Multispecific antigen binding molecules for cell targeting and uses thereof

Abstract

To provide multispecific antigen binding molecules that bind to both of T cell antigen (e.g., CD3) and a target antigen (e.g., tumor associated antigen, viral or bacterial antigen) and comprise a single polypeptide chain being multivalent (e.g., divalent) with respect to T cell antigen binding, and uses thereof.SOLUTION: A multispecific antigen binding molecule of the invention comprises; (a) a first polypeptide comprising, from N terminus to C terminus, (i) a first antigen binding domain that specifically binds to a T cell antigen, (ii) a first multimerizing domain, and (iii) a second antigen binding domain that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N terminus to C terminus, (i) a third antigen binding domain that specifically binds to a target antigen and (ii) a second multimerizing domain, where the first and the second multimerizing domains associate each other.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] Sequence Listing Reference This application incorporates by reference a Sequence Listing, which was created on Aug. 7, 2020, and which is submitted in computer readable format as file 10606WO01-Sequence.txt, which contains 64,570 bytes.

[0002] Technical Field The present invention relates to alternative formats of multivalent antigen binding proteins and methods of their use. Multivalent antigen binding proteins, including bispecific and multispecific molecules, comprise a first polypeptide chain having both N-terminal and C-terminal antigen binding domains that specifically bind to a T cell antigen (e.g., CD3), and a second polypeptide chain that comprises at least one antigen binding domain that binds to a target antigen (e.g., a tumor cell antigen). [Background technology]

[0003] Bispecific and multispecific antibodies and antigen-binding molecules are known in the art (see, for example, Non-Patent Document 1). Among such known formats are FcFc* (structure in FIG. 1A), a conventional bispecific antibody with a Fab antigen-binding domain on either arm of the antibody, and an Fc region with a modified CH3 domain that alters protein A binding affinity to allow for the isolation of heterodimers from homodimeric impurities (ibid., p. 184; Fig. 2, panel 7, last structure). This conventional bispecific antibody format was used to create bispecific antibodies in which one arm of the antibody targets a tumor cell antigen and the second arm targets a T cell antigen such as CD3. Another conventional format is IgG-HC-scFv (structure in FIG. 1B), a bispecific antibody in which two N-terminal Fab domains bind a first antigen and two scFv domains linked to the C-terminus of the Fc region bind a second antigen (ibid., p. 184; Fig. 2, panel 10, first structure). There is a need in the art for new and useful formats for bispecific or multispecific antigen-binding molecules that improve desired functionality. Although Brinkmann et al. generally refer to "building blocks" for the generation of homodimeric or heterodimeric antigen-binding molecules (p.183, Fig. 1), the possibilities are virtually limitless, and reportedly only those molecules shown in Fig. 2 (p.184) have been prepared. Furthermore, Brinkmann does not contemplate molecules that contain specific antigen-binding domains, particularly T cell antigen-binding domains at both the N-terminus and C-terminus of a single polypeptide chain that forms part of a multispecific molecule. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Brinkmann and Kontermann,MABS,9(2):182-212,2017 Summary of the Invention

[0005] In general, the invention provides multispecific antigen-binding molecules that bind both a T cell antigen (TCA) (e.g., CD3) and a target antigen (TA) (e.g., a tumor-associated antigen, a viral or bacterial antigen) and comprise a single polypeptide chain that is multivalent (e.g., bivalent) with respect to T cell antigen binding.

[0006] In one aspect, the invention provides a multispecific antigen binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first antigen binding domain that specifically binds to a T cell antigen, (ii) a first multimerization domain, and (iii) a second antigen binding domain that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third antigen binding domain that specifically binds to a target antigen, and (ii) a second multimerization domain, wherein the first and second multimerization domains associate with each other to form the molecule.

[0007] In some embodiments, the second polypeptide further comprises a fourth antigen binding domain at the C-terminus of the second multimerization domain. In some cases, the fourth antigen binding domain specifically binds to a target antigen. In some cases, the third antigen binding domain and the fourth antigen binding domain specifically bind to different target antigens. In some cases, the different target antigens are expressed (or present) on the surface of the same cell. In some cases, the different target antigens are expressed (or present) on the surface of different cells. Reference herein to a target antigen expressed (or present) on the surface of a cell includes both a protein expressed by the cell that is embedded in or spans the membrane of the cell, and a peptide presented in the context of a groove of a major histocompatibility complex (MHC) protein by the cell. In some cases, the third antigen binding domain and the fourth antigen binding domain specifically bind to the same target antigen. In some embodiments, the fourth antigen binding domain specifically binds to a T cell antigen. In some cases, the first antigen binding domain and the second antigen binding domain specifically bind to the same T cell antigen. In some embodiments, the first antigen binding domain and the second antigen binding domain specifically bind to different T cell antigens. In some embodiments, the first antigen binding domain specifically binds to a first T cell antigen that is a costimulatory molecule, and the second antigen binding domain specifically binds to a second T cell antigen that is a checkpoint inhibitor. In some embodiments, the costimulatory molecule is CD28, and the checkpoint inhibitor is PD-1. In some embodiments, the first, second, and fourth antigen binding domains specifically bind to the same T cell antigen. In some embodiments, the first, second, and fourth antigen binding domains specifically bind to different T cell antigens. In some embodiments, the first and fourth antigen binding domains specifically bind to the same T cell antigen. In some embodiments, the second and fourth antigen binding domains specifically bind to the same T cell antigen.

[0008] In various embodiments, one or more of the antigen binding domains are Fab. In various embodiments, one or more of the antigen binding domains are scFv. In some embodiments, the multispecific molecule contains both Fab and scFv antigen binding domains. In some embodiments, the first antigen binding domain and the third antigen binding domain are Fab. In some embodiments, the second antigen binding domain is an scFv. In some embodiments, the fourth antigen binding domain is an scFv. In some embodiments, the first, second, and third antigen binding domains are Fab. In some embodiments, the first and third antigen binding domains are Fab domains and the second antigen binding domain is an scFv domain. In some embodiments, the first, second, third, and fourth antigen binding domains are Fab domains. In some embodiments, the first, second, third, and fourth antigen binding domains are Fab domains. In some embodiments, the first and third antigen binding domains are Fab domains and the second and fourth antigen binding domains are scFv domains. In some cases, the first, second, third, and fourth antigen-binding domains are Fab domains. In some cases, the first and third antigen-binding domains are Fab domains and the second and fourth antigen-binding domains are scFv domains. In some cases, the first, second, third, and fourth antigen-binding domains are Fab domains.

[0009] In any embodiment in which the antigen binding domain is an scFv domain, the scFv domain may comprise a heavy chain variable region (HCVR) that comprises a cysteine ​​mutation at residue 44, and a light chain variable region that comprises a cysteine ​​mutation at residue 100 (Kabat numbering). In some embodiments, the scFv comprises a HCVR and a LCVR linked together via a polypeptide linker of 10-30 amino acids, optionally a (G4S)4 linker. In some embodiments, the scFv is connected to the C-terminus of the first and / or second multimerization domain via a linker of 5-25 amino acids, optionally a (G4S)3 linker.

[0010] In some embodiments, the T cell antigen is a T cell receptor complex antigen (i.e., any of the protein subunits that make up the T cell receptor complex). Optionally, the T cell antigen is CD3. Optionally, the T cell antigen is a costimulatory molecule on a T cell or a checkpoint inhibitor. In some embodiments, the T cell antigen is selected from the group consisting of CD27, CD28, 4-1BB, and PD-1. In some embodiments, the T cell antigen is selected from the group consisting of CD3, CD27, CD28, 4-1BB, and PD-1.

[0011] In some embodiments, the target antigen is a tumor-associated antigen. In some embodiments, the target antigen is a viral or bacterial antigen. In some embodiments, the target antigen is a fungal antigen or a parasitic antigen.

[0012] In some embodiments, the first and second multimerization domains are immunoglobulin Fc domains. Optionally, the first and second multimerization domains are human IgG1 or human IgG4 Fc domains. Optionally, the first and second multimerization domains comprise an immunoglobulin hinge domain, a CH2 domain, and a CH3 domain of a human IgG polypeptide (e.g., IgG1, IgG2, IgG3, or IgG4). Optionally, the first and second multimerization domains comprise a hinge domain, a CH2 domain, and a CH3 domain of a human IgG1 polypeptide. Optionally, the first and second multimerization domains comprise a hinge domain, a CH2 domain, and a CH3 domain of a human IgG4 polypeptide. In some embodiments, the first and second multimerization domains associate with each other via disulfide bonds.

[0013] In some embodiments, the first multimerization domain or the second multimerization domain comprises an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type Fc domain of the same isotype. Optionally, the amino acid substitution comprises an H435R modification, or an H435R and a Y436F modification (EU numbering). Optionally, the first multimerization domain comprises an H435R and a Y436F modification. Optionally, the second multimerization domain comprises an H435R and a Y436F modification. In some embodiments, the first polypeptide, the second polypeptide, or both the first and second polypeptides comprise an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type hinge domain of the same isotype. Fcγ It contains a modified hinge domain which reduces the binding affinity for the receptor.

[0014] In another aspect, the invention provides a multispecific antigen-binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to the target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule.

[0015] In some embodiments, the second Fab and the second scFv specifically bind to different target antigens. In some cases, the different target antigens are expressed on the surface of the same cell. In some embodiments, the second Fab and the second scFv specifically bind to the same target antigen.

[0016] In another aspect, the invention provides a multispecific antigen-binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to the target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule.

[0017] In some embodiments, the third Fab and the fourth Fab specifically bind to different target antigens. In some cases, the different target antigens are expressed on the surface of the same cell. In some embodiments, the third Fab and the fourth Fab specifically bind to the same target antigen.

[0018] In another aspect, the invention provides a multispecific antigen-binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a T cell antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule.

[0019] In another aspect, the invention provides a multispecific antigen-binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, and (ii) a second immunoglobulin Fc domain, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule.

[0020] In various embodiments, such as any of those described above or mentioned herein, the T cell antigen is a T cell receptor complex antigen (i.e., any of the protein subunits that make up the T cell receptor complex). In some embodiments, the T cell antigen is CD3. In some embodiments, the T cell antigen is a costimulatory molecule on a T cell or a checkpoint inhibitor. In some embodiments, the T cell antigen is selected from the group consisting of CD27, CD28, 4-1BB, and PD-1. In some embodiments, the T cell antigen is selected from the group consisting of CD3, CD27, CD28, 4-1BB, and PD-1.

[0021] In various embodiments, such as any of those described above or mentioned herein, the target antigen is a tumor-associated antigen. In some embodiments, the target antigen is a viral or bacterial antigen. In some embodiments, the target antigen is a fungal antigen or a parasitic antigen.

[0022] In some embodiments, such as any of those described above or mentioned herein, the first and second multimerization domains are immunoglobulin Fc domains. In some cases, the first and second multimerization domains are human IgG1 or human IgG4 Fc domains. In some cases, the first and second multimerization domains comprise the immunoglobulin hinge domain, CH2 domain, and CH3 domain of a human IgG polypeptide (e.g., IgG1, IgG2, IgG3, or IgG4). In some cases, the first and second multimerization domains comprise the hinge domain, CH2 domain, and CH3 domain of a human IgG1 polypeptide. In some cases, the first and second multimerization domains comprise the hinge domain, CH2 domain, and CH3 domain of a human IgG4 polypeptide. In some embodiments, the first and second multimerization domains associate with each other via disulfide bonds.

[0023] In some embodiments, such as any of those described above or mentioned herein, the first multimerization domain or the second multimerization domain comprises an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type Fc domain of the same isotype. In some cases, the amino acid substitution comprises an H435R modification, or an H435R and a Y436F modification (EU numbering). In some cases, the first multimerization domain comprises an H435R and a Y436F modification. In some cases, the second multimerization domain comprises an H435R and a Y436F modification. In some embodiments, the first polypeptide, the second polypeptide, or both the first and second polypeptides comprise an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type hinge domain of the same isotype. Fcγ It contains a modified hinge domain which reduces the binding affinity for the receptor.

[0024] In another aspect, the present invention provides a pharmaceutical composition comprising any one of the multispecific molecules described above or discussed herein and a pharma- ceutically acceptable carrier or diluent.

[0025] In another aspect, the present invention provides a method of treating cancer comprising administering to a subject in need of such treatment any one of the multispecific molecules described above or discussed herein.

[0026] In another aspect, the present invention provides a method of treating an infectious disease comprising administering any one of the multispecific molecules described above or discussed herein to a subject in need of such treatment. Optionally, the infectious disease is a bacterial infection. Optionally, the infectious disease is a viral infection. Optionally, the infectious disease is a fungal infection. Optionally, the infectious disease is a parasitic infection.

[0027] In various embodiments, the target antigen is present at a density of 10-10,000,000 copies per target cell. In various embodiments, the target antigen is present at a density of 100-10,000,000 copies per target cell. In various embodiments, the target antigen is present at a density of 100-1,000,000 copies per target cell. In some embodiments, the target antigen is present at a density of 50-10,000. In some embodiments, the target antigen is present at a density of 100-5000. In some embodiments, the target antigen is present at a density of 100-20,000. In some embodiments, the target antigen is present at a density of 500-1,000,000 copies per target cell. In some embodiments, the target antigen is present at a density of 1000-20,000 copies per target cell. In some embodiments, the target antigen is present at a density of greater than 20,000 copies per target cell. In various embodiments, the target antigen is administered at about 10, about 50, about 100, about 200, about 300, about 400, about 500, about 1000, about 2000, about 3000, about 4000, about 5000, about 6000, about 7000, about 8000, about 9000, about 10,000, about 15,000, about 20,000, about 25,000, about 50,000, about 75, A "low density antigen" is an antigen that is present at a density of about 5000 copies or less on a target cell. References to low density antigen include where a cell has 4000 copies or less, 3000 copies or less, 2000 copies or less, 1000 copies or less, 900 copies or less, 800 copies or less, 700 copies or less, 600 copies or less, 500 copies or less, 400 copies or less, 300 copies or less, 200 copies or less, 100 copies or less, or 50 copies or less of the target antigen.

[0028] In various embodiments, the multispecific molecule is administered in combination with a second therapeutic agent to treat a disease or disorder. In some cases, the second therapeutic agent comprises a bispecific antigen-binding molecule comprising a first antigen-binding domain that binds to a target antigen (TA) and a second antigen-binding domain that binds to a T cell antigen. In some cases, the target antigen is a tumor cell antigen. In some embodiments, the second therapeutic agent comprises a bispecific anti-TA x anti-CD28 antibody. In some embodiments, the second therapeutic agent comprises a bispecific anti-EGFR x anti-CD28 antibody. In some embodiments, the second therapeutic agent comprises an antibody that binds to a checkpoint inhibitor on T cells. In some embodiments, the second therapeutic agent comprises an anti-PD-1 antibody. In some cases, the multispecific molecule is administered in combination with two or more second therapeutic agents.

[0029] In another aspect, the invention provides the use of any one of the multispecific molecules discussed above or herein in the manufacture of a medicament for treating a disease or disorder (e.g., cancer, or an infectious disease) in a subject in need of such treatment.

[0030] In another aspect, the present invention provides the use of any one of the multispecific molecules described above or discussed herein in medicine, or to treat a disease or disorder (e.g., cancer, or an infectious disease).

[0031] In another aspect, the present invention provides multispecific molecules, as described above or discussed herein, for use in medicine or for treating a disease or disorder (e.g., cancer, or an infectious disease).

[0032] In any of the embodiments discussed above or herein, the target antigen may be a peptide in the context of a groove of a major histocompatibility complex (MHC) protein. In various embodiments, any of the features or components of the embodiments described above or discussed herein may be combined, and such combinations are encompassed within the scope of the present disclosure. Any particular value described above or discussed herein may be combined with another associated value described above or discussed herein to recite a range where those values ​​represent the upper and lower limits of the range, and such ranges are encompassed within the scope of the present disclosure.

[0033] Other embodiments will be apparent from review of the detailed description that follows. [Brief description of the drawings]

[0034] [Figure 1-1] Figures 1A and 1B show known bispecific antibody and antigen-binding molecule formats. Figures 1C, 1E, 1F, 1G, 1H, 1I, 1J, 1K, 1L, 1M, 1N, 1O, 1P, 1Q, 1R, and 1S show bispecific or multispecific antigen-binding molecule formats according to embodiments of the present invention. In each of these formats, a first polypeptide chain includes both N-terminal and C-terminal antigen-binding domains (e.g., Fab or scFv) that specifically bind to a T cell antigen (TCA) (e.g., CD3), and a second polypeptide chain includes at least one antigen-binding domain (e.g., Fab or scFv) that specifically binds to a target antigen (TA) (e.g., tumor cell antigen). Figure 1D shows a format in which two antigen-binding domains that specifically bind to a T cell antigen (e.g., CD3) are located on different polypeptide chains (the N-terminus on one polypeptide chain, and the C-terminus on the second polypeptide chain). [Figure 1-2] Same as above. [Diagram 2] 1A, 1B, and 1C, compared to a T cell only control (zero) and a positive control. None of the molecules activated T cells in the absence of target cells. [Diagram 3]The cytotoxic activity of molecules having each of the formats shown in Figures 1A, 1B, and 1C in the presence of human PBMC and target cells (A375) compared to the positive control inducing maximal cell killing. The CD3 binding domain of the molecules comprises the variable region of the 7221G anti-CD3 antibody. The molecule having the structure of Figure 1C was significantly more potent than the molecule having the structure of Figures 1A and 1B. [Figure 4A] Figures 4A, 4B, and 4C show the cytotoxic activity of molecules having each of the formats shown in Figures 1A, 1B, and 1C in the presence of human PBMCs and target cells (A375) in combination with an anti-PD-1 antibody (Figure 4A), a costimulatory bispecific EGFRxCD28 antibody (Figure 4B), or both an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody (Figure 4C), compared to the positive control inducing maximal cell killing. The CD3 binding domain of the molecule comprises the variable region of the 7221G anti-CD3 antibody. The molecule having the structure of Figure 1C was significantly more potent in combination with these additional antibodies than the molecules having the structures of Figures 1A and 1B. [Figure 4B] Same as above. [Figure 4C] Same as above. [Diagram 5] Figures 4A, 4B, and 4C show measured cytokine levels for a molecule having the structure of Figure 1C (right panel) compared to a molecule having the structure of Figure 1A (left panel) at the maximum antibody concentration shown. The CD3 binding domain of the molecule comprises the variable region of the 7221G anti-CD3 antibody. The molecule having the structure of Figure 1C does not exhibit a greater level of cytokine release, despite significantly greater cytotoxic activity. [Figure 6A]Figures 6A, 6B, 6C, and 6D show the binding of a molecule with the structure of Figure 1C and a modified version of this molecule (with an inactive domain - indicated by an X in the legend) to Raji cells (Figure 6A) or A375 cells (Figure 6C) overexpressing the MAGEA4 peptide, or to CD3+ Jurkat cells (Figures 6B and 6D). The CD3 binding domain of the molecule shown in Figures 6A and 6B comprises the variable region of the 7195P anti-CD3 antibody. The CD3 binding domain of the molecule shown in Figures 6C and 6D comprises the variable region of the 7221G anti-CD3 antibody. As shown in Figures 6A, 6B, 6C, and 6D, the presence of two active antigen binding domains improved binding to the target antigen, and similar binding was observed regardless of the source of the anti-CD3 binding domain. As shown in these figures, binding was most affected when the N-terminal Fab domain was removed. [Figure 6B] Same as above. [Figure 6C] Same as above. [Figure 6D] Same as above. [Figure 7A] Figures 7A and 7B show the cytotoxic activity of the same molecules shown in Figures 6A and 6B (Figure 7A) and Figures 6C and 6D (Figure 7B). Molecules with the structure of Figure 1C showed the greatest cytotoxic potency, followed by molecules with two active T cell antigen (e.g., CD3) binding domains. Similar patterns of cytotoxicity were observed regardless of the source of the anti-CD3 binding domains. [Figure 7B] Same as above. [Figure 8A] Figures 8A and 8B show the binding of a molecule having the structure of Figure 1C and modified versions of this molecule (with a C-terminal Fab domain or an inactive domain - indicated by an X in the legend) to Raji cells (Figure 8A) or CD3+ Jurkat cells (Figure 8B) overexpressing the MAGEA4 peptide. The CD3 binding domain of the molecule comprises the variable region of the 7195P anti-CD3 antibody. As shown in Figures 8A and 8B, the C-terminal scFv domain provided superior binding to the target antigen compared to the C-terminal Fab domain. [Figure 8B] Same as above. [Figure 9]The cytotoxic activity of the same molecules shown in Figures 8A and 8B is shown. The CD3 binding domain of the molecules comprises the variable region of the 7195P anti-CD3 antibody. The molecule having the structure of Figure 1C showed the greatest cytotoxic potency, followed by the molecule having the structure of Figure 1E. [Figure 10A] Figures 10A and 10B show the binding of molecules having the structures of Figures 1C and 1D to A375 cells overexpressing the MAGEA4 peptide (Figure 10A), or to CD3+ Jurkat cells (Figure 10B). The CD3 binding domain of the molecule comprises the variable region of the 7221G anti-CD3 antibody. The two molecules showed similar binding to both cell types compared to each other. [Figure 10B] Same as above. [Figure 11A] Figures 11A and 11B show the cytotoxic activity of the same molecules shown in Figures 10A and 10B against A375 cells from two different donor sources. The CD3 binding domain of the molecule comprises the variable region of the 7221G anti-CD3 antibody. The molecule with the structure of Figure 1C was more potent than the molecule with the structure of Figure 1D. [Figure 11B] Same as above. [Figure 12A] Figures 12A and 12B show the relative cytotoxic activity and potency of molecules having the structures of Figures 1C and 1F, respectively, compared to molecules having the structure of Figure 1A. The molecules were tested individually and in combination with a costimulatory bispecific EGFRxCD28 antibody and an anti-PD-1 antibody, as discussed in Example 7. The CD3 binding domain of the molecules comprises the variable region of the 7195P anti-CD3 antibody. The molecule having the structure of Figure 1F targets two different epitopes of the same target antigen with two TA antigen binding domains, while the molecule having the structure of Figure 1C targets the same epitope of the target antigen with two TA antigen binding domains. The molecule having the structure of Figure 1F was more potent than the molecule having the structure of Figure 1C, and both molecules were more potent than the molecule having the structure of Figure 1A. In both cases, the combination of these molecules with a costimulatory bispecific antibody and an anti-PD-1 antibody resulted in even greater cytotoxic potency, similar to the results shown in Figures 4A-4C. [Figure 12B] Same as above. [Figure 13] Relative binding affinities are shown for molecules having the structure of FIG. 1F, in which the CD3 binding domain is derived from an anti-CD3 antibody with strong, medium, or weak binding affinity to CD3. The "strong" binding domain is derived from the 7195P anti-CD3 antibody. The "medium" binding domain is derived from the 7221G anti-CD3 antibody. The "weak" binding domain is derived from the 7221G20 anti-CD3 antibody. For example, references to "strong / strong" refer to the Fab and scFc anti-CD3 binding domains, respectively. As expected, binding to CD3-positive Jurkat cells correlates with the strength of affinity of the anti-CD3 binding domain in the molecule. [Figure 14A] Figures 14A and 14B show the relative cytotoxic activity and potency of the molecules shown in Figure 13 in MAGEA4 positive A375 cells. The molecules were tested individually (Figure 14A) and in combination with a costimulatory bispecific EGFRxCD28 antibody and an anti-PD-1 antibody (Figure 14B) as discussed in Example 8. There is a clear correlation between the strength of the anti-CD3 binding domain and the potency of the molecule. The "control" is a positive control that targets the scaffold of all HLA molecules to provide maximum cytotoxicity for comparison with other molecules. [Figure 14B] Same as above. [Figure 15A] Figures 15A and 15B show the relative cytotoxic activity and potency of the molecules shown in Figure 13 in MAGEA4 positive ScaBER cells. The molecules were tested individually (Figure 15A) and in combination with a costimulatory bispecific EGFRxCD28 antibody and an anti-PD-1 antibody (Figure 15B) as discussed in Example 8. There is a clear correlation between the strength of the anti-CD3 binding domain and the potency of the molecule. The "control" is a positive control that targets the scaffold of all HLA molecules to provide maximum cytotoxicity for comparison with other molecules. [Figure 15B] Same as above. [Figure 16A]Figures 16A, 16B, and 16C show the relative binding affinity for molecules with the structures of Figures 1A (molecule C), 1C (molecule B), and IF (molecule A and D) to NYESO-1 positive cells (Figure 16A), MAGEA4 (peptide 1) positive cells (Figure 16B), and MAGEA4 (peptide 2) positive cells (Figure 16C). As expected, molecule D, which does not have a NYESO-1 binding domain, does not bind to NYESO-1 expressing cells (Figure 16A), and molecules lacking the related MAGEA4 binding domain do not bind to MAGEA4 expressing cells, as shown in Figures 16B and 16C. The CD3 binding domain of the molecule contains the variable region of the 7195P anti-CD3 antibody. The "HLA targeting bispecific" positive control binds HLA molecules and CD3. The "isotype control multispecific" is a molecule with the structure of Figure 1C with a binding domain to an unrelated target antigen. [Figure 16B] Same as above. [Figure 16C] Same as above. [Figure 17A]Figures 17A and 17B show the relative cytotoxic activity and potency of molecules with the structures of Figures 1C and 1F, respectively, compared to a positive control with the structure of Figure 1A, which binds HLA molecules and CD3. Isotype controls included a molecule with the structure of Figure 1C, which has a binding domain to an unrelated target antigen, and a molecule with the structure of Figure 1A, which has a binding domain to CD3 and an unrelated target antigen. The molecules were tested individually and in combination with a costimulatory bispecific EGFRxCD28 antibody and an anti-PD-1 antibody, as discussed in Example 9. The CD3 binding domain of the molecule comprises the variable region of the 7195P anti-CD3 antibody. The molecule with the structure of Figure 1F targets two different antigens (NYESO-1 and MAGEA4) with two TA antigen binding domains, while the molecule with the structure of Figure 1C targets a single antigen with both two TA antigen binding domains. The molecule with the structure of Figure 1F, which targets two different antigens, was more potent than the molecule with the structure of Figure 1C. In each case, the combination of these molecules with costimulatory bispecific antibodies and anti-PD-1 antibodies resulted in greater cytotoxic potency compared to the molecules alone, similar to the results shown in Figures 4A-4C. [Figure 17B] Same as above. [Figure 17C] Figures 17C and 17D show the relative T cell activation of the molecules discussed in connection with Figures 17A and 17B. [Figure 17D] Same as above. [Figure 18A]Figures 18A and 18B show the relative cytotoxic activity and potency of molecules with the structures of Figures 1C and 1F, respectively, compared to molecules with the structure of Figure 1A. The positive control is a molecule with the structure of Figure 1A, which binds a human leukocyte antigen (HLA) molecule and CD3. Isotype controls included a molecule with the structure of Figure 1C, which has a binding domain to an unrelated target antigen, and a molecule with the structure of Figure 1A, which has a binding domain to CD3 and an unrelated target antigen. The molecules were tested individually and in combination with a costimulatory bispecific EGFRxCD28 antibody and an anti-PD-1 antibody, as discussed in Example 9. The CD3 binding domain of the molecule comprises the variable region of the 7195P anti-CD3 antibody. As shown in Figure 18A, the molecule with the structure of Figure 1F (targeting two different epitopes of MAGEA4) is more potent than the molecule with the structure of Figure 1C (targeting a single epitope with both TA binding domains), and both molecules are more potent than the molecule with the structure of Figure 1A. Similarly, as shown in Figure 18B, the molecule with the structure of Figure 1F (targeting two different antigens) is more potent than the molecule with the structure of Figure 1C (targeting a single antigen with both TA-binding domains), and both molecules are more potent than the molecule with the structure of Figure 1A. In each case, combination of these molecules with a costimulatory bispecific antibody and an anti-PD-1 antibody resulted in even greater cytotoxicity potency compared to the molecules alone, similar to the results shown in Figures 4A-4C. [Figure 18B] Same as above. [Figure 18C] Figures 18C, 18D, 18E, and 18F show the relative T cell activation of the molecules discussed in connection with Figures 18A and 18B. [Figure 18D] Same as above. [Figure 18E] Same as above. [Figure 18F] Same as above. [Figure 19A]Figures 19A and 19B show the cytotoxic activity and potency, and T cell activation, respectively, of a molecule having the structure of Figure 1F compared to a combination of two molecules having the structure of Figure 1A, which binds to the same pair of target antigens as the molecule having the structure of Figure 1F. As shown in Figures 19A and 19B, the molecule having the structure of Figure 1F more potently kills tumor cells and increases T cell activation than the combination of two molecules having the structure of Figure 1A. [Figure 19B] Same as above. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0035] Before the present invention is described in further detail, it is to be understood that the present invention is not limited to the particular methods and experimental conditions described, as such methods and conditions may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, as the scope of the present invention will be limited only by the appended claims.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.As used herein, the term "about" when used in relation to a specific recited numerical value means that the value may vary by 1% or less from the recited value.For example, as used herein, the expression "about 100" includes 99 and 101 and all values ​​therebetween (for example, 99.1, 99.2, 99.3, 99.4, etc.).

[0037] Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred methods and materials are now described. All patents, applications, and non-patent publications mentioned herein are incorporated by reference in their entirety.

[0038] definition The term "T cells" refers to immune cells that express CD3, including CD4+ cells (helper T cells), CD8+ cells (cytotoxic T cells), regulatory T cells (Tregs), and tumor-infiltrating lymphocytes.

[0039] The term "T cell antigen" refers to a cell surface expressed protein present on a T cell, including a "costimulatory molecule." A "costimulatory molecule" refers to a protein expressed by a T cell that binds to a cognate ligand or receptor (e.g., on an antigen presenting cell) to provide a stimulatory signal that, in combination with the primary signal provided by engagement of the T cell's TCR with a peptide / MHC, stimulates the activity of the T cell. Stimulation of a T cell can include activation, proliferation, and / or survival of the T cell.

[0040] As used herein, the phrase "cell surface expression" or "cell surface molecule" refers to one or more proteins expressed on the surface of a cell, in vitro or in vivo, where at least a portion of the protein is exposed on the extracellular face of the cell membrane and is accessible to the antigen-binding portion of an antibody or antigen-binding domain of a multispecific antigen-binding molecule discussed herein.

[0041] The term "CD3", as used herein, refers to an antigen expressed on T cells as part of the multi-molecule T cell receptor (TCR) and consisting of a homodimer or heterodimer formed from the association of two of the four receptor chains, namely CD3-epsilon, CD3-delta, CD3-zeta, and CD3-gamma. All references herein to proteins, polypeptides, and protein fragments are intended to refer to the human version of the respective protein, polypeptide, or protein fragment, unless specifically identified as being from a non-human species. Thus, the term "CD3" refers to human CD3, unless specifically identified as being from a non-human species, e.g., "mouse CD3", "monkey CD3", etc.

[0042] As used herein, "antibodies that bind CD3" or "anti-CD3 antibodies" include antibodies and antigen-binding fragments thereof that specifically recognize a single CD3 subunit (e.g., epsilon, delta, gamma, or zeta), as well as antibodies and antigen-binding fragments thereof that specifically recognize a dimeric complex of two CD3 subunits (e.g., gamma / epsilon, delta / epsilon, and zeta / zeta CD3 dimers). The antigen-binding domains of the present invention may bind to soluble CD3 and / or cell surface expressed CD3. Soluble CD3 includes native CD3 protein, as well as recombinant CD3 protein variants that lack a transmembrane domain or are not associated with a cell membrane, such as monomeric and dimeric CD3 constructs.

[0043] As used herein, the term "cell surface expressed CD3" refers to one or more CD3 proteins expressed on the cell surface in vitro or in vivo, where at least a portion of the CD3 protein is exposed on the extracellular face of the cell membrane and is accessible to the antigen-binding portion of an antibody. "Cell surface expressed CD3" includes CD3 proteins contained within functional T cell receptors in the cell membrane. The term "cell surface expressed CD3" includes CD3 proteins expressed as part of homodimers or heterodimers on the surface of a cell (e.g., gamma / epsilon, delta / epsilon, and zeta / zeta CD3 dimers). The term "cell surface expressed CD3" also includes CD3 chains expressed by themselves on the surface of a cell, without other CD3 chain types (e.g., CD3-epsilon, CD3-delta, or CD3-gamma). Alternatively, "cell surface expressed CD3" can include or consist of CD3 proteins expressed on the surface of a cell that normally expresses CD3 protein. Alternatively, "cell surface expressed CD3" can include or consist of CD3 protein expressed on the surface of a cell that does not normally express human CD3 on its surface but has been artificially engineered to express CD3 on its surface.

[0044] The term "antigen binding domain" refers to that portion of a multispecific molecule or corresponding antibody that specifically binds to a given antigen (e.g., CD3 or a tumor-associated antigen). Reference to a "corresponding antibody" refers to the antibody from which the CDRs or variable regions (HCVRs and LCVRs) used in the multispecific molecule are derived. For example, the molecule of the Figure 1C structure discussed in the examples contains a Fab and scFv with variable regions derived from specific anti-CD3 and anti-MAGEA4 antibodies. These antibodies are the "corresponding antibodies" to the respective multispecific molecules.

[0045] The term "multispecific antigen-binding molecule" includes molecules that bind to two or more (e.g., three or four) different epitopes or antigens. In some cases, the multispecific antigen-binding molecule is bispecific. In some cases, the multispecific antigen-binding molecule is trispecific. In some cases, the multispecific antigen-binding molecule is tetraspecific.

[0046] The term "antibody" refers to any antigen-binding molecule or molecular complex that contains at least one complementarity determining region (CDR) that specifically binds to or interacts with a particular antigen (e.g., CD3 or target antigen (TA)). The term "antibody" includes immunoglobulin molecules that contain four polypeptide chains, two heavy (H) chains and two light (L) chains, interconnected by disulfide bonds, as well as multimers thereof (e.g., IgM). The term "antibody" also includes immunoglobulin molecules that consist of four polypeptide chains, two heavy (H) chains and two light (L) chains, interconnected by disulfide bonds. Each heavy chain contains a heavy chain variable region (referred to herein as HCVR or VVR). H The heavy chain constant region is made up of three domains: H 1. C H 2, and C H Each light chain comprises a light chain variable region (herein referred to as LCVR or V L The light chain constant region comprises one domain (C L 1) is included. H Area and V LThe regions can be further subdivided into regions of hypervariability, called complementarity determining regions (CDRs), interspersed with regions that are more conserved, called framework regions (FRs). H and V L Each of the CDRs consists of three CDRs and four FRs arranged from amino terminus to carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. In different embodiments of the present invention, the FRs of the anti-TA antibody or anti-CD3 antibody (or antigen-binding portion thereof) may be identical to the human germline sequence or may be naturally or artificially modified. An amino acid consensus sequence may be defined based on the parallel analysis of two or more CDRs.

[0047] The term "antibody" as used herein also includes antigen-binding fragments of complete antibody molecules. An "antigen-binding portion" of an antibody, an "antigen-binding fragment" of an antibody, and similar terms as used herein include naturally occurring, enzymatically obtainable, synthetic, or genetically engineered polypeptides or glycoproteins that specifically bind an antigen to form a complex. Antibody-binding fragments of antibodies can be derived from complete antibody molecules using any suitable standard techniques, such as, for example, proteolytic or recombinant genetic engineering techniques involving the manipulation and expression of DNA encoding antibody variable domains and, optionally, constant domains. Such DNA is known and / or readily available, for example, from commercial sources, DNA libraries (including, for example, phage antibody libraries), or can be synthesized. DNA can be sequenced and manipulated, for example, chemically or by using molecular biology techniques, to place one or more variable and / or constant domains in a suitable configuration, or to introduce codons, create cysteine ​​residues, modify, add, or delete amino acids, etc.

[0048] Non-limiting examples of antibody binding fragments include (i) Fab fragments, (ii) F(ab')2 fragments, (iii) Fd fragments, (iv) Fv fragments, (v) single chain Fv (scFv) molecules, (vi) dAb fragments, and (vii) minimal recognition units consisting of amino acid residues mimicking the hypervariable regions (e.g., isolated complementarity determining regions (CDRs) such as CDR3 peptides) of an antibody, or constrained FR3-CDR3-FR4 peptides. Domain-specific antibodies, single domain antibodies, domain deleted antibodies, chimeric antibodies, CDR grafted antibodies, diabodies, triabodies, tetrabodies, minibodies, nanobodies (e.g., monovalent nanobodies, bivalent nanobodies, etc.), small modular immunopharmaceuticals (SMIPs), and other engineered molecules such as shark variable IgNAR domains are also encompassed by the term "antigen binding fragment" as used herein.

[0049] Antigen-binding fragments of antibodies typically contain at least one variable domain. The variable domain may be of any size or amino acid composition and generally contains at least one CDR adjacent to or in frame with one or more framework sequences. H Domain is V L In the antigen-binding fragment associated with the domain, V H Domains and V L The domains may be positioned relative to each other in any suitable configuration. For example, the variable region is a dimer and the V H -V H , V H -V L , or V L -V L Alternatively, the antigen-binding fragment of the antibody may comprise a monomeric V H Domain or V L It may also include a domain.

[0050] 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 configurations of variable and constant domains that may be found in an antigen-binding fragment of an antibody of the invention include: (i) a V H -C H 1, (ii) V H -C H 2. (iii) V H -C H 3. (iv) V H -C H 1-C H 2. (v) V H -C H 1-C H 2-C H 3. (vi) V H -C H 2-C H 3. (vii) V H -C L , (viii) V L -C H 1, (ix) V L -C H 2. (x)V L -C H 3. (xi) V L -C H 1-C H 2. (xii) V L -C H 1-C H 2-C H 3. (xiii) V L -C H 2-C H 3, and (xiv) V L -C LIn any configuration of the variable and constant domains, including any of the exemplary configurations listed above, the variable and constant domains may be either directly linked to each other or linked by a complete or partial hinge or linker region. The hinge region may consist of at least two (e.g., 5, 10, 15, 20, 40, 60, or more) amino acids that provide a flexible or semi-flexible link between adjacent variable and / or constant domains in a single polypeptide molecule. Furthermore, antigen-binding fragments of the antibodies of the present invention may be linked to each other and / or to one or more monomeric V H Or V L The variable domain configurations and constant domain configurations may include homodimers or heterodimers (or other multimers) of any of the above listed variable domain configurations and constant domain configurations in non-covalent association with the domains (e.g., via disulfide bonds).

[0051] In certain embodiments of the present invention, the antibody is a human antibody. The term "human antibody" is intended to include antibodies with variable and constant regions derived from human germline immunoglobulin sequences. Human antibodies may include amino acid residues that are not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by in vitro random or site-specific mutagenesis or by in vivo somatic mutation), for example in the CDRs, particularly CDR3. However, the term "human antibody" as used herein is not intended to include antibodies in which CDR sequences derived from the germline of another mammalian species (e.g., mouse) are grafted onto human framework sequences.

[0052] The antibodies discussed herein may, in some embodiments, be recombinant human antibodies. The term "recombinant human antibody" is intended to include all human antibodies prepared, expressed, generated, or isolated by recombinant means, e.g., antibodies expressed using a recombinant expression vector transfected into a host cell, antibodies isolated from a recombinant combinatorial human antibody library, antibodies isolated from an animal (e.g., a mouse) that is transgenic for human immunoglobulin genes (see, e.g., Taylor et al. (1992) Nucl. Acids Res. 20:6287-6295), or antibodies prepared, expressed, generated, or isolated by any other means, including splicing of human immunoglobulin gene sequences to other DNA sequences. Such recombinant human antibodies have variable and constant regions derived from human germline immunoglobulin sequences. However, in certain embodiments, such recombinant human antibodies are subjected to in vitro mutagenesis (or in vivo somatic mutagenesis when animals transgenic for human Ig sequences are used) to thereby improve the V and V regions of the recombinant antibodies. H and V L The amino acid sequence of the region is H and V L These are sequences which are derived from and related to the sequences, but which may not naturally occur within the human antibody germline repertoire in vivo.

[0053] The antibody referred to herein may be an isolated antibody. "Isolated antibody" as used herein means an antibody that has been identified, separated and / or recovered from at least one component of its natural environment. For example, an antibody that has been separated or removed from at least one component of an organism, or from a tissue or cell in which it naturally occurs or is naturally produced, is an "isolated antibody". An isolated antibody also includes an antibody in situ within a recombinant cell. An isolated antibody is an antibody that has been subjected to at least one purification or isolation step. An isolated antibody may be substantially free of other cellular materials and / or chemicals.

[0054] The antibodies referred to herein may contain 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 germline sequences from which the antibodies were derived. Such mutations can be readily ascertained by comparing the amino acid sequences disclosed herein to germline sequences available, for example, from public antibody sequence databases.

[0055] The term "epitope" refers to an antigenic determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule known as the paratope. A single antigen may have more than one epitope. Thus, different antibodies may bind to different regions on an antigen and have different biological effects. Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are those generated by adjacent amino acid residues within a polypeptide chain. In certain circumstances, epitopes may include moieties of saccharides, phosphoryl groups, or sulfonyl groups on an antigen.

[0056] A "multimerization domain" or "multimerizing domain" is any macromolecule that has the ability to associate (covalently or non-covalently) with a second macromolecule of the same or similar structure or composition. For example, a multimerization domain may be found in the immunoglobulin C HThe multimerization domain may be a polypeptide comprising three domains. Non-limiting examples of multimerization domains are the Fc portion of an immunoglobulin, such as the Fc domain of an IgG selected from isotypes IgG1, IgG2, IgG3, and IgG4, as well as any allotype within each isotype group. In certain embodiments, the multimerization domain is an Fc fragment or amino acid sequence of 1 to about 200 amino acids in length containing at least one cysteine ​​residue. In other embodiments, the multimerization domain is a cysteine ​​residue or a short cysteine-containing peptide. Other multimerization domains include peptides or polypeptides that comprise or consist of a leucine zipper, a helix loop motif, or a coiled-coil motif. In some embodiments, the multimerization domain is an immunoglobulin Fc domain, and the multispecific antigen-binding molecules of the present invention are formed by the association of two such Fc domains via interchain disulfide bonds, as in conventional antibodies.

[0057] The term "nucleic acid" or "polynucleotide" refers to nucleotides and / or polynucleotides, such as deoxyribonucleic acid (DNA) or ribonucleic acid (RNA), oligonucleotides, fragments generated by polymerase chain reaction (PCR), and fragments generated by either ligation, cleavage, endonuclease action, and exonuclease action. Nucleic acid molecules can be composed of monomers that are naturally occurring nucleotides (such as DNA and RNA), or analogs of naturally occurring nucleotides (e.g., enantiomeric forms of naturally occurring nucleotides), or combinations of both. Modified nucleotides can have changes in the sugar moiety and / or the pyrimidine or purine base moiety. Sugar modifications include, for example, replacement of one or more hydroxyl groups with halogens, alkyl groups, amines, and azide groups, or the sugar can be functionalized as an ether or ester. Additionally, the entire sugar moiety can be replaced with sterically and electronically similar structures, such as azasugars and carbocyclic sugar analogs. Examples of modifications at the base moiety include alkylated purines and pyrimidines, acylated purines or pyrimidines, or other well-known heterocyclic substitutes. Nucleic acid monomers can be linked by phosphodiester bonds or analogs of such linkages. Nucleic acids can be either single-stranded or double-stranded.

[0058] The term "recombinant", as used herein, refers to all molecules that are prepared, expressed, produced, or isolated by recombinant means, such as multispecific (e.g., bispecific) molecules expressed using a recombinant expression vector transfected into a host cell, multispecific (e.g., bispecific) molecules isolated from an animal (e.g., a mouse) that is transgenic for human immunoglobulin genes (see, e.g., Taylor et al. (see, e.g., J. Med. Sci. 20:6287-6295), or any other means including splicing of human immunoglobulin and / or MHC gene sequences to other DNA sequences. Such recombinant multispecific molecules can include antigen-binding domains having variable and constant regions derived from human germline immunoglobulin sequences.

[0059] The term "subject" or "patient" as used herein includes all members of the animal kingdom, including non-human primates and humans. In one embodiment, a patient is a human having a disease or disorder, e.g., an infectious disease or cancer.

[0060] The term "substantial identity" or "substantially identical," when referring to a nucleic acid or a fragment thereof, indicates that when optimally aligned with appropriate nucleotide insertions or deletions with another nucleic acid (or its complementary strand), there is at least about 95%, more preferably at least about 96%, 97%, 98%, or 99% nucleotide sequence identity of the nucleotide bases as measured by any well-known algorithm of sequence identity, such as FASTA, BLAST, or Gap, as discussed below. A nucleic acid molecule having substantial identity to a reference nucleic acid molecule can, in certain cases, encode a polypeptide having the same or substantially similar amino acid sequence as the polypeptide encoded by the reference nucleic acid molecule.

[0061] As applied to polypeptides, the term "substantial similarity" or "substantially similar" means that two peptide sequences share at least 95% sequence identity, and even more preferably at least 98% or 99% sequence identity, when optimally aligned, such as by the programs GAP or BESTFIT, using a predefined gap weight. Preferably, residue positions that are not identical differ by conservative amino acid substitutions. A "conservative amino acid substitution" is one in which an amino acid residue is replaced by another amino acid residue having a side chain (R group) with similar chemical properties (e.g., charge or hydrophobicity). In general, conservative amino acid substitutions do not substantially change the functional properties of a protein. When two or more amino acid sequences differ from each other in conservative substitutions, the percent sequence identity or degree of similarity may be adjusted upwards to correct for the conservative nature of the substitution. Means for making this adjustment are well known to those skilled in the art. See, for example, Pearson (1994) Methods Mol. Biol. 24:307-331, which is incorporated herein by reference. Examples of amino acid groups having side chains with similar chemical properties include: (1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine, (2) aliphatic hydroxyl side chains: serine and threonine, (3) amide-containing side chains: asparagine and glutamine, (4) aromatic side chains: phenylalanine, tyrosine, and tryptophan, (5) basic side chains: lysine, arginine, and histidine, (6) acidic side chains: aspartic acid and glutamic acid, and (7) sulfur-containing side chains: cysteine ​​and methionine. Preferred conservative amino acid substitution groups are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamate-aspartate, and asparagine-glutamine. Alternatively, a conservative substitution is any change that has a positive value in the PAM250 log-likelihood matrix disclosed in Gonnet et al. (1992) Science 256:1443-1445, which is incorporated herein by reference. A "moderately conservative" substitution is any change that has a non-negative value in the PAM250 log-likelihood matrix.

[0062] Sequence similarity to polypeptides, also referred to as sequence identity, is typically measured using sequence analysis software. Protein analysis software matches similar sequences using similarity measures assigned to various substitutions, deletions, and other modifications, including conservative amino acid substitutions. For example, GCG software includes programs such as Gap and Bestfit, which can be used with default parameters to determine sequence homology or sequence identity between closely related polypeptides, such as homologous polypeptides from different species of organisms, or between a wild-type protein and its mutant protein. See, for example, GCG version 6.1. Polypeptide sequences can also be compared using FASTA, a program in GCG version 6.1, with default or recommended parameters. FASTA (e.g., FASTA2 and FASTA3) provides alignment and percent sequence identity of the best overlapping regions between the query sequence and the search sequence (Pearson (2000) (see above)). Another preferred algorithm for comparing the sequences of the present invention to a database containing a large number of sequences from different organisms is the computer program BLAST, particularly BLASTP or TBLASTN, with default parameters. See, e.g., 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.

[0063] The terms "vector" and "expression vector" include, but are not limited to, viral vectors, plasmids, RNA vectors, or linear or circular DNA or RNA molecules that may consist of chromosomal, non-chromosomal, semisynthetic, or synthetic nucleic acids. In some cases, vectors are capable of autonomous replication (episomal vectors) and / or expression of nucleic acids to which they are linked (expression vectors). Many suitable vectors are known to those of skill in the art and are commercially available. Viral vectors include negative-stranded RNA viruses such as retroviruses, adenoviruses, parvoviruses (e.g., adeno-associated viruses), coronaviruses, orthomyxoviruses (e.g., influenza viruses), positive-stranded RNA viruses such as rhabdoviruses (e.g., rabies and vesicular stomatitis viruses), paramyxoviruses (e.g., measles and Sendai), picornaviruses and alphaviruses, as well as double-stranded DNA viruses including adenoviruses, herpesviruses (e.g., herpes simplex virus types 1 and 2, Epstein-Barr virus, cytomegalovirus), and poxviruses (e.g., vaccinia, fowlpox, canarypox). Other viruses include, for example, Norwalk virus, togavirus, flavivirus, reovirus, papovavirus, hepadnavirus, and hepatitis virus. Examples of retroviruses include avian leukosis sarcoma, mammalian C, B, D viruses, HTLV-BLV group, and lentivirus.

[0064] Multispecific antigen binding molecules A multispecific antigen-binding molecule of the invention (e.g., bispecific or trispecific or tetraspecific) comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first antigen-binding domain that specifically binds a T cell antigen, (ii) a first multimerization domain, and (iii) a second antigen-binding domain that specifically binds a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third antigen-binding domain that specifically binds a target antigen, and (ii) a second multimerization domain, wherein the first and second multimerization domains associate with each other (e.g., via interchain disulfide bonds) to form the molecule.

[0065] In some embodiments, a multispecific antigen-binding molecule of the invention (e.g., bispecific or trispecific or tetraspecific) comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first antigen-binding domain that specifically binds a T cell antigen, (ii) a first multimerization domain, and (iii) a second antigen-binding domain that specifically binds a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third antigen-binding domain that specifically binds a target antigen, (ii) a second multimerization domain, and (iii) a fourth antigen-binding domain that specifically binds to the target antigen, wherein the first and second multimerization domains associate with each other (e.g., via interchain disulfide bonds) to form the molecule.

[0066] The antigen-binding domains referred to above and herein can be Fab domains, including a heavy chain variable region (HCVR), and a heavy chain CH1 domain paired with a light chain variable region (LCVR), and a CL domain. The antigen-binding domains referred to above and herein can also be single chain variable fragment (scFv) domains, including a HCVR and a LCVR connected together by a short peptide linker, for example, of about 10 to about 25 amino acids. Particular linkers include (G4S) nA linker is included, where n=1-10, or n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. Optionally, the linker between the HCVR and LCVR of each scFv is (G4S)4. Unless otherwise defined, the antigen binding domains of the multispecific molecules of the invention can be all Fab domains, all scFv domains, or a combination of Fab domains and scFv domains. Optionally, one or more of the antigen binding domains are Fab domains. Optionally, one or more of the antigen binding domains are scFv domains. Optionally, the first antigen binding domain and the third antigen binding domain are Fab domains. Optionally, the second antigen binding domain is an scFv domain. Optionally, the fourth antigen binding domain is an scFv domain. Optionally, the first and third antigen binding domains are Fab domains and the second and fourth antigen binding domains are scFv domains. In some cases, the first, second, and third antigen-binding domains are Fab domains. In some cases, the first, second, third, and fourth antigen-binding domains are Fab domains.

[0067] In various embodiments, the scFv domains are connected to the C-terminus of each multimerization domain via a linker peptide. In some cases, the linker is 1-10 amino acids long. In some embodiments, the linker is 1-20 amino acids long. In this regard, the linker can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acids long. In some embodiments, the linker can be 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 amino acids long. Ranges including the numbers discussed herein are also encompassed within the present disclosure, e.g., a linker 10-30 amino acids long. In some embodiments, the linker is a flexible linker. Suitable linkers can be easily selected and can be of any of a variety of lengths, such as 1 amino acid (e.g., Gly) to 20 amino acids, 2 to 15 amino acids, 3 to 12 amino acids, including 4 to 10 amino acids, 5 to 9 amino acids, 6 to 8 amino acids, or 7 to 8 amino acids, and can be 1, 2, 3, 4, 5, 6, or 7 amino acids. Exemplary flexible linkers include glycine polymers (G), glycine-serine polymers (GS), n , where n is an integer of at least 1 (e.g., 1 to 20), glycine-alanine polymers, alanine-serine polymers, and other flexible linkers known in the art. Particular linkers include (G4S) n A linker is included, where n=1-10, or n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In some cases, the linker between each scFv domain and the C-terminus of the respective multimerization domain is (G4S)3.

[0068] In those embodiments in which one or more antigen-binding domains are scFvs, the scFvs may be stabilized scFvs, in which one or more modifications are made to the HCVR and / or LCVR sequences to provide and maintain the proper conformation of the scFv. In some embodiments, the scFvs contain cysteine ​​mutations at residue 44 of the HCVR and residue 100 of the LCVR (Kabat numbering) to generate disulfide bonds between the variable regions (see Zhao et al., Int. J. Mol. Sci, 12:1-11, 2011, and Weatherill et al., Protein Engineering, Design and Selection, 25(7):321-329, 2012). In some embodiments, the scFv contains mutations at residue 39 of the HCVR and residue 38 of the LCVR (Kabat numbering) to modify the glutamine residue to glutamic acid or lysine residue to inhibit conformational isomerization (see Igawa et al., Protein Engineering, Design and Selection, 23(8):667-677, 2010).

[0069] In various embodiments, the LCVR (and optionally the CL) of any of the antigen-binding domains may be a cognate LCVR corresponding to the HCVR, or the LCVR may be a universal LCVR (and optionally a CL) common to multiple antigen-binding domains. In some embodiments, the light chain of the Fab domain is a common light chain. In some embodiments, the light chain of the Fab domain is a cognate light chain corresponding to the target antigen-binding domain, where the light chain is common to both Fab domains. In some embodiments, the LCVR of the scFv domain is a cognate LCVR. In some embodiments, the light chain of the Fab domain is a common light chain, and the LCVR of the scFv domain is a cognate LCVR.

[0070] In some embodiments, the multispecific antigen-binding molecule of the present invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1C.

[0071] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1E.

[0072] In some embodiments, the multispecific antigen-binding molecule of the present invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a first target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a second target antigen different from the first target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. IF.

[0073] In some embodiments, the multispecific antigen-binding molecule of the present invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a first target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a second target antigen different from the first target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1G.

[0074] In some embodiments, the multispecific antigen-binding molecule of the present invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a T cell antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1H.

[0075] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a T cell antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1I.

[0076] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, and (ii) a second immunoglobulin Fc domain, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1J.

[0077] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, and (ii) a second immunoglobulin Fc domain, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1K.

[0078] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1L.

[0079] In some embodiments, the multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1M.

[0080] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a first target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a second target antigen different from the first target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1N.

[0081] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a first target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a second target antigen different from the first target antigen, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1O.

[0082] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a first scFv that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a second scFv that specifically binds to a T cell (optionally which may bind to the first T cell antigen, the second T cell antigen, or a third T cell antigen), wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1P.

[0083] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, (ii) a second immunoglobulin Fc domain, and (iii) a fourth Fab that specifically binds to a T cell (optionally which may bind to the first T cell antigen, the second T cell antigen, or the third T cell antigen), wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form a molecule. An exemplary structure of such a molecule is shown in FIG. 1Q.

[0084] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a second Fab that specifically binds to a target antigen, and (ii) a second immunoglobulin Fc domain, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1R.

[0085] In some embodiments, a multispecific antigen-binding molecule of the invention comprises: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first Fab that specifically binds to a first T cell antigen, (ii) a first immunoglobulin Fc domain, and (iii) a second Fab that specifically binds to a second T cell antigen; and (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third Fab that specifically binds to a target antigen, and (ii) a second immunoglobulin Fc domain, wherein the first and second immunoglobulin domains associate with each other via disulfide bonds to form the molecule. An exemplary structure of such a molecule is shown in FIG. 1S.

[0086] Unless otherwise defined, and if present, the fourth antigen-binding domain can specifically bind to a target antigen or a T cell antigen. In some cases, the third antigen-binding domain and the fourth antigen-binding domain specifically bind to different target antigens (the same protein, or different epitopes on different proteins). In some cases, the different target antigens are expressed on the surface of the same target cell (e.g., tumor cell). In some cases, the third antigen-binding domain and the fourth antigen-binding domain specifically bind to the same target antigen (the same epitope on the same protein). In various embodiments, the first and second antigen-binding domains, and the fourth antigen-binding domain (if present and directed to a T cell antigen) can bind to the same or different T cell antigens, as shown in the figure. In some cases, the first, second, and fourth antigen-binding domains specifically bind to different T cell antigens (the same protein, or different epitopes on different proteins). In some cases, the first, second, and fourth antigen-binding domains specifically bind to the same T cell antigen (the same epitope on the same protein). In some cases, the different T cell antigens are a costimulatory molecule (e.g., CD28) and a checkpoint inhibitor (e.g., PD-1) on the surface of the T cell. In such embodiments, the multispecific molecules of the invention can provide a costimulatory signal to the T cell as well as prevent checkpoint inhibition. As used herein, reference to the "same" target antigen or T cell antigen does not necessarily mean that the antigen-binding domains bind to the same surface molecule, but rather that the antigen-binding domains have the same specificity (e.g., they each bind CD3 or TA). Similarly, reference to a "different" target antigen or T cell antigen means that it is different from another target antigen (e.g., MAGEA4 vs. EGFR) or another T cell antigen (e.g., CD28 vs. PD-1), or is a different epitope on the same protein.

[0087] In any of the embodiments discussed above or herein, the target antigen may be a tumor-associated antigen or an infectious disease-associated antigen (e.g., a viral antigen, a bacterial antigen, a fungal antigen, or an antigen expressed by a parasite). In some cases, the target antigen is a tumor-associated antigen. In some cases, the target antigen is an infectious disease-associated antigen. In some cases, the target antigen is a viral antigen. In some cases, the target antigen is a bacterial antigen. In some cases, the target antigen is a fungal antigen. In some cases, the target antigen is an antigen expressed by a parasite.

[0088] In some embodiments, the target antigen is a peptide in the context of a groove (PiG) of a major histocompatibility complex (MHC) protein. In some embodiments, the PiG is a peptide of about 5 to about 40 amino acid residues, about 6 to about 30 amino acid residues, about 8 to about 20 amino acid residues, or about 9, 10, or 11 amino acid residues. In some embodiments, the PiG is a fragment of a tumor-associated antigen, a viral antigen, a bacterial antigen, a fungal antigen, or a parasitic antigen. In various embodiments, the target antigen is a peptide in the context of a groove of any class, subtype, or allele of a human leukocyte antigen, including any of HLA-A, HLA-B, HLA-C, HLA-DR, HLA-DQ, or HLA-DP. In some embodiments, the target antigen is a peptide / MHC complex. In some embodiments, the peptide in the peptide / MHC complex is a fragment of a tumor-associated antigen, a fragment of a bacterial antigen, a fragment of a viral antigen, a fragment of a fungal antigen, or a fragment of a parasitic antigen.

[0089] In some cases, the antigen is a tumor-associated antigen or an antigen expressed by a tumor cell. In some embodiments, the tumor associated antigen is AFP, ALK, BAGE protein, BIRC5 (survivin), BIRC7, β-catenin, brc-abl, BRCA1, BORIS, CA9, carbonic anhydrase IX, caspase-8, CALR, CCR5, CD19, CD20 (MS4A1), CD22, CD40, CD70, CDK4, CEA, cyclin-B1, CYP1B1, EGFR, EGFRvIII, ErbB2 / Her2, ErbB3, ErbB4, ETV6-AML, EpCAM, EphA2, Fra-1, FOLR1, GAGE ​​protein (e.g., GAGE-1, -2), GD2, GD3, GloboH, glypican-3, GM3, gp100, Her2, HLA / B-raf, HLA / k-ras, HLA / MAGE-A3, hTERT , IL-10, LMP2, MAGE proteins (e.g., MAGE-1, -2, -3, -4, -6, and -12), MART-1, mesothelin, ML-IAP, Muc1, Muc2, Muc3, Muc4, Muc5, Muc16 (CA-125), MUM1, NA17, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, p15, p53, PAP, PAX3, P Selected from the group consisting of AX5, PCTA-1, PLAC1, PRLR, PRAME, PSMA (FOLH1), RAGE protein, Ras, RGS5, Rho, SART-1, SART-3, STEAP1, STEAP2, TAG-72, TGF-β, TMPRSS2, Thompson-Nouvelle antigen (Tn), TRP-1, TRP-2, tyrosinase, and uroplakin-3.

[0090] In some cases, the antigen is a viral or bacterial antigen. In some embodiments, the viral antigen is associated with or expressed by a virus selected from the group consisting of adenovirus, astrovirus, chikungunya, cytomegalovirus, dengue, ebola, EBV, hantavirus, HBsAg, hepatitis A, hepatitis B, hepatitis C, hepatitis D, hepatitis E, herpes, HIV, HPIV, HTLV, influenza, Japanese encephalitis virus, Lassa, measles, metapneumovirus, mumps, norovirus, oropouche, HPV, parvovirus, rotavirus, RSV, rubella, SARS, TBEV, Ustu, vaccinia, chickenpox, West Nile, yellow fever, and Zika, or the bacterial antigen is associated with or expressed by a virus selected from the group consisting of Methicillin-resistant Staphylococcus Aureus (MRSA), Clostridium Difficile, carbapenem-resistant Enterobacteriaceae, drug-resistant Neisseria Gonorrhoeae, multidrug-resistant Acinetobacter, drug-resistant Campylobacter, fluconazole-resistant Candida, extended-spectrum beta-lactamase-producing bacteria, vancomycin-resistant Enterococcus, multidrug-resistant Pseudomonas Aeruginosa, drug-resistant nontyphoidal Salmonella, drug-resistant Salmonella serotype typhi, drug-resistant Shigella, drug-resistant Streptococcus pneumoniae, drug-resistant tuberculosis, vancomycin-resistant Staphylococcus Aureus, erythromycin-resistant group A Streptococcus, and clindamycin-resistant group B Streptococcus.

[0091] In any of the embodiments discussed above or herein, the T cell antigen may be an antigen expressed on the surface of a T cell, a T cell receptor complex antigen, a costimulatory molecule on a T cell or a checkpoint inhibitor, CD3, CD27, CD28, 4-1BB, or PD-1. In some cases, the T cell antigen is a T cell receptor complex antigen. In some cases, the T cell antigen is CD3. In some cases, the T cell antigen is a costimulatory molecule on a T cell or a checkpoint inhibitor. In some cases, the T cell antigen is selected from the group consisting of CD27, CD28, 4-1BB, and PD-1. In some cases, the T cell antigen is selected from the group consisting of CD3, CD27, CD28, 4-1BB, and PD-1. In some cases, the T cell antigen is selected from the group consisting of CD28, ICOS, HVEM, CD27, 4-1BB, OX40, DR3, GITR, CD30, SLAM, CD2, 2B4, CD226, TIM1, and TIM2.

[0092] In certain embodiments where the T cell antigen is CD3, the CD3 binding domain binds to human CD3 and induces human T cell activation. In certain embodiments, the CD3 binding domain binds weakly to human CD3 and induces human T cell activation. In some embodiments, the CD3 binding domain binds weakly to human CD3 and induces tumor-associated antigen-expressing cell killing. In some embodiments, the CD3 binding domain binds or associates weakly with human and cynomolgus (monkey) CD3, but the binding interaction is not detectable by in vitro assays known in the art. In some embodiments, the CD3 binding domain binds to human CD3 with weak affinity. In some embodiments, the CD3 binding domain binds to human CD3 with moderate affinity. In some embodiments, the CD3 binding domain binds to human CD3 with high affinity. In some embodiments, the CD3 binding domain binds to human CD3 (e.g., at 25° C.) with a K of less than about 15 nM as measured by surface plasmon resonance (e.g., mAb capture or antigen capture format) or a substantially similar assay. DIn some embodiments, the CD3 binding domain binds to human CD3 with a K of greater than about 15 nM, greater than about 20 nM, greater than about 30 nM, greater than about 40 nM, greater than about 50 nM, greater than about 60 nM, greater than about 100 nM, greater than about 200 nM, or greater than about 300 nM, as measured by a surface plasmon resonance binding assay (e.g., mAb capture or antigen capture format) or a substantially similar assay. D In some embodiments, the antibodies or antigen-binding fragments of the invention bind to CD3 with a K value of less than about 5 nM, less than about 2 nM, less than about 1 nM, less than about 800 pM, less than about 600 pM, less than about 500 pM, less than about 400 pM, less than about 300 pM, less than about 200 pM, less than about 180 pM, less than about 160 pM, less than about 140 pM, less than about 120 pM, less than about 100 pM, less than about 80 pM, less than about 60 pM, less than about 40 pM, less than about 20 pM, or less than about 10 pM, as measured by surface plasmon resonance, e.g., using the assay format defined in Example 3 herein, or a substantially similar assay. D Combine with.

[0093] In some embodiments, the CD3 binding domain has an EC of less than about 50 nM, less than about 40 nM, less than about 30 nM, less than about 20 nM, less than about 10 nM, less than about 5 nM, less than about 4 nM, less than about 3 nM, less than about 2 nM, less than about 1 nM, less than 900 pM, less than 800 pM, less than 700 pM, less than 600 pM, or less than 500 pM as measured in an in vitro flow cytometry binding assay. 50 In some embodiments, the CD3 binding domain has an EC value of about 1 nM, 2 nM, 3 nM, 4 nM, 5 nM, 10 nM, 25 nM, 50 nM, 100 nM, 500 nM, or 1 μM or greater as measured by an in vitro flow cytometry binding assay. 50 Indicates the value.

[0094] In any of the embodiments, the CD3 binding domain can comprise any of the HCVR / LCVR or CDR (e.g., the six CDRs contained within a pair of HCVR / LCVR sequences) amino acid sequences of the anti-CD3 antibodies disclosed in WO2014 / 047231 (9250-WO) or WO2017 / 053856 (10151WO01), including the antibodies identified as 7195P, 7221G, 7221G5, and 7221G20. In various embodiments, anti-CD3 antibodies identified as "strong binders" have affinities for human CD3 in the single digit nanomolar range (e.g., 1-9 nM) as measured by surface plasmon resonance assay (e.g., at 25° C. in an antigen capture format with measurements performed on a T200 BIACORE instrument). In various embodiments, an anti-CD3 antibody identified as a "moderate binder" has an affinity for human CD3 in the double-digit nanomolar range (e.g., 10-99 nM, optionally 10-50 nM, or 10-25 nM) as measured by a surface plasmon resonance assay. In various embodiments, an anti-CD3 antibody identified as a "weak binder" has an affinity for human CD3 in the triple-digit nanomolar range (e.g., 100-999 nM, optionally 100-500 nM, or 500 nM to 1 μM) as measured by a surface plasmon resonance assay. In various embodiments, an anti-CD3 antibody identified as a "very weak binder" has an affinity for human CD3 that is greater than 10 μM or undetectable as measured by a surface plasmon resonance assay.

[0095] In any of the embodiments, the CD3 binding domain can comprise any of the HCVR / LCVR or CDR (e.g., the six CDRs contained within a pair of HCVR / LCVR sequences) amino acid sequences set forth in the table below (the "G" version is taken from WO 2017 / 053856). In some embodiments, the CD3 binding domain (e.g., in the Fab arm of a molecule having the structure of Figure 1C or 1F) comprises a cognate light chain corresponding to the target antigen binding domain. In other words, the cognate light chain of the target antigen binding domain is common to both the target antigen binding domain and the CD3 binding domain (e.g., in the N-terminal Fab domain of the structure of Figure 1C or 1F).

[0096] [Table 1]

[0097] [Table 2]

[0098] [Table 3]

[0099] [Table 4]

[0100] Each of the antibodies set forth in Table 1 comprises a common light chain variable region comprising the amino acid sequence set forth in Table 3. Each of the "G" designated antibodies may also be referred to herein with a "7221" prefix, e.g., 7221G, 7221G5, 7221G20, etc. In the scFv version of the antigen binding domain, the amino acid residue at position 44 of the heavy chain variable region may be replaced with a cysteine ​​residue, for example, as shown in SEQ ID NO: 169 (modified heavy chain corresponding to 7195P) or SEQ ID NO: 170 (modified heavy chain corresponding to 7221G).

[0101] The multispecific antigen-binding molecule of the present invention (e.g., bispecific or trispecific or tetraspecific) comprises two polypeptide chains, each of which comprises a multimerization domain that promotes the association of the two polypeptide chains (e.g., via interchain disulfide bonds) to form a single multispecific antigen-binding molecule. In any of the embodiments discussed above or herein, the first and second multimerization domains can be immunoglobulin Fc domains (e.g., human IgG isotypes). In some cases, the first and second multimerization domains associate with each other via disulfide bonds. In some embodiments, the first and second multimerization domains are human IgG1 or human IgG4 Fc domains. In some cases, the first and second multimerization domains comprise the hinge domain, CH2 domain, and CH3 domain of human IgG1 or human IgG4.

[0102] In some embodiments, the first multimerization domain or the second multimerization domain comprises an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type Fc domain of the same isotype (e.g., human IgG1 or human IgG4). In some cases, the amino acid substitution comprises an H435R modification, or an H435R and a Y436F modification (EU numbering). In some cases, the first multimerization domain comprises an H435R and a Y436F modification. In some cases, the second multimerization domain comprises an H435R and a Y436F modification.

[0103] In some embodiments, the first polypeptide, the second polypeptide, or both the first and second polypeptides have a sequence similar to that of a wild-type hinge domain of the same isotype (e.g., human IgG1 or human IgG4). Fcγ It contains a modified hinge domain which reduces the binding affinity for the receptor.

[0104] In various embodiments in which the multimerization domain comprises a heavy chain constant region including a hinge domain, the constant region can be chimeric, combining sequences from multiple immunoglobulin isotypes. For example, a chimeric Fc domain can be a human IgG1, human IgG2, or human IgG4 C H C derived from 2 regions H 2 sequences, and C derived from human IgG1, human IgG2, or human IgG4. H The chimeric Fc domain may comprise some or all of the three sequences. The chimeric Fc domain may also comprise a chimeric hinge region. For example, the chimeric hinge may comprise an "upper hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region combined with a "lower hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region. A specific example of a chimeric Fc domain that may be included in any of the antigen-binding molecules described herein is a chimeric Fc domain that comprises, from the N-terminus to the C-terminus, an "upper hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region, combined with a "lower hinge" sequence derived from a human IgG1 hinge region, a human IgG2 hinge region, or a human IgG4 hinge region. H Another example of a chimeric Fc domain that may be included in any of the antigen-binding molecules described herein comprises, from the N-terminus to the C-terminus, [IgG1 C H 1]-[IgG1 upper hinge]-[IgG2 lower hinge]-[IgG4 CH2]-[IgG1 CH3]. These and other examples of chimeric Fc domains that can be included in any of the antigen-binding molecules of the present invention are described in WO2014 / 121087 (8550-WO). Chimeric Fc domains with these general structural arrangements and variants thereof can have altered Fc receptor binding, which in turn affects Fc effector function.

[0105] In various embodiments in which the multimerization domain comprises a heavy chain constant region that includes a hinge domain, positions 233-236 in the hinge domain are G, G, G, and free; G, G, free, and free; G, free, free, and free; or all free, with positions numbered according to EU numbering. Optionally, the heavy chain constant region comprises, from N-terminus to C-terminus, a hinge domain, a CH2 domain, and a CH3 domain. Optionally, the heavy chain constant region comprises, from N-terminus to C-terminus, a CH1 domain, a hinge domain, a CH2 domain, and a CH3 domain. Optionally, the CH1 region, if present, the remainder of the hinge region, if present, the CH2 region, and the CH3 region are of the same human isotype. Optionally, the CH1 region, if present, the remainder of the hinge region, if present, the CH2 region, and the CH3 region are human IgG1. Optionally, the CH1 region, if present, the remainder of the hinge region, if present, the CH2 region, and the CH3 region are human IgG2. Optionally, the CH1 region, if present, the remainder of the hinge region, if present, the CH2 region, and the CH3 region are human IgG4. Optionally, the constant region has a CH3 domain modified to reduce binding to Protein A. These and other examples of multimerized heavy chain constant regions that can be included in any of the antigen binding molecules of the present invention are described in WO2016 / 161010 (10140WO01).

[0106] In an embodiment of the present invention, the association of one multimerization domain with another multimerization domain promotes the association between the two antigen-binding domains, thereby forming a bispecific antigen-binding molecule. The multimerization domain can be any polymer, protein, polypeptide, peptide, or amino acid that has the ability to associate with a second multimerization domain of the same or similar structure or composition. For example, the multimerization domain can be a multimerization domain of the immunoglobulin C. H A non-limiting example of a multimerizing component is the Fc portion of an immunoglobulin (C H 2-C H3 domains), such as the Fc domain of IgG selected from isotypes IgG1, IgG2, IgG3, and IgG4, as well as any allotypes within each isotype group.

[0107] In some embodiments, the first and second multimerization domains can be of the same IgG isotype, e.g., IgG1 / IgG1, IgG2 / IgG2, IgG4 / IgG4, etc. Alternatively, the first and second multimerization domains can be of different IgG isotypes, e.g., IgG1 / IgG2, IgG1 / IgG4, IgG2 / IgG4, etc.

[0108] In certain embodiments, the multimerization domain is an Fc fragment or an amino acid sequence of 1 to about 200 amino acids in length that contains at least one cysteine ​​residue. In other embodiments, the multimerization domain is a cysteine ​​residue or a short cysteine-containing peptide. Other multimerization domains include peptides or polypeptides that contain or consist of a leucine zipper, a helix loop motif, or a coiled-coil motif.

[0109] The multimerization domain, e.g., the Fc domain (with or without hinge), may contain one or more amino acid changes (e.g., insertions, deletions, or substitutions) compared to a wild-type, naturally occurring version of the Fc domain. For example, the present invention includes bispecific antigen-binding molecules that contain one or more modifications in the Fc domain that result in a modified Fc domain with modified binding interactions (e.g., enhanced or reduced) between Fc and FcRn. In one embodiment, the bispecific antigen-binding molecule comprises a C H 2 or C HThe FcRn-F ... In one embodiment, the modifications include the 428L (e.g., M428L) and 434S (e.g., N434S) modifications, the 428L, 259I (e.g., V259I), and 308F (e.g., V308F) modifications, the 433K (e.g., H433K) and 434 (e.g., 434Y) modifications, the 252, 254, and 256 (e.g., 252Y, 254T, and 256E) modifications, the 250Q and 428L modifications (e.g., T250Q and M428L), the 307 and / or 308 modifications (e.g., 308F or 308P).

[0110] The present invention also relates to a first IgC H 3 domain and second Ig C H The multispecific antigen-binding molecule comprises a first and a second IgC domain. H The three domains differ from each other by at least one amino acid, and the at least one amino acid difference reduces binding of the bispecific antibody to Protein A compared to a bispecific antibody lacking the amino acid difference. In one embodiment, the first Ig C H The 3 domain binds to protein A and the second Ig C H The C3 domain contains a mutation that reduces or abolishes Protein A binding, such as the H95R modification (according to IMGT exon numbering, H435R according to EU numbering). H3 may further include a Y96F modification (according to IMGT, Y436F by EU). See, e.g., U.S. Patent No. 8,586,713. H Further modifications that may be found within 3 include D16E, L18M, N44S, K52N, V57M, and V82I for IgG1 antibodies (D356E, L358M, N384S, K392N, V397M, and V422I by EU), N44S, K52N, and V82I for IgG2 antibodies (N384S, K392N, and V422I by IMGT, EU), and Q15R, N44S, K52N, V57M, R69K, E79Q, and V82I for IgG4 antibodies (Q355R, N384S, K392N, V397M, R409K, E419Q, and V422I by IMGT, EU).

[0111] Preparation of antigen-binding domains and construction of bispecific molecules Antigen-binding domains specific for a particular antigen can be prepared by any antibody production technique known in the art. Once obtained, the different antigen-binding domains specific for two or more different antigens (e.g., CD3 and a target antigen) can be appropriately positioned with respect to each other to produce the structure of the multispecific antigen-binding molecule of the present invention using conventional methods. In certain embodiments, one or more of the individual components (e.g., heavy and light chains or portions thereof) of the multispecific antigen-binding molecule of the present invention are derived from a chimeric antibody, a humanized antibody, or a fully human antibody. Methods for producing such antibodies are well known in the art. For example, one or more of the heavy and / or light chains of the multispecific antigen-binding molecule of the present invention can be prepared using VELOCIMMUNE™ technology. Using VELOCIMMUNE™ technology (or any other human antibody generation technique), a high affinity chimeric antibody against a particular antigen (e.g., CD3 or a target antigen) with a human variable region and a mouse constant region is first isolated. The antibody is characterized and selected for desirable characteristics, including affinity, selectivity, epitope, etc. The mouse constant regions are replaced with the desired human constant regions to generate fully human heavy and / or light chains that can be incorporated into the multispecific antigen-binding molecules of the invention.

[0112] Genetically engineered animals can be used to generate human multispecific antigen-binding molecules. For example, genetically modified mice can be used that cannot rearrange and express endogenous mouse immunoglobulin light chain variable sequences, and the mice express only one or two human light chain variable domains encoded by human immunoglobulin sequences operably linked to mouse kappa constant genes at the endogenous mouse kappa locus. Such genetically modified mice can be used to produce fully human multispecific antigen-binding molecules that include two different heavy chains associated with the same light chain that includes a variable domain derived from one of two different human light chain variable region gene segments. (See, for example, US2011 / 0195454). Fully human refers to an antibody, or an antigen-binding fragment thereof, or an immunoglobulin domain that includes an amino acid sequence that is encoded by DNA derived from a human sequence over the entire length of each polypeptide of the antibody, or an antigen-binding fragment thereof, or an immunoglobulin domain. In some examples, the fully human sequence is derived from an endogenous protein of a human. In other examples, the fully human protein or protein sequence includes a chimeric sequence in which each component sequence is derived from a human sequence. Without being bound by any theory, chimeric proteins or sequences are generally designed to minimize the creation of immunogenic epitopes at the junctions of the component sequences, e.g., as compared to any wild-type human immunoglobulin region or domain.

[0113] In various embodiments, the methods and techniques discussed above are used to generate antibodies against a T cell antigen and a target antigen, and the antigen binding domains (e.g., HCVRs, LCVRs, or CDRs) of these antibodies are used to produce multispecific antigen binding molecules having structures as discussed herein or as shown, for example, in Figures 1C and 1E-1S.

[0114] Binding properties of antigen-binding domains As used herein, the term "binding" in the context of binding of an antibody (e.g., a corresponding antibody), immunoglobulin, antigen-binding domain, or multispecific antigen-binding molecule to a given antigen, such as, for example, a cell surface protein or fragment thereof, typically refers to an interaction or association between at least two entities or molecular structures, e.g., an antigen-binding domain / antigen interaction.

[0115] For example, binding affinities, as determined by surface plasmon resonance (SPR) techniques, for example in a BIAcore 3000 instrument, using an antigen as the ligand and an antibody, Ig, antibody binding domain, or multispecific antigen-binding molecule as the analyte (or antiligand), are typically around 10 -7 M or less, e.g., about 10 -8 M or less, e.g., about 10 -9 K below M D Flow cytometry assays are also routinely used.

[0116] Thus, an antibody (e.g., a corresponding antibody), antigen-binding domain, or multispecific antigen-binding molecule of the invention has a K that is at least 10-fold lower than its affinity for binding to a non-specific antigen (e.g., BSA, casein). D According to the present invention, the antibody binds to a given antigen or cell surface molecule with an affinity corresponding to a K value that is 10-fold lower than that of a non-specific antigen. D Although the affinity of an antibody (e.g., a corresponding antibody), antigen-binding domain, or multispecific antigen-binding molecule corresponding to the value may be considered as undetectable binding, such an antibody may be paired with a second antigen-binding arm for the production of a bispecific antibody of the invention.

[0117] "K D The term "M" refers to the dissociation equilibrium constant of a particular antibody (or antigen binding domain)-antigen interaction, or the dissociation equilibrium constant of an antibody (or antigen binding domain) or antibody binding fragment that binds to an antigen. D There is an inverse relationship between the binding affinity and the K DThe smaller the value, the higher, i.e., the stronger, the affinity. Thus, the terms "higher affinity" or "stronger affinity" refer to a higher ability to form an interaction, i.e., a smaller K D Conversely, the term "lower affinity" or "weaker affinity" refers to a lower ability to form an interaction, i.e., a larger K D In some situations, the higher binding affinity (or K ) of a particular molecule (e.g., an antibody or antigen-binding domain) to an interaction partner molecule (e.g., antigen X) compared to the binding affinity of the molecule (e.g., an antibody or antigen-binding domain) to another interaction partner molecule (e.g., antigen Y). D ) is larger than K D The smaller the K value (lower, or weaker, affinity), the D The binding affinity may be expressed as a binding ratio determined by dividing by the affinity (higher, or stronger, affinity), for example, as a 5-fold or 10-fold greater binding affinity, as the case may be.

[0118] "k d The term "(sec-1 or 1 / s) refers to the dissociation rate constant of a particular antibody (or antigen binding domain)-antigen interaction, or the dissociation rate constant of an antibody or antibody binding domain. Its value is k off Also called value.

[0119] "k a The term "(M-1 x sec-1 or 1 / M) refers to the association rate constant of a particular antibody (or antigen binding domain)-antigen interaction, or the association rate constant of an antibody or antibody binding domain.

[0120] "K A The term "(M-1 or 1 / M)" refers to the association equilibrium constant of a particular antibody (or antigen binding domain)-antigen interaction, or the association equilibrium constant of an antibody or antibody binding domain. The association equilibrium constant is k a k d It is obtained by dividing by.

[0121] "EC50" or "EC50 The term "half-maximal effective concentration" refers to the concentration of an antibody (or antigen-binding domain or multispecific molecule) that induces a response halfway between the baseline and maximum after a particular exposure time. EC 50 essentially represents the concentration of an antibody (or antigen-binding domain or multispecific molecule) at which 50% of its maximal effect is observed. In certain embodiments, the EC 50 The value is equal to the concentration of a multispecific molecule of the invention that confers half-maximal binding to cells expressing CD3 or a target antigen (e.g., a tumor-associated antigen) as determined, for example, by a flow cytometric binding assay. Thus, reduced or weak binding is considered to be an EC 50 or half-maximal effective concentrations.

[0122] In one embodiment, the reduced binding is an increased EC that allows binding to half-maximal amounts of target cells. 50 It can be defined as the molecular concentration. In another embodiment, EC 50 Values ​​represent the concentration of the molecules of the invention that induce half-maximal depletion of target cells by T cell cytotoxic activity. Thus, an increase in cytotoxic activity (e.g., T cell-mediated tumor cell killing) is associated with an increase in EC 50 or a decrease in the half-maximal effective concentration value is observed.

[0123] pH dependent binding The present invention includes antigen-binding domains and bispecific antigen-binding molecules with pH-dependent binding properties. For example, the molecules of the present invention may exhibit reduced binding to T cell antigens or target antigens at acidic pH compared to neutral pH. Alternatively, the molecules of the present invention may exhibit enhanced binding to T cell antigens or target antigens at acidic pH compared to neutral pH. The term "acidic pH" includes pH values ​​below about 6.2, such as about 6.0, 5.95, 5, 9, 5.85, 5.8, 5.75, 5.7, 5.65, 5.6, 5.55, 5.5, 5.45, 5.4, 5.35, 5.3, 5.25, 5.2, 5.15, 5.1, 5.05, 5.0 or less. As used herein, the term "neutral pH" refers to a pH of about 7.0 to about 7.4. The expression "neutral pH" includes pH values ​​of about 7.0, 7.05, 7.1, 7.15, 7.2, 7.25, 7.3, 7.35, and 7.4.

[0124] In some cases, "reduced binding ... at acidic pH compared to neutral pH" refers to the K of a molecule (or antigen-binding domain) that binds to its antigen at neutral pH. D K value for a molecule (or antigen-binding domain) that binds to its antigen at acidic pH D For example, a molecule or antigen-binding domain may have an acidic / neutral K of about 3.0 or greater. D When a ratio is presented, for purposes of the present invention, a molecule or antigen-binding domain may be considered to exhibit "reduced binding to a T cell antigen or target antigen at acidic pH compared to neutral pH." In certain exemplary embodiments, the acidic / neutral K D The ratio can be about 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11.5, 12.0, 12.5, 13.0, 13.5, 14.0, 14.5, 15.0, 20.0, 25.0, 30.0, 40.0, 50.0, 60.0, 70.0, 100.0 or more.

[0125] Multispecific antibodies with pH-dependent binding properties can be obtained, for example, by screening a population of corresponding antibodies for reduced (enhanced) binding to a particular antigen at acidic pH compared to neutral pH. In addition, modification of the antigen-binding domain at the amino acid level can produce molecules with pH-dependent properties. For example, by replacing one or more amino acids in the antigen-binding domain (e.g., in the CDR) with histidine residues, molecules with reduced antigen binding at acidic pH versus neutral pH can be obtained.

[0126] Biological properties of multispecific antigen-binding molecules The present invention may include multispecific antigen-binding molecules and antigen-binding domains thereof that can simultaneously bind to a human T cell antigen (e.g., CD3) and a human target antigen(s) (e.g., a tumor-associated antigen).

[0127] The present invention can include multispecific antigen-binding molecules that bind to a human T cell antigen (e.g., CD3) and induce T cell activation in the presence of a target cell. For example, in some embodiments, the present invention includes multispecific antigen-binding molecules that bind to a human T cell antigen (e.g., CD3) and induce T cell cytotoxic activity in the presence of a cell expressing a target antigen(s) (e.g., a tumor-associated antigen).

[0128] The present invention can include multispecific antigen-binding molecules that bind to human T cell antigens (e.g., CD3) and induce T cell activation without increasing cytokine production, as compared to conventional bispecific anti-CD3 x anti-TA antibodies (e.g., FIG. 1A).

[0129] The present invention can include multispecific antigen-binding molecules that can deplete or reduce cell populations in which cells express a target antigen(s). The multispecific antigen-binding molecules of the present invention can induce T cell-mediated cytotoxicity more potently than molecules with a conventional bispecific antibody format (e.g., Figures 1A and 1B).

[0130] The present invention can include a multispecific antigen-binding molecule that binds to a human T cell antigen (e.g., CD3) and two different target antigens (e.g., a molecule having the structure of Figure 1F) and induces cytotoxic activity and / or T cell activation in the presence of cells expressing the two target antigens.

[0131] Many cancers express various intracellular antigens that are processed intracellularly by proteosomes, and associated peptides are presented on the surface of cells in the context of HLA molecules. Targeting peptides from different proteins can be used to increase the specificity of the multispecific molecules of the present invention. In some cases, cancers characterized by PiG antigens or low density cancer antigens escape traditional cancer therapy because the target copy number in tumors is often low. In addition, solid tumors characterized by PiG or low density cancer antigens may be more resistant to therapy and more difficult to treat because they are not cell surface antigens, but reside in the grooves within the cancer-associated peptides. Thus, the use of multispecific molecules of the present invention that target two different antigens (e.g., low density antigens) can effectively target PiG and / or low density cancer antigens to increase / enhance the efficacy of therapy in cancer, especially cancers characterized by solid tumors.

[0132] In various embodiments, the multispecific antigen binding molecules of the invention are capable of inducing T cell-mediated cytotoxicity in a cell population when the density of the target antigen ranges from about 100 copies per cell to about 1 million or more copies per cell. In some cases, the target antigen is present at a density of about 100, about 200, about 300, about 400, about 500, about 1000, about 2000, about 3000, about 4000, about 5000, about 6000, about 7000, about 8000, about 9000, about 10000, about 15000, about 20000, about 25000, about 30000, about 35000, about 40000, or about 50000. , about 45,000, about 50,000, about 75,000, about 100,000 (i.e., 100K), about 200K, about 300K, about 400K, about 500K, about 600K, about 700K, about 800K, about 900K, about 1 million, about 2 million, about 3 million, about 4 million, about 5 million, or about 10 million copies per cell.

[0133] Without intending to be bound by theory, the inventors hypothesize that the improved cytotoxic potency of the molecular format of the present invention is a function of the presence of two T cell antigen (e.g., CD3) binding domains on a single chain of the molecule. In particular, the shape of the molecular structure of the present invention is hypothesized to selectively induce lytic synapse formation at low concentrations without inducing stimulatory synapse formation, the latter being involved in cytokine production from cytotoxic T lymphocytes.

[0134] Epitope Mapping and Related Techniques The epitope on a T cell antigen (e.g., CD3) and / or a target antigen (e.g., a tumor-associated antigen) to which the antigen-binding molecule of the present invention binds may consist of a single contiguous sequence of three 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 a protein. Alternatively, an epitope may consist of multiple non-contiguous amino acids (or amino acid sequences) of a protein. The molecules of the present invention may, for example, interact with amino acids contained within a single CD3 chain (e.g., CD3-epsilon, CD3-delta, or CD3-gamma) or may interact with amino acids on two or more different CD3 chains. The term "epitope" as used herein refers to an antigenic determinant that interacts with a specific antigen-binding site in the variable region of an antigen-binding domain known as a paratope. A single antigen may have two or more epitopes. Thus, different antigen-binding domains may bind to different regions on an antigen and have different biological effects. Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are generated by adjacent amino acid residues in a polypeptide chain. In certain circumstances, epitopes may include sugar, phosphoryl, or sulfonyl moieties on an antigen.

[0135] Various techniques known to those skilled in the art can be used to determine whether the antigen-binding domain of a molecule "interacts with one or more amino acids" in a polypeptide or protein. Exemplary techniques include routine cross-blocking assays, as described, for example, in Antibodies, Harlow and Lane (Cold Spring Harbor Press, Cold Spring Harb., NY), alanine scanning mutation analysis, peptide blot analysis (Reineke, 2004, Methods Mol Biol 248:443-463), and peptide truncation analysis. In addition, methods such as epitope excision, epitope extraction, and chemical modification of antigens can be employed (Tomer, 2000, Protein Science 9:487-496). Another method that can be used to identify the amino acid in a polypeptide that the antigen-binding domain of a molecule interacts with is hydrogen / deuterium exchange, detected by mass spectrometry. Generally speaking, hydrogen / deuterium exchange involves deuterium-labeling the protein of interest, and then binding the molecule to the deuterium-labeled protein. The protein / molecule complex is then transferred to water to allow hydrogen-deuterium exchange to occur at all residues except those protected by the molecule (which remain deuterium-labeled). After dissociation of the molecule, the target protein is subjected to protease cleavage and mass spectrometry, thereby revealing the deuterium-labeled residues that correspond to the specific amino acids with which the molecule interacts. See, for example, Ehring (1999) Analytical Biochemistry 267(2):252-259; Engen and Smith (2001) Anal.Chem.73:256A-265A. X-ray crystallography of antigen / molecule complexes may also be used for epitope mapping purposes.

[0136] biological equivalent The present invention includes multispecific antigen-binding molecules that are biologically equivalent to any of the exemplary multispecific antigen-binding molecules described herein.Two antigen-binding proteins are considered biologically equivalent if, for example, they are pharmaceutical equivalents or pharmaceutical alternatives whose absorption rate and degree of absorption do not show significant differences when administered at the same molar dose, either in single or multiple doses, under similar experimental conditions.Some antigen-binding proteins will be considered equivalent or pharmaceutical alternatives if they can be considered biologically equivalent because their absorption degree is equivalent but their absorption rate is not equivalent, and such differences in absorption rate are intentional and reflected in the label, are not essential for achieving effective body drug concentration for long-term use, and are considered medically insignificant for the particular drug product tested.

[0137] In one embodiment, two antigen binding proteins are bioequivalent if there are no clinically significant differences in their safety, purity, or efficacy. In one embodiment, two antigen binding proteins are bioequivalent if a patient can be switched one or more times compared to therapy continued without switching between the reference product and the biological product without an expected increase in risk of adverse effects, including clinically significant changes in immunogenicity or diminished efficacy.

[0138] In one embodiment, two antigen binding proteins are biologically equivalent if they both act by a common mechanism or mode of action for a condition or condition of use, to the extent that such mechanism is known.

[0139] Bioequivalence may be demonstrated by in vivo and in vitro methods. Bioequivalence measurements include, for example, (a) in vivo studies in humans or other mammals where the concentration of the antigen binding protein or its metabolites is measured as a function of time in blood, plasma, serum, or other biological fluids, (b) in vitro studies that correlate with and reasonably predict human in vivo bioavailability data, (c) in vivo studies in humans or other mammals where the relevant acute pharmacological effects of the antigen binding protein (or its target) are measured as a function of time, and (d) well-controlled clinical studies that establish the safety, efficacy, or bioavailability or bioequivalence of the antigen binding protein.

[0140] Biologically equivalent variants of the exemplary multispecific antigen-binding molecules described herein can be constructed, for example, by making various substitutions of residues or sequences, or by deleting terminal or internal residues or sequences that are not required for biological activity. For example, cysteine ​​residues that are not essential for biological activity can be deleted or replaced with other amino acids to prevent the formation of unnecessary or incorrect intramolecular disulfide bridges during renaturation. In other contexts, biologically equivalent antigen-binding proteins can include variants of the exemplary multispecific antigen-binding molecules described herein that contain amino acid changes that modify the glycosylation properties of the molecule, for example, mutations that eliminate or remove glycosylation.

[0141] Species selectivity and species cross-reactivity According to certain embodiments of the present invention, antigen-binding molecules are provided that bind to human T cell antigens (e.g., CD3) but do not bind to the same antigens from other species. Antigen-binding molecules are also provided that bind to human target antigens (e.g., tumor antigens) but do not bind to the same target antigens from other species. The present invention also includes antigen-binding molecules that bind to human antigens and corresponding antigens from one or more non-human species.

[0142] According to certain exemplary embodiments of the present invention, antigen-binding molecules are provided that bind to human CD3 and / or human tumor antigens, and may or may not bind to one or more of mouse, rat, guinea pig, hamster, gerbil, pig, cat, dog, rabbit, goat, sheep, cow, horse, camel, cynomolgus monkey, marmoset, rhesus monkey, or chimpanzee CD3 and / or tumor antigens. For example, in certain exemplary embodiments of the present invention, multispecific antigen-binding molecules are provided, comprising a first antigen-binding domain that binds to human CD3 and cynomolgus monkey CD3, and a second antigen-binding domain that specifically binds to a human tumor antigen.

[0143] immune complex The present invention encompasses antigen-binding molecules conjugated to therapeutic moieties ("immunoconjugates"), such as cytotoxins, chemotherapeutic drugs, immunosuppressants, or radioisotopes. Cytotoxic agents include any agent that is detrimental to cells. Examples of cytotoxic and chemotherapeutic agents suitable for forming immunoconjugates are known in the art (see, for example, WO05 / 103081).

[0144] Therapeutic Formulations and Administration The present invention provides pharmaceutical compositions comprising the multispecific antigen-binding molecules of the present invention. The pharmaceutical compositions of the present invention are formulated with suitable carriers, excipients, and other agents that provide improved transfer, delivery, tolerance, etc. Many suitable formulations can be found in a formulary known to every pharmacist: Remington's Pharmaceutical Sciences, Mack Publishing Company, Easton, PA. These formulations include, for example, powders, pastes, ointments, jellies, waxes, oils, lipids, lipids (cationic or anionic) including vesicles (such as LIPOFECTIN™, Life Technologies, Carlsbad, CA), DNA complexes, anhydrous absorption pastes, oil-in-water and water-in-oil emulsions, emulsions carbowax (polyethylene glycols of various molecular weights), semi-solid gels, and semi-solid mixtures including carbowax. See also Powell et al. "Compendium of excipients for parenteral formulations" PDA (1998) J Pharm Sci Technol 52:238-311.

[0145] The dose of the antigen-binding molecule administered to a patient may vary depending on the age and size of the patient, the target disease, the pathology, the route of administration, and the like. The preferred dose is typically calculated according to body weight or body surface area. When the multispecific antigen-binding molecule of the present invention is used for therapeutic purposes in an adult patient, it may be advantageous to administer the multispecific antigen-binding molecule of the present invention intravenously at a single dose of usually about 0.01 to about 20 mg / kg body weight, more preferably about 0.02 to about 7, about 0.03 to about 5, or about 0.05 to about 3 mg / kg body weight. Depending on the severity of the pathology, the frequency and duration of treatment can be adjusted. The effective dosage and schedule for administering the multispecific antigen-binding molecule can be determined empirically, for example, by monitoring the progress of the patient by periodic evaluation and adjusting the dosage accordingly. Furthermore, interspecies scaling of dosage can be performed using methods well known in the art (e.g., Mordenti et al., 1991, Pharmaceut. Res. 8:1351).

[0146] A variety of delivery systems are known and can be used to administer the pharmaceutical compositions of the present invention, such as encapsulation in liposomes, microparticles, microcapsules, recombinant cells capable of expressing mutant viruses, receptor-mediated endocytosis (see, for example, Wu et al., 1987, J. Biol. Chem. 262:4429-4432). Methods of introduction include, but are not limited to, intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, intranasal, epidural, and oral routes. The compositions can be administered by any convenient route, such as by infusion or bolus injection, absorption through epithelial or mucocutaneous linings (e.g., oral mucosa, rectal and intestinal mucosa, etc.), and can be administered together with other biologically active agents. Administration can be systemic or local.

[0147] The pharmaceutical composition of the present invention can be delivered subcutaneously or intravenously using a standard needle and syringe. In addition, for subcutaneous delivery, a pen delivery device facilitates application when delivering the pharmaceutical composition of the present invention. Such a pen delivery device can be reusable or disposable. A reusable pen delivery device generally utilizes a replaceable cartridge containing the pharmaceutical composition. Once the pharmaceutical composition in the cartridge has been administered and the cartridge is empty, the empty cartridge can be easily discarded and easily replaced with a new cartridge containing the pharmaceutical composition. The pen delivery device can then be reused. In a disposable pen delivery device, there is no replaceable cartridge. Rather, the disposable pen delivery device is pre-filled with the pharmaceutical composition held in a reservoir within the device. Once the reservoir is emptied of the pharmaceutical composition, the entire device is discarded.

[0148] Numerous reusable pens and autoinjector delivery devices have application in the subcutaneous delivery of the pharmaceutical compositions of the present invention. Examples include the AUTOPEN™ (Owen Mumford, Inc., Woodstock, UK), the DISETRONIC™ pen (Disetronic Medical Systems, Bergdorf, Switzerland), the HUMALOG MIX 75 / 25™ pen, the HUMALOG™ pen, the HUMALIN 70 / 30™ pen (Eli Lilly and Co., Indianapolis, IN), the NOVOPEN™ I, II and III (Novo Nordisk, Copenhagen, Denmark), the NOVOPEN JUNIOR™ (Novo Nordisk, Copenhagen, Denmark), the BD™ pen (Becton Dickinson, Franklin Lakes, NJ), the OPTIPEN™, the OPTIPEN PRO™, the OPTIPEN™ IV ... Examples of disposable pen delivery devices that have application in subcutaneous delivery of pharmaceutical compositions of the present invention include, but are not limited to, SOLOSTAR pen (Sanofi-Aventis), FLEXPEN (Novo Nordisk), and KWIKPEN (Eli Lilly), SURECLICK autoinjector (Amgen, Thousand Oaks, CA), PENLET (Haselmeier, Stuttgart, Germany), EPIPEN (Dey, LP) and HUMIRA pen (Abbott Labs, Abbott Park IL).

[0149] In certain circumstances, the pharmaceutical composition can be delivered in a controlled release system. In one embodiment, a pump can be used (see Langer, supra; Sefton, 1987, CRC Crit. Ref. Biomed. Eng. 14:201). In another embodiment, a polymeric material can be used. See Medical Applications of Controlled Release, Langer and Wise (eds.), 1974, CRC Pres., Boca Raton, Florida. In yet another embodiment, the controlled release system can be placed in the vicinity of the target of the composition, thereby requiring only a fraction of the systemic dose (see, e.g., Goodson, 1984, Medical Applications of Controlled Release, supra, vol. 2, pp. 115-138). Other controlled release systems are discussed in the review by Langer, 1990, Science 249:1527-1533.

[0150] The injectable preparations may include dosage forms for intravenous injection, subcutaneous injection, intradermal injection, and intramuscular injection, drip infusion, and the like. These injectable preparations may be prepared by publicly known methods. For example, the injectable preparations may be prepared, for example, by dissolving, suspending, or emulsifying the antigen-binding molecule or its salt described above in a sterile aqueous or oily medium conventionally used for injection. The aqueous medium for injection may be, for example, physiological saline, glucose-containing isotonic solution, and other auxiliary agents, which may be used in combination with a suitable solubilizing agent such as alcohol (e.g., ethanol), polyalcohol (e.g., propylene glycol, polyethylene glycol), nonionic surfactant [e.g., polysorbate 80, HCO-50 (polyoxyethylene (50 mol) adduct of hydrogenated castor oil)]. As the oily medium, for example, sesame oil, soybean oil, and the like may be used in combination with a solubilizing agent such as benzyl benzoate, benzyl alcohol, and the like. The injections thus prepared are preferably filled into suitable ampoules.

[0151] Advantageously, the pharmaceutical compositions for oral or parenteral use described above are prepared into a suitable unit dose dosage form to suit the dose of the active ingredient. Such unit dose dosage forms include, for example, tablets, pills, capsules, injections (ampoules), suppositories, etc. The amount of the antigen-binding molecule contained is generally about 5 to about 500 mg per unit dose dosage form, and in particular, in the form of injection, the antigen-binding molecule is preferably contained in an amount of about 5 to about 100 mg, and for other dosage forms, about 10 to about 250 mg.

[0152] Therapeutic Uses of Antigen-Binding Molecules The present invention includes a method comprising administering to a subject in need thereof a therapeutic composition comprising a multispecific antigen-binding molecule that specifically binds to a T cell antigen (e.g., CD3) and a target antigen (e.g., a tumor-associated antigen). The therapeutic composition can comprise any of the multispecific antigen-binding molecules disclosed herein and a pharma- ceutically acceptable carrier or diluent. As used herein, the phrase "subject in need thereof" refers to a human or non-human animal that exhibits one or more symptoms or signs of cancer, or that would otherwise benefit from inhibition or reduction of target antigen activity or depletion of target antigen-positive cells (e.g., tumor cells).

[0153] The multispecific antigen-binding molecules of the present invention (and therapeutic compositions comprising the same) are particularly useful for treating any disease or disorder in which stimulating, activating, and / or targeting an immune response is beneficial. In particular, the multispecific antigen-binding molecules of the present invention can be used to treat, prevent, and / or ameliorate any disease or disorder associated with or mediated by target antigen expression or activity or proliferation of target antigen-positive cells. The mechanism of action by which the therapeutic methods of the present invention are achieved involves the killing of cells expressing the target antigen in the presence of T cells.

[0154] The multispecific antigen binding molecules of the present invention can be used to treat diseases or disorders related to target antigen expression, including, for example, cancer. Analytical / diagnostic methods known in the art, such as tumor scanning, can be used to determine whether a patient has tumor cells that are positive for the target antigen. In some cases, the cancer is selected from solid tumors, cervical cancer, head and neck squamous cell carcinoma, melanoma, prostate cancer, acute myeloid leukemia, pancreatic cancer, colon cancer, acute lymphocytic leukemia, non-Hodgkin's lymphoma, gastric cancer, post-transplant lymphoproliferative disorder, ovarian cancer, lung cancer, squamous cell carcinoma, non-small cell lung cancer, esophageal cancer, bladder cancer, nasopharyngeal cancer, uterine cancer, liver cancer, testicular cancer, or breast cancer.

[0155] The present invention also includes methods for treating residual cancer in a subject. As used herein, the term "residual cancer" refers to the presence or persistence of one or more cancerous cells in a subject after treatment with an anti-cancer therapy.

[0156] According to certain embodiments, the present invention provides methods for treating a disease or disorder associated with target antigen expression (e.g., cancer), comprising administering to a subject one or more of the multispecific antigen binding molecules described elsewhere herein after the subject has been determined to have a target antigen-positive cancer. For example, the present invention includes methods for treating cancer, comprising administering to a patient a multispecific antigen binding molecule 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 1 week, 2 weeks, 3 weeks, or 4 weeks, 2 months, 4 months, 6 months, 8 months, 1 year or more after the subject has received other immunotherapy or chemotherapy.

[0157] Combination Therapies and Formulations The present invention provides a method comprising administering a pharmaceutical composition comprising any of the exemplary multispecific antigen-binding molecules described herein in combination with one or more additional therapeutic agents. Exemplary additional therapeutic agents that may be combined or administered in combination with the antigen-binding molecules of the present invention include, for example, anti-tumor agents (e.g., chemotherapeutic agents). In certain embodiments, the second therapeutic agent may be a monoclonal antibody, an antibody-drug conjugate, a bispecific antibody conjugated to an anti-tumor agent, a checkpoint inhibitor, or a combination thereof. Other agents that may be beneficially administered in combination with the antigen-binding molecules of the present invention include 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, IL-9, IL-10, IL-11, IL-12, IL-13, IL-17, IL-18, or their respective receptors. The pharmaceutical compositions of the present invention (e.g., pharmaceutical compositions comprising the multispecific antigen-binding molecules disclosed herein) may also be administered as part of a therapeutic regimen comprising one or more therapeutic combinations selected from monoclonal antibodies capable of interacting with different antigens on the cell surface, bispecific antibodies with one arm that binds to an antigen on a tumor cell surface and the other arm that binds to an antigen on a T cell, antibody-drug conjugates, bispecific antibodies conjugated to an anti-tumor agent, checkpoint inhibitors, such as those targeting PD-1 or CTLA-4, or combinations thereof. In certain embodiments, the checkpoint inhibitor may be selected from a PD-1 inhibitor, such as pembrolizumab (Keytruda), nivolumab (Opdivo), or cemiplimab (REGN2810). In certain embodiments, the checkpoint inhibitor may be selected from a PD-L1 inhibitor, such as atezolizumab (Tecentriq), avelumab (Bavencio), or durvalumab (Imfinzi). In certain embodiments, the checkpoint inhibitor may be selected from a CTLA-4 inhibitor, such as ipilimumab (Yervoy). Other combinations that can be used in conjunction with the antibodies of the invention are described above.

[0158] The invention also includes therapeutic combinations comprising any of the antigen binding molecules described herein and one or more inhibitors of VEGF, Ang2, DLL4, EGFR, ErbB2, ErbB3, ErbB4, EGFRvIII, cMet, IGF1R, IL-10, B-raf, PDGFR-α, PDGFR-β, FOLH1 (PSMA), PRLR, STEAP1, STEAP2, TMPRSS2, MSLN, CA9, uroplakin, or any of the aforementioned cytokines, wherein the inhibitor is an aptamer, antisense molecule, ribozyme, siRNA, peptibody, nanobody, antibody, bispecific antibody, or antibody fragment (e.g., Fab fragment, F(ab')2 fragment, Fd fragment, Fv fragment, scFv, dAb fragment, or other engineered molecules such as diabodies, triabodies, tetrabodies, minibodies, and minimal recognition units). The antigen-binding molecules of the present invention may also be administered in combination with and / or co-formulated with antivirals, antibiotics, analgesics, corticosteroids and / or NSAIDs. The antigen-binding molecules of the present invention may also be administered as part of a treatment regimen that also includes radiation therapy and / or conventional chemotherapy.

[0159] The additional therapeutically active ingredient may be administered immediately prior to, simultaneously with, or immediately following administration of the antigen-binding molecule of the present invention. (For purposes of this disclosure, such administration regimes will be considered as administering the antigen-binding molecule "in combination" with the additional therapeutically active ingredient.

[0160] The present invention includes pharmaceutical compositions in which an antigen-binding molecule of the invention is co-formulated with one or more additional therapeutically active ingredients as described elsewhere herein. Dosing regimen According to certain embodiments of the invention, multiple doses of the multispecific antigen-binding molecule may be administered to a subject over a defined time course. The method according to this aspect of the invention comprises sequentially administering multiple doses of the antigen-binding molecule of the invention to the subject. As used herein, "sequentially administering" means that each dose of the antigen-binding molecule is administered to the subject at different time points, e.g., on different days separated by a predetermined interval (e.g., hours, days, weeks or months). The invention includes a method comprising sequentially administering to a patient a single initial dose of the antigen-binding molecule, followed by one or more secondary doses of the antigen-binding molecule, and then optionally one or more tertiary doses of the antigen-binding molecule.

[0161] The terms "primary dose", "secondary dose", and "tertiary dose" refer to the time sequence of administration of the antigen-binding molecule of the present invention. Thus, a "primary dose" is a dose administered at the beginning of a treatment regimen (also called a "baseline dose"); a "secondary dose" is a dose administered after the primary dose; and a "tertiary dose" is a dose administered after the secondary dose. The primary, secondary, and tertiary doses may all contain the same amount of antigen-binding molecule, but generally may differ from each other in terms of frequency of administration. However, in certain embodiments, the amount of antigen-binding molecule contained in the primary, secondary, and / or tertiary doses differs from each other (e.g., adjusted up or down as appropriate) during the course of treatment. In certain embodiments, two or more doses (e.g., two, three, four, or five) are administered as a "loading dose" at the beginning of a treatment regimen, followed by subsequent doses (e.g., "maintenance doses") administered on a less frequent basis.

[0162] In certain exemplary embodiments of the invention, each secondary dose and / or tertiary dose is 1 to 26 (e.g., 1, 1 and 1 / 2, 2, 2 and 1 / 2, 3, 3 and 1 / 2, 4, 4 and 1 / 2, 5, 5 and 1 / 2, 6, 6 and 1 / 2, 7, 7 and 1 / 2, 8, 8 and 1 / 2, 9, 9 and 1 / 2, 10, 10 and 1 / 2, 11, 11 and 1 / 2, 12, 12 and 1 / 2, 13, The phrase "immediately preceding dose" as used herein refers to a dose of an antigen-binding molecule in a series of multiple doses that is administered to a patient prior to administration of the immediately following dose with no intervening doses.

[0163] The method according to this aspect of the invention may include administering any number of secondary and / or tertiary doses of the antigen-binding molecule to the patient. For example, in certain embodiments, only a single secondary dose is administered to the patient. In other embodiments, two or more (e.g., two, three, four, five, six, seven, eight or more) secondary doses are administered to the patient. Similarly, in certain embodiments, only a single tertiary dose is administered to the patient. In other embodiments, two or more (e.g., two, three, four, five, six, seven, eight or more) tertiary doses are administered to the patient.

[0164] In embodiments involving multiple secondary doses, each secondary dose may be administered at the same frequency as the other secondary doses. For example, each secondary dose may be administered to the patient 1-2 weeks after the immediately preceding dose. Similarly, in embodiments involving multiple tertiary doses, each tertiary dose may be administered at the same frequency as the other tertiary doses. For example, each tertiary dose may be administered to the patient 2-4 weeks after the immediately preceding dose. Alternatively, the frequency with which the secondary and / or tertiary doses are administered to the patient may vary over the course of the treatment regimen. The frequency of administration may also be adjusted by the physician during the course of treatment depending on the needs of the individual patient after clinical testing. EXAMPLES

[0165] The following examples are presented to provide those skilled in the art with a complete disclosure and description of how to make and use the methods and compositions of the present invention, and are not intended to limit the scope of what the inventors regard as the invention. Efforts have been made to ensure accuracy with respect to the numbers used (e.g., amounts, temperatures, etc.), but some experimental error and deviation should be accounted for. Unless otherwise indicated, parts are parts by weight, molecular weight is average molecular weight, temperature is in degrees Celsius, and pressure is at or near atmospheric pressure.

[0166] Methods for binding by flow cytometry: In the examples below, the following flow cytometry methods were used to determine binding for various molecules. Flow cytometry analysis was used to measure binding of MAGEA4xCD3 multispecific molecules to RAJI / HLA-A2 / B2M / MAGEA4 (peptide a), A375 / hHLA-A2 / B2M / MAGEA4 (peptide b), RAJI / HLA-A2 / B2M / NY-ESO-1, and JURKAT cells, followed by detection with an APC-labeled anti-human IgG antibody. Briefly, 1x10 5 Cells / wells were incubated with serial dilutions of MAGEA4xCD3 multispecific molecules or an isotype control (human IgG4 stealth antibody that binds human antigens without cross-reactivity with human MAGEA4 or CD3) for 30 min at 4°C. After incubation, cells were washed twice with cold PBS containing 1% filtered FBS and PE-conjugated anti-human secondary antibody was added to the cells and incubated for an additional 30 min. Wells containing no antibody or only secondary antibody were used as controls. After incubation, cells were washed, resuspended in 200 μL cold PBS containing 1% filtered FBS and analyzed by flow cytometry on a BD FACS Canto II.

[0167] Methods for Cytotoxicity Assays: In the following examples, the following cytotoxicity assays were used to determine the cytotoxicity of various molecules. To monitor the killing of MAGEA4+ cells in the presence of MAGEA4xCD3 as a single agent or in combination with EGFRxCD28 bispecific antibody and / or PD-1 antibody, A375 cells, ScaBER cells, NCI-H1755 metastatic (liver-derived) cells, and NCI-H1755 cells were labeled with 1 uM of the fluorescent tracking dye Violet Cell Tracker. After labeling, cells were seeded overnight at 37°C. Separately, human PBMCs were plated at 1 x 10 6 Cells were seeded at 1000 ng / mL in supplemented RPMI medium and incubated overnight at 37°C to enrich for lymphocytes by depleting adherent macrophages, dendritic cells, and some monocytes. The next day, target cells were co-incubated with adherent cell-depleted naive PBMCs (effector / target cell 10:1 ratio), serial dilutions of MAGEA4xCD3 multispecific molecules, and fixed concentrations of EGFRxCD28 and / or anti-PD1 antibodies for 96 h at 37°C. Cells were removed from cell culture plates using trypsin-EDTA dissociation buffer and analyzed by FACS on a FACS BD LSRFortessa-X20. For FACS analysis, cells were stained with dead / live Near IR Reactive (Invitrogen) dye. Immediately before FACS analysis, 5E05 counting beads were added to each well. 1E05 beads were collected for each sample. To assess the specificity of killing, cells were gated on the live violet-labeled population. The percentage of the live population was recorded and used to calculate survival rates.

[0168] Example 1: T cell activation is dependent on the presence of target cells T cell activation was assessed for each of the molecular formats shown in Figures 1A, 1B, and 1C. T cell activation and upregulation of PD-1 markers was assessed by incubating cells with antibodies directly conjugated to CD2, CD4, CD8, CD25, and PD-1 and reporting the percentage of late-activated (CD25+ / CD8+) T cells and PD-1+ / CD4+ T cells among total T cells (CD2+).

[0169] As shown in Figure 2, an exemplary multispecific molecule of the invention (Figure 1C structure) did not activate T cells in the absence of target cells. "Zero" represents the T cell only control. Example 2: Cytotoxicity of multispecific molecules compared to conventional formats The cytotoxicity of the exemplary multispecific molecules of the present invention (FIG. 1C structure) was measured as discussed above and compared to that of conventional format molecules with the same antigen binding domain (FIGS. 1A and 1B). The CD3 binding domain used in this example has a moderate binding affinity to human CD3. The target antigen binding domain used in this example binds to MAGEA4 (melanoma associated antigen A4) peptide. The "control" is a positive control that targets the scaffold of all HLA molecules to provide maximum cytotoxicity for comparison with other formats.

[0170] As shown in FIG. 3, an exemplary multispecific molecule of the present invention (FIG. 1C structure) killed target cells more potently than molecules with a conventional bispecific format (FIG. 1A structure and FIG. 1B structure).

[0171] Example 3: Cytotoxicity of multispecific molecules compared to conventional formats of anti-PD-1 antibodies, costimulatory bispecific antibodies, or combinations of both The cytotoxicity of an exemplary multispecific molecule of the invention (FIG. 1C structure) was measured as discussed above and compared to the cytotoxicity of a conventional format molecule (FIGS. 1A and 1B) with the same antigen-binding domains in combination with an anti-PD-1 antibody, a costimulatory bispecific EGFRxCD28 antibody, or both an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody. The positive control, as well as the CD3 and target antigen-binding domains, were as discussed above in Example 2.

[0172] As shown in Figures 4A, 4B, and 4C, the addition of an anti-PD-1 antibody, a costimulatory bispecific EGFRxCD28 antibody, or both further enhanced the potency of an exemplary multispecific molecule of the invention (Figure 1C structure). The solid lines represent the cytotoxicity of the single agents (as shown in Figure 3), and the dashed lines represent the cytotoxicity of each combination.

[0173] In addition to cytotoxicity, assay well supernatants from the human PBMC assay were evaluated for Th1 / Th2 cytokine release using the BD Cytometric Bead Array Human Kit according to the manufacturer's protocol. As shown in Figure 5, the greater cytotoxicity of the exemplary multispecific molecule of the invention (Figure 1C structure) did not result in any greater cytokine release compared to the conventional bispecific antibody format (Figure 1A structure).

[0174] This series of experiments demonstrated that (a) at the highest concentration in the cytotoxicity assay, the molecule having the structure of FIG. 1C exhibited greater potency than the molecule having the structure of FIG. 1A with comparable levels of cytokine release; (b) the cytotoxicity EC50 of the molecule having the structure of FIG. 1C (single agent) was lower than that observed for the molecule having the structure of FIG. 1A (single agent); (c) at the highest concentration in the cytotoxicity assay, the molecule having the structure of FIG. 1A exhibited greater potency in combination with an anti-PD-1 antibody than the molecule having the structure of FIG. 1A (anti-PD-1 combination) with comparable levels of cytokine release; (d) the cytotoxicity EC50 of the molecule having the structure of FIG. 1C combined with an anti-PD-1 antibody was lower than that observed for the molecule having the structure of FIG. 1A (anti-PD-1 combination); and (e) at the highest concentration in the cytotoxicity assay, the molecule having the structure of FIG. 1A exhibited greater potency than the molecule having the structure of FIG. 1A with comparable levels of cytokine release. (f) the cytotoxicity EC50 of the molecule having the structure of FIG. 1C combined with the anti-EGFR×CD28 bispecific antibody was lower than that observed for the molecule having the structure of FIG. 1A (anti-EGFR×CD28 combination); (g) at the highest concentration in the cytotoxicity assay, the molecule having the structure of FIG. 1A combined with the anti-PD-1 antibody and the anti-EGFR×CD28 bispecific antibody showed greater potency than the molecule having the structure of FIG. 1A (triple combination) with comparable levels of cytokine release; and (h) the cytotoxicity EC50 of the molecule having the structure of FIG. 1C combined with the anti-PD-1 antibody and the anti-EGFR×CD28 bispecific antibody was lower than that observed for the molecule having the structure of FIG. 1A (triple combination).

[0175] Example 4: Potency of multispecific molecules is enhanced by two effector binding domains Binding of the exemplary multispecific molecule (FIG. 1C structure) to target cells overexpressing the MAGEA4 peptide and to CD3+ Jurkat cells was measured as discussed above. Binding to these cells was also assessed for modifications of the FIG. 1C structure in which one or more of the antigen binding domains were inactivated. Inactive domains are indicated by an "X" in the figure legend.

[0176] As shown in Examples 6A-6D, the binding data indicates that combinations of two antigen binding domains (e.g., a single Fab and a single scFv) bound to target cells with greater affinity (lower EC50) than molecules with a single Fab domain or a single scFv domain. As expected, isotype control molecules showed no binding. No distinction in binding patterns was observed regardless of the source of the anti-CD3 binding domain.

[0177] In addition to binding, the cytotoxicity of these molecules was also determined using the methods discussed above. As shown in Figures 7A and 7B, the exemplary multispecific molecule of the present invention (Figure 1C structure) showed the greatest cytotoxicity potency, followed by two modified molecules that contain two T cell antigen (CD3) binding domains, but only a single target antigen (MAGEA4) binding domain (scFv or Fab). Similarly, the same cytotoxicity pattern was observed regardless of the source of the anti-CD3 binding domain. The negative control (Figure 1A format) contained an irrelevant target antigen binding domain.

[0178] Example 5: C-terminal scFv domains enhance the potency of multispecific molecules compared to C-terminal Fab domains Binding of the exemplary multispecific molecule (FIG. 1C structure) to target cells overexpressing the MAGEA4 peptide and to CD3+ Jurkat cells was measured as discussed above. Binding to these cells was also assessed for modifications of the FIG. 1C structure in which the C-terminal scFv domain is replaced with a Fab domain (FIG. 1E structure), or in which the N-terminal Fab domain is inactivated. Inactive domains are indicated by an "X" in the figure legend.

[0179] Similar to the binding discussed in Example 4, the binding data shown in Figures 8A and 8B indicate that combinations of two antigen binding domains (e.g., a single Fab and a single scFv, or two Fabs) bound to target cells with greater affinity (lower EC50) than molecules with a single Fab domain or molecules with a single scFv domain. As shown in the tables of Figures 8A and 8B, molecules with the structures of Figures 1C and 1E bind with comparable binding titers.

[0180] In addition to binding, the cytotoxicity of these molecules was also determined using the methods discussed above. As shown in Figure 9, an exemplary multispecific molecule of the invention (Figure 1C structure) exhibited the greatest cytotoxic potency, followed by a modified molecule containing a C-terminal Fab domain instead of the two scFv domains.

[0181] Example 6: Single-chain bivalency for T cell antigens enhances the potency of multispecific molecules compared to multi-chain bivalency Binding of the exemplary multispecific molecule (FIG. 1C structure) to target cells overexpressing MAGEA4 peptides and to CD3+ Jurkat cells was measured as discussed above. Binding to these cells was also assessed for a molecule having the structure shown in FIG. 1D, in which the MAGEA4-binding domain and the CD3-binding domain are swapped such that the two sets of antigen-binding domains are located on two separate polypeptide chains.

[0182] As shown in Figures 10A and 10B, the binding data demonstrated similar binding of the two molecular structures to each of the two cell types. In addition to binding, the cytotoxicity of these molecules was also determined using the methods discussed above. As shown in Figures 11A and 11B, an exemplary multispecific molecule of the invention (Figure 1C structure) exhibited greater cytotoxic potency compared to the molecule having the structure of Figure 1D, confirming that the presence of two T cell antigen binding domains on a single polypeptide chain provides enhanced cytotoxic potency.

[0183] Example 7: Relative cytotoxicity of multispecific molecules targeting one or two antigens compared to traditional formats alone or in combination with anti-PD-1 antibodies and costimulatory bispecific antibodies The cytotoxicity of two exemplary multispecific molecules of the invention (structures in Figures 1C and 1F) was measured as discussed above and compared to that of the conventional format molecules (Figure 1A) alone or in combination with an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody. This example uses a positive control with greater specificity than that used in the previous example to show greater discrimination between the molecules with the structures in Figures 1C and 1F and their combination with a costimulatory bispecific antibody and an anti-PD-1 antibody. The CD3 antigen binding domain used in this example has a strong binding affinity for human CD3 and the target antigen binding domain (MAGEA4a) was as discussed in Example 2 above. The negative control (Figure 1A format) included an irrelevant target antigen binding domain. The second target antigen binding domain (MAGEA4b) used in this example for the molecule with the structure in Figure 1F binds to an epitope on MAGEA4 that is completely different from the epitope bound by the first target antigen binding domain.

[0184] As shown in Figures 12A and 12B, multispecific molecules targeting two different low density antigens on tumor cells show increased potency compared to multispecific molecules targeting only a single tumor antigen, and both molecules show greater potency than the conventional format molecule having the structure of Figure 1 A. The addition of an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody further enhanced the potency of an exemplary multispecific molecule of the invention (Figure 1C and 1F structures).

[0185] Example 8: Relative cytotoxicity of multispecific molecules correlates with affinity of the T cell antigen-binding domain Exemplary multispecific molecules having the structure of FIG. 1F (as shown in FIG. 13) were prepared with anti-CD3 binding domains of varying affinity. Five molecules were prepared according to the following parameters: Molecule A with CD3 arms 7195P(strong)fab and 7195P(strong)scfv; Molecule B with CD3 arms 7221G(medium)fab and 7221G(medium)scfv; Molecule C with CD3 arms 7221G20(weak)fab and 7221G20(weak)scfv; Molecule D with CD3 arms 7221G20 (weak) fab and 7221G (medium) scfv, and Molecule E with CD3 arms 7221G (medium) fab and 7195P (strong) scfv.

[0186] The range of binding titers to T cells from these five molecules was examined by flow cytometry and correlates with the potency of the CD3 binding domain as shown in FIG. 13 compared to the isotype control.

[0187] Cytotoxicity assays targeting two different MAGEA4+ cell lines (A375 and ScaBER) demonstrated that the potency of the molecules decreased when the strength of the effector arm (e.g., anti-CD3 binding domain) was reduced, either as single agents or in combination with an EGFRxCD28 bispecific antibody and an anti-PD1 antibody, as shown in Figures 14A, 14B, 15A, and 15B. Each of the molecules contained the same target antigen binding domains (for non-overlapping MAGEA4 peptide 1 and MAGEA4 peptide 2).

[0188] Example 9: Relative cytotoxicity of a multispecific molecule targeting two antigens compared to conventional formats alone or in combination with an anti-PD-1 antibody and a costimulatory bispecific antibody The cytotoxicity of three exemplary multispecific molecules of the invention (structures in Figures 1C and 1F) was measured as discussed above and compared to that of the conventional format molecule (Figure 1A) alone or in combination with an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody. This example uses a positive control with the structure in Figure 1A that binds CD3 and HLA. The CD3 antigen binding domain used in this example has a strong binding affinity to human CD3 (derived from 7195P), and the target antigen binding domain is directed against one or two non-overlapping MAGEA4 (melanoma associated antigen A4) peptides (MAGEA4Aa and MAGEA4b) or against a peptide of NY-ESO-1 (New York esophageal squamous cell carcinoma 1). Two isotype negative controls (formats in Figures 1A and 1C) containing irrelevant target antigen binding domains were also included.

[0189] As shown in Figures 16A, 16B, and 16C, the molecules bound to NY-ESO-1, MAGEA4a, or MAGEA4b expressing cells by flow cytometry, as expected.

[0190] As shown in Figures 17A and 17B, multispecific molecules targeting two different antigens (molecule A) or two different epitopes of a single antigen (molecule B) potently induced cytotoxicity in both metastatic non-small cell lung cancer (NSCLC) cells (Figure 17A) and NSCLC (Figure 17B), with multispecific molecules targeting two different antigens (molecule A) showing increased potency compared to multispecific molecules targeting two different epitopes of the same antigen (molecule B). The addition of an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody further enhanced the potency of an exemplary multispecific molecule of the invention (Figure 1F structure). The relative induction of T cell activation of these molecules was also evaluated and is shown in Figures 17C (metastatic NSCLC cells) and 17D (NSCLC cells).

[0191] The relative cytotoxic activity and potency of multispecific molecules targeting one or two antigens (different epitopes or different antigens) and having the structures of Figures 1C and 1F were compared to the cytotoxicity of conventional format molecules (Figure 1A) alone or in combination with anti-PD-1 antibody and costimulatory bispecific EGFRxCD28 antibody. Positive and isotype controls were as discussed above in this example. As shown in Figures 18A and 18B, the multispecific molecules were more potent than conventional format molecules, and multispecific molecules targeting two different epitopes (Figure 18A) or two different antigens (Figure 18B) were more potent than multispecific molecules targeting the same antigen with both target antigen binding domains. The relative induction of T cell activation of these molecules is shown in Figures 18C, 18D, 18E, and 18F.

[0192] Example 10: Relative cytotoxicity of multispecific molecules targeting two antigens compared to combinations of molecules in conventional formats targeting the same antigen The cytotoxicity of an exemplary multispecific molecule of the invention targeting two different antigens (Figure 1F structure) was measured as discussed above and compared to the cytotoxicity of a combination of a conventional format molecule targeting the same two antigens (Figure 1A structure) alone or in combination with an anti-PD-1 antibody and a costimulatory bispecific EGFRxCD28 antibody.

[0193] Cytotoxicity assays targeting MAGEA4-expressing SCaBER cells (bladder) demonstrated that multispecific molecules targeting both MAGEA4a and MAGEA4b (non-overlapping peptides of MAGEA4) were more potent than a combination of bispecific antibodies in a conventional format targeting the same two MAGEA4 peptides, as shown in Figure 19A. The addition of an anti-PD-1 antibody and a co-stimulatory bispecific EGFRxCD28 antibody further enhanced the potency of an exemplary multispecific molecule of the invention (Figure 1F structure). The relative induction of T cell activation by these same molecules is shown in Figure 19B.

[0194] The present invention is not to be limited in scope by the specific embodiments described herein. Indeed, various modifications of the invention in addition to those described herein will become apparent to those skilled in the art from the foregoing description. Such modifications are intended to be included within the scope of the appended claims. [Sequence table] [Table 5-1] [Table 5-2] [Table 5-3] [Table 5-4] [Table 5-5] [Table 5-6] [Table 5-7] [Table 5-8] [Table 5-9] [Table 5-10] [Table 5-11] [Table 5-12]

Table 5-13

Table 5-14

Table 5-15

Table 5-16

Table 5-17

Table 5-18

Table 5-19

Table 5-20

Table 5-21

Table 5-22

Table 5-23

Table 5-24

Table 5-25

Table 5-26

Table 5-27

Table 5-28

Table 5-29

Table 5-30

Table 5-31

Table 5-32

Table 5-33

Table 5-34

Table 5-35

Table 5-36

Table 5-37

Table 5-38

Table 5-39

Table 5-40

Table 5-41

Table 5-42

Table 5-43

Table 5-44

Table 5-45

Table 5-46

Table 5-47

Table 5-48

Table 5-49

Table 5-50

Table 5-51

Table 5-52

Table 5-53

Table 5-54

Table 5-55

Table 5-56

Claims

1. 1. A multispecific antigen-binding molecule comprising: (a) a first polypeptide comprising, from N-terminus to C-terminus, (i) a first antigen-binding domain that specifically binds a T cell antigen, (ii) a first multimerization domain, and (iii) a second antigen-binding domain that specifically binds a T cell antigen; (b) a second polypeptide comprising, from N-terminus to C-terminus, (i) a third antigen-binding domain that specifically binds to a target antigen, and (ii) a second multimerization domain; wherein said first and said second multimerization domains associate with one another to form said molecule.

2. The molecule of claim 1 , wherein the first antigen-binding domain and the second antigen-binding domain specifically bind to the same T cell antigen.

3. The molecule of claim 1 , wherein the first antigen-binding domain and the second antigen-binding domain specifically bind to different T cell antigens.

4. 4. The molecule of claim 3, wherein the first antigen-binding domain specifically binds to a first T cell antigen that is a costimulatory molecule and the second antigen-binding domain specifically binds to a second T cell antigen that is a checkpoint inhibitor.

5. The molecule of claim 4, wherein the costimulatory molecule is CD28 and the checkpoint inhibitor is PD-1.

6. The molecule of any one of claims 1 to 5, wherein one or more of the antigen binding domains is a Fab domain.

7. The molecule according to any one of claims 1 to 6, wherein the T cell antigen is a T cell receptor complex antigen.

8. The molecule of claim 7 , wherein the T cell antigen is CD3.

9. The molecule of any one of claims 1 to 6, wherein the T cell antigen is a costimulatory molecule on a T cell or a checkpoint inhibitor.

10. The molecule of any one of claims 1 to 6, wherein the T cell antigen is selected from the group consisting of CD27, CD28, 4-1BB, and PD-1.

11. The molecule of any one of claims 1 to 10, wherein the target antigen is a tumor-associated antigen.

12. The molecule of any one of claims 1 to 11, wherein the first and second multimerization domains are immunoglobulin Fc domains.

13. The molecule of claim 12 , wherein the first and second multimerization domains associate with each other via a disulfide bond.

14. The molecule of any one of claims 1 to 13, wherein the first multimerization domain and the second multimerization domain are human IgG1 or human IgG4 Fc domains.

15. 15. The molecule of any one of claims 12 to 14, wherein the first multimerization domain or the second multimerization domain comprises an amino acid substitution that reduces affinity for Protein A binding compared to a wild-type Fc domain of the same isotype.

16. 16. The molecule of claim 15, wherein the amino acid substitution comprises a H435R modification, or a H435R and a Y436F modification (EU numbering).

17. 17. The molecule of any one of claims 1-16, wherein the first polypeptide, the second polypeptide, or both the first and second polypeptides comprise a modified hinge domain that reduces binding affinity for an Fcγ receptor compared to a wild-type hinge domain of the same isotype.

18. A pharmaceutical composition comprising a molecule according to any one of claims 1 to 17 and a pharma- ceutically acceptable carrier or diluent.

19. 20. The pharmaceutical composition of claim 18 for treating cancer in a subject in need thereof.

20. 20. The pharmaceutical composition of claim 18 for treating an infection in a subject in need thereof.

Images