EA039356B1 - BISPECIFIC EGFR/c-Met ANTIBODIES - Google Patents

Abstract

Bispecific EGFR/c-Met antibodies and methods of making and using the molecules. Epidermal growth factor receptor (EGFR, ErbB1 or HER1) is a Type I transmembrane glycoprotein of 170 kDa that is encoded by the c-erbB1 proto-oncogene. EGFR is a member of the human epidermal growth factor receptor (HER) family of receptor tyrosine kinases (RTK) which includes HER2 (ErbB2), HER3 (ErbB3) and HER4 (ErbB4). EGFR signaling is initiated by ligand binding followed by induction of conformational change, homodimerization or heterodimerization of the receptor with other ErbB family members.

Description

Scope of the invention

The present invention relates to EGFR and/or c-Met bispecific antibodies and methods for making and using the molecules.

Prerequisites for the creation of the invention

The epidermal growth factor receptor (EGFR, ErbB1, or HER1) is a 170 kDa type I transmembrane glycoprotein encoded by the c-ErbB1 proto-oncogene. EGFR is a member of the human epidermal growth factor receptor (HER) family, which are receptor tyrosine kinases (RTKs), including HER2 (ErbB2), HER3 (ErbB3), and HER4 (ErbB4). EGFR signaling is triggered by ligand binding, followed by induction of a conformational change, homodimerization or heterodimerization of the receptor with other members of the ErbB family, and transautophosphorylation of the receptor (Ferguson et al., Annu Rev Biophys, 37:353-73, 2008), resulting in triggering signal transduction cascades that ultimately affect a wide range of cellular functions, including cell proliferation and survival. An increase in EGFR expression or kinase activity has been correlated with a number of human cancers, making EGFR an attractive target for therapeutic intervention (Mendelsohn et al., Oncogene, 19:6550-6565, 2000; Grunwald et al., J. Natl. Cancer Inst. 95:851-67, 2003; Mendelsohn et al., Semin Oncol 33:369-85, 2006). Both an increase in the copy number of the EGFR gene and an increase in protein expression have been associated with favorable responses to the EGFR tyrosine kinase inhibitor IRESSA™ (gefitinib) in non-small cell lung cancer (Hirsch et al., Ann Oncol. 18:752-60, 2007).

EGFR therapy includes the use of both small molecules and anti-EGFR antibodies, which are approved for the treatment of colorectal cancer, pancreatic cancer, head and neck cancer, and non-small cell lung cancer (NSCLC) (Baselga and Arteaga, J. Clin. Oncol. 23: 2445-2459 (20005; Gill et al., J. Biol. Chem., 259:7755-7760, 1984; Goldstein et al., Clin. Cancer Res., 1:131 1-1318; 1995; Prewett et al. , Clin Cancer Res. 4:2957-2966, 1998).

The effectiveness of anti-EGFR therapy may depend on the type of tumor and the EGFR mutation/amplification status of the tumor. Side effects of existing therapeutic agents may include dermal toxicity (De Roock et al., Lancet Oncol. 11:753-762, 2010; Linardou et al., Nat. Rev. Clin. Oncol., 6:352-366, 2009; Li and Perez-Soler, Targ Oncol 4:107-119, 2009). EGFR tyrosine kinase inhibitors (TKIs) are commonly used as 2nd line therapy for non-small cell lung cancer (NSCLC), but often fail within 12 months due to resistance-causing pathways (Riely et al., Clin. Cancer Res. 12 :839-44, 2006).

c-Met encodes a transmembrane tyrosine kinase receptor. It was first discovered as a proto-oncogene in 1984 after exposure to a carcinogen resulted in the constitutively active TPR-MET fusion protein (Cooper et al., Nature, 311:29-33, 1984). Activation of c-Met via its ligand, hepatocyte growth factor (HGF), causes stimulation of a wide range of cellular processes, including growth, motility, malignancy, metastasis, epithelial-mesenchymal transformation, angiogenesis/wound healing, and tissue regeneration (Christensen et al ., Cancer Lett. 225:1-26, 2005; Peters and Adjei, Nat. Rev. Clin. Oncol. 9:314-26, 2012). c-Met is synthesized as a single chain protein that is proteolytically cleaved into a 50 kDa alpha subunit and a 140 kDa beta subunit, which are linked by a disulfide bond (Ma et al., Cancer and Metastasis Reviews, 22:309-325, 2003). c-Met is structurally similar to other membrane receptors such as RON and Sea. The exact stoichiometry of HGF:c-Met binding is not known, but essentially it is believed that two HGF molecules bind to two c-Met molecules, resulting in receptor dimerization and autophosphorylation at tyrosine residues 1230, 1234 and 1235 (Stamos et al., The EMBO Journal, 23:2325-2335, 2004). Ligand-independent c-Met autophosphorylation is also possible due to gene amplification, mutation, or receptor overexpression.

c-Met is often amplified, mutated, or overexpressed in many types of cancer, including gastric, lung, colon, breast, bladder, head and neck, ovarian, prostate, thyroid, pancreatic, and CNS cancers. Missensmutations, typically located in the kinase domain, are frequently found in hereditary papillary renal cell carcinomas (PRCC) and in 13% of sporadic PRCCs (Schmidt et al., Oncogene, 18:2343-2350, 1999). c-Met mutations located in the semaphorin or near-membrane domains of c-Met are frequently found in gastric, head and neck, liver, ovarian, NSCLC, and thyroid cancers (Ma et al., Cancer and Metastasis Reviews, 22:309-325 , 2003; Sakakura et al., Chromosomes and Cancer, 1999. 24:299-305). In brain cancer, colorectal cancer, gastric and lung cancer, c-Met amplification is found, which often correlates with disease progression (Ma et al., Cancer and Metastasis Reviews, 22:309-325, 2003). Up to 4% and 20% of non-small cell lung cancer (NSCLC) and gastric cancer, respectively, show c-Met amplification (Sakakura et al., Chromosomes and Cancer, 1999, 24:299-305, Sierra and Tsao, Therapeutic Advances in Medical Oncology, 3 :S21-35, 2011). In lung cancer, even in the absence of gene amplification, c-Met overexpression is often observed (Ichimura et al., Jpn. J. Cancer Res., 87:1063-9, 1996). Moreover, almost half of the clinical samples of pulmonary adenocarcinomas showed high levels of c-Met and HGF, both of which are correlated with increased tumor growth rate, metastasis, and poor prognosis (Sierra and Tsao,

Therapeutic Advances in Medical Oncology, 3:S21-35, 2011; Siegfried et al., Ann. Thorac. Surg. 66:1915-8,

1998).

Nearly 60% of all tumors that have become resistant to EGFR tyrosine kinase inhibitors have increased c-Met expression, c-Met amplification, or the only known c-Met HGF ligand (Turke et al., Cancer Cell, 17:77-88, 2010 ), indicating the presence of a compensatory pathway for EGFR via c-Met. Amplification of c-Met was first discovered in cultured cells that became resistant to gefitinib, an EGFR kinase inhibitor, and showed improved survival through the Her3 pathway (Engelman et al., Science, 316:1039-43, 2007). This was further confirmed in clinical samples in which 9 of 43 patients with acquired resistance to erlotinib or gefitinib showed c-Met amplification compared to only 2 of 62 patients without treatment. Four of the nine treated patients also acquired the EGFR-activating T790M mutation, demonstrating concurrent resistance-causing pathways (Beat et al., Proc. Natl. Acad. Sci. USA, 104:20932-7, 2007).

The separate roles of both EGFR and c-Met in cancer are well known, making these targets attractive for combination therapy. Both receptors signal through the same anti-apoptotic and survival pathways (ERK and AKT). Thus, inhibition of this pair in combination may limit the activation potential of compensatory pathways, thus increasing overall efficacy. Combination therapies targeting EGFR and c-Met have been tested in clinical trials with Tarceva® (erlotinib) in combination with a monovalent anti-c-Met antibody in NSCLC (Spigel et al., 2011 ASCO Annual Meeting Proceedings 2011, Journal of Clinical Oncology: Chicago, IL. p. 7505) and Tarceva (erlotinib) in combination with ARQ-197, a small molecule c-Met inhibitor (Adjei et al., Oncologist, 16:788-99, 2011). Combination therapies or bispecific anti-EGFR/c-Met molecules are described, for example, in International Patent Publication Nos. No. US 2009/0042906.

Current approaches using small and high molecular weight therapeutic agents, antagonists to EGFR and/or c-Met signaling pathways, may be suboptimal due to the possible lack of specificity, potential side activity and dose-limiting toxicity that can occur with the use of small molecule inhibitors. Typical monospecific divalent antibodies can lead to clustering of membrane-bound receptors and unwanted activation of downstream signaling pathways. Monovalent antibodies having full-length heavy chains (half-arms) are highly complex and costly to manufacture.

Accordingly, there is a need for additional monospecific or bispecific EGFR and/or c-Met inhibitors for both therapeutic and diagnostic purposes.

Statement of the Invention

One embodiment of the present invention is an isolated epidermal growth factor receptor (EGFR)/hepatocyte growth factor receptor (c-Met) bispecific antibody comprising:

a first heavy chain (HC1) containing HC1 constant domain 3 (HC1 CH3) and HC1 variable region 1 (VH1);

a second heavy chain (HC2) containing the HC2 constant domain 3 (HC2 CH3) and the HC2 variable region 2 (VH2);

a first light chain (LC1) containing variable region 1 light chain (VL1); and a second light chain (LC2) containing a light chain variable region 2 (VL2), wherein VH1 and VL1 together form a first antigen-binding site that specifically binds to EGFR,

VH2 and VL2 together form a second antigen-binding site that specifically binds to c-Met,

HC1 contains at least one substitution in HC1 CH3, and HC2 contains at least one substitution in HC2 CH3, and the substitution in HC1 CH3 and the substitution in HC2 CH3 occur at different positions of amino acid residues if the numbering of the residues corresponds to the EU catalog.

In other embodiments, the present invention provides an EGFR/c-Met bispecific antibody, wherein the antibody inhibits phosphorylation of extracellular signal-associated kinases 1 and 2 (ERK1/2) in an NCI-H292, NCI-H1975, or SKMES-1 cell line with a value of an IC ₅₀ that is at least about 10 times less, at least about 20 times less, at least about 30 times less, at least about 40 times less, at least about 50 times less, or at least about 60-fold lower than the IC ₅₀ value for inhibition of ERK1/2 phosphorylation in NCI-H292, NCI-H1975, or SKMES-1 cell lines using a mixture of control monovalent anti-EGFR antibody containing heavy chain 3 (HC3) and light chain 3 (LC3), and a control monovalent antibody to c-Met containing heavy chain 4 (HC4) and light chain 4 (LC4), with HC3 and HC1, LC3 and LC1, HC4 and HC2, and also LC4 and LC2 have and identical amino acid sequences, respectively, wherein ERK1/2 phosphorylation is measured in whole cell lysates by sandwich ELISA using an anti-phospho-ERK1/2 antibody as the capture antibody and an electrochemiluminescent compound-conjugated antibody that binds to non-phosphorylated and phosphorylated ERK1/2 as detection antibody.

In other embodiments, the present invention provides an EGFR/c-Met bispecific antibody, wherein the antibody inhibits protein kinase B (AKT) phosphorylation at Ser473 in the NCI-H1975 cell line with an IC ₅₀ value that is at least about 70 times less than the IC value ₅₀ to inhibit AKT phosphorylation at Ser473 in the NCI-H1975 cell line using a mixture of a control monovalent antibody to EGFR containing HC3 and LC3 and a control monovalent antibody to c-Met containing HC4 and LC4, with HC3 and HC1, LC3 and LC1, HC4 and HC2 and LC4 and LC2 have identical amino acid sequences, respectively, with AKT phosphorylation at Ser473 measured in whole cell lysates by sandwich ELISA using an antibody that binds to non-phosphorylated and phosphorylated AKT as the capture antibody and an antibody conjugated to an electrochemiluminescent compound. to Ser473 phosphorylated AKT as a detector antibodies.

In other embodiments, the present invention provides EGFR/c-Met bispecific antibodies that bind to EGFR at SEQ ID NO: 73 at EGFR residues K489, I491, K467, and S492, and to c-Met at PEFRDSYPIKYVHAF residues (SEQ ID NO: 238) and FAQSKPDSAEPMDRSA (SEQ ID NO: 239).

In other embodiments, the present invention provides EGFR/c-Met bispecific antibodies that inhibit the growth of NCI-H292 or NCI-H1975 cells with an IC ₅₀ value that is at least about 300-fold less, at least about 400-fold less, at least about 500 times less, at least about 600 times less, at least about 700 times less, or at least about 800 times less than the IC ₅₀ value for growth inhibition of NCI-H292 or NCI- H1975 with cetuximab when NCI-H292 or NCI-H1975 cells are grown under low attachment conditions.

In other embodiments, the present invention provides EGFR/c-Met bispecific antibodies that inhibit tumor growth from HGF-expressing SKMES-1 cells in SCID Beige mice with a percentage (%) T/C at day 36 that is at least 500 times less compared to cetuximab when administered with a bispecific antibody and cetuximab at a dose of 20 mg/kg.

In other embodiments, the present invention provides EGFR/c-Met bispecific antibodies wherein HC1 CH3 contains a K409R or F405L substitution and HC2 CH3 contains a K409R or F405L substitution, with residue numbering according to the EU catalog.

In other embodiments, the present invention provides EGFR/c-Met bispecific antibodies comprising some of the CDR, VH1, VL1, VH2, VL2, HC1, LC1, HC2, and LC2 heavy and light chain sequences.

Another embodiment of the present invention is an isolated synthetic polynucleotide encoding the HC1, HC2, LC1 or LC2 of the present invention.

Another embodiment of the present invention is a vector containing a polynucleotide of the present invention.

Another embodiment of the present invention is a host cell containing the vector of the present invention.

Another embodiment of the present invention is a method for preparing an isolated EGFR/c-Met bispecific antibody comprising:

combining an isolated anti-EGFR monospecific bivalent antibody containing two heavy chains of SEQ ID NO: 199 and two light chains of SEQ ID NO: 200, and an isolated anti-c-Met monospecific bivalent antibody containing two heavy chains of SEQ ID NO: 201 and two light chains of SEQ ID NO: 202, in a mixture with a molar ratio of approximately 1:1;

introducing a reducing agent into the mixture;

incubating the mixture for about 90 minutes to about 6 hours;

removal of the reducing agent;

purification of an EGFR/c-Met bispecific antibody comprising a first heavy chain of SEQ ID NO: 199 and a second heavy chain of SEQ ID NO: 201, a first light chain of SEQ ID NO: 200 and a second light chain of SEQ ID NO: 202 , wherein the first heavy chain of SEQ ID NO: 199, together with the first light chain of SEQ ID NO: 200, forms a first binding domain that specifically binds to EGFR, and the second heavy chain of SEQ ID NO: 201, together with the second light chain of SEQ ID NO: 202 forms a second binding domain that specifically binds to c-Met.

Another embodiment of the present invention is a pharmaceutical composition comprising a bispecific antibody of the present invention and a pharmaceutically acceptable carrier.

Another embodiment of the present invention is a method of treating a subject having cancer, which comprises administering a therapeutically effective amount of an EGFR/c-Met bispecific antibody of the present invention to a patient in need of treatment for a period of time sufficient to treat the cancer.

Another embodiment of the present invention is a method for inhibiting the growth or proliferation of cells expressing EGFR and/or c-Met, which comprises contacting the cells with a bispecific antibody of the present invention.

Another embodiment of the present invention is a method of inhibiting the growth and metastasis of an EGFR and/or c-Met expressing tumor or cancer cells in a subject, which comprises administering to the subject an effective amount of a bispecific antibody of the present invention to inhibit the growth or metastasis of an EGFR and/or c-Met expressing tumors or cancer cells.

Brief description of the drawings

Fig. 1A and 1B. Alignment of the amino acid sequence of FN3 domains that bind to EGFR. Boxed loops are BC and FG at residues 22-28 and 75-86 of SEQ ID NO: 18. Some variants include temperature stability enhancing substitutions L17A, N46K and E86I (residue numbering is Tencon SEQ ID NO: 1).

Fig. 2. Sequence alignment of the Tencon27 framework (SEQ ID NO: 99) and the TCL14 library (SEQ ID NO: 100) having a randomized C-CD-F-FG alternate surface. The rest of the loop is enclosed in rectangles. Loops and strands are indicated above the sequences.

Fig. 3. Sequence alignment of FN3 domains binding to c-Met. Loop C and strand CD, as well as loop F and strand FG, are enclosed in rectangles and enclose residues 29-43 and 65-81.

Fig. 4. Inhibition of c-Met phosphorylation in NCI-H292 cells pretreated with monospecific or bispecific FN3 domain-containing molecules and stimulated with HGF. A significant increase in the potency of the EGFR/c-Met bispecific molecule (ECB1) was observed compared to the c-Met binding FN3 monospecific domain (P114AR5P74-A5, shown as element A5 in the figure) used alone or in combination with the FN3 binding domain. with EGFR (P54AR4-83v2, shown in the figure as an 83v2 element).

Fig. 5. Inhibition of EGFR and c-Met phosphorylation in cells pretreated with monospecific or bispecific molecules containing the FN3 domain. In cell lines with high levels of EGFR expression, NCI-H292 (FIG. 5A) and H596 (FIG. 5B), monospecific or bispecific anti-EGFR antibody molecules containing an FN3 domain can equally effectively reduce EGFR phosphorylation. In cell lines with low levels of EGFR expression relative to c-Met, NCI-H441 (FIG. 5C) EGFR/c-Met bispecific molecules increase the efficiency of inhibition of EGFR phosphorylation compared to the monospecific EGFR-binding FN3 domain alone. In cell lines with low levels of c-Met expression for EGFR, NCI-H292 (Fig. 5D) and H596 (Fig. 5E), inhibition of c-Met phosphorylation was significantly enhanced with the EGFR/c-Met bispecific molecule compared to the monospecific domain alone. FN3 binding to c-Met. The following molecules were used in the study: ECB5 bispecific (shown in the figure as element 17-A3), FN3 monospecific domain binding to EGFR, P53A1R5-17 (shown in the figure as element 17), EGFR/c-Met bispecific ECB3 molecule (shown in the figure as element 83-H9) and a monospecific FN3 domain that binds to c-Met, P114AR7P93-H9 (shown in the figure as element H9).

Fig. 6. Pharmacodynamic signaling in tumors isolated from mice treated with EGFR/c-Met bispecific molecules for 6 or 72 h. affinity of FN3 domains for EGFR and/or c-Met. Bispecific molecules were generated by linking an EGFR binding FN3 domain with high (shown as element 83, p54AR4-83v2) or moderate (shown as element 17v2, P53A1R5-17v2) affinity to the c-Met binding FN3 domain. , with high (shown in the figure as element A3, P114AR7P94-A3) or medium (shown in the figure as element A5, P114AR5P74-A5) affinity.

Fig. 7. Plasma (top) and tumor (bottom) accumulation of EGFR/c-Met bispecific molecules with different affinity values bound to the albumin-binding domain (ABD) shown after 6 h (left) and 72 h (right) after intraperitoneal (i/p) administration. At 6 h post-injection, tumor accumulation was maximal in mice treated with a bispecific molecule that retains an FN3 EGFR-binding domain (17v2) with medium affinity or an EGFR-binding domain (83v2) with high affinity. Bispecific molecules that included FN3 domains binding to EGFR or c-Met with medium or high affinity were as follows: 83v2-A5-ABD (ECB18; high/medium for EGFR/c-Met) 83v2-A3-ABD ( ECB38; high/high) 17v2-A5 (ECB28; medium/high) 17v2-A3-ABD (ECB39; medium/high). In FIG. 7 3v2 denotes p54AR4-83v2; 17v2 denotes p53A1R5-17v2; A3 is p114AR7P94-A3; and A5 is p114AR5P74-A5.

Fig. 8. H292-HGF tumor xenografts were implanted in SCID Beige mice. When tumors reached an average volume of approximately 80 mm ³ , mice were injected with EGFR/c-Met bispecific molecules (25 mg/kg) or vehicle PBS three times a week. All bispecific molecules reduced tumor growth, with tumor growth inhibition (TGI) dependent on the affinity values of the molecules for c-Met and EGFR (high EGFR - high c-Met corresponds to p54AR4-83v2-p114AR7P94-A3 (ECB38); high EGFR - medium c -Met corresponds to p54AR4-83v2-p114AR5P74-A5 (ECB18); average EGFR high c-Met corresponds to p53A1R5-17v2-p114AR7P94-A3 (ECB39); average EGFR - average c-Met corresponds to p53AlR5-17-p114AR5P74-A5 (ECB28) ).

Fig. 9. H292-HGF tumor xenografts were implanted in SCID Beige mice and treated with various therapies. Therapy has been shown to have antitumor activity (EGFR/c-Met bispecific molecule designates p54AR4-83v2-p114AR7P94-A3-ABD (ECB38); other therapies are crizotinib, erlotinib, cetuximab, and a combination of crizotinib and erlotinib).

Fig. 10. SKMES-HGF tumor xenografts were implanted in SCID Beige mice and various therapies were performed. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time. The EGFR/c-Met bispecific antibody EM1-mAb was administered intraperitoneally (i/p) twice a week at a dose of 20, 5, or 1 mg/kg; cetuximab was administered i/p twice a week at a dose of 20 mg/kg. The arrows in the figure indicate the days of administration. The numbers after the antibodies indicate the dose administered.

Fig. 11. HCC827 tumor xenografts were implanted in nude mice and mice were treated with erlotinib or EM1-mAb at indicated doses. EM1-mAb was administered twice a week, and erlotinib was administered once a day for four weeks. The arrows in the figure indicate the days of administration. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time.

Fig. 12. SNU-5 tumor xenografts were implanted in CB17/SCID mice and mice were treated with cetuximab 10 mg/kg or EM1-mAb 10 or 1 mg/kg. Antibodies were administered twice a week for four weeks. The arrows in the figure indicate the days of administration. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time.

Fig. 13. H1975-HGF tumor xenografts were implanted in nude mice and mice were treated with 10 mg/kg cetuximab, 10 mg/kg EM1-mAb, 50 mg/kg erlotinib, 15 mg/kg afatinib, or a combination of 10 mg/kg EM1-mAb and 15 mg/kg afatinib. Antibodies were administered twice a week, and low molecular weight compounds - once a day for three weeks. The arrows in the figure indicate the days of administration. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time.

Fig. 14. HCC827-ER1 tumor xenografts were implanted in nude mice and mice were treated with 10 mg/kg EM1-mAb, 25 mg/kg erlotinib, or a combination of the two drugs. EM1-mAb was administered twice weekly and erlotinib was administered once daily for 19 days. The arrows in the figure indicate the days of administration. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time.

Fig. 15. Mean levels of EGFR and c-Met in tumor lysates isolated from H1975 HGF tumor xenografts implanted in SCID Beige mice following a single dose of 20 mg/kg EM1-mAb. Receptor levels are shown as % relative to PBS control at the indicated time post-treatment.

Fig. 16. H1975-HGF tumor xenografts were implanted in nude mice and mice were treated with 10 mg/kg EM1-mAb or 10 mg/kg EM1-mAb IgG ₂ variant V234A/G237A/P238S/H268A/V309L/A330S/P331S, not binding to Fc receptor and has no effector functions. Antibodies were administered twice a week on the indicated days. The antitumor activity of the therapy is shown as the change in tumor size (mm ³ ) over time.

Detailed description of the invention

The term fibronectin type III (FN3) domain (FN3 domain) herein refers to a domain frequently found in proteins, including fibronectins, tenascin, intracellular cytoskeletal proteins, cytokine receptors, and prokaryotic enzymes (Bork and Doolittle, Proc. Nat. Acad. Sci. USA 89:8990-8994, 1992; Meinke et al., J. Bacteriol 175:1910-1918, 1993; Watanabe et al., J. Biol. Chem. 265:15659-15665, 1990). Examples of FN3 domains are 15 different FN3 domains present in human tenascin C, 15 different FN3 domains present in human fibronectin (FN), and non-naturally occurring synthetic FN3 domains as described, for example, in US Patent Publication No. 2010/0216708. Individual FN3 domains are designated by domain number and protein name, eg tenascin FN3 domain 3 (TN3) or fibronectin FN3 domain 10 (FN10).

The term substitution, substituted, mutation, or mutated as used herein refers to a change, deletion, or insertion of one or more amino acids or nucleotides in a polypeptide or polynucleotide sequence to create a variant of that sequence.

- 5 039356

The term randomization, randomized, diversified, or diversification as used herein refers to making at least one substitution, insertion, or deletion in the sequence of a polynucleotide or polypeptide.

The term variant as used herein refers to a polypeptide or polynucleotide that differs from a reference polypeptide or reference polynucleotide by one or more modifications, such as substitutions, insertions, or deletions.

The term specifically binds or specific binding as used herein refers to the ability of the FN3 domain, a bispecific agent that specifically binds to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention to bind to a given antigen with a dissociation constant (K _D ) about 1 x 10' ⁶ M or less, for example, about 1 x 10' ⁷ M or less, about 1 x 10' ⁸ M or less, about 1 x 10' ⁹ M or less, about 1 x 10'1° M or less, about 1 x 10-11 M or less , approximately 1x10 ¹² M or less or approximately 1x10'1 M or less. Typically, the FN3 domain, a bispecific agent that specifically binds to ECFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention binds a given antigen (i.e., EGFR or c-Met) to a KD that at least ten times less than its K _D for a non-specific antigen (eg, BSA or casein), as measured by surface plasmon resonance using, for example, a Proteon instrument (BioRad). Thus, an EGFR/c-Met bispecific molecule comprising an FN3 domain, a bispecific agent that specifically binds to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention specifically binds to each EGFR and to c-Met with a binding affinity (K _D ) of at least about 1 x 10' ⁶ M or less, for example, about 1 x 10' ⁷ M or less, about 1 x 10' ⁸ M or less, about 1 x 10' ⁹ M or less, about 1 x 10' ¹⁰ M or less, about 1 x 10 ^-11 M or less, about 1 x 10' ¹² M or less, or about 1 x 10' ¹³ M or less. However, an EGFR/c-Met bispecific molecule containing an FN3 domain, a bispecific agent that specifically binds to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention that specifically binds to a given antigen may be cross-reactive. to other related antigens, for example to the same given antigen of other species (homologues).

The term library in this document refers to a group of variants. The library may consist of polypeptide or polynucleotide variants.

The term stability as used herein refers to the ability of a molecule to maintain its folded state under physiological conditions so as to retain at least one of its normal functional activities, such as the ability to bind to a given antigen such as EGFR or c-Met.

The term epidermal growth factor receptor or EGFR as used herein refers to human EGFR, also known as HER1 or ErbB1 (Ullrich et al., Nature, 309:418425, 1984), with the amino acid sequence shown in SEQ ID NO: 73, and according to accession number NP 005219 in GenBank, as well as its naturally occurring variants. Such variants include the well-known EGFRvIII and other alternative splicing variants (e.g., defined as P00533-1 in SwissProt (wild type identical to SEQ ID NO: 73 and NP 005219), P00533-2 (F404L/L405S), P00533-3 ( 628-705:

CTGPGLEGCP...GEAPNQALLR^PGNESLKAML...SVIITASSCH and 706-1210 deletion), P00533-4 (C628S and 629-1210 deletion), variants GlnQ98, R266, K521, I674, G962 and P988 (Livingston et al., NIEHS- SNPs, Environmental Genome Project, NIEHS ES15478), T790M, L858R/T790M and del (E746, A750)).

The term EGFR ligand as used herein encompasses all (eg, physiological) ligands for EGFR, including EGF, TGFa, heparin-binding EGF (HB-EGF), amphiregulin (AR), and epiregulin (EPI).

The term epidermal growth factor (EGF) as used herein refers to the well-known 53 amino acid human EGF having the amino acid sequence shown in SEQ ID NO: 74.

The term hepatocyte growth factor receptor, or c-Met, as used herein, refers to human c-Met with the amino acid sequence shown in SEQ ID NO: 101, or GenBank Accession Number NP 001120972, and natural variants thereof.

The term hepatocyte growth factor (HGF) as used herein refers to the well-known human HGF with the amino acid sequence shown in SEQ ID NO: 102, which is cleaved to form a dimer of alpha and beta chains linked by a disulfide bond.

Interchangeably herein, the terms blocking binding or inhibiting binding refer to the ability of FN3 domains, the EGFR/c-Met bispecific molecule containing the FN3 domains of the present invention, to be a bispecific agent specifically binding to EGFR and c-Met, or to be specific to EGFR/ c-Met antibodies of the present invention block or inhibit EGFR ligand binding, eg EGF to EGFR and/or HGF to c-Met, and includes both partial and complete blocking/inhibition. Blocking/inhibition of an EGFR ligand, such as EGF with EGFR and/or HGF with c-Met, via FN3 domains or an EGFR/c-Met bispecific molecule containing FN3 domains, a bispecific agent specifically binding to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention partially or completely reduces the normal level of EGFR signaling and/or c-Met signaling compared to EGFR ligand binding to EGFR and/or HGF binding to c-Met without blocking or inhibition. An FN3 domain or an EGFR/c-Met bispecific molecule containing FN3 domains, a bispecific agent specifically binding to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention blocks the binding of an EGFR ligand, e.g., EGF to EGFR and/ or HGF with c-Met when inhibition is at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96 , 97, 98, 99 or 100%.

Binding inhibition can be measured using well-known methods, for example by measuring inhibition of biotinylated EGF binding on EGFR-expressing A431 cells that have been exposed to the FN3 domain, a bispecific molecule containing FN3 domains that bind to EGFR/c-Met, a bispecific agent that specifically binds to EGFR and c-Met, or an EGFR/c-Met bispecific antibody of the present invention, using FACS, as well as using the methods described herein, or by measuring the binding of biotinylated HGF to the c-Met extracellular domain using well-known methods and the methods described herein.

The term EGFR signaling as used herein refers to signal transduction induced by binding of an EGFR ligand to EGFR, resulting in autophosphorylation of at least one tyrosine residue in EGFR. An example of an EGFR ligand is EGF.

The term neutralizes EGFR signaling as used herein refers to the ability of FN3 domains, an EGFR/c-Met bispecific FN3-containing molecule, a bispecific EGFR and c-Met specific binding agent, or an EGFR/c-Met bispecific antibody of the present invention to inhibit EGFR- signaling induced by an EGFR ligand, such as EGF, at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100%.

The term c-Met signaling as used herein refers to signal transduction induced by binding of an HGF ligand to c-Met, resulting in autophosphorylation of at least one tyrosine residue in c-Met. Typically, at least one tyrosine residue at positions 1230, 1234, 1235, or 1349 is autophosphorylated after HGF binding.

The term neutralizes c-Met signaling as used herein refers to the ability of FN3 domains, an EGFR/c-Met bispecific FN3-containing molecule, an EGFR and c-Met specific binding agent, or an EGFR/c-Met bispecific antibody of the present invention to inhibit c-Met signaling induced by HGF by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95.96, 97, 98, 99 or 100%.

As used interchangeably herein, the terms overexpressing, overexpressing, and overexpressing refer to a cancerous or malignant cell that has measurably elevated levels of EGFR and/or c-Met on its surface compared to a normal cell from the same type of tissue. Such overexpression may be due to gene amplification or increased transcription or translation. Expression and overexpression of EGFR and/or c-Met can be measured using well known assays such as ELISA, immunofluorescence, flow cytometry or radioimmunoassay on live or lysed cells. Alternatively or additionally, the levels of nucleic acid molecules encoding EGFR and/or c-Met can be measured in the cell, for example, using fluorescence in situ hybridization, Southern blot, or PCR techniques. Overexpression of EGFR and/or c-Met means that the level of EGFR and/or c-Met on the cell surface is at least 1.5 times higher than that of a normal cell.

The term Tencon as used herein refers to a type III synthetic fibronectin (FN3) domain having the sequence shown in SEQ ID NO: 1 and described in US Patent Publication No. US 2010/0216708.

The term cancer cell or tumor cell as used herein means a cancerous, precancerous or transformed cell in vivo, ex vivo and in cell culture that has spontaneous or induced phenotypic changes that do not necessarily involve the production of new genetic material. Although transformation can cause infection with a transforming virus and the insertion of a new genomic nucleic acid or ingestion of an exogenous nucleic acid, it can also occur spontaneously or after exposure to a carcinogen, resulting in a mutation of the endogenous gene. Transformation/cancer manifests itself, for example, in morphological changes, cell immortalization, impaired growth control, foci formation, proliferation, malignancy, tumor-specific marker levels, invasiveness, tumor growth or suppression in suitable host animals such as nude mice and etc., in vitro, in vivo, and ex vivo ( ^Freshney , Culture of Animal Cells: A Manual of Basic Technique (3rd ed. 1994)).

The term vector means a polynucleotide capable of reproducing within a biological

- 7 039356 systems or which can be moved between such systems. Vector polynucleotides typically contain elements, such as origins of replication, a polyadenylation signal, or markers of choice, that function to promote duplication or retention of these polynucleotides in a biological system. Examples of such biological systems may include cell, virus, animal, plant, and reengineered biological systems using biological components capable of vector doubling. The vector-containing polynucleotide may be a DNA or RNA molecule or a hybrid thereof.

The term expression vector means a vector that can be used in a biological system or a reshaped biological system to directly translate the polypeptide encoded by the polynucleotide sequence present in the expression vector.

The term polynucleotide means a molecule containing a chain of nucleotides covalently linked through a sugar phosphate backbone or other equivalent covalent chemical structure. Double-stranded and single-stranded DNA and RNA molecules are typical examples of polynucleotides.

The term complementary DNA, or cDNA, refers to a known synthetic polynucleotide having the same arrangement of sequence elements as in native mature mRNA varieties, with contiguous exons and intervening introns removed, present in genomic DNA. The codons encoding the initial methionine may or may not be present in the cDNA. cDNA can be synthesized, for example, by reverse transcription or synthetic gene assembly.

Synthetic, or non-naturally occurring, or artificial, as used herein, refers to a polynucleotide or polypeptide molecule not naturally occurring.

The term polypeptide or protein refers to a molecule that contains at least two amino acid residues linked by a peptide bond to form a polypeptide. Small polypeptides containing less than about 50 amino acids may be referred to as peptides.

The term EGFR/c-Met bispecific molecule or EGFR/c-Met bispecific molecule containing an FN3 domain as used herein refers to a molecule containing an EGFR binding FN3 domain and a separate c-Met binding FN3 domain that are covalently linked. either directly or via a linker. An example of a bispecific molecule that binds to EGFR/c-Met contains a first FN3 domain that specifically binds to EGFR and a second FN3 domain that specifically binds to c-Met.

The term valence as used herein refers to the presence in a molecule of a specified number of antigen-specific binding sites. Thus, the terms monovalent, divalent, tetravalent, and hexavalent refer to the presence of one, two, four, and six antigen-specific binding sites, respectively, in a molecule.

The term mixture as used herein refers to a pattern or preparation of two or more FN3 domains not linked together by a covalent bond. The mixture may consist of two or more identical FN3 domains or different FN3 domains. A mixture as used herein also refers to a sample or preparation of two or more monovalent antibodies that are monovalent to EGFR and/or monovalent to c-Met.

The term bispecific agent specifically binding to EGFR and c-Met, as used herein, means a molecule containing a first domain specifically binding to EGFR and a second domain specifically binding to c-Met. An example of an agent that specifically binds to EGFR and c-Met is a bispecific antibody. Another example of a bispecific agent that specifically binds to EGFR and c-Met is a molecule containing an EGFR-binding FN3 domain and a separate FN3 domain that binds to c-Met. A bispecific agent that specifically binds to EGFR and c-Met can be formed from a single polypeptide or more than one polypeptide.

The term anti-EGFR/c-Met bispecific antibody, or EGFR/c-Met bispecific antibody, as used herein means a bispecific antibody having a first domain specifically binding to EGFR and a second domain specifically binding to c-Met. EGFR and c-Met specific binding domains are typically VH/VL pairs, and an anti-EGFR/c-Met bispecific antibody is monovalent in binding to EGFR and c-Met.

The term antibodies is intended to be used herein in a broad sense to include immunoglobulin molecules, including polyclonal antibodies, monoclonal antibodies, including murine, human, human-adapted, humanized and chimeric monoclonal antibodies, antibody fragments, bispecific or multispecific antibodies, dimeric, tetrameric or multimeric antibodies as well as single chain antibodies.

Immunoglobulins can fall into five major classes, namely IgA, IgD, IgE, IgG and IgM, depending on the amino acid sequence of the heavy chain constant domain. IgA and IgG are further classified into IgA, IgA2, IgGi, IgG2, IgG ₃ and IgG ₄ isotypes. Antibody light chains from any vertebrate species can be classified into one of two distinct types, namely kappa (k)

- 8 039356 and lambda (λ), depending on the amino acid sequences of their constant domains.

The term antibody fragments means the part of an immunoglobulin molecule that retains the heavy chain and/or light chain antigen-binding site, e.g. complementarity determining heavy chain regions (HCDR) 1, 2 and 3, complementarity determining light chain regions (LCDR) 1, 2 and 3 , a heavy chain variable region (VH) or a light chain variable region (VL). Antibody fragments include a Fab fragment, a monovalent fragment consisting of the VL, VH, CL, and CHI domains; a F(ab) ₂ fragment, a divalent fragment containing two Fab fragments linked by a disulfide bridge in the hinge region; Fd fragment consisting of VH and CHI domains; an Fv fragment consisting of the VL and VH domains of one antibody arm; a domain (dAb) fragment of an antibody (Ward et al. (1989), Nature, 341:544-546) consisting of a VH domain. The VH and VL domains can be engineered and linked together via a synthetic linker to form various types of single chain antibody constructs in which the VH/VL domains are paired intramolecularly or intermolecularly when the VH and VL domains are expressed in separate single chain antibody constructs , to form a monovalent antigen-binding site, such as a single chain Fv (scFv) or diabody; are described, for example, in PCT International Publication Nos. WO 1998/44001, WO 1988/01649, WO 1994/13804 and WO 1992/01047. These antibody fragments are prepared using techniques well known to those skilled in the art, and the fragments are screened for suitability in the same manner as for full length antibodies.

The phrase isolated antibody means an antibody or antibody fragment essentially free of other antibodies having different antigen specificity values (e.g., an isolated bispecific antibody that specifically binds to EGFR and c-Met is essentially free of antibodies that specifically bind to antigens other than human EGFR and c-Met). However, an isolated antibody that specifically binds to EGFR and c-Met may be cross-reactive with other antigens, such as human EGFR and/or c-Met orthologues such as EGFR and/or c-Met Macaca fascicularis (cynomolgus monkey). Moreover, the isolated antibody may be substantially free of other cellular material and/or chemicals.

The variable region of an antibody consists of a framework region separated by three antigen-binding sites. Antigen binding sites are defined using various terms: (i) complementarity determining regions (CDRs) - three in VH (HCDR1, HCDR2, HCDR3) and three in VL (LCDR1, LCDR2, LCDR3) - based on sequence variability (Wu and Kabat ( 1970), J. Exp. Med. 132: ^211-50 , 1970; Kabat et al Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md., 1991); (ii) hypervariable regions, HVR or HV - three in VH (H1, H2, H3) and three in VL (L1, L2, L3) - denote regions of antibody variable domains that are hypervariable in structure as defined by Chothia and Lesk (Chothia and Lesk Mol Biol 196:901-17, 1987). Other terms include IMGT-CDR (Lefranc et al., Dev. Comparat Immunol. 27:55-77, 2003) and Usable Specificity-Determining Residues (SDRUs) (Almagro Mol. Recognit. 17:132-43, 2004). The International ImMunoGeneTics (IMGT) database (http://www_imgt_org) provides a standardized numbering and definition of antigen-binding sites. The correspondence between CDR, HV and IMGT boundaries is described in Lefranc et al., Dev. Comparat Immunol. 27:55-77, 2003.

Chothia residues herein are VL and VH antibody residues numbered according to Al-Lazikani (Al-Lazikani et al., J. Mol. Biol. 273:927-48, 1997).

The framework or framework sequences are the remaining variable region sequences that are different from the defining antigen binding sites. Since antigen-binding sites, as described above, may be defined by various terms, the exact amino acid sequence of the framework depends on the definition of antigen-binding site.

A humanized antibody means an antibody in which the antigen binding sites are derived from a non-human species and the variable region frameworks are derived from human immunoglobulin sequences. Humanized antibodies may include substitutions in the framework regions, whereby the framework may not be an exact copy of the expressed human immunoglobulin or germline gene sequences.

The term human adapted antibodies or human framework adapted (HFA) antibodies means humanized antibodies adapted according to the methods described in US Patent Publication No. US 2009/0118127. Human-adapted antibodies are humanized by selecting human recipient scaffolds based on maximum CDR and FR similarities, length and sequence similarity of the CDR1 and CDR2 loops, and part of the light chain CDR3 loops.

The term human antibody means an antibody having heavy and light chain variable regions in which both the backbone and antigen-binding sites are derived from sequences of human origin. If the antibody contains a constant region, then the constant region is also derived from sequences of human origin.

A human antibody contains heavy or light chain variable regions that are derived from sequences of human origin if the antibody variable regions are derived from a system that uses human germline immunoglobulin or redistributed immunoglobulin genes. Such systems include human immunoglobulin gene libraries on phage display and transgenic non-human animals, such as mice, carrying human immunoglobulin loci as described herein. The human antibody may contain amino acid differences from the human germline or rearranged immunoglobulin sequences due, for example, to naturally occurring somatic mutations or intentional substitutions at the backbone or antigen binding sites. Typically, a human antibody is at least about 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97.98, 99, or 100% identical to the amino acid sequence encoded by the human germline or redistributed immunoglobulin gene. In some instances, the human antibody may contain consensus framework sequences derived from human framework sequence assays, for example, as described in Knappik et al., J. Mol. Biol. 296:5786, 2000), or synthetic HCDR3s included in human immunoglobulin gene libraries on phage display, for example, as described in Shi et al., J. Mol. Biol. 397:385-96, 2010 and International Patent Publication No. WO 2009/085462). Antibodies in which the antigen-binding sites are derived from a non-human species do not fit the definition of a human antibody.

The isolated humanized antibodies may be synthetic. Human antibodies, although derived from human immunoglobulin sequences, may be generated using systems such as phage display, including synthetic CDRs and/or synthetic scaffolds, or may be subjected to in vitro mutagenesis to improve the properties of the antibody, resulting in antibodies in vivo not included in the repertoire of human germline antibodies in vivo.

The term recombinant antibody as used herein includes all antibodies produced, expressed, generated, or isolated by recombinant means, such as antibodies isolated from an animal (e.g., mouse) that is transgenic or transchromosomal for human immunoglobulin genes, or from a hybridoma derived therefrom (described further). below), antibodies isolated from a host cell transformed to express the antibody, antibodies isolated from a recombinant combinatorial antibody library, and antibodies produced, expressed, generated, or isolated by any other means that involve splicing human immunoglobulin genes with other DNA sequences, or antibodies generated in vitro using Fab shoulder exchange.

The term monoclonal antibody as used herein means a preparation of antibody molecules of a single molecule composition. A monoclonal antibody composition exhibits a single binding specificity and affinity for a particular epitope or, in the case of a bispecific monoclonal antibody, dual binding specificity for two separate epitopes.

The term substantially identical as used herein means that the amino acid sequences of the two antibody variable regions being compared are identical or have minor differences. Minor differences are substitutions of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 amino acids in the antibody variable region sequence that do not adversely affect the properties of the antibody. Amino acid sequences substantially identical to those of the variable regions described herein are within the scope of the present invention. In some embodiments, sequence identity may be approximately 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or greater. Percent identity can be determined, for example, by pairwise alignment using the default settings for the AlignX module in Vector NTI v. 9.0.0 (Invitrogen, Carlsbad, California, USA). The protein sequences of the present invention can be used as a search sequence when performing searches in public or proprietary databases, for example, to identify related sequences. Examples of programs used to perform such searches are the XBLAST or BLASTP programs (http_//www_ncbi_nlm/nih_gov) or the GenomeQuest™ package (GenomeQuest, Westboro, Massachusetts, USA) using default settings.

The term epitope as used herein refers to the portion of an antigen to which an antibody specifically binds. Typically, epitopes are composed of reactive (eg, polar, nonpolar, or hydrophobic) surface groupings of residues, eg, amino acid side chains or polysaccharides, and they may have specific three-dimensional structural features as well as specific charge characteristics. An epitope may be formed from contiguous and/or non-contiguous amino acids forming a conformational space block. In the case of a non-contiguous epitope, amino acids from different parts of the linear antigen sequence come close to each other in three-dimensional space due to the folding of the protein molecule.

The term in combination with in this document means that two or more therapeutic

- 10 039356 agents can be administered to the subject in a mixture, simultaneously as separate agents or sequentially as separate agents in any order.

The numbering of amino acid residues in the constant region of the antibody in the present description is carried out in accordance with the EU catalog, as described in Kabat et al., Sequences of Proteins of Immunological Interest, 5 ^th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991), unless expressly stated otherwise.

Compositions of the present invention

The present invention provides bispecific agents that specifically bind to EGFR and c-Met. The present invention provides polypeptides and polynucleotides encoding the bispecific agents of the present invention or their complementary nucleic acids, vectors, host cells, and methods for their preparation and use.

Monospecific and bispecific for EGFR and/or c-Met containing FN3 domain binding molecules.

Monospecific to EGFR containing FN3 domain binding molecules.

The present invention provides type III fibronectin (FN3) domains that specifically bind to the epidermal growth factor receptor (EGFR) and block the binding of epidermal growth factor (EGF) to EGFR, and therefore can be widely used in therapy and diagnosis. The present invention provides polynucleotides encoding the FN3 domains of the present invention, or complementary nucleic acids thereto, vectors, host cells, and methods for their preparation and use.

The FN3 domains of the present invention bind to EGFR with high affinity, inhibit EGFR signaling, may have beneficial effects in terms of specificity and reduced side toxicity compared to small molecule EGFR inhibitors, and have improved tissue penetration compared to conventional antibody-based therapeutics.

One embodiment of the present invention is an isolated type III fibronectin (FN3) domain that specifically binds to the epidermal growth factor receptor (EGFR) and blocks the binding of epidermal growth factor (EGF) to EGFR.

The FN3 domains of the present invention can block the binding of EGF to EGFR with an IC ₅₀ value of less than about 1 x 10' ⁷ M, less than about 1 x 10' ⁸ M, less than about 1 x 10'9 M, less than about 1 x ^{10 10} M, less than about 1 x 10- 11 M or less than about 1 x 10 ^-12 M in a competitive assay using A431 cells and detecting the amount of fluorescence from bound biotinylated EGF using streptavidin phycoerythrin conjugate at 600 nm in A431 cells incubated with or without FN3 domains of the present invention. Examples of FN3 domains can block the binding of EGF to EGFR with an IC ₅₀ value ranging from about 1x10'9 to about ^1x10'7 M, for example EGFR binding FN3 domains with the amino acid sequence of SEQ ID NOs: 18-29, 107-110 or 122-137. The FN3 domains of the present invention can block the binding of EGF to EGFR by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96%, 97%, 98%, 99% or 100% compared to the binding of EGF to EGFR in the absence of the FN3 domains of the present invention using the same assay conditions.

The FN3 domains of the present invention can inhibit EGFR signaling by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75,80, 85, 90, 91, 92, 93, 94, 95, 96, 97%, 98%, 99% or 100% compared to the level of signaling in the absence of FN3 domains of the present invention using the same assay conditions.

Binding of a ligand, such as EGF to EGFR, stimulates receptor dimerization, autophosphorylation, activation of the intrinsic receptor, the cytoplasmic tyrosine kinase domain, and the initiation of multiple transduction and transactivation signaling pathways involved in the regulation of DNA synthesis (gene activation) and the passage of the cell cycle or division. Inhibition of EGFR signaling may result in inhibition of one or more downstream EGFR signaling pathways, and thus EGFR neutralization may have various effects, including inhibition of cell proliferation and differentiation, angiogenesis, cell motility, and metastasis.

EGFR signaling can be measured by various known methods, for example by measuring receptor autophosphorylation at any of the tyrosine residues Y1068, Y1148 and Y1173 (Downward et al., Nature, 311:483-5, 1984) and/or phosphorylation of natural or synthetic substrates. Phosphorylation can be detected by well known methods such as ELISA or Western blotting using an antibody specific for phosphotyrosine. Examples of assays can be found in Panek et al., J. Pharmacol. Exp. Thera. 283:1433-44, 1997 and Batley et al., Life Sci. 62:143-50, 1998, as well as in the assays described herein.

In one embodiment, the FN3 domain of the present invention inhibits EGF-induced EGFR phosphorylation at the EGFR tyrosine residue at position 1173 with an IC ₅₀ value of less than about 2.5x10'6 M, e.g., less than about ^1x10'6 M, less than about 11 039356 specifically 1 x 10 ^-7 M, less than about 1 x 10 ^-8 M, less than about 1 x 10 ^-9 M, less than about 1 x ^{10 -10} M, less than about 1 x 10 ^-11 M, or less than about

1x10 ^-12 M when measured in A431 cells using 50 ng/ml human EGF.

In one embodiment, the FN3 domain of the present invention inhibits EGF-induced EGFR phosphorylation at the EGFR tyrosine residue at position 1173 with an IC ₅₀ value of about 1.8 x 10 ^-8 to about 2.5 x 10 ^-6 M as measured in A431 cells using 50 ng/ ml of human EGF. Such examples of FN3 domains are those with the amino acid sequence of SEQ ID NOs: 18-29, 107-110, or 122-137.

In one embodiment, the FN3 domain of the present invention binds human EGFR with a dissociation constant (K _D ) of less than about 1 x 10 ^-8 M, e.g., less than about 1 x 10 ^-9 M, less than about 1 x 10 ^-10 M, less than about 1 x 10 ^-11 M, less than about 1 x 10 ^-12 M, or less than about 1 x 10 ^-13 M as determined by surface plasmon resonance or kinetic exclusion (Kinexa) methods known to those skilled in the art. In some embodiments, the FN3 domain of the present invention binds to human EGFR with a KD ranging from about 2 x ^{10 -10} to about 1 x 10 ^-8 M. The affinity of the FN3 domain for EGFR can be determined experimentally using any suitable method (see, e.g., Berzofsky , et al., Antibody-Antigen Interactions, Fundamental Immunology, Paul, WE, Ed., Raven Press: New York, NY (1984); Kuby, Janis Immunology, WH Freeman and Company: New York, NY (1992); and methods described in this document). The measured affinity of interaction in a particular FN3 domain-antigen pair may vary when measured under different conditions (eg, osmolarity, pH). Thus, measurements of affinity and other antigen-binding parameters (eg, KD, K o _n , K _off ) are preferably performed using standardized solutions of the protein backbone and antigen and a standardized buffer, such as the buffer described herein.

Examples of FN3 domains of the present invention that bind to EGFR include the FN3 domains of SEQ ID NO: 18-29, 107-110 or 122-137.

In one embodiment, the FN3 domain that specifically binds to EGFR contains an amino acid sequence at least 87% identical to the amino acid sequence of SEQ ID NO: 27.

In one embodiment, the FN3 domain that specifically binds to EGFR contains:

an FG loop containing the sequence HNVYKDTNX9RGL (SEQ ID NO: 179) or the sequence LGSYVFEHDVML (SEQ ID NO: 180), wherein X ₉ is M or I; and loop BC containing the sequence X1X ₂ X ₃ X ₄ X ₅ X ₆ X ₇ X ₈ (SEQ ID NO: 181);

wherein X ₁ is A, T, G or D;

X ₂ is A, D, Y or W;

X ₃ represents P, D or N;

X ₄ is L or absent;

X5 is D, H, R, G, Y or W;

X ₆ is G, D or A;

X ₇ is A, F, G, H or D;

X8 is Y, F or L.

The FN3 domains of the present invention, which specifically bind to EGFR and inhibit EGFR autophosphorylation, may contain as a structural element an FG loop that contains the sequence HNVYKDTNX ₉ RGL (SEQ ID NO: 179) or the sequence LGSYVFEHDVML (SEQ ID NO: 180), wherein X ₉ is M or I. Such FN3 domains may further comprise an 8 or 9 amino acid BC loop as described by the sequence X1X ₂ X ₃ X ₄ X ₅ X ₆ X ₇ X ₈ (SEQ ID NO: 181) and may inhibit autophosphorylation EGFR with an IC ₅₀ value of less than about 2.5 x 10 ^-6 M or with an IC ₅₀ value in the range of about 1.8 x 10 ^-8 to about 2.5 x 10 ^-6 M when measured in A431 cells using 50 ng/ml human EGF .

The FN3 domains of the present invention that specifically bind to EGFR and inhibit EGFR autophosphorylation further comprise the sequence

LPAPKNLVVSEVTEDSLRLSWX1X2X3X4X5X6X7X8DSFLIQYQESEKVGEAINLTVP

GSERSYDLTGLKPGTEYTVSIYGVHNVYKDTNX9RGLPLSAEFTT (SEQ ID NO: 182) or sequence

LPAPKNLVVSEVTEDSLRLSWXiX2X3X4X 5X _{& X7Xs DSFLIQYQESEKVGEAINLTVP}

GSERSYDLTGLKPGTEYTVSIYGVLGSYVFEHDVMLPLSAEFTT (SEQ ID NO: 183), wherein X1 is A, T, G, or D;

X ₂ is A, D, Y or W;

- 12 039356

X ₃ represents P, D or N;

X is L or absent;

X ₅ is D, H, R, G, Y or W;

X ₆ is G, D or A;

X7 is A, F, G, H or D;

X ₈ is Y, F or L;

X ₉ is M or I.

EGFR-binding FN3 domains can be generated and tested for their ability to inhibit EGFR autophosphorylation using known methods and the methods described herein.

Another embodiment of the present invention is an isolated FN3 domain that specifically binds to EGFR, the FN3 domain containing the sequence shown in SEQ ID NOs: 18-29, 107-110, or 122-137.

In some embodiments, the FN3 domains that bind to EGFR contain an initiator methionine (Met) linked to the N-terminus or a cysteine (Cys) linked to the C-terminus of a particular FN3 domain, for example, to facilitate expression and/or conjugation to molecules, prolonging the half-life.

Another embodiment of the present invention is an isolated type III fibronectin (FN3) domain that specifically binds to EGFR and blocks the binding of EGF to EGFR, the FN3 domain being isolated from a library developed based on the sequence of Tencon SEQ ID NO: 1.

Monospecific c-Met containing FN3 domain binding molecules.

The present invention provides fibronectin type III (FN3) domains that specifically bind to the hepatocyte growth factor (c-Met) receptor, block the binding of hepatocyte growth factor (HGF) to c-Met, and therefore can be widely used in therapy and diagnostics. The present invention provides polynucleotides encoding the FN3 domains of the present invention, or complementary nucleic acids thereto, vectors, host cells, and methods for their preparation and use.

The FN3 domains of the present invention bind to c-Met with high affinity, inhibiting c-Met signaling, and may have a beneficial effect in terms of specificity and reduced side toxicity compared to small molecular weight c-Met inhibitors and have improved tissue penetration compared to conventional therapeutics. antibody based agents. The FN3 domains of the present invention are monovalent, and therefore unwanted clustering and receptor activation that is possible with other divalent molecules is prevented.

One embodiment of the present invention is an isolated type III fibronectin (FN3) domain that specifically binds to the hepatocyte growth factor (c-Met) receptor and blocks the binding of hepatocyte growth factor (HGF) to c-Met.

The FN3 domains of the present invention can block HGF binding to c-Met with an IC ₅₀ value of less than about 1x10'7 M, less than about ^1x10'8 M, less than about ^1x10'9 M, less than about ^1x10'10 M, less than about 1x10'11 M or less than about 1x10' ¹² M in an assay to detect inhibition of binding of biotinylated HGF to c-Met-Fc fusion protein in the presence of FN3 domains of the present invention. Exemplary FN3 domains can block HGF binding to c-Met with an IC ₅₀ value ranging from about 2x10' ¹⁰ to about 6x10' ⁸ M. FN3 domains of the present invention can block HGF binding to c-Met by at least 30, 35, 40 , 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, or 100% compared to HGF binding to c-Met in the absence of FN3 domains of the present invention using the same assay conditions.

The FN3 domain of the present invention described herein can inhibit c-Met signaling by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% compared to the level of signaling in the absence of FN3 domains of the present invention using the same assay conditions.

Binding of a ligand, such as HGF to c-Met, stimulates receptor dimerization, autophosphorylation, activation of the intrinsic receptor, the cytoplasmic tyrosine kinase domain, and the initiation of multiple transduction and transactivation signaling pathways involved in the regulation of DNA synthesis (gene activation) and cell cycle progression or division. Inhibition of c-Met signaling may result in inhibition of one or more downstream c-Met signaling pathways, and thus c-Met neutralization may have various effects, including inhibition of cell proliferation and differentiation, angiogenesis, cell motility, and metastasis.

c-Met signaling can be measured by various known methods, for example by measuring

- 13 039356 receptor autophosphorylation of at least one of the tyrosine residues Y1230, Y1234, Y1235 or Y1349 and/or phosphorylation of natural or synthetic substrates. Phosphorylation can be detected, for example, with a phosphotyrosine-specific antibody, ELISA or Western blot analysis. Assays for tyrosine kinase activity are described, for example, in Panek et al., J. Pharmacol. Exp. Thera. 283:1433-44, 1997 and Batley et al., Life Sci. 62:143-50, 1998, as well as in the assays described herein.

In one embodiment, the FN3 domain of the present invention inhibits HGF-induced c-Met phosphorylation at c-Met residue at position 1349 with an IC ₅₀ value of less than about 1x10'6 M, less than about ^1x10'7 M, less than about 1x10'8 M , less than about 1x10' ⁹ M, less than about 1x10' ¹⁰ M, less than about 1x10 ^-11 M, or less than about 1x10' ¹² M when measured in NCI-H441 cells using 100 ng/ml recombinant human HGF.

In one embodiment, the FN3 domain of the present invention inhibits HGF-induced c-Met phosphorylation at the c-Met tyrosine residue at position Y1349 with an IC50 value ranging from about 4x10'9 to about 1x10'6 M as measured in NCI-H441 cells using 100 ng/ml recombinant human HGF.

In one embodiment, the FN3 domain of the present invention binds human cMet with a dissociation constant (KD) that is equal to or less than about 1x10' ⁷ M, 1x10'8 M, 1x10'9 M, 1x10' ¹⁰ M, 1x10 ^-11 M, 1x10' ¹² M, 1x10' ¹³ M, 1x10' ¹⁴ M, or 1x10' ¹⁵ M according to surface plasmon resonance or Kinexa data known to those skilled in the art. In some embodiments, the FN3 domain of the present invention binds to human c-Met with a KD ranging from about ^3x10'10 to about 5x10'8 M. The affinity of the FN3 domain for c-Met can be determined experimentally using any suitable method (see e.g. , Berzofsky, et al., Antibody-Antigen Interactions, Fundamental Immunology, Paul, WE, Ed., Raven Press: New York, NY (1984); Kuby, Janis Immunology, WH Freeman and Company: New York, NY (1992) ; and the methods described herein). The measured affinity of interaction in a particular FN3 domain-antigen pair may vary when measured under different conditions (eg, osmolarity, pH). Thus, measurements of affinity and other antigen-binding parameters (eg, KD, K _on , K _off ) are preferably performed using standardized solutions of the protein backbone and antigen and a standardized buffer, such as the buffer described herein.

Examples of FN3 domains of the present invention that bind to c-Met include FN3 domains with the amino acid sequence of SEQ ID NO: 32-49 or 111-114.

In one embodiment, the FN3 domain that specifically binds to c-Met contains an amino acid sequence at least 83% identical to the amino acid sequence of SEQ ID NO: 41.

In one embodiment, the FN3 domain that specifically binds to c-Met comprises a C strand and a CD loop containing the sequence DSFXi ₀ IRYXnE X ₁₂ X ₁₃ X ₁₄ X ₁₅ GX ₁₆ (SEQ ID NO: 184), wherein X ₁₀ is W, F or V;

X ₁₁ is D, F or L;

X ₁₂ is V, F or L;

X ₁₃ is V, L or T;

X ₁₄ is V, R, G, L, T or S;

X ₁₅ is G, S, A, T or K;

X ₁₆ is E or D; and strand F and loop FG containing the sequence TEYX ₁₇ VX ₁₈ IX ₁₉ X ₂₀ V KGGX ₂₁ X ₂₂ SX ₂₃ (SEQ ID NO: 185), wherein X ₁₇ is Y, W, I, V, G, or A;

X ₁₈ is N, T, Q or G;

X ₁₉ is L, M, N or I;

X20 is G or S;

X21 is S, L, G, Y, T, R, H, or K;

X22 is I, V or L;

X23 is V, T, H, I, P, Y, or L.

The FN3 domains of the present invention, which specifically bind to c-Met and inhibit c-Met autophosphorylation, further contain the sequence

LPAPKNLVVSRVTEDSARLSWTAPDAAF DSFXioIRYXnE X12X13X14X15GX16

AIVLTVPGSERSYDLTGLKPGTEYX17VX18IX19X20VKGGX21X22SX23PLSAEFTT (SEQ ID NO: 186), wherein X ₁₀ is W, F or V;

X11 is D, F or L;

X ₁₂ is V, F or L;

- 14 039356

X ₁₃ is V, L or T;

X ₁₄ is V, R, G, L, T or S;

X ₁₅ is G, S, A, T or K;

X ₁₆ is E or D;

X ₁₇ is Y, W, I, V, G or A;

X ₁₈ is N, T, Q or G;

X ₁₉ is L, M, N or I;

X ₂₀ is G or S;

X ₂₁ is S, L, G, Y, T, R, H, or K;

X22 is I, V or L;

X ₂₃ is V, T, H, I, P, Y, or L.

Another embodiment of the present invention is an isolated FN3 domain that specifically binds to c-Met, the FN3 domain containing the sequence shown in SEQ ID NOs: 32-49 or 111-114.

Another embodiment of the present invention is an isolated type III fibronectin (FN3) domain that specifically binds to c-Met and blocks HGF binding to c-Met, the FN3 domain being isolated from a library developed based on the sequence of Tencon SEQ ID NO: 1.

Isolation of EGFR or c-Met FN3 domains from a library based on the Tencon sequence.

Tencon (SEQ ID NO: 1) is a non-naturally occurring type III fibronectin (FN3) domain developed from a consensus sequence of fifteen human tenascin-C FN3 domains (Jacobs et al., Protein Engineering, Design, and Selection, 25 :107-117, 2012; Patent Publication No. 2010/0216708). The Tencon crystal structure shows six surface-exposed loops connecting seven beta-strands characteristic of FN3 domains, with beta-strands designated as A, B, C, D, E, F, and G, and loops designated as AB, BC, CD , DE, EF, and FG (Bork and Doolittle, Proc. Natl. Acad. Sci. USA, 89:8990-8992, 1992; US Pat. No. 6,673,901). These loops or selected residues within each loop can be randomized to construct a library of type III fibronectin (FN3) domains that can be used to select new EGFR or c-Met binding molecules. In table. 1 shows the positions and sequences of each loop and beta strand in Tencon (SEQ ID NO: 1).

Thus, a library designed from the Tencon sequence may have a randomized FG loop or randomized BC and FG loops, such as the TCL1 or TCL2 libraries described below. The Tencon BC loop is 7 amino acids long, so 1, 2, 3, 4, 5, 6, or 7 amino acids can be randomized in a Tencon sequence-based library diversified along the BC loop. The Tencon FC loop is 7 amino acids long, so a Tencon sequence library with FG loop diversification can randomize 1, 2, 3, 4, 5, 6, or 7 amino acids. Additional diversity in the loops of the Tencon libraries can be achieved by insertion and/or deletion of residues in the loops. For example, the FG and/or BC loops can be expanded by 1-22 amino acids or reduced by 1-3 amino acids. The Tencon FG loop is 7 amino acids long, while the corresponding loop in antibody heavy chains ranges from 4 to 28 residues. To maximize diversity, the FG loop can be diversified in sequence as well as in length to fit the antibody CDR3 length range of 4-28 residues. For example, the FG loop can be further diversified in length by extending the loop by an additional 1, 2, 3, 4, or 5 amino acids.

A library designed based on the Tencon sequence may also have alternative randomized surfaces formed lateral to the FN3 domain and containing two or more beta strands and at least one loop. One such alternative surface is formed by the amino acids in the C and F beta strands and in the CD and FG loops (C-CD-F-FG surface). The development of a library based on Tencon's alternative C-CD-F-FG surface is shown in FIG. 1, and the detailed creation of such libraries is described in US Patent Publication No. US 2013/0226834.

Libraries developed from the Tencon sequence also include libraries based on Tencon variants, such as Tencon variants having substitutions at residue positions 11, 14, 17, 37, 46, 73, or 86 (residue numbering according to SEQ ID NO: 1), and such variants exhibit improved thermal stability. Examples of Tencon variants are described in US Patent Publication No. 2011/0274623 and include Tencon27 (SEQ ID NO: 99) having the substitutions E11R, L17A, N46V and E86I compared to Tencon in SEQ ID NO: 1.

- 15 039356

Libraries based on Tencon and other FN3 sequences can be randomized to selected residue positions using a random or predetermined set of amino acids. For example, library variants having random substitutions can be generated using NNK codons that code for all 20 naturally occurring amino acids. In other diversification schemes, DVK codons can be used to code for the amino acids Ala, Trp, Tyr, Lys, Thr, Asn, Lys, Ser, Arg, Asp, Glu, Gly, and Cys. Alternatively, NNS codons can be used to obtain all 20 amino acid residues and simultaneously reduce the frequency of stop codons. Libraries of FN3 domains with a shifted distribution of amino acids at diversifiable positions can be synthesized, for example, using Slonomics® technology (http:_//www_sloning_com). The technology utilizes a library of prefabricated double-stranded triplets that act as versatile building blocks sufficient for thousands of gene synthesis processes. The library of triplets contains all possible combinations of sequences needed to build any desired DNA molecule. The well-known IUB code is used to designate codons.

The FN3 domains of the present invention that specifically bind to EGFR or c-Met can be isolated by preparing an FN3 library such as the Tencon library using a cis display to ligate DNA fragments encoding scaffold proteins with a DNA fragment encoding RepA to generate a pool of protein-DNA complexes formed after translation in vitro, with each protein stably associated with its encoding DNA (US patent No. 7842476; Odegrip et al., Proc. Natl. Acad. Sci. USA, 101, 2806-2810, 2004) , and analysis of the library for specific binding to EGFR and/or c-Met by any method known in the art or described in the example. Examples of well-known methods that can be used are ELISA, sandwich ELISA, and competitive and non-competitive assays (see, for example, Ausubel et al., eds, 1994, Current Protocols in Molecular Biology, Vol. 1, John Wiley & Sons , Inc., New York). The identified FN3 domains that specifically bind to EGFR or c-Met are further characterized by their ability to block EGFR ligand, e.g. EGF, binding to EGFR or HGF binding to c-Met, as well as their ability to inhibit EGFR signaling and/or c- Met using the methods described herein.

The FN3 domains of the present invention that specifically bind to EGFR or c-Met can be constructed using any FN3 domain as a template to create a library and select molecules in the library that specifically bind to EGFR or c-Met using the methods provided therein. Examples of suitable FN3 domains for use include tenascin C (TN3) FN3 3 (SEQ ID NO: 75), Fibcon (SEQ ID NO: 76) and fibronectin FN3 10 (FN10) (SEQ ID NO: 77) . Standard cloning and expression techniques are used to clone libraries into a vector or synthesize double-stranded cDNA library cassettes to express libraries or translate them in vitro. For example, ribosome display (Hanes and Pluckthun, Proc. Natl. Acad. Sci. USA, 94, 4937-4942, 1997), mRNA display (Roberts and Szostak, Proc. Natl. Acad. Sci. USA, 94, 12297-12302, 1997) or other cell free systems (US Pat. No. 5,643,768). FN3 domain variant libraries can be expressed as fusion proteins that display on the surface of, for example, any suitable bacteriophage. Methods for displaying fused polypeptides on the surface of a bacteriophage are well known (US Patent Publication No. 2011/0118144; International Patent Publication No. WO 2009/085462; US Pat. Nos. 6969108; 6172197; 5223409; 6582915; 6472147).

FN3 domains of the present invention that specifically bind to EGFR or c-Met can be modified to improve their properties, for example, improve thermal stability and reversibility of thermal folding and unfolding. To improve apparent thermal stability

- 16 039356 proteins and enzymes have been used in several ways, including rational configuration based on comparison with highly similar thermally stable sequences; directed evolution and composition of consensus sequences (Lehmann and Wyss, Curr. Opin. Biotechnol., 12,371-375, 2001). High thermal stability can increase expressed protein yield, solubility or activity, reduce immunogenicity, and minimize the need for a cold line for production. Residues that can be substituted to improve Tencon thermal stability (SEQ ID NO: 1) are residue positions 11, 14, 17, 37, 46, 73, or 86 as described in US Patent Publication No. 2011/0274623. Substitutions corresponding to these residues can be introduced into FN3 domains or into bispecific molecules of the present invention containing FN3 domains.

Another embodiment of the present invention is an isolated FN3 domain that specifically binds to EGFR and blocks the binding of EGF to EGFR, which contains the sequence shown in SEQ ID NOS: 18-29, 107-110, 122-137, which further contains substitutions in one or more residue positions corresponding to positions 11, 14, 17, 37, 46, 73 and 86 in Tencon (SEQ ID NO: 1).

Another embodiment of the present invention is an isolated FN3 domain that specifically binds to c-Met and blocks HGF binding to c-Met, comprising the sequence shown in SEQ ID NOs: 32-49, 111-114, which further comprises substitutions in one or more residue positions corresponding to positions 11, 14, 17, 37, 46, 73 and 86 in Tencon (SEQ ID NO: 1).

Examples of substitutions are substitutions E11N, E14P, L17A, E37P, N46V, G73Y and E86I (numbering according to SEQ ID NO: 1).

In some embodiments, the FN3 domains of the present invention contain substitutions corresponding to Tencon's L17A, N46V, and E86I substitutions (SEQ ID NO: 1).

The FN3 domains specifically binding to EGFR (FIG. 1) have an extended FG loop compared to Tencon (SEQ ID NO: 1). Therefore, the residues corresponding to residues 11, 14, 17, 37, 46, 73 and 86 in Tencon (SEQ ID NO: 1) are residues 11, 14, 17, 37, 46, 73 and 91 in the EGFR FN3 domains, shown in FIG. 1A and 1B, except for the FN3 domain of SEQ ID NO: 24, in which the corresponding residues are residues 11, 14, 17, 38, 74 and 92 due to the insertion of one amino acid into the BC loop.

Another embodiment of the present invention is an isolated FN3 domain that specifically binds to EGFR and blocks the binding of EGF to EGFR, which contains the amino acid sequence shown in SEQ ID NO: 18-29,107-110 or 122-137, optionally having substitutions corresponding to substitutions L17A , N46V and E86I at Tencon (SEQ ID NO: 1).

In another embodiment, the present invention provides an isolated FN3 domain that specifically binds to c-Met and blocks the binding of HGF to c-Met, comprising the amino acid sequence shown in SEQ ID NOs: 32-49 or 111-114, which optionally has substitutions, corresponding to Tencon's replacements L17A, N46V and E86I (SEQ ID NO: 1).

The measurement of protein stability and protein lability can be considered as one or different aspects of protein integrity. Proteins are sensitive or labile to denaturation caused by heat, ultraviolet or ionizing radiation, changes in environmental osmolarity and pH when in liquid solution, mechanical shear caused by filtration through small pores, ultraviolet radiation, ionizing radiation such as gamma irradiation, chemical or thermal dehydration or any other impact or force that can lead to the destruction of the protein structure. The stability of the molecule can be determined using standard methods. For example, the stability of a molecule can be determined by measuring the thermal melting point (TM) - the temperature in degrees Celsius (°C) at which half of the molecules unfold using standard methods. As a rule, the higher the TM, the more stable the molecule is. The ability of a protein to maintain a particular three-dimensional structure can be changed, in addition to heating, by changing the chemical environment.

In one embodiment, the FN3 domains of the present invention that bind to EGFR or c-Met show a stability increased by at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65%, 70%, 75%, 80%, 85%, 90%, or 95% or more of the same pre-design domain, as measured by the increase in TM.

Chemical denaturation can similarly be measured in a variety of ways. Denaturing chemicals include guanidinium hydrochloride, guanidinium thiocyanate, urea, acetone, organic solvents (DMF, benzene, acetonitrile), salts (ammonium sulfate, lithium bromide, lithium chloride, sodium bromide, calcium chloride, sodium chloride); reducing agents (e.g. dithiothreitol, beta-mercaptoethanol, dinitrothiobenzene as well as hydrides such as sodium borohydride), non-ionic and ionic detergents, acids (e.g. hydrochloric acid (HCl), acetic acid

- 17 039356 (CH3COOH), halogenated acetic acids), hydrophobic molecules (eg phospholipids) and directed denaturants. Quantification of the degree of denaturation can be based on the loss of a functional property, such as the ability to bind to a target molecule, or on physicochemical properties, such as a tendency to aggregate, access to residues previously inaccessible to the solvent, or on the destruction or formation of disulfide bonds.

In one embodiment, the FN3 domain of the present invention binding to EGFR or c-Met exhibits a stability increased by at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65 , 70, 75, 80, 85, 90, or 95% or more compared to the same scaffold prior to construction, as measured by using guanidinium hydrochloride as the chemical denaturant. The increase in stability can be measured as a function of the decrease in tryptophan fluorescence upon treatment with increasing concentrations of guanidinium hydrochloride using well known methods.

The FN3 domains of the present invention can be designed as monomers, dimers, or multimers, for example, as a means of increasing valency and thus avidity to bind to a target molecule, or to create bi- or multispecific scaffolds that simultaneously bind two or more different target molecules. Dimers and multimers can be created by linking monospecific, bi- or multispecific protein scaffolds, for example by including an amino acid linker, for example a linker containing polyglycine, glycine and serine or alanine and proline. Examples of linkers include (GS) ₂ , (SEQ ID NO: 78), (GGGGS) ₅ (SEQ ID NO: 79), (AP) 2 (SEQ ID NO: 80), (AP) ₅ (SEQ ID NO: 81 ), (AP) ₁₀ (SEQ ID NO: 82), (AP) ₂₀ (SEQ ID NO: 83) and A(EAAAK) ₅ AAA (SEQ ID NO: 84) linkers. Dimers and multimers can be connected to each other in the direction from the N-terminus to the C-terminus. The use of naturally occurring as well as artificial peptide linkers to link polypeptides into novel linked fused polypeptides is known in the literature (Hallewell et al., J. Biol. Chem. 264, 5260-5268, 1989; Alfthan et al., Protein Eng. 8, 725-731, 1995; Robinson & Sauer, Biochemistry, 35, 109-116, 1996; U.S. Patent No. 5,856,456).

Bispecific agents that specifically bind to EGFR and c-Met.

Bispecific agents of the present invention that specifically bind to EGFR and c-Met may provide an advantage in terms of specificity and reduced side toxicity compared to small molecule inhibitors of EGFR and/or c-Met. The present invention is based at least in part on the unexpected finding that bispecific agents specifically binding to EGFR and c-Met provide a significantly increased synergistic inhibitory effect compared to a mixture of monospecific agents binding to EGFR and c-Met. Molecules can be tuned to a specific affinity for both EGFR and c-Met to maximize tumor penetration and retention. Bispecific agents that specifically bind to EGFR and c-Met provide more effective inhibition of EGFR and/or c-Met signaling pathways and more effective tumor growth inhibition than cetuximab (Erbitux®).

Bispecific agents specifically binding to EGFR and c-Met can be formed by any polypeptide or multimeric polypeptide that contains an EGFR binding domain and a c-Met binding domain. EGFR and c-Met binding domains can be antibody antigen-binding sites, an antibody VH/VL pair, or another type of binding molecule, e.g. any combination of them.

EGFR and c-Met binding polypeptides can be derived from existing monospecific EGFR and c-Met binding polypeptides or can be isolated de novo.

FN3 domain containing EGFR/c-Met bispecific molecules.

One embodiment of the present invention is an isolated FN3 domain-containing bispecific molecule that comprises a first type III fibronectin (FN3) domain and a second FN3 domain, wherein the first FN3 domain specifically binds to epidermal growth factor receptor (EGFR) and blocks epidermal growth factor binding (EGF) to EGFR, and the second FN3 domain specifically binds to the hepatocyte growth factor receptor (c-Met) and blocks the binding of hepatocyte growth factor (HGF) to c-Met.

EGFR/c-Met bispecific molecules of the present invention containing an FN3 domain can be generated by covalently linking any FN3 domain that binds to EGFR and any FN3 domain of the present invention that binds to c-Met, either directly or through a linker. Therefore, the first FN3 domain of the bispecific molecule may have the characteristics described above for EGFR binding FN3 domains, and the second FN3 domain of the bispecific molecule may have the characteristics described above for c-Met binding FN3 domains.

In one embodiment, the first FN3 domain of the EGFR/c-Met bispecific molecule containing the FN3 domain inhibits EGF-induced EGFR tyrosine phosphorylation.

- 18 039356 an EGFR residue at position 1173 with an IC ₅₀ value of less than about 2.5x10 ^-6 M when measured in A431 cells using 50 ng/ml human EGF, and the second FN3 domain is bispecific to

EGFR/c-Met of an FN3 domain-containing molecule inhibits HGF-induced cMet phosphorylation at the c-Met tyrosine residue at position 1349 with an IC ₅₀ value of less than about

1.5x10 ^-6 M when measured in NCI-H441 cells using 100 ng/ml human HGF.

In another embodiment, the first FN3 domain of an EGFR/c-Met bispecific molecule comprising an FN3 domain inhibits EGF-induced EGFR phosphorylation at the EGFR tyrosine residue at position 1173 with an IC ₅₀ value ranging from about 1.8x10-8 to about 2.5x10 ^-6 M as measured in A431 cells using 50 ng/ml human EGF, and the second FN3 domain of the EGFR/c-Met bispecific molecule containing the FN3 domain inhibits HGF-induced c-Met phosphorylation at the c-Met tyrosine residue at position 1349 with an IC ₅₀ value ranging from about 4x10 ^-9 to about 1.5x10 ^-6 M when measured in NCI-H441 cells using 100 ng/ml human HGF.

In another embodiment, the first FN3 domain of the EGFR/c-Met bispecific molecule containing the FN3 domain binds to human EGFR with a dissociation constant (K _D ) of less than about 1 x 10 ^-8 M and the second FN3 domain of the EGFR/c-Met bispecific a molecule containing an FN3 domain binds to human c-Met with a K _D of less than about 5 x 10 ^-8 M.

In a bispecific molecule that binds to both EGFR and c-Met, the first FN3 domain binds to human EGFR with a K _D ranging from about 2x10 ^-10 to about 1x10 ^-8 M, and the second FN3 domain binds to human c-Met with K _D in the range from about 3x10 ^-10 to about 5x10 ^-8 M.

The affinity of the EGFR/c-Met bispecific molecule for EGFR and c-Met can be determined as described above for monospecific molecules.

The first FN3 domain in the EGFR/c-Met bispecific molecule of the present invention can block the binding of EGF to EGFR with an IC ₅₀ value ranging from about 1 x 10 ^-9 to about 1.5 x 10 ^-7 M in an assay using A431 cells and with fluorescence value detection from bound biotinylated EGF using a streptavidin-phycoerythrin conjugate at 600 nm in A431 cells incubated with or without the first FN3 domain. The first FN3 domain of the EGFR/c-Met bispecific molecule of the present invention can block the binding of EGF to EGFR by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99%, or 100% compared to EGF binding to EGFR in the absence of FN3 first domains using the same assay conditions.

The second FN3 domain of the EGFR/c-Met bispecific molecule of the present invention can block the binding of HGF to c-Met with an IC ₅₀ value ranging from about 2 x ^{10 -10} to about 6 x 10 ^-8 M in an assay for detecting inhibition of biotinylated HGF binding to c fusion protein -Met-Fc in the presence of a second FN3 domain. The second FN3 domain of the EGFR/c-Met bispecific molecule can block HGF binding to c-Met by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, or 100% compared to HGF binding to c-Met in the absence of a second FN3 domain using the same assay conditions.

The EGFR/c-Met bispecific molecule of the present invention can inhibit EGFR and/or c-Met signaling by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95 .

EGFR and c-Met signaling can be measured using a variety of well known methods as described above for monospecific molecules.

The EGFR/c-Met bispecific molecules of the present invention containing a first FN3 domain specifically binding to EGFR and a second FN3 domain specifically binding to c-Met provide significantly greater synergistic inhibition of EGFR and c-Met signaling and tumor cell proliferation along compared with the synergistic inhibition observed in the case of a mixture of the first and second domain of FN3. Synergistic inhibition can be analyzed, for example, by measuring inhibition of ERK phosphorylation by EGFR/c-Met bispecific FN3 domain-containing molecules and by a mixture of two monospecific molecules, one of which binds to EGFR and the other to c-Met. The EGFR/c-Met bispecific molecules of the present invention can inhibit ERK phosphorylation with an IC50 value that is at least about 100 times less, such as at least 500, 1000, 5000, or 10,000 times less than the IC50 value of the mixture. two monospecific FN3 domains, indicating at least 100-fold higher potency of EGFR/c-Met bispecific molecules containing an FN3 domain than a mixture of two monospecific FN3 domains. Examples of EGFR/c-Met bispecific molecules containing an FN3 domain can inhibit ERK phosphorylation with an IC ₅₀ value of approximately 5x10 ^-9 M or less. ERK phosphorylation can be measured using standard methods and the methods described herein.

The EGFR/c-Met bispecific molecule of the present invention containing an FN3 domain can inhibit the proliferation of NCI-H292 cells with an IC50 value that is at least 30 times lower than the IC50 value of NCI- _H292 cell growth inhibition by a mixture of the first FN3 domain and a second FN3 domain, with cell proliferation induced by a medium containing 10% FBS supplemented with 7.5 ng/ml HGF. The bispecific molecule of the present invention can inhibit the proliferation of tumor cells with an IC ₅₀ value that is about 30, 40, 50, 60, 70, 80, 90, 100, 150, 200, 300, 400, 500, 600, 700, 800, or about 1000 times less than the IC ₅₀ value of inhibition of tumor cell proliferation by a mixture of the first FN3 domain and the second FN3 domain. Inhibition of tumor cell proliferation can be measured using standard methods and the methods described herein.

Another embodiment of the present invention is an FN3 domain-containing bispecific molecule that contains a first type III fibronectin (FN3) domain and a second FN3 domain, the first FN3 domain specifically binding to the epidermal growth factor receptor (EGFR) and blocking the binding of epidermal growth factor ( EGF) with EGFR, and the second FN3 domain specifically binds to the hepatocyte growth factor receptor (c-Met) and blocks the binding of hepatocyte growth factor (HGF) to c-Met, and the first FN3 domain contains:

an FG loop containing the sequence HNVYKDTNX9RGL (SEQ ID NO: 179) or the sequence LGSYVFEHDVML (SEQ ID NO: 180), wherein X ₉ is M or I; and an BC loop containing the sequence X1X2X3X4X5X6X7X8 (SEQ ID NO: 181), where X1 is A, T, G or D;

X ₂ represents A, D, Y or W;

X ₃ represents P, D or N;

X ₄ is L or absent;

X ₅ is D, H, R, G, Y or W;

X ₆ is G, D or A;

X ₇ is A, F, G, H or D;

X ₈ is Y, F or L; and the second FN3 domain contains:

strand C and a CD loop containing the sequence DSFX ₁₀ IRYX ₁₁ EX ₁₂ X ₁₃ X ₁₄ X ₁₅ GX ₁₆ (SEQ ID NO: 184), wherein X10 is W, F or V;

X11 is D, F or L;

X ₁₂ is V, F or L;

X13 is V, L or T;

X ₁₄ is V, R, G, L, T or S;

X ₁₅ is G, S, A, T or K;

X ₁₆ is E or D; and strand F and loop FG containing the sequence TEYX ₁₇ VX ₁₈ IX ₁₉ X ₂₀ VKGGX ₂₁ X ₂₂ SX ₂₃ (SEQ ID NO: 185), where X ₁₇ is Y, W, I, V, G, or A;

X ₁₈ is N, T, Q or G;

X ₁₉ is L, M, N or I;

X ₂₀ is G or S;

X21 is S, L, G, Y, T, R, H, or K;

X ₂₂ represents I, V or L; and

X ₂₃ is V, T, H, I, P, Y, or L.

In another embodiment, the bispecific molecule contains a first EGFR-binding FN3 domain that contains the sequence:

LPAPKNLVVSEVTEDSLRLSWX1X2X3X4X5X6X7X8DSFLIQYQESEKVGEAINLTVP

GSERSYDLTGLKPGTEYTVSIYGVHNVYKDTNX9RGLPLSAEFTT (SEQ ID NO: 182) or sequence

LPAPKNLVVSEVTEDSLRLSWX^XsXiXsXoXvXsDSFLIQYQESEKVGEAINLTVP

GSERSYDLTGLKPGTEYTVSIYGVLGSYVFEHDVMLPLSAEFTT (SEQ ID NO: 183), wherein in SEQ ID NO: 182 and 183:

X ₁ represents A, T, G or D;

X2 is A, D, Y or W;

X ₃ represents P, D or N;

- 20 039356

X ₄ is L or absent;

X ₅ is D, H, R, G, Y or W;

X ₆ is G, D or A;

X7 is A, F, G, H or D;

X ₈ is Y, F or L;

X ₉ is M or I.

In another embodiment, the bispecific molecule contains a c-Met binding FN3 domain that contains the sequence

LPAPKNLVVSRVTEDSARLSWTAPDAAF DSFXioIRYXnE X12X13X14X15GX16

AIVLTVPGSERSYDLTGLKPG TEUXPUX^EXzeXgoUKOOXgAggZXgz PLSAEFTT (SEQ ID

NO: 186), wherein X ₁₀ is W, F or V;

X ₁₁ is D, F or L;

X ₁₂ is V, F or L;

X ₁₃ is V, L or T;

X ₁₄ is V, R, G, L, T or S;

X ₁₅ is G, S, A, T or K;

X ₁₆ is E or D;

X ₁₇ is Y, W, I, V, G or A;

X18 is N, T, Q or G;

X ₁₉ is L, M, N or I;

X ₂₀ is G or S;

X21 is S, L, G, Y, T, R, H, or K;

X22 is I, V or L;

X ₂₃ is V, T, H, I, P, Y, T or L.

Examples of EGFR/c-Met bispecific molecules containing FN3 domains comprise the amino acid sequence shown in SEQ ID NOs: 50-72, 106, 118-121, or 138-167.

The EGFR/c-Met bispecific molecules of the present invention have some structural characteristics related to their functional characteristics, such as inhibition of EGFR autophosphorylation, for example, the FG loop of the first FN3 domain binding to EGFR, which contains the sequence HNVYKDTNX9RGL (SEQ ID NO: 179) or the sequence LGSYVFEHDVML (SEQ ID NO: 180) wherein X ₉ is M or I.

In one embodiment, the EGFR/c-Met bispecific molecules of the present invention containing an FN3 domain:

inhibit EGF-induced EGFR phosphorylation at the EGFR tyrosine residue at position 1173 with an IC ₅₀ value of less than about 8 x 10' ⁷ M as measured in H292 cells using 50 ng/ml human EGF;

inhibit HGF-induced c-Met phosphorylation at the c-Met tyrosine residue at position 1349 with an IC ₅₀ value of less than about 8.4x10'7 M as measured in NCI-H441 cells using 100 ng/ml human HGF;

inhibit HGF-induced proliferation of NCI-H292 cells with an IC ₅₀ value of less than about 9.5x10' ⁶ M, wherein cell proliferation is induced with 10% FBS containing 7.5 ng HGF;

bind to EGFR with a K _D of less than about 2.0 x 10' ⁸ M, or bind to c-Met with a K _D of less than about 2.0 x 10' ⁸ M.

In another embodiment, EGFR/c-Met bispecific molecules of the present invention containing an FN3 domain:

inhibit EGF-induced EGFR phosphorylation at the EGFR tyrosine residue at position 1173 with an IC ₅₀ value ranging from about ^4.2x10'9 to 8x10'7 M as measured in H292 cells using 50 ng/ml human EGF;

inhibit HGF-induced c-Met phosphorylation at the c-Met tyrosine residue at position 1349 with an IC ₅₀ value ranging from approximately ^2.4x10'8 to 8.4x10'7 M as measured in NCI-H441 cells using 100 ng/mL human HGF;

inhibit HGF-induced proliferation of NCI-H292 cells with an IC50 value in the range from approximately 2.3x10' ⁸ to 9.5x10' ⁶ M, and cell proliferation is induced with 10% FBS containing 7.5 ng of HGF;

bind to EGFR with a K _D of about 2x10' ¹⁰ to about 2.0x10' ⁸ M, or bind to c-Met with a K _D of about 1x10' ⁹ to about 2.0x10' ⁸ M.

In one embodiment, the EGFR/c-Met bispecific molecules comprise an EGFR-binding FN3 domain containing the sequence

- 21 039356

LPAPKNLVVSEVTEDSLRLSWX1X2X3X4X5X6X7X8DSFLIQYQESEKVGEAINLTVP

GSERSYDLTGLKPGTEYTVSIYGV HNVYKDTNX9RGL PLSAEFTT (SEQ ID NO: 182), wherein X ₁ is D;

X ₂ is D;

X ₃ is P;

X ₄ missing;

X ₅ represents H or W;

X ₆ is A;

X ₇ is F;

X ₈ is Y;

X ₉ is M or I; and a c-Met binding FN3 domain with the sequence

LPAPKNLVVSRVTEDSARLSWTAPDAAF DSFXioIRYXnE X12X13X14X15GX16

AIVLTVPGSERSYDLTGLKPG TEYX17VX18IX19X20VKGGX21X22SX23 PLSAEFTT (SEQ ID NO: 186), wherein X ₁₀ is W;

X ₁₁ is F;

X12 is F;

X ₁₃ is V or L;

X ₁₄ is G or S;

X ₁₅ is S or K;

X ₁₆ is E or D;

X ₁₇ is V;

X ₁₈ is N;

X ₁₉ is L or M;

X20 is G or S;

X21 is S or K;

X ₂₂ is I;

X ₂₃ is R.

Examples of EGFR/c-Met bispecific molecules are those having the sequence shown in SEQ ID NOS: 57, 61, 62, 63, 64, 65, 66, 67 and 68.

The bispecific molecules of the present invention may further comprise substitutions at one or more residue positions in the first FN3 domain and/or the second FN3 domain corresponding to positions 11, 14, 17,37,46,73 and 86 in Tencon (SEQ ID NO: 1), as described above and a substitution at position 29. Examples of substitutions are substitutions E11N, E14P, L17A, E37P, N46V, G73Y, E86I and D29E (numbering according to SEQ ID NO: 1). One skilled in the art will recognize that other amino acids may be used as substitutions, for example, amino acids from the family of amino acids having related side chains, as described below. Generated variants can be tested for their stability and binding to EGFR and/or c-Met using the methods described herein.

In one embodiment, the EGFR/c-Met bispecific FN3 domain-containing molecule comprises a first FN3 domain specifically binding to EGFR and a second FN3 domain specifically binding to c-Met, wherein the first FN3 domain contains the sequence

LPAPKNLWSX24VTX25DSX26RLSWDDPX27AFYX28SFLIQYQX29SEKVGEAIX30LTVPGSERSYDL

TGLKPGTEYTVSIYX31VHNVYKDTNX32RGLPLSAX33FTT (SEQ ID NO: 187), wherein X ₂₄ is E, N or R;

X ₂₅ is E or P;

X ₂₆ is L or A;

X27 is H or W;

X ₂₈ is E or D;

X29 is E or P;

X ₃₀ is N or V;

X31 is G or Y;

X ₃₂ is M or I;

X ₃₃ represents E or I; and the second FN3 domain contains the sequence

LPAPKNLVVSX34VTX3 ₅ DSX36RLSWTAPDAAFDSFWIRYFX37FX38X39X4oGX4iAIX ₄ 2LTVPGSERSY

DLTGLKPGTEYVVNIX43X44VKGGX45ISPPLSAX46FTT (SEQ ID NO: 188);

moreover, X ₃₄ represents E, N or R;

X ₃₅ is E or P;

- 22 039356

X ₃₆ is L or A;

X ₃₇ is E or P;

X ₃₈ is V or L;

X ₃₉ is G or S;

X ₄₀ is S or K;

X ₄₁ represents E or D;

X ₄₂ is N or V;

X ₄₃ is L or M;

X44 is G or S;

X4 ₅ is S or K;

X ₄₆ is E or I.

In other embodiments, the EGFR/c-Met bispecific FN3 domain-containing molecule comprises a first FN3 domain that contains an amino acid sequence of at least 87, 88, 89, 90, 91, 92, 93, 94, 95, 96 , 97, 98, or 99% identical to the amino acid sequence of SEQ ID NO: 27, and a second FN3 domain that contains an amino acid sequence of at least 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identical to the amino acid sequence of SEQ ID NO: 41.

The EGFR/c-Met bispecific molecules of the present invention containing the FN3 domain can be designed with specific affinity for EGFR and c-Met to maximize tumor uptake.

Another embodiment of the present invention is an isolated FN3 domain-containing bispecific molecule that contains a first type III fibronectin (FN3) domain and a second FN3 domain, wherein the first FN3 domain specifically binds to the epidermal growth factor receptor (EGFR) and blocks epidermal growth factor binding (EGF) with EGFR, and the second FN3 domain specifically binds to the hepatocyte growth factor receptor (c-Met) and blocks the binding of hepatocyte growth factor (HGF) to c-Met, and the first FN3 domain and the second FN3 domain are isolated from a library developed on based on the Tencon SEQ ID NO: 1 sequence.

The EGFR/c-Met bispecific molecule of the present invention containing an FN3 domain can be generated by covalently linking the EGFR binding FN3 domain and the c-Met binding FN3 domain of the present invention using known methods. The FN3 domains can be connected via a linker, for example a linker containing polyglycine, glycine and serine or alanine and proline. Examples of linkers include (GS) ₂ , (SEQ ID NO: 78), (GGGGS) ₅ (SEQ ID NO: 79), (AP) ₂ (SEQ ID NO: 80), (AP) 5 (SEQ ID NO: 81 ), (AP)u (SEQ ID NO: 82), (AP)20 (SEQ ID NO: 83), A(EAAAK)5AAA (SEQ ID NO: 84) linkers. The use of naturally occurring as well as artificial peptide linkers to link polypeptides into novel linked fused polypeptides is known in the literature (Hallewell et al., J. Biol. Chem. 264, 5260-5268, 1989; Alfthan et al., Protein Eng. 8,725-731, 1995; Robinson & Sauer, Biochemistry, 35, 109-116, 1996; U.S. Patent No. 5,856,456). The EGFR/c-Met bispecific molecules of the present invention can be joined together by the C-terminus of the first FN3 domain to the N-terminus of the second FN3 domain, or by the C-terminus of the second FN3 domain to the N-terminus of the first FN3 domain. Any FN3 domain that binds to EGFR can be covalently linked to a FN3 domain that binds to c-Met. Examples of EGFR binding FN3 domains are domains having the amino acid sequence shown in SEQ ID NOs: 18-29, 107-110 or 122-137, and examples of c-Met binding FN3 domains are domains having the amino acid sequence shown in SEQ ID NOs: 32-49 and 111-114. The EGFR-binding FN3 domains to be linked to the bispecific molecule may additionally contain a methionine (Met) initiator at their N-terminus.

Variants of EGFR/c-Met bispecific FN3 domain-containing molecules are within the scope of the present invention. For example, an EGFR/c-Met bispecific molecule containing an FN3 domain can be substituted if the resulting variant retains similar selectivity and potency for EGFR and c-Met compared to the parent molecule. Examples of modifications are, for example, conservative substitutions that result in variants with characteristics that are similar to those of the parent molecule. Conservative are substitutions made within a family of amino acids with a related structure of their side chains. Genetically encoded amino acids can be divided into four families: (1) acidic (aspartate, glutamate); (2) basic (lysine, arginine, histidine); (3) non-polar (alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan) and (4) uncharged polar (glycine, asparagine, glutamine, cysteine, serine, threonine, tyrosine). Phenylalanine, tryptophan and tyrosine are sometimes combined into one group of aromatic amino acids. Alternatively, the set of amino acids can be divided into the following groups: (1) acidic (aspartate, glutamate); (2) basic (lysine, arginine, histidine); (3) aliphatic (glycine, alanine, valine, leucine, isoleucine, serine, threonine), wherein serine and threonine may optionally be combined into a separate group of aliphatic hydroxyl amino acids; (4) aromatic (phenylalanine, tyrosine, tryptophan); (5) amide (asparagine, glutamine) and (6) sulfur-containing

- 23 ⁰³⁹³⁵⁶ (cysteine and methionine) (Stryer (ed.), Biochemistry, 2nd ed, WH Freeman and Co., 1981). It is possible to introduce non-conservative substitutions into the EGFR/c-Met bispecific molecule containing the FN3 domain, which include substitutions of amino acid residues with other classes of amino acids to improve the properties of the bispecific molecules. To quickly determine whether a functional homolog can be obtained by changing the amino acid sequence of a polypeptide or fragment thereof, the ability of a modified polypeptide or fragment to elicit a response similar to an unmodified polypeptide or fragment can be assayed using the assays described herein. Peptides, polypeptides or proteins with more than one substitution can also be easily tested in a similar manner.

The EGFR/c-Met bispecific molecules of the present invention containing an FN3 domain can be designed, for example, in the form of dimers or multimers as a means of increasing valence and thus avidity of binding to the target molecule. Multimers can be created by joining one or more EGFR binding FN3 domains and one or more c-Met binding FN3 domains to form molecules containing at least three separate FN3 domains that are at least bispecific to either EGFR , or to c-Met, for example, by incorporating an amino acid linker using known methods.

Another embodiment of the present invention is an FN3 domain-containing bispecific molecule that contains a first type III fibronectin (FN3) domain and a second FN3 domain, the first FN3 domain specifically binding to the epidermal growth factor receptor (EGFR) and blocking the binding of epidermal growth factor ( EGF) with EGFR, and the second FN3 domain specifically binds to the hepatocyte growth factor receptor (c-Met) and blocks the binding of hepatocyte growth factor (HGF) to c-Met, and the molecule contains the amino acid sequence shown in SEQ ID NO: 50-72 , 106 or 138-165.

Fragments that prolong the half-life.

EGFR/c-Met bispecific molecules containing an FN3 domain, or monospecific FN3 domains of the present invention binding to EGFR or c-Met, may include other subunits, for example, through covalent interaction. In one aspect of the present invention, the EGFR/c-Met bispecific molecules of the present invention comprising an FN3 domain further comprise a half-life prolonging moiety. Examples of half-life-prolonging fragments are albumin, albumin variants, albumin-binding proteins and/or domains, transferrin and fragments and analogs thereof, and Fc regions. An example of an albumin binding domain is shown in SEQ ID NO: 117.

All or part of an antibody constant region can be attached to the molecules of the present invention to give them antibody-like properties, especially those properties associated with the Fc region, such as Fc effector functions such as binding to Clq, complement dependent cytotoxicity (CDC), binding to Fc receptor, antibody-dependent cell-mediated cytotoxicity (ADCC), phagocytosis, suppression of inhibition of cell surface receptors (e.g., B-cell receptor; BCR), and these can be further modified by modifying the residues that in Fc are responsible for these functions (see. review in Strohl, Curr Opin Biotechnol 20, 685-691, 2009).

Additional moieties can be incorporated into the bispecific molecules of the present invention, such as polyethylene glycol (PEG) molecules such as PEG-5000 or PEG-20000, fatty acids and fatty acid esters of various chain lengths, e.g. laurate, myristate, stearate, arachidate, behenate , oleate, arachidonate, octandioic acid, tetradecanedioic acid, octadecanedioic acid, docosandioic acid and the like, polylysine, octane, carbohydrates (dextran, cellulose, oligo- or polysaccharides) to obtain the desired properties. These fragments may be direct fusions to scaffold-encoding sequences and may be generated using standard cloning and expression techniques. Alternatively, well-known chemical coupling techniques can be used to attach fragments to the recombinantly formed molecules of the present invention.

For example, a legal moiety can be added to the bispecific or monospecific molecules of the present invention by adding a cysteine residue to the C-terminus of the molecule and attaching a pegyl group to the cysteine using well known methods. Examples of C-terminal cysteine bispecific molecules are those with the amino acid sequence shown in SEQ IN NO:170-178.

Monospecific and bispecific molecules of the present invention, which include additional fragments, can be compared for functionality using several known assays. For example, changing the properties of monospecific and/or bispecific molecules due to the incorporation of Fc domains and/or Fc domain variants can be analyzed in Fc receptor binding assays using soluble receptor forms such as FcyRI, FcyRII, FcyRIII, or FcRn receptors, or using known cellular assays to measure e.g. ADCC

- 24 039356 or CDC, or evaluation of the pharmacokinetic properties of the molecules of the present invention in in vivo models.

Polynucleotides, vectors, host cells.

The present invention provides nucleic acids encoding FN3 domains that bind to EGFR or c-Met, or EGFR/c-Met bispecific molecules of the present invention containing an FN3 domain, as isolated polynucleotides, or as parts of expression vectors, or as parts of linear DNA sequences, including linear DNA sequences used for transcription/translation in vitro, vectors compatible with the expression, secretion and/or presentation of the compositions or the results of their directed mutagenesis in prokaryotes, eukaryotes or filamentous phages. Some examples of polynucleotides are described herein, but other polynucleotides that, given the degeneracy of the genetic code or codon preference in a given expression system, encode EGFR or cMet binding FN3 domains or EGFR/c-Met bispecific molecules of the present invention, containing an FN3 domain are also within the scope of the present invention.

One embodiment of the present invention is an isolated polynucleotide encoding an FN3 domain specifically binding to EGFR having the amino acid sequence of SEQ ID NOs: 18-29, 107-110, or 122-137.

One embodiment of the present invention is an isolated polynucleotide comprising the polynucleotide sequence of SEQ ID NOs: 97-98 or 168-169.

One embodiment of the present invention is an isolated polynucleotide encoding an FN3 domain specifically binding to c-Met having the amino acid sequence shown in SEQ ID NOs: 32-49 or 111-114.

One embodiment of the present invention is an isolated polynucleotide encoding an EGFR/c-Met bispecific FN3 domain-containing molecule having the amino acid sequence of SEQ ID NOs: 50-72, 106, 118-121, or 138-165.

One embodiment of the present invention is an isolated polynucleotide comprising the polynucleotide sequence of SEQ ID NOs: 115-116 or 166-167.

The polynucleotides of the present invention can be prepared by chemical synthesis, such as solid phase synthesis of polynucleotides on an automatic polynucleotide synthesizer, and assembly into complete single or double stranded molecules. Alternatively, the polynucleotides of the present invention can be obtained by other techniques such as PCR followed by standard cloning. Techniques for producing or obtaining polynucleotides of a given known sequence are well known in the art.

The polynucleotides of the present invention may contain at least one non-coding sequence such as a promoter or enhancer sequence, an intron, a polyadenylation signal, a cis sequence to facilitate binding to RepA, and the like. The polynucleotide sequences may also contain additional sequences encoding additional amino acids that encode, for example, a marker or tag sequence such as a histidine tag or a HA tag to facilitate protein purification or detection, a signal sequence, a fusion protein partner such as RepA , Fc or bacteriophage coating protein such as pIX or pIII

Another embodiment of the present invention is a vector containing at least one polynucleotide of the present invention. Such vectors may be plasmid vectors, viral vectors, baculovirus expression vectors, transposon vectors, or any other vectors suitable for introducing the polynucleotides of the present invention into a given organism or genetic environment in any way. Such vectors may be expression vectors containing elements of the nucleotide sequence that can control, regulate, cause or permit the expression of the polypeptides encoded by such a vector. Such elements may contain transcriptional enhancer binding sites, RNA polymerase initiation sites, ribosome binding sites, and other sites that facilitate the expression of encoded polypeptides in a given expression system. Such expression systems may be cellular or cell-free systems well known in the art.

Another embodiment of the present invention is a host cell containing the vector of the present invention. An FN3 monospecific domain binding to EGFR or c-Met, or an EGFR/c-Met bispecific molecule of the present invention containing FN3 domains, can optionally be generated by a cell line, a mixed cell line, an immortalized cell, or a clonal population of immortalized cells, as well known in the field. See, for example, Ausubel, et al., ed., Current Protocols in Molecular Biology, John Wiley & Sons, Inc., NY, NY (1987-2001); Sambrook, et al., Molecular Cloning: A Laboratory Manual, 2nd ^Edition , Cold Spring Harbor, NY (1989); Harlow and Lane, Antibodies, a Laboratory Manual, Cold Spring Harbor, NY (1989); Colligan et al., eds., Current Protocols in Immunology, John Wiley & Sons, Inc., NY (1994-2001); Colligan et al., Current Protocols in Protein Science, John Wiley & Sons, NY, NY, (1997-2001).

- 25 039356

The host cell selected for expression may be mammalian or may be selected from COS-1, COS-7, HEK293, BHK21, CHO, BSC-1, He G2, SP2/0, HeLa, myeloma, lymphoma, yeast, insect cells. or plants or any derivatives, their immortalized or transformed cells. Alternatively, the host cell may be selected from a species or organism not capable of polypeptide glycosylation, for example a prokaryotic cell or an organism such as BL21, BL21(DE3), BL21-GOLD(DE3), XL1-Blue, JM109, HMS174, HMS174( DE3) and any of the natural or engineered strains of E. coli, Klebsiella or Pseudomonas.

Another embodiment of the present invention is a method for producing an isolated FN3 domain of the present invention specifically binding to EGFR or c-Met, or an isolated EGFR/c-Met bispecific molecule of the present invention comprising an FN3 domain, comprising culturing the isolated host cell of the present invention in conditions such that an isolated FN3 domain specifically binding to EGFR or c-Met or an isolated EGFR/c-Met bispecific molecule containing an FN3 domain is expressed, and the domain or molecule is purified.

An FN3 domain specifically binding to EGFR or c-Met, or an isolated EGFR/c-Met bispecific molecule of the present invention containing an FN3 domain, can be purified from recombinant cell cultures by well-known methods, for example, protein A purification, ammonium sulfate precipitation or ethanol, acid extraction, anion or cation exchange chromatography, phosphocellulose chromatography, hydrophobic interaction chromatography, affinity chromatography, hydroxylapatite chromatography, lectin chromatography, or high performance liquid chromatography (HPLC).

Antibodies bispecific to EGFR/c-Met.

Bispecific antibodies to EGFR/c-Met can be generated de novo or can be constructed from existing monospecific antibodies to EGFR and c-Met.

Examples of antibodies to EGFR that can be used to construct bispecific molecules are, for example, panitumumab (ABX-EGF), nimotuzumab, necitumumab, matuzumab, as well as those described, for example, in US patent No. US 7595378, US 7247301, US patent publication No. US 2011/0256142, US Patent No. US 5891996; US 5212290, US 5558864, US 7589180. For example, an antibody VH domain having the amino acid sequence shown in SEQ ID NO: 189 or 191 and an antibody VL domain having the amino acid sequence shown in SEQ ID NO: 190 or 192 can be used.

Examples of c-Met antibodies that can be used to construct bispecific molecules include, for example, rilotumumab, onartuzumab, ficlatuzumab, and the antibodies described in, for example, PCT International Publication No. WO 2011/110642, US Patent Publication No. US 2004/0166544 , in PCT International Publication No. WO 2005/016382 or in PCT International Publication No. WO 2006/015371. For example, an antibody VH domain having the amino acid sequence shown in SEQ ID NO: 193 or 195 and an antibody VL domain having the amino acid sequence shown in SEQ ID NO: 194 or 196 can be used. by their United States Adopted Names (USAN) names, are available from the American Medical Association at http://_www_ama-assn_org or from the CAS registry.

EGFR and c-Met monospecific variable domains can be selected de novo, for example, from a phage display library where phage is engineered to express human immunoglobulins or portions thereof such as Fab, single chain antibodies (scFv), or unpaired or paired antibody variable regions ( Knappik et al., J. Mol. Biol. 296:57-86, 2000; Krebs et al., J. Immunol. Meth. 254:67-84, 2001; Vaughan et al., Nature Biotechnology, 14:309- 314, 1996; Sheets et al., PITAS (USA), 95:61576162, 1998; Hoogenboom and Winter, J. Mol. Biol. 227:381, 1991; Marks et al., J. Mol. Biol. 222:581 , 1991) and then designed in a bispecific format. EGFR and c-Met monospecific variable domains can be isolated, for example, from phage display libraries expressing light and heavy chain variable regions as fusion proteins to the bacteriophage pIX coat protein, as described in Shi et al. (2010), J. Mol. Biol. 397:385-96 and PCT International Publication No. wO 09/085462). Antibody libraries are screened for binding to the extracellular domains of human EGFR or c-Met, and the resulting positive clones are further characterized and Fab isolated from clone lysates. Such use of phage display methods for the isolation of human antibodies is known in the art. See, for example, US Pat. Nos. 5,223,409; 5,403,484; 5,571,698; 6521404; 6544731; 6555313;

Bispecific antibody formats.

The antibodies of the present invention have two or more antigen-binding sites and are bispecific. The bispecific antibodies of the present invention include antibodies having a full-length antibody structure.

- 26 039356

The term full-length antibody as used herein means an antibody having two full-length antibody heavy chains and two full-length antibody light chains. A full-length antibody heavy chain (HC) consists of the well-known heavy chain variable and constant domains VH, CH1, CH2, and CH3. A full-length antibody light chain (LC) consists of the well-known light chain variable and constant domains VL and CL. A full-length antibody may lack a C-terminal lysine (K) in either one or both heavy chains.

The term Fab-arm or half-molecule means one heavy chain-light chain pair specifically binding to an antigen.

Full-length bispecific antibodies of the present invention can be generated, for example, by exchanging Fab arms (or exchanging hemimolecules) between two monospecific divalent antibodies by introducing substitutions into the heavy chain CH3 interface in each hemimolecule that promote the formation of a heterodimer from two antibody hemimolecules having different specificities, either in vitro in a cell-free medium or using co-expression. The Fab-arm exchange reaction is the result of a disulfide bond isomerization reaction and a dissociation-association of CH3 domains. The disulfide bonds of the heavy chains in the hinge regions of the original monospecific antibodies are restored. The obtained free cysteines of one of the initial monospecific antibodies form a disulfide bond of heavy chains with cysteine residues of the second initial monospecific antibody molecule, and at the same time the CH3 domains of the initial antibodies are released? and there is a re-formation by dissociation-association. Fab-arm CH3 domains can be designed to promote heterodimerization rather than homodimerization. The resulting product is a bispecific antibody having two Fab arms, or half molecules, each of which binds to a separate epitope, i.e. an epitope on EGFR; and an epitope on c-Met.

The term homodimerization as used herein refers to the interaction of two heavy chains having identical CH3 amino acid sequences. The term homodimer as used herein refers to an antibody having two heavy chains with identical CH3 amino acid sequences.

The term heterodimerization as used herein refers to the interaction of two heavy chains having non-identical CH3 amino acid sequences. The term heterodimer as used herein refers to an antibody having two heavy chains with non-identical CH3 amino acid sequences.

To create full-length bispecific antibodies, you can use the protrusion in the cavity (see, for example, international publication PCT No. WO 2006/028936). Briefly, selected amino acids that form an interface between CH3 domains in human IgG can be mutated at positions that affect CH3 domain interactions to promote heterodimer formation. A short side chain amino acid (trough) is inserted into the heavy chain of an antibody specifically binding to the first antigen, and a long side chain amino acid (ledge) is inserted into the heavy chain of an antibody that specifically binds to the second antigen. After co-expression of the two antibodies, a heterodimer is formed as a result of the preferential interaction of the heavy chain in the pit and the heavy chain in the ridge. Examples of pairs of substitutions in CH3 forming a ridge and a trough are as follows (indicated as modified position in the first CH3 domain of the first heavy chain/modified position in the second CH3 domain of the second heavy chain): T366Y/F405A, T366W/F405W, F405W/Y407A, T394W /Y407T, T394S/Y407A, T366W/T394S, F405W/T394S and T366W/T366S_L368A_Y407V.

Other strategies can be used, such as promoting heavy chain heterodimerization using electrostatic interactions by introducing substitutions of positively charged residues on one CH3 surface and negatively charged residues on the other CH3 surface, as described in US Patent Publication No. US 2010/0015133; US Patent Publication No. US 2009/0182127, US Patent Publication No. US 2010/028637, or US Patent Publication No. US2011/0123532. In other strategies, heterodimerization can be promoted by the following substitutions (indicated modified position in the first CH3 domain of the first heavy chain/modified position in the second CH3 domain of the second heavy chain):

T366L_K392M_T394W/F405A_Y407V, L351Y_Y407A/T366A_K409F, L351Y_Y407A/T366V_K409F,

Y407A/T366A_K409F or T350V_L351Y_F405A_Y407V/T350V_T366L_K392L_T394W as described in US Patent Publication No. US2012/0149876; U.S. Patent Publication No. US 2013/0195849.

In addition to the methods described above, the bispecific antibodies of the present invention can be generated in vitro in a cell-free environment by introducing asymmetric mutations in the CH3 regions of two monospecific homodimeric antibodies and generating a bispecific heterodimeric antibody from the two original monospecific homodimeric antibodies under reducing conditions, which promotes isomerization of the disulfide bond into in accordance with the methods described in US International Patent Publication No. WO 2011/131746. In these methods, a first monospecific divalent antibody (eg, an anti-c-Met antibody) and a second monospecific divalent antibody (eg,

- 27 039356 antibody to EGFR) are designed so that they have some substitutions in the CH3 domain, which contribute to the stability of the heterodimer; the antibodies are incubated together under reducing conditions to allow the cysteines in the hinge region to undergo disulfide bond isomerization; as a result, a bispecific antibody is created by exchanging Fab arms. The incubation conditions can be optimally returned to non-reducing conditions. Examples of suitable reducing agents include 2-mercaptoethylamine (2-MEA), dithiothreitol (DTT), dithioerythritol (DTE), glutathione, tris-(2-carboxyethyl)phosphine (TCEP), L-cysteine and beta-mercaptoethanol, preferably the reducing agent is selected from the group consisting of: 2-mercaptoethylamine, dithiothreitol and tris-(2-carboxyethyl)phosphine. For example, an incubation of at least 90 minutes at a temperature of at least 20° C. in the presence of at least 25 mM 2-MEA or in the presence of at least 0.5 mM dithiothreitol at pH 5-8 can be used, e.g. at pH 7.0 or at pH 7.4.

Antibodies bispecific to EGFR/c-Met.

The EGFR/c-Met bispecific antibodies of the present invention may provide an advantage in specificity and reduced side toxicity over small molecule inhibitors of EGFR and/or c-Met. The present invention is based at least in part on the surprising discovery that the EGFR/c-Met bispecific antibodies of the present invention provide a substantially enhanced synergistic inhibitory effect compared to a mixture of EGFR-binding and c-Met-binding monospecific antibodies or published bispecific antibodies. EGFR/c-Met antibodies. Depending on the assay, a synergistic effect ranging from about 14-fold to more than about 800-fold was observed. EGFR/c-Met bispecific antibodies of the present invention provide more effective inhibition of EGFR and c-Met signaling pathways and more effective tumor growth inhibition than cetuximab (Erbitux®) EGFR/c-Met bispecific antibodies of the present invention inhibit EGFR signaling in tumors and /or tumor cell lines with activating EGFR mutations and/or mutations in EGFR, which are known to be resistant to treatment with tyrosine kinase inhibitors, such as gefitinib, and inhibit the c-Met signaling pathway, which has been shown to increase activity and provide compensatory signaling when treatment with EGFR tyrosine kinase inhibitors for cancers such as NSCLC. The EGFR/c-Met bispecific antibodies of the present invention, in addition to direct inhibition of EGFR and c-Met signaling, exhibit antitumor activity through enhanced antibody-dependent cellular cytotoxicity (ADCC) and degradation of EGFR and c-Met receptors. Unlike existing EGFR therapies (cetuximab and panitumumab), the EGFR/c-Met bispecific antibodies of the present invention induce the killing of tumor cells bearing KRAS mutations via enhanced ADCC.

US International Patent Publication No. WO 2010/115551 describes an EGFR/c-Met bispecific antibody (BSAB01) constructed in IgG-scFv format using an EGFR-binding VH/VL pair from cetuximab and a c-Met binding VH/VL pair from an antibody. 5D5 (MetMab, onartuzumab), currently in Phase III trials. BSAB01 exhibits an approximately two-fold (additive) increase in inhibition of A431 cell proliferation compared to parental antibodies (Example 7, Fig. 8b in WO 2010/115551) and moderate additive inhibition of proliferation of Ovarc-8 cells (Fig. 10a, Example 16 in WO 2010/ 115551) compared to the combination of the two parent antibodies (15% inhibition vs. 10%). Thus, surprisingly and unexpectedly, the present invention provides EGFR/c-Met bispecific antibodies showing a significant synergistic effect in inhibition of EGFR and c-Met signaling, cancer cell survival and tumor growth. Without wishing to be bound by any theory, it is believed that the significant synergistic effect of the bispecific antibodies of the present invention is at least in part due to the epitope specificity of both of the EGFR and c-Met binding arms, possibly leading to inhibition of signaling not only via EGFR homodimers. and c-Met, but also via EGFR/HERx heterodimers.

One embodiment of the present invention is an isolated epidermal growth factor receptor (EGFR)/hepatocyte growth factor receptor (c-Met) bispecific antibody comprising:

a) a first heavy chain (HC1) containing HC1 constant domain 3 (HC1 CH3) and HC1 variable region 1 (VH1);

b) a second heavy chain (HC2) containing the HC2 constant domain 3 (HC2 CH3) and the HC2 variable region 2 (VH2);

c) a first light chain (LC1) containing light chain variable region 1 (VL1); and a second light chain (LC2) containing a light chain variable region 2 (VL2), wherein VH1 and VL1 together form a first antigen-binding site that specifically binds to EGFR, and VH2 and VL2 together form a second antigen-binding site that specifically binds to c- Met, and HC1 contains at least one substitution in HC1 CH3, and HC2 contains at least

- 28 039356 at least one substitution in HC2 CH3, and the substitution in HC1 CH3 and the substitution in HC2 CH3 occur at different positions of amino acid residues, if the numbering of the residues corresponds to the EU catalog.

In some embodiments described herein, an EGFR/c-Met bispecific antibody inhibits phosphorylation of extracellular signal-associated kinases 1 and 2 (ERK1/2) in an NCI-H292, NCI-H1975, or SKMES-1 cell line with an IC value ₅₀ which is at least about 10 times less, at least about 20 times less, at least about 30 times less, at least about 40 times less, at least about 50 times less, or at least approximately 60-fold lower than the IC50 value for inhibition of ERK1/2 phosphorylation in NCI- _H292 , NCI-H1975, or SKMES-1 cell lines using a mixture of a control monovalent anti-EGFR antibody containing heavy chain 3 (HC3) and light chain 3 (LC3), and a control monovalent antibody to c-Met containing a heavy chain 4 (HC4) and a light chain 4 (LC4), moreover, HC3 and HC1, LC3 and LC1, HC4 and HC2, and LC4 and LC2 have identical amino acid p sequences, respectively, and ERK1/2 phosphorylation is measured in whole cell lysates by sandwich ELISA using an anti-phospho-ERK1/2 antibody as the capture antibody, and an electrochemiluminescent compound-conjugated antibody that binds to non-phosphorylated and phosphorylated ERK1/2 as detection antibody. The EGFR/c-Met bispecific antibodies of the present invention provide synergistic, more pronounced inhibition of EGFR and c-Met signaling compared to the combination of monospecific anti-EGFR antibodies and monospecific anti-c-Met antibodies as measured by inhibition of ERK1/2 phosphorylation. Such an example of an EGFR/c-Met bispecific antibody is the EM1-mAb of the present invention.

The term EGFR monospecific control antibody as used herein means an antibody having a first Fab arm that binds to EGFR and is identical in amino acid sequence to the EGFR-binding Fab arm of the test EGFR/c-Met bispecific antibody and has a second Fab arm that is inert and binding to an unrelated/unrelated antigen, human immunodeficiency virus (HIV) gp120. The second Fab arm has a light chain having the sequence of SEQ ID NO: 209 and a heavy chain having the sequence of SEQ ID NO: 198 when the EGFR-binding Fab arm in the EGFR/c-Met bispecific antibody tested contains replacement for F405L. The second Fab arm has a light chain having the sequence of SEQ ID NO: 209 and a heavy chain having the sequence of SEQ ID NO: 197 when the EGFR-binding Fab arm in the EGFR/c-Met bispecific antibody tested contains replacement for K409R.

The term c-Met control monospecific antibody as used herein means an antibody having a first Fab arm that binds to c-Met and is identical in amino acid sequence to the c-Met binding Fab arm of the test EGFR/c-Met bispecific antibody, and having a second Fab arm that is inert and binds to an unrelated/unrelated antigen, HIV gp120. The second Fab arm has a light chain having the sequence of SEQ ID NO: 209 and a heavy chain having the sequence of SEQ ID NO: 198 in cases where the c-Met binding Fab arm in the EGFR/c-Met bispecific tested the antibody contains the substitution F405L. The second, inert Fab arm has a light chain having the sequence of SEQ ID NO: 209 and a heavy chain having the sequence of SEQ ID NO: 197, in cases where the c-Met binding Fab arm in the tested EGFR/c bispecific -Met antibody contains a K409R substitution.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits ERK1/2 phosphorylation with an IC ₅₀ value of about 2x10'9 M or less, about 1x10'9 M or less, or about ^1x10'10 M or less .

In some embodiments described herein, ERK1 is phosphorylated at Thr202 and Tyr204, and ERK2 is phosphorylated at Thr185 and Tyr197.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits protein kinase B (AKT) phosphorylation at Ser473 in the NCI-H1975 cell line with an IC ₅₀ value of at least about 70 times less than IC ₅₀ for inhibition of AKT phosphorylation at Ser473 in the NCI-H1975 cell line with a mixture of a control monovalent antibody to EGFR containing HC3 and LC3, and a control monovalent antibody to c-Met containing HC4 and LC4, with HC3 and HC1, LC3 and LC1, HC4 and HC2 and LC4 and LC2 have identical amino acid sequences, respectively, with AKT phosphorylation at Ser473 measured in whole cell lysates by sandwich ELISA using an antibody that binds to non-phosphorylated and phosphorylated AKT as a capture antibody and is conjugated to an electrochemiluminescent compound antibody to Ser473 phosphorylated AKT as a detection antibody.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits protein kinase B (AKT) phosphorylation at Thr308 in the NCI-H1975 cell line with an IC ₅₀ value of at least about 100-fold lower than

- 29 039356 with an IC ₅₀ value for inhibiting AKT phosphorylation at Thr308 in the NCI-H1975 cell line with a mixture of a control monovalent antibody to EGFR containing HC3 and LC3 and a control monovalent antibody to c-Met containing HC4 and LC4, with HC3 and HC1 , LC3 and LC1, HC4 and HC2, and LC4 and LC2 have identical amino acid sequences, respectively, wherein AKT phosphorylation at Thr308 is measured in whole cell lysates by sandwich ELISA using an antibody that binds to unphosphorylated and phosphorylated AKT as a capture antibody, and an antibody conjugated with an electrochemiluminescent compound to AKT phosphorylated at Thr308 as a detection antibody.

The EGFR/c-Met bispecific antibodies of the present invention provide synergistic, more pronounced inhibition of EGFR and c-Met signaling compared to the combination of EGFR monospecific antibodies and c-Met monospecific antibodies, as measured by inhibition of AKT phosphorylation. Such an example of an EGFR/c-Met bispecific antibody is the EM1-mAb of the present invention.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits AKT phosphorylation at Ser473 or Thr308 with an IC ₅₀ value of approximately 1x10'9 M or less.

In some embodiments described herein, an EGFR/c-Met bispecific antibody binds to EGFR of SEQ ID NO: 73 at EGFR residues K489, I491, K467, and S492 and to c-Met at PEFRDSYPIKYVHAF residues (SEQ ID NO: 238) and FAQSKPDSAEPMDRSA (SEQ ID NO: 239). An example of such a bispecific antibody is EM1-mAb. The EM-1 bispecific antibody binds to EGFR and c-Met at separate epitopes compared to the BSAB01 antibody, as described above in US International Patent Publication No. WO 2010/115551. The original EGFR-binding arm (cetuximab) BSAB01 binds to EGFR amino acid residues R353, Q384, Q408, H409, F412, S418, S440, K443, K465, I467, S468, and N473 in mature EGFR, corresponding to residues R367, Q408, Q432, H433, F436, S442, S464, K467, K489, I491, S492 and N497 full-length EGFR of SEQ ID NO: 73 (Li et al., Cancer Cell, 7:301-311, 2005). The parent c-Met-binding arm BSAB01 (mAb 5D5) binds to c-Met at residues 325-340 PGAQLARQIGASLNDD (SEQ ID NO: 240).

Epitope mapping of the EGFR-binding antibody (2F8) parent of the EM1-mAb is described in US Patent Publication No. US 2011/0256142A1. Cetuximab and the parent antibody 2F8 bind to overlapping but distinct epitopes.

Epitope mapping can be performed by standard methods. For example, in the case where the structures of both individual components are known, in silico analysis of the attachment of one protein to another protein can be used to identify sites capable of interacting. Selective displacement of hydrogen with deuterium (H/D) in the antigen-antibody complex can be performed to map regions of the antigen capable of binding to the antibody. Antigen segment and point mutagenesis can be used to locate amino acids important for antibody binding.

In some embodiments described herein, the EGFR/c-Met bispecific antibody neutralizes EGFR and c-Met signaling.

The EGFR/c-Met bispecific antibody of the present invention can neutralize EGFR and c-Met signaling by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95 .

Binding of a ligand, such as EGF to EGFR, stimulates receptor dimerization, autophosphorylation, activation of the intrinsic receptor, the cytoplasmic tyrosine kinase domain, and the initiation of multiple transduction and transactivation signaling pathways involved in the regulation of DNA synthesis (gene activation) and cell cycle progression or division. Neutralization of EGFR signaling can lead to inhibition of one or more downstream EGFR signaling pathways, and thus EGFR neutralization can have various effects, including inhibition of cell proliferation and differentiation, angiogenesis, cell motility and metastasis, and inhibition of downstream signaling pathways.

EGFR signaling and EGFR signaling neutralization can be measured by various well known methods, for example by measuring receptor autophosphorylation at any of the tyrosine residues Y1068, Y1148 and Y1173 (Downward et al., Nature, 311:483-5, 1984) and/or phosphorylation of natural or synthetic substrates; and measurement of inhibition of autophosphorylation and/or phosphorylation of natural or synthetic substrates by the bispecific antibodies of the present invention. Phosphorylation can be detected by well known methods such as ELISA or Western blotting using an antibody specific for phosphotyrosine. Examples of assays can be found in Panek et al., J. Pharmacol. Exp. Thera. 283:1433-44, 1997 and Batley et al., Life Sci. 62:143-50, 1998, as well as in this document.

Binding of a ligand, such as HGF to c-Met, stimulates receptor dimerization, autophosphorylation, activation of the receptor's cytoplasmic tyrosine kinase domain, and initiation of multiple transduction and transactivation signaling pathways involved in the regulation of DNA synthesis (gene activation) and cell cycle or division progression. . Inhibition of c-Met signaling can lead to inhibition in one or more of the downstream c-Met signaling pathways, and thus c-Met neutralization can have various effects, including inhibition of cell proliferation and differentiation, angiogenesis, cell motility, and metastasis.

c-Met signaling and c-Met signaling neutralization can be measured by various well known methods, for example by measuring receptor autophosphorylation of at least one of the Y1230, Y1234, Y1235 or Y1349 tyrosine residues and/or phosphorylation of natural or synthetic substrates. Phosphorylation can be detected, for example, using an antibody specific for phosphotyrosine, ELISA or Western blotting. Examples of assays can be found in Panek et al., J. Pharmacol. Exp. Thera. 283:1433-44, 1997 and Batley et al., Life Sci. 62:143-50, 1998, as well as in this document.

EGFR and c-Met signaling can be measured using various well known methods described herein, for example by measuring inhibition of ERK1/2 and AKT phosphorylation. Inhibition of ERK1 phosphorylation at Thr202 and Tyr204 and ERK2 phosphorylation at Thr185 and Tyr187, as well as inhibition of AKT phosphorylation at Ser473 or Thr308 can be measured, for example, in NCI-H1975 cell lysates by sandwich ELISA with capture antibodies coated on a solid support, and detection antibodies conjugated to an electrochemiluminescent compound, such as the Meso Scale Discover (MSD) SULFO-TAG tag, followed by signal detection on a plate reader.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits the growth of NCI-H292 or NCI-H1975 cells with an IC ₅₀ value of at least about 300 times less, at least about 400 times less , at least about 500 times less, at least about 600 times less, at least about 700 times less, or at least about 800 times less than the IC ₅₀ value of growth inhibition of NCI-H292 or NCI cells -H1975 with cetuximab if NCI-H292 or NCI-H1975 cells are grown under low attachment conditions.

Cell growth inhibition can be analyzed by known methods. Cells can be seeded onto plates coated with hydrogels or biomimetic polymers (e.g., Corning Ultra Low Attachment Plates) to prevent or reduce cell attachment, and the effect of antibodies on 7.5 ng/mL HGF-induced cell growth can be assessed by measuring the percentage of viable cells. cells after incubation for 72 h using standard methods.

The EGFR/c-Met bispecific antibodies of the present invention provide synergistic and more pronounced inhibition of EGFR and/or c-Met expressing cancer cells compared to the combination of EGFR monospecific antibodies and c-Met monospecific antibodies, as well as standard treatment with cetuximab. An example of such an EGFR/c-Met bispecific antibody is the EM1-mAb of the present invention. The EGFR/c-Met bispecific antibodies of the present invention inhibit cancer cells that express wild-type EGFR and wild-type c-Met, as well as cancer cells that express the L858R/T790M mutant EGFR, and this mutation was found to contribute to the development of resistance to treatment with small molecule tyrosine kinase inhibitors (TKIs), such as gefitinib. Thus, the EGFR/cMet bispecific antibodies of the present invention may provide an advantage in a wider patient population compared to cetuximab and TKI.

In some embodiments described herein, the EGFR/c-Met bispecific antibody inhibits tumor growth from HGF-expressing SKMES-1 cells in SCID Beige mice with a T/C percentage (%) that is at least 500 times less compared to cetuximab when administered with a bispecific antibody and cetuximab at a dose of 20 mg/kg.

Tumor xenograft models using SCID Beige mice are well known. SKMES-1 cells can be engineered to express human HGF using standard methods. Typically, SCID Beige mice can be subcutaneously inoculated with SKMES-1 cells expressing human HGF embedded in an extracellular matrix such as Culturex by injecting them into the dorsal flank of each animal. One week after implantation, mice can be divided into groups with equivalent tumor volumes, and thereafter administered, for example, three times a week, EGFR/c-Met bispecific antibodies of the present invention, control or comparison antibodies, or small molecule compounds. Tumor volumes can be recorded twice a week, and tumor growth inhibition (TGI) can be observed by calculating the percentage (%) value of T/C. The value of % T/C is an indicator of antitumor efficacy. T and C are the average volumes in the treated and control groups, respectively, on a given day.

The EGFR/c-Met bispecific antibodies of the present invention provide significantly higher tumor-killing efficacy in vivo compared to standard

- 31 039356 treatment with cetuximab and therefore may provide an advantage in the patient population over cetuximab.

In some embodiments described herein, the bispecific

An EGFR/c-Met antibody contains HC1 and HC2 from the IgG1, IgG ₂ , IgG3 or IgG4 isotype.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises HC1 and HC2 from the IgG1 isotype.

In some embodiments described herein, an EGFR/c-Met bispecific antibody in HC1 CH3 contains at least one, two, three, four, five, six, seven, or eight substitutions, and in HC2 CH3 contains at least one, two, three, four, five, six, seven or eight substitutions for residue positions 350, 366, 368, 370, 399, 405, 407 or 409, if the numbering of the residues corresponds to the EU catalog.

In some embodiments described herein, an EGFR/c-Met bispecific antibody contains at least one, two, three, or four substitutions in HC1 CH3 and at least one, two, three, or four substitutions in HC2 CH3. according to residue positions 350, 370, 405 or 409, if the numbering of the residues corresponds to the EU catalog.

The antibody domains and numbering are well known. Two CH3 domains (or CH3 regions) are non-identical if they differ from each other by at least one amino acid substitution. The IgG1 CH3 region usually consists of residues 341-446 in IgG1 (numbering of residues corresponds to the EU catalog). An example of an IgG1 constant region is shown in SEQ ID NO: 203. The CH3 domain spans residues 224-329 of SEQ ID NO: 203 and corresponds to EC residues 341-446.

In some embodiments described herein, an EGFR/c-Met bispecific antibody contains at least one substitution in HC1 CH3 and at least one substitution in HC2 CH3 at residue positions 405 or 409 if the residue numbering is in the catalog EU.

In other embodiments described herein, the EGFR/c-Met bispecific antibody contains a K409R or F405L substitution in HC1 CH3 and a K409R or F405L substitution in HC2 CH3 if the residue numbering is in accordance with the EU catalogue.

In some embodiments described herein, the EGFR/c-Met bispecific antibody contains the F405L substitution in HC1 CH3 and the K409R substitution in HC2 CH3.

In some embodiments described herein, substitutions in HC1 CH3 and HC2 CH3 are substitutions in positions 366, 368, 370, 399, 405, 407 or 409 (numbering follows the EU catalog). These positions correspond to the linear positions of residues 248, 250, 252, 281, 287, 289 and 291, respectively, in the heavy chain constant region of SEQ ID NOS: 203 and 204.

In some embodiments described herein, at position 409 in HC1 CH3 there is an amino acid substitution other than Lys, Leu, or Met, and at position 405 in HC2 CH3 there is an amino acid substitution other than Phe.

In some embodiments described herein, at position 405 in HC1 CH3 there is an amino acid substitution other than Phe, and at position 409 in HC2 CH3 there is an amino acid substitution other than Lys, Leu, or Met.

In some embodiments described herein, there is an amino acid substitution at position 409 in HC1 CH3 other than Lys, Leu, or Met, and at position 405 in HC2 CH3 there is an amino acid substitution other than Phe, Arg, or Gly.

In some embodiments described herein, at position 405 in HC1 CH3 there is an amino acid substitution other than Phe, Arg or Gly, and at position 409 in HC2 CH3 there is an amino acid substitution other than Lys, Leu or Met.

In some embodiments described herein, HC1 CH3 contains Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 contains an amino acid other than Phe at position 405 and contains Lys at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Phe at position 405 and Lys at position 409, and HC2 CH3 has Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has a substitution other than Phe, Arg, or Gly at position 405 and Lys in position 409.

In some embodiments described herein, HC1 CH3 has a substitution other than Phe, Arg, or Gly at position 405 and Lys at position 409, and HC2 CH3 has Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 405 position 409.

In some embodiments described herein, HC1 CH3 has Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has Leu at position 405 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has Leu at position 405 and Lys at position 409, and HC2 CH3 has Phe at position 405 and an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has Phe at position 405 and Arg at position 409, and HC2 CH3 has an amino acid other than Phe, Arg or Gly at position 405 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Phe, Arg, or Gly at position 405 and Lys at position 409, and HC2 CH3 has Phe at position 405 and Arg at position 409.

In some embodiments described herein, HC1 CH3 has Phe at position 405 and Arg at position 409, and HC2 CH3 has Leu at position 405 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has Leu at position 405 and Lys at position 409, and HC2 CH3 has Phe at position 405 and Arg at position 409.

In some embodiments described herein, HC1 CH3 has Phe at position 405 and Lys at position 409, and HC2 CH3 has Leu at position 405 and Arg at position 409.

In some embodiments described herein, HC1 CH3 has Leu at position 405 and Arg at position 409, and HC2 CH3 has Phe at position 405 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405.

In some embodiments described herein, HC1 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405, and HC2 CH3 has an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has Arg at position 409 and HC2 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405.

In some embodiments described herein, HC1 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405, and HC2 CH3 has Arg at position 409.

In some embodiments described herein, HC1 CH3 has Lys at position 370, Phe at position 405, and Arg at position 409, and HC2 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405.

In some embodiments described herein, HC1 CH3 has Lys at position 409, Thr at position 370, and Leu at position 405, and HC2 CH3 has Lys at position 370, Phe at position 405, and Arg at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln, Arg, Ser, or Thr at position 407.

In some embodiments described herein, CH3 HC1 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln, Arg, Ser, or Thr at position 407, and CH3 HC2 has an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has Ala, Gly, His, Ile, Leu, Met, Asn, Val, or Trp at position 407 .

In some embodiments described herein, CH3 HC1 has Ala, Gly, His, Ile, Leu, Met, Asn, Val, or Trp at position 407, and CH3 HC2 has an amino acid other than Lys, Leu, or Met at position 409 .

In some embodiments described herein, HC1 CH3 has an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has Gly, Leu, Met, Asn, or Trp at position 407.

In some embodiments described herein, HC1 CH3 has Gly, Leu, Met, Asn, or Trp at position 407 and HC2 CH3 has an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has Tyr at position 407 and an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln , Arg, Ser or Thr at position 407 and Lys at position 409.

In some embodiments described herein, CH3 HC1 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln, Arg, Ser, or Thr at position 407 and Lys at position 409, and CH3 HC2 has Tyr at position 407 and an amino acid other than Lys, Leu or Met at position 409.

In some embodiments described herein, HC1 CH3 has Tyr at position 407 and an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has Ala, Gly, His, Ile, Leu, Met, Asn, Val or Trp at position 407 and Lys at position 409.

In some embodiments described herein, CH3 HC1 has Ala, Gly, His, Ile, Leu, Met, Asn, Val, or Trp at position 407 and Lys at position 409, and CH3 HC2 has Tyr at

- 33 039356 position 407 and an amino acid other than Lys, Leu or Met at position 409.

In some embodiments described herein, CH3 HC1 has Tyr at position 407 and an amino acid other than Lys, Leu, or Met at position 409, and CH3 HC2 has Gly,

Leu, Met, Asn or Trp at position 407 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has Gly, Leu, Met, Asn, or Trp at position 407 and Lys at position 409, and HC2 CH3 has Tyr at position 407 and an amino acid other than Lys, Leu, or Met at position 407. position 409.

In some embodiments described herein, HC1 CH3 has Tyr at position 407 and Arg at position 409, and HC2 CH3 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln, Arg, Ser, or Thr at position 407 and Lys at position 409.

In some embodiments described herein, CH3 HC1 has an amino acid other than Tyr, Asp, Glu, Phe, Lys, Gln, Arg, Ser, or Thr at position 407 and Lys at position 409, and CH3 HC2 has Tyr at position 407 and Arg at position 409.

In some embodiments described herein, HC1 CH3 has Tyr at position 407 and Arg at position 409, and HC2 CH3 has Ala, Gly, His, Ile, Leu, Met, Asn, Val, or Trp at position 407 and Lys at position 407. position 409.

In some embodiments described herein, HC1 CH3 has Ala, Gly, His, Ile, Leu, Met, Asn, Val, or Trp at position 407 and Lys at position 409, and HC2 CH3 has Tyr at position 407 and Arg at position 407. position 409.

In some embodiments described herein, HC1 CH3 has Tyr at position 407 and Arg at position 409, and HC2 CH3 has Gly, Leu, Met, Asn or Trp at position 407 and Lys at position 409.

In some embodiments described herein, HC1 CH3 has Gly, Leu, Met, Asn, or Trp at position 407 and Lys at position 409, and HC2 CH3 has Tyr at position 407 and Arg at position 409.

In some embodiments described herein, HC1 CH3 has an amino acid other than Lys, Leu, or Met at position 409, and HC2 CH3 has (i) an amino acid other than Phe, Leu, and Met at position 368, or ( ii) Trp at position 370, or (iii) an amino acid other than Asp, Cys, Pro, Glu, or Gln at position 399.

In some embodiments described herein, HC1 CH3 has (i) an amino acid other than Phe, Leu, and Met at position 368, or (ii) Trp at position 370, or (iii) an amino acid other than Asp, Cys, Pro, Glu, or Gln at position 399, and HC2 CH3 has an amino acid other than Lys, Leu, or Met at position 409.

In some embodiments described herein, HC1 CH3 has Arg, Ala, His or Gly at position 409 and HC2 CH3 contains (i) Lys, Gln, Ala, Asp, Glu, Gly, His, Ile, Asn, Arg , Ser, Thr, Val, or Trp at position 368, or (ii) Trp at position 370, or (iii) Ala, Gly, Ile, Leu, Met, Asn, Ser, Thr, Trp, Phe, His, Lys, Arg, or Tyr at position 399.

In some embodiments described herein, CH3 HC1 has (i) Lys, Gln, Ala, Asp, Glu, Gly, His, Ile, Asn, Arg, Ser, Thr, Val, or Trp at position 368, or (ii ) Trp at position 370, or (iii) Ala, Gly, Ile, Leu, Met, Asn, Ser, Thr, Trp, Phe, His, Lys, Arg, or Tyr at position 399, and HC2 CH3 has Arg, Ala, His or Gly at position 409.

In some embodiments described herein, HC1 CH3 has Arg at position 409 and HC2 CH3 has (i) Asp, Glu, Gly, Asn, Arg, Ser, Thr, Val, or Trp at position 368, or (ii) Trp at position 370, or (iii) Phe, His, Lys, Arg, or Tyr at position 399.

In some embodiments described herein, HC1 CH3 has (i) Asp, Glu, Gly, Asn, Arg, Ser, Thr, Val, or Trp at position 368, or (ii) Trp at position 370, or (iii) Phe, His, Lys, Arg or Tyr at position 399 and HC2 CH3 has Arg at position 409.

In some embodiments described herein, HC1 CH3 contains a K409R substitution or an F405L substitution, and HC2 CH3 contains a K409R substitution or a F405L substitution, with residue numbering following the EU catalogue.

In some embodiments described herein, HC1 CH3 contains the F405L substitution and HC2 CH3 contains the K409R substitution.

As a rule, substitutions are introduced into the molecule, for example, into the constant domain of an antibody, at the DNA level using standard methods.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises VH1 and VL1, wherein:

VH1 contains the heavy chain complementarity determining region (HCDR) 1 (HCDR1), HCDR2 (HCDR2) and HCDR3 (HCDR3) with the amino acid sequences of SEQ ID NOs: 210, 211 and 212, respectively; and

VL1 contains the light chain complementarity determining region (LCDR) 1 (LCDR1), LCDR2 (LCDR2) and LCDR3 (LCDR3) with the amino acid sequences of SEQ ID NOS: 213, 214 and 215, respectively.

- 34 039356

In some embodiments described herein, the bispecific

EGFR/c-Met antibody contains VH2 and VL2, and:

VH2 contains HCDR1, HCDR2 and HCDR3 with amino acid sequences

SEQ ID NOs: 216, 217 and 218, respectively; and

VL2 contains LCDR1, LCDR2 and LCDR3 with amino acid sequences

SEQ ID NOs: 219, 220 and 221, respectively.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises VH1, VL1, VH2, and VL2 with the amino acid sequences of SEQ ID NOs: 189, 190, 193, and 194, respectively.

In some embodiments described herein, an EGFR/c-Met bispecific antibody comprises HC1, LC1, HC2, and LC2 with the amino acid sequences of SEQ ID NOs: 199, 200, 201, and 202, respectively, optionally having a C-terminal lysine deleted from HC1, HC2, or both of HC1 and HC2.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises VH1 and VL1, wherein:

VH1 contains HCDR1, HCDR2 and HCDR3 with amino acid sequences

SEQ ID NOs: 222, 223 and 224, respectively; and

VL1 contains LCDR1, LCDR2 and LCDR3 with amino acid sequences

SEQ ID NOs: 225, 226 and 227, respectively.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises VH2 and VL2, wherein:

VH2 contains HCDR1, HCDR2 and HCDR3 with amino acid sequences

SEQ ID NOs: 228, 229 and 230, respectively; and

VL2 contains LCDR1, LCDR2 and LCDR3 with amino acid sequences

SEQ ID NOs: 231, 232 and 233, respectively.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises VH1, VL1, VH2, and VL2 with the amino acid sequences of SEQ ID NOs: 191, 192, 195, and 196, respectively.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises HC1, LC1, HC2, and LC2 with the amino acid sequences of SEQ ID NOs: 234, 235, 236, and 237, respectively, optionally having HC1, HC2, or from both HC1 and HC2 C-terminal lysine.

In some embodiments described herein, EGFR/c-Met bispecific antibodies can block EGF binding to EGFR and HGF binding to c-Met with an IC ₅₀ value of less than about 1 x 10 ^-8 M, less than about 1 x 10' ⁹ M , less than about 1 x 10' ¹⁰ M, less than about 1 x 10 ^-11 M, or less than about 1 x 10' ¹² M in a competition assay using extracellular domains of recombinant human EGFR or recombinant human c-Met coated on plates and incubated with EGFR bispecific /c-Met antibodies of the present invention or without them. The EGFR/c-Met bispecific antibodies described herein can block ECF binding to EGFR and HGF binding to c-Met by at least 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 , 80, 85, 90, 91, 92, 93, 94, 95 , 96, 97, 98, 99 or 100% compared to EGF binding to EGFR and HGF binding to c-Met in the absence of EGFR/c-Met bispecificity antibodies of the present invention described herein using the same assay conditions.

In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises HC1, LC1, HC2, and LC2, wherein HC1, LC1, HC2, and LC2 are encoded by synthetic polynucleotides containing the sequence of SEQ ID NOs: 205, 206, 207 and 208 respectively.

The EGFR/c-Met bispecific antibodies of the present invention can be generated using the techniques described herein, for example, using CH3 construction and antibody generation by Fab arm exchange in vitro. An example of a bispecific antibody can be generated from two monospecific antibodies by combining approximately 1-20 mg/mL of each antibody in a 1:1 molar ratio in PBS at pH 7.0-7.4 in a buffer solution having a final concentration of 75 mM 2-mercaptoethanolamine ( 2-MEA), incubating for 2-6 hours at 25-37°C, followed by removal of 2-MEA by dialysis, diafiltration, tangential flow filtration and vortex cell filtration. The yield of the bispecific antibody may be greater than about 80%, greater than about 90%, greater than about 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%.

In some embodiments described herein, methods for preparing an isolated EGFR/c-Met bispecific antibody are provided, comprising:

combining an isolated anti-EGFR monospecific bivalent antibody containing two heavy chains of SEQ ID NO: 199 and two light chains of SEQ ID NO: 200, and an isolated anti-c-Met monospecific bivalent antibody containing two heavy chains of SEQ ID NO: 201 and two light chains of SEQ ID NO: 202, in a mixture with a molar ratio of approximately 1:1;

introducing a reducing agent into the mixture;

incubating the mixture for about 90 minutes to about 6 hours;

removal of the reducing agent;

purification of an EGFR/c-Met bispecific antibody comprising a first heavy chain of SEQ ID NO: 199 and a second heavy chain of SEQ ID NO: 201, a first light chain of SEQ ID NO: 200 and a second light chain of SEQ ID NO: 202 , wherein the first heavy chain of SEQ ID NO: 199, together with the first light chain of SEQ ID NO: 200, forms a first binding domain that specifically binds to EGFR, and the second heavy chain of SEQ ID NO: 201, together with the second light chain of SEQ ID NO: 202 forms a second binding domain that specifically binds to c-Met.

In some embodiments described herein, the reducing agent is 2-mercaptoethanolamine (2-MEA).

In some embodiments described herein, 2-MEA is present at a concentration of about 25 to about 75 mM.

In some embodiments described herein, the incubation step is carried out at a temperature of from about 25 to about 37°C.

In some embodiments described herein, an isolated EGFR/-c-Met bispecific antibody is provided, comprising HC1, LC1, HC2, and LC2, wherein HC1 contains the sequence of SEQ ID NO: 199, LC1 contains the sequence of SEQ ID NO: 200, HC2 contains the sequence of SEQ ID NO: 201 and LC2 contains the sequence of SEQ ID NO: 202, and HC1, LC1, HC2 and/or LC2 additionally contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 conservative amino acid substitutions.

In some embodiments described herein, an isolated EGFR/-c-Met bispecific antibody is provided, comprising HC1, LC1, HC2, and LC2, wherein HC1 contains the sequence of SEQ ID NO: 234, LC1 contains the sequence of SEQ ID NO: 235, HC2 contains the sequence of SEQ ID NO: 236 and LC2 contains the sequence of SEQ ID NO: 237, and HC1, LC1, HC2 and/or LC2 additionally contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 conservative amino acid substitutions.

EGFR/c-Met bispecific antibodies in which the amino acid sequences of HC1, LC1, HC2, and LC2 do not differ significantly from the antibodies described herein are within the scope of the present invention. Typically, this includes one or more conservative amino acid substitutions for an amino acid having similar charge, hydrophobic, or stereochemical characteristics at antigen-binding sites or scaffolds without adversely altering antibody properties. Can also be made conservative substitutions that improve the properties of the antibody, such as stability or affinity. For example, it is possible to introduce 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 amino acid substitutions in VH1, VL1, VH2 and/or VL2. For example, conservative amino acid substitutions may include the replacement of native amino acid residues with non-native residues, such as those accompanied by little or no change in polarity or charge of the amino acid residue at that position. Moreover, any native residue in the polypeptide can also be replaced by alanine, as previously described in relation to alanine-scanning mutagenesis (MacLennan et al., Acta Physiol. Scand. Suppl. 643:55-67, 1998; Sasaki et al. , Adv Biophys 35:1-24, 1998). Desired amino acid substitutions can be determined by those skilled in the art when such substitutions are needed. For example, amino acid substitutions can be used to identify important residues in the sequence of a molecule or to increase or decrease the affinity of the molecules described in the present invention. Examples of conservative amino acid substitutions are described above.

Amino acid substitutions can be performed, for example, using PCR mutagenesis (US patent No. 4,683,195). Libraries of variants can be generated using well-known methods, for example, by using random (NNK) or non-random codons, such as DVK codons encoding 11 amino acids (Ala, Cys, Asp, Glu, Gly, Lys, Asn, Arg, Ser, Tyr , Trp), as well as screening libraries for variants with desired properties.

In some embodiments described herein, amino acid substitutions can be introduced into the constant region of an antibody. For example, different IgG1 allotypes can be used in the EGFR/c-Met bispecific antibodies of the present invention, eg the well-known Glm17 allotype, Glm3 allotype, Glm1 allotype, or a combination thereof.

In some embodiments described herein, the pharmacokinetic properties of EGFR/c-Met bispecific antibodies can be improved by introducing substitutions in the Fc domain that modulate the half-life of the antibody. In some embodiments described herein, the EGFR/c-Met bispecific antibody comprises an M252Y/S254T/T256E substitution in HC1 and/or HC2, with residue numbering according to the EU catalog. The M252Y/S254T/T256E substitutions have been shown to increase the half-life of the antibody (Dall'Acqua et al., J. Biol. Chem.

- 36 039356

281:23514-24, 2006).

EGFR/c-Met bispecific antibodies having conservative substitutions and/or additional substitutions in their Fc region were tested for characterization using the methods described herein.

In some embodiments described herein, the immunoeffector properties of EGFR/c-Met bispecific antibodies can be enhanced or attenuated by Fc modifications using techniques known to those skilled in the art. For example, Fc effector functions such as C1q binding, complement dependent cytotoxicity (CDC), antibody dependent cell mediated cytotoxicity (ADCC), antibody dependent cell mediated phagocytosis (ADCP), downregulation of cell surface receptors (e.g. B cell receptor; BCR ) and the like can be provided and/or controlled by modifying residues in the Fc responsible for these functions.

Antibody-dependent cellular cytotoxicity, or ADCC, is a cell-mediated response in which non-specific cytotoxic cells expressing Fc receptors (FcRc) (e.g., natural killer (NK) cells, neutrophils, and macrophages) recognize bound antibody on a target cell and then elicit target cell lysis.

The ability of monoclonal antibodies to induce ADCC can be enhanced by designing their oligosaccharide component. Human IgG1 or IgG3 are N-glycosylated at Asn297 by most glycans in the well-known 2-antennary forms GO, GOF, G1, G1F, G2 or G2F. Antibodies produced by N-engineered CHO cells typically have a glycan fucose content of at least about 85%. Removal of central fucose from 2-antennary complex oligosaccharides attached to Fc regions enhances antibody ADCC through improved binding to FcyRIIIa without altering antigen binding or complement-dependent (CDC) activity. Such mAbs can be prepared in a variety of ways that have been reported to result in successful expression of relatively highly defucosylated antibodies bearing 2-anthenary complex-type Fc-oligosaccharides, for example by controlling culture osmolality (Konno et al., Cytotechnology, 64:249- 65, 2012), by using a variant of the CHO cell line Lec13 as host cells (Shields et al., J. Biol. Chem. 277:26733-26740, 2002), by using a variant of the CHO cell line EB66 as host cells (Olivier et al., MAbs; 2(4), 2010; Epub ahead of print; PMID:20562582), by using the YB2/0 hybridoma cell line as host cells (Shinkawa et al., J. Biol Chem 278: 3466-3473, 2003), by introducing a small interfering RNA specific for the α 1,6-fucosyltransferase (FUT8) gene (Mori et al., Biotechnol. Bioeng. 88:901-908, 2004) or by co-expression in-1, 4-M-acetylglucosaminyltransferase III and Golgi α-mannosidase II or an effective alpha-mannosidase I inhibitor , kifunensin (Ferrara et al., J. Biol Chem 281:50325036, 2006, Ferrara et al., Biotechnol Bioeng 93:851-861, 2006; Xhou et al., Biotechnol Bioeng 99:652-65, 2008).

In some embodiments described herein, ADCC caused by EGFR/c-Met bispecific antibodies can also be enhanced by introducing some substitutions in the Fc region of the antibody. Examples of substitutions are, for example, substitutions at amino acid positions 256, 290, 298, 312, 356, 330, 333, 334, 360, 378 or 430 (EC numbering) as described in US Patent No. US6737056.

In some embodiments described herein, an EGFR/c-Met bispecific antibody of the present invention has a 2-anthenary glycan structure with a fucose content ranging from about 1 to about 15%, e.g., 15, 14, 13, 12, 11 10 , 9, 8, 7, 6, 5, 4, 3, 2 or 1%. In some embodiments, the EGFR/c-Met bispecific antibody has a 2-anthenary glycan structure with about 50%, 40%, 45%, 40%, 35%, 30%, 25%, or 20% fucose.

The fucose content refers to the amount of fucose monosaccharide within the sugar chain at position Asn297. The relative amount of fucose is the percentage of fucose-containing structures relative to all glycostructures. It can be characterized and quantified in a variety of ways, for example 1) using MALDI-TOF (Matrix-assisted laser desorption time-of-flight ionization) of an N-glycosidase F treated sample (e.g. complex, hybrid, oligomannose and high mannose structures) as described in international US Patent Publication No. WO 2008/077546; 2) by enzymatic release of Asn297 glycans followed by derivatization and detection/quantification by HPLC (FASLC) with fluorescence detection and/or HPLC-MS (FASLC-MS); 3) by analyzing the intact protein of the native or reconstituted mAb with/without Asn297 glycan treatment with Endo S or another enzyme that cleaves between the first and second GlcNAc monosaccharides, leaving fucose attached to the first GlcNAc; 4) by cleaving the mAb into its constituent peptides by an enzymatic method (eg, trypsin or Lys-C endopeptidase), followed by separation, detection and quantification by HPLC-MS (SELC-MS);

- 37 039356

5) by separating the oligosaccharides from the mAb protein by specific enzymatic deglycosylation of PNGase F at Asn 297. The oligosaccharides thus released can be fluorophore-labeled, separated and identified by various auxiliary methods that allow accurate characterization of glycan structures by laser desorption ionization mass spectrometry using a matrix (MALDI) by comparing the experimental masses with the theoretical masses, determine the degree of sialylation by ion-exchange HPLC (GlycoSep C), separate and quantify the oligosaccharide forms by the regular phase HPLC hydrophilicity criterion (GlycoSep N), and separate and quantify the oligosaccharide forms by high performance capillary electrophoresis with laser -induced fluorescence (HPCE-LIF).

The term low fucose or low fucose as used herein refers to antibodies with a fucose content of approximately 1-15%.

The term normal fucose or normal fucose as used herein refers to antibodies with a fucose content of greater than about 50%, typically greater than about 80% or greater than 85%.

In some embodiments, the present invention provides synthetic nucleic acids encoding the heavy chains and light chains of the EGFR/c-Met-binding bispecific antibodies of the present invention, as isolated polynucleotides, or as parts of expression vectors, or as parts of linear DNA sequences, including linear DNA sequences used for in vitro transcription/translation, vectors compatible with the expression, secretion and/or presentation of the compositions or the results of their site-directed mutagenesis in prokaryotes, eukaryotes or filamentous phages.

In some embodiments, the invention provides an isolated polynucleotide comprising the polynucleotide sequence of SEQ ID NO: 205, 206, 207, or 208.

The polynucleotides of the present invention can be prepared by chemical synthesis, such as solid phase synthesis of polynucleotides on an automatic polynucleotide synthesizer, and assembly into complete single or double stranded molecules. Alternatively, the polynucleotides of the present invention can be obtained by other techniques such as PCR followed by standard cloning. Techniques for producing or obtaining polynucleotides of a given known sequence are well known in the art.

The polynucleotides of the present invention may contain at least one non-coding sequence such as a promoter or enhancer sequence, an intron, a polyadenylation signal, a cis sequence to facilitate binding to RepA, and the like. The polynucleotide sequences may also contain additional sequences encoding additional amino acids that encode, for example, a marker or tag sequence such as a histidine tag or a HA tag to facilitate protein purification or detection, a signal sequence, a fusion protein partner such as RepA , Fc or bacteriophage coating protein such as pIX or pIII.

In some embodiments described herein, a vector is provided that contains a polynucleotide of the present invention. Such vectors may be plasmid vectors, viral vectors, baculovirus expression vectors, transposon vectors, or any other vectors suitable for introducing a polynucleotide of the present invention into a given organism or genetic environment in any way. For example, polynucleotides encoding the heavy and light chains of the bispecific antibodies of the present invention can be introduced into expression vectors. The light and heavy chains can be cloned in the same or different expression vectors. The DNA segments encoding the immunoglobulin chains may be operably linked in the expression vector(s) to control sequences for expression of the immunoglobulin polypeptides. Such control sequences include signal sequences, promoters (eg, naturally associated or heterologous promoters), enhancer elements, and transcriptional stop sequences, and can be chosen to be compatible with the host cell chosen for expression of the antibody. After introducing the vector into an appropriate host, the host can be incubated under conditions suitable for high-level expression of the proteins encoded by the introduced synthetic polynucleotides.

In general, suitable expression vectors can replicate in host organisms either as episomes or as an integral part of the host's chromosomal DNA. Typically, expression vectors contain selection markers, eg, ampicillin resistance, hygromycin resistance, tetracycline resistance, kanamycin resistance, or neomycin resistance, to enable detection of cells transformed with the desired DNA sequences.

In some embodiments described herein, a host cell containing the vector of the present invention is provided. The term host cell refers to the cell into which the vector is introduced. It should be understood that the term host cell means not only a particular cell of a subject, but also the descendants of such a cell. Since some modifications may occur in subsequent generations due to either mutation or environmental influences, such progeny may not be identical to the original cell, but may also be covered by the term host cell. Such host cells can be eukaryotic, prokaryotic, plant or archaeal cells.

Examples of eukaryotic cells may be mammalian, insect, avian or other animal cells. Mammalian eukaryotic cells include immortalized cell lines such as hybridomas or myeloma cell lines, e.g. (ECACC), Salisbury, Wiltshire, UK, ECACC No. 85110503), FO (ATCC CRL-1646) and Ag653 (ATCC CRL-1580) mouse cell lines. The human myeloma cell line U266 (ATTC CRL-TIB-196) is commonly used. Other useful cell lines include Chinese Hamster Ovary (CHO) derived lines, such as CHO-K1SV (Lonza Biologics, Walkersville, MD, USA), CHO-Ki (ATCC CRL-61) or DG44.

The use of EGFR/c-Met bispecific FN3 domain containing molecules, EGFR/c-Met bispecific antibodies and EGFR binding or c-Met FN3 domains of the present invention.

EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, c-Met binding FN3 domains, or EGFR-c-Met bispecific antibodies of the present invention can be used to diagnose, track, modulate, treat, attenuate, assisting in the prevention of development or alleviation of the symptoms of human disease or specific pathologies of cells, tissues, organs, fluid, or the host itself. The methods of the present invention can be used to treat any animal patient within any classification. Examples of such animals include mammals such as humans, rodents, dogs, cats, and farm/domestic animals.

One aspect of the present invention is a method of inhibiting the growth or proliferation of cells expressing EGFR and/or c-Met, comprising contacting the cells with an isolated EGFR/c-Met bispecific molecule containing an FN3 domain, an FN3 domain that binds to EGFR, a domain FN3 binding to c-Met or EGFR/c-Met bispecific antibody of the present invention.

Another aspect of the present invention is a method of inhibiting the growth or metastasis of EGFR and/or c-Met expressing tumor or cancer cells in a subject, comprising administering to the subject an isolated EGFR/c-Met bispecific molecule containing an FN3 domain, an FN3 domain that binds to EGFR, a cMet binding FN3 domain, or an EGFR/c-Met bispecific antibody of the present invention so as to inhibit the growth or metastasis of EGFR and/or c-Met expressing tumor or cancer cells.

Another aspect of the present invention is a method of treating a subject having cancer comprising administering a therapeutically effective amount of an isolated EGFR/c-Met bispecific molecule comprising an FN3 domain, an EGFR binding FN3 domain, a c-Met binding FN3 domain, or a bispecific to the EGFR/c-Met antibody of the present invention to a patient in need of treatment for a period of time sufficient for the treatment of cancer.

An EGFR/c-Met bispecific molecule comprising an FN3 domain, an EGFR-binding FN3 domain, or an FN3 c-Met binding domain, or an EGFR/c-Met bispecific antibody of the present invention can be used to treat any disease or a disorder characterized by abnormal activation or production of EGFR, c-Met, EGF, soluble EGF, soluble cMet, or another EGFR or HGF ligand, or a disorder associated with EGFR or c-Met expression, which may or may not be associated with malignancies or cancer, in which abnormal activation and/or production of EGFR, c-Met, EGF, or another EGFR or HGF ligand occurs in cells or tissues of a subject having a disease or disorder, or a subject predisposed to the disease or disorder.

The FN3 domains of the present invention, which specifically bind to c-Met and block the binding of HGF to c-Met, can be used to treat tumors, including cancers and benign tumors. Cancers treatable by c-Met binding FN3 domains of the present invention include those associated with c-Met overexpression. Examples of cancers treatable with the FN3 domains of the present invention include epithelial cell cancers, breast cancer, ovarian cancer, lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer. , stomach cancer, ovarian cancer, pancreatic cancer, skin cancer, oral cancer, esophageal cancer, vaginal cancer, cervical cancer, spleen cancer, testicular cancer, gastric cancer and thymus cancer.

The FN3 domains of the present invention, which specifically bind to EGFR and block the binding of EGF to EGFR, can be used to treat tumors, including cancers and benign tumors. Cancers treatable by the FN3 domains of the present invention include those associated with EGFR overexpression or variants thereof. Examples of cancers treatable by the FN3 domains of the present invention include epithelial cell cancers, breast cancer, ovarian cancer, lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer, ovarian cancer, pancreatic cancer, skin cancer, oral cancer, esophageal cancer, vaginal cancer, cervical cancer, spleen cancer, testicular cancer, stomach cancer and thymus cancer. EGFR/c-Met bispecific molecules containing an FN3 domain or EGFR/c-Met bispecific antibodies of the present invention can be used to treat tumors, including cancers and benign tumors. Examples of cancers treatable with the FN3 domain-containing EGFR/c-Met bispecific molecule of the present invention or the EGFR/c-Met bispecific antibody of the present invention are cancers overexpressing EGFR and/or c-Met, cancers with increased activity and/or level of EGFR expression (eg, EGFR activating mutation, EGFR gene amplification, or ligand-mediated EGFR activation) and increased c-Met activity and/or expression level (eg, c-Met activating mutation, c-Met gene amplification or ligand-mediated c-Met activation).

Examples of EGFR-activating mutations that may be associated with cancer include point mutations, deletion mutations, insertion mutations, inversions, or gene amplifications that result in an increase in at least one biological activity of EGFR, such as an increase in tyrosine kinase activity, the formation of receptor homodimers, or heterodimers. , improved ligand binding, and the like. Mutations can be localized to any part of the EGFR gene or regulatory regions associated with the EGFR gene and include mutations in exon 18, 19, 20, or 21 or mutations in the kinase domain. Examples of EGFR activating mutations are G719A, L861X (X is any amino acid), L858R, E746K, L747S, E749Q, A750P, A755V, V765M, L858P or T790M substitutions, E746-A750 deletion, R748-P753 deletion, Ala insertion between M766 and A767, an SVA (Ser, Val, Ala) insert between S768 and V769, and an NS (Asn, Ser) insert between P772 and H773. Other examples of EGFR activating mutations are known in the art (see, for example, US Patent Publication No. US 2005/0272083). Information on EGFR and other ErbB receptors, including homo- and heterodimeric receptors, receptor ligands, autophosphorylation sites, and signaling molecules involved in ErbB-mediated signaling, is known in the art (see, for example, Hynes and Lane, Nature Reviews Cancer, 5: 341-354, 2005).

Examples of c-Met activating mutations include point mutations, deletion mutations, insertion mutations, inversions, or gene amplifications that result in an increase in at least one biological activity of the c-Met protein, such as increased tyrosine kinase activity, formation of receptor homodimers or heterodimers, improved binding with a ligand, etc. Mutations can be localized to any part of the c-Met gene or regulatory regions associated with the gene, such as mutations in the c-Met kinase domain. Examples of c-Met activating mutations are those at residue positions N375, V13, V923, R175, V136, L229, S323, R988, S1058/T1010 and E168. Methods for detecting mutations or amplifications of the EGFR and c-Met genes are well known.

Examples of cancers treatable with the bispecific molecules of the present invention, such as the EGFR/c-Met bispecific antibodies of the present invention, are epithelial cell cancer, breast cancer, ovarian cancer, lung cancer, non-small cell lung cancer (NSCLC), pulmonary adenocarcinoma, small cell lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer, pharyngeal cancer, nose cancer, pancreatic cancer, skin cancer, oral cavity cancer, tongue cancer, esophageal cancer, vaginal cancer , cervical cancer, spleen cancer, testicular cancer, stomach cancer, thymus cancer, colon cancer, thyroid cancer, liver cancer (hepatocellular carcinoma (HCC)) or sporadic or hereditary papillary renal carcinoma (PRCC).

Another aspect of the present invention is a method of treating a subject having cancer comprising administering a therapeutically effective amount of an isolated EGFR/c-Met bispecific antibody of the present invention to a patient in need thereof for a period of time sufficient to treat the cancer, wherein the subject is homozygous for phenylalanine in position 158 in CD16 (genotype FcyRIIIa-158F/F) or heterozygous for valine and phenylalanine at position 158 in CD16 (genotype FcyRHIa-158F/V). CD16 is also known as the Fc gamma receptor IIIa (FcyRHIa) or low affinity receptor for the Fc region of immunoglobulin gamma, isoform III-A. The valine/phenylalanine (V/F) polymorphism in the FcyRIIIa protein at position 158 has been shown to negatively affect the affinity of FcyRIIIa for human IgG. The receptor with the FcyRIIIa-158F/F or FcyRHIa158F/V polymorphisms shows reduced Fc binding and thus reduced ADCC compared to FcyRHIa-158V/V. The absence or low amount of fucose in N-linked oligosaccharides increases the ability of antibodies to induce ADCC due to improved antibody binding to human FcyRHIa (CD16) (Shields et al., J. Biol. Chem. 277:26733-40, 2002). Antibodies of the present invention

- 40 039356 have a reduced fucose content in the range of about 1 to about 10%. In some embodiments, the EGFR/c-Met bispecific antibody has a glycan structure with a fucose content of about 50, 40, 45, 40, 35, 30, 25, 20, 15, 10, 9, 8, 7, 6, 5, 4 , 3, 2 or 1%. Thus, the antibodies of the present invention may be more effective in treating patients with FcyRIIIa-158F/F or FcyRIIIa-158F/V genotypes. Patients can be analyzed for their FcyRIIIa polymorphism using standard methods.

In some of the methods described herein, the antibodies of the present invention can be used to treat a subject having a cancer that is resistant or has become resistant to treatment with one or more EGFR inhibitors. Examples of EGFR inhibitors to which cancer can become resistant are the anti-EGFR antibodies cetuximab (Erbitux®), pantinumumab (Vectibix®), matuzumab, nimotuzumab, small molecule EGFR inhibitors Tarceva® (erlotinib), IRESSA (gefitinib), EKB-569 ( pelitinib, an irreversible EGFR tyrosine kinase inhibitor), pan-ErbB inhibitors and other receptor tyrosine kinase inhibitors, lapatinib (EGFR and HER2 inhibitor), pelitinib (EGFR and HER2 inhibitor), vandetanib (ZD6474, ZACTIMA™, EGFR, VEGFR2 and RET TKI), PF00299804 (dacomitinib, irreversible pan-ErbB tyrosine kinase inhibitor), CI-1033 (irreversible pan-ErbB tyrosine kinase inhibitor), afatinib (BIBW2992, irreversible pan-ErbB tyrosine kinase inhibitor), AV-412 (dual inhibitor of EGFR and ErbB2), EXEL-7647 ( inhibitor of EGFR, ErbB2, GEVGR and EphB4), CO-1686 (irreversible mutant selective EGFR tyrosine kinase inhibitor), AZD9291 (irreversible mutant selective tyrosine kinase inhibitor of EGFR) and HKI-272 (neratinib, an irreversible inhibitor of EGFR/ErbB 2). The methods described herein can be used to treat cancers that are resistant to treatment with gefitinib, erlotinib, afatinib, CO-1686, AZD9291, and/or cetuximab. An example of an antibody that can be used is EM1-mAb.

Another aspect of the present invention is a method of treating a subject having cancer comprising administering to a patient in need of treatment a therapeutically effective amount of an isolated EGFR/c-Met bispecific antibody of the present invention for a period of time sufficient to treat the cancer, wherein the subject is or has acquired resistance. to treatment with erlotinib, gefitinib, afatinib, CO-1686, AZD9291, or cetuximab.

Various qualitative and/or quantitative methods can be used to determine whether a subject has resistance, acquired resistance, or a predisposition to acquire resistance to treatment with an EGFR inhibitor. Symptoms that may be associated with resistance to an EGFR inhibitor include, for example, worsening or no improvement in the health of the patient, an increase in tumor size, cessation or slowing down of tumor growth inhibition, and/or spread of cancer cells in the body from one site to other organs, tissues or cells. An indicator that a subject has developed resistance or is predisposed to developing resistance to EGFR inhibitors can also be the reappearance or worsening of various cancer-related symptoms such as anorexia, cognitive impairment, depression, dyspnea, fatigue, hormonal disturbances, neutropenia, pain, peripheral neuropathy and sexual dysfunction. The symptoms associated with cancer can vary depending on the type of cancer. For example, symptoms associated with cervical cancer may include abnormal bleeding, unusually heavy vaginal discharge, pelvic pain not associated with a normal menstrual cycle, pain in the bladder or when urinating, and bleeding between normal periods, after sexual intercourse, douching, or pelvic examination. Symptoms associated with lung cancer may include a persistent cough, coughing up blood, shortness of breath, wheezing and chest pain, loss of appetite, weight loss without weight loss, and fatigue. Symptoms of liver cancer may include loss of appetite and weight, abdominal pain, especially in the upper right side of the abdomen, which may radiate to the back and shoulder, nausea and vomiting, general weakness and fatigue, liver enlargement, bloating (ascites), and yellow discoloration. skin and whites of the eyes (jaundice). An oncologist can easily identify the symptoms associated with a particular type of cancer.

Other methods for determining the development of resistance to an EGFR inhibitor include assaying EGFR phosphorylation, ERK1/2 phosphorylation, and/or AKT phosphorylation in cancer cells, where increased phosphorylation may be an indication that a subject has acquired resistance or a predisposition to develop resistance to an EGFR inhibitor. Methods for determining EGFR, ERK1/2 and/or AKT phosphorylation are well known and are described herein. Identification of a patient with developed resistance to an EGFR inhibitor may include detection of increased c-Met expression or increased c-Met activity, for example, associated with increased circulating HGF, an activating c-Met gene mutation, or amplification of the c-Met gene.

Another embodiment of the present invention is a method of treating NSCLC in a patient having an NSCLC tumor or metastasizing tumor having an EGFR activating mutation or EGFR gene amplification, comprising administering to the patient therapeutically

- 41 039356 an effective amount of an EGFR/c-Met bispecific antibody of the present invention.

The EGFR/c-Met bispecific antibodies of the present invention can be used to treat non-small cell lung cancer (NSCLC), including squamous cell carcinoma, adenocarcinoma, and large cell carcinoma. In some embodiments, the NSCLC cells have an epithelial phenotype. In some embodiments, the NSCLC has acquired resistance to treatment with one or more EGFR inhibitors.

In NSCLC, specific EGFR gene mutations are associated with high response rates (70-80%) to EGFR tyrosine kinase inhibitors (EGFR-TKIs). EGFR-TKI sensitivity has been associated with a 5 amino acid deletion in exon 19 or a L858R point mutation in EGFR (Nakata and Gotoh, Expert. Opin. Ther. Targets, 16:771-781, 2012). Such mutations result in ligand-independent activation of EGFR kinase activity. EGFR-activating mutations are seen in 10-30% of patients with NSCLC, and are significantly more common in East Asians, women, never-smokers, and patients with adenocarcinoma histology (Janne and Johnson, Clin. Cancer Res. 12( 14 Suppl):4416s4420s, 2006). EGFR gene amplification also correlates significantly with response after EGFR-TKI treatment (Cappuzzo et al., J. Natl. Cancer Inst. 97:643-55, 2005).

Although most NSCLC patients with EGFR mutations initially respond to EGFR TKI therapy, virtually all develop resistance preventing a sustained response. In 50-60% of patients, resistance develops due to a point mutation of the second localization in the kinase domain of EGFR (T790M). Nearly 60% of all tumors that become resistant to EGFR tyrosine kinase inhibitors increase c-Met expression, amplify the c-Met gene, or increase the level of the only known ligand, HGF (Turke et al., Cancer Cell, 17:77-88, 2010) .

Other embodiments of the present invention are a method of treating a patient having cancer comprising administering to a patient in need of treatment a therapeutically effective amount of an EGFR/c-Met bispecific antibody of the present invention for a period of time sufficient to treat the cancer, wherein the cancer is associated with an EGFR activating mutation. , EGFR gene amplification, elevated circulating HGF, c-Met activating mutation, or c-Met gene amplification.

In some embodiments, the EGFR activating mutation is a G719A, G719X (X is any amino acid), L861X (X is any amino acid), L858R, E746K, L747S, E749Q, A750P, A755V, V765M, L858P, or T790M substitution, E746 deletion -A750, R748-P753 deletion, Ala (A) insertion between M766 and A767, Ser, Val and Ala (SVA) insertion between S768 and V769, and Asn and Ser (NS) insertion between P772 and H773.

Other embodiments of the present invention are methods of treating a patient having cancer comprising administering to a patient in need of treatment a therapeutically effective amount of an EGFR/c-Met bispecific antibody of the present invention for a period of time sufficient to treat the cancer, wherein the cancer is associated with an EGFR L858R mutation. , T790M or deletion of residues E746-A750 (del (E746, A750)), EGFR amplification or c-Met amplification.

In some embodiments, the cancer is associated with wild-type EGFR and wild-type c-Met.

In some embodiments, the cancer is associated with wild-type EGFR and c-Met amplification.

In some embodiments, cancer is associated with mutations in EGFR L858R and T790M and wild-type c-Met.

In some embodiments, the cancer is associated with a deletion in EGFR del (E764, A750) and wild-type c-Met.

In some embodiments, the cancer is associated with a deletion in EGFR del (E764, A750) and c-Met amplification.

In some embodiments, the cancer is associated with a deletion in EGFR del (E764, A750), amplification of EGFR, and amplification of c-Met.

In some embodiments, the patient has NSCLC associated with mutations in EGFR L858R and T790M and wild-type c-Met.

In some embodiments, the patient has NSCLC associated with amplification of EGFR and wild-type c-Met.

In some embodiments, the patient has NSCLC associated with EGFR amplification and c-Met amplification.

In some embodiments, the patient has an NSCLC associated with a deletion in EGFR del (E764, A750) and wild-type c-Met.

In some embodiments, the patient has an NSCLC associated with a deletion in EGFR del (E764, A750) and c-Met amplification.

In some embodiments, patients are treated with an EM1-mAb of the present invention. The EM1-mAb of the present invention demonstrates efficacy in in vivo animal tumor models when tumors are associated with L858R, T790M, del(E746, A750) mutations in EGFR, EGFR amplification, wild-type c-Met and/or c-Met amplification. Amplification of EGFR or c-Met can be assessed by standard methods, for example, by determining the number of copies of the EGFR gene or c-Met

- 42 039356 using Southern blotting, FISH or comparative genomic hybridization (CGH).

Other embodiments of the present invention are a method of treating a patient having cancer comprising administering to a patient in need of treatment a therapeutically effective amount of an EGFR/c-Met bispecific antibody of the present invention for a period of time sufficient to treat the cancer, wherein the cancer is associated with an EGFR L858R mutation. , T790M or deletion of residues E746-A750 (del (E746, A750)), EGFR amplification or c-Met amplification and mutant KRAS.

In some embodiments, the mutant KRAS has a G12V substitution. KRAS belongs to the RAS family of proto-oncogenes encoding guanosine triphosphatases (GTPases) and mediates EGFR signal transduction downstream of the receptor. Thus, tumors with mutations in the proto-oncogenic KRAS, such as those with an activating mutation in G12V or G12C, are expected to be unresponsive to treatment with EGFR antibodies. Clinical studies of anti-EGFR antibodies such as cetuximab and panitumumab have demonstrated that patients with KRAS-mutated colorectal tumors do not respond to treatment with these agents (Van Cutsem et al., N. Eng. J. Med. 360:1408-1417 , 2009; Lievre et al., J. Clin. Oncol. 26:374-379, 2008; Amado et al., J. Clin. Oncol. 26:1626-1634m 2008). The EGFR/c-Met bispecific antibodies of the present invention mediate the killing of a KRAS mutant cell line by efficient ADCC and thus, unlike existing anti-EGFR antibody therapies, may be effective in treating patients whose cancer is associated with mutations that activate KRAS. An example of such an antibody is EM1-mAb.

The terms "treat" or "treatment" refer to both therapeutic treatment and prophylactic or preventive measures, the aim being to prevent or slow down (reduce) an undesired physiological change or disorder, such as the development or spread of cancer. For the purposes of the present invention, advantageous or desirable clinical results include, but are not limited to, relief of symptoms, reduction in the degree of disease, stabilization of the condition (i.e., no worsening), delay or slowing of the progression of the disease, improvement or temporary improvement in the condition, and remission (complete or partial) both discoverable and undetectable. Treatment can also mean prolongation of life compared to that expected in the absence of treatment. Subjects in need of treatment include those who already have the condition or disorder, as well as those who are predisposed to the condition or disorder, or those whose condition or disorder is to be prevented.

The term "therapeutically effective amount" refers to an amount effective at the dosages and for periods of time necessary to achieve the desired therapeutic result. The therapeutically effective amount of the EGFR/c-Met bispecific antibody of the present invention may vary depending on factors such as the disease state, age, sex, and weight of the individual, as well as the ability of the EGFR/c-Met bispecific antibody of the present invention to elicit the desired response in the individual. . Examples of performance measures of an EGFR/c-Met therapeutic agent that may decrease or decrease with resistance include, for example, improved patient health, reduction or reduction in tumor size, cessation or slowing of tumor growth, and/or absence of cancer cell metastasis to other sites. in the body.

Introduction/pharmaceutical compositions.

The present invention provides pharmaceutical compositions comprising an EGFR/c-Met bispecific antibody of the present invention and a pharmaceutically acceptable carrier. For therapeutic use, EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, c-Met binding FN3 domains, or EGFR/c-Met bispecific antibodies of the present invention can be formulated into pharmaceutical compositions, containing an effective amount of the domain, molecule or antibody as the active ingredient in a pharmaceutically acceptable carrier. The term carrier refers to a diluent, adjuvant, excipient or carrier with which the active compound is administered. Such carriers may be liquids such as water and oils, including oils derived from petroleum, animal, vegetable or synthetic oils such as peanut oil, soybean oil, mineral oil, sesame oil, and the like. For example, 0 .4% saline solution and 0.3% glycine solution. These solutions are sterile and essentially free of solid impurities. Sterilization is then carried out using standard well known sterilization techniques (eg filtration). The compositions may contain pharmaceutically acceptable adjuvants necessary for the respective physiological conditions, such as pH adjusting and buffering agents, stabilizing, thickening, wetting and coloring agents, and the like. The concentration of the molecules or antibodies of the present invention in such a pharmaceutical composition can vary considerably, i. e. from less than about 0.5%, typically at least about 1% and up to 15 or 20% by weight, and is determined primarily based on the required dose, fluid volumes, viscosity values, and the like. according to the specific route of administration chosen. Suitable carriers and formulations, including other human proteins, such as human serum albumin, are described, for example, in

- 43 039356

Remington: The Science and Practice of Pharmacy, 21st Edition, Troy, DB ed., ^Lipincott Williams and Wilkins,

Philadelphia, PA 2006, Part 5, Pharmaceutical Manufacturing, p. 691-1092, see especially p. 958-989.

The route of administration for the therapeutic use of EGFR/c-Met bispecific FN3 domain-containing molecules, EGFR-binding FN3 domains, c-Met-binding FN3 domains, or EGFR/c-Met bispecific antibodies of the present invention may be any suitable a method for delivering the agent to the host, such as parenteral administration, for example, intradermal, intramuscular, intraperitoneal, intravenous or subcutaneous, pulmonary, transmucosal (oral, intranasal, intravaginal, rectal), administration using a formulation in the form of a tablet, capsule, solution, powder, gel, particles, and administration of the contents in a syringe, implantable device, osmotic pump, cassette, micropump, or other means known to those skilled in the art and also known in the art. Specific administration to the area can be provided, for example, by delivery to the joint, bronchi, abdominal cavity, inside the capsule, cartilage, cavity, cell, cerebellum, cerebral ventricle, colon, cervix, stomach, liver, myocardium, bone, pelvis, pericardium, abdominal cavity, pleura, prostate, lungs, rectum, kidney, retina, spine, articular bag, chest, uterus, vessel, inside the bladder, into damaged tissue, vaginally, rectally, buccally, sublingually, intranasally or transdermally.

Thus, the pharmaceutical composition of the present invention for intramuscular injection can be prepared with a content of 1 ml of sterile buffered water and from about 1 ng to about 100 mg/kg, for example, from about 50 ng to about 30 mg/kg, or more preferably from about 5 to approximately 25 mg/kg of EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, or c-Met binding FN3 domains of the present invention.

The EGFR/c-Met bispecific antibodies of the present invention can be administered to a patient by any suitable route, for example parenterally, by intravenous (IV) infusion or bolus injection, intramuscularly, subcutaneously, or intraperitoneally. Intravenous infusion can be given for as little as 15 minutes, but more often for 30, 60, 90 minutes, or even 2 or 3 hours. . The dose administered to a patient having cancer is sufficient to alleviate or at least partially inhibit the disease being treated (therapeutically effective amount), and may sometimes be from 0.1 to 10 mg/kg of body weight, for example 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 mg/kg, but may be even higher, such as 15, 20, 30, 40, 50, 60, 70, 80, 90 or 100 mg/kg. It is also possible to provide a fixed single dose, for example 50, 100, 200, 500 or 1000 mg, or the dose may depend on the surface area of the patient's body, for example 400, 300, 250, 200 or 100 mg/m ² . For the treatment of cancer, 1 to 8 doses are usually administered (eg, 1, 2, 3, 4, 5, 6, 7, or 8), but 10, 12, 20 or more doses may be administered. Administration of the EGFR/c-Met bispecific antibody of the present invention may be repeated after one day, two days, three days, four days, five days, six days, one week, two weeks, three weeks, one month, five weeks, six weeks, seven weeks, two months, three months, four months, five months, six months or more. Repeated courses of treatment in the form of chronic use are also possible. Repeated administration can be carried out at the same dose or at a different dose.

For example, a pharmaceutical composition containing the EGFR/c-Met bispecific antibodies of the present invention for intravenous infusion can be formulated to contain about 200 ml of sterile Ringer's solution and about 8 to about 2400 mg, about 400 to about 1600 mg or about 400 to about 800 mg of an EGFR/c-Met bispecific antibody for administration to an 80 kg patient. Methods for preparing compositions for parenteral administration are well known and described in more detail, for example, in Remington's Pharmaceutical Science, 15th ^ed ., Mack Publishing Company, Easton, PA.

EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, c-Met binding FN3 domains, and EGFR/c-Met bispecific antibodies of the present invention can be lyophilized for storage and reconstituted in a suitable carrier. This technique has been shown to be effective for standard protein preparations, and well known lyophilization and reconstitution techniques can be used.

EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, c-Met binding FN3 domains, and EGFR/c-Met bispecific antibodies of the present invention can be administered in combination with a second therapeutic agent simultaneously, sequentially or separately. The second therapeutic agent may be a chemotherapeutic agent or a targeted cancer therapy agent.

The EGFR/c-Met bispecific antibody of the present invention can be administered together with any one or more of the chemotherapeutic drugs or other anticancer therapeutics known in the art. Chemotherapeutic agents are chemical compounds suitable for the treatment of cancer, include growth inhibitory or other cytotoxic agents, and include alkylating agents, antimetabolites, microtubule inhibitors, topoisomerase inhibitors, receptor tyrosine kinase inhibitors, growth inhibitors

- 44 039356 vessels, etc. Examples of chemotherapeutic agents include alkylating agents such as thiotepa and cyclophosphamide (CYTOXAN®); alkylsulfonates such as bisulfan, improsulfan and piposulfan; aziridines such as benzodopa, carboquine, meturedopa and uredopa; ethyleneimines and methylamelamines, including altretamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramine, and trimethylolomelamin; nitrogen mustards such as chlorambucil, chlornaphasine, chlorphosphamine, estramustine, ifosfamide, mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembichin, phenesterin, prednimustine, trofosfamide, uracil mustard; nitrosoureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, ranimustine; antibiotics such as aclacinomycins, actinomycin, autramycin, azaserin, bleomycins, cactinomycins, calicheamicin, carabicin, carminomycin, carsinophilin, chromomycins, dactinomycin, daunorubicin, detorubicin, 6-diazo-5-oxo-b-norleucine, doxorubicin, epirubicin, esorubicin, idarubicin, marcellomycin, mitomycins, mycophenolic acid, nogalamycin, olivomycins, peplomycin, potfiromycin, puromycin, quelamycin, rhodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-FU; folic acid analogs such as denopterin, methotrexate, pteropterin, trimetrexate; purine analogs such as fludarabine, 6-mercaptopurine, thiamiprin, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxyfluridine, enocytabine, floxuridine; androgens such as calusterone, dromostanolone propionate, epitiostanol, mepitiostane, testolactone; antiadrenergics such as aminoglutethimide, mitotane, trilostane; folic acid compensator such as frolinic acid; aceglatone; aldophosphamide glycoside; aminolevulinic acid; amsacrine; bestrabucil; bisantrene; edatraksat; defofamine; demecolcin; diazichion; elfornitin; elliptinium acetate; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidamine; mitoguazone; mitoxantrone; mopidamol; nitracrine; pentostatin; phenamet; pyrarubicin; podophyllic acid; 2-ethylhydrazide; procarbazine; PSK®; razoxane; sisofiran; spirogermanium; tenuazonic acid; triazichion; 2,2',2-trichlorotriethylamine; urethane; vindesine; dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; hacytosine; arabinoside (Ara-C); cyclophosphamide; tirotep; members of the taxoid or taxane family such as paclitaxel (TAXOL® docetaxel (TAXOTERE®) and its analogs; chlorambucil; gemcitabine; 6-thioguanine; mercaptopurine; methotrexate; platinum analogs such as cisplatin and carboplatin; vinblastine; platinum; etoposide (VP- 16); ifosfamide; mitomycin C; mitoxantrone; vincristine; vinorelbine; navelbin; novantrone; teniposide; daunomycin; aminopterin; xeloda; ibandronate; CPT-11; topoisomerase inhibitor RFS 2000; difluoromethylornithine (DMFO); retinoic acid; esperamycins; capecitabine; inhibitors receptor tyrosine kinases and/or angiogenesis such as sorafenib (NEXAVAR®), sunitinib (SUTENT®), pazopanib (VOTRIENT™), toceranib (PALLADIA™), vandetanib (ZACTIMA™), cediranib (RECENTIN®), regorafenib (BAY 73- 4506), axitinib (AG013736), lestaurtinib (CEP-701), erlotinib (TARCEVA®), gefitinib (IRESSA™), BIBW 2992 (TOVOK™), lapatinib (TYKERB®), neratinib (HKI-272), etc. ., and pharmaceutically acceptable salts, acids or derivatives of any of the above compounds. this definition also includes antihormonal agents that regulate or inhibit the effect of hormones on a tumor, such as antiestrogen substances, including, for example, tamoxifen, raloxifene, aromatase-inhibiting 4(5)-imidazoles, 4-hydroxytamoxifen, trioxifene, keoxifene, LY 117018, onapristone and toremifene (FARESTON®); as well as antiandrogenic agents such as flutamide, nilutamide, bicalutamide, leuprolide and goserelin; and pharmaceutically acceptable salts, acids and derivatives of any of the above compounds. Other conventional cytotoxic chemicals described in Wiemann et al., 1985, in Medical Oncology (Calabresi et aL, eds.), Chapter 10, McMillan Publishing, are also applicable to the methods of the present invention.

Examples of agents that can be used in combination with EGFR/c-Met bispecific molecules containing an FN3 domain, EGFR binding FN3 domains, c-Met binding FN3 domains, and EGFR/c-Met bispecific antibodies of the present invention , include tyrosine kinase inhibitors and targeted cancer drugs such as Iressa® (gefitinib) and Tarceva (erlotinib), as well as other HER2, HER3, HER4 or VEGF antagonists. Examples of HER2 antagonists include CP-724-714, HERCEPTIN™ (trastuzumab), OMNITARG™ (pertuzumab), TAK-165, lapatinib (an EGFR and HER2 inhibitor), and GW-282974. Examples of HER3 antagonists include antibodies to Her3 (see, for example, US Patent Publication No. US 2004/0197332). Examples of HER4 antagonists include siRNA to HER4 (see, for example, Maatta et al., Mol. Biol. Cell, 17:67-79, 2006. An example of a VEGF antagonist is bevacizumab (Avastin™).

When a small molecule compound is used in combination with an EGFR/c-Met bispecific antibody of the present invention, it is generally administered more frequently, preferably once a day, but it is also possible to administer 2, 3, 4 or more times a day, as well as every two days, once a week or at some other interval. Small molecule drugs are often administered orally, but parenteral administration, such as IV infusion, bolus injection, subcutaneous or intramuscular administration, is also possible. Doses of small molecule drugs are usually

- 45 039356 may be from 10 to 1000 mg or about 100, 150, 200 or 250 mg.

When the EGFR/c-Met bispecific antibody of the present invention is administered in combination with a second therapeutic agent, the combination may take place at any convenient time frame. For example, the EGFR/c-Met bispecific antibody and the second therapeutic agent can be administered to a patient on the same day and even in the same intravenous infusion. However, the EGFR/c-Met bispecific antibody and the second therapeutic agent may also be administered on alternating days or alternating weeks, biweekly intervals or months, etc. In some methods, the EGFR/c-Met bispecific antibody and the second therapeutic agent are administered close enough in time to each other that they are simultaneously present (eg, in serum) at detectable levels in the patient being treated. In some methods, the entire course of treatment with the EGFR/c-Met bispecific antibody consists of administering a number of doses over a period of time followed or preceded by a course of treatment with a second therapeutic agent that also consists of administering a number of doses. In some methods, treatment with a second EGFR/c-Met bispecific antibody is initiated if the patient is or develops resistance to the second therapeutic agent administered initially. The patient may receive only one course or multiple treatments with one or both of the EGFR/c-Met bispecific antibody and the second therapeutic agent. A recovery period of 1, 2, or several days or weeks can be used between administration of the EGFR/c-Met bispecific antibody and the second therapeutic agent. Once a suitable second therapeutic agent treatment regimen has been established, this regimen may be used in combination with the EGFR/c-Met bispecific antibody of the present invention. For example, Tarceva® (erlotinib) is taken as a 100 or 150 mg pill once a day, and Iressa® (gefitinib) is taken as a 250 mg tablet once a day.

The EGFR/c-Met bispecific antibody, optionally in combination with a second therapeutic agent, can be administered with any form of radiation therapy, including external beam radiation, intensity-modulated radiation therapy (IMRT), as well as any form of radiosurgery, including Gamma Knife, Cyberknife, Linac, and interstitial irradiation (eg, implanted radioactive seeds, GliaSite balloon), and/or surgical treatment. Combination with radiation therapy is especially appropriate for the treatment of head and neck cancer and brain tumors.

Although the present invention has been described in general terms, embodiments of the present invention are described in more detail in the following examples, which should not be construed as limiting the scope of the claims.

Example 1: Creating Tencon Libraries.

Tencon (SEQ ID NO: 1) is a type III fibronectin (FN3) framework immunoglobulin-like domain developed from the consensus sequence of the 15 human tenascin-C FN3 domains (Jacobs et al., Protein Engineering, Design, and Selection, 25: 107-117, 2012; US Patent Publication No. 2010/0216708). Tencon's crystal structure shows six surface loops connecting seven beta strands. These loops or selected residues within each loop can be randomized to create a library of type III fibronectin (FN3) domains that can be used to select new molecules that bind to specific targets.

Tencon:

LPAPKNLVVSEVTEDSLRLSWTAPDAAFDSFLIQYQESEKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIY GVKGGHRSNPLSAEFTT (SEQ ID NO 1):

Creation of the TCL1 library.

A library designed to randomize only the FG loop in Tencon (SEQ ID NO: 1), TCL1, was created for use with a cis display system (Jacobs et al., Protein Engineering, Design, and Selection, 25:107-117, 2012 ). This system generates a single-stranded DNA that includes the sequences for the Tac promoter, the Tencon library code sequence, the RepA code sequence, the cis element, and the ori element. When expressed in an in vitro transcription/translation system, a Tencon-RepA fusion protein complex is formed that is cis-linked to the DNA that encodes it. The complexes binding to the target molecule are then isolated and amplified by polymerase chain reaction (PCR) as described below.

Creation of the TCL1 library for use with cis display is achieved by successive PCR cycles to produce the final linear double-stranded DNA molecules from the two halves; The 5-fragment contains the Tencon promoter and sequences, while the 3-fragment contains the repA gene and the cis and ori elements. These two halves are combined with a restriction fragment to obtain a complete construct. The TCL1 library was designed to include random amino acids only in the FG loop in Tencon, KGGHRSN (SEQ ID NO: 86). When creating this library, NNS codons were used, which made it possible to insert all 20 amino acids and one stop codon into the FG loop. The TCL1 library contains six separate sub-libraries, each with a different randomized FG loop length, from 7 to 12 residues, to further increase diversity. The Tencon-based library configuration is shown in Table. 2.

- 46 039356

table 2

Library BC Loop Configuration FG loop configuration WT Tencon TAPDAAFD* KGGHRSN** TCL1 TAPDAAFD* XXXXXXXXX xxxxxxxxx xxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx TCL2 ########

* TAPDAAFD: residues 22-28 in SEQ ID NO: 1;

* *KGGHRSN: SEQ ID NO: 86.

X denotes degenerate amino acids encoded by NNS codons.

# denotes the designed distribution of amino acids described in the text.

To generate the TCL1 library, serial PCR cycles were performed to attach the Tac promoter, introduce degenerate elements into the FG loop, and add the necessary restriction sites for final assembly. First, a DNA sequence containing the promoter sequence and the Tencon FG loop 5'-loop sequence was created in two steps. DNA corresponding to the full sequence of the Tencon gene was used as a template for PCR, with primers POP2220 (SEQ ID NO: 2) and TC5'toFG (SEQ ID NO: 3). The PCR product obtained from this reaction was used as a template for the next round of PCR amplification with primers 130mer (SEQ ID NO: 4) and Tc5'toFG to complete the addition of the 5' fragment sequence and promoter sequence to Tencon. Diversification was then introduced into the FG loop by amplifying the DNA product obtained in the first step with forward primer POP2222 (SEQ ID NO: 5) and reverse primers TCF7 (SEQ ID NO: 6), TCF8 (SEQ ID NO: 7), TCF9 (SEQ ID NO: 8), TCF10 (SEQ ID NO: 9), TCF11 (SEQ ID N NO: 10) or TCF12 (SEQ ID NO: 11) containing degenerate nucleotides. Performed at least eight PCR reactions of 100 μl for each sub-library to minimize the number of PCR cycles and maximize library diversity. At least 5 μg of this PCR product was gel purified and used in the next step with primers POP2222 (SEQ ID NO: 5) and POP2234 (SEQ ID NO: 12), resulting in 6xHis-Tega and NotI restriction site attachment to 3 ' End of the Tencon sequence. This PCR reaction was performed using only 15 PCR cycles and at least 500 ng of template DNA. The resulting PCR product was gel purified, digested with NotI restriction enzyme, and purified on a Qiagen column.

The 3'-fragment of the library is a constant DNA sequence containing display elements, including the PspOMI restriction site encoding the repA gene region and cis- and ori-elements. PCR was performed using a plasmid (pCR4Blunt) (Invitrogen) containing this DNA fragment, with primers M13 Forward (forward) and M13 Reverse (reverse). The resulting PCR products were digested with PspOMI overnight and purified from the gel. To ligate the 5-part of the library DNA with the 3' DNA fragment containing the repA gene, 2 pmol of the 5' DNA fragment was ligated with an equimolar amount of the 3' repA DNA fragment in the presence of NotI and PspOMI enzymes and T4 ligase. After ligation overnight at 37° C., a small portion of the ligated DNA was run through the gel to check the effectiveness of the ligation. The ligated library product was divided into twelve PCR amplifications and PCR was performed in 12 cycles with the primer pair POP2250 (SEQ ID NO: 13) and DidLigRev (SEQ ID NO: 14). The DNA yield for each sublibrary in the TCL1 library was in the range of 32-34 μg.

For library quality analysis, a small portion of the working library was amplified with primers Tcon5new2 (SEQ ID NO: 15) and Tcon6 (SEQ ID NO: 16) and cloned into the modified pET vector by ligation-free cloning. Plasmid DNA was transformed into BL21-GOLD (DE3) competent cells (Stratagene) and 96 randomly selected colonies were sequenced using the T7 promoter primer. Duplicate sequences were not found. Overall, approximately 70-85% of the clones had the complete promoter and Tencon coding sequence without the frameshift mutation. Fraction of functional sequencing that excludes clones

- 47 039356 with stop codons, ranged from 59 to 80%.

Creation of the TCL2 library.

A TCL2 library was created in which both the BC loop and the FG loop were randomized in Tencon, and the distribution of amino acids at each position was tightly controlled. In table. 3 shows the distribution of amino acids at desired loop positions in the TCL2 library. The developed distribution of amino acids had two purposes. First, the library was targeted to residues predicted to be structurally significant for Tencon folding and stability based on Tencon crystal structure analysis and/or homology modeling. For example, position 29 was only intended for a set of hydrophobic amino acids, as this residue is embedded in the hydrophobic core of the folded Tencon. The second level of development included shifting the distribution of amino acids to a distribution of residues that are preferentially found in the heavy chain of HCDR3 antibodies to efficiently produce compounds with high binding affinity (Birtalan et al., J. Mol. Biol. 377:1518-28, 2008; Olson et al., Protein Sci. 16:476-84, 2007). Thus, the developed distribution from Table. 3 denotes the following distribution: 6% alanine, 6% arginine, 3.9% asparagine, 7.5% aspartic acid, 2.5% glutamic acid, 1.5% glutamine, 15% glycine, 2.3% histidine, 2.5% isoleucine, 5% leucine, 1.5% lysine, 2.5% phenylalanine, 4% proline, 10% serine, 4.5% threonine, 4% tryptophan, 17.3% tyrosine and 4% valine. This distribution lacks methionine, cysteine, and stop codons.

Table 3

Residual position* Remains wild type Distribution in TCL2 library 22 T Designed Distribution 23 BUT Designed Distribution 24 R 50% R + developed distribution 25 D Designed Distribution 26 BUT 20% A + 20% G + developed distribution 27 BUT Designed Distribution 28 F 20% F, 20% I, 20% L, 20% V, 20% Y 29 D 33% D, 33% E, 33% T 75 To Designed Distribution 76 G Designed Distribution 77 G Designed Distribution 78 H Designed Distribution 79 R Designed Distribution 80 S 100% S 81 N Designed Distribution 82 R 50% R + developed distribution

*Residue numbering is based on Tencon SEQ ID NO: 1.

The 5' fragment of the TCL2 library contained the Tencon promoter and coding region (SEQ ID NO: 1), which were chemically synthesized as a library pool (Sloning Biotechnology). This DNA pool contained at least ^1x10 different elements. A BsaI restriction site was included at the end of the fragment for ligation with RepA.

The 3-fragment of the library was a constant DNA sequence containing display elements, including the 6xHis tag encoding the repA region and the cis element. DNA was generated in a PCR reaction using the existing DNA template (above) and primers LS1008 (SEQ ID NO: 17) and DidLigRev (SEQ ID NO: 14). To assemble the final TCL2 library, a total of 1 μg of BsaI digested 5-fragment of Tencon library DNA was ligated with 3.5 μg of the 3' fragment obtained by restriction digestion with the same enzyme. After overnight ligation, the DNA was purified on a Qiagen column and the DNA was quantified by measuring absorbance at 260 nm. The ligated library product was PCR amplified for 12 cycles with primer pair POP2250 (SEQ ID NO: 13) and DidLigRev (SEQ ID NO: 14). A total of 72 reactions were performed, each of which used 50 ng of ligated DNA products as a template. The total yield of the working TCL2 DNA library was approximately 100 μg. A small portion of the working library was subcloned and sequenced as described above for the TCL1 library.

- 48 039356

Duplicate sequences were not found. Approximately 80% of the sequences contained the complete promoter and Tencon coding sequences without frameshift mutations.

Creation of the TCL14 library.

The upper (BC, DE, and FG) and lower (AB, CD, and EF) loops, for example, the marked binding surfaces in the FN3 domains, are separated by beta strands that form the center of the FN3 structure. Alternate surfaces found on both sides of the FN3 domains and having different shapes than the surfaces formed only by loops are formed on one side of the FN3 domain by two antiparallel beta strands, the C and F beta strands and the CD and FG loops, and are referred to herein as the C surface. -CD-F-FG.

A library was generated randomizing an alternate Tencon surface by randomizing selected surface-exposed residues related to strands C and F, as well as portions of the CD and FG loops, as shown in FIG. 1. A variant of Tencon, Tencon27 (SEQ ID NO: 99) was used to create the library, having the following replacements compared to Tencon (SEQ ID NO: 1): E11R L17A, N46V, E86I. For a full description of the methods used to develop this library, see US Patent Publication No. US 2013/0226834.

Example 2 Selection of Type III Fibronectin (FN3) Domains Binding to EGFR and Inhibiting EGF Binding.

Library screening.

A cis display was used to select EGFR binding domains from the TCL1 and TCL2 libraries. Recombinant human extracellular domain EGFR fused to IgG1 Fc (R&D Systems) was biotinylated by standard methods and used for panning (residues 25-645 of full-length EGFR of SEQ ID NO: 73). For in vitro transcription and translation (ITT), 2-6 μg of library DNA were incubated with 0.1 mM of the complete set of amino acids, components of the 1x S30 premix, and 30 μl of S30 extract (Promega) in a total volume of 100 μl and incubated at 30°C . After 1 h, 450 μl of blocking solution (PBS pH 7.4 supplemented with 2% bovine serum albumin, 100 μg/ml herring sperm DNA and 1 mg/ml heparin) was added and the reaction mixture was incubated on ice for 15 min. EGFR-Fc:EGF complexes were assembled at 1:1 and 10:1 molar ratios of EGFR to EGF by mixing recombinant human EGF (R&D Systems) with biotinylated recombinant EGFR-Fc in blocking solution for 1 hour at room temperature. For binding, 500 µl of blocked ITT reactions were mixed with 100 µl of EGFR-Fc:EGF complexes and incubated for 1 h at room temperature, after which the bound complexes were recovered with neutravidin or streptavidin magnetic beads (Seradyne). Unbound library elements were removed by successive washes with PBST and PBS. After washing, the DNA was eluted from the bound complexes by heating to 65°C for 10 min, amplified by PCR and attached to the DNA fragment encoding RepA by restriction digestion and ligation for additional panning rounds. Compounds with high binding affinity were isolated by successively lowering the concentration of EGFR-Fc target in each cycle from 200 to 50 nM and increasing the stringency of the washing conditions. In cycles 4 and 5, unbound and loosely bound FN3 domains were removed by washing in the presence of a 10-fold molar excess of non-biotinylated EGFR-Fc overnight in PBS.

After panning, the selected FN3 domains were PCR amplified with Tcon5new2 (SEQ ID NO: 15) and Tcon6 (SEQ ID NO: 16) oligonucleotides, subcloned into a pET vector, modified to include a ligase-free cloning site, and transformed into BL21-GOLD (DE3) cells ( Stratagene) to express the soluble product in E. coli using standard molecular biology techniques. A gene sequence encoding a C-terminal polyhistidine tag was added to each FN3 domain to allow purification and detection. Cultures were grown to achieve an optical density of 0.6-0.8 in 2YT medium supplemented with 100 μg/ml carbenicillin in 1 ml 96-well blocks at 37°C, then IPTG was added to 1 mM, at which point the temperature was reduced to 30 °C. Cells were harvested approximately 16 hours later by centrifugation and frozen at -20°C. Cell lysis was performed by incubating each pellet in 0.6 ml BugBuster® HT lysis buffer (Novagen EMD Biosciences) with shaking at room temperature for 45 minutes.

Selection of FN3 domains that bind to EGFR on cells.

To analyze the ability of different FN3 domains to bind to EGFR in a more physiological context, their ability to bind to A431 cells was measured. A431 cells (American Type Culture Collection, cat. no. CRL-1555) overexpress EGFR ~2 x 10 ⁶ receptors per cell. Cells were seeded at 5000/well in opaque black 96-well plates, allowed to adhere overnight at 37° C. in a humidified atmosphere with 5% CO ₂ . Lysates of FN3-expressing bacteria were diluted 1000-fold in FACS staining buffer (Becton Dickinson) and incubated for 1 hour at room temperature in three plates. The lysates were removed and the cells were washed 3 times with 150 µl/well of FACS staining buffer added to the wells. Cells were incubated with 50 µl/well of His-Alexa488 anti-pentahistidine antibody conjugate (Qiagen) diluted 1:100 in FACS staining buffer for 20 min at room temperature. Cells were washed 3 times with 150 µl/well FACS staining buffer, after which 100 µl FACS staining buffer was added to the cells and fluorescence was recorded at 488 nm using an Acumen eX3 reader. Bacterial lysates containing FN3 domains were screened for their ability to bind to A431 cells (1320 crude bacterial lysates from TCL1 and TCL2 libraries) and 516 positive clones were found with binding >10-fold greater than background signal. For binding, 300 lysates from the TCL14 library were screened to obtain 58 unique EGFR-binding FN3 domain sequences.

Selection of FN3 domains that inhibit EGF binding to EGFR on cells.

To better characterize the EGFR binding mechanism, the ability of the various FN3 domain clones found to bind to EGFR in competition with EGF was measured using A431 cells in parallel with an A431 binding screening assay. A431 cells were seeded at 5000/well in opaque black 96-well plates, allowed to adhere overnight at 37°C in a humidified atmosphere with 5% CO ₂ . Cells were incubated with 50 μl/well diluted 1:1000 bacteria lysate for 1 hour at room temperature in three plates. Biotinylated EGF (Invitrogen, Cat # E-3477) was added to each well to a final concentration of 30 ng/mL and incubated for 10 minutes at room temperature. Cells were washed 3 times with FACS staining buffer at 150 μl/well. Cells were incubated with 50 μl/well streptavidin phycoerythrin conjugate (Invitrogen) diluted 1:100 in FACS staining buffer for 20 min at room temperature. Cells were washed 3 times with 150 µl/well FACS staining buffer, after which 100 µl FACS staining buffer was added to the cells and fluorescence was recorded at 600 nm using an Acumen eX3 reader.

Screening for bacterial lysates containing FN3 domains was performed in the EGF competition assay described above. 1320 crude bacterial lysates from the TCL1 and TCL2 libraries were screened, yielding 451 positive clones with >50% inhibition of EGF binding.

Expression and purification of identified EGFR-binding FN3 domains.

His-tagged FN3 domains were purified from E. coli lysates using His MultiTrap™ HP plates (GE Healthcare) and eluted in buffer containing 20 mM sodium phosphate, 500 mM sodium chloride, and 250 mM imidazole, pH 7.4. Purified assay samples were buffer exchanged with PBS, pH 7.4 using PD MultiTrap™ G-25 plates (GE Healthcare).

Analysis by size exclusion chromatography.

Size exclusion chromatography (SEC) was used to assess the state of aggregation of FN3 domains that bind to EGFR. Aliquots (10 μl) of each purified FN3 domain were injected into a Superdex 75 5/150 column (GE Healthcare) at a flow rate of 0.3 ml/min in PBS pH 7.4 mobile phase. The elution from the column was monitored by absorbance at 280 nm. FN3 domains showing high levels of aggregation, in accordance with the SEC, were excluded from additional analysis.

Dissociation rate of selected EGFR-binding FN3 domains from EGFR-Fc.

Selected EGFR-binding FN3 domains were screened to identify domains with low dissociation constants (k _off ) when binding to EGFR-Fc on a ProteOn XPR-36 instrument (Bio-Rad) to simplify the selection of compounds with high binding affinity. Goat anti-human Fc IgG (R&D systems) at a concentration of 5 μg/ml was directly immobilized by amine coupling (at pH 5.0) in all 6 horizontally aligned ligand channels on the chip at a flow rate of 30 μl/min in PBS containing 0.005% Tween-20. Immobilization density values averaged approximately 1500 response units (RU) with less than 5% scatter between different channels. EGFR-Fc was surface-captured with anti-human IgG Fc antibodies at a density of approximately 600 RU in vertical ligand orientation. All tested FN3 domains were normalized to a concentration of 1 μm and tested for their binding in a horizontal orientation. For FN3 domains, all 6 analyte channels were used to maximize screening throughput. The dissociation phase was monitored for 10 min at a flow rate of 100 μl/min. Binding signals between spots were used as references for monitoring non-specific binding between analytes and the surface with immobilized IgG, and these signals were subtracted from all binding signals. Processed binding data was locally fitted with a simple 1:1 Langmuir binding model to obtain kf for each FN3 domain binding to captured EGFR-Fcs.

Inhibition of EGF-stimulated EGFR phosphorylation.

Purified EGFR-binding FN3 domains were tested for their ability to inhibit EGF-stimulated EGFR phosphorylation in A431 cells at a single concentration. Phosphorylation of EGFR was monitored using the EGFR phospho (Tyr1173) kit (Meso Scale Discovery). Cells were plated at 20,000/well in clear 96-well culture-treated plates.

- 50 039356 tissue (Nunc), into RPMI 100 µl/well (Gibco) containing GlutaMAX™ supplemented with 10% fetal bovine serum (FBS) (Gibco) and allowed to adhere overnight at 37°C in a humid atmosphere containing 5% CO2. The culture medium was completely removed and the cells were kept overnight under fasting conditions in 100 μl/well of FBS-free medium at 37°C in a humid atmosphere containing 5% CO2. The cells were then treated with 100 µl/well of pre-warmed (37°C) nutrient-free medium containing EGFR-binding FN3 domains at a concentration of 2 µM for 1 h at 37°C in a humidified atmosphere containing 5% CO2. Controls were treated with nutrient-free media only. Cells were stimulated by adding and gently mixing 100 µl/well of warmed (37°C) nutrient-free medium containing 100 ng/ml recombinant human EGF (R&D Systems, Cat # 236-EG) to final concentrations of 50 ng/ml EGF , as well as 1 μM EGFR-binding FN3 domain and incubation at 37°C, 5% CO2 for 15 min. One set of control wells was left unstimulated as negative controls. The medium was completely removed and the cells were lysed with 100 μl/well Complete Lysis Buffer (Meso Scale Discovery) for 10 min at room temperature with shaking as described in the manufacturer's instructions. Assay plates designed to measure EGFR phosphorylated at tyrosine residue 1173 (Meso Scale Discovery) were blocked with the supplied blocking solution according to the manufacturer's instructions at room temperature for 1.5-2 hours. The plates were then washed 4 times using 200 µl /well 1X Tris Wash Buffer (Meso Scale Discovery). Cell lysate aliquots (30 µl/well) were transferred to assay plates that were sealed with plate sealing film (VWR) and incubated at room temperature with shaking for 1 hour. Assay plates were washed 4 times with 200 µl/well Tris Wash Buffer, after which 25 µl of ice-cold detection antibody solution (Meso Scale Discovery) was added to each well, trying to avoid the appearance of bubbles. The plates were incubated at room temperature with shaking for 1 hour, then washed 4 times with 200 μl/well Tris Wash Buffer. Signals were detected by adding 150 µl/well of read buffer (Meso Scale Discovery) and read on a SECTOR® Imager 6000 (Meso Scale Discovery) using the manufacturer's default parameters for this assay. Percent inhibition of the EGF-stimulated positive control signal was calculated for each FN3 domain binding to EGFR.

Inhibition of EGF-stimulated EGFR phosphorylation was measured for 232 identified clones from the TCL1 and TCL2 libraries. 22 of these clones inhibited EGFR phosphorylation by >50% at 1 μM. After the removal of clones that were either poorly expressed or were considered multimers according to size exclusion chromatography, nine clones were selected for further biological characterization. The sequence of the loops of the sun and FG of these clones are shown in table. 4. Eight of the nine selected clones shared a common FG loop sequence (HNVYKDTNMRGL; SEQ ID NO: 95) and significant similarity in BC loop sequences was observed between several clones.

Table 4

FN3 domain Sun loop Hinge FG clone ID SEQ ID NO: Subsequence SEQ ID NO: Hinge FG SEQ ID NO: P53A1R5-17 eighteen ADPHGFYD 87 HNVYKDTTNMRGL 95 P54AR4-17 19 TYDRDGYD 88 HNVYKDTTNMRGL 95 P54AR4-47 twenty WDPFSFYD 89 HNVYKDTTNMRGL 95 P54AR4-48 21 DDPRGFYE 90 HNVYKDTTNMRGL 95 P54AR4-73 22 TWPYADLD 91 HNVYKDTTNMRGL 95 P54AR4-74 23 GYNGDHFD 92 HNVYKDTTNMRGL 95 P54AR4-81 24 DYDLGVYD 93 HNVYKDTTNMRGL 95 P54AR4-83 25 DDPWDFYE 94 HNVYKDTTNMRGL 95 P54CR4-31 26 TAPDAAFD 85 LGSYVFEHDVM 96

Example 3 Characterization of EGFR binding FN3 domains that inhibit EGF binding.

Large scale expression, purification and removal of endotoxin.

FN3 domains shown in Table. 4 was subjected to large-scale expression to provide more material for detailed characterization. An overnight culture containing each of the EGFR-binding FN3 domain variants was inoculated into 0.8 L of Terrific broth medium supplemented with 100 μg/ml ampicillin in a 1/80 dilution of the overnight culture with fresh medium and incubated with shaking at 37°C. FROM. When an optical density of ~1.2-1.5 at 600 nm was reached, culture was induced by adding IPTG to a final concentration of mM and lowering the temperature to 30°C. After 4 hours, the cells were collected by centrifugation and the cell pellet was stored at -80° C. for as long as necessary.

For cell lysis, the thawed pellet was resuspended in 1X BugBuster® solution supplemented with 25 U/ml Benzonase® (Sigma-Aldrich) and 1000 U/ml rLysozyme™ (Novagen EMD Biosciences) at a ratio of 5 ml BugBuster® per 1 g draft. Lysing was performed for 1 hour at room temperature with gentle shaking, then centrifuged at 56,000xg for 50 minutes at 4°C. The supernatant was collected and filtered through a 0.2 μm filter, after which it was loaded onto a 5 ml HisTrap FF column previously equilibrated with buffer A (50 mM Tris-HCl, pH 7.5, 500 mM NaCl, 10 mM imidazole ) using a GE Healthcare AKTAexplorer 100s chromatography system. The column was washed with 20 column volumes of buffer A and additionally washed with 16% buffer B (50 mM Tris-HCl, pH 7.5, 500 mM NaCl, 250 mM imidazole) with 6 column volumes. The FN3 domains were eluted with 50% buffer B, 10 column volumes, followed by a gradient of 50-100% buffer B, 6 column volumes. Fractions containing the FN3 domain protein were pooled, concentrated using a Millipore 10K MWCO concentrator and filtered, then applied to a HiLoad™ 16/60 Superdex™ 75 column (GE Healthcare) pre-equilibrated with PBS. The monomeric protein peak eluting from the size-exclusion column was selected.

Endotoxins were removed using a batch approach using ActiClean Etox resin (Sterogene Bioseparations). Before removing the endotoxin, the resin was pre-treated with 1N. NaOH for 2 hours at 37° C. (or overnight at 4° C.) and washed abundantly with PBS until the pH stabilizes at ~7 as measured with pH indicator paper. The purified protein was filtered through a 0.2 μm filter, then 1 ml of Etox resin was added at a ratio of 10 ml of protein to 1 ml of resin. The binding of endotoxin to the resin was carried out at room temperature for at least 2 hours with gentle rotation. The resin was removed by centrifugation at 500xg for 2 min and the protein supernatant was collected. Endotoxin levels were measured using EndoSafe®-PTS™ cassettes and analyzed on an EndoSafe®-MCS reader (Charles River). If endotoxin levels exceeded 5 EU/mg after the first Etox treatment, the above procedure was repeated until endotoxin levels were reduced to >5 EU/mg. When the endotoxin level was above 5 EU/mg and stabilized after two successive treatments with Etox, anion exchange or hydrophobic chromatography conditions were developed for the protein to remove the remaining endotoxins.

Determination of the affinity of selected EGFR-binding FN3 domains for EGFR-Fc (EGFR-Fc affinity).

The binding affinity of selected EGFR-binding FN3 domains to the recombinant extracellular domain of EGFR was further characterized by surface plasmon resonance methods using the Proteon instrument (BioRad). The assay scheme (chip preparation, EGFR-Fc capture) was similar to that described above for dissociation rate analysis. Selected EGFR-binding FN3 domains were tested at 1 μM in a 3-fold dilution series in horizontal orientation. A buffer sample was also injected to monitor initial stability. The dissociation phase for all concentrations of each EGFR-binding FN3 domain was monitored at a flow rate of 100 μl/min for 30 min (for domains with k _off ~10 ^-2 s ^-1 according to dissociation rate screening data) or 1 h (for domains with koff ~10 ^-3 s ^-1 or less). Two sets of reference data were subtracted from these responses: 1) signals between spots to correct for non-specific interactions between the EGFR-binding FN3 domain and the immobilized IgG surface; 2) Buffer channel signals to correct for baseline shift due to dissociation of the trapped surface from EGFR-Fc over time. Processed binding data at all concentrations for each FN3 domain was globally fitted with a simple 1:1 Langmuir binding model to obtain kinetic constants (kon, koff) and affinity constant (K _D ). In table. 5 shows the kinetic constants for each of the constructs with affinities ranging from 200 pM to 9.6 nM.

Binding of selected EGFR-binding FN3 domains to EGFR on cells (A431 Cell Binding Assay).

A431 cells were seeded at 5000/well in opaque black 96-well plates, allowing them to adhere overnight at 37°C in a humidified atmosphere with 5% CO ₂ . Purified EGFR-binding FN3 domains (1.5 nM to 30 μM) were added to cells (50 μl) for 1 hour at room temperature in three plates. The supernatant was removed and the cells were washed 3 times with 150 µl/well of FACS staining buffer added to the wells. Cells were incubated with 50 μl/well of His-Alexa488 anti-pentahistidine antibody conjugate (Qiagen) diluted 1:100 in FACS staining buffer for 20 min at room temperature. Cells were washed 3 times with 150 µl/well FACS staining buffer, after which 100 µl FACS staining buffer was added to the cells and fluorescence was recorded at 488 nm using a reader

- 52 039356

Acumen ex3. The raw fluorescence signal was plotted against the logarithm of the molar concentration of the FN3 domain and fitted with a sigmoid dose-response curve with variable slope using GraphPad Prism 4 (GraphPad Software) to calculate EC ₅₀ values. In table. 5 shows EC ₅₀ for each of the constructs ranging from 2.2 nM to >20 μM.

Inhibition of EGF binding to EGFR on cells using selected EGFR-binding FN3 domains (EGFR competition assay on A431 cells).

A431 cells were plated at 5000/well on opaque black 96-well plates, allowing them to adhere overnight at 37°C in a humidified atmosphere with 5% CO ₂ . Purified EGFR-binding FN3 domains (1.5 nM to 30 μM) were added to cells (50 μl/well) for 1 hour at room temperature in three plates. Biotinylated EGF (Invitrogen, Cat # E-3477) was added to each well at a final concentration of 30 ng/mL and incubated for 10 minutes at room temperature. Cells were washed 3 times with FACS staining buffer at 150 µl/well. Cells were incubated with 50 μl/well streptavidin-phycoerythrin conjugate (Invitrogen) diluted 1:100 in FACS staining buffer for 20 min at room temperature. Cells were washed 3 times with 150 µl/well FACS staining buffer, after which 100 µl FACS staining buffer was added to the cells and fluorescence was recorded at 600 nm using an Acumen eX3 reader. The raw fluorescence signal was plotted against the logarithm of the molar concentration of the FN3 domain and fitted with a sigmoid dose-response curve with variable slope using GraphPad Prism 4 (GraphPad Software) to calculate IC ₅₀ values. In table. 5 shows IC ₅₀ values ranging from 1.8 to 121 nM.

Inhibition of EGF-stimulated EGFR phosphorylation (phospho-EGRF assay).

Selected FN3 domains that specifically inhibited EGF-stimulated EGFR phosphorylation were more fully analyzed by measuring inhibition IC50 values. Inhibition of EGF-stimulated EGFR phosphorylation was assayed at different concentrations of the FN3 domain (from 0.5 nM to 10 μM) as described above under Inhibition of EGF-stimulated EGFR phosphorylation. The electrochemiluminescence signal was plotted against the logarithm of the molar concentration of the FN3 domain, and IC ₅₀ values were determined by fitting a variable slope sigmoid dose-response curve data using GraphPad Prism 4 (GraphPad Software). In table. 5 shows IC ₅₀ values ranging from 18 nM to >2.5 μM.

Growth inhibition of human tumor cells (NCI-H292 and NCI-H322 cell growth assay).

Inhibition of EGFR-dependent cell growth was analyzed by measuring cell viability of human tumor cell lines overexpressing EGFR, NCI-H292 and NCI-H322 (American Type Culture Collection, cat. no. CRL-1848 and CRL-5806, respectively) after exposure to FN3 domains, binding to EGFR. Cells were plated at 500/well (NCI-H292) or 1000/well (NCI-H322) in opaque white tissue culture treated 96-well plates (Nunc) in RPMI medium (Gibco) 100 µl/well containing GlutaMAX ™ and 10 mM HEPES supplemented with 10% heat-inactivated fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Gibco) and allowed to attach overnight at 37° C. in a humid atmosphere containing 5% CO ₂ . Cells were treated with phosphate buffered saline (PBS) containing varying concentrations of EGFR-binding FN3 domains at 5 μl/well. Controls were treated with 5 μl/well PBS alone or 25 mM EDTA in PBS. Cells were incubated at 37°C, 5% CO2 for 120 h. Viable cells were detected by adding 75 µl/well CellTiter-Glo® reagent (Promega) followed by shaking on a plate shaker for 2 min and incubation in the dark at room temperature for an additional 10 minutes. Plates were read on a SpectraMax M5 plate reader (Molecular Devices) in luminescence mode with a read time of 0.5 s/well compared to blank medium. The percent growth of PBS-treated cells was plotted against the logarithm of the molar concentration of the FN3 domain. IC ₅₀ values were determined by fitting the data with a variable slope sigmoid dose response equation using GraphPad Prism 4 (GraphPad Software). In table. 5 shows IC50 values ranging from 5.9 nM to 1.15 µM and from 9.2 nM to >3.1 µM using NCI-H292 and NCI-H322 cells, respectively. In table. 5 provides a summary of the biological properties of EGFR-binding FN3 domains for each assay.

- 53 039356

Table 5

clone ID FN3 domain SEQ ID NO: Affinity for EGFR-Fc (nM) Binding with cells A431 Competition from EGF to A431 cells Phospho- EGFR Growth NCI- H292 Growth NCI- H322 EC ₅ o (nM) 1C ₅ o (nM) 1C ₅ o (nM) 1С ₅₀ (nM) 1C ₅ o (nM) P53A1R5-17 eighteen 1.89 4.0 9.8 >2500 86 65 P54AR4-17 19 9.62 16 21 184 But But P54AR4-47 twenty 2.51 8.6 7.1 295 44 39 P54AR4-48 21 7.78 12 9.8 170 But But P54AR4-73 22 0.197 9.4 4.6 141 83 73 P54AR4-74 23 But 77 But But But But P54AR4-81 24 But 84 121 But But But P54AR4-83 25 0.255 2.2 1.8 eighteen 5.9 9.2 P54CR4-31 26 0.383 > 20,000 55 179 1150 >3073

Example 4 Construction of EGFR-binding FN3 domains.

A subset of EGFR-binding FN3 domains was designed to increase the conformational stability of each molecule. Mutations L17A, N46V and E86I, which were shown to improve the stability of the FN3 domain (described in US patent publication No. US 2011/0274623), were introduced into clones P54AR4-83, P54CR4-31 and P54AR4-37 by DNA synthesis. New mutants P54AR5-83v2, P54CR431-v2 and P54AR4-37v2 were expressed and purified as described above. Differential scanning calorimetry in PBS was used to analyze the stability of each mutant for comparison with the corresponding parent molecule. In table. 6 shows that each variant molecule was substantially stabilized with an average increase in T _m =18.5°C.

Table 6

FN3 domain clone SEQID NO: _Tm (°C) P54AR4-83 25 50.6 P54AR4-83v2 27 69.8 P54CR4-31 26 60.9 P54CR4-31v2 28 78.9 P54AR4-37 22 45.9 P54AR4-37v2 29 64.2

Example 5 Selection of Type III Fibronectin (FN3) Domains Binding to c-Met and Inhibiting HGF Binding Panning on human c-Met.

The TCL14 library was screened for the biotinylated extracellular domain of human c-Met (bt-c-Met) to identify FN3 domains capable of specific binding to c-Met. For selection, 3 μg of the TCL14 library was transcribed and translated in vitro (IVTT) in E. coli S30 linear extract (Promega, Madison, Wisconsin, USA) and the expressed library was blocked with Cis Block (2% BSA (Sigma-Aldrich, USA)). Louis, Missouri, USA), 100 μg/ml herring sperm DNA (Promega) and 1 mg/ml heparin (Sigma-Aldrich)). For selection, bt-c-Met was added at concentrations of 400 nM (cycle 1), 200 nM (cycles 2 and 3), and 100 nM (cycles 4 and 5). Bound library elements were extracted using neutravidin magnetic beads (Thermo Fisher, Rockford, IL, USA) (runs 1, 3, and 5) or streptavidin magnetic beads (Promega) (runs 2 and 4), and unbound library elements was removed by washing the beads 5-14 times with 500 μl PBS-T followed by 2x washing with 500 μl PBS.

Additional rounds of selection were performed to identify FN3 domain molecules with improved affinity values. Briefly, the results of cycle 5 were obtained as described above and subjected to additional iterative selection cycles with the following changes: bt-c-Met incubation decreased from 1 h to 15 min, bead capture time decreased from 20 to 15 min, bt- c-Met was reduced to 25 nM (cycles 6 and 7) or 2.5 nM (cycles 8 and 9) and an additional wash was performed in the presence of an excess of non-biotinylated c-Met for 1 hour. The purpose of these changes was to simultaneously select binding compounds with a potentially higher rate of association and a lower rate of dissociation, resulting in a substantially lower K _D value.

The results of cycles 5, 7, and 9 were cloned by PCR into a modified pET15 vector (EMD Biosciences, Gibbstown, NJ, USA) containing a ligase-free cloning site (pET15-LIC) using TCON6 primers (SEQ ID NO: 30) and TCON5 E86I short (SEQ ID NO: 31), and the proteins were expressed as proteins with a His6 tag at the C-terminus after transformation and IPTG induction (final concentration 1 mM, 30°C for 16 h) using standard protocols . Cells were harvested by centrifugation and then lysed with a Bugbuster HT (EMD Biosciences) supplemented with 0.2 mg/ml hen egg white lysozyme (Sigma-Aldrich). Bacterial lysates were purified by centrifugation and the supernatants were transferred to new 96-well deep-well Plates (Nunc) and incubated overnight at 4°C. The plates were washed twice using 300 μl/well Tris buffer solution with 0.05% Tween 20 (TBS-T, Sigma-Aldrich) in a Biotek plate washer. Assay plates were blocked with StartingBlock T20 (200 μl/well, Thermo Fisher Scientific, Rockland, IL, USA) for 1 h at room temperature (RT) with shaking and washed twice again with 300 μl TBS-T. FN3 domain lysates were diluted in StartingBlock T20 (1:10 to 1:100,000) using a Hamilton STARplus robotic system. Lysates (50 μl/well) were incubated on assay plates for 1 hour at room temperature with shaking. Without washing the plates, bt-HGF (1 μg/ml in StartingBlock T20, 50 μl/well, biotinylated) was added for 30 minutes at room temperature with shaking. Control wells containing Tencon27 lysates were supplemented with either Starting Block T20 or diluted bt-HGF. Plates were then washed four times with 300 µl/well TBS-T and incubated with 100 µl/well Streptavidin-HRP (1:2000 in TBS-T, Jackson Immunoresearch, West Grove, PA, USA) for 30-40 min. at room temperature with shaking. The plates were washed again with TBS-T four times. A chemiluminescent substrate, POD Chemiluminescence Substrate (50 µl/well, Roche Diagnostics, Indianapolis, IN, USA), prepared according to the manufacturer's instructions, was added to the plate for signal generation and chemiluminescence was recorded for approximately 3 minutes on a Molecular Devices instrument. M5 using SoftMax Pro. Percent inhibition was determined using the following calculations: 100 - ((RLU_sample - mean RLU control without bt-HGF^{)/(mean RL}U control bt-HGF^- mean RLU control without bt-HGF) *¹⁰⁰⁾. Percentage values 50% or more inhibition was considered successful.

High throughput expression and purification of FN3 domains.

His-tagged FN3 domains were purified from E. coli lysates using His MultiTrap™ HP plates (GE Healthcare) and eluted in buffer containing 20 mM sodium phosphate, 500 mM sodium chloride, and 250 mM imidazole, pH 7.4. Purified assay samples were buffer exchanged with PBS, pH 7.4 using PD MultiTrap™ G-25 plates (GE Healthcare).

Determination of IC ₅₀ for inhibition of HGF binding to c-Met.

Selected FN3 domains were further characterized in an HGF competition assay. Dose-response curves were generated for the purified FN3 domains using the assay described above (initial concentrations, 5 μM). Percent inhibition values were calculated. % inhibition was plotted against the logarithm of the molar concentration of the _FN3 domain and IC50 values were determined by fitting the data of a sigmoid dose-response curve with variable slope using GraphPad Prism 4.

In cycle 5, 35 unique sequences were identified showing activity at 1:10 dilutions with IC ₅₀ values ranging from 0.5 to 1500 nM. Cycle 7 generated 39 unique sequences with activity at 1:100 dilutions and IC ₅₀ values ranging from 0.16 to 2.9 nM. In cycle 9, 66 unique sequences were identified, where variants that were active at 1:1000 dilutions were considered successful. In cycle 9 (Table 8), IC ₅₀ values of only 0.2 nM were observed.

Determination of the affinity of selected c-Met binding FN3 domains for c-Met-Fc (EGFR-Fc affinity).

Affinity values for selected c-Met binding FN3 domains were determined as described in Example 3 for determining the affinity of selected EGFR binding FN3 domains, except that c-Met-Fc was used in the assays.

Example 6 Characterization of FN3 Domains Binding to c-Met and Inhibiting HGF Binding.

FN3 domains were expressed and purified as described in Example 2 above. Size exclusion chromatography and kinetic analysis were performed as described in Examples 1 and 2 above, respectively. In table. 7 shows the sequences of the C strand, CD loop, F strand, and FG loop, as well as the complete amino acid sequence of SEQ ID NO: for each domain.

- 55 039356

Table 7

Clone Loop C Strand CD Loop F HEAVY Name SEQID NO: P114AR5P74-A5 32 FDSFWIRYDE VVVGGE TEYYVNILGV KGGSISV P114AR5P75-E9 33 FDSFFIRYDE FLRSGE TEYWVTILGV KGGLVST P114AR7P92-F3 34 FDSFWIRYFE FLGSGE TEYIVNIMGV KGGSISH P114AR7P92-F6 35 FDSFWIRYFE FLGSGE TEYYVNILGV KGGGLSV P114AR7P92-G8 36 FDSFVIRYFE FLGSGE TEYVVQILGV KGGYISI P114AR7P92-H5 37 FDSFWIRYLE FLLGGE TEYVVQIMGV KGGTVSP P114AR7P93-D11 38 FDSFWIRYFE FLGSGE TEYVVGINGV KGGYISY P114AR7P93-G8 39 FDSFWIRYFE FLGSGE TEYGVTINGV KGGRVST P114AR7P93-H9 40 FDSFWIRYFE FLGSGE TEYVVQIIGV KGGHISL P114AR7P94-A3 41 FDSFWIRYFE FLGSGE TEYVVNIMGV KGGKISP P114AR7P94-E5 42 FDSFWIRYFE FLGSGE TEYAVNIMGV KGGRVSV P114AR7P95-B9 43 FDSFWIRYFE FLGSGE TEYVVQILGV KGGSISV P114AR7P95-D3 44 FDSFWIRYFE FLGSGE TEYVVNIMGV KGGSISY P114AR7P95-D4 45 FDSFWIRYFE FLGSGE TEYVVQILGV KGGYISI P114AR7P95-E3 46 FDSFWIRYFE FLGSGE TEYVVQIMGV KGGTVSP P114AR7P95-F10 47 FDSFWIRYFE FTTAGE TEYVVNIMGV KGGSISP P114AR7P95-G7 48 FDSFWIRYFE LLSTGE TEYVVNIMGV KGGSISP P114AR7P95-H8 49 FDSFWIRYFE FVSKGE TEYVVNIMGV KGGSISP

Loop C residues correspond to residues 28-37 of the indicated SEQ ID NO. CD strand residues correspond to residues 38-43 of the indicated SEQ ID NO. Loop F residues correspond to residues 65-74 of the indicated SEQ ID NO. FG loop residues correspond to residues 75-81 of the indicated SEQ ID NO.

Binding of selected c-Met binding FN3 domains to c-Met on cells (H441 cell binding assay).

NCI-H441 cells (Cat. No. HTB-174, American Type Culture Collection, Manassas, VA, USA) were plated at 20,000/well in poly-D-lysine-coated black 96-well clear bottom plates (BD Biosciences , San Jose, California, USA) and allowed them to attach overnight at 37° C., 5% CO ₂ . Purified FN3 domains (50 μl/well; 0-1000 nM) were added to cells for 1 hour at 4°C in two plates. Supernatants were removed and cells were washed three times with FACS staining buffer (150 μl/well, BD Biosciences, Cat # 554657). Cells were incubated with biotinylated anti-HIS antibody (1:160 dilution in FACS staining buffer, 50 µl/well, R&D Systems, Cat # BAM050) for 30 min at 4°C. Cells were washed three times with FACS staining buffer (150 µl/well) and then incubated with anti-mouse IgG _r Alexa488 conjugated antibody (1:80 diluted in FACS staining buffer, 50 µl/well, Life Technologies, cat. no. A21121) for 30 minutes at 4°C. Cells were washed three times with FACS staining buffer (150 μl/well) and left in FACS staining buffer (50 μl/well). Total fluorescence was determined on an Acumen eX3 reader. The raw fluorescence signal was plotted against the logarithm of the molar concentration of the FN3 domain and fitted with a sigmoid dose-response curve with variable slope using GraphPad Prism 4 (GraphPad Software) to calculate EC ₅₀ values. The FN3 domains were found to show a range of binding activities with EC ₅₀ values from 1.4 to 22.0 nM, as shown in Table 1. eight.

Inhibition of HGF-stimulated c-Met phosphorylation.

Purified FN3 domains were tested for their ability to inhibit HGF-stimulated c-Met phosphorylation at NCI-H441 using the c-Met phospho(Tyr1349) kit from Meso Scale Discovery (Gaithersburg, MD, USA). Cells were plated at 20,000/well in clear 96-well tissue culture-treated plates in RPMI 100 μl/well (containing Glutamax and HEPES, Life Technologies) supplemented with 10% fetal bovine serum (FBS; Life Technologies) and allowed to they attach overnight at 37°C in a humid atmosphere containing 5% CO2. The culture medium was completely removed and the cells were kept under starvation conditions in serum-free RPMI medium (100 μl/well) at 37°C, 5% CO2. Then, fresh serum-free RPMI medium (100 µl/well) containing FN3 domains at a concentration of 20 µM and below was added to the cells for 1 h at 37°C, 5% CO ₂ . Controls were treated with medium only. Cells were stimulated with 100 ng/ml recombinant human HGF (100 µl/well, R&D Systems, Cat # 294-HGN) and incubated at 37°C, 5% CO2 for 15 min. One set of control wells was left unstimulated as negative controls. The medium was completely removed and the cells were lysed with Complete Lysis Buffer (50 µl/well, Meso Scale Discovery) for 10 min at room temperature with shaking according to the manufacturer's instructions. Assay plates for measuring phosphorylated c-Met were blocked with the supplied blocking solution according to the manufacturer's instructions at room temperature for 1 hour. The plates were then washed three times with Tris Wash Buffer (200 μl/well, Meso Scale Discovery). Cell lysates (30 µl/well) were loaded onto assay plates and incubated at room temperature with shaking for 1 hour. The assay plates were then washed four times with Tris Wash Buffer, after which ice-cold detection antibody solution (25 µl/well) was added to each well. hole,

- 56 039356

Meso Scale Discovery) for 1 hour at room temperature with shaking. The plates were washed four times with Tris Wash Buffer. Signals were detected by adding 150 Read Buffer (150 µl/well, Meso Scale Discovery) and reading on a SECTOR® Imager 6000 (Meso Scale Discovery) using the manufacturer's default parameters for this assay. The electrochemiluminescence signal was plotted against the logarithm of the molar concentration of the FN3 domain and IC ₅₀ values were determined by fitting the data of a sigmoid dose-response curve with a variable slope using GraphPad Prism 4. The FN3 domains were found to inhibit c-Met phosphorylation with IC ₅₀ values in the range from 4.6 to 1415 nm, as shown in table. eight.

Inhibition of growth or viability of human tumor cells.

Inhibition of c-Met-dependent cell growth was analyzed by measuring the viability of U87-MG cells (American Type Culture Collection Cat # HTB-14) after exposure to c-Met binding FN3 domains. Cells were plated at 8000/well in opaque white 96-well tissue culture-treated plates (Nunc) in RPMI 100 μl/well supplemented with 10% FBS and allowed to adhere overnight at 37°C, 5% CO2. 24 hours after seeding, the medium was aspirated and fresh serum-free RPMI medium was added to the cells. 24 hours after treatment with serum-free medium, cells were treated by adding serum-free medium containing FN3 domains that bind to c-Met (30 μl/well). Cells were incubated at 37° C., 5% CO ₂ for 72 hours. Viable cells were detected by adding 100 μl/well of CellTiter-Glo® reagent (Promega) followed by 10 minutes of shaking on a plate shaker. Plates were read on a SpectraMax M5 plate reader (Molecular Devices) set to luminescence mode with a read time of 0.5 s/well. Build a plot of the original luminescence units (RLU) from the logarithm of the molar concentration of the FN3 domain. IC ₅₀ values were determined by fitting the data with a variable slope sigmoid dose-response curve using GraphPad Prism 4. 8 shows IC ₅₀ values ranging from 1 to >1000 nM. The characteristics of the FN3 domains that bind to c-Met are summarized in Table 1. eight.

Table 8

Clone Affinity (Kd, nM) Competition with HGF, IC50 (nM) Binding to H441 cells (EC50, nM) pMet inhibition in cells H441 (IC50, nM) Inhibition of U87-MG cell proliferation (IC50, nM) Name SEQID NO P114AR5P74-A5 32 10.1 5.2 18.7 1078 464.4 P114AR5P75-E9 33 45.8 51.9 But 1415 1193.9 P114AR7P92-F3 34 0.4 0.2 1.5 8.3 2.7 P114AR7P92-F6 35 3.1 2.2 4.9 165.3 350.5 P114AR7P92-G8 36 1.0 1.6 5.9 155.3 123.9 P114AR7P92-H5 37 11.6 But 22.0 766.4 672.3 P114AR7P93-D11 38 But But 2.3 16 14.4 P114AR7P93-G8 39 6.9 one 3.8 459.5 103.5 P114AR7P93-H9 40 3.3 2.9 12.9 288.2 269.9 P114AR7P94-A3 41 0.4 0.2 1.4 5 9.3 P114AR7P94-E5 42 4.2 0.7 3.4 124.3 195.6 P114AR7P95-B9 43 0.5 0.3 But 9.8 17.4 P114AR7P95-D3 44 0.3 0.2 1.5 4.6 17 P114AR7P95-D4 45 0.4 But 1.4 19.5 19.4 P114AR7P95-E3 46 1.5 But 3.2 204.6 209.2 P114AR7P95-F10 47 4.2 1.4 4.4 187.6 129.7 P114AR7P95-G7 48 20.0 But 11.3 659.3 692 P114AR7P95-H8 49 3.7 But 4.1 209.8 280.7

Thermal stability of FN3 domains binding to c-Met.

Differential scanning calorimetry in PBS was used to analyze the stability of each FN3 domain. The results of the experiment are shown in table. 9.

Table 9

Clone Thermal stability (Tm, °C) Name SEQ. ID NO: P114AR5P74-A5 32 74.1 P114AR5P75-E9 33 But P114AR7P92-F3 34 81.5 P114AR7P92-F6 35 76.8

- 57 039356

Example 7 Design and characterization of EGFR/c-Met bispecific molecules.

Creation of molecules bispecific to EGFR/c-Met.

Many of the combinations of EGFR and c-Met binding FN3 domains described in Examples 1-6 were combined into bispecific molecules capable of binding to both EGFR and c-Met. In addition, EGFR binding FN3 domains having the amino acid sequence shown in SEQ ID NO: 107-110 and c-Met binding FN3 domains having the amino acid sequence shown in SEQ ID NO: : 111-114. Synthetic genes encoding the amino acid sequences described in SEQ ID NOs: 50-72 and 106 (Table 10) were designed to adhere to the following format: cEGFR-binding FN3 domain, then peptide linker, then cEGFR-binding FN3 domain. -Met. A polyhistidine tag was inserted at the C-terminus to allow purification. In addition to these molecules described in Table. 10, the linker between the two FN3 domains varied in length, sequence composition, and structure as listed in Table 10. 11. A number of other linkers are contemplated to connect such FN3 domains. EGFR/c-Met bispecific molecules were expressed and purified from E. coli as described for EGFR monospecific or c-Met FN3 domains using IMAC steps and gel permeation chromatography.

- 58 039356

Table 10

Bispecific to EGFR/c-Met molecule FN3 domain, EGFR-binding FN3 domain, contacting c-MET Linker clone ID SEQID NO' clone ID SEQID NO' clone ID SEQID NO' Subsequence SEQID NO' ERU1 fifty P54AR4-83V2 27 Р114AR5P74-A5 32 (GGGGS)4 79 ERU2 51 P54AR4-83V2 27 Р114AR7P94-A3 41 (GGGGS)4 79 EEAS 52 P54AR4-83V2 27 Р114AR7P93-H9 40 (GGGGS)4 79 ERU4 53 P54AR4-83V2 27 Р114AR5P75-E9 33 (GGGGS)4 79 ERU5 54 R53A1 R5-17V2 107 Р114AR7P94-A3 41 (GGGGS)4 79 ERU6 55 R53A1 R5-17V2 107 Р114AR7P93-H9 40 (GGGGS)4 79 ERU7 56 R53A1 R5-17V2 107 Р114AR5P75-E9 33 (GGGGS)4 79 ERU15 57 P54AR4-83V2 27 Р114AR7P94-A3 41 (AP)5 81 ERU27 58 P54AR4-83V2 27 Р114AR5P74-A5 32 (AP)5 81 ECB60 59 R53A1 R5-17V2 107 Р114AR7P94-A3 41 (AP)5 81 ERU37 60 R53A1 R5-17V2 107 Р114AR5P74-A5 32 (AP)5 81 ERU94 61 P54AR4-83V22 108 P114AR7P94-A3V22 111 (AP)5 81 ERU95 62 P54AR4-83v22 108 P114AR9P121-A6v2 112 (AP)5 81 ERU96 63 P54AR4-83v22 108 Р114AR9P122-A7V2 113 (AP)5 81 ERU97 64 P54AR4-83v22 108 P114AR7P95-C5V2 114 (AP)5 81 ECB106 65 P54AR4-83v23 109 109 109 109 110 110 110 Р114AR7P94-A3V22 111 (AP)5 81 ECB107 66 P54AR4-83v23 P114AR9P121-A6v2 112 (AP)5 81 ECB108 67 P54AR4-83v23 P114AR9P122-A7V2 113 (AP)5 81 ECB109 68 P54AR4-83V23 Р114AR7P95-C5V2 114 (AP)5 81 ECB118 69 P53A1 R5-17v22 Р114AR7P94-A3V22 111 (AP)5 81 ECB119 70 P53A1 R5-17V22 P114AR9P121-A6v2 112 (AP)5 81 ECB120 71 P53A1 R5-17v22 P114AR9P122-A7v2 113 (AP)5 81 ECB121 72 P53A1 R5-17v22 110 P114AR7P95-C5V2 114 (AP)5 81 ECB91 106 P54AR4-83v22 108 Р114AR7P95-C5V2 114 (AP)5 81 ERU18 118 P54AR4-83v2 27 Р114AR5P74-A5 32 (AP)5 81 ERU28 119 P53A1 R5-17v2 107 Р114AR5P74-A5 32 (AP)5 81 ECB38 120 P54AR4-83v2 27 Р114AR7P94-A3 41 (AP)5 81 ERU39 121 P53A1 R5-17v2 107 Р114AR7P94-A3 41 (AP)5 81

Table 11

Linker SEQID NO: Linker length in amino acids Structure GS 78 2 disordered GGGGS 105 5 disordered (GGGGS) ₄ 79 twenty disordered (AP)2 80 four Rigid (AP)5 81 5 Rigid (AR)yu 82 twenty Rigid (AP)20 83 40 Rigid A(EAAAK) ₅ AAA 84 29 a-spiral

Bispecific to EGFR/c-Met molecules increase the efficiency compared to single monospecific molecules, indicating avidity.

NCI-H292 cells were seeded in 96-well plates in RPMI medium containing 10% FBS. After 24 h, the medium was replaced with serum-free RPMI. 24 h after serum-free medium, cells were treated with various concentrations of FN3 domains: high affinity FN3 monospecific for EGFR (P54AR4-83v2), low affinity FN3 monospecific for c-Met (P114AR5P74-A5), a mixture of two monospecific for EGFR and for c-Met FN3 domains or EGFR/c-Met bispecific molecules comprising a low c-Met affinity FN3 domain coupled to a high EGFR affinity FN3 domain (ECB1). Cells were treated for 1 h with monospecific or bispecific molecules and then stimulated with EGF, HGF, or a combination of EGF and HGF for 15 min at 37°C, 5% CO2. Cells were lysed with MSD lysis buffer and cell signaling was analyzed using appropriate MSD assay plates according to the manufacturer's instructions as described above.

- 59 039356

The FN3 domain with low affinity for c-Met inhibited c-Met phosphorylation with an IC ₅₀ value of 610 nM (FIG. 4). As expected, the EGFR-binding FN3 domain was unable to inhibit c-Met phosphorylation, and the mixture of monospecific molecules looked identical to a single c-Met-binding FN3 domain. However, the EGFR/c-Met bispecific molecule inhibited c-Met phosphorylation with an IC ₅₀ of 1 nM (FIG. 4), i. providing a shift in efficiency by more than 2 orders of magnitude relative to one monospecific c-Met molecule.

The potential of the EGFR/c-Met bispecific molecule to increase inhibition of c-Met and/or EGFR phosphorylation via an avidity effect was evaluated in a variety of cell types with different densities and c-Met to EGFR ratios (FIG. 5). NCI-H292, NCI-H441, or NCI-H596 cells were plated in 96-well plates in RPMI medium containing 10% FBS. After 24 h, the medium was replaced with serum-free RPMI. 24 hours after serum-free medium, cells were treated with various concentrations of FN3 monospecific EGFR binding domain, FN3 monospecific c-Met binding domain, or EGFR/c-Met bispecific molecule (ECB5, derived from P53A1R5-17v2 and P114AR7P94-A3) . Cells were treated for 1 h with monospecific or bispecific molecules and then stimulated with EGF, HGF, or a combination of EGF and HGF for 15 min at 37°C, 5% CO ₂ . Cells were lysed with MSD lysis buffer and cell signaling was analyzed using appropriate MSD assay plates according to the manufacturer's instructions as described above.

In FIG. 5A-5C show EGFR inhibition by a monospecific EGFR-binding FN3 domain compared to an EGFR/c-Met bispecific molecule in three different cell lines. For avidity analysis in the EGFR phosphorylation assay, the FN3 domain binding to EGFR with a medium affinity (1.9 nM) (P53A1R5-17v2) was compared with an EGFR/c-Met bispecific molecule containing the same FN3 domain binding to EGFR, coupled to a high affinity (0.4 nM) c-Met binding FN3 domain (P114AR7P94-A3). In NCI-H292 and H596 cells, inhibition of EGFR phosphorylation for monospecific and bispecific molecules was comparable (FIGS. 5A and 5B), probably due to the fact that these cell lines had a high ratio of EGFR to c-Met. To test this theory, the inhibition of EGFR phosphorylation in NCI-H441 cells containing more c-Met receptors than EGFR was evaluated. Treatment of NCI-H441 cells with the EGFR/c-Met bispecific molecule reduced the IC ₅₀ for inhibition of EGFR phosphorylation compared to the EGFR-binding FN3 monospecific domain by 30-fold (FIG. 5C).

The potential for increased potency with the EGFR/c-Met bispecific molecule was assessed in a c-Met phosphorylation assay using a molecule with high affinity for EGFR (0.26 nM) and medium affinity for c-Met (10.1 nM). In both NCI-H292 and NCI-H596 cells, inhibition of c-Met phosphorylation by the bispecific molecule was increased compared to the c-Met binding FN3 monospecific domain by 134-fold and 1012-fold, respectively (Fig. 3D and 3E).

It has been confirmed that the increased efficiency of inhibiting EGFR and c-Met phosphorylation by EGFR/c-Met bispecific molecules is reflected in increased inhibition of signaling and proliferation. In these experiments, a mixture of EGFR-binding FN3 domains and c-Met binding FN3 domains was compared to an EGFR/c-Met bispecific molecule. As described in Table. 12 and 13, IC50 values for ERK phosphorylation (Table 12) and NCI- _H292 cell proliferation (Table 13) were reduced when cells were treated with an EGFR/c-Met bispecific molecule compared to monospecific binding compounds. The IC50 for inhibition of ERK phosphorylation with the EGFR/c-Met bispecific molecule was 143-fold lower than with a mixture of two FN3 monospecific EGFR and c-Met domains, demonstrating the effect of avidity on the performance of the molecules in this assay. In table. 12 FN3 monospecific domains binding to EGFR and c-Met do not completely inhibit activity, and therefore the IC50 values shown should be considered lower limits. Proliferation assays were performed using different combinations of EGFR and c-Met binding FN3 domains, either as a mixture or linked in a bispecific format. The IC ₅₀ for inhibition of proliferation for the EGFR/c-Met bispecific molecule was 34-236 times lower than when using a mixture of monospecific parent FN3 domains binding to EGFR and c-Met. This confirms that the avidity effect observed at the receptor level (FIGS. 4 and 5) translates into improved inhibition of cell signaling (Table 12) and proliferation (Table 13).

- 60 039356

Table 12

Specificity of a molecule containing FN3 domains clone name Type of 1C ₅ o (nM) (ERK phosphorylation) EGFR P54AR4-83v2 Monospecific > 10,000 c-Met P114AR5P74-A5 Monospecific 2366 EGFR or cMet P54AR4-83v2+ P114AR5P74-A5 A mixture of monospecific molecules 798.4 EGFR and c-Met ECB1 Bispecific molecule 5.6

Table 13

FN3 domain, binding to EGFR (affinity) FN3 domain, contacting c-Met (affinity) ICso for a mixture of monospecific molecules (nM) IC50 FOR bispecific molecule (nM) Multiplicity of increase 1C ₅ o for a mixture of monospecific / bispecific P54AR4-83V2 (0.26 nM) Р114ARP94-A3 (0.4 nM) 36.5 1.04 35 P54AR4-83v2 (0.26 nM) Р114AR7P93-H9 (3.3 nM) 274.5 8.05 34 P54AR4-83V2 (0.26 nM) Р114AR5P74-A5 (10.1 nM) 1719 7.29 236

Tumor xenografts in vivo: pharmacokinetics/pharmacodynamics (PK/PD).

To determine the efficacy of FN3 monospecific and bispecific domain molecules in vivo, human HGF-secreting tumor cells were constructed (mouse HGF does not bind to human c-Met). Human HGF was stably expressed in NCI-H292 cells using lentivirus infection (human HGF expressing lentiviral DNA vector (Cat. No. X16322) and lentiviral packaging kit manufactured by Genecopoeia). After infection, HGF expressing cells were selected using 4 μg/ml puromycin (Invitrogen). Human HGF protein was detected in pooled cell conditioned medium using assay plates manufactured by MesoScale Discovery.

SCID Beige mice were subcutaneously inoculated with NCI-H292 cells expressing human HGF (2.0×10 ⁶ cells in Cultrex (Trevigen) in a volume of 200 μl) into the dorsal flank of each animal. The tumors were measured twice a week until the tumor volume was 150-250 mm ³ . Mice were then treated ip with a single dose of EGFR/c-Met bispecific molecules (linked to an albumin-binding domain to increase half-life) or vehicle PBS. Tumors were removed 6 or 72 hours after injection and immediately frozen in liquid nitrogen. Blood samples were collected by cardiac puncture in 3.8% citrate containing protease inhibitors. Immediately after collection, blood samples were centrifuged and the resulting plasma was transferred to sample tubes and stored at -80°C. Tumors were weighed, cut into small pieces and lysed in Lysing Matrix A (LMA) tubes containing RIPA buffer with protease/phosphatase inhibitors HALT (Pierce), 50 mM sodium fluoride (Sigma), 2 mM activated sodium orthovanadate (Sigma) and 1 mM PMSF (MesoScale Discovery). Lysates were removed from the LMA matrix and centrifuged to remove insoluble protein. Soluble tumor protein was quantified in the BCA protein assay and diluted to equivalent protein levels in tumor lysis buffer.

Phosphorylated c-Met, EGFR and ERK were measured using MesoScale Discovery assay plates (according to manufacturer's protocol and above description).

In FIG. 6 shows the results of the experiments. Each EGFR/c-Met bispecific molecule significantly reduced c-Met, EGFR, and ERK phosphorylation levels at both 6 and 72 hours. The data presented in FIG. 6 show the importance of simultaneous inhibition of both c-Met and EGFR, as well as how the affinity of the EGFR/c-Met bispecific molecule for each receptor affects downstream ERK inhibition. Molecules containing FN3 domains that bind to EGFR with high affinity (P54AR4-83v2; shown in the figure as element 8, K _D = 0.26 nM) inhibited EGFR phosphorylation to a greater extent than molecules containing FN3 domains that bind to EGFR with medium affinity (P53A1R5-17v2; shown in the figure as element 17, K _D = 1.9 nM), both at 6 h and at 72 h. All four tested bispecific molecules completely inhibited ERK phosphorylation at the time point of 6 h regardless of affinity. After a time point of 72 h, molecules containing a high affinity c-Met binding domain (P114AR7P94-A3; shown in the figure as element A3, K _D = 0.4 nM) significantly inhibited ERK phosphorylation compared to the FN3 binding domain with c-Met with medium affinity (P114AR5P74-A5; shown in the figure as element A5; K _D =10.1 nm; Fig. 6).

The concentration of each EGFR/c-Met bispecific molecule was measured 6 and 72 hours after administration in the blood and in the tumor (FIG. 7). Interestingly, a bispecific molecule with a moderate affinity EGFR binding domain (P53A1R5-17v2; K _D = 1.9 nM) but with a high affinity c-Met binding FN3 domain (P114AR7P94-A3; K _D = 0 ,4 nM), showed significant pain

- 61 039356 neck accumulation at 6 h relative to other molecules, although the difference disappeared by 72 h. affinity for EGFR does not bind to normal tissue as strongly as the FN3 domain, which binds to EGFR with high affinity. Therefore, more free FN3 domain is left for tumor binding, binding to EGFR with medium affinity. Therefore, determining the appropriate affinity values for each receptor may allow the discovery of a therapeutic agent with reduced systemic toxicity values and increased tumor uptake.

Studies on the efficacy of EGFR/c-Met bispecific molecules against tumors.

SCID Beige mice were subcutaneously inoculated with NCI-H292 cells expressing human HGF (2.0x106 cells in Cultrex (Trevigen), 200 μl) into the dorsal flank of each animal. One week after implantation, mice were divided into groups with equivalent tumor volumes (mean tumor volume = 77.9 ± 1.7 mm ³ ). Mice were injected with bispecific molecules three times a week and tumor volume was recorded twice a week.

Tumor growth inhibition (TGI) was observed using four different bispecific molecules with different affinities for c-Met and EGFR. In FIG. 8 shows the benefit of inhibition of both c-Met and EGFR as tumor growth was observed in mice treated with a high affinity EGFR binding FN3 domain compared to a medium affinity EGFR FN3 domain when the FN3 binding domain with c-Met, had medium affinity (open and solid triangles, P54AR4-83v2-P114AR5P74-A5 versus P53A1R5-17-P114AR5P74-A5). In addition, the data show the significance of having a high affinity c-Met binding FN3 domain, since bispecific molecules containing FN3 domains that bind to EGFR with either high or moderate affinity, but a high c-Met binding FN3 domain affinity showed the highest efficiency (dashed gray and black lines, P54AR4-83v2-P114AR7P94-A3 and P53A1R5-17v2-P114AR7P94-A3).

Efficacy of the bispecific molecule and other EGFR and c-Met inhibitors.

The in vivo therapeutic efficacy of the EGFR/c-Met bispecific molecule (ECB38) and small molecule inhibitors crizotinib (c-Met inhibitor) and erlotinib (EGFR inhibitor), cetuximab (anti-EGFR antibody), each alone, as well as the combination of crizotinib and erlotinib was evaluated. in a subcutaneous xenotransplantation model of H292-HGF human lung cancer cells in SCID/Beige mice.

H292-HGF cells were cultured in vitro in RPMI1640 medium supplemented with fetal bovine serum (10% v/v) and L-glutamine (2 mM) at 37°C in an air atmosphere with 5% CO2. Cells were regularly passaged twice a week with trypsin-EDTA treatment. Cells that reached the exponential growth phase were harvested and counted for tumor inoculation.

For tumor development, each mouse was subcutaneously inoculated with tumor cells in the right flank region of H292-HGF (2x106) in 0.1 ml PBS with Cultrex (1:1). Treatment was started when the average tumor size reached 139 mm ³ . The administration of the test product and the number of animals in each experimental group are shown in the following experimental design table. The date of inoculation with tumor cells was designated as day 0. In table. 14 shows the treatment groups.

Table 14

Group N Treatment Dose (mg/kg) Method of administration Planned schedule Actual Graph one ten control with carrier 0 i/n QD x 3 weeks QD x 3 weeks 2 ten Bispecific to EGFR/c-Met molecule 25 i/n 3 times a week x 3 weeks 3 times a week x 3 weeks 3 ten Crizotinib fifty on QD x 3 weeks QD x 17 days four ten Erlotinib fifty on QD x 2 weeks QD x 3 weeks 5 ten Crizotinib fifty on QD x 3 weeks QD x 3 weeks 6 ten Cetuximab 1 mg/mouse i/n Q4d x 6 Q4d x 6

N: number of animals;

p/o: oral administration;

i / p: intraperitoneal injection 3 times a week: administration of doses on the 1st, 3rd and 5th day of the week;

QD: once a day;

Q4d: once every four days;

the interval for the combination of crizotinib and erlotinib was 0.5 h;

the volume of administration was adjusted according to body weight (10 l/g);

a: No dose was given 14 days after grouping.

All animals were weighed and tumor volumes measured prior to treatment. Since tumor volume can affect the effectiveness of any given treatment, mice were divided into groups in

- 62 039356 according to the plan with randomized blocks based on their tumor volumes. This ensures that all groups at baseline will be comparable. For the distribution of experimental animals into groups used a plan with randomized blocks. First, the experimental animals were divided into homogeneous blocks according to their initial tumor size. Secondly, within each block, the experimental animals were randomized according to the types of treatment. Distribution of experimental animals using a randomized block design ensures that all animals have the same probability of receiving a given type of treatment, and hence bias is reduced.

During regular follow-up, the animals were checked for any effects of tumor growth and treatments on normal behavior, e.g., on mobility, visual assessment of food and water intake, weight gain/loss (body weight measured twice a week), blurred eyes/ wool and any other anomalous effects.

The final evaluation criterion is whether the tumor growth can be slowed down or tumor-bearing mice can be cured. Tumor size was measured twice a week in two directions using a caliper; the volume was expressed in mm ³ using the formula V=0.5 axb ² where a and b represent the major and minor tumor diameters, respectively. Tumor size was then used to calculate both T-C and T/C values. The T-C value was calculated, where T is the time (in days) required for the average-sized tumor in the treatment group to reach 1000 mm ³ , and C is the time (in days) required for the average-sized tumor in the control group reached the same size. The value of T/C (in percent) was an indicator of antitumor efficacy; T and C were the mean tumor volume in the treatment group and the control group, respectively, on a given day. Complete tumor regression (CR) was defined as the reduction of the tumor to a limit that is not detectable by palpation (62.5 mm ³ ). Tumor partial regression (PR) was defined as tumor reduction relative to baseline tumor volume. For CR or PR to be considered stable, it was required that CR or PR be determined for at least three or more consecutive tumor measurements.

Animals whose body weight loss exceeded 20% or whose average tumor size in the group exceeded 2000 mm ³ were euthanized. The study was terminated two weeks after the final dose.

In table. 15 shows summary statistics including the mean and standard error of the mean (SEM) for tumor volume in each group and at each time point. Statistical analyzes of differences in tumor volume between groups were performed using one-way analysis of variance (ANOVA) followed by separate comparisons using the Games Howell test (not assuming equal variances). All data was analyzed using SPSS 18.0. A p<0.05 value was considered statistically significant.

Table 15

Day Tumor volume (mm ³ )a Carrier Bispecific to EGFR/c-Met molecule 25 mg/kg Crizotinib 50 mg/kg Erlotinib 50 mg/kg Crizotinib; erlotinib 50 mg/kg; 50 mg/kg Cetuximab 1 mg/mouse 7 139±7 137±7 140±9 141±8 139±8 139±10 9 230±20 142±7 217±20 201 ± 19 134±9 168 ± 13 13 516±45 83±6 547 ± 43 392 ± 46 109 ± 10 212±20 16 808 ± 104 44±7 914±92 560±70 127±15 252 ± 28 twenty 1280 ± 209 30±6 1438±239 872±136 214±30 371±48 23 1758±259 23±7 2102±298 1122±202 265±40 485±61 27 2264±318 21±5 — 1419±577 266 ± 42 640 ± 82 thirty — 23±6 — 1516±623 482±61 869±100

The mean tumor size in the vehicle-treated group (Group 1) reached 1758 mm ³ on day 23 after tumor inoculation. Treatment with the EGFR/c-Met bispecific molecule at a dose level of 25 mg/kg (Group 2) resulted in complete tumor regression (CR) in all mice that was stable at >3 consecutive tumor measurements (mean tumor volume 23 mm ³ , T/C value=1%, p=0.004, compared to the vehicle group, on day 23).

Treatment with crizotinib as a single agent at a dose level of 50 mg/kg (group 3) showed no antitumor activity; the mean tumor size was 2102 mm ³ on day 23 (T/C=120%, p=0.944 compared to the vehicle group).

Treatment with erlotinib as a single agent at a dosage level of 50 mg/kg (group 4) showed little antitumor activity, but no significant differences from the vehicle group were found; the mean tumor size was 1122 mm ³ on day 23 (T/C value=64%, p=0.429 compared with the vehicle group) with a slowdown in tumor growth at 1000 mm ³ for 4 days compared with the vehicle group.

- 63 039356

The combination of crizotinib (50 mg/kg, group 5) and erlotinib (50 mg/kg, group 5) showed significant antitumor activity; the mean tumor size was 265 mm ³ on day 23 (T/C=15%, p=0.008) with slower tumor growth at 1000 mm ³ on day 17 compared to the vehicle group.

Cetuximab at a dosage level of 1 mg/mouse as the only agent (Group 6) showed significant antitumor activity; the mean tumor size was 485 mm ³ on day 23 (T/C=28%, p=0.018) with slower tumor growth at 1000 mm ³ on day 17 compared to the vehicle group. In FIG. 15 and in table. 16 shows the antitumor activity values of various therapeutic agents.

Table 16

Treatment Tumor size (mm ³ ) on day 23 T/C (%) T - C (day) at 1000 mm ³ Meaning R Carrier 1758+259 - - - Bispecific to EGFR/c-Met molecule (25 mg/kg) 23+7 one — 0.004 Crizotinib (50 mg/kg) 2102+298 120 -one 0.944 Erlotinib (50 mg/kg) 1122+202 64 four 0.429 Crizotinib + erlotinib (50 mg/kg + 50 mg/kg) 265±40 fifteen 17 0.008 Cetuximab (1 mg/mouse) 485+61 28 17 0.018

Moderate to severe weight loss was observed in the vehicle group, which may be due to increased tumor exposure; 3 mice died and 1 mouse was euthanized with >20% weight loss by day 23. Group 2 showed little toxicity of the EGFR/c-Met bispecific molecule; 3 mice were euthanized with >20% weight loss during the treatment period; body weight gradually recovered after discontinuation of treatment during the 2-week follow-up period. More severe weight loss was observed in the crizotinib or erlotinib monotherapy group compared to the vehicle group, indicating treatment-related toxicity. The combination of crizotinib and erlotinib was essentially tolerated during the administration phase, but significant weight loss was observed by the end of the study, which may be due to the resumption of rapid tumor growth during the non-treatment period. Cetuximab monotherapy in the present study was well tolerated; weight loss was observed only towards the end of the study due to continued tumor growth.

Thus, the EGFR/c-Met bispecific molecule at 25 mg/kg (3 times a week for 3 weeks) provided a complete response in the xenotransplantation model of human lung cancer H292-HGF cells into SCID/Beige mice. Treatment was tolerated by 7 out of 10 mice, and severe weight loss was observed in 3 out of 10 mice. In FIG. 9 shows the effect of various therapeutic agents on tumor size at time points after treatment.

Example 8 Extending the half-life of EGFR/c-Met bispecific molecules.

Many methods have been described for reducing renal filtration and thus prolonging the serum half-life of proteins, including modification with polyethylene glycol (PEG) or other polymers, binding to albumin, fusion with protein domains that bind to albumin or other serum proteins, genetic fusion with albumin, fusion with IgG Fc domains and fusion with long unstructured amino acid sequences.

The EGFR/c-Met bispecific molecules were modified with PEG to increase the hydrodynamic radius by incorporating free cysteine at the C-terminus of the molecule. Most commonly, the free thiol group of a cysteine residue is used to attach PEG molecules that have been functionalized with maleimide or iodoacetimide groups using standard methods. Various forms of PEG can be used to modify the protein, including linear PEGs of 1000, 2000, 5000, 10000, 20000 or 40000 kDa. Branched PEG molecules of this molecular weight can also be used for modification. In some instances, PEG groups can also be attached via primary amines in EGFR/c-Met bispecific molecules.

In addition to PEGylation, the half-life of EGFR/c-Met bispecific molecules was extended by forming these proteins as fusion molecules with a naturally occurring 3-helical serum albumin-binding domain (ABD) or consensus albumin-binding domain (ABDCon). These protein domains were attached to the C-terminus of the c-Met binding FN3 domain via any of the linkers described in Table 1. 12. The ABD or ABDCon domain can also be placed in the primary sequence between the EGFR binding FN3 domain and the c-Met binding FN3 domain.

Example 9 Characterization of EGFR/c-Met Bispecific Molecules.

The selected EGFR/c-Met bispecific molecules were characterized by their affinity as

- 64 039356

EGFR and c-Met in their ability to inhibit EGFR and c-Met autophosphorylation and in their effect on HGF cell proliferation. The binding affinity of the EGFR/c-Met bispecific molecules to the recombinant extracellular domain of EGFR and/or c-Met was further assessed by surface plasmon resonance using the Proteon instrument (BioRad) according to the protocol described in Example 3. Characterization results are shown in Table 1. 17.

Table 17

kd (EGFR, nM) kd (c-Met, nM) pMet inhibition in H441 cells (IC ₅₀ , nM) H292 - pEGFR inhibition in H292 cells (IC ₅₀ , nM) H292-HGF Proliferation inhibition in HGF-induced H292 cells (1C ₅₀ , nM) ECB15 0.2 2.6 N/A 4.2 23 ECB94 one 4.3 53.8 5.1 29.6 ECB95 1.1 6.2 178.8 13.6 383.4 ECB96 1.6 22.1 835.4 24.7 9480 ECB97 1.3 1.7 24.2 16.6 31.0 ECB106 16.7 5.1 53.3 367.4 484.5 ECB107 16.9 9 29.9 812.3 2637 ECB108 15.3 25.5 126.2 814.4 11372 ECB109 17.3 2.1 26 432 573.6

Example 10 Generation of EGFR/c-Met Bispecific Antibodies.

Several EGFR and c-Met monospecific antibodies were expressed as IgG1, kappa having Fc substitutions K409R or F405L in their Fc regions (EC numbering). Monospecific antibodies were expressed in two CHO cell lines, with one line having a reduced ability to fucosylate, which made it possible to obtain antibodies with a fucose content in the antibody polysaccharide chain of 1-15%.

Monospecific antibodies were purified using standard methods using a protein A column (HiTrap MabSelect SuRe column). After elution, the pools were dialyzed in D-PBS, pH 7.2.

EGFR/c-Met bispecific antibodies were generated by combining an anti-EGFR monospecific mAb and an anti-c-Met monospecific mAb during in vitro Fab arm exchange (as described in WO 2011/131746). Briefly, approximately 1-20 mg/ml of each antibody in a 1:1 molar ratio in PBS, pH 7-7.4 and 75 mM 2-mercaptoethanolamine (2-MEA) were mixed together and incubated at 25-37°C for 2 -6 h, after which the 2-MEA was removed by dialysis, diafiltration, tangential flow filtration and vortex cell filtration using standard methods.

Several anti-EGFR monospecific antibodies and anti-c-Met antibodies were pooled during in vitro Fab-arm exchange in a matrix to generate bispecific antibodies, which were subsequently further characterized. The generated bispecific antibodies were ranked using a four-step strategy using the following assays.

Step 1: Binding to NCI-H441, NCI-H1975 and A549 cells in a FACS (fluorescence excited cell sorting) assay.

Step 2: Inhibition of pMet phosphorylation in A549 cells.

Stage 3: inhibition of proliferation of NCI-H1975, KP4 and NCI-H441 cells.

Step 4: Inhibition of EGFR phosphorylation in A549 and SNU-5 cells.

It should be noted that the characteristics of the original antibodies were not preserved in the bispecific antibody. For example, the presence of certain EGFR-binding arms in a bispecific antibody resulted in a loss or decrease in inhibition or an increase in c-Met phosphorylation. Pairs were selected on the basis of characterization studies.

A monospecific bivalent anti-EGFR antibody, E1-K409R, was generated containing the VH and VL portions of the anti-EGFR antibody 2F8 having a VH of SEQ ID NO: 189 and a VL of SEQ ID NO: 190 (the 2F8 antibody is described in International Patent Publication No. WO 2002/ 100348) and an IgG1 constant region with a K409R substitution.

A monospecific bivalent anti-EGFR antibody, E1-F405L, was created containing the VH and VL regions of the anti-EGFR antibody 2F8 having a VH of SEQ ID NO: 189 and a VL of SEQ ID NO: 190 (the 2F8 antibody is described in International Patent Publication No. WO 2002/ 100348) and IgG1 constant region with F405L substitution.

A monospecific bivalent anti-EGFR antibody, E2-K409R, containing the VH and VL regions of an anti-EGFR antibody 018 having a VH of SEQ ID NO: 191 and a VL of SEQ ID NO: 192 was created (antibody 018 is described in International Patent Publication No. WO 2009/ 030239) and IgG1 constant region with K409R substitution.

A monospecific bivalent anti-EGFR antibody, E2-F405L, was generated containing the VH and VL regions of an anti-EGFR 018 antibody having a VH of SEQ ID NO: 191 and a VL of SEQ ID NO: 192 (antibody 018

- 65 039356 described in International Patent Publication No. WO 2009/030239) and an IgG1 constant region with the F405L substitution.

A monospecific bivalent anti-c-Met antibody, M1-K409R, was generated containing the VH and VL regions of an anti-c-Met 069 antibody having a VH of SEQ ID NO: 193 and a VL of SEQ ID NO: 194 (antibody 069 described in WO 2011/ 110642), and an IgG ₁ constant region with a K409R substitution.

A monospecific bivalent anti-c-Met antibody, M1-F405L, was generated containing the VH and VL regions of an anti-c-Met 069 antibody having a VH of SEQ ID NO: 193 and a VL of SEQ ID NO: 194 (antibody 069 described in WO 2011/ 110642), and an IgG ₁ constant region with the F405L substitution.

A monospecific anti-c-Met antibody, M2-K409R, was created containing the VH and VL regions of an anti-c-Met 058 antibody having a VH of SEQ ID NO: 195 and a VL of SEQ ID NO: 196 (antibody 058 is described in WO 2011/110642 ), and an IgG ₁ constant region with a K409R substitution.

A monospecific anti-c-Met antibody, M2-F405L, was created containing the VH and VL regions of an anti-c-Met 058 antibody having a VH of SEQ ID NO: 195 and a VL of SEQ ID NO: 196 (antibody 058 is described in WO 2011/110642 ), and an IgG ₁ constant region with the F405L substitution.

The VH, VL, HC and LC sequences of the antibodies are shown below:

> SEQ ID NO: 189, mAb to EGFR El VH

QVQLVESGGGVVQPGRSLRLSCAASGFTFSTYGMHWVRQAPGKGLEWVAVIWDDGSYKYYGD SVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDGITMVRGVMKDYFDYWGQGTLVTVS S > SEQ ID NO: 190, mAb to EGFR El VL

AIQLTQSPSSLSASVGDRVTITCRASQDISSALVWYQQKPGKAPKLLIYDASSLESGVPSRFSGSES

GTDFTLTISSLQPEDFATYYCQQFNSYPLTFGGGTKVEIK > SEQ ID NO: 191, mAb to EGFR E2 VH

EVQLVESGGG LVQPGGSLRL SCAASGFTFS SYWMNWVRQA PGKGLEWVAN IKKDGSEKYY

VDSVKGRFTI SRDNAKNSLY LQMNSLRAED TAVYYCARDL GWGWGWYFDL

- 66 039356

WGRGTLVTVSS > SEQ ID NO: 192, mAb to EGFR E2 VL

EIVLTQSPAT LSLSPGERAT LSCRASQSVS SYLAWYQQKP GQAPRLLIYD ASNRATGIPARFSGSGSGTD FTLTISSLEP EDFAVYYCQQ RSNWPPTFGQ GTKVEIK > SEQ ID NO: 193, mAb to c-Met Ml VH

QVQLVQSGAEVKKPGASVKVSCETSGYTFTSYGISWVRQAPGHGLEWMGWISAYNGYTNYAQ KLQGRVTMTTDTSTSTAYMELRSLRSDDTAVYYCARDLRGTNYFDYWGQGTLVTVSS > SEQ ID NO: 194, mAb to c-Met M1VL

SEQ ID NO: 195, mAb to c-Met M2 VH

SEQ ID NO: 196 c-Met mAb M2 VL

QLTQSPSSLSASVGDRVTITCRASQGLSSALAWYRQKPGKAPKLLIYDASSLESGVPSRFSGSGSG TDFTLTISSLQPEDFATYYCQQFTSYPQITFGQGTRLEIK > SEQ ID NO: 199, EMl-mAb Hl (EGFR antibody, 405L)

QVQLVESGGGVVQPGRSLRLSCAASGFTFSTYGMHWVRQAPGKGLEWVAVIWDDGSYKYYGD SVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDGITMVRGVMKDYFDYWGQGTLVTVS SASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYS LSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKP KDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLH QDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYP SDIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYT QKSLSLSPGK > SEQ ID NO: 200, EM-1 mAb LI

AIQLTQSPSSLSASVGDRVTITCRASQDISSALVWYQQKPGKAPKLLIYDASSLESGVPSRFSGSES GTDFTLTISSLQPEDFATYYCQQFNSYPLTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVC LLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVT HQGLSSPVTKSFNRGEC > SEQ ID NO: 201, EM-1 mAb H2 (K409R, антитело к c-Met)

- 67 039356

QVQLVQSGAEVKKPGASVKVSCETSGYTFTSYGISWVRQAPGHGLEWMGWISAYNGYTNYAQ KLQGRVTMTTDTSTSTAYMELRSLRSDDTAVYYCARDLRGTNYFDYWGQGTLVTVSSASTKGP SVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTV PSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMIS RTPEVTCWVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNG KEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPSDIAVEW ESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSP GK > SEQ ID NO: 202, EM-1 mAb L2

DIQMTQSPSSVSASVGDRVTITCRASQGISNWLAWFQHKPGKAPKLLIYAASSLLSGVPSRFSGSG SGTDFTLTISSLQPEDFATYYCQQANSFPITFGQGTRLEIKRTVAAPSVFIFPPSDEQLKSGTASVVC LLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVT HQGLSSPVTKSFNRGEC > SEQ ID NO: 234, E2-mAb HCl (EGFR-F405L)

EVQLVESGGG LVQPGGSLRL SCAASGFTFS SYWMNWVRQA PGKGLEWVAN IKKDGSEKYY VDSVKGRFTI SRDNAKNSLY LQMNSLRAED TAVYYCARDL

GWGWGWYFDLWGRGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN SGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTH TCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKT KPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPP SREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTVDKSRW QQGNVFSCSVMHEALHNHYTQKSLSLSPGK > SEQ ID NO: 235, E2-mAb LC1 (EGFR)

EIVLTQSPAT LSLSPGERAT LSCRASQSVS SYLAWYQQKP GQAPRLLIYD ASNRATGIPARFSGSGSGTD FTLTISSLEP EDFAVYYCQQ RSNWPPTFGQ GTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTE QDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC > SEQ ID NO: 236, E2-mAb HC2 (c-Met-K409R)

EVQLVESGGGLVKPGGSLKLSCAASGFTFSDYYMYWVRQTPEKRLEWVATISDDGSYTYYPDS VKGRFTISRDNAKNNLYLQMSSLKSEDTAMYYCAREGLYYYGSGSYYNQDYWGQGTLVTVSS ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSL SSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPK DTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQ DWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQQGNVFSCSVMHEALHNHYTQ KSLSLSPGK > SEQ ID NO: 237

E2-mAb LC2 (c-Met)

QLTQSPSSLSASVGDRVTITCRASQGLSSALAWYRQKPGKAPKLLIYDASSLESGVPSRFSGSGSG TDFTLTISSLQPEDFATYYCQQFTSYPQITFGQGTRLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCL LNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTKTLSKPVNGSFNTH

Generated monospecific antibodies to EGFR and c-Met were mixed to exchange Fab arms in vitro in a matrix and characterized by various assays. The bispecific EM1-mAb comprises an EGFR binding arm of mAb E1-F405L and a c-Met binding arm of mAb M1-K409R. The bispecific EM2-mAb comprises an EGFR binding arm mAb E2-F405L and a c-Met binding arm mAb M2-K409R. The bispecific EME-mAb comprises an EGFR binding arm mAb E1-K409R and a c-Met binding arm mAb M1-F405L. The bispecific EM4-mAb comprises an EGFR binding arm mAb E2-K409R and a c-Met binding arm mAb M2-F405L. EM1-mAb and EME-mAb had comparable characteristics.

The bispecific antibody EM1-mAb was cultured in a CHO cell line having reduced glycoprotein fucosylation capacity and therefore had a fucose content of approximately 1-15%. Removal of central fucose from 2-antennary complex-type oligosaccharides attached to Fc regions significantly enhances antibody ADCC through improved binding to FcYRIIIa without altering antigen binding or complement-dependent (CDC) activity. Such mAbs can be generated by a variety of methods that reportedly result in the successful expression of relatively highly defucosylated therapeutic antibodies that contain the complex-type 2-antennary oligosaccharides described above in Fc regions.

Example 11 Purification of EGFR/c-Met Bispecific Antibodies.

The bispecific EM1-mAb was further purified after Fab-arm exchange in vitro using hydrophobic interaction chromatography to minimize residual parental anti-c-Met and EGFR antibodies using standard methods.

Example 12 Characterization of EGFR/c-Met Bispecific Antibodies.

Various assays tested the characteristics of the EGFR/c-Met bispecific antibody EM1-mAb, including inhibition of EGF-stimulated EGFR phosphorylation, HGF-stimulated c-Met phosphorylation, ERK1/2 phosphorylation, AKT phosphorylation, inhibition of ligand binding, and cell viability. The characteristics of the EM1-mAb were compared with control monovalent EGFR- or c-Met-binding antibodies, as well as known EGFR inhibitors such as erlotinib (CAS 183321-74-6; tyrosine kinase inhibitor) and cetuximab (CAS 205923-56-4).

Since the parent antibodies for the EM-1 mAb are bivalent, control monovalent antibodies to EGFR and c-Met were generated in a bispecific format in combination with the unrelated/unbound antigen-binding Fab arm to more accurately compare synergy and avidity. bispecific mAb EM-1 with a mixture of appropriate control monovalent molecules.

To create control monovalent antibodies to EGFR and c-Met, a monospecific antibody to HIV gp120, gp120-K409R, containing a heavy chain of SEQ ID NO: 198 and a light chain of SEQ ID NO: 209 was created. A monospecific antibody to HIV gp120, gp120-F405L containing the heavy chain of SEQ ID NO: 197 and the light chain of SEQ ID NO: 209.

The control monovalent anti-EGFR mAb, E1-F405L-gp120-K409R, was generated by in vitro Fab-arm exchange between E1-F405L and gp120-K409R, and the control monovalent anti-c-Met mAb, M1-K409R-gp120-F405L, was generated by Fab-arm exchange in vitro between M1-K409R and gp120-F405L and purified as previously described.

The following cell lines were used to characterize bispecific antibodies: NCI-H292 (American Type Culture Collection (ATCC), cat. no. CRL-1848), NCI-H1975 (ATCC, cat. no. CRL-5908), SKMES-1 (ATCC , cat. no. NTV-58), A431 (ATCC, cat. no. CRL-1555), NCI-H441 (ATCC, cat. no. NTV-174), NCI-H3255 (DCTD tumor/linear cell storage, NCI , Frederick, NCI-Navy Medical oncology Cell Line supplement J. Cell Biochem suppl 24, 1996; Tracy S. Cancer Res. 64:7241-4, 2004; Shimamura T. Cancer Res. 65:6401-8, 2005) and HCC-827 (ATCC, cat. no. CRL-2868). NCI-H292 and SKMES-1 cells express both wild-type EGFR and wild-type c-Met. NCI-3255 expresses the L858R EGFR mutant and exhibits EGFR and c-Met amplification. H1975 expresses the L858R/T790M EGFR mutant and wild-type c-Met. HCC-827 expresses Δ (E746, A750) EGFR and exhibits EGFR amplification. The cell lines NCI-H292, NCI-H975, NCI-H441, and NCIH3255 are referred to interchangeably herein as H292, H975, H441, and H3255, respectively.

Binding of EGFR/c-Met bispecific antibodies to EGFR and c-Met on cells (A431 cell binding assay).

The EGFR/c-Met bispecific EM1-mAb was tested for EGFR and c-Met binding on cells using the protocol described in Example 3 (A431 Cell Binding Assay) and Example 6 (H441 Cell Binding Assay). Cetuximab and a control monovalent anti-EGFR antibody, E1-F405L-gp120-K409R, were used as controls for A431 cells. Cetuximab had an EC ₅₀ value of 5.0 nM. In table. 18 shows the EC ₅₀ values for binding. EM1-mAb showed a 1.9-fold (A431 cells) and 2.3-fold (H441 cells) reduction in binding compared to bivalent monospecific parental control antibodies. Cetuximab was comparable to divalent baseline control antibodies. EM1-mAb exhibited higher EC ₅₀ binding values than parent mAbs due to univalent binding of EGFR and c-Met. The EM1-mAb had EC ₅₀ values similar to the bispecific monovalent mAbs single arm E1/inert arm and E2/inert arm.

- 69 039356

Table 18

EC50 (nM) when bound to cells EMl-mAb Initial SHAL El-F405L-gpl20-K409R (A431 cells) or Ml - K409R -gpl20-F405L (H441 cells) A431 (analysis binding to EGFR) 9.6+3 5.1+0.3 10.1±0.6 H441 (analysis c-Met binding) 1.5 + 0.7 0.65 + 0.1 1.0 + 0.3

Inhibition of ligand binding to the receptor.

Bispecific antibodies were tested for their ability to block the binding of EGF to the extracellular domain of EGFR and HGF to the extracellular domain of c-Met in an enzyme immunoassay (ELISA). Recombinant human EGFR-Fc (R&D Systems, cat. no. 344-ER-050) or human HGF (R&D Systems, cat. no. 294-HGN-025/CF) was applied to MSD HighBind plates (Meso Scale Discovery, Meso Scale Discovery, New York). Gaithersburg, Maryland, USA) for 2 hours at room temperature. Buffer MSD Blocker A (Meso Scale Discovery, Gaithersburg, MD, USA) was added to each well and incubated for 2 hours at room temperature. The plates were washed three times with 0.1 M HEPES buffer, pH 7.4, followed by the addition of a mixture of either fluorescent dye (MSD) labeled EGF or biotinylated HGF proteins with varying concentrations of antibodies. The ruthenium-labeled EGF protein was incubated for 30 min at room temperature (RT) with increasing concentrations of various antibodies ranging from 1 nM to 4 μM. After 2 hours of culture with gentle shaking at room temperature, the plates were washed three times with 0.1 M HEPES buffer (pH 7.4). MSD Read Buffer T was diluted and added and signals were analyzed using a SECTOR Imager 6000. HGF inhibition assays were performed similarly to EGF/EGFR inhibition assays, except that 10 nM biotinylated HGF was incubated for 30 min at RT with increasing concentrations of different antibodies - from 1 nM to 2 μM.

EM1-mAb inhibited the binding of EGF to EGFR with an IC ₅₀ value of 10.7 ± 1.2 nM in the presence of 50 nM EGF and an IC ₅₀ value of 10.6 ± 1.5 nM in the presence of 80 nM EGF. The parent divalent antibody inhibited the binding of EGF to EGFR with an IC ₅₀ value of 0.14±1.5 nM in the presence of 50 nM EGF and an IC ₅₀ value of 1.7±1.4 nM in the presence of 80 nM EGF. The EM1 mAb had weaker inhibition of EGF binding to the extracellular domain of EGFR due to the univalent binding of the EM1 mAb compared to the bivalent parent mAb.

EM1-mAb inhibited HGF binding to c-Met with an IC ₅₀ value of 29.9±1.5 nM. The parent divalent antibody inhibited HGF binding to c-Met with an IC ₅₀ value of 14.8±1.6 nM. The EM1 mAb had weaker inhibition of HGF binding to the c-Met extracellular domain due to the univalent binding of the EM1 mAb compared to the parent divalent mAb.

Inhibition of EGF-stimulated EGFR phosphorylation and HGF-stimulated c-Met phosphorylation.

Antibodies were tested to determine IC50 values for inhibition of EGFR and c-Met phosphorylation. Inhibition of EGF-stimulated EGFR phosphorylation and HGF-stimulated c-Met phosphorylation were evaluated at various antibody concentrations (final 0.035-700 nM) as described in Example 2 (Inhibition of EGF-stimulated EGFR phosphorylation) and Example 6 (Inhibition of HGF-stimulated c-Met phosphorylation) . In some experiments, both EGF and HGF were added to cells so that the same cell lysate could be used to detect both EGFR phosphorylation and c-Met phosphorylation.

A control anti-EGFR mAb, E1-F405L-gp120-K409R, monovalent to EGFR, and the original low-fucose divalent anti-EGFR antibody were used as control antibodies. In table. 19 shows the IC ₅₀ values in the assays.

- 70 039356

Table 19

pEGFR (1C ₅₀ , nM) pMet (1C ₅₀ , nM) cell line H292 H1975 H292 H1975 Antibody EMl-mAb 8.6-29 1.5 0.55-0.83 0.64 El-F405L-gpl20-K409R 10.9-13.1 But 0.7-4 But Anti-EGFR parent mAb (F405L)* 1.5 But No effect But

*Antibody was low in fucose.

Enhanced inhibition of pERK and pAKT by EM1-mAb compared to monovalent antibody mixture (pERK assay (anti-pAKT mAb assay)).

The potential for enhancing the efficacy of the EGFR/c-Met bispecific antibody was assessed by analyzing the effects of the mAb on pERK and pAKT signaling downstream. In these experiments, a mixture of anti-EGFR monovalent control antibodies and anti-c-Met monovalent control antibodies was compared with the bispecific EM1-mAb. Cells were plated in clear 96-well tissue culture-treated plates (Nunc) in 100 µl/well RPMI medium containing GlutaMAX and 25 mM Hepes (Invitrogen) supplemented with 1 mM sodium pyruvate (Gibco), 0.1 mM NEAA (Gibco), 10% heat-inactivated fetal calf serum (Gibco) and 7.5 ng/ml HGF (R&D Systems, Cat # 294-HGN) and allowed to attach overnight at 37° C. in a humidified atmosphere containing 5 %CO2. Cells were not exposed to serum-free medium. Cells were treated for 30 minutes (pERK assay) or 1 hour (pAkt assay) with various concentrations (0.11-700 nM final) of monovalent control antibodies or EM1-mAb.

Cells were analyzed for pERK or pAKT levels using the kits below and following the manufacturer's instructions, Meso Scale Discovery: Phospho-ERK1/2 (Thr202/Tyr204; Thr185/Tyr187) Assay Whole Cell Lysate Kit (Cat. No. K151DWD, Phospho-Akt (Ser473) Assay Whole Cell Lysate Kit (Cat. No. K151CAD), Phospho-Akt (Thr308) Assay Whole Cell Lysate Kit (Cat. No. K151DYD) In the pERK assay, cells were lysed and whole cell lysates were added to plates, coated with an anti-phospho-ERK1/2 antibody (recognizing ERK1 phosphorylated at Thr202 and Tyr204, and ERK2 phosphorylated at Thr185 and Tyr187) and phosphorylated ERK1/2 were detected using an anti-ERK1/2 antibody conjugated with the MSD SULFO reagent -TAG™ In the pAKT Ser473 assay, the capture antibody was anti-total AKT and the detection antibody was anti-pAKT Ser473 conjugated with the MSD SULFO-TAG™ reagent. AKT and detection antibody - anti-pAKT Thr308 antibody conjugated with MSD SULFO-TAG™ reagent.

Plates were read on a SECTOR® Imager 6000 (Meso Scale Discovery) using the manufacturer's default settings for this assay. Electrochemiluminescence signal was plotted against the logarithm of antibody concentration and IC50 values were determined by fitting sigmoid dose-response curve data with variable slope using _GraphPad Prism 5 software. Cell lines NCI-H292, H1975 and SKMES-1 were used in these assays.

The IC ₅₀ for inhibition of ERK phosphorylation by the bispecific EM1-mAb was approximately 14-63-fold lower relative to a mixture of two monovalent control antibodies, depending on the cell line tested (Table 20). The increased potency of the EM1-mAb compared to a mixture of two monovalent control antibodies indicates a cooperative or avidity effect due to improved binding of the EM1-mAb to these cell lines. The IC50 for inhibition of AKT phosphorylation at _Ser475 (pAKTS475) and Thr308 (pAKTT308) in the NCI-H1975 cell line was approximately 75- and 122-fold lower, respectively, compared to a mixture of two monovalent control antibodies (Table 21). The increased potency of EM1-mAb compared to a mixture of two monovalent control antibodies indicates a cooperative or avidity effect due to improved binding of EM1-mAb to these cell lines. Thus, the bispecific nature of EM1-mAb resulted in an enhanced effect on downstream signaling effectors.

- 71 039356

Table 20

1C ₅₀ (nM) pERK Antibody cell line Bispecific antibody EM1mAb Mix E1- F405L-gpl20K409R and M1K409R-gpl20- F405L The fold change for bispecific antibodies compared to a mixture of two monovalent control antibodies H292 0.64 34.94 55 H1975 1.67 106 63 SKMES-1 0.54 7.63 fourteen

Table 21

Antibody 1С ₅₀ (nM) pAKTS473 1C ₅₀ (nM) pAKTT308 Bispecific antibody EML-mAb 0.87 0.96 Mix of El-F405L-gpl20-K409R and M1-K409Rgpl20-F405L 65 117 Multiplicity of change for a mixture of two monovalent antibodies in comparison with bispecific 75 122

Antibody inhibition of growth or viability of human tumor cells.

Inhibition of c-Met-dependent cell growth was assessed by measuring the viability of various tumor cells after exposure to bispecific antibodies EM1-mAb. The studies used NCI-H292, SKMES-1, NCI-H1975 and NCI-H3255 cells.

Cells were cultured in a standard 2D format and in a low attachment format. Erlotinib and cetuximab were used as controls. In table. 22 summarizes the IC50 values in the assays.

Antibody inhibition of growth or viability of human tumor cells is a standard two-dimensional format.

Cell growth inhibition was analyzed by measuring the viability of NCI-H292 and NCI-H1975 after exposure to antibodies in two formats. Using a standard two-dimensional format, cells were plated on white opaque 96-well tissue culture-treated plates (PerkinElmer) in RPMI medium containing GlutaMAX and 25 mM Hepes (Invitrogen) supplemented with 1 mM sodium pyruvate (Gibco), 0. 1 mM NEAA (Gibco) and 10% heat-inactivated fetal calf serum (Gibco) then allowed to attach overnight at 37° C., 5% CO2. Cells were treated with various concentrations of antibodies (final concentration 0.035-700 nM) as well as HGF (7.5 ng/ml, R&D Systems, Cat # 294-HGF), then incubated at 37°C, 5% CO2, for 72 hours Some wells were left untreated with either HGF or antibodies as controls. Viable cells were detected using the CellTiter-Glo® reagent (Promega) and data analyzed as described in Example 3 Growth Inhibition of Human Tumor Cells (NCI-H292 and NCI-H322 Cell Growth Assay).

Antibody inhibition of growth or viability of human tumor cells - low attachment format.

To evaluate survival under low attachment conditions, cells were plated in 96-well Ultra Low Attachment plates (Corning Costar) in 50 µl/well of RPMI medium (Invitrogen) containing GlutaMAX and 25 mM Hepes supplemented with 1 mM sodium pyruvate (Gibco ), 0.1 mM NEAA (Gibco) and 10% heat-inactivated fetal bovine serum (Gibco), after which it was allowed to attach overnight at 37° C., 5% CO ₂ . Cells were treated with various concentrations of antibodies (final concentration 0.035-700 nM) as well as HGF (7.5 ng/ml, R&D Systems cat. no. 294-HGN), then incubated at 37°C, 5% CO2, for 72 h Some wells were left untreated with either HGF or antibodies as controls. Viable cells were detected using the CellTiter-Glo® reagent (Promega) and data were analyzed as described above under Inhibition of Human Tumor Cell Growth (NCI-H292 and NCI-H322 Cell Growth Assay) in Example 3, except that lysates were transferred to opaque white 96-well tissue culture-treated plates (PerkinElmer) before luminescence was recorded.

In standard two-dimensional culture, EM1-mAb inhibited the growth of NCI-H292 with an IC ₅₀ of 31 nM and under low attachment conditions with an IC _{50 of} 0.64 nM. EM-1 mAb inhibited growth of NCI-H1975 cells with IC ₅₀ >700 nM and 0.8-1.35 nM in standard 2D culture and low attachment, respectively. In NCI-H292 cells expressing both EGFR and wild-type c-Met, EM1-mAb had a potency that was 22-fold higher in standard two-dimensional culture and approximately 330-fold higher under low attachment culture conditions compared to cetuximab. In NCI-H1975 cells expressing the EGFR L858R mutant, T790M, and wild-type c-Met, mAb EM-1 was at least 518-fold

- 72 039356 more effective than cetuximab in conditions of low culture attachment. In table. 22 shows a summary of the assays.

Table 22

cell line Status EGFR c-Met status EMl-mAb, standard two-dimensional culture, 1C ₅₀ (nM) EMl-mAb, low attachment, 1C ₅₀ (nM) Cetuximab, standard two-dimensional culture, 1C ₅₀ (nM) Cetuximab, low attachment, 1C ₅₀ (nM) NCI-H292 d.t. D.t. 31 0.64 >700 212 NCI-H1975 L858R, T790M D.t. >700 0.8-1.35 >700 >700

The combination of erlotinib and EM1-mAb effectively inhibits the growth of cell lines with mutant EGFR.

Cell growth inhibition by the combination of erlotinib and EM1-mAb was assessed in both standard two-dimensional culture conditions and in a low-attachment format. NCI-H3255 and HCC-827 cells were seeded as described above under Antibody Inhibition of Growth or Viability of Human Tumor Cells. Along with antibody treatment, HGF (7.5 ng/mL, R&D Systems, Cat # 294-HGN) was added to the cells. Cells were treated with various concentrations of antibodies (final - 0.11-700 nM), or erlotinib (final - 0.46-3000 nM), or a combination of erlotinib with antibody using increasing amounts of each at a fixed ratio (for example, the lowest concentration of the combination = the lowest antibody concentration (0.11 nM) + the lowest concentration of erlotinib (0.46 nM)). Some wells were left untreated with either HGF or antibodies as controls. Cells were incubated at 37°C, 5% CO2 for 72 h, after which viable cells were detected using the CellTiter-Glo® reagent (Promega) and analyzed as described above under Inhibition of Human Tumor Cell Growth (NCI Cell Growth Assay). -H292 and NCI-H322). In table. 23 summarizes the results of the experiment. In the table, IC ₅₀ values for the combinations are given relative to either the antibody or erlotinib, whichever is indicated in brackets.

In both NCI-H3255 and HCC-827 cells (cell lines with mutant EGFR), the addition of EM1-mAb to erlotinib both increased the potency of cell viability inhibition and was more effective in producing fewer total viable cells. In NCI-H3255 cells using a standard two-dimensional format, the IC50 for erlotinib alone was 122 nM, while for the combination it was 49 nM. Similarly, in HCC-827 cells, the IC ₅₀ for erlotinib alone was 27 nM, while for the combination it was 15 nM. In addition, the combination of erlotinib and EM1-mAb was also more effective than the combination of erlotinib and cetuximab. Thus, in this assay, in the presence of HGF, the addition of EM1-mAb increased the efficacy of erlotinib.

NCI-H3255 cells express the EGFR L858R mutant and amplified c-Met. HCC827 cells express EGFR mutants with deletions at amino acid positions 746 and 750, as well as wild-type c-Met. The EM1-mAb had greater effects on the viability of HCC-827 and NCI-3255 in the presence of erlotinib than erlotinib alone, either under standard or low culture attachment conditions.

Table 23

Samples IC ₅₀ mAb EM1 + erlotinib (nM) Erlotinib IC50 (nM) NCI-H3255 Standard 2D Culture 49.0 122 NCI-H3255 low attachment culture 10.6 47.1 HCC-827 Standard 2D Culture 14.6 27.4 HCC-827 low attachment culture 3.5 9.5

Example 13 Antibody-mediated cellular cytotoxicity (ADCC) of EM1-mAb in cell lines in vitro.

ADCC assays were performed as previously described (Scallon et al., Mol. Immunol. 44:1524-1534, 2007).

- 73 039356

Briefly, human peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll gradients and used as effector cells for ADCC assays. NCI-H292, NCI-H1975 or NCI-H441 cells were used as target cells in a ratio of 1 target cell to 50 effector cells. Target cells were prelabeled with BATDA (PerkinElmer) for 20 min at 37°C, washed twice and redissolved in DMEM supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine (all from Invitrogen). Target cells (1 x 10 ⁴ cells) and effector cells (0.5 x 10 ⁶ cells) were pooled and 100 μl of cells were added to the wells of a 96-well U-bottomed plate. An additional 100 μl of wild-type wLb constructs with or without protease resistance was added. All samples were processed in duplicate. The plates were centrifuged at 200xg for 3 minutes, incubated at 37°C for 2 hours, and then centrifuged again at 200xg for 3 minutes. A total of 20 μl of supernatant per well was taken and cell lysis was measured by adding 200 μl of europium-based DELPHIA reagent (PerkinElmer). Fluorescence was measured using an Envision 2101 Multilabel Reader (PerkinElmer). The data were normalized to the maximum cytotoxicity of 0.67% Triton X-100 (Sigma Aldrich) and the minimum control, which was determined by the spontaneous release of BATDA from target cells in the absence of any antibody, using the equation below: (experimental release - spontaneous release)/ (maximum release - spontaneous release) x100%. The data were fitted with a sigmoid dose response model using GraphPad Prism v5.

The ADCC results for the EM1 mAb and comparators are summarized in Table 1. 24 (NCI-H292 cells), tab. 25 (NCI-H1975 cells) and tab. 26 (NCI-H441 cells), and in the table. 27 (NCI-H1993 cells) lists EC ₅ o values and maximum lysis achieved. NCI-H292 cells express wild-type (wt) EGFR, c-Met wt. and KRAS d.t.; NCI-H1975 cells express mutant EGFR (L858R T790M), c-Met d.t. and KRAS d.t.; NCI-H441 cells express EGFR d.t., c-Met d.t. and a KRAS (G12V) mutation, and NCI-H1993 cells express EGFR d.t., amplified c-Met, and KRAS d.t. KRAS is also known as GTPase KRas and also as a homologue of the rat Kirsten's sarcoma viral oncogene V-Ki-ras2.

EM 1-mAb has higher ADCC response efficiency than cetuximab and the normal fucose version of EMl-mAb, as indicated by lower EC ₅ o EMI mAb has higher potency in terms of maximum lysis achieved than cetuximab and bispecific mAb with normal fucose content. As shown in the profiles in Table. 24-27, EM-1 mAb exerts ADCC activity on cells having mutant and wild type EGFR, wild type with normal and amplified c-Met levels, and wild type KRAS and mutation.

Table 24

mAb Efficiency (EUzo, mcg/ml) ^R2 Efficiency (maximum lysis achieved) EMI mAb 0.0058 0.93 19% EGFR bispecific x normal fucose c-Met mAb 0.22 0.85 13% Cetuximab 0.0064 0.94 12%

Table 25

mAb Efficiency (EUzo, mcg/ml) ^R2 Efficiency (maximum lysis achieved) EMI mAb 0.022 0.91 19% EGFR bispecific x normal fucose c-Met mAb 1.8 0.79 13% Cetuximab 0.029 0.70 eleven%

-74039356

Table 26

mAb Efficiency (ECeo, mcg/ml) ^R2 Efficiency (maximum lysis achieved) EMI mAb 0.022 0.97 24% EGFR bispecific x normal fucose c-Met mAb 0.52 0.87 7.9% Cetuximab 0.013 0.85 fifteen%

Table 27

mAb Efficacy (EUzo µg/ml) ^R2 Efficiency (maximum lysis achieved) EMI mAb 0.0013 0.95 27% Bispecific to EGFR χ c-Met normal fucose mAb 0.054 0.87 17% Cetuximab 0.0042 0.76 21%

Example 14 Antitumor Efficacy Studies of EM1-mAb.

The efficacy of the EM1 mAb on tumor growth was evaluated as described in Example 7 Efficacy Study of EGFR/c-Met Bispecific Molecules on Tumor. Briefly, NCI-H292-HGF cells were implanted subcutaneously (SC) with Cultrex at 2x106 female SCID Beige mice. Mice were divided by tumor volume 7 days after implantation into 5 groups, 10 mice per group. The introduction was started after the average tumor volume in the group reached 62-66 mm ³ (small tumors). PBS or therapeutic agent was administered intraperitoneally (i/p) 2 times a week.

SKMES-HGF, human squamous cell carcinoma cells transfected with human HGF (liver growth factor) was also used in the efficacy evaluation. These cells were implanted sc at 10x106 in female SCID Beige mice. These mice were divided by tumor volume 12 days after implantation into 5 groups, 8 mice per group. The first study was started after the average tumor volume in the group reached 98-101 mm ³ (large tumors). PBS or therapeutic mAbs were administered i/p 2 times a week for 4 weeks. In the study of larger tumors, mice were divided into 2 groups after the tumors reached a volume of approximately 200-300 mm ³ . These mice were then treated with either cetuximab (20 mg/kg) or EM1-mAb (20 mg/kg), ip, twice a week (3 weeks).

Summary data are shown in table. 28. In FIG. 10 shows the efficiency of the molecules over time. EM1-mAb had an improved tumor suppression profile compared to cetuximab in the H292-HGF small tumor model and in the SKMES-HGF small and large tumor models.

Table 28

sample and time cell line Dose in Partial Complete mg per kg regression regression EM1 for 35 days H292-HGF, twenty 10/10 10/10 small tumor 5 10/10 10/10 one 0/10 0/10 Cetuximab on day 35 H292-HGF, twenty 0/10 0/10 small tumor EM1 on day 67 SKMES-HGF, twenty 0/8 8/8 small tumor 5 1/8 6/8 one 2/8 4/8 Cetuximab on day 67 twenty 0/8 6/8 EM1 for 70 days SKMES-HGF, twenty 4/7 3/7 Cetuximab on day 35 large tumor twenty 0/7 0/7

- 75 039356

In table. 29 shows tumor sizes in the SKMES-HGF tumor treatment groups, and Table. 30 shows antitumor activity.

EM1-mAb inhibited tumor growth in the SKMES-HGF model by 97% or more at multiple doses up to 1 mg/kg. Although cetuximab was initially very effective (88% TGI at 20 mg/kg), cetuximab-treated tumors grew back after administration ended. In contrast, tumors treated with EM1-mAb 5 or 20 mg/kg did not grow during the study (>2 months).

Table 29

Day Tumor volume (mm ³ ) Carrier Bispecific EM1, 20 mg/kg Bispecific EM1, 5 mg/kg Bispecific EMI, 1 mg/kg Cetuximab, 20mg/kg one 99±6 99±7 101±6 101±6 98±5 eight 146 ± 14 48 ± 10 49±9 49±10 60±8 fifteen 192±21 9 ± 1 22 ± 10 41 ± 13 44±23 22 326±43 3±2 17 ± 12 33±15 42±23 29 577±55 2 ± 1 15±9 38 ± 17 85±60 36 994± 114 0.2±0.1 13±9 26 ± 14 125 ± 62 fifty — 0.04±0.04 10±7 18 ± 9 423 ± 115 57 |- 0.1±0.2 3±2 21 ± 10 650 ± 116 67 - 0±0 8±7 34±22 1257±151

Table 30

Treatment Tumor size (mm ³ ) on day 36 T/C (%) T - C (day) at 1000 mm ³ P value Carrier 994± 114 — — — Bispecific EM1, 20 mg/kg 0.19±0.12 0.02 -- Bispecific EM1, 5 mg/kg 13±9 1.3 -- Bispecific EMI, 1 mg/kg 26 ± 14 2.6 -- Cetuximab (20 mg/kg) 125 ± 62 13 31

Example 15 Inhibition of EM1-mAb cell migration in vitro.

Way.

The effect of EM-mAb and control monovalent antibodies on the inhibition of tumor cell migration was analyzed in NIH-1650 cells. The EGFR mutant H1650 cell line (pulmonary bronchial alveolar carcinoma cells having a mutation in exon 19 [E746, A750 deletion]) were cultured in tissue culture flasks under normal culture conditions (37° C., 5% CO ₂ , 95% humidity). All media and supplements were as recommended by the cell supplier (American Type Culture Collection, Manassas, VA, USA)

Spheroids were generated by seeding H1650 lung tumor cells, at 10,000 cells/well, into Ultra Low Adherence (ULA) U-bottom 96-well plates (Corning, Tewkesbury, USA) at 200 μl/well. These plates stimulate the spontaneous formation of a single spheroid from cells within 24 h (incubation at 37°C, 5% CO ₂ ). The spheroids were grown for four days under normal cultivation conditions.

Round bottom 96 well plates (BD Bioscience) were coated with 0.1% gelatin (EMD Millipore, Billerica, USA) in sterile water for 1 hour at 37°C. In compound evaluation studies at day 4, spheroids of 10,000 tumor cells (H1650 and NCI-H1975) were transferred to round bottom coated plates and treated with EM1-mAb, a low-fucose control monovalent anti-EGFR mAb, E1-F405L-gp120-K409R , control monovalent anti-c-Met mAbs, M1-K409R-gp120-F405L, having low fucose content, and a combination of two monovalent antibodies E1-F405L-gp120-K409R and M1-K409R-gp120-F405L (produced with low fucose content) in dilution series with 20 ng/ml HGF (R&D systems). Controls were treated with vehicle which was an IgG ₁ kappa isotype control (concentration equal to the highest for drug-treated cells). Compound exposure was analyzed after 48 hours by measuring the area covered by migrating cells using bright field imaging on a fully automated Operetta multi-parameter imaging system (Perkin Elmer) with a 2x objective. Inhibition of cell migration (total area) due to treatment exposure was analyzed by normalizing the data using dividing by the value of the medium-only control to give the percentage of cell migration relative to the control. Thus, a value less than 1 would mean inhibition of cell migration.

- 76 039356

Results.

EM1-mAb demonstrated potent synergistic inhibition of cell migration in H1650 (EGFR L858R mutant) and H1975 (EGFR L858R/T790M mutant) cells compared to the combination of anti-EGRF and c-Met monovalent control antibodies, E1-F405L-gp120-K409R and M1- K409R-gp120-F405L. In H1650 cells, the six highest concentrations of EM1-mAb significantly inhibited cell migration (p < 0.001) compared to the isotype control. The EC ₅₀ value for EM1-mAb was 0.23 nM, while the EC ₅₀ value for the monospecific control antibody combination was 4.39 nM. Therefore, EM1-mAb was approximately 19 times more effective at inhibiting H1650 cell migration compared to the monovalent control antibody combination. The level of cell migration inhibition of EM1-mAb was higher than that of the combination of monospecific control mAbs in H1650 and H1975 cells. In table. 31 shows the EC ₅₀ values for analysis.

Table 31

Samples H1650 H1975 EU ₅₀ (nM) Inhibition at 30 nM Inhibition at 30 nM I eat 1-mAb 0.23 64% 38% Mix of El-F405L-gpl20-K409R* and Ml-K409R-gpl20-F405L* 4.39 59% twenty% E1-F405L - gp!20-K409R* 5.44 fifteen% 7% Ml-K409R-gpl20-F405L* 7.36 43 ten%

* Antibodies were low in fucose.

Example 16 Epitope Mapping of Anti-c-Met-069 and 5D5 Antibodies.

The mAb binding epitope to c-Met 069 was mapped using linear and constrained CLIPS peptide technology. Peptides were scanned by SEMA, PSI and Ig domains of human c-Met. Linear peptides and CLIPS peptides were synthesized using the amino acid sequence of the aforementioned c-Met using standard Fmoc chemistry and deprotected with acceptor trifluoroacetic acid. Staggered peptides were synthesized on chemical scaffolds to reconstruct conformational epitopes using the technology of chemically linked peptides on scaffolds (CLIPS) (Timmerman et al., J. Mol. Recognition. 20:283, 2007). Linear and staggered peptides were attached to PEPSCAN plates and screened using PEPSCAN-targeted ELISA (Slootstra et al., Molecular Diversity, 1, 87-96, 1996). The c-Met 069 mAb binding epitope is a contiguous epitope consisting of c-Met amino acids 239-253 PEFRDSYPIKYVHAF (SEQ ID NO:238) and 346-361 FAQSKPDSAEPMDRSA (SEQ ID NO:239). The amino acid sequence of c-Met is shown in SEQ ID NO: 201.

Similar methods were used to map the epitope of mAb 5D5 (MetMab, onartuzumab). mAb 5D5 binds to c-Met at residues 325-340 PGAQLARQIGASLNDD (SEQ ID NO: 240).

Example 17 In Vivo Antitumor Efficacy Studies of EM1-mAb.

The efficacy of EM1-mAb on tumor growth was tested as described in Example 7 Efficacy studies of EGFR/c-Met bispecific molecules on tumor and in Example 14 using additional tumor cell lines with an EGFR mutation or amplifications of EGFR and/or c- Met. Briefly, SNU-5, H1975, HCC827 cells, H1975 cells expressing human HGF, or an HCC827 cell clone selected for its increased resistance to erlotinib (HCC827-ER1 cells) were implanted subcutaneously (sc) in female nude mice, except SNU-5 cells that were implanted in CR17/SCID mice. Mice were injected intraperitoneally with PBS or EM1-mAb, cetuximab (CAS 205923-56-4), erlotinib (CAS 183321-74-6), afatinib (CAS 439081-18-2), or a combination of mAb EM-1 and afatinib and mAb EM- 1 and erlotinib at the indicated dosage and according to the scheme shown in table. 32. Antitumor efficacy was measured as %TGI (tumor growth inhibition), calculated as 100 - %T/C (T = average tumor size in the treatment group; C = average tumor size in the control group on the indicated day, as described in example 7 ).

In tumors with primary activating EGFR mutations (without resistance to EGFR tyrosine kinase inhibitors (TKI)): (tumor HCC827, EGFR del (E746, A750)), EM1-mAb at a dose of 10 mg/kg inhibited tumor growth by 82%. Erlotinib in this model had similar efficacy as the combination of erlotinib and EM1-mAb. In FIG. 11 shows the efficacy of therapeutic agents over time in the HCC827 tumor model.

In tumors with wild-type EGFR and c-Met gene amplification (SNU-5 gastric cancer model), EM1-mAb demonstrated antitumor efficacy with complete tumor regression (98% TGI at day 34, p<0.01 vs. vehicle using single-factor dispersion

- 77 039356 analysis (ANOVA) followed by individual comparisons using the Games-Howell test). The antitumor activity of cetuximab (mAb to EGFR) in this model was less - 49% on day 34. In FIG. 12 shows the effectiveness of therapeutic agents over time in the model

SNU-5.

EM1-mAb was tested in a non-small cell lung cancer (NSCLC) model containing a primary activating EGFR mutation and an EGFR T790M mutation conferring resistance to 1st generation EGFR tyrosine kinase inhibitors in tumors (Model H1975). EM1-mAb inhibited tumor growth with a TGI value of 57% in the athymic mouse H1975 cell line implantation model (p<0.0001 vs. PBS vehicle using Prism 3.03 log-rank analysis). As expected, erlotinib was not effective in this T790M mutation model. Afatinib was as effective as EM1-mAb (57% TGI). Cetuximab and the combination of EM1-mAb with afatinib were most effective, causing tumor regression with tumor growth inhibition of 91% and 96%, respectively (p<0.0001 for both cetuximab versus PBS and EM1-mAb + afatinib versus PBS + afatinib vehicle group using log-rank analysis in Prism 3.03 software). c-Met signaling pathways were not activated in this model because mouse HGF does not bind to human c-Met.

EM1-mAb was tested in several models designed to express human HGF in a lentivirus transduction system. This allows ligand activation of the cMet pathway to be modeled in vivo, since murine HGF does not activate human c-Met on implanted human tumor cells. The results of the SKMES-HGF model are shown in Example 14 and FIG. 10, and the TGI values in % are summarized in Table. 32. EM1-mAb inhibited tumor growth in the H1975-HGF model by 71% (p<0.0001 compared to PBS vehicle using Prism 3.03 log-rank analysis). Afatinib, erlotinib, and cetuximab were less effective in this model. The combination of EM1-mAb and afatinib was very effective (96% TGI, p<0.0001, compared with the PBS + afatinib vehicle group using Prism 3.03 log-rank analysis). In FIG. 13 shows the performance of the molecules over time in the H1975-HGF model. Thus, erlotinib, afatinib, and cetuximab lose their antitumor efficacy in tumor models in which the c-Met pathway is activated.

EM1-mAb was tested in a tumor model characterized by a primary EGFR-activating mutation and increased resistance to the 1st generation EGFR TKI (erlotinib) due to amplification of the c-Met gene (model HCC827-ER1). EM1-mAb at 10 mg/kg resulted in partial regression of implanted HCC827-ER1 tumors with a TGI of 86% at day 25 and was more effective than erlotinib alone (65% TGI at day 25). The combination of EM1-mAb and erlotinib did not provide an additional increase in efficacy. In FIG. 14 shows the efficiency of the molecules over time.

Thus, EM1-mAb is effective in wild-type EGFR tumor models, with primary EGFR-activating mutations, with the EGFR T790M mutation associated with EGFR therapeutic resistance, and in models in which c-Met is activated by ligand- dependent (autocrine expression of HGF) or ligand-independent (amplification of the c-Met gene) in a manner. Combining EM1-mAb with erlotinib or afatinib may improve efficacy in some tumor models.

- 78 039356

Table 32

Type of tumor EGFR c-Met Treatment % TGI (days of study); compared to PBS vehicle administration (dose in mg/kg), scheme SKMES-HGF, lung squamous cell carcinoma D.t. D.t. EM1-mAb (20), BIWx4 weeks Cetuximab (20), BIW x4 weeks 100 (36) 88 (36) SNU-5 stomach cancer Λ τ. ampl. EM1-mAb (10), BIWx4 weeks Cetuximab (10), BIW x4 weeks 98 (34) 49 (34) H1975 NSCLC L858R; T790M D.t. EM1-mAb (10), BIWx3 weeks Cetuximab (10), BIW x3 weeks 57 (18) 91 (18) Erlotinib (50), QDx21 days 9 (18) Afatinib (15), QD x 21 days 57 (18) EM1-mAb (10), BIW x3 weeks + afatinib (15), QD x 21 days 96 (18) H1975-HGF NSCLC L858R; T790M D.t. EM1-mAb (10), BIWx3 weeks Cetuximab (10), BIW x3 weeks 71 (16) 42 (16) Erlotinib (50), QDx21 days 20 (16) Afatinib (15), QD x 21 days 29 (16) EM1-mAb (10), BIW x3 weeks + afatinib (15), QD x 21 days 96 (16) HCC827 NSCLC del (E746, A750); ampl. D.t. EM1-mAb (10), BIWx4 weeks 82 (35) Erlotinib (25), QD x 28 days 79 (35) EM1-mAb (10), BIW x3 weeks + erlotinib (25), QD x28 days 78 (35) HCC827-ER1 NSCLC del (E746, A750); ampl. ampl. EM1-mAb (10), BIWx4 weeks 86 (25) Erlotinib (25), QD x 28 days 65 (25) EM1-mAb (10), BIW x3 weeks + erlotinib (25), QD x28 days 87 (25)

BIW = twice a week;

QD = once a day; d.t. = wild type; ampl. = amplified.

Example 18 EM1-mAb induces degradation of EGFR and c-Met in vivo.

To demonstrate EM1-mAb involvement of both EGFR and c-Met in the tumor, H1975-HGF tumors were sampled at various time points following a single dose of 20 mg/kg EM1-mAb. Tumor lysates were prepared, normalized to total protein, and samples run on SDS-PAGE gels. Gels were transferred to nitrocellulose and analyzed by Western blotting for either EGFR (mouse anti-human EGFR mAb (EGF-R2); Santa Cruz Biotechnology, Cat # sc-73511) or c-Met ( mouse monoclonal antibody to human Met (L41G3; Cell Signaling Technology, Cat # 3148). EGFR levels normalized to GAPDH; c-Met levels normalized to actin. Receptor levels in EM1-mAb-treated tumors were compared to receptor levels in PBS-treated tumors, yielding % total receptors. Treatment with EM1-mAb reduced the total levels of EGFR and c-Met receptors in H1975-HGF tumors to values ranging from 20% to 60% of control, depending on the analyzed time point. In FIG. 15 shows mean receptor levels versus PBS over time. pEGFR, pc-Met and pAKT also decreased 72 hours after a single dose of EM1.

Example 19 Antitumor activity in comparison of IgG1 and IgG2 isoform variants bispecific to EGFR/c-Met mAb.

- 79 039356

To better understand the contribution of effector function to the potency observed in the H1975-HGF model, a comparison was made between EM1-mAb and an EM1-mAb variant having an IgG2 Fc with effector-reducing substitutions V234A/G237A/P238S/H268A/V309L/A330S/P331S in IgG2 (substitutions are described in the application for international patent No. WO 2011/066501) (numbering corresponds to the EU catalog). The IgG ₂ antibody with V234A/G237A/P238S/H268A/V309L/A330S/P331S substitutions does not interact with Fc receptors or effector cells (such as natural killer cells and macrophages). Thus, any loss of activity observed in the IgG ₂ variant V234A/G237A/P238S/H268A/V309L/A330S/P331S EM1-mAb may reflect the contribution of effector-mediated mechanisms such as ADCC and/or ADCP to antitumor activity. 32 days after implantation of tumor cells in the above H1975-HGF model, a loss of antitumor activity of the IgG ₂ variant was observed.

V234A/G237A/P238S/H268A/V309L/A330S/P331S EM1-mAb compared to parent EM1-mAb, indicating that effector-mediated mechanisms contribute to EM-1 mAb function. In FIG. 16 shows the antitumor activity of the molecules.

Sequence listing

SEQID NO: Type of View Description Subsequence one PRT Artificial Tencon LPAPKNLWSEVTEDSLRLSWTAPDAAFDSFLIQY QESEKVGEAINLT VPGSERSYDLTGLKPGTEYTVSIYGVKGGHRSNP LSAEFTT 2 DNA Artificial POP2220 GGAAAACAGGATCTACCATGCTGCCGGCGCCGA AAAACCTGGTTGT TTCTGAAGTTACC 3 DNA Artificial TC5'toFG AACACCGT AGAT AGAAACGGT four DNA Artificial 130mer CGGCGGTT AGAACGCGGCT ACAATT AAT ACAT AACCCCAT CCCCC TGTTGACAATTAATCATCGGCTCGTATAATGTG TGGAATTGTGAGC GGATAACAATTTCACACAGGAAACAGGATCTAC CATGCTG 5 DNA Artificial POP2222 CGGCGGTTAGAACGCGGCTAC 6 DNA Artificial TCF7 GGTGGTGAATTCCGCAGACAGCGGSNNNSNNSN NSNNSNNSNNSNN AACACCGTAGATAGAAACGGT 7 DNA Artificial TCF8 GGTGGTGAATTCCGCAGACAGCGGSNNSNNSN NSNNNSNNSNNSNN S NNAACACCGT AGAT AGAAACGGT eight DNA Artificial TCF9 GGTGGTGAATTCCGCAGACAGCGGSNNNSNNSN SNNSNNSNNSNN SNNS NNAACACCGTAGATAGAAACGGT 9 DNA Artificial TCF10 GGTGGTGAATTCCGCAGACAGCGGSNNNSNNSN NSNNSNNSNNSNN SNNSNNS NNAACACCGTAGATAGAAACGGT ten DNA Artificial TCF11 GGTGGTGAATTCCGCAGACAGCGGSNNNSNNSN SNNSNNSNNSNN SNNSNNSNNS NNAACACCGTAGATAGAAACGGT eleven DNA Artificial TCF12 GGTGGTGAATTCCGCAGACAGCGGSNNSNNSN NSNNNSNNSNNSNN SNNSNNSNNSNNSNNAACACCGTAGATAGAAA CGGT 12 DNA Artificial POP2234 AAGAT CAGTTGCGGCCGCT AGACT AGAACCGCTGCCAT GGTGATG GTGATGGTGACCGCCGGTGGTGAATTCCGCAG ACAG 13 DNA Artificial POP2250 CGGCGGTTAGAACGCGGCTACAATTAATAC fourteen DNA Artificial DidLigRev CATGATTACGCCAAGCTCAGAA fifteen DNA Artificial Tcon5new 2 GAGCCGCCGCCACCGGTTTAATGGTGATGGTG ATGGT GACCACCGGTGGTGAATTCCGCAGACAG 16 DNA Artificial Tsopb AAGAAGGAGAACCGGTATGCTGCCGGCGCCG AAAAAC 17 DNA Artificial LS1008 TTTGGGAAGCTTCTAGGTCTCGGCGGTCACCA TCACC ATCACCATGGCAGCGGTTCTAGTCTAGCGGCC CCAAC TGATCTTCACCAAAC eighteen PRT Artificial P53A1R5- 17 no met LPAPKNLWSEVTEDSLRLSWADPHGFYDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT

- 80 039356

19 PRT Artificial P54AR4- 17 no met LPAPKNLWSEVTEDSLRLSWTYDRDGYDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT twenty PRT Artificial P54AR4- 47 no met LPAPKNLWSEVTEDSLRLSWGYNGDHFDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT 21 PRT Artificial P54AR4- 48 no met LPAPKNLWSEVTEDSLRLSWDDPRGFYESFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT 22 PRT Artificial P54AR4- 37 without met LPAPKNLWSEVTEDSLRLSWTWPYADLDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT 23 PRT Artificial 54AR4-74 without met LPAPKNLWSEVTEDSLRLSWGYNGDHFDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT 24 PRT Artificial P54AR4- 81 without met LPAPKNLWSEVTEDSLRLSWDYDLGVYFDSFLI QYQE SEKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIY GVHN VYKDTNMRGLPLSAEFTT 25 PRT Artificial P54AR4- 83 no met LPAPKNLWSEVTEDSLRLSWDDPWAFYESFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT 26 PRT Artificial P54CR4- 31 without met LPAPKNLWSEVTEDSLRLSWTAPDAAFDSFLIQY QESE KVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYGV LGSY VFEHDVMLPLSAEFTT 27 PRT Artificial P54AR483v2 without met LPAPKNLWSEVTEDSARLSWDDPWAFYESFLIQ YQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAIFTT 28 PRT Artificial P54CR4- 31v2 without met LPAPKNLWSEVTEDSARLSWTAPDAAFDSFLIQ YQESE KVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYGV LGSY VFEHDVMLPLSAIFTT 29 PRT Artificial P54AR473v2 without met LPAPKNLWSEVTEDSLRLSWTWPYADLDSFLIQ YQES EKVGEAINLTVPGSERSYDLTGLKPGTEYTVSIYG VHNV YKDTNMRGLPLSAEFTT thirty DNA Artificial TCON6 AAG AAG GAG AAC CGG TAT GCT GCC GGC GCC GAA AAA C 31 DNA Artificial TCON5 E86lshort GAG CCG CCG CCA CCG GTT TAA TGG TGA TGG TGA TGG TGA CCA CCG GTG GTG AAG ATC GCA GAC AG

- 81 039356

32 PRT Artificial P114AR5 Р74А5 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YDEV VVGGEAIVLTVPGSERSYDLTGLKPGTEYYVNILG VKGG SISVPLSAIFTT 33 PRT Artificial P114AR5 R75E9 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFFIR YDEFL RSGEAIVLTVPGSERSYDLTGLKPGTEYWVTILGV KGGL VSTPLSAIFTT 34 PRT Artificial P114AR7 Р92F3 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYIVNIMGV KGGSI SHPLSAIFTT 35 PRT Artificial P114AR7 P9 2-F6 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVNILGV KGGGL SVPLSAIFTT 36 PRT Artificial P114AR7 P9 2-G8 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFVIR YFEFLG SGEAIVLTVPGSERSYDLTGLKPGTEYVVQILGVK GGYISI PLSAIFTT 37 PRT Artificial P114AR7 R9 2-H5 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YLEFLL GGEAIVLTVPGSERSYDLTGLKPGTEYWQIMGV KGGTVS PPLSAIFTT 38 PRT Artificial P114AR7 P9 3-D11 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVGING VKGGYI SYPLSAIFTT 39 PRT Artificial P114AR7 P9 3-G8 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTDLKPGTEYGVTINGV KGGRV STPLSAIFTT 40 PRT Artificial P114AR7 R9 3-H9 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQNGV KGGHIS LPLSAIFTT 41 PRT Artificial P114AR7 R9 4-AZ LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMG VKGGKI SPPLSAIFTT 42 PRT Artificial P114AR7 R9 4-E5 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYAVNIMG VKGGRV SVPLSAIFTT 43 PRT Artificial P114AR7 R9 5-B9 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQILGV KGGSI SVPLSAIFTT

- 82 039356

44 PRT Artificial P114AR7 P9 5-D3 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMG VKGGSI SYPLSAIFTT 45 PRT Artificial P114AR7 P9 5-D4 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQILGV KGGYI SIPLSAIFTT 46 PRT Artificial P114AR7 P9 5-EZ LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQIMG VKGGTV SPPLSAIFTT 47 PRT Artificial P114AR7 P9 5-F10 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFTT AGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMGV KGGSIS PPLSAIFTT 48 PRT Artificial P114AR7 P9 5-G7 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFELLS TGEAIVLTVPGSERSYDLTGLKPGTEYWNIMGVK GGSIS PPLSAIFTT 49 PRT Artificial P114AR7 R9 5-H8 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFV SKGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMG VKGGSI SPPLSAIFTT fifty PRT Artificial ERU1 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YDEVV VGGEAIVLTVPGSERSYDLTGLKPGTEYYVNILGV KGGSIS VPLSAIFTT 51 PRT Artificial ERU2 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSL PAPKNLVVSRVTEDSARLSWTAPDAAFDSFWIRY FEFLG SGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMGV KGGKIS PPLSAIFTT 52 PRT Artificial EEAS MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQNGV KGGHIS LPLSAIFTT

- 83 039356

53 PRT Artificial ERU4 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFFIR YDEFLR SGEAIVLTVPGSERSYDLTGLKPGTEYWVTILGVK GGLVS TPLSAIFTT 54 PRT Artificial ERU5 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVNIMG VKGGKI SPPLSAI FTT 55 PRT Artificial ERU6 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQNGV KGGHIS LPLSAIFTT 56 PRT Artificial ERU7 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTGGGGSGGGGSGGGGSG GGGSM LPAPKNLWSRVTEDSARLSWTAPDAAFDSFWIR YFEFL GSGEAIVLTVPGSERSYDLTGLKPGTEYVVQNGV KGGHIS LPLSAIFTT 57 PRT Artificial ERU15 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTAPAPAPAPAPLPAPKNLVV SRVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNIMGVKGGKISPPLSAIFTT 58 PRT Artificial ERU27 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAIFTTAPAPAPAPAPLPAPKNLVV SRVTED SARLSWTAPDAAFDSFWIRYDEVWGGEAIVLTV PGSER SYDLTGLKPGTEYYVNILGVKGGSISVPLSAIFTT

- 84 039356

59 PRT Artificial ECB60 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAI FTTAPAPAPAPAPMLPAPKNLVVSRVT EDSARLSWTAPDAAFDSFWIRYFEFLGSGEAIVL TVPGSE RSYDLTGLKPGTEYVVNIMGVKGGKISPPLSAIFT T 60 PRT Artificial ERU37 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNMRGLPLSAI FTTAPAPAPAPAPLPAPKNLVVSRVTED SARLSWTAPDAAFDSFWIRYDEVWGGEAIVLTV PGSER SYDLTGLKPGTEYYVNILGVKGGSISVPLSAIFTT 61 PRT Artificial ERU94 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILGVKGGKISPPLSAIFTT 62 PRT Artificial ERU95 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVGSGEAIVLTV PGSER SYDLTGLKPGTEYWNILGVKGGSISPPLSAIFTT 63 PRT Artificial ERU96 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVSKGDAIVLTV PGSERS YDLTGLKPGTEYVVNILGVKGGSISPPLSAIFTT 64 PRT Artificial ERU97 MLPAPKNLVVSEVTEDSARLSWDDPWAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILSVKGGSISPPLSAIFTT 65 PRT Artificial ECB106 MLPAPKNLVVSEVTEDSARLSWDDPHAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILGVKGGKISPPLSAIFTT

- 85 039356

66 PRT Artificial ECB107 MLPAPKNLVVSEVTEDSARLSWDDPHAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVGSGEAIVLTV PGSER SYDLTGLKPGTEYWNILGVKGGSISPPLSAIFTT 67 PRT Artificial ECB108 MLPAPKNLVVSEVTEDSARLSWDDPHAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVSKGDAIVLTV PGSERS YDLTGLKPGTEYVVNILGVKGGSISPPLSAIFTT 68 PRT Artificial ECB109 MLPAPKNLVVSEVTEDSARLSWDDPHAFYESFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILSVKGGSISPPLSAIFTT 69 PRT Artificial ECB118 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFLGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILGVKGGKISPPLSAIFTT 70 PRT Artificial ECB119 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVGSGEAIVLTV PGSER SYDLTGLKPGTEYWNILGVKGGSISPPLSAIFTT 71 PRT Artificial ECB120 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVGEAIVLTVPGSERSYDLTGLKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SARLSWTAPDAAFDSFWIRYFEFVSKGDAIVLTV PGSERS YDLTGLKPGTEYVVNILGVKGGSISPPLSAIFTT 72 PRT Artificial ECB121 MLPAPKNLVVSEVTEDSARLSWADPHGFYDSFLI QYQES EKVG EAIVLTVPGS E RSYDLTG LKPGTEYTVSIYG VHNVY KDTNIRGLPLSAIFTTAPAPAPAPAPLPAPKNLVVS RVTED SAR LSWTAPDAAFDS FWIRYFE F LGSGEAIVLTVP GSERS YDLTGLKPGTEYVVNILSVKGGSISPPLSAIFTT

SEQ ID NO: 73, PRT, Homo Sapiens, EGFR (includes 24 amino acid signal sequence. Mature protein starts at residue 25)

- 86 039356

MRPSGTAGAA LLALLAALCP ASRALEEKKV CQGTSNKLTQ LGTFEDHFLS LQRMFNNCEV

VLGNLEITYV QRNYDLSFLK TIQEVAGYVL IALNTVERIP LENLQIIRGN MYYENSYALA

121 VLSNYDANKT GLKELPMRNL QEILHGAVRF SNNPALCNVE SIQWRDIVSS DFLSNMSMDF

181 QNHLGSCQKC DPSCPNGSCW GAGEENCQKL TKIICAQQCS GRCRGKSPSD CCHNQCAAGC

241 TGPRESDCLV CRKFRDEATC KDTCPPLMLY NPTTYQMDVN PEGKYSFGAT CVKKCPRNYV

301 VTDHGSCVRA CGADSYEMEE DGVRKCKKCE GPCRKVCNGI GIGEFKDSLS INATNIKHFK

361 NCTSISGDLH ILPVAFRGDS FTHTPPLDPQ ELDILKTVKE ITGFLLIQAW PENRTDLHAF

421 ENLEIIRGRT KQHGQFSLAV VSLNITSLGL RSLKEISDGD VIISGNKNLC YANTINWKKL

481 FGTSGQKTKI ISNRGENSCK ATGQVCHALC SPEGCWGPEP RDCVSCRNVS RGRECVDKCN

541 LLEGEPREFV ENSECIQCHP ECLPQAMNIT CTGRGPDNCI QCAHYIDGPH CVKTCPAGVM

601 GENNTLVWKY ADAGHVCHLC HPNCTYGCTG PGLEGCPTNG PKIPSIATGM VGALLLLLW

661 ALGIGLFMRR RHIVRKRTLR RLLQERELVE PLTPSGEAPN QALLRILKET EFKKIKVLGS

721 GAFGTVYKGL WIPEGEKVKI PVAIKELREA TSPKANKEIL DEAYVMASVD NPHVCRLLGI

781 CLTSTVQLIT QLMPFGCLLD YVREHKDNIG SQYLLNWCVQ IAKGMNYLED RRLVHRDLAA

841 RNVLVKTPQH VKITDFGLAK LLGAEEKEYH AEGGKVPIKW MALESILHRI YTHQSDVWSY

901 GVTVWELMTF GSKPYDGIPA SEISSILEKG ERLPQPPICT IDVYMIMVKC WMIDADSRPK

961 FRELIIEFSK MARDPQRYLV IQGDERMHLP SPTDSNFYRA LMDEEDMDDDV VDADEYLIPQ

1021 QGFFSSPSTS RTPLLSSLSA TSNNSTVACI DRNGLQSCPI KEDSFLQRYS SDPTGALTED

1081 SIDDTFLPVP EYINQSVPKR PAGSVQNPVY HNQPLNPAPS RDPHYQDPHS TAVGNPEYLN

1141 TVQPTCVNST FDSPAHWAQK GSHQISLDNP DYQQDFFPKE AKPNGIFKGS TAENAEYLRV

1201 APQSSEFIGA

74 PRT Homo sapiens EGF NSDSECPLSHDGYCLHDGVCMYIEALDKYACNCVVGYIG ERCQYRDLKWWELR

SEQ ID NO: 75, PRT, Homo Sapiens, tenascin-C mgamtqllag vflaflalat eggvlkkvir hkrqsgvnat Ipeenqpwf nhvyruklpv gsqcsvdles asgekdlapp sepsesfqeh tvdgenqivf thnruprra cgcaaapdvk

121 ellsrleele nlvsslreqc tagagcclqp atgrldtrpf csgrgnfste gcgcvcepgw

181 kgpncsepec pgnchlrgrc idgqcicddg ftgedcsqla cpsdcndqgk cvngvcicfe

241 gyagadcsre icpvpcseeh gtcvdglcvc hdgfagddcnkplclnncyn rgrcvenecv

301 cdegftgedc selicpndcf drgrcingtc yceegftged cgkptcphac htqgrceegq

- 87 039356

361 cvcdegfagv dcsekrcpad chnrgrcvdg rcecddgftg adcgelkcpn gcsghgrcvn 421 gqcvcdegyt gedcsqlrcp ndchsrgrcv egkcvceqgf kgydcsdmsc pndchqhgrc 481 vngmcvcddg ytgedcrdrq cprdcsnrgl cvdgqcvced gftgpdcael scpndchgqg 541 rcvngqcvch egfmgkdcke qrcpsdchgq grcvdgqcic hegftgldcg qhscpsdcnn 601 Igqcvsgrci cnegysgedc sevsppkdlv vtevteetvn lawdnemrvt eylwytpth 661 egglemqfrv pgdqtstiiq elepgveyfi rvfailenkk sipvsarvat ylpapeglkf

721 ksiketsvev ewdpldiafe tweufmmn kedegeitks Irrpetsyrq tglapgqeye

781 islhivknnt rgpglkrvtt trldapsqie vkdvtdttal itwfkplaei dgieltygik

841 dvpgdrttid Itedenqysi gnlkpdteye vslisrrgdm ssnpaketft tgldapmlr 901 rvsqtdnsit lewmgkaai dsynkyapi sggdhaevdv pksqqattkt tltglrpgte

961 ygigvsavke dkesnpatm aateldtpkd Iqvsetaets Itllwktpla kfdiyrlnys

1021 Iptgqwvgvq Ipmttsyvl rglepgqeyn vlltaekgrh kskparvkas teqapelenl 1081 tvtevgwdgl rlnwtaadqa yehfiiqvqe ankveaaml tvpgslravd ipglkaatpy 1141 tvsiygviqg yrtpvlsaea stgetpnlge vwaevgwda Iklnwtapeg ayeyffiqvq 1201 eadtveaaqn Itvpgglrst dlpglkaath ytitirgvtq dfsttplsve vlteevpdmg 1261 nltvtevswd alrlnwttpd gtydqftiqv qeadqveeah nltvpgslrs meipglragt 1321 pytvtlhgev rghstrplav ewtedlpql gdlavsevgw dglrlnwtaa dnayehfviq 1381 vqevnkveaa qnltlpgslr avdipgleaa tpyrvsiygv irgyrtpvls aeastakepe 1441 ignlnvsdit pesfnlswma tdgifetfti eudsnrlle tveymsgae rtahisglpp 1501 stdfivylsg lapsirtkti satattealp llenltisdi npygftvswm asenafdsfl

1561 vtwdsgkll dpqeftlsgt qrklelrgli tgigyevmvs gftqghqtkp

1621 pevdnllvsd atpdgfrlsw tadegvfdnf vlkirdtkkq sepleitlla pertrditgl

1681 reateyeiel ygiskgrrsq tvsaiattam gspkevifsd itensatvsw raptaqvesf 1741 ntyvpitgg tpsmvtvdgt ktqtrlvkli pgveylvsn amkgfeesep vsgsfttald 1801 gpsglvtam tdsealarwq paiatvdsyv isytgekvpe itrtvsgntv eyaltdlepa

1861 teytlnfae kgpqksstit akfttdldsp rdltatevqs etalltwrpp rasvtgyllv

1921 yesvdgtvke vivgpdttsy sladlspsth ytakiqalng plrsnnuqti fttigllypf

1981 pkdcsqamln gdttsglyti ylngdkaeal evfcdmtsdg ggwivflrrk ngrenfyqnw 2041 kayaagfgdr reefwlgldn Inkitaqgqy elrvdlrdhg etafavydkf svgdaktryk 2101 Ikvegysgta gdsmayhngr sfstfdkdtd saitncalsy kgafwymch rvnlmgrygd

2161 nnhsqgvnwf hwkghehsiq faemklrpsn fmlegrrkr a

76 PRT Artificial fibcon Ldaptdlqvtnvtdtsitvswtppsatitgyritytpsngpgepkelt vppsstsv titgltpgveywslyalkdnqespplvgtqtt 77 PRT Artificial 10th domain FN3 fibronectin (FN10) VSDVPRDLEVVAATPTSLLISWDAPAVTVRYY RITYGETGGNSPV QEFTVPGSKSTATISGLKPGVDYTITVYAVTG RGDSPASSKPISINY RT 78 PRT Artificial Linker GSGS 79 PRT Artificial Linker GGGGSGGGGSGGGGSGGGGSGGGGS 80 PRT Artificial Linker APAP 81 PRT Artificial Linker APAPAPAPAP 82 PRT Artificial Linker APAPAPAPAPAPAPAPAPAP 83 PRT Artificial Linker APAPAPAPAPAPAPAPAPAPAPAPAPAPAPAP ARARA PAP 84 PRT Artificial Linker AEAAAKEAAAKEAAAKEAAAKEAAAKAAA 85 PRT Artificial Sun loop Tencon TAPDAAFD

- 88 039356

86 PRT Artificial GF loop Tencon KGGHRSN 87 PRT Artificial P53A1R5-17 Sun loop ADPHGFYD 88 PRT Artificial P54AR4-17 Sun loop TYDRDGYD 89 PRT Artificial P54AR4-47 Sun loop WDPFSFYD 90 PRT Artificial P54AR4-48 Sun loop DDPRGFYE 91 PRT Artificial P54AR4-73 Sun loop TWPYADLD 92 PRT Artificial P54AR4-74 Sun loop GYNGDHFD 93 PRT Artificial P54AR4-81 Sun loop DYDLGVYD 94 PRT Artificial P54AR4-83 Sun loop DDPWDFYE 95 PRT Artificial FG hinges EGFR HNVYKDTTNMRGL 96 PRT Artificial FG hinges EGFR LGSYVFEHDVM 97 DNA Artificial > EGFR part of ECB97; P54AR4-83v22 Atgttgccagcgccgaagaacctggtagttagcgaggttactga ggac agcgcgcgtctgagctgggacgatccgtgggcgttctacgaga gctttct gatccagtatcaagagagcgagaaagtcggtgaagcgattgtg ctgac cgtcccgggctccgagcgttcctacgacctgaccggtttgaagc cgggt accgagtatacggtgagcatctacggtgttcacaatgtctataag gaca ctaatatccg cggtctg cctctg ag eg ccattttcaccacc 98 DNA Artificial > EGFR part of ECB15; P54AR4-83V2 Atgctgccagcccctaagaatctggtcgtgagcgaagtaaccg agga cagcgcccgcctgagctgggacgacccgtgggcgttctatgag tctttcc tgattcagtatcaagaaagcgaaaaagttggcgaagcgatcgt cctga ccgtcccg gg tag eg ag eg ctcctacg atetg accgg cctg a a accgg gtacggagtacacggtgtccatttacggtgttcacaatgtgtataa agac acccaacatgcgtggcctgccgctgtcggcgattttcaccacc 99 PRT Artificial Tencon 27 LPAPKNLWSRVTEDSARLSWTAPDAAFDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VKGGHRSNPLSAIFTT 100 PRT Artificial Library TCL14 LPAPKNLWSRVTEDSARLSWTAPDAAFDSF XIXYX EXXXXGEAIVLTVPGSERSYDLTGLKPGTEYX VXIXG VKGGXXSXPLSAIFTT

> SEQ ID NO: 101

PRT

Homo sapiens c-Met

- 89 039356 mkapavlapg ilvllftlvq rsngeckeal aksemnvnmk yqlpnftaet piqnvilheh hiflgatnyi yvlneedlqk vaeyktgpvl ehpdcfpcqd csskanlsgg vwkdninmal

121 wdtyyddql iscgsvnrgt cqrhvfphnh tadiqsevhc ifspqieeps qcpdcwsal

181 gakvlssvkd rfinffvgnt inssyfpdhp Ihsisvrrlk etkdgfmflt dqsyidvlpe

241 frdsypikyv hafesnnfiy fltvqretld aqtfhtmr fcsinsglhs ymemplecil

301 tekrkkrstk kevfrulqaa yvskpgaqla rqigaslndd ilfgvfaqsk pdsaepmdrs

361 amcafpikyv ndffnkivnk nnvrclqhfy gpnhehcfnr tllmssgce arrdeyrtef

421 ttalqrvdlf mgqfsevllt sistfikgdl tianlgtseg rfmqvwsrs gpstphvnfl

481 Idshpvspev ivehtlnqng ytlvitgkki tkiplnglgc rhfqscsqcl sappfvqcgw

541 chdkcvrsee clsgtwtqqi clpaiykvfp nsapleggtr Iticgwdfgf rmnkfdlkk

601 trvllgnesc tltlsestmn tlkctvgpam nkhfnmsiu snghgttqys tfsyvdpvit

661 sispkygpma ggtlltltgn ylnsgnsrhi siggktctlk svsnsilecy tpaqtistef

721 avklkidlan retsifsyre dpivyeihpt ksfistwwke plruvsflfc fasggstitg

781 vgknlnsvsv prmvinvhea gmftvacqh rsnseucct tpslqqlnlq Iplktkaffm

841 Idgilskyfd liyvhnpvfk pfekpvmism gnenvleikg ndidpeavkg evlkvgnksc

901 eruhlhseav Ictvpndllk Inselniewk qaisstvlgk vivqpdqnft gliagwsis

961 talllllgff Iwlkkrkqik dlgselvryd arvhtphldr Ivsarsvspt temvsnesvd

1021 yratfpedqf pnssqngscr qvqypltdms piltsgdsdi sspllqntvh idlsalnpel

1081 vqavqhwig psslivhfne vigrghfgcv yhgtlldndg kkihcavksl nntdigevs

1141 qfltegumk dfshpnvlsl Igiclrsegs plwlpymkh gdlmfime thnptvkdli

1201 gfglqvakgm kylaskkfvh rdlaamcml dekftvkvad fglardmydk eyysvhnktg

1261 aklpvkwmal eslqtqkftt ksdvwsfgvl Iwelmtrgap pypdvntfdi tvyllqgrrl

1321 Iqpeycpdpl yevmlkcwhp kaemrpsfse

1381 apypsllsse dnaddevdtr pasfwets

- 90 039356

102 PRT Homo sapiens HGF QRKRRNTIHEFKKSAKTTLIKIDPALKIK TKKVNTADQCANRCTRNKGLPFTCKAFVFDK ARKQCLWFPFNSMS SGVKKEFFGHEFDLYE N KD Yl RNCIIG KG RSYKGTVSITKSGIKCQPWS SMIPHEHSFLPSSYRG KDLQENYCRNP RGEEGGPWCFTSNPEVRYEVCDIPQCSEVEC MTCNGESYRGLMDH TESGKICQRWDHQTP HRHKFLPERYPDKGFDDNYCRNPDGQPRPW CYTLDPHTRWEYCAIK TCADNTMNDTDVPL ETTECQGQGEGYRGTVNTWNGIPCQRWDSQ YPHEHDMTPENFKC KDLRENYCRNPDGS ESPWCFTTDPNIRVGYCSQIPNCDMSHGQDC YRGNGKNYMGNLSQT RSGLTCSMWDKNME DLHRHIFWEPDASKLNENYCRNPDDAHGPW CYTGNPLIPWDYCPIS RCEGDTTPTIVNL DHPVISCAKTKQLRVVNGIPTRTNIGWMVSLR YRNHICGGSLIKESW VLTARQCFPSRD LKDYEAWLGIHDVHGRGDEKCKQVLNVSQLV YGPEGSDLVLMKLAR PAVLDD FVSTID LP NYGCTIPEKTSCSVYGWGYTGLINYDGLLRVA HLYIMGNEKCSQHHRG KVTLNESEICAG AEKIGSGPCEGDYGGPLVCEQHKMRMVLGVI VPGRGCAIPNRPGIFV RVAYYAKWIHKII LTYKVPQS 103 DNA Artificial > c-MET part ECB97 P114AR7P95C5v2 Ctgccggctccgaagaacttggtggtgagccgtgttaccgaag atagc gcacgcctgagctggacggcaccggatgcggcgttcgatagct tctgg attcgctattttgagtttctgggtagcggtgaggcaattgttctgacg gtgcc gggctctgaacgctcctacgatttgaccggtctgaaaccgggca ccga gtatgtggtgaacattctgagcgttaagggcggtagcatcagccc accg ctgagcgcgatcttcacgactggtggttgc 104 DNA Artificial > c-MET part ECB15 P114AR7P94AZ Ctgccggcaccgaagaacctggttgtcagccgtgtgaccgagg atag eg ca eg tttg ag ctg g a ccg ctccg g atg cag cctttg a cag ett etgga ttcgttactttgaatttctgggtagcggtgaggcgatcgttctgacg gtgeeg ggctctgaacgcagctatgatttgacgggcctgaagccggggtac tgagt acgtggttaacatcatggggcgttaagggtggtaaaatcagcccg ccatt gtccgcgatctttaccacg 105 PRT Artificial Linker GGGGS

- 91 039356

106 PRT Artificial ECB91 mlpapknlwsevtedsarlswddpwafyesfliqyqesekvg eaivltvpgse rsydltglkpgteytvsiygvhnvykdtnirglplsaifttapapapa papLPAP KNLWSRVTEDSARLSWTAPDAAFDSFWIRYF EFLGSGEAIVLTV PGSERSYDLTGLKPGTEYWNILSVKGGSISPP LSAIFTT 107 PRT Artificial P53A1R517v2 Ipapknlwsevtedsarlswadphgfydsfliqyqesekvgeai vltvpgsersy dltglkpgteytvsiygvhnvykdtnmrglplsaiftt 108 PRT Artificial P54AR483v22 Ipapknlwsevtedsarlswddpwafyesfliqyqesekvgea ivltvpgsers ydltglkpgteytvsiygvhnvykdtnirglplsaiftt 109 PRT Artificial P54AR483v23 Ipapknlwsevtedsarlswddphafyesfliqyqesekvgeai vltvpgsersy dltglkpgteytvsiygvhnvykdtnirglplsaiftt 110 PRT Artificial P53A1R517v22 Ipapknlwsevtedsarlswadphgfydsfliqyqesekvgeai vltvpgsersy dltglkpgteytvsiygvhnvykdtnirglplsaiftt 111 PRT Artificial P114AR7P94- A3v22 Ipapknlwsrvtedsarlswtapdaafdsfwiryfeflgsgeaivl tvpgsersyd Itglkpgteywnilgvkggkispplsaiftt 112 PRT Artificial P114AR9P121 -A6v2 LPAPKN LWSRVTE DSARLSWT APDAAF DSFWIRYFEFVGSGEAI VLTVPGSERSYDLTGLKPGTEYWNILGVKGGS ISPPLSAIFTT 113 PRT Artificial P114AR9P122 -A7v2 LPAPKNLVVSRVTEDSARLSWTAPDAAFDSF WIRYFEFVSKGDA IVLTVPGSERSYDLTGLKPGTEYVVNILGVKGG SISPPLSAIFTT 114 PRT Artificial P114AR7P95- C5v2 LPAPKN LWSRVTE DSARLSWT APDAAF DSFWIRYFEFLGSGEAI VLTVPGSERSYDLTGLKPGTEYVVNILSVKGG SISPPLSAIFTT 115 DNA Artificial ECB97 atgttgccagcgccgaagaacctggtagttagcgaggttactga ggac agcgcgcgtctgagctgggacgatccgtgggcgttctacgaga gctttct g atccagtatca agagagcgagaaagtcggtgaagcg attgtg ctgac cgtcccgggctccgagcgttcctacgacctgaccggtttgaagc cgggt accgagtatacggtgagcatctacggtgttcacaatgtctataag gaca ctaatatccgcggtctgcctctgagcgccattttcaccaccgcacc ggc accggctccggctcctgccccgctgccggctccgaagaacttgg tggtg agccgtgttaccgaagatagcgcacgcctgagctggacggca ccgga tg egg cgttcg atag ettetgg atteg ctattttg agtttctg ggtag c ggtga ggcaattgttctgacggtgccgggctctgaacgctcctacgatttg accg gtctgaaaccgggcaccgagtatgtggtgaacattctgagcgtta aggg cggtagcatcagcccaccgctgagcgcgatcttcacgactggtg gttgc

- 92 039356

116 DNA Artificial ERU15 atgctgccagcccctaagaatctggtcgtgagcgaagtaaccg aggac agcgcccgcctgagctgggacgacccgtgggcgttctatgagtc tttcct gattcagtatcaagaaagcgaaaaagttggcgaagcgatcgtc ctgac cgtcccgggtagcgagcgctcctacgatctgaccggcctgaaa ccggg tacggagtacacggtgtccatttacggtgttcacaatgtgtataaa gaca cca a catg eg tg g cctg ccg ctg t eg g eg atttt ca cca ccg eg cctgc gccagcgcctgcaccggctccgctgccggcaccgaagaacct ggttgt cag ccgtgtg accg agg atag eg cacgtttg ag ctg g accg et cegga tgcagcctttgacagcttctggattcgttactttgaatttctgggtagc ggtg aggcgatcgttctgacggtgccgggctctgaacgcagctatgatt tgacg ggcctgaagccgggtactgagtacgtggttaacatcatgggcgtt aagg gtggtaaa atcag cccg ccattgtccg eg atctttaccacg 117 PRT Artificial albumin-binding domain tidewllkeakekaieelkkagitsdyyfdlinkaktvegvnalkde ilka 118 PRT Artificial ERU18 mlpapknlwsevtedsarlswddpwafyesfliqyqesekvge aivltv pgsersydltglkpgteytvsiygvhmvykdtnmrglplsaiftta papapa paplpapknlwsrvtedsarlswtapdaafdsfwirydewvgg eaivlt vpgsersydltglkpgteyyvnilgvkggsisvplsaifttapapap apapl aeakvlanreldkygvsdyyknlinnaktvegvkalldeilaalp 119 PRT Artificial ERU28 mlpapknlwsevtedsarlswadphgfydsfliqyqesekvge aivltv pgsersydltglkpgteytvsiygvhnvykdtnmrglplsaifttap apapa paplpapknlwsrvtedsarlswtapdaafdsfwirydevwg geaivlt vpgsersydltglkpgteyyvnilgvkggsisvplsaifttapapap apapl aeakvlanreldkygvsdyyknlinnaktvegvkalldeilaalp 120 PRT Artificial ECB38 mlpapknlwsevtedsarlswddpwafyesfliqyqesekvg eaivltv pgsersydltglkpgteytvsiygvhnvykdtnmrglplsaifttap apapa paplpapknlwsrvtedsarlswtapdaafdsfwiryfeflgsge aivltv pgsersydltglkpgteywnimgvkggkispplsaifttapapa papapl aeakvlanreldkygvsdyyknlinnaktvegvkalldeilaalp 121 PRT Artificial ERU39 mlpapknlwsevtedsarlswadphgfydsfliqyqesekvge aivltv pgsersydltglkpgteytvsiygvhnvykdtnmrglplsaifttap apapa paplpapknlwsrvtedsarlswtapdaafdsfwiryfeflgsge aivltv pgsersydltglkpgteywnimgvkggkispplsaifttapapa papapl aeakvlanreldkygvsdyyknlinnaktvegvkalldeilaalp

- 93 039356

122 PRT Artificial R53A1 R5-17 wthMet MLPAPKNLVVSEVTEDSLRLSWADPHGFYDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 123 PRT Artificial P54AR4-17 with Met MLPAPKNLVVSEVTEDSLRLSWTYDRDGYDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 124 PRT Artificial P54AR4-47 with Met MLPAPKNLVVSEVTEDSLRLSWGYNGDHFDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 125 PRT Artificial P54AR4-48 with Met MLPAPKNLVVSEVTEDSLRLSWDDPRGYES FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 126 PRT Artificial P54AR4-73 with Met MLPAPKNLVVSEVTEDSLRLSWTWPYADLDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 127 PRT Artificial 54AR4-74 with Met MLPAPKNLVVSEVTEDSLRLSWGYNGDHFDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 128 PRT Artificial P54AR4-81 with Met MLPAPKNLVVSEVTEDSLRLSWDYDLGVYFD SFLIQ YQESEKVGEAINLTVPGSERSYDLTGLKPGTE YTVSI YGVHNVYKDTNMRGLPLSAEFTT 129 PRT Artificial P54AR4-83 with Met MLPAPKNLVVSEVTEDSLRLSWDDPWAFYES FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 130 PRT Artificial P54CR4-31 with Met MLPAPKNLVVSEVTEDSLRLSWTAPDAAFDSF LIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVLGSYVFEHDVMLPLSAEFTT 131 PRT Artificial P54AR4-83v2 with Met MLPAPKNLVVSEVTEDSARLSWDDPWAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNM RGLPLSAIFTT 132 PRT Artificial P54CR4-31v2 with Met MLPAPKNLVVSEVTEDSARLSWTAPDAAFDS FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVLGSYVFEHDVMLPLSAIFTT 133 PRT Artificial P54AR4-73v2 with Met MLPAPKNLVVSEVTEDSLRLSWTWPYADLDS FLIQY QESEKVGEAINLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNMRGLPLSAEFTT 134 PRT Artificial P53A1R517v2 with Met mlpapknlwsevtedsarlswadphgfydsfliqyqesekvge aivltvpgser sydltglkpgteytvsiygvhnvykdtnmrglplsaiftt

- 94 039356

135 PRT Artificial P54AR4- 83v22 with Met mlpapknlwsevtedsarlswddpwafyesfliqyqesekvg eaivltvpgse rsydltglkpgteytvsiygvhnvykdtnirglplsaiftt 136 PRT Artificial P54AR4- 83v23 with Met mlpapknlwsevtedsarlswddphafyesfliqyqesekvge aivltvpgser sydltglkpgteytvsiygvhnvykdtnirglplsaiftt 137 PRT Artificial P53A1R517v22cMet mlpapknlwsevtedsarlswadphgfydsfliqyqesekvge aivltvpgser sydltglkpgteytvsiygvhnvykdtnirglplsaiftt 138 PRT Artificial ECB1 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFWIRYDEWVGGEAIVLTVPGSERSYDLTG LKPG TEYYVNILGVKGGSISVPLSAI FTT 139 PRT Artificial ECB2 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSLPAPKNLVVSRVTEDSARLSWTAPD AAFD SFWIRYFEFLGSGEAIVLTVPGSERSYDLTGLK PGT EYVVNIMGVKGGKISPPLSAIFTT 140 PRT Artificial EER without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFWIRYFEFLGSGEAIVLTVPGSERSYDLTGL KPG TEYWQIIGVKGGHISLPLSAI FTT 141 PRT Artificial ECB4 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFFIRYDEFLRSGEAIVLTVPGSERSYDLTGL KPGT EYWVTILGVKGGLVSTPLSAIFTT

- 95 039356

142 PRT Artificial ECB5 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFWIRYFEFLGSGEAIVLTVPGSERSYDLTGL KPG TEYWNIMGVKGGKISPPLSAIFTT 143 PRT Artificial ECB6 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFWIRYFEFLGSGEAIVLTVPGSERSYDLTGL KPG TEYWQ11GVKGGHISLPSAIFTT 144 PRT Artificial ECB7 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTGGGGSGGGGS GGGG SGGGGSMLPAPKNLWSRVTEDSARLSWTAP DAAF DSFWIRYFEFLGSGEAIVLTVPGSERSYDLTGL KPG TEYWQ 11 GVKGGHISLPLSAI FTT 145 PRT Artificial ECB15 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF LGSG EAIVLTVPGSERSYDLTGLKPGTEYWNIMGV KGGKI SPPLSAIFTT 146 PRT Artificial ECB27 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYDEV WGG EAIVLTVPGSERSYDLTGLKPGTEYYVNILGVK GGSI SVPLSAIFTT

- 96 039356

147 PRT Artificial ECB60 met without LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG LPAPK NLVVSRVTEDSARLSWTAPDAAFDSFWIRYFE FLGS GEAIVLTVPGSERSYDLTGLKPGTEYWNIMG VKGG KISPPLSAIFTT 148 PRT Artificial ECB37 met without LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNMRGLPLSAIFTTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYDEV WGG EAIVLTVPGSERSYDLTGLKPGTEYYVNILGVK GGSI SVPLSAIFTT 149 PRT Artificial ECB94 met without LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAIFTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GKIS PPLSAIFTT 150 PRT Artificial ECB95 met without LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV GSG EAIVLTVPGSERSYDLTGLKPGTEYWNILGVK GGSI SPPLSAIFTT 151 PRT Artificial ECB96 met without LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV SKGD AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GSIS PPLSAIFTT

- 97 039356

152 PRT Artificial ECB97 without met LPAPKNLVVSEVTEDSARLSWDDPWAFYESF LIQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILSVKG GSISP PLSAIFTT 153 PRT Artificial ECB106 without met LPAPKNLVVSEVTEDSARLSWDDPHAFYESFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GKIS PPLSAIFTT 154 PRT Artificial ECB107 without met LPAPKNLVVSEVTEDSARLSWDDPHAFYESFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV GSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGSI SPPLSAIFTT 155 PRT Artificial ECB108 without met LPAPKNLVVSEVTEDSARLSWDDPHAFYESFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV SKGD AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GSIS PPLSAIFTT 156 PRT Artificial ECB109 without met LPAPKNLVVSEVTEDSARLSWDDPHAFYESFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAI FTTAPAPAPAPAP L PAP KNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILSVKG GSISP PLSAIFTT

- 98 039356

157 PRT Artificial ECB118 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPLPLP APKNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GKIS PPLSAIFTT 158 PRT Artificial ECB119 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPLPLP APKNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV GSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGSI SPPLSAIFTT 159 PRT Artificial ECB120 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPLPLP APKNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFV SKGD AIVLTVPGSERSYDLTGLKPGTEYVVNILGVKG GSIS PPLSAIFTT 160 PRT Artificial ECB121 without met LPAPKNLVVSEVTEDSARLSWADPHGFYDSFL IQYQ ESEKVGEAIVLTVPGSERSYDLTGLKPGTEYT VSIYG VHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPLPLP APKNL WSRVTEDSARLSWTAPDAAFDSFWIRYFEFL GSGE AIVLTVPGSERSYDLTGLKPGTEYVVNILSVKG GSISP PLSAIFTT 161 PRT Artificial ECB91 without met Ipapknlwsevtedsarlswddpwafyesfhqyqesekvgeai vltvpgsers ydltglkpgteytvsiygvhnvykdtnirglplsaiftttapapapapap apLPAPK NLVVSRVTEDSARLSWTAPDAAFDSFWIRYFE FLGSGEAIVLTVP GSERSYDLTGLKPGTEYVVNILSVKGGSISPPL SAIFTT 162 PRT Artificial ECB18 without met Ipapknlwsevtedsarlswddpwafyesfhqyqesekvgea ivltvp gsersydltglkpgteytvsiygvhnvykdtnmrglplsaifttapa papap aplpapknlwsrvtedsarlswtapdaafdsfwirydevwgg eaivltv pgsersydltglkpgteyyvmlgvkggsisvplsaifttapapapa papla eakvlanreldkygvsdyyknhnnaktvegvkalldeilaalp

- 99 039356

163 PRT Artificial ECB28 met without Ipapknlwsevtedsarlswadphgfydsfhqyqesekvgeai vltvpg sersydltglkpgteytvsiygvhnvykdtnmrglplsaifttapap apapa plpapknlvvsrvtedsarlswtapdaafdsfwirydevwgge aivltvp gsersydltglkpgteyyvmlgvkggsisvplsaiftttapapapapap aplae akvlanreldkygvsdyyknhnnaktvegvkalldeilaalp 164 PRT Artificial ECB38 met without Ipapknlwsevtedsarlswddpwafyesfhqyqesekvgea ivltvp gsersydltglkpgteytvsiygvhnvykdtnmrglplsaifttapa papap aplpapknlwsrvtedsarlswtapdaafdsfwiryfeflgsgea ivltvp gsersydltglkpgteywmmgvkggkispplsaifttapapap apapla eakvlanreldkygvsdyyknhnnaktvegvkalldeilaalp 165 PRT Artificial ECB39 met without Ipapknlwsevtedsarlswadphgfydsfhqyqesekvgeai vltvpg sersydltglkpgteytvsiygvhnvykdtnmrglplsaifttapap apapa plpapknlwsrvtedsarlswtapdaafdsfwiryfeflgsgeai vltvpg sersydltglkpgteywmmgvkggkispplsaifttapapapa paplae akvlanreldkygvsdyyknhnnaktvegvkalldeilaalp 166 DNA Artificial ECB97 met without ttgccagcgccgaagaacctggtagttagcgaggttactgagga cage gcgcgtctgagctgggacgatccgtgggcgttctacgagagcttt etgat ccagtatcaagagagcgagaaagtcggtgaagcgattgtgctg accgt cccgggctccgagcgttcctacgacctgaccggtttgaagccgg gtacc gagtatacggtgagcatctacggtgttcacaatgtctataaggac actaa tateeg eggtetg cctctg ag eg ccatttt caeca ccg caccgg c accg gctccggctcctgccccgctgccggctccgaagaacttggtggt gagee gtgttaccgaagatagcgcacgcctgagctggacggcaccgg atgeg gcgttcgatagcttctggattcgctattttgagtttctgggtagcggt gaggc aattgttctgacggtgccgggctctgaacgctcctacgatttgacc ggtet gaaaccgggcaccgagtatgtggtgaacattctgagcgttaag ggcggt agcatcagcccaccgctgagcgcgatcttcacgactggtggttg c

- 100 039356

167 DNA Artificial ECB15 without met ctgccagcccctaagaatctggtcgtgagcgaagtaaccgagg acag cgcccgcctgagctgggacgacccgtgggcgttctatgagtcttt cctga ttcagtatcaagaaagcgaaaaagttggcgaagcgatcgtcct gaccg tcccgggtagcgagcgctcctacgatctgaccggcctgaaacc gggta cggagtacacggtgtccatttacggtgttcacaatgtgtataaag acacc aacatgcgtggcctgccgctgtcggcgattttcaccaccgcgcct gcgc cagcgcctgcaccggctccgctgccggcaccgaagaacctgg ttgtca gccgtgtgaccgaggatagcgcacgtttgagctggaccgctcc ggatg cagcctttgacagcttctggattcgttactttgaatttctgggtagcg gtgag gcgatcgttctgacggtgccgggctctgaacgcagctatgatttg acggg cctgaagccgggtactgagtacgtggttaacatcatgggcgttaa gggtg gtaa aatcag cccg ccattgtccg eg atctttacca eg 168 DNA Artificial > EGFR part of ECB97; P54AR4- 83v22 without met ttgccagcgccgaagaacctggtagttagcgaggttactgagga cage gcgcgtctgagctgggacgatccgtgggcgttctacgagagcttt etgat ccagtatcaagagagcgagaaagtcggtgaagcgattgtgctg accgt cccgggctccgagcgttcctacgacctgaccggtttgaagccgg gtacc gagtatacggtgagcatctacggtgttcacaatgtctataaggac actaa tateeg eggtetg cctctg ag eg ccatttt caeca cc 169 DNA Artificial > EGFR part of ECB15; P54AR4-83v2 without met ctgccagcccctaagaatctggtcgtgagcgaagtaaccgagg acag cgcccgcctgagctgggacgacccgtgggcgttctatgagtcttt cctga ttcagtatcaagaaagcgaaaaagttggcgaagcgatcgtcct gaccg tcccgggtagcgagcgctcctacgatctgaccggcctgaaacc gggta cggagtacacggtgtccatttacggtgttcacaatgtgtataaag acacc aacatgcgtggcctgccgctgtcggcgattttcaccacc 170 PRT Artificial ECB94 with C-terminal cysteine mlpapknlwsevtedsAr'lswddpwafyes FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF LGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGKI SPPLSAIFTTC

- 101 039356

171 PRT /!artificial ECB95 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDDPWAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAI FTTAPAPAPAPAPLPAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF VGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGSI SPPLSAIFTTC 172 PRT Artificial ECB96 with C-terminal cysteine MLPAPKNLWSEVTEDSARLSWDDPWAFYESF LIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF VSKG DAIVLTVPGSERSYDLTGLKPGTEYWNILGVK GGSI SPPLSAIFTTC 173 PRT Artificial ECB97 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDDPWAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF LGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILSVK GGSIS PPLSAIFTTC 174 PRT Artificial ECB106 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDPPHAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF LGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGKI SPPLSAIFTTC 175 PRT Artificial ECB107 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDPPHAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF VGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILGVK GGSI SPPLSAIFTTC 176 PRT Artificial ECB108 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDPPHAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF VSKG DAIVLTVPGSERSYDLTGLKPGTEYWNILGVK GGSI SPPLSAIFTTC 177 PRT Artificial ECB109 with C-terminal cysteine MLPAPKNLVVSEVTEDSARLSWDPPHAFYES FLIQY QESEKVGEAIVLTVPGSERSYDLTGLKPGTEY TVSIY GVHNVYKDTNIRGLPLSAIFTTAPAPAPAPAPL PAPKN LVVSRVTEDSARLSWTAPDAAFDSFWIRYFEF LGSG EAIVLTVPGSERSYDLTGLKPGTEYVVNILSVK GGSIS PPLSAIFTTC 178 PRT Artificial ECB91 with C-terminal cysteine mlpapknlwsevtedsarlswddpwafyesfliqyqesekvg eaivltvpgse rsydltglkpgteytvsiygvhnvykdtnirglplsaifttapapapa papLPAP KNLWSRVTEDSARLSWTAPDAAFDSFWIRYF EFLGSGEAIVLTV PGSERSYDLTGLKPGTEYVVNILSVKGGSISP PLSAIFTTC

- 102 039356 > SEQ ID NO: 179

PRT

Artificial

FG loop of the FN3 domain that binds to EGFR HNVYKDTNXgRGL;

wherein Xg is M or I>SEQIDNO: 180

PRT

Artificial

FG loop of EGFR-binding FN3 domain LGSYVFEHDVML (SEQ ID NO: 180) > SEQ ID NO: 181

PRT

Artificial

Loop BC of the FN3 EGFR-binding domain

- 103 039356

X1X2X3X4X5X6X7X8 (SEQ ID NO: 181); and:

Xi is A, T, G or D;

X2 is A, D, Y or W;

Xs is P, D or N;

X ₄ is L or absent;

Xs is D, H, R, G, Y or W;

Xe is G, D or A;

X7 is A, F, G, H or D; and Xs is Y, F or L.

> SEQID NO: 182

PRT

Artificial

EGFR-binding FN3 domain

LPAPKNLVVSEVTEDSLRLSWX1X2X3X4X5X6X7X8DSFLIQYQESEKVGEAINLTVP GSERSYDLTGLKPGTEYTVSIYGVHNVYKDTNXgRGLPLSAEFTT (SEQ ID NO: 182),

Xi is A, T, G or D;

X2 is A, D, Y or W;

Xs is P, D or N;

X ₄ is L or absent;

Xs is D, H, R, G, Y or W;

Xb is G, D or A;

x? is A, F, G, H or D;

Xs is Y, F or L; and Xg is M or I.

> SEQIDNO: 183

PRT

Artificial

EGFR-binding FN3 domain

LPAPKNLVVSEVTEDSLRLSWX1X2X3X4X5X6X7X8DSFLIQYQESEKVGEAINLTVP GSERSYDLTGLKPGTEYTVSIYGVLGSYVFEHDVMLPLSAEFTT (SEQ ID NO: 183), wherein:

Xi is A, T, G or D;

Xr is A, D, Y or W;

Xs is P, D or N;

X ₄ is L or absent;

- 104 039356

Xs is D, H, R G, Y or W;

Xb is G, D or A;

x? is A, F, G, H or D; and Xx is Y, F or L.

>SEQIDNO: 184

PRT

Artificial

Sequence of the C strand and CD loop of the c-Met-binding FN3 domain

DSFXioIRYXnE X12X13X14X15GX16 (SEQ ID NO: 184), wherein:

Hu is W, F or V;

Xn is D, F or L;

X12 is V, F or L;

X13 is V, L or T;

X14 is V, R, G, L, T or S;

X15 is G, S, A, T or K; and

X16 is E or D; and >SEQIDNO: 185

PRT

Artificial

FN3 strand and FG loop sequence of c-Met binding domain FN3

TEYX17VX18IX19X20V KGGX21X22SX23 (SEQ ID NO: 185), wherein:

Xi7 is Y, W, I, V, G or A;

Xis is N, T, Q or G;

Hu is L, M, N, or I;

X20 is G or S;

X21 is S, L, G, Y, T, R, H, or K;

X22 is I, V or L; and

X23 is V, T, Η, I, P, Y, or L.

>SEQIDNO: 186

PRT

Artificial FN3 domain binding to c-Met

LPAPKNLVVSRVTEDSARLSWTAPDAAF DSFXioIRYXnE X12X13X14X15GX16

AIVLTVPGSERSYDLTGLKPGTEYXi7VXi ₈ IXi ₉ X2oVKGGX2iX22SX23PLSAEFTT (SEQ ID NO: 186), wherein:

- 105 039356

Hu is W, F or V; and

Xn is D, F or L;

X12 is V, F or L;

Xn is V, L or T;

X14 is V, R, G, L, T or S;

X15 is G, S, A, T or K;

X16 is E or D;

X17 is Y, W, I, V, G, or A;

Hu is N, T, Q or G;

Hu is L, M, N, or I;

X20 is G or S;

X21 is S, L, G, Y, T, R, H, or K;

X22 is I, V or L; and

X23 is V, T, Η, I, P, Y, or L.

>SEQIDNO: 187

PRT

Artificial

FN3 domain binding to EGFR, an EGFR/c-Met bispecific molecule containing an FN3 domain

LPAPKNLVVSX24VTX25DSX26RLSWDDPX27AFYX28SFLIQYQX29SEKVGEAIX30LTVPGSERSYDL

TGLKPGTEYTVSIYX31VHNVYKDTNX32RGLPLSAX33FTT (SEQ ID NO: 187), wherein:

X24 is E, N or R;

X25 is E or P;

X26 is L or A;

X27 is H or W;

X28 is E or D;

X29 is E or P;

Xzo is N or V;

Xz1 is G or Y;

X32 is M or I; and

Xs is E or I;

- 106 039356

X34 is E, N or R;

X35 is E or P;

Xsb is L or A;

X37 is E or P;

Xs8 is V or L;

Xe is G or S;

X40 is S or K;

X41 is E or D;

X42 is N or V;

X43 is L or M;

X44 is G or S;

X45 is S or K; and

X46 is E or I.

>SEQIDNO: 189

VH mAb to EGFR El

QVQLVESGGGVVQPGRSLRLSCAASGFTFSTYGMHWVRQAPGKGLEWVAVIWDDGSYKYYGD SVKGRFTISRDNSKNTLY

LQMNSLRAEDTAVYYCARDGITMVRGVMKDYFDYWGQGTLVTVSS > SEQ ID NO: 190

VL mAb to EGFR El

AIQLTQSPSSLSASVGDRVTITCRASQDISSALVWYQQKPGKAPKLLIYDASSLESGVPSRFSGSES GTDFTLTISSLQP

EDFATYYCQQFNSYPLTFGGGTKVEIK > SEQ ID NO: 191

VHmAb _K EGFRE2

EVQLVESGGG LVQPGGSLRL SCAASGFTFS SYWMNWVRQA PGKGLEWVAN IKKDGSEKYY

VDSVKGRFTI SRDNAKNSLY LQMNSLRAED TAVYYCARDL GWGWGWYFDL WGRGTLVTVS

121 S > SEQ ID NO: 192

VLmAb _K EGFRE2

EIVLTQSPAT LSLSPGERAT LSCRASQSVS SYLAWYQQKP GQAPRLLIYD ASNRATGIPA 61 RFSGSGSGTD FTLTISSLEP EDFAVYYCQQ RSNWPPTFGQ GTKVEIK > SEQ ID NO: 193

VH mAb to c-Met Ml

- 107 039356

SEQ ID NO: 194

VL mAb to c-Met Ml

SEQ ID NO: 195

VH mAb to c-Met M2

SEQ ID NO: 196 c-Met mAb M2 VL

QLTQSPSSLSASVGDRVTITCRASQGLSSALAWYRQKPGKAPKLLIYDASSLESGVPSRFSGSGSG TDFTLTISSLQPEDFATYYCQQFTSYPQITFGQGTRLEIK > SEQ ID NO: 197

Gpl20 тяжелая цепь c F405L qvqlvqsgaevkkpgasvkvscqasgyrfsnfvihwvrqapgqrfewmgwinpyngnkefsakfqdrvtftadtsantaymelrslrsadtavy ycarvgpyswddspqdnyymdvwgkgttvivssastkgpsvfplapsskstsggtaalgclvkdyfpepvtvswnsgaltsgvhtfpavlqssgl yslsswtvpssslgtqtyicnvnhkpsntkvdkrvepkscdkthtcppcpapellggpsvflfppkpkdtlmisrtpevtcvvvdvshedpevkfh wyvdgvevhnaktkpreeqynstyrvvsvltvlhqdwlngkeykckvsnkalpapiektiskakgqprepqvytlppsreemtknqvsltclvk gfypsdiavewesngqpennykttppvldsdgsfllyskltvdksrwqqgnvfscsvmhealhnhytqkslslspgk > SEQ ID NO: 198

Gpl20 тяжелая цепь c K409R qvqlvqsgaevkkpgasvkvscqasgyrfsnfvihwvrqapgqrfewmgwinpyngnkefsakfqdrvtftadtsantaymelrslrsadtavy ycarvgpyswddspqdnyymdvwgkgttvivssastkgpsvfplapsskstsggtaalgclvkdyfpepvtvswnsgaltsgvhtfpavlqssgl yslsswtvpssslgtqtyicnvnhkpsntkvdkrvepkscdkthtcppcpapellggpsvflfppkpkdtlmisrtpevtcvvvdvshedpevkfh wyvdgvevhnaktkpreeqynstyrvvsvltvlhqdwlngkeykckvsnkalpapiektiskakgqprepqvytlppsreemtknqvsltclvk gfypsdiavewesngqpennykttppvldsdgsfflysrltvdksrwqqgnvfscsvmhealhnhytqkslslspgk > SEQ ID NO: 199

Hl mAb EMI (antibody to EGFR, 405L)

QVQLVESGGGVVQPGRSLRLSCAASGFTFSTYGMHWVRQAPGKGLEWVAVIWDDGSYKYYGD SVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDGITMVRGVMKDYFDYWGQGTLVTVS

- 108 039356

SASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYS LSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKP KDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLH QDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYP SDIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYT QKSLSLSPGK > SEQ ID NO: 200

LI mAb EM-1

AIQLTQSPSSLSASVGDRVTITCRASQDISSALVWYQQKPGKAPKLLIYDASSLESGVPSRFSGSES GTDFTLTISSLQPEDFATYYCQQFNSYPLTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVC LLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKGLSSYACEVT HQ1

H2 mAb EM-1 (K409R, c-Met antibody)

QVQLVQSGAEVKKPGASVKVSCETSGYTFTSYGISWVRQAPGHGLEWMGWISAYNGYTNYAQ KLQGRVTMTTDTSTSTAYMELRSLRSDDTAVYYCARDLRGTNYFDYWGQGTL VTVSS ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSL SSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPK DTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQ DWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQQGNVFSCSVMHEALHNHYTQ KSLSLSPGK > SEQ ID NO: 202

L2mAbEM-l

DIQMTQSPSSVSASVGDRVTITCRASQGISNWLAWFQHKPGKAPKLLIYAASSLLSGVPSRFSGSG SGTDFTLTISSLQPEDFATYYCQQANSFPITFGQGTRLEIK

RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS TYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC > SEQ ID NO: 203

Hl, constant region

ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSL SSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPK DTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQ

- 109 039356

SEQ ID NO: 204

H2, constant region

ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSL SSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPK DTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQ DWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQ KSLSLSPGK > SEQ ID NO: 205

EMl-mAb Hl, кДНК pdr000015499 caggtgcagctggtcgagagcggcggaggggtggtgcagcccggcagaagcctgaggctgtcctgcgccgccagcggcttcaccttcagcacctac ggcatgcactgggtgcggcaggccccaggcaagggcctggagtgggtggccgtgatctgggacgacggcagctacaagtactacggcgacagcgt gaagggcaggttcaccatcagcagggacaacagcaagaacaccctgtacctgcagatgaacagcctgagggccgaggacaccgccgtgtactactg cgccagggacggcatcaccatggtgcggggcgtgatgaaggactacttcgactactggggccagggcaccctggtgaccgtgagcagcgccagcac caagggcccaagcgtgttccccctggcccccagcagcaagagcaccagcggcggcacagccgccctgggctgcctggtgaaggactacttccccga gccagtgaccgtgtcctggaactctggcgccctgacctccggcgtgcacaccttccccgccgtgctgcagagcagcggcctgtacagcctgagcagcg tggtgaccgtgcccagcagcagcctgggcacccagacctacatctgcaacgtgaaccacaagcccagcaacaccaaggtggacaagagagtggagc ccaagagctgcgacaagacccacacctgccccccctgcccagccccagagctgctgggcggacccagcgtgttcctgttcccccccaagcccaagga caccctgatgatcagcaggacccccgaggtgacctgcgtggtggtggacgtgagccacgaggacccagaggtgaagttcaactggtacgtggacggc gtggaggtgcacaacgccaagaccaagcccagagaggagcagtacaacagcacctacagggtggtgtccgtgctgaccgtgctgcaccaggact gg ctgaacggcaaggaatacaagtgcaaggtctccaacaaggccctgccagcccccatcgaaaagaccatcagcaaggccaagggccagccacggga gccccaggtgtacaccctgccccccagccgggaggagatgaccaagaaccaggtgtccctgacctgtctggtgaagggcttctaccccagcgacatc gccgtggagtgggagagcaacggccagcccgagaacaactacaagaccacccccccagtgctggacagcgacggcagcttcctcctgtacagcaag ctgaccgtggacaagtccaggtggcagcagggcaacgtgttcagctgcagcgtgatgcacgaggccctgcacaaccactacacccagaagtccctga gcctgagccccggcaaatga > SEQ ID NO: 206

EMl-mAb LI, кДНК pDROOOO 15499 atccagctgacccagagccccagcagcctgagcgccagcgtgggcgaccgggtgaccatcacctgccgggccagccaggacatcagcagcgccct ggtctggtatcagcagaagcccggcaaggcccccaagctgctgatctacgacgccagctccctggaaagcggcgtgcccagccggttcagcggcag cgagagcggcaccgacttcaccctgaccatcagcagcctgcagcccgaggacttcgccacctactactgccagcagttcaacagctaccccctgacctt tggcggcggaacaaaggtggagatcaagcgtacggtggccgctcccagcgtgttcatcttcccccccagcgacgagcagctgaagagcggcaccgc cagcgtggtgtgcctgctgaacaacttctacccccgggaggccaaggtgcagtggaaggtggacaacgccctgcagagcggcaacagccaggagag

- 110 039356 cgtcaccgagcaggacagcaaggactccacctacagcctgagcagcaccctgaccctgtccaaggccgactacgagaagcacaaggtgtacgcctg cgaggtgacccaccagggcctgtccagccccgtgaccaagagcttcaacaggggcgagtgctga > SEQ ID NO: 207

EMI-mAb H2, кДНК pDR000016584 caggttcagctggtgcagtctggagctgaggtgaagaagcctggggcctcagtgaaggtctcctgcgagacttctggttacacctttaccagctatggtat cagctgggtgcgacaggcccctggacacgggcttgagtggatgggatggatcagcgcttacaatggttacacaaactatgcacagaagctccagggca gggtcaccatgaccacagacacatccacgagcacagcctacatggagctgaggagcctgagatctgacgacacggccgtgtattactgtgcgagagat ctgagaggaactaactactttgactactggggccagggaaccctggtcaccgtctcctcagcctccaccaagggcccaagcgtgttccctctggccccc agcagcaagagcacatctggcggaacagccgccctgggctgcctggtgaaagactacttccccgagcccgtgaccgtgtcctggaactctggcgccct gaccagcggcgtgcacacctttccagccgtgctgcagagcagcggcctgtacagcctgtccagcgtggtgaccgtgcccagcagctccctgggcacc cagacctacatctgcaacgtgaaccacaagcccagcaacaccaaggtggacaagcgggtggaacccaagagctgcgacaagacccacacctgtccc ccctgccctgcccctgaactgctgggcggaccctccgtgttcctgttccccccaaagcccaaggacaccctgatgatcagccggacccccgaagtgac ctgcgtggtggtggacgtgtcccacgaggaccctgaagtgaagttcaattggtacgtggacggcgtggaagtgcacaacgccaagaccaagcccaga gaggaacagtacaacagcacctaccgggtggtgtccgtgctgacagtgctgcaccaggactggctgaacggcaaagagt acaagtgcaaggtctcca acaaggccctgcctgctcccatcgagaaaaccatcagcaaggccaagggccagccccgcgagcctcaggtgtacacactgcctcccagccgggaag agatgaccaagaaccaggtgtccctgacctgtctggtgaaaggcttctaccccagcgatatcgccgtggaatgggagagcaacggacagcccgagaa caactacaagaccaccccccctgtgctggacagcgacggctccttcttcctgtactctcggctgaccgtggacaagagccggtggcagcagggaaacg tgttcagctgcagcgtgatgcacgaggccctgcacaaccactacacccagaagtccctgagcctgagccccgggaagtga > SEQ ID NO: 208

EMI-mAb L2, кДНК pDR000016584 gacatccagatgacccagtccccctcctccgtgtccgcctctgtgggcgacagagtgaccatcacctgtcgggcctcccagggcatctccaactggctg gcctggttccagcacaagcccggcaaggcccccaagctgctgatctacgccgcctcctccctgctgtccggcgtgccctccagattctccggctctggct ccggcaccgacttcaccctgaccatctccagcctgcagcccgaggacttcgccacctactactgccagcaggccaactccttccccatcaccttcggcca gggcacccggctggaaatcaagcgtacggtggccgctcccagcgtgttcatcttcccccccagcgacgagcagctgaagagcggcaccgccagcgt ggtgtgcctgctgaacaacttctacccccgggaggccaaggtgcagtggaaggtggacaacgccctgcagagcggcaacagccaggagagcgtcac cgagcaggacagcaaggactccacctacagcctgagcagcaccctgaccctgtccaaggccgactacgagaagcacaaggtgtacgcctgcgaggt gacccaccagggcctgtccagccccgtgaccaagagcttcaacaggggcgagtgctga > SEQ ID NO: 209

Gpl20 light chain

Eivltqspgtlslspgeratfscrsshsirsrrvawyqhkpgqaprlvihgvsnrasgisdrfsgsgsgtdftltitrvepedfalyycqvygassytfgqg tklerkrtvaapsvfifppsdeqlksgtasvvcllnnfypreakvqwkvyvdnalqsgnsqesvteqdskdstyslsskfltlskadyvqt

- 111 039356 > SEQ ID NO 210

El HC1 HCDR1

TYGMH > SEQ ID NO 211

El HC1 HCDR2

VIWDDGSYKYYGDSVKG > SEQ ID NO 212

El HC1 HCDR3

DGITMVRGVMKDYFDY > SEQ ID NO 213

El LC1 LCDR1

RASQDISSALV > SEQ ID NO 214

El LC1 LCDR2

DASSLES > SEQ ID NO 215

El LC1 LCDR3

QQFNSYPLT > SEQ ID NO 216

El HC2 HCDR1

SYGIS > SEQ ID NO 217

El HC2 HCDR2

WISAYNGYTNYAQKLQG > SEQ ID NO 218

E1HC2 HCDR3

DLRGTNYFDY > SEQ ID NO 219

El LC2 LCDR1

- 112 039356

RASQGISNWLA > SEQ ID NO: 220 El LC2 LCDR2

AASSLLS > SEQ ID NO: 221 El LC2 LCDR3

QQANSFPIT > SEQ ID NO: 222 E2-mAb HCl, HCDR1 SYWMN > SEQ ID NO: 223 E2-mAb HCl, HCDR2

NIKKDGSEKYYVDSVKG > SEQ ID NO: 224 E2-mAb HCl, HCDR3 DLGWGWGWYFDL > SEQ ID NO: 225 E2-mAB LC1 LCDR1

RASQSVSSYLA > SEQ ID NO: 226 E2-mAb LC1, LCDR2 DASNRAT > SEQ ID NO: 227 E2-mAb LC1, LCDR3 QQRSNWPPT > SEQ ID NO: 228 E2-mAb HC2, HCDR1

DYYMY

- 113 039356 > SEQ ID NO: 229

E2-mAb HC2, HCDR2

TISDDGSYTYYPDSVKG > SEQ ID NO: 230

E2-mAb HC2, HCDR3

EGLYYYGSGSYYNQDY > SEQ ID NO: 231

E2-mAb LC2, LCDR1

RASQGLSSALA > SEQ ID NO: 232

E2-mAb LC2, LCDR2

DASSLES > SEQ ID NO: 233

E2-mAb LC2, LCDR3

QQFTSYPQIT > SEQ ID NO: 234

E2-mAb, HC1 (EGFR-F405L)

EVQLVESGGG LVQPGGSLRL SCAASGFTFS SYWMNWVRQA PGKGLEWVAN IKKDGSEKYY VDSVKGRFTI SRDNAKNSLY LQMNSLRAED TAVYYCARDL GWGWGWYFDL WGRGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFP AVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLG GPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTY RVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVS LTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFLLYSKLTVDKSRWQQGNVFSCSVM HEALHNHYTQKSLSLSPGK > SEQ ID NO: 235

E2-mAb LC1, (EGFR)

EIVLTQSPAT LSLSPGERAT LSCRASQSVS SYLAWYQQKP GQAPRLLIYD ASNRATGIPA RFSGSGSGTD FTLTISSLEP EDFAVYYCQQ RSNWPPTFGQ GTKVEIK RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS TYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC > SEQ ID NO: 236

E2-mAb HC2 (c-Met-K409R)

EVQLVESGGGLVKPGGSLKLSCAASGFTFSDYYMYWVRQTPEKRLEWVATISDDGSYTYYPDS

VKGRFTISRDNAKNNLYLQMSSLKSEDTAMYYCAREGLYYYGSGSYYNQDYWGQGTLVTVSS

ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSL

- 114 039356

SSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPK DTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQ DWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQQGNVFSCSVMHEALHNHYTQ KSLSLSPGK > SEQ ID NO: 237

E2-mAb LC2 (c-Met)

QLTQSPSSLSASVGDRVTITCRASQGLSSALAWYRQKPGKAPKLLIYDASSLESGVPSRFSGSGSG TDFTLTISSLQPEDFAIYYCQQFTSYPQITFGQGTRLEIK

RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDS lYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC > SEQ ID NO: 238 contiguous c-Met epitope in mAb 069

PEFRDSYPIKYVHAF > SEQ ID NO: 239 contiguous c-Met epitope in mAb 069

FAQSKPDSAEPMDRSA > SEQ ID NO: 240 epitope mAb 5D5

PGAQLARQIGASLNDD

Claims (50)

1. Isolated bispecific antibody to epidermal growth factor receptor (EGFR)/hepatocyte growth factor receptor (c-Met), containing: a) a first heavy chain (HC1) containing HC1 constant domain 3 (HC1 CH3) and HC1 variable region 1 (VH1) containing the complementarity-determining regions of the heavy chain HCDR1, HCDR2 and HCDR3 with the amino acid sequences of SEQ ID NOS: 210, 211 and 212 respectively; b) a second heavy chain (HC2) containing HC2 constant domain 3 (HC2 CH3) and HC2 variable region 2 (VH2) containing HCDR1, HCDR2 and HCDR3 with the amino acid sequences of SEQ ID NOs: 216, 217 and 218, respectively; c) a first light chain (LC1) comprising a light chain variable region 1 (VL1) comprising the complementarity determining light chain regions LCDR1, LCDR2 and LCDR3 with the amino acid sequences of SEQ ID NOs: 213, 214 and 215, respectively; and d) a second light chain (LC2) containing a light chain variable region 2 (VL2) containing LCDR1, LCDR2 and LCDR3 with the amino acid sequences of SEQ ID NOs: 219, 220 and 221, respectively, where the pair of VH1 and VL1 forms a first antigen-binding site that binds specifically to EGFR, the VH2 and VL2 pair form a second antigen-binding site that specifically binds to c-Met; where the antibody is IgG ₁ and has a F405L substitution in the EGFR binding arm and a K409R substitution in the c-Met binding arm in the CH3 constant domain to enhance heterodimerization to form a bispecific antibody. 2. The bispecific antibody of claim 1, wherein the antibody inhibits the phosphorylation of extracellular signal-associated kinases 1 and 2 (ERK1/2) in the NCI-H292, NCI-H1975, or SKMES-1 cell line with an IC ₅₀ value that is at least approximately 10 times less, at least about 20 times less, at least about 30 times less, at least about 40 times less, at least about 50 times less, or at least about 60 times less compared to IC50 value for inhibition of ERK1/2 phosphorylation in NCI- _H292 , NCI-H1975, or SKMES-1 cell lines with a control monovalent anti-EGFR antibody containing heavy chain 3 (HC3) and light chain 3 (LC3) , and a control monovalent antibody to c-Met containing heavy chain 4 (HC4) and light chain 4 (LC4); HC3 and HC1, LC3 and LC1, HC4 and HC2, and LC4 and LC2 have identical amino acid sequences, respectively; ERK1/2 phosphorylation is measured in whole cell lysates by sandwich ELISA using an anti-phospho-ERK1/2 antibody as capture antibody and an electrochemiluminescent compound-conjugated antibody that binds to unphosphorylated and phosphorylated ERK1/2 as detection antibody. 3. The bispecific antibody of claim 1 or 2, wherein the antibody inhibits ERK1/2 phosphorylation with an IC ₅₀ value of about 2x10' ⁹ M or less, about 1x10' ⁹ M or less, or about 1x10'1° M or less. 4. A bispecific antibody according to claim 2 or 3, wherein ERK1 is phosphorylated at Thr202 and Tyr204 and ERK2 is phosphorylated at Thr185 and Tyr197. 5. The bispecific antibody of any one of claims 1 to 4, wherein the antibody inhibits protein kinase B (AKT) phosphorylation at Ser473 in the NCI-H1975 cell line with an IC ₅₀ value of at least about 70 times less than the IC ₅₀ value for inhibition of AKT phosphorylation at Ser473 in the NCI-H1975 cell line with a mixture of a control monovalent anti-EGFR antibody containing HC3 and LC3 and a control monovalent anti-c-Met antibody containing HC4 and LC4; HC3 and HC1, LC3 and LC1, HC4 and HC2, and LC4 and LC2 have identical amino acid sequences, respectively; AKT phosphorylation at Ser473 was measured in whole cell lysates by sandwich ELISA using an antibody that binds to unphosphorylated and phosphorylated AKT as capture antibody and an electrochemiluminescent compound-conjugated antibody to Ser473 phosphorylated AKT as detection antibody. 6. The bispecific antibody of any one of claims 1-5, wherein the antibody inhibits AKT phosphorylation at Thr308 in the NCI-H1975 cell line with an IC50 value of at least about 100-fold lower than the IC50 value for inhibiting AKT phosphorylation at Thr308 at Thr308. cell line NCI-H1975 with a mixture of a control monovalent antibody to EGFR containing HC3 and LC3, and a control monovalent antibody to c-Met containing HC4 and LC4; HC3 and HC1, LC3 and LC1, HC4 and HC2, and LC4 and LC2 have identical amino acid sequences, respectively; AKT phosphorylation at Thr308 is measured in whole cell lysates by sandwich ELISA using an antibody that binds to unphosphorylated and phosphorylated AKT as capture antibody and an electrochemiluminescent compound-conjugated antibody to Thr308 phosphorylated AKT as detection antibody. 7. A bispecific antibody according to claim 5 or 6, wherein the antibody inhibits AKT phosphorylation at Ser473 or Thr308 with an IC ₅₀ value of approximately 1x10' ⁹ M or less. 8. The bispecific antibody according to any one of claims 1 to 7, wherein the bispecific antibody binds to EGFR having the amino acid sequence of SEQ ID NO: 73 at EGFR residues K489, I491, K467 and S492 and c-Met at PEFRDSYPIKYVHAF residues (SEQ ID NO: 238) and FAQSKPDSAEPMDRSA (SEQ ID NO: 239). 9. A bispecific antibody according to any one of claims 1 to 8, wherein the antibody inhibits the growth of NCI-H292 or NCI-H1975 cells with an IC ₅₀ value of at least about 300 times less, at least about 400 times less, at least at least about 500 times less, at least about 600 times less, at least about 700 times less, or at least about 800 times less than the IC50 value for cell growth inhibition NCI-H292 or NCI-H1975 cetuximab if NCI-H292 or NCI-H1975 cells are grown under low attachment conditions. 10. A bispecific antibody according to any one of claims 1 to 9, wherein the antibody inhibits tumor growth from HGF-expressing SKMES-1 cells in SCID Beige mice with a percentage (%) T/C value at day 36 at least 500 times less than compared with cetuximab when administered with a bispecific antibody and cetuximab at a dose of 20 mg/kg. 11. A bispecific antibody according to any one of claims 1-10, wherein the antibody neutralizes EGFR and c-Met signaling. 12. The bispecific antibody of claim 1, wherein VH1, VL1, VH2, and VL2 comprise the amino acid sequences of SEQ ID NOs: 189, 190, 193, and 194, respectively. 13. A bispecific antibody according to claim 12 containing 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 conservative amino acid substitutions in HC1, LC1, HC2, or LC2 . 14. Bispecific antibody according to claim 13, containing the substitution M252Y/S254T/T256E in HC1 and/or in - 116 039356 HC2, where the numbering of residues corresponds to the EU index. 15. A bispecific antibody according to any one of claims 1-14, wherein the antibody has a 2-antennary glycan structure with a fucose content ranging from about 1% to about 15%. 16. The bispecific antibody of claim 2, wherein HC1, LC1, HC2, and LC2 are encoded by synthetic polynucleotides containing the sequences of SEQ ID NOs: 205, 206, 207, and 208, respectively. 17. An isolated synthetic polynucleotide encoding HC1, HC2, LC1 or LC2 according to claim 16. 18. The isolated polynucleotide according to claim 17, containing the polynucleotide sequences of SEQ ID NO: 205, 206, 207 or 208. 19. A vector containing a polynucleotide according to claim 17 or 18. 20. An isolated host cell containing the vector of claim 19. 21. A method for producing an isolated EGFR/c-Met bispecific antibody according to claim 12, comprising: a) combining an isolated anti-EGFR monospecific bivalent antibody containing two heavy chains of SEQ ID NO: 199 and two light chains of SEQ ID NO: 200 and an isolated monospecific bivalent anti-c-Met antibody containing two heavy chains of SEQ ID NO: 201 and two light chains of SEQ ID NO: 202, mixed in a molar ratio of approximately 1:1; b) adding a reducing agent to the mixture; c) incubating the mixture for about 90 minutes to about 6 hours; d) removal of the reducing agent; e) purification of an anti-EGFR/c-Met bispecific antibody comprising a first heavy chain containing SEQ ID NOs: 210, 211 and 212 and a second heavy chain containing SEQ ID NOs: 216, 217 and 218, a first light chain containing SEQ ID NOs: 213, 214 and 215, and a second light chain comprising SEQ ID NOs: 219, 220 and 221, wherein the first heavy chain and first light chain pair form a first binding domain that specifically binds to EGFR and a second heavy chain pair chain and a second light chain forms a second binding domain that specifically binds to c-Met. 22. The method of claim 21 wherein the reducing agent is 2-mercaptoethanolamine (2-MEA). 23. The method of claim 22 wherein 2-MEA is present at a concentration of about 25 to about 75 mM. 24. The method of claim 23, wherein the incubation step is carried out at a temperature of from about 25°C to about 37°C. 25. A pharmaceutical composition comprising a bispecific antibody according to any one of claims 1 to 16 and a pharmaceutically acceptable carrier. 26. A method of treating a cancer patient, comprising administering a therapeutically effective amount of an EGFR/c-Met bispecific antibody according to claim 2 to a patient for a period of time sufficient to treat cancer, wherein the cancer is epithelial cell cancer, breast cancer, cancer ovarian cancer, lung cancer, non-small cell lung cancer (NSCLC), pulmonary adenocarcinoma, small cell lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer, pharyngeal cancer, nose cancer, pancreatic cancer , skin cancer, oral cancer, tongue cancer, esophageal cancer, vaginal cancer, cervical cancer, spleen cancer, testicular cancer, stomach cancer, thymus cancer, colon cancer, thyroid cancer, liver cancer, hepatocellular carcinoma (HCC), or sporadic or hereditary papillary renal carcinoma (PRCC). 27. The method of claim 26, wherein the cancer is associated with an EGFR activating mutation, EGFR gene amplification, elevated circulating HGF, c-Met activating mutation, c-Met gene amplification, or KRAS mutation. 28. The method of claim 27 wherein the EGFR activating mutation is G719A, G719X (X is any amino acid), L861X (X is any amino acid), L858R, E746K, L747S, E749Q, A750P, A755V, V765M, L858P or T790M, E746-A750 deletion, R748-P753 deletion, Ala (A) insertion between M766 and A767, Ser, Val and Ala (SVA) insertion between S768 and V769, and Asn and Ser (Ns) insertion between P772 and H773. 29. The method of claim 28 wherein the EGFR activating mutation is a substitution for L858R, del (E476, A750) and/or T790M. 30. The method of claim 27, wherein the KRAS mutation has a G12V or G12C substitution. 31. The method of claim 30, wherein the KRAS mutation has a G12V substitution. 32. The method of claim 27, wherein the patient is or has become resistant to treatment with erlotinib, gefitinib, afatinib, CO-1686, AZD9192, or cetuximab. 33. The method of claim 26 wherein the subject is homozygous for phenylalanine at position 158 in CD16 or heterozygous for valine and phenylalanine at position 158 in CD16. 34. The method of claim 26, comprising administering a second therapeutic agent. 35. The method of claim 34 wherein the second therapeutic agent is a chemotherapeutic agent or a targeted cancer therapy agent. - 117 039356 36. The method of claim 35 wherein the chemotherapeutic agent is cisplatin or vinblastine. 37. The method of claim 35, wherein the chemotherapeutic agent or targeted cancer therapy agent is an EGFR, c-Met, HER2, HER3, HER4, or VEGFR tyrosine kinase inhibitor. 38. The method of claim 37, wherein the tyrosine kinase inhibitor is erlotinib, gefitinib, or afatinib. 39. The method of claim 34 wherein the second therapeutic agent is administered simultaneously or separately. 40. The method of claim 39, wherein the second therapeutic agent is administered sequentially. 41. A method for inhibiting the growth or proliferation of cells expressing EGFR and/or c-Met, which includes bringing the cells into contact with a bispecific antibody according to claims 2, 12. 42. A method for inhibiting the growth or metastasis of an EGFR- and/or c-Met-expressing tumor or cancer cells in a patient, comprising administering to the patient an effective amount of a bispecific antibody according to claims 2, 12. 43. The method of claim 42, wherein the tumor expressing EGFR and/or c-Met is epithelial cell cancer, breast cancer, ovarian cancer, lung cancer, non-small cell lung cancer (NSCLC), pulmonary adenocarcinoma, small cell lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer, pharyngeal cancer, nose cancer, pancreatic cancer, skin cancer, oral cancer, tongue cancer, esophageal cancer, vaginal cancer, cervical cancer uterus, spleen cancer, testicular cancer, gastric cancer, thymus cancer, colon cancer, thyroid cancer, liver cancer, hepatocellular carcinoma (HCC) or sporadic or hereditary papillary renal carcinoma (PRCC). 44. The method of claim 43, wherein the EGFR and/or c-Met expressing tumor is associated with an EGFR activating mutation, EGFR gene amplification, elevated circulating HGF, c-Met activating mutation, c-Met gene amplification, or KRAS mutation. 45. The method of claim 44 wherein the EGFR activating mutation is G719A, G719X (X is any amino acid), L861X (X is any amino acid), L858R, E746K, L747S, E749Q, A750P, A755V, V765M, L858P or T790M, E746-A750 deletion, R748-P753 deletion, Ala (A) insertion between M766 and A767, Ser, Val and Ala (SVA) insertion between S768 and V769, and Asn and Ser (NS) insertion between P772 and H773. 46. The method of claim 45 wherein the EGFR activating mutation is a replacement for L858R, del (E476, A750) and/or T790M. 47. The method of claim 44, wherein the KRAS mutation has a G12V or G12C substitution. 48. The method of claim 47, wherein the KRAS mutation has a G12V substitution. 49. The use of a bispecific antibody according to any one of claims 1-16 for therapy. 50. The use of a bispecific antibody according to any one of claims 1 to 16 for the treatment of cancer, wherein the cancer is epithelial cell cancer, breast cancer, ovarian cancer, lung cancer, non-small cell lung cancer (NSCLC), pulmonary adenocarcinoma, small cell lung cancer, colorectal cancer, anal cancer, prostate cancer, kidney cancer, bladder cancer, head and neck cancer, pharyngeal cancer, nose cancer, pancreatic cancer, skin cancer, oral cavity cancer, tongue cancer, esophageal cancer, vaginal cancer, cervical cancer , spleen cancer, testicular cancer, gastric cancer, thymus cancer, colon cancer, thyroid cancer, liver cancer, hepatocellular carcinoma (HCC), or sporadic or hereditary papillary renal carcinoma (PRCC).