000) diluted in MPBS. 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates. The binding of both IgG YU250-A02 and IgG YU250-H10 to EGFR in ELISA was efficiently inhibited by cetuximab in a concentration-dependent manner, as described below and shown in Figure 12a.

[0211] IC, which is determined in competitive assays. 50 value: scFv-huFc YU250-A02:1.29±0.26nM scFv-huFc YU250-H10:1.20±0.17nM.

[0212] Furthermore, competition for the binding of Fab-His YU250-A02 or Fab-His YU250-H10 (as defined in Example 4) to EGFR by cetuximab, IgG GC1118 (SEQ ID NOs. 11 and 12), matsuzumab (SEQ ID NOs. 13 and 14), nesitumumab (SEQ ID NOs. 15 and 16), and panitumumab (SEQ ID NOs. 17 and 18) was analyzed by ELISA using immobilized EGFR-moIgG2a-Fc fusion protein (SEQ ID NOs. 33) containing the extracellular domain (aa25-645) of human EGFR. The EGFR-moIgG2a-Fc fusion protein was coated on polystyrene microtiter plates at 3 μg / mL in PBS. The remaining regions were blocked with PBS and 2% skim milk (MPBS). Plates were incubated with 200 nM Fab-His YU250-A02 or Fab-His YU250-H10 in MPBS. After washing, 1 nM IgG YU250-A02, IgG YU250-H10 (sequence as defined in Example 4), cetuximab, GC1118, matsuzumab, necitumumab, or panitumumab were incubated with 200 nM Fab-His YU250-A02 or Fab-His YU250-H10 in MPBS. After washing, conjugated IgG was detected by incubation with HRP-conjugate anti-human Fc antibody (1:5,000) diluted in MPBS. As a control, conjugated Fab-His was detected with HRP-conjugate anti-His antibody (1:1,000) diluted in MPBS. As detection substrates, 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used. Fab-His YU250-A02 and YU250-H10 efficiently blocked the binding of matsuzumab, nesitumumab, and panitumumab, while only partial inhibition was observed for cetuximab. IgG GC1118 was able to bind regardless of inhibition by Fab-His YU250-A02 or Fab-His YU250-H10 (Figure 12b, c).

[0213] In conclusion, anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10 bind to epitopes identical to or overlapping with those of cetuximab, matsuzumab, necitumumab, and panitumumab, while anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10 have different epitopes from GC1118.

[0214] [Example 13] The anti-EGFR antibodies YU250-A02 and YU250-H10 utilize unique interactions with EGFR compared to cetuximab, GC1118, matsuzumab, necitumumab, and panitumab. Site-directed mutagenesis was used to introduce mutations into domain III of EGFR-moIgG2a-Fc fusion proteins (SEQ ID NOs: 34, 35, 36, 37, 38) containing the extracellular domains of human EGFR (aa25-645). The mutated antigens were produced by transient transfection of HEK293-6E cells as described above, purified from the supernatant by protein A affinity chromatography (eluted at 100 mM glycine buffer pH 3.5), and dialyzed against PBS. The EGFR-moIgG2a-Fc fusion proteins were coated onto polystyrene microtiter plates at 3 μg / mL in PBS. The remaining regions were blocked with PBS and 2% skim milk (MPBS). The plates were incubated with a series of serial dilutions of anti-EGFR IgG in MPBS. Binding IgG was detected with HRP-conjugated anti-human Fc antibody (1:5,000) diluted in MPBS. For detection, 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used.

[0215] All antibodies are similar to EGFR wt, F436A, and I462A. 50Binding was observed at a value. For YU250-A02, slightly reduced binding to G434A was observed, while all other antibodies bound to this mutant. For YU250-A02 and YU250-H10, no binding to Q435P was observed, while all other antibodies showed binding to varying degrees. Furthermore, binding to the double mutant F436A / I462A was reduced for IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4), reduced for nesitumumab, and unaffected or only slightly affected for cetuximab, GC1118, matsuzumab, and panitumumab (Figure 13).

[0216] Therefore, the binding of anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10 to EGFR can be distinguished from the binding of cetuximab, GC118, matsuzumab, necitumumab, and panitumumab by the mutations Q435P and F436A / I462A, which completely suppress the binding of YU250-A02 and YU250-H10, respectively. Binding to these mutants was observed with all other antibodies, but was reduced for necitumumab.

[0217] Furthermore, mutations at positions F381, H383, F381 / H383, S484, and K487 (SEQ ID NOs. 39, 40, 41, 42, and 43) were studied using alanine substitution (Figure 14). Notably, the double mutation F381A / H383A completely inhibited GC118 binding, while no effect was observed on IgG YU250-A02, IgG YU250-H10, and cetuximab. In addition, mutations S484A and K487A inhibited matsuzumab binding, while no effect was observed on IgG YU250-A02, IgG YU250-H10, and cetuximab.

[0218] [Example 14] The anti-EGFR antibodies YU250-A02 and YU250-H10 bind to clinically relevant EGFR mutations. In further experiments, the inventors studied the binding of IgG YU250-A02, IgG YU250-H10 (with sequences as described in Example 4) and other EGFR antibodies to clinically observed EGFR mutants (SEQ ID NOs. 44, 45, 46, 47, 48, 49, 50), including V441D, V441G, S442R, S464L, G465R, K489E, and S492R (Strickler et al., Cancer Discovery 2018) (Figure 15). IgG YU250-A02 and IgG YU250-H10 were able to bind to V441D, V441G, S442R, S464L, G465R, K489E, and S492R. Therefore, anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10 are capable of binding to seven clinically relevant EGFR mutants. Notably, cetuximab did not show binding to V441D, S464L, G465R, and S492R. Necitumumab did not show binding to V441D and S442R, but showed reduced binding to V441G, G465R, and K489E. Panitumumab did not show binding to S442R, S464L, and G465R, but showed slightly reduced binding to K489E. Therefore, anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10 exhibit distinct binding patterns compared to cetuximab, nesitumumab, and panitumumab, and are capable of binding to several clinically relevant EGFR mutants.

[0219] The binding properties demonstrated in Examples 13 and 14 are summarized in Figure 16, further highlighting the specific and unique binding characteristics of anti-EGFR antibodies containing the variable domains of YU250-A02 or YU250-H10.

[0220] [Example 15] Affinity measurement of anti-EGFR Fab fragments binding to EGFR using biolayer interferometry. The affinity of monovalent anti-EGFR Fab is demonstrated by Octet® QK. e The determination was made by biolayer interferometry using the ForteBio system. For this purpose, His-tagged anti-EGFR Fabs were used as antigens, containing the anti-EGFR binding moiety MKU011-A7 (SEQ ID NOs. 4 and 5), YU250-A02 (SEQ ID NOs. 6 and 8), YU250-H10 (SEQ ID NOs. 5 and 9), or cetuximab (SEQ ID NOs. 32 and 62) and the His-tagged extracellular domain of EGFR wt, or the mutant Q435P, F436A / I462A, V441D, S464L, G465R, or S492R. A FAB2G sensor was used for measurement and activated by incubation for 10 minutes or more in RT in PBS supplemented with 2% BSA and 0.05% Tween20. Baseline measurements were performed in PBS supplemented with 2% BSA and 0.05% Tween20 for 120 seconds at 1,000 rpm RT. Then, Fab antibody was loaded onto the sensor by incubation for 180 seconds at RT in a 2.5 μg / mL solution diluted in PBS supplemented with 2% BSA and 0.05% Tween20. Baseline measurements were then performed again in PBS supplemented with 2% BSA and 0.05% Tween20 for 120 seconds at 1,000 rpm RT. Antigen binding was measured using 500nM, 158nM, 50nM, 15.8nM, 5nM, 1.58nM, 0.5nM, and 0nM solutions diluted in PBS supplemented with 2% BSA and 0.05% Tween20. Antigen dissociation was then measured by incubation for 300 seconds at 1,000 rpm in PBS supplemented with 2% BSA and 0.05% Tween20. Analysis was performed using Data Analysis HT software (version 11.1.2.48). In short, insignificant binding traces with responses less than 0.05 nm were excluded from the analysis. Furthermore, approximately 10-fold K was excluded. D Only traces with the specified antigen concentration were used to calculate the 1:1 binding model.

[0221] [Table 8]

[0222] All Fabs are 10 -7 Less than M K D It bound to EGFR wt. For Fab MKU011-A7, YU250-A02 and YU250-H10, it bound to antigens containing mutation Q435P or double mutation F436A / I462A (10 -5 K that exceeds M D ) was not detected. For these two variants, Fab cetuximab was 10 -5 Less than M K D The antigens of Fab MKU011-A7, YU250-A02, and YU250-H10, including mutants V441D, S464L, G465R, or S492R, were bound. -5 Less than M K D Binding at these four variants. Fab cetuximab did not bind to these four variants (10 -5 K that exceeds M D Figure 18 shows an overview of the locations of the mutations examined in EGFR domain III.

[0223] These data demonstrate that antibodies containing anti-EGFR binding sites of MKU011-A7, YU250-A02, or YU250-H10 bind to EGFR variants with cetuximab escape mutations V441D, S464L, G465R, or S492R. Binding to Q435P or F436A / I462A EGFR variants was not detected for antibodies containing anti-EGFR binding sites of MKU011-A7, YU250-A02, or YU250-H10, indicating that the epitopes of these antibodies include Q435, F436, and I462.

[0224] [Example 16] The anti-EGFR antibodies YU250-A02 and YU250-H10 cross-react to mouse EGFR. The binding of anti-EGFR antibodies IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4), as well as various other EGFR antibodies including cetuximab, GC1118, matsuzumab, nesitumumab, and panitumumab, to human and mouse EGFR was analyzed by ELISA using immobilized EGFR-His fusion protein (SEQ ID NO: 53) containing the extracellular domain of human EGFR (aa25-645) or immobilized EGFR-His fusion protein (SEQ ID NO: 51) containing the extracellular domain of mouse EGFR (aa25-647-Sino Biological[51091-M08H]) (Figure 19). Antigens were coated on polystyrene microtiter plates at 2 μg / mL in PBS. The remaining areas were blocked with PBS and 2% skim milk (MPBS). Next, the plates were incubated with a series of serial dilutions of 10 nM anti-EGFR IgG in MPBS (Figure 19a, b) or anti-EGFR IgG antibodies YU250-A02 and YU250-H10 (Figure 19c, d). Binding IgG was detected with HRP-conjugated anti-huFc antibody (1:5,000) diluted in MPBS. 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates. Anti-EGFR antibodies IgG YU250-A02 and IgG YU250-H10 bound to human EGFR and exhibited similar EC2 binding. 50 The antibodies showed cross-reactivity with mouse EGFR at 10 nM (Table 6). Only GC1118 and, to some extent, nesitumumab showed cross-reactivity with mouse EGFR at 10 nM. All other antibodies (cetuximab, matsuzumab, panitumumab) did not show cross-reactivity with mouse EGFR at 10 nM.

[0225] [Table 9]

[0226] [Example 17] Anti-EGFR IgG YU250-H10 is more potent than cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab in inhibiting the proliferation of human cancer cell lineage A431. The binding of anti-EGFR IgG to cells was analyzed by flow cytometry in A431 cells (Figure 20a). For this purpose, 100,000 cells were collected and incubated for 1 hour in a U-bottom 96-well plate with a series of serial dilutions of the antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA to collect the cells), the cells were incubated for 1 hour with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgG YU250-A02 and YU250-H10 bound to human cancer cell lineage A431 with similar efficacy to cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab (EC 50 The difference in values ​​is less than 2x (Table 7).

[0227] Proliferation inhibition was analyzed for A431 cells (Figure 20b). 2,000 cells per 96-well plate were seeded in RPMI + 10% FCS + P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 5 days until cell count quantification using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). Anti-EGFR IgG YU250-H10 resulted in a statistically significant inhibition of proliferation in serum-starved A431 cells at 10 nM compared to untreated cells.

[0228] Inhibition of EGFR signaling was analyzed in A431 cells by Western blotting (Figure 20c). 200,000 cells per 6 wells were seeded in RPMI + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, followed by stimulation with 50 ng / mL EGF (biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]). Equivalent volumes were loaded onto 4-12% NuPAGE® gels (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4) resulted in reduced EGFR phosphorylation at position Y1068 in serum-starved A431 cells stimulated with 50 ng / mL EGF at 100 nM, compared to untreated cells. Treatment with anti-EGFR IgG YU250-H10 alone resulted in inhibition of the MAPK pathway in serum-starved A431 cells stimulated with 50 ng / mL EGF at 100 nM, compared to untreated cells.

[0229] [Table 10]

[0230] [Example 18] Anti-EGFR IgG YU250-A02 and YU250-H10 exhibit potent inhibitory activity against the proliferation of the human cancer cell lineage DiFi compared to GC1118. The binding of anti-EGFR IgG to cells was analyzed for DiFi by flow cytometry (Figure 21a). 100,000 cells were collected and incubated for 1 hour in a U-bottom 96-well plate with a series of serial dilutions of the antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA), the cells were incubated with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]) for 1 hour. After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgG YU250-A02 and YU250-H10 bound to the human cancer cell lineage DiFi, as well as cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab, as analyzed by flow cytometry (EC 50 The difference in values ​​is less than 2x (Table 8).

[0231] Proliferation inhibition was analyzed for DiFi (Figure 21b). 2,000 cells per 96 wells were seeded in RPMI + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of cell starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 5 days until cell count quantification using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). In contrast to anti-EGFR IgG GC1118, anti-EGFR IgG YU250-A02 and YU250-H10 resulted in statistically significant inhibition of proliferation for serum-starved DiFi at 10 nM compared to untreated cells.

[0232] Inhibition of EGFR signaling was analyzed for DiFi by Western blotting (Figure 21c). 200,000 cells per 6 wells were seeded in RPMI + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, followed by stimulation with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) or 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma [05-829]), as well as secondary antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).In contrast to anti-EGFR IgG GC1118, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4) resulted in MAPK pathway inhibition in serum-starved and unstimulated DiFi cells at 100 nM, comparable to the inhibitory effects of cetuximab, matsuzumab, necitumumab, and panitumab. In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in reduced EGFR phosphorylation at position Y1068 in serum-starved DiFi cells stimulated with 50 ng / mL of EGF at 100 nM compared to untreated cells.

[0233] [Table 11]

[0234] [Example 19] Anti-EGFR IgG YU250-A02 and YU250-H10 exhibit potent inhibitory activity against the proliferation of the human cancer cell lineage FaDu compared to GC1118 and matsuzumab. The binding of anti-EGFR IgG to FaDu was analyzed by flow cytometry (Figure 22a). 100,000 cells were collected and incubated for 1 hour in a U-bottom 96-well plate with a series of serial dilutions of the antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA), the cells were incubated for 1 hour with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgG YU250-A02 and YU250-H10 bound to the human cancer cell lineage FaDu, as well as cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab, as analyzed by flow cytometry (EC 50The difference in values ​​is less than 2x (Table 9).

[0235] Growth inhibition was analyzed for FaDu (Figure 22b). 2,000 cells per 96 wells were seeded in DMEM + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of cell starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 5 days until cell count quantification using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). In contrast to anti-EGFR IgG GC1118 and matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (sequences as defined in Example 4) resulted in statistically significant inhibition of serum-starved FaDu growth at 10 nM compared to untreated cells.

[0236] Inhibition of EGFR signaling was analyzed for FaDu by Western blotting (Figure 22c). 200,000 cells per 6 wells were seeded in DMEM + 10% FCS + P / S and incubated for 24 hours, followed by a further 24 hours of starvation in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, and then stimulated with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in a more potent reduction of EGFR phosphorylation at position Y1068 in serum-starved FaDu stimulated with 50 ng mL of EGF at 100 nM compared to untreated cells.

[0237] [Table 12]

[0238] [Example 20] Anti-EGFR IgG YU250-A02 and YU250-H10 exhibit comparable inhibitory activity to cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab in the proliferation of the human cancer cell lineage LIM1215. The binding of anti-EGFR IgG to cells was analyzed for LIM1215 by flow cytometry (Figure 23a). 100,000 cells were harvested and incubated for 1 hour in a U-bottom 96-well plate with a series of serial dilutions of the antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA), the cells were incubated for 1 hour with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgG YU250-A02 and IgG YU250-H10 (sequences as defined in Example 4) bound to human cancer cell lineage LIM215 as well as cetuximab, GC1118, matsuzumab, necitumumab, and panitumumab (EC 50 The difference in values ​​is less than 2x (Table 10).

[0239] Proliferation inhibition was analyzed for LIM1215 (Figure 23b). For this purpose, 2,000 cells per 96 wells were seeded in RPMI + 10% FCS + P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 10 nM anti-EGFR IgG for 5 days until cell count quantification using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). Anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in statistically significant inhibition of proliferation in serum-starved LIM1215 cells at 10 nM, comparable to cetuximab, GC1118, matsuzumab, nesitumumab, and panitumumab, compared to untreated cells.

[0240] Inhibition of EGFR signaling was analyzed in LIM1215 cells by Western blotting (Figure 23c). Therefore, 200,000 cells per 6 wells were seeded in RPMI + 10% FCS + P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, followed by stimulation with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-062]).In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4) resulted in a more potent reduction of EGFR phosphorylation at position Y1068 in serum-starved LIM1215 cells stimulated with 50 ng / mL EGF at 100 nM, compared to untreated cells.

[0241] [Table 13]

[0242] [Example 21] Anti-EGFR IgG YU250-A02 and YU250-H10 exhibit more potent inhibitory activity than matsuzumab in the proliferation of EGF or HB-EGF-stimulated human cancer cell lines DiFi and LIM1215. Inhibition of cell proliferation was analyzed for stimulated DiFi (Figure 24a / c) and LIM1215 (Figure 24b / d). 2,000 cells per 96 wells were seeded in RPMI + 10% FCS + P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, followed by stimulation with 1 ng / mL EGF (Figure 24a / b) or 3 ng / mL HB-EGF (Figure 24c / d). After 5 days, cell count quantification was performed using the CellTiter-Glo® 2.0 cell viability assay (Promega[G9242]). In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4) at 100 nM resulted in statistically significant inhibition of proliferation in serum-starved DiFi and LIM1215 cells stimulated with 1 ng / mL EGF, as well as in serum-starved LIM1215 cells stimulated with 3 ng / mL HB-EGF.

[0243] [Example 22] Anti-EGFR IgG YU250-A02 and YU250-H10 exhibit cell binding to clinically observed EGFR mutations G465R and S492R, which are located in the epitopes of the approved antibodies cetuximab, nesitumumab, and panitumumab. Stable mouse fibroblast cell lines were generated using a retroviral MSCV-based vector system. Human EGFR (aa1-1,210) sequences were cloned into a pMSCV-LTR-puromycin vector, and transduced cells were selected using 1 μg / mL promycin (Sigma[P8833-25MG]) in DMEM+10%FCS+P / S. Empty vector controls were generated containing EGFR wild-type (wt) sequences (SEQ ID NO: 54), two clinically observed EGFR variants (G465R (SEQ ID NO: 55) and S492R (SEQ ID NO: 56)), and multiple cloning sites in place of EGFR, and EGFR expression was analyzed by Western blotting (Figure 25a). 200,000 cells per 6 wells were seeded in DMEM + 10% FCS + P / S, incubated for 24 hours, and then lysed (150 mM NaCl, 1% Nonident P-40, 10 mM NaF, 20 mM β-glycerophosphate, 1 mM EDTA, 1 mM Na3VO3, 0.5 mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)). After cell lysis, the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]) and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).

[0244] The binding of anti-EGFR IgG to cells was analyzed by flow cytometry for NIH-3T3+EGFR wt (Figure 25b), NIH-3T3+EGFR G465R (Figure 25c), and NIH-3T3+EGFR S492R (Figure 25d). For this purpose, 100,000 cells were collected and incubated for 1 hour in a U-bottom 96-well plate with 10 nM antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending cells in 175 μL of PBA to collect cells), the cells were incubated for 1 hour with PE conjugate anti-huFc antibody (1:500 dilution - dianova [109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Anti-EGFR IgG YU250-A02 and IgG YU250-H10 (sequences as defined in Example 4) bound to NIH-3T3+EGFR wt, NIH-3T3+EGFR G465R, and NIH-3T3+EGFR S492R, similar to IgG GC1118 and matsuzumab, while nesitumumab's binding to NIH-3T3+EGFR G465R was significantly reduced. No binding to NIH-3T3+EGFR G465R was observed for cetuximab and panitumumab, and cetuximab also did not bind to NIH-3T3+EGFR S492R.

[0245] [Example 23] Anti-EGFR IgG YU250-A02 and YU250-H10 effectively inhibit EGFR phosphorylation and MAPK pathway activation in stable mouse fibroblast lines expressing clinically observed EGFR mutations G465R and S492R. Inhibition of EGFR signaling was analyzed by Western blotting for NIH-3T3+EGFR wt (Figure 26a / b), NIH-3T3+EGFR G465R (Figure 26c), and NIH-3T3+EGFR S492R (Figure 26d). For this purpose, 200,000 cells per 6 wells were seeded in DMEM+10%FCS+P / S, incubated for 24 hours, and then starved for another 24 hours in a medium containing 0.2% FCS. Serum-starved cells were treated with 100 nM anti-EGFR IgG for 1 hour, and then stimulated with 50 ng / mL EGF (Biotechne[236-EG-200]) for 15 minutes. After cell lysis (150mM NaCl, 1% Nonident P-40, 10mM NaF, 20mM β-glycerophosphate, 1mM EDTA, 1mM Na3VO3, 0.5mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)), the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]), pEGFR Y1068 (Cell Signaling [CS#3777]), ERK (Cell Signaling [CS#9107]), pERK T202 / Y204 (Cell Signaling [CS#9101]), and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).

[0246] In contrast to anti-EGFR IgG matsuzumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 (as defined in Example 4) resulted in a more potent reduction of EGFR phosphorylation at position Y1068 and a more potent inhibition of the MAPK pathway in serum-starved NIH-3T3+ EGFR wt stimulated with 50 ng / mL EGF at 100 nM compared to untreated cells.

[0247] In contrast to anti-EGFR IgG cetuximab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in reduced EGFR phosphorylation at position Y1068 and inhibition of the MAPK pathway in serum-starved NIH-3T3+EGFR G465R and NIH-3T3+EGFR S492R cells stimulated with 50 ng / mL EGF at 100 nM, compared to untreated cells.

[0248] In contrast to anti-EGFR IgG nesitumumab and panitumumab, anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in reduced EGFR phosphorylation at position Y1068 and inhibition of the MAPK pathway in serum-starved NIH-3T3+ EGFR G465R cells stimulated with 50 ng / mL EGF at 100 nM, compared to untreated cells.

[0249] In contrast to anti-EGFR IgG nesitumumab and panitumumab, which still bound to NIH-3T3+EGFR S492R with at least 80% of the signal intensity compared to IgG GC1118 analyzed by flow cytometry at 10 nM (Figure 26d), anti-EGFR IgG YU250-A02 and IgG YU250-H10 resulted in reduced EGFR phosphorylation at position Y1068 in serum-starved NIH-3T3+EGFR S492R stimulated with 50 ng / mL EGF at 100 nM compared to untreated cells, which can also be described as inhibition of the MAPK pathway.

[0250] [Example 24] Various forms of bispecific EGFR×CD3 antibodies against YU250-H10 efficiently induce target cell death in co-culture assays with PBMCs isolated from healthy donors. The binding of bispecific EGFR×CD3 antibodies to cells was analyzed by flow cytometry in CD3-expressing Jurcut cells (Figure 27a), as well as in EGFR-expressing cancer cells FaDu (Figure 27b) and LIM1215 (Figure 27c). For this purpose, 100,000 cells were collected and incubated for 1 hour in a series of serial dilutions of the bispecific antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA) in a U-bottom 96-well plate. After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending cells in 175 μL of PBA to collect cells), the cells were incubated for 1 hour with either a PE conjugate anti-His antibody for scDb and scDb-scFv (1:250-Miltenyi[130-120-718]) or a PE conjugate anti-huFc antibody for eIg and Fab-eIg (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). The variable domain of the anti-EGFR antibody YU250-H10 and the variable domain sequence of the anti-human CD3 antibody huU3 (humanized version of UCHT1) (as described in SEQ ID NO: 96) were measured. H V as described in sequence number 97 L Bispecific antibodies containing scDb (SEQ ID NO: 19), scDb-scFv (SEQ ID NO: 20), eIg (SEQ ID NOs: 9, 21, and 23, 24; examples of this form are shown in Figures 33B and 33C), and Fab-eIg (SEQ ID NOs: 9, 21, and 22, 24; examples of this form are shown in Figures 33D and 33E) bound to CD3-expressing Jurcut cells and EGFR-expressing FaDu and LIM1215 cells.

[0251] The cytotoxicity of bispecific EGFR×CD3 antibodies was analyzed using EGFR-expressing FaDu (Figure 27d) and LIM1215 (Figure 27e) cells isolated as PBMCs from healthy donors and induced by activation. For this purpose, 20,000 target cells (FaDu or LIM1215) per 96-well plate were seeded in DMEM+10%FCS+P / S (FaDu) or RPMI+10%FCS+P / S (LIM1215) and incubated for 24 hours. PBMCs isolated from healthy donors by low-density gradient centrifugation (PromoCell Lymphocyte Separation Medium 1077 [C-44010]) and stored in FCS+10%DMSO at -80°C were thawed for 24 hours before being used in the cytotoxicity assay and cultured in RPMI+10%FCS+P / S. Target cells were pre-incubated for 15 minutes with a series of serial dilutions of a bispecific antibody diluted in RPMI + 10% FCS + P / S, followed by the addition of 200,000 PBMCs per 96 wells (effector cell to target cell ratio (E:T) -10:1). Viable cells were quantified 72 hours after treatment by measuring the optical density at 550 nm using crystal violet staining (TECAN Spark®).

[0252] A bispecific trivalent EGFR×CD3 antibody (scDb-scFv: 1.68±0.25nM / Fab-eIg: 0.76±0.11nM) containing the variable domain of the anti-EGFR antibody YU250-H10 showed increased binding to FaDu compared to a bivalent EGFR×CD3 antibody (scDb: 1.68±0.25nM / eIg: 2.17±0.26nM), resulting in more potent cytotoxicity in a co-culture assay with PBMCs isolated from healthy donors.

[0253] A bispecific trivalent EGFR×CD3 antibody containing the variable domain of the anti-EGFR antibody YU250-H10 (scDb-scFv: 0.31±0.04 nM / Fab-eIg: 0.27±0.04 nM) showed increased binding to LIM1215 compared to a bivalent EGFR×CD3 antibody (scDb: 1.02±0.08 nM / eIg: 0.85±0.07 nM), resulting in more potent cytotoxicity in a co-culture assay with PBMCs isolated from healthy donors.

[0254] [Table 14]

[0255] [Example 25] The anti-EGFR antibodies MKU011-A7 and YU250-H10, in eIg and Fab-eIg forms, exhibit bispecific EGFR×CD3 antibody efficacy in co-culture assays with PBMCs isolated from healthy donors, resulting in efficient target cell death. The binding of bispecific EGFR×CD3 antibodies to cells in two distinct forms of eIg and Fab-eIg containing the variable domains of MKU011-A7 or YU250-H10 was analyzed by flow cytometry for FaDu (Figure 28a), HCT-116 (Figure 28c), and SW620 (Figure 28e). For this purpose, 100,000 cells were harvested and incubated for 1 hour in a U-bottom 96-well plate with a series of serial dilutions of the bispecific antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA to harvest the cells), the cells were incubated for 1 hour with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]). After three washes of each well, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec). Bispecific EGFR×CD3 antibodies, including formal eIg and Fab-eIg containing the variable domain sequence of anti-EGFR antibody MKU011-A7 or YU250-H10 and anti-human CD3 antibody huU3 (UCHT1 humanized version), bound to EGFR-expressing FaDu, HCT-116, and SW620 cells.

[0256] The cytotoxicity of bispecific EGFR×CD3 antibodies in EGFR-expressing FaDu cells (Figure 28b), HCT-116 cells (Figure 28d), and SW620 cells (Figure 28f) was induced by activation of T cells isolated as PBMCs from healthy donors. For this purpose, 20,000 target cells (FaDu, HCT-116, or SW620) per 96 wells were seeded in DMEM+10%FCS+P / S (FaDu and SW620) or RPMI+10%FCS+P / S (HCT-116) and incubated for 24 hours. PBMCs were isolated from healthy donors by low-density gradient centrifugation (PromoCell Lymphocyte Separation Medium 1077 [C-44010]), stored at -80°C in FCS + 10% DMSO, thawed for 24 hours, and then used in cytotoxicity assays. They were cultured in RPMI + 10% FCS + P / S. Target cells were pre-incubated for 15 minutes with a series of serial dilutions of a bispecific antibody diluted in RPMI + 10% FCS + P / S, followed by the addition of 200,000 PBMCs per 96 wells (effector cell to target cell ratio (E:T) -10:1). Viable cells were quantified 72 hours after treatment by measuring the optical density at 550 nm using crystal violet staining (TECAN Spark®).

[0257] Bispecific bivalent EGFR×CD3 eIg antibodies containing the variable domain of YU250-H10 (SEQ ID NOs: 9, 21, 23, 24) (FaDu: 0.78±0.06 nM / HCT-116: 0.46±0.03 nM) show increased binding to FaDu and HCT-116 compared to bivalent EGFR×CD3 eIg antibodies containing the variable domain of MKU011-A7 (SEQ ID NOs: 4, 21, 23, 24) (FaDu: 3.28±0.17 nM / HCT-116: 4.28±0.24 nM), resulting in more potent cytotoxicity in co-culture assays with PBMCs isolated from healthy donors.

[0258] The bispecific trivalent EGFR×CD3 Fab-eIg antibody containing the variable domain YU250-H10 (SEQ ID NOs: 9, 21, 22, 24) (FaDu: 0.32±0.04nM / HCT-116: 0.13±0.02nM) exhibits similar binding affinity to the trivalent EGFR×CD3 Fab-eIg antibody containing the variable domain MKU011-A7 (SEQ ID NOs: 4, 21, 22, 24) (FaDu: 0.40±0.05nM / HCT-116: 0.16±0.02nM), and yields comparable cytotoxicity in co-culture assays with PBMCs isolated from healthy donors (EC related to cell binding). 50 (The difference in values ​​is less than 2x).

[0259] [Table 15]

[0260] [Example 26] The bispecific trivalent EGFR×CD3 antibody containing the anti-EGFR antibody YU250-H10 efficiently mediates target cell death in stable mouse fibroblast lineages expressing clinically observed EGFR mutations G465R and S492R. A bispecific Fab-eIg antibody was generated as a control by combining the binding site of anti-EGFR cetuximab with a humanized version of the anti-human CD3 antibody UCHT1 (huU3). The individual chain sequences of each bispecific antibody were cloned and expressed in suspension-culture-adapted HEK293-6E cells. The proteins were purified from the supernatant of transiently transfected cells by protein A affinity chromatography. Protein integrity was confirmed by size exclusion chromatography after preparative FPLC using a TSKgel SuperSW mAb HR column (Tosoh) (Figure 29a).

[0261] The cytotoxicity of bispecific EGFR×CD3 Fab-eIg antibodies against NIH-3T3+ empty vector, NIH-3T3+EGFR wt, NIH-3T3+EGFR G465R, and NIH-3T3+EGFR S492R was induced by activation of T cells isolated as PBMCs from healthy donors (Figure 29b). 5,000 target cells per 96 wells were seeded in DMEM+10%FCS+P / S and incubated for 24 hours. PBMCs isolated from healthy donors by low-density gradient centrifugation (PromoCell Lymphocyte Separation Medium 1077 [C-44010]), stored in FCS+10%DMSO at -80°C, thawed for 24 hours, and then used in cytotoxicity assays, cultured in RPMI+10%FCS+P / S. Target cells were pre-incubated for 15 minutes with a 10 nM bispecific antibody diluted in RPMI + 10% FCS + P / S, followed by the addition of 500,000 PBMCs per 96 wells (effector cell to target cell ratio (E:T) -10:1). Viable cells were quantified 72 hours after treatment by measuring the optical density at 550 nm using crystal violet staining (TECAN Spark®).

[0262] At 10 nM, a bispecific EGFR×CD3 Fab-eIg antibody containing the variable domain of YU250-H10 and the variable domain sequence of the human anti-CD3 antibody huU3 (humanized version of UCHT1) (SEQ ID NOs: 9, 21, 22, and 24) resulted in more than 50% cell death of NIH-3T3+EGFR wt, NIH-3T3+EGFR G465R, and NIH-3T3+EGFR S492R cells compared to untreated cells. At 10 nM, a bispecific EGFR×CD3 Fab-eIg antibody containing the anti-EGFR antibody cetuximab and the variable domain sequence of the anti-human CD3 antibody huU3 (humanized version of UCHT1) resulted in more than 50% cell death of NIH-3T3+EGFR wt cells compared to untreated cells, but failed to kill cells expressing EGFR mutations G465R and S492R.

[0263] [Table 16]

[0264] [Example 27] As an alternative to in vivo studies in syngeneic mouse models, the bispecific EGFR×CD3 antibody for YU250-H10 binds to CD3+ mouse spleen cells isolated from stable CT-26+EGFR wt and BALB / c. Stable mouse cancer cell lines were generated using a retroviral MSCV-based vector system. For this purpose, the human EGFR (aa1-1,210) sequence (SEQ ID NO: 54) was cloned into a pMSCV-LTR-puromycin vector, and transduced cells were selected using 1 μg / mL puromycin (Sigma [P8833-25MG]) in RPMI + 10 mM HEPES + 1 mM sodium + 0.1 mM non-essential amino acids + 10% FCS + P / S. Cell lines were generated for EGFR wild-type (wt) sequences and empty vector controls containing multiple cloning sites instead of EGFR. EGFR expression was verified by flow cytometry (Figure 30a) and Western blotting (Figure 30b). Binding of mouse anti-EGFR antibody (Biolegend [352902]) to CT-26+ empty vector and CT-26+ EGFR wt was analyzed by flow cytometry. For this purpose, 100,000 cells were collected and incubated with mouse anti-EGFR antibody diluted 1:100 in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA) in a U-bottom 96-well plate at 4°C for 1 hour. After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA to collect the cells), the cells were incubated with FITC conjugate anti-moF(ab')2 antibody contained in QIFIKIT® (Agilent [K0078]) at 4°C for 1 hour. After three washes of each well, fluorescence was measured using a MACSQuant® analyzer (Miltenyi Biotec), and receptor expression was quantified using calibration beads contained in QIFIKIT® (Agilent [K0078]).For Western blotting, 200,000 cells per 6 wells were seeded in RPMI + 10 mM HEPES + 1 mM sodium + 0.1 mM non-essential amino acids + 10% FCS + P / S, incubated for 24 hours, and then lysed (150 mM NaCl, 1% Nonident P-40, 10 mM NaF, 20 mM β-glycerophosphate, 1 mM EDTA, 1 mM Na3VO3, 0.5 mM PMSF, 0.25% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 7.5, cOmplete® protease inhibitor (Roche)). After cell lysis, the concentration of the total cell lysate was determined (DC Protein Assay - Bio-Rad [5000116]), and an equivalent volume was loaded onto a 4-12% NuPAGE® gel (Thermo Fisher [WG1402BOX]), and Western blotting was performed using the iBlot 2 Dry Blotting System (Thermo Fisher [IB23001]). The blots were blocked (0.5% (v / v) Roche blocking reagent (Merck) and 0.05% (v / v) Tween-20 in PBS), and the signals were detected using Fusion Solo S (Vilber) with primary antibodies induced against EGFR (Santa Cruz [sc-03]) and α-tubulin (Sigma [05-829]), as well as secondary HRP conjugate antibodies induced against mouse IgG (dianova [115-035-062]) or rabbit IgG (dianova [111-035-144]).

[0265] The binding of bispecific antibodies containing binding sites for anti-EGFR antibody YU250-H10 and Armenian hamster anti-mouse CD3 antibody 2C11 (SEQ ID NOs. 120, 121) was analyzed by flow cytometry on CT-26+EGFR wt (Figure 30c). For this purpose, 100,000 cells were collected and incubated in a series of serial dilutions of the bispecific antibodies diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA) in a U-bottom 96-well plate at 4°C for 1 hour. After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending cells in 175 μL of PBA to collect cells), the cells were incubated with PE-conjugated anti-huFc antibody (1:500-dianova[109-115-098]) at 4°C for 1 hour. After washing each well three times, fluorescence was measured using MACSQuant® VYB (Miltenyi Biotec).

[0266] The binding of bispecific antibodies containing binding sites for the anti-EGFR antibody YU250-H10 and the Armenian hamster anti-mouse CD3 antibody 2C11 was analyzed by flow cytometry in CD3+ spleen cells (Figure 30d). For this purpose, mouse spleen cells were isolated from BALB / c mice, and 100,000 cells were incubated for 1 hour at 4°C in a series of serial dilutions of the bispecific antibody diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA) in a U-bottom 96-well plate. After three washes of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending the cells in 175 μL of PBA to collect the cells), the cells were incubated for 1 hour at 4°C with PE conjugate anti-huFc antibody (1:500-dianova[109-115-098]). After washing each well three times, cells were stained with anti-moCD3-FITC antibody (Miltenyi Biotec [130-119-798]) at 4°C for 30 minutes. After washing each well once, fluorescence was measured using a MACSQuant® analyzer (Miltenyi Biotec).

[0267] Bispecific EGFR×CD3 antibodies, formal eIg (SEQ ID NOs. 25, 26, 27, and 28) and Fab-eIg (SEQ ID NOs. 25, 26, 27, and 29), containing the anti-EGFR antibody YU250-H10 and the anti-mouse CD3 antibody 2C11 variable domain sequence, bound to CD3+ mouse spleen cells isolated from CT-26+EGFR wt and BALB / c.

[0268] [Example 28] As an alternative to in vivo studies in syngeneic mouse models, a bispecific EGFRxCD3 antibody against YU250-H10 activates and induces proliferation of T cells in mouse spleen cells isolated from BALB / c. CD4 + / CD8 +Initial activation (Figure 31a / b) and proliferation (Figure 31c / d) of mouse T cells were analyzed by flow cytometry for bispecific antibodies containing binding sites for anti-EGFR antibody YU250-H10 and Armenian hamster anti-mouse CD3 antibody 2C11. 10,000 CT-26-EGFR-wt cells per 96-well plate were seeded in RPMI + 10 mM HEPES + 1 mM sodium + 0.1 mM non-essential amino acids + 10% FCS + P / S and incubated for 24 hours. Target cells were pre-incubated for 15 minutes with a series of serial dilutions of the bispecific antibody diluted in RPMI + 10% FCS + 50 μM β-ME + P / S, after which mouse splenocytes isolated from BALB / c cells (effector-to-target cell ratio (E:T) - 10:1) were added per 96-well plate. For the analysis of T cell proliferation, mouse splenocytes were stained with the CellTrace® CFSE cell proliferation kit (Thermo Fisher [C34554]). After 24 hours (initial activation) or 144 hours (proliferation), mouse splenocytes were collected by transferring them to a U-bottom 96-well plate, centrifugation at 1,500 rpm / 4°C for 3 minutes, and discarding the culture medium. For initial activation, mouse splenocytes were stained with anti-moCD3-FITC (Miltenyi Biotec [130-119-798]), anti-moCD4-VioBlue (Miltenyi Biotec [130-118-696]), anti-moCD8-PE (Miltenyi Biotec [130-123-781]), and anti-moCD69-APC (Miltenyi Biotec [130-115-576]) antibodies. For proliferation analysis, mouse splenocytes were stained with anti-moCD3-PE-Vio770 (Miltenyi Biotec [130-116-494]), anti-moCD4-VioBlue (Miltenyi Biotec [130-118-696]), and anti-moCD8-PE (Miltenyi Biotec [130-123-781]) antibodies. The antibodies were diluted 1:100 in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA) and incubated at 4°C for 30 minutes.After one wash of each well (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending cells in 175 μL of PBA to collect cells), fluorescence was measured using a MACSQuant® analyzer (Miltenyi Biotec).

[0269] Bispecific EGFR×CD3 assays containing YU250-H10 and anti-mouse CD3 antibody 2C11 variable domain sequence, formal eIg (SEQ ID NOs. 25, 26, 27, and 28) and Fab-eIg (SEQ ID NOs. 25, 26, 27, and 29), were activated in a co-culture assay with CT-26+EGFR wt (CD69). + ), and also CD4 derived from BALB / c + / CD8 + Mouse T cell proliferation was induced. Fab-eIg activity was significantly increased compared to eIg.

[0270] [Example 29] The bispecific EGFR×CD3 Fab-eIg substitute YU250-H10 inhibits tumor growth in a syngeneic mouse model using CT-26+EGFR wt in BALB / c. The pharmacokinetics (Figure 32a) of bispecific EGFR×CD3 eIg (SEQ ID NOs. 25, 26, 27, and 28) or Fab-eIg (SEQ ID NOs. 25, 26, 27, and 29) generated by combining the binding site of anti-EGFR YU250-H10 with Armenian hamster anti-mouse CD3 antibody 2C1 were analyzed by intravenous injection of 25 μg (eIg) or 30 μg (Fab-eIg) into the tail vein of 8-12 week old BALB / c mice. Serum was prepared by incubation of blood samples on ice for 20 minutes, followed by centrifugation at 4°C / 13,200 rpm for 30 minutes, and stored at -20°C until analysis. The antibody concentration of Fab-eIg was quantified by ELISA using immobilized EGFR-moIgG2a-Fc fusion protein (SEQ ID NOs. 33) containing the extracellular domain of human EGFR (aa25-645). The EGFR-moIgG2a-Fc fusion protein was coated onto polystyrene microtiter plates at 3 μg / mL in PBS. The remaining region was blocked with PBS and 2% skim milk (MPBS). The plates were incubated with serum diluted in MPBS. Binding antibodies were detected with HRP-conjugated anti-huFc antibody (1:5,000) diluted in MPBS. 1% TMB and 0.06% H2O2 diluted in 100 mM sodium phosphate buffer (pH 6.0) were used as detection substrates. The antibody concentration for eIg was quantified by flow cytometry using mouse splenocytes isolated from BALB / c. For this purpose, 100,000 cells were incubated for 1 hour at 4°C in a U-bottom 96-well plate with serum diluted in PBS + 2% (v / v) FCS + 0.02% NaN3 (PBA). After washing each well three times (centrifugation at 1,500 rpm / 4°C for 3 minutes, discarding the supernatant, and resuspending cells in 175 μL of PBA to collect cells), the cells were incubated with PE conjugate anti-huFc antibody (1:500-dianova[109-115-098]) at 4°C for 1 hour. After washing each well three times, the cells were stained with anti-moCD3-FITC antibody (Miltenyi Biotec[130-119-798]) at 4°C for 30 minutes.After one wash of each well, fluorescence was measured using a MACSQuant® analyzer.

[0271] The inhibition of tumor growth was analyzed in a syngeneic mouse model using various doses of bispecific EGFR×CD3 eIg (25 μg / 2.5 μg / 0.25 μg) or Fab-eIg (30 μg / 3 μg / 0.3 μg) (Figure 32b). For this purpose, 1 × 10⁻¹⁶ 6 Individual CT-26+EGFR wt cells were injected into the left and right flanks of 8-12 week old BALB / c mice. Tumor volume was measured using a sliding gauge. Mice were treated twice (11 and 15 days after tumor cell inoculation), and tumor volume at 17 days is shown. Treatment of syngeneic mouse models (CT-26+EGFR wt in BALB / c mice) with 30 μg or 3 μg of EGFR×CD3 Fab-eIg substitute containing the anti-EGFR antibody YU250-H10 combined with Armenian hamster anti-mouse CD3 antibody 2C11 resulted in a statistically significant reduction in tumor growth compared to treatment with PBS.

[0272] [Example 30] The bivalent and bispecific antibody Dab-Fc targets EGFR and HER3. The bispecific divalent Dab-Fc molecule contains a diabody (Db) domain specific to EGFR(H10) and HER3(3-43) (Schmitt et al., 2017, mAbs 9, 831-843) in IgG1 C H 1 / C L The heterodimer was produced by combining it with the heterodimer Fc moiety (knob-into-hole technique). Thus, the bispecific molecule is composed of two different polypeptide chains, the first chain being the heavy chain V H H10 variable domain, light chain V L Variable domains 3-43, steady domain C H Composed of 1 and a heterodimer Fc portion (hole) (V H H10-V L 3-43-C H 1-Fc ホール ), the second chain is heavy chain VH Variable domains 3-43, light chain V L Variable domain of H10, steady-state domain C L (V) is composed of k and the heterodimer Fc portion (knob) H 3-43-V L H10-C L k-Fc ノブ The bispecific bivalent molecule exhibits one antigen-binding site for EGFR and one binding site for HER3. The resulting binding molecule is schematically shown in Figure 34A.

[0273] The bispecific bivalent molecule was expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of two plasmids encoding both heavy chains. The protein secreted into the cell culture supernatant was purified using protein A affinity chromatography. SDS-PAGE analysis revealed two major bands at approximately 55 kDa, corresponding to the two heavy chains, under reducing conditions. Under non-reducing conditions, a single major band at approximately 200 kDa was observed, corresponding to an intact antibody composed of two different polypeptide chains (Figure 34B). The purity, integrity, and homogeneity of the bispecific bivalent molecule were confirmed by size exclusion chromatography (Figure 34C). Binding of the bispecific molecule to the extracellular domain (ECD) of EGFR (aa20-643 in SEQ ID NO: 33) and the extracellular domain (ECD) of HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). Subsequently, the plates were incubated with serial dilutions of bispecific antibodies. After washing, the bound antibodies were detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific bivalent antibodies exhibited EC values ​​in the sub-nanomolecal and nanomolar ranges. 50The values ​​(0.4 nM for EGFR-moFc and 1.4 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 34D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific bivalent antibody within the expected range.

[0274] [Example 31] The bivalent and bispecific antibody eIg 1+1 targets EGFR and HER3. The bispecific divalent eIg 1+1 molecule has an IgE-modified Fv domain (Schmitt et al., 2017, mAbs 9, 831-843) that is specific to HER3(3-43). H Two domains (hetEHD2) and an EGFR(H10)-specific Fv domain are incorporated into IgG1 C H 1 / C L The heterodimer was generated by combining it with the heterodimer Fc moiety (knob-into-hole technique). Therefore, the bispecific molecule is composed of four different polypeptide chains and EGFR V L H10-C L λ-targeting light chain (SEQ ID NO: 9), HER3 V L Light chains targeting 3-43-hetEHD2 (C247S, N275Q), EGFR V H H10-C H 1-Fc ホール A heavy chain targeting HER3 V (SEQ ID NO: 21), HER3 V H 3-43-hetEHD2(C337S)-Fc ノブ It consists of a heavy chain that targets EGFR. The bispecific bivalent eIg molecule exhibits one antigen-binding site for EGFR and one binding site for HER3. The resulting binding molecule is schematically shown in Figure 35A.

[0275] Bispecific divalent eIg 1+1 molecules were expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of four plasmids encoding both heavy and light chains. The proteins secreted into the cell culture supernatant were purified using protein A affinity chromatography. SDS-PAGE analysis revealed four major bands under reducing conditions: one band of approximately 55 kDa corresponding to the heavy chain containing the CH1 domain (SEQ ID NO: 21), and a glycosylated form of the heavy chain containing hetEHD2 (V H 3-43-hetEHD2(C337S)-Fc ノブ One band of approximately 65 kDa, corresponding to the EGFR molecule, and two bands of approximately 25 kDa, corresponding to the two light chains of the antibody, were identified. Under non-reducing conditions, one major band of approximately 200 kDa, corresponding to the intact antibody composed of four different polypeptide chains, was observed (Figure 35B). The purity, integrity, and homogeneity of the bispecific bivalent eIg 1+1 molecule were confirmed by size exclusion chromatography (Figure 35C). Binding of the bispecific molecule to the extracellular domain (ECD) of EGFR (aa20-643 in SEQ ID NO: 33) and the extracellular domain (ECD) of HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). Next, the plates were incubated with serial dilutions of the bispecific antibody. After washing, the conjugated antibody was detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific bivalent antibody was detected in sub-nanomole and nanomolar range EC 50The values ​​(0.5 nM for EGFR-moFc and 2.7 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 35D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific bivalent antibody within the expected range.

[0276] [Example 32] The tetravalent, bispecific antibody Dab-Fc targets EGFR and HER3. The bispecific tetravalent Fab-eIg molecule modifies the variable domain of the anti-EGFR antibody (H10) with IgG1 C H 1 / C L A modified IgE molecule is formed by combining a heterodimer and the variable domain of an anti-HER3 antibody (3-43; Schmitt et al., 2017, mAbs 9, 831-843) with the homodimerizing Fc portion. H It was generated by combining it with two domains (hetEHD2). The tetravalent bispecific molecule is composed of three different polypeptide chains: light chain V L 3-43-hetEHD2(C247S, N275Q), light chain V L H10-CLλ (SEQ ID NO: 9) and heavy chain V H H10-CH1-V H 3-43-hetEHD2(C337S)-Fc. The two Fab moieties in the heavy chain were separated using a linker containing 10 amino acids ((GGSGG)2; SEQ ID NO: 132). The bispecific tetravalent Fab-eIg molecule exhibits two antigen-binding sites for EGFR and two binding sites for HER3. The resulting binding molecule is schematically shown in Figure 36A.

[0277] The bispecific tetravalent Fab-eIg molecule was expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of three plasmids encoding the heavy chain and both light chains. The protein secreted into the cell culture supernatant was purified using protein A affinity chromatography. SDS-PAGE analysis revealed three major bands under reducing conditions: one band of approximately 80 kDa corresponding to the heavy chain, and two bands of approximately 26 kDa corresponding to the two light chains of the antibody (Figure 36B). Under non-reducing conditions, one major band of approximately 250 kDa corresponding to the intact antibody composed of three different polypeptide chains was observed. The purity, integrity, and homogeneity of the bispecific tetravalent Fab+eIg molecule were confirmed by size exclusion chromatography (Figure 36C). The binding of this bispecific tetravalent Fab+eIg molecule to the extracellular domains of EGFR (aa20-643 in SEQ ID NO: 33) and HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). Subsequently, the plates were incubated with serial dilutions of the bispecific antibody. After washing, the bound antibody was detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific tetravalent antibody was detected in sub-nanomolar and nanomolar range EC 50 The values ​​(0.5 nM for EGFR-moFc and 1.7 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 36D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific tetravalent antibody within the expected range.

[0278] [Example 33] The tetravalent, bispecific antibody Db-Ig 2+2 targets EGFR and HER3. The bispecific tetravalent Db3-43×H10-Ig molecule was generated by combining Db molecules specific to EGFR(H10) and HER3(3-43) with the constant domain of an IgG antibody. Therefore, the Db3-43×H10-Ig molecule consists of two different polypeptides: light chain V H 3-43×V L H10-C L and heavy chain V H H10×V L hu3-43-C H 1-C H 2-C H It consists of three components. The bispecific Db3-43×H10-Ig exhibits two antigen-binding sites for EGFR and two antigen-binding sites for HER3 (Figure 37A).

[0279] Db3-43×H10-Ig was expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of two plasmids encoding the light and heavy chains. The protein secreted into the cell culture supernatant was then C HThe protein was purified using 1-CaptureSelect affinity chromatography. SDS-PAGE analysis revealed two bands of approximately 65 kDa and 35 kDa, corresponding to the heavy and light chains, under reducing conditions, and one major band of approximately 220 kDa, corresponding to the intact bispecific tetravalent Db3-43×H10-Ig molecule, under non-reducing conditions (Figure 37B). The purity, integrity, and homogeneity of Db3-43×H10-Ig were confirmed by size exclusion chromatography (Figure 37C). The binding of Db3-43×H10-Ig to the extracellular domains of EGFR (aa20-643 in SEQ ID NO: 33) and HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. The EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). Next, the plates were incubated with serial dilutions of the bispecific antibody. After washing, the bound antibody was detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific tetravalent antibody was detected in the subnanomolecular EC range. 50 The values ​​(0.9 nM for EGFR-moFc and 0.8 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 37D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific tetravalent antibody within the expected range.

[0280] [Example 34] The trivalent, bispecific antibody Db / Fab-Fc targets EGFR and HER3. Bispecific trivalent Db3-43×3-43 / FabH10-Fc kihThe molecule was generated by combining a HER3(3-43)-specific Db molecule and an EGFR-specific Fab molecule (H10) with a CH1 / CL heterodimer gG1 type using the heterodimer Fc moiety (knob-into-hole technique). Therefore, Db3-43×3-43 / FabH10-Fc kih The molecule consists of four different polypeptides: i)V H 3-43×VL3-43-C L λ (light chain of the diamond body), ii) V H 3-43×V L 3-43-C H 1-C H 2-C H 3 ホール (Heavy chain of the diamond body), iii)V H H10-C H 1-C H 2-C H 3c ノブ (Fab portion heavy chain, SEQ ID NO: 123), and iv)V L H10-C L It is composed of λ (light chain of the Fab portion, SEQ ID NO: 9). Bispecific antibody Db3-43×3-43 / FabH10-Fc kih It has two antigen-binding sites for HER3 and one antigen-binding site for EGFR (Figure 38A).

[0281] Db3-43×3-43 / FabH10-Fc kihThe protein was expressed in HEK293-6E cells transiently transfected after co-administration of four plasmids encoding the light and heavy chains of the diabody and Fab moieties using polyethyleneimine as the transfection reagent. The protein secreted into the cell culture supernatant was purified using protein A affinity chromatography. SDS-PAGE analysis revealed four bands under reducing conditions: approximately 70 kDa (diabody moiety heavy chain), 50 kDa (Fab moiety heavy chain), 35 kDa (diabody moiety light chain), and 20 kDa (Fab moiety light chain). Under non-reducing conditions, a single major band of approximately 200 kDa was observed (Figure 38B). The purity, integrity, and homogeneity of the Db3-43×3-43 / FabH10-Fckih molecule were confirmed by size exclusion chromatography (Figure 38C).

[0282] Db3-43×3-43 / FabH10-Fc kih The binding of the EGFR and HER3 mouse Fc fusion proteins to the extracellular domains of EGFR (aa20-643 in SEQ ID NO: 33) and HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). Subsequently, the plates were incubated with serial dilutions of bispecific antibodies. After washing, the bound antibodies were detected using HRP-conjugated anti-human Fc antibodies and TMB and H2O2 as substrates. The bispecific trivalent antibodies were found in the sub-nanomolelic EC range. 50 The values ​​(0.1 nM for EGFR-moFc and 0.1 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 38D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific trivalent antibody within the expected range.

[0283] [Example 35] The bivalent and bispecific antibody Fab / scFv-Fc targets EGFR and HER3. The bispecific bivalent Fab / scFv-Fc molecule modifies the variable domain of the anti-EGFR antibody (H10) with IgG1 C H 1 / C L Heterodimers were generated by combining them with single-chain variable fragments of IgG 3-43 (scFv3-43; Schmitt et al., 2017, mAbs 9, 831-843) further combined with the heterodimer and the Fc moiety for heterodimerization (knob-into-hole technique). Thus, the divalent bispecific molecule was generated by combining it with three different polypeptide chains: light chain V L H10-C L λ (SEQ ID NO: 9), heavy chain V H H10-C H 1-Fc ホール (Sequence ID 21) and heavy chain scFv3-43-Fc ノブ It is composed of the following. The scFv3-43 portion of the chain was separated by a linker containing three amino acids (AAA). The bispecific bivalent Fab / scFv-Fc molecule exhibits one antigen-binding site for EGFR and one binding site for HER3. The resulting binding molecule is schematically shown in Figure 39A.

[0284] The bispecific bivalent Fab / scFv-Fc molecule was expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of three plasmids encoding both heavy and light chains. The protein secreted into the cell culture supernatant was purified using protein A affinity chromatography. SDS-PAGE analysis revealed three bands under reducing conditions: two bands in the range of 60 kDa–55 kDa corresponding to the heavy chain, and one band of approximately 24 kDa corresponding to the antibody light chain (Figure 39B). Under non-reducing conditions, one major band of approximately 170 kDa corresponding to the intact antibody composed of three different polypeptide chains was observed. The purity, integrity, and homogeneity of the bispecific bivalent Fab / scFv-Fc molecule were confirmed by size exclusion chromatography (Figure 39C).

[0285] The binding of this bispecific bivalent Fab / scFv-Fc molecule to the extracellular domains of EGFR (aa20-643 in SEQ ID NO: 33) and HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR and HER3 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). The plates were then incubated with serial dilutions of the bispecific antibody. After washing, the bound antibody was detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific bivalent antibody was detected in the sub-nanomolelic EC range. 50 The values ​​(0.3 nM for EGFR-moFc and 0.2 nM for HER3-moFc) showed concentration-dependent binding to EGFR and HER3 (EGFR-moFc and HER3-moFc) fused to the mouse Fc portion (Figure 39D). These experiments confirmed the binding of both EGFR and HER3 antigens of the bispecific bivalent antibody within the expected range.

[0286] [Example 36] The bivalent and bispecific antibody eIg 1+1 targets EGFR and MET. The bispecific bivalent antibody eIg 1+1 molecule contains a MET-specific Fv domain (5D5; Jin H, Yang R, Zheng Z, Romero M, Ross J, Bou-Reslan H, Carano RA, Kasman I, Mai E, Young J, Zha J, Zhang Z, Ross S, Schwall R, Colbern G, Merchant M. MetMAb, the one-armed 5D5 anti-c-Met antibody, inhibits orthotopic pancreatic tumor growth and improves survival. Cancer Res. 2008 Jun 1; 68(11): 4360-8. doi: 10.1158 / 0008-5472.CAN-07-5960. PMID: 18519697) modified from IgE C H Two domains (hetEHD2) and an EGFR(H10)-specific Fv domain are incorporated into IgG1 C H 1 / C L The heterodimer was generated by combining it with the heterodimer Fc moiety (knob-into-hole technique). Therefore, the bispecific molecule consists of four different polypeptide chains: light chain V L H10-C L λ (SEQ ID NO: 9), light chain V L 5D5-hetEHD2(C247S, N275Q)(Sequence ID 126), heavy chain V H H10-C H 1-Fc ノブ (Sequence ID 124) and heavy chain V H 5D5-hetEHD2(C337S)-Fc ホール It consists of (SEQ ID NO: 125). The bispecific bivalent eIg molecule exhibits one antigen-binding site for EGFR and one binding site for MET. The resulting binding molecule is schematically shown in Figure 40A.

[0287] Bispecific bivalent eIg 1+1 molecules were expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of four plasmids encoding both heavy and light chains. Proteins secreted into the cell culture supernatant were purified using protein A affinity chromatography. SDS-PAGE analysis revealed four bands under reducing conditions: two bands of approximately 55 kDa most likely corresponding to the two heavy chains, and two bands of approximately 25 kDa most likely corresponding to the two light chains of the antibody. Under non-reducing conditions, one major band of approximately 200 kDa was observed, most likely corresponding to the intact antibody composed of four different polypeptide chains (Figure 40B). Additionally, smaller bands of approximately 190 kDa and 100 kDa were also observed under non-reducing conditions. The purity, integrity, and homogeneity of the bispecific divalent eIg 1+1 molecules were confirmed by size exclusion chromatography (Figure 40C). Binding of the bispecific molecules to the extracellular domains of EGFR (aa20-643 in SEQ ID NO: 33) and HER3 (aa21-643 in SEQ ID NO: 129) was determined by ELISA. EGFR mouse Fc and MET human Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). The plates were then incubated with serial dilutions of the bispecific antibody. After washing, the bound antibody was detected using HRP-conjugated anti-human Fc antibody and TMB and H2O2 as substrates. The bispecific divalent antibody was detected in the sub-nanomolecular EC range. 50 The values ​​(1.1 nM for EGFR-moFc and 4.1 nM for MET-huFc) showed concentration-dependent binding to EGFR fused to the mouse Fc portion (EGFR-moFc) and MET fused to the human Fc portion (MET-huFc) (Figure 40D). These experiments confirmed the binding of both bispecific bivalent antibodies to the antigens EGFR and HER3.

[0288] [Example 37] The bivalent and bispecific antibody eIg 1+1 targets EGFR and HER2. The bispecific bivalent eIg 1+1 molecule has a HER2-specific Fv domain (4D5; Lewis GD, Figari I, Fendly B, Wong WL, Carter P, Gorman C, Shepard HM. Differential responses of human tumor cell lines to anti-p185HER2 monoclonal antibody. Cancer Immunol Immunother. 1993 Sep; 37(4): 255-63. doi: 10.1007 / BF01518520. PMID: 8102322) modified from IgE C H Two domains (hetEHD2) and an EGFR(H10)-specific Fv domain are incorporated into IgG1 C H 1 / C L The heterodimer was generated by combining it with the heterodimer Fc moiety (knob-into-hole technique). Therefore, the bispecific molecule consists of four different polypeptide chains: light chain V L H10-C L λ (SEQ ID NO: 9), light chain V L 4D5-hetEHD2(C247S, N275Q)(Sequence ID 128), heavy chain V H H10-C H 1-Fc ノブ (Sequence ID 124), and heavy chain V H 4D5-hetEHD2(C337S)-Fc ホール It consists of (SEQ ID NO: 127). The bispecific bivalent eIg molecule has one antigen-binding site for EGFR and one binding site for HER2. The resulting binding molecule is schematically shown in Figure 41A.

[0289] The bispecific bivalent eIg 1+1 molecule was expressed in HEK293-6E cells transiently transfected with polyethyleneimine as the transfection reagent after co-administration of four plasmids encoding both heavy and light chains. The protein secreted into the cell culture supernatant was purified using protein A affinity chromatography. SDS-PAGE analysis revealed four major bands under reducing conditions: two bands of approximately 55 kDa most likely corresponding to the two heavy chains, and two bands of approximately 25 kDa most likely corresponding to the two light chains of the antibody. Under non-reducing conditions, one major band of approximately 200 kDa was observed, most likely corresponding to the intact antibody composed of four different polypeptide chains (Figure 41B). The purity, integrity, and homogeneity of the bispecific bivalent eIg 1+1 molecule were confirmed by size exclusion chromatography (Figure 41C). The binding of bispecific molecules to the extracellular domains (ECDs) of EGFR (aa20-643 in SEQ ID NO: 33) and HER2 (aa23-652 in SEQ ID NO: 131) was determined by ELISA. EGFR and HER2 mouse Fc fusion proteins were coated onto polystyrene microtiter plates at a concentration of 3 μg / ml diluted in PBS. The remaining binding sites were blocked with PBS and 2% skim milk (MPBS). The plates were then incubated with serial dilutions of bispecific antibodies. After washing, the bound antibodies were detected using HRP-conjugated anti-human Fc antibodies and TMB and H2O2 as substrates. The bispecific bivalent antibodies were found to have sub-nanomolecular EC2 levels. 50 The values ​​(6.2 nM for EGFR-moFc and 14.7 nM for HER2-moFc) showed concentration-dependent binding to EGFR and HER2 (EGFR-moFc and HER2-moFc) fused to the mouse Fc region (Figure 41D). These experiments confirmed the binding of both bispecific bivalent antibodies to the EGFR and HER2 antigens.

Claims

1. An antigen-binding protein comprising a first antibody variable domain and a second antibody variable domain, wherein the variable domains are a) 10 determined by biolayer interferometry (BLI) －５ It specifically binds to human epidermal growth factor receptor (EGFR) variants having one or more mutations among V441D, S464L, G465R, and S492R at a KD concentration of less than molal, b) Does not specifically bind to human EGFRs having mutation Q435P and / or double mutation F436A / I462A. An antigen-binding protein that forms a binding site, and its binding to human EGFR inhibits ligand-induced activation of human EGFR.

2. The antigen-binding protein according to claim 1, wherein the variable domain forms a binding site that specifically binds to domain III (SEQ ID NO: 61) of EGFR.

3. An antigen-binding protein according to claim 1 or 2, selected from the group consisting of antibodies or their antigen-binding fragments and chimeric antigen receptors (CARs), preferably selected from the group consisting of Fab, Fab', (Fab')2, Fv, disulfide-linked Fv, BsFv, dsFv, (dsFv)2, dsFv-dsFv', scFv, scFv dimer, single-chain domain antibody, diabody, ds-diabody, and bivalent domain antibody.

4. The antigen-binding protein according to claim 3, wherein the antibody is selected from the group consisting of chimeric antibodies, humanized antibodies, bispecific antibodies, and polyspecific antibodies, and preferably the antibody is a human antibody.

5. An antigen-binding protein according to any one of claims 1 to 4, wherein the first antibody-variable domain comprises light chain complementarity-determining region (CDRL) 1 (CDRL1), CDRL2, CDRL3 of the variable light chain region of SEQ ID NO: 83, 84, or 85, and / or the second variable domain comprises CDRH1, CDRH2, and CDRH3 of the heavy chain-variable region of SEQ ID NO: 74 or 75, wherein the CDRs are identified by the Kabat definition, Chothia definition, IMGT definition, or contact definition, and each CDR may contain one or two amino acid substitutions.

6. (a) The first antibody variable domain is CDRL1 of sequence number 86, CDRL2 of Sequence ID 87, and CDRL3 of sequence number 88 Includes, (b) The second antibody variable domain is CDRH1 of sequence number 71, CDRH2 of Sequence ID No. 72, and CDRH3 of Sequence ID 73 It includes, preferably, (a) The first antibody variable domain is CDRL1 of sequence number 76 or 77, CDRL2 of sequence number 78 or 79, and CDRL3 of sequence number 80, 81, or 82 Includes, (b) The second antibody variable domain is CDRH1 of sequence number 71, CDRH2 of Sequence ID No. 72, and CDRH3 of Sequence ID 73 The antigen-binding protein according to any one of claims 1 to 5, wherein each CDR may contain one or two amino acid substitutions.

7. The first antibody variable domain comprises one or more of the framework (FR) sequence of SEQ ID NO: 83, 84, or 85, or variants of the FR sequence having at least 90% sequence identity with the FR sequence of SEQ ID NO: 83, 84, or 85, and / or the second antibody variable domain comprises one or more of the FR sequence of SEQ ID NO: 74 or 75, or variants of the FR sequence having at least 90% sequence identity with the FR sequence of SEQ ID NO: 74 or 75, wherein the FR is identified by Kabat definition, Chothia definition, IMGT definition, or contact definition, preferably, The first variable domain comprises one of the amino acid sequences according to SEQ ID NO: 83, 84, or 85, and the second variable domain comprises one of the amino acid sequences according to SEQ ID NO: 74 or 75, more preferably, Antigen-binding proteins, in combination: a) Sequence ID 74 and Sequence ID 83, b) Sequence ID 75 and Sequence ID 84, or c) Sequence IDs 74 and 85 The antigen-binding protein according to any one of claims 1 to 6, comprising one of the above, wherein each CDR may comprise one or two amino acid substitutions.

8. The antigen-binding protein has the following sequence: a) CDRH1, CDRH2, and CDRH3 of sequence numbers 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of sequence numbers 76, 78, and 80, b) CDRH1, CDRH2, and CDRH3 of Sequence IDs 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of Sequence IDs 77, 79, and 81, or c) CDRH1, CDRH2, and CDRH3 of Sequence IDs 71, 72, and 73, and CDRL1, CDRL2, and CDRL3 of Sequence IDs 77, 79, and 82 The antigen-binding protein according to any one of claims 1 to 7, wherein each CDR may contain one or two amino acid substitutions.

9. (i) V of sequence number 83, 84 or 85 Ｌ and V of sequence number 74 or 75 Ｈ It competes for binding to IgG1 antibodies, or (ii) V of sequence number 83, 84 or 85 Ｌ and V of sequence number 74 or 75 Ｈ IgG1 antibodies that bind to the same epitopes as the IgG1 antibody containing the antibody. an antigen-binding protein according to any one of claims 1 to 8.

10. A nucleic acid (one or more) encoding an antigen-binding protein according to any one of claims 1 to 9.

11. A vector comprising one or more nucleic acids according to claim 10.

12. Recombinant cells expressing an antigen-binding protein according to any one of claims 1 to 9, one or more nucleic acids according to claim 10, or a vector according to claim 11.

13. A pharmaceutical composition comprising an antigen-binding protein according to any one of claims 1 to 9, one or more nucleic acids according to claim 10, a vector according to claim 11, or recombinant cells according to claim 12, and a pharmaceutically acceptable excipient.

14. An antigen-binding protein according to any one of claims 1 to 9, a nucleic acid (one or more) according to claim 10, a vector according to claim 11, a recombinant cell according to claim 12, or a pharmaceutical composition according to claim 13, for use in pharmaceuticals.

15. Cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, cervical cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, more preferably, A pharmaceutical composition according to any one of claims 1 to 9, an antigen-binding protein according to claim 10, a vector according to claim 11, a recombinant cell according to claim 12, or a pharmaceutical composition according to claim 13, for use in the treatment of cancer selected from the group consisting of head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

16. A method for treating cancer, comprising administering to a patient in need of treatment a therapeutically effective amount of an antigen-binding protein according to any one of claims 1 to 9, one or more nucleic acids according to claim 10, a vector according to claim 11, recombinant cells according to claim 12, or a pharmaceutical composition according to claim 13.

17. The method according to claim 16, wherein the cancer is selected from the group consisting of lung cancer, prostate cancer, breast cancer, colon cancer, rectal cancer, head cancer, neck cancer, esophageal and gastric cancer, liver cancer, glioblastoma, cervical cancer, ovarian cancer, bladder cancer, kidney cancer, and pancreatic cancer, and more preferably the cancer is selected from the group consisting of head and neck squamous cell carcinoma, non-small cell lung cancer, metastatic colorectal cancer, and recurrent or metastatic squamous cell carcinoma of the head and neck.

Images