EA042597B1 - ANTIBODIES TO MET, BISPECIFIC ANTIGEN-BINDING MOLECULES THAT BIND MET, AND METHODS OF THEIR APPLICATION - Google Patents

Description

The technical field to which the invention belongs

The invention relates to antibodies, bispecific antibodies and their antigen-binding fragments, as well as antibody-drug conjugates of such antibodies that specifically bind the hepatocyte growth factor receptor (c-Met or MET) and modulate signal transduction involving MET, and methods for their use.

Sequence listing

An official copy of the sequence listing is filed simultaneously with this description electronically via EFS-Web as an ASCII sequence listing with the file name 10316WO01_Sequence_Listing_ST25_2017_11_06.TXT, created on November 6, 2017, and approximately 136 kB in size. The sequence listing contained in this ASCII document is part of this specification and is incorporated herein by reference in its entirety.

Background of the Invention

Hepatocyte growth factor (HGF) (also known as scattering factor [SF]) is a heterodimeric paracrine growth factor that exerts its activity by interacting with the HGF receptor (HGFR). HGFR is a product of the c-Met oncogene, and is also known as MET. MET is a receptor tyrosine kinase consisting of a transmembrane beta chain linked via a disulfide bridge to an extracellular alpha chain. Binding of HGF to MET activates the kinase catalytic activity of MET, which results in phosphorylation of the Tyr 1234 and Tyr 1235 beta chain and subsequent activation of downstream signaling pathways.

Overexpression, activation, or amplification of MET and/or HGF has been shown to be involved in the development of non-small cell lung carcinoma (NSCLC), a form of carcinoma of the stomach, ovary, pancreas, thyroid, head and neck breast, colon, and kidney (Sierra and Tsao , Ther. Adv. Med. Oncol., 3(1 Suppl): S21-S35, 2011). MET amplification is believed to be a key factor in oncogenesis in NSCLC forms and esophagogastric malignancies. In addition, mutations causing deletion of exon 14 MET have been described as oncogenic factors in NSCLC variants. Tumor cell lines characterized by MET gene amplification are highly dependent on MET for growth and survival. Preclinical data suggest that MET signaling is involved in the development of resistance to targeted therapies in many tumor types such as NSCLC, colorectal cancer, and head and neck squamous cell carcinoma (HNSCC).

Both preclinical and recent clinical results suggest that tumors with these genetic alterations respond to MET inhibitors, supporting the role of MET as a factor in cancer development. Various anti-MET monovalent blocking antibodies are under clinical development for the treatment of various forms of cancer (see US Pat. Such antibodies include onartuzumab (MetMab) and emibetuzumab, (Xiang et al., Clin. Cancer Res. 19(18): 5068-78, 2013, and Rosen et al, Clin. Cancer Res., Published October 10, 2016, doi : 10.1158/1078-0432.CCR-16-1418). Some of these antibodies block ligand-dependent signaling involving MET, but are not as effective in blocking ligand-independent MET activation.

There is still a significant unmet medical need for improved anti-cancer drugs that effectively block both ligand-dependent and ligand-independent MET signaling.

Brief summary of the invention

Provided herein are antibodies, antigen-binding antibody fragments, combinations of bivalent monospecific antibodies, and bispecific antibodies that bind the human c-Met receptor protein (MET x MET). The antibodies are useful, inter alia, for targeting tumor cells that express MET. Anti-MET antibodies and their antigen-binding portions may be used alone in unmodified form, or they may be included as part of an antibody-drug conjugate or bispecific antibody.

Other embodiments will be apparent from a review of the following detailed description.

Brief description of the drawings

Fig. 1 is a table illustrating the components of 272 exemplary MET x MET bispecific antibodies described herein. Each numbered cell of the table corresponds to a unique bispecific antibody containing a D1 antigen-binding domain and a D2 antigen-binding domain, where the D1 antigen-binding domain contains an immunoglobulin variable domain (HCVR/LCVR amino acid sequence pair) or a CDR from the corresponding anti-MET antibody shown on the Y-axis, and where the antigen-binding domain the D2 domain contains the immunoglobulin variable domain (HCVR/LCVR amino acid sequence pair) or the CDR from the corresponding anti-MET antibody shown on the x-axis.

- 1 042597

Fig. 2 is a schematic representation of an assay using a luciferase-based reporter system used to assess antibody-induced MET signaling pathway activation or antibody blocking of HGF-induced signaling pathway activation in HEK293T cells containing the luciferase SRE reporter gene construct.

Fig. 3(A-B) are line graphs showing relative light units (RLU) showing SRE-luciferase expression versus logarithm of antibody concentration in moles per liter. Solid squares () correspond to the original bivalent monospecific H4H13306P2 antibody, solid pyramids (▲) correspond to the original bivalent monospecific H4H13312P2 antibody, solid circles (•) correspond to the monovalent antibody, solid diamonds (♦) correspond to the isotype control, and filled inverted pyramids (▼) correspond to the absence of ligand . In FIG. 3A shows only the antibody without HGF ligand. In FIG. 3B shows antibodies and HGF ligand.

Fig. 4(A-B) are line graphs showing relative light units (RLU) showing SRE-luciferase expression versus logarithm of antibody concentration in moles per liter. Solid squares () represent monovalent anti-MET antibody, solid circles (•) represent bispecific MET x MET antibody, and filled diamonds (♦) represent parent H4H13312P2 antibody. In FIG. 4A shows only the antibody without HGF ligand. In FIG. 4B shows antibodies and HGF ligand.

Fig. 5 is a bar graph showing the relative growth of Met amplified SNU5 gastric cancer cells as a function of treatment with bivalent anti-MET 1-18 monospecific human antibodies, control antibody, and anti-MET monovalent antibody. For comparison purposes, antibody 8 (abscissa) is the parent H4H13306P2 antibody and antibody 11 (abscissa) is the parent H4H13312P2.

Fig. 6(A-B) are histograms showing the relative growth of Met amplified cells as a function of treatment with MET x MET bispecific antibody, control antibody, and anti-MET monovalent antibody. In FIG. 6A shows relative growth of SNU5 cells as a function of treatment with control antibody, monovalent antibody at 0.1, 1, and 10 μg/mL, and MET x MET bispecific antibody at 0.1, 1, and 10 μg/mL. In FIG. 6B shows the relative growth of EBC-1 cells as a function of treatment with control antibody and MET x MET bispecific antibody at 0.1 and 1 μg/ml.

In FIG. 7(A-B) shows immunoblotograms of pMET (phosphorylated MET), MET, pErk (phosphorylated Erk), and tubulin (as a loading control) isolated from Hs746T cells after treatment with control antibody and MET x MET bispecific antibody (Fig. 7A), and expression of MET (and tubulin as a load control) in Hs746T cells after treatment with the MET x MET bispecific antibody for 0, 2, and 6 hours (FIG. 7B).

In FIG. 8 shows an immunoblotogram of pMET, MET, pErk and tubulin (as a loading control) isolated from Hs746T cells after treatment with control antibody, MET x MET bispecific antibody, monospecific bivalent parent antibody 1, anti-MET monospecific bivalent parent antibody 2, and parent antibody combination 1 and 2.

In FIG. 9 shows an immunoblotogram of MET expression (and tubulin as a loading control) in Hs746T cells after treatment with control antibody and MET x MET bispecific antibody for 2, 6, and 18 hours.

In FIG. 10(A-B) shows immunoblotograms of pMET, MET, pErk and tubulin (as a loading control) isolated from SNU5 cells after treatment with control antibody and MET x MET bispecific antibody (FIG. 10A); and expression of MET (and tubulin as a load control) in SNU5 cells after treatment with control antibody and anti-MET monovalent antibody (FIG. 10B).

In FIG. 11 shows an immunoblotogram of pMET, MET, pErk and tubulin (as a loading control) isolated from EBC-1 cells after treatment with a control antibody and a MET x MET bispecific antibody.

Fig. 12 is a line graph showing the change in EBC-1 tumor volume in cubic millimeters versus time in days after implantation of EBC-1 cells in animals treated with control antibody (solid square), monovalent anti-MET antibody (solid circle •), or bispecific antibody MET x MET (filled diamond ♦).

Fig. 13(A-B) are histograms showing the relative growth of Met amplified cells as a function of treatment with MET x MET bispecific antibody, control antibody, and anti-MET monovalent antibody. In FIG. 13A shows the relative growth of Hs746T cells as a function of treatment with control antibody, MET x MET bispecific antibody, parent 1 MET x MET monospecific antibody, parent 2 MET x MET monospecific antibody, and combination of parent antibodies 1 and 2. FIG. 13B shows relative growth of Hs746T cells as a function of treatment with control antibody, monovalent antibody at 1, 10, and 25 μg/mL, and MET x MET bispecific antibody at 1, 10, and 25 μg/mL.

- 2 042597

Fig. 14 is a bar graph showing relative growth of NCI-H596 cells as a function of treatment with control antibody (C), MET x MET bispecific antibody (MM), parent 1 MET x MET monospecific antibody (M1), parent 2 MET x monospecific antibody MET (M2), a combination of parent antibodies 1 and 2 (M1M2) and hepatocyte growth factor (HGF) as a MET agonist.

Fig. 15 is a line graph showing the change in Hs746T tumor volume in cubic millimeters versus time in days after implantation of Hs746T cells in animals treated with control antibody (solid square ), anti-MET monovalent antibody (solid circle •), or MET x bispecific antibody. MET (filled diamond ♦).

Fig. 16A is a line graph showing the change in volume of SNU5 tumor in cubic millimeters versus time in days after implantation of SNU5 cells in animals treated with control antibody (solid square ), monovalent antibody to MET at 1 mg/mL (solid circle •) , anti-MET monovalent antibody at 10 mg/mL (open circle o), MET x MET bispecific antibody at 1 mg/mL (solid diamond ♦), or MET x MET bispecific antibody at 10 mg/mL (open diamond ◊).

Fig. 16B is an immunoblotogram of pMET, MET, and tubulin (load control) isolated from SNU5 tumor cells excised from mouse xenograft model after treatment with control antibody, 10 mg/kg anti-MET monovalent antibody, and 10 mg/kg MET x MET bispecific antibody.

Fig. 17 is a line graph showing the change in volume of U87MG tumor in cubic millimeters versus time in days after implantation of U87-MG cells in animals treated with control antibody (solid square ), anti-MET monovalent antibody (solid circle •), or bispecific antibody. MET x MET (filled diamond ♦).

Fig. 18 is a line graph showing the change in U118MG tumor volume in cubic millimeters versus time in days after implantation of U118-MG cells in animals treated with control antibody (solid square), anti-MET monovalent antibody (solid circle), or bispecific antibody MET x MET (open diamond ◊).

Fig. 19 is a schematic representation of the synthesis of maytansinoid 6.

Fig. 20 is a schematic representation of the synthesis of maytansinoid intermediate 1.

Detailed disclosure of the invention

Before the present invention is described, it should be understood that the present invention is not limited to the specific experimental methods and conditions described, as such methods and conditions may vary. It should also be understood that the terminology used herein is only intended to describe specific embodiments, and is not intended to be limiting, as the scope of the present invention will only be limited by the appended claims.

Unless otherwise indicated, all technical and scientific terms used herein have the same meanings as are commonly understood by one of ordinary skill in the art to which the present invention pertains. In the context of this document, the term approximately when applied to a specific specified numerical value means that the value may differ from the specified value by no more than 1%. For example, in the context of this document, the expression approximately 100 includes 99 and 101 and all values in between (eg, 99.1, 99.2, 99.3, 99.4, etc.).

While any methods and materials similar or equivalent to those described herein may be used to practice or test the present invention, the methods and materials described herein are preferred. All patents, applications and non-patent publications mentioned in the present description are incorporated herein by reference in their entirety.

Protein MET.

MET expressions, c-Met, etc. as used herein, refers to a human transmembrane receptor tyrosine kinase comprising: (1) the amino acid sequence set forth under SEQ ID NO: 145 and/or having the amino acid sequence set forth under NCBI accession number NM_001127500.2, which is a full-length proprotein precursor of the isoform a, (2) the amino acid sequence set forth under SEQ ID NO: 146 and/or having the amino acid sequence set forth under NCBI accession number NM_000236.2, which is a full-length isoform b proprotein precursor, (3) the amino acid sequence set forth under SEQ ID NO: 147, and/or having the amino acid sequence set forth under NCBI accession number NM_001311330.1, which is a full-length proprotein precursor of isoform c, and/or (3) a mature protein containing the cytoplasmic alpha subunit (SEQ ID NO: 148) , common to all three

- 3 042597 isoforms, and a transmembrane subunit beta (SEQ ID NO: 149, 150 or 151 isoforms a, b and c, respectively). The MET expression includes both monomeric and multimeric MET molecules. In the context of this document, the expression human monomeric MET means a MET protein or a portion thereof that does not contain or lack any multimerization domains and that normally exists as a single MET molecule without a direct physical link to another MET molecule. An exemplary monomeric MET molecule is a molecule referred to herein as hMET.mmh containing the amino acid sequence of SEQ ID NO: 152 (see, for example, Example 3 herein). In the context of this document, the expression dimeric human MET means a structure containing two MET molecules connected to each other by a linker, a covalent bond, a non-covalent bond, or a multimerization domain, such as the Fc domain of an antibody. An exemplary dimeric MET molecule is a molecule referred to herein as hMET.mFc containing the amino acid sequence of SEQ ID NO: 153 (see, for example, Example 3 herein).

All proteins, polypeptides, and protein fragments referenced herein are intended to refer to the human version of the corresponding protein, polypeptide, or protein fragment, unless explicitly stated to be of a non-human species. Thus, the expression MET means human MET, unless it is indicated that it refers to a species other than human, for example, mouse MET, monkey MET, etc.

In the context of this document, cell surface expressed MET refers to one or more MET proteins or its extracellular domain(s) that(s) are(are) expressed on the cell surface in vitro or in vivo such that at least a portion of the MET protein is exposed. on the extracellular side of the cell membrane and accessible to the antigen-binding site of the antibody. Cell surface-expressed MET may comprise or consist of a MET protein expressed on a cell surface that normally expresses the MET protein. Alternatively, cell surface-expressed MET may comprise or consist of a MET protein expressed on the surface of a cell that does not normally express human MET on its surface, but has been artificially engineered to express MET on its surface.

Anti-MET antibodies and their antigen-binding fragments.

In one aspect, anti-MET antibodies (eg, anti-MET monospecific antibodies) are provided. Exemplary anti-MET antibodies according to this aspect are listed in Table 1. 1 and 2 in this document. In table. 1 shows amino acid sequence identifiers for heavy chain variable regions (HCVR), light chain variable regions (LCVR), heavy chain complementarity determining regions (HCDR1, HCDR2, and HCDR3), and light chain complementarity determining regions (LCDR1, LCDR2, and LCDR3) of exemplary anti-MET antibodies. from which the bispecific antigen-binding molecules (herein used interchangeably with the bispecific antigen-binding protein) disclosed herein can be derived. In table. 2 shows the nucleic acid sequence identifiers HCVR, LCVR, HCDR1, HCDR2 HCDR3, LCDR1, LCDR2, and LCDR3 of exemplary anti-MET antibodies.

Also provided herein are antibodies or antigen-binding fragments thereof that specifically bind MET and exert an agonistic effect (e.g., activation) on a MET signaling pathway in cells, as well as the use of such antibodies in therapeutic conditions in which activation of signaling involving MET will be beneficial or therapeutically useful. Non-limiting examples of such anti-MET agonist antibodies include the antibody referred to herein as H4H14636D, as well as antibodies and antigen-binding fragments thereof containing heavy and light chain CDRs (SEQ ID NOs: 28, 30, 32, 140, 142, 144) and/or their heavy and light chain variable domains (SEQ ID NO: 26/138).

Provided herein are antibodies or antigen-binding fragments thereof that specifically bind METs containing an HCVR containing an amino acid sequence selected from any of those shown in Table 1. 1 amino acid sequences of HCVR, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Provided herein are antibodies or antigen-binding fragments thereof that specifically bind METs containing an LCVR containing an amino acid sequence selected from any of those shown in Table. 1 amino acid sequences of an LCVR, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Provided herein are antibodies or antigen-binding fragments thereof that specifically bind MET, comprising a pair of amino acid sequences HCVR and LCVR (HCVR/LCVR) comprising any of those listed in Table 1. 1 amino acid follower

- 4 042597 HCVR paired with any of those given in the table. 1 amino acid sequences of LCVR. In accordance with certain variants of the implementation of antibodies or antigennegative fragments contain a pair of amino acid sequences HCVR/LCVR, contained within any of those given in table. 1 exemplary anti-MET antibodies. In certain embodiments, the HCVR/LCVR amino acid sequence pair is selected from the group consisting of: SEQ ID NO: 2/138, 10/138, 18/138, 26/138, 34/138, 42/138, 50/138, 58 /138, 66/138, 74/138, 82/138, 90/138, 98/138, 106/138, 114/138, 122/138 and 130/138.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a heavy chain CDR1 (HCDR1) containing an amino acid sequence selected from any of those given in Table 1. 1 amino acid sequences of HCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a heavy chain CDR2 (HCDR2) containing an amino acid sequence selected from any of those listed in Table 1. 1 amino acid sequences of HCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a heavy chain CDR3 (HCDR3) containing an amino acid sequence selected from any of those shown in Table 1. 1 amino acid sequences of HCDR3, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a light chain CDR1 (LCDR1) containing an amino acid sequence selected from any of those shown in Table 1. 1 amino acid sequences of LCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are antibodies or their antigen-binding fragments that specifically bind MET, containing a pair of amino acid sequences HCDR1 and LCDR1 (HCDR1/LCDR1), providing any of those given in table. 1 amino acid sequences HCDR1 paired with any of those given in table. 1 amino acid sequences of LCDR1. In accordance with certain variants of the implementation of antibodies or antigennegative fragments contain a pair of amino acid sequences HCDR3/LCDR3 contained within any of those given in table. 1 exemplary anti-MET antibodies. In certain embodiments, the HCDR1/LCDR1 amino acid sequence pair is selected from the group consisting of: SEQ ID NO: 4/140, 12/140, 20/140, 28/140, 36/140, 44/140, 52/140, 60 /140, 68/140, 76/140, 84/140, 92/140, 100/140, 108/140, 116/140, 124/140 and 132/140.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a light chain CDR2 (LCDR2) containing an amino acid sequence selected from any of those given in Table 1. 1 amino acid sequences of LCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are antibodies or their antigen-binding fragments that specifically bind MET, containing a pair of amino acid sequences HCDR2 and LCDR2 (HCDR2/LCDR2), providing any of those given in table. 1 amino acid sequences HCDR2 paired with any of those given in table. 1 LCDR2 amino acid sequences. In accordance with certain variants of the implementation of antibodies or antigennegative fragments contain a pair of amino acid sequences HCDR2/LCDR2 contained within any of those given in table. 1 exemplary anti-MET antibodies. In certain embodiments, the HCDR2/LCDR2 amino acid sequence pair is selected from the group consisting of: SEQ ID NO: 6/142, 14/142, 22/142, 30/142, 38/142, 46/142, 54/142, 62 /142, 70/142, 78/142, 86/142, 94/142, 102/142, 110/142, 118/142, 126/142 and 134/142.

Also provided are antibodies or antigen-binding fragments thereof that specifically bind METs containing a light chain CDR3 (LCDR3) containing an amino acid sequence selected from any of those given in Table 1. 1 amino acid sequences of LCDR3, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity

- 5 042597 news.

Also provided herein are antibodies or antigen-binding fragments thereof that specifically bind MET, containing a pair of amino acid sequences HCDR3 and LCDR3 (HCDR3/LCDR3), including any of those listed in table. 1 amino acid sequences HCDR3 paired with any of those given in table. 1 LCDR3 amino acid sequences. In accordance with certain variants of the implementation of antibodies or antigennegative fragments contain a pair of amino acid sequences HCDR3/LCDR3 contained within any of those given in table. 1 exemplary anti-MET antibodies. In certain embodiments, the HCDR3/LCDR3 amino acid sequence pair is selected from the group consisting of: SEQ ID NO: 8/144, 16/144, 24/144, 32/144, 40/144, 48/144, 56/144, 64 /144, 72/144, 80/144, 88/144, 96/144, 104/144, 112/144, 120/144, 128/144 and 136/144.

Also provided herein are antibodies or antigen-binding fragments thereof that specifically bind METs containing a set of six CDRs (i.e., HCDR1-HCDR2-HCDR3LCDR1-LCDR2-LCDR3) contained within any of those shown in Table 1. 1 exemplary anti-MET antibodies. In certain embodiments, the set of amino acid sequences HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 is selected from the group consisting of: SEQ ID NO: 4-6-8-140-142-144, 12-14-16-140-142 -144, 20-22-24-140-142-144, 28-30-32-140-142-144, 36-38-40-140142-144, 44-44-48-140-142-144, 52 -54-56-140-142-144 60-62-64-140-142-144 68-70-72-140-142-144 76-7880-140-142-144 84-86-88 -140-142-144, 92-94-96-140-142-144, 100-102-104-140-142-144, 108-110-112-140142-144, 116-118-120-140-142 -144, 124-126-128-140-142-144 and 132-134-136-140-142-144.

In a similar embodiment, antibodies or antigen-binding fragments thereof that specifically bind MET comprise a set of six CDRs (i.e. HCDR1-HCDR2-HCDR3LCDR1-LCDR2-LCDR3) contained within the HCVR/LCVR amino acid sequence pair as defined for any from those given in table. 1 exemplary anti-MET antibodies. For example, antibodies or antigen-binding fragments thereof that specifically bind MET comprise the HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3 amino acid sequence set contained within the HCVR/LCVR amino acid sequence pair selected from the group consisting of: SEQ ID NO: 4-6-8-140-142-144 12-14-16-140-142-144 20-22-24-140-142-144 2830-32-140-142-144 36-38- 40-140-142-144, 44-44-48-140-142-144, 52-54-56-140-142-144, 60-62-64-140-142144, 142-144, 76-78-80-140-142-144, 84-86-88-140-142-144, 92-94-96-140-142-144, 100-102104-140-142-144, 108-110-112-140-142-144, 116-118-120-140-142-144, 124-126-128-140-142-144 and 132-134136-140-142-144.

Methods and techniques for identifying CDRs within the amino acid sequences of HCVRs and LCVRs are well known in the art and can be used to identify CDRs within the amino acid sequences of certain HCVRs and/or LCVRs disclosed herein. Illustrative conventional approaches that can be used to identify CDR boundaries include, for example, Kabat detection, Chothia detection, and antibody model detection. In general terms, the Kabat definition is based on sequence variability, the Chothia definition is based on the location of structural loop regions, and the antibody model definition is a compromise between the Kabat and Chothia approaches. See, for example, Kabat, Sequences of Proteins of Immunological Interest, National Institutes of Health, Bethesda, Maryland. (1991); Al-Lazikani et al., J. Mol. Biol. 273:927-948 (1997); and Martin et al., Proc. Natl. Acad. sci. USA 86:9268-9272 (1989). Public databases can also be used to identify CDR sequences within an antibody.

Also provided herein are nucleic acid molecules encoding anti-MET antibodies or portions thereof. For example, the present invention relates to nucleic acid molecules encoding any of those listed in table. 1 amino acid sequences of HCVR; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding HCVR, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 LCVR amino acid sequences; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding an LCVR, or a sequence substantially similar to it, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 amino acid sequences of HCDR1; in certain embodiments, the nucleic acid molecule comprises a polynucleotide sequence selected from any of

- 6 042597 given in table. 2 nucleic acid sequences encoding HCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 amino acid sequences of HCDR2; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding HCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 amino acid sequences of HCDR3; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding HCDR3, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 LCDR1 amino acid sequences; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding LCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 LCDR2 amino acid sequences; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding LCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding any of those listed in table. 1 LCDR3 amino acid sequences; according to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding LCDR3, or a sequence substantially similar to it, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also presented are nucleic acid molecules encoding HCVR, where HCVR contains a set of three CDRs (i.e. HCDR1-HCDR2-HCDR3), where the set of amino acid sequences HCDR1-HCDR2-HCDR3 is as defined for any of those given in table. 1 exemplary anti-MET antibodies.

Also provided are nucleic acid molecules encoding LCVR, where the LCVR contains a set of three CDRs (i.e., LCDR1-LCDR2-LCDR3), where the set of amino acid sequences of LCDR1LCDR2-LCDR3 is as defined for any of those given in table. 1 exemplary anti-MET antibodies.

Also presented are nucleic acid molecules encoding both HCVR and LCVR, where HCVR contains an amino acid sequence from any of those given in table. 1 amino acid sequences of HCVR, and where LCVR contains an amino acid sequence from any of those given in table. 1 amino acid sequences of LCVR. According to certain embodiments, the implementation of the nucleic acid molecule contains a polynucleotide sequence selected from any of those given in table. 2 nucleic acid sequences encoding HCVR, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it, and a polynucleotide sequence selected from any of those given in the table. 2 nucleic acid sequences encoding an LCVR, or a sequence substantially similar to it, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it. In certain embodiments in accordance with this aspect, the nucleic acid molecule encodes HCVR and LCVR, where both HCVR and LCVR come from the same table. 1 antibody to MET.

- 7 042597

Also provided herein are recombinant expression vectors capable of expressing a polypeptide containing the variable region of the heavy or light chains of an anti-MET antibody. For example, recombinant expression vectors contain any of the nucleic acid molecules mentioned above, ie. nucleic acid molecules encoding any of the HCVR, LCVR and/or CDR sequences shown in Table. 1. Also included within the scope of this disclosure are host cells into which such vectors have been introduced, as well as methods for producing antibodies or parts thereof by culturing the host cells under conditions that allow the production of antibodies or antibody fragments and isolating the antibodies and fragments thus obtained. antibodies.

This document provides antibodies to MET with a modified glycosylation profile. In some embodiments, modification to remove unwanted glycosylation sites or lack of a fucose moiety present in the oligosaccharide chain may be useful, e.g., to enhance the function of antibody-dependent cellular cytotoxicity (ADCC) (see Shield et al. (2002) JBC 277:26733). According to other applications, in order to modify complement-dependent cytotoxicity (CDC), modification can be performed in relation to galactosylation.

Bispecific antigen-binding molecules MET x MET

The present inventors have observed that certain monospecific antigen-binding MET-binding molecules that block the binding of HGF to MET tend to strongly activate MET signaling (an undesirable consequence for the therapeutic molecule). The present inventors surprisingly found, however, that bispecific antigen-binding molecules that simultaneously bind to two distinct epitopes on the extracellular domain of the MET protein are effective in blocking ligand binding to MET while conferring little agonistic effect on MET signaling.

Accordingly, provided herein are bispecific antigen-binding molecules comprising a first antigen-binding domain (also referred to herein as D1) and a second antigen-binding domain (also referred to herein as D2). Simultaneous binding of two separate MET epitopes by a bispecific antigen-binding molecule results in effective ligand blocking with minimal activation of MET signaling.

Bispecific antigen-binding molecules that comprise a first antigen-binding domain (D1) that specifically binds a first human MET epitope and a second antigen-binding domain (D2) that specifically binds a second human MET epitope may be referred to herein as bispecific antibodies MET x MET, MET x MET, or other similar terms. In some embodiments, the first human MET epitope comprises amino acids 192-204 of SEQ ID NO: 155. In some embodiments, the second human MET epitope comprises amino acids 305-315 and 421-455 of SEQ ID NO: 155. In some embodiments, the first human MET epitope contains amino acids 192-204 of SEQ ID NO: 155; and the second human MET epitope contains amino acids 305-315 and 421-455 of SEQ ID NO: 155.

In certain embodiments, the D1 and D2 domains of the MET x MET bispecific antibody do not compete with each other. The lack of competition between D1 and D2 for binding to MET means that the respective monospecific antigen-binding proteins from which D1 and D2 were derived do not compete with each other for binding to human MET. Illustrative antigen-binding protein competition assays are known in the art, non-limiting examples of which are described elsewhere herein.

In certain embodiments, D1 and D2 bind to different (eg, non-overlapping or partially overlapping) epitopes on the MET, as described elsewhere herein.

Bispecific antigen-binding molecules MET x MET can be constructed using the antigen-binding domains of two separate anti-MET monospecific antibodies. For example, a series of monoclonal monospecific antibodies to MET can be obtained using standard methods known in the art. Individual antibodies thus obtained can be tested in pairs against each other for cross-competition for binding to the MET protein. If two different anti-MET antibodies are able to bind to MET at the same time (i.e. do not compete with each other), then the antigen-binding domain from the first anti-MET antibody and the antigen-binding domain from the second non-competitive anti-MET antibody can be constructed as a single bispecific MET x antibody. MET according to this disclosure.

In accordance with the present disclosure, the bispecific antigen binding molecule may be a single polyfunctional polypeptide, or it may be a multimeric complex of two or more polypeptides that are covalently or non-covalently linked to each other. As will be appreciated from the present disclosure, any antigen-binding construct that is capable of simultaneously binding two separate, non-identical epitopes of the MET molecule is considered to be a bispecific antigen-binding molecule. Any of the bispecific antigen-binding molecules described herein, or variants thereof, may be constructed using standard molecular biology techniques (eg, recombinant DNA and protein engineering techniques), as will be known to those skilled in the art.

Antigen binding domains

The bispecific antigen-binding molecules of the present disclosure comprise two separate antigen-binding domains (D1 and D2). As used herein, the expression antigen-binding domain means any peptide, polypeptide, nucleic acid molecule, backbone molecule, peptide-containing molecule, or polypeptide-containing construct that is capable of specifically binding to a particular antigen of interest (e.g., human MET). The term specifically binds, etc. in the context of this document means that the antigen-binding domain forms a complex with a specific antigen, characterized by a dissociation constant (KD) of 500 pM or less, and does not bind other unrelated antigens under normal testing conditions. Unrelated antigens are proteins, peptides, or polypeptides that share less than 95% amino acid identity with respect to each other.

Illustrative categories of antigen-binding domains that can be used in the context of the present disclosure include antibodies, antigen-binding portions of antibodies, peptides that specifically interact with a particular antigen (e.g., peptibodies), receptor molecules that specifically interact with a particular antigen, proteins containing a ligand-binding portion of a receptor , which specifically binds a particular antigen, antigen-binding scaffold proteins (e.g., DARPin, HEAT repeat proteins, ARM repeat proteins, tetratricopeptide repeat proteins, and other scaffold proteins based on naturally occurring repeat proteins, etc. , [see, for example, Boersma and Pluckthun, 2011, Curr Opin Biotechnol 22:849-857 and references cited therein]), and aptamers and parts thereof.

Methods for determining whether two molecules specifically bind to each other are well known in the art and include, for example, equilibrium dialysis, surface plasmon resonance, and the like. For example, an antigen-binding domain used in the context of the present disclosure includes polypeptides that bind a particular antigen (e.g., target molecule [T] or internalizing effector protein [E]), or a portion thereof, with a K _D of less than about 500 pM, less than about 400 less than about 300 pM, less than about 200 pM, less than about 100 pM, less than about 90 pM, less than about 80 pM, less than about 70 pM, less than about 60 pM, less than about 50 pM, less than about 40 pM, less than about 30 pM, less than about 20 pM, less than about 10 pM, less than about 5 pM, less than about 4 pM, less than about 2 pM, less than about 1 pM, less than about 0.5 pM, less than about 0.2 pM, less than about 0.1 pM, or less than about 0.05 pM, as measured in an analysis based on the surface plasmon resonance phenomenon.

The term surface plasmon resonance in the context of this document refers to an optical phenomenon that allows analysis of interactions in real time by detecting changes in protein concentrations on a matrix in the form of a biosensor, for example, using the BIAcore™ system (Biacore Life Sciences, a division of GE Healthcare, Piscataway, New Jersey).

The term K _D in the context of this document means the equilibrium dissociation constant of a particular protein-protein interaction (eg, antibody-antigen interaction). Unless otherwise indicated, the K _D values disclosed herein refer to the K _D values determined by analysis based on the phenomenon of surface plasmon resonance at 25°C.

As indicated above, the antigen-binding domain (D1 and/or D2) may comprise or consist of an antibody or an antigen-binding fragment of an antibody. The term antibody as used herein means any antigen-binding molecule or molecular complex containing at least one complementarity determining region (CDR) that specifically binds or interacts with a particular antigen (eg, human MET). The term antibody includes immunoglobulin molecules containing four polypeptide chains, two heavy (H) chains and two light (L) chains connected to each other by disulfide bonds, as well as their multimers (eg, IgM). Each heavy chain contains a heavy chain variable region (abbreviated herein as HCVR or VH) and a heavy chain constant region. The heavy chain constant region contains three domains, CH1, C _H 2 and C _H 3. Each light chain contains a light chain variable region (abbreviated herein as LCVR or VL) and a light chain constant region. The light chain constant region contains one domain (C _L 1). The VH and VL regions can further be divided into hypervariable regions called complementarity determining regions (CDRs) interspersed with regions that are more conserved called framework regions (FRs). Each V _H and V _L consists of three CDRs and four FRs, arranged from amino to carboxy in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. According to times

- 9 042597 personal embodiments The FRs of the antibodies (or antigen-binding portion thereof) provided herein may be identical to human germline sequences or may be naturally or artificially modified. Amino acid consensus sequence can be determined based on a comparative analysis of two or more CDRs.

The D1 and/or D2 components of the bispecific antigen-binding molecules provided herein may contain or consist of antigen-binding fragments of full antibody molecules. The terms antigen-binding portion of an antibody, antigen-binding fragment of an antibody, etc. as used herein, any naturally occurring, enzymatically produced, synthetic, or genetically engineered polypeptide or glycoprotein that specifically binds an antigen to form a complex is included. Antigen-binding antibody fragments can be generated, for example, from whole antibody molecules using any suitable standard techniques, such as proteolytic cleavage techniques or recombinant DNA genetic engineering techniques involving the manipulation and expression of DNA encoding the variable and optionally constant domains of antibodies. Such DNA is known and/or readily available, for example from commercial sources, DNA libraries (including, for example, antibody phage libraries), or it can be synthesized. The DNA can be sequenced and chemically manipulated using molecular biology techniques, for example, to place one or more variable and/or constant domains in an appropriate configuration, or to include codons, introduce cysteine residues, modify, add or remove amino acids, etc.

Non-limiting examples of antigen-binding fragments include: (i) Fab fragments; (ii) F(ab') ₂ fragments; (iii) Fd fragments; (iv) Fv fragments; (v) single chain Fv molecules (scFv); (vi) dAb fragments; and (vii) minimal recognition units consisting of amino acid residues that mimic the hypervariable region of an antibody (eg, a distinguished complementarity determining region (CDR), such as a CDR3 peptide), or a sterically restricted FR3-peptide. CDR3-FR4. Other engineered molecules such as domain-specific antibodies, single domain antibodies, domain deleted antibodies, chimeric antibodies, CDR grafted antibodies, diabodies, tribodies, tetrabodies, minibodies, nanobodies (e.g., monovalent nanobodies, bivalent nanobodies, etc.) , small modular protein (SMIP) immunopharmaceuticals, and shark IgNAR variable domains are also encompassed within the context of this document by the expression antigen-binding fragment.

An antigen binding fragment of an antibody will typically contain at least one variable domain. The variable domain may be of any size or amino acid composition, and will typically contain at least one CDR that is adjacent to or in frame with one or more framework sequences. In antigen-binding fragments with a VH domain associated with a VL domain, the V _H and V _L domains may be positioned relative to each other in any suitable configuration. For example, the variable region may be dimeric and contain V _H -V _H , V _H -V _L or V _L -V _L dimers. Alternatively, the antigen-binding fragment of an antibody may contain a monomeric V _H or VL domain.

In certain embodiments, an antigen-binding fragment of an antibody may comprise at least one variable domain covalently linked to at least one constant domain. Non-limiting exemplary variable and constant domain configurations that can be found within an antigen-binding fragment of an antibody of the present disclosure include: (i) VH-CH1; (ii) VH-CH2; (iii) VH-CH3; (iv) VH-CH1-CH2; (v) VH-CH1-CH2-CH3; (vi) Vh-Ch2-Ch3; (vii) Vh-Cl; (viii) Vl-Ch1; (ix) Vl-Ch2; (x) Vl-Ch3; (xi) Vl-Ch1-Ch2; (xii) Vl-Ch1-Ch2-Ch3; (xiii) VL-CH2-CH3; and (xiv) VL-CL. In any configuration of the variable and constant domains, including any of the above illustrative configurations, the variable and constant domains can either be directly connected to each other, or can be connected through a full or partial hinge region or a linker region. The hinge region may be at least 2 (e.g., 5, 10, 15, 20, 40, 60 or more) amino acids that provide a flexible or semi-flexible connection of adjacent variable and/or constant domains within a single polypeptide molecule. Moreover, the antigen-binding moiety may comprise a homodimer or heterodimer (or other multimer) of any of the above variable and constant domain configurations non-covalently linked to each other and/or to one or more monomeric V _H or V _L domains (e.g., via disulfide(disulfide) bond(s)).

The bispecific antigen-binding molecules provided herein may comprise or consist of human antibodies and/or recombinant human antibodies, or fragments thereof. As used herein, the term human antibody includes antibodies with variable and constant regions derived from human germline immunoglobulin sequences. Human antibodies, however, may contain amino acid residues not encoded by human germline immunoglobulin sequences (for example, due to mu

- 10 042597 mutations introduced by random or site-specific mutagenesis in vitro or due to somatic mutation in vivo), for example, in CDR and, in particular, CDR3. However, in the context of this document, the term human antibody is intended not to include antibodies in which CDR sequences derived from the germline of another mammalian species, such as a mouse, have been grafted onto human framework sequences.

The bispecific antigen-binding molecules of the present disclosure may comprise or consist of recombinant human antibodies or antigen-binding fragments thereof. The term recombinant human antibody as used herein is intended to include all human antibodies produced, expressed, generated or isolated by recombinant methods, such as antibodies expressed using a recombinant expression vector transfected into a host cell (described further below). ), antibodies isolated from a recombinant combinatorial human antibody library (described further below), antibodies derived from an animal (e.g., mouse) that is transgenic for human immunoglobulin genes (see, e.g., Taylor et al. (1992) Nucl. Acids Res. 20:6287-6295), or antibodies produced, expressed, constructed, or isolated by other means, which involve combining human immunoglobulin gene sequences with other DNA sequences. Such recombinant human antibodies contain variable and constant regions derived from human germline immunoglobulin sequences. However, in certain embodiments, such human recombinant antibodies are subjected to in vitro mutagenesis (or, in the case of an animal transgenic for human Ig sequences, in vivo somatic mutagenesis), and hence the amino acid sequences of the VH and VL regions of the recombinant antibodies are sequences that, although derived from, or related to, human germline VH and V _L sequences, may not naturally exist within the in vivo human germline antibody repertoire.

Methods for making bispecific antibodies are known in the art and can be used to construct the bispecific antigen-binding molecules disclosed herein. Illustrative bispecific formats that can be used in the context of the present disclosure include, without limitation, for example, scFv-based or diabody bispecific formats, IgG-scFv fusion bispecific formats, Ig dual variable domains (DVD), Quadroma, antibodies with ridge-to-pit, common light chain antibody (e.g., ridge-to-bottom common light chain, etc.), CrossMab, CrossFab, SEED-derived antibody ((SEED)body), antibody derived by leucine zipper mediated dimerization, Duobody, IgG1/IgG2, IgG with dual acting Fab (DAF), and Mab ² bispecific formats (see e.g. Klein et al. 2012, mAbs 4:6, 1-11 and references cited therein , for an overview of the formats listed above).

Exemplary antigen-binding domains (D1 and D2) that may be included in the MET x MET bispecific antigen-binding molecules provided herein include antigen-binding domains derived from any of the anti-MET antibodies disclosed herein. For example, the present disclosure includes bispecific antigen-binding molecules MET x MET containing antigennegative domain D1 or D2 containing HCVR containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of HCVR, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Also provided herein are bispecific antigen-binding molecules MET x MET containing a D1 or D2 antigen-binding domain containing an LCVR containing an amino acid sequence selected from any of those shown in Table 1. 1 LCVR amino acid sequences, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity with it.

Presented herein are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a pair of amino acid sequences HCVR and LCVR (HCVR/LCVR) comprising any of those listed in Table 1. 1 amino acid sequences HCVR paired with any of those given in table. 1 amino acid sequences of LCVR. In accordance with certain variants of implementation of the present invention relates to bespecifically antigennegative molecules MET x MET containing antigennegative domain D1 or D2 containing a pair of amino acid sequences HCVR/LCVR contained within any of those given in table. 1 exemplary anti-MET antibodies.

Also provided herein are MET x MET bispecific antigen-binding molecules containing a D1 or D2 antigen-binding domain containing a heavy chain CDR1

- 11 042597 (HCDR1), containing the amino acid sequence selected from any of those given in table. 1 amino acid sequences of HCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a heavy chain CDR2 (HCDR2) containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of HCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a heavy chain CDR3 (HCDR3) containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of HCDR3, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a light chain CDR1 (LCDR1) containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of LCDR1, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a light chain CDR2 (LCDR2) containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of LCDR2, or a sequence substantially similar to it, characterized by at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a light chain CDR3 (LCDR3) containing an amino acid sequence selected from any of those given in table. 1 amino acid sequences of an LCDR3, or a sequence substantially similar to it, having at least 90%, at least 95%, at least 98%, or at least 99% sequence identity.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a pair of amino acid sequences HCDR3 and LCDR3 (HCDR3/LCDR3), providing any of those given in table. 1 amino acid sequences HCDR3 paired with any of those given in table. 1 LCDR3 amino acid sequences. In accordance with certain variants of implementation of the present disclosure relates to antibodies or their antigennegative fragments containing a pair of amino acid sequences HCDR3/LCDR3 contained within any of those given in table. 1 illustrative antibodies to MET.

Also presented are bispecific antigen-binding molecules MET x MET containing an antigen-binding domain D1 or D2 containing a set of six CDRs (i.e. HCDR1-HCDR2-HCDR3LCDR1-LCDR2-LCDR3) contained within any of those given in table. 1 exemplary anti-MET antibodies.

In a similar embodiment, the present disclosure relates to MET x MET bispecific antigen-binding molecules comprising a D1 or D2 antigen-binding domain comprising a set of six CDRs (i.e. HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3) contained within an amino acid pair sequences HCVR/LCVR, as defined for any of those given in table. 1 exemplary anti-MET antibodies.

The bispecific MET x MET antigen-binding molecules provided herein may contain a D1 antigen-binding domain derived from any of the anti-MET antibodies of Table 1. 1, and antigennegative domain D2, derived from any other antibody to MET table. 1. Non-limiting examples of MET x MET bispecific antibodies of the present disclosure are shown in FIG. 1. FIG. 1 is a table illustrating the components of 272 exemplary MET x MET bispecific antibodies. Each numbered cell of the table (numbered 1-272) corresponds to a unique bispecific antibody containing a D1 antigen-binding domain and a D2 antigen-binding domain, where the D1 antigen-binding domain contains an immunoglobulin variable domain (a pair of HCVR/LCVR amino acid sequences) or a CDR from the corresponding anti-MET antibody shown along the y-axis, and where the antigen-binding domain

- 12 042597

D2 contains an immunoglobulin variable domain (a pair of amino acid sequences HCVR/LCVR) or a CDR from the corresponding antibody to MET, shown on the x-axis. Thus, for example, shown in table. MET x MET bispecific antigen-binding molecule number 10 comprises a D1 antigen-binding domain containing the HCVR/LCVR pair or the 6-CDR set from exemplary anti-MET antibody H4H13290P2 and a D2 antigen-binding domain containing the HCVR/LCVR pair or the 6-CDR set from exemplary anti-MET antibody H4H13321P2. Additional examples of MET x MET bispecific antibodies provided herein are described in Example 4 herein.

As a non-limiting illustrative example, the present disclosure includes MET x MET bispecific antigen-binding molecules comprising a D1 antigen-binding domain and a D2 antigen-binding domain, wherein the D1 antigen-binding domain comprises a pair of amino acid sequences HCVR/LCVR SEQ ID NO: 58/138 or a set of heavy and light chain CDRs (HCDR1HCDR2-HCDR3-LCDR1-LCDR2-LCDR3) comprising SEQ ID NO: 60-62-64-140-142-144, and wherein the D2 antigen-binding domain contains a pair of amino acid sequences HCVR/LCVR SEQ ID NO: 82/138 or a set Heavy and light chain CDRs (HCDR1-HCDR2-HCDR3-LCDR1-LCDR2LCDR3) providing for SEQ ID NO: 84-86-88-140-142-144. An exemplary MET x MET bispecific antibody with these sequence characteristics is the bispecific antibody designated H4H14639D, also referred to as bispecific antibody #122, which contains D1 derived from H4H13306P2 and D2 derived from H4H13312P2 (see Example 4, Table 5). in this document).

As a further non-limiting illustrative example, the present disclosure includes MET x MET bispecific antigen-binding molecules comprising a D1 antigen-binding domain and a D2 antigen-binding domain, wherein the D1 antigen-binding domain comprises the amino acid sequence pair HCVR/LCVR SEQ ID NO: 18/138 or the severe and mild CDR set chains (HCDR1-HCDR2-HCDR3-LCDR1-LCDR2-LCDR3), providing SEQ ID NO: 20-22-24140-142-144, and where the antigen-binding domain D2 contains a pair of amino acid sequences HCVR/LCVR SEQ ID NO: 82/138 or a set of heavy and light chain CDRs (HCDR1-HCDR2-HCDR3LCDR1-LCDR2-LCDR3) comprising SEQ ID NO: 84-86-88-140-142-144. An exemplary MET x MET bispecific antibody with these sequence characteristics is the bispecific antibody designated H4H14635D, also referred to as bispecific antibody #42, which contains D1 derived from H4H13295P2 and D2 derived from H4H13312P2 (see Example 4, Table 5). in this document).

Components that provide multimerization.

In certain embodiments, the bispecific antigen-binding molecules provided herein may also contain one or more multimerization components. Components that provide multimerization, can function to maintain the association of antigen-binding domains between themselves (D1 and D2). As used herein, a multimerizing component is a macromolecule, protein, polypeptide, peptide, or amino acid that has the ability to associate with a second multimerizing component with the same or similar structure or composition. For example, the multimerization component may be a polypeptide containing the CH3 domain of an immunoglobulin. A non-limiting example of a multimerizing component is the Fc portion of an immunoglobulin, such as the Fc domain of an IgG, selected from the IgG1, IgG2, IgG3, and IgG4 isotypes, as well as any allotype within each isotype group. In certain embodiments, the multimerization component is an Fc fragment or an amino acid sequence of 1 to about 200 amino acids in length containing at least one cysteine residue. In other embodiments, the multimerizing component is a cysteine residue or a short cysteine-containing peptide. Other multimerization domains include peptides or polypeptides containing or consisting of a leucine zipper, a helix-loop motif, or a double helix motif.

In certain embodiments, the bispecific antigen-binding molecules provided herein comprise two multimerization domains, M1 and M2, where D1 is connected to M1 and D2 is connected to M2, and where the association of M1 to M2 promotes physical association of D1 and D2 with each other in within a single bispecific antigen-binding molecule. According to certain embodiments, M1 and M2 are identical to each other. For example, M1 may be an Fc domain with a specific amino acid sequence, and M2 is an Fc domain with the same amino acid sequence as M1. Alternatively, M1 and M2 may differ from each other in one or more amino acid positions. For example, M1 may contain a first C _H 3 domain of an immunoglobulin (Ig), and M2 may contain a second C _H 3 domain of an Ig, where the first and second C _H 3 domains of Ig differ from each other by at least

- 13 042597 at least one amino acid, and where at least one difference in amino acid provides a decrease in the ability of the targeting construct to bind to protein A compared to a reference construct having identical M1 and M2 sequences. In one embodiment, the C _H 3 domain of Ig M1 binds protein A, and the C _H 3 domain of Ig M2 provides a mutation that reduces or eliminates the ability to bind to protein A, such as the H95R modification (according to IMGT mutation numbering in exons; H435R according to EU numbering). CH3 M2 can additionally provide for the modification Y96F (according to IMGT; Y436F according to EU). Additional modifications that can be found within CH3 M2 include: D16E, L18M, N44S, K52N, V57M and V82I (according to IMGT; D356E, L358M, N384S, K392N, V397M and V422I according to EU) in the case of the IgG1 Fc domain; N44S, K52N and V82I (IMGT; N384S, K392N and V422I according to EU) in the case of the IgG2 Fc domain; and Q15R, N44S, K52N, V57M, R69K, E79Q and V82I (as per IMGT; Q355R, N384S, K392N, V397M, R409K, E419Q and V422I as per EU) in the case of the IgG4 Fc domain.

The bispecific antigen-binding molecules of the present disclosure may be isolated. An isolated bespecifically antigen-binding molecule as used herein means a bispecific antigen-binding molecule that has been identified and separated from and/or extracted from at least one component of its natural environment. For example, a bispecific antibody that has been separated from and/or extracted from at least one component of an organism or from a tissue or cell in which the antibody is produced is, for the purposes of this disclosure, an isolated bispecific antibody. An isolated bispecific antigen-binding molecule also includes molecules in situ in a recombinant cell. Isolated bispecific antigen-binding molecules are molecules that have been subjected to at least one purification or isolation step. In accordance with certain embodiments, the isolated bispecific antigen-binding molecule may be substantially free of other cellular material and/or chemicals.

The bispecific antigen-binding molecules or antigen-binding domains (D1 and/or D2) thereof disclosed herein may include one or more amino acid substitutions, insertions and/or deletions in the framework and/or CDR regions of the heavy and light chain variable domains compared to the corresponding sequences. the germline from which the antigen-binding proteins or antigen-binding domains were derived. Such mutations can be readily identified by comparing the amino acid sequences disclosed herein with germline sequences available, for example, from public antibody sequence databases. The present disclosure includes the bispecific antigen-binding molecules disclosed herein, or antigen-binding domains (D1 and/or D2) thereof, which are derived from any of the amino acid sequences disclosed herein, wherein one or more amino acids within one or more framework and/or CDR regions are subjected to mutations with a replacement for the corresponding(s) residue(s) of the germline sequence from which the antibody is derived, or the corresponding(s) residue(s) of another germline sequence, or are subject to a conservative amino acid substitution of the corresponding residue(s) of the germline (such changes with respect to the sequence in this document are collectively referred to as germline mutations).

One skilled in the art, from the heavy and light chain variable region sequences disclosed herein, will readily be able to generate a plurality of bispecific antigen-binding molecules or antigen-binding domains (D1 and/or D2) thereof that involve one or more single germline mutations or combinations thereof. In certain embodiments, all of the frame and/or CDR residues within the V _H and/or V _L domains are back-mutated to residues found in the original germline sequence from which the antibody is derived. In other embodiments, only certain residues are mutated back to the original germline sequence, such as only mutated residues found within 8 amino acids of FR1 or within the last 8 amino acids of FR4, or only mutated residues found within CDR1, CDR2, or CDR3 . In other embodiments, one or more of the framework and/or CDR residues are mutated with the corresponding residue(s) of a different germline sequence (i.e., a germline sequence that differs from the germline sequence from which the antibody originates).

In addition, the bispecific antigen-binding molecules or antigen-binding domains (D1 and/or D2) thereof of the present disclosure may involve any combination of two or more germline mutations within the framework and/or CDR regions, for example, where certain individual residues are mutated to the corresponding residue of a particular germline sequence, while certain other residues that differ from those in the original germline sequence are retained or mutated to the corresponding residue of a different germline sequence. After receiving the bis

- 14 042597 specific antigen-binding molecules or their antigen-binding domains (D1 and/or D2), which provide one or more germline mutations, can be easily tested for one or more desired properties, such as improved binding specificity, increased binding affinity, improved or enhanced antagonistic or agonistic biological properties (as appropriate), reduced immunogenicity, etc. Bispecific antigen-binding molecules or their antigen-binding domains (D1 and/or D2) obtained in such a general way, are covered by the scope of this disclosure.

Options.

The invention also includes anti-MET antibodies and bispecific antigen-binding molecules containing variants of any of the HCVR, LCVR and/or CDR amino acid sequences disclosed herein. Illustrative variants included in this aspect include variants of any of the HCVR, LCVR and/or CDR amino acid sequences disclosed herein with one or more conservative substitutions. For example, the present disclosure includes anti-MET antibodies and MET x MET bispecific antigen binding molecules having the amino acid sequences HCVR, LCVR, and/or CDR, e.g., 10 or less, 8 or less, 6 or less, 4 or less, etc. conservative amino acid substitutions compared to any of the amino acid sequences of HCVR, LCVR and/or CDR presented in table. 1 in this document.

Exemplary variants included in this aspect of the present disclosure also include variants having a significant degree of sequence identity with any of the HCVR, LCVR and/or CDR amino acid sequences disclosed herein. In the context of this document, with regard to amino acid sequences, the term significant identity or substantially identical means that two amino acid sequences, when optimally aligned, for example, using the GAP or BESTFIT programs using standard gap opening penalties, are characterized by at least 95 , 98 or 99% sequence identity. In certain embodiments, positions of residues that are not identical are distinguished by conservative amino acid substitutions. A conservative amino acid substitution is a substitution in which an amino acid residue is replaced with another amino acid residue with a side chain (R group) with similar chemical properties (eg, charge or hydrophobicity). In general, a conservative amino acid substitution will not significantly alter the functional properties of a protein. In cases where two or more amino acid sequences differ from each other by conservative substitutions, the percent sequence identity or degree of similarity can be increased to correct for the conservative nature of the substitution. Methods for making such adjustments are well known to the person skilled in the art. See, for example, Pearson (1994) Methods Mol. Biol. 24: 307331, incorporated herein by reference. Examples of groups of amino acids that have side chains with similar chemical properties include (1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine; (2) aliphatic side chains with a hydroxyl group: serine and threonine; (3) amide-containing side chains: asparagine and glutamine; (4) aromatic side chains: phenylalanine, tyrosine and tryptophan; (5) main side chains: lysine, arginine and histidine; (6) acidic side chains: aspartate and glutamate; and (7) sulfur side chains: cysteine and methionine. Preferred groups for conservative amino acid substitutions are: valine-leucine isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamate-aspartate, and asparagine-glutamine. Alternatively, a conservative substitution is any change with a positive value in the PAM250 log probability table disclosed in Gonnet et al. (1992) Science 256: 1443-1445, incorporated herein by reference. Average conservative substitution is any change with a non-negative value in the PAM250 log probability table.

Sequence identity between two different amino acid sequences is typically determined using sequence analysis software. Sequence analysis software allows similar sequences to be matched by similarity, taking into account various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, the GCG software contains programs such as GAP and BESTFIT that can be used using default parameters to determine sequence homology or sequence identity between closely related polypeptides, such as homologous polypeptides from different species, or between a wild-type protein and its mutant variant. See, for example, GCG version 6.1. Polypeptide sequences can also be compared using FASTA using the default or recommended parameters of the program in GCG version 6.1. FASTA (eg, FASTA2 and FASTA3) provides alignments and percent sequence identity of regions of best overlap between given and target sequences (Pearson (2000), supra). Another preferred algorithm when comparing the present to

- 15 042597 A sequence document with a database containing a plurality of sequences from different organisms is a BLAST computer program, in particular BLASTP or TBLASTN, using default parameters. See, for example, Altschul et al. (1990) J. Mol. Biol. 215:403-410 and Altschul et al. (1997) Nucleic Acids Res. 25:3389-402, each of which is incorporated herein by reference.

Anti-MET antibodies and MET x MET bispecific antigen-binding molecules containing Fc variants.

In accordance with certain embodiments provided herein, anti-MET antibodies and MET x MET bispecific antigen-binding proteins are provided that contain an Fc domain containing one or more mutations that increase or decrease antibody binding to the FcRn receptor, e.g., at acidic pH, compared to neutral pH. For example, the present disclosure includes anti-MET antibodies and MET x MET bispecific antigen-binding proteins involving a mutation in the CH2 or _CH3 region of the Fc domain, where the mutation(s) increase(s) the affinity of the Fc domain for FcRn in an acidic environment ( for example, in the endosome, where the pH is in the range from about 5.5 to about 6.0). Such mutations can lead to an increase in the serum half-life of the antibody when administered to an animal. Non-limiting examples of such Fc modifications include, for example, a modification at position 250 (eg, E or Q); 250 and 428 (for example, L or F); 252 (eg L/Y/F/W or T), 254 (eg S or T), and 256 (eg S/R/Q/E/D or T); or a modification at position 428 and/or 433 (eg H/L/R/S/P/Q or K) and/or 434 (eg H/F or Y); or a modification at position 250 and/or 428; or a modification at position 307 or 308 (eg, 308F, V308F) and 434. In one embodiment, the modification includes modification to 428L (eg, M428L) and 434S (eg, N434S); modification 428L, 259I (for example, V259I) and 308F (for example, V308F); modification 433K (for example, H433K) and modification 434 (for example, 434Y); modification 252, 254 and 256 (for example, 252Y, 254T and 256E); modification 250Q and 428L (for example, T250Q and M428L); and modification 307 and/or 308 (eg, 308F or 308P).

For example, the present disclosure includes anti-MET antibodies and MET x MET bispecific antigen-binding proteins containing an Fc domain containing one or more pairs or groups of mutations selected from the group consisting of: 250Q and 248L (eg, T250Q and M248L); 252Y, 254T and 256E (for example, M252Y, S254T and T256E); 428L and 434S (eg M428L and N434S); and 433K and 434F (eg H433K and N434F). All possible combinations of the aforementioned Fc domain mutations and other mutations within the antibody variable domains disclosed herein are included within the scope of this disclosure.

Biological characteristics of the antigen-binding molecules presented herein.

Provided herein are antibodies and antigen-binding fragments thereof that bind monomeric human MET with high affinity. For example, the present disclosure includes anti-MET antibodies that bind monomeric human MET (e.g., hMET.mmh) with a KD of less than about 230 nM as measured by surface plasmon resonance at 25°C or 37°C, for example, using the assay format defined in Example 3 herein, or a substantially similar assay. In certain embodiments, anti-MET antibodies are provided that bind monomeric human MET at 37° C. with a KD of less than about 230 nM, less than about 200 nM, less than about 150 nM, less than about 100 nM, less than about 50 nM, less than about 25 nM, less than about 20 nM, less than about 10 nM, less than about 8 nM, less than about 6 nM, less than about 5 nM, less than about 4 nM, or less than about 3 nM, as measured by surface plasmon resonance, for example, using the assay format defined in Example 3 herein, or a substantially similar assay.

The present disclosure also includes antibodies and antigen-binding fragments thereof that bind monomeric human MET (e.g., hMET.mmh) with a dissociation half-life (t1/2) of greater than about 1 min, as measured by surface plasmon resonance at 25°C or 37 °C, for example, using the analysis format defined in Example 3 herein, or substantially similar analysis. According to certain embodiments, anti-MET antibodies are provided that bind monomeric human MET at 37° C. with a t1/2 of greater than about 1 minute, greater than about 2 minutes, greater than about 4 minutes, greater than about 6 minutes, greater than more than about 8 minutes, more than about 10 minutes, more than about 12 minutes, more than about 14 minutes, more than about 16 minutes, more than about 18 minutes, or more than about 20 minutes or longer, as measured by the surface plasmon resonance method , for example, using the analysis format defined in Example 3 herein, or a substantially similar analysis.

This document provides antibodies and their antigen-binding fragments that bind

- 16 042597 is a dimeric human MET (eg hMET.mFc) with high affinity. For example, the present disclosure includes anti-MET antibodies that bind dimeric human MET with a K _D of less than about 3 nM as measured by surface plasmon resonance at 25°C or 37°C, e.g. using the assay format defined in Example 3 herein, or a largely similar analysis. In certain embodiments, anti-MET antibodies are provided that bind dimeric human MET at 37° C. with a KD of less than about 3 nM, less than about 2 nM, less than about 1 nM, less than about 0.9 nM, less than about 0.8 nM, less than about 0.7 nM, less than about 0.6 nM, less than about 0.5 nM, less than about 0.4 nM, less than about 0.3 nM, or less than about 0 .25 nM as measured by surface plasmon resonance, for example using the assay format defined in Example 3 herein, or a substantially similar assay.

Also provided herein are antibodies and antigen-binding fragments thereof that bind dimeric human MET (e.g., hMET.mFc) with a dissociation half-life (t1/2) of greater than about 4 min as measured by surface plasmon resonance at 25 or 37° C, for example using the analysis format defined in Example 3 herein, or a substantially similar analysis. According to certain embodiments, anti-MET antibodies are provided that bind dimeric human MET at 37° C. with a t1/2 of greater than about 4 minutes, greater than about 5 minutes, greater than about 10 minutes, greater than about 20 minutes, greater than more than about 30 minutes, more than about 40 minutes, more than about 50 minutes, more than about 60 minutes, more than about 70 minutes, more than about 80 minutes, more than about 90 minutes, more than about 100 minutes, more than about 105 minutes or longer as measured by surface plasmon resonance, for example, using the assay format defined in Example 3 herein, or a substantially similar assay.

Also provided herein are MET x MET bispecific antigen-binding proteins that bind dimeric human MET (e.g., hMET.mFc) with a dissociation half-life (t1/2) of greater than about 10 min as measured by surface plasmon resonance at 25 or 37 °C, for example using the assay format defined in Example 5 herein, or a substantially similar assay. In certain embodiments, MET x MET bispecific antigen-binding proteins are provided that bind dimeric human MET at 37° C. with a t1/2 of greater than about 10 minutes, greater than about 20 minutes, greater than about 30 minutes, greater than about 40 minutes. more than about 50 minutes, more than about 60 minutes, more than about 70 minutes, more than about 80 minutes, more than about 90 minutes, more than about 100 minutes, more than about 200 minutes, more than about 300 minutes, more more than about 400 minutes, more than about 500 minutes, more than about 600 minutes, more than about 700 minutes, more than about 800 minutes, more than about 900 minutes, more than about 1000 minutes, more than about 1100 minutes, or longer, as measured using the surface plasmon resonance method, for example, using the analysis format defined in Example 5 herein, or a substantially similar analysis.

Also provided herein are anti-MET antibodies and MET x MET bispecific antigen-binding proteins that block the interaction between HGF and MET, for example, in an in vitro ligand binding assay. In accordance with certain embodiments provided herein, MET x MET bispecific antigen-binding proteins are provided that block HGF binding to human MET-expressing cells and induce minimal or no MET activation, in the absence of HGF signaling. For example, the present disclosure relates to MET x MET bispecific antigen binding proteins that have a degree of MET agonist activity in a cellular MET activity assay using a reporter system that is less than 50%, less than 40%, less than 30%, less than 20%, less than 10%, less than 5%, less than 3%, less than 2%, or less than 1% MET agonist activity observed in an equivalent activity assay using a reporter system using a monospecific antibody containing D1 or D2 alone.

The antibodies and antigen-binding proteins of the present disclosure may have one or more of the aforementioned biological characteristics, or any combination thereof. The above list of biological characteristics of antibodies is not exhaustive. Other biological characteristics of the antibodies provided herein will be apparent to those skilled in the art from the overview of the present disclosure, including the illustrative examples herein.

Antibody drug conjugates (ADC).

This document provides antibody drug conjugates (ADC)

- 17 042597 containing an anti-MET antibody or a MET x MET bispecific antigen-binding protein conjugated to a therapeutic moiety such as a cytotoxic agent, a chemotherapeutic drug or a radioisotope.

Cytotoxic agents include any agent that adversely affects cell growth, viability, or proliferation, including, without limitation, tubulin-interacting agents and DNA-damaging agents. Examples of suitable cytotoxic agents and chemotherapeutic agents that can be conjugated with anti-MET antibodies in accordance with this aspect of the present disclosure include, for example, 1-(2-chloroethyl)-1,2-dimethanesulfonyl hydrazine, 1,8dihydroxy-bicyclo[7.3. 1] trideca-4,9-dien-2,6-diin-13-one, 1-dihydrotestosterone, 5-fluorouracil, 6-mercaptopurine, 6-thioguanine, 9-aminocamptothecin, actinomycin D, amanitins, aminopterin, anvidin, anthracycline, anthramycin (AMS), auristatins, bleomycin, busulfan, butyric acid, calicheamycins (eg, calicheamycin γ ₁ ), camptothecin, carminomycins, carmustine, cemadotins, cisplatin, colchicine, combretastatins, cyclophosphamide, cytarabine, cytochalasin B, dactinomycin, daunorubicin, decarbazine, diacetoxypentyldoxorubicin , dibromomannite, dihydroxyanthracindione, disorazoles, dolastatin (eg, dolastatin 10), doxorubicin, duocarmycin, echinomycins, eleutherobins, emetine, epothilones, esperamycin, estramustines, ethidium bromide, etoposide, fluorouracils, geldanamycins, gramicidin D, verecorticoids, irinotecane inhibitors (KSP), leptomycins, leurosins, lidocaine, lomustine (CCNU), maytansinoids, chlormethine hydrochloride, melphalan, mercaptopurines, metopterins, methotrexate, mithramycin, mitomycin, mitoxantrone, N8acetylspermidine, podophyllotoxins, procaine, propranolol, pteridines, puromycin, pyrrolobenzodiazepines (BDP ), rhizoxins, streptozotocin, thallisomycins, taxol, tenoposide, tetracaine, thioepu chlorambucil, tomaimycins, topotecans, tubulizin, vinblastine, vincristine, vindesine, vinorelbines and derivatives of any of the above. In certain embodiments, the cytotoxic agent that is conjugated to the anti-MET antibody is a maytansinoid such as DM1 or DM4, a tomaimycin derivative, or a dolastatin derivative. In certain embodiments, the cytotoxic agent that is conjugated to the anti-MET antibody is auristatin, such as MMAE, MMAF, or derivatives thereof. Other cytotoxic agents known in the art are included within the scope of this disclosure, including, for example, protein toxins such as ricin, C. difficile toxin, pseudomonas exotoxin, ricin, diphtheria toxin, botulinum toxin, bryodin, saporine, toxins of the American pokeweed (i.e. . phytolacca toxin and phytolaccagenin) and others, such as those set forth in Sapra et al., Pharmacol. & Therapeutics, 2013, 138:452-469.

In certain embodiments, the cytotoxic agent is a maytansinoid, such as a maytansine derivative. Suitable maytansinoids include DM1, DM4 or derivatives, stereoisomers or isotopologues thereof. Suitable maytansinoids also include, without limitation, those disclosed in WO 2014/145090 A1, WO 2015/031396 A1, US 2016/0375147 A1 and US 2017/0209591 A1, incorporated herein by reference in its entirety.

In some embodiments, the maytansinoid has the following structure:

where A is an optionally substituted arylene or heteroarylene. In some embodiments, the maytansinoid has the following structure:

where A is an optionally substituted arylene or heteroarylene. In some embodiments, the maytansinoid has the following structure:

- 18 042597

where n is an integer from 1 to 12 and R ¹ is alkyl. In some embodiments, the maytansinoid is

- 19 042597

- 20 042597

- 21 042597

or

In some embodiments, the maytansinoid is

In some embodiments, the maytansinoid is

Also provided herein are antibody-radionuclide conjugates (ARC) containing anti-MET antibodies conjugated to one or more radionuclides. Illustrative radionuclides that can be used in the context of this aspect of the present disclosure include, without limitation, for example, ²²⁵ Ac, ²¹² Bi, ²¹³ Bi, ¹³¹ I, ¹⁸⁶ Re, ²²⁷ Th, ²²² Rn, ²²³ Ra, ²²⁴ Ra, and ⁹⁰ Y.

Certain embodiments provided herein provide ADCs comprising an anti-MET antibody or a MET x MET bispecific antigen-binding protein conjugated to a cytotoxic agent (eg, any of the cytotoxic agents disclosed above) via a linker molecule. Linkers are any group or moiety that connects, attaches or binds the antibodies described herein.

- 22 042597 lo or antigen-binding proteins with a therapeutic component, for example a cytotoxic agent. Suitable linkers can be found, for example, in Antibody-Drug Conjugates and Immunotoxins; Phillips G.L., Ed.; Springer Verlag: New York, 2013; Antibody-Drug Conjugates; Ducry, L., Ed.; Humana Press, 2013; Antibody-Drug Conjugates; Wang J., Shen W.-C., and Zaro J.L., Eds.; Springer International Publishing, 2015, the contents of which are incorporated herein by reference in their entirety. In general, suitable binding agent linkers for the antibody-based conjugates described herein are those that are stable enough to utilize the circulating half-life of the antibody and yet capable of releasing their load following antigen-mediated internalization of the conjugate. Linkers may be cleavable or non-cleavable. Cleavable linkers include linkers that are cleaved by intracellular metabolic processes after internalization, for example, cleaved by hydrolysis, reduction or enzymatic reaction. Non-cleavable linkers include linkers that release the attached load through lysosomal degradation of the antibody after internalization. Suitable linkers include, without limitation, acid-labile linkers, hydrolysis-labile linkers, enzyme-cleavable linkers, reduction-labile linkers, self-cleaving linkers, and non-cleavable linkers. Suitable linkers also include, without limitation, those that are peptides, glucuronides, succinimide thioethers, polyethylene glycol (PEG) units, hydrazones, malcaproyl units, dipeptide units, valine-citrulline units, and para-aminobenzyl (PAB) units, or which contain .

Any linker molecule or linker technology known in the art can be used to create or construct the ADC of the present disclosure. In certain embodiments, the linker is a cleavable linker. In accordance with other embodiments, the linker is a non-cleavable linker. Illustrative linkers that can be used in the context of the present disclosure include linkers that contain, for example, MC (6-maleimidocaproyl), MP (maleimidopropanoyl), val-cit (valinecitrulline), val-ala (valine-alanine), a dipeptide site in protease-cleavable linker, ala-phe (alanine-phenylalanine), dipeptide site in protease-cleavable linker, PAB (p-aminobenzyloxycarbonyl), SPP (N-succinimidyl-4-(2-pyridylthio)pentanoate), SMCC (N-succinimidyl-4 -(Nmaleimidomethyl)cyclohexane-1-carboxylate), SIAB (N-succinimidyl-(4-iodo-acetyl)aminobenzoate) and variants and combinations thereof, or consist of the foregoing. Additional examples of linkers that can be used in the context of this disclosure are provided, for example, in US 7754681 and in Ducry, Bioconjugate Chem., 2010, 27:5-13 and the references cited therein, the contents of which are incorporated herein by reference throughout their completeness.

In certain embodiments, the linkers are stable under physiological conditions. In certain embodiments, the linkers are cleavable, eg, capable of releasing at least a portion of the load in the presence of an enzyme or at a particular pH range or value. In some embodiments, the linker contains an enzyme-cleavable moiety. Illustrative enzyme-cleavable moieties include, without limitation, peptide bonds, ester bonds, hydrazones, and disulfide bonds. In some embodiments, the linker comprises a cathepsin-cleavable linker.

In some embodiments, the linker provides a non-cleavable moiety.

Suitable linkers also include, without limitation, those chemically linked to two cysteine residues of a single binding agent, such as an antibody. Such linkers can serve to mimic the disulfide bonds of an antibody that are broken as a result of the conjugation process.

In some embodiments, the linker provides one or more amino acids. Suitable amino acids include natural, artificial, standard, non-standard, proteinogenic, non-proteinogenic and L- or D-a-amino acids. In some embodiments, the linker contains alanine, valine, glycine, leucine, isoleucine, methionine, tryptophan, phenylalanine, proline, serine, threonine, cysteine, tyrosine, asparagine, glutamine, aspartic acid, glutamic acid, lysine, arginine, histidine, or citrulline, their derivative, or their combination. In certain embodiments, one or more amino acid side chains are linked to the side chain described below. In some embodiments, the linker contains valine and citrulline. In some embodiments, the linker contains lysine, valine, and citrulline. In some embodiments, the linker contains lysine, valine, and alanine. In some embodiments, the linker contains valine and alanine.

In some embodiments, the linker contains a self-cleaving group. The self-cleaving group may be any such group known to those skilled in the art. According to specific embodiments, the self-cleaving group is

- 23 042597 is p-aminobenzyl (PAB) or a derivative thereof. Suitable derivatives include p aminobenzyloxycarbonyl (PABC). Those skilled in the art will appreciate that a self-cleaving group is capable of carrying out a chemical reaction that releases the remaining linker atoms from the load.

In some embodiments, the linker is

Where

A is bonded to an antibody or antigen-binding protein (eg via a lysine residue) and is bonded to a cytotoxic agent (eg DM1). In some embodiments, the linker is

A O

Oh o _R

V-N' 1

W o’ where

A is a link to an antibody or antigen-binding protein (for example, via a lysine residue) and । P' is a link to a cytotoxic agent (eg DM1). In certain embodiments, the linker is

In certain embodiments, the linker is

A O

In some embodiments, the linker is derived from maleimidylmethyl 4-transcyclohexanecarboxysuccinate

- 24 042597

In some embodiments, the linker is

Where

is a link to an antibody or antigen-binding protein (eg, via a lysine residue) and

is a link with a cytotoxic agent (for example, a compound characterized by the following formula:

The present disclosure provides for ADCs in which a linker connects an anti-MET antibody or a MET x MET bispecific antigen-binding protein to a drug or cytotoxin via attachment at a specific amino acid within the antibody or antigen-binding molecule. Exemplary amino acid attachments that can be used in the context of this aspect include, for example, lysine (see, for example, US 5208020; US 2010/0129314; Hollander et al, Bioconjugate Chem., 2008, 19:358-361; WO 2005 /089808; US 5714586; US 2013/0101546; and US 2012/0585592), cysteine (see, for example, US 2007/0258987; WO 2013/055993; WO 2013/055990; WO 2013/053873; WO 2013; WO 2011/130598; US 2013/0101546; and US 7750116), selenocysteine (see e.g. WO 2008/122039; and Hofer et al, Proc. Natl. Acad. Sci., USA, 2008, 105:12451-12456 ), formylglycine (see e.g. Carrico et al, Nat. Chem. Biol., 2007, 3:321-322; Agarwal et al, Proc. Natl. Acad. Sci., USA, 2013, 110:46-51 , and Rabuka et al, Nat. Protocols, 2012, 10:1052-1067), artificial amino acids (see, for example, WO 2013/068874 and WO 2012/166559) and acidic amino acids (see, for example, WO 2012/05982 ). Linkers can also be conjugated to an antigen binding protein by attachment to carbohydrates (see, for example, US 2008/0305497, WO 2014/065661 and Ryan et al, Food & Agriculture Immunol., 2001, 13:127-130) and disulfide linkers ( see, for example, WO 2013/085925, WO 2010/010324, WO 2011/018611 and Shaunak et al, Nat Chem Biol 2006 2:312-313). For direct conjugation to specific antibody or antigen-binding protein residues, site-specific conjugation techniques can also be used (see, for example, Schumacher et al. J. Clin. Immunol. (2016) 36 (Suppl 1): 100). Site-specific conjugation techniques include, but are not limited to, glutamine conjugation with transglutaminase (see, for example, Schibli, Angew Chemie Inter Ed. 2010, 49, 9995).

In certain embodiments, the present disclosure relates to an ADC wherein an anti-MET antibody or a MET x MET bispecific antigen-binding protein as described herein is conjugated to a linker-drug composition as set forth in International Patent Application Publication WO 2014/145090 , (for example, compound 7, also referred to herein as M0026 and shown below), the disclosure of which is incorporated herein by reference in its entirety:

- 25 042597

Also provided herein are antibody-drug conjugates comprising monospecific anti-MET antibodies and bispecific MET x MET antibodies disclosed herein, wherein said anti-MET antibody or bispecific MET x MET antibody is conjugated to a cytotoxic agent. In certain embodiments, the cytotoxic agent is a maytansinoid. In certain embodiments, the maytansinoid is a compound having the following formula:

₀ HON ^OCH ^ f G ''Ί °h _sn , o, I o _k - ' 11

ZhS! N ₇ DOS ₃

CH ₃ O H ₃ cCI, ΗN X < N-OH, - f V Ό

O CH ₃ where η is an integer from 1 to 12 and R ¹ is alkyl. In certain embodiments, the maytansinoid is

In certain embodiments, the cytotoxic agent is a maytansinoid, wherein the maytansinoid is covalently linked to the antibody via a non-cleavable linker. In certain embodiments, the cytotoxic agent is a maytansinoid, wherein the maytansinoid is covalently linked to the antibody via a cleavable linker. In one embodiment, the antibody is conjugated to

where is the link to the antibody.

Where

In one embodiment, the antibody is conjugated to

-26042597 where

c is the link to the antibody.

In one embodiment, the antibody is conjugated

c is the link to the antibody.

In one embodiment, the antibody is conjugated

where is the link to the antibody.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is an anti-MET antibody or the bispecific antigen-binding protein MET x MET described herein;

L is a linker;

Pay is a cytotoxic agent; and n is an integer from 1 to 10.

In some embodiments, Ab is an anti-MET antibody comprising a CDR within the HCVR/LCVR amino acid sequence pair SEQ ID NO: 82/138. In some embodiments, Ab is an anti-MET antibody comprising the amino acid sequence of HCVR SEQ ID NO: 82 and the amino acid sequence of LCVR SEQ ID NO: 138.

In some embodiments, Ab is a MET x MET bispecific antigen binding protein comprising a CDR within the D1HCVR amino acid sequence of SEQ ID NO: 58 and a CDR within the D2-HCVR amino acid sequence of SEQ ID NO: 82. In some aspects, the MET x MET bispecific antigen binding protein further comprises a CDR within the amino acid sequence of LCVR SEQ ID NO: 138. In some embodiments, Ab is a MET x MET bispecific antigen-binding protein comprising the amino acid sequence of D1-HCVR of SEQ ID NO: 58 and the amino acid sequence of D2-HCVR of SEQ ID NO: 82 In some aspects, the MET x MET bispecific antigen binding protein further comprises the amino acid sequence LCVR of SEQ ID NO: 138.

In some embodiments, Ab is a MET x MET bispecific antigen binding protein comprising a CDR within the D1HCVR amino acid sequence of SEQ ID NO: 18 and a CDR within the D2-HCVR amino acid sequence of SEQ ID NO: 82. In some aspects, the MET x MET bispecific antigen binding protein further comprises a CDR within the amino acid sequence of LCVR SEQ ID NO: 138. In some embodiments, Ab is a MET x MET bispecific antigen-binding protein comprising the amino acid sequence D1-HCVR of SEQ ID NO: 18 and am

- 27 042597 no acid sequence D2-HCVR SEQ ID NO: 82.

In some embodiments, L is a cleavable linker. In some embodiments, L is a non-cleavable linker. In some embodiments, L provides a dipeptide. In some embodiments, L provides a PAB fragment.

In some embodiments, L provides for a fragment characterized by the following structure:

In some embodiments, L provides for a fragment characterized by the following structure:

In some embodiments, L provides for a fragment characterized by the following structure:

In some embodiments, L provides for a fragment characterized by the following structure:

In some embodiments, Pay is a maytansinoid. In some embodiments, Pay is

where R ¹ represents alkyl.

In some embodiments, Pay is

- 28 042597

In some embodiments, Pay is

In some embodiments, n is an integer from 2 to 5. In some embodiments, -L-Pay is where

is bound to the antibody.

In some embodiments, -L-Pay is where

is bound to the antibody.

In some embodiments, -L-Pay is

Where

is bound to the antibody.

- 29 042597 where

In some embodiments, -L-Pay is

is bound to the antibody.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay] η, where Ab is an antibody to MET containing the amino acid sequence of HCVR SEQ ID NO: 82 and the amino acid sequence of LCVR SEQ ID NO: 138;

L-Pay is where

is the link with the antibody; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is an antibody to MET containing the amino acid sequence of HCVR SEQ ID NO: 82 and the amino acid sequence of LCVR SEQ ID NO: 138;

L-Pay is where

is the link with the antibody; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is an anti-MET antibody containing the amino acid sequence of HCVR SEQ ID NO: 82 and the amino acid sequence of LCVR SEQ ID NO: 138;

L-Pay is

- 30 042597

Where

is the link with the antibody; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is an antibody to MET containing the amino acid sequence of HCVR SEQ ID NO: 82 and the amino acid sequence of LCVR SEQ ID NO: 138;

L-Pay is

Where

is the link with the antibody; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay] _n where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 58 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

where is the relationship with the antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay] _n where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 58 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

- 31 042597 where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 58 and the amino acid sequence of D2-HCVR SEQ ID NO: 82; L-Pay is

Where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is a MET x MET bispecific antigen-binding protein containing the amino acid sequence of D1-HCVR SEQ ID NO: 58 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

Where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay] _n where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 18 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

- 32 042597

Where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay]n, where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 18 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

Where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab- [L-Pay] η, where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 18 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

L-Pay is

where is the relationship with the antigen-binding protein; and n is an integer from 2 to 5.

In some embodiments, the conjugates have the following structure:

Ab-[L-Pay] n where Ab is a bispecific antigen-binding protein MET x MET containing the amino acid sequence of D1-HCVR SEQ ID NO: 18 and the amino acid sequence of D2-HCVR SEQ ID NO: 82;

- 33 042597

L-Pay is

Where

is a link to an antigen-binding protein; and n is an integer from 2 to 5.

The antibody-drug conjugates described herein can be prepared using conjugation conditions known to those skilled in the art (see, for example, Doronina et al. Nature Biotechnology 2003, 21, 7, 778, which is incorporated herein by reference throughout its fullness). In some embodiments, an anti-MET antibody or MET x MET bispecific antigen-binding protein antibody-drug conjugate is prepared by contacting an anti-MET antibody or MET x MET bispecific antigen-binding protein described herein with a compound containing the desired linker and cytotoxic agent, where the specified linker contains a fragment that is reactive against the antibody or antigen-binding protein, for example, against the desired residue of the antibody or antigen-binding protein.

In some embodiments, provided herein are methods for preparing an antibody-drug conjugate comprising contacting an anti-MET antibody or MET x MET bispecific antigen-binding protein described herein with a compound of the following formula A ¹ :

and water diluent.

In some embodiments, the compound of formula A ¹ is present in a stoichiometric excess. In some embodiments, the compound of formula A ¹ is present in a 5-6 fold stoichiometric excess. In some embodiments, the aqueous diluent comprises HEPES. In some embodiments, the aqueous diluent includes DMA.

In some embodiments, a compound of formula A ¹ is a compound of formula A ² or A ³ :

- 34 042597

In some embodiments, the compound of formula A ² is stoichiometrically pure A ³ . In some embodiments, a compound of formula A ¹ provides a compound of formula A ¹ or A ² , wherein the compound A ¹ or A ² is present in a diastereomeric excess of greater than 50%. In certain embodiments, the diastereomeric excess is greater than 70%. In certain embodiments, the diastereomeric excess is greater than 90%. In certain embodiments, the diastereomeric excess is greater than 95%.

The term diastereomeric excess refers to the difference between the mole fraction of the desired one diastereoisomer and the remaining diastereoisomers in the composition. The diastereomeric excess is calculated as follows:

(number of one diastereoisomer)-(number of other diastereoisomers)/1.

For example, a composition that contains 90% 1 and 10% 2, 3, 4 or mixtures thereof has a diastereomeric excess of 80% [(90-10)/1]. A composition that contains 95% 1 and 5% 2, 3, 4 or mixtures thereof is characterized by a diastereomeric excess of 90% [(95-5)/1]. A composition that contains 99% 1 and 1% 2, 3, 4 or mixtures thereof is characterized by a diastereomeric excess of 98% [(99-1)/1]. The diastereomeric excess can similarly be calculated for any of 1, 2, 3, or 4.

According to some embodiments, the compound of formula A ¹ is obtained by bringing the compound of formula (a):

in contact with a compound of formula (b)

in the presence of silica gel and diluent. In some embodiments, the diluent comprises an organic solvent and water.

This document also presents a product obtained using the following method:

(i) bringing the compound of formula (a):

- 35 042597

(a) in contact with a compound of formula (b):

ABOUT

(b) in the presence of silica gel and a diluent to synthesize an intermediate; and (ii) contacting an anti-MET antibody or MET x MET bispecific antigen-binding protein described herein with an intermediate and an aqueous diluent.

In some embodiments, provided herein are methods for preparing an antibody-drug conjugate comprising contacting an anti-MET antibody or a MET x MET bispecific antigen-binding protein described herein with a compound of the following formula B:

In where LG is a leaving group, and an aqueous diluent.

In some embodiments, the compound of formula B is present in a stoichiometric excess. In some embodiments, the compound of formula B is present in a 5-6 fold stoichiometric excess. In some embodiments, the aqueous diluent comprises HEPES. In some embodiments, the aqueous diluent includes DMA. In some embodiments, -C(O)-LG is an ester, such as NHS- or trifluorophenyl ether.

In some embodiments, the compound of formula B is a compound of formula B ¹ :

At ¹ .

In some embodiments, a compound of formula B ¹ is prepared by bringing a compound of formula C

- 36 042597

in contact with N-hydroxysuccinimide (NHS), a peptide bond condensing reagent, and an organic diluent. Suitable peptide bonding reagents include those that activate, i. make reactive, carboxylic acid fragments to be able to react with the nucleophile. In certain embodiments, the peptide bond forming reagent is N-(3-dimethylaminopropyl)-N'ethylcarbodiimide hydrochloride (EDC). In some embodiments, the organic solvent is dichloromethane.

In some embodiments, a compound of formula C is prepared by bringing a compound of formula D

in contact with adipic acid, a peptide bonding agent and an organic solvent. In certain embodiments, the peptide bonding reagent is 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline (EEDQ). In certain embodiments, the organic solvent comprises dichloromethane. Compound D can be prepared as described in WO 2014/145090.

Epitope mapping and related technologies.

The epitope to which antibodies and antigen-binding domains bind may be one contiguous sequence of 3 or more (e.g., 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 , 17, 18, 19, 20 or more) amino acids of the MET protein. Alternatively, the corresponding epitope may consist of a plurality of discontinuous amino acids (or amino acid sequences) of MET. In some embodiments, the epitope is located on or near the ligand-binding domain of MET. In other embodiments, the epitope is located outside the ligand-binding domain of the MET, for example, at a location on the surface of the MET where the antibody, when bound to such an epitope, does not prevent HGF from binding to the MET.

As described elsewhere herein, the individual antigen-binding domains (D1 and D2) of the MET x MET bispecific antigen-binding molecules can bind to distinct, or non-overlapping, or partially overlapping epitopes relative to each other. As used herein, the expression partially overlapping epitopes means that the first and second epitopes have less than 5, less than 4, less than 3, or only one amino acid in common, as determined by any epitope mapping technique known in the art (e.g., X-ray diffraction analysis, alanine scanning mutagenesis). , hydrogen-deuterium exchange [HDX], domain exchange, etc.). Domains D1 and D2 may be non-competing with each other. For example, in certain embodiments, binding of the D1 domain of a particular MET x MET bispecific antigen binding molecule to its epitope on MET does not (or only minimally) inhibits binding of the D2 domain of the MET x MET bispecific antigen binding molecule to its epitope on MET. Due to the non-overlapping (or, for the most part, partially overlapping) nature of the respective epitopes of the D1 and D2 components, bispecific antigen-binding molecules MET x MET are able to bind to a single MET molecule on the cell surface.

Various techniques known to those skilled in the art can be used to determine the epitope on the MET with which the antibodies and antigen-binding domains of the present disclosure interact. Exemplary techniques that can be used to determine an epitope or

- 37 042597 binding domain of a particular antibody or antigen binding domain include, for example, point mutagenesis (e.g., alanine scanning mutagenesis, arginine scanning mutagenesis, etc.), peptide blot analysis (Reineke, 2004, Methods Mol. Biol. 248:443-463 ), lipid bilayer-mediated protease protection assay, and peptide cleavage assay. In addition, methods such as the epitope excision method, the epitope isolation method, and the chemical modification of antigens can be used (Tomer, 2000, Protein Science 9:487-496). Another method that can be used to identify amino acids within a polypeptide with which an antibody interacts is hydrogen-deuterium exchange, detected using mass spectrometry. In general terms, the hydrogen deuterium exchange method involves labeling a protein of interest with deuterium, followed by binding of an antibody to the deuterium labeled protein. The protein/antibody complex is then transferred to water to allow hydrogen-deuterium exchange of all residues except for antibody-protected residues (which remain labeled with deuterium). After dissociation of the antibody, the target protein is subjected to protease digestion and mass spectrometric analysis, thus revealing deuterium-labeled residues that correspond to the specific amino acids with which the antibody interacts. See, for example, Ehring (1999) Analytical Biochemistry 267(2):252-259; Engen and Smith (2001) Anal. Chem. 73:256A-265A. X-ray diffraction analysis can also be used to identify amino acids within a polypeptide with which an antibody interacts.

Also provided herein are anti-MET antibodies (including bispecific antibodies) that bind to the same epitope as any of the specific illustrative antibodies or antigen-binding domains described herein (e.g., antibodies containing any of the amino acid sequences shown in Table 1). .1 in this document). Similarly, also provided herein are anti-MET antibodies that compete for binding to MET with any of the specific exemplary antibodies described herein (eg, antibodies comprising any of the amino acid sequences shown in Table 1 herein). In some embodiments, the human MET epitope to which anti-MET antibodies bind comprises amino acids 192-204, amino acids 305-315, and/or amino acids 421-455 of SEQ ID NO: 155. In some embodiments, the first human MET epitope comprises amino acids 192- 204 SEQ ID NO: 155; and the second human MET epitope contains amino acids 305-315 and 421-455 of SEQ ID NO: 155.

Using standard methods known in the art and exemplified herein, it can be readily determined whether an antibody binds to the same epitope as a reference anti-MET antibody or competes with it for binding. For example, to determine if a test antibody binds to the same epitope as the reference anti-MET antibody provided herein, the reference antibody is provided to bind to the MET protein. The ability of the test antibody to bind to the MET molecule is then evaluated. If a test antibody is capable of binding to MET after saturation binding using a reference anti-MET antibody, it can be concluded that the test antibody binds to a different epitope than that of the reference anti-MET antibody. Conversely, if a test antibody is unable to bind to a MET molecule after saturation binding using a reference anti-MET antibody, then the test antibody may bind to the same epitope as the epitope bound by the reference anti-MET antibody. Additional standard experiments (e.g., peptide mutations and binding assays) can then be performed to confirm whether the observed lack of binding of the test antibody is due to binding to the same epitope as the epitope for the reference antibody, or whether steric blocking is responsible for the lack of observed binding ( or some other event). Experiments of this type can be performed using ELISA, RIA, Biacore, flow cytometry, or any other quantitative or qualitative antibody binding assay known in the art. In certain embodiments, two antibodies bind to the same (or overlapping) epitope if, for example, a 1-, 5-, 10-, 20-, or 100-fold excess of one antibody results in inhibition of the binding of the other by at least 50%, and preferably 75, 90 or even 99%, as measured in a competitive binding assay (see, for example, Junghans et al., Cancer Res. 1990:50:1495-1502). Alternatively, two antibodies are said to bind to the same epitope if substantially all amino acid mutations in the antigen that reduce or eliminate binding of one antibody reduce or eliminate binding of the other. Two antibodies are considered to bind overlapping epitopes if only a subset of amino acid mutations that reduce or eliminate binding of one antibody reduce or eliminate binding of the other.

To determine if an antibody competes for binding (or cross-competes for binding) with a reference anti-MET antibody, the bidirectional binding method described above is performed: subsequent evaluation of the binding of the test antibody to the MET molecule. When implementing the method in the second direction, links are provided

- 38 042597 Saturation of the test antibody with the MET molecule, followed by evaluation of the binding of the reference antibody to the MET molecule. If, when the method is carried out in both directions, only the first (saturation binding) antibody is capable of binding to the MET molecule, then it is concluded that the test antibody and the reference antibody compete for binding to the MET. As will be appreciated by one of skill in the art, an antibody that competes for binding with a reference antibody need not bind to the same epitope as the reference antibody, but may sterically block binding of the reference antibody by binding to an overlapping or contiguous epitope.

Obtaining human antibodies.

Anti-MET antibodies and MET x MET bispecific antibodies provided herein may be fully human antibodies. Methods for making monoclonal antibodies, including fully human monoclonal antibodies, are known in the art. Any of such known methods can be used in the context of the present disclosure to generate human antibodies that specifically bind to human METs.

Using, for example, VELOCIMMIINE™ technology, or any other similar known method for producing fully human monoclonal antibodies, high affinity chimeric anti-MET antibodies with a human variable region and a mouse constant region are first isolated. As described in the Experimental section below, antibodies are characterized and selected for desired characteristics, including affinity, ligand blocking activity, selectivity, epitope, and so on. If necessary, the mouse constant regions are replaced with the desired human constant region, eg wild-type IgG1 or IgG4 or modified IgG1 or IgG4, to produce a fully human anti-MET antibody. While the constant region chosen may vary according to the particular application, the characteristics of high affinity antigen binding and target specificity are inherent in the variable region. In certain cases, fully human anti-MET antibodies are isolated directly from antigen-positive B cells.

Bioequivalents.

The anti-MET antibodies and antibody fragments provided herein encompass proteins with amino acid sequences that differ from those of the described antibodies but retain the ability to bind human MET. Such variant antibodies and antibody fragments have one or more amino acid additions, deletions, or substitutions from the original sequence, but have biological activity that is substantially equivalent to that of the described antibodies. Similarly, anti-MET antibody-encoding DNA sequences of the present disclosure encompass sequences that include one or more nucleotide additions, deletions, or substitutions compared to the disclosed sequence, but which encode an anti-MET antibody or antibody fragment that are essentially bioequivalents of an anti-MET antibody. MET or antibody fragment of the present disclosure. Examples of such variant amino acid and DNA sequences are discussed above.

Two antigen-binding proteins or antibodies are considered to be bioequivalent if, for example, they are pharmaceutical equivalents or pharmaceutical variants, the rate and extent of absorption of which is not significantly different when administered at the same molar dose under similar experimental conditions, either as a single dose or in multiple dose form. Some antibodies will be considered equivalents or pharmaceutical options if they are equivalent in their extent of absorption, but not in their rate of absorption, and may still be considered bioequivalent because such differences in absorption rate are intentional and are reflected in the instructions for use of the medicinal product, not are essential to achieve effective drug concentrations in the body, for example, with long-term use, and are not considered medically significant for a particular investigational drug.

In one embodiment, two antigen-binding proteins are considered bioequivalent if there are no clinically significant differences in their safety, purity, and potency.

In one embodiment, two antigen-binding proteins are bioequivalent if the patient can be switched one or more times from treatment with a reference drug to treatment with a biological drug, and vice versa, without an expected increase in the risk of adverse effects, including a clinically significant change in immunogenicity or reduced efficacy, compared with ongoing therapy without such a transfer.

In one embodiment, two antigen-binding proteins are bioequivalent if they both function by a common mechanism or mechanisms of action for a condition or conditions, provided that such mechanisms are known.

Bioequivalence can be demonstrated using in vivo and in vitro methods. The determination of bioequivalence includes, for example, (a) an in vivo test in humans or other mammals, in

- 39 042597 which measure the concentration of an antibody or its metabolites in blood, blood plasma, blood serum or other biological fluid depending on time; (b) an in vitro test that correlates with in vivo bioavailability data in humans and is sufficiently predictive of such data; (c) an in vivo test in humans or other mammals, in which the corresponding immediate pharmacological effect of an antibody (or its target) is measured as a function of time; and (d) a clinical trial under highly controlled conditions that establishes the safety, efficacy, or bioavailability or bioequivalence of the antibody.

Bioequivalent variants of the anti-MET antibodies provided herein can be constructed, for example, by performing various residue or sequence substitutions, or deleting terminal or internal residues or sequences not necessary for biological activity. For example, cysteine residues not essential for biological activity can be removed or replaced with other amino acids to prevent the formation of extra or inappropriate intramolecular disulfide bridges upon renaturation. In other instances, bioequivalent antibodies may include variants of an anti-MET antibody that include amino acid changes that modify the glycosylation characteristics of the antibody, such as mutations that exclude or abolish glycosylation.

Interspecies selectivity and interspecies cross-reactivity.

The present disclosure, in accordance with certain embodiments, relates to anti-MET antibodies (and antigen-binding molecules containing the antigen-binding domains of an anti-MET antibody) that bind to human MET but not to another species of MET. The present disclosure also includes anti-MET antibodies (and antigen-binding molecules containing the antigen-binding domains of an anti-MET antibody) that bind to human METs and to METs of one or more non-human species. For example, anti-MET antibodies and antigen-binding molecules can bind to human METs, and may or may not bind, as appropriate, one or more of the mouse, rat, guinea pig, hamster, gerbil, porcine, cat, dog, rabbit, goat METs. , sheep, cow, horse, camel, cynomolgus monkey, monkey, rhesus monkey or chimpanzee. According to certain exemplary embodiments, anti-MET antibodies and antigen-binding molecules are provided that specifically bind human MET and cynomolgus monkey (ie, Macaca fascicularis) MET. Other anti-MET antibodies and antigen-binding molecules bind human MET but do not bind, or weakly bind, cynomolgus monkey MET.

polyspecific antibodies.

As described elsewhere herein, the present disclosure relates to bispecific antigen-binding molecules containing two different antigen-binding domains, where a first antigen-binding domain (D1) binds a first epitope on MET, and where a second antigen-binding domain (D2) binds a second epitope on MET. In certain embodiments, the first and second epitopes on the MET to which the D1 and D2 domains bind are distinct, or non-overlapping, or partially overlapping. In accordance with this aspect, the D1 domain may contain any of the amino acid sequences HCVR/LCVR or CDR presented in table. 1 herein, and the D2 domain may contain any other of the HCVR/LCVR or CDR amino acid sequences shown in Table 1. 1 herein (provided that the binding specificity of the D1 domain is different from the binding specificity of the D2 domain and/or the antigen-binding protein from which D1 was derived does not compete for MET binding with the antigen-binding protein from which D2 was derived). In some embodiments, the human MET epitope to which anti-MET antibodies bind comprises amino acids 192-204, amino acids 305-315, and/or amino acids 421-455 of SEQ ID NO: 155. In some embodiments, the first human MET epitope comprises amino acids 192- 204 SEQ ID NO: 155; and the second human MET epitope contains amino acids 305-315 and 421-455 of SEQ ID NO: 155.

In accordance with a separate aspect of the present disclosure, conventional bispecific antibodies are also provided, wherein one arm of the bispecific antibody binds to an epitope on human MET and the other arm of the bispecific antibody binds to a second non-MET antigen. The MET binding arm may contain any of the HCVR/LCVR or CDR amino acid sequences shown in Table 1. 1 in this document. In certain embodiments, the MET binding arm binds human MET and blocks the binding of HGF to MET. In other embodiments, the MET binding arm binds human MET but does not block HGF binding to MET.

An exemplary bispecific antibody format that can be used in the context of the present disclosure involves the use of a first C _H 3 immunoglobulin (Ig) domain and a second C _H 3 Ig domain, wherein the first and second Ig C _H 3 domains differ from each other by at least at least one amino acid, and where the difference in at least one amino acid causes a decrease in the ability of the bispecific antibody to bind to protein A compared to a bispecific antibody not characterized by such a difference in amino acid. In one embodiment, the first C _H 3 Ig domain binds protein A, and the second Ig C _H 3 domain contains a mutation that reduces or eliminates the ability to bind to protein A, such as a modification of H95R (according to IMGT-numbering of mutations in exons; H435R according to EU-numbering). The second C _H 3 may further provide for the Y96F modification (as per IMGT; Y436F as per EU). Additional modifications that can be found within the second CH3 include: D16E, L18M, N44S, K52N, V57M and V82I (according to IMGT; D356E, L358M, N384S, K392N, V397M and V422I according to EU) in the case of IgG1 antibodies; N44S, K52N and V82I (IMGT; N384S, K392N and V422I according to EU) in the case of IgG2 antibodies; and Q15R, N44S, K52N, V57M, R69K, E79Q and V82I (as per IMGT; Q355R, N384S, K392N, V397M, R409K, E419Q and V422I as per EU) for IgG4 antibodies. Variations on the bispecific antibody format described above are included within the scope of this disclosure.

Other exemplary bispecific formats that can be used in the context of the present disclosure include, without limitation, for example, scFv-based or diabody bispecific formats, IgG-scFv fusion bispecific formats, Ig dual variable domains (DVD), Quadroma, antibodies tab-to-pit, common light chain antibodies (e.g., tab-to-pit common light chain antibodies, etc.), CrossMab, CrossFab, SEED-derived antibody ((SEED)body), antibody, produced by leucine zipper mediated dimerization, Duobody, IgG1/IgG2, IgG with dual acting Fab (DAF) and Mab ² bispecific formats (see e.g. Klein et al. 2012, mAbs 4:6, 1-11 and cited therein links, for an overview of the formats listed above). Bispecific antibodies can also be constructed using peptide/nucleic acid conjugation, for example, where non-naturally occurring amino acids with cross-reactivity are used to generate site-specific antibody-oligonucleotide conjugates, which then self-assemble into multimeric complexes with defined composition, valence, and geometry. (See, for example, Kazane et al., J. Am. Chem. Soc. [Epub: Dec. 4, 2012]).

Therapeutic composition and administration.

Provided herein are pharmaceutical compositions comprising anti-MET antibodies or MET x MET bispecific antigen-binding molecules of the present invention. Pharmaceutical compositions may be formulated with suitable carriers, excipients, and other means that provide improved transfer, delivery, tolerability, and so on.

Therapeutic applications of antibodies.

Provided herein are methods comprising administering to a subject in need thereof a therapeutic composition comprising an anti-MET antibody or a MET x MET bispecific antigen-binding molecule (e.g., an anti-MET antibody comprising any of the HCVR/LCVR or CDR sequences shown in Table 1 herein). document, or a MET x MET bispecific antigen-binding molecule containing any of the D1 and D2 components shown in Table 5 herein). The therapeutic composition may contain any of the anti-MET antibodies or MET x MET bispecific antigen-binding molecules disclosed herein and a pharmaceutically acceptable carrier or diluent.

Anti-MET antibodies and MET x MET bispecific antigen-binding molecules are useful, inter alia, in the treatment, prevention and/or amelioration of any disease or disorder mediated by or associated with MET expression, signaling or activity, or which is treatable by blocking the interaction between MET and HGF, or by inhibiting MET activity and/or signaling in another way, and/or by promoting receptor internalization and/or reducing the number of receptors on the cell surface.

For example, the anti-MET antibodies and MET x MET bispecific antigen binding molecules of the present disclosure are useful in the treatment of tumors that express (or overexpress) MET. For example, anti-MET antibodies and MET x MET bispecific antigen-binding molecules can be used to treat primary and/or metastatic tumors arising in the brain and meninges, oropharynx, lung and bronchial tree, gastrointestinal tract, male and female reproductive tract, muscle , bones, skin and appendages, connective tissue, spleen, organs of the immune system, blood cells and bone marrow, liver and urinary tract, and specialized sense organs such as the eye. In certain embodiments, anti-MET antibodies and MET x MET bispecific antigen-binding molecules are used to treat one or more of the following cancers: acute myeloid leukemia, adult T-cell leukemia, forms of astrocytoma, bladder cancer, breast cancer, cervical cancer, cholangiocarcinoma , chronic myeloid leukemia, colon and rectal cancer, endometrial cancer, esophageal cancer, gastric cancer (eg, MET amplification gastric cancer), glioblastoma, head and neck cancer (eg, head and neck squamous cell carcinoma [HNSCC]), Kaposi's sarcoma , kidney cancer, forms of leiomyosarcoma, liver cancer, lung cancer (eg, non-small cell lung cancer [NSCLC]), forms of lymphoma, forms of malignant glioma, malignant mesothelioma, melanoma, mesothelioma, MFH/fibrosarcoma, multiple myeloma, nasopharyngeal cancer, osteosarcoma, cancer ovary,

- 41 042597 pancreatic carcinoma, prostate cancer, renal cell carcinoma, rhabdomyosarcoma, small cell lung cancer, synovial sarcoma, thyroid cancer and Wilms tumor.

In the context of the treatments described herein, anti-MET antibodies and MET x MET bispecific antigen-binding molecules can be administered as monotherapy (i.e., as the sole therapeutic agent) or in combination with one or more additional therapeutic agents (examples of which are described elsewhere). of this document).

Combined therapeutic drugs and formulations.

Provided herein are compositions and therapeutic formulations comprising any of the anti-MET antibodies and MET x MET bispecific antigen-binding molecules described herein in combination with one or more additional therapeutically active ingredients, and methods of treatment comprising administering such combinations to those in need thereof. subjects.

Anti-MET antibodies and MET x MET bispecific antigen-binding molecules can be formulated with and/or administered in combination with one or more additional therapeutically active moieties selected from the group consisting of: a MET antagonist (e.g., an anti-MET antibody [e.g., onartuzumab, emibetuzumab and H4H14639D] or a small molecule MET inhibitor), an EGFR antagonist (eg, an anti-EGFR antibody [eg, cetuximab or panitumumab] or a small molecule EGFR inhibitor [eg, gefitinib or erlotinib]), an antagonist of another EGFR family member such as Her2/ErbB2, ErbB3 or ErbB4 (eg, an anti-ErbB2 antibody [eg, trastuzumab or T-DM1 {KADCYLA®}], an anti-ErbB3 or anti-ErbB4 antibody, or a small molecule inhibitor of ErbB2, ErbB3, or ErbB4 activity), an EGFRvIII antagonist (eg, an anti-EGFRvIII antibody), an antagonist IGF1R (eg, anti-IGF1R antibody), B-raf inhibitor (eg, vemurafenib, sorafenib, GDC-0879, PLX-4720), PDGFR-α inhibitor (eg, anti-PDGFR-α antibody), PDGFR-β inhibitor (eg, anti-PDGFR-β antibody or small molecule kinase inhibitor such as, for example, imatinib mesylate or sunitinib malate), PDGF ligand inhibitor (for example, antibody, aptamer, siRNA, etc. binding to PDGF-A, -B, -C or -D), a VEGF antagonist (e.g. VEGF trap protein (VEGFTrap) such as aflibercept, see e.g. US 7087411 (also referred to herein as VEGF inhibitory fusion protein), anti-VEGF antibody (e.g. bevacizumab) , a small molecule kinase inhibitor of the VEGF receptor (e.g. sunitinib, sorafenib or pazopanib)), a DLL4 antagonist (e.g. an anti-DLL4 antibody disclosed in US 2009/0142354 such as REGN421), an Ang2 antagonist (e.g. an anti-Ang2 antibody disclosed in US 2011/0027286, such as H1H685P), FOLH1 antagonist (for example, anti-FOLH1 antibody), STEAP1 or STEAP2 antagonist (for example, anti-STEAP1 antibody or anti-STEAP2 antibody), TMPRSS2 antagonist (for example, anti-TMPRSS2 antibody), MSLN antagonist (for example, anti-MSLN antibody), CA9 antagonist (eg, anti-CA9 antibody), uroplakin antagonist (eg, anti-uroplakin [eg, UPK3A] antibody), MUC16 antagonist (eg, anti-MUC16 antibody), Tn antigen antagonist (eg, anti-Tn ), CLEC12A antagonist (eg, anti-CLEC12A antibody), TNFRSF17 antagonist (eg, anti-TNFRSF17 antibody), LGR5 antagonist (eg, anti-LGR5 antibody), monovalent CD20 antagonist (eg, monovalent anti-CD20 antibody such as rituximab), bispecific antibody to CD20 x CD3, a PD-1 blocking agent (eg, an anti-PD-1 antibody such as pembrolizumab or nivolumab), etc. Other agents that can be successfully administered in combination with the antibodies provided herein include, for example, tamoxifen, aromatase inhibitors, and cytokine inhibitors, including small molecule cytokine inhibitors and antibodies that bind to cytokines such as IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-8, IL9, IL-11, IL-12, IL-13, IL-17, IL-18, or their respective receptors.

To illustrate, a PD-1 inhibitor, such as an anti-PD-1 antibody, can be combined with an anti-Met antibody-drug conjugate described herein. The target patient population includes, in particular, those patients with tumors that overexpress mutant c-Met, for example, a patient with non-small cell lung cancer expressing c-Met.

Provided herein are compositions and therapeutic formulations comprising any of the anti-MET antibodies and MET x MET bispecific antigen-binding molecules described herein in combination with one or more chemotherapeutic agents. Examples of chemotherapeutic agents include alkylating agents such as thiotepa and cyclophosphamide (Cytoxan™); alkylsulfonates such as busulfan, improsulfan and piposulfan; aziridines such as benzodopa, carbokwon, meturedopa and uredopa; ethyleneimines and methylamelamines, including altretamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramide, and trimethylomelamin; nitrogen mustards such as chlorambucil, chlornaphasine, chlorphosphamide, estramustine, ifosfamide, chlormethine, lormethine hydrochloride, melphalan, novembihin, phenesterin, prednimustine, trofosfamide, uracil mustard; nitrosoureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, ranimustine; antibiotics such as aclacinomycins, actinomycin, autramycin, azaserine, bleomycins, cactinomycin, calicheamicin, carabicin, carminomycin, carzinophilin, chromomycins, dactinomycin,

- 42 042597 daunorubicin, detorubicin, 6-diazo-5-oxoT-norleucine, doxorubicin, epirubicin, esorubicin, idarubicin, marcellomycin, mitomycins, mycophenolic acid, nogalamycin, olivomycins, peplomycin, potfiromycin, puromycin, quelamycin, streptonzoigrin, streptonzoigrin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-fluorouracil (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; adrenal depressants such as aminoglutethimide, mitotane, trilostane; a folic acid supplement such as frolinic acid; aceglatone; aldophosphamide glycoside; aminolevulinic acid; amsacrine; bestrabucil; bisantrene; edatraksat; defofamine; demecolcin; diazichon; eflornithine; elliptinium acetate; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidamine; mitoguazone; mitoxantrone; mopidamol; nitracrine; pentostatin; phenamet; pyrarubicin; podophyllic acid; 2-ethylhydrazide; procarbazine; PSK™; razoxane; schizophyllan; spirogermanium; tenuazonic acid; triazichon; 2,2',2-trichlorotriethylamine; urethane; vindesine; dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; hacytosine; arabinoside (AraC); cyclophosphamide; thiotepa; taxanes such as paclitaxel (Taxol™, Bristol-Myers Squibb Oncology, Princeton, NJ) and docetaxel (Taxotere™; Aventis Antony, France); chlorambucil; gemcitabine; 6-thioguanine; mercaptopurine; methotrexate; platinum analogues such as cisplatin and carboplatin; vinblastine; platinum; etoposide (VP-16); ifosfamide; mitomycin C; mitoxantrone; vincristine; vinorelbine; navelbin; novantron; teniposide; daunomycin; aminopterin; xeloda; ibandronate; SRT-11; topoisomerase inhibitor RFS 2000; difluoromethylornithine (DMFO); retinoic acid; esperamycins; capecitabine; and pharmaceutically acceptable salts, acids or derivatives of any of the above. Also included in this definition are antihormonal agents that regulate or inhibit the action of hormones against tumors, such as antiestrogen, including, for example, tamoxifen, raloxifene, 4(5)-imidazoles that inhibit aromatase, 4-hydroxytamoxifen, trioxifene, keoxifene, LY 117018 , onapristone and toremifene (Fareston); and antiandrogens such as flutamide, nilutamide, bicalutamide, leuprolide and goserelin; and pharmaceutically acceptable salts, acids or derivatives of any of the above.

Anti-MET antibodies and MET x MET bispecific antigen-binding molecules can also be administered and/or formulated in combination with antivirals, antibiotics, analgesics, corticosteroids, steroids, oxygen, antioxidants, COX inhibitors, cardioprotectors, chelators, IFN-gamma and/or NSAIDs.

Additional(e) therapeutically active(s) component(s), for example any of the above funds or their derivatives, you can enter immediately before, simultaneously with or immediately after the introduction of antibodies to MET or bispecific antigen-binding molecule MET x MET; (For the purposes of this disclosure, such administration regimens are considered to be administration of the antibody in combination with an additional therapeutically active ingredient). The present disclosure includes pharmaceutical compositions in which an anti-MET antibody or MET x MET bispecific antigen-binding molecule is co-formulated with one or more additional therapeutically active ingredients as described elsewhere herein.

Introduction schemes.

In certain embodiments, multiple doses of an anti-MET antibody or a MET x MET bispecific antigen-binding molecule (or a pharmaceutical composition comprising a combination of an anti-MET antibody or a MET x MET bispecific antigen-binding molecule and any of the additional therapeutically active agents referred to herein) may be administered. subject for a specified period of time. Methods in accordance with this aspect involve sequentially administering to a subject multiple doses of an anti-MET antibody or a MET x MET bispecific antigen-binding molecule as provided herein. As used herein, sequential administration means that each dose of an antibody is administered to a subject at different points in time, eg, on different days separated by a predetermined interval (eg, hours, days, weeks, or months). The present disclosure includes methods that involve sequentially administering to a patient a single initial dose of anti-MET antibody or MET x MET bispecific antigen-binding molecule, followed by one or more second doses of anti-MET antibody or MET x MET bispecific antigen-binding molecule, and optionally followed by one or multiple doses of third line anti-MET antibody or bispecific antigen-binding molecule MET x MET.

The terms initial dose, second line doses, and third line doses refer to the time sequence of administration of the anti-MET antibody or the MET x MET bispecific antigen-binding molecule. Thus, the initial dose is the dose that is administered at the beginning of the course of treatment (also called the initial dose); second-order doses are doses that

- 43 042597 administered after the initial dose; and third line doses are doses that are administered after the second line doses. The initial, second, and third line doses may contain the same amount of anti-MET antibody or MET x MET bispecific antigen-binding molecule, but may generally differ from each other based on the frequency of administration. However, in certain embodiments, the amounts of antibodies contained in the initial, second, and third line doses differ from each other (eg, adjusted upwards or downwards as necessary) over the course of treatment. In certain embodiments, two or more (eg, 2, 3, 4, or 5) doses are administered at the beginning of the course of treatment as loading doses, followed by doses that are administered less frequently (eg, maintenance doses).

Diagnostic applications of antibodies.

The anti-MET antibody or MET x MET bispecific antigen-binding molecule of the present disclosure can also be used to detect and/or measure MET or MET-expressing cells in a sample, for example, for diagnostic purposes. For example, an anti-MET antibody or fragment thereof can be used to diagnose a condition or disease characterized by abnormal expression (eg, overexpression, underexpression, lack of expression, etc.) of MET. Exemplary diagnostic assays for MET may involve, for example, contacting a patient sample with an anti-MET antibody or a MET x MET bispecific antigen-binding molecule, where the antibody is labeled with a detectable label or reporter molecule. Alternatively, an unlabeled anti-MET antibody or a MET x MET bispecific antigen-binding molecule can be used for diagnostic applications in combination with a secondary antibody that is labeled for detection. The detectable label or reporter molecule may be a radioisotope such as ³ H, ¹⁴ C, ³² P, ³⁵ S, or ¹²⁵ I; a fluorescent or chemiluminescent substance such as fluorescein or rhodamine; or an enzyme such as alkaline phosphatase, beta-galactosidase, horseradish peroxidase or luciferase. Specific illustrative assays that can be used to detect or measure MET in a sample include enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), immuno-PET (e.g., ⁸⁹ Zr, ⁶⁴ Cu, etc.), and cell sorting with fluorescence activated (FACS).

Samples that can be used in MET diagnostic assays in accordance with the present disclosure include any tissue or body fluid sample that can be obtained from a patient. In general, MET levels in a particular sample obtained from a healthy patient (eg, a patient who is not suffering from a disease or condition associated with abnormal MET levels or activity) will be measured first to establish a baseline, or standard, MET level. This baseline MET level can then be compared to MET levels measured in samples obtained from individuals suspected of having a MET-associated disease or condition.

Examples

The following examples are provided to provide those skilled in the art with a complete disclosure and description of how to make and use the methods and compositions provided herein, and are not intended to limit the scope of what the present inventors consider to be their invention. Efforts have been made to ensure accuracy with respect to the values used (e.g., amounts, temperature, etc., however, some errors and variations in experiments should be considered. Unless otherwise indicated, parts are parts by weight, molecular weight is average molecular weight , the temperature is given in degrees Celsius and the pressure is atmospheric or near atmospheric.

Example 1. Obtaining antibodies to MET.

Anti-MET antibodies were obtained by immunizing a genetically engineered mouse containing DNA encoding the human immunoglobulin heavy chain and kappa light chain variable regions with an immunogen containing a recombinant construct of a human MET extracellular domain fused to human Fc (R&D Systems, part number 358-MT, Minneapolis, MN). Mice used for immunization express the universal light chain. That is, the antibodies produced in this mouse have different heavy chain variable regions, but nearly identical light chain variable domains.

The humoral immune response was monitored using a MET-specific immunoassay. When the desired immune response was achieved, the splenocytes were collected and fused with mouse myeloma cells to preserve the viability of the former and to obtain hybridoma cell lines. Hybridoma cell lines were screened and selected to identify cell lines that produce MET-specific antibodies. Several chimeric anti-MET antibodies (ie, antibodies having human variable domains and mouse constant domains) have been generated using this technique. In addition, several fully human anti-MET antibodies have been isolated directly from antigen-positive B cells without fusion with myeloma cells, as described in US

- 44 042597

2007/0280945 A1.

Certain biological properties of the exemplary anti-MET antibodies prepared according to the methods in this example, and the bispecific antibodies constructed therefrom, are detailed in the examples below.

Example 2 Amino acid and nucleic acid sequences for heavy and light chain variable regions.

In table. 1 provides the amino acid sequence identifiers for the heavy and light chain variable regions and CDRs of selected anti-MET antibodies described herein. (As noted above, all antibodies prepared in Example 1 have the same light chain variable region, and therefore also the same light chain CDR sequences). The corresponding nucleic acid sequence identifiers are shown in Table 1. 2.

Table 1. Amino acid sequence identifiers

SEQID NO: Antibody designation HCVR HCDR1 HCDR2 HDDR3 LCVR LCDR1 LCDR2 LCDR3 H4N13290R2 2 4 6 8 138 140 142 144 Н4Н13291Р2 10 12 14 16 138 140 142 144 H4N13295R2 18 20 22 24 138 140 142 144 H4N13299R2 26 28 thirty 32 138 140 142 144 H4H13300R2 34 36 38 40 138 140 142 144 Н4Н13301Р2 42 44 46 48 138 140 142 144 Н4Н13302Р2 50 52 54 56 138 140 142 144 Н4Н13306Р2 58 60 62 64 138 140 142 144 H4N13309R2 66 68 70 72 138 140 142 144 Н4Н13311Р2 74 76 78 80 138 140 142 144 Н4Н13312Р2 82 84 86 88 138 140 142 144 H4H13313R2 90 92 94 96 138 140 142 144 H4H13316R2 98 100 102 104 138 140 142 144 Н4Н13318Р2 106 108 110 112 138 140 142 144 H4N13319R2 114 116 118 120 138 140 142 144 Н4Н13325Р2 122 124 126 128 138 140 142 144 Н4Н13331Р2 130 132 134 136 138 140 142 144

- 45 042597

Table 2. Nucleic acid sequence identifiers

SEQID NO: Antibody designation HCVR HCDR1 HCDR2 HDDR3 LCVR LCDR1 LCDR2 LCDR3 H4N13290R2 1 3 5 7 137 139 141 143 Н4Н13291Р2 9 AND 13 15 137 139 141 143 H4N13295R2 17 19 21 23 137 139 141 143 H4N13299R2 25 27 29 31 137 139 141 143 H4H13300R2 33 35 37 39 137 139 141 143 Н4Н13301Р2 41 43 45 47 137 139 141 143 Н4Н13302Р2 49 51 53 55 137 139 141 143 Н4Н13306Р2 57 59 61 63 137 139 141 143 H4N13309R2 65 67 69 71 137 139 141 143 Н4Н13311Р2 73 75 77 79 137 139 141 143 Н4Н13312Р2 81 83 85 87 137 139 141 143 H4H13313R2 89 91 93 95 137 139 141 143 H4H13316R2 97 99 101 103 137 139 141 143 Н4Н13318Р2 105 107 109 111 137 139 141 143 H4N13319R2 FROM 115 117 119 137 139 141 143 Н4Н13325Р2 121 123 125 127 137 139 141 143 Н4Н13331Р2 129 131 133 135 137 139 141 143

Antibodies are generally named herein according to the following notation: a prefix corresponding to Fc (eg H4H) followed by a numeric identifier (eg 13290, 13291, 13295, etc.) followed by the suffix P2, as shown in table. 1 and 2. Thus, in accordance with such a notation, an antibody herein may be named, for example, H4H13290P2, H4H13291P2, H4H13295P2, etc. The prefix in antibody designations used herein indicates the specific isotype in the Fc region of the antibody. In particular, the H4H antibody contains human IgG4 Fc (all variable regions are fully human, as indicated by the first H in the antibody designation). As one skilled in the art will appreciate, an antibody with a particular Fc isotype can be converted into an antibody with an Fc of a different isotype (e.g., an antibody with a mouse IgG4 Fc can be converted into an antibody with a human IgG1 Fc, etc.), but in any case, variable domains (including CDRs) - which are designated by the numeric identifiers shown in Tables 1 and 2 - will remain the same and binding properties are expected to be identical or substantially similar, regardless of the origin of the Fc domain.

Example 3 Surface Plasma Resonance Binding Affinities and Kinetic Constants of Anti-MET Human Monoclonal Antibodies (Monospecific).

The binding affinities and kinetic constants of human antibodies to MET were determined using the surface plasmon resonance method (Biacore 4000 or T-200) at 37°C. Anti-MET antibodies tested in this example were bivalent monospecific MET binding molecules. Antibodies expressed as human IgG4 (designated H4H) were captured on the surface of a Biacore CM4 or CM5 sensor derivatized by amino coupling to a mouse anti-human Fc monoclonal antibody (GE, BR-1008-39). Various concentrations of soluble monomeric (human (h) Met.mmh; SEQ ID NO: 152; macaca fascicularis (mf) Met.mmh; SEQ ID NO: 154) or dimeric (hMet.mFc; SEQ ID NO: 153) Met proteins by injections covered the surface with captured anti-MET antibody at a flow rate of 30 or 50 µl/min. The binding of hMET.mmh or hMET.mFc to the captured monoclonal antibody was monitored for 4 or 5 min, and the dissociation of hMET.mmh or hMET.mFc in HBS-ET running buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% v/v surfactant P20) or PBS-P (0.01 M sodium phosphate, pH 7.4, 0.15 M NaCl, 0.05% v/v ./vol surfactant P20) was monitored for 10 minutes.

The kinetic rate constants of association (k _a ) and dissociation (kd) were determined by fitting real-time sensogram curves to a 1:1 s binding model at

- 46 042597 change of curve fitting software Scrubber 2.0c. The dissociation constant (KD) for equilibrium binding and the dissociation half-life (t1/2) were calculated from the kinetic rate constants as „ kd ... , . 1p(21

Lb (M) \u003d - and (½ (min) \u003d ^v 'ka ^{y 7} 60 * kd

The parameters of the kinetics of binding of monospecific antibodies to MET with the monomeric and dimeric MET protein are shown in Table 1 below. 3.

Table 3. Binding affinities in the Biacore assay for anti-MET monospecific mAbs at 37°C

Binding at 37°C/antibody capture format

Antibody Analyte ka (Ms ¹ ) kd (s' ¹ ) Kd (molar) T½ (min) H4N13290R2 hMet.mmh 2.53E+05 8.03E-04 ZD7E-09 14.4 hMET.mFc 6.15E+05 ZD5E-04 5DZE-10 36.6 mfMet.mmh l,23E+05 6.33E-04 5D6E-09 18.2 Н4Н13291Р2 hMet.mmh 2.55E+04 2.38E-03 9.34E-08 4.8 hMET.mFc 3.33E+05 3.39E-04 l,02E-09 34 mfMet.mmh 3.70E+04 l,39E-03 3.76E-08 8.3 H4N13295R2 hMet.mmh l,67E+04 5.40E-04 3.24E-08 21.4 hMET.mFc 2.28E+05 2.64E-04 l,16E-09 43.8 mfMet.mmh l,65E+04 9.79E-04 5.93E-08 H.8 H4N13299R2 hMet.mmh 9D0E+04 7.80E-04 8.57E-09 14.8 hMET.mFc 3.57E+05 ZD4E-04 8.78E-10 36.8 mfMet.mmh 1DZE+05 8.84E-04 7.86E-09 13.1 H4H13300R2 hMet.mmh 3.35E+04 2.43E-03 7.25E-08 4.8 hMET.mFc 2.65E+05 2.95E-04 l,12E-09 39.1 mfMet.mmh 5JE+04 l,94E-03 3.77E-08 6.0 Н4Н13301Р2 hMet.mmh 7.57E+04 6.22E-03 8.22E-08 1.9 hMET.mFc 7.05E+05 1D4E-03 l,62E-09 10.1 mfMet.mmh 6.85E+04 5.30E-03 7.74E-08 2.2 Н4Н13302Р2 hMet.mmh 5.24E+04 2.46E-03 4.70E-08 4.7 hMET.mFc 2.51E+05 5.84E-04 2.33E-09 19.8 mfMet.mmh 3.56E+04 2.92E-03 8.20E-08 4.0 Н4Н13306Р2 hMet.mmh l,52E+05 l,66E-02 l,09E-07 0.7 hMET.mFc l,21E+06 2.60E-03 2D5E-09 4.4 mfMet.mmh l,21E+06 3,11E-O2 2.58E-08 0.4 H4N13309R2 hMet.mmh 9.20E+04 5.87E-04 6.38E-09 19.7 hMET.mFc 4.06E+05 2.67E-04 6.57E-10 43.3 mfMet.mmh l,23E+05 6.33E-04 5D6E-09 18.2 Н4Н13311Р2 hMet.mmh 4.48E+04 5.19E-03 1D6E-07 2.2 hMET.mFc 3.02E+05 4.68E-04 l,55E-09 24.7 mfMet.mmh 7.61E+04 6.04E-03 7.94E-08 1.9 Н4Н13312Р2 hMet.mmh 7D9E+04 l,63E-02 2.27E-07 0.7 hMET.mFc 6D4E+05 l,71E-03 2.79E-09 6.7 mfMet.mmh l,47E+05 7.72E-03 5.24E-08 1.5 H4H13313R2 hMet.mmh 8.78E+04 5.70E-03 6.49E-08 2 hMET.mFc 7.50E+05 8.93E-04 1D9E-09 12.9 mfMet.mmh 5D0E+04 4.08E-03 8.00E-08 2.8 H4H13316R2 hMet.mmh 7.82E+04 l,51E-03 l,93E-08 7.6 hMET.mFc 2.93E+05 l,08E-04 3.67E-10 107.4 mfMet.mmh NB NB NB NB Н4Н13318Р2 hMet.mmh 3.30E+04 2.92E-03 8.83E-08 4 hMET.mFc 3.52E+05 l,65E-04 4.67E-10 70.2 mfMet.mmh NB NB NB NB H4N13319R2 hMet.mmh ZD1E+04 2.38E-03 7.65E-08 4.9

- 47 042597

Binding at 37°C/antibody capture format

Antibody Analyte ka (Ms ¹ ) kd (s' ¹ ) Kd (molar) T½ (min) hMET.mFc 3.82E+05 5.42E-04 l,42E-09 21.3 mfMet.mmh 2.66E+04 1D5E-03 4.33E-08 10.0 Н4Н13325Р2 hMet.mmh 9.53E+04 2.36E-03 2.48E-08 4.9 hMET.mFc 3.06E+05 l,85E-04 6.05E-10 62.4 mfMet.mmh NB NB NB NB Н4Н13331Р2 hMet.mmh 2.61E+05 8.73E-04 3.35E-09 13.2 hMET.mFc 6.39E+05 l,56E-04 2.44E-10 74.1 mfMet.mmh l,61E+05 l,04E-03 6.47E-09 H.1

NB - binding under the conditions used was not observed.

As shown in Table. 3, several antibodies showed high binding affinity for the human and monkey MET protein.

Example 4 Anti-MET antibodies bind to isolated epitopes on the Met receptor.

To evaluate whether two anti-MET antibodies were able to compete with each other for binding to their respective epitopes on MET, a binding competition assay was performed using real-time biolayer interferometric analysis (BLI) without labeling on the OCTET® HTX biosensor (ForteBio Corp., Menlo Park, California).

Briefly, approximately 0.25 nM of the extracellular domain of human MET expressed with the C-terminal myc-myc-hexahistidine tag (hMet.mmh) was first captured on the surface of penta-His antibody-coated OCTET® biosensors (ForteBio Corp., no. 18- 5079) by immersing biosensors for 5 min in wells containing 20 µg/ml hMET.mmh solution. The captured antigen biosensors were then saturated with the first anti-MET monoclonal antibody (hereinafter referred to as mAb-1) by immersion in wells containing a 50 μg/ml mAb-1 solution for 5 min. The biosensors were then immersed in wells containing a 50 µg/mL solution of a second anti-MET monoclonal antibody (hereinafter referred to as mAb-2) for 3 min. All biosensors were washed in OCTET® FIEPES-Buffered Saline-EDTA Polysorbate 20 (HBS-EP) buffer between each run of the experiment. The real-time binding response was monitored throughout the experiment and at the end of each stage, the binding response was recorded. The binding response of mAb-2 to anti-MET antibody pre-complexed with mAb-1 was compared and competitive/non-competitive binding patterns of different anti-MET monoclonal antibodies were determined using 50% inhibition threshold. In table. 4 explicitly defines the relationship of antibodies competing in both directions, regardless of the order of binding.

- 48 042597

Table 4. Cross-competition of anti-MET antibodies for binding to hMET.mmh

First mAb (mAb-Ι) captured using anti-pentaHis coated Octet biosensors mAb-2 antibodies that compete with mAb-Ι First mAb (mAb-Ι) captured using anti-pentaHis coated Octet biosensors mAb-2 antibodies that compete with mAb-Ι Н4Н13301Р2 Н4Н13302Р2 H4H13300R2 Н4Н13291Р2 Н4Н13302Р2 Н4Н13301Р2 H4N13295R2 H4N13290R2 Н4Н13306Р2 Н4Н13311Р2 H4H13316R2 Н4Н13318Р2 Н4Н13306Р2 H4N13290R2 H4N13319R2 H4H13316R2 Н4Н13311Р2 Н4Н13291Р2 H4H13316R2 H4N13290R2 H4N13295R2 Н4Н13306Р2 H4H13300R2 Н4Н13325Р2 Н4Н13318Р2 Н4Н13331Р2 H4N13319R2 Н4Н13325Р2 H4H13316R2 Н4Н13318Р2 Н4Н13291Р2 Н4Н13331Р2 H4N13295R2 Н4Н13312Р2 Н4Н13331Р2 H4H13300R2 Н4Н13291Р2 H4N13295R2 Н4Н13311Р2 H4H13300R2 H4N13319R2 Н4Н13311Р2 H4N13319R2 Н4Н13291Р2 Н4Н13318Р2 H4N13295R2 H4N13319R2 H4H13300R2 H4N13295R2 Н4Н13291Р2 Н4Н13311Р2 H4H13300R2 Н4Н13318Р2 Н4Н13311Р2 Н4Н13331Р2 H4H13316R2 Н4Н13318Р2 Н4Н13325Р2 H4N13319R2 Н4Н13312Р2

Example 5 Construction of bispecific antibodies with two different antigen-binding domains specific for different MET epitopes.

This example describes the construction of bispecific antibodies containing two different antigen-binding domains (D1 and D2), where D1 and D2 originate from different antibodies to MET and therefore bind to separate epitopes on the extracellular domain of MET.

The individual anti-MET antibody antigen-binding domains used to construct the bispecific antibodies of this example were derived from the various bivalent anti-MET monospecific antibodies described in Examples 1-3 herein. All anti-MET antibodies described herein contain the same (common) light chain (comprising the amino acid sequence of the light chain variable region [LCVR] SEQ ID NO: 138 and the amino acid sequences of the light chain CDRs [LCDR1, LCDR2 and LCDR3] of SEQ ID NO: 140, 142 and 144). In addition, all of the bispecific antibodies illustrated in this example contain the D2 arm derived from the exemplary anti-MET antibody H4H13312P2. Thus, both antigen-binding domains (D1 and D2) of all of the bispecific antibodies described in this example contain this common light chain variable region, and all of the D2 binding arms contain the heavy chain variable region from H4H13312P2; however, bispecific antibodies differ from each other in the context of their D1 heavy chain variable regions (HCVR) and heavy chain CDRs (HCDR). Components bispecific antibodies from this example are shown in table. 5.

- 49 042597

Table 5. Brief description of the components of the MET x MET bispecific antibody

SEQ ID NO: (amino acid sequences) Bispecific antibody First antigen binding domain (D1) Second antigen binding domain (D2) D1HCVR D1HCDR1 D1HCDR2 D1HCDR3 D2HCVR D2HCDR1 D2HCDR2 D2HCDR3 H4H14634D (No. 10) H4N13290R2 Н4Н13312Р2 2 4 6 8 82 84 86 88 H4H14635D (No. 42) H4N13295R2 Н4Н13312Р2 18 20 22 24 82 84 86 88 Bispecific antibody First antigen binding domain (D1) Second antigen binding domain (D2) D1HCVR D1HCDR1 D1HCDR2 D1HCDR3 D2HCVR D2HCDR1 D2HCDR2 D2HCDR3 H4H14636D (No. 74) H4N13299R2 Н4Н13312Р2 26 28 thirty 32 82 84 86 88 H4H14637D (No. 90) Н4Н13301Р2 Н4Н13312Р2 42 44 46 48 82 84 86 88 H4H14638D (No. 106) Н4Н13302Р2 Н4Н13312Р2 50 52 54 56 82 84 86 88 H4H14639D (No. 122) Н4Н13306Р2 Н4Н13312Р2 58 60 62 64 82 84 86 88 H4H14640D (No. 138) H4N13309R2 Н4Н13312Р2 66 68 70 72 82 84 86 88 H4H14641D (No. 187) H4H13313R2 Н4Н13312Р2 90 92 94 96 82 84 86 88 H4H16445D (No. 26) Н4Н13291Р2 Н4Н13312Р2 10 12 14 16 82 84 86 88 H4H16446D (No. 58) H4H13300R2 Н4Н13312Р2 34 36 38 40 82 84 86 88 H4H16447D (No. 154) Н4Н13311Р2 Н4Н13312Р2 74 76 78 80 82 84 86 88 H4H16448D (#219) Н4Н13318Р2 Н4Н13312Р2 106 108 BY 112 82 84 86 88 H4H16449D (#235) H4N13319R2 Н4Н13312Р2 114 116 118 120 82 84 86 88

* the number in brackets under the identifiers of the bispecific antibodies (for example, No. 10) indicates the number of the bispecific antibody indicated in the table. bispecific antibodies MET x MET in FIG. 1.

Example 6 Surface Plasma Resonance Binding Affinities and Kinetic Constants of Bispecific Human Monoclonal Antibodies MET x MET.

The binding affinities and kinetic constants of the MET x MET bispecific antibodies constructed according to Example 4 herein were determined using surface plasmon resonance (Biacore 4000 or T-200) at 37°C. Bispecific antibodies expressed as human IgG4 (designated H4H) were captured on the surface of a Biacore CM4 or CM5 sensor derivatized by amino coupling to a mouse anti-human Fc monoclonal antibody (GE, BR-1008-39). Various concentrations of soluble monomeric protein MET (hMet.mmh, SEQ ID NO: 152) were sprayed onto the surface with

- 50 042597 captured anti-MET x MET bispecific antibody at a flow rate of 30 or 50 µl/min. Binding of the analyte to the captured bispecific antibody was monitored for 4 or 5 min, and dissociation of the analyte in running HBS-ET buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05 % v/v surfactant P20) or PBS-P (0.01 M sodium phosphate, pH 7.4, 0.15 M NaCl, 0.05% v/v surfactant P20) monitored for 10 min.

The kinetic rate constants of association (k _a ) and dissociation (kd) were determined as described in Example 3.

The parameters of the kinetics of binding of bispecific antibodies to MET with the monomeric protein MET (hMET.mmh) are shown in Table. 6.

Table 6. Binding affinities in the Biacore assay for anti-MET bispecific mAbs at 37 °C

Binding at 37°C/antibody capture format Bispecific antibody Analyte ka (Ms ¹ ) kd (s ¹ ) Kd (molar) ΤΆ (min) H4H14634D hMet.mmh n.a. < IE-5 N.d. > 1155 H4H14635D hMet.mmh N.D. 8.21Е-05 N.d. 140.6 H4H14636D hMet.mmh n.a. < IE-5 n.a. > 1155 H4H14637D hMet.mmh n.a. 3.26E-04 N.d. 35.4 H4H14638D hMet.mmh N.d. 1.65Е-04 N.d. 70.2 H4H14639D hMet.mmh N.d. 1.63E-04 N.d. 70.8 H4H14640D hMet.mmh N.d. < 1Е-5 N.d. > 1155 H4H14641D hMet.mmh N.d. 3.27E-04 N.d. 35.3 H4H16445D hMet.mmh N.d. 3.93E-04 N.d. 29.4 H4H16446D hMet.mmh N.d. 1.03Е-04 N.d. 111.8 H4H16447D hMet.mmh N.d. 8.48Е-04 N.d. 13.6 H4H16448D hMet.mmh N.d. 5.92E-04 N.d. 19.5 H4H16449D hMet.mmh N.d. 2.94Е-04 N.d. 39.3

As shown in Table. 6, the MET x MET bispecific antibodies described herein have T1/2 values up to greater than 1155 min.

As shown in Table. 7, the dissociation rate for the bispecific antibody H4H14639D is significantly lower than the dissociation rates of each of its parent antibodies, H4H13306P2 and H4H13312P2.

Table 7. Binding affinities in the Biacore assay for bispecific anti-MET mAb and monospecific parent antibodies at 37°C

Binding at 37°C/antibody capture format Antibody Analyte kd (s' ¹ ) TI (min) Н4Н13306Р2 hMet.mmh l,66E-02 0.7 Н4Н13312Р2 hMet.mmh 8.40E-03 1.4 H4H14639D hMet.mmh l,63E-04 70.8

Example 7 Anti-MET antibodies block HFG-mediated Met activation in a biological assay using the SRE-luciferase reporter system.

The ability of anti-MET antibodies to block hepatocyte growth factor (HGF) mediated MET activation was studied in an assay using a luciferase-based reporter system. The growth factor HGF binds to the extracellular domain of its c-Met (MET) receptor, triggering rapid homodimerization and activating several downstream signaling cascades. Anti-MET antibodies tested in this example were bivalent monospecific MET binding molecules or bispecific anti-MET antibodies in which each arm of the bispecific antibody binds to a different and distinct epitope on MET.

- 51 042597

The developed cellular assay using a luciferase-based reporter system (FIG. 2) was used to determine the ability of anti-MET antibodies to activate MET signaling (FIG. 3, panel A; Table 8, columns 3 and 4) and block ligand-mediated MET activation (Figure 3, panel B; Table 8, columns 1 and 2). Briefly, the CIGNAL™ Lenti SRE Reporter (luc) Kit (SABiosciences, Hilden, Germany) was used to generate HEK293/SRE-Luc cells. HEK293 (human embryonic kidney) cells were chosen because they endogenously express c-Met. The serum response element (SRE) dependent luciferase (Luc) reporter gene is stably integrated into the genome of HEK293/SRE-Luc cells (see Dinter et al., PLoS ONE 10(2): e0117774, 2015). HEK293/SRE-Luc cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), penicillin/streptomycin/glutamine and 1 μg/ml puromycin.

Then 2.0 x 105 HEK293/SRE-Luc cells were plated in luciferase assay medium in 96 well plates and incubated overnight at 37° C. in 5% CO ₂ . Dose-response curves for hepatocyte growth factor (HGF) were generated by adding serially diluted HGF (0.01 pM to 1.0 nM) to cells and reading the luciferase signal after incubation at 37°C for four to six hours in the absence of antibodies . To obtain inhibition curves for antibodies, cells were pre-incubated for one hour at 37° C. with serially diluted anti-human MET antibodies (1.1 pM to 200 nM). Then HGF was added at a concentration of 73 or 100 pM for another four to six hours before reading the signal. The ability of antibodies to activate c-Met in the absence of ligand was also separately assessed.

Luciferase activity was detected using the ONE-Glo™ Luciferase Assay System (Promega, Madison, WI) and emitted light was measured on a Victor or Envision luminometer (Perkin Elmer, Shelton, Conn.) and expressed as relative light units (RLU). EC ₅₀ /IC ₅₀ values were determined from a four parameter logistic model equation from a 12 point response curve using GRAPFIPAD PRISM®. Percent HGF blocking and MET activation fold (mAb only) were recorded for the highest antibody dose. The results are shown in table. 8.

- 52 042597

Table 8 Anti-MET Antibody Blocking HGF-Mediated Signaling and SRE-Luc Activation in the Absence of Ligand

Blocking activity against HEK293/SRE-Luc (pre-binding to mAb for 1 hour) Ligand independent (HGF) activation of HEK293/SRE-Luc Antibody ID % inhibition 1C ₅ o (M) Multiplicity of answer EU ₅ o (M)

Bivalent monospecific antibodies to MET

Antibodies expressed with hlgGl-Fc H1H13301R2 42 3,ZE-09 1.4 ND H1H13316R2 86 4.0E-11 1.7 ND Antibodies expressed with h!gG4-Fc Н4Н13312Р2 48 7.7E-11 10.9 1.2E-10 Н4Н13325Р2 69 1,ZE-11 4.3 1.9E-10 H4H13316R2 74 7.8E-12 2.3 4.7E-11 Н4Н13302Р2 45 1.6E-09 1.8 ND H4H13313R2 47 2,ZE-09 1.2 ND Н4Н13301Р2 40 1.5E-09 1.6 ND H4N13295R2 70 5.5E-11 2.8 3.0E-10 Н4Н13306Р2 67 ND 9.8 1,ZE-11 Н4Н13291Р2 61 1,ZE-10 2.7 3.9E-10 H4N13319R2 67 5.2E-11 4.8 1.8E-10 H4N13309R2 77 2.0E-10 9.2 3.9E-10 Н4Н13318Р2 77 1.0E-10 3.1 ND H4H13300R2 69 1.2E-10 2.8 4.8E-10 H4N13290R2 56 <2.0E-12 9.8 < 2.0Е-12 Н4Н13311Р2 62 3.5E-11 5.2 3.0E-10 Н4Н13331Р2 75 < 1.0Е-11 7.1 2,ZE-12 H4N13299R2 51 ND 14.4 3.7E-12

Bispecific antibodies to MET (h!gG4-Fc)

- 53 042597

H4H14639D 95 2.4E-11 1.8 5.7E-11 H4H14640D 89 5.2E-10 2.5 6.8E-09 H4H14634D 85 9.7E-12 3.4 9.0E-ll H4H14635D 85 l,9E-10 2.2 l,4E-09 H4H14638D 79 Ι,ΙΕ-09 2.6 5.9E-09 H4H14641D 75 2.7E-09 4.4 8.4E-08 H4H14636D 74 ND 2.8 2.8E-10 H4H14637D 73 ND 2.1 4DE-09 H4H16445D 81 5.2E-10 4.3 l,0E-09 H4H16446D 83 l,0E-09 4.0 l,4E-09 H4H16447D 76 8.6E-10 5.8 l,4E-09 H4H16448D 87 6.2E-10 4.3 9,LE-10 H4H16449D 85 3.2E-10 4.2 4.2E-10

NT - not tested; ND - EC50/IC50 was not determined due to the presence of non-sigmoid curves or incomplete blocking.

As shown in Table. 8, most antibodies inhibited SRE reporter activation with IC50 values ranging from <2.0 pM to approximately 1.0 nM. Several exemplary anti-MET monospecific bivalent antibodies, such as H4H13306P2 and H4H13309P2, were potent inhibitors of SRE-luc activation, with inhibition percentages of 67% and 77%, respectively. Bispecific antibodies to MET (MET x MET) were generally characterized by a higher degree of inhibition of SRE-luc activation. For example, the MET x MET bispecific antibody H4H14639D exhibited 95% inhibition. In addition, several blocking antibodies were weakly activating in the absence of ligand, with fold activation responses ranging from 0.8 to 14.4 above baseline.

Also, as shown in FIG. 3, bivalent monospecific antibodies H41413306P2 and H4H13312P2 each activate Met signaling in the absence of HGF ligand (panel A) and also block HGF activation of Met (panel B).

The effect of a MET x MET bispecific antibody (eg, H4H14639D) on HGF-dependent and HGF-independent MET activation was also evaluated using the HEK293/SRELuc system. SRE-dependent luciferase activity was measured in HEK293T cells treated with anti-MET antibody H4H14639D (MET x MET bispecific antibody), anti-MET monovalent antibody and H4H13312P2 based on the parental H4H14639D antibody at various concentrations to determine the level of HGF-independent MET agonist activity. While MET agonist activity was observed for the parent monospecific anti-MET bivalent antibody, neither the monovalent nor the bispecific MET x MET antibody showed MET agonist activity (FIG. 4, panel A).

SRE-dependent luciferase activity was measured in FIEK293T cells treated with anti-MET antibody H4H14639D (MET x MET bispecific antibody), monovalent anti-MET antibody and H4H13312P2 based on the original H4H14639D antibody at various concentrations to determine the level of inhibition or blocking of HGF-dependent agonist action in regarding MET. While some HGF blocking activity was observed for the parent anti-MET monospecific bivalent antibody, a higher degree of HGF blocking was observed for both the monovalent and bispecific MET x MET antibody (FIG. 4, panel B).

The MET x MET bispecific antibody blocks HGF signal transduction and has low MET agonist activity.

Example 8 Anti-MET antibodies inhibit the growth of Met-amplified cells.

Next, the selected anti-MET antibodies were tested for their ability to inhibit the growth of Met-amplified SNU5 cells. Briefly, 2.5 x 10 ³ human gastric carcinoma (SNU5) cells were plated in complete growth medium in the presence of anti-MET antibodies at concentrations ranging from 1.5 pM to 100 nM. Complete SNU5 growth medium contained Iskov's Dulbecco's medium, 10% FBS, and penicillin/streptomycin/glutamine. Cells were incubated for 5 days and viable cell count was determined using the CELLTITER-GLO® Luminescent Cell Viability Assay kit (Promega, Madison, Wisconsin) according to the manufacturer's instructions.

- 54 042597

As shown in Table. 9, several anti-MET antibodies, such as H4H13312P2 and H4H13325P2, blocked SNU5 growth by more than 50%, with overall IC ₅₀ values ranging from 44 to 780 pM.

In FIG. 5 shows the relative growth of SNU5 cells treated with various anti-MET bivalent monospecific antibodies (ie conventional antibodies). A subset of conventional anti-MET antibodies inhibits the growth of Met-amplified SNU5 gastric cancer cells (FIG. 5). SNU5 cells in 96-well plates were treated with each antibody at 10 μg/ml and cell growth was determined 5 days later by reconstitution of ALAMARBLUE® reagent (Thermo Fisher Scientific, Waltham, MA). A monovalent anti-MET antibody (bar 2, Figure 5) was generated using the MetMab heavy and light chain variable region sequences as set forth in US Patent Application 7,892,550 B2, which is incorporated herein by reference in its entirety. Conventional antibody 8 is H4H13306P2 and reference antibody 11 is H4H13312P2, which were used to construct the MET x MET H4H14639D bispecific antibody.

In a separate growth assay, the blocking activity of a MET x MET bispecific antibody (i.e., H4H14639D) was assessed in both SNU5 and the EBC-1 non-small cell lung cancer (NSCLC) cell line, which also has an amplified Met gene and overexpresses MET (Lutterbach et al., Cancer Res. 67(5): 2081-2088, 2007). Complete growth medium for EBC-1 cells contained MEM with Earle's salts, 10% fetal calf serum (FBS), penicillin/streptomycin/glutamine, and non-essential amino acids for MEM. H4H14369D had the strongest inhibition of MET activity as measured by the SRE-luciferase assay. In this experiment, 3.0 x 10 ³ SNU5 or EBC-1 cells were seeded in complete growth medium in the presence of H4H14639D at concentrations ranging from 15 pM to 100 nM. Cells were incubated for 3 days at 37°C in 5% CO2. Cells were then fixed in 4% formaldehyde and stained with 3 μg/ml Hoechst 33342 nuclear labeling dye. Images were acquired on an IMAGEXPRESS® Micro XL (Molecular Devices, Sunnyvale, CA) and nuclei counts were determined using METAXPRESS® image analysis software (Molecular Devices, Sunnyvale, CA). Background nuclei from cells treated with 40 nM digitonin were subtracted from all wells and viability was expressed as a percentage relative to untreated controls. EC ₅₀ values were determined based on a four parameter logistic model equation from a 10 point response curve (GRAPHPAD PRISM®). The values of the EU ₅₀ and the destruction of cells in percent are shown in table. 9.

Table 9. Anti-MET antibody blocking the growth of SNU5

Antibody % growth inhibition IC50 (M) Antibody % growth inhibition IC50 (M) Н4Н13312Р2 69 7.8E-10 Н4Н13291Р2 24 ND Н4Н13325Р2 57 4.4E-11 H4N13319R2 23 1.0E-10 H4H13316R2 53 1.0E-10 H4N13309R2 22 1.0E-10 Н4Н13302Р2 40 Ι, ΙΕ-10 Н4Н13318Р2 18 5DE-11 H4H13313R2 34 4.4E-11 H4H13300R2 16 ND Н4Н13301Р2 33 7.4E-11 H4N13290R2 12 ND H1H13301R2 33 1.0E-10 Н4Н13311Р2 8 ND H1H13316R2 thirty 2.0E-10 Н4Н13331Р2 5 ND H4N13295R2 thirty ND H4N13299R2 -8 ND Н4Н13306Р2 28 7.1E-11

ND - IC ₅₀ was not determined due to the presence of non-sigmoidal curves or incomplete blocking.

As shown in Table. 10 below, the MET x MET H4H14639D bispecific antibody inhibited the growth of EBC-1 and SNU5 cells by 37% and 40%, with IC _50s of 0.82 and 0.3 nM, respectively.

Table 10 Bispecific anti-Met antibody blocks the growth of EBC-1 and SNU5

mAb ICso (nM) % growth inhibition EBC-1 SNU5 EVS-1 SNU5 H4H14639D 0.82 0.30 37 40

SNU5 (stomach) cells in 96-well plates were treated with control antibody, anti-MET monovalent antibody, or MET x MET bispecific antibody at 0.1, 1, or 10 μg/ml. Cell growth was determined 5 days later by reconstitution of ALAMARBLUE® reagent (Thermo Fisher Scientific, Waltham, MA). The MET x MET bispecific antibody provided a significant reduction in the relative growth of SNU5 cells compared to control and monovalent antibody (FIG. 6, panel A).

Similarly, the effect of the bispecific antibody MET x MET on the growth of EBC-1 cells was evaluated. 2500 EBC-1 cells were seeded in a 96-well plate and cultured in Dulbecco's medium supplemented with 10% FBS. Cells were treated with control antibody or bispecific antibody MET x MET at 0.1 or 1 μg/ml and then incubated with 5% CO ₂ at 37°C. After 5 days, relative cell growth was determined by measuring the reduction of the ALAMARBLUE® indicator dye to its highly fluorescent form on a SPECTRAMAX® M3 plate reader (Molecular Devices, LLC, Sunnyvale, CA). The results are shown in table. 11 and in FIG. 6, panel B. The MET x MET bispecific antibody (H4H14639D) provided a significant reduction in the relative growth of EBC-1 cells compared to the control antibody (FIG. 6, panel B).

Several anti-MET antibodies, both bivalent monospecific and bivalent MET x MET, are potent inhibitors of SRE-Luc activation and inhibit the growth of Met-amplified cell lines and MET-overexpressing cell lines.

Table 11 Bispecific Anti-Met Antibody Blocks EBC-1 Cell Growth

Relative cell growth (n=3) Standard deviation Control 1,000 0.045 0.1 µg/ml H4H14639D 0.397 0.032 1 μg/ml H4H14639D 0.462 0.028

Example 9 A MET x MET bispecific antibody induces a slight and transient activation of a MET signaling pathway in NSCLC NCI-H596 cells.

The effect of the MET x MET bispecific antibody on the MET signaling pathway in human adenosquamous lung carcinoma cells was evaluated in vitro.

250,000 NCI-H596 cells were plated in a 12-well plate and cultured in RPMI supplemented with 10% FBS. Cells were treated with hepatocyte growth factor (HGF) at 50 ng/ml or bispecific antibody MET x MET H4H14639D at 10 μg/ml in duplicate. The cells were then incubated in 5% CO2 at 37°C. After 0, 2, 6, or 18 hours, cell lysates were prepared, the protein content was normalized, and immunoblot analysis was performed. MET phosphorylation and ERK phosphorylation were quantified using the ImageJ image processing program (T. Collins, BioTechniques 43: S25-S30, 2007). Phosphorylation levels were normalized to tubulin as a load control and expressed as fold change from treatment control. The results are shown in table. 12.

Table 12 Phosphorylation of MET and ERK

Processing (hours) Phospho-MET (mean ± SD) Phospho-ERK (mean ± SD) Control (hFc) (18) 1.0±0.5 1.0±0.3 HGF(2) 202.3 ± 38.7 16.7 ± 1.6 HGF(6) 38.9±4.9 12.4±3.9 HGF(18) 59.2 ± 24.4 12.4±0.9 H4H14639D (2) 69.7 ± 7.0 2.2±0.9 H4H14639D (6) 9.9 ± 7.4 0.3±0.4 H4H14639D (18) 1.4±0.1 0.1±0.1

HGF treatment of NCI-H596 cells induced strong MET and ERK activation, which peaked at 2 hours and persisted after 18 hours. Negligible levels of MET and ERK phosphorylation were detected when treated with the H4H14636D bispecific antibody, which returned to baseline levels at 18 or 6 hours, respectively. .

Example 10 MET x MET bispecific antibody induces MET degradation and inhibits signaling pathway activation more strongly than monospecific antibodies in Hs746T gastric cancer cells.

The effect of the MET x MET bispecific antibody on the activity of human gastric carcinoma MET cells was evaluated in vitro. 250,000 Hs746T human gastric carcinoma cells (H. Smith, J.

- 56 042597

Nat'l. Cancer Inst. 62(2): 225-230, 1979) were plated in a 12-well plate and cultured in Dulbecco's Modified Medium supplemented with 10% FBS. Cells were treated with (1) 5 µg/mL hFc control molecule, (2) 5 µg/mL stock bivalent anti-MET monospecific antibody H4H13306P2, (3) 5 µg/mL stock bivalent anti-MET monospecific antibody H4H13312P2, (4) combination 2.5 µg/ml H4H13306P2 and 2.5 µg/ml H4H13312P2 or (5) 5 µg/ml MET x MET H4H14639D bispecific antibody. The cells were then incubated with 5% CO ₂ at 37°C. After 18 hours, cell lysates were prepared, the protein content was normalized, and immunoblot analysis was performed. MET expression, MET phosphorylation, and ERK phosphorylation were quantified using ImageJ image processing software (T. Collins, BioTechniques 43: S25-S30, 2007). The results are shown in table. 13 and in FIG. 7, panel A showing primary immunoblot data. In panel B of Fig. 7 shows the results of MET protein expression in cells that were treated with MET x MET bispecific antibody at 10 μg/ml for 0, 2, or 6 hours. Total MET levels in Hs747T cells decreased over time after treatment with MET x MET bispecific antibody. Similar results were obtained for the NCI-H820 human papillary adenocarcinoma cell line with amplified MET gene (Bean et al., MET amplification occurs with or without T790M mutations in EGFR mutant lung tumors with acquired resistance to getfitnib or erlotinib, Proc. Natl. Acad. Sci. 2007 Dec 26, 104(52): 20932-20937).

Table 13 Relative MET Protein Levels and MET/ERK Signaling Pathway Activation

Molecule Relative MET protein level (mean ± SD) Relative phosphoMET level (mean ± SD) Relative phosphoERK level (mean ± SD) Control (hFc) 1.00±0.06 1.00±0.06 1.00±0.03 Н4Н13306Р2 0.61±0.09 0.57 ± 0.02 0.41±0.03 Н4Н13312Р2 1.15±0.19 0.93±0.04 0.39±0.11 Н4Н13306Р2+Н4Н13312Р2 1.06±0.02 1.07±0.10 1.04±0.23 H4H14639D 0.41±0.02 0.20±0.01 0.04±0.01

The bispecific antibody, H4H14639D, induced MET degradation more than its original conventional antibodies. Phosphorylation of both MET and ERK was inhibited more efficiently by treatment with H4H14636D compared to the parent antibody or combination of parent antibodies.

Hs746T gastric cancer cells were treated with control antibody, MET x MET bispecific antibody H4H14639D, parental anti-MET antibody H4H13306P2, parental anti-MET antibody H4H13312P2, and a combination of parental antibodies 1 and 2, each antibody at 10 μg/mL or combination of parental antibodies at 5 μg /ml each, for 18 hours. MET expression (MET) and signaling pathway activation (pMET and pErk) were determined by immunoblotting using the indicated antibodies (Fig. 8). The MET x MET bispecific antibody inhibits MET signaling pathway activation more effectively than its parent antibody in Hs746T gastric cancer cells.

Example 11 MET x MET bispecific antibody induces MET degradation more strongly than monospecific antibodies in NCI-H596 lung cancer cells.

The effect of the MET x MET bispecific antibody and the parent anti-MET bivalent monospecific antibodies on the expression levels of the hepatocyte growth factor receptor (HGFR or MET) on the surface of human lung adenosquamous carcinoma cells was evaluated. 250,000 NCI-H596 human lung adenosquamous carcinoma cells were plated in a 12-well plate and cultured in RPMI medium supplemented with 10% FBS. Cells were treated with (1) 5 µg/mL hFc control molecule, (2) 5 µg/mL stock anti-MET bivalent mono-specific antibody H4H13306P2, (3) 5 µg/mL stock bivalent anti-MET H4H13312P2 mono-specific antibody, (4) combination 2.5 µg/ml H4H13306P2 and 2.5 µg/ml H4H13312P2 or (5) 5 µg/ml MET x MET H4H14639D bispecific antibody. The cells were then incubated with 5% CO ₂ at 37°C. After 18 hours, cell lysates were prepared, the protein content was normalized, and immunoblot analysis was performed. MET expression was quantified using ImageJ image processing software (T. Collins, BioTechniques 43: S25S30, 2007). The results are shown in table. 14.

- 57 042597

Table 14 Relative MET Protein Level

Molecule Relative MET level Control (hFc) 1 ± 0.03 Н4Н13306Р2 0.50±0.01 Н4Н13312Р2 0.35±0.04 Н4Н13306Р2+Н4Н13312Р2 0.61±0.04 H4H14639D 0.24 ± 0.01

Lung cancer cells NCI-H596 (with a 14 MET exon skipping mutation) were also treated with control or bispecific antibodies MET x MET at 10 μg/ml for 2, 6, or 18 h. which showed bispecific antibody MET x MET induced degradation of MET with increasing duration of treatment.

The bispecific antibody, H4H14636D, induces MET degradation more than its parent conventional antibodies in NCI-H596 lung cancer cells.

Example 12 MET x MET bispecific antibodies induce MET degradation and inhibit signaling pathway activation more strongly than monospecific antibodies in SNU5 gastric cancer cells.

The effect of a bivalent anti-MET monospecific antibody and several MET x MET bispecific antibodies on expression levels of the hepatocyte growth factor receptor (HGFR or MET) on the surface of gastric carcinoma cells was evaluated. SNU5 human gastric carcinoma cells were seeded in Iscove's medium containing 20% FBS and penicillin-streptomycin-glutamine. Через 24 ч после высевания клетки обрабатывали контрольным hFc, исходным бивалентным моноспецифическим антителом к MET H4H13312P2 или биспецифическими антителами МЕТ х MET (H4H14634D, H4H14635D, H4H14636D, H4H14637D, H4H14638D, H4H14639D, H4H14640D, H4H14641D) в течение 18 ч. Затем готовили клеточные лизаты и analyzed by Western blotting. Immunoblots were studied for MET and tubulin. The level of MET protein expression was quantified and normalized against tubulin as a load control. The results are presented in table. 15 and in FIG. 10, panel B.

Table 15 Relative MET Protein Level

Molecule Relative MET level Molecule Relative MET level Control (hFc) 1 H4H14637D 0.49 Н4Н13312Р2 0.62 H4H14638D 0.35 H4H14634D 0.45 H4H14639D 0.27 H4H14635D 0.27 H4H14640D 0.18 H4H14636D 0.50 H4H14641D 0.31

SNU5 cancer cells were treated with a control antibody, or a MET x MET bispecific antibody, or an anti-MET monovalent antibody at 10 μg/ml for 18 h as described above. MET expression (FIG. 10, panels A and B) and signaling pathway activation (ie, pMET and pERK; panel A) were determined by immunoblotting using the indicated antibodies. The immunoblotograms are shown in Fig. 10.

Treatment of SNU5 cells with MET x MET bispecific antibodies induced more MET degradation than treatment with anti-MET bivalent monospecific antibody (H4H13312P2) (Figure 10, panel B), anti-MET monovalent antibody, or control hFc. Treatment of SNU5 cells with the MET x MET bispecific antibody provided inhibition of downstream signaling pathway effectors involving MET. Similar results were obtained for the non-small cell lung cancer cell line NCI-H1993 with amplified MET gene (Kubo et al., MET gene amplification or EGFR mutation activate MET in lung cancers untreated with EGFR tyrosine kinase inhibitors, Int. J. Cancer 2009 Apr 15; 124(8): 1778-1784).

Example 13 MET x MET bispecific antibody induces MET degradation, inhibits signaling pathway activation, and inhibits tumor growth more strongly than monospecific antibodies in EBC-1 cells.

Met-amplified EBC-1 human lung squamous cell carcinoma cells (Lutterbach et al., Lung cancer cell lines harboring MET gene amplification are dependent on Met for growth and survival, Cancer Res. 2007 Mar 1;67(5):2081-8) treated with control antibody or 10 μg/ml bispecific MET x MET antibody for 18 hours as described above. The expression level of MET and the activation of the signaling pathway involving MET, as determined by the expression of pMET and pErk, were determined by immunoblotting using the indicated antibodies. The immunoblotograms are shown in Fig. eleven.

Treatment of EBC-1 cells, which are characterized by MET gene amplification, with MET x MET bispecific antibodies induced more MET degradation than treatment with control antibody. Treatment of EBC-1 cells with the MET x MET bispecific antibody provided inhibition of downstream signaling pathway effectors involving MET.

In another experiment, C.B.-17 SCID mice were implanted subcutaneously in the flank with 5 million EBC-1 cells. After tumor volumes reached approximately 150 mm ³ , mice were randomized into groups of 6 and treated twice a week with control antibody at 25 mg/kg or MET x MET H4H14639D bispecific antibody at 25 mg/kg. Tumor growth was monitored for 30 days after implantation and tumor volume (mm ³ ) was measured over time for each experimental group. The results are shown in table. 16 and in FIG. 12, which shows that the MET x MET bispecific antibody significantly inhibits the growth of EBC-1 tumors.

Table 16. Relative growth rates of EBC-1 tumors

Treatment Tumor growth (mm ³ ) from start of treatment (mean ± SEM) 25 mg/kg control 1394±226 25 mg/kg H4H14639D 89±47

Example 14 MET x MET bispecific antibody inhibits in vitro growth of Hs746T gastric cancer cells more strongly than monospecific antibodies.

The effect of the MET x MET bispecific antibody on the growth of human gastric carcinoma cells was evaluated in vitro. 2500 Hs746T human gastric carcinoma cells (H. Smith, J. Nat'l. Cancer Inst. 62(2): 225-230, 1979) were plated in 96 plates and cultured in Dulbecco's Modified Medium supplemented with 10% FBS. Cells were treated with (1) single anti-MET bivalent monospecific antibodies (H4H13306P2 or H4H13312P2) at 5 µg/mL, (2) a combination of two bivalent anti-MET monospecific parent antibodies (H4H13306P2 and H4H13312P2) at 2.5 µg/mL each, or (3) a bispecific antibody containing one binding arm from H4H13306P2 and the other binding arm from H4H13312P2 (H4H14639D), at 5 μg/ml. The cells were then incubated with 5% CO2 at 37°C. After 5 days, relative cell growth was determined by measuring the reduction of the indicator dye, ALAMAR BLUE® (ThermoFischer Scientific, Waltham, MA) to its highly fluorescent form on a SPECTRAMAX® M3 plate reader (Molecular Devices, Sunnyvale, CA). Increasing fluorescence levels correlate with cell growth. in table. 17 shows the relative growth of Hs746T cells for each antibody treatment, normalized to control growth of Hs746T cells (no treatment). The bispecific antibody, H4H14639D, inhibits Hs746T cell proliferation more strongly than its parent monospecific antibodies, alone or in combination.

Table 17. Normalized growth of Hs746T cells

Relative cell growth (n=3) Standard deviation Control 1 0.133497801 H4H14639D 0.647408139 0.019090432 Н4Н13306Р2 1.623312821 0.189647479 Н4Н13312Р2 0.852680493 0.01728527 Н4Н13 3 06Р2±Н4Н 13 312Р2 1.767720125 0.077445717

Hs746T gastric cancer cells were treated with control antibody, MET x MET bispecific antibody H4H14639D, parental anti-MET H4H13306P2 antibody, parental anti-MET H4H13312P2 antibody, and parental antibody combination 1 and 2, each antibody at 2 μg/mL. Cell growth was determined after 5 days by reconstitution of ALAMAR BLUE® reagent (FIG. 13, panel A). The MET x MET bispecific antibody inhibited cell growth compared to the parent antibody alone or in combination and inhibited MET signaling pathway activation more effectively than its parent antibody in Hs746T gastric cancer cells.

Hs746T gastric cancer cells in 96-well plates were treated with 25 μg/ml control antibody, 1, 10, or 25 μg/ml anti-MET monovalent antibody, or 1, 10, or 25 μg/ml MET x MET bispecific antibody. Growth of Hs746T gastric cancer cells was determined 5 days later by reconstitution of ALAMARBLUE® reagent (FIG. 13, panel B). The bispecific antibody MET x MET strongly inhibited

- 59 042597 growth of cells with amplified Met.

Example 15 MET x MET bispecific antibody does not induce growth of NCIH596 lung cancer cells in vitro.

The effect of the MET x MET bispecific antibody on the growth of human non-small cell lung cancer (NSCLC) cells (NCI-H596) was evaluated in vitro. 10,000 NCI-H596 adenosquamous lung carcinoma cells (Nair et al., J. Nat'l. Cancer Inst. 86(5): 378-383, 1994) were seeded in 96-well plates on a layer of 0.66% agar in medium, supplemented with 1% fetal bovine serum (FBS). Cells were cultured in RPMI 1640 supplemented with 1% FBS with 0.3% agarose. Cells were treated with (1) single stock anti-MET bivalent monospecific antibodies (H4H13306P2 or H4H13312P2) at 5 µg/mL, (2) a combination of two stock bivalent anti-MET monospecific antibodies (H4H13306P2 and H4H13312P2) at 2.5 µg/mL each, (3 ) a bispecific antibody containing one binding arm from H4H13306P2 and the other binding arm from H4H13312P2 (H4H14639D), at 5 μg/ml, or (4) 100 ng/ml of hepatocyte growth factor (HGF). The cells were then incubated with 5% CO ₂ at 37°C. Two weeks later, relative cell growth was determined by measuring the reduction of the indicator dye, ALAMAR BLUE® (Thermo Fischer Scientific, Waltham, MA) to its highly fluorescent form on a SPECTRAMAX® M3 plate reader (Molecular Devices, Sunnyvale, CA). Increasing fluorescence levels correlate with cell growth. in table. 18 and in FIG. 14 shows the relative growth of NCI-H596 cells for each antibody treatment, normalized to control growth of NCI-H596 cells (no treatment). Treatment of lung cancer cells with NCI-H596 HGF resulted in a strong induction of growth in soft agar. The MET x MET bispecific antibody (MM in Fig. 14) H4H14639D did not result in a significant change in growth compared to control-treated cells. A slight induction of cell growth was observed in the case of each of the original bivalent monospecific antibodies H4H13306P2 (M1) or H4H13312P2 (M2) alone, or in combination (H4H13306P2 and H4H13312P2) (M1M2).

Table 18 Normalized growth of NCI-H596 cells

Relative cell growth (n=3) Standard deviation Control 1 0.030074808 H4H14639D 1.070339237 0.075103746 Н4Н13306Р2 2.9593578 0.337877264 Н4Н13312Р2 1.686580346 0.145670753 Н4Н13306Р2+Н4Н13312Р2 1.693724668 0.168651046 HGF 7.87655937 0.46057617

Example 16 MET x MET bispecific antibody inhibits in vitro growth of SNU5 gastric cancer cells more strongly than monospecific antibodies.

The effect of the MET x MET bispecific antibody on the growth of human gastric carcinoma cells was evaluated in vitro. 2500 SNU5 human gastric carcinoma cells (Ku and Park, Cancer Res. Treat. 37(1): 1-19, 2005) were plated in a 96-well plate and cultured in Iskov's Modified Dulbecco's medium supplemented with 20% FBS. Cells were treated with (1) single anti-MET bivalent monospecific antibodies (H4H13306P2 or H4H13312P2) at 5 µg/mL, (2) a combination of two bivalent anti-MET monospecific antibodies (H4H13306P2 and H4H13312P2) at 2.5 µg/mL each, or (3) bispecific an antibody containing one binding arm from H4H13306P2 and the other binding arm from H4H13312P2 (H4H14639D), at 5 μg/ml. The cells were then incubated with 5% CO2 at 37°C. After 5 days, relative cell growth was determined by measuring the reduction of the indicator dye, ALAMAR BLUE® (Thermo Fischer Scientific, Waltham, MA) to its highly fluorescent form on a SPECTRAMAX® M3 plate reader (Molecular Devices, Sunnyvale, CA). Increasing fluorescence levels correlate with cell growth. in table. 19 shows the relative growth of SNU5 cells for each antibody treatment, normalized to control SNU5 cell growth (no treatment). The bispecific antibody, H4H14639D, inhibited SNU5 cell proliferation more strongly than its parent monospecific antibody.

- 60 042597

Table 19. Normalized growth of SNU5 cells

Relative cell growth (n=3) Standard deviation Control 1 0.070814765 H4H14639D 0.271100069 0.01324024 Н4Н13306Р2 0.766317547 0.061930288 Н4Н13312Р2 0.431990234 0.033183065 Н4Н13306Р2+Н4Н13312Р2 0.331287005 0.012042949

Example 17 MET x MET bispecific antibody induces regression of Hs746T tumor xenograft.

The effect of the MET x MET bispecific antibody on human gastric carcinoma was evaluated in an immunodeficient mouse model. Three million Hs746T human gastric carcinoma cells were implanted subcutaneously in the flank of CB-17 SCID mice (Bancroft et al., J. Immunol. 137(1): 4-9, 1986). After tumor volumes reached approximately 200 mm ³ , mice were randomized into groups of six and treated twice a week with control antibody at 25 mg/kg or MET x MET bispecific antibody (H4H14639D) at 25 mg/kg. Tumor growth was monitored for 16 days post-implantation for the control group if the control-treated tumors reached the size limit according to the protocol. In the case of the H4H14639 treatment group, tumor growth was monitored for 30 days after implantation.

Treatment of tumors with the MET x MET bispecific antibody resulted in the induction of tumor size regression in 21 days compared to the start of treatment. For control-treated tumors, an approximately 12-fold mean increase in volume over 16 days of growth was observed (Table 20). The tumor volume over time, in which Hs746T tumor regression was observed due to the MET x MET bispecific antibody, is shown in FIG. 15.

Table 20 Growth of gastric tumor Hs746T

Antibody (mg/kg) Tumor growth ( ^mm3 ) since start of treatment (mean ± SEM) Control (10) 1164 ± 138 H4H14639D (25) -215±8.3

Example 18 MET x MET bispecific antibody induces regression of SNU5 tumor xenograft.

The effect of the MET x MET bispecific antibody on human gastric carcinoma was evaluated in an immunodeficient mouse model. Ten million SNU5 human gastric carcinoma cells were subcutaneously implanted in the flank of CB-17 SCID mice. After tumor volumes reached approximately 500 mm ³ , mice were randomized into groups of five and treated twice a week with control antibody at 10 mg/kg or MET x MET bispecific antibody (H4H14639D) at either 1 mg/kg or 10 mg/kg. Tumor growth was monitored for 81 days post-implantation if control-treated tumors reached the size limit of the protocol.

For tumors in mice treated with 1 or 10 mg/kg of MET x MET antibody, a mean reduction in size of approximately 95% or 98%, respectively, was demonstrated. For control-treated tumors, an approximately 12-fold mean increase in volume from the start of treatment was observed (Table 21).

Table 21 Growth of gastric tumor SNU5

Antibody (mg/kg) Tumor growth ( ^mm3 ) since start of treatment (mean ± SEM) Control (10) 1123 ± 194 H4H14639D (1) -477±43 H4H14639D (10) -492±18

Subcutaneously implanted SNU5 tumors were treated twice weekly with control antibody, anti-MET monovalent antibody at 1 or 10 mg/kg, or MET x MET bispecific antibody at 1 or 10 mg/kg. Strong and robust regression of Met-amplified SNU5 tumors (ie, reduction in tumor volume) was observed over time in mice treated with the MET x MET bispecific antibody (FIG. 16, panel A). Protein was isolated from the tumors at the end of the study, and MET expression and signaling pathway activation by phosphorylation were determined using immunoblotting.

- 61 042597 MET (expression of pMET). MET x MET treated mice (in tumors) showed reduced levels of MET and pMET expression compared to controls (FIG. 16, panel B). The bispecific antibody MET x MET is a strong inhibitor of tumors with amplification of the MET gene.

Example 19 MET x MET bispecific antibody induces regression of U87-MG tumor xenograft.

The effect of the bispecific antibody MET x MET on human glioblastoma was evaluated in a model using immunodeficient mice. Five million U87-MG human glioblastoma cells were implanted subcutaneously in the flank of CB-17 SCID mice (Vordermark and Brown, Int. J. Radiation Biol. 56(4): 1184-1193, 2003). U87-MG glioblastoma xenograft models are under the control of autocrine HGF signaling. After tumor volumes reached approximately 100 mm ³ , mice were randomized into groups of six and treated with a control antibody or a MET x MET bispecific antibody (H4H14639D). Each mouse was injected with 25 mg/kg antibody (control or MET x MET) twice a week. Tumor growth was monitored for 29 days post-implantation if control-treated tumors reached the size limit of the protocol.

Tumors in mice treated with MET x MET antibody showed a mean size reduction of approximately 38%, while control-treated tumors showed an approximately 19-fold mean increase in volume over 29 days of growth (Table 22). The tumor volume over time, in which U87-MG tumor regression was observed due to the MET x MET bispecific antibody, is shown in FIG. 17.

Table 22. Tumor growth in glioblastoma

Antibody (mg/kg) Tumor growth ( ^mm3 ) since start of treatment (mean ± SEM) Control (25) 1777±98 H4H14639D (25) -38±18

Example 20 MET x MET bispecific antibody inhibits growth of U118-MG tumor xenograft.

The effect of the bispecific antibody MET x MET on human glioblastoma was evaluated in a model using immunodeficient mice. U118-MG glioblastoma xenograft models are under the control of autocrine HGF signaling. Five million U118-MG human glioblastoma cells were implanted subcutaneously in the flank of CB-17 SCID mice (Olopade et al., Cancer Research 52: 2523-2529, 1992). After tumor volumes reached approximately 100 mm ³ , mice were randomized into groups of six and treated with a control antibody or a MET x MET bispecific antibody (H4H14639D). Each mouse was injected with 25 mg/kg antibody (control or MET x MET) twice a week. Tumor growth was monitored for 72 days after implantation.

Anti-MET antibody inhibited tumor growth by 99% over a 72 day time period (Table 23).

Table 23 Tumor growth in glioblastoma

Antibody (mg/kg) Tumor growth ( ^mm3 ) since start of treatment (mean ± SEM) % Tumor Growth Inhibition Compared to Control Control (25) 1228 ± 123 - H4H14639D (25) and ± 18 99.1

In another experiment, subcutaneously implanted U118-MG glioblastomas in mice were treated twice weekly with 25 mg/kg of a control antibody, an anti-MET monovalent antibody, or a MET x MET bispecific antibody. Tumor volume (mm ³ ) was measured in dynamics for each experimental group. The results are shown in FIG. 18, which shows that the MET x MET bispecific antibody inhibits the growth of U118-MG tumors.

Example 21 Synthesis of maytansinoids

Maytansine-3-N-methyl-L-alanine-N-Me-beta-alanine-carbamyl-(p-amino)benzyl-citrulline-valinadipoyl succinate (compound 1 in Fig. 20) was synthesized from compound 2 (Fig. 19 ) as described below.

Maytansine-3-N-methyl-L-alanine-Fmoc-N-Me-beta-alanine (compound 3, Fig. 19). Deacetyl-maytansine (compound 2, Fig. 19, 0.433 g, 0.666 mmol), Fmoc-N-Me-beta-Ala (0.434 g, 1.33 mmol) and HATU (0.757 g, 1.99 mmol) were weighed dry flask, dissolved in anhydrous DMF (9 ml) and treated with 4-methylmorpholine (0.300 ml, 2.73 mmol). The flask was sealed with a rubber septum cap, purged with argon, and the reaction mixture was stirred at ambient temperature. After 3 days, the mixture was evaporated to an oil, dissolved in acetonitrile and water, and purified with flash.

- 62 042597 chromatography on a 275 g C18 silica gel column (30-90% acetonitrile in water for 20 min, 0.05% acetic acid in both phases). Lyophilization of product fractions gave the title compound as a white solid. The crude product was purified on an 80 g silica gel column (EtOAc - 5:5:1 EtOAc:DCM:MeOH over 17 min). Pure fractions were combined, evaporated and dried in vacuo overnight to give the title compound as a white solid (0.424 g, 66%). MS (ESI, pol): calculated for C ₅₁ H ₆₁ ClN ₄ O ₁₂ , 956.4; found 956.9 (M+H), 979.0 (M+Na), 939.0 (M-H2O+H).

N-tert-butoxycarbonyl-N-methyl-beta-alanine ester of succinic acid (compound 4, Fig. 19). The title compound was prepared from commercial Boc-N-Me-beta-Ala-OH using a method well known in the art (cf. Widdison et al., J. Med. Chem., 2006, 49 (14), 4401). 1H NMR (300 MHz, CDCl ₃ ): δ 3.62 (bm, 2H), 2.88 (m, 9H), 1.47 (s, 9H).

Maytansine-3-N-methyl-L-alanine-Boc-N-Me-beta-alanine (compound 5, Fig. 19).

Method A: The product from the previous step (compound 4, Fig. 19, 0.453 g, 1.51 mmol) and deacetyl-maytansine (compound 2, Fig. 19, 0.304 g, 0.468 mmol) were dissolved in acetonitrile:water 3:1 (8 ml), treated with 1M aqueous NaHCO ₃ (0.5 ml) and stirred at ambient temperature for 18 h. After completion of the reaction, as determined by TLC, the reaction mixture was then stirred with brine for 10 min and three times extracted with ethyl acetate (EtOAc). The combined organic layers were then dried over Na ₂ SO ₄ , filtered, and the filtrate was concentrated and dried in vacuo to a golden syrup, which was purified by flash column chromatography on a 20 g silica gel cartridge (0-10% MeOH in EtOAc for 15 min) to give the title compound as a white solid (0.084 g, 43%). MS (ESI, pol): calculated for C41H59ClN ₄ O12, 834.4; found 835.2 (M+H), 857.2 (M+Na), 817.4 (M-H2O+H).

Method B: Boc-N-Me-beta-Ala-OH (0.294 g, 1.45 mmol) was dissolved in anhydrous DMF (5 ml), treated with pentafluorophenyl diphenylphosphinate (FDPP, 0.555 g, 1.44 mmol) and the reaction mixture was stirred at ambient temperature for 30 min. The mixture was then transferred to a larger flask containing a mixture of deacetyl-maytansine (compound 2, Fig. 19, 0.462 g, 0.711 mmol) and diisopropylethylamine (DIEA, 0.250 ml, 1.44 mmol) in anhydrous DMF (7 ml), the flask was sealed closed with a lid with a rubber membrane, purged with argon and the reaction mixture was again stirred at ambient temperature. After 24 h, the reaction mixture was concentrated in vacuo to an oil, dissolved in ethyl acetate (EtOAc, 2 ml) and purified on a 40 g silica gel cartridge (EtOAc - 5:5:1 EtOAc/DCM/MeOH over 15 min) to give the title compound as a pale yellow solid (0.468 g, 79%). MS (ESI, pol): calculated for C ₄₁ H ₅₉ ClN ₄ O ₁₂ , 834.4; found 857.2 (M+Na), 817.2 (M-H2O+H).

Maytansine-3-N-methyl-L-alanine-N-Me-beta-alanine (compound 6, Fig. 19).

Method A: maytansine-N-Me-L-Ala-Boc-N-Me-beta-Ala (compound 5, Fig. 19, 0.464 g, 0.555 mmol) was dissolved in acetonitrile/water/trifluoroacetic acid 3:1:1 (7 ml), the flask was sealed with a rubber septum cap, purged with argon, and the reaction mixture was stirred at ambient temperature for 24 h, then closed and stored at -20°C for 3 days. The crude reaction mixture was warmed to ambient temperature for 2 h, concentrated briefly in vacuo, purified on a 100 g C18 RediSep Gold column (20-80% acetonitrile in water for 25 min, 0.1% TFA in both solvents). ) and the combined pure fractions were partially evaporated at ambient temperature, frozen in a dry ice bath and lyophilized to give the title compound as a pale yellow solid (0.295 g, 63%). MS (ESI, pol): calculated for C ₃₆ H ₅₁ ClN ₄ O ₁₀ , 734.3; found 735.7 (M+H), 1471.3 (2M+H).

Method B: maytansine-N-Me-L-Ala-Fmoc-beta-Ala (compound 3, Fig. 19, 0.422 g, 0.441 mmol) was dissolved in 5% piperidine in DMF (6.00 ml, 3.04 mmol) , the reaction flask was sealed with a rubber septum cap, purged with argon, and the mixture was stirred at ambient temperature. After 3 hours, the reaction was terminated by LCMS, so the reaction mixture was concentrated in vacuo, sealed and stored at -20° C. overnight. The crude product was warmed to ambient temperature, treated with acetonitrile and 10% aq. acetic acid (3 ml each) and purified by flash chromatography on a 275 g C18 silica gel column (1090% acetonitrile in water for 20 min, 0.05% acetic acid in both solvents). Lyophilization of product fractions gave the title compound as a white solid. The solid was triturated three times with dry diethyl ether, filtered, the solids were washed off the frit with DCM and the filtrate was evaporated and dried in vacuo to give the title compound as a white solid (0.311 g, 89%). MS (ESI, pol): calculated for C ₃₆ H ₅₁ ClN ₄ O ₁₀ , 734.3; found 735.0 (M+H).

Maytansine-3-N-methyl-L-alanine-N-Me-beta-alanine-carbamyl-(p-amino)benzyl-citrulline-valine Fmoc (compound 7, Fig. 20).

Stage 1: product from the previous stage (compound 6, Fig. 19, 0.310 g, 0.390 mmol), 1-hyd

- 63 042597 roxy-7-azabenzotriazole (HOAT, 0.164 g, 1.20 mmol), sodium bicarbonate (0.138 g, 1.64 mmol) and Fmocvaline-citrulline-(p-amino) benzyl-(p-nitrophenyl) carbonate ( 0.595 g, 0.776 mmol, obtained by a method known in the art, cf. Gangwar et al., US Pat. stirred at ambient temperature. After 24 h, the reaction mixture was partially evaporated in vacuo to a volume of approximately 2-3 ml, treated with 10% aq. acetonitrile and water (approximately 1 ml each), dissolved in acetonitrile (approximately 6 ml) and purified by flash chromatography on a 275 g C18 silica gel column (30-90% acetonitrile in water for 20 min, 0.05% acetic acid in both solvents). Partial evaporation, freezing and lyophilization gave the title compound as a white solid (0.362 g, 68%). MS (ESI, pol): calculated for C ₇₀ H ₈₈ ClN ₉ O ₁₇ , 1361.6; found 1362.1 (M+H), 1384.1 (M+Na), 1344.1 (M-H ₂ O+H).

Step 2: The product from the previous step (0.360 g, 0.264 mmol) was dissolved in 5% piperidine in DMF (7 ml), the reaction flask was sealed with a rubber septum cap, purged with argon, and the mixture was stirred at ambient temperature. After 3 h the reaction mixture was evaporated in vacuo, the residue was treated with 10% aq. acetonitrile (2 ml), dissolved in acetonitrile (4 ml) and purified by flash chromatography on a 275 g C18 silica gel column (10-70% acetonitrile in water over 20 min, 0.05% acetonitrile in both solvents ). Pure fractions were pooled, stored at -20°C overnight, partially evaporated in vacuo at 25-30°C, frozen on dry ice and lyophilized for 6 days to give the title compound as a pale yellow solid (0.303 g, 95%). MS (ESI, pol.): calculated for C ₁₅ H ₇₈ ClN ₉ O ₁₅ , 1139.5; found 1140.1 (M+H), 1162.0 (M+Na).

Maytansine-3-N-methyl-L-alanine-N-Me-beta-alanine-carbamyl-(p-amino)benzyl-citrulline-valinadipic acid (compound 8, Fig. 20). The product from the previous step (compound 7, Fig. 20, 0.205 g, 0.171 mmol), adipic acid (0.258 g, 1.77 mmol) and 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline (EEDQ, 0.215 g, 0.869 mmol ) was dissolved in dry DCM (10 ml) and anhydrous methanol (5 ml), the reaction flask was sealed with a rubber septum cap, purged with argon, and the mixture was stirred at ambient temperature. After 21 hours, the reaction mixture was evaporated in vacuo, the residue was dissolved in a few milliliters of acetonitrile/water and purified by flash chromatography on a 150 g C18 silica gel column (20-80% acetonitrile in water for 17 min, 0.05% acetic acid in both solvents). Partial evaporation, freezing and lyophilization of pure fractions for 18 hours gave the title compound as a white solid (0.140 g, 65%). MS (ESI, pol): calculated for C ₆₁ H ₈₆ ClN ₉ O ₁₈ , 1267.6; found 1268.9 (M+H), 1290.9 (M+Na).

Maytansine-3-N-methyl-L-alanine-N-Me-beta-alanine-carbamyl-(p-amino)benzyl-citrulline-valinadipoyl-succinate (compound 1, Fig. 20). The product from the previous step (compound 8, Fig. 20, 0.061 g, 0.048 mmol), N-hydroxysuccinimide (0.063 g, 0.55 mmol) and N-(3-dimethylaminopropyl)-N'ethylcarbodiimide hydrochloride (EDC-HCl, 0.071 g, 0.37 mmol) was dissolved in dry DCM (7 mL), the reaction flask was sealed with a rubber septum cap, purged with argon, and the mixture was stirred at ambient temperature. After 5 days, the reaction mixture was evaporated in vacuo, the residue was dissolved in a few ml of acetonitrile/water and purified by flash chromatography on a 100 g C18 silica gel column (30-90% acetonitrile in water over 15 min, 0.05% acetonitrile in both solvents). Partial evaporation, freezing and lyophilization of the purest product fractions over 18 hours gave the title compound as a white solid (0.044 g, 67%). MS (ESI, pol): calculated for C ₆₅ H ₈₉ ClN ₁₀ O ₂₀ , 1364.6; found 1365.7 (M+H), 1387.7 (M+Na), 1347.7 (M-H2O+H). 1H-NMR (500 MHz; CDCl ₃ ): δ 7.56 (d, J=8.3 Hz, 2H), 7.20 (d, J=8.7 Hz, 1H), 6.80 (s, 1H), 6.71 (m, 1H), 6.62 (d, J=10.0 Hz, 1H), 6.39 (dd, J=15.1, 11.3 Hz, 1H), 5, 68 (dd, J=15.3, 9.1 Hz, 1H), 5.38-5.32 (m, 1H), 5.03 (t, J=15.1 Hz, 1H), 4.88 (d, J=12.3 Hz, 1H), 4.73 (d, J=11.3 Hz, 1H), 4.61 (dd, J=9.1, 3.6 Hz, 1H), 4 .26 (d, J=7.0 Hz, 1H), 4.17 (t, J=7.1 Hz, 1H), 3.95 (s, 3H), 3.61 (d, J=11, 7 Hz, 1H), 3.57 (d, J=12.4 Hz, 1H), 3.46 (d, J=9.1 Hz, 2H), 3.33 (s, 3H), 3.27 (t, J=6.9 Hz, 1H), 3.173.07 (m, 5H), 2.97 (dd, J=16.6, 9.9 Hz, 1H), 2.88 (d, J= 11.7 Hz, 3H), 2.84 (s, 4H), 2.77 (s, 2H), 2.66 (s, 2H), 2.62 (t, J=4.8 Hz, 2H) , 2.56 (d, J=13.1 Hz, 1H), 2.32 (t, J=6.6 Hz, 2H), 2.15 (d, J=14.0 Hz, 1H), 2 .10 (q, J=6.8 Hz, 1H), 1.92 (s, 4H), 1.75 (m, 5H), 1.61 (s, 3H), 1.52 (s, 3H) , 1.27 (d, J=6.3 Hz, 3H), 1.22 (dt, J=12.7, 6.3 Hz, 6H), 0.95 (t, J=5.9 Hz, 7H), 0.78 (s, 3H).

DM1 was synthesized as a single diastereoisomer based on the procedures described in WO 2015/031396 (eg, example 2, paragraph [00106]), incorporated herein by reference in its entirety.

Example 22 Antibody Conjugation and Conjugate Characterization

Antibody conjugation.

Antibodies (H4H14639D, H4H13312P, H4H14635D and isotopic control; 10-20 mg/ml) in 50 mM HEPES, 150 mM NaCl, pH 8.0 and 10-15% v/v. DMA was conjugated with a 5-6 fold excess

- 64 042597

SMCC diastereoisomer of DM1 prepared as described in Example 21 (maytansinoid A) or maytansine3-N-methyl-L-alanine-N-Me-beta-alanine-carbamyl-(n-amino)benzyl-citrulline-valine-adinoyl- succinate (compound 1, Fig. 20) (maytansinoid B) for 2 hours at ambient temperature. The conjugates were purified by size exclusion chromatography or extensive ultrafiltration and subjected to sterilizing filtration. Protein concentrations were determined using UV spectral analysis. All conjugates used were found to be >90% monomeric by size exclusion HPLC and <1% unconjugated linker load was found by RP-HPLC. All conjugated antibodies were analyzed by UV analysis for linker loading values according to Hamblett et al. (American Association for Cancer Research. 2004 Oct 15;10(20):7063-70) and/or by difference in masses, initial compared to those after conjugation. The ratio of antibodies and load are presented in table. 24.

Table 24. Percent Yield and Load Ratios and Antibody Ratios for Each of the Antibody Drug Conjugates

Antibody Exit (%) DAR (MS) DAR (UV) H4H14639O-maytansinoid A 60 3.8 3.7 H4H14639O-maytansinoid B 50 2.4 2.4 H4H13312P-maytansinoid A 60 4.1 4.1 H4H13312P-maytansinoid B 50 2.3 2.5 Isotype control REOLN945-maytansinoid B 70 2.3 2.5 Isotype control REGN1945-maytansinoid A 80 3.7 3.7

Characterization of conjugates using liquid chromatography with mass spectrometry.

To determine the loading of the linker load on the antibody, the conjugates were subjected to deglycosylation and analyzed by LC-MS.

For analysis, 50 μg of the conjugate was diluted with milli-Q water to a final concentration of 1 mg/mL. 10 µl of PNGase F solution [a solution of PNGase F prepared by adding 150 µl of PNGase F stock solution (New England Biolabs, catalog number P0704L) and 850 µl of milli-Q water and mixing thoroughly] was added to the diluted conjugate solution and then incubated at 37 °C during the night. 5 µl of each sample was injected into LC-MS (Waters Synat G2-Si) and eluted with a gradient of 20-40% with a mobile phase rate of 0.1 ml/minute for 25 min (mobile phase A: 0.1% vol./ v/v FA in H ₂ O, mobile phase B: 0.1% v/v FA in acetonitrile). LC separation was performed on an Acquity BEN C4 column from Waters (1.0 x 50 mm, 1.7 μm) at 80°C.

Mass spectrometric spectra were deconvoluted using Masslynx software and the drug to antibody ratio (DAR) was calculated using the following equations.

1. Relative percentage (%) of drug (Dn) by peak intensity distribution (PI)

(n=0, 1, 2, 3, ..., i).

2. Calculation of the average DAR

DAR = Z(lxDl%+2xD2%+3xD3%+......+ixDi%)

Example 23 Surface Plasma Resonance Binding Affinities and Kinetic Constants of Conjugated Human Monoclonal (Monospecific and Bispecific) Anti-MET Antibodies.

Equilibrium dissociation constants (K _D values) for MET binding to anti-MET antibodies conjugated to either the MCC diastereoisomer DM1 (maytansinoid A) or maytansine-3-Nmethyl-L-alanine-N-Me-beta-alanine-carbamyl- (p-amino)benzyl-citrulline-valine-adipoyl-succinate (compound 1, Fig. 20) (maytansinoid B) was determined using real-time surface plasmon resonance analysis using a biosensor on a Biacore 2000 instrument. Surface The Biacore sensor was derivatized by amino coupling to a mouse anti-human Fc monoclonal antibody (GE Healthcare, No. BR-1008-39) to capture ADCs on wasp

- 65 042597 new anti-MET antibodies and original unmodified antibodies expressed with human constant regions. Biacore binding studies were performed in HEPES buffered saline (HBS) running buffer EP (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% v/v surfactant). active substance P20). Human MET was produced in-house, expressing the C-terminal myc-myc-hexahistidine tag (hMET-mmh). By injecting various concentrations (3-fold dilution) of hMET-mmh (ranging from 30 to 1.1 nM) prepared in HBS-EP running buffer, coated the surface with captured anti-MET ADCs or antibody at a flow rate of 40 µl /min Binding of hMET-mmh to each of the captured ADCs and monoclonal antibodies was monitored for 4 min. Dissociation of hMET-mmh was then monitored for 6 min in HBS-EP running buffer. The surface with antibodies to human Fc was restored by a short injection of 20 mm H ₃ RHO ₄ . All experiments on the kinetics of binding was carried out at 25°C.

The kinetic rate constants of association (k _a ) and dissociation (kd) were determined by fitting real-time sensogram curves to a 1:1 binding model using Scrubber 2.0c curve fitting software. All sensograms were double-corrected to reference values by subtracting the buffer injection sensogram signal from the corresponding analyte sensogram, thereby eliminating artifacts associated with the dissociation of the antibody from the capture surface. The dissociation constants (KD) for equilibrium binding and the dissociation half-life (t1/2) were calculated from the kinetic rate constants as

Kd(M) = ^ and 1½ (min) = ka 60*kd

Binding kinetic parameters for maytansinoid A or maytansinoid B conjugated anti-Met monospecific and bispecific antibodies are shown in Table 1 below. 25, with some experiments performed in duplicate.

Table 25: Biacore Assay Binding Affinities for Anti-MET Conjugated Mono- and Bispecific Monoclonal Antibodies at 25°C

Antibody Captured mAbs (RU) Bound antigen (OE) ka (1/Ms) kd (1/s) Kd (M) 1½ (min) Н4Н13312Р2 148.1±1.2 12.3 2.59E+05 5.35Е-03 2.07Е-08 2.2 Н4Н13312Р2 142.7±0.3 12.1 1.87E+05 4.85Е-03 2.59Е-08 2.4 H4H13312R2 maytansinoid A 232.6±0.5 P,9 1.82E+05 7.18E-03 3.94Е-08 1.6 H4H13312R2 maytansinoid B 263.0±2.6 10.9 1.80E+05 6.32E-03 3.51E-08 1.8 H4H14639D 283.6±4.4 82.8 5.90E+05 1.56Е-03 2.64Е-09 7.4 H4H14639D maytansinoid A 207.7±0.8 55.8 4.95E+05 1.81E-03 3.65Е-09 6.4 H4H14639D maytansinoid B 227.5±0.4 55.4 4.83E+05 1.87Е-03 3.86E-09 6.2 H4H14639D maytansinoid A 284.0±1.1 62.8 4.70E+05 1.76Е-03 3.74Е-09 6.6 H4H14639D maytansinoid B 268.7±0.7 72.8 4.91E+05 1.45Е-03 2.95Е-09 8.0

Example 24 In Vitro Activity Scores of Anti-MET Antibody Drug Conjugates (ADC)

To determine the relative cell killing activity of the anti-MET antibody drug conjugates (ADC) described herein, cell killing assays were performed using several cell lines expressing various levels of endogenous MET. Cell lines EBC-1 (Riken Cell Bank; no. RBRC-RCB1965), MKN-45 (JCRB; no. JCRB0254), NCI-H1993 (ATCC; no. CRL-5909), and J.RT3 (ATCC; no. TIB-153) maintained in RPMI+ 10% FBS + 1X penicillin/streptomycin/L-glutamine (P/S/G), SNU-5 (ATCC; no. CRL-5973) maintained in Iscove's medium + 10% FBS + 1X P/S/G, Hs746t (ATCC; # HTB-135) and HEK293 (ATCC; # 003041) maintained in DME + 10% FBS + 1X P/S/G, MDA-MB-231 (ATCC; # HTB-26) maintained in Leibovitz L- 15 + 10% FBS + 1X P/S/G + 1X non-essential amino acids (NEAA) no CO2, U87MG (ATCC; no. HTB-14) maintained in MEM with Earle's salts + 15% FBS + 1X P/S/G + 1X NEAA, T47D (ATCC; no. HTB133) maintained in RPM1 1640 + 10% FBS + 1X P/S/G + 10 mM HEPES + 1 mM sodium pyruvate + 10

- 66,042,597 μg/ml bovine insulin and A549 (ATCC; #CCL-185) were maintained in Kaighn's F-12 (HAM's F-12K) medium + 10% FBS + 1X P/S/G.

Initially, the relative binding of antibodies to MET was assessed using unconjugated antibodies H4H14635D, H4H14639D and H4H13312P2 in the entire panel of cell lines by flow cytometry. Briefly, 1 x 106 cells were incubated with 10 μg/ml H4H14635D, H4H14639D, H4H13312P2 or isotype control antibody (REGN1945) for 30 min on ice in PBS + 2% FBS (FACS buffer). After one wash with FACS buffer, cells were incubated with 10 μg/ml Alexa647-conjugated anti-human Ig secondary antibody (Jackson ImmunoResearch, no. 109-606170) for 30 minutes on ice. After one additional wash with FACS buffer, samples were fixed with Cytofix (BD Biosciences, no. 554655), filtered with FACS buffer, and analyzed on an iQue flow cytometer (Intelicyte). Mean Fluorescence Intensity (MFI) data were determined using FlowJo software (FlowJo LLC). FACS binding was expressed as fold MFI binding compared to isotype control levels and the results are shown in Table 1. 26. Relative binding for the three anti-MET antibodies was comparable in each cell line, and ranged from 447-fold to 7-fold compared to isotype controls. No detectable binding of any of the 3 anti-MET antibodies tested was observed for T47D, HEK293, or J.RT3 cells.

To measure the in vitro cytotoxicity of anti-MET ADCs, the number of nuclei was assessed 3 or 6 days after ADC treatment. Briefly, cells were seeded in 96-well collagen-coated plates (Greiner, VWR; no. 82050-812) at 750-3000 cells/well in complete growth medium and grown overnight at 37°C, 5% CO ₂ . To obtain cell viability curves, serially diluted ADCs, unconjugated antibodies, or loose loads were added to cells at final concentrations ranging from 100 to 0.01 nM (based on toxin concentration) and incubated for 3 or 6 days at 37°C at 5% CO2. Cells were then treated with 3 μg/ml Hoechst 33342 nuclear dye (Invitrogen, #H3570) while fixing with 4% formaldehyde. Images were acquired on an ImageXpress micro XL (Molecular Devices, Sunnyvale, CA) and core numbers were determined using MetaXpress image analysis software (Molecular Devices, Sunnyvale, CA). Background nuclei from cells treated with 40 nM digitonin were subtracted from all wells and viability was expressed as a percentage of untreated controls. EC ₅₀ values were determined based on a four parameter logistic model equation from a 10 point response curve (GraphPad Prism). The no-treatment condition for each dose-response curve is also included in the analysis, and is presented as the lowest dose. The values of the EU ₅₀ and the destruction of cells in percent are shown in table. 27 and 28.

As shown in Table. 27, H4H14639D anti-MET drug conjugate maytansinoid A specifically reduces cell viability in EBC-1, SNU-5, MKN-45, NCI-H1993 and Hs746t cells with amplified Met with EC ₅₀ values ranging from 0.35 nM up to 0.96 nM. The percentage of destroyed cells (max. % destruction) ranged from 73% to 100%. H4H14639D-maytansinoid A also provided specific killing of 84% of A549 cells with an EC ₅₀ value of 13.91 nM. EC ₅₀ values for H4H14639D-maytansinoid A exceeded 37 nM for low expression (MDA-MB-231 and U87MG) and no expression (T47D, HEK293 and J.RT3) cell lines. Similarly, the conjugated isotype control antibody ensured the destruction of all cell lines with EC ₅₀ values greater than 35 nM. The dimethyl disulphide version of DM1 (MeS-DM1) killed all tested lines with EC ₅₀ values ranging from 0.07 to 2.86 nM.

In a separate experiment, three anti-MET antibodies (H4H14639D, H4H14635D, and H4H13312P2) were conjugated to maytansinoid A or maytansinoid B loads, and in vitro cytotoxicity was assessed in EBC-1, Hs746t, A549, and T47D cells 6 days post-treatment. As shown in Table. 28, all anti-MET antibody drug conjugates produced a strong and specific reduction in cell viability among Met positive cells, with _EC50 values of less than 10 pM for EBC-1 cells, 0.82 nM for Hs746T cells, and 3.5 nM in the case of A549 cells. The percentage of cells killed was greater than 95% for EBC-1 cells, greater than 86% for Hs746T cells, and greater than 72% for A549 cells. T47D cells (Met-negative) did not undergo specific killing by ADC based on anti-MET antibodies. Similarly, conjugated isotype control antibodies produced reduced cell viability among all cell lines tested, with _EC50 values greater than 5 nM for EBC-1 cells, greater than 33 nM for Hs746T cells, and greater than 90 nM for A549 and T47D cells. Unconjugated H4H14639D reduced cell viability among EBC-1, Hs746t and A549 cells, but at a lower percentage than conjugated antibodies. Unconjugated H4H14635D and H4H13312P2 had virtually no effect on viability in any of the cell lines tested. The dimethyl disulfide version of DM1 (MeS-DM1) killed all tested lines with EC ₅₀ values ranging from 0.12 to 1.39 nM. In contrast, M24 (load released from

- 67 042597 maytansinoid B), ensured the destruction of cells with IC50 values > 100 nM.

Table 26 FACS binding of non-conjugated anti-MET antibodies to tumor cell lines

Binding in FACS (MFT fold versus isotype control)

REGN1945

Line Only

cells Unstained H4H14635D H4H14639D H4H13312P2

EVS-1 0.7 0.6 1 263 252 147 SNU-5 1 1.2 1 477 454 235 MKN-45 1 0.8 1 183 156 94 NCIN1993* 1 2' ND ND 188 188 Hs746t 0.8 1.1 1 39 34 27 MDAM B-231 1 h 5.6 1 eleven 12 7 U87MG 1.6 1.7 1 18 18 10 T47D 1 0.9 1 1.3 1 1.4 A549 0.7 : 0.5 1 12 10 7 HEK293 0.2 '0.2 1 1.8 1.8 1.2 J.RT3 0.8 1 1 1.6 1.4 v.v *Expressed in fold as compared to unstained for NCI-H1993.

Table 27. IC50 and max. % ADC kill based on anti-MET antibodies in 3-day in vitro cytotoxicity assay

Antibody drug conjugate EVS-1 SNU-5 MKN-45 NCI-H1993 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction DM1 (MeS-DMl) 2.22 90 1.22 99 2.73 85 2.86 81 H4H14639D 0.82 37 0.30 40 ND 0 ND 0 H4H14639D maytansinoid A 0.96 89 0.40 100 0.35 86 0.41 94 REGN1945 maytansinoid A 35.06 65 >100 14 >100 39 49.42 68

Antibody drug conjugate Hs746t MDA-MB-231 U87MG ICso (nM) Max. % destruction ICso (nM) Max. % destruction ICso (nM) Max. % destruction DM1 (free drug) 1.46 81 1.53 89 0.61 89 H4H14639D 0.42 7 >100 6 >100 6 H4H14639D maytansinoid A 0.56 73 >100 48 100 58 REGN1945 maytansinoid A 33.22 44 >100 42 94.71 58

- 68 042597

Antibody drug conjugate T47D A549 HEK293 J.RT3 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction DM1 (free drug) 1.33 91 2.56 97 0.15 95 0.07 100 H4H14639D >100 0 >100 37 ND 0 >100 5 H4H14639D maytansinoid A >100 6 13.91 84 40.90 65 37.82 59 REGN1945 maytansinoid A >100 1 >100 63 >100 44 39.79 70

Table 28. IC ₅₀ and max. % ADC kill based on anti-MET antibodies in a 6-day in vitro cytotoxicity assay

Antibody drug conjugate EVS-1 Hs746t T47D A549 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction 1С ₅₀ (nM) Max. % destruction DM1 (MeS-DMl) 0.12 62 1.39 88 0.24 96 0.49 90 M24 (load released from maytansinoid B) >100 32 >100 10 >100 0 >100 10 H4H14639D 0.37 66 0.44 35 >100 0 0.17 29 H4H14639D maytansinoid A 0.27 97 0.82 87 >100 3 6.01 86 H4H14639D maytansinoid B 0.01 96 0.86 90 >100 0 3.54 80 Н4Н13312Р2 >100 thirty >100 0 >100 0 >100 7 H4H13312R2 maytansinoid A 0.39 95 1.59 87 >100 6 18.30 89 H4H13312R2 maytansinoid B 0.07 95 0.89 90 >100 3 27.10 85 H4H14635D >100 eleven >100 7 >100 7 >100 0 H4H14635D maytansinoid A 0.76 96 1.76 86 >100 92 6.78 91 H4H14635D maytansinoid B 0.26 96 2.32 89 >100 2 21.40 72 REGN1945 >100 0 >100 0 >100 0 >100 1 REGN1945 maytansinoid A 28.08 93 33.06 76 >100 14 93.40 49 REGN1945 maytansinoid B 5.01 97 >100 0 >100 1 >100 15

Example 25 Efficacy against gastric cancer cells in vivo.

C.B.-17 SCID mice were implanted subcutaneously in the flank with 3 million Hs746T gastric cancer cells. After tumor volumes reached approximately 150 mm ³ , mice were randomized into groups of 6 and treated with control antibodies REGN1945-maytansinoid B or REGN1945 maytansinoid A at 10 mg/kg or H4H14639D-maytansinoid A or H4H14639D-maytansinoid B at 3 or 10 mg/kg. kg. All antibodies were administered three times at a frequency of once a week. Tumor growth was monitored for 37 days after implantation.

The effect of H4H14639D-maytansinoid A or H4H14639D-maytansinoid B on the growth of human tumor xenografts in immunodeficient mice was evaluated and the results are shown in Table 1. 29. Tumors treated with control antibodies, REGN1945-maytansinoid B or REGN1945-maytansinoid A, grew to reach protocol limits in 20 days. Tumors treated with H4H14639D-mitenzinoid A at 3 mg/kg grew to protocol limits in 27 days. Growth of tumors treated with H4H14639D maytansinoid B at 3 mg/kg was inhibited throughout the experiment. Treatment of tumors with H4H14639D-maytansinoid A or H4H14639D-maytansinoid B at 10 mg/kg resulted in the induction of tumor size regression compared to the start of treatment.

- 69 042597

Table 29. Tumor growth in SCID mice treated with anti-Met-C antibody conjugates

Antibody (mg/kg) Tumor growth (mm ³ ) since start of treatment (mean ± SD) KEOLP945-maytansinoid A 10 mg/kg 1244±199 KEOLP945-maytansinoid B 10 mg/kg 1345±121 H4H14639O-maytansinoid A 3 mg/kg 832±15 H4H14639O-maytansinoid A 10 mg/kg -148±0.17 H4H14639O-maytansinoid B 3 mg/kg 19 ± 147 H4H14639O-maytansinoid B 10 mg/kg -137±0

Example 26 Efficacy against lung cancer cells in vivo.

C.B.-17 SCID mice were implanted subcutaneously in the flank with 5 million EBC1 lung cancer cells. After tumor volumes reached approximately 170 mm ³ , mice were randomized into groups of 6 and treated with control antibody REGN1945-maytansinoid B at 15 mg/kg or H4H14639P-maytansinoid B at 2.5, 5, 10, or 15 mg/kg. Antibodies were injected twice with a frequency of once a week. Tumor growth was monitored for 73 days after implantation.

The effect of H4H14639D on the growth of human tumor xenografts in immunodeficient mice was evaluated. Tumors treated with the control antibody, REGN1945-maytansinoid B, grew to protocol limits in 24 days (IAUCC protocols required animals with tumors greater than 2 cm in diameter, approximately 1500 mm ³ to be sacrificed). Treatment of tumors with H4H14639D-maytansinoid B at 2.5, 5, 10, or 15 mg/kg resulted in the induction of tumor size regression compared to the start of treatment. The results are shown in table. thirty.

Table 30. Tumor growth in SCID mice treated with anti-Met-C antibody conjugates

Antibody (mg/kg) Tumor growth (mm ³ ) since start of treatment (mean ± SD) KEOLP945-maytansinoid B 15 mg/kg 1106±165 H4H14639O-maytansinoid B 2.5 mg/kg -142±24 H4H14639O-maytansinoid B 5 mg/kg -163±0 H4H14639O-maytansinoid B 10 mg/kg -173±0 H4H14639O-maytansinoid B 15 mg/kg -179±0

Example 26 Efficacy against NSCLC tumors from patients in vivo.

Nude nu/nu mice were subcutaneously implanted in the flank with Met-expressing NSCLC CTG0165 tumors from a patient. After tumor volumes reached approximately 150 mm ³ , mice were randomized into groups of 6 and treated with control antibodies REGN1945-maytansinoid B or REGN1945-maytansinoid A at 10 mg/kg or H4H14639D-maytansinoid A or H4H14639D-maytansinoid B at 3 or 10 mg/kg. kg. All antibodies were administered three times at a frequency of once a week. Tumor growth was monitored for 61 days after implantation.

The effect of H4H14639D-maytansinoid A or H4H14639D-maytansinoid B on the growth of human tumor xenografts in immunodeficient mice was evaluated. Tumors treated with control antibodies, REGN1945-maytansinoid A or REGN1945-maytansinoid B, grew to reach protocol limits in 27 days. Growth of tumors treated with H4H14639E-maytansinoid A or H4H14639D-maytansinoid B at 3 mg/kg was inhibited for 27 days. Treatment of tumors with H4H14639D-maytansinoid A or H4H14639E-maytansinoid B at 10 mg/kg resulted in the induction of tumor size regression compared to the start of treatment. The data are presented in table. 31.

- 70 042597

Table 31. Tumor growth in nude mice treated with anti-Met-C antibody conjugates

Antibody (mg/kg) Tumor growth (mm ³ ) since start of treatment (mean ± SD) KEOH1945-maytansinoid A 10 mg/kg 967±136 KEOH1945-maytansinoid B 10 mg/kg 1537±373 H4H14639O-maytansinoid A 3 mg/kg 154±227 H4H14639O-maytansinoid A 10 mg/kg -141±2.3 H4H14639O-maytansinoid B 3 mg/kg 517±362 H4H14639O-maytansinoid B 10 mg/kg -145±2

Example 27 Epitope mapping based on hydrogen-deuterium exchange (H/D) for anti-MET antibodies H4H13312P2, H4H13306P2 and H4H14639D binding to human MET.

Experiments were carried out to determine specific regions of the human hepatocyte growth factor receptor ectodomain (SEQ ID NO: 155: human MET isoform 1 (Uniprot ID: P08581) expressed with a myc-myc-hexahistidine (.mmh) tag, hereinafter referred to as hMet), with which antibodies to MET H4H13312P2, H4H13306P2 and H4H14639D interact. H4H13312P2 and H4H13306P2 are bivalent monospecific antibodies to MET; H4H14639D is a bispecific antibody containing two heavy chains that bind to distinct epitopes on Met, each of H4H13312P2 and H4H13306P2, respectively, and a universal light chain. (See example 5).

Epitope mapping based on hydrogen-deuterium exchange (H/D) with mass spectrometry (HDX-MS) was used to determine epitopes bound by the antibodies mentioned above. A general description of the HDX process is set forth in, for example, Ehring (1999) Analytical Biochemistry 267(2):252259; and Engen and Smith (2001) Anal. Chem. 73:256A-265A.

Experiment procedure.

For HDX mapping of the epitope(s) to hMET bound by the anti-MET antibodies H4H13312P2, H4H13306P2, and H4H14639D, individual antibodies were separately covalently bound to NHS-activated Sepharose 4 Fast Flow beads (GE Healthcare, Pittsburgh, Pennsylvania). To confirm the epitopes bound by antibodies to MET, two methods were used - free antigen and in the composition of the complex, described below.

For the experimental condition, the free hMET antigen was deuterated for 5.0 min or 10.0 min in PBS buffer prepared with _D2O . ice-cold buffer to stop the reaction at low pH. The eluted sample was manually loaded into a Waters H/DX-MS system consisting of integrated real-time tracking peptide digestion, capture, 9.0-minute liquid chromatography (LC) separation, and MS data collection with Synapt G2-SL systems

In the case of the experimental condition in the complex, hMET was first bound on the beads with H4H13312P2 or H4H13306P2 and then deuterated for 5.0 minutes or 10.0 minutes by incubation in PBS buffer prepared with D2O. The deuterated hMET was eluted and analyzed using the Waters H/DX-MS system mentioned above.

To identify hMET peptides after pepsin digestion, LC-MSE data from a non-deuterated sample was processed and searched for human MET using ProteinLynx Global Server (PLGS) software from Waters. The identified peptides were imported into the DynamX 3.0 program and filtered according to the following two criteria: 1) the minimum number of products per amino acid is 0.3; 2) the limit number of repetitions in the file is 3.0. The DynamX 3.0 software then automatically calculated the difference in deuterium uptake for each identified peptide between free antigen and complexed conditions for both deuterium times of 5 and 10 min. The individual isotopic peak for each peptide, obtained using the DynamX centroid calculation software, was also manually examined to ensure the accuracy of the deuterium uptake calculation.

In general, deuteration deltas above 0.2 were used as a cut-off point to determine a particular binding epitope.

Results.

Using pepsin digestion with real-time tracking

- 71 042597 using a Waters Enzymate™ PEPSIN (2.1 x 30 mm, 5 µm) column, combined with 9.0 minute LC-MSE data collection, a total of 162 pepsin-cleaved MET peptides were reproducibly identified human with traceable deuterium uptake for experiments with both conditions free antigen and as part of a complex when using beads with the H4H13312P2 antibody. These peptides are characterized by 55.7% sequence overlap. Among these peptides, only five were found to have a significantly reduced degree of deuterium uptake after H4H13312P2 binding (as part of a complex) compared to deuteration of the antigen alone (free antigen). The centroid values of these five peptides under both experimental conditions are illustrated in Table 1. 32. The region corresponding to residues 192-204 overlapping for these five peptides, based on HDX data, was determined to be the binding epitope for antibody H4H13312P2.

Table 32. Pepsin-cleaved hMET peptides with reduced deuterium uptake after binding to H4H13312P2

Remains of hMET 5 min deuteration 10 min deuteration "As part of the complex" Mass centroid MN "Free antigen" Mass centroid MN Δ "As part of the complex" Mass centroid MN "Free antigen" Mass centroid MN Δ 192-202 1351.25 1351.83 -0.58 1351.39 1352.27 -0.88 192-203 1482.34 1482.94 -0.60 1482.50 1483.40 -0.90 192-204 1629.84 1630.71 -0.87 1630.01 1631.10 -1.09 193-202 1252.07 1252.79 -0.72 1252.25 1253.08 -0.83 193-203 1383.22 1383.79 -0.57 1383.40 1384.17 -0.77

For the HDX experiment, performed using H4H13306P2 antibody beads, a total of 98 pepsin-cleaved hMET peptides with deuterium uptake monitored were reproducibly identified in both free antigen and complexed experiments. These 98 peptides have 52.1% sequence overlap. Among these peptides, twelve were found to have a significantly reduced degree of deuterium uptake after H4H13306P2 binding (as part of a complex) compared to deuteration of the antigen alone (free antigen).

The centroid values of these twelve peptides under both experimental conditions are illustrated in Table 1. 33. Regions corresponding to residues 305-315 and residues 421-455, overlapping for these peptides, based on HDX data, were identified as the binding epitope for antibody H4H13306P2.

Table 33. Pepsin-cleaved hMET peptides with reduced deuterium uptake after binding to H4H13306P2

Remains of hMET 5 min deuteration 10 min deuteration "As part of the complex" Mass centroid MH+ "Free antigen" Mass centroid MH+ Δ "As part of the complex" Mass centroid MH+ "Free antigen" Mass centroid MH+ Δ 305-312 818.20 818.83 -0.63 818.31 819.13 -0.82 305-315 1161.50 1162.58 -1.08 1161.80 1162.95 -1.15 306-313 818.48 818.97 -0.49 818.71 819.28 -0.57 421-431 1206.24 1206.75 -0.51 1206.28 1206.95 -0.67 421-435 1581.28 1581.84 -0.56 1581.41 1582.09 -0.68 421-438 1941.58 1942.15 -0.57 1941.71 1942.39 -0.68 422-438 1794.58 1795.04 -0.46 1794.72 1795.34 -0.62 439-447 963.90 964.83 -0.93 963.97 965.24 -1.27 439-455 1846.58 1847.79 -1.21 1847.24 1847.85 -0.61 439-456 1960.24 1961.32 -1.08 1960.83 1961.42 -0.59 441-455 1586.30 1587.71 -1.41 1587.33 1587.79 -0.46 442-455 1487.50 1488.50 -1.00 1487.92 1488.54 -0.62

The same method as described above was used to determine binding epitopes for the anti-MET H4H14639D bispecific antibody. Binding epitopes on hMET for H4H14639D determined using this method correspond to the epitopes determined for the original antibodies.

Anti-MET antibody binding epitope H4H13312P2: AA 192-204: VRRLKETKDGFMF (SEQ ID NO: 156) SEQ ID NO: 155.

Anti-MET antibody binding epitope H4H13306P2: AA 305-315: LARQIGASLND (SEQ ID NO: 157) SEQ ID NO: 155 and AA 421-455: FIKGDLTIANLGTSEGRFMQVVVSRSGPSTPHVNF (SEQ ID NO: 158) SEQ ID NO: 155.

-

Claims (9)

The scope of the invention is not limited to the specific embodiments described herein. In fact, various modifications of the present invention in addition to those described herein will be apparent to those skilled in the art from the foregoing description and the accompanying drawings. Such modifications are intended to be covered by the scope of the appended claims. CLAIM 1. Bispecific antigennegative molecule containing the first antigennegative domain (D1); and a second antigen-binding domain (D2), wherein D1 specifically binds the first human MET epitope; and wherein D2 specifically binds a second human MET epitope; and where D1 contains three heavy chain complementarity determining regions (HCDR1, HCDR2 and HCDR3) within a heavy chain variable region (HCVR) containing the amino acid sequence of SEQ ID NO: 58 and three light chain complementarity determining regions (LCDR1, LCDR2 and LCDR3) within the light chain variable region (LCVR) containing the amino acid sequence of SEQ ID NO: 138; and where D2 contains three heavy chain complementarity determining regions (HCDR1, HCDR2 and HCDR3) within a heavy chain variable region (HCVR) containing the amino acid sequence of SEQ ID NO: 82 and three light chain complementarity determining regions (LCDR1, LCDR2 and LCDR3) within the light chain variable region (LCVR) containing the amino acid sequence of SEQ ID NO: 138. 2. Bispecific antigennegative molecule according to claim 1, where D1 contains HCDR1 containing the amino acid sequence of SEQ ID NO: 60; HCDR2 containing the amino acid sequence of SEQ ID NO: 62; HCDR3 containing the amino acid sequence of SEQ ID NO: 64; LCDR1 containing the amino acid sequence of SEQ ID NO: 140; LCDR2 containing the amino acid sequence of SEQ ID NO: 142; and LCDR3 containing the amino acid sequence of SEQ ID NO: 144. 3. Bispecific antigennegative molecule according to claim 1, where D1 contains HCVR containing the amino acid sequence of SEQ ID NO: 58 or an amino acid sequence that is at least 95% identical to it; and LCVR containing the amino acid sequence of SEQ ID NO: 138 or an amino acid sequence that is at least 95% identical to it. 4. Bispecific antigennegative molecule according to claim 1, where D1 contains HCVR containing the amino acid sequence of SEQ ID NO: 58; and LCVR containing the amino acid sequence of SEQ ID NO: 138. 5. Bispecific antigennegative molecule according to claim 1, where D2 contains HCDR1 containing the amino acid sequence of SEQ ID NO: 84; HCDR2 containing the amino acid sequence of SEQ ID NO: 86; HCDR3 containing the amino acid sequence of SEQ ID NO: 88; LCDR1 containing the amino acid sequence of SEQ ID NO: 140; LCDR2 containing the amino acid sequence of SEQ ID NO: 142; and LCDR3 containing the amino acid sequence of SEQ ID NO: 144. 6. Bispecific antigennegative molecule according to claim 1, where D2 contains HCVR containing the amino acid sequence of SEQ ID NO: 82 or an amino acid sequence that is at least 95% identical to it; and LCVR containing the amino acid sequence of SEQ ID NO: 138 or an amino acid sequence that is at least 95% identical to it. 7. Bispecific antigennegative molecule according to claim 1, where D2 contains HCVR containing the amino acid sequence of SEQ ID NO: 82; and LCVR containing the amino acid sequence of SEQ ID NO: 138. 8. Bispecific antigennegative molecule containing the first antigennegative domain (D1); and a second antigen-binding domain (D2), wherein D1 specifically binds the first human MET epitope; and wherein D2 specifically binds a second human MET epitope; and where D1 contains HCDR1 containing the amino acid sequence of SEQ ID NO: 60; HCDR2 containing the amino acid sequence of SEQ ID NO: 62; HCDR3 containing the amino acid sequence of SEQ ID NO: 64; LCDR1 containing the amino acid sequence of SEQ ID NO: 140; LCDR2 containing the amino acid sequence of SEQ ID NO: 142; and LCDR3 containing the amino acid sequence of SEQ ID NO: 144; And D2 contains HCDR1 containing the amino acid sequence of SEQ ID NO: 84; HCDR2 containing the amino acid sequence of SEQ ID NO: 86; HCDR3 containing the amino acid sequence of SEQ ID NO: 88; LCDR1 containing the amino acid sequence - 73 042597 SEQ ID NO: 140; LCDR2 containing the amino acid sequence of SEQ ID NO: 142; and LCDR3 containing the amino acid sequence of SEQ ID NO: 144. 9. Bispecific antigennegative molecule according to claim 1, where (a) the first human MET epitope contains amino acids 192-204 of SEQ ID NO: 155; (b)