Bispecific binding proteins and uses thereof

Abstract

To provide bispecific molecules or proteins that bind to two epitopes and are bivalent for binding to each of the two epitopes as cancer immunotherapeutics.SOLUTION: Provided is a protein comprising: a first binding domain (BD1) that binds to a first epitope; a second binding domain (BD2) that binds to a second epitope; and an Fc region comprising CH2 and CH3 domains. The Fc region comprises the BD2 at a solvent exposed loop in the CH2 domain, the CH3 domain, or at an interface of the CH2 and CH3 domains. The protein is bivalent for binding to each of the first and second epitopes. Also provided is a method for inducing an immune response in a subject, as well as a method for treating or preventing cancer in a subject by administering the proteins, nucleic acid molecules and / or compositions to the subject.SELECTED DRAWING: Figure 1-1

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority and the benefit of U.S. Provisional Patent Application No. 62 / 332,788, filed on May 6, 2016, the disclosure of which is hereby incorporated by reference in its entirety.

[0002] Sequence Listing This application is being provided with a sequence listing in electronic format. The sequence listing is provided as a file named "IOBS_100_ST25.txt", created on May 2, 2017, and having a size of 244 kb. The information in the electronic format of the sequence listing is hereby incorporated by reference in its entirety.

[0003] Technical Field The present invention relates to bispecific binding proteins and their use.

Background Art

[0004] Cancer remains a major global health burden. Despite advances in cancer treatment, there continues to be an unmet need for more effective and less toxic therapies, particularly for patients with progressive disease or cancer that is resistant to existing therapies.

[0005] The role of the immune system, particularly T - cell - mediated cytotoxicity, in tumor control is well recognized. There is increasing evidence that T cells control tumor growth and the survival of cancer patients in both the early and late stages of the disease. However, it is difficult to enhance and sustain tumor - specific T - cell responses in cancer patients. The continued progress and success of cancer immunotherapeutic agents that stimulate or enhance the innate immune response against cancer have made such agents an attractive treatment option compared to therapies using non - specific chemotherapeutic agents and / or radiation.

[0006] Due to their potential usefulness as cancer immunotherapy (IO) drugs against cancer, several molecular targets have been identified. Some molecular targets being studied for their potential as therapeutic agents in the field of cancer immunotherapy include cytotoxic T lymphocyte antigen 4 (CTLA-4 or CD152), programmed death ligand 1 (PD-L1 or B7-H1 or CD274), programmed death-1 (PD-1), OX40 (CD134 or TNFRSF4), and T cell inhibitory receptor T cell immunoglobulin and mucin-domain containing 3 (TIM3). Some of these targets have been successfully utilized in treatment (e.g., PD-1 and CTLA-4), but many patients are unresponsive to the developed therapeutic agents. Also, treatment regimens involving higher doses and / or combination with immunotherapy drugs can be considered, but such treatments may carry an increased risk of side effects, which tend to increase with higher doses and cumulative exposure and appear to be additive when used in combination immunotherapy. Some common side effects include hypophysitis, thyroiditis, adrenal insufficiency, enteritis, dermatitis, pneumonitis, hepatitis, pancreatitis, sensory motor neuropathy, and arthritis. Furthermore, since immunotherapy drugs typically involve high costs, treatment methods involving combination of immunotherapy drugs can be prohibitively expensive for patients.

[0007] Therefore, there is still a need to identify candidate targets for IO therapeutic agents, develop new therapeutic agents against existing targets, and develop treatment strategies that avoid the drawbacks of currently used immunotherapy drugs (such as lack of patient responsiveness and increased risk of side effects associated with combination treatments). IO therapeutic agents (e.g., binding proteins) having bispecificity for combinations of target molecules, particularly those that exhibit excellent binding affinity for target molecules when compared to the binding affinity for combinations of individual monospecific binding proteins, provide a particularly desirable class of molecules for therapeutic efficacy. SUMMARY OF THE INVENTION MEANS FOR SOLVING THE PROBLEM

[0008] The present invention provides a bispecific molecule or protein that binds to two epitopes (e.g., a first and a second epitope) and is bivalent for binding to each of the first and second epitopes. The present invention also provides a method of inducing an immune response in a subject, as well as a method of treating or preventing cancer in a subject (e.g., a human subject) by administering a protein, nucleic acid molecule, and / or composition to the subject.

[0009] In one aspect, the present invention provides a protein comprising a first binding domain (BD1) that binds to a first epitope, a second binding domain (BD2) that binds to a second epitope, and an Fc region having C H 2 domain and C H 3 domain, wherein the Fc region comprises BD2 in a solvent-exposed loop at the interface of the C H 2 domain, C H 3 domain, or C H 2 and C H 3 domains, and the protein is bivalent for binding to each of the first and second epitopes.

[0010] In another aspect, the present invention provides a composition comprising a protein or antibody according to any aspect described herein and a pharmaceutically acceptable carrier.

[0011] In another aspect, the present invention provides a method of treating or preventing cancer in a subject, the method comprising administering to the subject (e.g., a human subject) a protein or antibody according to any aspect described herein. In various embodiments, the cancer is one or more of ovarian cancer, breast cancer, colorectal cancer, prostate cancer, cervical cancer, uterine cancer, testicular cancer, bladder cancer, head and neck cancer, melanoma, pancreatic cancer, renal cell cancer, and lung cancer.

[0012] In another aspect, the present invention provides a method of inducing an immune response in a subject, the method comprising administering to the subject (e.g., a human subject) a protein or antibody according to any aspect described herein.

[0013] In another aspect, the present invention provides a nucleic acid molecule having a nucleotide sequence encoding a protein or antibody according to any aspect described herein.

[0014] In another aspect, the present invention provides a vector containing a nucleic acid molecule according to any aspect described herein.

[0015] In another aspect, the present invention provides a host cell containing a vector according to any aspect described herein.

[0016] In one aspect, the present invention provides a bispecific binding protein that binds to PD-1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 1 and a second peptide having the amino acid sequence of SEQ ID NO: 2.

[0017] In another aspect, the present invention provides a bispecific binding protein that binds to PD-1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 3 and a second peptide having the amino acid sequence of SEQ ID NO: 4.

[0018] In another aspect, the present invention provides a bispecific binding protein that binds to PD-1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 5 and a second peptide having the amino acid sequence of SEQ ID NO: 6.

[0019] In one aspect, the present invention provides a bispecific binding protein that binds to PD-1 and CTLA-4 and has a first heavy chain having the amino acid sequence of SEQ ID NO: 9, a first light chain having the amino acid sequence of SEQ ID NO: 7, a second heavy chain having the amino acid sequence of SEQ ID NO: 12, and a second light chain having the amino acid sequence of SEQ ID NO: 4.

[0020] In one aspect, the present invention provides a bispecific binding protein that binds to PD-L1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 14 and a second peptide having the amino acid sequence of SEQ ID NO: 15.

[0021] In another aspect, the present invention provides a bispecific binding protein that binds to PD-L1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 16 and a second peptide having the amino acid sequence of SEQ ID NO: 17.

[0022] In another aspect, the present invention provides a bispecific binding protein that binds to PD-L1 and CTLA-4 and has a first peptide having the amino acid sequence of SEQ ID NO: 18 and a second peptide having the amino acid sequence of SEQ ID NO: 19.

[0023] In one aspect, the present invention provides a bispecific binding protein that binds to PD-1 and TIM3 and has a first peptide having the amino acid sequence of SEQ ID NO: 22 and a second peptide having the amino acid sequence of SEQ ID NO: 23.

[0024] In another aspect, the present invention provides a bispecific binding protein that binds to PD-1 and TIM3 and has a first peptide having the amino acid sequence of SEQ ID NO: 24 or SEQ ID NO: 91 and a second peptide having the amino acid sequence of SEQ ID NO: 23 or SEQ ID NO: 92.

[0025] In one aspect, the present invention provides a bispecific binding protein that binds to PD-1 and TIM3 and has a first heavy chain having the amino acid sequence of SEQ ID NO: 9, a first light chain having the amino acid sequence of SEQ ID NO: 7, a second heavy chain having the amino acid sequence of SEQ ID NO: 27 or SEQ ID NO: 30, and a second light chain having the amino acid sequence of SEQ ID NO: 26 or SEQ ID NO: 28.

[0026] In one aspect, the present invention provides a bispecific binding protein that binds to OX40 and PD-L1 and has a first peptide having the amino acid sequence of SEQ ID NO: 34 and a second peptide having the amino acid sequence of SEQ ID NO: 32.

[0027] In another aspect, the present invention provides a bispecific binding protein that binds to OX40 and PD-L1 and has a first peptide having the amino acid sequence of SEQ ID NO: 35 and a second peptide having the amino acid sequence of SEQ ID NO: 32.

[0028] In another aspect, the present invention provides a bispecific binding protein that binds to OX40 and PD-L1 and has a first peptide having the amino acid sequence of SEQ ID NO: 36 or SEQ ID NO: 94 and a second peptide having the amino acid sequence of SEQ ID NO: 32 or SEQ ID NO: 93.

[0029] In another aspect, the present invention provides an antibody or antigen-binding fragment thereof that binds to TIM3 and has a heavy chain having CDR1, CDR2, and CDR3 and a light chain having CDR1, CDR2, and CDR3, wherein heavy chain CDR1 comprises SEQ ID NO: 88, heavy chain CDR2 comprises SEQ ID NO: 80, heavy chain CDR3 comprises SEQ ID NO: 81, and light chain CDR1 comprises SEQ ID NO: 82, light chain CDR2 comprises SEQ ID NO: 83, and light chain CDR3 comprises SEQ ID NO: 84.

[0030] In other aspects, the present invention provides a composition having a bispecific binding protein and a pharmaceutically acceptable carrier, a nucleic acid molecule having a nucleotide sequence encoding a bispecific binding protein, a method of treating or preventing cancer in a subject by administering a bispecific binding protein, and a method of enhancing an immune response in a subject by administering a bispecific binding protein.

[0031] In various embodiments of any of the aspects described herein, the Fc region is C H 2 domain, C H 3 domain, or C H 2 and CH It contains BD2 in the solvent-exposed loop in the amino acid sequence at the interface of the 3 domains.

[0032] In various embodiments of any of the aspects described herein, the solvent-exposed loop contains an amino acid sequence derived from the C H 2 domain. In certain embodiments, the solvent-exposed loop contains the amino acid sequence ISRTP (SEQ ID NO: 39).

[0033] In various embodiments of any of the aspects described herein, the solvent-exposed loop contains an amino acid sequence derived from the C H 3 domain. In certain embodiments, the solvent-exposed loop contains the amino acid sequence SNG.

[0034] In various embodiments of any of the aspects described herein, the solvent-exposed loop contains an amino acid sequence derived from the C H 2 domain and the C H 3 domain interface. In certain embodiments, the solvent-exposed loop contains the amino acid sequence AKGQP (SEQ ID NO: 40), the protein according to claim 7.

[0035] In various embodiments of any of the aspects described herein, BD2 is a single-chain variable fragment (scFv) or contains it.

[0036] In various embodiments of any of the aspects described herein, BD1 is a binding domain that is one or more of a Fab domain, scFv, single-domain antibody, and antibody variable domain, or contains it. In certain embodiments, BD1 contains a Fab domain.

[0037] In various embodiments of any of the aspects described herein, the Fab domain is linked to the Fc domain via an antibody hinge region. In certain embodiments, the Fc domain is or comprises one or more domains of an Fc domain derived from IgG1, IgG2, IgG3, IgG4, IgA, IgM, IgE, or IgD. In certain embodiments, the Fc domain comprises a mutant Fc domain. In some embodiments, the Fc domain is non-glycosylated, deglycosylated, and / or non-fucosylated, or has low fucosylation.

[0038] In various embodiments of any of the aspects described herein, the protein further comprises a protein linker L1 between BD2 and the Fc domain. In various embodiments of any of the aspects described herein, the protein further comprises a first protein linker L1 and a second protein linker L2 between BD2 and the Fc domain. In various embodiments of any of the aspects described herein, BD2 is bound to the Fc domain via the protein linker L1. In various embodiments of any of the aspects described herein, BD2 is bound to the Fc domain via two protein linkers L1 and L2. In certain embodiments, L1 and L2 are independently selected from (G4S)2 (SEQ ID NO: 41), (G4S)3 (SEQ ID NO: 42), and (G4S)4 (SEQ ID NO: 43).

[0039] In various embodiments of any of the aspects described herein, the protein has the following polypeptide domains from the N-terminus to the C-terminus: V H 1-C H 1-C H 2 (N-terminus)-BD2-C H 2 (C-terminus)-C H and comprises a chimeric heavy chain having 3, and BD1 comprises a Fab domain, V H 1 comprises the heavy chain variable domain of the Fab domain, and C H 1 comprises the heavy chain constant domain 1 of the Fab.

[0040] In various embodiments of any of the aspects described herein, the protein comprises, from the N-terminus to the C-terminus, the following polypeptide domains: V H 1-C H 1-C H 2-BD2-C H 3 and comprises a chimeric heavy chain having, and BD1 comprises a Fab domain, V H 1 comprises the heavy chain variable domain of the Fab domain, and C H 1 comprises the heavy chain constant domain 1 of the Fab.

[0041] In various embodiments of any of the aspects described herein, the protein comprises, from the N-terminus to the C-terminus, the following polypeptide domains: V H 1-C H 1-C H 2-C H 3(N-terminus)-BD2-C H 3(C-terminus) and comprises a chimeric heavy chain having, and BD1 comprises a Fab domain, V H 1 comprises the heavy chain variable domain of the Fab domain, and C H 1 comprises the heavy chain constant domain 1 of the Fab.

[0042] In various embodiments of any of the aspects described herein, BD2 is or comprises an scFv. In certain embodiments, the scFv comprises, from the N-terminus to the C-terminus, V H 2-polypeptide linker-V L 2 or V L 2 polypeptide linker-V H 2 and comprises, V H 2 comprises the heavy chain variable domain of the scFv, and V L 2 comprises the light chain variable domain of the scFv.

[0043] In various embodiments of any of the aspects described herein, the protein further comprises a protein linker L1 between BD2 and the Fc region. In various embodiments of any of the aspects described herein, the protein further comprises a first protein linker L1 and a second protein linker L2 between BD2 and the Fc region.

[0044] In various embodiments of any of the aspects described herein, BD2 is attached via a linker (L1) to the C of the Fc region H 2 domain, C H 2 domain, or C H 2 and C H 3 domain interface.

[0045] In various embodiments of any of the aspects described herein, BD2 is attached via two protein linkers L1 and L2 to the C of the Fc region H 2 domain, C H 3 domain, or C H 2 and C H 3 domain interface. In various embodiments, L1 and L2 are independently selected from protein linkers having a length of 1 to 25 amino acids. In certain embodiments, L1 and L2 are independently selected from (G4S)2 (SEQ ID NO: 41), (G4S)3 (SEQ ID NO: 42), and (G4S)4 (SEQ ID NO: 43).

[0046] In various embodiments of any of the aspects described herein, the first and second epitopes are different. In various embodiments of any of the aspects described herein, the first and second epitopes are the same.

[0047] In the description of the present disclosure, certain aspects of the present disclosure are shown in the figures. However, the present disclosure is not limited to the exact arrangements and means of the aspects shown in the figures.

Brief Description of the Drawings

[0048]

Figure 1-1

Figure 1-2

Figure 2-1

Figure 2-2

Figure 2-3

Figure 3

Figure 4

Figure 5

Figure 6-1

Figure 6-2

Figure 7

Figure 8

Figure 9

Figure 10-1

Figure 10-2

Figure 11-1

Figure 11-2

Figure 12-1

Figure 12-2

Figure 13-1

Figure 13-2

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23-1

Figure 23-2

Figure 24

Figure 25

Figure 26

Figure 27A

Figure 27B

Figure 28-1

Figure 28-2

Figure 29

Figure 30

Figure 31

Figure 32

Figure 33

Figure 34

Figure 35

Figure 36-1

Figure 36-2

Figure 37

Figure 38

Figure 39

Figure 40-1

Figure 40-2

Figure 41

Figure 42

Figure 43

Figure 44

Figure 45

Figure 46

Figure 47-1

Figure 47-2

Figure 48-1

Figure 48-2

Figure 49

Figure 50

Figure 51

Figure 52-1

Figure 52-2

Figure 53

Figure 54

Figure 55

Figure 56

Figure 57

Figure 58

Figure 59

Figure 60

Figure 61-1

Figure 61-2

Figure 62

Figure 63

Figure 64

Figure 65

Figure 66

Figure 67

Figure 68

Figure 69

Figure 70

Figure 71

Figure 72

Figure 73-1

Figure 73-2

Figure 74-1

Figure 74-2

Figure 75

Figure 76

Figure 77-1

Figure 77-2

Figure 78

Figure 79

Figure 80

Figure 81

Figure 82

Figure 83

Figure 84

Figure 85

Figure 86

Figure 87

Figure 88

Figure 89

Figure 90

DETAILED DESCRIPTION OF THE INVENTION

[0049] Before continuing to describe the present disclosure in more detail, it should be understood that the present disclosure is not limited to specific compositions or process steps and may thus vary. It should be noted that, as used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise.

[0050] Unless otherwise noted, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. For example, Concise Dictionary of Biomedicine and Molecular Biology, Juo, Pei-Show, 2nd ed., 2002, CRC Press; The Dictionary of Cell and Molecular Biology, 3rd ed., 1999, Academic Press; and Oxford Dictionary Of Biochemistry And Molecular Biology, Revised, 2000, Oxford University Press serve as general dictionaries for many of the terms used in this invention for those of ordinary skill in the art.

[0051] In this specification, amino acids may be denoted by their generally known three-letter symbols or by the one-letter symbols recommended by the IUPAC-IUB Biochemical Nomenclature Commission. Nucleotides, likewise, may be denoted by their generally recognized one-letter symbols.

[0052] The numbering of amino acids in the variable domains, complementarity determining regions (CDRs), and framework regions (FRs) of an antibody, unless otherwise specified, follows the Kabat definition as described in Kabat et al. Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991). Using this numbering system, the actual linear amino acid sequence can contain fewer or additional amino acids corresponding to deletions or insertions in the FR or CDR of the variable domain. For example, the heavy chain variable domain can contain an insertion of one amino acid after residue 52 of H2 (residue 52a according to Kabat), and residues inserted after heavy chain FR residue 82 (e.g., residues 82a, 82b, and 82c according to Kabat). The Kabat numbering of residues can be determined for an individual antibody by aligning the antibody's sequence with the sequence of the "standard" Kabat-numbered sequence in the region of homology. Alignment of the framework residues often requires the insertion of "spacer" residues of this numbering system used in the Fv region for maximum alignment. Further, the particular type of individual residue at any given Kabat site number can vary depending on the antibody chain due to interspecies differences or allelic differences.

[0053] As used herein, the term "antibody" is also referred to as immunoglobulin and includes monoclonal antibodies (including full-length monoclonal antibodies), polyclonal antibodies, multispecific antibodies formed from at least two different epitope-binding fragments (e.g., bispecific antibodies, e.g., International Publication No. WO 2009 / 018386 pamphlet, International Application PCT / US Patent Application Publication No. US 2012 / 045229, which are incorporated herein by reference in their entirety), BiSAb, human antibodies, humanized antibodies, camelized antibodies, single-chain Fv (scFv), single-chain antibodies, single-domain antibodies, domain antibodies, Fab fragments, F(ab')2 fragments, antibody fragments exhibiting desired biological activity (e.g., antigen-binding portions), disulfide-bonded Fv (dsFv), and anti-idiotype (anti-Id) antibodies (e.g., anti-Id antibodies against the antibodies of the present invention), intrabodies, as well as epitope-binding fragments of any of the foregoing. In particular, an antibody includes an immunoglobulin molecule and an immunologically active fragment of the immunoglobulin molecule, i.e., a molecule containing at least one antigen-binding site. An antibody also includes a peptide fusion with an antibody or a portion thereof, e.g., a fusion protein with an Fc domain. The immunoglobulin molecule may be of any isotype (e.g., IgG, IgE, IgM, IgD, IgA, and IgY), subtype (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2), or allotype (e.g., Gm, e.g., G1m(f, z, a, or x), G2m(n), G3m(g, b, or c), Am, Em, and Km(1, 2, or 3)). The antibody may be derived from any mammal, including but not limited to humans, monkeys, pigs, horses, rabbits, dogs, cats, mice, etc., or other animals, e.g., birds (e.g., chickens).

[0054] CTLA-4 Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) is expressed on activated T cells and acts as a co-inhibitor that blocks T cell responses after CD28-mediated T cell activation. CTLA-4 is thought to be part of a central inhibitory pathway that regulates the magnitude of early activation of naive and memory T cells after TCR engagement and affects both anti-tumor immunity and autoimmunity. CTLA-4 is expressed only on T cells, and the expression of its ligands CD80 (B7.1) and CD86 (B7.2) is mainly limited to antigen-presenting cells, T cells, and other immune mediator cells. Antagonistic anti-CTLA-4 antibodies that block the CTLA-4 signaling pathway have been reported to enhance T cell activation. One such antibody, ipilimumab, was approved by the FDA in 2011 for the treatment of metastatic melanoma. The use of anti-CTLA-4 antibodies for treating infectious diseases and tumors and upregulating adoptive immune responses has been proposed (see U.S. Patent Nos. 6,682,736; 7,109,003; 7,132,281; 7,411,057; 7,824,679; 8,143,379; 7,807,797; 8,491,895; 8,883,984; and U.S. Patent Application Publication No. 20150104409; these are hereby incorporated by reference in their entirety).

[0055] PD-L1 Programmed death ligand 1 (PD-L1) is also part of a complex system of receptors and ligands involved in the regulation of T cell activation. In normal tissues, PD-L1 is expressed on T cells, B cells, dendritic cells, macrophages, mesenchymal stem cells, bone marrow-derived mast cells, and various non-hematopoietic cells. Its normal function is to regulate the balance between T cell activation and tolerance through its interaction with two of its receptors: programmed death 1 (also known as PD-1 or CD279) and CD80 (also known as B7-1 or B7.1). PD-L1 is also expressed by tumors and acts at multiple sites to promote the escape of tumors from detection and elimination by the host immune system. PD-L1 is frequently expressed in a variety of cancers. In some cancers, the expression of PD-L1 is associated with low survival rates and poor prognosis. Antibodies that block the interaction between PD-L1 and its receptor can reduce the PD-L1-dependent immunosuppressive effect and enhance the cytotoxic activity of anti-tumor T cells in vitro. Durvalumab is a human monoclonal antibody against human PD-L1 that can block the binding of both PD-1 and CD80 receptors to PD-L1. The use of anti-PD-L1 antibodies for treating infections and tumors and enhancing adoptive immune responses has been reported (see U.S. Patent Nos. 8,779,108 and 9,493,565, which are hereby incorporated by reference in their entirety).

[0056] PD-1 Programmed death-1 (“PD-1”) is a type I membrane protein member of the extended CD28 / CTLA-4 family of T cell regulators, approximately 31 kDa (see Ishida, Y. et al. (1992) Induced Expression Of PD-1, A Novel Member Of The Immunoglobulin Gene Superfamily, Upon Programmed Cell Death, ”EMBO J. 11: 3887-3895).

[0057] PD-1 is expressed on activated T cells, B cells, and monocytes (Agata, Y. et al. (1996) “Expression of the PD-1 Antigen on the Surface of Stimulated Mouse T and B Lymphocytes,” Int. Immunol. 8(5):765-772; Martin-Orozco, N. et al. (2007) “Inhibitory Costimulation and Anti-Tumor Immunity,” Semin. Cancer Biol. 17(4):288-298). PD-1 is a receptor that causes downregulation of the immune system after activation by binding to PDL-1 or PDL-2 (Martin-Orozco, N. et al. (2007) “Inhibitory Costimulation and Anti-Tumor Immunity,” Semin. Cancer Biol. 17(4):288-298) and acts as a cell death inducer (Ishida, Y. et al. (1992) “Induced Expression of PD-1, A Novel Member of The Immunoglobulin Gene Superfamily, Upon Programmed Cell Death,” EMBO J. 11:3887-3895; Subudhi, S.K. et al. (2005) “The Balance of Immune Responses: Costimulation Verse Coinhibition,” J. Molec. Med. 83:193-202). This process is exploited in many tumors via overexpression of PD-L1, leading to suppression of the immune response.

[0058] PD-1 is a well-validated target for immune-mediated therapies in oncology, and has yielded promising results especially from clinical trials for the treatment of melanoma and non-small cell lung cancer (NSCLC). Antagonistic inhibition of the PD-1 / PD-L1 interaction enhances T cell activation and the recognition and elimination of tumor cells by the host immune system. The use of anti-PD-1 antibodies for treating infections and tumors and augmenting adoptive immune responses has been proposed (see U.S. Patent Nos. 7,521,051; 7,563,869; 7,595,048).

[0059] OX40 OX40 (CD134; TNFRSF4) is a tumor necrosis factor receptor found primarily on activated CD4 + and CD8 + T cells, regulatory T (Treg) cells, and natural killer (NK) cells (Croft et al., 2009, Immunol Rev. 229:173-91). OX40 has one known endogenous ligand, OX40 ligand (OX40L; CD152; TNFSF4), which exists in a trimeric form and can cluster OX40, thereby resulting in potent cell signaling events within T cells. Id. Signaling through OX40 on activated CD4 + and CD8 + T cells results in increased cytokine production, granzyme and perforin release, and expansion of the effector and memory T cell pools (Jensen et al., 2010, Semin Oncol. 37:524-32). In addition, OX40 signaling in Treg cells inhibits Treg expansion, halts Treg induction, and blocks Treg suppressive function (Voo et al., 2013, J Immunol. 191:3641-50; Vu et al., 2007, Blood. 110:2501-10).

[0060] Immunohistochemical studies and early flow cytometry analysis have demonstrated that OX40 is expressed on T cells infiltrating a variety of human cancers (Baruah et al., 2011, Immunobiology 217:668-675; Curti et al, 2013, Cancer Res. 73:7189-98; Ladanyi et al, 2004, Clin Cancer Res. 10:521-30; Petty et al, 2002, Am J Surg. 183:512-8; Ramstad et al, 2000, Am J Surg. 179:400-6; Sarff et al, 2008, Am J Surg. 195:621-5; discussion 625; Vetto et al, 1997, Am J Surg. 174:258-65). Without intending to be bound by any particular theory, OX40 expression in tumor-infiltrating lymphocytes has been correlated with longer survival in several human cancers, suggesting that the OX40 signal may play a role in the establishment of an antitumor immune response (Ladanyi et al., 2004, Clin Cancer Res. 10:521-30; Petty et al., 2002, Am J Surg. 183:512-8).

[0061] In a variety of non-clinical mouse tumor models, the use of agonists of OX40, including antibodies and OX40 ligand fusion proteins, has been successful and has yielded promising results (Kjaergaard et al., 2000, Cancer Res. 60:5514-21; Ndhlovu et al., 2001, J Immunol. 167:2991-9; Weinberg et al., 2000, J Immunol. 164:2160-9). Co-stimulation of T cells by OX40 has, in some cases, promoted sustained anti-tumor activity and provided protection that persists for an extended period against subsequent tumor challenges (Weinberg et al., 2000, J Immunol. 164:2160-9). It has been shown that Treg cell inhibition and co-stimulation of effector T cells are required for the inhibition of tumor growth by OX40 agonists (Piconese et al., 2008, J Exp Med. 205:825-39). To enhance the anti-tumor effects of OX40 agonist therapy, many strategies and techniques have been explored through combination with vaccines, chemotherapy, radiotherapy, and immunotherapy (Jensen et al., 2010, Semin Oncol. 37:524-32; Melero et al., 2013, Clin Cancer Res. 19:997-1008). The use of anti-OX40 antibodies to treat infectious diseases and tumors and up-modulate adoptive immune responses has been proposed (see U.S. Patent Application Publication No. 20160137740, which is hereby incorporated by reference in its entirety).

[0062] TIM3 The T cell inhibitory receptor Tim-3 (T cell immunoglobulin and mucin-domain containing 3) is expressed on IFNγ-producing CD4+ helper 1 (Th1) and CD8+ T cell cytotoxic 1 (Tc1) T cells, and thus plays a role in the regulation of anti-tumor immunity. It was initially identified as a T cell inhibitory receptor that functions as an immune checkpoint receptor, particularly to limit the duration and magnitude of Th1 and Tc1 T cell responses. Further studies have identified that the Tim-3 pathway can cooperate with the PD-1 pathway to promote the development of a severely dysfunctional phenotype in CD8+ T cells in cancer. It is also expressed on regulatory T cells (Tregs) in certain cancers. Given the involvement of the TIM3 pathway in the major immune cell populations immunosuppressed in several cancers, this represents an attractive candidate for cancer immunotherapy. See Anderson, A.C., Cancer Immunol Res., (2014) 2:393-398; and Ferris, R.L., et al., J Immunol. (2014) 193:1525-1530.

[0063] A. Bispecific binding protein Adding multiple binding sites to a molecule with specificity for a single binding domain can greatly enhance the molecule's capabilities (e.g., therapeutic, diagnostic, etc.). For example, a bispecific antibody can bind to two or more regions of the same target biomolecule, conferring greater specificity than a monospecific polypeptide that binds to only one epitope of the target. Alternatively, a bispecific antibody can bind to multiple target biomolecules, such as targets present in a complex or targets for which blockade and / or clustering is desirable. In a third scenario, the same bispecific antibody can exhibit different functions depending on the localization and / or expression of its target molecule.

[0064] Described herein are novel binding proteins. One such structure of these novel binding proteins is called "DuetMab". DuetMab has the following basic structure: a modified heavy chain (wherein the CH1 region of the modified heavy chain has substitutions of native cysteine with non-cysteine amino acids and substitutions of native non-cysteine amino acids with cysteine amino acids); a corresponding modified light chain (wherein the CL region of the modified light chain also has substitutions of native cysteine with non-cysteine amino acids and substitutions of native non-cysteine amino acids with cysteine amino acids); a second Fc region having a second heavy chain; and an Fc region having a corresponding second modified light chain (wherein the modified heavy chain is directly linked to the corresponding modified light chain, on individual target binding arms, the second heavy chain is directly linked to the second corresponding light chain, and the substituted cysteine of the modified heavy chain obtained by substitution of native non-cysteine amino acids with cysteine amino acids and the substituted cysteine of the corresponding modified light chain obtained by substitution of native non-cysteine amino acids with cysteine amino acids can form a disulfide bond). The disclosure regarding DuetMab can be found, for example, in U.S. Patent No. 9,527,927, which is hereby incorporated by reference in its entirety.

[0065] Another exemplary structure of these novel binding proteins is referred to as “BiSAb” or “BiSAbs”. Schematic diagrams of exemplary BiSAb, as well as specific examples of particular BiSAb, are provided herein. More generally, a BiSAb is a polypeptide containing two binding units, each of which binds to an epitope (e.g., binding unit 1 binds to a first epitope and binding unit 2 binds to a second epitope). A basic BiSAb is bivalent for binding to each of the two epitopes (e.g., the polypeptide includes two binding units 1 (“BD1” or “BU1”) and two binding units 2 (“BD2” or “BU2”). Thus, when binding units 1 and 2 bind to different epitopes, the BiSAb has the multispecificity of a conventional bispecific antibody and the bivalency of a conventional antibody molecule. In embodiments where binding units 1 and 2 bind to the same epitope, the BiSAb has the single specificity of a normal antibody but is tetravalent. In addition to the binding units, the BiSAb also includes a linker polypeptide and an Fc portion. The present disclosure relates to a broad set of bispecific binding proteins, such as proteins including BiSAb and BiSAb cores, which target molecules that modulate the immune response. Generally, the novel binding protein platforms and exemplary bispecific binding proteins (BiSAb) described herein include binding units / domains, linker polypeptides, and Fc portions. The present disclosure also provides nucleic acid molecules encoding such BiSAb, as well as vectors and host cells that contain such nucleic acids and can be used in methods for producing and using such BiSAb. BiSAb, binding proteins including BiSAb cores, and various portions of BiSAb are described in further detail herein.

[0066] In some embodiments, a BiSAb may comprise two heavy-chain - light-chain pairs derived from a specific binding protein (i.e., an antibody), where the heavy and light chains each comprise variable regions (e.g., VL and VH), which together form a first binding unit, and the heavy chains each further comprise a second binding unit (e.g., an scFv domain bound to Fc or Fab). When the first and second binding units bind to different epitopes, each heavy-chain - light-chain pair is bispecific, and two pairs are bivalent together for each of the epitopes. When the first and second binding units bind to the same epitope, each heavy-chain - light-chain pair is monospecific, and two pairs are tetravalent together for that epitope. In some embodiments, the two heavy-chain - light-chain pairs are identical. In some embodiments, the two heavy-chain - light-chain pairs are not identical.

[0067] In certain embodiments, the domains of the BiSAb may be based on known immunoglobulin domains. Immunoglobulin molecules such as monoclonal antibodies (mAbs) are widely used as diagnostic and therapeutic agents, and methods for generating binding fragments of mAbs are known in the art. Monoclonal antibodies, such as any immunoglobulin molecule, are composed of heavy and light chain peptide subunits, which each contain variable and constant domains that confer binding specificity (variable domain) and isotype (constant domain), respectively.

[0068] The BiSAb disclosed herein has an overall structure similar to that of a general antibody, but is distinguishable by the presence of additional binding units that are linked at positions within the Fab domain and at positions remote from the Fab domain and within the hinge or Fc region (e.g., at the interfaces of such regions such as the CH2, CH3, or CH4 regions or the CH2-CH3 interface). Thus, unlike a normal antibody that is bivalent for binding to a single epitope, the BiSAb is bivalent for binding to two epitopes. However, as described herein, the BiSAb can still maintain many desirable properties of normal antibodies, such as the ability to bind to C1q and FcRn and the ability to bind to Fcγ receptors (e.g., the ability to mediate antibody- and complement-dependent cytotoxicity).

[0069] The binding domains described herein may include antigen-binding fragments containing only portions of mAb molecules such as Fab, F(ab’)2, Fab’, scFv, di-scFv, sdAb fragments, etc. This is because these fragments have been found to be useful as diagnostic or therapeutic agents. In addition, specific residues in the variable domains can also be modified to improve the binding specificity and / or stability of the antibody and antibody fragments. Other residues that do not directly participate in antigen binding can be substituted to “humanize” regions of non-human antibodies and reduce the immunogenicity of the mAb.

[0070] Unlike conventional antibodies, BiSAb is, for example, bivalent (or tetravalent for binding to one epitope) with respect to binding to two different epitopes, but many parts of BiSAb are derived from or similar to conventional antibody parts. Any mAb domains and / or fragments known in the art may be used for the BiSAb described herein. In particular, BiSAb may include Fab fragments and / or scFv fragments or variants thereof. Exemplary and non-limiting variants of scFv include, but are not limited to, tandem di-scFv, tandem tri-scFv, diabody and tribody or triabody.

[0071] The present disclosure generally relates to novel binding proteins, of which BiSAb is an example. Another example includes binding proteins that include one or more additional binding units in addition to the BiSAb core and / or binding proteins that include an extended BiSAb core. When describing BiSAb or the features of BiSAb herein, it should be understood that such description generally applies to the novel binding proteins of the present disclosure, regardless of whether such binding proteins include two binding units or three or more binding units. Thus, the term BiSAb is illustrative of the binding proteins described herein, and any such reference to BiSAb may be used to describe a binding protein that includes a BiSAb core, where context permits.

[0072] Novel BiSAb Structural Platform In one aspect, the present disclosure provides a BiSAb binding protein having a structural platform generally shown by the schematic diagrams of FIGS. 1A-1F. These figures are illustrative, and thus insertions between additional residues are also included within the scope of the disclosed binding proteins. FIGS. 1A-1C depict the Fc region of an antibody at the CH2-CH3 interface of IgG1 modeled using PyMOL and show some exemplary BiSAb of the present disclosure. Three surface-exposed loops were identified at or near the CH2-CH3 interface, which may be able to withstand the insertion of a second binding moiety (e.g., scFv) without compromising the structural integrity or stability of the IgG or the second binding moiety. FIG. 1A is a schematic diagram of one such representative loop, ISRTP (SEQ ID NO: 39), identified within the CH2 region near the CH2-CH3 interface. FIG. 1D also shows the representative construct IS-scFv-RTP, where the scFv is inserted between the S and R of the ISRTP loop. FIG. 1B is a schematic diagram of the representative loop AKGQP (SEQ ID NO: 40) identified at the CH2-CH3 interface. FIG. 1E shows the representative construct AK-scFv-GQP, where the scFv is inserted between the K and G of the AKGQP loop. FIG. 1C is a schematic diagram of the representative loop SNG identified within the CH3 region downstream of the CH2-CH3 interface. FIG. 1F also shows the representative construct S-scfv-NG, where the scFv is inserted between the S and N of the SNG loop. The examples described herein provide an illustration of constructs oriented like SN-scFv-G, where in this case the scFv is inserted between the N and G of the SNG loop.

[0073] Accordingly, one aspect of the present disclosure relates to a BiSAb comprising two identical heavy chain-light chain pairs, where each heavy chain-light chain pair is bispecific and the two identical pairs are bivalent together for each epitope. Each heavy chain-light chain pair may include a binding domain (BD) (binding unit 1) that may include a Fab domain that binds to a first epitope, a second binding domain (BD2) (or binding unit 2, which may be, for example, an scFv) that binds to a second epitope, and an Fc region. In some embodiments, the second binding domain may be bound to the Fab domain. In some embodiments, the Fc region of the BiSAb may be bound to a second binding domain (BD2) (binding unit 2; depicted as an scFv in FIGS. 1D-1F) that binds to the second epitope.

[0074] In some embodiments, the present disclosure provides a BiSAb having a general platform structure comprising two chimeric heavy chains, each of which comprises a heavy chain variable region (VH1), a heavy chain constant region (CH1), a hinge or polypeptide linker region, and an Fc region comprising a CH2 domain and a CH3 domain, where optionally, a second binding domain (BD2) flanked on one or both sides by a polypeptide linker (L1 and / or L2) is bound to a solvent-exposed loop within the Fc region in the order of (i) the CH2 region, (ii) the CH2 and CH3 regions, or (iii) the CH3 region. The BiSAb of this aspect of the present disclosure also includes two normal antibody light chains, each of which comprises a light chain variable region (VL1) and a light chain constant region (CL) and forms part of the first binding domain (BD1). The binding domain (BD2) of the specific BiSAb shown in FIGS. 1D-1F is an scFv.

[0075] Figures 1D - 1F provide useful schematic diagrams of BiSAb, which may also be referred to herein as the BiSAb "core". As shown in Figure 1D, the polypeptide chain includes a VH1 domain, a CH1 domain, a hinge / linker, a partial N - terminal CH2 domain, an optional linker (referred to herein as L1 or the first polypeptide linker), a binding unit 2 (such as VL2 and VH2 of an scFv), another optional linker (e.g., L2 or the second polypeptide linker), the remaining C - terminal CH2 domain, and a CH3 domain of a heavy chain. This heavy chain may include an alternative binding protein and / or BD2 having a conventional light chain region, and is thus referred to herein as a chimeric heavy chain. A BiSAb includes two such chimeric heavy chains, which may be the same or different. Note that the variable heavy chain domain (VH) of binding unit 1 is VH1. In one embodiment, this is the variable heavy chain of a Fab that binds to a first epitope. Similarly, the variable light chain domain (VL) of binding unit 1 is VL1. In one embodiment, this is the variable light chain of a Fab that binds to a first epitope. On the other hand, in an embodiment where binding unit 2 is an scFv that binds to a second epitope, the domains of binding unit 2 are denoted by the number "2", such as VH2 and VL2.

[0076] Similarly, as shown in Figure 1E, the polypeptide chain includes a VH1 domain, a CH1 domain, a hinge / linker, a CH2 domain, an optional linker (referred to herein as L1 or the first polypeptide linker), a binding unit 2 (such as VL2 and VH2 of an scFv), another optional linker (e.g., L2 or the second polypeptide linker), and a CH3 domain of a heavy chain. In this embodiment, the BiSAb includes a second binding domain shown as an scFv that binds to Fc with a sequence at the interface of the CH2 and CH3 regions.

[0077] As shown in Figure 1F, the polypeptide chain includes a VH1 domain, a CH1 domain, a hinge / linker, a CH2 domain, a partial CH3 domain, an optional linker (referred to herein as L1 or the first polypeptide linker), a binding unit 2 (such as VL2 and VH2 of the scFv), another optional linker (for example, L2 or the second polypeptide linker), and a heavy chain having a CH3 domain.

[0078] In these embodiments, the BiSAb typically includes a typical or modified antibody hinge region in the chimeric heavy chain sequence. Non-limiting examples of amino acid sequences containing a hinge region include EPKSCDKTHTCPPCP (SEQ ID NO: 44); EPKSCDKT (SEQ ID NO: 45); EPKSCGKT (SEQ ID NO: 46); EPKSC (SEQ ID NO: 47).

[0079] Although a general format has been described for aspects related to the specific structural platforms of some BiSAb molecules disclosed herein, the various parts and exemplary functional characteristics of the disclosed BiSAb will be described in further detail below. In other embodiments, the present disclosure contemplates and provides other BiSAb binding proteins including alternative structural formats and arrangements briefly described herein and in other disclosures incorporated herein by reference in their entirety (see, for example, US Patent Application Publication No. 20090155275 and US Patent No. 9,580,509).

[0080] 1. Binding Unit The BiSAb of the present disclosure comprises at least two binding units or binding domains (binding unit / domain 1 and binding unit / domain 2). In certain embodiments, each binding unit binds to a different epitope, whether on the same target molecule or on different targets. Since the binding units of the BiSAb exist as pairs (two binding units 1 and two binding units 2), the BiSAb exhibits bivalent binding to each epitope. From the teachings herein, it will be understood that if each binding unit binds to the same epitope, the BiSAb exhibits tetravalent binding to the epitope.

[0081] In certain embodiments, the first binding unit is a Fab fragment, such as a Fab fragment of a normal monoclonal antibody, or an antigen-binding fragment produced by recombinant techniques comprising a variable light chain (VL1), a constant light chain (CL), a variable heavy chain (VH1), and a constant heavy chain portion (CH1). Optionally, the light and heavy chains of the Fab may be linked to each other via one or more disulfide bonds, such as a suitable antibody hinge region. The Fab binds to a first epitope.

[0082] In certain embodiments, the Fab is derived from or based on the sequence of a normal monoclonal antibody, such as a normal mouse, humanized, or human antibody. In certain embodiments, a BiSAb containing a Fab derived from or based on the sequence of a normal monoclonal antibody retains one or more functional activities of a normal antibody (e.g., retains at least 80% or more (80%, 85%, 90%, 95%, 97%, 98%, 99% or 100%) of the functional activity). For example, in certain embodiments, a BiSAb containing such a Fab retains one or more of the affinity for an antigen, inhibitory activity, immune system modulation activity, activation or induction of an immune response, and / or cell (e.g., cancer cell) killing activity of a normal antibody.

[0083] In one aspect, the BiSAb of the present disclosure includes a binding unit 2, and the binding unit 2 includes a binding domain that binds to a second epitope. The binding unit 2 (or binding domain 2 (BD2)) may bind to the BiSAb using any suitable strategy. As used herein, BD2 that is "bound to" the BiSAb (e.g., within the Fc region in some embodiments and within the Fab region in other embodiments) means that these two molecules have an interaction between them such that they retain the orientation for target binding and the binding to the Fc or Fab portion of the BiSAb structure. Examples of such interactions include covalent bonds via an amino acid linker, covalent bonds by recombinant expression of BD2 within the CH2, CH3, or the interface of CH2 and CH3, or in the Fab region, hinge region, or Fc region of the CH4 region, as well as non-covalent interaction such as van der Waals forces and hydrogen bond interactions within these same regions. Non-limiting examples of binding domains (or "BD" or "binding unit") included within the scope of the present disclosure include antibody variable regions, antibody fragments, scFv, single-chain diabodies, or other binding domains known in the art. The binding domain also includes a bispecific single-chain diabody or a single-chain diabody designed to bind two different epitopes. In one aspect, the epitope-binding domains useful for constructing the multispecific epitope-binding domains of the present disclosure are exemplified in U.S. Patent Application Publication Nos. 20100298541 and 20130079280, which are incorporated herein by reference for all purposes.

[0084] In some embodiments, the BiSAb can include a binding domain that includes an scFv. Thus, in some embodiments, binding unit 2 includes an scFv. It will be appreciated that an scFv includes a polypeptide chain having a variable heavy chain domain (VH) linked to a variable light chain domain (VL) via a flexible polypeptide linker. FIGS. 1D-1F show schematic diagrams of exemplary BiSAbs where the BD (herein shown as binding unit 2) is an scFv having the domains described herein that can be represented by VL2 and VH2. In some embodiments, the polypeptide linker between VH2 and VL2 includes a protease cleavage site. The VH domain and VL domain of the scFv may be derived from the same antibody or from different antibodies. In some embodiments, the VH or VL of the scFv may include one or more CDRs that bind to the target of interest, and the remainder of the VH domain or VL domain is derived from a different antibody or is a synthetic domain. In some embodiments, the scFv includes at least one CDR of an antibody, such as an antibody known in the art that binds to the target of interest. In some embodiments, the scFv includes at least two CDRs of an antibody. In some embodiments, the scFv includes at least three CDRs of an antibody. In some embodiments, the scFv includes at least four CDRs of an antibody. In some embodiments, the scFv includes at least five CDRs of an antibody. In some embodiments, the scFv includes at least six CDRs of an antibody.

[0085] In some embodiments, the BD may comprise a ligand binding domain of a receptor or a receptor binding domain of a ligand. In some embodiments, the BD comprises a sequence having binding affinity for one or more epitopes on a target selected from the group consisting of CTLA-4, PD-1, PD-L1, OX40, and TIM3 as described above. In some embodiments, the binding domain exhibits specific binding activity for a target selected from the group consisting of CTLA-4, PD-1, PD-L1, OX40, and TIM3. The BiSAb disclosed herein may comprise any combination of binding domains having binding affinity or specific binding activity for the molecular targets disclosed herein. For example, the BiSAb disclosed herein may comprise CTLA-4 and PD-1; CTLA-4 and PD-L1; CTLA-4 and TIM3; PD-1 and PD-L1; PD-L1 and OX40; PD-1 and TIM3; PD-L1 and TIM3; and combinations of binding domains that enable bispecific binding to targets including TIM3. BiSAb comprising binding domains that bind to specific target combinations are exemplified in the examples and include, but are not limited to, combinations of PD-1 / CTLA-4; PD-L1 / CTLA-4; PD-1 / OX40; PD-L1 / OX40; and PD-1 / TIM3.

[0086] In some further embodiments, the BiSAb exhibits a binding activity (e.g., binding affinity and / or binding specificity) greater than the binding activity of the parental monospecific binding sequences used to generate the BiSAb, for at least one of the target molecules. In similar embodiments, the BiSAb may exhibit a binding activity (e.g., binding affinity and / or binding specificity) greater than the binding activities of both of the parental monospecific binding sequences used to generate the BiSAb, for both of the target molecules. In yet another embodiment, the BiSAb may exhibit a binding activity (e.g., binding affinity and / or binding specificity) greater than the binding activity of the combination of the parental monospecific binding sequences used to generate the BiSAb, for both of the target molecules. Whether alone or in combination, the enhancement of the binding properties of the BiSAb to the parental monospecific binding sequences provides unexpected advantages compared to the use of monospecific therapeutic agents targeting the same molecule, even when used in combination.

[0087] In some embodiments, the disclosure relates to an antibody or antigen-binding fragment thereof that binds to a target selected from the group consisting of CTLA-4, PD-1, PD-L1, OX40, and TIM3. In such embodiments, the antibody or antigen-binding fragment thereof may comprise a heavy chain sequence and a light chain sequence, or a portion of the heavy chain sequence and the light chain sequence that includes the CDR1, CDR2, and CDR3 sequences of the heavy chain and light chain sequences. In other embodiments, the antibody or antigen-binding fragment thereof may comprise a heavy chain variable (HCv) region sequence and a light chain variable (LCv) region sequence, or a portion of the HCv and LCv that includes the CDR1, CDR2, and CDR3 sequences of the heavy chain and light chain sequences. In yet another embodiment, the antibody or antigen-binding fragment thereof may comprise the CDR1, CDR2, and CDR3 sequences of the heavy chain and light chain sequences. In some embodiments, the antibody may be a chimeric, humanized, or human antibody. In some embodiments, the antibody may be a polyclonal or monoclonal antibody. In another embodiment, the antibody is a monoclonal antibody.

[0088] In some embodiments, bispecific binding proteins (BiSAb) disclosed herein can be generated using all or a domain comprising the antigen-binding region of the foregoing "parent" antibody. The non-limiting embodiments exemplified in the Examples provide an explanation of how antibody sequences can be identified and combined to generate BiSAb that exhibit bispecific binding to a combination of molecular targets.

[0089] One or more methods can be used alone or in combination to enhance the stability of BiSAb containing scFv molecules. One method that can be used alone or in combination with one or more of the other methods described herein is to manipulate the length and / or composition of the linker that connects the scFv domains so as to stabilize the scFv portion.

[0090] Another method that can be used is to introduce at least two amino acid substitutions (also referred to as modifications or mutations) into the VH domain and / or VL domain of the scFv so as to promote the formation of disulfide bonds (see, for example, Brinkmann et al., 1993, PNAS, 90:7538-42; Zhu et al., 1997, Prot. Sci. 6:781-8; Reiter et al., 1994, Biochem. 33:5451-9; Reiter et al., 1996, Nature 14:1239-45; Luo et al., 1995, J. Biochem. 118:825-31; Young et al., 1995, FEBS Let. 377:135-9; Glockshuber et al., 1990, Biochem. 29:1362-7). This method may be used alone or in combination with one or more of the other methods described herein.

[0091] In one aspect, one or more mutations can be introduced into the VH and VL domains of the scFv to promote the formation of inter-chain disulfide bonds between the VH and VL domains upon expression of the BiSAb containing the scFv. In another aspect, two mutations are introduced into the same domain of the chain. In one aspect, two mutations are introduced into different chains. In one aspect, multiple complementary mutations are introduced to promote the formation of multiple disulfide bonds or other stabilizing interactions. In one aspect, cysteine is introduced to promote the formation of disulfide bonds. Exemplary amino acids that may be mutated to cysteine include amino acids 43, 44, 45, 46, 47, 103, 104, 105, and 106 of VH2, and amino acids 42, 43, 44, 45, 46, 98, 99, 100, and 101 of VL2. The foregoing numbering is based on Kabat numbering that identifies positions (not with respect to the positions of the amino acids of the full-length sequence of the BiSAb or the sequence numbers provided herein) specific to only VH2 and VL2 of the scFv. Exemplary combinations of amino acid positions that may be mutated to cysteine residues include VH44-VL100, VH105-VL43, VH105-VL42, VH44-VL101, VH106-VL43, VH104-VL43, VH44-VL99, VH45-VL98, VH46-VL98, VH103-VL43, VH103-VL44, and VH103-VL45. In some aspects, amino acid 44 of VH and amino acid 100 of VL are mutated to cysteine.

[0092] Another method that may be used alone or in combination with one or more of the other methods described herein is to select the order of the scFv. In certain embodiments, the orientation of the VH domain relative to the VL domain is optimized for stability. In certain embodiments, the scFv is in the VH-linker-VL orientation. In certain embodiments, the scFv is in the VL-linker-VH orientation. In embodiments related to the novel BiSAb formats disclosed herein, the orientation of the domains within the scFv can determine how the scFv binds to the Fc portion of the BiSAb. This is described in more detail below in relation to the polypeptide linker. Briefly, however, since the BD (e.g., scFv) is interconnected with the CH2, CH3, or the interface of CH2 and CH3 by optional polypeptide linkers (L1) and (L2), the order of the domains determines which portion of the scFv is interconnected with L1 and which portion of the scFv is interconnected with L2.

[0093] A further method that may be used alone or in combination with other methods is to mutate one or more surface residues of the scFv to introduce one or more stabilizing mutations. In some embodiments, one, two, three, four, five, six or more than six residues of one or both of the VH domain and / or VL domain of the scFv are mutated. In one embodiment, only the VH domain of the scFv is modified. In one embodiment, only the VL domain of the scFv is modified. In one embodiment, both the VH domain and the VL domain of the scFv are modified. The same number of modifications may be made to each domain, or different numbers of modifications may be made to each domain. In one embodiment, one or more of the modifications are conservative amino acid substitutions of residues present in the unmodified parental scFv. In other embodiments, one or more of the modifications are non-conservative amino acid substitutions of residues present in the unmodified parental scFv. When multiple substitutions are made in one or both of the VH domain or VL domain of the scFv, the substitutions are each independently conservative or non-conservative substitutions. In one embodiment, the substitutions are all conservative substitutions. In one embodiment, the substitutions are all non-conservative. In one embodiment, at least one of the substitutions is conservative. In one embodiment, at least one or the substitutions is non-conservative.

[0094] Yet another method, which may be used alone or in combination with other methods, is to mutate one or more residues of the VH domain and / or VL domain of the scFv to introduce one or more amino acid substitutions to match the most frequent residues at said specific positions of the consensus sequence of the VH domain and / or VL domain of known antibodies. In certain embodiments, substitutions are introduced at one, two, three, four, five, six or more positions in one or both of the VH domain and / or VL domain of the scFv. The same number of changes may be made to each domain, or different numbers of changes may be made to each domain. In certain embodiments, one or more changes to match a certain consensus sequence are conservative amino acid substitutions of residues present in the unmodified VH sequence and / or VL sequence. In other embodiments, one or more changes are non-conservative amino acid substitutions of residues present in the unmodified VH sequence and / or VL sequence. When multiple substitutions are made in one or both of the VH domain or VL domain of the scFv, the substitutions are each independently conservative or non-conservative substitutions. In certain embodiments, the substitutions are all conservative substitutions. In certain embodiments, the substitutions are all non-conservative substitutions. In certain embodiments, at least one of the substitutions is conservative. In certain embodiments, at least one or more substitutions are non-conservative.

[0095] Note that any of the modifications described as being useful for modifying or stabilizing the scFv portion are applicable to modifying the Fab portion. For example, the variable domains of the Fab portion of a BiSAb can be modified to improve stability, antigen binding and isotypes. In addition, the Fab portion or the scFv portion can be modified to reduce immunogenicity.

[0096] In one aspect, the binding unit 2 (BD) is an scFv derived from a normal monoclonal antibody comprising a variable light chain (VL2) and a variable heavy chain (VH2) linked to each other by a flexible linker such as a glycine-serine linker. Optionally, the variable light chain and variable heavy chain of the scFv may be further linked to each other via one or more disulfide bonds and, as described above, may contain one or more mutations or modifications. The scFv binds to a second epitope. In one aspect, the second epitope is different from the first epitope to which binding unit 1 binds. In other aspects, the second epitope is the same as the first epitope to which binding unit 1 binds. In one aspect, the scFv is derived from or based on the sequence of a normal monoclonal antibody such as a normal mouse antibody, a humanized antibody or a human antibody. In one aspect, a BiSAb comprising an scFv derived from or based on the sequence of a normal monoclonal antibody retains one or more functional activities of a normal antibody (e.g., retains at least 80% or more (80%, 85%, 90%, 95%, 97%, 98%, 99% or 100%) of the functional activity). For example, in one aspect, a BiSAb comprising such an scFv retains one or more of the affinity for an antigen of a normal antibody, inhibitory activity or cell killing activity.

[0097] In one aspect, a BiSAb comprises any combination of binding unit 1 and binding unit 2, including any of binding unit 1 and binding unit 2 described herein. For example, in one aspect, the present disclosure provides a polypeptide comprising a Fab that binds to a specific target (e.g., binds to an epitope on a specific target), such as a Fab comprising a specific amino acid sequence or encoded by a specific nucleotide sequence, and / or an scFv that binds to a specific target (e.g., binds to an epitope on a specific target), such as an scFv comprising a specific amino acid sequence or encoded by a specific nucleotide sequence.

[0098] As detailed above, the binding units 1 and 2 may be bound to the BiSAb by covalent bonding via the linker polypeptides 1 (L1, L2). Generally, this binding may be through the mutual linkage via the heavy chain C H 2 domain, heavy chain C H 3 domain, or via the chimeric heavy chain of the BiSAb such that it is at the interface of the heavy chain C H 2 domain and C H 3 domain, or in some embodiments within the hinge region or Fab domain. L1 and L2 may independently differ in length and sequence from each other, and exemplary structures are described herein. The present disclosure contemplates BiSAbs including any combination of a binding unit and a linker polypeptide, such as any combination of a specific binding unit that binds to a desired target and the specific L1 and L2 polypeptide linkers described herein.

[0099] 2. Fc Region As used herein, "Fc region" encompasses domains derived from the constant region of an immunoglobulin, preferably a human immunoglobulin, including fragments, analogs, variants, mutants or derivatives of the constant region. Suitable immunoglobulins include IgG1, IgG2, IgG3, IgG4 and other classes, such as IgA, IgD, IgE and IgM. The Fc region may be a native sequence Fc region or a modified Fc region. The Fc region of an immunoglobulin generally includes two constant domains, C H 2 domain and C H 3 domain, and optionally includes C H 4 domain. The BiSAb of the present disclosure includes an Fc domain of the same class as the hinge portion of one or both of L1 or L2.

[0100] a. Modified Fc Region To alter the effector function and / or half-life of the bispecific antibodies (BiSAb) of the present disclosure, a modified Fc region (also referred to herein as a "mutated Fc region") may be used. In the Fc region, one or more modifications may be made to alter the functional and / or pharmacokinetic properties of the molecule. As a result of such modifications, the C1q binding and complement-dependent cytotoxicity (CDC) of IgG, or the FcγR binding and antibody-dependent cell-mediated cytotoxicity (ADCC), or antibody-dependent cell phagocytosis (ADCP) are decreased or enhanced. The present disclosure encompasses BiSAb that have been modified to fine-tune the effector function by enhancing or attenuating the function or providing a desired effector function. Thus, in one aspect of the present disclosure, the BiSAb comprises a mutated Fc region (i.e., an Fc region that has been modified as discussed below). The BiSAb comprising a mutated Fc region is also referred to herein as "Fc-mutated BiSAb". As used herein, "native" refers to the unmodified parental sequence, and herein the BiSAb comprising a native Fc region is referred to as "native Fc BiSAb". The Fc-mutated BiSAb can be produced by many methods well known to those skilled in the art. Non-limiting examples include isolation of the antibody coding region (e.g., from a hybridoma) and introduction of one or more desired substitutions in the Fc region. Alternatively, the antigen-binding portion (e.g., the variable region) of the BiSAb may be subcloned into a vector encoding the mutated Fc region. In one aspect, the mutated Fc region induces effector function at a level equivalent to that of the native Fc region. In another aspect, the mutated Fc region induces a higher level of effector function compared to native Fc. In another aspect, the mutated Fc region induces a lower level of effector function compared to native Fc. Some specific aspects of the mutated Fc region are detailed below. Methods for measuring effector function are well known in the art.

[0101] Generally, effector functions are modified by changes in the Fc region, including, but not limited to, amino acid substitutions, additions, deletions, and modifications in post-translational modifications of Fc amino acids (e.g., glycosylation). Using the methods described below, the effector functions of the BiSAb of the present disclosure, the ratio of the binding of the Fc region to FcR (e.g., affinity and specificity) can be finely tuned to obtain a BiSAb with desired properties.

[0102] It will be understood that the Fc region as used herein includes a polypeptide containing the constant region of the antibody molecule, excluding the immunoglobulin domain of the first constant region. For this reason, Fc refers to the last two constant region immunoglobulin domains of IgA, IgD and IgG, the last three constant region immunoglobulin domains of IgE and IgM, and optionally all or part of the N-terminus of the flexible hinge of these domains. In the case of IgA and IgM, Fc may include the J chain. In the case of IgG, Fc includes the immunoglobulin domains C gamma 2 and C gamma 3 (Cγ2 and Cγ3), and optionally the lower hinge portion between C gamma 1 (Cγ1) and C gamma 2 (Cγ2). The boundaries of the Fc region may vary, but as used herein, the Fc region of the human IgG heavy chain includes residues A231 to its carboxyl terminus, numbered according to the EU index described in Kabat. Fc may refer to this isolated region or to this region in relation to an antibody, antibody fragment or Fc fusion protein. Polymorphisms have been observed at many different Fc positions, including, but not limited to, positions 270, 272, 312, 315, 356 and 358 of IgG1 numbered by the EU index, and thus there may be some differences between this sequence and prior art sequences.

[0103] In one aspect, the present disclosure includes Fc mutant BiSAb with modified binding to Fc ligands (e.g., Fc receptors, C1q) compared to native Fc BiSAb. Examples of binding include binding specificity, equilibrium dissociation constant (K d) Dissociation rate and association rate (k off and k on ), binding affinity and / or avidity, among others, although not limited thereto. It is known in the art that the equilibrium dissociation constant (K d ) is defined as k off / k on . In some embodiments, a BiSAb comprising an Fc mutant region with a low K d may be more desirable than a BiSAb with a high K d . However, in some cases, the value of k on or k off may be more important than the value of K d . One of ordinary skill in the art can determine which kinetic parameters are most important for individual applications. For example, modifications that reduce binding to one or more positive regulators (e.g., FcγRIIIA) and / or enhance binding to inhibitory Fc receptors (e.g., FcγRIIB) are considered suitable for reducing ADCC activity. Thus, from the ratio of binding affinities to various receptors (e.g., the ratio of equilibrium dissociation constants (K d )) it can be indicated whether the ADCC activity of the Fc mutant BiSAb of the present disclosure is enhanced or reduced. In addition, modifications that reduce binding to C1q are also considered suitable for reducing or eliminating CDC activity.

[0104] In one aspect, the binding affinity for one or more Fc receptors, including but not limited to, FcRn, FcγRI (CD64) (including isoforms FcγRIA, FcγRIB, and FcγRIC); FcγRII (CD32, e.g., isoforms FcγRIIA, FcγRIIB, and FcγRIIC); and FcγRIII (CD16, e.g., isoforms FcγRIIIA and FcγRIIIB) is modified for the Fc mutant BiSAb as compared to the native Fc BiSAb.

[0105] In one aspect, the Fc mutant BiSAb has enhanced affinity for Fc ligands. In other aspects, the Fc mutant BiSAb has reduced affinity for Fc ligands compared to the native Fc BiSAb.

[0106] In certain aspects, the Fc mutant BiSAb has enhanced binding to the Fc receptor FcγRIIIA. In another certain aspect, the Fc mutant BiSAb has enhanced binding to the Fc receptor FcγRIIB. In a further certain aspect, the Fc mutant BiSAb has enhanced binding to both the Fc receptors FcγRIIIA and FcγRIIB. In one aspect, the Fc mutant BiSAb with enhanced binding to FcγRIIIA does not simultaneously enhance binding to the FcγRIIB receptor compared to the native Fc BiSAb. In certain aspects, the Fc mutant BiSAb has reduced binding to the Fc receptor FcγRIIIA. In a further certain aspect, the Fc mutant BiSAb has reduced binding to the Fc receptor FcγRIIB. In another certain aspect, the Fc mutant BiSAb has enhanced binding to the Fc receptor FcRn. In yet another certain aspect, the Fc mutant BiSAb with modified affinity for FcγRIIIA and / or FcγRIIB has enhanced binding to the Fc receptor FcRn. In yet another certain aspect, the Fc mutant BiSAb with modified affinity for FcγRIIIA and / or FcγRIIB has modified binding to C1q compared to the native Fc BiSAb.

[0107] In another aspect, the Fc mutant BiSAb increases or decreases affinity for C1q compared to the native Fc BiSAb. In yet another certain aspect, the Fc mutant BiSAb with modified affinity for C1q has enhanced binding to the Fc receptor FcRn. In yet another certain aspect, the Fc mutant BiSAb with modified affinity for C1q has modified binding to FcγRIIIA and / or FcγRIIB compared to the native Fc BiSAb.

[0108] It is recognized that antibodies can induce the attack and destruction of target antigens by a plurality of processes collectively referred to as antibody effector functions in the art. One of these processes is called "antibody-dependent cell-mediated cytotoxicity" or "ADCC", which refers to a form of cytotoxicity where secreted Ig binds to Fcγ receptors (FcγR) present on certain cytotoxic cells (e.g., natural killer (NK) cells, neutrophils, and macrophages), enabling these cytotoxic effector cells to specifically bind to target cells bearing the antigen and subsequently kill the target cells with cytotoxins. High-affinity IgG antibodies specific for the surface of the target cell "arm" the cytotoxic cells and are essential for such killing. Lysis of the target cell occurs extracellularly and requires direct cell-to-cell contact but not complement. Another process encompassed by the term effector function is complement-dependent cytotoxicity (hereinafter referred to as "CDC"), which refers to a biochemical phenomenon in which the complement system, via antibodies, destroys target cells. The complement system is a complex protein system found in normal plasma that, in combination with antibodies, destroys pathogens and other foreign cells. Yet another process encompassed by the term effector function is antibody-dependent cell phagocytosis (ADCP), which refers to a cellular reaction in which non-specific cytotoxic cells expressing one or more effector ligands recognize antibodies bound to target cells and subsequently cause phagocytosis of the target cells.

[0109] Fc mutant BiSAb is intended to be characterized by in vitro functional assays related to the function of effector cells via one or more FcγRs. In certain embodiments, Fc mutant BiSAb has binding and effector cell functions (such as those described and disclosed herein) similar to in vitro assays in an in vivo model. However, the present disclosure does not exclude Fc mutant BiSAb that does not exhibit the desired phenotype in in vitro assays but does exhibit the desired phenotype in vivo.

[0110] The serum half-life of a protein containing an Fc region can be extended by enhancing the binding affinity of the Fc region for FcRn. The term "antibody half-life," as used herein, refers to the pharmacokinetic property of an antibody that is an indicator of the average survival time of the antibody molecule after administration. Antibody half-life can be expressed, for example, as the time required to remove 50 percent of a known amount of immunoglobulin from the patient's body (or other mammal) or a particular compartment thereof, as seen in serum (i.e., plasma half-life) or in other tissues. The half-life may vary depending on the immunoglobulin or class of immunoglobulin. Generally, extending the antibody (or BiSAb) half-life results in an increase in the mean residence time (MRT) of the administered BiSAb in the circulating blood.

[0111] Extending the half-life can reduce the dosage of the agent administered to the patient and may also reduce the dosing frequency. To extend the serum half-life of a BiSAb, a salvage receptor binding epitope, such as that described in U.S. Patent No. 5,739,277, may be incorporated into the BiSAb (particularly an antibody fragment). As used herein, the term "salvage receptor binding epitope" refers to an epitope of the Fc region of an IgG molecule (e.g., IgG1, IgG2, IgG3, or IgG4) that is involved in the in vivo serum half-life of the IgG molecule. Alternatively, a BiSAb of the present disclosure with an extended half-life may be prepared by modifying amino acid residues that have been confirmed to be involved in the interaction between Fc and the FcRn receptor (see, e.g., U.S. Patent Nos. 6,821,505 and 7,083,784). Further, the half-life of a BiSAb of the present disclosure may be extended by conjugation to PEG or albumin by techniques widely utilized in the art.

[0112] It is contemplated that either insertion of another binding domain into the Fc region and / or subsequent binding by an antigen as described herein may affect Fc activity. For example, the binding antigen may increase or decrease the binding affinity and activity for FcgR, Clq, and FcRn. This may form an antigen-dependent switch for modulating various antibody-dependent processes. In one aspect, antigen binding may reduce the interaction with FcRn such that the free BiSAb interacts with FcRn to have a normal half-life, but enables rapid clearance / endocytosis of the BiSAb-Ag complex. Further, this may allow the BD2-antigen-mediated interaction to affect the clearance of the antigen bound to BD1. In another aspect, the BiSAb may comprise an Fc region directly inserted into BD2 (Fc-BD2).

[0113] In one aspect, the present disclosure provides an Fc variant comprising a modification (e.g., amino acid substitution, amino acid insertion, amino acid deletion) at one or more positions selected from the group consisting of 221, 225, 228, 234, 235, 236, 237, 238, 239, 240, 241, 243, 244, 245, 247, 250, 251, 252, 254, 255, 256, 257, 262, 263, 264, 265, 266, 267, 268, 269, 279, 280, 284, 292, 296, 297, 298, 299, 305, 308, 313, 316, 318, 320, 322, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 339, 341, 343, 370, 373, 378, 392, 416, 419, 421, 428, 433, 434, 435, 436, 440 and 443, where the Fc region is numbered according to the EU index described in Kabat. Optionally, the Fc region may include modifications at additional and / or alternative positions known to those of skill in the art (see, e.g., U.S. Patent Nos. 5,624,821; 6,277,375; 6,737,056; 7,083,784; 7,317,091; 7,217,797; 7,276,585; 7,355,008). Further, useful amino acid positions and specific substitutions are exemplified in Table 2 and Tables 6-10 of U.S. Patent No. 6,737,056; the tables shown in Figure 41 of U.S. Patent Application Publication No. 2006 / 024298; the tables shown in Figures 5, 12 and 15 of U.S. Patent Application Publication No. 2006 / 235208; the tables shown in Figures 8, 9 and 10 of U.S. Patent Application Publication No. 2006 / 0173170 and the tables shown in Figures 8-10, 13 and 14 of International Publication No. 09 / 058492.

[0114] In certain embodiments, the present disclosure provides that the Fc region is numbered 221K, 221Y, 225E, 225K, 225W, 228P, 234D, 234E, 234N, 234Q, 234T, 234H, 234Y, 234I, 234V, 234F, 235A, 235D, 235R, 235W, 235P, 235S, 235N, 235Q, 235T, 235H, 235Y, 235I, 235V, 235E, 235F, 236E, 237L, 237M, 237P, 239D, 239E, 239N, 239Q, 239F, 239T, 239H, 239Y, 240I, 240A, 240T, 240M, 241W, 241L, 241Y, 241E, 241R, 243W, 243L, 243Y, 243R, 243Q, 244H, 245A, 247L, 247V, 247G, 250E, 250Q, 251F, 252L, 252Y, 254S, 254T, 255L, 256E, 256F, 256M, 257C, 257M, 257N, 262I, 262A, 262T, 262E, 263I, 263A, 263T, 263M, 264L, 264I, 264W, 264T, 264R, 264F, 264M, 264Y, 264E, 265A, 265G, 265N, 265Q, 265Y, 265F, 265V, 265I, 265L, 265H, 265T, 266I, 266A, 266T, 266M, 267Q, 267L, 268E, 269H, 269Y, 269F, 269R, 270E, 280A, 284M, 292P, 292L, 296E, 296Q, 296D, 296N, 296S, 296T, 296L, 296I, 296H, 296G, 297S, 297D, 297E, 298A, 298H, 298I, 298T, 298F, 299I, 299L, 299A, 299S, 299V, 299H, 299F, 299E, 305I, 308F, 313F, 316D, 318A, 318S, 320A, 320S, 322A, 322S, 325Q, 325L, 325I, 325D, 325E, 325A, 325T, 325V, 325H, 326A, 326D, 326E, 326G, 326M, 326V, 327G, 327W, 327N, 327L, 328S, 328M, 328D, 328E, 328N, 328Q, 328F, 328I, 328V, 328T, 328H, 328A, 329F, 329H, 329Q,Provided is an Fc variant comprising at least one substitution selected from the group consisting of 330K, 330G, 330T, 330C, 330L, 330Y, 330V, 330I, 330F, 330R, 330H, 331G, 331A, 331L, 331M, 331F, 331W, 331K, 331Q, 331E, 331S, 331V, 331I, 331C, 331Y, 331H, 331R, 331N, 331D, 331T, 332D, 332S, 332W, 332F, 332E, 332N, 332Q, 332T, 332H, 332Y, 332A, 333A, 333D, 333G, 333Q, 333S, 333V, 334A, 334E, 334H, 334L, 334M, 334Q, 334V, 334Y, 339T, 370E, 370N, 378D, 392T, 396L, 416G, 419H, 421K, 428L, 428F, 433K, 433L, 434A, 424F, 434W, 434Y, 436H, 440Y and 443W. Optionally, the Fc region may include additional and / or alternative amino acid substitutions known to those skilled in the art, including, but not limited to, those exemplified in Table 2 and columns 6-10 of U.S. Patent No. 6,737,056, which is hereby incorporated by reference in its entirety; the tables shown in Figure 41 of U.S. Patent Application Publication No. 2006 / 024298; the tables shown in Figures 5, 12 and 15 of U.S. Patent Application Publication No. 2006 / 235208; the tables shown in Figures 8, 9 and 10 of U.S. Patent Application Publication No. 2006 / 0173170 and the tables shown in Figures 8, 9 and 10 of U.S. Patent Application Publication No. 20090041770, all of which are hereby incorporated by reference in their entirety.

[0115] In certain embodiments, the disclosure provides an Fc variant BiSAb comprising at least one modification (e.g., amino acid substitution, amino acid insertion, amino acid deletion) at one or more positions selected from the group consisting of 228, 234, 235 and 331 numbered according to the EU index described in Kabat. In one embodiment, the modification is at least one substitution selected from the group consisting of 228P, 234F, 235E, 235F, 235Y and 331S numbered according to the EU index described in Kabat.

[0116] In another specific embodiment, the present disclosure provides an Fc mutant BiSAb, wherein the Fc region is the Fc region of IgG4 and comprises at least one modification at one or more positions selected from the group consisting of 228 and 235 numbered by the EU index described in Kabat. In yet another specific embodiment, the Fc region is the Fc region of IgG4, and the non-natural amino acid is selected from the group consisting of 228P, 235E, and 235Y numbered by the EU index described in Kabat.

[0117] In another specific embodiment, the present disclosure provides an Fc mutant BiSAb, wherein the Fc region comprises at least one non-natural amino acid at one or more positions selected from the group consisting of 239, 330, and 332 numbered by the EU index described in Kabat. In one embodiment, the modification is at least one substitution selected from the group consisting of 239D, 330L, 330Y, and 332E numbered by the EU index described in Kabat. See U.S. Patent No. 7,317,091, which is incorporated herein by reference in its entirety.

[0118] In a specific embodiment, the present disclosure provides an Fc mutant BiSAb, wherein the Fc region comprises at least one non-natural amino acid at one or more positions selected from the group consisting of 252, 254, and 256 numbered by the EU index described in Kabat. In one embodiment, the modification is at least one substitution selected from the group consisting of 252Y, 254T, and 256E numbered by the EU index described in Kabat. See U.S. Patent No. 7,083,784, which is incorporated herein by reference in its entirety.

[0119] In one aspect, the present disclosure provides an Fc mutant BiSAb, wherein the Fc region contains a non-natural amino acid at position 428 numbered by the EU index described in Kabat. In one aspect, the modification at position 428 is selected from the group consisting of 428T, 428L, 428F, and 428S numbered by the EU index described in Kabat. See U.S. Patent No. 7,670,600, which is incorporated herein by reference in its entirety. In another aspect, the Fc mutant BiSAb may further contain a non-natural amino acid at position 434 numbered by the EU index described in Kabat. In one aspect, the modification at 434 is selected from the group consisting of 434A, 434S, and 434F numbered by the EU index described in Kabat. In other aspects, the present disclosure provides an Fc mutant BiSAb, wherein the Fc region contains non-natural amino acids at positions 428 and 434 numbered by the EU index described in Kabat. In certain aspects, the Fc region includes 428L and 434S. See U.S. Patent No. 8,088,376.

[0120] In one aspect, the effector functions induced by IgG antibodies strongly depend on the sugar moieties attached to the Fc region of the protein (Claudia Ferrara et al., 2006, Biotechnology and Bioengineering 93:851-861). Thus, glycosylation modification of the Fc region can be performed to enhance or reduce effector functions (e.g., Umana et al, 1999, Nat. Biotechnol 17:176-180; Davies et al., 2001, Biotechnol Bioeng 74:288-294; Shields et al, 2002, J Biol Chem 277:26733-26740; Shinkawa et al., 2003, J Biol Chem 278:3466-3473; U.S. Patent No. 6,602,684; No. 6,946,292; No. 7,064,191; No. 7,214,775; No. 7,393,683; No. 7,425,446; No. 7,504,256; POTELLIGENT (trademark) technology (Biowa, Inc. Princeton, N.J.); GLYCOMAB (trademark) glycosylation technology (GLYCART biotechnology AG, Zurich, Switzerland)). Thus, in one aspect, the Fc region of the BiSAb of the present disclosure includes altered glycosylation of amino acid residues. In another aspect, as a result of altering the glycosylation of amino acid residues, the effector function is reduced. In another aspect, as a result of altering the glycosylation of amino acid residues, the effector function is enhanced. In certain aspects, the Fc region suppresses fucosylation. In another aspect, the Fc region is defucosylated (see, e.g., U.S. Patent Application Publication No. 2005 / 0226867).In one aspect, such BiSabs with enhanced effector functions, particularly ADCC, are expressed in host cells (e.g., CHO cells, Lemna minor) engineered to produce highly fucosylated polypeptides having ADCC exceeding 100 - fold compared to polypeptides produced by parental cells (Mori et al., 2004, Biotechnol Bioeng 88:901 - 908; Cox et al., 2006, Nat Biotechnol., 24:1591 - 7).

[0121] Adding sialic acid to the oligosaccharides on IgG molecules can enhance their anti - inflammatory activity and alter their cytotoxicity (Keneko et al., Science, 2006, 313:670 - 673; Scallon et al., Mol. Immuno. 2007 Mar;44(7):1524 - 34). From the above - mentioned studies, it is demonstrated that IgG molecules with increased sialylation have anti - inflammatory properties, while IgG molecules with decreased sialylation enhance immunostimulatory activity (e.g., enhance ADCC activity). Therefore, BiSabs may be modified with a sialylation profile appropriate for a particular application (U.S. Patent Application Publication No. 2009 / 0004179 and International Publication No. 2007 / 005786 pamphlet).

[0122] In one aspect, the Fc region of the BiSabs of the present disclosure contains a modified sialylation profile compared to the native Fc region. In one aspect, the Fc region of the BiSabs of the present disclosure contains an increased sialylation profile compared to the native Fc region. In another aspect, the Fc region of the BiSabs of the present disclosure contains a decreased sialylation profile compared to the native Fc region.

[0123] In one aspect, the Fc variants of the present disclosure are other known Fc variants, such as Ghetie et al., 1997, Nat Biotech. 15:637-40; Duncan et al, 1988, Nature 332:563-564; Lund et al., 1991, J. Immunol 147:2657-2662; Lund et al, 1992, Mol Immunol 29:53-59; Alegre et al, 1994, Transplantation 57:1537-1543; Hutchins et al., 1995, Proc Natl. Acad Sci U S A 92:11980-11984; Jefferis et al, 1995, Immunol Lett. 44:111-117; Lund et al., 1995, Faseb J 9:115-119; Jefferis et al, 1996, Immunol Lett 54:101-104; Lund et al, 1996, J Immunol 157:4963-4969; Armour et al., 1999, Eur J Immunol 29:2613-2624; Idusogie et al, 2000, J Immunol 164:4178-4184; Reddy et al, 2000, J Immunol 164:1925-1933; Xu et al., 2000, Cell Immunol 200:16-26; Idusogie et al, 2001, J Immunol 166:2571-2575; Shields et al., 2001, J Biol Chem 276:6591-6604; Jefferis et al, 2002, Immunol Lett 82:57-65; Presta et al., 2002, Biochem Soc Trans 30:487-490); U.S. Patent No. 5,624,821; No. 5,885,573; No. 5,677,425; No. 6,165,745; No. 6,277,375; No. 5,869,046; No. 6,121,022; No. 5,624,821; No. 5,648,260; No. 6,528,624; No. 6,194,551; No. 6,737,056; No. 7,122,637; No. 7,183,387; No. 7,332,581; No. 7,335,742; No. 7,371,826; No. 6,821,505; No. 6,180,377; No. 7,317,091; No. 7,355,008, and may be combined with those disclosed therein. Other modifications and / or substitutions and / or additions and / or deletions of the Fc domain will also be readily apparent to those skilled in the art.

[0124] It should be noted that the polypeptides shown in the BiSAb format containing native Fc retain the ability to bind to FcRn and C1q and mediate ADCC, as shown in the examples. Thus, in certain embodiments, the BiSAb retains the ability to bind to FcRn and / or C1q and / or one or more Fcγ receptors (FcγR). For example, in certain embodiments, the BiSAb retains the ability to bind to FcRn and / or C1q and / or one or more FcγR by at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or 100% compared to a normal antibody that binds to one of the epitopes to which the BiSAb binds. In certain embodiments, the BiSAb is made from the binding domains of one or two normal antibodies, and the activity is compared with one or both of these normal antibodies.

[0125] Alternatively, a heavy chain heterodimer may be produced using a modified Fc region to obtain a BiSAb containing two different heavy chain-light chain pairs. To facilitate heterodimer formation, the interface between a pair of Fc regions is manipulated to maximize the proportion of heterodimers recovered from recombinant cell culture. In certain embodiments, the interface includes at least a portion of the CH3 domain. In this method, "protrusions" are created by substituting one or more small amino acid side chains at the interface of the first antibody molecule with larger side chains (e.g., tyrosine or tryptophan). By substituting larger amino acid side chains with smaller amino acid side chains (e.g., alanine or threonine), "cavities" of the same or similar size as the larger side chains are formed as compensation at the interface of the second antibody molecule. This becomes the mechanism for increasing the yield of heterodimers over other unwanted end products such as homodimers. Examples of CH3 modifications include, for example, Y407V / T366S / L368A in one heavy chain and T366W in the other heavy chain; S354C / T366W in one heavy chain and Y349C / Y407V / T366S / L368A in the other heavy chain. Other modifications that give a protrusion to one chain and a cavity to the other chain are described in U.S. Patent No. 7,183,076; U.S. Patent Application Publication No. 2014 / 0348839; and Merchant et al., 1998, Nat. Biotech 16:677-681. Some non-limiting examples of modifications that can result in a protrusion-cavity configuration are shown in Table 1a. Other modifications that can be used to form heterodimers include, but are not limited to, modifications that change the charge polarity at both ends of the interface of the Fc dimer such that heterodimerization occurs when electrostatically complementary Fc regions are co-expressed. Modifications that change the charge polarity include, but are not limited to, those shown in Table 1b below (see also International Publication No. 20090182127 pamphlet; Gunasekaran et al., 2010, JBC 285:19637-46).Furthermore, Davis et al. (2010, Prot. Eng. Design & Selection 23:195-202) described a heterodimeric Fc platform using a chain-exchanged engineered domain (SEED) CH3 region that is a derivative of the human IgG domain and the IgA CH3 domain (see also International Publication No. WO 2007 / 110205 pamphlet).

[0126]

Table 1

[0127]

Table 2

[0128] One of ordinary skill in the art will appreciate that in some embodiments, the Fc fusion protein may form dimers due to the homodimeric nature of the molecule containing the Fc region. In some embodiments, mutations (e.g., chimeric mutations, complementary mutations, dock-and-lock mutations, knob-into-hole mutations, strand exchange operation domain (SEED) mutations, etc.) may be performed on the Fc region of the binding protein (e.g., BiSAb) to promote and / or maintain heterodimerization. See, for example, U.S. Patent No. 7,183,076; Merchant et al. (1998) Nat. Biotech 16:677-681; Ridgway et al. (1996) Protein Engineering 9:617-621; Davis et al. (2010) Prot. Eng. Design & Selection 23:195-202; International Publication No. 2007 / 110205 pamphlet; International Publication No. 2007 / 147901 pamphlet; Gunasekaran et al. (2010) JBC 285:19637-46. All of these are hereby incorporated by reference herein. Thus, the binding protein can be engineered to form heterodimers comprising, for example, a first binding protein, binding domain, or BiSAb fused to a first Fc region or fragment thereof, and a second (i.e., different) binding protein, binding domain, or BiSAb fused to a second Fc region or fragment thereof, wherein the first and second Fc regions, or fragments thereof, are engineered to heterodimerize.

[0129] 3. Glycosylation Glycosylation can not only change the effector function of polypeptides, but also change the affinity of antibodies (or BiSAb) for target antigens through glycosylation modification of variable regions. In one aspect, the glycosylation pattern of the variable region of the present BiSAb is modified. For example, non-glycosylated BiSAb may be prepared (i.e., BiSAb lacks glycosylation). Glycosylation can be modified, for example, to enhance the affinity of BiSAb for the target antigen. Such sugar chain modifications can be achieved, for example, by changing one or more glycosylation sites within the BiSAb sequence. For example, one or more amino acid substitutions may be made to remove the glycosylation sites in one or more variable region frameworks, thereby removing the glycosylation at that site. Such non-glycosylation can enhance the affinity of BiSAb for the antigen. Such approaches are described in more detail in U.S. Patent Nos. 5,714,350 and 6,350,861. Also, one or more amino acid substitutions may be made to remove the glycosylation sites present in the Fc region (e.g., asparagine 297 of IgG). Furthermore, non-glycosylated BiSAb may be prepared using bacterial cells lacking the necessary glycosylation machinery.

[0130] 4. Polypeptide linker The linker can be used to link the domains / regions of the BiSAb chimeric heavy chain to adjacent molecules. As described herein, the BiSAb can include one, two, or more linker polypeptides (e.g., L1 and L2). In addition, the BiSAb may include another linker, such as a flexible linker that links the variable heavy and light chains of the scFv to each other. Furthermore, the BiSAb may include another linker, such as a flexible linker that links the variable heavy and light chains of the scFv to each other, and another linker that links other binding units to the BiSAb core structure.

[0131] As an illustrative and non-limiting example of a linker, there is a polypeptide chain comprising at least 4 residues. Some of such linkers are flexible, hydrophilic, and may have little or no secondary structure of their own (linker portion or flexible linker portion). Linkers of at least 4 amino acids may be used to interconnect domains and / or regions located in the vicinity after the molecules have assembled. Also, longer or shorter linkers may be used. For this reason, the linker may be about 1 residue, 2 residues, 3 residues, 4 residues, 5 residues, 6 residues, 7 residues, 8 residues, 9 residues, 10 residues, 11 residues, 12 residues, 13 residues, 14 residues, 15 residues, 16 residues, 17 residues, 18 residues, 19 residues, 20 residues, 25 residues, 30 residues, 35 residues, 40 residues, 45 residues, or about 50 residues in length. When multiple linkers are used to interconnect parts of a molecule, the linkers may be the same or different (e.g., having the same or different lengths and / or amino acid sequences).

[0132] The linker may be a cleavable linker that includes at least one bond that is selectively cleaved by a cleavage reagent. Using a cleavable linker may facilitate removal of all or part of the linker sequence. The linker may be engineered to include a protease cleavage site such that cleavage occurs within the linker or at at least one end of the linker. For example, thrombin sites may be engineered at each of the two ends on both sides of the linker. Also, depending on the type of linker used, cleavage may be mediated by agents such as TCEP, TFA, and DTT. The linker may be designed such that the cleavage reagent removes all residues derived from the linker from the cleavage product. Other exemplary and non-limiting linkers include prodrug linkers that can be selectively cleaved in vivo, for example, in the presence of endogenous enzymes or other endogenous factors, or simply in the aqueous fluids present in the body or within cells in the body. When the BiSAb includes two or more polypeptide linkers, the linkers may each be different, or at least one of the linkers may be different from the others. In some embodiments, the BiSAb includes a cleavable linker. In certain embodiments, the BiSAb includes a scFv that includes a cleavable linker between VH2 and VL2.

[0133] The linker facilitates formation of the desired structure. The linker contains (Gly-Ser) n residues and, to enhance solubility, Glu residues or Lys residues may be somewhat evenly dispersed throughout. Alternatively or in addition, the linker contains no serine residues, and such linkers may be preferred when the linker is subject to O-linked glycosylation. In some embodiments, for example, when dimerization of the linker is used to bring the domains of the BiSAb into a properly folded structure, the linker may contain cysteine residues. In some embodiments, the BiSAb includes at least two polypeptide linkers that link the domains of the polypeptide. In other embodiments, the BiSAb includes at least three polypeptide linkers. In other embodiments, the BiSAb includes four or more polypeptide linkers.

[0134] In some embodiments, the polypeptide linker comprises a portion of the Fc portion. For example, in some embodiments, the polypeptide linker may comprise a portion of the immunoglobulin hinge domain of an IgG1 antibody, an IgG2 antibody, an IgG3 antibody, and / or an IgG4 antibody. In some embodiments, the polypeptide linker comprises a portion of a mutant immunoglobulin hinge domain of IgG1, IgG2, IgG3, and / or IgG4. In some embodiments, the polypeptide linker comprises at least 5 amino acid residues, 7 amino acid residues, 8 amino acid residues, or 15 amino acid residues of the immunoglobulin hinge region / domain of an IgG1 antibody, an IgG2 antibody, an IgG3 antibody, and / or an IgG4 antibody. In some embodiments, the polypeptide linker comprises at least 5 amino acid residues, 7 amino acid residues, 8 amino acid residues, or 15 amino acid residues of a modified immunoglobulin hinge region / domain of an IgG1 antibody, an IgG2 antibody, an IgG3 antibody, and / or an IgG4 antibody.

[0135] The polypeptide linker may comprise all or a portion of a hinge region that naturally contains three cysteines. In one embodiment, the selected hinge region is cleaved, or otherwise modified or substituted, such that only one or two of the cysteine residues remain relative to the complete and / or native hinge region. Similarly, in certain other embodiments, the polypeptide linker may comprise a mutated portion or otherwise modified portion of a hinge region in which the number of cysteine residues has been reduced by amino acid substitution or deletion, for example, the mutated hinge region or otherwise modified hinge region comprises 0, 1, or 2 cysteine residues as described herein.

[0136] Accordingly, a mutated hinge domain or a hinge domain modified by other means may be derived from or constructed using a wild-type immunoglobulin hinge domain containing one or more cysteine residues. In certain embodiments, the mutated portion or the portion modified by other means of the hinge region may contain no cysteine residues or may contain only one cysteine residue, and the mutated hinge region or the hinge region modified by other means is a wild-type immunoglobulin hinge region containing one or more cysteine residues or two or more cysteine residues, respectively, or is derived therefrom. In the mutated portion or the portion modified by other means of the hinge region, the cysteine residues of the wild-type immunoglobulin hinge region are preferably deleted or substituted with amino acids that cannot form disulfide bonds. In some embodiments, the mutated portion or the portion modified by other means of the hinge region is a human IgG wild-type hinge region that may contain any of the four human IgG isotype subclasses IgG1, IgG2, IgG3, or IgG4, or is derived therefrom.

[0137] In some embodiments, the polypeptide linker comprises a portion of the hinge region that contains a cysteine residue (EU residue 220) that forms a disulfide bond with the immunoglobulin light chain. In some embodiments, the polypeptide linker comprises a modified portion of the hinge region that contains an amino acid substitution at EU residue C220. In some embodiments, the polypeptide linker comprises the amino acid substitution C220V.

[0138] In some embodiments, the polypeptide linker comprises an amino acid substitution that prevents hinge-related spontaneous self-cleavage. In some embodiments, the polypeptide linker comprises an amino acid substitution at the position of EU residue D221. In some embodiments, the polypeptide linker comprises the amino acid substitution D221G. In some embodiments, the polypeptide linker lacks amino acid D221.

[0139] As described above, some embodiments include one or more polypeptide linkers that include, or consist of, a gly-ser linker. As used herein, the term "gly-ser linker" refers to a peptide consisting of glycine and serine residues. Exemplary gly-ser linkers include an amino acid sequence of the formula (Gly4Ser)n, where n is a positive integer (e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10). Some preferred non-limiting examples of gly-ser linkers include (Gly4Ser)2 (SEQ ID NO: 41) and (Gly4Ser)4 (SEQ ID NO: 42), as well as (Gly4Ser)3 (SEQ ID NO: 43). In still other embodiments, two or more gly-ser linkers are incorporated in series into the polypeptide linker. In some embodiments, the polypeptide linker includes at least a portion of a hinge region (e.g., derived from an IgG1 molecule, an IgG2 molecule, an IgG3 molecule, or an IgG4 molecule), and a series of gly-ser amino acid residues, such as a gly-ser linker such as (Gly4Ser)n, where n is 2, 3, or 4).

[0140] In some embodiments, a linker (e.g., such as L1 and / or L2 and / or L3) includes both a hinge portion and a linker portion, e.g., a linker portion including a gly-ser linker. In other embodiments, L1 and / or L2 includes only a hinge portion or only a linker portion, e.g., only a gly-ser linker. In other embodiments, L1 and L2 include a gly-ser linker portion. In some embodiments, the gly-ser linkers within the BiSAb are of the same length, whereas in other embodiments, the gly-ser linker portions within the BiSAb (e.g., L1 and L2) are of different lengths. When the BiSAb includes an scFv, the heavy and light chains of the scFv may be linked to the BiSAb (e.g., BD1, Fab, Fc, etc.) by a flexible linker. This flexible linker generally does not include a hinge portion, but rather is a gly-ser linker or other flexible linker. The length and amino acid sequence of the flexible linker that links the domains of the scFv to each other can be readily selected and optimized (e.g., (Gly4Ser)n, (SEQ ID NO: 48), wherein n is 2, 3, or 4 or more).

[0141] Regardless of the various binding units and domains (e.g., binding domain / units (e.g., Fab-scFv), or the polypeptide linker used to interconnect the binding domain / units to the Fc (e.g., scFv via L1 and L2)), the BiSAb may optionally include an additional polypeptide linker. The length and sequence of such additional polypeptide linker are independently selected. For example, the BiSAb may further include a flexible polypeptide linker that interconnects the variable heavy and light chains of the scFv. This flexible polypeptide linker may include a gly-ser linker. Generally, this linker does not include a hinge portion.

[0142] Here, it is considered that changes in the length of the linker that franks BD2 can affect the orientation of the BD2 antigen-binding site and the spacing relative to the remaining BiSAb molecule. For example, a short N-terminal linker and a long C-terminal linker can generate an orientation in which the binding site is conformationally arranged in one direction, while a long N-terminal linker and a short C-terminal linker can affect the opposite conformational orientation. Thus, the length of the linker can be modulated to orient the BD2 antigen-binding site and can have an important impact on the generation or avoidance of steric interactions between BD1 and BD2 and / or between BD2 and other entities that bind to antibody molecules within the Fc or other domains.

[0143] 5. Specific Structure of BiSAb As described above, one aspect of the present disclosure relates to a BiSAb structural arrangement (platform) comprising two heavy chain-light chain pairs (shown in FIGS. 1A-1F). In some embodiments of this aspect, the polypeptide sequence of the BiSAb chimeric heavy chain may comprise a polypeptide sequence comprising an antibody heavy chain variable domain (VH1), a polypeptide sequence comprising an antibody heavy chain constant domain 1 (CH1), a portion of the Fc domain, a polypeptide sequence comprising a first polypeptide linker (L1), a polypeptide sequence comprising a binding domain (BD2), a polypeptide sequence comprising a second polypeptide linker (L2), and a polypeptide sequence comprising the remaining Fc domain. In some aspects, the Fc domain comprises the C H 2 domain and the C H 3 domain. Thus, one embodiment has the following orientation from the N-terminus to the C-terminus: VH1-C H 1-C H 2 (N-terminus)-L1-BD2-L2-C H 2 (C-terminus)-C H 3; VH1-C H 1-C H 2-L1-BD2-L2-C H 3; and VH1-C H 1-C H 2-C H 3 (N-terminus)-L1-BD2-L2-C HProvided is a BiSAb chimeric heavy chain that may contain a polypeptide sequence at the 3 (C-terminus). The polypeptide sequence of the BiSAb light chain may include a light chain variable domain (VL1) and a light chain constant domain (CL). Thus, the BiSAb light chain may contain a polypeptide sequence with the following orientation from the N-terminus to the C-terminus: VL1-CL. Note that VH1, VL1, and CL are used to represent parts of "Binding Unit 1" (BD1) that binds to the first epitope. BD2 is used to represent parts of "Binding Unit 2" that binds to the second epitope.

[0144] In an embodiment where the binding domain is an scFv, the BiSAb chimeric heavy chain may contain a polypeptide sequence including an antibody heavy chain variable domain (VH1), a polypeptide sequence including an antibody heavy chain constant domain 1 (CH1), a polypeptide sequence including a first polypeptide linker (L1), a polypeptide sequence including an antibody light chain variable domain (VL2), a polypeptide sequence including a flexible linker, a polypeptide sequence including an antibody heavy chain variable domain (VH2), a polypeptide sequence including a second polypeptide linker (L2), and a polypeptide sequence including an antibody Fc domain. Therefore, the chimeric heavy chain of BiSAb containing scFv as BD2 may have the following orientation from the N-terminus to the C-terminus: VH1-CH1-CH2 (N-terminus)-L1-VL2-L3-VH2-L2-CH2 (C-terminus)-CH3; VH1-CH1-CH2-L1-VL2-L3-VH2-L2-CH3; VH1-CH1-CH2-CH3 (N-terminus)-L1-VL2-L3-VH2-L2-CH3 (C-terminus); VH1-CH1-CH2 (N-terminus)-L1-VH2-L3-VL2-L2-CH2 (C-terminus)-CH3; VH1-CH1-CH2-L1-VH2-L3-VL2-L2-CH3; and VH1-CH1-CH2-CH3 (N-terminus)-L1-VH2-L3-VL2-L2-CH3 (C-terminus) and may contain a polypeptide sequence.

[0145] The chimeric heavy chain is a polypeptide chain comprising an amino acid sequence (e.g., the amino acid sequence of each polypeptide domain). The chimeric heavy chain is a polypeptide chain comprising an amino acid sequence (e.g., the amino acid sequence of each polypeptide domain). Note that VH1, VL1, and CL are used to indicate a part of binding unit 1, and VH1 and VL1 indicate the part that binds to its first epitope. VH2 and VL2 are used to indicate a part of binding unit 2 that binds to the second epitope. In certain embodiments, additional scFv binding domains are present at the N-terminus and / or C-terminus of the polypeptide constituting the BiSAb core (where BiSAb further comprises binding units (BD) 3 and / or 4 and / or 5). In certain embodiments, two or more scFv binding domains are present within the BiSAb core. The additional scFv comprises an antibody heavy chain variable region represented by VH3, VH4, VH5, respectively, and a corresponding antibody light chain variable region represented by VL3, VL4, VL5.

[0146] 6. Labels, Conjugates, and Moieties In one feature, drugs and other molecules may target the BiSAb by site-specific conjugation. For example, the BiSAb may include a cysteine engineering domain (including cysteine introduced into the binding unit and / or Fc domain) from which a free thiol group for the conjugation reaction can be obtained. In one aspect, the BiSAb may be engineered to include specific conjugation sites. In some aspects, the present disclosure provides Fc mutant BiSAb in which the Fc region includes amino acid substitutions at one or more of positions 239, 282, 289, 297, 312, 324, 330, 335, 337, 339, 356, 359, 361, 383, 384, 398, 400, 440, 422, and 442 numbered by the EU index. In some aspects, the Fc region includes substitutions in one or more of the following groups: a) positions 289 and 440; b) positions 330 and 440; c) positions 339 and 440; d) positions 359 and 440; e) positions 289 and 359; f) positions 330 and 359; g) positions 339 and 359; h) positions 289 and 339; i) positions 330 and 339; j) positions 289 and 330; k) positions 339 and 442; l) positions 289, 339, and 442; m) positions 289, 330, and 339; n) positions 330, 339, and 442; and o) positions 289, 330, and 442. In other aspects, the present disclosure provides BiSAb in which the CH1 domain of the Fab arm includes substitutions at one or more of positions 131, 132, 134, 135, 136, and 139 numbered by the EU index. In one aspect, the substitutions include substitutions of amino acids selected from cysteine, lysine, tyrosine, histidine, selenocysteine, and selenomethionine. In certain aspects, the substitution is cysteine. Methods for making stable cysteine-engineered antibodies are described in U.S. Patent No. 7,855,275, U.S. Patent Application Publication No. 20110033378, and U.S. Patent Application Publication No. 20120213705, the entire contents of which are incorporated herein by reference.

[0147] 7. Exemplary Targets Aspects and embodiments relating to the various DuetMab and BiSAb platforms described herein can be made to bind to any one or more desired targets, but the BiSAb disclosed herein preferably targets a specific pair of target molecules (e.g., binding unit 1 binds to one of the targets and binding unit 2 binds to the other target). As described above and as illustrated in the following exemplary embodiments, the antibodies, DuetMab, and BiSAb disclosed herein target molecules that modulate the immune response in a recipient subject or immune cells in culture. In some embodiments, the binding domain exhibits specific binding activity for a target selected from the group consisting of CTLA-4, PD-1, PD-L1, OX40, and TIM3. DuetMab and BiSAb can include combinations of different binding domains in various orders and orientations, and these domains have a binding affinity for the targets disclosed herein or specifically bind to them. For example, the DuetMab and BiSAb disclosed herein can include combinations of binding domains that enable bispecific binding to targets such as CTLA-4 and PD-1; CTLA-4 and PD-L1; and CTLA-4; CTLA-4 and TIM3; PD-1 and PD-L1; PD-L1 and OX40; PD-1 and TIM3; PD-L1 and TIM3. DuetMab and BiSAb containing binding domains that bind to specific target combinations are exemplified in the examples and include non-limiting combinations of PD-1 / CTLA-4; PD-L1 / CTLA-4; PD-1 / TIM3; and PD-L1 / OX40. In one embodiment, the BiSAb has enhanced binding properties compared to the binding properties of the combination of individual monospecific binding proteins used to generate the BiSAb.

[0148] In one aspect, the DuetMab or BiSAb of the present disclosure binds to two different epitopes on the same target (e.g., binding unit 1 binds to a first epitope on the target and binding unit 2 binds to a second epitope on the same target).

[0149] In some embodiments, due to the multimeric nature of the DuetMab or BiSAb of the present disclosure, a labeling or therapeutic agent can target a specific cell type or molecular target. For example, one functional domain of the DuetMab or BiSAb can bind to a cell surface target while another functional domain of the same DuetMab or BiSAb can bind to a hapten or labeling agent useful for detection. Similarly, one functional domain can bind to a cell target while a second functional domain can bind to a toxin. Since both binding reactions are mediated by a single molecule, the toxin can be positioned in the vicinity of the cell target where the cytotoxic function affects the cell.

[0150] B. Nucleic acid molecules encoding BiSAb The present disclosure provides nucleic acid molecules encoding BiSAb. One aspect of the present disclosure provides nucleic acid molecules encoding any of the BiSAb of the present disclosure. The nucleic acid molecule can encode any of the BiSAb molecules disclosed herein, as well as the heavy and / or light chains of any of the individual binding domains (e.g., scFv) disclosed herein. Those skilled in the art will recognize that the nucleotide sequence of such polynucleotide molecules can vary, taking into account nucleic acid codon degeneracy and codon frequency for a particular organism, as is well known in the art.

[0151] C. Vectors and host cells for producing BiSAb and subsequent purification The present disclosure relates to a method for producing a BiSAb. In certain aspects, a recombinant nucleic acid molecule encoding all or a portion of the BiSAb disclosed herein may be operably linked to one or more regulatory nucleotide sequences of an expression construct. Nucleic acid sequences encoding the light chain and the chimeric heavy chain of the BiSAb may be cloned into the same expression vector in any orientation (e.g., the light chain before or vice versa of the heavy chain), or may be cloned into two different vectors. When expression is performed using one vector, the two coding genes may have their own gene elements (e.g., promoter, RBS, leader, stop, polyA, etc.), or may be cloned with a set of gene elements but ligated to a cis - tron element. The regulatory nucleotide sequences are generally made appropriate for the host cell used for expression. Many types of expression vectors and suitable regulatory sequences appropriate for various host cells are known in the art. Typically, the one or more regulatory nucleotide sequences may include, but are not limited to, a promoter sequence, a leader or signal sequence, a ribosome binding site, transcription start and termination sequences, translation start and termination sequences, and enhancer or activator sequences. The present disclosure contemplates a constitutive promoter or an inducible promoter known in the art. The promoter may be a native promoter or a hybrid promoter combining elements of two or more promoters. The expression construct may be present on a cell episome, such as a plasmid, or the expression construct may be inserted into the chromosome.

[0152] In one aspect, the expression vector includes a selectable marker gene that enables the selection of transformed host cells. Selectable marker genes are well known in the art and vary depending on the host cells used. In one aspect, the present disclosure relates to an expression vector comprising a nucleotide sequence encoding a polypeptide and operably linked to at least one regulatory sequence. Regulatory sequences are known in the art and are selected to induce the expression of the encoded polypeptide. Thus, the term regulatory sequence includes promoters, enhancers and other expression control elements. Exemplary and non-limiting regulatory sequences are described in Goeddel; Gene Expression Technology: Methods in Enzymology, Academic Press, San Diego, CA (1990). It should be understood that the design of the expression vector may vary depending on factors such as the choice of host cell to be transformed and / or the type of protein for which expression is desired. Furthermore, the copy number of a particular vector, the ability to control the copy number, and the expression of any other proteins encoded by the vector, such as antibiotic markers, should also be considered.

[0153] The present disclosure further relates to methods for making the BiSAb of the present disclosure. For example, one or more expression vectors encoding the BiSAb (e.g., a single vector encoding a chimeric heavy and light chain, or two vectors, a vector encoding a chimeric heavy chain and a vector encoding a light chain) may be transfected into host cells and cultured under appropriate conditions to allow for expression of the polypeptide. The BiSAb may be secreted and isolated from a mixture of cells and medium containing the polypeptide. Alternatively, the BiSAb may be retained in the cytoplasmic or membrane fraction, and the cells may be harvested, lysed, and the protein isolated. Cell cultures contain host cells, medium, and other by-products. Media suitable for cell culture are well known in the art. To purify the protein, the BiSAb may be isolated from the cell culture medium, host cells, or both, using techniques known in the art such as ion exchange chromatography, gel filtration chromatography, ultrafiltration, electrophoresis, and immunoaffinity purification. In certain embodiments, the BiSAb is made as a fusion protein that includes a domain that facilitates its purification.

[0154] A cloned gene or a part thereof may be ligated into a vector suitable for expression in prokaryotic cells, eukaryotic cells (yeast, birds, insects or mammals) or both to produce a recombinant nucleic acid. Expression media for producing recombinant polypeptides include plasmids and other vectors. For example, suitable vectors include the following types of plasmids for expression in prokaryotic cells, such as E. coli: plasmids derived from pBR322, plasmids derived from pEMBL, plasmids derived from pEX, plasmids derived from pBTac, and plasmids derived from pUC. In one embodiment, a mammalian expression vector includes both prokaryotic sequences that facilitate propagation of the vector in bacteria and one or more eukaryotic transcription units for expression in eukaryotic cells. Vectors derived from pcDNAI / amp, pcDNAI / neo, pRc / CMV, pSV2gpt, pSV2neo, pSV2-dhfr, pTk2, pRSVneo, pMSG, pSVT7, pko-neo, and pHyg are examples of mammalian expression vectors suitable for transfection of eukaryotic cells. Some of these vectors are modified with bacterial plasmid sequences, such as those derived from pBR322, to facilitate replication and drug resistance selection in both prokaryotic and eukaryotic cells. Alternatively, for transient expression of proteins in eukaryotic cells, viruses such as derivatives of bovine papillomavirus (BPV-1) or Epstein-Barr virus (pHEBo, pREP-derived, and p205) may be used. Various methods used for plasmid preparation and transformation of host organisms are known in the art. In addition to other expression systems suitable for both prokaryotic and eukaryotic cells, for general recombinant procedures, see Molecular Cloning A Laboratory Manual, 2nd Ed., ed. by Sambrook, Fritsch and Maniatis (Cold Spring Harbor Laboratory Press, 1989) Chapters 16 and 17. Optionally, it may be desirable to express the recombinant polypeptide by using a baculovirus expression system.Examples of such baculovirus expression systems include vectors derived from pVL (such as pVL1392, pVL1393, and pVL941), vectors derived from pAcUW (such as pAcUW1), and vectors derived from pBlueBac (pBlueBac III containing β-gal).

[0155] Once the molecule is produced, it may be purified by any method known in the art for purifying proteins, immunoglobulin molecules, or other multimeric molecules, such as chromatography (e.g., ion exchange chromatography, affinity chromatography (particularly by affinity for specific antigen proteins A or G), and size column chromatography), centrifugation, solubility differences, or any other multimeric molecule technique for protein purification. Further, the molecules disclosed herein may be fused with heterologous polypeptide sequences (e.g., affinity tags) routinely used to facilitate purification.

[0156] Regardless of how the BiSAb is made and purified, a binding assay, such as a dual ELISA assay, may be performed (before and / or after purification) to confirm the functional binding activity of the BiSAb. Such binding assays are generally well known in the art.

[0157] D. Pharmaceutical Formulations In one aspect, the disclosure provides a pharmaceutical composition. Such a pharmaceutical composition may be a composition comprising a nucleic acid molecule encoding a BiSAb. Also, such a pharmaceutical composition may be a composition comprising a DuetMab, a BiSAb, a combination of DuetMabs, or a combination of BiSAbs and a pharmaceutically acceptable excipient. In one aspect, the pharmaceutical composition of the disclosure is used as a drug.

[0158] E. Use As described herein, DuetMab and BiSAb can be used to bind to targets associated with cancer and cell proliferative diseases or disorders, which diseases or disorders can be responsive to immunotherapy, for example, by inhibiting immunosuppressive activity and / or inducing an immune response related to a target molecule. For example, abnormal signal transduction and / or an inhibited immune response can contribute to unwanted cell proliferation and cancer. Thus, the DuetMab, BiSAb, and antibodies disclosed herein can be used to treat unwanted cell proliferation and / or cancer that targets, is associated with, or results from an inhibited, reduced, or insufficient immune response. In particular, the tumor growth curve of a tumor and / or the volume of a tumor can be decreased by administration of a DuetMab or BiSAb that induces and / or stimulates an immune response in a subject, such as a human patient afflicted with cancer.

[0159] Accordingly, the present disclosure also relates to various methods including the administration of the binding proteins disclosed herein to a subject in need thereof. In one aspect, the present disclosure relates to a method of inducing an immune response in a subject having or at risk of developing cancer, which method includes administering to the subject a binding protein disclosed herein. In some embodiments, the method activates an immune response against cancer in the subject. In some embodiments, the method enhances an immune response against cancer in the subject. In some embodiments, the method activates an inhibited immune response pathway in the subject, and this activation increases an immune response targeting cancer in the subject. In some embodiments, the method enhances an immune response pathway targeting cancer in the subject.

[0160] In another aspect, the present disclosure relates to a method of treating cancer in a subject in need thereof, which comprises administering to the subject a binding protein disclosed herein. In one embodiment, the method of treating cancer comprises stopping or slowing the growth of the cancer in the subject. In one embodiment, the method of treating cancer comprises stopping or slowing the metastasis of the cancer to other sites in the subject. In one embodiment, the method of treating cancer comprises killing cancer cells in the subject. In one embodiment, the method of treating cancer comprises stopping the proliferation and / or spread of cancer cells in the subject.

[0161] In various embodiments of the foregoing aspects, the method relates to treating a patient for a tumor disease and / or a cancer disease. In multiple embodiments, the cancer is selected from the group of cancers that are sensitive to an induced immune response in the subject. In some embodiments, the cancer is one or more of ovarian cancer, breast cancer, colorectal cancer, prostate cancer, cervical cancer, uterine cancer, testicular cancer, bladder cancer, head and neck cancer, melanoma, pancreatic cancer, renal cell cancer, or lung cancer. In some embodiments, the cancer is a gastrointestinal or gastrointestinal cancer (e.g., anal cancer; cholangiocarcinoma; extrahepatic bile duct cancer; appendiceal cancer; carcinoid tumor, colon cancer; colorectal cancer including pediatric colorectal cancer; esophageal cancer including pediatric esophageal cancer; bladder cancer; gastric (stomach) cancer including pediatric gastric cancer; hepatocellular carcinoma (e.g., hepatocellular carcinoma tumor) including adult (primary) hepatocellular carcinoma and pediatric hepatocellular carcinoma; pancreatic cancer including pediatric pancreatic cancer; sarcoma, rhabdomyosarcoma; islet cell carcinoma; rectal cancer; and small intestine cancer); lung cancer (e.g., non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC); head and neck cancer (e.g., lip and oral cavity cancer; oral cancer including pediatric oral cancer; hypopharyngeal cancer; pharyngeal cancer including pediatric pharyngeal cancer; metastatic squamous neck cancer of unknown primary; oral cancer; nasal and paranasal cavity cancer; nasopharyngeal cancer including pediatric nasopharyngeal cancer; oropharyngeal cancer; parathyroid cancer; pharyngeal cancer; salivary gland cancer including pediatric salivary gland cancer; laryngeal cancer; and thyroid cancer); ovarian cancer and breast cancer.

[0162] In the foregoing method, the amount of the binding protein administered to the subject is effective in inducing an immune response, increasing the immune response, arresting or slowing cancer growth, arresting or slowing cancer metastasis, killing cancer cells, and / or arresting or slowing the proliferation and / or spread of cancer cells in the subject.

[0163] In embodiments of the foregoing method, the binding protein includes Duetmab and BiSAb disclosed herein. In some embodiments of the foregoing method, the binding protein includes an antibody or an antigen-binding fragment thereof, as disclosed herein.

[0164] As used herein, the term "subject" is intended to include humans and non-human animals, particularly mammals. Examples of subjects include human patients having a disorder described herein, such as cancer, or human patients having a normal patient. "Non-human animals" include all vertebrates, such as non-mammals (e.g., chickens, amphibians, reptiles, etc.) and non-human primates, domestic and / or agricultural animals (e.g., sheep, dogs, cats, cows, pigs, etc.) and mammals such as rodents (e.g., mice, rats, hamsters, guinea pigs, etc.). In certain embodiments, the subject is a human patient.

[0165] "Treatment" or "treating" refers to both therapeutic treatment and prophylactic or preventive measures. Subjects in need of treatment include those already having a disorder and those prone to having a disorder or in whom a disorder should be prevented. When used with respect to a disease or subject in need of treatment, the term thus includes, but is not limited to, arresting or slowing the progression of the disease, remission of the disease, prevention of symptoms, reduction in the severity of the disease and / or symptoms, or shortening of the length of the disease as compared to an untreated subject. In multiple embodiments, the treatment method can alleviate one or more clinical indications of the particular disease to be treated.

[0166] The examples described below are provided to illustrate specific aspects and embodiments of the present disclosure described above and should not be construed as limiting the scope of the description or the subject matter claimed in the appended patent.

Example

[0167] Materials and Methods Immune Response Modulation Assay To evaluate the potential immune responses induced by some of the immunotherapeutic molecules described herein, a cytomegalovirus (CMV) antigen recall assay was used. The reagents for the assay included the following: - CMV-reactive cryopreserved peripheral blood mononuclear cells (PBMC); - AIM V® Medium (Life Technologies, cat#12055-091); - Phosphate Buffered Saline (PBS, Life Technologies, cat#20012-043); - PepTivator® CMV pp65 Peptide Pool (Miltenyi Biotec, cat#130-093-438, 50 μg / ml); - Ovalbumin (Thermo scientific cat#77120, 1 mg / ml); - Coater, 96-well plate non-TC treated (Corning, cat#3788); and - Immunotherapeutic molecule.

[0168] General Assay Protocol: The day before performing the assay, the cryopreserved PBMC were thawed in warm AIM V medium. The cells were washed twice in a Coater round well plate. The cell concentration was adjusted to 1×10 6 cells / mL.

[0169] Aliquots of cells (100 μL) were dispensed into individual wells, leaving the outer rows of the plate empty. The cells were left to stand overnight.

[0170] The next day, 100 μL of AIMV medium containing 2× PepTivator CMV peptide pool (final 0.1 μg / ml - 0.05 μg / ml) and 2× immunotherapeutic drug molecule was added to the wells.

[0171] After 72 hours, 25 μL of the supernatant from each well was transferred to a pre-blocked and washed MSD plate (anti-human IFNγ). After adding the standard solution, the plate was incubated at room temperature for 2 hours. After the incubation time, the MSD plate was washed three times. After washing, 25 μL of SULFO-TAG detection antibody was added and reacted at room temperature for 1 hour. The plate was washed again, and after adding 150 μL of 2× MSD reading buffer, the reading was performed.

[0172] Stahphylococcal enterotoxin A / B (SEA / SEB) assay protocol To determine the effect of DuetMab or BiSAb on the IL-2 immune response, the reagents used in either the SEB or SEA assay protocol include the following: - Leukocyte cones (NHSBT code NC24; from Addenbrookes Hospital); - 50 ml Falcon tubes (BD352070); - Ficoll-Paque PLUS (GE Healthcare 17-1440-02); - Anti-CD3 (clone OKT3; 1 mg / ml; eBioscience; cat no: 16-0037-85); - Ammonium chloride solution (Stemcell Technologies 07850); - A stock solution of 1 mg / ml Stahphylococcal enterotoxin A (SEA; Sigma, S-9399) or Stahphylococcal enterotoxin B (SEB; Sigma, S-4881) stored at -20°C: - Media (all from Life Technologies): RPMI 1640 supplemented with 10% v / v heat-inactivated FCS (90005M), 100 U / mL penicillin + 100 μg / ml streptomycin (15140-122), and glutamax (61870); - V-bottom plates (Greiner BioOne 651201); - 96-well flat-bottom plates (Corning Costar 7107).

[0173] Reagents for IL-2 DELFIA ELISA include the following: - FLUONUNC Maxisorp ELISA plates (Nunc 437958); - Europium-labeled streptavidin, SA-EU (Perkin-Elmer 1244-360); - DELFIA® assay buffer (Perkin-Elmer, #4002-0010); - DELFIA® enhancement solution (Perkin-Elmer 4001-0010); RT before use; - Assay diluent: DELFIA wash buffer (0.05% Tween®-20, 20 mM Tris, 150 mM NaCl; pH 7.2 - 7.4) supplemented with 0.1% BSA and sterile filtered; - Dried milk (Marvel; Premier Foods): - Sample diluent (PRMI1640 + 10% FCS + 1% penicillin / streptomycin (same as above)); - PBS (ThermoFisher 14190235); - PBS-Tween® (0.01% Tween®-20 in PBS); - Human IL-2 ELISA kit (Duoset DY202, R&D Systems); - Biotek plate washer with automatic plate loader (EL406).

[0174] General assay protocol PBMCs were isolated from human blood leukocyte cones (NHS Blood and Transplant Service code NC24) using density gradient centrifugation (Ficoll-Paque PLUS; GE Healthcare), followed by lysing red blood cells in ammonium chloride solution (Stemcell Technologies). Flat-bottom 96-well plates (Corning Costar 7107) were coated with anti-human CD3 (clone OKT3 at 0.5 μg / ml in PBS; eBioscience) at 37 °C for 2 hours. Next, PBMCs in medium (RPMI1640-Glutamax supplemented with 10% v / v heat-inactivated fetal bovine serum and 100 U / 100 μg / ml streptomycin / penicillin respectively (Life Technologies)) were added at 0.2×10 6 cells per well. PBMCs were further stimulated by the addition of Staphylococcal Enterotoxin A or B (SEB; Sigma Aldrich) in the range of 0.0088 - 0.1 μg / ml, and then the candidate DuetMab or BiSAb was added to the final test concentration. After 3 days of culture at 37 °C and 5% CO2, the supernatant was removed from the cells and IL-2 secretion was determined using a commercially available ELISA according to the manufacturer's instructions (R&D Systems Duoset product code DY202). See Figure 90.

[0175] Mixed lymphocyte reaction (MLR) assay protocol (fresh blood) Furthermore, an MLR cell assay was also used to obtain an in vitro correlation of T cell function in response to the DuetMabs and BiSAbs disclosed herein. Reagents used to perform the MLR assay from fresh blood samples included the following: - 8 mL CPT heparin tube; - AIM-V Medium (serum-free) Gibco #12055-091, without additives; - 50 ml conical tube; - 2 ml cryopreservation container; - ACK Lysis Buffer (Gibco #A10492-01); - 96-well tissue culture-treated round bottom plate BD falcon #3077; - PHA (Roche) at 1 mg / ml (final concentration of 10 ug / mL) as a positive control.

[0176] General assay protocol PBMCs were prepared from blood samples introduced into CPT heparin tubes. At 25 °C, the tubes were centrifuged at 2700 rpm for 20 minutes without using the brake. The upper layer of serum was aspirated. The remaining material was gently collected with a pipette, and all above the CPT tube plug was recovered and introduced into a 50 ml conical tube. AIM-V medium was added to the cells to wash the cells (with the brake on, 1500 rpm, 5 minutes at 25 °C for 3 times). All remaining red blood cells were lysed using red blood cell lysis buffer (for example, about 3 ml of buffer for about 5 minutes). The remaining cells were washed twice with AIM-V medium (with the brake on, 1500 rpm, 5 minutes at 25 °C). If necessary, the pellet was consolidated in a single tube, resuspended in AIM-V medium, and a cell count was performed.

[0177] To perform the MLR assay, in a 96-well plate, cells were plated at 200,000 cells / donor / well in 50 ul (total 400,000 / 100 ul) of AIM-V medium per donor. The candidate molecule was added at (4×) 50 ul per well and diluted with serum-free AIM-V medium. After 72 hours, the plate was imaged and 30 ul of the supernatant was removed for the human TH1 / TH2 (MSD) cytokine assay.

[0178] Human TH1 / TH2 MSD 10-plex protocol Using this assay, the amount of cytokine present in the culture supernatant in response to the administration of DuetMab and BiSAb disclosed herein was determined. To perform this assay, a blocker was prepared by dissolving 200 mg of blocker B in 20 ml of PBS per plate. 150 μl of the dissolved blocker was added to each well. After sealing the plate, it was shaken at room temperature for 2 hours or overnight at 4°C. The wells were washed three times with PBST buffer. A calibration substance was prepared by diluting 10 μl of the frozen calibration substance blend with 1 ml of diluent, and this was further serially diluted four-fold. To separate the wells, 25 μl of the calibration substance (standard) and 25 μl of the sample were added. The wells were incubated at room temperature for 2 hours while shaking. After incubation, the wells were washed three times with PBST.

[0179] The detection antibody was prepared, diluted to the required concentration, and then added to each well. After incubation at room temperature for 2 hours while shaking, the wells were washed (three times) with PBST. Read buffer was added to each well before reading with the MSD device.

[0180] Tumor-specific killing assay protocol Human gp100 209-217The human CD8+ T cell line (JR6C12) with reactivity against the peptide was kindly provided by Dr. Steven Rosenberg (National Cancer Institute, Bethesda, MD, USA). JR6C12 cells were co-cultured with CFSE (CellTrace CFSE proliferation kit, ThermoFisher)-labeled human melanoma cell line (Mel624) at a 1:1 ratio (20,000 JR6C12 + 20,000 Mel624) in a 96-well flat-bottom plate at 37 °C for 18 hours. At the 0 time point of the co-culture, the candidate molecule was added at a concentration of 69 nM. After 18 hours, the wells were visualized by bright-field microscopy. The supernatant was collected for MSD analysis, and the adherent cells were trypsinized, washed (2×) with PBS, and then subjected to live staining (Zombie UV Fixable Viability kit, Biolegend). The uptake of the live stain by CFSE-labeled cells was evaluated by flow cytometry using LSR Fortessa (BD).

[0181] Example 1. Determination of candidate Fc positions for binding domain binding Using the open-source software PyMOL molecular visualization system, the antibody structure was examined in the CH2 and CH3 regions, as well as at or near the CH2-CH3 interface, to find candidate regions for binding domain binding, such as exposed surface loops. Such regions would be compatible with the insertion of a second binding domain (e.g., scFv) without compromising the structural integrity or stability of the IgG or the second binding domain itself. Three regions were identified from this analysis (represented as spheres in Figures 1A - 1C). Figures 1D, 1E, and 1F depict embodiments of the binding domain and show the binding of the second binding domain (having scFv for illustrative purposes) in each of the loops found in Figures 1A, 1B, and 1C, respectively.

[0182] Figure 2A provides a more detailed schematic of one of the identified representative loops found in the CH2 region near the CH2-CH3 interface and containing the sequence ISRTP (SEQ ID NO: 39). A binding domain can be inserted within this amino acid sequence to generate any number of representative constructs, such as the inserted scFv domains exemplified in the Examples (e.g., I-scFv-SRTP, IS-scFv-RTP, ISR-scFv-TP, or ISRT-scFv-P, scFv-ISRTP, and ISRTP-scFv, where “-scFv-” indicates the point in the native loop sequence to which the binding domain can be attached. Figure 2B is a similar schematic representing a loop found at the CH2-CH3 interface and containing the amino acid sequence AKGQP (SEQ ID NO: 40). Representative constructs described herein may include a binding domain (e.g., an scFv domain) attached to this loop sequence as described herein, such as A-scFv-KGQP, AK-scFv-GQP, AKG-scFv-QP, AKGQ-scFv-P, scFv-AKGQ, AKGQ-scFv, where “-scFv-” indicates the point in the native loop sequence to which the binding domain can be attached. Figure 2C provides a schematic of a representative loop found downstream of the CH2-CH3 interface within the CH3 region and containing the amino acid sequence SNG. Representative constructs for this loop sequence are discussed with respect to the exemplary embodiments for the other two loop regions above and include scFv-SNG, S-scFv-NG, SN-scFv-G, and SNG-scFv.

[0183] Example 2. Generation and Characterization of Bispecific Binding Proteins Comprising a Series of Parent Antibodies and Combinations of Binding Units A series of monoclonal antibodies were prepared and characterized. Using combinations of antigen-binding sequences (e.g., CDR, HCv, LCv, HC, LC) derived from these "parent" antibodies, a series of bispecific binding proteins were prepared, which were found to have bispecific binding activity against the combined target antigens. The bispecific binding proteins were designed to have the specific structural platform motif (i.e., "BiS5") disclosed herein.

[0184] The parent antibody sequences are listed in the following table.

[0185] [Table 3]

[0186] [Table 4]

[0187] [Table 5]

[0188] [Table 6]

[0189] [Table 7]

[0190] [Table 8]

[0191] [Table 9]

[0192] Example 2(a) PD-1 / CTLA-4 Bispecific Binding Protein Although not bound by a particular theory, there is strong clinical and preclinical theoretical rationale for the combination of PD-1 and CTLA-4 blockade. Therefore, it would be desirable to maximize the risk / benefit ratio of the combination of PD-1 and CTLA-4 (Figure 4).

[0193] Using the parental sequences identified above in Table 2, the following bispecific binding proteins that bind to PD-1 and CTLA-4 were generated. Proteins identified as BiS2, BiS3, and BiS5 were generated using the sequences identified below and evaluated for simultaneous antigen binding activity using the Octet binding assay described below.

[0194]

Table 10

[0195]

Table 11

[0196] Octet Binding Assay (BiS2, BiS3, and BiS5) To evaluate the binding of the bispecific binding molecules disclosed herein, an Octet QK equipped with a Ni-NTA biosensor chip and 10× kinetics buffer were used (ForteBio, Menlo Park, CA). For this particular series of bispecific binding proteins, His-tagged PD-L1-Fc, his-tagged PD-1-Fc and CTLA-4-Fc (human recombinant proteins) were purchased from R&D Systems (Minneapolis, MN). All binding assays were performed at 25°C.

[0197] Before analysis, the sample plate was stirred at 1000 rmp. The Ni-NTA biosensor chip was pre-wetted with 1× kinetic buffer for 5 minutes. The 1× kinetic buffer also served as the running buffer for baseline determination and the dilution buffer for the antigen and bispecific antibody. The Ni-NTA biosensor chip was immersed in 100 nM his-tagged PD-L1-Fc (see (b) below) or his-tagged PD-1-Fc for about 1 minute for antigen capture. The antigen-coated biosensor chips were each immersed in 10 μg / ml bispecific antibody for about 5 minutes and then transferred into the column of wells containing 100 nM CTLA-4-Fc antigen for 2 minutes. The results of the binding are shown in Figure 3.

[0198] Using the parental sequences identified above in Table 2, bispecific binding proteins in the DuetMab format that bind to PD-1 and CTLA-4 were prepared. PD-1 / CTLA-4 DuetMab was prepared using the sequences shown in Table 4 below and evaluated as described below, including comparison with PD-1 / CTLA-4 BiS5.

[0199]

Table 12

[0200]

Table 13

[0201]

Table 14

[0202]

Table 15

[0203] Octet Binding Assay (DuetMab) The simultaneous binding test for two individual antigens was performed by Octet analysis. After loading biotinylated human PD-1 onto streptavidin sensors, sequential interactions were first performed with DuetMab PD-1 / CTLA-4 and then with soluble CTLA-4 antigen. Streptavidin (SA) biosensors (ForteBio) were used to capture 5 μg / mL of biotinylated human PD-1 in PBS pH 7.2, 3 mg / mL of BSA, and 0.05% (v / v) Tween® 20 (assay buffer). After the wash step, the loaded biosensors were subjected to sequential association and dissociation interactions with a sample well containing 133 nM of DuetMab PD-1 / CTLA-4 bispecific antibody and then with a well containing 200 nM of human CTLA-4 antigen. The results of the binding are shown in Figure 5.

[0204] The intrinsic kinetics of the PD-1 / CTLA-4 DuetMab bispecific antibody were also evaluated by BiaCore. Binding experiments were performed using a BIAcore T200 instrument (BIAcore). Mouse anti-huIgG-Fab was immobilized on a CM5 chip to a target response of 2000 RU to capture the antibody. To achieve a captured antibody of approximately 100 response units, 100 nM of DuetMab or mAb was flowed at 20 μL / min for 5 minutes. Next, the antigen was continuously injected for 5 minutes at a flow rate of 50 μl / min. Kinetic parameters (k on and k off ) as well as the dissociation constant (KD) were calculated from non-linear fitting using BIAevaluation4.1 software. The binding results are shown in Table 5.

[0205]

Table 16

[0206] Reporter gene assay The results from the reporter gene assay showed that the PD-1 / CTLA-4 bispecific binding protein inhibited the PD-1 and CTLA-4 pathways (Figures 6A - D). The BiS5 binding protein maintained the PD-1 titer compared to the parent, but had a ~3-fold lower titer compared to anti-CTLA-4 IgG. The DuetMab antibody showed a ~9-fold decrease in PD-1 titer and a ~16-fold decrease for CTLA-4 (compared to IgG4P). In the art, there is a need for molecules that target CTLA-4 poorly but maintain functional activity (as shown in the SEB assay). Thus, the PD-1 / CTLA-4 bispecific binding protein has the potential to provide a safety benefit to patients.

[0207] Staphylococcal Enterotoxin B (SEB) assay The results from the SEB assay revealed that DuetMab and BiS5Ab had equivalent activity in the SEB primary immune cell assay (Figure 7), and DuetMab showed high activity compared to the DummyDuet control arm (Figure 8A). DuetMab showed activity nearly equivalent to the combination of parent molecules and high activity compared to LO115 or the CTLA-4 antibody MEDI1123 (tremelimumab) (Figure 8B). Finally, BiS5 and DuetMab showed high activity compared to a combination of novel isotype controls (Figures 9A - B). Data were obtained from 4 donors through 2 independent experiments, and the use of IFNγ was required for these specific assays.

[0208] Mixed lymphocyte reaction (MLR) assay The MLR assay was performed to test the PD-1 / CTLA-4 bispecific molecule. PD-1 / CTLA-4 DuetMab and BiS5Ab had equivalent activities in the mixed lymphocyte reaction (MLR) assay (Figures 10A - C). PD-1 / CTLA-4 DuetMab had high activity compared to the combination of the DummyDuet / isotype control arm (Figures 11A - D). PD-1 / CTLA-4 DuetMab had almost equivalent activity compared to the combination of the parental antibody controls (Figures 12A - D). Finally, PD-1 / CTLA-4 DuetMab had almost equivalent activity compared to the competing PD-1 / CTLA-4 combination and had higher activity than anti-PD-1 alone (e.g., pembrolizumab and nivolumab) (Figures 13A - D).

[0209] Pharmacokinetics and pharmacodynamics (PK / PD) studies Studies were conducted to investigate single-dose pharmacokinetics / pharmacodynamics (PK / PD) in cynomolgus monkeys. The study design is shown in Figure 14. DuetMab showed distinct pharmacodynamics (PD) in cynomolgus monkeys, and a robust PD response was observed for both molecules (Figures 15A - B). Thus, the viability of the PD-1 / CTLA-4 bispecific binding protein in the in vivo environment was confirmed.

[0210] T cell-dependent antibody response (TDAR) DuetMab or BiS5 bispecific molecules at indicated doses (0.5, 5, 50 mg / kg) were administered intravenously (saphenous vein or cephalic vein) to cynomolgus monkeys. Keyhole limpet hemocyanin (KLH) protein was reconstituted with an appropriate amount of sterile injection solution under sterile conditions. The low-dose KLH solution was administered subcutaneously to the back of each animal twice (on days 1 and 29). Blood samples were collected from all animals for further analysis. Evaluation of KLH-specific IgM and IgG antibody titers was performed. Anti-KLH antibodies in monkey sera were detected using ELISA.

[0211] T cell-dependent antibody responses (TDAR) were observed in cynomolgus monkeys administered with PD-1 / CTLA-4 DuetMab (Figure 16A) and PD-1 / CTLA-4 BiS5Ab (Figure 16B).

[0212] CHO cells expressing various levels of human PD-1 and / or CTLA-4 To test PD-1 / CTLA-4 bispecific molecules, a model system was created using stable CHO cells expressing various levels of human PD-1 and / or CTLA-4 (Figure 17). Fluorescently labeled soluble PD-1 and CTLA-4 proteins were used to detect free antigen-binding arms on the cell-binding DuetMab by flow cytometry. The results of this assay reveal that PD-1 / CTLA-4 DuetMab binds simultaneously to PD-1 and CTLA-4 on the surface of the same cell (Figures 18A - C).

[0213] CTLA-4 is constantly internalized into clathrin-coated pits and represents only a small fraction of the receptors expressed on the cell surface at any given time. Recycling of cell surface CTLA-4 is rapid, with less than 80% of surface CTLA-4 internalized within 5 minutes. Therefore, experiments were conducted to examine whether cooperative binding to a combination of anti-PD-1 and anti-CTLA-4 antibodies discriminates PD-1 / CTLA-4 DuetMab in a state where CTLA-4 is saturated on cells expressing excess PD-1 (Figures 19A - C). Target anti-PD-1 and anti-CTLA-4 mAbs were used to independently determine receptor occupancy of each target antigen.

[0214] The parental monoclonal antibodies were found to bind to and occupy their target receptors without exerting a measurable effect on non-target receptors (Figures 20A-D). PD-1 / CTLA-4 DuetMab saturated CTLA-4 on CHO cells expressing excess PD-1 at a concentration approximately 250-fold lower compared to the combination of monoclonal antibodies (Figures 21A-D). PD-1 / CTLA-4 DuetMab saturated CTLA-4 on CHO cells expressing excess PD-1 at a concentration approximately 500-fold lower compared to cells expressing only CTLA-4 (Figures 22A-F). As determined by quantification of doublet formation within the total CHO population pre-mixed, PD-1 / CTLA-4 DuetMab preferentially cis-bound to PD-1 and CTLA-4 on the surface of the same cells (Figures 23A-B). However, PD-1 / CTLA-4 DuetMab can also trans-bind to single-expressing cells. PD-1 / CTLA-4 DuetMab exhibited the internalization properties of the parental anti-CTLA-4 antibody, tremelimumab (Figures 24A-D). Without being bound by a particular theory, the action exhibited by this molecule may induce down-regulation of PD-1. The internalization properties of PD-1 / CTLA-4 DuetMab were also observed in stable CHO cells expressing 10-fold excess PD-1 (Figure 25B).

[0215] Example 2(b) PD-L1 / CTLA-4 Bispecific Binding Protein Using the parental sequences identified above in Table 2, the following bispecific binding proteins that bind to PD-1 and CTLA-4 were generated. The proteins identified as BiS2, BiS3, and BiS5 were generated using the sequences shown in Table 6 below, and the sequences identified below were evaluated for simultaneous antigen binding activity using the Octet binding assay as described above in Section 2(a) (Figure 26).

[0216]

Table 17

[0217]

Table 18

[0218] Example 2(c) PD-1 / TIM3 Bispecific Binding Protein Using the parental sequences identified above in Table 2, the following bispecific binding proteins (Table 7) that bind to PD-1 and TIM3 were prepared. The proteins identified as BiS3, BiS5, and DuetMab were prepared using the sequences identified below and evaluated for simultaneous binding assays by Octet analysis. Briefly, biotinylated human TIM3-IgV domain was captured using a streptavidin (SA) biosensor (ForteBio) at 2 μg / ml in PBS pH 7.2, 3 mg / ml BSA, 0.05% (v / v) Tween® 20 (assay buffer). After the wash step, the loaded biosensor was subjected to sequential binding and dissociation interactions with a sample well carrying 200 nM bispecific antibody and then a well carrying 200 nM PD-1 antigen. After loading the biotinylated human TIM3-IgV domain onto the streptavidin sensor, it was subjected to sequential interactions with the bispecific molecule and then the PD-1 antigen. The binding results are shown in FIGS. 27A - 27B.

[0219] [Table 19]

[0220] [Table 20]

[0221] [Table 21]

[0222] Tumor-Specific Killing Activity Assay Rosenberg Clone Melanoma Killing Assay Using the Rosenberg Clone: JR6C12 and the melanoma cell line: Me1324, the general cell killing activities of the TIM3 / PD-1 bispecific binding molecule and the parental TIM3 antibody were tested.

[0223] General assay protocol JR6C12 is a human CD8+ T cell line that functions as an effector, is expanded from melanoma patients, and is specific for the gp100 - melanoma antigen. To evaluate its therapeutic ability, Mel624 tumor cells were fluorescently labeled and then added together with the effector (JR6C12) and candidate antibodies that bind to TIM3 and / or PD-1. The cells were co-cultured for 16 hours. The multiple panels shown in Figure 28A provide figures visually showing that adding TIM3 62 either in combination with anti-PD1 or as a PD-1 / TIM3 bispecific molecule (as described in Table 7) enhances T cell activation and tumor killing.

[0224] Furthermore, as shown in Figures 28B - 28C, the PD-1 / TIM3 bispecific molecule shows the greatest tumor killing efficacy compared to anti-TIM3, anti-PD-1, or isotype control monotherapy, as evaluated by (b) tumor cell vital stain uptake and (c) IFNγ secretion.

[0225] In addition to clone 62, another bispecific binding protein that binds to PD-1 and TIM3 in the DuetMab format was prepared using the parental sequence identified above in Table 2. The PD-1 / TIM3 DuetMab was prepared using the sequences shown in Table 8 below. The sequence of the TIM3 arm was obtained from O13-1, an affinity matured variant of clone 62, and the sequence of the anti-PD-1 arm was obtained from LO115, which is the same as the PD-1 arm used in the aforementioned PD-1 / CTLA-4 DuetMab bispecific antibody. The PD-1(LO115) / TIM3(O13-1) bispecific antibody was evaluated as discussed below, including comparison with PD-1 / TIM BiS3 and BiS5.

[0226]

Table 22

[0227]

Table 23

[0228]

Table 24

[0229]

Table 25

[0230] Octet Binding Assay (DuetMab, TIM3 Arm Affinity Matured Variant) The simultaneous binding test for two individual antigens, PD-1 and TIM3, was performed by Octet assay. After loading biotinylated human TIM3 onto streptavidin sensors, first, PD-1 / TIM3 DuetMab and then soluble PD-1 antigen were interacted continuously. Biotinylated human TIM3 was captured using 5 μg / ml streptavidin (SA) biosensors (ForteBio) in PBS pH7.2, 3 mg / ml BSA, 0.05% (v / v) Tween® 20 (assay buffer). After the washing step, the loaded biosensors were subjected to consecutive association and dissociation interactions with a sample well carrying 200 nM DuetMab PD-1 / CTLA-4 bispecific antibody having TIM3 arm (O13-1), an affinity matured variant of clone 62 TIM antibody, and then a well carrying 200 nM human PD-1 antigen. The binding results are shown in Figure 29.

[0231] Using BiaCore, the intrinsic kinetics of the PD-1 / TIM3 DuetMab bispecific antibody were also evaluated. Binding experiments were performed using a BIAcore T200 instrument (BIAcore). Mouse anti-huIgG-Fab was immobilized on a CM5 chip up to a target response of 2000 RU to capture the antibody. By flowing 100 nM DuetMab or mAb at 20 μL / min for 5 minutes, a response of approximately 100 response units of the capture antibody was achieved. Next, the antigen was sequentially injected at a flow rate of 50 μL / min for 5 minutes. Kinetic parameters (k on and k off ) and the dissociation constant (KD) were calculated from non-linear fitting using BIAevaluation 4.1 software. The binding results are shown in Table 9.

[0232]

Table 26

[0233] PD-1 / TIM3 bispecific antibodies, including BiS3, BiS5, and DuetMab, bind to CHO cells overexpressing human TIM3 or human PD-1 (Figure 30 and Table 25), and PD-1 and TIM3 expression (DMF4) are shown in Figure 31.

[0234]

Table 27

[0235] CMV Ag recall assay PD-1 / TIM3 bispecific antibodies, including BiS3, BiS5, and DuetMab, increased CD8+ T cell proliferation in the CMV antigen recall assay compared to isotype treatment (Figure 32A - C).

[0236] Mixed leukocyte reaction (MLR) assay PD-1 / TIM3 bispecific antibodies, including BiS3, BiS5, and DuetMab, increased interferon (IFNγ) secretion with activities exceeding those of single agents and combination therapies in the mixed leukocyte reaction (MLR) assay (Figures 33A - D). PD-1 / TIM3 bispecific antibodies, including BiS3, BiS5, and DuetMab, showed similar activities to parental LO115 IgG1 in the jurkat NFκB reporter cell line that mainly expresses PD-1 (87% PD-1 single positive) (Figures 34A - C).

[0237] In summary, three bispecific formats (DuetMab, BiS3, and BiS5) were generated for PD-1 / TIM3. The bispecific formats all showed in vitro functionality equivalent to or superior to anti-PD-1, suggesting that these molecules may offer excellent advantages over existing cancer immunotherapy strategies.

[0238] Example 2(d) OX40 / PD-L1 Bispecific Binding Protein Using the parental sequences identified above in Table 2, the following bispecific binding proteins that bind to PD-1 and OX40 were generated. Proteins identified as BiS2, BiS3, and BiS5 were generated using the sequences shown in Table 10 below and evaluated for simultaneous binding using the Octet binding assay described below.

[0239]

Table 28

[0240]

Table 29

[0241] Octet Binding Assay To evaluate the binding of the bispecific binding molecules disclosed herein, an Octet QK equipped with a Ni-NTA biosensor chip and 10× kinetics buffer were used (ForteBio, Menlo Park, CA). For this particular series of bispecific binding proteins, His-tagged PD-L1-Fc, His-tagged PD-1-Fc and hOX40-Fc (human recombinant proteins) were purchased from R&D Systems (Minneapolis, MN). All binding assays were performed at 25 °C.

[0242] Before analysis, the sample plate was stirred at 1000 rmp. The Ni-NTA biosensor chip was pre-wetted with 1× kinetics buffer for 5 minutes. The 1× kinetics buffer also served as the running buffer for baseline determination and the dilution buffer for the antigen and bispecific antibody. The Ni-NTA biosensor chip was immersed in 100 nM His-tagged PD-L1-Fc (see (b) below) or His-tagged PD-1-Fc for about 1 minute for antigen capture. The antigen-coated biosensor chips were each immersed in 10 μg / ml bispecific antibody for about 5 minutes and then transferred into the column of wells containing 100 nM hOX40-Fc antigen for 2 minutes. The binding results showed that the BiS2 / BiS3 OX40Ab / PD-L1 molecule binds to both PD-L1-His and hOX40-Fc, and that BiS2 OX40Ab / PD-L1 binds with higher affinity than BiS3 OX40Ab / PD-L1. BiS2 OX40Ab / PD-1 was used as a control (Figure 35).

[0243] Staphylococcal Enterotoxin B (SEB) assay The SEB assay using the protocol described above showed that the OX40 / PD-L1 bispecific molecule is active in both the BiS2 and BiS3 formats (Figures 36A - B).

[0244] PD-L1 reporter assay Materials: - Cell lines and culture conditions: - Human PD-1 Jurkat NFAT luciferase clone 2 receptor - PD-L1-expressing CHO scFv OKT3 (UBC) (All cells were maintained at 37 °C in RPMI 1640 medium (RPMI complete medium) containing 10% FBS and 1× pen / strep antibodies in a humidified tissue culture incubator). - RPMI-1640, Life Technologies cat♯ A1049101 - Heat-inactivated fetal bovine serum (FBS), Life Technologies cat♯ 26010074 - Complete RPMI medium: RPMI-1640 containing 10% FBS - 100× Penicillin / Streptomycin, Life Technologies cat♯ 15140-122 - 96-well TC-treated flat-bottom culture plates, Costar 3903, VWR cat♯ 29444-010 - SteadyGlo Luciferase Assay System, Promega, cat# E2510 - Test antibodies - EnVision Multilabel Plate Reader, Perkin Elmer

[0245] Methods For the two-cell bioactivity assay for the neutralization of PD-1 inhibition, PD-L1-expressing CHO scFv OKT3 cells were trypsinized, neutralized in warm RPMI complete medium, and then collected in 50 mL conical tubes. The cells were pelleted at 380 g for 5 minutes at RT and then resuspended in fresh RPMI complete medium and counted on a Vi cell counter. The PD-L1 expressing CHO scFv OKT3 cells were adjusted to 0.4e6 / mL and 25 μL (10,000 cells) per well were plated as shown in the plate layout. The cells were allowed to adhere to the plate for 3 hours. Then, 50 μL of RPMI containing the test reagent (2× final concentration) was aliquoted and added onto the CHO cells and further incubated for 1 hour. This incubation gives the test reagent time to bind to PD-L1 on the CHO cell surface. After 1 hour, PD-1-expressing Jurkat NFAT luciferase reporter cells were collected in 50 mL conical tubes, pelleted at 380 g for 5 minutes at RT, and resuspended in fresh warm RPMI complete medium. The cells were adjusted to 1.2e6 / mL and 25 μL (30,000) cells were plated into the wells containing the PD-L1-expressing CHO scFv OKT3 cells and the test item.

[0246] The cells and test reagent were incubated for an additional 18 hours for the activation of PD-1 Jurkat reporter cells. Then, SteadyGlo luciferase reagent was prepared and 100 μL was aliquoted and added to each well. Complete lysis was achieved by gently shaking (200 rpm orbital shaker) for 15 minutes at RT. After lysis, luciferase activity was measured on an Envision Multilabel Plate Reader using the US96 luminescence protocol. Using Graphpad Prism software, luciferase RLU was plotted against log[test reagent] and the EC 50 value of PD-L1 antagonism was determined using non-linear regression analysis, four-parameter fit of a sigmoidal dose-response curve.

[0247] Results: OX40 / PD-L1 BiS2 / 3 was tested against the PD-L1 / PD-1 parent and NIP228 (G4P) control using a 5-point dose titration starting at 100 nM (PD-L1). All OX40-PD-L1 BisAbs were active and were shown to have stronger agonism than the PD-L1(4736) parent (Figure 37). The BiS2 and BiS3 formats functioned similarly.

[0248] CMV Ag recall assay In the CMV Ag recall assay (using the protocol described above), the BiS2 and BiS3 molecules showed equivalent activity compared to the combination (Figure 38).

[0249] All of the binding and immune response assays described above provide illustrative data that the bispecific binding molecules disclosed herein exhibit specific binding to both target molecules and, in some cases, exhibit higher activity than combinations of individual monospecific parent binding molecules (antibodies), and thus can induce or enhance an immune response. Additionally, they are shown to have cell killing activity against cancer cell lines. Thus, the data demonstrate that these molecules and bispecific platform structures are excellent candidates for cancer immunotherapeutics.

[0250] Octet binding assay (OX40(SLR) / PD-L1 BiS5) To evaluate the binding of the bispecific binding molecules disclosed herein, an Octet QK equipped with a Ni-NTA biosensor chip and 10× kinetics buffer were used (ForteBio, Menlo Park, CA). For this particular series of bispecific binding proteins, His-tagged PD-L1-Fc, his-tagged PD-1-Fc, and hOX40-Fc (human recombinant proteins) were purchased from R&D Systems (Minneapolis, MN). All binding assays were performed at 25°C. The binding results demonstrate that the BiS5 OX40Ab / PD-L1 molecule binds to both PD-L1-His and hOX40-Fc (Figure 39).

[0251] PD-L1 / OX40 BiS5 bound to CHO cells expressing human or cynomolgus monkey OX40 and PD-L1 / B7H1 (Figures 40A - F). Binding of the PD-L1 / OX40 BiS5 construct was also measured by flow cytometry (HyperCyt) (Figure 42). OX40 IgG4P and the OX40 / PD-L1 bispecific molecule bound to Jurkat OX40 receptor cells. PD-L1 IgG and the OX40 / PD-L1 bispecific molecule bound to NCI H358 and CHOK1 B7H1 (PD-L1) / OKT3 cells. All IgG and bispecific molecules bound to HEK CD32a cells.

[0252] PD-L1 and OX40 Receptor Assays In the PD-L1 receptor assay (using the protocol described above), all PD-L1scFv-containing bispecific molecules and the positive control IgG showed activity (Figures 42A - B). The single-arm OX40 control and the isotype control did not show activity in this assay. EC 50 and the Hill slope were consistent with values obtained in previous assays for the anti-PD-L1 parental control and the PD-L1 Bis2, Bis3, and Bis5 constructs.

[0253] In the OX40 reporter gene assay using HEK CD32a cells, the bispecific constructs had equivalent activity to each other and Fc-mediated agonism was observed (Figures 43A - B). OX40 / PD-L1 Bis5 N434A IgG1 had equivalent EC 50 activity to OX40 IgG4P and MEDI0562 (OX40 IgG1).

[0254] In the OX40 reporter gene assay using CHOK1 PD-L1 against expressing cells, the OX40 / PD-L1 bispecific molecule shows equivalent agonism (Figures 44A - B). OX40 / PD-L1 Bis5 N434A IgG1 has an EC 50 activity equivalent to other Fc mutants of the tested OX40 / PD-L1 Bis5 bispecific Mab. Agonism by OX40 IgG or PD-L1 IgG was not detected. Thus, PD-L1-mediated OX40 agonism was demonstrated.

[0255] PD-L1-mediated OX40 agonism with tumor cells was detected using the OX40 / PD-L1 bispecific molecule (Figures 45A - B). The OX40 / PD-L1 bispecific molecule showed an equal agonism-dose curve in this assay. Since no agonism by OX40 IgG was observed, the benefit of using the bispecific molecule over the combination of OX40 IgG and PD-L1 IgG was demonstrated. No agonism was detected with NCI H358 PD-L1 KO cells (Figures 46A - D), which indicates that the NCI H358 agonism observed in the cells is PD-L1 specific.

[0256] Staphylococcal Enterotoxin B (SEB) assay In the SEB assay, the OX40 / PD-L1 bispecific molecule had higher activity than the combination of individual antibodies against OX40 and PD-L1 (Figures 47A - D). In particular, the G4P construct had higher activity than the G1 construct. The wild type, YTE-containing mutant, and N434A mutant had equivalent activity.

[0257] Treg suppression assay The OX40 / PD-L1 bispecific molecule was tested by performing a Treg suppression assay (Figures 4A - D). The OX40 / PD-L1 bispecific molecule only in the presence of PD-L1, CD4+T effwas active above (Figures 49 and 50). Without being bound by a particular theory, this was indicative of cross-linking of OX40 in trans. The OX40 / PD-L1 bispecific molecule reg suppressed the inhibitory effect, but only when cross-linked by binding to plate-immobilized PD-L1.

[0258] Mixed lymphocyte reaction (MLR) assay To test the OX40 / PD-L1 bispecific molecule, an MLR assay was performed (Figures 51A - B). The OX40 / PD-L1 bispecific molecule had higher activity than a combination of individual antibodies against OX40 and PD-L1 (Figures 52A - E).

[0259] Antibody-dependent cell-mediated cytotoxicity (ADCC) assay The ADCC assay was performed to test the OX40 / PD-L1 bispecific molecule. It is an ADCC assay using freshly isolated NK cells as effector cells and CHOK1 PD-L1 B7H1 and CHOK1 OX40 overexpressing cells as target cells respectively at an effector:target molecule (E:T) ratio of 20:1. Target cell lysis was analyzed using the release of europium from labeled target cells after 5 hours. In the ADCC assay, OX40 / PD-L1 BiS2 and BiS5 mediated ADCC against PD-L1 or OX40 expressing CHO cells (Figures 53A - B and 54).

[0260] Using newly isolated NK cells as effector cells and PD-L1 and OX40 overexpressing CHO K1 as target cells respectively, a CD107a mobilization assay was performed at an E:T ratio of 10:1. The mobilization of CD107a on the cell surface of NK cells was analyzed by flow cytometry after 4 hours. BiS2 and BiS5 OX40 / PD-L1 bispecific molecules increased the CD107a mobilization of NK cells against PD-L1 and OX40 expressing CHO cells in the antibody-dependent cell-mediated cytotoxicity (ADCC) assay (Figure 55). BiS2 and BiS5 OX40 / PD-L1 bispecific molecules increased the CD107a mobilization of NK cells against activated allogeneic T cells that upregulated OX40 and PD-L1 (Figure 56). BiS5 OX40 / PD-L1 increased the CD107a mobilization of NK cells from two different donors against activated allogeneic T cells (Figures 57A - B).

[0261] Pharmacokinetics and pharmacodynamics (PK / PD) studies A study was designed to compare the PK / PD of OX40 / PD-L1 bispecific molecules (Figure 58). The serum concentration-time profiles of PD-L1 / OX40 bispecific molecules were compared in cynomolgus monkeys (Figures 59 and Table 11). The mean T 1 / 2 of the Bis5 OX40 / PD-L1 IgG1 N434A molecule was higher than that of the WT Bis5 molecule; the clearance rate was lower for the Bis5 OX40 / PD-L1 IgG1 N434A molecule compared to the WT Bis5 molecule. Both molecules similarly reduced soluble PD-L1 in serum and induced a significant increase in the percentages of Ki67+ total memory CD4+ T cells, total memory CD8+ T cells, and NK cells.

[0262]

Table 30

[0263] The OX40 / PD-L1 bispecific molecule reduced serum soluble PD-L1 concentration below the assay LLOQ (Figure 60). The N434A mutation improved the pharmacokinetics of Bis5 OX40 / PD-L1-G1. In particular, CL was reduced by nearly half; correspondingly, a two-fold increase in T1 / 2 and AUCinf was seen; also, Cmax and Vss were not affected. This was consistent with the reported effect of this mutation on the PK of monoclonal antibodies. Thus, progress towards mAb-like PK for Bis5 OX40 / PD-L1-G1 IO BisAb was achieved. The serum concentration of Bis5 OX40 / PD-L1-G1 BisAb was below the lower limit of quantification (BLOQ) at the 2-week time point, which is thought to be related to ADA.

[0264] A substantial and statistically significant increase was observed in total memory CD4, total memory CD8, and NK cell proliferation (percentage of Ki67+ cells) when the PD-L1 OX40 Bis5 group was compared to the control (anti-PcrV-Psl control) Ab group (Figures 61A - F). A trend towards a significant difference was observed between PD-1 LO115 and the PD-L1 OX40 Bis5 group in total memory CD4, total memory CD8, and NK cell proliferation (Ki67+). There was no statistically significant difference in proliferation between the PD-L1 OX40 Bis5 N434A (half-life extended) version and the G1 version. The PD-L1 OX40 Bis5 N434A and IgG1 versions are biologically active bispecific molecules.

[0265] Example 2(e). OX40 / PD-1 Bispecific Binding Protein Using the parental sequences identified above in Table 2, the following bispecific binding proteins that bind to PD-1 and OX40 were generated. The proteins identified as BiS2 and BiS3 were generated using the sequences shown in Table 24 below and evaluated for simultaneous antigen binding assays using the Octet binding assay described below.

[0266]

Table 31

[0267] The PD-1 / OX40 BiS2 monoclonal antibody (mAb) is a bispecific antibody engineered to simultaneously bind human and cynomolgus PD-1 as well as human and cynomolgus OX40 (Figure 62; the PD-1 binding protein is shown in gray and the OX40 binding protein is shown in light gray). Without being bound by a particular theory, the proposed mechanism of action suggests dual signaling to T cells after cis-binding to both OX40 and PD-1, agonism of the T cell co-stimulatory surface receptor OX40, and blockade of immunosuppressive PD-1 (Figure 63).

[0268] Octet binding assay Shows the simultaneous binding activity of two different lots of PD-1(LO115) / OX40 BiS2 mAb to PD1-His and human OX40-Fc (Figure 64).

[0269] OX40 reporter assay PD-1(LO115) / OX40 BiS2 mAb showed activity comparable to other OX40 agonists (Figure 65A - B). The protein was stored at 4°C, used immediately, frozen / thawed three times, stored at 4°C for 7 days, and then stored at 40°C for 7 days. Activity was recorded as relative luminescence to mAb concentration. On day 0, the EC 50 of PD1(LO115) / OX40 BiS2 mAb at 4°C was approximately 2 nM.

[0270] PD-1 / PD-L1 reporter assay PD-1(LO115) / OX40 BiS2 mAb showed activity comparable to other PD-1 agonists (Figure 66A - B). The protein was stored at 4°C, used immediately, frozen / thawed three times, stored at 4°C for 7 days, and then stored at 40°C for 7 days. Activity was recorded as relative luminescence to mAb concentration. On day 0, for PD1(LO115) / OX40 BiS2 mAb at 4°C, the EC 50It was approximately 1 nM. Two sets of primary human in vitro potency assays were performed: antigen recall T cell assay and T cell co-stimulation using Staphylococcal Enterotoxin B (SEB).

[0271] Staphylococcal Enterotoxin B (SEB) Assay In the SEB assay, the PD-1(LO115) / OX40 BiS2 mAb induced an increase in the level of IL-2 detected in the cell supernatant after 3 days of culture (Figure 67). Thus, the PD-1 / OX40 BiS2 mAb can simultaneously bind to its human target antigen and co-stimulate T cells in vitro.

[0272] In the antigen recall assay, the PD-1 / OX40 BiS2 mAb drove an increase in the level of interferon (IFN)γ compared to the parental mAb and combinations of parental mAbs (Figures 68 and 69).

[0273] CMV Ag Recall Assay As a result of the CMV Ag recall assay (using the protocol described above), the BiS2 and BiS3 molecules did not show equal activity compared to the combination (Figure 70). The data indicate that the PD-1 / OX40 BiS2 IgG4P mAb is active in vitro and in vivo. PD-1 / OX40 BiS3, which has a different structure from PD-1 / OX40 BiS2, was not detectably active. Thus, PD-1 / OX40 BiS3 (inactive) is different from BiS2 (active).

[0274] Pharmacokinetics and Pharmacodynamics (PK / PD) Studies The cynomolgus monkey was considered to be a pharmacologically appropriate non-clinical species for testing the functional activity of the PD-1 / OX40 BiS2 mAb. The pharmacokinetics (PK) and pharmacodynamics (PD) of the PD-1 / OX40 BiS2 mAb were evaluated in non-GLP (Good Laboratory Practices) cynomolgus monkey studies. In cynomolgus monkeys (n = 3; male) after single intravenous (IV) administration over a dose range of 0.1 mg / kg to 30 mg / kg, PD-1(LO115) / OX40 BiS2 mAb PK and PD (percentage of Ki67-positive CD4+ and CD8+ total memory T cells) were evaluated. PBMC were collected before dosing and at 1, 8, 11, and 15 days after dosing, cryopreserved and thawed, and then analyzed by flow cytometry. Briefly, PD-1(LO115) / OX40 BiS2 mAb showed nearly linear PK with a short half-life of 0.6 - 1.7 days (Figure 70; Table 12).

[0275]

Table 32

[0276] The mean peak concentration (C max ) increased almost proportionally to the dose from 2.0 μg / mL at 0.1 mg / kg to 607 μg / mL at 30 mg / kg. AUC∞ increased almost proportionally to the dose from 1.7 μg·day / mL at 0.1 mg / kg to 577 μg·day / mL at 30 mg / kg. The mean serum clearance was in the range of 41.8 mL / day / kg to 60.2 mL / day / kg. The steady-state volume of distribution was in the range of 43.2 mL / kg to 85.6 mL / kg. The PD results (Figure 71) showed a dose-dependent increase in CD4+ total memory T cell proliferation (Ki67) and an increase in CD8+ total memory T cell proliferation (Ki67). A representative standard curve for the quantification of PD-1 / OX40 in cynomolgus monkey serum is shown (Figure 72).

[0277] Example 3. Physical and Chemical Stability of the BiSAb Construct A series of experiments were conducted to evaluate the physical and chemical stability of the BiSAb constructs described herein as compared to other bispecific binding protein structures and platforms. In particular, a series of stability tests described below were performed to reveal and analyze the effects of various pH ranges on the stability of BiSAb (e.g., hydrolysis, fragmentation, aggregation, thermal stability). For different exemplary embodiments of various BiSAb formats, as the data show, the BiSAb disclosed herein (identified as "BiS5" in the following tests and having a D / H format as shown in Table 13) exhibited unexpected and surprising physical and chemical stability compared to all other BiSAb structural motifs.

[0278]

Table 33

[0279] Example 3.1 A further comparison was performed between the BiS format ("BiS5") disclosed herein and another BiS format, identified as "BiS4", which contains two binding domains (scFv domains) linked in the hinge region (e.g., between the Fc and Fab regions). BiS4 and BiS5 proteins were expressed in Chinese hamster ovary (CHO) and purified by conventional chromatography methods. As noted above, these two formats have similar Fab and scFv sequences, and their main difference is the position of the scFv domain (for BiS4, as described herein, the scFv is located within the hinge region; for this particular BiS5, the scFv is present in the SNG loop within the C H 3 domain). The purified BiS molecules were added to PBS buffer and the protein concentration was determined using a NanoDrop ND-1000 (Thermo Scientific, Wilmington, Delaware) with an extinction coefficient of 1.54 M -1 cm -1 .

[0280] pH Screen and Short-Term Stability Test For the pH screen test, BiS4 and BiS5 antibodies were concentrated to about 12 mg / mL and dialyzed against six different pH conditions: 20 mM sodium succinate (pH 5.0), histidine / histidine HCl (pH 5.5, 6.0, and 6.5), and sodium phosphate (pH 7.0 and 7.5) (all containing 240 mM sucrose). Dialysis was performed using Slide-A-Lyzer dialysis cassettes (10 kDa molecular weight cut-off (MWCO), Thermo-Fisher, Rockford, Illinois). After completion of dialysis, 0.02% polysorbate 80 was added and the final protein concentration was adjusted to about 10 mg / mL. The BiS4 and BiS5 formulations were sterilized using a 0.22 μm filter (Millipore, Billerica, Massachusetts) in a pre-sterilized clean bench. One milliliter aliquots were dispensed into 3 mL type I borosilicate glass vials (West Pharmaceutical Services, Exton, Pennsylvania). Samples were stored at 40 °C and analyzed by SEC at zero time point and after 1, 2, 3, and 4 weeks of storage.

[0281] Differential Scanning Calorimetry (DSC) Using a VP-Capilary DSC connected to a temperature-controlled autosampler (Malvern Instruments Ltd., Westborough, Massachusetts), a differential scanning calorimetry thermogram was obtained for the zero-time sample. To obtain the thermogram, a protein concentration of 1 mg / mL was used with a scan rate of 90 °C / h over a temperature range of 20 °C to 100 °C. After subtracting the buffer from the thermograms of BiS4 and BiS5 at different pH conditions in the range of 5.0 to 7.5, baseline correction was performed. Data analysis was carried out using a DSC plugin for the Origin 7 SR4 software package. The experimental results were fitted to a multi-state model with three transitions to calculate the melting temperature (T m ) value. The heat capacity (C p)The value reached 500 cal mol -1 °C -1 The point at which it reached was regarded as the starting temperature (T onset ).

[0282] High-performance size-exclusion chromatography (HP-SEC) To separate monomer from aggregate and fragment species based on size, a 7.8×30 cm 2 , 5 μm, 250 Å, Agilent high-performance liquid chromatography system equipped with a diode array detector capable of recording UV absorption spectra from 200 to 400 nm and a corresponding guard column, Tosho TSKgel G3000SWxl (TOSOH Biosciences, King of Prussia, Pennsylvania) was used to analyze the stability samples. To separate the species, a mobile phase containing 0.1 M disodium hydrogen phosphate anhydrous, 0.1 M sodium sulfate, 0.01% sodium azide, pH 6.8 and a flow rate of 1 mL / min was used. The amount of protein injected was approximately 250 μg. The separation of BiS4 and BiS5 was monitored using the absorption spectrum at 280 nm. The peak areas of soluble aggregates (multimers and dimers), monomers, and fragments were quantified. Next, the percentage of each of these species was calculated and plotted against the incubation time to create a kinetic plot. By calculating the slope of each kinetic plot, a pH profile curve for the rate of monomer loss, fragmentation, and aggregation per month was created.

[0283] Thermal stability of BiS4 and BiS5 Obtained using capillary DSC and from the analysis of thermograms created under six different pH conditions, the effect of pH on the thermal stability of BiS4 and BiS5 was evaluated. Figures 74A and 74B show the DSC thermograms of BiS4 and BiS5 up to pH 5.0 - 7.5 superimposed, respectively. As shown in Figure 73, each thermogram has transition temperatures T m 1, T m 2, and T mShows three heat denaturation events by 3. The first transition (T m 1) is considered to be associated with the simultaneous denaturation of C H 2 and the scFv domain, and the second (T m 2) and the third (T m 3) transitions are associated with the denaturation of C H 3 and the F ab domain. For any of the formats, with an increase in pH up to 6.5, T onset 、T m 1、T m 2, and T m 3 increases were observed (Figure 73A, 73B, 73E and Table 14 below). For BiS4 and BiS5, no differences were observed in T onset 、T m 2, and T m 3 under all pH conditions (Figure 73E and Table 15), which indicates that the presence of scFv in either the hinge region or the C H 3 domain does not affect the thermal stability of C H 3 and F ab . Interestingly, a slight increase in T m 1 was observed for BiS5 under all pH conditions, which indicates that when scFv is located within the C H 3 domain, it shows an increase in the thermal stability of either scFv, C H 2 or both.

[0284]

Table 34

[0285] Physical and Chemical Stability of BiS4 and BiS5 The physical and chemical stability of the BiS4 and BiS5 formats was evaluated at different pH values (5.0 - 7.5) at 40°C for up to 4 weeks. Using the HP-SEC chromatogram at "zero time", the total area, monomer, aggregate, and fragment content of the HP-SEC chromatograms at other time points were compared. Representative chromatograms of BiS4 and BiS5 at pH 7.5 zero time compared to 4 weeks are shown in Figure 74A. All samples mainly contained low levels of soluble aggregates and contained monomers with or without fragments. At zero time (solid line), most of the samples were monomers, with no significant differences other than a slight difference in peak height between the two samples, presumably due to some concentration differences (Figure 74A). The dotted line shows the overlay of the HP-SEC chromatograms of both formats under the same pH conditions after storage at 40°C for 4 weeks. Under accelerated temperature stress conditions, both formats showed additional peaks, early eluting peaks (multimeric species), a decrease in monomers, and an increase in fragment levels (Figure 74A). The loss of monomers due to fragmentation was more pronounced in BiS4 compared to BiS5, indicating that BiS5 is more chemically stable. Based on their structures, possible fragmentation sites, and retention times, the small fragment peak (RT ~10.8 minutes) is presumed to be Fab, and the large fragment peak (RT ~9.8 minutes) and shoulder peak (RT ~8.7 minutes) are presumed to be Fab containing scFv, and its corresponding high molecular weight fragment (HMWF) containing Fab, scFv, and Fc, respectively.

[0286] To further better evaluate the effect of the position of the scFv on the physical and chemical stability of BiS4 and BiS5, the percentage of the total area of each species (%) was plotted as a bar graph for zero time and 4 weeks at pH 7.5 and 40 °C (Figure 74B). As shown in Figure 74B, at zero time, the monomer purity of BiS4 and BiS5 is similar. Samples incubated at 40 °C for up to 4 weeks showed significant differences in the type and extent of the fragments formed. In the case of BiS4, shoulder peaks (RT ~8.7 min) of 11.8%, 7.2% and 3.5%, a large fragment (RT ~9.8 min) and a small fragment (RT ~10.8 min) were formed respectively (Figure 74B). Surprisingly, the BiS5 sample showed only a small fragment (RT ~10.8 min) of 1.4%, probably because the scFv was tethered to the Fc from both sides of the domain.

[0287] Figures 75A - 75C show the kinetics of aggregation, fragmentation and monomer loss of BiS4 and BiS5 incubated at 40 °C for the pH 7.5 samples. The BiS4 sample showed a faster monomer loss rate compared to BiS5 (Figure 75A). The monomer loss rates of BiS4 and BiS5 at pH 7.5 were 27.4% per month and 4.5% per month respectively (Figure 75A). In the case of BiS4, most of the monomer loss was due to fragmentation, which was 23.9% per month, and to a lesser extent, 3.5% / month due to aggregation (Figures 75B and 75C). Interestingly, in the case of BiS5, aggregation seems to be at a slightly higher rate (2.8% / month) compared to the fragmentation rate (1.7% / month) (Figures 75B - 75C).

[0288] Further analysis of the effect of pH on the physical and chemical stability, monthly monomer loss, fragmentation and aggregation rates of the BiS4 and BiS5 formats was performed by plotting the above values against six pH conditions (Figs. 76A - 76C). Throughout all six pH conditions in the range of pH 5.0 - 7.5, the monomer loss rate was lower for the BiS5 format than for BiS4 (Fig. 76A), suggesting that the BiS5 format disclosed herein has unexpectedly superior physical and chemical stability compared to other bispecific protein formats. In the case of BiS4, most of the monomer degradation was due to fragmentation at lower pH conditions (Fig. 76B). BiS5 showed a lower fragmentation rate than Bis4 at all pH conditions tested. Surprisingly, the fragmentation rate of BiS5 appears to be flat over a wide pH range and lower than Bis4. Without being bound to any particular theory, the relatively low fragmentation rate observed with BiS5 may be due to the G4S linker at either end of the scFv that links it to Fc. Fragmentation at one of the G4S linkers that link to Fc may still be linked to Fc via the other G4S linker and thus may not release the scFv. In BiS4 and BiS5, the aggregation rates appear to be similar throughout all pH conditions tested (Fig. 76C), suggesting that the position of the scFv has a minimal effect on the aggregation kinetics, which, as measured using capillary-type DSC, is the T between the two formats at all pH conditions onsetThis is also supported by the fact that no change was observed (Figure 73E and Table 14 above). At pH 7.5 and 40 °C (time = 0), none of the molecules showed obvious fragmentation (Figure 77A), but obvious fragmentation was observed for BiS4 and slight fragmentation was observed for BiS5 under the same conditions after storage at 40 °C for two weeks (Figure 77B). For both BiS4 and BiS5, fragmentation and aggregation decreased at low (5.5) pH, but BiS5 had excellent performance in both fragmentation and aggregation at any pH value (Figure 78). This series of experiments demonstrates that BiS5 disclosed herein has better chemical stability than BiS4 and has similar physical stability to BiS4.

[0289] Example 3.2 Further tests were conducted to evaluate the physical and chemical stability of various embodiments of the bispecific binding proteins disclosed herein and identified as constructs A - H (e.g., Table 13 above and related examples). As described below, these constructs were analyzed using DSC, accelerated storage stability, and FcRn and FcgR binding assays.

[0290] Differential scanning calorimetry analysis DSC experiments for this dataset were performed using a Microcal VP-DSC scanning microcalorimeter (Microcal). All solutions and samples used in DSC were filtered using a 0.22 μm filter and degassed before loading into the calorimeter. Antibodies used in DSC tests were >98% monomer as determined by analytical SEC. All samples were dialyzed thoroughly in 25 mM histidine-HCl (pH 6.0) (at least 3 buffer exchanges) before DSC analysis. The buffer from this dialysis was used as the standard buffer for the next DSC experiment. Baseline measurements (buffer vs. buffer) were subtracted from the sample measurements before sample measurement. Dialyzed samples (concentration 1 mg / ml) were added to the sample wells and DSC measurements were performed at a scanning rate of 1 °C / min. Data analysis and deconvolution were performed using Origin™ DSC software provided by Microcal. Deconvolution analysis was performed using a non-two-state model and the best fit was obtained using 100 iteration cycles. T onset is defined as the qualitative temperature at which it becomes apparent that the thermogram has a non-zero slope, and T m is defined as the temperature at which half of the molecules in the assembly are denatured, which is calculated as the temperature value corresponding to the maximum of each peak in the thermogram.

[0291] The results for different constructs are shown in Figure 79. In general, constructs A, C, and D containing 2F4 as scFv have lower T M 1 than constructs E, G, and H containing LC10 as scFv. Without being bound to a particular theory, the difference in TM1 values is thought to be due to the inherently superior thermal stability of the LC10 scFv domain relative to the 2F4 variable domain. These data suggest that constructs A-D having 2F4 as scFv will be less thermally stable than constructs E-H having LC10 as scFv.

[0292] Accelerated storage stability analysis The concentration of the construct was normalized to 1 mg / mL. 1 mL of each bispecific construct or IgG control was aliquoted and introduced into 1.5 ml Eppendorf tubes. The samples were incubated in a static incubator at 45 °C for 2 weeks. Samples were analyzed at 3, 7, and 14 days to evaluate stability. Appearance inspection was performed at each time point to record any increase in turbidity or precipitation. Samples were filtered using a 0.2 um spin column, 120 ul of the sample was aliquoted and introduced into HPLC, and it was confirmed that there were no bubbles at the bottom of the vial. Next, TSK-GEL G3000SW XL (300×7.8 mm) Tosoh Bioscience column was used to test the samples by Agilent 1100 series HPLC-SEC to check for aggregation and degradation. 60 μL of the sample was injected and run at a flow rate of 1 mL / min. Monomer retention time (min), total peak area, % monomer, % aggregate, % fragment, % monomer loss were obtained and used for analytical SEC analysis. The results are summarized in Table 15.

[0293]

Table 35

[0294] As described herein, the positions of the scFv domains in the aforementioned construct are as follows ("-" indicates scFv): A and E are IS-RTP; B and F are AK-GQP; C and G are S-NG; D and H are SN-G. Various T M values are associated with the following domains: T M 1 = CH2 / scFv; T M 2 = Fab; T M3 = CH3. The data tend to show that constructs A and C with 2F4 scFv inserted into the ISRTP(A) and SNG(C) loops are more prone to aggregation than constructs E and G with LC10 scfv inserted at the same position. This observation suggests that the sequence identity and behavior of the scFv domain can have an impact on the stability of bispecific binding protein constructs. Furthermore, from the above, it can be predicted that construct D containing 2F4 scFv has a similar low stability to A and C, but inserting 2F4 scFv into the SNG loop seems to stabilize the molecule and reduce the tendency to form aggregates. Overall, this accelerated stability test indicates that the scFv sequence and position within the Fc region can play a fairly important role in the stability of the BiSAb construct.

[0295] FcRn and FcγR Binding Analysis Binding experiments were performed using a BIAcore 3000 instrument (BIAcore). 1000 RU IsdH(Fab) antigen was immobilized on a CM5 chip to capture the antibody. The antibody was captured by flowing 100 nM of the BiSAb construct or mAb control at 20 μL / min for 5 minutes. 5 μM huFcRn or FcγRI, IIa, IIb, IIIa - 158V or IIIA158F was flowed at 5 μL / min for 20 minutes. FcRn binding was performed in PBS + 5 μM EDTA at pH 6.0 and also in PBS + 5 μM EDTA at pH 7.4.

[0296] Constructs A, C, D, E, G, and H were evaluated for FcRn binding. Representative data for each of the bispecific constructs E and H, as well as for 2F4 IgG binding to FcRn, are shown in Figure 80. Constructs with an scFv downstream of the CH2-CH3 interface were found to retain FcRn binding (e.g., constructs D and H). Constructs with an scFv located within the ISRTP loop upstream of the CH2-CH3 interface appear to abolish detectable FcRn binding (e.g., constructs A and E). The ISRTP loop is within the region of the known half-life extending YTE mutation in Fc (M252Y / S254T / T256E), which has been found to be important for FcRn binding.

[0297] Constructs A, C, D, E, G, and H were tested for binding to FcγRI, FcγRIIa, FcγRIIb, FcγRIIIa-158F, and FcγRIIIa-158V. Representative data for the binding of constructs E, G, and H to FcγRIIIa-158V are shown in Figure 81. All constructs tested retained binding to FcγR, but with different affinities (Figure 81, inset). Table 16 shows the observed binding trends of various constructs to FcγR.

[0298]

Table 36

[0299] Compared to other constructs (C, D, G, and H) with an scFv inserted into the SNG loop downstream of the CH2-CH3 interface, the constructs (A and E) with an scFv inserted into the ISRTP loop upstream of the CH2-CH3 interface consistently show lower FcγR binding.

[0300] Attempts to evaluate whether the FcRn binding of constructs A and E can be improved or restored were carried out by introducing a half-life extension loop (N3) into the Fc region. Figure 82 shows representative data indicating that for construct E, neither BiS5Ab E nor construct E with the N3 loop introduced (BiS5Ab E+N3) was able to bind to FcRn. Furthermore, even when LC10 scFv was inserted into the N3 loop (N3 scFv) and the ISRTP loop was maintained intact, FcRn binding decreased to levels below the detectable level. These data indicate that, at least in the case of construct E and, by extension, for each of the constructs disclosed herein, both the ISRTP loop and the N3 loop (if present) need to be maintained intact and unmodified in order to retain FcRn binding.

[0301] Example 3.3 In addition to the comparison of BiS4 with the bispecific binding construct (BiS5) disclosed herein, tests were carried out to evaluate three other BiS structural motif platforms, identified as BiS1, BiS2, and BiS3 (see Figure 83). As can be understood with reference to Figure 83, these platforms vary with respect to the position of one binding domain (shown as the scFv domain). Of the five motifs, only BiS4 and BiS5 contain two linker portions as binding points to large proteins, while the others (BiS1, BiS2, and BiS3) are joined by a single linker.

[0302] Briefly, representative molecules of each construct were analyzed for stability using the techniques described in Examples 3.1 and 3.2 above. Samples of each construct were added to buffers at pH 5.0, 5.5, 6.0, 6.5, 7.0, and 7.5 and stored at 40 °C for a period of two months. Next, the samples were analyzed using HP-SEC for fragmentation rate (Figure 84), aggregation rate (Figure 85), and monomer loss rate (Figure 86). Under these conditions, analysis revealed that the bispecific binding protein formats disclosed herein (“BiS5”; and the D / H format shown above in Table 13) had superior physical and chemical stability compared to all other formats at all pH conditions.

[0303] Also, SEC data was used to map various peaks to the corresponding fragments of the BiS molecules (Figure 87). The mapping was based on estimates such as where fragmentation occurs in the hinge and linker regions of the molecule, the size of the fragments, the theoretical fragmentation, and how the expected fragment species are adjusted with respect to the fragment species observed in other formats. Although there was good resolution among the low molecular weight fragments (LMWF) in each format, there was poor or no resolution between the monomer and high molecular weight fragments (HMWF) in all formats. Based on the information in Table 17, it was concluded that the HP-SEC technique grossly underestimated the fragmentation of the BiS format compared to monoclonal antibodies. Alternative analyses were developed as described below.

[0304]

Table 37

[0305] (i) If small and medium-sized fragments are detected during degradation, the corresponding large fragments should also be present; (ii) During the stability test period, secondary fragmentation (fragmentation of fragments) should not occur significantly; and (iii) Based on the presumption that fragmentation occurs in the linker region and / or hinge region, an alternative analysis was developed to calculate the fragmentation rate of HMWF using the molar extinction coefficient. The fragmentation rate was determined based on the following relationship.

Number

[0306]

Table 38

[0307] Furthermore, to determine whether disulfide bonds affected fragmentation and stability, an analysis of the fragmentation rate was performed using constructs under reducing conditions. Representative data for this assay are shown in (Figure 86). Under reducing conditions, a higher fragmentation rate was observable in all BiS formats except BiS1 (Table 19). It was concluded that the higher fragmentation rate under reducing conditions supported that the scFv portion in the BiS5 construct was tethered to the CH3 region (Figure 89).

[0308]

Table 39

[0309]

Table 40

[0310] The characteristics of the stabilized disulfide bonds disclosed above were further investigated. The results are shown in Tables 23 and 24 below. Bispecific antibodies corresponding to two different specificities were prepared in BiS4 and BiS5 (scFv is inserted into the SN-G loop) formats with or without a stabilized disulfide bond in the scFv. Accelerated stability testing revealed that the BiS4 construct without a stabilized disulfide bond had a substantial monomer loss due to degradation that was blocked by the introduction of a stabilized VL-VH disulfide bond. These results indicate that the removal of the stabilized disulfide bond in the scFv of the BiS5 construct had no significant effect on its stability.

[0311]

Table 41

[0312]

Table 42

[0313] Based on all the data above, the bispecific binding protein formats disclosed herein appear to be the most stable among all the formats tested. Furthermore, BiSAb5 was found to be the most stable in terms of minimizing both fragmentation and aggregation at the low pH values tested (e.g., 5.0, 5.5, and 6.0). Thus, the unexpected and surprising stability characteristics of the BisAb disclosed herein provide additional advantages over other structural platforms and formats used in the production of bispecific binding molecules.

[0314] Incorporation by reference All publications and patents mentioned in this specification are hereby incorporated by reference in their entirety as if each individual publication or patent was specifically and individually indicated to be incorporated by reference.

[0315] While specific embodiments of the present disclosure have been considered, the above description is illustrative and not limiting. Many variations of the present disclosure will become apparent to those skilled in the art upon review of this specification and the following claims. The full scope of the present disclosure should be determined with reference to the claims, their full scope of equivalents, and this specification with its variations. Hereinafter, embodiments of the present invention will be shown. (1) A protein comprising a first binding domain (BD1) that binds to a first epitope, a second binding domain (BD2) that binds to a second epitope, C H 2 and C H 3 domains and an Fc region, wherein the Fc region contains BD2 in a solvent-exposed loop at the interface of the C 2 domain, the C H 3 domain, or the C H 2 and C H 3 domains, and H the protein is bivalent for binding to each of the first and second epitopes. (2) The protein according to (1), wherein the Fc region contains BD2 in a solvent-exposed loop in the amino acid sequence at the interface of the C 2 domain, the C H 3 domain, or the C H 2 and C H 3 domains. H (3) The protein according to (2), wherein the solvent-exposed loop contains an amino acid sequence derived from the C 2 domain. H (4) The protein according to (3), wherein the solvent-exposed loop contains the amino acid sequence ISRTP (SEQ ID NO: 39). (5) The protein according to (2), wherein the solvent-exposed loop contains an amino acid sequence derived from the C 3 domain. H (6) The protein according to (5), wherein the solvent-exposed loop contains the amino acid sequence SNG. (7) The solvent-exposed loop contains the C (6) The protein according to (5), wherein the solvent-exposed loop contains the amino acid sequence SNG. (7) The solvent-exposed loop contains the C H2 domains and said C H The protein according to (2), comprising the amino acid sequence derived from the boundary surface of the 3 domains. (8) The protein according to (7), wherein the solvent-exposed loop comprises the amino acid sequence AKGQP (SEQ ID NO: 40). (9) BD2 comprises a single-chain variable fragment (scFv), the protein according to any one of (1) to (8). (10) BD1 comprises a binding domain selected from the group consisting of a Fab domain, an scFv, a single-domain antibody, and an antibody variable domain, the protein according to any one of (1) to (8). (11) BD1 comprises a Fab domain, the protein according to any one of (1) to (8). (12) The protein according to (11), wherein the Fab domain is linked to the Fc region via an antibody hinge region. (13) The protein according to any one of (1) to (8), wherein the Fc region comprises a domain selected from the group consisting of Fc regions derived from IgG1, IgG2, IgG3, IgG4, IgA, IgM, IgE, and IgD. (14) The protein according to (13), wherein the Fc region comprises a mutant Fc region. (15) The protein according to (13), wherein the Fc region is non-glycosylated. (16) The protein according to (13), wherein the Fc region is deglycosylated. (17) The protein according to (13), wherein the Fc region has hypofucosylation or is non-fucosylated. (18) The protein according to any one of (1) to (8), further comprising a protein linker L1 between BD2 and the Fc region. (19) The protein according to any one of (1) to (8), further comprising a first protein linker L1 and a second protein linker L2 between BD2 and the Fc region. (20) The protein according to any one of (1) to (8), wherein BD2 is bound to the Fc region via a protein linker L1. (21) BD2 is the protein according to any one of (1) to (8), which is bound to the Fc region via two protein linkers L1 and L2. (22) L1 and L2 are the proteins according to any one of (18) to (21), which are independently selected from (G4S)2 (SEQ ID NO: 41), (G4S)3 (SEQ ID NO: 42), and (G4S)4 (SEQ ID NO: 43). (23) From the N-terminus to the C-terminus, the following polypeptide domains: V H 1-C H 1-C H 2 (N-terminus)-BD2-C H 2 (C-terminus)-C H 3 comprising a chimeric heavy chain containing, and BD1 contains a Fab domain, V H 1 contains the heavy chain variable domain of the Fab domain, and C H 1 contains the heavy chain constant domain 1 of the Fab, the protein according to (1). (24) From the N-terminus to the C-terminus, the following polypeptide domains: V H 1-C H 1-C H 2-BD2-C H 3 comprising a chimeric heavy chain containing, and BD1 contains a Fab domain, V H 1 contains the heavy chain variable domain of the Fab domain, and C H 1 contains the heavy chain constant domain 1 of the Fab, the protein according to (1). (25) From the N-terminus to the C-terminus, the following polypeptide domains: V H 1-C H 1-C H 2-C H 3 (N-terminus)-BD2-C H 3 (C-terminus) comprising a chimeric heavy chain containing, and BD1 contains a Fab domain, V H1 includes the heavy chain variable domain of the Fab domain and C H 1 is the protein according to (1), which includes the heavy chain constant domain 1 of the Fab. (26) BD2 is the protein according to any one of (23) to (25), which includes an scFv. (27) The scFv, from the N-terminus to the C-terminus, V H 2-polypeptide linker-V L 2 or V L 2-polypeptide linker-V H includes 2, V H 2 includes the heavy chain variable domain of the scFv, and V L 2 includes the light chain variable domain of the scFv, which is the protein according to (26). (28) The protein according to any one of (23) to (27), which further includes a protein linker L1 between BD2 and the Fc region. (29) The protein according to any one of (23) to (27), which further includes a first protein linker L1 and a second protein linker L2 between BD2 and the Fc region. (30) BD2 is bound to the interface of the C H 2 domain, the C H 2 domain, or the C H 2 and C H 3 domain of the protein according to any one of (23) to (25). (31) BD2 is bound to the interface of the C H 2 domain, the C H 2 domain, or the C H 2 and C H 3 domain of the protein according to any one of (23) to (25) through two protein linkers L1 and L2. (32) L1 and L2 are independently selected from protein linkers having a length of 1 to 25 amino acids, which is the protein according to any one of (28) to (31). (33)L1 and L2 are the proteins according to any one of (28) to (31), independently selected from (G4S)2 (SEQ ID NO: 41), (G4S)3 (SEQ ID NO: 42), and (G4S)4 (SEQ ID NO: 43). (34) The protein according to any one of (1) to (33), wherein the first and second epitopes are different. (35) The protein according to any one of (1) to (34), wherein the first and second epitopes are the same. (36) A bispecific binding protein that binds to PD-1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 1 and a second peptide comprising the amino acid sequence of SEQ ID NO: 2. (37) A bispecific binding protein that binds to PD-1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 3 and a second peptide comprising the amino acid sequence of SEQ ID NO: 4. (38) A bispecific binding protein that binds to PD-1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 5 and a second peptide comprising the amino acid sequence of SEQ ID NO: 6. (39) A bispecific binding protein that binds to PD-1 and CTLA-4, comprising a first heavy chain comprising the amino acid sequence of SEQ ID NO: 9, a first light chain comprising the amino acid sequence of SEQ ID NO: 7, a second heavy chain comprising the amino acid sequence of SEQ ID NO: 12, and a second light chain comprising the amino acid sequence of SEQ ID NO: 4. (40) A bispecific binding protein that binds to PD-L1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 14 and a second peptide comprising the amino acid sequence of SEQ ID NO: 15. (41) A bispecific binding protein that binds to PD-L1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 16 and a second peptide comprising the amino acid sequence of SEQ ID NO: 17. (42) A bispecific binding protein that binds to PD-L1 and CTLA-4, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 18 and a second peptide comprising the amino acid sequence of SEQ ID NO: 19. A bispecific binding protein that binds to PD-1 and TIM3, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 22 or SEQ ID NO: 89 and a second peptide comprising the amino acid sequence of SEQ ID NO: 23 or SEQ ID NO: 90. (44) A bispecific binding protein that binds to PD-1 and TIM3, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 24 or SEQ ID NO: 91 and a second peptide comprising the amino acid sequence of SEQ ID NO: 23 or SEQ ID NO: 92. (45) A bispecific binding protein that binds to PD-1 and TIM3, comprising a first heavy chain comprising the amino acid sequence of SEQ ID NO: 9, a first light chain comprising the amino acid sequence of SEQ ID NO: 7, a second heavy chain comprising the amino acid sequence of SEQ ID NO: 27 or SEQ ID NO: 30, and a second light chain comprising the amino acid sequence of SEQ ID NO: 26 or SEQ ID NO: 28. (46) A bispecific binding protein that binds to OX40 and PD-L1, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 34 and a second peptide comprising the amino acid sequence of SEQ ID NO: 32. (47) A bispecific binding protein that binds to OX40 and PD-L1, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 35 and a second peptide comprising the amino acid sequence of SEQ ID NO: 32. (48) A bispecific binding protein that binds to OX40 and PD-L1, comprising a first peptide comprising the amino acid sequence of SEQ ID NO: 36 or SEQ ID NO: 94 and a second peptide comprising the amino acid sequence of SEQ ID NO: 32 or SEQ ID NO: 93. (49) An antibody or antigen-binding fragment thereof that binds to TIM3, comprising a heavy chain comprising CDR1, CDR2 and CDR3 and a light chain comprising CDR1, CDR2 and CDR3, wherein the heavy chain CDR1 comprises SEQ ID NO: 79, the heavy chain CDR2 comprises SEQ ID NO: 80, the heavy chain CDR3 comprises SEQ ID NO: 81, the light chain CDR1 comprises SEQ ID NO: 82, the light chain CDR2 comprises SEQ ID NO: 83, and the light chain CDR3 comprises SEQ ID NO: 84. An antibody or antigen-binding fragment thereof according to (49), comprising a heavy chain variable region and a light chain variable region, wherein the heavy chain variable region comprises SEQ ID NO: 85 and the light chain variable region comprises SEQ ID NO: 86. An antibody or antigen-binding fragment thereof according to (49), wherein the heavy chain comprises SEQ ID NO: 87 and the light chain comprises SEQ ID NO: 88. A composition comprising a protein or antibody according to any one of (1) to (51) and a pharmaceutically acceptable carrier. A nucleic acid molecule comprising a nucleotide sequence encoding a protein or antibody according to any one of (1) to (51). A vector comprising the nucleic acid molecule according to (53). A host cell comprising the vector according to (54). A method for treating or preventing cancer in a subject, the method comprising administering to the subject a protein or antibody according to any one of (1) to (51). The method according to (56), wherein the cancer is one or more of ovarian cancer, breast cancer, colorectal cancer, prostate cancer, cervical cancer, uterine cancer, testicular cancer, bladder cancer, head and neck cancer, melanoma, pancreatic cancer, renal cell cancer, and lung cancer. A method for enhancing an immune response in a subject, the method comprising administering to the subject a protein or antibody according to any one of (1) to (51).

[0316] [Sequence Listing] TIFF2025090743000045.tif199141TIFF2025090743000046.tif199141TIFF2025090743000047.tif199141TIFF2025090743000048.tif199141TIFF2025090743000049.tif199141TIFF2025090743000050.tif199141TIFF2025090743000051.tif199141TIFF2025090743000052.tif199141TIFF2025090743000053.tif199141TIFF2025090743000054.tif199141TIFF2025090743000055.tif199141TIFF2025090743000056.tif199141TIFF2025090743000057.tif199141TIFF2025090743000058.tif199141TIFF2025090743000059.tif199141TIFF2025090743000060.tif199141TIFF2025090743000061.tif199141TIFF2025090743000062.tif199141TIFF2025090743000063.tif199141TIFF2025090743000064.tif199141TIFF2025090743000065.tif199141TIFF2025090743000066.tif199141TIFF2025090743000067.tif199141TIFF2025090743000068.tif199141TIFF2025090743000069.tif199141TIFF2025090743000070.tif199141TIFF2025090743000071.tif199141TIFF2025090743000072.tif199141TIFF2025090743000073.tif199141TIFF2025090743000074.tif199141TIFF2025090743000075.tif199141TIFF2025090743000076.tif199141TIFF2025090743000077.tif199141TIFF2025090743000078.tif199141TIFF2025090743000079.tif199141TIFF2025090743000080.tif199141TIFF2025090743000081.tif199141TIFF2025090743000082.tif199141TIFF2025090743000083.tif199141TIFF2025090743000084.tif199141TIFF2025090743000085.tif199141TIFF2025090743000086.tif199141TIFF2025090743000087.tif199141TIFF2025090743000088.tif199141TIFF2025090743000089.tif199141TIFF2025090743000090.tif199141TIFF2025090743000091.tif199141TIFF2025090743000092.tif199141TIFF2025090743000093.tif199141TIFF2025090743000094.tif199141TIFF2025090743000095.tif199141TIFF2025090743000096.tif199141TIFF2025090743000097.tif199141TIFF2025090743000098.tif199141TIFF2025090743000099.tif199141TIFF2025090743000100.tif199141TIFF2025090743000101.tif199141TIFF2025090743000102.tif199141TIFF2025090743000103.tif199141TIFF2025090743000104.tif199141TIFF2025090743000105.tif199141TIFF2025090743000106.tif199141TIFF2025090743000107.tif199141TIFF2025090743000108.tif199141TIFF2025090743000109.tif199141TIFF2025090743000110.tif199141TIFF2025090743000111.tif199141TIFF2025090743000112.tif199141TIFF2025090743000113.tif199141TIFF2025090743000114.tif199141TIFF2025090743000115.tif199141TIFF2025090743000116.tif199141TIFF2025090743000117.tif199141TIFF2025090743000118.tif199141TIFF2025090743000119.tif199141TIFF2025090743000120.tif199141TIFF2025090743000121.tif199141TIFF2025090743000122.tif199141TIFF2025090743000123.tif199141TIFF2025090743000124.tif199141TIFF2025090743000125.tif199141TIFF2025090743000126.tif199141TIFF2025090743000127.tif199141TIFF2025090743000128.tif199141TIFF2025090743000129.tif199141TIFF2025090743000130.tif199141TIFF2025090743000131.tif199141TIFF2025090743000132.tif199141TIFF2025090743000133.tif199141TIFF2025090743000134.tif199141TIFF2025090743000135.tif199141TIFF2025090743000136.tif199141TIFF2025090743000137.tif199141TIFF2025090743000138.tif199141TIFF2025090743000139.tif199141TIFF2025090743000140.tif199141TIFF2025090743000141.tif199141TIFF2025090743000142.tif199141TIFF2025090743000143.tif199141TIFF2025090743000144.tif199141TIFF2025090743000145.tif199141TIFF2025090743000146.tif199141TIFF2025090743000147.tif199141TIFF2025090743000148.tif199141TIFF2025090743000149.tif199141TIFF2025090743000150.tif199141TIFF2025090743000151.tif199141TIFF2025090743000152.tif199141TIFF2025090743000153.tif199141TIFF2025090743000154.tif199141TIFF2025090743000155.tif199141TIFF2025090743000156.tif199141TIFF2025090743000157.tif199141TIFF2025090743000158.tif199141TIFF2025090743000159.tif199141TIFF2025090743000160.tif199141TIFF2025090743000161.tif199141TIFF2025090743000162.tif199141TIFF2025090743000163.tif199141TIFF2025090743000164.tif199141TIFF2025090743000165.tif199141TIFF2025090743000166.tif199141TIFF2025090743000167.tif199141TIFF2025090743000168.tif199141TIFF2025090743000169.tif199141TIFF2025090743000170.tif199141TIFF2025090743000171.tif199141TIFF2025090743000172.tif199141TIFF2025090743000173.tif199141TIFF2025090743000174.tif199141TIFF2025090743000175.tif199141TIFF2025090743000176.tif199141TIFF2025090743000177.tif199141TIFF2025090743000178.tif199141TIFF2025090743000179.tif199141TIFF2025090743000180.tif199141TIFF2025090743000181.tif199141TIFF2025090743000182.tif199141TIFF2025090743000183.tif199141TIFF2025090743000184.tif199141.

Claims

[Claim 1] A bispecific binding protein that binds to OX40 and PD-L1, comprising a first peptide having the amino acid sequence of SEQ ID NO: 35 and a second peptide having the amino acid sequence of SEQ ID NO: 32.

Images