4-1BB and OX40 binding proteins and related compositions and methods, antibodies to 4-1BB, antibodies to OX40

Abstract

The present disclosure provides antibodies that specifically bind to 4-1BB and / or OX40, including bispecific antibodies that bind to 4-1BB and OX40, and compositions comprising such antibodies. Also provided are methods for treating disorders such as cancer using such antibodies and compositions.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims the benefit of U.S. Provisional Application Nos. 62 / 885,751, filed Aug. 12, 2019; 62 / 902,318, filed Sep. 18, 2019; 62 / 911,010, filed Oct. 4, 2019; and 63 / 056,115, filed Jul. 24, 2020, each of which is hereby incorporated by reference in its entirety.

[0002] Reference to a Sequence Listing Submitted Electronically via EFS - WEB The content of the sequence listing submitted electronically (name: 4897_004PC04_Seqlisting_ST25.txt; size: 366,599 bytes; and creation date: Aug. 12, 2020) is hereby incorporated by reference in accordance with 37 C.F.R.§1.52(e)(5).

[0003] The present disclosure relates to antibodies, particularly bispecific antibodies that bind to 4 - 1BB and OX40, and compositions containing them, which are useful for treating diseases, including solid tumor cancers, to enhance the immune response.

Background Art

[0004] 4-1BB (CD137) and OX40 are members of the TNF receptor (TNFR) family (Bremer, ISRN Oncol.: 371854 (2013)). These receptors are not constitutively present on naive T cells or NK cells: their expression is induced by T cell stimulation through the T cell receptor (TCR) or other stimuli in NK cells. 4-1BB is mainly upregulated in CD8 T cells and NK cells, while OX40 is mainly upregulated in CD4 T cells. The function of these receptors is to provide co-stimulatory signals to T cells and NK cells. Activation of these receptors is naturally induced by trimerization through interaction with the 4-1BB ligand (4-1BBL) or OX40 ligand (OX40L) trimers, leading to signal transduction and initiation of specific cellular functions. 4-1BB enhances the effector functions of CD8 T cells and NK cells through increased expression of IFN-γ, granzyme, and anti-apoptosis genes, resulting in the generation of more and better effector CD8 T cells and NK cells. OX40 enhances the effector functions of CD4 T cells by enhancing their ability to generate IL-2 and clonal expansion of memory CD4 T cells.

[0005] 4-1BB and OX40 are often expressed on tumor-infiltrating lymphocytes, and indeed, their expression has been used to identify tumor-specific T cells. Human solid tumors are often infiltrated by lymphocytes, mainly CD8+ T cells and CD4+ T cells. The accumulation of tumor-infiltrating lymphocytes is often associated with improved survival among patients affected by various malignancies. (Ye et al., OncoImmunology 2: e27184 (2013); Montler et al., Clin Transl Immunology 5: e70 (2016)).

[0006] Trimerization of the 4-1BB receptor and the OX40 receptor can be induced via monoclonal antibodies. In some published examples, monoclonal antibodies have been developed to bind to Fc gamma receptors through their Fc regions and induce signaling by inducing higher-order clustering of the receptors (Mayes et al., Nature Reviews Drug Discovery 17:509 (2018)).

[0007] Preclinical results in various induced spontaneous tumor models suggest that targeting 4-1BB with agonist antibodies may lead to tumor clearance and durable antitumor immunity. Urelumab and utomilumab are agonist anti-4-1BB monoclonal antibodies, and clinical trials are ongoing in indications including the treatment of solid tumors. Despite early signs of efficacy, the clinical development of urelumab has been halted due to inflammatory hepatotoxicity at doses above 1 mg / kg. Utomilumab is less potent than urelumab but has an improved safety profile compared to urelumab (Chester et al., Blood 131:39-57 (2018)). There is a need for an effective treatment that targets 4-1BB without causing hepatotoxicity as observed with urelumab or other systemic injuries.

[0008] OX40 agonists have been reported to increase T cell infiltration into tumors. Another advantage of targeting OX40 is that OX40 signaling can block Treg-mediated suppression of the anti-tumor immune response. In several preclinical mouse cancer models, including 4T-1 breast cancer, B16 melanoma, Lewis lung carcinoma, and some chemically induced sarcomas, injection of OX40 agonists has resulted in a therapeutic response. (Ohsima et al., J. Immunology 159:3838-3848 (1997); Imura et al., J. Exp. Med. 183:2185-2195 (1996); Maxwell et al., J. Immunology 164:107-112 (2000); Gough et al., J. Immunotherapy 33(8):798-809 (2010)).

[0009] Mouse anti-human OX40 mAb (clone 9B12) was the first OX40 agonist reagent tested in a clinical trial of 30 patients with progressive solid tumors. In this Phase I trial, no patients showed an objective response according to RECIST criteria, but some immune responses, such as Ki67 staining by CD4+ T cells and CD8+ T cells that had experienced the antigen in the blood, were increased, indicating enhanced T cell activation. In addition, upregulation of OX40 by tumor-infiltrating Tregs was detected. Overall, the agonist anti-OX40 mAb 9B12 showed sufficient tolerance with mild to moderate side effects. (Curti et al., Cancer Res. 73(24):7189-7198 (2013)).

[0010] In some cases, researchers generated protein constructs containing multiple binding domains (>2) for 4-1BB or OX40, or fusions of multiple OX40L and 4-1BBL extracellular domains to induce agonism. In other examples, there are bispecific proteins containing binding domains (including plural ones) for 4-1BB or OX40 and binding domains (including plural ones) for tumor-specific antigens. Binding and clustering via tumor antigen binding induces clustering and signaling of 4-1BB and OX40. However, none of these constructs are expected to stimulate the functions of tumor-infiltrating lymphocytes, namely CD8+ T cells, CD4+ T cells, and NK cells, and minimize or eliminate off-target activation of effector cells (i.e., activation through binding to FcγR1, FcγRIIa, FcγRIIb, FcγRIIa, and FcγRIIIb). Therefore, a bispecific antibody that binds to and stimulates both 4-1BB and OX40 is required to selectively enhance the activity of tumor-infiltrating lymphocytes (with minimal or no effect on circulating lymphocytes).

Summary of the Invention

[0011] As shown herein, bispecific proteins that bind to 4-1BB and OX40 (e.g., ADAPTIR™ bispecific antibodies) act by binding to one receptor and inducing signaling of the other receptor, and vice versa. Conveniently, this elicits agonism of both receptors using a single therapeutic protein. The Fc region of the bispecific protein may contain modifications that eliminate binding to Fc gamma receptors and related complement proteins, such that the activity of the bispecific protein is strictly dependent on the presence of both receptors, either on the same cell or on different cells. Activity is not observed in the absence of one or both receptors. Importantly, the bispecific constructs provided herein result in a dose-dependent increase in the proliferation of T cells and NK cells, whereas combinations of monospecific constructs targeting 4-1BB and OX40 cannot do so.

[0012] In certain cases, the bispecific antibodies provided herein comprise a polypeptide that, in order from amino terminus to carboxyl terminus, (i) a first single-chain variable fragment (scFv), (ii) optionally, a linker that is a hinge region, (iii) an immunoglobulin constant region, and (iv) a second scFv, wherein (a) the first scFv comprises a human 4-1BB antigen-binding domain and the second scFv comprises a human OX40 antigen-binding domain, or (b) the first scFv comprises a human OX40 antigen-binding domain and the second scFv comprises a human 4-1BB antigen-binding domain.

[0013] In certain cases, the antibodies provided herein comprise a human 4-1BB antigen-binding domain, and the 4-1BB antigen-binding domain competitively inhibits the binding of an antibody comprising a heavy-chain variable domain (VH) comprising SEQ ID NO: 17 and a light-chain variable domain (VL) comprising SEQ ID NO: 18 to human 4-1BB.

[0014] In certain cases, the antibodies provided herein comprise a human 4-1BB antigen-binding domain, and the 4-1BB antigen-binding domain specifically binds to the same epitope of human 4-1BB as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18.

[0015] In certain cases, the antibodies provided herein comprise a human 4-1BB antigen-binding domain, and the human 4-1BB antigen-binding domain comprises six complementarity-determining regions (CDRs) in a VH of SEQ ID NO: 17 and a VL of SEQ ID NO: 18, or six CDRs in a VH of SEQ ID NO: 19 and a VL of SEQ ID NO: 20.

[0016] In certain cases, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs.

[0017] In certain cases, the antibodies provided herein comprise a human 4-1BB antigen-binding domain, and the human 4-1BB antigen-binding domain comprises a VH and a VL, and the VH comprises the amino acid sequence of SEQ ID NO: 17.

[0018] In certain cases, the antibodies provided herein comprise a human 4-1BB antigen-binding domain, and the human 4-1BB antigen-binding domain comprises a VH and a VL, and the VL comprises the amino acid sequence of SEQ ID NO: 18.

[0019] In certain cases, the antibodies provided herein comprise a human OX40 antigen-binding domain, and the OX40 antigen-binding domain competitively inhibits the binding of an antibody comprising a VH comprising SEQ ID NO: 29 and a VL comprising SEQ ID NO: 28 to human OX40.

[0020] In certain cases, the antibodies provided herein comprise a human OX40 antigen-binding domain, and the OX40 antigen-binding domain specifically binds to the same epitope of human OX40 as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28.

[0021] In certain cases, the antibodies provided herein comprise a human OX40 antigen-binding domain, and the human OX40 antigen-binding domain comprises six CDRs in VH of SEQ ID NO: 29 and VL of SEQ ID NO: 28.

[0022] In certain cases, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs.

[0023] In certain cases, the antibodies provided herein comprise a human OX40 antigen-binding domain, and the human OX40 antigen-binding domain comprises VH and VL, and VH comprises the amino acid sequence of SEQ ID NO: 29.

[0024] In certain cases, the antibodies provided herein comprise a human OX40 antigen-binding domain, and the human OX40 antigen-binding domain comprises VH and VL, and VL comprises the amino acid sequence of SEQ ID NO: 28.

[0025] In certain cases, the antibody is monospecific.

[0026] In certain cases, the antibody is an IgG antibody. In certain cases, the IgG antibody is an IgG 1 antibody.

[0027] In certain cases, the antibody further comprises a heavy chain constant region and a light chain constant region, and optionally, the heavy chain constant region is a human IgG 1 heavy chain constant region, and optionally, the light chain constant region is a human IgGκ light chain constant region.

[0028] In certain cases, the antibody is a single-chain Fv (scFv). In certain cases, the antibody comprises Fab, Fab’, F(ab’) 2 , scFv, disulfide-linked Fv, or scFv-Fc.

[0029] In certain cases, an antibody comprising a 4-1BB binding domain is bispecific. In certain cases, the bispecific antibody comprises a human OX40 antigen-binding domain. In certain cases, the human OX40 antigen-binding domain: (a) competitively inhibits the binding of an antibody comprising a VH comprising SEQ ID NO: 29 and a VL comprising SEQ ID NO: 28 to human OX40; (b) specifically binds to the same epitope of human OX40 as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (c) comprises six CDRs in the VH of SEQ ID NO: 29 and the VL of SEQ ID NO: 28, and optionally, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs; (d) comprises a VH and a VL, the VH comprising the amino acid sequence of SEQ ID NO: 29, and / or (e) comprises a VH and a VL, the VL comprising the amino acid sequence of SEQ ID NO: 28.

[0030] In certain cases, an antibody comprising an OX40 binding domain is bispecific. In certain cases, the bispecific antibody comprises a human 4-1BB antigen-binding domain. In certain cases, the human 4-1BB antigen-binding domain: (a) competitively inhibits the binding of an antibody comprising a VH comprising SEQ ID NO: 17 and a VL comprising SEQ ID NO: 18 to human 4-1BB; (b) specifically binds to the same epitope of human 4-1BB as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18; (c) comprises six CDRs in the VH of SEQ ID NO: 17 and the VL of SEQ ID NO: 18 or six CDRs in the VH of SEQ ID NO: 19 and the VL of SEQ ID NO: 20, and optionally, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs; (d) comprises a VH and a VL, the VH comprising the amino acid sequence of SEQ ID NO: 17, and / or (e) comprises a VH and a VL, the VL comprising the amino acid sequence of SEQ ID NO: 18.

[0031] In certain cases, the bispecific antibodies provided herein comprise (a) a human 4-1BB antigen-binding domain and (b) a human OX40 antigen-binding domain, wherein the 4-1BB antigen-binding domain comprises (i) a VH-CDR1 comprising the amino acid sequence of GYTFTSYW (SEQ ID NO: 5); (ii) a VH-CDR2 comprising the amino acid sequence of IYPGSSTT (SEQ ID NO: 6); (iii) a VH-CDR3 comprising the amino acid sequence of ASFSDGYYAYAMDY (SEQ ID NO: 7); (iv) a light chain variable domain (VL)-CDR1 comprising the amino acid sequence of QDISNY (SEQ ID NO: 8); (v) a VL-CDR2 comprising the amino acid sequence of YTS (SEQ ID NO: 9); and (vi) a VL-CDR3 comprising the amino acid sequence of QQGYTLPYT (SEQ ID NO: 10), and the OX40 antigen-binding domain comprises (i) a VH-CDR1 comprising the amino acid sequence of GFTLSYYG (SEQ ID NO: 11); (ii) a VH-CDR2 comprising the amino acid sequence of ISHDGSDK (SEQ ID NO: 12); (iii) a VH-CDR3 comprising the amino acid sequence of SNDQFDP (SEQ ID NO: 13); (iv) a VL-CDR1 comprising the amino acid sequence of NIGSKS (SEQ ID NO: 14); (v) a VL-CDR2 comprising the amino acid sequence of DDS (SEQ ID NO: 15); and (vi) a VL-CDR3 comprising the amino acid sequence of QVWDSSSDHVV (SEQ ID NO: 16).

[0032] In certain cases of the bispecific antibodies provided herein, the human 4-1BB antigen-binding domain (a) competitively inhibits the binding of an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18 to human 4-1BB, (b) specifically binds to the same epitope of human 4-1BB as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18, (c) comprises six CDRs in the VH of SEQ ID NO: 17 and the VL of SEQ ID NO: 18 or six CDRs in the VH of SEQ ID NO: 19 and the VL of SEQ ID NO: 20, and optionally, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs, (d) comprises a VH and a VL, wherein the VH comprises the amino acid sequence of SEQ ID NO: 17, and / or (e) comprises a VH and a VL, wherein the VL comprises the amino acid sequence of SEQ ID NO: 18.

[0033] In certain cases of the bispecific antibodies provided herein, the human OX40 antigen-binding domain: (a) competitively inhibits the binding of an antibody comprising a VH comprising SEQ ID NO: 29 and a VL comprising SEQ ID NO: 28 to human OX40; (b) specifically binds to the same epitope of human OX40 as an antibody comprising a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (c) comprises six CDRs in the VH of SEQ ID NO: 29 and the VL of SEQ ID NO: 28, and optionally, the CDRs are IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs; (d) comprises a VH and a VL, the VH comprising the amino acid sequence of SEQ ID NO: 29; and / or (e) comprises a VH and a VL, the VL comprising the amino acid sequence of SEQ ID NO: 28.

[0034] In certain cases, the human 4-1BB binding domain comprises a VH comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 17, 19, 21, 23, 32, and 143. In certain cases, the human 4-1BB binding domain comprises a VH comprising the amino acid sequence of any one of SEQ ID NOs: 17, 19, 21, 23, 32, and 143.

[0035] In certain cases, the human 4-1BB binding domain comprises a VL comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 18, 20, 22, and 24. In certain cases, the human 4-1BB binding domain comprises a VL comprising the amino acid sequence of any one of SEQ ID NOs: 18, 20, 22, and 24.

[0036] In certain cases, the human 4-1BB binding domain comprises (a) VH comprising the amino acid sequence of SEQ ID NO: 17 and VL comprising the amino acid sequence of SEQ ID NO: 18, (b) VH comprising the amino acid sequence of SEQ ID NO: 19 and VL comprising the amino acid sequence of SEQ ID NO: 20, (c) VH comprising the amino acid sequence of SEQ ID NO: 21 and VL comprising the amino acid sequence of SEQ ID NO: 22, (d) VH comprising the amino acid sequence of SEQ ID NO: 23 and VL comprising the amino acid sequence of SEQ ID NO: 24, (e) VH comprising the amino acid sequence of SEQ ID NO: 32 and VL comprising the amino acid sequence of SEQ ID NO: 18, or (f) VH comprising the amino acid sequence of SEQ ID NO: 143 and VL comprising the amino acid sequence of SEQ ID NO: 20.

[0037] In certain cases, the human 4-1BB binding domain comprises VH comprising the amino acid sequence of SEQ ID NO: 17 and VL comprising the amino acid sequence of SEQ ID NO: 18.

[0038] In certain cases, the human 4-1BB binding domain comprises VH and VL on the same polypeptide chain. In certain cases, the VH of the human 4-1BB binding domain is N-terminal to the VL of the human 4-1BB binding domain. In certain cases, the VH of the human 4-1BB binding domain is C-terminal to the VL of the human 4-1BB binding domain. In certain cases, the human 4-1BB binding domain comprises a linker between VH and VL. In certain cases, the linker comprises the amino acid (Gly 4 Ser) n and n = 1-5 (SEQ ID NO: 117). In certain cases, n = 3-5 or n = 4-5. In certain cases, n = 4.

[0039] In certain cases, the human 4-1BB binding domain comprises an scFv comprising any one of the amino acid sequences of SEQ ID NOs: 42, 44, 58, 63, 77, and 145.

[0040] In certain cases, the human 4-1BB binding domain comprises an scFv comprising the amino acid sequence of SEQ ID NO: 58.

[0041] In certain cases, the human 4-1BB binding domain can bind to cynomolgus 4-1BB.

[0042] In certain cases, the human 4-1BB binding domain can stimulate human 4-1BB activity.

[0043] In certain cases, the human 4-1BB binding domain comprises humanized VH and VL sequences.

[0044] In certain cases, the human OX40 binding domain comprises a VH comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 25, 27, 29, 31, and 33. In certain cases, the human OX40 binding domain comprises a VH comprising an amino acid sequence of any one of SEQ ID NOs: 25, 27, 29, 31, and 33.

[0045] In certain cases, the human OX40 binding domain comprises a VL comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 26, 28, 30, and 34-41. In certain cases, the human OX40 binding domain comprises a VL comprising an amino acid sequence of any one of SEQ ID NOs: 26, 28, 30, and 34-41.

[0046] In certain cases, the human OX40 binding domain comprises (a) a VH comprising the amino acid sequence of SEQ ID NO: 25 and a VL comprising the amino acid sequence of SEQ ID NO: 26, (b) a VH comprising the amino acid sequence of SEQ ID NO: 27 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (c) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 26, (d) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (e) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (f) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (g) a VH comprising the amino acid sequence of SEQ ID NO: 33 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (h) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 34, (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35, (j) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 36, (k) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 37, (l) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 34, (m) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 35, (n) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 36, (o) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 37, (p) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 38, (q) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 39, (r) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 40, or (s) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 41.

[0047] In certain cases, the human OX40 binding domain comprises (a) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (b) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, or (c) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35.

[0048] In certain cases, the human OX40 binding domain comprises VH and VL in the same polypeptide chain. In certain cases, the VH of the human OX40 binding domain is N-terminal to the VL of the human OX40 binding domain. In certain cases, the VH of the human OX40 binding domain is C-terminal to the VL of the human OX40 binding domain. In certain cases, the human OX40 binding domain comprises a linker between VH and VL. In certain cases, the linker comprises the amino acid sequence (Gly 4 Ser) n where n = 1-5 (SEQ ID NO: 117). In certain cases, n = 3-5. In certain cases, n = 4.

[0049] In certain cases, the human OX40 binding domain comprises a scFv comprising any one of the amino acid sequences of SEQ ID NOs: 46, 47, 52, 54, 56, 59-62, 64-76, and 146. In certain cases, the human OX40 binding domain comprises a scFv comprising any one of the amino acid sequences of SEQ ID NOs: 59, 62, or 66.

[0050] In certain cases, the human OX40 binding domain is capable of binding to cynomolgus OX40.

[0051] In certain cases, the human OX40 binding domain is capable of stimulating human OX40 activity.

[0052] In certain cases, the human OX40 binding domain comprises mouse or rat VH and VL sequences.

[0053] In certain cases, the human 4-1BB binding domain comprises a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18, and the human OX40 binding domain comprises (a) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (b) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, or (c) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35.

[0054] In certain cases, the human 4-1BB binding domain comprises a scFv comprising the amino acid sequence of SEQ ID NO: 58, and the human OX40 binding domain comprises a scFv comprising any one of the amino acid sequences of SEQ ID NO: 59, 62, or 66.

[0055] In certain cases, the human 4-1BB binding domain and the human OX40 binding domain are on the same polypeptide. In certain cases, the human 4-1BB binding domain is N-terminal to the human OX40 binding domain. In certain cases, the human 4-1BB binding domain is C-terminal to the human OX40 binding domain.

[0056] In certain cases, the antibody comprises an immunoglobulin constant region. In certain cases, the immunoglobulin constant state comprises the immunoglobulin CH2 and CH3 domains of IgG1, IgG2, IgG3, IgG4, IgA1, IgA2, or IgD. In certain cases, the immunoglobulin constant region comprises the immunoglobulin CH2 and CH3 domains of IgG1. In certain cases, the antibody does not comprise a CH1 domain.

[0057] In certain cases, the immunoglobulin constant region contains one, two, three or more amino acid substitutions compared to the wild-type immunoglobulin constant region to prevent binding to FcγR1, FcγRIIa, FcγRIIb, FcγRIIa, and FcγRIIIb. In certain cases, the immunoglobulin constant region contains one, two, three or more amino acid substitutions compared to the wild-type immunoglobulin constant region to prevent or reduce Fc-mediated T cell activation. In certain cases, the immunoglobulin constant region contains one, two, three or more amino acid substitutions compared to the wild-type immunoglobulin constant region to prevent or reduce CDC activation. In certain cases, the immunoglobulin constant region contains one, two, three or more amino acid substitutions compared to the wild-type immunoglobulin constant region to prevent or reduce ADCC activity. In certain cases, the immunoglobulin constant region contains a human IgG1 CH2 domain comprising the substitutions E233P, L234A, L235A, G237A, and K322A and the deletion of G236 according to the EU numbering system.

[0058] In certain cases, the antibody contains a linker between the immunoglobulin constant region and the human 4-1BB binding domain and / or between the immunoglobulin constant region and the human OX40 binding domain. In certain cases, the linker between the immunoglobulin constant region and the human 4-1BB binding domain and / or between the immunoglobulin constant region and the human OX40 binding domain contains 10-30 amino acids, 15-30 amino acids, or 20-30 amino acids. In certain cases, the linker between the immunoglobulin constant region and the human 4-1BB binding domain or between the immunoglobulin constant region and the human OX40 binding domain contains the amino acid sequence (Gly 4 Ser) n where n = 1-5 (SEQ ID NO: 117). In certain cases, n = 1.

[0059] In certain cases, the antibody comprises a dimer of two polypeptides, each polypeptide comprising, in order from amino terminus to carboxyl terminus, a first scFv, a hinge region, an immunoglobulin constant region, and a second scFv, wherein (a) the first scFv comprises a human 4-1BB antigen-binding domain and the second scFv comprises a human OX40 antigen-binding domain, or (b) the first scFv comprises a human OX40 antigen-binding domain and the second scFv comprises a human 4-1BB antigen-binding domain. In certain cases, the dimer is a homodimer.

[0060] In certain cases, the first scFv comprises a human 4-1BB binding domain and the second scFv comprises a human OX40 antigen-binding domain.

[0061] In certain cases, the hinge is an IgG 1 hinge. In certain cases, the hinge comprises amino acids 1-15 of SEQ ID NO: 115.

[0062] In certain cases, the hinge and the immunoglobulin constant region comprise the amino acid sequence of SEQ ID NO: 115.

[0063] In certain cases, the antibody comprises a linker between the immunoglobulin constant region and the human OX40 binding domain, the linker comprising the amino acid sequence (Gly 4 Ser) n where n = 1-5 (SEQ ID NO: 117). In certain cases, n = 1.

[0064] In certain cases, the human 4-1BB binding domain comprises a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18, and the human OX40 binding domain comprises (a) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (b) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, or (c) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35.

[0065] In certain cases, the human 4-1BB binding domain comprises the amino acid sequence of SEQ ID NO: 58, and the human OX40 binding domain comprises any one of the amino acid sequences of SEQ ID NO: 59, 62, or 66.

[0066] In certain cases, the bispecific antibodies provided herein comprise a human 4-1BB antigen-binding domain and a human OX40 antigen-binding domain, and the antibody comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 78-100 and 144. In certain cases, the antibody is a homodimer comprising two polypeptides, and each polypeptide comprises the same amino acid sequence selected from the group consisting of SEQ ID NOs: 78-100 and 144.

[0067] In certain cases, the bispecific antibodies provided herein comprise a human 4-1BB antigen-binding domain and a human OX40 antigen-binding domain, and the antibody comprises the amino acid sequence of SEQ ID NO: 78. In certain cases, the antibody is a homodimer comprising two polypeptides, and each polypeptide comprises the amino acid sequence of SEQ ID NO: 78.

[0068] In certain cases, the bispecific antibodies provided herein comprise a human 4-1BB antigen-binding domain and a human OX40 antigen-binding domain, and the antibody comprises the amino acid sequence of SEQ ID NO: 81. In certain cases, the antibody is a homodimer comprising two polypeptides, and each polypeptide comprises the amino acid sequence of SEQ ID NO: 81.

[0069] In certain cases, the bispecific antibodies provided herein comprise a human 4-1BB antigen-binding domain and a human OX40 antigen-binding domain, and the antibody comprises the amino acid sequence of SEQ ID NO: 90. In certain cases, the antibody is a homodimer comprising two polypeptides, and each polypeptide comprises the amino acid sequence of SEQ ID NO: 90.

[0070] In certain cases, the human 4-1BB binding domain and the human OX40 binding domain are on separate polypeptides. In certain cases, the human 4-1BB binding domain contains VH and VL on a separate polypeptide. In certain cases, the human OX40 binding domain contains VH and VL on a separate polypeptide.

[0071] In certain cases, the antibody is a knob-in-hole (KIH) antibody, an IgG1 antibody containing mutations matched in the CH3 domain, two modified Fv fragments with exchanged VH, a diabody, scFv×scFv, scFv-Fc-scFv, a quadroma, a CrossMab Fab, a CrossMab VH-VL, or a strand-exchange engineered domain body (SEEDbody).

[0072] In certain cases, the antibody can bind to human 4-1BB and human OX40 simultaneously.

[0073] In certain cases, the antibody is CD8 + T, CD4 + T, and / or can promote the dose-dependent expansion of NK cells.

[0074] In certain cases, the antibody can increase the secretion of IFN-γ, IL-2, and / or TNF-α from stimulated PBMCs.

[0075] In certain cases, the antibody is agonistic to human 4-1BB and human OX40.

[0076] In certain cases, the antibody is isolated.

[0077] In certain cases, the antibody is a monoclonal antibody.

[0078] In certain cases, the antibody further comprises a detectable label.

[0079] In certain cases, the polynucleotides provided herein encode the antibodies provided herein. In certain cases, the vectors provided herein contain the polynucleotides provided herein, and optionally, the vector is an expression vector.

[0080] In certain cases, the host cells provided herein contain the polynucleotides provided herein or the vectors provided herein.

[0081] In certain cases, the host cells provided herein contain a combination of the polynucleotides provided herein that encode the antibodies provided herein. In certain cases, the polynucleotide is encoded by one vector. In certain cases, the polynucleotide is encoded by multiple vectors.

[0082] In certain cases, the host cell is selected from the group consisting of CHO, HEK293, or COS cells.

[0083] In certain cases, a method for producing an antibody that specifically binds to human 4-1BB and human OX40 provided herein includes culturing the host cells provided herein, thereby producing the antibody, and optionally further recovering the antibody.

[0084] In certain cases, a method for detecting 4-1BB and OX40 in a sample provided herein includes contacting the sample with an antibody according to any one of claims 1 to 106, and optionally, the sample contains cells.

[0085] In certain cases, the pharmaceutical composition provided herein contains the antibody provided herein and a pharmaceutically acceptable excipient.

[0086] In certain cases, the methods provided herein for enhancing the proliferation of NK cells include contacting the NK cells with an antibody provided herein or a pharmaceutical composition provided herein.

[0087] In certain cases, the methods provided herein for enhancing the proliferation of T cells include contacting the T cells with an antibody provided herein or a pharmaceutical composition provided herein.

[0088] In certain cases, the methods provided herein for enhancing the proliferation of NK cells and T cells include contacting the NK cells and T cells with an antibody provided herein or a pharmaceutical composition provided herein.

[0089] In certain cases, the methods provided herein for stimulating the T cell co-stimulatory pathway include contacting the T cells with an antibody provided herein or a pharmaceutical composition provided herein.

[0090] In certain cases, the T cells are CD4+ T cells. In certain cases, the T cells are CD8+ T cells.

[0091] In certain cases, the cells are within a subject, and the contacting includes administering the antibody or pharmaceutical composition to the subject.

[0092] In certain cases, the methods provided herein for enhancing the immune response in a subject include administering to the subject an effective amount of an antibody provided herein or a pharmaceutical composition provided herein.

[0093] In certain cases, the methods provided herein for treating cancer in a subject include administering to the subject an effective amount of an antibody provided herein or a pharmaceutical composition provided herein. In certain cases, the cancer is selected from the group consisting of melanoma, renal cancer, pancreatic cancer, lung cancer, intestinal cancer, prostate cancer, breast cancer, liver cancer, brain cancer, or blood cancer.

[0094] In certain cases, the subject is a human.

Brief Description of the Drawings

[0095]

FIG. 1A - B

FIG. 2A - C

FIG. 3A - B

FIG. 4

FIG. 5A - B

FIG. 6

FIG. 7

FIG. 8A - B

FIG. 9

FIG. 10A - D

FIG. 11A - B

FIG. 12A - D

FIG. 13

FIG. 14A - D

FIG. 15A - B

FIG. 16

FIG. 17

FIG. 18A - B

FIG. 19

FIG. 20A - D

FIG. 21

FIG. 22A - D

FIG. 23A - B

FIG. 24

FIG. 25

FIG. 26

FIG. 27

FIG. 28A - B

FIG. 29

FIG. 30

FIG. 31

FIG. 32

FIG. 33

FIG. 34

Mode for Carrying Out the Invention

[0096] To facilitate the understanding of the present disclosure, some terms and phrases are defined below. I. Terms

[0097] As used herein, the term "4-1BB" refers to mammalian 4-1BB polypeptides, including but not limited to native 4-1BB polypeptides and isoforms of 4-1BB polypeptides. "4-1BB" encompasses full-length, unprocessed 4-1BB polypeptides as well as forms of 4-1BB polypeptides resulting from intracellular processing. As used herein, the term "CD137" is to be understood as being replaceable with the term "4-1BB". As used herein, the term "human 4-1BB" refers to a polypeptide comprising the amino acid sequence of SEQ ID NO: 1. As used herein, the term "cynomolgus 4-1BB" refers to a polypeptide comprising the amino acid sequence of SEQ ID NO: 2. "4-1BB polynucleotide", "4-1BB nucleotide", or "4-1BB nucleic acid" refers to a polynucleotide encoding 4-1BB.

[0098] As used herein, the term "OX40" refers to mammalian OX40 polypeptides, including but not limited to native OX40 polypeptides and isoforms of OX40 polypeptides. "OX40" encompasses full-length, unprocessed OX40 polypeptides as well as forms of OX40 polypeptides resulting from intracellular processing. As used herein, the term "human OX40" refers to a polypeptide comprising the amino acid sequence of SEQ ID NO: 3. As used herein, the term "cynomolgus OX40" refers to a polypeptide comprising the amino acid sequence of SEQ ID NO: 4. "OX40 polynucleotide", "OX40 nucleotide", or "OX40 nucleic acid" refers to a polynucleotide encoding OX40.

[0099] As used herein, the term "tumor infiltrating lymphocyte" or "TIL" refers to lymphocytes that directly oppose and / or surround tumor cells. Tumor infiltrating lymphocytes are typically non-circulating lymphocytes and include CD8+ T cells, CD4+ T cells, and NK cells. Tumor infiltrating lymphocytes can express OX40 and 4-1BB.

[0100] As used herein, the terms "antibody" and "antibodies" are technical terms and can be used interchangeably herein to refer to a molecule or complex of molecules having at least one antigen-binding site that specifically binds an antigen.

[0101] Antibodies can include, for example, monoclonal antibodies, recombinantly produced antibodies, human antibodies, humanized antibodies, reshaped antibodies, chimeric antibodies, immunoglobulins, synthetic antibodies, tetrameric antibodies comprising two heavy chains and two light chain molecules, antibody light chain monomers, antibody heavy chain monomers, antibody light chain dimers, antibody heavy chain dimers, antibody light chain-heavy chain pairs, intracellular antibodies, heteroconjugate antibodies, single domain antibodies, monovalent antibodies, single chain antibodies or single chain Fv (scFv), camelized antibodies, affibodies, Fab fragments, F(ab’) 2 fragments, disulfide-linked Fv (sdFv), anti-idiotype antibodies (anti-Id) (including, for example, anti-anti-Id antibodies), bispecific antibodies, and multispecific antibodies. In certain embodiments, the antibodies described herein refer to a polyclonal antibody population. Antibodies can be of any type of immunoglobulin molecule (e.g., IgG, IgE, IgM, IgD, IgA, or IgY), any class (e.g., IgG 1 IgG 2 IgG 3 IgG 4 IgA 1 or IgA 2 ), or any subclass (e.g., IgG 2a or IgG 2b ). In certain embodiments, the antibodies described herein are IgG antibodies, or a class thereof (e.g., human IgG 1 IgG 2 or IgG 4 ) or subclass. In a specific embodiment, the antibody is a humanized monoclonal antibody. In another specific embodiment, the antibody is a human monoclonal antibody that is, for example, an immunoglobulin. In certain embodiments, the antibodies described herein are IgG 1 IgG 2 or IgG 4It is an antibody.

[0102] A "bispecific" antibody is an antibody that has two different antigen-binding sites (excluding the Fc region) that bind to two different antigens. Bispecific antibodies can include, for example, recombinantly produced antibodies, human antibodies, humanized antibodies, reshaped antibodies, chimeric antibodies, immunoglobulins, synthetic antibodies, tetrameric antibodies comprising two heavy chains and two light chain molecules, antibody light chain monomers, heteroconjugate antibodies, linked single-chain antibodies or linked single-chain Fv (scFv), camelized antibodies, affibodies, linked Fab fragments, F(ab') 2 fragments, chemically linked Fv, and disulfide-linked Fv (sdFv). Bispecific antibodies can be of any type of immunoglobulin molecule (e.g., IgG, IgE, IgM, IgD, IgA, or IgY), any class (e.g., IgG 1 IgG 2 IgG 3 IgG 4 IgA 1 IgA 2 ), or any subclass (e.g., IgG 2a IgG 2b ). In certain embodiments, the bispecific antibodies described herein are IgG antibodies, or a class thereof (e.g., human IgG 1 IgG 2 IgG 4or a subclass thereof. In certain embodiments, the bispecific antibodies described herein comprise two polypeptides that are optionally identical polypeptides, each polypeptide comprising, in order from amino terminus to carboxyl terminus, a first scFv antigen-binding domain, a linker (optionally, the linker is a hinge region), an immunoglobulin constant region, and a second scFv antigen-binding domain. This particular type of antibody is exemplified by the ADAPTIR™ technology, which is shown in FIG. 27. The bispecific antibody can be, for example, monovalent for each target (e.g., an IgG molecule having one arm that targets one antigen and another arm that targets a second antigen) or bivalent for each target (e.g., a bispecific variable domain antibody, an IgG-scFv, an scFv-Fc-scFv, or an ADAPTIR™ antibody that comprises a dimer, wherein each polypeptide of the dimer comprises two different antigen-binding domains).

[0103] As used herein, the terms “antigen-binding domain,” “antigen-binding region,” “antigen-binding site,” and like terms refer to the portion of an antibody molecule that contains the amino acid residues that confer its specificity for an antigen (e.g., a complementarity determining region (CDR)) on the antibody molecule. The antigen-binding region can be derived from any animal species, such as rodents (e.g., mouse, rat, or hamster) and humans. An antigen-binding domain that binds to 4-1BB can be referred to herein, for example, as a “4-1BB binding domain.” An antigen-binding domain that binds to OX40 can be referred to herein, for example, as an “OX40 binding domain.” As used herein, a “human 4-1BB binding domain” or a “human 4-1BB antigen-binding domain” refers to an antigen-binding domain that specifically binds to human 4-1BB but may also bind to non-human 4-1BB (e.g., mouse, rodent, or non-human primate 4-1BB). Similarly, a “human OX40 binding domain” or a “human OX40 antigen-binding domain” refers to an antigen-binding domain that specifically binds to human OX40.

[0104] As used herein, the terms "4-1BB / OX40 antibody", "anti-4-1BB / OX40 antibody" or "4-1BB×OX40 antibody" refer to bispecific antibodies that include an antigen-binding domain that binds to 4-1BB (e.g., human 4-1BB) and an antigen-binding domain that binds to OX40 (e.g., human OX40).

[0105] A "monoclonal" antibody refers to a homogeneous population of antibodies that are involved in the highly specific recognition and binding of a single antigen determinant, or epitope. This is in contrast to polyclonal antibodies, which typically include different antibodies directed against different antigen determinants. The term "monoclonal" antibody encompasses both whole immunoglobulin molecules and full-length immunoglobulin molecules, as well as Fab, Fab’, F(ab)2, Fv, single-chain (scFv), fusion proteins containing antibody portions, and any other modified immunoglobulin molecule containing an antigen recognition site. Further, a "monoclonal" antibody refers to such antibodies made by any number of means including, but not limited to, hybridomas, phage selection, recombinant expression, and transgenic animals.

[0106] The term "chimeric" antibody refers to an antibody whose amino acid sequence is derived from two or more species. Typically, the variable regions of both the light and heavy chains correspond to the variable regions of an antibody derived from one species of mammal (e.g., mouse, rat, rabbit, etc.) having the desired specificity, affinity, and capacity, while the constant regions are homologous to the sequences in an antibody derived from another (usually human) species and avoid inducing an immune response in that species.

[0107] The term "humanized" antibody refers to the form of a non-human (e.g., murine) antibody that contains minimal non-human (e.g., murine) sequences. Typically, a humanized antibody is a human immunoglobulin in which the residues from the complementarity determining regions (CDRs) are replaced by residues from the CDRs of a non-human species (e.g., mouse, rat, rabbit, hamster) having the desired specificity, affinity, and capacity ("CDR grafting") (Jones et al., Nature 321:522-525 (1986); Riechmann et al., Nature 332:323-327 (1988); Verhoeyen et al., Science 239:1534-1536 (1988)). In some cases, residues of the Fv framework region (FR) of the human immunoglobulin are replaced with corresponding residues in an antibody from a non-human species having the desired specificity, affinity, and capacity. The humanized antibody can further be modified by substitution of additional residues either within the Fv framework region and / or within the replaced non-human residues to refine and optimize the specificity, affinity, and / or capacity of the antibody. Generally, a humanized antibody contains substantially all of the variable domain that contains all or substantially all of the CDR regions corresponding to a non-human immunoglobulin, and at least one, typically two or three, but all or substantially all of the FR regions are those of a human immunoglobulin consensus sequence. A humanized antibody can also contain an immunoglobulin constant region or domain (Fc), typically at least a portion of a human immunoglobulin. Examples of methods used to generate humanized antibodies are described in U.S. Patent No. 5,225,539; Roguska et al., Proc. Natl. Acad. Sci., USA, 91(3):969-973 (1994), and Roguska et al., Protein Eng. 9(10):895-904 (1996).

[0108] The term "human" antibody means an antibody having an amino acid sequence derived from a human immunoglobulin locus, and such antibodies are made using any technique known in the art.

[0109] The variable regions typically differ greatly in sequence between antibodies, generally being part of the light and heavy chains, typically the amino-terminal approximately 110 to 125 amino acids in the mature heavy chain and approximately 90 to 115 amino acids in the mature light chain, which are used in the binding and specificity of a particular antibody for a particular antigen. The variability in sequence is concentrated in regions called complementarity-determining regions (CDRs), and the more highly conserved regions within the variable domains are called framework regions (FRs). Without being bound by a particular mechanism or theory, the CDRs of the light and heavy chains are thought to be primarily responsible for the interaction and specificity of the antibody with the antigen. In some embodiments, the variable region is a human variable region. In some embodiments, the variable region comprises rodent or mouse CDRs and human framework regions (FRs). In certain embodiments, the variable region is a primate (e.g., non-human primate) variable region. In some embodiments, the variable region comprises rodent or mouse CDRs and primate (e.g., non-human primate) framework regions (FRs).

[0110] The terms “VH” and “VH domain” are used interchangeably to refer to the variable region of the heavy chain of an antibody.

[0111] The terms “VL” and “VL domain” are used interchangeably to refer to the variable region of the light chain of an antibody.

[0112] The terms "Kabat numbering" and similar terms are recognized in the art and refer to a system for numbering the amino acid residues in the variable regions of the heavy and light chains of an antibody, or in its antigen-binding portion. In one aspect, the CDRs of an antibody can be determined according to the Kabat numbering system (see, e.g., Kabat EA & Wu TT (1971) Ann NY Acad Sci 190:382-391 and Kabat EA et al., (1991) Sequences of Proteins of Immunological Interest, Fifth Edition, U.S. Department of Health and Human Services, NIH Publication No. 91-3242). Using the Kabat numbering system, the CDRs within the antibody heavy chain molecule are typically, optionally, amino acids 31 to 35 (CDR1), which can include one or two additional amino acids (referred to as 35A and 35B in the Kabat numbering scheme) following 35, amino acids 50 to 65 (CDR2), and amino acids 95 to 102 (CDR3). Using the Kabat numbering system, the CDRs within the antibody light chain molecule are typically amino acids 24 to 34 (CDR1), amino acids 50 to 56 (CDR2), and amino acids 89 to 97 (CDR3). In a specific embodiment, the CDRs of the antibodies described herein are determined according to the Kabat numbering scheme.

[0113] Chothia (Chothia and Lesk, J. Mol. Biol. 196:901-917 (1987)) instead refers to the position of the structural loop. The ends of the Chothia CDR-H1 loop numbered using the Kabat numbering convention vary between H32 and H34 depending on the length of the loop (this is because the Kabat numbering scheme positions insertions at H35A and H35B; if neither 35A nor 35B is present, the loop ends at 32; if only 35A is present, the loop ends at 33; if both 35A and 35B are present, the loop ends at 34). In a specific embodiment, the CDRs of the antibodies described herein are determined according to the Chothia numbering scheme.

[0114] The AbM hypervariable regions represent an intermediate between the Kabat CDRs and the Chothia structural loops and are used by Oxford Molecular's AbM antibody modeling software. In a specific embodiment, the CDRs of the antibodies described herein are determined according to the AbM numbering scheme.

[0115] [Table 1]

[0116] The IMGT numbering convention is described in Brochet, X, et al, Nucl. Acids Res. 36:W503-508 (2008). In a specific embodiment, the CDRs of the antibodies described herein are determined according to the IMGT numbering convention. As used herein, unless otherwise specified, the positions of amino acid residues in the variable region of an immunoglobulin molecule are numbered according to the IMGT numbering convention.

[0117] As used herein, the terms "constant region" or "constant domain" are interchangeable and have a meaning common in the art. A constant region is the carboxyl-terminal portion of a light and / or heavy chain that is not directly involved in binding of the antibody to an antigen, e.g., an antibody, but can exhibit various effector functions such as interaction with an Fc receptor. The constant regions of immunoglobulin molecules generally have more conserved amino acid sequences compared to immunoglobulin variable domains. An immunoglobulin "constant region" or "constant domain" can include the CH1 domain, hinge, CH2 domain, and CH3 domain or a subset of these domains, e.g., the CH2 domain and CH3 domain. In certain embodiments provided herein, the immunoglobulin constant region does not include the CH1 domain. In certain embodiments provided herein, the immunoglobulin constant region does not include the hinge. In certain embodiments provided herein, the immunoglobulin constant region includes the CH2 domain and CH3 domain.

[0118] The "Fc region" or "Fc domain" refers to a polypeptide sequence corresponding to or derived from the portion of an antibody source that is involved in binding to antibody receptors on cells and the C1q component of complement. Fc is an abbreviation for "fragment crystallizable" and refers to a fragment of an antibody that readily forms protein crystals. Originally, discrete protein fragments, which have been explained by proteolytic digestion, can define the overall general structure of an immunoglobulin protein. The "Fc region" or "Fc domain" includes the CH2 domain, CH3 domain, and optionally all or part of the hinge. The "Fc region" or "Fc domain" can refer to a single polypeptide or two disulfide-bonded polypeptides. For a review of immunoglobulin structure and function, see Putnam, The Plasma Proteins, Vol. V (Academic Press, Inc., 1987), pp. 49-140; and Padlan, Mol. Immunol. 31:169-217, 1994. As used herein, the term Fc includes modifications of naturally occurring sequences.

[0119] As used herein, an "immunoglobulin dimerization domain" or "immunoglobulin heterodimerization domain" refers to an immunoglobulin domain of a polypeptide chain that preferentially interacts or associates with different immunoglobulin domains of a second polypeptide chain, and the interaction of different immunoglobulin heterodimerization domains substantially contributes to or efficiently promotes the heterodimerization (i.e., the formation of a dimer between two different polypeptide chains, also referred to as a "heterodimer") of the first polypeptide chain and the second polypeptide chain. The interaction between immunoglobulin heterodimerization domains "substantially contributes to or efficiently promotes" the heterodimerization of the first polypeptide chain and the second polypeptide chain if there is a statistically significant reduction in the dimerization between the first polypeptide chain and the second polypeptide chain in the absence of the immunoglobulin heterodimerization domain of the first polypeptide chain and / or the immunoglobulin heterodimerization domain of the second polypeptide chain. In certain embodiments, when the first polypeptide chain and the second polypeptide chain are co-expressed, at least 60%, at least about 60% to about 70%, at least about 70% to about 80%, at least 80% to about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% of the first polypeptide chain and the second polypeptide chain form heterodimers with each other. Representative immunoglobulin heterodimer domains include, as provided therein, immunoglobulin CH1 domains, immunoglobulin CL domains (e.g., Cκ or Cλ isotypes), or derivatives thereof that include wild-type immunoglobulin CH1 and CL domains and modified (or mutated) immunoglobulin CH1 and CL domains.

[0120] The "wild-type immunoglobulin hinge region" refers to the native upper hinge and central hinge amino acid sequences that are inserted between and connect the CH1 and CH2 domains (for IgG, IgA, and IgD) or the CH1 and CH3 domains (for IgE and IgM) that are identified in the heavy chains of naturally occurring antibodies. In certain embodiments, the wild-type immunoglobulin hinge region sequence is human and can include the human IgG hinge region. A "modified wild-type immunoglobulin hinge region" or "modified immunoglobulin hinge region" is (a) a wild-type immunoglobulin hinge region having up to 30% amino acid changes (e.g., up to 25%, 20%, 15%, 10%, or 5% amino acid substitutions or deletions), or (b) having a length of from about 5 amino acids (e.g., about 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acids) to up to about 120 amino acids (e.g., having a length of from about 10 to about 40 amino acids or from about 15 to about 30 amino acids or from about 15 to about 20 amino acids or from about 20 to about 25 amino acids) and having up to about 30% amino acid changes (e.g., up to about 25%, 20%, 15%, 10%, 5%, 4%, 3%, 2%, or 1% amino acid substitutions or deletions or combinations thereof), and refers to a portion of a wild-type immunoglobulin hinge region having an IgG core hinge region as disclosed in U.S. Patent Publication No. 2013 / 0129723 and U.S. Patent Publication No. 2013 / 0095097. As provided herein, a "hinge region" or "hinge" can be located between an antigen-binding domain (e.g., 4-1BB or OX40 binding domain) and an immunoglobulin constant region.

[0121] As used herein, "linker" refers to a polypeptide that can link moieties, such as two compounds, such as polypeptides. Non-limiting examples of linkers include flexible linkers comprising glycine-serine (e.g., (Gly4Ser)) repeats, and (a) the interdomain region of a transmembrane protein (e.g., a type I transmembrane protein); (b) the stalk region of a type II C-type lectin; or (c) a linker derived from an immunoglobulin hinge. As provided herein, a linker can refer to, for example, (1) the polypeptide region between the VH and VL regions in a single-chain Fv (scFv) or (2) the polypeptide region between an immunoglobulin constant region and an antigen-binding domain. In certain embodiments, the linker is composed of from 5 to about 35 amino acids, such as from about 15 to about 25 amino acids. In some embodiments, the linker is composed of at least 5 amino acids, at least 7 amino acids or at least 9 amino acids.

[0122] As used herein, the term "heavy chain" when used in reference to an antibody can refer to any distinct type, such as alpha (α), delta (δ), epsilon (ε), gamma (γ), and mu (μ), based on the amino acid sequence of the constant region, which give rise to antibodies of the IgA, IgD, IgE, IgG, and IgM classes, respectively, including subclasses of IgG, such as IgG 1 、IgG 2 、IgG 3 、およびIgG 4 。

[0123] As used herein, the term "light chain" when used in reference to an antibody can refer to any distinct type, such as kappa (κ) or lambda (λ), based on the amino acid sequence of the constant region. Light chain amino acid sequences are well known in the art. In a specific embodiment, the light chain is a human light chain.

[0124] As used herein, the term "EU numbering system" refers to the EU numbering provisions for the constant regions of antibodies, as described in Edelman, G.M. et al., Proc. Natl. Acad. USA, 63, 78-85 (1969) and Kabat et al, Sequences of Proteins of Immunological Interest, U.S. Dept. Health and Human Services, 5th edition, 1991, each of which is hereby incorporated by reference in its entirety. As used herein, unless otherwise specified, the positions of amino acid residues in the constant regions of immunoglobulin molecules are numbered according to the EU nomenclature (Ward et al., 1995 Therap. Immunol. 2:77-94).

[0125] As used herein, the term "dimer" refers to a biological entity consisting of two subunits that are related to each other through one or more forms of intramolecular forces, including covalent bonds (e.g., disulfide bonds), and other interactions (e.g., electrostatic interactions, salt bridges, hydrogen bonds, and hydrophobic interactions), and that is stable under appropriate conditions (e.g., physiological conditions, in an aqueous solution suitable for expressing, purifying, and / or storing a recombinant protein, or under conditions for non-denaturing and / or non-reducing electrophoresis). As used herein, "heterodimer" or "heterodimeric protein" refers to a dimer formed from two different polypeptides. As used herein, "homodimer" or "homodimeric protein" refers to a dimer formed from two identical polypeptides.

[0126] As used herein, "antibody-dependent cell-mediated cytotoxicity" and "ADCC" refer to a cell-mediated process in which non-specific cytotoxic cells expressing FcγR (e.g., monocytic cells such as natural killer (NK) cells and macrophages) recognize an antibody (or other protein capable of binding FcγR) bound to a target cell and subsequently cause lysis of the target cell. In principle, any effector cell that activates FcγR can be made to operate to mediate ADCC. Primary cells for mediating ADCC are NK cells that express only FcγRIII, while monocytes can express FcγRI, FcγRII, and FcγRIII depending on their state of activation, localization, or differentiation. For a review of FcγR expression on hematopoietic cells, see, e.g., Ravetch et al., Annu. Rev. Immunol., 9:457-92 (1991).

[0127] As used herein with respect to a, the term "having ADCC activity" means that a polypeptide (e.g., one containing an immunoglobulin hinge region and an immunoglobulin constant region having CH2 and CH3 domains such as those derived from IgG (e.g., IgG1)) can mediate antibody-dependent cell-mediated cytotoxicity (ADCC) through the binding of a cell-lytic Fc receptor (e.g., FcγRIII) to a cell-lytic immune effector cell (e.g., an NK cell) expressing an Fc receptor.

[0128] As used herein, "complement-dependent cytotoxicity" and "CDC" refer to a process in which components in normal serum ("complement") together with C1q complement-binding protein bound to an antibody or other target antigen cause lysis of a target cell expressing the target antigen. Complement consists of a group of serum proteins that act in concert and in an ordered manner to exert their effects.

[0129] As used herein, the terms "classical complement pathway" and "classical complement system" are synonymous and refer to a specific pathway for activation of the complement. The classical pathway requires an antigen-antibody complex for initiation and involves the activation of nine major protein components designated C1 through C9 in a regular manner. For some steps in the activation process, the products are enzymes that catalyze subsequent steps. This cascade provides amplification and activation of a large amount of complement by a relatively small initial signal.

[0130] As used herein in connection with a polypeptide, the term "having CDC activity" means that the polypeptide (e.g., an immunoglobulin hinge region and an immunoglobulin constant region having CH2 and CH3 domains such as those derived from IgG (e.g., IgG1)) can mediate complement-dependent cytotoxicity (CDC) through binding of C1q complement protein and activation of the classical complement system. In one embodiment of the invention, the recombinant polypeptide is modified to attenuate CDC activity.

[0131] "Binding affinity" generally refers to the total strength of non-covalent interactions between a single binding site of a molecule (e.g., an antibody) and its binding partner (e.g., an antigen). Unless otherwise indicated, as used herein, "binding affinity" refers to the intrinsic binding affinity that reflects a 1:1 interaction between members of a binding pair (e.g., an antibody and an antigen). The affinity of molecule X for its partner Y can generally be represented by the dissociation constant (K D ). Affinity can be measured and / or expressed by many methods known in the art, including but not limited to the equilibrium dissociation constant (K D ) and the equilibrium association constant (K A ). K D is calculated from the quotient of k off / k on , and K A is calculated from the quotient of k on / k off . k onrefers to, for example, the association rate constant of an antibody to an antigen, k off refers to, for example, the dissociation of an antibody from an antigen. k on and k off can be determined by techniques known to those skilled in the art such as BIAcore® or KinExA.

[0132] As used herein, the terms "immunologically specifically bind", "immunologically specifically recognize", "specifically bind", and "specifically recognize" are similar terms in the context of antibodies. These terms indicate that an antibody binds to an epitope via its antigen-binding domain and that the binding involves some complementarity between the antigen-binding domain and the epitope. Thus, it is possible to have an antibody that "specifically binds" to human 4-1BB and / or OX40, but the degree of binding to non-related non-4-1BB and / or OX40 proteins is less than about 10% of the binding of the antibody to 4-1BB and / or OX40, as measured, for example, by radioimmunoassay (RIA).

[0133] Binding domains can be classified as "high-affinity" binding domains and "low-affinity" binding domains. A "high-affinity" binding domain refers to those binding domains having a K -7 value of less than 10 -8 M, less than 10 -9 M, less than 10 -10 M, or less than 10 D M. A "low-affinity" binding domain refers to those binding domains having a KD greater than 10 -7 M, greater than 10 -6 M, or greater than 10 -5 M. "High-affinity" and "low-affinity" binding domains bind their targets but do not significantly bind other components present in a test sample.

[0134] As used herein, an antibody is "capable of binding" if it specifically binds its target (i.e., human 4-1BB or human OX40) under conditions that one of ordinary skill in the art would consider necessary for binding when in proximity to the target. "Human 4-1BB antigen-binding domain" should be understood to mean a binding domain that specifically binds to human 4-1BB. "Human OX40 antigen-binding domain" should be understood to mean a binding domain that specifically binds to OX40.

[0135] As used herein, "epitope" is a term in the art and refers to the localized region of an antigen to which an antibody can specifically bind. An epitope can be, for example, contiguous amino acids of a polypeptide (linear or contiguous epitope), or an epitope can come together from two or more non-contiguous regions of a polypeptide or polypeptides (conformational, non-linear, discontinuous, or non-contiguous epitope). In certain embodiments, the epitope to which an antibody binds can be determined, for example, by NMR spectroscopy, X-ray crystallographic studies, ELISA assays, hydrogen / deuterium exchange combined with mass spectrometry (e.g., liquid chromatography electrospray mass spectrometry), array-based oligopeptide scanning assays, and / or mutagenesis mapping (e.g., site-directed mutagenesis mapping). For X-ray crystallography, crystallization can be achieved using any method known in the art (e.g., Giege R et al., (1994) Acta Crystallogr D Biol Crystallogr 50(Pt 4):339-350; McPherson A (1990) Eur J Biochem 189:1-23; Chayen NE (1997) Structure 5:1269-1274; McPherson A (1976) J Biol Chem 251:6300-6303).Antibody:antigen crystals can be studied using well-known X-ray diffraction techniques and refined using computer software such as X-PLOR (distributed by Yale University, 1992, Molecular Simulations, Inc.; see, for example, Meth Enzymol (1985) vols. 114 & 115, eds Wyckoff HW et al.,; U.S. Patent Publication No. 2004 / 0014194), and BUSTER (Bricogne G (1993) Acta Crystallogr D Biol Crystallogr 49(Pt 1):37-60; Bricogne G (1997) Meth Enzymol 276A:361-423, ed Carter CW; Roversi P et al., (2000) Acta Crystallogr D Biol Crystallogr 56(Pt 10):1316-1323). Mutagenesis mapping studies can be achieved using any method known to those skilled in the art. For an explanation of mutagenesis techniques, including alanine scanning mutagenesis techniques, see, for example, Champe M et al., (1995) J Biol Chem 270:1388-1394 and Cunningham BC & Wells JA (1989) Science 244:1081-1085.

[0136] An antibody that "binds to the same epitope" as a reference antibody refers to an antibody that binds to the same amino acid residues on the antigen as the reference antibody. The ability of an antibody to bind to the same epitope as a reference antibody can be determined by a hydrogen / deuterium exchange assay (see Coales et al. Rapid Commun. Mass Spectrom. 2009;23:639-647).

[0137] An antibody that "binds to the same conformational epitope" as a reference antibody refers to an antibody that binds to the conformation or structure of an antigen as the reference antibody. The ability of an antibody to bind to the same conformational epitope as a reference antibody can be determined by methods known in the art, including, for example, comparison of the structures of the antibody(ies) complexed with the antigen as determined by hydrogen / deuterium exchange assays (see Coales et al. Rapid Commun. Mass Spectrom. 2009;23:639-647), X-ray crystallography, and alanine scanning. An antibody that binds to the same linear epitope as a reference antibody refers to an antibody that binds to the same linear amino acid sequence on the antigen as the reference antibody. For linear epitopes, peptide mapping experiments such as pepspot analysis can be used to determine binding to the same linear epitope.

[0138] An antibody is said to "competitively inhibit" the binding of a reference antibody to an epitope if the antibody preferentially binds to that epitope or an overlapping epitope to such an extent that it blocks the binding of the reference antibody to the epitope to some degree. Competitive inhibition can be determined by any method known in the art, such as a competitive ELISA assay, surface plasmon resonance (SPR), or biolayer interferometry (BLI). An antibody can be said to competitively inhibit the binding of a reference antibody to a given epitope if it prevents or reduces the binding of the reference antibody to its target by at least 50%. In certain embodiments, the antibody competitively inhibits the binding of the reference antibody to a given epitope by at least 90%, at least 80%, at least 70%, or at least 60%. In certain embodiments, the reference antibody is an anti-4-1BB antibody, an anti-OX40 antibody, an anti-4-1BB bispecific or multispecific antibody, or an anti-OX40 bispecific or multispecific antibody. For example, the reference antibody can be an anti-4-1BB × anti-OX40 bispecific antibody. A reference antibody having a human 4-1BB antigen-binding domain can potentially comprise the heavy chain variable domain (VH) of SEQ ID NO: 17 and the light chain variable domain (VL) of SEQ ID NO: 18. A reference antibody having a human OX40 antigen-binding domain can potentially comprise the heavy chain of SEQ ID NO: 29 and the VL of SEQ ID NO: 28.

[0139] The terms "polypeptide", "peptide", and "protein" are used interchangeably herein and refer to polymers of amino acids of any length. The polymers can be linear or branched, and they may contain modified amino acids and may be interrupted by non-amino acids. The term also encompasses amino acid polymers that have been modified naturally or by any other manipulation or modification, such as disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, or conjugation to a labeling component. Also included in the definition are polypeptides containing, for example, one or more analogs of amino acids (including, e.g., non-natural amino acids), as well as other modifications known in the art. Since the polypeptides of the present invention are antibody-based, it is understood that in certain embodiments, the polypeptides can occur as single chains or related chains.

[0140] As used herein, the terms "nucleic acid", "nucleic acid molecule", or "polynucleotide" refer to deoxyribonucleotides or ribonucleotides in either single-stranded or double-stranded form and their polymers. Unless otherwise restricted, the term encompasses nucleic acids containing nucleotide analogs that have binding properties similar to the reference nucleic acid and are metabolized in a manner similar to naturally occurring nucleotides. Unless otherwise specified, a particular nucleic acid sequence also implicitly encompasses its conservatively modified variants (e.g., degenerate codon substitutions), complementary sequences, and sequences explicitly indicated. Specifically, degenerate codon substitutions can be achieved by generating sequences in which the third position of one or more selected (or all) codons is substituted with a mixture of bases and / or deoxyinosine residues (Batzer et al. (1991) Nucleic Acid Res. 19:5081; Ohtsuka et al. (1985) J. Biol. Chem. 260:2605-2608; Cassol et al. (1992); Rossolini et al. (1994) Mol. Cell. Probes 8:91-98). The term nucleic acid is used interchangeably with gene, cDNA, and mRNA encoded by a gene. As used herein, the terms "nucleic acid", "nucleic acid molecule", or "polynucleotide" are intended to include DNA molecules (e.g., cDNA or genomic DNA), RNA molecules (e.g., mRNA), analogs of DNA or RNA generated using nucleotide analogs, and their derivatives, fragments, and homologs.

[0141] As used herein, the term "expression vector" refers to a linear or circular nucleic acid molecule containing one or more expression units. In addition to one or more expression units, an expression vector can also contain additional nucleic acid segments such as, for example, one or more origins of replication or one or more selectable markers. Expression vectors generally are derived from plasmid or viral DNA or can contain elements of both.

[0142] "Percent identity" refers to the degree of identity between two sequences (e.g., amino acid sequence or nucleic acid sequence). Percent identity can be determined by aligning the two sequences and introducing gaps to maximize the identity between the sequences. The alignment can be generated using programs known in the art. For the purposes herein, the alignment of nucleotide sequences can be performed using the blastn program set to default parameters, and the alignment of amino acid sequences can be performed using the blastp program set to default parameters (see the National Center for Biotechnology Information (NCBI) on the World Wide Web, ncbi.nlm.nih.gov).

[0143] As used herein, "conservative amino acid substitution" refers to an amino acid residue that has been replaced with an amino acid residue having a similar side chain. Families of amino acid residues having side chains are defined in the art. These families include amino acids having basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine, tryptophan), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine), beta-branched side chains (e.g., threonine, valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). In certain embodiments, one or more amino acid residues within the CDR(s) or framework region(s) of an antibody can be replaced with amino acid residues having similar side chains.

[0144] As used herein, a polypeptide or amino acid sequence "derived from" a specified polypeptide refers to the origin of the polypeptide. In certain embodiments, a polypeptide or amino acid sequence derived from a particular sequence (often referred to as the "starting" or "parent" or "parental" sequence) has an amino acid sequence that is essentially identical to the starting sequence or a portion thereof, where the portion consists of at least 10 - 20 amino acids, at least 20 - 30 amino acids, or at least 30 - 50 amino acids, or at least 50 - 150 amino acids, or is otherwise distinguishable to one of ordinary skill in the art as having its origin in the starting sequence. For example, a binding domain can be derived from an antibody, such as a Fab, F(ab’) 2 , Fab’, scFv, single domain antibody (sdAb), etc.

[0145] A polypeptide derived from another polypeptide can have one or more mutations in relation to the starting polypeptide, such as having one or more amino acid residues that are substituted with another amino acid residue, or having one or more insertions or deletions of one or more amino acid residues. A polypeptide can include non-naturally occurring amino acid sequences. Such changes will necessarily have less than 100% sequence identity or similarity to the starting polypeptide. In one embodiment, the variant will have an amino acid sequence with about 60% to less than 100% amino acid sequence identity or similarity to the amino acid sequence of the starting polypeptide. In another embodiment, the variant will have an amino acid sequence with about 75% to less than 100%, about 80% to less than 100%, about 85% to less than 100%, about 90% to less than 100%, about 95% to less than 100% amino acid sequence identity or similarity to the amino acid sequence of the starting polypeptide.

[0146] As used herein, the term "host cell" can be any type of cell, e.g., a primary cell, a cell in culture, or a cell from a cell line. In certain embodiments, the term "host cell" refers to a cell transfected with a nucleic acid molecule and the progeny or potential progeny of such a cell. The progeny of such a cell may not be identical to the parent cell transfected with the nucleic acid molecule, e.g., due to mutations that may occur in subsequent generations or environmental influences in the integration of the nucleic acid molecule into the host cell genome.

[0147] An "isolated" polypeptide, antibody, polynucleotide, vector, cell, or composition is a polypeptide, antibody, polynucleotide, vector, cell, or composition that is in a form not found in nature. Isolated polypeptides, antibodies, polynucleotides, vectors, cells or compositions include those that have been purified to the extent that they are no longer in the form found in nature. In some embodiments, an isolated antibody, polynucleotide, vector, cell, or composition is substantially pure. As used herein, "substantially pure" refers to a material that is at least 50% pure (i.e., free of contaminants). In some cases, the material is at least 90% pure, at least 95% pure, at least 98% pure, or at least 99% pure.

[0148] The term "pharmaceutical formulation" refers to a preparation that is in a form that enables the biological activity of the active ingredient to be effective and that contains no additional ingredients that are toxic to an unacceptable degree to the subject to which the formulation is administered. The formulation can be sterilized.

[0149] As used herein, the term "pharmaceutically acceptable" refers to molecular entities and compositions that do not generally produce allergic or other serious adverse reactions when administered using routes well known in the art. Molecular entities and compositions that are approved by a regulatory agency of the Federal or State government or listed in the U.S. Pharmacopeia or other generally recognized pharmacopeias for use in animals and more particularly in humans are considered "pharmaceutically acceptable."

[0150] As used herein, the terms "administer," "administering," "administration," etc. refer to methods that can be used to effect delivery of a drug, e.g., a 4-1BB / OX40 antibody, to the desired site of its biological action (e.g., intravenous administration). Administration techniques that can be used with the drugs and methods described herein can be found, for example, in Goodman and Gilman, The Pharmacological Basis of Therapeutics, current edition, Pergamon; and Remington’s, Pharmaceutical Sciences, current edition, Mack Publishing Co., Easton, Pa.

[0151] As used herein, the terms "subject" and "patient" are used interchangeably. A subject can be an animal. In some embodiments, the subject is a mammal such as a non-human animal (e.g., a cow, pig, horse, cat, dog, rat, mouse, monkey or other primate, etc.). In some embodiments, the subject is a human. As used herein, the term "patient in need" or "subject in need" refers to a patient who has a risk of, or is suffering from, a disease, disorder or condition that is suitable for treatment or amelioration using, for example, the 4-1BB / OX40 antibodies provided herein. A patient in need can be, for example, a patient diagnosed with cancer.

[0152] The term "therapeutically effective amount" refers to an amount of an agent, e.g., an anti-4-1BB / OX40 antibody, that is effective to treat a disease or disorder in a subject. In the case of cancer, a therapeutically effective amount of an agent can reduce the number of cancer cells; reduce the size or amount of a tumor; inhibit (i.e., delay to some extent, stop in some embodiments) cancer cell infiltration into visceral organs; inhibit (i.e., delay to some extent, stop in some embodiments) tumor metastasis; inhibit tumor growth to some extent; alleviate to some extent one or more of the symptoms associated with cancer; and / or result in a favorable response such as an increased progression-free survival (PFS), disease-free survival (DFS), or overall survival (OS), complete response (CR), partial response (PR), or in some cases, stable disease (SD), reduction in progressive disease (PD), shortened time to progression (TTP), or any combination thereof.

[0153] Terms such as "treating" or "treatment" or "for treating" or "alleviating" or "for alleviating" refer to therapies that cure, slow down, relieve symptoms of, and / or halt the progression of a diagnosed pathological condition or disorder. Thus, those in need of treatment include those already diagnosed with or suspected of having a disorder. In certain embodiments, a patient exhibits one or more of the following: a decrease or complete absence of the number of cancer cells; a decrease in tumor size; inhibition or absence of invasion of cancer cells into peripheral organs, including, for example, the spread of cancer to soft tissue and bone; inhibition or absence of tumor metastasis; inhibition or absence of tumor growth; alleviation of one or more symptoms associated with a particular cancer; a reduced disease rate and mortality rate; an improvement in quality of life; a decrease in tumorigenicity, tumorigenic frequency, or tumorigenic potential of a tumor; a decrease in the number or frequency of cancer stem cells within a tumor; differentiation of tumorigenic cells into a non-tumorigenic state; an increased progression-free survival (PFS), disease-free survival (DFS), or overall survival (OS), a decrease in complete response (CR), partial response (PR), stable disease (SD), progressive disease (PD), a shortened time to progression (TTP), or any combination thereof, then the subject has had a successful "treatment" of cancer according to the methods of the invention.

[0154] The terms "cancer" and "cancerous" refer to or describe a physiological state in a mammal in which a population of cells is characterized by unregulated cell growth. Examples of cancer include, but are not limited to, melanoma, renal cancer, pancreatic cancer, lung cancer, intestinal cancer, prostate cancer, breast cancer, liver cancer, brain cancer, and blood cancer. Cancer can be a primary tumor, or it can be a progressive or metastatic cancer.

[0155] Cancer can be a solid tumor cancer. The term "solid tumor" generally refers to a mass of abnormal tissue that does not normally contain cysts or liquid regions. Examples of solid tumors are sarcomas, carcinomas, and lymphomas. Leukemia (cancer of the blood) generally does not form solid tumors. Solid tumors can include tumor-infiltrating lymphocytes that express OX40 and 4-1BB.

[0156] As used herein, the terms "a" and "an" refer to "one or more" of the listed components, unless otherwise specified.

[0157] Unless otherwise specified or clear from the context, as used herein, the term "or" is understood to be inclusive. The term "and / or" as used in phrases such as "A and / or B" herein is intended to include "A and B", "A or B", "A", and "B". Similarly, the term "and / or" as used in phrases such as "A, B, and / or C" is intended to include the following embodiments: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C: A (alone); B (alone); and C (alone).

[0158] Whenever an embodiment is described herein in terms of the word "comprising", other similar embodiments described in terms of the terms "consisting of" and / or "consisting essentially of" are also provided and are understood to be part of the disclosure of this application. In this disclosure, terms such as "comprise", "comprising", "include", and "having" can have the meanings ascribed to them in US patent law and European patent law and can mean "include", "including", etc.; "consisting essentially of" or "consisting essentially" similarly has meanings based on US patent law and European patent law. As far as US patent law is concerned, the term is unrestricted and it should be recognized that the presence of more than the listed ones, excluding prior art embodiments, allows for more presence than the listed ones as long as the basic or novel features of the listed ones are not changed. As far as European patent law is concerned, it should be recognized that the use of "consisting essentially of" or "substantially comprising" means that there may be present certain additional components, i.e., components that do not substantially affect the essential features of the compound or composition.

[0159] As used herein, the term “about” or “approximately” when used to modify a numerical value or numerical range indicates that a deviation of up to 5% above or 5% below the value or range is within the intended meaning of the recited value or range.

[0160] Any domain, component, composition, and / or method provided herein can be combined with any one or more of the other domains, components, compositions, and / or methods provided herein. II.4-1BB and OX40 antibodies

[0161] Provided herein are 4-1BB antibodies, OX40 antibodies, and 4-1BB×OX40 bispecific antibodies. The 4-1BB antibodies and 4-1BB×OX40 bispecific antibodies include an antigen-binding domain that specifically binds to human 4-1BB (i.e., a human 4-1BB antigen-binding domain). The OX40 antibodies and 4-1BB×OX40 bispecific antibodies include an antigen-binding domain that binds to human OX40 (i.e., a human OX40 antigen-binding domain). The 4-1BB×OX40 bispecific antibodies can include a human 4-1BB binding domain and a human OX40 binding domain. The 4-1BB×OX40 bispecific antibodies are monovalent for each target, i.e., include one human 4-1BB binding domain and one human OX40 binding domain. The bispecific antibodies can also be divalent for one or both of the target proteins, i.e., include two 4-1BB binding domains and / or two OX40 binding domains. A representative 4-1BB×OX40 bispecific antibody format is shown in Figure 27. A. 4-1BB binding domain

[0162] Provided herein is an antigen-binding domain that binds to human 4-1BB (i.e., 4-1BB binding domain) that can be used to associate a 4-1BB×OX40 bispecific antibody. In addition to binding to human 4-1BB, the 4-1BB binding domain can bind to 4-1BB from other species, such as cynomolgus monkey and / or mouse 4-1BB. In certain cases, the 4-1BB binding domain binds to human 4-1BB and cynomolgus monkey 4-1BB.

[0163] The 4-1BB binding domain can comprise six complementarity-determining regions (CDRs), namely variable heavy chain (VH) CDR1, VH CDR2, VH CDR3, variable light chain (VL) CDR1, VL CDR2, and VL CDR3. The 4-1BB binding domain can comprise a variable heavy chain (VH) and a variable light chain (VL). The VH and VL can be separate polypeptides or (e.g., in an scFv) part of the same polypeptide.

[0164] In certain embodiments, the 4-1BB binding domains described herein comprise a combination of six CDRs listed in Tables A and B (e.g., SEQ ID NOs: 5-10 or SEQ ID NOs: 5, 119, 7, 120, 121, and 122).

[0165] [Table 2]

[0166] [Table 3]

[0167] The 4-1BB×OX40 bispecific antibody that is monovalent for 4-1BB can include a single 4-1BB binding domain in combination with the six CDRs listed in Table A and Table B above (e.g., SEQ ID NOs: 5-10 or SEQ ID NOs: 5, 119, 7, 120, 121, and 122). The 4-1BB×OX40 bispecific antibody that is bivalent for 4-1BB can include two 4-1BB binding domains, each including a combination of the six CDRs listed in Table A and Table B above (e.g., SEQ ID NOs: 5-10 or SEQ ID NOs: 5, 119, 7, 120, 121, and 122).

[0168] As described herein, 4-1BB binding can include the VH of the antibodies listed in Table C.

[0169] [Table 4]

[0170] As described herein, the 4-1BB binding domain can include a VH having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or 100% sequence identity to the sequences in Table C, and optionally, the VH can include the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5-7, respectively, or the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5, 119, and 7, respectively.

[0171] As described herein, the 4-1BB binding domain can include a VH that includes the CDRs of the VH sequences in Table C, e.g., IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs.

[0172] As described herein, the 4-1BB binding domain can include the VL of the antibodies listed in Table D.

[0173]

Table 5

[0174] As described herein, the 4-1BB binding domain can include a VL having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or 100% sequence identity to the sequence of Table D, and optionally, the VL can include the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 8-10, respectively, or the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 120-122, respectively.

[0175] As described herein, the 4-1BB binding domain can include a VL that includes a CDR of the VL sequence of Table D, such as an IMGT-defined CDR, Kabat-defined CDR, Chothia-defined CDR, or AbM-defined CDR.

[0176] As described herein, the 4-1BB binding domain can comprise a VH listed in Table C and a VL listed in Table D. A 4-1BB×OX40 bispecific antibody that is monovalent for 4-1BB can comprise a single 4-1BB binding domain that comprises a VH listed in Table C and a VL listed in Table D. A 4-1BB×OX40 bispecific antibody that is bivalent for 4-1BB can comprise two 4-1BB binding domains that each comprise a VH listed in Table C and a VL listed in Table D. The VH listed in Table C and the VL listed in Table D can be different polypeptides or can be on the same polypeptide. When the VH and VL are on the same polypeptide, they can be in either orientation (i.e., VH-VL or VL-VH), and they can be connected by a linker (e.g., a glycine-serine linker). In certain embodiments, the VH and VL are connected by a glycine-serine linker that is at least 15 amino acids in length (e.g., 15-50 amino acids, 15-40 amino acids, 15-30 amino acids, 15-25 amino acids, or 15-20 amino acids). In certain embodiments, the VH and VL are connected by a glycine-serine linker that is at least 20 amino acids in length (e.g., 20-50 amino acids, 20-40 amino acids, 20-30 amino acids, or 20-25 amino acids).

[0177] As described herein, the 4-1BB binding domain can comprise a VH that comprises a CDR of the VH sequence of Table C, e.g., an IMGT-defined CDR, a Kabat-defined CDR, a Chothia-defined CDR, or an AbM-defined CDR, and a VL that comprises a CDR of the VL sequence of Table D, e.g., an IMGT-defined CDR, a Kabat-defined CDR, a Chothia-defined CDR, or an AbM-defined CDR.

[0178] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 17 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 17, optionally a VH comprising the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5-7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 18 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally a VL comprising the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 8-10, respectively).

[0179] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 19 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 19, optionally a VH comprising the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5, 119, and 7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 20 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 20, optionally a VL comprising the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 120-122, respectively).

[0180] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 21 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 21, optionally a VH comprising the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5, 19, and 7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 22 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 22, optionally a VL comprising the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 120-122, respectively).

[0181] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 23 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 23, optionally a VH comprising the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 5, 119, and 7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 24 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 24, optionally a VL comprising the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 120-122, respectively).

[0182] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 32 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 32, optionally comprising a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 5-7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 18 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally comprising a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 8-10, respectively).

[0183] In certain embodiments, the 4-1BB binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 143 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 143, optionally comprising a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 5, 119, and 7, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 20 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 20, optionally comprising a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 120-122, respectively).

[0184] In certain embodiments, the 4-1BB binding domains (e.g., scFv) described herein bind to human 4-1BB and include one of the amino acid sequences set forth in Table E.

[0185] [Table 6]

[0186] As described herein, a 4-1BB×OX40 bispecific antibody that is monovalent for 4-1BB can include a single 4-1BB binding domain that includes the sequences listed in Table E. A 4-1BB×OX40 bispecific antibody that is bivalent for 4-1BB can include two 4-1BB binding domains, each of which includes the sequences listed in Table E.

[0187] As described herein, the 4-1BB binding domain can include an amino acid sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99% identical to the sequences of Table E, and optionally, the sequence includes the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 5-7, respectively, and the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 8-10, respectively, or the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 5, 119, and 7, respectively, and the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 120-122, respectively.

[0188] In certain embodiments, the 4-1BB binding domains provided herein competitively inhibit the binding of an antibody that includes the VH sequence of Table C (e.g., VH including SEQ ID NO: 17) and the VL sequence of Table D (e.g., VL including SEQ ID NO: 18) to human 4-1BB.

[0189] In certain embodiments, the 4-1BB binding domain provided herein specifically binds to the same epitope of human 4-1BB as an antibody comprising a VH sequence (e.g., VH comprising SEQ ID NO: 17) of Table C and a VL sequence (e.g., VL comprising SEQ ID NO: 18) of Table D.

[0190] In certain embodiments, the 4-1BB binding domain provided herein is capable of stimulating 4-1BB. In certain embodiments, the 4-1BB binding domain provided herein in a 4-1BB×OX40 bispecific antibody stimulates only 4-1BB in the presence of both 4-1BB and OX40. B. OX40 Binding Domain

[0191] Provided herein is an antigen-binding domain (i.e., OX40 binding domain) that binds to human OX40 and can be used to associate a 4-1BB×OX40 bispecific antibody. In addition to binding to human OX40, the OX40 binding domain can bind to OX40 from other species, such as cynomolgus monkey and / or mouse OX40. In certain cases, the OX40 binding domain binds to human OX40 and cynomolgus monkey OX40.

[0192] The OX40 binding domain can comprise six complementarity-determining regions (CDRs), namely, variable heavy chain (VH) CDR1, VH CDR2, VH CDR3, variable light chain (VL) CDR1, VL CDR2, and VL CDR3. The OX40 binding domain can comprise a variable heavy chain (VH) and a variable light chain (VL). The VH and VL can be separate polypeptides or (e.g., in an scFv) part of the same polypeptide.

[0193] In certain embodiments, the OX40 binding domain described herein comprises the six CDRs listed in Tables F and G.

[0194]

Table 7

[0195]

Table 8

[0196] The 4-1BB×OX40 bispecific antibody that is monovalent for OX40 can include a single OX40 binding domain having the six CDRs listed in Tables F and G above. The 4-1BB×OX40 bispecific antibody that is bivalent for OX40 can include two OX40 binding domains each having the six CDRs listed in Tables F and G above.

[0197] As described herein, OX40 binding can include the VH of the antibodies listed in Table H.

[0198]

Table 9

[0199] As described herein, the OX40 binding domain can include a VH having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or 100% sequence identity to the sequence of Table H, and optionally, the VH can include the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 11-13, respectively.

[0200] As described herein, the OX40 binding domain can include a VH that includes the CDRs of the VH sequence of Table H, such as IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs.

[0201] As described herein, the OX40 binding domain can include the VL of the antibodies listed in Table I.

[0202]

Table 10

[0203] As described herein, the OX40 binding domain can include a VL having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or 100% sequence identity to the sequences of Table I, and optionally, the VL can include the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 14-16, respectively.

[0204] As described herein, the OX40 binding domain can include a VL that includes the CDRs of the VL sequences of Table I, such as, for example, the IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs.

[0205] As described herein, the OX40 binding domain can include a VH listed in Table H and a VL listed in Table I. A 4-1BB×OX40 bispecific antibody that is monovalent for 4-1BB can include a single 4-1BB binding domain that includes a VH listed in Table H and a VL listed in Table I. A 4-1BB×OX40 bispecific antibody that is bivalent for 4-1BB can include two 4-1BB binding domains, each of which includes a VH listed in Table H and a VL listed in Table I. The VH listed in Table H and the VL listed in Table I can be different polypeptides or can be on the same polypeptide. When the VH and VL are on the same polypeptide, they can be in either orientation (i.e., VH-VL or VL-VH), and they can be connected by a linker (e.g., a glycine-serine linker). In certain embodiments, the VH and VL are connected by a glycine-serine linker that is at least 15 amino acids in length (e.g., 15-50 amino acids, 15-40 amino acids, 15-30 amino acids, 15-25 amino acids, or 15-20 amino acids). In certain embodiments, the VH and VL are connected by a glycine-serine linker that is at least 20 amino acids in length (e.g., 20-50 amino acids, 20-40 amino acids, 20-30 amino acids, or 20-25 amino acids).

[0206] As described herein, the OX40 binding domain can include a VH that includes CDRs of the VH sequence of Table H, e.g., IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs, and a VL that includes CDRs of the VL sequence of Table I, e.g., IMGT-defined CDRs, Kabat-defined CDRs, Chothia-defined CDRs, or AbM-defined CDRs.

[0207] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 25 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 25, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 26 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0208] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 27 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 27, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 28 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 28, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0209] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 29, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 28 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 28, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0210] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 29, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of any one of SEQ ID NOs: 26, 30, and 34-37 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to any one of SEQ ID NOs: 28 and 34-37, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0211] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 31 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 31, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of any one of SEQ ID NOs: 28, 30, and 34-41 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to any one of SEQ ID NOs: 28, 30, and 34-41, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0212] In certain embodiments, the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 33 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 33, optionally a VH comprising the VH CDR1, VH CDR2, and VH CDR3 sequences of SEQ ID NOs: 11-13, respectively), and (ii) a VL comprising the amino acid sequence of SEQ ID NO: 28 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 28, optionally a VL comprising the VL CDR1, VL CDR2, and VL CDR3 sequences of SEQ ID NOs: 14-16, respectively).

[0213] In certain embodiments, the OX40 binding domains (e.g., scFv) described herein bind to human OX40 and include one of the amino acid sequences set forth in Table J.

[0214] [Table 11] TIFF0007691138000012.tif42162

[0215] As described herein, a 4-1BB×OX40 bispecific antibody that is monovalent for OX40 can include a single OX40 binding domain that includes the sequences listed in Table J. A 4-1BB×OX40 bispecific antibody that is bivalent for OX40 can include two OX40 binding domains, each of which includes the sequences listed in Table J.

[0216] As described herein, the OX40 binding domain can include an amino acid sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to the sequences of Table J, and optionally, the sequence includes the VH CDR1 sequence, VH CDR2 sequence, and VH CDR3 sequence of SEQ ID NOs: 11-13, respectively, and the VL CDR1 sequence, VL CDR2 sequence, and VL CDR3 sequence of SEQ ID NOs: 14-16, respectively.

[0217] In certain embodiments, the OX40 binding domains provided herein competitively inhibit the binding of an antibody comprising the VH sequence of Table H (e.g., VH comprising SEQ ID NO: 29) and the VL sequence of Table I (e.g., VL comprising SEQ ID NO: 28) to human OX40.

[0218] In certain embodiments, the OX40 binding domains provided herein specifically bind to the same epitope of human OX40 as an antibody comprising a VH sequence of Table H (e.g., VH comprising SEQ ID NO: 29) and a VL sequence of Table I (e.g., VL comprising SEQ ID NO: 28) against human OX40.

[0219] In certain embodiments, the OX40 binding domains provided herein can stimulate OX40. "Can" means that the OX40 binding domain can perform the activity, but can do so only under appropriate conditions that can be understood by those skilled in the art. In certain embodiments, the OX40 binding domain provided herein in a 4-1BB×OX40 bispecific antibody stimulates only OX40 in the presence of both 4-1BB and OX40. C. 4-1BB and / or OX40 Binding Domain

[0220] In the 4-1BB or OX40 binding domain, the VH CDR or the VH and VL CDRs or the VL can be separate polypeptides or can be on the same polypeptide. The VH CDR or the VH and VL CDRs or the VL are on the same polypeptide and they can be in either orientation (i.e., VH-VL or VL-VH).

[0221] When the VH CDR or the VH and VL CDRs or the VL are on the same polypeptide, they can be connected by a linker (e.g., a glycine-serine linker). The VH can be located at the N-terminus of the linker sequence and the VL can be located at the C-terminus of the linker sequence. Alternatively, the VL can be located N-terminal to the linker sequence and the VH can be located C-terminal to the linker sequence.

[0222] The use of peptide linkers for joining the VH and VL regions is well known in the art and there are many publications within this particular field. In some embodiments, the peptide linker is a 15-mer consisting of three repeats of the Gly-Gly-Gly-Gly-Ser amino acid sequence ((Gly 4 Ser) 3 )(SEQ ID NO: 116). Other linkers have been used and phage display technology, as well as selected infective phage technology, have been used to diversify and select appropriate linker sequences (Tang et al., J. Biol. Chem. 271, 15682-15686, 1996; Hennecke et al., Protein Eng. 11, 405-410, 1998). In certain embodiments, the VH and VL regions are joined by a peptide linker having an amino acid sequence comprising the formula (Gly 4 Ser) n where n = 1-5 (SEQ ID NO: 117). In certain embodiments, n = 3-10. In certain embodiments, n = 3-5. In certain embodiments, n = 4-10. In certain embodiments, n = 4-5. In certain embodiments, n = 4. Other suitable linkers can be obtained by optimizing a simple linker (e.g., (Gly 4 Ser) n ) through random mutagenesis, where n = 1-5 (SEQ ID NO: 117).

[0223] The 4-1BB and / or OX40 binding domain(s) can be a humanized binding domain. The 4-1BB and / or OX40 binding domain(s) can be a rat binding domain. The 4-1BB and / or OX40 binding domain(s) can be a mouse binding domain. In certain embodiments, the 4-1BB×OX40 bispecific antibody comprises a humanized 4-1BB binding domain and a rat OX40 binding domain. In certain embodiments, the 4-1BB×OX40 bispecific antibody comprises a humanized 4-1BB binding domain and a mouse OX40 binding domain. In certain embodiments, the 4-1BB×OX40 bispecific antibody comprises a humanized 4-1BB binding domain and a humanized OX40 binding domain.

[0224] The 4-1BB and / or OX40 binding domain(s) can be a scFv. In certain embodiments, all of the 4-1BB and OX40 binding domains within the 4-1BB×OX40 bispecific antibody are scFvs. In certain embodiments, one 4-1BB binding and one OX0 binding domain within the 4-1BB×OX40 bispecific antibody are scFvs. In certain embodiments, at least one 4-1BB or OX40 binding domain within the 4-1BB×OX40 bispecific antibody is a scFv. In certain embodiments, the polypeptide comprises a 4-1BB binding domain (e.g., scFv) and an OX40 binding domain (e.g., scFv).

[0225] The 4-1BB and / or OX40 binding domain(s) can comprise VH and VL in separate polypeptide chains. In certain embodiments, all of the 4-1BB and OX40 binding domains within the 4-1BB×OX40 bispecific antibody comprise VH and VL in separate polypeptide chains. In certain embodiments, at least one 4-1BB or OX40 binding domain within the 4-1BB×OX40 bispecific antibody comprises VH and VL in separate polypeptide chains. D. 4-1BB×OX40 Bispecific Antibody

[0226] Provided herein are bispecific antibodies that bind to human 4-1BB and human OX40 (4-1BB×OX40 bispecific antibodies). Such bispecific antibodies comprise at least one 4-1BB binding domain and at least one human OX40 binding domain. The 4-1BB binding domain within the bispecific antibody can be any human 4-1BB binding domain, including, for example, any of the 4-1BB binding domains discussed above. The OX40 binding domain within the bispecific antibody can be any human OX40 binding domain, including, for example, any of the OX40 binding domains discussed above.

[0227] In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein are capable of binding to 4-1BB and OX40 simultaneously.

[0228] In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein are capable of stimulating the T cell costimulatory pathway. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein are capable of stimulating 4-1BB only in the presence of OX40. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein are capable of stimulating OX40 only in the presence of 4-1BB.

[0229] In certain embodiments, the 4-1BB×OX40 antibodies provided herein can enhance natural killer (NK) cell proliferation. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein can enhance T cell proliferation. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein can enhance CD8 T cell proliferation. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein can enhance CD4 T cell proliferation. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein can enhance both CD8 T cell proliferation and CD4 T cell proliferation. In certain embodiments, the 4-1BB×OX40 bispecific antibodies provided herein can enhance both NK cell proliferation and T cell proliferation.

[0230] In certain embodiments, the 4-1BB×OX40 bispecific antibody co-stimulates 4-1BB and OX40. In certain embodiments, the 4-1BB×OX40 bispecific antibody provides synergistic co-stimulation of T cells. In certain embodiments, the 4-1BB×OX40 bispecific antibody provides synergistic tumor lysis. In certain embodiments, the 4-1BB×OX40 bispecific antibody provides a synergistic effect that enhances the anti-tumor immune response.

[0231] In certain embodiments, the 4-1BB×OX40 bispecific antibody enhances T cell activation and / or prolongs T cell survival.

[0232] In certain embodiments, the 4-1BB×OX40 bispecific antibody comprises two 4-1BB binding domains and two OX40 binding domains. The 4-1BB×OX40 bispecific antibody comprises two antigen-binding domains that bind to the same target (e.g., 4-1BB or OX40), and those two antigen-binding domains can comprise the same amino acid sequence(s) or different amino acid sequences. In certain embodiments, the two 4-1BB binding domains comprise the same amino acid sequence(s). In certain embodiments, the two OX40 binding domains comprise the same amino acid sequence(s). In certain embodiments, the two 4-1BB binding domains comprise the same amino acid sequence(s) and the two OX40 binding domains comprise the same amino acid sequence(s).

[0233] 4-1BB×OX40 bispecific antibodies as provided herein can be prepared by chemically conjugating two different monoclonal antibodies or by fusing two hybridoma cell lines to produce a hybrid-hybridoma. Other multivalent formats that can be used include, for example, quadromas, Kλ bodies, dAbs, diabodies, TandAbs, nanobodies, small modular immunopharmaceuticals (SMIP™), DOCK-AND-LOCK® (DNL®), CrossMab Fab, CrossMab VH-VL, strand-exchanged engineered domain bodies (SEED bodies), Affibodies, Fynom ers, knotted domains, Albu-dabs, two modified Fv fragments with exchanged VHs (e.g., dual affinity retargeting molecules (DARTs)), scFv×scFv (e.g., BiTE), DVD-Ig, Covx-bodies, peptibodies, scFv-Ig, SVD-Ig, dAb-Ig, Knobs-in-Holes, IgG1 antibodies containing mutations that coincide within the CH3 domain (e.g., DuoBody antibodies), and triomAbs. Representative bispecific formats are discussed in Garber et al., Nature Reviews Drug Discovery 13:799-801 (2014), which is incorporated herein by reference in its entirety. Additional representative bispecific formats are discussed in Liu et al., Front. Immunol. 8:38 doi:10.2289 / fimmu.2017.00038, and Brinkmann and Kontermann, MABS 9:2,182-212 (2017), each of which is incorporated herein by reference in its entirety. In certain embodiments, the bispecific antibody can be an F(ab’) 2 fragment. The F(ab’) 2 fragment contains the two antigen-binding arms of a tetrameric antibody molecule joined by disulfide bonds in the hinge region.

[0234] The 4-1BB×OX40 bispecific antibodies disclosed herein can incorporate a multivalent binding protein scaffold. Multivalent binding proteins using scaffolds are disclosed, for example, in PCT Application Publication WO2007 / 146968, US Patent Application Publication No. 2006 / 0051844, PCT Application Publication WO2010 / 040105, PCT Application Publication WO2010 / 003108, US Patent No. 7,166,707, and US Patent No. 8,409,577, each of which is incorporated herein by reference in its entirety. The 4-1BB×OX40 bispecific antibody can include two binding domains (the domains can be designed to specifically bind the same or different targets), a hinge region, a linker (e.g., a carboxyl-terminal or amino-terminal linker), and an immunoglobulin constant region. The 4-1BB×OX40 bispecific antibody can be a homodimeric protein comprising two identical disulfide-bonded polypeptides.

[0235] In one embodiment, the 4-1BB×OX40 bispecific antibody comprises two polypeptides, each peptide comprising, in order from amino-terminus to carboxyl-terminus, a first antigen-binding domain, a linker (e.g., the linker is a hinge region), an immunoglobulin constant region, and a second antigen-binding domain. Figure 27 shows the 4-1BB×OX40 bispecific antibody in this arrangement. This arrangement is also referred to herein as the ADAPTIR™ format.

[0236] In some embodiments, the 4-1BB×OX40 bispecific antibody comprises a polypeptide that, in order from the amino terminus to the carboxyl terminus, comprises a 4-1BB binding domain (e.g., scFv), a linker (e.g., the linker is a hinge region), an immunoglobulin constant region, a linker, and an OX40 binding domain (e.g., scFv). In certain embodiments, the 4-1BB binding domain (e.g., scFv) comprises, in order from the amino terminus to the carboxyl terminus, a VH, a linker (e.g., a glycine-serine linker), and a VL. In certain embodiments, the linker between the 4-1BB binding domain and the immunoglobulin constant region is a hinge, and the hinge is an IgG 1 hinge. In certain embodiments, the immunoglobulin constant region comprises a CH2 domain and a CH3 domain. In certain embodiments, the OX40 binding domain (e.g., scFv) comprises, in order from the amino terminus to the carboxyl terminus, a VL, a linker (e.g., a glycine-serine linker), and a VH.

[0237] Thus, in some embodiments, the 4-1BB×OX40 bispecific antibody comprises a polypeptide that, in order from the amino terminus to the carboxyl terminus, comprises the VH of the 4-1BB binding domain, a linker (e.g., a glycine-serine linker), the VL of the 4-1BB binding domain, an IgG1 hinge, an immunoglobulin constant region comprising a CH2 domain and a CH3 domain, a linker (e.g., a glycine-serine linker), the VL of the OX40 binding domain, a linker (e.g., a glycine-serine linker), and the VH of the OX40 binding domain. In some embodiments, the 4-1BB×OX40 bispecific antibody comprises a dimer of such a polypeptide.

[0238] In some embodiments, the 4-1BB×OX40 bispecific antibody comprises a protein scaffold as generally disclosed, for example, in U.S. Patent Application Publication Nos. 2003 / 0133939, 2003 / 0118592, and 2005 / 0136049. The 4-1BB×OX40 bispecific antibody can comprise a dimer (e.g., a homodimer) of two peptides, each comprising, in order from amino terminus to carboxyl terminus: a first antigen-binding domain, a linker (e.g., the linker is a hinge region), and an immunoglobulin constant region. In other embodiments, the 4-1BB×OX40 bispecific antibody comprises a protein scaffold as generally disclosed, for example, in U.S. Patent Application Publication No. 2009 / 0148447. The 4-1BB / OX40 antibody can comprise a dimer (e.g., a homodimer) of two polypeptides, each comprising, in order from amino terminus to carboxyl terminus: an immunoglobulin constant region, a linker (e.g., the linker is a hinge region), and a first antigen-binding domain.

[0239] In some embodiments, the 4-1BB×OX40 bispecific antibody comprises two antigen-binding domains that are scFvs and two antigen-binding domains that comprise VH and VL on separate polypeptides. In such embodiments, the scFv can be fused to the N-terminus or C-terminus of the polypeptide comprising VH. The scFv can also be fused to the N-terminus or C-terminus of the polypeptide comprising VL.

[0240] Additional exemplary bispecific antibody molecules of the invention include: (i) antibodies having two arms, each containing two different antigen-binding regions, one with specificity for 4-1BB and one with specificity for OX40; (ii) antibodies having one antigen-binding region or arm specific for 4-1BB and a second antigen-binding region or arm specific for OX40; (iii) single-chain antibodies having a first specificity for 4-1BB and a second specificity for OX40 via, for example, two scFvs tandemly linked by an additional peptide linker; (iv) dual variable domain antibodies (DVD-Ig) (Wu et al., Generation and Characterization of a Dual Variable Domain Immunoglobulin (DVD-Ig™) Molecule, In: Antibody Engineering, Springer Berlin Heidelberg (2010)) in which each light and heavy chain contains two variable domains in tandem through short peptide linkages; (v) chemically conjugated bispecific F(ab’) 2 fragments; (vi) Tandab, which is a fusion of two single-chain diabodies and results in a tetravalent bispecific antibody having two binding sites for each of the target antigens; (vii) Flexibody, which is a combination of an scFv and a diabody that results in a multivalent molecule; (viii) so-called “dock-and-lock” molecules based on the “dimerization and docking domain” in protein kinase A that can generate a trivalent bispecific protein consisting of two identical Fab fragments bound to different Fab fragments when applied to Fab; (ix) so-called scorpion molecules containing, for example, two scFvs fused to both ends of a human Fab arm; and (x) diabodies.

[0241] Examples of different classes of bispecific antibodies include IgG-like molecules with complementary CH3 domains to enforce heterodimerization; recombinant IgG-like dual-targeting molecules where each of the two sites of the molecule contains a Fab fragment or a part of a Fab fragment of two different antibodies; IgG fusion molecules where a full-length IgG antibody is fused to an additional Fab fragment or a part of a Fab fragment; Fc fusion molecules where a single-chain Fv molecule or a stabilized diabody is fused to a heavy-chain constant domain, an Fc region or a part thereof; Fab fusion molecules where different Fab fragments are fused together; ScFv-based and diabody-based heavy-chain antibodies (e.g., domain antibodies, nanobodies) where different single-chain Fv molecules or different diabodies or different heavy-chain antibodies (e.g., domain antibodies, nanobodies) are fused to each other or to another protein or carrier molecule, but are not limited thereto.

[0242] Examples of Fab fusion bispecific antibodies include F(ab) 2 (Medarex / AMGEN), bispecific or bis-Fab (Genentech), Dock-and-Lock (DNL) (Immunomedics), bivalent bispecific (Biotecnol) and Fab-Fv (UCB-Celltech), but are not limited thereto. Examples of ScFv-based and diabody-based domain antibodies include bispecific T cell engaging antibodies (BITE) (Micromet, tandem diabody (Tandab) (Affimed), dual affinity retargeting technology (D.A.R.T) (Macrogenics), single-chain diabody (academic), TCR-like antibodies (AIT, ReceptorLogics), human serum albumin ScFv fusion (Merimac) and COMBODY (Epigen Biotech), dual-targeting nanobodies (Ablynx), and dual-targeting heavy-chain only domain antibodies, but are not limited thereto.

[0243] As provided herein, the 4-1BB×OX40 bispecific antibody can comprise a 4-1BB VH CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 5-7, respectively, a 4-1BB VL CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 8-10, respectively, an OX40 VH CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 11-13, respectively, and an OX40 VL CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 14-16, respectively.

[0244] As provided herein, the 4-1BB×OX40 bispecific antibody can comprise a 4-1BB VH CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 5, 119, and 7, respectively, a 4-1BB VL CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 120-122, respectively, an OX40 VH CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 11-13, respectively, and an OX40 VL CDR1 sequence, CDR2 sequence, and CDR3 sequence of SEQ ID NOs: 14-16.

[0245] As provided herein, the 4-1BB×OX40 bispecific antibody can comprise any combination of the 4-1BB VH and VL sequences and the OX40 VH and VL sequences provided herein.

[0246] For example, a 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain, the 4-1BB binding domain comprising a VH comprising the amino acid sequence of SEQ ID NO: 17 and a VL comprising the amino acid sequence of SEQ ID NO: 18, and the OX40 binding domain comprising (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (ii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (iii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (iv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (v) a VH comprising the amino acid sequence of SEQ ID NO: 33 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (vi) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 34, (vii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35, (viii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 36, (ix) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 37, (x) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 34, (xi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 35, (xii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 36, (xiii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 37, (xiv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 38, (xv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 39, (xvi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 40, (xvii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 41, or (xviii) a VH comprising the amino acid sequence of SEQ ID NO: 25 and a VL comprising the amino acid sequence of SEQ ID NO: 26.In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide containing one 4-1BBscFv and one OX40scFv). In some embodiments, one polypeptide contains both VH sequences and another polypeptide contains both VL sequences.

[0247] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain, the 4-1BB binding domain comprising a VH comprising the amino acid sequence of SEQ ID NO: 32 and a VL comprising the amino acid sequence of SEQ ID NO: 18, and the OX40 binding domain comprising (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (ii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (iii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (iv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30, (v) a VH comprising the amino acid sequence of SEQ ID NO: 33 and a VL comprising the amino acid sequence of SEQ ID NO: 28, (vi) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (vii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35; (viii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 36; (ix) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 37; (x) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (xi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 35, (xii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 36, (xiii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 37, (xiv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 38, (xv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 39, (xvi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 40, (xvii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 41, or (xviii) a VH comprising the amino acid sequence of SEQ ID NO: 25 and a VL comprising the amino acid sequence of SEQ ID NO: 26.In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide containing one 4-1BBscFv and one OX40scFv). In some embodiments, one polypeptide contains both VH sequences and another polypeptide contains both VL sequences.

[0248] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain, wherein the 4-1BB binding domain comprises a VH comprising the amino acid sequence of SEQ ID NO: 23 and a VL comprising the amino acid sequence of SEQ ID NO: 24, and the OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (ii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 30; (iii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (iv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30; (v) a VH comprising the amino acid sequence of SEQ ID NO: 33 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (vi) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (vii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35; (viii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 36; (ix) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 37; (x) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (xi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 35; (xii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 36; (xiii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 37; (xiv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 38; (xv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 39; (xvi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 40; (xvii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 41; or (xviii) a VH comprising the amino acid sequence of SEQ ID NO: 25 and a VL comprising the amino acid sequence of SEQ ID NO: 26.In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide comprising one 4-1BBscFv and one OX40scFv). In some embodiments, one polypeptide comprises both VH sequences and another polypeptide comprises both VL sequences.

[0249] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain. The 4-1BB binding domain comprises a VH comprising the amino acid sequence of SEQ ID NO: 19 and a VL comprising the amino acid sequence of SEQ ID NO: 20. The OX40 binding domain comprises (i) a VH comprising the amino acid sequence of SEQ ID NO: 20 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (ii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 30; (iii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (iv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 30; (v) a VH comprising the amino acid sequence of SEQ ID NO: 33 and a VL comprising the amino acid sequence of SEQ ID NO: 28; (vi) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (vii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 35; (viii) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 36; (ix) a VH comprising the amino acid sequence of SEQ ID NO: 29 and a VL comprising the amino acid sequence of SEQ ID NO: 37; (x) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 34; (xi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 35; (xii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 36; (xiii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 37; (xiv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 38; (xv) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 39; (xvi) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 40; (xvii) a VH comprising the amino acid sequence of SEQ ID NO: 31 and a VL comprising the amino acid sequence of SEQ ID NO: 41; or (xviii) a VH comprising the amino acid sequence of SEQ ID NO: 25 and a VL comprising the amino acid sequence of SEQ ID NO: 26.In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide comprising one 4-1BBscFv and one OX40scFv). In some embodiments, one polypeptide comprises both VH sequences and another polypeptide comprises both VL sequences.

[0250] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain. The 4-1BB binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 17 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 17, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 5-7 respectively) and a VL having the amino acid sequence of SEQ ID NO: 18 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 8-10 respectively). The OX40 binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 29 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 17, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of 11-13 respectively) and a VL having the amino acid sequence of SEQ ID NO: 28 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 28, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 14-16 respectively). In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide comprising one 4-1BBscFv and one OX40scFv).In some embodiments, one polypeptide comprises both VH sequences and another polypeptide comprises both VL sequences.

[0251] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain. The 4-1BB binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 17 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 17, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 5-7, respectively) and a VL having the amino acid sequence of SEQ ID NO: 18 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 8-10, respectively). The OX40 binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 31 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 31, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 11-13, respectively) and a VL having the amino acid sequence of SEQ ID NO: 30 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 30, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 14-16, respectively). In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide comprising one 4-1BBscFv and one OX40scFv).In some embodiments, one polypeptide comprises both VH sequences and another polypeptide comprises both VL sequences.

[0252] The 4-1BB×OX40 bispecific antibody can comprise a 4-1BB binding domain and an OX40 binding domain. The 4-1BB binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 17 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 17, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 5-7 respectively) and a VL having the amino acid sequence of SEQ ID NO: 18 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 18, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 8-10 respectively). The OX40 binding domain can comprise a VH having the amino acid sequence of SEQ ID NO: 29 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 29, optionally, the VH comprises a VH CDR1 sequence, a VH CDR2 sequence, and a VH CDR3 sequence of SEQ ID NOs: 11-13 respectively) and a VL having the amino acid sequence of SEQ ID NO: 35 (or a sequence that is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99% identical to SEQ ID NO: 35, optionally, the VL comprises a VL CDR1 sequence, a VL CDR2 sequence, and a VL CDR3 sequence of SEQ ID NOs: 14-16 respectively). In some embodiments, both VH sequences and both VL sequences are on a single polypeptide chain (e.g., a single polypeptide comprising one 4-1BBscFv and one OX40scFv).In some embodiments, one polypeptide comprises both VH sequences and another polypeptide comprises both VL sequences.

[0253] As provided herein, the 4-1BB×OX40 bispecific antibody can comprise any combination of the 4-1BB scFv sequences and OX40 scFv sequences provided herein. For example, the 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 59. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 60. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 61. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 62. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 63 and 59. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 63 and 60. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 63 and 61. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 63 and 62. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 44 and 59. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 44 and 64. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 64. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 65. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 66. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 67. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 68. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 69. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 70. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 71. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 72. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NOs: 58 and 73.The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NO: 58 and 74. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NO: 58 and 75. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NO: 58 and 76. The 4-1BB×OX40 bispecific antibody can comprise the scFvs of SEQ ID NO: 145 and 146. Such scFv pairs can be on the same polypeptide or on separate polypeptides. When the scFv pair is on the same polypeptide, the 4-1BB scFv can be N-terminal to the OX40 scFv or the 4-1BB scFv can be C-terminal to the OX40 scFv.

[0254] As provided herein, an antibody or polypeptide comprising any CDR, VH, VL, and / or scFv sequence provided herein can further comprise a hinge. The hinge can be located, for example, between the 4-1BB binding domain (e.g., scFv) and the immunoglobulin constant region. The hinge can also be located between the OX40 binding domain (e.g., scFv) and the immunoglobulin constant region. In some embodiments, the polypeptide comprises, in order from amino terminus to carboxyl terminus, an antigen binding domain (e.g., scFv), a hinge region, and an immunoglobulin constant region.

[0255] The hinge can be an immunoglobulin hinge, such as a human IgG hinge. In some embodiments, the hinge is a human IgG 1 hinge. In some embodiments, the hinge is a human IgG 1 comprises the sequence of amino acids 216-230 (by EU numbering) or a sequence that is at least 90% identical thereto. For example, the hinge is a human IgG 1can include substitutions with amino acid C220 according to EU numbering. When derived from non-human sources, the hinge can be humanized. In some embodiments, the hinge includes amino acids 1-15 of SEQ ID NO: 115. Non-limiting examples of the hinge are provided in Tables K and L below.

[0256] In certain embodiments, the hinge comprises or is a sequence that is at least 80%, at least 81%, at least 82%, at least 83%, at least 84%, at least 85%, at least 86%, at least 87%, at least 88%, at least 89%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or at least 99% identical to a wild-type immunoglobulin hinge region such as a wild-type human IgG1 hinge, a wild-type human IgG2 hinge, or a wild-type human IgG4 hinge.

[0257] A representative modified immunoglobulin hinge includes an immunoglobulin human IgG1 hinge region having one, two, or three cysteine residues found within the wild-type human IgG1 hinge substituted with one, two, or three different amino acid residues (e.g., serine or alanine). The modified immunoglobulin hinge can additionally have a proline substituted with another amino acid (e.g., serine or alanine). For example, the modified human IgG1 hinge described above can additionally have a proline located at the carboxyl terminus relative to the three cysteines of the wild-type human IgG1 hinge region substituted with another amino acid residue (e.g., serine, alanine). In one embodiment, the proline in the core hinge region is not substituted.

[0258] In certain embodiments, the hinge comprises from about 5 to 150 amino acids, from 5 to 10 amino acids, from 10 to 20 amino acids, from 20 to 30 amino acids, from 30 to 40 amino acids, from 40 to 50 amino acids, from 50 to 60 amino acids, from 5 to 60 amino acids, from 5 to 40 amino acids, from 8 to 20 amino acids, or from 10 to 15 amino acids. The hinge is primarily flexible but can also impart higher rigidity characteristics or can comprise primarily an α-helix structure with a minimal β-sheet structure. The length of the hinge sequence can affect the binding affinity of the binding domain to which the hinge is directly or indirectly connected (via another region or domain) as well as one or more activities of the Fc region moiety to which the hinge or linker is directly or indirectly connected.

[0259] In certain embodiments, the hinge is stable in plasma and serum and is resistant to proteolytic cleavage. The first lysine within the IgG1 upper hinge region can be mutated to minimize proteolytic cleavage. For example, the lysine can be substituted with methionine, threonine, alanine, or glycine or can be deleted.

[0260] In some embodiments, the 4-1BB×OX40 bispecific antibody does not comprise a hinge. For example, in some embodiments, the 4-1BB×OX40 bispecific antibody comprises a linker instead of a hinge.

[0261] As provided herein, an antibody or polypeptide comprising any CDR, VH, VL, scFv, and / or hinge provided herein can further comprise an immunoglobulin constant region. The immunoglobulin constant region can be located, for example, between a hinge and a 4-1BB binding domain (e.g., 4-1BB binding scFv). The immunoglobulin constant region can also be located between a hinge and an OX40 binding domain (e.g., OX40 binding scFv). In some embodiments, the polypeptide comprises, in order from amino terminus to carboxyl terminus, a hinge region, an immunoglobulin constant region, and an antigen binding domain (e.g., scFv).

[0262] In some embodiments, the immunoglobulin constant region comprises the immunoglobulin CH2 and CH3 domains of IgG1, IgG2, IgG3, IgG4, IgA1, IgA2, or IgD, and optionally, the IgG is human. In some cases, the immunoglobulin constant region comprises the immunoglobulin CH2 and CH3 domains of IgG1 (e.g., human IgG1). In some embodiments, the polypeptide does not comprise a CH1 domain.

[0263] In some embodiments, the immunoglobulin constant region comprises one, two, three, four, five, or more amino acid substitutions and / or deletions that prevent binding to FcγR1, FcγRIIa, FcγRIIb, FcγRIIa, and FcγRIIIb.

[0264] In certain embodiments, the immunoglobulin constant region comprises one, two, three, or more amino acid substitutions that prevent or reduce Fc-mediated T cell activation.

[0265] In some embodiments, the immunoglobulin constant region comprises one, two, three, four or more amino acid substitutions and / or deletions that prevent or reduce CDC activity and / or ADCC activity. In some embodiments, the immunoglobulin constant region comprises one, two, three, four, five or more amino acid substitutions and / or deletions that prevent or attenuate FcγR or C1q interaction.

[0266] The present invention includes antibodies (e.g., the antibody of SEQ ID NO: 81) having a human 4-1BB antigen-binding domain comprising the CDRs of VH of SEQ ID NO: 17 and the CDRs of VL of SEQ ID NO: 18 and a human OX40 antigen-binding domain comprising the CDRs of VH of SEQ ID NO: 31 and the CDRs of VL of SEQ ID NO: 30. In this embodiment, the human 4-1BB antigen-binding domain and the human OX40 antigen-binding domain can be separated by a "null" constant region comprising mutations that prevent binding to FcγR1, FcγRIIa, FcγRIIb, FcγRIIa, and FcγRIIIb. Such a "null" constant region allows the bispecific antibodies of the present invention to activate tumor-infiltrating lymphocytes while not activating or minimizing the activation of other effector cells. The presence of the constant region extends the half-life of the bispecific antibody compared to similar bispecific antibodies without a constant region.

[0267] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain comprising the substitutions L234A, L235A, G237A, and K322A according to the EU numbering system.

[0268] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain comprising one or more of the following substitutions according to the EU numbering system: E233P, L234A, L234V, L235A, G237A, E318A, K320A, and K322A, and / or a deletion of G236.

[0269] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain that, according to the EU numbering system, contains one or more of the following substitutions: E233P, L234A, L234V, L235A, G237A, and K322A, and / or a deletion of G236.

[0270] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain that, according to the EU numbering system, contains the substitutions L234A, L235A, G237A, E318A, K320A, and K322A.

[0271] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain that, according to the EU numbering system, contains the substitutions L234A, L235A, G237A, and K322A.

[0272] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain that, according to the EU numbering system, contains the substitutions E233P, L234V, L235A, G237A, and K322A.

[0273] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain that, according to the EU numbering system, contains the substitutions E233P, L234V, L235A, G237A, and K322A, and a deletion of G236.

[0274] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain with substitutions E233P, L234A, L235A, G237A, and K322A according to the EU numbering system. For example, the present invention includes bispecific antibodies comprising, from amino terminus to carboxyl terminus, a first scFv, an immunoglobulin hinge, an IgG1 CH2 domain with substitutions E233P, L234A, L235A, G237A, and K322A according to the EU numbering system, IgG1 CH3, and a second scFv. In one embodiment, the first scFv specifically binds to human 4-1BB and the second scFv specifically binds to human OX40. In one embodiment, the first scFv specifically binds to human OX40 and the second scFv specifically binds to human OX40.

[0275] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH2 domain with substitutions E233P, L234A, L235A, G237A, and K322A and a deletion of G236 according to the EU numbering system. For example, the present invention includes bispecific antibodies comprising, from amino terminus to carboxyl terminus, a first scFv, an immunoglobulin hinge, an IgG1 CH2 with substitutions E233P, L234A, L235A, G237A, and K322A and a deletion of G236 according to the EU numbering system, IgG1 CH3, and a second scFv. In one embodiment, the first scFv specifically binds to human 4-1BB and the second scFv specifically binds to human OX40. In one embodiment, the first scFv specifically binds to human OX40 and the second scFv specifically binds to human OX40.

[0276] In certain embodiments, the immunoglobulin constant region comprises a human IgG1 CH3 domain.

[0277] In certain embodiments, the immunoglobulin constant region comprises amino acids 16-231 of SEQ ID NO: 111, 112, or 114 or amino acids 16-230 of SEQ ID NO: 113 or 115. In certain embodiments, the immunoglobulin constant region comprises amino acids 16-230 of SEQ ID NO: 115.

[0278] Additional immunoglobulin constant regions that can be present in the 4-1BB×OX40 antibodies provided herein are discussed in more detail below.

[0279] In some embodiments, the hinge and immunoglobulin constant region comprise the amino acid sequence of any one of SEQ ID NOs: 111-115. In some embodiments, the hinge and immunoglobulin constant region comprise the amino acid sequence of SEQ ID NO: 115.

[0280] In some embodiments, the 4-1BB×OX40 bispecific antibody does not comprise an immunoglobulin constant region. In some embodiments, the 4-1BB×OX40 bispecific antibody does not comprise a hinge and does not comprise an immunoglobulin constant region.

[0281] As provided herein, an antibody or polypeptide comprising any CDR, VH, VL, scFv, hinge, and / or immunoglobulin constant region provided herein can further comprise a linker. The linker can be located, for example, between the immunoglobulin constant region and the C-terminal binding domain. For example, the linker can be located between the immunoglobulin constant region and the C-terminal 4-1BB binding domain. The linker can also be located between the immunoglobulin constant region and the C-terminal OX40 binding domain. In some embodiments, the polypeptide comprises, in order from the amino terminus to the carboxyl terminus, an immunoglobulin constant region, a linker, and an antigen binding domain.

[0282] In some embodiments, the linker (e.g., between the immunoglobulin constant region and the antigen-binding domain) comprises from 3 to 30 amino acids, from 3 to 15 amino acids, or from about 3 to 10 amino acids. In some embodiments, the linker (e.g., between the immunoglobulin constant region and the antigen-binding domain) comprises from 5 to 30 amino acids, from 5 to 15 amino acids, or from about 5 to 10 amino acids. In some embodiments, the linker (e.g., between the immunoglobulin constant region and the antigen-binding domain) comprises the amino acid sequence (Gly 4 Ser) n where n = 1-5 (SEQ ID NO: 117), and optionally n = 1. In some embodiments, the linker (e.g., between the immunoglobulin constant region and the antigen-binding domain) comprises the amino acid sequence GGGSPS (SEQ ID NO: 118). In some embodiments, the linker (e.g., between the immunoglobulin constant region and the antigen-binding domain) comprises the amino acid sequence of SEQ ID NO: 109 or 110.

[0283] Non-limiting examples of linkers are given in Tables K and L below.

[0284]

Table 12

[0285] In some embodiments, the 4-1BB×OX40 antibody comprises, in order from the amino terminus to the carboxyl terminus, (i) a VH comprising the amino acid sequence of SEQ ID NO: 17, (ii) a linker (e.g., a glycine-serine linker), (iii) a VL comprising the amino acid sequence of SEQ ID NO: 18, (iv) an IgG comprising a C220S substitution according to EU numbering 1An immunoglobulin constant region comprising a hinge, (v) a CH2 domain containing the following substitutions according to the EU numbering system: E233P, L234A, L234V, L235A, G237A, and K322A, and a deletion of G236, and a wild-type CH3 domain, (vi) a VL containing the amino acid sequence of SEQ ID NO: 28, (vii) a linker (e.g., a glycine-serine linker), and (viii) a VH containing the amino acid sequence of SEQ ID NO: 29. In some embodiments, the 4-1BB×OX40 antibody comprises a dimer of such a polypeptide.

[0286] In some embodiments, the 4-1BB×OX40 antibody, from amino terminus to carboxyl terminus in order, (i) a VH containing the amino acid sequence of SEQ ID NO: 17, (ii) a linker (e.g., a glycine-serine linker), (iii) a VL containing the amino acid sequence of SEQ ID NO: 18, (iv) an IgG containing a C220S substitution according to EU numbering 1 An immunoglobulin constant region comprising a hinge, (v) a CH2 domain containing the following substitutions according to the EU numbering system: E233P, L234A, L234V, L235A, G237A, and K322A, and a deletion of G236, and a wild-type CH3 domain, (vi) a VL containing the amino acid sequence of SEQ ID NO: 30, (vii) a linker (e.g., a glycine-serine linker), and (viii) a VH containing the amino acid sequence of SEQ ID NO: 31. In some embodiments, the 4-1BB×OX40 antibody comprises a dimer of such a polypeptide.

[0287] In some embodiments, the 4-1BB×OX40 antibody, from amino terminus to carboxyl terminus in order, (i) a VH containing the amino acid sequence of SEQ ID NO: 17, (ii) a linker (e.g., a glycine-serine linker), (iii) a VL containing the amino acid sequence of SEQ ID NO: 18, (iv) an IgG containing a C220S substitution according to EU numbering 1A polypeptide comprising a hinge, (v) a CH2 domain containing the following substitutions according to EU numbering: E233P, L234A, L234V, L235A, G237A, and K322A, and a deletion of G236, and a wild-type CH3 domain, (vi) a VL containing the amino acid sequence of SEQ ID NO: 35, (vii) a linker (e.g., a glycine-serine linker), and (viii) a VH containing the amino acid sequence of SEQ ID NO: 29. In some embodiments, the 4-1BB×OX40 antibody comprises a dimer of such a polypeptide.

[0288] In some embodiments, the 4-1BB×OX40 bispecific antibody comprises any one of the amino acid sequences of SEQ ID NOs: 78-100.

[0289] [Table 13]

[0290] In some embodiments, the 4-1BB×OX40 bispecific antibody comprises the amino acid sequence of SEQ ID NO: 78. In some embodiments, the 4-1BB×OX40 bispecific antibody comprises the amino acid sequence of 81. In some embodiments, the 4-1BB×OX40 bispecific antibody comprises the amino acid sequence of SEQ ID NO: 90. In some embodiments, the 4-1BB×OX40 bispecific antibody consists essentially of the amino acid sequence of SEQ ID NO: 78. In some embodiments, the 4-1BB×OX40 bispecific antibody consists essentially of the amino acid sequence of 81. In some embodiments, the 4-1BB×OX40 bispecific antibody consists essentially of the amino acid sequence of SEQ ID NO: 90. In some embodiments, the 4-1BB×OX40 bispecific antibody consists of the amino acid sequence of SEQ ID NO: 78. In some embodiments, the 4-1BB×OX40 bispecific antibody consists of the amino acid sequence of 81. In some embodiments, the 4-1BB×OX40 bispecific antibody consists of the amino acid sequence of SEQ ID NO: 90.

[0291] In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, and comprises two polypeptides, each polypeptide comprising an amino acid sequence selected from the group consisting of SEQ ID NOs: 78-100.

[0292] In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, and comprises two identical polypeptides, each polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% or more identical to the amino acid sequence of SEQ ID NO: 78. In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer comprising two polypeptides, each polypeptide comprising the amino acid sequence of SEQ ID NO: 78. In some embodiments, the bispecific antibody that binds to human 4-1BB and human OX40 is a dimer consisting of or consisting essentially of two polypeptides, each polypeptide comprising the amino acid sequence of SEQ ID NO: 78.

[0293] In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, and comprises two identical polypeptides, each polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% or more identical to the amino acid sequence of SEQ ID NO: 81.

[0294] In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, and comprises two polypeptides, each polypeptide comprising the amino acid sequence of SEQ ID NO: 81. In some embodiments, the bispecific antibody that binds to human 4-1BB and human OX40 is a dimer consisting of or consisting essentially of two polypeptides, each polypeptide comprising the amino acid sequence of SEQ ID NO: 81.

[0295] In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, comprises two identical polypeptides, and each polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% or more identical to the amino acid sequence of SEQ ID NO: 90. In some embodiments, the 4-1BB×OX40 bispecific antibody is a homodimer capable of binding to human 4-1BB and human OX40, comprises two polypeptides, and each polypeptide comprises the amino acid sequence of SEQ ID NO: 90. In some embodiments, the bispecific antibody that binds to human 4-1BB and human OX40 is a dimer consisting of or consisting essentially of two polypeptides, and each polypeptide comprises the amino acid sequence of SEQ ID NO: 90.

[0296] The bispecific antibodies of the present invention are capable of lysing tumor cells. "Capable" means that the bispecific antibody can perform its activity under appropriate laboratory conditions. Tumor lysis can be determined in vitro and in vivo using methods known in the art. For example, tumor lysis can be evaluated by co-culturing PBMCs (or purified T cells) and tumor cells with an anti-CD3×anti-tumor associated antigen (TAA) bispecific molecule (CD3×TAA-induced antibody). The CD3×TAA-induced antibody is a polyclonal stimulator of T cells, which signals the T cells and results in upregulation of 4-1BB and OX40. In this type of experiment, the CD3×TAA-induced antibody is added to the culture at sub-optimal concentrations, while the addition of the anti-4-1BB and anti-OX40 bispecific antibody (e.g., the antibody comprising SEQ ID NO: 81) to the culture further enhances target cell lysis in a dose-dependent manner induced by the CD3×TAA-induced antibody. Similarly, lysis of target cells can also be evaluated using a chromium 51 release assay.

[0297] Tumor lysis can also be evaluated using a syngeneic tumor model using host mice expressing human 4-1BB and human OX40 (e.g., mice expressing human 4-1BB and human OX40 under the control of the corresponding endogenous mouse promoter gene, e.g., Biocytogen, female B-hOX40 / h4-1BB mice from China (C57BL / 6-Tnfrsf4 tm1(TNFRSF4) CD137 tm1(CD137) / Bcgen)). For example, a mouse can be inoculated with a syngeneic tumor strain such as MB49 tumor cells or MC38 tumor cells. When tumor growth becomes visible, for example, on about day 6, an anti-4-1BB and anti-OX40 bispecific antibody or a control antibody can be administered (e.g., intraperitoneally). A decrease in tumor size in mice treated with the anti-4-1BB and anti-OX40 bispecific antibody compared to mice treated with the control antibody suggests that the bispecific antibody can lyse tumor cells. Tumor lysis can also be evaluated in a xenograft model of immunodeficient mice transplanted with human T cells administered in combination with a CD3 bispecific inducing antibody to prime the T cells.

[0298] In one embodiment, the antibody of the present invention is thermostable. The antibody of the present invention exhibits improved stability over many prior art antibodies (e.g., those disclosed in US Patent Publication No. 2018 / 0118841 and US Patent Publication No. 2015 / 0307620). Tm is a measure of thermostability and can be determined by methods known in the art (e.g., according to any of the methods described in the Examples). In one embodiment, the bispecific antibody of the present invention has a Tm of about 63, 64, 65, 66, 67, 68 or 69. For example, the present invention includes a bispecific antibody comprising a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 17 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 18, and a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 31 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 30, wherein the bispecific antibody has a Tm of 64-68.

[0299] In another embodiment, the antibodies of the present invention have a theoretical pI of less than 7.5, less than 7.6, less than 7.7, less than 7.8, less than 7.9, or less than 8. The theoretical pI can be determined by methods known in the art (e.g., according to any of the methods described in the Examples). In one embodiment, the present invention includes a VH containing an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 17 and a VL containing an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 18 for the human 4-1BB binding domain, and a VH containing an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 31 and a VL containing an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 30 for the human OX40 binding domain, and the bispecific antibody has a pI of less than 7.8. E.4-1BB and OX40 Monospecific Antibodies

[0300] As used herein, monospecific antibodies that bind to either human 4-1BB or human OX40 are provided. The anti-4-1BB antibodies provided herein can include one or more of any of the 4-1BB binding domains described herein. The anti-OX40 antibodies provided herein can include one or more of any of the anti-OX40 binding domains described herein.

[0301] In some embodiments, the anti-4-1BB antibodies or anti-OX40 antibodies provided herein are IgG antibodies. In some embodiments, the anti-4-1BB antibodies or anti-OX40 antibodies provided herein are IgG 1 antibodies.

[0302] In some embodiments, the anti-4-1BB antibody comprises the six CDRs of SEQ ID NOs: 5-10, the six CDRs of SEQ ID NOs: 5, 119, 7, 120, 121, and 122, or a combination of the 4-1BB-binding VH and VL sequences provided herein and a heavy chain constant region. In some embodiments, the anti-4-1BB antibody comprises the six CDRs of SEQ ID NOs: 5-10, the six CDRs of SEQ ID NOs: 5, 119, 7, 120, 121, and 122, or a combination of the 4-1BB-binding VH and VL sequences provided herein and a light chain constant region. In some embodiments, the anti-4-1BB antibody comprises the six CDRs of SEQ ID NOs: 5-10, the six CDRs of SEQ ID NOs: 5, 119, 7, 120, 121, and 122, or a combination of the 4-1BB-binding VH and VL sequences provided herein and, a heavy chain constant region, and a light chain constant region.

[0303] In some embodiments, the anti-OX40 antibody comprises the six CDRs of SEQ ID NOs: 11-16, or a combination of the OX40-binding VH and VL sequences provided herein and a heavy chain constant region. In some embodiments, the anti-OX40 antibody comprises the six CDRs of SEQ ID NOs: 11-16, or a combination of the OX40-binding VH and VL sequences provided herein and a light chain constant region. In some embodiments, the anti-OX40 antibody comprises the six CDRs of SEQ ID NOs: 11-16, or a combination of the OX40-binding VH and VL sequences provided herein and, a heavy chain constant region, and a light chain constant region.

[0304] The constant region of the anti-4-1BB antibody or OX40 antibody can be any constant region discussed herein. The constant regions present in these antibodies are discussed in more detail below.

[0305] In some embodiments, the anti-4-1BB antibody or anti-OX40 antibody is a Fab, Fab’, F(ab’) 2, scFv, disulfide-linked Fv, or scFv-Fc. In some embodiments, the anti-4-1BB antibody or anti-OX40 antibody is Fab, Fab’, F(ab’) 2 , scFv, disulfide-linked Fv, or scFv-Fc. For example, the present invention includes anti-4-1BB antibodies or anti-OX40 antibodies in the SMIP format (i.e., scFv-Fc), as disclosed in U.S. Patent No. 9,005,612. SMIP antibodies can include, from amino-terminus to carboxy-terminus, an scFv and a modified constant domain comprising an immunoglobulin hinge and CH2 / CH3 regions. The present invention also includes anti-4-1BB antibodies or anti-OX40 antibodies in the PIMS format, as disclosed in published U.S. Patent Application No. 2009 / 0148447. PIMS antibodies can include, from amino-terminus to carboxy-terminus, an immunoglobulin hinge and CH2 / CH3 regions, and a modified constant domain comprising an scFv.

[0306] The anti-4-1BB antibody can be monovalent for 4-1BB (i.e., comprising one 4-1BB binding domain), divalent for 4-1BB (i.e., comprising two 4-1BB binding domains), or can have three or more 4-1BB binding domains.

[0307] The anti-OX40 antibody can be monovalent for OX40 (i.e., comprising one OX40 binding domain), divalent for OX40 (i.e., comprising two OX40 binding domains), or can have three or more OX40 binding domains. F. Constant Region

[0308] As discussed above, the antibodies provided herein, including the single specificity antibodies that bind 4-1BB or OX40 and the 4-1BB×OX40 bispecific antibodies, can include an immunoglobulin constant region. In certain embodiments, the immunoglobulin constant region does not interact with Fc gamma receptors.

[0309] In certain embodiments, the antibodies described herein that immunospecifically bind to 4-1BB and / or OX40 comprise a VH domain and a VL domain comprising any of the amino acid sequences described herein, and the constant region comprises the amino acid sequence of a IgG, IgE, IgM, IgD, IgA, or IgY immunoglobulin molecule, or a human IgG, IgE, IgM, IgD, IgA, or IgY immunoglobulin molecule. In another specific embodiment, the antibodies described herein that immunospecifically bind to 4-1BB and / or OX40 comprise a VH domain and a VL domain comprising any of the amino acid sequences described herein, and the constant region comprises the amino acid sequence of a IgG, IgE, IgM, IgD, IgA, or IgY immunoglobulin molecule, any class of immunoglobulin molecule (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2), or any subclass (e.g., IgG2a and IgG2b). In certain embodiments, the constant region comprises the amino acid sequence of a human IgG, IgE, IgM, IgD, IgA, or IgY immunoglobulin molecule, any class of immunoglobulin molecule (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2), or any subclass (e.g., IgG2a and IgG2b).

[0310] In one embodiment, the heavy chain constant region is human IgG 1 is a heavy chain constant region, and the light chain constant region is a human IgGκ light chain constant region.

[0311] In some embodiments, the constant region comprises one, two, three or more amino acid substitutions that prevent binding to FcγR1, FcγRIIa, FcγRIIb, FcγRIIa, and FcγRIIIb.

[0312] In certain embodiments, the constant region comprises one, two, three or more amino acid substitutions that prevent or reduce Fc-mediated T cell activation.

[0313] In some embodiments, the constant region comprises one, two, three or more amino acid substitutions that prevent or reduce CDC activity and / or ADCC activity.

[0314] In some embodiments, one, two, or more mutations (e.g., amino acid substitutions) are introduced into the antibody or antigen-binding fragment thereof described herein (e.g., numbering according to the Kabat numbering system (e.g., EU index in Kabat)) in the CH2 domain (residues 231-340 of human IgG 1 and / or the CH3 domain (residues 341-447 of human IgG 1 and / or the Fc region of the hinge region, altering one or more functional properties of the antibody or antigen-binding fragment thereof such as serum half-life, complement binding reaction, Fc receptor binding, and / or antibody-dependent cell cytotoxicity activity.

[0315] In certain embodiments, one, two, or more mutations (e.g., amino acid substitutions) are introduced into the hinge region (CH1 domain) of the Fc region, such that the number of cysteine residues in the hinge region is altered (e.g., increased or decreased) as described, for example, in U.S. Patent No. 5,677,425. The number of cysteine residues in the hinge region of the CH1 domain can be altered (e.g., increased or decreased) to, for example, facilitate the association of the light and heavy chains or to alter the stability of the antibody or antigen-binding fragment thereof.

[0316] In some embodiments, one, two, or more mutations (e.g., amino acid substitutions) are introduced into the Fc region (e.g., numbered according to the Kabat numbering system (e.g., EU index in Kabat) of the antibody or antigen-binding fragment thereof described herein, the CH2 domain (residues 231-340 of human IgG1) and / or the CH3 domain (residues 341-447 of human IgG1)) and / or the hinge region, to increase or decrease the affinity of the antibody or antigen-binding fragment thereof for Fc receptors (e.g., activating Fc receptors) on the surface of effector cells. Mutations in the Fc region that increase or decrease the affinity for Fc receptors, and techniques for introducing such mutations into Fc receptors or fragments thereof, are known to those of skill in the art. Examples of mutations in Fc receptors that can alter the affinity of an antibody or antigen-binding fragment thereof for Fc receptors are described, for example, in Smith P et al., (2012) PNAS 109:6181-6186, U.S. Patent No. 6,737,056, and International Publications WO02 / 060919; WO98 / 23289; and WO97 / 34631, which are incorporated herein by reference.

[0317] In certain embodiments, one, two, or more amino acid mutations (i.e., substitutions, insertions or deletions) are introduced into the IgG constant domain, or its FcRn-binding fragment (preferably, the Fc or hinge Fc domain fragment), to alter (e.g., decrease or increase) the in vivo half-life of an antibody or its antigen-binding fragment. For examples of mutations that alter (e.g., decrease or increase) the in vivo half-life of an antibody or its antigen-binding fragment, see, e.g., International Publication Nos. WO02 / 060919; WO98 / 23289; and WO97 / 34631; and U.S. Patent Nos. 5,869,046, 6,121,022, 6,277,375 and 6,165,745. In some embodiments, one, two or more amino acid mutations (i.e., substitutions, insertions, or deletions) are introduced into the IgG constant domain, or its FcRn-binding fragment (preferably, the Fc or hinge Fc domain fragment), to shorten the in vivo half-life of an antibody or its antigen-binding fragment. In other embodiments, one, two or more amino acid mutations (i.e., substitutions, insertions or deletions) are introduced into the IgG constant domain, or its FcRn-binding fragment (preferably, the Fc or hinge Fc domain fragment), to extend the in vivo half-life of an antibody or its antigen-binding fragment. In certain embodiments, the antibody or its antigen-binding fragment may have one or more amino acid mutations (e.g., substitutions) in the second constant (CH2) domain (residues 231-340 of human IgG1) and / or the third constant (CH3) domain (residues 341-447 of human IgG1), as numbered by the EU index according to Kabat (Kabat EA et al., (1991), supra). In certain embodiments, the constant region of IgG1, as numbered by the EU index according to Kabat, includes a substitution of methionine (M) to tyrosine (Y) at position 252, a substitution of serine (S) to threonine (T) at position 254, and a substitution of threonine (T) to glutamic acid (E) at position 256. See U.S. Patent No. 7,658,921, which is incorporated herein by reference.This type of mutant IgG, also referred to as "YTE mutation", has been shown to exhibit a four-fold extended half-life compared to the wild-type version of the same antibody (see Dall’Acqua WF et al., (2006) J Biol Chem 281:23514-24). In certain embodiments, the antibody or antigen-binding fragment thereof comprises an IgG constant domain having one, two, three, or more amino acid substitutions at positions 251-257, 285-290, 308-314, 385-389, and 428-436, numbered according to the EU index in Kabat.

[0318] In further embodiments, one, two, or more amino acid substitutions are introduced into the Fc region of the IgG constant domain to alter the effector function(s) of the antibody or antigen-binding fragment thereof. For example, by numbering according to the EU index in Kabat, one or more amino acids selected from amino acid residues 234, 235, 236, 237, 297, 318, 320, and 322 can be replaced with different amino acid residues, such that the antibody or antigen-binding fragment thereof has an altered affinity for effector ligands but retains the antigen-binding ability of the parental antibody. Effector ligands whose affinity is altered can be, for example, Fc receptors or the C1 component of complement. This approach is described in more detail in U.S. Patent Nos. 5,624,821 and 5,648,260. In some embodiments, deletion or inactivation of the constant domain (through point mutations or other means) may reduce the binding of the Fc receptor of the circulating antibody or antigen-binding fragment thereof, thereby enhancing tumor localization. See, for example, U.S. Patent Nos. 5,585,097 and 8,591,886 for descriptions of mutations that delete or inactivate the constant domain and thereby enhance tumor localization. In certain embodiments, one or more amino acid substitutions can be introduced into the Fc region to remove potential glycosylation sites thereon, which may reduce Fc receptor binding (see, for example, Shields RL et al., (2001) J Biol Chem 276:6591-604).

[0319] In certain embodiments, one or more amino acids selected from amino acid residues 329, 331, and 322 in the constant region can be replaced with different amino acid residues such that the antibody or its antigen-binding fragment has altered C1q binding and / or reduced or abolished complement-dependent cytotoxicity (CDC) by EU-index numbering at Kabat. This endeavor is described in more detail in U.S. Patent No. 6,194,551 (Idusogie et al). In some embodiments, one or more amino acid residues within positions 231 to 238 in the N-terminal region of the CH2 domain are altered, thereby altering the ability of the antibody to fix complement. This endeavor is further described in International Publication WO94 / 29351. In certain embodiments, the Fc region is modified to enhance the ability of the antibody or its antigen-binding fragment to mediate antibody-dependent cell-mediated cytotoxicity (ADCC) and / or to enhance the affinity of the antibody or its antigen-binding fragment for the Fcγ receptor by mutating (e.g., introducing amino acid substitutions) one or more amino acids at the following positions: 238, 239, 248, 249, 252, 254, 255, 256, 258, 265, 267, 268, 269, 270, 272, 276, 278, 280, 283, 285, 286, 289, 290, 292, 293, 294, 295, 296, 298, 301, 303, 305, 307, 309, 312, 315, 320, 322, 324, 326, 327, 328, 329, 330, 331, 333, 334, 335, 337, 338, 340, 360, 373, 376, 378, 382, 388, 389, 398, 414, 416, 419, 430, 434, 435, 437, 438, or 439 by EU-index numbering at Kabat. This endeavor is further described in International Publication WO00 / 42072.

[0320] In certain embodiments, the antibodies or antigen-binding fragments thereof described herein comprise a constant domain of IgG1 having a mutation (e.g., substitution) at position 267, position 328, or a combination thereof, numbered according to the EU index in Kabat. In certain embodiments, the antibodies or antigen-binding fragments thereof described herein comprise a constant domain of IgG1 having a mutation (e.g., substitution) selected from the group consisting of S267E, L328F, and combinations thereof. In certain embodiments, the antibodies or antigen-binding fragments thereof described herein comprise a constant domain of IgG1 having an S267E / L328F mutation (e.g., substitution). In certain embodiments, the antibodies or antigen-binding fragments thereof described herein that comprise a constant domain of IgG1 having an S267E / L328F mutation (e.g., substitution) have improved binding affinity for FcγRIIA, FcγRIIB, or both FcγRIIA and FcγRIIB.

[0321] In certain embodiments, any of the constant region mutations or modifications described herein can be introduced into one or both of the heavy chain constant regions of the antibodies or antigen-binding fragments thereof described herein that have two heavy chain constant regions. III. Antibody Production

[0322] Antibodies that immunospecifically bind to human 4-1BB and / or human OX40 can be produced by any method known in the art for the synthesis of antibodies, e.g., by chemical synthesis or by recombinant expression techniques. The methods described herein use conventional techniques in the related fields within the scope of techniques possessed by those skilled in the art, including, but not limited to, molecular biology, microbiology, genetic analysis, recombinant DNA, organic chemistry, biochemistry, PCR, oligonucleotide synthesis and modification, nucleic acid hybridization, unless otherwise specified. These techniques are described, for example, in the references cited herein and are fully explained therein.For example, see Maniatis T et al., (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press; Sambrook J et al., (1989), Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor Laboratory Press; Sambrook J et al., (2001) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; Ausubel FM et al., Current Protocols in Molecular Biology, John Wiley & Sons (1987 and updated annually); Current Protocols in Immunology, John Wiley & Sons (1987 and updated annually); Gait (ed.) (1984) Oligonucleotide Synthesis: A Practical Approach, IRL Press; Eckstein (ed.) (1991) Oligonucleotides and Analogues: A Practical Approach, IRL Press; Birren B et al., (eds.) (1999) Genome Analysis: A Laboratory Manual, Cold Spring Harbor Laboratory Press.

[0323] Bispecific antibodies as provided herein can be prepared by expressing a polynucleotide in a host cell, the polynucleotide encoding a polypeptide that, in order from amino terminus to carboxyl terminus, comprises a first scFv, a hinge region, an immunoglobulin constant region, and a second scFv, wherein (a) the first scFv comprises a human 4-1BB antigen-binding domain and the second scFv comprises a human OX40 antigen-binding domain, or (b) the first scFv comprises a human OX40 antigen-binding domain and the second scFv comprises a human 4-1BB antigen-binding domain. The polypeptide can be expressed in the host cell as a dimer.

[0324] Bispecific antibodies as provided herein can be prepared by chemically conjugating two different monoclonal antibodies or by fusing two hybridoma cell lines to produce a hybrid-hybridoma. Bispecific bivalent antibodies, and methods for making them, are described, for example, in U.S. Pat. Nos. 5,731,168, 5,807,706, 5,821,333, and U.S. Patent Application Publication Nos. 2003 / 020734 and 2002 / 0155537; each of which is incorporated herein by reference in its entirety. Bispecific tetravalent antibodies, and methods for making them, are described, for example, in International Application Publication Nos. WO02 / 096948 and WO00 / 44788; the disclosures of both of which are incorporated herein by reference in their entirety. Generally, see International Application Publication Nos. WO93 / 17715, WO92 / 08802, WO91 / 00360, and WO92 / 05793; Tutt et al., J. Immunol. 147:60-69 (1991); U.S. Pat. Nos. 4,474,893; 4,714,681; 4,925,648; 5,573,920; and 5,601,819; and Kostelny et al., J. Immunol. 148:1547-1553 (1992); each of which is incorporated herein by reference in its entirety.

[0325] Bispecific antibodies as described herein can be generated according to the DuoBody technology platform (Genmab A / S), for example, as described in International Publications WO2011 / 131746, WO2011 / 147986, WO2008 / 119353, and WO2013 / 060867, and in Labrijn AF et al., (2013) PNAS 110(13):5145-5150. The DuoBody technology can be used to combine half of a first monospecific antibody comprising two heavy chains and two light chains and half of a second monospecific antibody comprising two heavy chains and two light chains. The resulting heterodimer comprises one heavy chain and one light chain from the first antibody paired with one heavy chain and one light chain from the second antibody. If both monospecific antibodies recognize different epitopes for different antigens, the resulting heterodimer is a bispecific antibody.

[0326] The DuoBody technology requires that each monospecific antibody comprises a heavy chain constant region having a single point mutation within the CH3 domain. The point mutation enables a stronger interaction between the CH3 domains in the resulting bispecific antibody than between the CH3 domains in either of the monospecific antibodies. The single point mutation in each monospecific antibody is, for example, as described in International Publication WO2011 / 131746, at residue 366, 368, 370, 399, 405, 407, or 409 in the CH3 domain of the heavy chain constant region according to the EU numbering system. Further, the single point mutation is located at a different residue in one monospecific antibody compared to the other monospecific antibody. For example, one monospecific antibody can comprise the mutation F405L (i.e., a mutation from phenylalanine to leucine at residue 405) according to the EU numbering system, while the other monospecific antibody can comprise the mutation K409R (i.e., a mutation from lysine to arginine at residue 409). The heavy chain constant region of the monospecific antibody is IgG 1 IgG 2 IgG3 or IgG 4 an isotype (e.g., human IgG 1 isotype), and bispecific antibodies produced by DuoBody technology can retain Fc-mediated effector functions.

[0327] Another method for generating bispecific antibodies is referred to as the "knobs-into-holes" strategy (see, e.g., International Publication WO2006 / 028936). In this technique, mispairing of Ig heavy chains is reduced by mutating selected amino acids that form the interface of the CH3 domain in IgG. At positions within the CH3 domain where the two heavy chains interact directly, amino acids with small side chains (holes) are introduced into the sequence of one heavy chain, and amino acids with large side chains (knobs) are introduced into the paired interacting residues located in the other heavy chain. In some embodiments, the compositions of the invention have immunoglobulin chains in which the CH3 domain is modified by mutating selected amino acids that interact at the interface between two polypeptides to preferentially form bispecific antibodies. Bispecific antibodies can be composed of immunoglobulin chains of the same subclass (e.g., IgG1 or IgG3) or different subclasses (e.g., IgG1 and IgG3, or IgG3 and IgG4).

[0328] In one embodiment, the bispecific antibody that binds to 4-1BB and OX40 has a T366W mutation in the "knob chain" (numbering according to the EU numbering system), and T366S, L368A, Y407V mutations in the "hole chain", and optionally, for example, a Y349C mutation in the "knob chain" and an E356C or S354C mutation in the "hole chain"; an R409D, K370E mutation in the "knob chain" and a D399K, E357K mutation in the "hole chain"; an R409D, K370E mutation in the "knob chain" and a D399K, E357K mutation in the "hole chain"; a T366W mutation in the "knob chain" and T366S, L368A, Y407V mutations in the "hole chain"; an R409D, K370E mutation in the "knob chain" and a D399K, E357K mutation in the "hole chain"; a Y349C, T366W mutation in one chain and an E356C, T366S, L368A, Y407V mutation in the paired chain; a Y349C, T366W mutation in one chain and an S354C, T366S, L368A, Y407V mutation in the paired chain; a Y349C, T366W mutation in one chain and an S354C, T366S, L368A, Y407V mutation in the paired chain; a Y349C, T366W mutation in one chain and an S354C, T366S, L368A, Y407V mutation in the paired chain, including additional inter-chain disulfide bridges between the CH3 domains by introducing these mutations.

[0329] The bispecific antibody that binds to 4-1BB and OX40 can, in some cases, comprise heterodimers of chains of IgG4 and IgG1, IgG4 and IgG2, IgG4 and IgG2, IgG4 and IgG3, or IgG1 and IgG3. Such heterodimeric heavy chain antibodies can always be engineered, for example, by modifying selected amino acids that form the interface of the CH3 domains in human IgG4 and IgG1 or IgG3 so as to promote heterodimeric heavy chain formation.

[0330] The bispecific antibodies described herein can be generated by any technique known to those of skill in the art. For example, the F(ab’) described herein 2Fragments can be produced by proteolytic cleavage of the immunoglobulin molecule using an enzyme such as pepsin.

[0331] In one aspect, provided herein is a method for producing an antibody that immunospecifically binds to human 4-1BB and / or human OX40, comprising culturing the cells or cell populations described herein. In one aspect, provided herein is a method for producing an antibody that immunospecifically binds to human 4-1BB and / or human OX40, comprising expressing (e.g., recombinantly expressing) the antibody using the cells or host cells described herein (e.g., cells or host cells comprising a polynucleotide encoding the antibody described herein). In certain embodiments, the cells are isolated cells. In certain embodiments, the exogenous polynucleotide has been introduced into the cells. In certain embodiments, the method further comprises the step of purifying the antibody from the cells or host cells.

[0332] Monoclonal antibodies are known in the art and can be produced using, for example, hybridoma technology including those taught in Harlow E & Lane D, Antibodies: A Laboratory Manual, (Cold Spring Harbor Laboratory Press, 2nd ed. 1988); Hammerling GJ et al.,: Monoclonal Antibodies and T-Cell Hybridomas 563 681 (Elsevier, N.Y., 1981). The term "monoclonal antibody" as used herein is not limited to antibodies produced through hybridoma technology. For example, monoclonal antibodies can be recombinantly produced from host cells that ectopically express the antibodies described herein. The monoclonal antibodies described herein can be made, for example, by the hybridoma method as described in Kohler G & Milstein C (1975) Nature 256:495 or, for example, can be isolated from phage libraries using techniques as described herein. Other methods for the preparation of clonal cell lines and the monoclonal antibodies expressed thereby are well known in the art (see, for example, Chapter 11 of Short Protocols in Molecular Biology, (2002) 5th Ed., Ausubel FM et al., supra).

[0333] Furthermore, the antibodies described herein can also be generated using various phage display methods known in the art. In phage display methods, proteins are presented on the surface of phage particles that carry the polynucleotide sequences encoding them. In particular, the DNA sequences encoding the VH and VL domains are amplified from an animal cDNA library (e.g., a human or mouse cDNA library of diseased tissue). The DNA encoding the VH and VL domains is recombined by PCR together with an scFv linker and cloned into a phagemid vector. The vector is electroporated into Escherichia coli (E. coli), and the E. coli is infected with helper phage. The phages used in these methods are typically filamentous phages including fd and M13, and the VH and VL domains are usually recombinantly fused to either phage gene III or gene VIII. Phages expressing antibodies that bind to a particular antigen can be selected or identified using the antigen, e.g., a labeled antigen or an antigen bound or captured on a solid surface or beads.Examples of phage display methods that can be used to create the antibodies described herein include Brinkman U et al., (1995) J Immunol Methods 182:41-50; Ames RS et al., (1995) J Immunol Methods 184:177-186; Kettleborough CA et al., (1994) Eur J Immunol 24:952-958; Persic L et al., (1997) Gene 187:9-18; Burton DR & Barbas CF (1994) Advan Immunol 57:191-280; PCT application PCT / GB91 / 001134; International Publications WO90 / 02809, WO91 / 10737, WO92 / 01047, WO92 / 18619, WO93 / 11236, WO95 / 15982, WO95 / 20401, and WO97 / 13844; and U.S. Patents Nos. 5,698,426, 5,223,409, 5,403,484, 5,580,717, 5,427,908, 5,750,753, 5,821,047, 5,571,698, 5,427,908, 5,516,637, 5,780,225, 5,658,727, 5,733,743, and 5,969,108.

[0334] As described in the above references, after phage selection, the antibody coding region can be isolated from the phage and used to generate antibodies, including human antibodies, which can be expressed in any desired host, including, for example, mammalian cells, insect cells, plant cells, yeast, and bacteria, as described below. Fab, Fab’ and F(ab) 2Techniques for recombinantly producing antibodies such as fragments can also be employed using methods known in the art, such as those disclosed in PCT Publication WO92 / 22324; Mullinax RL et al., (1992) BioTechniques 12(6):864-9; Sawai H et al., (1995) Am J Reprod Immunol 34:26-34; and Better M et al., (1988) Science 240:1041-1043.

[0335] In one aspect, to generate an antibody, a VH nucleotide sequence or a VL nucleotide sequence, a restriction site, and a PCR primer containing a flanking sequence for protecting the restriction site can be used to amplify a VH sequence or a VL sequence from a template, such as an scFv clone. Using cloning techniques known to those skilled in the art, the PCR-amplified VH domain can be cloned into a vector expressing a VH constant region, and the PCR-amplified VL domain can be cloned into a vector expressing a VL constant region, such as a human kappa or lambda constant region. The VH domain and the VL domain can also be cloned into one vector expressing the required constant region. Subsequently, the heavy chain conversion vector and the light chain conversion vector are co-transfected into a cell line, and using techniques known to those skilled in the art, a stable cell line or a transient cell line expressing an antibody, such as IgG, is generated.

[0336] A humanized antibody can bind to a given antigen and includes a framework region having substantially the amino acid sequence of a human immunoglobulin and CDRs having substantially the amino acid sequence of a non-human immunoglobulin (e.g., a mouse immunoglobulin). In certain embodiments, a humanized antibody also includes an immunoglobulin constant region (Fc), typically at least a portion of a human immunoglobulin. The antibody can also include the CH1 region, hinge region, CH2 region, CH3 region, and CH4 region of the heavy chain. Humanized antibodies include any class of immunoglobulin, including IgM, IgG, IgD, IgA, and IgE, as well as IgG1 、IgG 2 、IgG 3 and IgG 4It can be selected from any isotype including. Humanized antibodies can be produced using various techniques known in the art, including but not limited to CDR grafting (European Patent EP239400; International Publication WO91 / 09967; and US Patents 5,225,539, 5,530,101, and 5,585,089), veneering or resurfacing (European Patents EP592106 and EP519596; Padlan EA (1991) Mol Immunol 28(4 / 5):489-498; Studnicka GM et al., (1994) Prot Engineering 7(6):805-814; and Roguska MA et al., (1994) PNAS 91:969-973), chain shuffling (US Patent 5,565,332), and, for example, US Patent 6,407,213, US Patent 5,766,886, International Publication WO93 / 17105; Tan P et al., (2002) J Immunol 169:1119-25; Caldas C et al., (2000) Protein Eng.13(5):353-60; Morea V et al., (2000) Methods 20(3):267-79; Baca M et al., (1997) J Biol Chem 272(16):10678-84; Roguska MA et al., (1996) Protein Eng 9(10):895 904; Couto JR et al., (1995) Cancer Res.55(23 Supp):5973s-5977s; Couto JR et al., (1995) Cancer Res 55(8):1717-22; Sandhu JS(1994) Gene 150(2):409-10 and Pedersen JT et al., (1994) J Mol Biol 235(3):959-73. Also, see US Application Publication US2005 / 0042664A1 (February 24, 2005), which is incorporated herein by reference in its entirety. IV. Polynucleotide Encoding Antibody

[0337] In certain embodiments, the present disclosure encompasses polynucleotides encoding 4-1BB and / or OX40, or polypeptides of such antibodies, such as VH, VL, VH having VL (e.g., in scFv), heavy chain, light chain, heavy chain having scFv, light chain having scFv, fusion proteins containing scFv, linkers (e.g., the linker is a hinge), immunoglobulin constant regions, and antibodies that bind to scFv, constant regions, or constant regions having scFv.

[0338] Accordingly, provided herein are polynucleotides or combinations of polynucleotides encoding the six CDRs of SEQ ID NOs: 5-10. The polynucleotides can include nucleotide sequences set forth as nucleotides 76-99, 151-174, 289-330, 502-519, 571-579, and 688-714 of SEQ ID NO: 147, respectively.

[0339] Also provided herein are polynucleotides or combinations of polynucleotides encoding the six CDRs of SEQ ID NOs: 5, 119, 7, 120, 121, and 122.

[0340] Also provided herein are polynucleotides or combinations of polynucleotides encoding the six CDRs of SEQ ID NOs: 11-16. The polynucleotides can include nucleotide sequences set forth as nucleotides 1912-1935, 1987-2010, 2125-2145, 1528-1545, 1597-1605, and 1714-1746 of SEQ ID NO: 147, respectively.

[0341] Also provided herein are polynucleotides or combinations of polynucleotides encoding the six CDRs of SEQ ID NOs: 5-11 and the six CDRs of SEQ ID NOs: 11-16.

[0342] Also provided herein are polynucleotides or combinations of polynucleotides encoding the six CDRs of SEQ ID NOs: 5, 119, 7, 120, 121, and 122 and the six CDRs of SEQ ID NOs: 11-16.

[0343] Also provided herein are polynucleotides encoding a VH provided herein, for example, a VH comprising the amino acid sequence of SEQ ID NO: 17, 19, 21, 23, 25, 27, 29, 31-33, or 143. The polynucleotide can comprise a nucleotide sequence exemplified as nucleotides 1-363 of SEQ ID NO: 147; 1837-2178 of SEQ ID NO: 147; nucleotides 1-363 of SEQ ID NO: 148; 1837-2178 of SEQ ID NO: 148; nucleotides 1-363 of SEQ ID NO: 149; or 1837-2178 of SEQ ID NO: 149.

[0344] Also provided herein are polynucleotides encoding a VL provided herein, for example, a VL comprising the amino acid sequence of SEQ ID NO: 18, 20, 22, 24, 26, 28, 30 or 34-41. The polynucleotide can comprise a nucleotide sequence exemplified as nucleotides 424-744 of SEQ ID NO: 147; 1453-1776 of SEQ ID NO: 147; 424-744 of SEQ ID NO: 148; 1453-1776 of SEQ ID NO: 148; 424-744 of SEQ ID NO: 149; or 1453-1776 of SEQ ID NO: 149.

[0345] Also provided herein are polynucleotides encoding a 4-1BB binding sequence (e.g., scFv) provided herein, for example, a 4-1BB binding sequence comprising the amino acid sequence of SEQ ID NO: 42-45, 58, 63, 77, or 101. The polynucleotide can comprise a nucleotide sequence exemplified as nucleotides 1-744 of SEQ ID NO: 147; 1-744 of SEQ ID NO: 148; or 1-744 of SEQ ID NO: 149.

[0346] Also provided herein are polynucleotides encoding OX40-binding sequences (e.g., scFv) provided herein, such as OX40-binding sequences comprising the amino acids of SEQ ID NOs: 46-57, 59-76, or 102. The polynucleotide can comprise a nucleotide sequence such as nucleotides 1453-2181 of SEQ ID NO: 147; 1453-2181 of SEQ ID NO: 148; or 1453-2181 of SEQ ID NO: 149.

[0347] Also provided herein are polynucleotides encoding antibodies provided herein, such as 4-1BB×OX40 bispecific antibodies, e.g., antibodies comprising the amino acid sequences of SEQ ID NOs: 78-100. The polynucleotide can comprise a nucleotide sequence represented by any one of SEQ ID NOs: 147-149.

[0348] In certain embodiments, the polynucleotide encodes a polypeptide comprising, in order from amino terminus to carboxyl terminus, a first scFv, a linker (e.g., the linker is a hinge region), an immunoglobulin constant region, and a second scFv, wherein (a) the first scFv comprises a human 4-1BB antigen-binding domain and the second scFv comprises a human OX40 antigen-binding domain, or (b) the first scFv comprises a human OX40 antigen-binding domain and the second scFv comprises a human 4-1BB antigen-binding domain.

[0349] Also provided are vectors comprising the polynucleotides disclosed herein, as discussed in more detail below.

[0350] The polynucleotides of the invention can be in the form of RNA or DNA. DNA includes cDNA, genomic DNA, and synthetic DNA, and can be double-stranded or single-stranded, and the single-stranded can be either the coding strand or the non-coding (antisense) strand in some cases. In some embodiments, the polynucleotide is cDNA or DNA lacking another endogenous intron.

[0351] In some embodiments, the polynucleotide is a non-naturally occurring polynucleotide. In some embodiments, the polynucleotide is recombinantly produced.

[0352] In certain embodiments, the polynucleotide is isolated. In certain embodiments, the polynucleotide is substantially pure. In some embodiments, the polynucleotide is purified from natural components.

[0353] In some embodiments, the polynucleotides provided herein are codons optimized for expression in a particular host (changing the codons in human mRNA to those preferred by a bacterial host such as E. coli). V. Cells and Vectors

[0354] Vectors and cells comprising the polynucleotides described herein are also provided herein.

[0355] In certain aspects, cells (e.g., host cells) that express (e.g., recombinantly) an antibody described herein that specifically binds to 4-1BB and / or OX40 and that comprises the relevant polynucleotide and expression vector are provided herein. A vector (e.g., an expression vector) comprising a polynucleotide comprising a nucleotide sequence encoding an antibody that specifically binds to 4-1BB and / or OX40 for recombinant expression in a host cell, e.g., a mammalian host cell, is provided herein. Also provided herein is a host cell comprising such a vector for recombinant expression of an antibody that specifically binds to 4-1BB and / or OX40 described herein. In certain embodiments, a method for producing an antibody that specifically binds to 4-1BB and / or OX40 described herein is provided herein, which comprises expressing such an antibody in a host cell.

[0356] The recombinant expression of an antibody that specifically binds to 4-1BB and / or OX40 as described herein involves the construction of an expression vector comprising a polynucleotide encoding the antibody or its polypeptide (e.g., scFv, a linker (e.g., the linker is a hinge), an immunoglobulin constant region; a heavy or light chain; a polypeptide comprising one or more variable domains; a polypeptide comprising one or more antigen-binding domains (e.g., scFv) optionally fused to a linker (e.g., the linker is a hinge), an immunoglobulin constant region and / or a linker, etc., a fusion protein). Once a polynucleotide encoding the antibody or its polypeptide as described herein is obtained, the vector for the production of the antibody or its polypeptide can be produced by recombinant DNA techniques using techniques well known in the art. Thus, methods for preparing a protein by expressing a polynucleotide are described herein, wherein the nucleotide sequence encoding the antibody or its fragment is provided. An expression vector containing a coding sequence for the antibody or its polypeptide and appropriate transcriptional and translational control signals can be constructed using methods well known to those skilled in the art. These methods include, for example, in vitro recombinant DNA techniques, synthetic techniques, and in vivo gene recombination. Also provided are replicable vectors containing a nucleotide sequence encoding an antibody or its fragment operably linked to a promoter. Such vectors can, for example, contain a nucleotide sequence encoding the constant region of the antibody molecule (see, for example, International Publication Nos. WO86 / 05807 and WO89 / 01036; and U.S. Patent No. 5,122,464), and the variable domain of the antibody can be cloned into such vectors for expression of the entire heavy chain, the entire light chain, or both the entire heavy chain and the entire light chain. Additional variable domains, 4-1BB binding domains (e.g., scFv), and / or OX40 binding domains can also be cloned into such vectors for expression of a fusion protein comprising a heavy or light chain fused to an additional variable domain, 4-1BB binding domain (e.g., scFv), and / or OX40 binding domain.

[0357] To direct the recombinant protein into the secretory pathway of the host cell, a secretory signal sequence (also known as a leader sequence) can be provided in the expression vector. The secretory signal sequence can be the native one of the recombinant protein or can be derived from another secreted or de novo synthesized protein. The secretory signal sequence can be operably linked to the DNA sequence encoding the polypeptide. The secretory signal sequence is generally located 5' to the DNA sequence encoding the polypeptide of interest, although some signal sequences can be located elsewhere in the DNA sequence of interest (see, for example, Welch et al., U.S. Patent No. 5,037,743; Holland et al., U.S. Patent No. 5,143,830).

[0358] The expression vector can be transferred into cells (e.g., host cells) by conventional techniques, and the resulting cells can then be cultured by conventional techniques to produce the antibodies or polypeptides thereof described herein (e.g., scFv, linker (e.g., the linker is a hinge), immunoglobulin constant region; heavy or light chain; polypeptide containing one or more variable domains; polypeptide containing one or more antigen-binding domains (e.g., scFv) optionally fused to a hinge, immunoglobulin constant region and / or linker, etc., a fusion protein). Accordingly, host cells containing a polynucleotide encoding an antibody or polypeptide thereof described herein operably linked to a promoter are provided herein for expression of such sequences within the host cell.

[0359] In certain embodiments, for the expression of a multi-polypeptide antibody, the vectors encoding all the polypeptides can be co-expressed within the host cell for the expression of the whole antibody individually.

[0360] In certain embodiments, the host cell comprises a vector comprising a polynucleotide encoding the entire polypeptide of the antibody described herein. In specific embodiments, the host cell comprises a plurality of different vectors encoding the entire polypeptide of the antibody described herein.

[0361] The vector or combination of vectors can comprise polynucleotides encoding two or more polypeptides that interact to form the antibodies described herein, such as a first polynucleotide encoding a heavy chain and a second polynucleotide encoding a light chain; a first polynucleotide encoding a fusion protein comprising a heavy chain and an scFv, having a second polynucleotide encoding a light chain; a first polynucleotide encoding a fusion protein comprising a light chain and an scFv, having a second polynucleotide encoding a heavy chain; a first polynucleotide encoding a fusion protein comprising a heavy chain and a VH, having a second polynucleotide encoding a fusion protein comprising a light chain and a VL, and the like. When two polypeptides are encoded by polynucleotides in two separate vectors, the vectors can be transfected into the same host cell.

[0362] Using various host expression vector systems, the antibodies or polypeptides thereof described herein (e.g., scFv, linker (e.g., the linker is a hinge), immunoglobulin constant region; heavy or light chain; polypeptide containing one or more variable domains; polypeptide containing one or more antigen-binding domains (e.g., scFv), optionally fused to a hinge, immunoglobulin constant region and / or linker, etc.) can be expressed. Such host expression systems represent a medium capable of producing the coding sequence of interest and subsequently purifying it, but also represent cells that can express the antibodies or polypeptides thereof described herein in situ when transformed or transfected with the appropriate nucleotide coding sequence.These include microorganisms such as bacteria (e.g., Escherichia coli (E. coli) and Bacillus subtilis) transformed with a recombinant bacteriophage DNA, plasmid DNA or cosmid DNA expression vector containing an antibody coding sequence; yeast (e.g., Saccharomyces pichia) transformed with a recombinant yeast expression vector containing an antibody coding sequence; insect cell lines infected with a recombinant virus expression vector (e.g., baculovirus) containing an antibody coding sequence; plant cell lines (e.g., green algae such as Chlamydomonas reinhardtii) infected with a recombinant virus expression vector (e.g., cauliflower mosaic virus, CaMV; tobacco mosaic virus, TMV) or transformed with a recombinant plasmid expression vector (e.g., Ti plasmid); or mammalian cell lines (e.g., COS (e.g., COS1 or COS), CHO, BHK, MDCK, HEK293, NS0, PER.C6, VERO, CRL7O3O, HsS78Bst, HeLa, and NIH3T3, HEK-293T, HepG2, SP210, R1.1, B-W, L-M, BSC1, BSC40, YB / 20 and BMT10 cells) holding a recombinant expression construct containing a promoter derived from the genome of a mammalian cell (e.g., metallothionein promoter) or a mammalian virus (e.g., adenovirus late promoter; vaccinia virus 7.5K promoter), but not limited thereto.

[0363] When the antibody or polypeptide thereof described herein (e.g., scFv, linker (e.g., the linker is a hinge), immunoglobulin constant region; heavy or light chain; polypeptide containing one or more variable domains; polypeptide containing one or more antigen-binding domains (e.g., scFv) optionally fused to a hinge, immunoglobulin constant region and / or linker, etc.) is produced by recombinant expression, it can be purified by any method known in the art for antibody purification, e.g., chromatography (e.g., ion exchange, affinity, especially protein A, and affinity for a specific antigen after sizing column chromatography), centrifugation, by differences in solubility, or by any other standard technique for protein purification. Further, the antibodies described herein can be fused to the heterologous polypeptide sequences described herein (e.g., FLAG tag, his tag, or avidin) or others known in the art to facilitate purification. VI. Compositions and Kits

[0364] Compositions are provided herein that contain an antibody having the desired purity described herein in a physiologically acceptable carrier, excipient, or stabilizer (Remington’s Pharmaceutical Sciences (1990) Mack Publishing Co., Easton, PA). The acceptable carrier, excipient, or stabilizer is non-toxic to the recipient at the dosages and concentrations employed.

[0365] The pharmaceutical composition can be formulated for a particular route of administration to a subject. For example, the pharmaceutical composition can be formulated for parenteral, e.g., intravenous administration. Compositions used for in vivo administration can be sterilized. This can be readily accomplished, for example, by filtration through sterile filtration membranes.

[0366] The pharmaceutical composition described herein is, in one embodiment, for use as a medicament. The pharmaceutical composition described herein may be useful for enhancing an immune response. The pharmaceutical composition described herein may be useful for enhancing the proliferation of natural killer (NK) cells and / or T cells (e.g., CD4 T cells and / or CD8 T cells) in a subject. The pharmaceutical composition described herein may be useful for stimulating the T cell co-stimulation pathway in a subject.

[0367] The pharmaceutical composition described herein may be useful for treating a health condition such as cancer. Examples of cancers that can be treated as described herein include, but are not limited to, melanoma, renal cancer, pancreatic cancer, lung cancer, bowel cancer, prostate cancer, breast cancer, liver cancer, brain cancer, and blood cancers such as lymphoma. In certain cases, the cancer is a solid tumor. VII. Methods and Uses

[0368] The antibodies of the disclosure that bind to 4-1BB and / or OX40 are useful in a variety of applications including, but not limited to, therapeutic treatment methods such as the treatment of cancer. In certain embodiments, the medicament is useful for inhibiting tumor growth and / or reducing tumor volume. The method of use can be an in vitro or in vivo method. The invention includes the use of any of the disclosed antibodies (and pharmaceutical compositions comprising the disclosed antibodies) for use in therapy.

[0369] The disclosure provides a method of treating cancer in a subject comprising administering to the subject a therapeutically effective amount of an antibody that binds to 4-1BB and / or OX40. The invention includes the use of any of the disclosed antibodies for the treatment of cancer.

[0370] In certain embodiments, the cancer is a cancer including, but not limited to, melanoma, renal cancer, pancreatic cancer, lung cancer, colon / intestinal cancer, stomach cancer, prostate cancer, ovarian cancer, breast cancer, liver cancer, brain cancer, and blood cancer. The cancer can be a primary tumor or a progressive or metastatic cancer. In some cases, the cancer is a solid tumor. For example, the present disclosure includes the use of bispecific antibodies for the treatment of sarcomas, carcinomas, and lymphomas. The present invention includes, for example, treating a human subject having a sarcoma, carcinoma, or lymphoma by administering a therapeutically effective amount of a pharmaceutical composition of the present invention (e.g., a pharmaceutical composition comprising a bispecific antibody that specifically binds to human 4-1BB and human OX40 and comprises an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group of SEQ ID NOs: 78-100 and 144).

[0371] The present invention includes a method of treating a human subject having a tumor or cancerous tissue comprising tumor infiltrating lymphocytes. The present invention includes treating a human subject having a tumor comprising lymphocytes expressing 4-1BB and OX40. In one embodiment, the present invention includes administering to a human subject having a solid tumor a therapeutically effective amount of a pharmaceutical composition comprising an anti-4-1BB × anti-OX40 bispecific antibody, wherein the human 4-1BB binding domain comprises a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 17 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 18, and the human OX40 binding domain comprises a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 31 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 30. For example, the present invention includes administering to a human subject having a tumor an effective amount of a pharmaceutical composition comprising an anti-4-1BB × anti-OX40 bispecific antibody, wherein the human 4-1BB binding domain comprises an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 58, and the human OX40 binding domain comprises an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 62. In one embodiment, the present invention includes administering to a human subject having a tumor a therapeutically effective amount of a pharmaceutical composition comprising an anti-4-1BB × anti-OX40 bispecific antibody comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 81.

[0372] The present disclosure provides a method of enhancing an immune response in a subject, comprising administering to the subject a therapeutically effective amount of an antibody that binds to 4-1BB and / or OX40.

[0373] The present disclosure provides a method of stimulating a T cell co-stimulation pathway in a subject, comprising administering to the subject a therapeutically effective amount of an antibody that binds to 4-1BB and / or OX40.

[0374] The present disclosure provides a method for enhancing the proliferation of NK cells and / or T cells (e.g., CD4+ T cells and / or CD8+ T cells) in a subject, which comprises administering to the subject a therapeutically effective amount of an antibody that binds to 4-1BB and / or OX40. The present disclosure provides a method for enhancing the proliferation of NK cells, CD4+ T cells, and CD8+ T cells in a subject, which comprises administering to the subject a therapeutically effective amount of an antibody that binds to 4-1BB and OX40. For example, the present invention includes administering a therapeutically effective amount of a pharmaceutical composition comprising a bispecific antibody that specifically binds to human 4-1BB and human OX40 and has an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 78 to 100 and 144, to enhance the proliferation of NK cells, CD4+ T cells, and CD8+ T cells in a subject.

[0375] The present invention includes a method for increasing the number of tumor-infiltrating lymphocytes in a subject by administering to the subject a therapeutically effective amount of the pharmaceutical composition of the present invention. For example, the present invention includes a method for increasing the number of tumor-infiltrating lymphocytes in a subject by administering a pharmaceutical composition comprising a bispecific antibody that specifically binds to human 4-1BB and human OX40 and has an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to an amino acid sequence selected from the group consisting of SEQ ID NOs: 78 to 100 and 144.

[0376] The present invention also includes a method for improving the expression of granzyme by tumor-infiltrating lymphocytes in a subject by administering to the subject a therapeutically effective amount of the antibody or pharmaceutical composition of the present invention. For example, the present invention includes a method for improving the expression of granzyme by tumor-infiltrating lymphocytes in a subject by administering a therapeutically effective amount of any antibody or pharmaceutical composition provided herein.

[0377] In certain embodiments, the subject is human.

[0378] Administration of an antibody that binds to 4-1BB and / or OX40 can be parenteral administration, including intravenous administration.

[0379] In some embodiments, an antibody that binds to 4-1BB and / or OX40, or a pharmaceutical composition comprising the same, for use as a medicament is provided herein. In some aspects, an antibody that binds to 4-1BB and / or OX40, or a pharmaceutical composition comprising the same, for use in a method for the treatment of cancer is provided herein. For example, the present invention includes a pharmaceutical composition comprising a bispecific antibody comprising a human 4-1BB binding domain, comprising a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 17 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 18, and the human OX40 binding domain comprises a VH comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 31 and a VL comprising an amino acid sequence that is at least 85%, 90%, 95%, or 99% identical to the amino acid sequence of SEQ ID NO: 30.

[0380] In one aspect, an antibody that binds to 4-1BB and / or OX40 provided herein is useful, for example, for detecting the presence of 4-1BB and / or OX40 in a biological sample. As used herein, the term "detecting" encompasses quantitative detection or qualitative detection. In certain embodiments, the biological sample includes cells or tissues. In certain embodiments, a method for detecting the presence of 4-1BB and / or OX40 in a biological sample includes contacting the biological sample with an antibody that binds to 4-1BB and / or OX40 provided herein under conditions that permit binding of the antibody, and detecting whether a complex is formed between the antibody and 4-1BB and / or OX40.

[0381] In certain embodiments, the antibodies provided herein that bind to 4-1BB and / or OX40 are labeled. Labels include, but are not limited to, labels or moieties that are directly detectable (such as fluorescent labels, chromophore labels, high electron density labels, chemiluminescent labels, and radioactive labels), as well as moieties such as enzymes or ligands that are indirectly detected through, for example, enzymatic reactions or molecular interactions.

[0382] Embodiments of the present disclosure can be further defined by reference to the following non-limiting examples, which are described in detail in the preparation of specific antibodies of the present disclosure and methods for using the antibodies of the present disclosure. It will be apparent to those skilled in the art that many modifications can be made to both the materials and methods without departing from the scope of the present disclosure. Examples

[0383] It is understood that the examples and embodiments described herein are for illustrative purposes only, and in light thereof, various modifications or changes may be proposed to those skilled in the art and are included within the spirit and scope of the present application. Example 1. Generation of OX40- and 4-1BB-expressing CHO cells and recombinant OX40 and 4-1BB extracellular domain proteins

[0384] Nucleotide sequences defining the full-length and extracellular domain (ECD) of human and cynomolgus monkey OX40 and 4-1BB are obtained from the Genbank database and listed in Table 1.

[0385] [Table 14]

[0386] The human and cynomolgus macaque ECDs contained C-terminal tags for the purposes of purification, detection, and biotin-based labeling. DNA containing the nucleotide sequences of Table 1 was synthesized and inserted into an expression vector suitable for mammalian cell expression and secretion. The non-human primate OX40 and 4-1BB ECD proteins were used to evaluate the cross-reactivity and affinity of the binding domains for species used in potential toxicology assessments. These proteins were also utilized for immunization and screening to isolate the binding domains for both targets. HEK-293 cells grown in suspension culture were transiently transfected using a DNA expression vector encoding the ECD. After several days of culture, the conditioned media were clarified via centrifugation and sterile filtration. Protein purification was performed using a combination of appropriate affinity purification steps (typically immobilized metal affinity chromatography, protein A, or protein G chromatography), followed by size exclusion chromatography (SEC) to remove aggregated, clipped products and other host cell contaminants. SEC was also used to buffer exchange the protein into phosphate-buffered saline (PBS). The final purity of the samples was determined by analytical SEC and was typically greater than 90% final purity. Protein batches were sterile filtered and stored at 4 °C until needed. Example 2. Generation of CHO cell lines expressing full-length OX40 and 4-1BB for screening

[0387] Plasmid DNA encoding two targets (OX40: OXF001(hu) and OXF004(cyno); 4-1BB: FOB005(hu) and FOB006(cyno)) was digested with PvuI, ethanol-precipitated, and the OX40 construct was dissolved in ultrapure water and then in Maxcyte electroporation buffer. Using a MaxCyte instrument for electroporation, the linearized DNA was transfected into CHO-K1SV cells (CDACF-CHO-K1SV cells (ID code 269-W3), Lonza Biologics). The transfected cells were transferred from the electroporation cuvette to a T150 culture flask to rest and then gently suspended in 40 mL of CD CHO medium supplemented with 6 mM L-glutamine in the T150 flask. The flask was placed in an incubator at 37 °C, 5% CO 2 and allowed to recover for 24 hours before being placed under selection conditions. The day after transfection, the cells were centrifuged at 1000 RPM for 5 minutes and resuspended in CD CHO medium containing 1×GS supplement and 50 μM MSX. After the bulk population was recovered from the initial selection, cells were evaluated for surface expression using a commercial reagent and representative vials were frozen. To obtain clones with various expression levels, cells were sorted by flow cytometry, seeded by limiting dilution, and grown for 2 weeks. Clones from the FOB005 and FOB006 selection pools were identified by imaging using CLD Cell Metric (Solentim) at 3 hours, 24 hours, 48 hours, 7 days, and 14 days after seeding. Only wells with high-quality images and single cells identified at 3 hours after seeding were selected for further expansion and characterization of surface expression by flow cytometry (Figures 1A and 1B). Clones from the OXF001 and OXF004 selection pools were imaged at 14 days after seeding. Only wells with high-quality images at 14 days after seeding were selected for further expansion and characterization of surface expression by flow cytometry (Figures 2A - 2C). All clones were frozen in the bank in a maximum of 30 vials per clone. Example 3. General Expression and Purification of 4-1BB and OX40 Binding Molecules and Antibodies

[0388] The monospecific and bispecific 4-1BB and OX40 binding molecules disclosed herein were produced by transient transfection of either human HEK293 or Chinese hamster ovary (CHO) cells. Cells, cell debris, and insoluble materials were clarified from the culture by centrifugation and / or filtration. Recombinant proteins were purified from the clarified conditioned media using Protein A affinity chromatography. Preparative size exclusion chromatography (Prep SEC) was typically performed to further purify and homogenize the protein and to buffer exchange into PBS. Protein purity was verified by analytical size exclusion chromatography (analytical SEC) on an Agilent HPLC following each of the Protein A and Prep SEC purification steps. To ensure that the presence of endotoxin did not confound the results of the in vitro activity assays, the endotoxin amount was determined by using an Endosafe PTS instrument according to the manufacturer's instructions. The resulting protein was buffer exchanged into PBS as part of the SEC purification process, concentrated to up to 1 mg / mL, sterile filtered, and stored at 4 °C until needed or as specified. Protein concentration was determined from the absorbance at 280 nm using the theoretical extinction coefficient calculated from the amino acid sequence. Example 4. Generation of OX40 Antibodies by Hybridoma

[0389] Anti-OX40 specific antibodies were isolated from a hybridoma library generated after immunizing OmniRat and OmniMouse (Ligand, San Diego, CA) with DNA encoding human OX40 protein (Aldevron Freiburg, Germany). The bispecificity of individual clones was confirmed by testing binding using flow cytometry against CHO cells transfected with human and cynomolgus monkey mutants of OX40, and further confirmed by the absence of binding to untransfected CHO cells. The variable heavy chain (VH) domain sequence and variable light chain (VL) domain sequence for the selected hybridoma clones were obtained by RT-PCR after isolating total RNA. Briefly, total RNA was isolated from the hybridoma clone cell bank using Qiagen's RNeasy Plus kit (Qiagen, Venlo, the Netherlands). Subsequently, 200 ng of total RNA was used in a First Stand cDNA synthesis reaction using Superscript IV (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer's protocol. PCR was performed using 2 μl of cDNA as a template and a specific primer mix defined by ligand / OMT for amplification of either the VH region or the VL region. The PCR product for each clone was directly sequenced using a reverse primer for the constant domain and the standard Sanger sequencing method. Subsequently, the sequence was converted to scFv by amplifying the variable domain using specific primers containing overlapping sequences and associated with a mammalian expression vector using the NEBuilder HiFi DNA Assembly cloning kit (New England Biolabs, Beverly, MA). Example 5. Surface Plasmon Resonance (SPR) Method for Determining the Binding Affinity of OX40 and 4-1BB Binding Domains to the Extracellular Domains of Recombinant Human, Mouse, and Cynomolgus Monkey

[0390] SPR binding affinity studies of single - specificity and bispecific proteins that bind to the extracellular domains (ECDs) of recombinant monomeric human and cynomolgus monkey OX40 and 4 - 1BB were performed at 25 °C in HBS - EP+ containing 0.2% BSA buffer on either a Biacore T200 or Biacore 8K system. Mouse anti - human IgG (GE, BR - 1008 - 39) at 25 μg / ml in 10 mM sodium acetate pH 5.0 was immobilized at a density of approximately 10,000 response units (RU) on each flow cell of a CM5 research grade sensor chip (GE) by standard amine coupling chemistry. Each binding protein at approximately 100 nM in HBS - EP+ containing 0.2% BSA buffer was captured on the flow cell with immobilized anti - human IgG at a flow rate of 10 μL / min for 20 seconds, leaving one flow cell surface unchanged as a reference. Using the single - cycle kinetics mode, five different concentrations of ECD were continuously injected into each flow cell at 30 μL / min for 300 seconds, followed by a 600 - second dissociation period. Regeneration was achieved by injection of 3M MgCl 2 for 30 seconds at a flow rate of 30 μL / min, followed by stabilization for 1 minute with HBS - EP+ containing 0.2% BSA buffer.

[0391] Sensorgrams obtained from kinetic SPR measurements were analyzed by double - subtraction. The signal from the reference flow cell was subtracted from the analyte binding response obtained from the flow cell with the immobilized ligand or the captured ligand. Then, the buffer reference responses were averaged over multiple injections. Then, the average buffer reference response was subtracted from the analyte binding response, and the final double - reference data were analyzed using Biacore T200 evaluation software (2.0, GE), and the data were globally fitted to derive kinetic parameters. The entire sensorgram was fitted using a simple 1:1 binding model. Example 6. Screening of the OX40 binding domain in the ADAPTIR™ format using a control scFv against the N - terminus and OX40scFv against the C - terminus for cell binding and activity

[0392] Based on the cell-binding data obtained by screening hybridoma supernatants, the selected anti-OX40 antibody was converted to scFv and incorporated at the C-terminal position into the ADAPTIR™ bispecific format (N-terminal scFv-IgG1 Fc-C-terminal scFv) and screened for cell binding and activity in an OX40 reporter assay. At the N-terminal position, a control anti-tumor antigen scFv was used and maintained constant across the set. Both orientations (VH-VL and VL-VH) of the anti-OX40 scFv variable domains, as well as the lengths of two different linkers (either a single Gly4Ser linker or a series of three Gly4Ser repeats) used to connect the Fc region to the C-terminal scFv, were evaluated.

[0393] Flow cytometry was used to quantify and confirm the binding of OX40-specific scFv to human and cynomolgus monkey OX40 expressed on the surface of transfected cells. Binding studies were performed on CHO-K1 cells stably expressing full-length human or cynomolgus monkey OX40 protein, developed in-house and subsequently cloned. Typically, 100,000 cells were incubated for 40 minutes at 4 °C in 50 μl of PBS buffer containing 0.2% BSA and 2 mM EDTA with a dilution of the bispecific construct, followed by washing. Subsequent incubation was performed for 30 minutes at 4 °C with a PE-labeled, minimally cross-reactive secondary antibody, goat anti-human IgG Fcγ, F(ab’)2 (Jackson ImmunoResearch). The signal from the bound molecules was detected using an LSR-II or FACSymphony A3 flow cytometer (BD Biosciences) and analyzed by FlowJo flow cytometry analysis software. The mean fluorescence intensity (MFI) of the bound molecules on the cells was determined after exclusion of doublets. EC 50 Nonlinear regression analysis to determine EC values was performed with GraphPad Prism7® graphing and statistics software.

[0394] Figures 3A and 3B show the binding of four different bispecific anti-ROR2 43×anti-OX40 constructs (OXF169, OXF170, OXF171, and OXF172) to CHO cells stably expressing either human or cynomolgus monkey OX40 protein. The observed EC 50 is reported in Table 2.

[0395]

Table 15

[0396] To compare the activities of different OX40-binding bispecific constructs for inducing target-dependent activation of OX40, a luciferase reporter assay was used. 30,000 Jurkat cells transfected to express human OX40, which carry a luciferase reporter gene under the control of an NFκB promoter (generated in-house), were cultured in 96-well plates with 120,000 ROR-expressing MDA-MB-231 cancer (target) cells. Five-fold dilutions of the bispecific constructs were added. The cells were cultured in a total volume of 100 μL of RPMI 1640 medium supplemented with 5% fetal bovine serum, sodium pyruvate, antibiotics, and non-essential amino acids. The plates were incubated at 37 °C, 5% CO 2 in a humidified incubator for 5 hours. 100 μL of Bio-Glow buffer (Promega) was added to each well, mixed, and incubated for 10 minutes. Luminescence was measured with a MicroBeta 2 2450 microplate counter (PerkinElmer). Nonlinear regression analysis to determine EC 50 values was performed with GraphPad Prism6® graphing and statistics software. The results are shown in Figure 4, with the y-axis showing values in relative light units (RLU).

[0397] Figures 3A and 3B and Table 2 show the activity of ROR243×OX40 constructs (OXF169, OXF170, OXF171 and OXF172) using an OX40-expressing NFκB Jurkat reporter strain. MDA-MB-231 target cells were used for crosslinking. The EC 50 values observed for activity were in the range between 4.4 pM and 12.2 pM.

[0398] In summary, a series of ADAPTIR™ constructs were generated from the same anti-OX40 hybridoma clone (including VH of SEQ ID NO: 25 and VL of SEQ ID NO: 26). Based on the reporter assay, the linker length did not appear to affect activity (OXF171 vs. OXF172, or OXF169 vs. OXF170). However, when the anti-OX40 scFv was in the VL-VH orientation, the activity appeared to be higher (Table 2, Figures 3 and 4). Example 7. Optimization of the Thermal Stability of OXF171 Anti-OX40 scFv

[0399] To enhance the thermal stability of the anti-OX40 binding domain (BZG-12C3scFv in the VL-VH orientation) used in OXF171, an optimization campaign was carried out. A random mutagenesis phage library was generated for OXF171scFv using error-prone PCR, and clones with enhanced stability were enriched using panning from the phage display library under mild denaturing conditions. A combination of well-established molecular biology and phage display protocols was used. Briefly, the gene encoding the anti-OX40scFv binding domain used in OXF171 was used as a template for an error-prone PCR reaction using a commercially available mutagenesis kit (GeneMorph II Random Mutagenesis Kit, Agilent Technologies, USA) according to the manufacturer's protocol. The PCR product was digested with restriction enzymes and ligated into a phagemid vector to create a pIII phage coat protein N-terminal fusion library. This library was transformed into Escherichia coli (E. coli) SS320 / M13KO7 competent cells to generate a phage library. Five rounds of panning were performed on the library using biotinylated OX40 ECD (SEQ ID NO: 107) as the bait. Increased stringency of panning was used for each successive round by lowering the antigen concentration and increasing the wash time. As a method for selecting more stable binders, standard PBS-T washes were replaced with guanidine hydrochloride or magnesium chloride washes, or an additional round of panning was performed using phage pre-incubated at high temperature. Following the final round of panning, the phage output was seeded and prepared for bulk cloning of the scFv pool into an expression vector prepared for mammalian expression and screening. Approximately 600 individual colonies were picked and sequenced. Plasmid DNA was isolated for approximately 200 unique sequences and used for high-throughput 293 transient transfection (culture volume approximately 0.6 mL). After culturing for 3 days, the cell supernatant was purified and the thermal stability was measured using differential scanning fluorimetry (DSF).Two amino acid changes with improved thermal stability compared to the parental OXF171 sequence were identified: H40N in the variable light chain and V55A in the variable heavy chain. Next, the H40N and V55A mutations were combined either individually or with framework germline mutations A51V in the variable light chain (to revert to IGLV3-21*02) and D92N and L101V in the variable heavy chain (to revert to IGHV3-30*03). The combination of H40N with the germline mutations is present in molecules OXF01099, FXX01055, and FXX01079. The combination of the germline mutations H40N and V55A is present in molecules OXF01115, FXX01047, and FXX01066. Example 8. Generation and evaluation of biophysical characteristics of OXF171 anti-OX40 scFv mutants

[0400] Following phage panning, the isolated scFvs were sequenced and incorporated into single-specificity constructs by binding the anti-OX40 scFvs to the C-terminus of the wild-type IgG1 Fc region (wtFc anti-OX40 scFv). Following transient expression and purification from Chinese hamster ovary cells, these constructs were characterized for expression, thermal stability by differential scanning calorimetry (DSC), binding affinity to human OX40 ECD by SPR, cell binding, and activity in the OX40 reporter assay.

[0401] DSC was performed to determine the midpoint temperature (Tm) of temperature-induced unfolding of the anti-OX40 scFvs using a MicroCal VP-capillary DSC system (Malvern Instrument). A perfect match of buffer, PBS pH 7.4, was used as a reference. 500 μl of a 0.5 mg / ml solution of each protein sample was loaded into the instrument along with the reference and heated from 25 °C to 100 °C at a rate of 1 °C per minute. The melting curves were analyzed using Origin 7 platform software MicroCal VP-Capillary DSC automated analysis software to derive the Tm values. Surface plasmon resonance was used to determine the binding affinity of the OX40 binding domain as described in Example 5.

[0402] As shown in Table 3, significant increases in expression were obtained with two variants (OXF01099 and OXF01115) compared to the unmodified parental construct (OXF01022). The thermal stability was improved from 55.7 °C to above 60 °C while retaining a similar binding affinity to the parental binding domain. The improved expression and Tm values suggest that the variants have improved stability and solubility, which are considered beneficial properties for therapeutic protein agents.

[0403]

Table 16

[0404] Example 9. Evaluation of Cell Binding and In Vitro Activity of OXF171 Anti - OX40 scFv Variants

[0405] Binding studies were used to confirm the binding of the preferred anti - OX40 variants to human and cynomolgus monkey OX40. As shown in Figures 5A and 5B, the binding of various anti - OX40 constructs (OXF01122, OXF01099, and OXF01115) to CHO cells stably expresses either human OX40 or cynomolgus monkey OX40 protein. There were no detectable differences in binding between the parental and anti - OX40 scFv variants.

[0406] To compare the ability of different OX40 - binding constructs to induce target - dependent activation of OX40, a luciferase reporter assay was used. The experimental setup was modified as described in Example 6. CHO - K1 (FcγRI) cells expressing CD64 were used to cross - link the wild - type Fc of these constructs. Figure 6 shows that the activities of the anti - OX40 constructs are similar. A summary of the binding and reporter assays is shown in Table 4.

[0407]

Table 17

[0408] Example 10. Generation of 4-1BB antibody by immunization of wild-type mice

[0409] 4-1BB specific antibodies were isolated from a hybridoma library generated after immunizing BALB / c mice and NZB / W mice with recombinant human 4-1BB protein antigen (ImmunoPrecise Antibodies Victoria, B.C. CAN). Supernatants from hybridoma clones were assayed by ELISA and wells identified for specific binding were confirmed using flow cytometry against CHO cells transfected with human and cynomolgus monkey 4-1BB. Positive clones were selected for expansion and viable cells were frozen for RNA extraction and variable domain analysis. Supernatants were stored for additional analysis.

[0410] After isolating total RNA, the variable heavy (VH) and light (VL) chain domain sequences for the selected hybridoma clones were obtained by RT-PCR. Briefly, total RNA was isolated from the hybridoma clone cell bank using the Qiagen RNeasy Plus kit (Qiagen, Venlo, Netherlands), and 400 ng of total RNA was used in a First Stand cDNA synthesis reaction using oligo dT and Superscript IV (Thermo Fisher Scientific, Waltham, MA) according to the manufacturer's protocol. After cDNA synthesis, variable region cDNA was amplified using 1 μL of cDNA and a series of primer mixes for mouse IgG VH, Vκ, and Vλ (Novagen mouse Ig-primerset, EMD Millipore Temecula, CA). The PCR products for each clone were directly sequenced using reverse (constant domain) PCR primers and the standard Sanger sequencing method. Subsequently, the sequences were converted to scFv by amplifying the variable domains using specific primers containing overlapping sequences and were ligated into a mammalian expression vector using the NEBuilder HiFi DNA Assembly cloning kit (New England Biolabs, Beverly, MA). Example 11. Humanization of Clone 6, 41BB Antibody in scFv Format

[0411] After evaluation of the hybridoma-derived antibody, clone 6 was selected for humanization and further optimization. The main objective was to minimize potential immunogenicity and to optimize the binding and stability of the binding domain by eliminating mouse-derived sequences as much as possible. The clone 6 antibody, a 41BB mouse monoclonal antibody (VH SEQ ID NO: 19; VL SEQ ID NO: 20; see also Figure 7), was humanized in three steps. In stage 1, the BioLuminate software package release 2018-2 (Schrodinger, LLC, New York, USA) was used. A homology model of mouse clone 6 was created based on PDB ID 1JV5, and human frameworks that were most geometrically appropriate and homologous for CDR grafting were identified using the software's default and modified settings. 19 CDR-grafted molecules were produced and tested for binding to cells expressing full-length human 41BB or cyno-41BB (data not shown). The graft based on molecule FOB01143 (FOBW006HLH20), SEQ ID NO: 43, PDB ID 5I17, was shown to have binding characteristics similar to the parental mAb clone 6 (data not shown). In stage 2, framework residues were mutated in sets and combinations of sets, and the mouse residues of FOB01143 were converted to human germline sequences IGHV1-46*01 and IGHJ4*01 for the heavy chain and IGKV3D-7*01 and IGKJ1*01 for the light chain. Molecule FOB01188 (FOBW006HLH26), SEQ ID NO: 45, was confirmed to retain the best combination of binding, functional, and developability characteristics (data not shown). In stage 3, individual residues different from the human V gene germline were mutated to the germline, respectively, and a set of mutant molecules was first characterized for binding to human 41BB and cyno-41BB recombinant proteins (SEQ ID NOs: 1 and 2, respectively) using a Biacore 8K (GE Healthcare Life Sciences, USA), and then characterized for stability by measuring Tm and Tagg using an Uncle instrument (Anchored Antibodies, USA) (data not shown).By incorporating germline amino acid changes from all benign mice into the FOB01188 molecule, the final molecule FOBW006HLH40 was created. The sequence of this molecule is 92% identical to IGHV1-46*01 and 94% identical to IGKV3D-7*01. All non-germline residues are essential for any stability bond. The progression from the mouse to the humanized sequence in the amino acid alignment (mouse clone 6 to humanized FOBW006HLH40) is shown in Figure 7. Example 12. Generation and Biophysical Evaluation of a Partially Humanized Version of Anti-41BB Clone 6

[0412] By binding the scFv sequence in the VH-VL direction to the C-terminus of the wild-type IgG1 Fc, different humanized versions of clone 6 scFv were produced as single-specificity DNA constructs. Following transient expression and purification, these constructs were characterized for thermal stability by differential scanning fluorimetry (DSF) and binding affinity to human and cyno 41BB EDC. DSF was performed on the samples and tested in triplicate against a 7500 Fast real-time PCR system (Thermo Fisher Scientific) in dPBS at 0.125 mg / mL with SYPRO Orange (Life Technologies) added to a final concentration of 5×. The samples were heated from 25°C to 95°C at a scan rate of 0.9°C / min. The average transition midpoint value (Tm) was determined using ProteoStat® ProProtein Thermal Shift™ software v1.0 (Thermo Fisher Scientific). The binding affinity was performed as described above.

[0413] This data verified that the conjugation and thermal stability were not adversely affected by deletion of the mouse sequence. The humanized construct (FOB01188) was compared to a chimeric molecule consisting of human IgG1Fc and mouse scFv sequence (FOB01143). The Tm values of both mouse and humanized scFv were both 69 °C, suggesting that the stability of the molecule did not change (Table 5). The binding affinity determined by SPR showed that tighter binding was achieved for both human and cyno 4-1BB ECD (Table 5) as a result of the humanization process.

[0414]

Table 18

[0415] Example 13. Cell Binding and In Vitro Activity of a Partially Humanized Version of Anti-41BB Clone 6 scFv

[0416] Using the general method described in Example 6, Jurkat cells stably expressing full-length human or cynomolgus 4-1BB protein were used to verify that the human modification made to FOB01143, which resulted in construct FOB01188, did not negatively inhibit the binding affinity for either human 4-1BB or cynomolgus 4-1BB protein (Table 6, Figures 8A and 8B).

[0417] To compare the activities of different 4-1BB binding constructs for inducing target-dependent activity of 4-1BB, a luciferase reporter assay was used. The experimental setup was as described in Example 6, except that CHO-K1 expressing CD64 (FcγRI) was used as the target cell to crosslink 4-1BB via binding to the wild-type Fc of these constructs. A human 4-1BB expressing NFκB reporter strain was generated in-house and utilized here to determine the activity. The EC observed for the activities of both constructs 50The value was 28 pM. These data (Table 6, Figure 9) showed that the activity was not affected by deleting the mouse sequence.

[0418]

Table 19

[0419] Example 14: Assembly of 41BB×OX40 Bispecific Protein

[0420] Subsets of the 41BB and OX40 binding domains were ligated to the bispecific proteins FXX01047, FXX01055, FXX01066, and FXX01079 (see Table 7; refer to SEQ ID NOs: 86, 87, 78, 88). The individual binding domains were amplified by PCR and associated with the DNA fragment encoding Fc and the linearized expression vector using standard molecular biology techniques. Example 15: Production and Biophysical Characterization of 4-1BB and OX40 Bispecific Proteins with Additional Humanized Mutations and Altered Positions of the OX40 and 4-1BB Binding Domains

[0421] After transient expression and purification in CHO cells, the 4-1BB×OX40 bispecific proteins were examined for the effect of incorporation into the anti-4-1BB scFv with additional human sequences (FOB01188), and similarly, the preferred orientation was determined. These comparisons were performed with the set of constructs described in Table 7.

[0422]

Table 20

[0423] Comparison of constructs shows that when the anti-OX40 binding domain is located at the N-terminus of the protein, transient CHO expression levels are improved (Table 8, FXX01055 vs FXX01047 and FXX01079 vs FXX01066). The presence of additional human residues in FXX01066 and FXX01079 results in good expression in both directions of the target binding domain when comparing the protein pairs without these changes. FXX01066 and FXX01079 also had better resistance to aggregation based on the % change in purity after 1 week of storage at 4°C and 40°C, determined by integrating the product peak area in analytical SEC. Comparison of constructs with the same orientation but different in terms of the inclusion of humanized mutations shows that less aggregation is formed with a larger number of human constructs (FXX01066 vs FXX01047, FXX01079 vs FXX01055), supporting that they are more stable. The position of the target binding domain affected the amount of degradation products measured after the initial ProA purification step was performed. ProA eluate samples were also analyzed by analytical size exclusion ultra-high performance liquid chromatography (analytical SE-UPLC) because of the high resolution of this method. Analysis was performed using a Waters ACQUITY UPLC instrument with two BEH SEC columns (200Å, 1.7μm, 4.6mm × 300mm) connected in tandem and using a potassium phosphate / potassium chloride electrophoresis buffer. Generally, 10 μg was injected and a 75-minute method was run at a flow rate of 0.15 mL / min. Following integration to obtain peak areas, the data showed that constructs with an anti-4-1BB scFv at the N-terminal position were more resistant to the formation of clipped products. This is based on the higher proportion of low molecular weight contaminants present in FXX01055 and FXX01079 compared to FXX01047 and FXX01066.

[0424]

Table 21

[0425] The binding affinities of these four variants for the human extracellular domains of OX40 and 4-1BB were determined (Table 9). The binding affinity for OX40 was not significantly affected by either the anti-target scFv contained in FXX01066 and FXX01079 or the relative position of the additional humanized mutations. When the anti-4-1BB scFv was located at the N-terminus of the bispecific construct, it was determined that the affinity for 4-1BB was stronger.

[0426]

Table 22

[0427] Example 16: Cell Binding and In Vitro Activity of 4-1BB and OX40 Bispecific Proteins with Additional Humanized Mutations and Altered Positions of the OX40 Binding Domain and 4-1BB Binding Domain

[0428] Flow cytometry was used to quantify and confirm the binding of 4-1BB×OX40 bispecific proteins using cell lines expressing either human or cynomolgus OX40 and either human or cynomolgus 4-1BB. As shown in FIGS. 10A-10D and Table 10, the variants did not present differences in binding due to either additional humanization of the binding domains or position. Since FXX01047 and FXX01066 have slightly lower EC50s than FXX01055 and FXX01079, the presence of the anti-4-1BB scFv at the N-terminus is preferred. In addition, cynomolgus OX40 binding is reduced in FXX01079.

[0429] To compare the activities of the 4-1BB×OX40 bispecific proteins, two luciferase reporter cell lines were utilized in a separate assay. OX40-expressing cells or 4-1BB-expressing cells were used for binding and crosslinking was induced via the anti-receptor binding domain at the other end of the bispecific antibody. To quantify 4-1BB activity, the human 4-1BB NFκB luciferase reporter cell line was incubated with OX40-expressing CHO-K1 target cells. Conversely, to examine OX40 activity, the human OX40 NFκB luciferase reporter cell line was added together with target 4-1BB-expressing Jurkat cells. In both assays, 30,000 reporter and target cells were added to the diluted 4-1BB×OX40 protein and incubated for 5 hours in reporter medium containing 5% FBS. As shown in Figure 11A, the 4-1BB activities of FXX01047, FXX01055, FXX01066, and FXX01079 could not be distinguished and were not affected by the scFv position or sequence modification in the anti-4-1BB binding domain. The OX40 reporter assay preferred that the anti-OX40 scFv be positioned at the C-terminus (Figure 11B), but showed that changes to the sequence did not affect the activity. A summary of the human binding and reporter assays is shown in Table 10.

[0430]

Table 23

[0431] Example 17: Evaluation of the orientation of 4-1BB scFv and framework sequence modification of anti-OX40 scFv to change the isoelectric point (pI)

[0432] To further optimize the bispecific molecule, the domain order was evaluated in the anti-41BB scFv, and germline-derived mutations were introduced to increase the isoelectric point of the anti-OX40 scFv. The anti-41BB clone 6 mAb in the scFv format was humanized at all stages in the VH-VL format. A set of mutants was generated to evaluate the behavior of anti-41BB in the VL-VH and VH-VL directions. As part of this set of constructs, modifications to change the pI of the anti-OX40 scFv were also included. These were constructed by first analyzing the highly homologous germline human frameworks of IGHV3-30*03 and IGLV3-21*02 and identifying charge changes away from the CDRs and surface-exposed positions. T86R is present in IGHV3-30*13, and Q17K is present in IGLV3-21*01 and was included to increase the overall pI of the protein. Side-specific mutagenesis was performed to generate the T86R and Q17K changes individually and in combination. Example 18: Generation and characterization of an anti-4-1BB × anti-OX40 bispecific antibody for the evaluation of 4-1BB scFv orientation and anti-OX40 scFv framework sequence modifications to change the isoelectric point (pI)

[0433] Transient expression was used to produce the protein, which was purified by ProA chromatography and size-exclusion SEC. Following purification, the protein concentration of each sample was adjusted to 1 mg / mL in PBS and tested through some evaluation of stability and binding affinity. The orientation of the anti-41BB scFv did not significantly affect the binding affinity for human 41BB ECD as measured by SPR when it was at the N-terminal position of the bispecific construct (Table 11). Similarly, changes made in the framework region of the anti-OX40 binding domain did not alter the strong binding to human OX40 ECD. Evaluation of protein stability did not show a significant difference as a result of these changes (data not shown).

[0434] [Table 24]

[0435] Example 19: Cell binding and in vitro activity of 4-1BB and OX40 bispecific proteins to evaluate the orientation of 4-1BBscFv and the framework sequence modification of anti-OX40scFv for changing the isoelectric point (pI)

[0436] Cell binding studies were completed to show that the ADAPTIR™ scFv binding domains bound well to cells expressing human or cynomolgus 4-1BB or OX40. The binding studies were performed using the flow cytometry-based staining procedure described above. These data show minor variations in binding for either human (Figures 12A and 12B) or cynomolgus (Figures 12C and 12D) for proteins FXX01066, FXX01099, FXX01101, FXX01102, FXX01104, FXX01105, FXX01107, and FXX01108. In addition, these proteins did not show any non-specific binding to parental CHO-K1 SV (Figure 13). Thus, there was no hindrance to binding with the addition of pI changes to anti-OX40scFv or with alternative orientations of anti-4-1BBscFv.

[0437] Using the activity assay, when crosslinking either OX40 or 4-1BB in the 4-1BB or OX40 reporter assay respectively, the ability to induce NFκB signaling was shown. In this set of experiments, both human and cynomolgus monkeys expressing the NFκB reporter strain of 4-1BB or OX40 were evaluated in this screen. The cynomolgus monkey reporter strain was generated and cloned in-house. The data in Figure 14A shows a slight increase in the maximum activity in human 4-1BB activity in protein over that generated by the parental FXX01066. In contrast, the activity in the cynomolgus monkey 4-1BB reporter (Figure 14C) shows variation in the maximum RLU, and thus the construct with the anti-4-1BB scFv in the VLVH direction with the T86R pI mutation was slightly lower than any construct in the VHVL direction and significantly lower than the construct without the T86R pI mutation. Figures 14B and 14D show no differences in EC50 or maximum RLU in the human or cynomolgus monkey OX40 reporter assay. A summary of the human binding and reporter assays is shown in Table 12.

[0438] To determine the non-specific activity induced by these various constructs, 4-1BB and OX40 reporter assays were performed. Instead of using the OX40-expressing cell line or 4-1BB-expressing cell line (respectively) for crosslinking, parental CHO-K1 SV cells were used. Without crosslinking, these constructs should not induce NFκB signaling. Without crosslinking, 4-1BB in the VLVH direction significantly induces NFκB signaling (Figure 15A). The non-specific activity was significantly lower when 4-1BB was in the VHVL direction. These proteins did not induce non-specific activity in the human OX40 reporter assay when crosslinked with parental CHO-K1 SV (Figure 15B).

[0439]

Table 25

[0440] Example 20: Anti-4-1BB × antibody OX40 ADAPTIR™ bispecific treatment is synergistic compared to treatment with anti-4-1BB plus anti-OX40 monospecific proteins in vitro

[0441] Costimulation of OX40 during clonal expansion has been shown to promote the survival of activated T cells (Rogers, P.R., et al., Immunity, 15(3):445-55 (2001) and Weatherill, A.R., et al., Cell Immunol, 209(1):63-75 (2001)). Therefore, the ability of anti-4-1BB × anti-OX40 ADAPTIR™ constructs to increase the number of T cells and NK cells in vitro was examined. Peripheral blood mononuclear cells (PBMCs) were isolated from normal donors and incubated with serial dilutions of ADAPTIR™ in the presence of α-CD3 (signal 1), which upregulates 4-1BB and OX40 expression (signal 2). In this particular experiment, anti-OX40 monospecific construct (OXF01070) with an scFv at the N-terminus of the Fc region of wild-type IgG1 and anti-4-1BB monospecific construct (FOB01173) with an scFv at the C-terminus of the Fc region of wild-type IgG1 were compared for their ability to induce PBMC proliferation. Monospecific treatments were compared to bispecific treatments using the OX40 and 4-1BB synergy. PBMCs were isolated from human blood using standard density gradient separation and stained with 5 μM CellTrace™ Violet (Molecular Probes) as recommended by the manufacturer. 120,000 PBMCs were incubated with 10-fold serial dilutions of the test molecule (ranging from 10 μM to 1 pM), and the cell mixture was added to a final volume of 200 μl / well in complete RPMI 1640 medium supplemented with 10% FBS and 5 ng / ml of α-CD3 in 96-well plates. The plates were incubated in a humidified incubator at 37 °C, 5% CO 2 for 24 to 6 days.

[0442] NK cell proliferation and T cell proliferation were evaluated by flow cytometry. Cells were fluorescently labeled using 7AAD (Sigma), PE / Cy7-αhCD25, APC / Cy7-αhCD5, BV605-αhCD56, BV650αhCD8, and BV510-α-hCD4 (Biolegend) and incubated at 4 °C for 30 minutes. Cells were washed twice, resuspended, and acquired on a BD FACSymphony™ flow cytometer. All samples were analyzed using FlowJo software, and the percentages of NK cells, CD8 + T cells, and CD4 + T cells that had been expanded via dilution of CellTrace™ Violet (CTV) were calculated. Graphs were plotted using GraphPad Prism 7.0.

[0443] As shown in Figure 16, the 4-1BB×OX40 bispecific proteins FXX01047 and FXX01055 promote dose-dependent expansion of CD8 + T cells, CD4 + T cells, and NK cells (gray symbols). EC 50 values were in the range of 13 nM to 41 nM for the three cell subsets. This experiment clearly shows that the single-specificity constructs OXF01070 or FOB01173 alone are unable to promote proliferation (uncolored symbols). Importantly, combinations of the two single-specificity constructs were not sufficient to induce T cell or NK cell proliferation (black diamonds). These results are clinically relevant since 4-1BB and OX40 single-specificity treatments containing wild-type Fc are severely impaired. Potent proliferation is promoted only in the synergistic bispecific format of the 4-1BB×OX40 antibody. Example 21: Human T Cell Proliferation in Response to Anti-4-1BB×Anti-OX40 ADAPTIR™ Bispecific Protein Treatment In Vitro

[0444] For functionality, additional bispecific constructs were analyzed in primary PBMC assays. The method was similar to that used in Example 20 with modifications. Cells were additionally stained with PE / Cy7-αhCD25 (Biolegend). All samples were analyzed to calculate the percentages of expanded NK cells, CD8 + T cells, and CD4 + T cells and the activation status via parental CD25+.

[0445] Additionally, cytokine secretion for IFN-γ, IL-2, and TNF-α was evaluated using a multiplex-based assay (Milliplex) from 72-hour supernatants diluted 1:3 in assay buffer prior to analysis on Magpix. EC 50 was determined by non-linear regression using GraphPad Prism7. Constructs were tested using PBMCs from two individual healthy donors.

[0446] These data show that anti-CD3-stimulated PBMCs treated with the 4-1BB×OX40 construct can reliably increase the percentages of CD8 + T cells and CD4 + T cells in a dose-dependent manner over 96 hours of culture (Figure 17). Additionally, cytokines from stimulated T cells induce NK cell proliferation (Figure 18A) and activation (Figure 18B). Furthermore, supernatants removed from the cultures at 72 hours reveal significant dose-dependent secretion of IFN-γ, IL-2, and TNF-α when the construct is added exogenously (Figure 19). There were no differences in the in vitro functionality of the 4-1BB×OX40 bispecific proteins tested. In summary, these results show dose-dependent in vitro NK cell proliferation, T cell proliferation, and cytokine production when the 4-1BB×OX40 construct is added to stimulated PBMCs. The levels of maximum proliferation induced by the top construct were similar in both the maximum percentage of proliferating cells and the concentration at peak proliferation induced by ADAPTIR™. Example 22. Evaluation of additional framework sequence modifications to optimize anti-OX40 scFv

[0447] After analysis of the experimental data and in combination with modeling of the surface properties of the binding domain, several variants were constructed and tested. Mutations were designed to mimic the sequence and structure of the human germline framework. Briefly, the parental IGLV3-21*01 amino acid sequence IPE (IMGT numbering 71-74) was mutated to IPA, VPN, VPS, and IPK. Bispecific molecules FXX01110 to 01121 (SEQ ID NOs: 89-100) represent additional variants of the OX40 domain. Example 23. Characterization of a bispecific anti-4-1BB × anti-OX40 ADAPTIR™ construct with additional modifications to anti-OX40 scFv

[0448] Following transient transfection and purification using the methods described above, additional constructs containing changes that altered the calculated isoelectric point (pI) were evaluated for their impact on expression level, purity, and stability characteristics. Changes in pI were generated via amino acid changes to the anti-OX40 scFv. The isoelectric point was calculated using the algorithm of the Genedata Biologics Platform®. Theoretical pI values may vary depending on the method. These values were used to look for general trends since the pI, a measure of the net charge of the protein, can affect the solubility and stability of the protein under different conditions. Expression levels were calculated based on the mass recovered from Protein A purification from the volume of supernatant generated (assuming 100% of the protein was captured by Protein A). As shown in Table 13 below, no specific trends were observed with the changes made to the amino acid sequences that altered expression and pI. Among this set of proteins, FXX01111 showed the highest expression level. Following the Protein A affinity capture step, size exclusion chromatography (SEC) was performed to remove high molecular weight aggregates (HMW) and, if present, some low molecular weight species (LMW), while simultaneously buffer-exchanging the sample into PBS. Samples of each construct were analyzed by analytical size-exclusion HPLC (SE-HPLC) and size-exclusion UPLC (SE-UPLC) to evaluate the homogeneity of the product. All constructs shown in Table 13 had minimal HMW products detected by either SE-HPLC or SE-UPLC and had high purity levels. Using the higher resolution SE-UPLC method, no significant peak areas for clipped / low molecular weight species were measured for these proteins.

[0449]

Table 26

[0450] Following purification and purity measurement, samples of each protein were stored under multiple conditions, and their properties of aggregating when formulated with only PBS at 1 mg / mL without any additional excipients under different storage conditions were evaluated. This included storage at 4 °C and 40 °C. In addition, each variant was tested for aggregate formation after freezing and then thawing of the protein from a -20 °C freezer. The rate of change of the peak area of the product was calculated and reflected the increase in aggregated protein present in the samples immediately after purification and after the treatments listed in Table 14 below. All constructs showed minimal change after 1 week at 4 °C, and larger changes were detected in samples stored under accelerated stability conditions at 40 °C. The samples showed only minor changes after one cycle of freezing / thawing from storage at -20 °C. No correlation was seen between these values and the theoretical pI.

[0451]

Table 27

[0452] Tm1 (midpoint of the first melting transition) and Tagg, the onset temperature of aggregation based on dynamic light scattering, were measured for these constructs using an Uncle instrument from Antidrugs. All constructs shown in Table 15 had Tm1 and Tagg values exceeding 60 °C, suggesting high-temperature stability. No specific trend was seen in the heat stability values correlating with the produced pI of the protein.

[0453]

Table 28

[0454] Using the BIACORE 8K SPR system, the affinity of these constructs for human and cyno 4-1BB, and human and cyno OX40 ECD was determined using the methods described above. The single-specificity binding affinity was determined by capturing the ADAPTIR™ bispecific construct on the chip and injecting the monovalent ECD of the target at multiple concentrations. The affinity of the OX40 scFv was not measurably affected by the changes made to alter the pI as all values remained in the nM range (Table 16). Binding to human and cyno 4-1BB was not affected as the anti-4-1BB scFv did not vary across this set of constructs.

[0455]

Table 29

[0456] Example 24: Characterization of Fc receptor binding to bispecific anti-4-1BB × anti-OX40 ADAPTIR™ constructs with different calculated isoelectric points

[0457] Fcγ receptor binding to the bispecific protein was measured by surface plasmon resonance at room temperature using a Biacore 8K instrument. The bispecific protein with a modified Fc region to reduce binding to the Fcγ receptor was directly immobilized on the surface of a CM5 sensor chip to a surface density of 4000 RU. The bispecific ADAPTIR™ protein with wild-type Fc was similarly immobilized on the surface as a positive control and used as a comparator to assess the reduction in binding due to changes incorporated into the Fc region amino acid sequence. Before injecting onto the surface of the prepared sensor chip at 30 μL / min for 60 seconds, the Fcγ receptor (purchased from R&D Systems) was diluted in HBS-EP+ buffer to either 100 nM (FcγRI) or 2 μM (for all Fcγ receptors except FcγRI). The maximum RU value during the association phase of each injection was used to compare the extent of the binding value of the modified Fc to the wild type and is reported in Table 17 below. There is a significant reduction in binding to all receptors tested, with some apparent residual binding to the FcγRIIA and RIIB / C receptors. As expected, different FXX bispecific antibodies share the same Fc sequence and thus show similar levels of binding.

[0458]

Table 30

[0459] Example 25: Evaluation of cell binding and functional activity of an anti-4-1BB × anti-OX40 ADAPTIR™ construct with an altered isoelectric point (pI)

[0460] Human and cynomolgus monkey binding studies were completed, demonstrating that our optimized ADAPTIR™ scFv binding domains specifically bound to cells using our standard flow cytometry-based staining procedure. These data show minimal changes in either human (Figures 20A and 20B) or cynomolgus monkey (Figures 20C and 20D) binding to either 4-1BB or OX40 expressed by the cloned cell lines for these constructs. Additionally, when tested for nonspecificity, these constructs did not bind to parental CHO-K1 SV (Figure 21). Thus, there is no unwanted binding alteration with the addition of these pI changes to the anti-OX40 scFv.

[0461] All screened constructs were subjected to a reporter assay using both human and cynomolgus monkey-expressing 4-1BB or OX40 reporter strains, demonstrating functionality. As shown in Figure 22, there were only minor differences in EC50 or maximal RLU in these reporter assays. Additionally, these proteins did not induce nonspecific activity in human 4-1BB (Figure 23A) or OX40 (Figure 23B) reporter assays when crosslinked to parental CHO-K1 SV. Thus, the functionality of these optimized ADAPTIR™ constructs was not altered by the introduction of pI mutations. Table 18 summarizes EC50 and maximal binding / induction in human binding and reporter assays.

[0462]

Table 31

[0463] Example 26: Human T Cell Proliferation and Cytokine Production in Response to Anti-4-1BB × Anti-OX40 ADAPTIR™ Bispecific Protein Treatment In Vivo

[0464] For functional differences, the ADAPTIR™ bispecific construct was analyzed in a primary PBMC assay. Similar to the method described above, isolated PBMCs were treated with 10 μg / mL αCD3 with titrated test constructs for 96 hours. After 96 hours, all samples were analyzed via flow cytometry to calculate the percentages of expanded NK cells, CD8 + T cells, and CD4 + T cells. As shown in Figure 24, each anti-4-1BB × anti-OX40 ADAPTIR™ construct was able to potently enhance the number of expanded CD8 + T cells and CD4 + T cells in a dose-dependent manner. Although there are minor changes in the function of these antibodies, they do not correlate with differences in the calculated pI. Interestingly, there is enhancement of NK cell proliferation and activation as measured by CTV dilution and CD25 expression, respectively (Figure 25). Furthermore, as shown in Figure 26, supernatants harvested from 72-hour cultures treated with anti-4-1BB × anti-OX40 bispecific protein all promoted dose-dependent secretion of IFN-γ, IL-2, and TNF-α. Changes in the calculated pI of the construct do not alter their ability to induce cytokine secretion. In summary, these results indicate dose-dependent in vitro NK cell proliferation, T cell proliferation, and cytokine production when anti-4-1BB × anti-OX40 constructs are added to stimulated PBMCs. Example 27: Fc Mutations to Eliminate Binding to Fc Receptors and Complement

[0465] In ADAPTIR™ bispecific molecules, it can be advantageous to introduce mutations into the Fc region to eliminate the ability to interact and signals through interaction with Fc receptors and complement. Table 19 below shows different mutations that may be made to the Fc region contained in ADAPTIR™ bispecific constructs (Null2, K322A Fc, TSC1004, TSC1005, TSC1006, and TSC1007) compared to the sequence of wild-type Fc (WT).

[0466]

Table 32

[0467] Example 28: Influence of Fc Mutations on SPR Analysis of Binding to Fc Receptors and Complement

[0468] Fc mutations potentially integrated into the ADAPTIR™ bispecific construct were analyzed for binding to common Fc receptors (human and cyno) and C1Q, a component of the complement activation system. SPR experiments were performed at 25 °C in HBS-EP+ buffer on a Biacore T200 system.

[0469] For these experiments, all four flow cells of the CM5 sensor chip were coated with goat F(ab’) 2Using anti-human Fc (Jackson ImmunoResearch), it was immobilized so that the response level became approximately 4000 RU. The Fc mutants and wild-type Fc were diluted to 100 nM in HBS-EP+ and captured on the surface of the chip at a flow rate of 10 μL / min for 120 seconds. Fcγ receptors (human and cyno, all purchased from R&D Systems) and complement (C1Q, purchased from Quidel Corporation and Complement Technology) were diluted in HBS-EP+ and then flowed as analytes at 30 μL / min for 120 seconds, followed by dissociation for 120 seconds. Regeneration was achieved by flowing 10 mM glycine pH 1.7 at 30 μL / min for 30 seconds, followed by stabilization for 60 seconds. Flow cell 1 was always left as a blank (without captured protein) for the purpose of background signal subtraction. The blank-subtracted sensorgrams of each captured Fc mutant were examined for the presence of binding to either Fcγ receptor protein or complement. Since there are polymorphisms in human Fc receptor IIA and RIIIA, both sequences were tested for binding. As shown below, all sets of mutations resulted in cleavage of C1Q, RIIA H167 mutant, RIIB / C, and RIIIB binding. Some mutant mutants showed residual binding to RI, RIIA R167, or RIIIA V176.

[0470]

Table 33

[0471] These Fc mutants were also evaluated for binding to the recombinant Fc receptor extracellular domain from cynomolgus monkeys, and it was verified that these mutations also disrupt binding in species that can be used to evaluate the toxicity of ADAPTIR (trademark) protein therapeutics. As shown by the results in the following table, the set of mutations disrupts binding to cynoRIIB and cynoRIII. Two mutants, TSC1004 and TSC1007, showed some residual binding to the cynoRI receptor.

[0472]

Table 34

[0473] Example 29: Differential Scanning Calorimetry (DSC) to Evaluate the Effect of Fc Mutations on Thermal Stability

[0474] The thermal stability of different Fc regions that can be used to construct ADAPTIR (trademark) bispecific antibodies was evaluated using differential scanning calorimetry (DSC). DSC measures the change in heat capacity associated with the the...

Claims

**Claim 1** A bispecific antibody comprising a polypeptide that, in order from the amino terminus to the carboxyl terminus, contains: (i) a first single-chain variable region fragment (scFv); (ii) a linker that is an immunoglobulin hinge region; (iii) an immunoglobulin constant region; (iv) a linker; and (v) a second scFv, wherein (a) the first scFv contains a human 4-1BB antigen-binding domain and the second scFv contains a human OX40 antigen-binding domain, or (b) the first scFv contains a human OX40 antigen-binding domain and the second scFv contains a human 4-1BB antigen-binding domain, the human 4-1BB antigen-binding domain contains six CDRs defined by IMGT in VH of SEQ ID NO: 17 and VL of SEQ ID NO: 18, or contains six CDRs defined by IMGT in VH of SEQ ID NO: 19 and VL of SEQ ID NO: 20, and the human OX40 antigen-binding domain contains six CDRs defined by IMGT in VH of SEQ ID NO: 29 and VL of SEQ ID NO: 28, A bispecific antibody. **Claim 2** (i) The human 4-1BB antigen-binding domain contains VH and VL, VH contains an amino acid sequence that is at least 95% identical to the amino acid sequence of SEQ ID NO: 17, and VL contains an amino acid sequence that is at least 95% identical to the amino acid sequence of SEQ ID NO: 18, and (ii) The human OX40 antigen-binding domain contains VH and VL, VH contains an amino acid sequence that is at least 95% identical to the amino acid sequence of SEQ ID NO: 29 or SEQ ID NO: 31, and VL contains an amino acid sequence that is at least 95% identical to the amino acid sequence of SEQ ID NO: 28 or SEQ ID NO: 30, The antibody according to claim 1. **Claim 3** (i) The human 4-1BB binding domain contains (a) VH containing the amino acid sequence of SEQ ID NO: 17 and VL containing the amino acid sequence of SEQ ID NO: 18, (b) VH containing the amino acid sequence of SEQ ID NO: 19 and VL containing the amino acid sequence of SEQ ID NO: 20, (c) VH containing the amino acid sequence of SEQ ID NO: 21 and VL containing the amino acid sequence of SEQ ID NO: 22, (d) VH containing the amino acid sequence of SEQ ID NO: 23 and VL containing the amino acid sequence of SEQ ID NO: 24, (e) VH containing the amino acid sequence of SEQ ID NO: 32 and VL containing the amino acid sequence of SEQ ID NO: 18, or (f) VH containing the amino acid sequence of SEQ ID NO: 143 and VL containing the amino acid sequence of SEQ ID NO: 20, and / or (ii) The human 4-1BB binding domain comprises VH and VL in the same polypeptide chain, the VH of the human 4-1BB binding domain is N-terminal to the VL of the human 4-1BB binding domain, or the VH of the human 4-1BB binding domain is C-terminal to the VL of the human 4-1BB binding domain, or The human 4-1BB binding domain comprises a linker between the VH and VL, the linker comprising the amino acid (Gly ４ Ser) ｎ and n=1-5 (SEQ ID NO:117). the antibody according to claim 1. **Claim 4** (i) the human 4-1BB binding domain is capable of binding to cynomolgus 4-1BB, or the human 4-1BB binding domain is capable of stimulating human 4-1BB activity, and / or (ii) the human 4-1BB binding domain comprises humanized VH and VL sequences, and / or (iii) The human OX40-binding domain comprises (a) VH comprising the amino acid sequence of SEQ ID NO: 25 and VL comprising the amino acid sequence of SEQ ID NO: 26; (b) VH comprising the amino acid sequence of SEQ ID NO: 27 and VL comprising the amino acid sequence of SEQ ID NO: 28; (c) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 26; (d) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 30; (e) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 28; (f) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 30; (g) VH comprising the amino acid sequence of SEQ ID NO: 33 and VL comprising the amino acid sequence of SEQ ID NO: 28; (h) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 34; (i) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 35; (j) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 36; (k) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 37; (l) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 34; (m) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 35; (n) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 36; (o) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 37; (p) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 38; (q) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 39; (r) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 40; or (s) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO:

41. The antibody according to any one of claims 1 to 3.

5. The human OX40-binding domain comprises VH and VL in the same polypeptide chain, and the VH of the human OX40-binding domain is N-terminal to the VL of the human OX40-binding domain, or the VH of the human OX40-binding domain is C-terminal to the VL of the human OX40-binding domain, or The human OX40 binding domain includes a linker between VH and VL, and the linker may include the amino acid sequence (Gly ４ Ser) ｎ and n = 1-5 (SEQ ID NO: 117). The antibody according to any one of claims 1 to 4. **Claim 6** The antibody according to any one of claims 1 to 5, wherein the 4-1BB binding domain comprises an scFv comprising any one amino acid sequence of SEQ ID NOs: 42, 44, 58, 63, 77, and 145. **Claim 7** The antibody according to any one of claims 1 to 6, wherein the OX40 binding domain comprises an scFv comprising any one amino acid sequence of SEQ ID NOs: 46, 47, 52, 54, 56, 59-62, 64-76, and 146. The antibody according to any one of claims 1 to 6. **Claim 8** (i) the OX40 binding domain is capable of binding to cynomolgus OX40, and / or (ii) the human OX40 binding domain is capable of stimulating human OX40 activity, and / or (iii) the OX40 binding domain comprises VH and VL sequences of mouse or rat, and / or (iv) the human 4-1BB binding domain comprises VH comprising the amino acid sequence of SEQ ID NO: 17 and VL comprising the amino acid sequence of SEQ ID NO: 18, and the OX40 binding domain comprises (a) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 28, (b) VH comprising the amino acid sequence of SEQ ID NO: 31 and VL comprising the amino acid sequence of SEQ ID NO: 30, or (c) VH comprising the amino acid sequence of SEQ ID NO: 29 and VL comprising the amino acid sequence of SEQ ID NO: 35, and / or (v) the human 4-1BB binding domain comprises an scFv comprising the amino acid sequence of SEQ ID NO: 58, and the human OX40 binding domain comprises an scFv comprising any one amino acid sequence of SEQ ID NOs: 59, 62, or 66, and / or (vi) the human 4-1BB binding domain and the human OX40 binding domain are on the same polypeptide, or (vii) the human 4-1BB binding domain is N-terminal to the human OX40 binding domain, or the human 4-1BB binding domain is C-terminal to the human OX40 binding domain, The antibody according to any one of claims 1 to 7. **Claim 9** The immunoglobulin constant region comprises the immunoglobulin CH2 domain and CH3 domain of IgG1, IgG2, IgG3, IgG4, IgA1, IgA2, or IgD. The antibody according to any one of claims 1 to 8. **Claim 10** A polynucleotide encoding the antibody according to any one of claims 1 to 9. **Claim 11** A vector comprising the polynucleotide according to claim 10.

12. (i) the polynucleotide according to claim 10 or the vector according to claim 11, or (ii) a combination of polynucleotides encoding the antibody according to any one of claims 1 to 9, or a combination of polynucleotides encoding the antibody according to any one of claims 1 to 9, wherein the polynucleotides are encoded on one vector or on a plurality of vectors, comprising a host cell.

13. A method for producing an antibody specifically binding to human 4-1BB and human OX40, comprising culturing the host cell according to claim 12 to produce the antibody.

14. A method for detecting 4-1BB and OX40 in a sample, comprising contacting the sample with the antibody according to any one of claims 1 to 9, wherein the sample contains cells.

15. A pharmaceutical composition comprising the antibody according to any one of claims 1 to 9 and a pharmaceutically acceptable excipient.

16. (a) NK cells, or (b) T cells, or T cells that are CD4+ T cells or CD8+ T cells, or (c) NK cells and T cells, or NK cells and T cells that are CD4+ T cells or CD8+ T cells, or (d) T cells, or T cells that are CD4+ T cells or CD8+ T cells contacting with the antibody according to any one of claims 1 to 9 or the pharmaceutical composition according to claim 15, comprising (a) for enhancing NK cell proliferation, or (b) for enhancing T cell proliferation, or (c) for enhancing NK cell proliferation and T cell proliferation, or (d) for stimulating the T cell co-stimulation pathway, an in vitro method.

17. (a) for enhancing the immune response in a subject, or (b) for increasing the number of tumor-infiltrating lymphocytes in a subject, or (c) for enhancing the expression of granzyme by effector cells in a subject, or (d) for reducing the number of tumor cells in a subject, or (e) for treating cancer in a subject, An antibody according to any one of claims 1 to 9 or a pharmaceutical composition according to claim 15 for use in a method, the method comprising administering to a subject an effective amount of an antibody according to any one of claims 1 to 9 or a pharmaceutical composition according to claim 15, the antibody or pharmaceutical composition.

18. (i) the cancer is a solid tumor cancer, and / or (ii) the cancer is a sarcoma, carcinoma, or lymphoma, and / or (iii) the cancer is selected from the group consisting of melanoma, renal cancer, pancreatic cancer, lung cancer, gastric cancer, colon / intestinal cancer, prostate cancer, ovarian cancer, breast cancer, liver cancer, brain cancer, or blood cancer, and / or (iv) the subject is human, or the subject is human and expresses 4-1BB and OX40 on tumor infiltrating lymphocytes, An antibody or pharmaceutical composition for use according to claim 17(e).

19. An antibody according to any one of claims 1 to 9 or a pharmaceutical composition according to claim 15 for use in therapy.

Images