Bispecific and trispecific binding proteins to PD-L1, CD137 and / or TGFβ and uses thereof

Abstract

The present disclosure provides binding proteins that bind PD-L1 and CD137 (PD-L1 / CD137 bispecifics), binding proteins that bind PD-L1 and TGFβ (PD-L1 / TGFβ bispecifics), binding proteins that bind PD-L1, TGFβ, and CD137 (PD-L1 / TGFβ / CD137 trispecifics), and binding proteins that bind CD137, TGFβ, and PD-L1 (CD137 / TGFβ / PD-L1 trispecifics). The present disclosure also provides pharmaceutical compositions comprising these binding proteins, and methods of their use to treat and / or prevent cancer.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This international patent application claims the benefit of U.S. Provisional Patent Application No. 63 / 240,404, filed September 3, 2021, the entire contents of which are incorporated herein by reference.

[0002] Description of sequence listing The sequence listing associated with this application is provided in XML format in lieu of a paper copy and is hereby incorporated by reference herein. The name of the XML file containing the sequence listing is 122863-5008_Sequence_Listing_ST.26.km. The text file is approximately 108000 bytes, was created on or about August 28, 2022, and has been submitted electronically via EFS-Web.

[0003] The present disclosure is in the field of immunotherapy and relates to binding proteins that bind PD-L1 and CD137 (PD-L1 / CD137 bispecific), binding proteins that bind PD-L1 and TGFβ (PD-L1 / TGFβ bispecific), and binding proteins that bind PD-L1, TGFβ, and CD137 (PD-L1 / TGFβ / CD137 trispecific). The present disclosure also relates to binding proteins that bind CD137, TGFβ, and PD-L1 (CD137 / TGFβ / PD-L1 trispecific). The present disclosure is also directed to polynucleotide sequences encoding these binding proteins and cells that produce them. The present disclosure further relates to pharmaceutical compositions comprising these binding proteins and methods of their use to modulate the PD-1 / PD-L1 and / or CD137 axis for immunotherapy. [Background technology]

[0004] Immune checkpoints refer to a set of inhibitory pathways that immune cells have to regulate and control the persistence of immune responses while maintaining self-tolerance. Upregulation of immune checkpoint molecules (e.g., PD-1, PD-L1, CTLA-4, TIM-3, Lag-3, VISTA, B7-H3, TIGIT, CD73, LAIR1) in the tumor microenvironment has been recognized as an important mechanism to limit the antitumor activity of effector T cells.

[0005] Programmed cell death-1 (PD-1) is one of the main immune T cell receptor checkpoint receptors targeted for cancer immunotherapy. PD-1 and its ligands programmed death ligand-1 and programmed death ligand-2 (PD-L1 and PD-L2, respectively) act as co-inhibitors that regulate the balance between T cell activation, tolerance, and immunopathology. In the absence of cancer, the PD-1 / PD-L1 axis functions to prevent excessive inflammation in normal tissues and helps maintain immune tolerance to self-antigens. In the presence of cancer, subversion of the PD-1 pathway provides a major mechanism of tumor immune resistance in both tumor and peripheral tissues. This axis is repurposed to promote cancer development and progression by enhancing tumor cell survival.

[0006] The PD-1 ligands, PD-L1 (also known as cluster of differentiation 274 (CD274)) or B7 homolog 1 (B7-H1), and PD-L2 (also known as B7-DC or CD273), are normally expressed on the surface of dendritic cells or macrophages. PD-L1 is also overexpressed on tumor cells or on non-transformed cells in the tumor microenvironment (TME). The interaction of PD-1 and PD-L1 leads to inhibition of T cell receptor (TCR) signaling and CD28 costimulation, ultimately leading to T cell inactivation and loss of proliferative capacity. PD-L1 expression in the tumor microenvironment allows cancer cells to exploit the PD-1 / PD-L1 checkpoint pathway as an evasion mechanism to prevent or evade the patient's antigen-specific T cell immune response.

[0007] Blockade of the PD-1 receptor or its ligands with antibodies deprives cancer cells of their evasive strategies and enhances or promotes antitumor immune responses. To date, six PD-1 / PD-L1 immune checkpoint inhibitors (ICIs) have been approved by the US Food and Drug Administration (FDA): three PD-1 inhibitors (nivolumab, pembrolizumab, and cemiplimab) and three PD-L1 inhibitors (atezolizumab, durvalumab, and avelumab).

[0008] Cancer immunotherapy using therapeutic monoclonal antibodies that neutralize PD-1 / PD-L1 checkpoint inhibition of the immune response has revolutionized the treatment of a wide variety of malignancies. However, some initial responders eventually develop acquired resistance to monotherapy and experience recurrent disease, and a significant number of patients demonstrate primary resistance and fail to respond to PD-1 / PD-L1 immune checkpoint inhibition. In light of the reported prevalence of checkpoint inhibitor treatment resistance, there is an unmet need for additional therapeutic agents for patients with refractory or recurrent cancers.

[0009] CD137 (4-1BB or TNFRSF9) is a member of the TNF receptor superfamily. CD137 is expressed on both innate and adaptive immune cells. It plays multifaceted roles in the tumor microenvironment (TME). It is predominantly upregulated on T cells in the TME and provides costimulation for the activation, proliferation, and survival of CD8 and CD4 T cells. It also regulates other cell functions in the TME, such as encouraging NK cells to interact with CD8 T cells and enhancing tumor antigen presentation on DC cells. CD137 agonists can enhance antitumor immune responses and can act synergistically with other antitumor agents, including PD1 / PD-L1 blocking antibodies.

[0010] Secretion of TGFβ is another major contributor to immune evasion and tumor progression. TGFβ promotes tumor progression and immune suppression by preventing T cell proliferation and decreasing the effector functions of both T cells and NK cells. It also enhances the function of T regulatory cells and induces epithelial-mesenchymal transition (EMT).

[0011] Because PD1, CD137, and TGFβ regulate independent immunosuppressive or immunostimulatory pathways, it is possible that targeting PD1 and CD137, PD1 and TGFβ, or PD1, CD137 and TGFβ may increase antitumor efficacy, particularly for immune-excluded and immune-inactive tumors. Summary of the Invention

[0012] The present disclosure addresses the above needs by providing multispecific binding proteins, including, for example, binding proteins that bind PD-L1 and CD137 (PD-L1 / CD137 bispecific), binding proteins that bind PD-L1 and TGFβ (PD-L1 / TGFβ bispecific), and binding proteins that bind PD-L1, TGFβ, and CD137 (PD-L1 / TGFβ / CD137 trispecific). In an exemplary embodiment, a PD-L1 / CD137 bispecific is a binding protein that binds PD-L1 and CD137 and comprises (a) an antibody scaffold module in an IgG format (e.g., a Y-shaped symmetric or asymmetric antibody comprising two heavy chains and two light chains) that comprises a first antigen-binding site that binds PD-L1 and a second antigen-binding site that binds PD-L1; and (b) at least one first binding module that comprises a third antigen-binding site that binds CD137. In an exemplary embodiment, a PD-L1 / TGFβ bispecific is a binding protein that binds PD-L1 and TGFβ and comprises (a) an antibody scaffold module in an IgG format that comprises a first antigen binding site that binds PD-L1 and a second antigen binding site that binds PD-L1; (b) at least one first binding module that comprises a third antigen binding site that binds TGFβ. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific is a binding protein that comprises (a) an antibody scaffold module in an IgG format that comprises a first antigen binding site that binds PD-L1 and a second antigen binding site that binds PD-L1; (b) at least one first binding module that comprises a third antigen binding site that binds TGFβ; and (c) at least one second binding module that comprises a fourth antigen binding site that binds CD137.

[0013] The present disclosure also relates to binding proteins that bind human CD137, including binding proteins that bind CD137, TGFβ, and PD-L1 (CD137 / TGFβ / PD-L1 trispecifics). More specifically, the present disclosure relates to binding proteins that bind CD137 and their use to stimulate CD137 and promote sustained anti-tumor immune responses. In an exemplary embodiment, the binding protein may bind to CD137 and comprises an antibody scaffold module in an IgG format comprising a first antigen binding site that binds CD137 and a second antigen binding site that binds CD137. In an exemplary embodiment, the CD137 / TGFβ / PD-L1 trispecific is a binding protein that comprises (a) an antibody scaffold module in an IgG format comprising a first antigen binding site that binds CD137 and a second antigen binding site that binds CD137; (b) at least one first binding module comprising a third antigen binding site that binds TGFβ; and (c) at least one second binding module comprising a fourth antigen binding site that binds PD-L1.

[0014] Despite the advances in immunotherapy resulting from the use of PD-1 / PD-L1 blockade, there is a need for binding proteins that can reverse immune suppression in the tumor microenvironment by binding to both PD-L1 and co-stimulatory targets on T cells and / or neutralizing TGFβ. Three TGFβ isoforms, TGFβ1, TGFβ2, and TGFβ3, are highly expressed in many tumors, which promote cancer progression primarily by suppressing both the innate and adaptive immune systems, and their serum concentrations correlate with poor prognosis. In the tumor microenvironment, TGFβ promotes tumor progression by inducing stromal modification, angiogenesis, and epithelial-mesenchymal transition (EMT). TGFβ signaling induces differentiation of Treg cells and drives metastasis through myeloid cells. In addition, TGFβ1 can directly inhibit T cell and NK cell function.

[0015] Binding proteins can bind two or three epitopes of a single antigen or to more than one target antigen simultaneously, allowing for multiple mechanistic functions and potential synergistic effects that cannot be achieved by monospecific therapeutic antibodies or fusion proteins. Advantageously, the use of binding proteins that bind two or three antigens may have a lower risk of toxicity than that associated with the use of multiple therapeutic agents.

[0016] In broad terms, the disclosed bispecifics and trispecifics disclosed herein are based on a native IgG scaffold, with or without modifications to promote heavy chain heterodimerization, and are characterized by two or three binding specificities (e.g., PD-L1, CD137 and / or TGFβ) contributed by antigen binding sites derived from fusion proteins prepared from antibody Fab or scFv fragments and / or the extracellular domain (ECD) of the TGFβRII receptor appended to the N- or C-terminus of the IgG heavy or light chain in the IgG scaffold.

[0017] In some embodiments, the PD-L1 / CD137 bispecific, PD-L1 / TGFβ bispecific, or PD-L1 / TGFβ / CD137 trispecific exhibits one or more of the following structural features, alone or in combination: (a) with anti-CD137 in scFv format, (b) having anti-PD-L1 in scFv format; (c) having an scFv fused to the N-terminus of an antibody light chain in an IgG scaffold; (d) having an scFv fused to the N-terminus of an antibody heavy chain in an IgG scaffold; (e) having an scFv fused to the C-terminus of an antibody light chain in an IgG scaffold; (f) having an scFv fused to the C-terminus of an antibody heavy chain in an IgG scaffold; (g) having an scFv of CD137 present in a monovalent or bivalent format; (h) comprises human TGFβRII fused to the C-terminus of the light chain in an IgG scaffold; or (i) It contains human TGFβRII fused to the C-terminus of the heavy chain in an IgG scaffold.

[0018] In some embodiments, the PD-L1 / CD137 bispecifics, PD-L1 / TGFβ bispecifics, and PD-L1 / TGFβ / CD137 trispecifics are + We provide an immunotherapy that targets CD8 T cells and reactivates dysfunctional tumor-infiltrating lymphocytes (TILs). Such binding proteins bind to PD-L1, and are involved in the pathogenesis of exhaustive TCD8 T cell-associated cancers that contribute to PD-1 / PD-L1 resistance. + It may be particularly beneficial as a therapeutic agent in solid tumors characterized by a microenvironment enriched in T cells and / or regulatory T cells. Indeed, by blocking the PD-1 / PD-L1 signaling axis, the immunosuppressive signals present in the TME are reduced and antitumor immunity is enhanced, which in turn produces durable clinical responses that prolong patient survival.

[0019] In some embodiments, the disclosed binding proteins that bind to PD-L1 are bispecific tetravalent molecules specific for PD-L1 and either CD137 or human TGFβ.

[0020] In other embodiments, the disclosed binding proteins that bind to PD-L1 are molecules that are specific for both PD-L1 and CD137 and human TGFβ.

[0021] In some embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics are designed to provide immunotherapy targeting CD137-expressing immune cells in the tumor microenvironment. Binding proteins that bind to CD137 may be particularly useful as therapeutic agents in solid tumors. In fact, activating the CD137 signaling axis enhances T effector function present in the TME and enhances anti-tumor immunity, which in turn can produce sustained clinical responses that prolong patient survival.

[0022] In other embodiments, the disclosed binding proteins bind to both CD137 and PD-L1 and human TGFβ.

[0023] According to some embodiments, the disclosed binding proteins that bind to PD-L1 comprise a set of six complementarity determining region (CDR) sequences selected from the group consisting of the three CDRs of an anti-PD-L1 antibody heavy chain (HC) variable region sequence selected from SEQ ID NOs: 1 and 3, and the three CDRs of a light chain variable region sequence selected from SEQ ID NOs: 2 and 4.

[0024] In some embodiments, a binding protein that binds to PD-L1 comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:10.

[0025] In some embodiments, the binding protein that binds to PD-L1 comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0026] In some embodiments, the binding protein that binds to PD-L1 comprises a heavy chain variable region sequence as shown in SEQ ID NO:1 or SEQ ID NO:3, or an analogue or derivative thereof having at least 90% sequence identity to SEQ ID NO:1 or SEQ ID NO:3.

[0027] In another embodiment, the binding protein that binds to PD-L1 comprises a light chain variable region sequence as shown in SEQ ID NO:2 or SEQ ID NO:4, or an analog or derivative thereof having at least 90% sequence identity to SEQ ID NO:2 or SEQ ID NO:4.

[0028] In other embodiments, the binding protein that binds to PD-L1 comprises a heavy chain variable region sequence shown in SEQ ID NO:1 or SEQ ID NO:3, and a light chain variable region sequence shown in SEQ ID NO:2 or SEQ ID NO:4.

[0029] In some embodiments, the binding protein that binds to PD-L1 comprises a heavy chain variable region sequence and a light chain variable region sequence selected from the following combinations: (a) a heavy chain variable region sequence comprising SEQ ID NO:1 and a light chain variable region sequence comprising SEQ ID NO:2; or (b) a heavy chain variable region sequence comprising SEQ ID NO:3 and a light chain variable region sequence comprising SEQ ID NO:4.

[0030] In some embodiments, the binding protein that binds to PD-L1 is (a) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:10; or (b) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0031] In some embodiments, a binding protein that binds to PD-L1 is provided, wherein the first and second antigen-binding sites bind to PD-L1 and the antibody scaffold module comprises: (i) a heavy chain variable region sequence set forth in SEQ ID NO:1, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:2, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66; or (ii) a heavy chain variable region sequence as set forth in SEQ ID NO:3, a heavy chain constant region sequence as set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence as set forth in SEQ ID NO:4, and a light chain constant region sequence as set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0032] In some embodiments, a binding protein that binds to PD-L1 is provided, wherein the first and second antigen-binding sites bind to PD-L1 and the antibody scaffold module comprises: The heavy chain sequence shown in SEQ ID NO:45 and the light chain sequence shown in SEQ ID NO:40.

[0033] In some embodiments, the binding protein that binds PD-L1 comprises one or more of the heavy chain variable region CDRs disclosed in Table 1 and / or one or more of the light chain variable region CDRs disclosed in Table 2.

[0034] In some embodiments, binding proteins that bind PD-L1 exhibit one or more of the following structural and functional characteristics, alone or in combination: (a) is specific for human PD-L1; (b) cross-reacts with cynomolgus monkey PD-L1; (c) disrupting the interaction of PD-1 and PD-L1; or (d) Reversing PD-1 / PD-L1 checkpoint-mediated inhibition of T cells.

[0035] In some embodiments, the binding proteins that bind PD-L1 specifically bind to human cells that express endogenous levels of PD-L1 and to host cells that have been engineered to overexpress human PD-L1. The binding proteins that bind PD-L1 also have a sub-nanomolar EC 50 The antibodies can bind to cells overexpressing human or cynomolgus PD-L1 with significant immunological effects.

[0036] In some embodiments, the binding protein that binds to PD-L1 cross-reacts with cynomolgus monkey PD-L1 (cynoPD-L1) and does not demonstrate cross-reactive binding with mouse PD-L1 (mu-PD-L1).

[0037] In some embodiments, the binding protein that binds to PD-L1 disrupts the human PD-1 / PD-L1 binding interaction.

[0038] In some embodiments, the binding protein that binds to PD-L1 disinhibits PD-1 / PD-L1 checkpoint-mediated inhibition of T cells.

[0039] In some embodiments, the binding protein that binds PD-L1 further comprises an engineered Fc region to eliminate / minimize cross-linking activity with Fc gamma R, which silences or eliminates Fc-mediated effector function of T cells.

[0040] The present disclosure also provides an isolated polynucleotide sequence encoding at least one of the above binding proteins that binds to PD-L1.

[0041] The present disclosure also provides a vector comprising at least one of the above polynucleotide sequences.

[0042] The present disclosure also provides a cell comprising one of the above polynucleotide sequences or one of the above vectors.

[0043] The present disclosure also provides pharmaceutical compositions comprising, or consisting of, at least one binding protein that binds to PD-L1, and optionally a pharma- ceutically acceptable diluent, carrier, vehicle, and / or excipient, such pharmaceutical compositions may be used for the treatment of cancer.

[0044] The present disclosure also relates to a method for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the disclosed binding proteins that binds to PD-L1, alone or in combination with another therapeutic agent.

[0045] According to some embodiments, the binding protein that binds to CD137 comprises a set of six complementarity determining region (CDR) sequences selected from the group consisting of the three CDRs of an anti-CD137 antibody heavy chain (HC) variable region sequence selected from SEQ ID NOs: 16, 18, and 20, and the three CDRs of a light chain variable region sequence selected from SEQ ID NOs: 17, 19, and 21.

[0046] In some embodiments, a binding protein that binds CD137 comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26.

[0047] In some embodiments, a binding protein that binds CD137 comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:27, CDR2: SEQ ID NO:28, and CDR3: SEQ ID NO:29; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:30, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:31.

[0048] In some embodiments, a binding protein that binds CD137 comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO: 32, CDR2: SEQ ID NO: 33, and CDR3: SEQ ID NO: 34; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO: 35, CDR2: SEQ ID NO: 36, and CDR3: SEQ ID NO: 37.

[0049] In some embodiments, the binding protein that binds to CD137 comprises a heavy chain variable region sequence as set forth in SEQ ID NO: 16, 18, or 20, or an analog or derivative thereof having at least 90% sequence identity to SEQ ID NO: 16, 18, or 20.

[0050] In other embodiments, the binding protein that binds to CD137 comprises a light chain variable region sequence as shown in SEQ ID NO: 17, 19, or 21, or an analog or derivative thereof having at least 90% sequence identity to SEQ ID NO: 17, 19, or 21.

[0051] In other embodiments, the binding protein that binds CD137 comprises a heavy chain variable region sequence shown in SEQ ID NO: 16, 18, or 20 and a light chain variable region sequence shown in SEQ ID NO: 17, 19, or 21.

[0052] In some embodiments, the binding proteins that bind comprise heavy chain and light chain variable region sequences selected from the following combinations: (a) a heavy chain variable region sequence comprising SEQ ID NO:16 and a light chain variable region sequence comprising SEQ ID NO:17; (b) a heavy chain variable region sequence comprising SEQ ID NO: 18 and a light chain variable region sequence comprising SEQ ID NO: 19; and (c) a heavy chain variable region sequence comprising SEQ ID NO:20 and a light chain variable region sequence comprising SEQ ID NO:21.

[0053] In some embodiments, a binding protein that binds to CD137 is provided, comprising: (a) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26; (b) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:27, CDR2: SEQ ID NO:28, and CDR3: SEQ ID NO:29; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO:30, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:31; or (c) a heavy chain variable region sequence comprising CDR1: SEQ ID NO: 32, CDR2: SEQ ID NO: 33, and CDR3: SEQ ID NO: 34; and / or a light chain variable region sequence comprising CDR1: SEQ ID NO: 35, CDR2: SEQ ID NO: 36, and CDR3: SEQ ID NO: 37.

[0054] In some embodiments, a binding protein that binds to CD137 is provided, wherein the first and second antigen binding sites bind to CD137, and wherein the antibody scaffold module comprises: (i) a heavy chain variable region sequence set forth in SEQ ID NO: 16, a heavy chain constant region sequence set forth in SEQ ID NO: 60, SEQ ID NO: 61, SEQ ID NO: 62, SEQ ID NO: 63, or SEQ ID NO: 64, a light chain variable region sequence set forth in SEQ ID NO: 17, and a light chain constant region sequence set forth in SEQ ID NO: 65 or SEQ ID NO: 66; (ii) a heavy chain variable region sequence set forth in SEQ ID NO: 18, a heavy chain constant region sequence set forth in SEQ ID NO: 60, SEQ ID NO: 61, SEQ ID NO: 62, SEQ ID NO: 63, or SEQ ID NO: 64, a light chain variable region sequence set forth in SEQ ID NO: 19, and a light chain constant region sequence set forth in SEQ ID NO: 65 or SEQ ID NO: 66; or (iii) a heavy chain variable region sequence set forth in SEQ ID NO:20, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:21, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0055] In some embodiments, a binding protein that binds to CD137 is provided, wherein the first and second antigen binding sites bind to CD137, and wherein the antibody scaffold module comprises a heavy chain sequence set forth in SEQ ID NO: 75 and a light chain sequence set forth in SEQ ID NO: 76.

[0056] In some embodiments, the anti-CD137 antibody comprises one or more of the heavy chain variable region CDRs disclosed in Table 3 and / or one or more of the light chain variable region CDRs disclosed in Table 4.

[0057] In some embodiments, binding proteins that bind CD137 exhibit one or more of the following structural and functional characteristics, either alone or in combination: (a) specific for human CD137; (b) cross-reacts with cynomolgus monkey CD137; (c) disrupting (e.g., reducing or preventing) human CD137L binding to CD137; (d) exhibits fast on and fast off properties for CD137; (e) has cross-linking-dependent agonistic activity for CD137 signaling; or (f) Activates T cells in a cross-linking dependent manner.

[0058] In some embodiments, binding proteins that bind CD137 specifically bind to human cells that express endogenous levels of CD137 and to host cells engineered to overexpress human CD137. In some embodiments, binding proteins that bind CD137 have an EC in the range of 0.2-1.1 nM (e.g., 0.2 nM, 0.3 nM, 0.4 nM, 0.5 nM, 0.6 nM, 0.7 nM, 0.8 nM, 0.9 nM, 1.0 nM, or 1.1 nM). 50 The antibody binds to cells overexpressing human or cynomolgus CD137 with high antibody titers.

[0059] In some embodiments, binding proteins that bind to CD137 have fast on and fast off kinetic properties.

[0060] In some embodiments, a binding protein that binds CD137 cross-reacts with cynomolgus monkey CD137 (cynoCD137) and does not demonstrate cross-reactive binding with mouse CD137 (mu-CD137).

[0061] In some embodiments, the binding protein that binds to CD137 disrupts the CD137 ligand / CD137 binding interaction.

[0062] In some embodiments, the binding proteins that bind CD137 have cross-linking-dependent agonist activity for CD137 signaling.

[0063] In some embodiments, binding proteins that bind to CD137 activate T cells in a cross-linking dependent manner.

[0064] In some embodiments, the binding protein that binds CD137 further comprises an Fc region engineered to eliminate / minimize cross-linking activity with Fc gamma R, which silences or eliminates Fc-mediated effector function of T cells.

[0065] The present disclosure also provides an isolated polynucleotide sequence encoding at least one of the above binding proteins that binds to CD137.

[0066] The present disclosure also provides a vector comprising at least one of the above polynucleotide sequences.

[0067] The present disclosure also provides a cell comprising one of the above polynucleotide sequences or one of the above vectors.

[0068] The present disclosure also provides pharmaceutical compositions comprising or consisting of at least one binding protein that binds CD137 and, optionally, a pharma- ceutically acceptable diluent, carrier, vehicle, and / or excipient, such pharmaceutical compositions may be used for the treatment of cancer.

[0069] The present disclosure also relates to methods for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the disclosed binding proteins that binds CD137, alone or in combination with another therapeutic agent.

[0070] In an exemplary embodiment, a PD-L1 / CD137 bispecific is a binding protein that binds to PD-L1 and CD137, and comprises (a) an antibody scaffold module in an IgG format that comprises a first antigen binding site that binds to PD-L1 and a second antigen binding site that binds to PD-L1; and (b) at least one first binding module that comprises a third antigen binding site that binds to CD137.

[0071] In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / CD137 bispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and a light chain variable region sequence comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO: 10. In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / CD137 bispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and a light chain variable region sequence comprising CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0072] In exemplary embodiments, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises a heavy chain variable region sequence shown in SEQ ID NO:1 or SEQ ID NO:3; and a light chain variable region sequence shown in SEQ ID NO:2 or SEQ ID NO:4.

[0073] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises a heavy chain variable region sequence shown in SEQ ID NO: 1 and a light chain variable region sequence shown in SEQ ID NO: 2. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises a heavy chain variable region sequence shown in SEQ ID NO: 3 and a light chain variable region sequence shown in SEQ ID NO: 4.

[0074] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 1, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:2, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO:3, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:4, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0075] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises the heavy chain sequence set forth in SEQ ID NO: 42 and the light chain sequence set forth in SEQ ID NO: 40. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / CD137 bispecific comprises the heavy chain sequence set forth in SEQ ID NO: 45 and the light chain sequence set forth in SEQ ID NO: 40.

[0076] In an exemplary embodiment, a PD-L1 / CD137 bispecific has one first link module. In an exemplary embodiment, a PD-L1 / CD137 bispecific has two first link modules. In an exemplary embodiment, a PD-L1 / CD137 bispecific has an antibody scaffold module comprising a heavy chain sequence comprising a C-terminus and an N-terminus, where the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the first link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the first link module and the antibody scaffold module are covalently attached to each other directly or through an interlinker. In an exemplary embodiment, the first link module and antibody scaffold module of the PD-L1 / CD137 bispecific are covalently attached to each other through an interlinker, which is the N-to-C-terminal sequence shown in SEQ ID NO: 58. In an exemplary embodiment, the first link module and antibody scaffold module of the PD-L1 / CD137 bispecific are covalently attached to each other through an interlinker, which is the N-to-C-terminal sequence shown in SEQ ID NO: 59. In an exemplary embodiment, the first link module of the PD-L1 / CD137 bispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the PD-L1 / CD137 bispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, when there is more than one first link module in a PD-L1 / CD137 bispecific, each is covalently attached to a different antibody scaffold module sequence or to a different end of the antibody scaffold module.

[0077] In an exemplary embodiment, the first binding module in the PD-L1 / CD137 bispecific is an scFv, which comprises a heavy chain variable region sequence and a light chain variable sequence, where the sequences are covalently attached to each other directly or through an scFv fusion linker. In an exemplary embodiment, the scFv fusion linker comprises glycine and serine. In an exemplary embodiment, the scFv fusion linker comprises the sequence Gly-Gly-Gly-Ser. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker is the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:59. In an exemplary embodiment, the scFv fusion linker is the sequence set forth in SEQ ID NO:59. In an exemplary embodiment, the first binding module in the PD-L1 / CD137 bispecific comprises the sequence set forth in SEQ ID NO:53. In an exemplary embodiment, the first link module in the PD-L1 / CD137 bispecific comprises the sequence set forth in SEQ ID NO: 54. In an exemplary embodiment, the first link module in the PD-L1 / CD137 bispecific comprises the sequence set forth in SEQ ID NO: 55. In an exemplary embodiment, the first link module in the PD-L1 / CD137 bispecific comprises the sequence set forth in SEQ ID NO: 56.

[0078] In an exemplary embodiment, the first link module in a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence comprising: CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and a light chain variable region sequence comprising: CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26. In an exemplary embodiment, the first link module in a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence comprising: CDR1: SEQ ID NO:27, CDR2: SEQ ID NO:28, and CDR3: SEQ ID NO:29; and a light chain variable region sequence comprising: CDR1: SEQ ID NO:30, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:31. In an exemplary embodiment, the first binding module in the PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence comprising: CDR1: SEQ ID NO:32, CDR2: SEQ ID NO:33, and CDR3: SEQ ID NO:34; and a light chain variable region sequence comprising: CDR1: SEQ ID NO:35, CDR2: SEQ ID NO:36, and CDR3: SEQ ID NO:37.

[0079] In an exemplary embodiment, the first link module in a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence set forth in SEQ ID NO: 16 and a light chain variable region sequence set forth in SEQ ID NO: 17. In an exemplary embodiment, the first link module in a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence set forth in SEQ ID NO: 18 and a light chain variable region sequence set forth in SEQ ID NO: 19. In an exemplary embodiment, the first link module in a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence set forth in SEQ ID NO: 20 and a light chain variable region sequence set forth in SEQ ID NO: 21.

[0080] In an exemplary embodiment, a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, the heavy chain sequence of the antibody scaffold module and first link module set forth in SEQ ID NO:38; and the light chain sequence of the antibody scaffold module set forth in SEQ ID NO:40. In an exemplary embodiment, a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, the heavy chain sequence of the antibody scaffold module and first link module set forth in SEQ ID NO:44; and the light chain sequence of the antibody scaffold module set forth in SEQ ID NO:40. In an exemplary embodiment, a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, the heavy chain sequence of the antibody scaffold module and first link module set forth in SEQ ID NO:45; and the light chain sequence of the antibody scaffold module and first link module set forth in SEQ ID NO:46. In an exemplary embodiment, a PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, the heavy chain sequence of the antibody scaffold module and first link module set forth in SEQ ID NO:47; and the light chain sequence of the antibody scaffold module set forth in SEQ ID NO:40. In an exemplary embodiment, the PD-L1 / CD137 bispecific comprises, from N-terminus to C-terminus, a heavy chain sequence of an antibody scaffold module and a first binding module set forth in SEQ ID NO:50; and a light chain sequence of an antibody scaffold module set forth in SEQ ID NO:40.

[0081] In an exemplary embodiment, the antibody scaffold module of the PD-L1 / CD137 bispecific further comprises a constant region. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / CD137 bispecific comprises at least one Fc silencing mutation. In an exemplary embodiment, the Fc silencing mutation in the constant region of the antibody scaffold module of the PD-L1 / CD137 bispecific is L234AL235A or N297A. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / CD137 bispecific comprises a knob-in-hole (KiH) mutation.

[0082] According to some embodiments, PD-L1 / CD137 bispecifics (e.g., 1923Ab8, 1923Ab11, 1923Ab12, 1923Ab13, and 1923Ab18) are capable of effectively blocking the interaction between PD-L1 and its receptor PD-1, and between CD137 and its ligand. The disclosed PD-L1 / CD137 bispecifics comprise an amino acid sequence derived from one of the binding proteins that bind to PD-L1 (e.g., 1923Ab2 or 1923Ab3) disclosed herein as a PD-L1 antibody scaffold module, and an amino acid sequence derived from one of the binding proteins that bind to CD137 (e.g., 1923Ab4, 1923Ab5, or 1923Ab6) disclosed herein as a first binding module for CD137.

[0083] In some embodiments, a PD-L1 / CD137 bispecific comprises a binding protein that binds to PD-L1 and disrupts the PD-1 / PD-L1 binding interaction, disabling PD-1 / PD-L1 checkpoint-mediated inhibition of T cells. As exemplified herein, in a non-limiting example, an antibody scaffold module that binds to PD-L1 can comprise a single-chain variable fragment (e.g., a fusion protein of the variable regions of the heavy (VH) and light (VL) chains of one of the disclosed binding proteins that binds to PD-L1, connected with a linker peptide) (scFv).

[0084] In some embodiments, the PD-L1 / CD137 bispecific comprises a CD137 binding module comprised of a fragment derived from one of the disclosed binding proteins that bind to CD137. In one embodiment, the CD137 binding module can be in the form of a binding fragment derived from one of the binding proteins disclosed herein that bind to CD137. As exemplified herein, in a non-limiting example, the CD137 binding module can comprise a single chain variable fragment (e.g., a fusion protein of the variable regions of the heavy (VH) and light (VL) chains of one of the disclosed binding proteins that bind to CD137, connected with a linker peptide). Alternatively, the CD137 binding module can be in the form of an IgG molecule.

[0085] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab8 comprises an antibody scaffold module with two Fabs from 1923Ab3 that binds PD-L1, a human IgG1 Fc comprising two Fc constant chains with L234A L235A mutations (SEQ ID NO: 61), and an scFv fragment (VH precedes VL) derived from 1923Ab4 (SEQ ID NO: 53) attached to the C-terminus of each of the two Fc constant chains.

[0086] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab11 comprises an antibody scaffold module with two Fabs from 1923Ab3 that binds PD-L1, a human IgG1 Fc comprising two Fc constant chains with L234A L235A mutations (SEQ ID NO: 61), and an scFv fragment (VL precedes VH) derived from 1923Ab4 (SEQ ID NO: 54) attached to the C-terminus of each of the two Fc constant chains.

[0087] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab12 comprises an antibody scaffold module with two Fabs from 1923Ab3 that binds PD-L1, a human IgG1 Fc comprising two constant chains with L234A L235A mutations (SEQ ID NO: 61), and a disulfide bond stabilized scFv fragment (VH precedes VL) derived from 1923Ab4 (SEQ ID NO: 56) attached to the N-terminus of each of the light chains in the Fab.

[0088] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab13 comprises an antibody scaffold module with two Fabs from 1923Ab3 that binds PD-L1, a human IgG1 Fc comprising two constant chains with L234A L235A mutations (SEQ ID NO: 61), and a disulfide bond stabilized scFv fragment (VH precedes VL) derived from 1923Ab4 (SEQ ID NO: 56) attached to the N-terminus of each of the heavy chains in the Fab.

[0089] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab18 comprises an antibody scaffold module with two Fabs from 1923Ab3 that binds PD-L1, a human IgG1 Fc comprising two Fc constant chains with L234A L235A mutations (SEQ ID NO: 61), and a disulfide bond stabilized scFv fragment (VH precedes VL) derived from 1923Ab4 (SEQ ID NO: 55) attached to the C-terminus of each of the Fc constant chains.

[0090] In some embodiments, the PD-L1 / CD137 bispecific comprises a variable heavy chain sequence and a variable light chain sequence selected from the following combinations: (a) a heavy chain sequence comprising SEQ ID NO:38 and a light chain sequence comprising SEQ ID NO:40; (b) a heavy chain sequence comprising SEQ ID NO:44 and a light chain sequence comprising SEQ ID NO:40; (c) a heavy chain sequence comprising SEQ ID NO:45 and a light chain sequence comprising SEQ ID NO:46; (d) a heavy chain sequence comprising SEQ ID NO:47 and a light chain sequence comprising SEQ ID NO:40; and (e) a heavy chain sequence comprising SEQ ID NO:50 and a light chain sequence comprising SEQ ID NO:40.

[0091] In some embodiments, the PD-L1 / CD137 bispecific comprises a heavy chain sequence selected from the group consisting of SEQ ID NOs: 38, 42, 44, 45, 47, and 50, or a fragment thereof having at least 90% sequence identity to SEQ ID NO: 38, 42, 44, 45, 47, or 50.

[0092] In another embodiment, the PD-L1 / CD137 bispecific comprises a light chain sequence selected from the group consisting of SEQ ID NOs: 40 and 46, or a fragment derivative thereof having at least 90% sequence identity to SEQ ID NO: 40 or 46.

[0093] In some embodiments, the CD137 binding module is an scFv subunit that blocks the CD137 / CD137 ligand interaction and has cross-linking dependent agonist activity for CD137 signaling and T cell activation, hi some embodiments, the CD137 binding module is an scFv subunit stabilized with a disulfide bond.

[0094] In some embodiments, the PD-L1 / CD137 bispecific comprises an antibody scaffold module with an IgG format in which a CD137 scFv binding module is fused to the C-terminus of the heavy or light chain of the antibody scaffold module to create the PD-L1 / CD137 bispecific.

[0095] In some embodiments, the disclosed PD-L1 / CD137 bispecifics further comprise an engineered Fc region to eliminate / minimize cross-linking activity with Fc gamma R, which silences or eliminates Fc-mediated effector function of T cells.

[0096] The present disclosure also provides pharmaceutical compositions comprising, or alternatively consisting of, at least one of the PD-L1 / CD137 bispecifics disclosed herein and, optionally, a pharma- ceutically acceptable diluent, carrier, vehicle and / or excipient. Such pharmaceutical compositions may be used for the treatment of cancer.

[0097] The present disclosure also relates to a method for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the disclosed PD-L1 / CD137 bispecifics, alone or in combination with another therapeutic agent.

[0098] The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / CD137 bispecific sequences described herein.The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / CD137 bispecific sequences described herein.

[0099] The present disclosure also provides vectors comprising the PD-L1 / CD137 bispecific polynucleotides described herein.

[0100] The present disclosure also provides vectors comprising at least one of the PD-L1 / CD137 bispecific polynucleotide sequences described herein.

[0101] The present disclosure also provides a cell comprising one of the PD-L1 / CD137 bispecific polynucleotide sequences described herein, or one of the above vectors.

[0102] In an exemplary embodiment, a PD-L1 / TGFβ bispecific is a binding protein that binds to PD-L1 and TGFβ, and comprises (a) an antibody scaffold module in an IgG format that comprises a first antigen-binding site that binds to PD-L1 and a second antigen-binding site that binds to PD-L1; and (b) at least a first binding module that comprises a third antigen-binding site that binds to TGFβ.

[0103] In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / TGFβ bispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and a light chain variable region sequence comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:10. In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / TGFβ bispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and a light chain variable region sequence comprising CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0104] In exemplary embodiments, the antibody scaffold module of a PD-L1 / TGFβ bispecific comprises a heavy chain variable region sequence shown in SEQ ID NO:1 or SEQ ID NO:3 and a light chain variable region sequence shown in SEQ ID NO:2 or SEQ ID NO:4.

[0105] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 1 and a light chain variable region sequence set forth in SEQ ID NO: 2. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 3 and a light chain variable region sequence set forth in SEQ ID NO: 4.

[0106] In an exemplary embodiment, an antibody scaffold module of a PD-L1 / TGFβ bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 1, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:2, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66. In an exemplary embodiment, an antibody scaffold module of a PD-L1 / TGFβ bispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO:3, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:4, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0107] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ bispecific comprises the heavy chain sequence set forth in SEQ ID NO: 42 and the light chain sequence set forth in SEQ ID NO: 40. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ bispecific comprises the heavy chain sequence set forth in SEQ ID NO: 45 and the light chain sequence set forth in SEQ ID NO: 40.

[0108] In an exemplary embodiment, a PD-L1 / TGFβ bispecific has one first link module. In an exemplary embodiment, a PD-L1 / TGFβ bispecific has two first link modules. In an exemplary embodiment, a PD-L1 / TGFβ bispecific has an antibody scaffold module comprising a heavy chain sequence comprising a C-terminus and an N-terminus, where the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the first link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the first link module and the antibody scaffold module are covalently attached to each other directly or through an interlinker. In an exemplary embodiment, the first link module and the antibody scaffold module of the PD-L1 / TGFβ bispecific are covalently attached to each other through an interlinker, where the interlinker is an N-terminus to C-terminus sequence as shown in SEQ ID NO:58. In an exemplary embodiment, the first link module and the antibody scaffold module of the PD-L1 / TGFβ bispecific are covalently attached to one another through an interlinker, which is an N-terminal to C-terminal sequence as shown in SEQ ID NO: 59. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ bispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ bispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, when there is more than one first link module in a PD-L1 / TGFβ bispecific, each is covalently attached to a different antibody scaffold module sequence or a different end of the antibody scaffold module.

[0109] In an exemplary embodiment, the first binding module in the PD-L1 / TGFβ bispecific comprises the extracellular domain of TGFβRII. In an exemplary embodiment, the extracellular domain of TGFβRII sequence is set forth in SEQ ID NO:67.

[0110] In an exemplary embodiment, a PD-L1 / TGFβ bispecific has two first link modules. In an exemplary embodiment, the heavy chain sequence of the antibody scaffold module and first link module of the PD-L1 / TGFβ bispecific comprises the sequence set forth in SEQ ID NO:51; and wherein the light chain sequence of the antibody scaffold module of the PD-L1 / TGFβ bispecific comprises the sequence from N-terminus to C-terminus as set forth in SEQ ID NO:40.

[0111] In an exemplary embodiment, the antibody scaffold module of the PD-L1 / TGFβ bispecific further comprises a constant region. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / TGFβ bispecific comprises at least one Fc silencing mutation. In an exemplary embodiment, the Fc silencing mutation in the constant region of the antibody scaffold module of the PD-L1 / TGFβ bispecific is L234A L235A or N297A. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / TGFβ bispecific comprises a knob-in-hole (KiH) mutation.

[0112] In some embodiments, the present disclosure provides PD-L1 / TGFβ bispecifics (e.g., 1923Ab20) that can effectively block the interaction between PD-L1 and its receptor PD-1, and sequester TGFβ. The disclosed PD-L1 / TGFβ bispecifics include all or part of an amino acid sequence derived from one of the binding proteins that bind PD-L1 disclosed herein (e.g., 1923Ab2 or 1923Ab3) as a PD-L1 antibody scaffold module, and an amino acid sequence that includes the full extracellular domain of TGFβRII or a truncation of TGFβRII (e.g., an N-terminal or C-terminal truncation), provided that the sequence is capable of binding to and neutralizing the biological activity of TGFβ. In some embodiments, the first binding module added to the binding protein that binds PD-L1 is a recombinant TGFβ binding protein derived from the ECD of the human TNFβRII receptor.

[0113] In some embodiments, the PD-L1 / TGFβ bispecific 1923Ab20 comprises an antibody scaffold module with two Fabs derived from 1923Ab3 that binds to PD-L1, a human IgG1 Fc with two Fc constant chains with L234A L235A mutations (SEQ ID NO:61), and two polypeptides encoding the extracellular domain of TGFβRII (SEQ ID NO:67), each attached to the C-terminus of each of the Fc constant chains.

[0114] In some embodiments, the PD-L1 / TGFβ bispecific comprises the heavy and / or light chain sequences disclosed in Table 5. In other embodiments, the PD-L1 / TGFβ bispecific comprises the heavy and / or light chain sequences disclosed in Table 6.

[0115] In some embodiments, the PD-L1 / TGFβ bispecific comprises a heavy chain sequence comprising SEQ ID NO:51 and a light chain sequence comprising SEQ ID NO:40.

[0116] In some embodiments, the PD-L1 / TGFβ bispecific comprises a heavy chain sequence selected from the group consisting of SEQ ID NOs: 42, 45, and 51, or a fragment thereof having at least 90% sequence identity to SEQ ID NO: 42, 45, or 51.

[0117] In another embodiment, the PD-L1 / TGFβ bispecific comprises a light chain sequence selected from the group consisting of SEQ ID NOs: 39 and 40, or a fragment derivative thereof having at least 90% sequence identity to SEQ ID NOs: 39 and 40.

[0118] In some embodiments, the PD-L1 / TGFβ bispecific exhibits one or more of the following characteristics, alone or in combination: (a) is specific for human PD-L1 and binds to human TGFβ; (b) cross-reacts with cynomolgus monkey PD-L1; (c) disrupting the interaction of PD-1 and PD-L1; (d) deactivating T cell PD-L1-mediated checkpoint inhibitory signals; or (e) sequestering human TGFβ;

[0119] The present disclosure also provides pharmaceutical compositions comprising, or alternatively consisting of, at least one of the PD-L1 / TGFβ bispecifics disclosed herein, and optionally a pharma- ceutically acceptable diluent, carrier, vehicle and / or excipient. Such pharmaceutical compositions may be used for antibody-based immunotherapy of cancer.

[0120] The disclosure also relates to a method for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the disclosed PD-L1 / TGFβ bispecifics, alone or in combination with another therapeutic agent.

[0121] The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / TGFβ bispecific sequences described herein.The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / TGFβ bispecific sequences described herein.

[0122] The disclosure also provides vectors comprising the PD-L1 / TGFβ bispecific polynucleotides described herein.

[0123] The present disclosure also provides vectors comprising at least one of the PD-L1 / TGFβ bispecific polynucleotide sequences described herein.

[0124] The present disclosure also provides a cell comprising one of the PD-L1 / TGFβ bispecific polynucleotide sequences described herein, or one of the above vectors.

[0125] The present disclosure also provides binding proteins that bind PD-L1, TGFβ, and CD137, comprising (a) an antibody scaffold module in an IgG format comprising a first antigen-binding site that binds PD-L1 and a second antigen-binding site that binds PD-L1; (b) at least one first binding module comprising a third antigen-binding site that binds TGFβ; and (c) at least one second binding module comprising a fourth antigen-binding site that binds CD137. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific is constructed in the form of a recombinant protein that comprises an antibody scaffold module that binds PD-L1, a first binding module that comprises an amino acid sequence derived from a TGFβ receptor II binding protein capable of binding to and neutralizing the activity of human TGFβ, and a second binding module that binds CD137.

[0126] In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / TGFβ / CD137 trispecific comprise: (i) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and a light chain variable region sequence comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:10. In an exemplary embodiment, the first and second antigen-binding sites of the PD-L1 / TGFβ / CD137 trispecific comprise: (i) a heavy chain variable region sequence comprising CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and a light chain variable region sequence comprising CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0127] In exemplary embodiments, the antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence shown in SEQ ID NO:1 or SEQ ID NO:3; and a light chain variable region sequence shown in SEQ ID NO:2 or SEQ ID NO:4.

[0128] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 1 and a light chain variable region sequence set forth in SEQ ID NO: 2. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 3 and a light chain variable region sequence set forth in SEQ ID NO: 4.

[0129] In an exemplary embodiment, an antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 1, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:2, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66. In an exemplary embodiment, an antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO:3, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:4, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0130] In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence set forth in SEQ ID NO: 42 and the light chain sequence set forth in SEQ ID NO: 40. In an exemplary embodiment, the antibody scaffold module of a PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence set forth in SEQ ID NO: 45 and the light chain sequence set forth in SEQ ID NO: 40.

[0131] In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has one first link module. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has two first link modules. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has an antibody scaffold module comprising a heavy chain sequence comprising a C-terminus and an N-terminus, where the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the first link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the first link module and the antibody scaffold module are covalently attached to each other directly or through a first link module interlinker. In an exemplary embodiment, the first link module and antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific are covalently attached to each other through a first link module interlinker, which is an N-terminal to C-terminal sequence as set forth in SEQ ID NO: 58. In an exemplary embodiment, the first link module and antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific are covalently attached to each other through a first link module interlinker, which is an N-terminal to C-terminal sequence as set forth in SEQ ID NO: 59. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, when there is more than one first binding module in a PD-L1 / TGFβ / CD137 trispecific, each is covalently attached to a different antibody scaffold module sequence or to a different end of the antibody scaffold module.

[0132] In an exemplary embodiment, the first binding module in the PD-L1 / TGFβ / CD137 trispecific comprises the extracellular domain of TGFβRII. In an exemplary embodiment, the extracellular domain of TGFβRII in the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO:67.

[0133] In an exemplary embodiment, the second binding module in the PD-L1 / TGFβ / CD137 trispecific is an scFv that comprises a heavy chain variable region sequence and a light chain variable sequence, where the sequences are covalently attached to each other directly or through an scFv fusion linker. In an exemplary embodiment, the scFv fusion linker comprises glycine and serine. In an exemplary embodiment, the scFv fusion linker comprises the sequence Gly-Gly-Gly-Ser. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker is the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:59. In an exemplary embodiment, the scFv fusion linker is the sequence set forth in SEQ ID NO:59. In an exemplary embodiment, the second binding module in the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO:53. In an exemplary embodiment, the second link module in the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 54. In an exemplary embodiment, the second link module in the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 55. In an exemplary embodiment, the second link module in the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 56.

[0134] In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific has one second link module. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific has two second link modules. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific has an antibody scaffold module comprising a heavy chain sequence comprising a C-terminus and an N-terminus, where the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the second link module and the antibody scaffold module are covalently attached to each other directly or through a second link module interlinker.

[0135] In an exemplary embodiment, the second link module and the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific are covalently attached to each other through a second link module interlinker, which is shown in SEQ ID NO: 58. In an exemplary embodiment, the second link module and the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific are covalently attached to each other through a second link module interlinker, which is shown in SEQ ID NO: 59.

[0136] In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence and the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence.

[0137] In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence and the first link module of the PD-L1 / TGFβ / CD137 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence.

[0138] In an exemplary embodiment, when there is more than one second link module in a PD-L1 / TGFβ / CD137 trispecific, each is covalently attached to a different antibody scaffold module sequence or to a different terminus of the antibody scaffold module (e.g., one link module may be attached to the C-terminus of a heavy chain in an antibody scaffold module and the other link module may be attached to the N-terminus of the same heavy chain). In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has one second link module, in which one second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has one second link module, in which one second link module is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific has two second link modules, where one second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence and the other second link module is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific is an scFv.

[0139] In an exemplary embodiment, the second binding module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and a light chain variable region sequence comprising CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26. In an exemplary embodiment, the second binding module of a PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO:27, CDR2: SEQ ID NO:28, and CDR3: SEQ ID NO:29; and a light chain variable region sequence comprising CDR1: SEQ ID NO:30, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:31. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence comprising CDR1: SEQ ID NO: 32, CDR2: SEQ ID NO: 33, and CDR3: SEQ ID NO: 34; and a light chain variable region sequence comprising CDR1: SEQ ID NO: 35, CDR2: SEQ ID NO: 36, and CDR3: SEQ ID NO: 37. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 16 and a light chain variable region sequence set forth in SEQ ID NO: 17. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 18 and a light chain variable region sequence set forth in SEQ ID NO: 19. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 20 and a light chain variable region sequence set forth in SEQ ID NO: 21.

[0140] In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 53. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 54. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 55. In an exemplary embodiment, the second link module of the PD-L1 / TGFβ / CD137 trispecific comprises the sequence set forth in SEQ ID NO: 56.

[0141] In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific has two first link modules and two second link modules. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence of an antibody scaffold module and a second link module set forth in SEQ ID NO:38; and the light chain sequence of an antibody scaffold module and a first link module set forth in SEQ ID NO:39. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence of an antibody scaffold module and a second link module set forth in SEQ ID NO:50; and the light chain sequence of an antibody scaffold module and a first link module set forth in SEQ ID NO:39. In an exemplary embodiment, a PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence of an antibody scaffold module and a first link module set forth in SEQ ID NO:51; and the light chain sequence of an antibody scaffold module and a second link module set forth in SEQ ID NO:52.

[0142] In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific has two first link modules and one second link module. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific comprises the heavy chain sequence of the antibody scaffold module and the second link module set forth in SEQ ID NO:41; the light chain sequence of the antibody scaffold module and the first link module set forth in SEQ ID NO:39, wherein the heavy chain sequence of the antibody scaffold module comprises the sequence set forth in SEQ ID NO:42; the light chain sequence of the antibody scaffold module and the first link module comprises the sequence set forth in SEQ ID NO:39. In an exemplary embodiment, the PD-L1 / TGFβ / CD137 trispecific antibody has the following structure from N-terminus to C-terminus: the heavy chain sequence of the antibody scaffold module and the second link module comprises the sequence set forth in SEQ ID NO:43; the light chain sequence of the antibody scaffold module and the first link module comprises the sequence set forth in SEQ ID NO:39, and the heavy chain sequence of the antibody scaffold module comprises the sequence set forth in SEQ ID NO:42; and the light chain sequence of the antibody scaffold module and the first link module comprises the sequence set forth in SEQ ID NO:39.

[0143] In an exemplary embodiment, the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific further comprises a constant region. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific comprises at least one Fc silencing mutation. In an exemplary embodiment, the Fc silencing mutation in the constant region of the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific is L234A L235A or N297A. In an exemplary embodiment, the constant region of the antibody scaffold module of the PD-L1 / TGFβ / CD137 trispecific comprises a knob-in-hole (KiH) mutation.

[0144] According to some embodiments, the disclosure provides PD-L1 / TGFβ / CD137 trispecifics (e.g., 1923Ab7, 1923Ab9, 1923Ab10, 1923Ab17, and 1923Ab19) that 1) block the interaction between PD-L1 and its receptor PD-1; 2) have cross-linking-dependent agonist activity of CD137 signaling; and / or 3) neutralize the immunosuppressive activity of TGFβ.

[0145] In some embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics comprise as an antibody scaffold module an amino acid sequence derived from one of the binding proteins that bind to PD-L1 (e.g., 1923Ab2 or 1923Ab3), as a first link module a TGFβ-binding amino acid sequence derived from the ECD of the human TGFβRII receptor disclosed herein, and as a second link module for CD137 an amino acid sequence derived from one of the binding proteins that bind to CD137 disclosed herein (e.g., 1923Ab4, 1923Ab5, or 1923Ab6).

[0146] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific comprises an antibody scaffold module that disrupts the PD-1 / PD-L1 binding interaction and disinhibits PD-1 / PD-L1 checkpoint-mediated inhibition of T cells. The PD-L1 antibody scaffold module can be in the form of an IgG molecule (e.g., 1923Ab7, 1923Ab9, 1923Ab10, 1923Ab17, 1923Ab19).

[0147] In some embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics further comprise a second binding module of CD137 that blocks CD137 / CD137 ligand interaction and has agonistic activity for CD137 signaling. In alternative embodiments, the second binding module of CD137 may comprise an scFv (e.g., 1923Ab7, 1923Ab9, 1923Ab10, 1923Ab17, and 1923Ab19). In alternative embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics may comprise a second binding module (e.g., a CD137 scFv) fused to the C-terminus of the heavy or light chain of the PD-L1 antibody scaffold module. In alternative embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics may comprise a second binding module (e.g., a CD137 scFv) fused to the N-terminus of the heavy or light chain of the PD-L1 antibody scaffold module.

[0148] In some embodiments, the second link module that binds CD137 contributes monovalent binding to CD137. In alternative embodiments, the second link module that binds CD137 contributes bivalent binding to CD137.

[0149] In some embodiments, the second binding module is a CD137 scFv and may be stabilized with a disulfide bond.

[0150] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific comprises a TGFβ first link module. In some embodiments, the disclosed TGFβ first link module comprises the extracellular domain of the human TGFβRII receptor. The TGFβ first link module functions to neutralize the biological activity of TGFβ present in the tumor microenvironment. In alternative embodiments, the TGFβ first link module comprises a truncated version (e.g., an N-terminal or C-terminal truncation) of the human TNFβRII receptor that is capable of binding to human TGFβ.

[0151] In an alternative embodiment, the disclosed PD-L1 / TGFβ / CD137 trispecific comprises a PD-L1 antibody scaffold module in which a TGFβ first binding module is attached to the C-terminus of the PD-L1 antibody scaffold module or to the N-terminus of the heavy chain or light chain via a linker.

[0152] In some embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics comprise antibody scaffold modules with a molecular design that is symmetric (e.g., 1923Ab7, 1923Ab17, and 1923Ab19). In alternative embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics are characterized by an asymmetric design (e.g., 1923Ab9 and 1923Ab10). Knobs-into-holes (KIH) technology can be applied to direct the heterodimerization of the heavy chains of the asymmetric disclosed PD-L1 / TGFβ / CD137 trispecifics.

[0153] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab7 comprises a PD-L1 antibody scaffold module with two Fabs from 1923Ab3, a human IgG1 Fc with two Fc constant chains with L234A L235A mutations (SEQ ID NO: 61), two second link modules in the form of a CD137 scFv (VH precedes VL) derived from 1923Ab4, each attached separately to the C-terminus of each of the two Fc constant chains (SEQ ID NO: 53), and two TGFβ first link modules as a polypeptide encoding the extracellular domain of TGFβRII (SEQ ID NO: 67), each attached separately to the C-terminus of each light chain in the Fab.

[0154] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab9 comprises a PD-L1 antibody scaffold module with two Fabs from 1923Ab3, a heterodimeric human IgG1 Fc with two Fc constant chains with L234A L235A mutations and knob-in-hole (KiH) mutations (e.g., the Fc constant chains are shown in SEQ ID NOs: 62 and 63), one second link module in the form of a CD137 scFv (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the knob Fc constant chain (SEQ ID NO: 53), and two TGFβ first link modules as a polypeptide encoding the extracellular domain of TGFβRII (SEQ ID NO: 67), each attached separately to the C-terminus of each light chain in the Fab.

[0155] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab10 comprises a PD-L1 antibody scaffold module with two Fabs from 1923Ab3, a heterodimeric human IgG1 Fc having two Fc chains with L234A L235A mutations and knob-in-hole (KiH) mutations (e.g., the Fc constant chains are shown in SEQ ID NOs: 62 and 63), one second link module in the form of a CD137 scFv (VL precedes VH) derived from 1923Ab4 attached to the C-terminus of the knob Fc constant chain (SEQ ID NO: 54), and two TGFβ first link modules as a polypeptide encoding the extracellular domain of TGFβRII (SEQ ID NO: 67), each attached separately to the C-terminus of each light chain in the Fab.

[0156] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab17 comprises a PD-L1 antibody scaffold module with two Fabs from 1923Ab3, a human IgG1 Fc with two Fc chains with L234A L235A mutations (SEQ ID NO: 61), two second link modules in the form of a disulfide bond stabilized CD137 scFv (VH precedes VL) derived from 1923Ab4, each attached separately to the C-terminus of each of the Fc chains (SEQ ID NO: 55), and two TGFβ first link modules as a polypeptide encoding the extracellular domain of TGFβRII, each attached separately to the C-terminus of each light chain in the Fab (SEQ ID NO: 67).

[0157] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab19 comprises a PD-L1 antibody scaffold module with two Fabs from 1923Ab3, a human IgG1 Fc with two Fc chains with L234A L235A mutations (SEQ ID NO: 61), two TGFβ first binding modules as polypeptides encoding the extracellular domain of TGFβRII (SEQ ID NO: 67), each attached separately to the C-terminus of each light chain in the Fab, and two CD137 second binding modules disulfide bond stabilized scFvs (VH precedes VL) (SEQ ID NO: 56) derived from 1923Ab4, each attached separately to the C-terminus of each of the Fc chains.

[0158] In some embodiments, a PD-L1 / TGFβ / CD137 trispecific comprises a PD-L1 antibody scaffold module derived from a heavy and / or light chain sequence disclosed in Table 7. In alternative embodiments, a PD-L1 / TGFβ / CD137 trispecific comprises a combination of two heavy and light chain sequences paired according to Table 7.

[0159] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific comprises heavy and light chain sequences selected from the following combinations: (a) a heavy chain sequence comprising SEQ ID NO:38 and a light chain sequence comprising SEQ ID NO:39; (b) a heavy chain sequence comprising SEQ ID NO:48 and a light chain sequence comprising SEQ ID NO:49; (c) a heavy chain sequence comprising SEQ ID NO:50 and a light chain sequence comprising SEQ ID NO:39; and (d) a heavy chain sequence comprising SEQ ID NO:51 and a light chain sequence comprising SEQ ID NO:52.

[0160] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific comprises two different variable heavy chain sequences (designed to heterodimerize using a knobs-into-holes format) and variable light chain sequences selected from the following combinations: (a) a first heavy chain sequence comprising SEQ ID NO:41, a second heavy chain sequence comprising SEQ ID NO:42, and a light chain sequence comprising SEQ ID NO:39; and (b) a first heavy chain sequence comprising SEQ ID NO:43, a second heavy chain sequence comprising SEQ ID NO:42, and a light chain sequence comprising SEQ ID NO:39.

[0161] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific comprises a heavy chain sequence selected from the group consisting of SEQ ID NOs: 38, 41, 42, 43, 48, 50, and 51, or an analog or derivative thereof having at least 90% sequence identity to SEQ ID NO: 38, 41, 42, 43, 48, 50, or 51.

[0162] In another embodiment, the PD-L1 / TGFβ / CD137 trispecific comprises a light chain sequence selected from the group consisting of SEQ ID NOs: 39, 49, and 52, or an analog or derivative thereof having at least 90% sequence identity to SEQ ID NO: 39, 49, or 52.

[0163] In some embodiments, the disclosed PD-L1 / TGFβ / CD137 trispecifics further comprise an engineered Fc region to eliminate / minimize cross-linking activity with Fc γR, which silences or eliminates Fc-mediated effector function of T cells.

[0164] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific exhibits one or more of the following functional characteristics, alone or in combination: (a) is capable of binding to human PD-L1, CD137 and TGFβ; (b) cross-reacts with cynomolgus monkey PD-L1 and CD137; (c) disrupting (e.g., reducing or preventing) the interaction of PD-1 and PD-L1; (d) disrupts (e.g., reduces or prevents) human CD137L binding to CD137; (e) exhibits fast on and fast off properties for CD137; (f) derepressing T cell PD-L1-mediated checkpoint inhibitory signals; (g) inhibiting TGFβ signaling and neutralizing its biological activity; (h) has PD-L1-dependent agonistic activity on CD137 signaling; (i) activates T cells in a PD-L1-dependent manner; and (j) Killing PD-L1-expressing tumor cells by activating CD8 T cells.

[0165] The present disclosure also provides pharmaceutical compositions comprising, or alternatively consisting of, at least one of the PD-L1 / TGFβ / CD137 trispecifics disclosed herein, and optionally a pharma- ceutically acceptable diluent, carrier, vehicle and / or excipient. Such pharmaceutical compositions may be used for the treatment of cancer.

[0166] The present disclosure also relates to a method for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the PD-L1 / TGFβ / CD137 trispecifics disclosed herein, alone or in combination with another therapeutic agent.

[0167] The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / TGFβ / CD137 trispecific sequences described herein.The present disclosure also provides isolated polynucleotide sequences encoding at least one of the PD-L1 / TGFβ / CD137 trispecific sequences described herein.

[0168] The present disclosure also provides vectors comprising the PD-L1 / TGFβ / CD137 trispecific polynucleotides described herein.

[0169] The present disclosure also provides vectors comprising at least one of the PD-L1 / TGFβ / CD137 trispecific polynucleotide sequences described herein.

[0170] The present disclosure also provides a cell comprising one of the PD-L1 / TGFβ / CD137 trispecific polynucleotide sequences described herein, or one of the above vectors.

[0171] The present disclosure also provides binding proteins that bind CD137, TGFβ, and PD-L1, comprising: (a) an antibody scaffold module in an IgG format comprising a first antigen binding site that binds CD137 and a second antigen binding site that binds CD137; (b) at least one first binding module comprising a third antigen binding site that binds TGFβ; and (c) at least one second binding module comprising a fourth antigen binding site that binds PD-L1. In some embodiments, the present disclosure also provides CD137 / TGFβ / PD-L1 trispecifics constructed in the form of a recombinant protein comprising an antibody scaffold module that binds CD137, a first binding module comprising an amino acid sequence derived from a TGFβ receptor II binding protein capable of binding to and neutralizing the activity of human TGFβ, and a second binding module that binds PD-L1.

[0172] In an exemplary embodiment, the first and second antigen-binding sites of the CD137 / TGFβ / PD-L1 trispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and a light chain variable region sequence comprising CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26. In an exemplary embodiment, the first and second antigen-binding sites of the CD137 / TGFβ / PD-L1 trispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:27, CDR2: SEQ ID NO:28, and CDR3: SEQ ID NO:29; and a light chain variable region sequence comprising CDR1: SEQ ID NO:30, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:31. In an exemplary embodiment, the first antigen-binding site and the second antigen-binding site of the CD137 / TGFβ / PD-L1 trispecific comprise a heavy chain variable region sequence comprising CDR1: SEQ ID NO:32, CDR2: SEQ ID NO:33, and CDR3: SEQ ID NO:34; and a light chain variable region sequence comprising CDR1: SEQ ID NO:35, CDR2: SEQ ID NO:36, and CDR3: SEQ ID NO:37.

[0173] In exemplary embodiments, the antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 16, SEQ ID NO: 18, or SEQ ID NO: 20. In exemplary embodiments, the antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a light chain variable region sequence set forth in SEQ ID NO: 17, SEQ ID NO: 19, or SEQ ID NO:21.

[0174] In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 16 and a light chain variable region sequence set forth in SEQ ID NO: 17. In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 18 and a light chain variable region sequence set forth in SEQ ID NO: 19. In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 20 and a light chain variable region sequence set forth in SEQ ID NO: 21.

[0175] In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO: 16, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:17, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66. In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO:18, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:19, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66. In an exemplary embodiment, the antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain variable region sequence set forth in SEQ ID NO:20, a heavy chain constant region sequence set forth in SEQ ID NO:60, SEQ ID NO:61, SEQ ID NO:62, SEQ ID NO:63, or SEQ ID NO:64, a light chain variable region sequence set forth in SEQ ID NO:21, and a light chain constant region sequence set forth in SEQ ID NO:65 or SEQ ID NO:66.

[0176] In an exemplary embodiment, the antibody scaffold portion of the CD137 / TGFβ / PD-L1 trispecific comprises the heavy chain sequence shown in SEQ ID NO:75 and the light chain sequence shown in SEQ ID NO:76.

[0177] In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has one first link module. In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has two first link modules. In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has an antibody scaffold module comprising a heavy chain sequence comprising a C-terminus and an N-terminus, where the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the first link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the first link module and the antibody scaffold module are covalently attached to each other directly or through a first link module interlinker. In an exemplary embodiment, the first link module and antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific are covalently attached to each other through a first link module interlinker, the first link module interlinker being shown in SEQ ID NO: 58. In an exemplary embodiment, the first link module and antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific are covalently attached to each other through a first link module interlinker being shown in SEQ ID NO: 59.

[0178] In an exemplary embodiment, the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, when there is more than one first link module in the CD137 / TGFβ / PD-L1 trispecific, each is covalently attached to a different antibody scaffold module sequence or to a different end of the antibody scaffold module sequence.

[0179] In an exemplary embodiment, the first binding module in the CD137 / TGFβ / PD-L1 trispecific comprises the extracellular domain of TGFβRII. In an exemplary embodiment, the extracellular domain of TGFβRII in the CD137 / TGFβ / PD-L1 trispecific comprises the sequence set forth in SEQ ID NO:67.

[0180] In an exemplary embodiment, the second binding module in a CD137 / TGFβ / PD-L1 trispecific is an scFv that comprises a heavy chain variable region sequence and a light chain variable sequence, where the sequences are covalently attached to one another either directly or through an scFv fusion linker. In an exemplary embodiment, the scFv fusion linker comprises glycine and serine. In an exemplary embodiment, the scFv fusion linker comprises the sequence Gly-Gly-Gly-Ser. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker is the sequence set forth in SEQ ID NO:58. In an exemplary embodiment, the scFv fusion linker comprises the sequence set forth in SEQ ID NO:59. In an exemplary CD137 / TGFβ / PD-L1 trispecific embodiment in which a CD137 antibody provides the scaffold moiety, the second binding module is an scFv that binds PD-L1 comprising SEQ ID NO:57.

[0181] In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has one second link module. In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has two second link modules. In an exemplary embodiment, an antibody scaffold module of a CD137 / TGFβ / PD-L1 trispecific comprises a heavy chain sequence comprising a C-terminus and an N-terminus, wherein the antibody scaffold module comprises a light chain sequence comprising a C-terminus and an N-terminus, and the second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence, the C-terminus of the antibody scaffold module light chain sequence, the N-terminus of the antibody scaffold module heavy chain sequence, the N-terminus of the antibody scaffold module light chain sequence, or a combination thereof, and optionally the second link module and the antibody scaffold module are covalently attached to each other directly or through a second link module interlinker.

[0182] In an exemplary embodiment, the second link module and the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific are covalently attached to each other through a second link module interlinker and a second link module interlinker set forth in SEQ ID NO: 58. In an exemplary embodiment, the second link module and the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific are covalently attached to each other through a second link module interlinker set forth in SEQ ID NO: 59.

[0183] In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence and the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence.

[0184] In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module light chain sequence and the first link module of the CD137 / TGFβ / PD-L1 trispecific is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence.

[0185] In an exemplary embodiment, when there is more than one second link module in a CD137 / TGFβ / PD-L1 trispecific, each is covalently attached to a different antibody scaffold module sequence or to a different end of an antibody scaffold module. In an exemplary embodiment, a CD137 / TGFβ / PD-L1 trispecific has one second link module, and one second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence.

[0186] In an exemplary embodiment, the CD137 / TGFβ / PD-L1 trispecific has one second link module, where one second link module is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the CD137 / TGFβ / PD-L1 trispecific has two second link modules, where one second link module is covalently attached to the C-terminus of the antibody scaffold module heavy chain sequence and the other second link module is covalently attached to the C-terminus of the antibody scaffold module light chain sequence. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific is an scFv.

[0187] In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence comprising: CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and a light chain variable region sequence comprising: CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO: 10. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence comprising: CDR1: SEQ ID NO:11, CDR2: SEQ ID NO:12, and CDR3: SEQ ID NO:13; and a light chain variable region sequence comprising: CDR1: SEQ ID NO:14, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:15.

[0188] In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence set forth in SEQ ID NO: 1 and a light chain variable region sequence set forth in SEQ ID NO: 2. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific comprises, from N-terminus to C-terminus, a heavy chain variable region sequence set forth in SEQ ID NO: 3 and a light chain variable region sequence set forth in SEQ ID NO: 4. In an exemplary embodiment, the second link module of the CD137 / TGFβ / PD-L1 trispecific comprises the sequence set forth in SEQ ID NO: 57.

[0189] In an exemplary embodiment, the CD137 / TGFβ / PD-L1 trispecific has two first link modules and two second link modules. In an exemplary embodiment, the CD137 / TGFβ / PD-L1 trispecific has a structure in which the heavy chain sequence of the antibody scaffold module and the second link module comprises SEQ ID NO:48; and the light chain sequence of the antibody scaffold module and the first link module comprises SEQ ID NO:49.

[0190] In an exemplary embodiment, the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific further comprises a constant region. In an exemplary embodiment, the constant region of the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific comprises at least one Fc silencing mutation. In an exemplary embodiment, the Fc silencing mutation in the constant region of the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific is L234A L235A or N297A. In an exemplary embodiment, the constant region of the antibody scaffold module of the CD137 / TGFβ / PD-L1 trispecific comprises a knob-in-hole (KiH) mutation.

[0191] According to some embodiments, the disclosure provides a CD137 / TGFβ / PD-L1 trispecific (1923Ab16) that 1) blocks the interaction between CD137 and its ligand; 2) disrupts (e.g., reduces or prevents) the interaction of PD-1 and PD-L1; and 3) neutralizes the immunosuppressive activity of TGFβ.

[0192] In some embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics comprise as an antibody scaffold module an amino acid sequence derived from one of the binding proteins that bind to CD137 (e.g., 1923Ab4, 1923Ab5, or 1923Ab6), as a first link module, a TGFβ-binding amino acid sequence derived from the ECD of the human TGFβRII receptor disclosed herein, and as a second link module for PD-L1 an amino acid sequence derived from one of the binding proteins that bind to PD-L1 disclosed herein (e.g., 1923Ab2, or 1923Ab3).

[0193] In some embodiments, the CD137 / TGFβ / PD-L1 trispecific comprises an antibody scaffold module that blocks the CD137 / CD137 ligand interaction and has agonistic activity for CD137 signaling. The CD137 antibody scaffold module can be in the form of an IgG molecule (e.g., 1923Ab16) or a binding fragment thereof.

[0194] In some embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics further comprise a second binding module of PD-L1 that disrupts the PD-1 / PD-L1 binding interaction and disinhibits PD-1 / PD-L1 checkpoint-mediated inhibition of T cells. In alternative embodiments, the second binding module of PD-L1 may comprise an scFv (e.g., 1923Ab16). In alternative embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics may comprise a second binding module of a PD-L1 scFv fused to the C-terminus of the heavy or light chain of the CD137 antibody scaffold module.

[0195] In some embodiments, the second link module of PD-L1 contributes to monovalent binding to PD-L1.In alternative embodiments, the second link module of PD-L1 contributes to bivalent binding to PD-L1.

[0196] In some embodiments, the second binding module of the scFv of PD-L1 may be stabilized by a disulfide bond.

[0197] In some embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics comprise a tumor microenvironment regulator, exemplified herein as a TGFβ first binding module. In some embodiments, the disclosed TGFβ first binding module comprises the extracellular domain of the human TGFβRII receptor. The TGFβ first binding module functions to neutralize the biological activity of TGFβ present in the tumor microenvironment. In alternative embodiments, the TGFβ first binding module may comprise a truncated version (e.g., an N-terminal or C-terminal truncation) of the human TNFβRII receptor that is capable of binding to human TGFβ.

[0198] In an alternative embodiment, the disclosed CD137 / TGFβ / PD-L1 trispecific comprises a CD137 antibody binding module, in which a first binding module of TGFβ is attached to the C-terminus or to the N-terminus of the heavy or light chain of the CD137 antibody binding module via a linker.

[0199] In some embodiments, the CD137 / TGFβ / PD-L1 trispecific has a molecular design that is symmetric. In alternative embodiments, the CD137 / TGFβ / PD-L1 trispecific is characterized by an asymmetric design. Knobs-into-holes (KIH) technology can be applied to direct the heterodimerization of the heavy chains of the asymmetric disclosed CD137 / TGFβ / PD-L1 trispecifics.

[0200] In some embodiments, the CD137 / TGFβ / PD-L1 trispecific 1923Ab16 comprises a CD137 antibody scaffold module with two Fabs from 1923Ab4, a human IgG1 Fc with two Fc chains with L234A L235A mutations (SEQ ID NO: 61), two PD-L1 second binding modules scFv (VH precedes VL) derived from 1923Ab3 (SEQ ID NO: 57), each attached separately to the C-terminus of each of the Fc chains, and two TGFβ first binding modules as polypeptides encoding the extracellular domain of TGFβRII (SEQ ID NO: 67), each attached separately to the C-terminus of each light chain in the Fab.

[0201] In some embodiments, the disclosed CD137 / TGFβ / PD-L1 trispecifics further comprise an engineered Fc region to eliminate / minimize cross-linking activity with Fc γR, which silences or eliminates Fc-mediated effector function of T cells.

[0202] In some embodiments, the CD137 / TGFβ / PD-L1 trispecific exhibits one or more of the following functional characteristics, alone or in combination: (a) is capable of binding to human PD-L1, CD137 and TGFβ; (b) cross-reacts with cynomolgus monkey PD-L1 and CD137; (c) disrupting (e.g., reducing or preventing) the interaction of PD-1 and PD-L1; (d) disrupting (e.g., reducing or preventing) human CD137L binding to CD137; (e) exhibits fast-on and fast-off properties for CD137; (f) derepressing T cell PD-L1-mediated checkpoint inhibitory signals; (g) inhibiting TGFβ signaling and neutralizing its biological activity; (h) has PD-L1-dependent agonistic activity on CD137 signaling; (i) activates T cells in a PD-L1-dependent manner; and (j) Killing PD-L1-expressing tumor cells by activating CD8 T cells.

[0203] The present disclosure also provides pharmaceutical compositions comprising, or alternatively consisting of, at least one CD137 / TGFβ / PD-L1 trispecific disclosed herein, and optionally a pharma- ceutically acceptable diluent, carrier, vehicle and / or excipient. Such pharmaceutical compositions may be used for antibody-based immunotherapy of cancer.

[0204] The present disclosure also relates to a method for treating cancer in a patient comprising administering to the patient a therapeutically effective amount of at least one of the CD137 / TGFβ / PD-L1 trispecifics disclosed herein, alone or in combination with another therapeutic agent.

[0205] The present disclosure also provides isolated polynucleotide sequences encoding at least one of the CD137 / TGFβ / PD-L1 trispecific sequences described herein.The present disclosure also provides isolated polynucleotide sequences encoding at least one of the CD137 / TGFβ / PD-L1 trispecific sequences described herein.

[0206] The present disclosure also provides vectors comprising the CD137 / TGFβ / PD-L1 trispecific polynucleotides described herein.

[0207] The present disclosure also provides vectors comprising at least one of the CD137 / TGFβ / PD-L1 trispecific polynucleotide sequences described herein.

[0208] The present disclosure also provides a cell comprising one of the CD137 / TGFβ / PD-L1 trispecific polynucleotide sequences described herein, or one of the above vectors. [Brief description of the drawings]

[0209] The foregoing summary, as well as the following detailed description of the present disclosure, will be better understood when read in conjunction with the accompanying drawings. For the purpose of illustrating the present disclosure, there are shown in the drawings embodiments which are presently preferred. It should be understood, however, that the present disclosure is not limited to the precise arrangements, examples, and instrumentalities shown.

[0210] [Figure 1A] Figures 1A-K provide the amino acid sequences of the VH and VL domains of human binding proteins that bind PD-L1 or bind CD137, the HC and LC sequences of the PD-L1 / CD137 bispecific, the PD-L1 / TGFβ bispecific, and the PD-L1 / CD137 / TGFβ trispecific, and the scFv subunits used to prepare the disclosed trispecifics. The CDR sequences (Kabat numbering) of anti-PD-L1 and anti-CD137 are underlined in their respective variable domain sequences. Sequence identifiers are provided. [Figure 1B] Same as above. [Figure 1C] Same as above. [Figure 1D] Same as above. [Figure 1E] Same as above. [Figure 1F] Same as above. [Figure 1G] Same as above. [Figure 1H] Same as above. [Figure 1I] Same as above. [Figure 1J] Same as above. [Figure 1K] Same as above.

[0211] [Diagram 2] Figures 2A-B show the binding activity of PD-L1 monospecifics, A) 1923Ab2 and B) 1923Ab3, to human, mouse and cynomolgus PD-L1 proteins by ELISA.

[0212] [Diagram 3]Figure 3 shows the binding activity of PD-L1 monospecifics in human PD-L1-expressing HEK293T by imaging binding assay.

[0213] [Figure 4] Figures 4A-B show the activity of PD-L1 monospecifics, A) 1923Ab2 and B) 1923Ab3, to block the interaction of PD-L1 and PD-1 by a PD-1 / PD-L1 blockade reporter assay.

[0214] [Figure 5-1] 5A-C show the binding activity of CD137 monospecifics to human, mouse, and cynomolgus CD137-expressing HEK293T cells by image binding assay. [Figure 5-2] Same as above.

[0215] [Figure 6] FIG. 6 shows a comparison of the binding activity of CD137 antibodies to human CD137-expressing HEK293T cells by image binding assay.

[0216] [Figure 7] FIG. 7 shows the activity of CD137 monospecifics to block CD137L binding to CD137 by biolayer interferometry (BLI).

[0217] [Figure 8] FIG. 8 shows the effect of cross-linking CD137 monospecifics on HEK293T CD137 reporter cells expressing human CD137 and an NFκB luciferase reporter.

[0218] [Figure 9] FIG. 9 shows the effect of cross-linking CD137 monospecifics on human PBMC stimulated with anti-CD3 to induce IFNγ secretion.

[0219] [Figure 10-1]Figure 10A shows a schematic diagram of four human / mouse hybrid CD137 expression constructs. The human CRD regions were replaced by their mouse counterparts and transiently expressed on HEK293T cells. Figures 10B-F show the binding activity of Urelumab-NR (PC2), Urelumab-NR (PC3) and 1923Ab4 to human CD137 wild type (10B) and human / mouse hybrid CD137 proteins msCRD1 (10C), msCRD2 (10D), msCRD3 (10E) and msCRD4 (10F). [Figure 10-2] Same as above. [Figure 10-3] Same as above.

[0220] [Figure 11-1] Figure 11A shows a sequence alignment of the CRD4 domain of human and mouse CD137. Five expression constructs of human CD137 were generated by changing the human amino acid sequence to the mouse amino acid sequence as indicated by M1-M5 and transiently expressing them on HEK293T cells. Figures 11B-G show the binding activity of urelumab-NR (PC2) and 1923Ab4 to human CD137 WT (11B) and five mutant human CD137 proteins M1 (11C), M2 (11D), M3 (11E), M4 (11F) and M5 (11G). [Figure 11-2] Same as above. [Figure 11-3] Same as above.

[0221] [Figure 12] Figure 12 shows the epitope region of 1923Ab4 on human CD137 (depicted as a shaded bar) identified by HDX-MS. The boxed region defines the CD137 cysteine-rich domain (CRD) region.

[0222] [Figure 13-1]Figures 13A-13L illustrate the structural features of the disclosed bispecifics and trispecifics: 1923Ab7 (A); 1923Ab8 (B), 1923Ab9 (C), 1923Ab10 (D), 1923Ab11 (E), 1923Ab12 (F), 1923Ab13 (G), 1923Ab16 (H), 1923Ab17 (I), 1923Ab18 (J), 1923Ab19 (K) and 1923Ab20 (L). [Figure 13-2] Same as above. [Figure 13-3] Same as above.

[0223] [Figure 14] Figures 14A-14B depict the heavy and light chains that comprise the disclosed bispecific (A) and the disclosed trispecific (B).

[0224] [Figure 15] FIG. 15 provides a detailed description of the structural and functional subcomponents of the antibody scaffold module and binding module used to construct the binding proteins.

[0225] [Figure 16] 16A-16B show the binding activity of bispecific and trispecific antibodies in human CD137-expressing HEK293T by image binding assay (A) and flow cytometry (B).

[0226] [Figure 17] Figures 17A-17B show the agonistic activity of bispecifics and trispecifics on CD137 signaling using HEK293T CD137 reporter cells (A) and Jurkat T CD137 reporter cells (B).

[0227] [Figure 18A]Figures 18A-18B show target cell-dependent activation of CD137 signaling by bispecifics and trispecifics using Jurkat T CD137 reporter cells in the presence of target cells (A) or in the absence of target cells (B). [Figure 18B] Same as above.

[0228] [Figure 19] 19A-19B show target cell-dependent activation of CD137 signaling for (A) trispecifics and (B) bispecifics generated by fusing the scFv of CD137 with different regions of the antibody. CD137 signaling activity was assessed using Jurkat T cell CD137 reporter cells.

[0229] [Figure 20] Figure 20 shows the activity of bispecifics and trispecifics to block the interaction of PD-L1 and PD-1 by PD-1 / PD-L1 blockade reporter assay.

[0230] [Figure 21] FIG. 21 shows the inhibitory activity of the trispecific to block TGFβ-induced signaling by a TGFβ blockade reporter assay.

[0231] [Figure 22] 22A-22B show the effect of bispecifics and trispecifics on human PBMC stimulated with anti-CD3 to induce IFNγ secretion.

[0232] [Figure 23] Figures 23A-23B show (A) T cell-mediated killing activity and (B) induction of IFNγ secretion of the bispecifics and trispecifics on human CD8 T cells co-cultured with NUGC4 tumor cells expressing endogenous PD-L1.

[0233] [Figure 24]Figure 24 shows the antigen-specific T cell activation activity of bispecifics and trispecifics on human PBMC stimulated with CMV lysate in a CMV recall assay. IFNγ secretion levels were measured as an index of T cell activation.

[0234] [Diagram 25] Figure 25 shows tumor growth in MC38-h-PD-L1 tumor-bearing hCD137 and hPD-L1 double knock-in mice upon treatment with 1923Ab18 or vehicle control. One-way ANOVA of tumor sizes from the different treatment groups on day 21 is plotted.

[0235] [Figure 26-1] Figures 26A-26E show analysis of tumor-infiltrating lymphocytes from MC38-h-PD-L1 tumor-bearing hCD137 and hDP-L1 double knock-in mice upon treatment with 1923Ab18 or vehicle control. The percentages of (A) CD3+CD45+, (B) CD4+CD3+, (C) CD8+CD3+, (D) Tregs among CD3+, and (E) CD8 / Treg ratios are summarized. [Figure 26-2] Same as above. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0236] PD-1 and its ligands, programmed death ligand-1 and programmed death ligand-2 (PD-L1 and PD-L2), act as corepressors that regulate the balance between T cell activation, tolerance, and immunopathology. Targeting the PD-1 / PD-L1 signaling axis is an area of ​​significant therapeutic discovery. The present disclosure provides binding proteins that bind to PD-L1, including binding proteins that bind to PD-L1 (PD-L1 monospecifics), binding proteins that bind to PD-L1 and CD137 (PD-L1 / CD137 bispecifics), binding proteins that bind to PD-L1 and TGFβ (PD-L1 / TGFβ bispecifics), and binding proteins that bind to PD-L1, TGFβ, and CD137 (PD-L1 / TGFβ / CD137 trispecifics), that can be used for the treatment of cancer. Advantageously, the binding proteins disclosed herein allow for inhibition of PD-L1, resulting in lower dose formulations, less frequent and / or more effective dosing, resulting in reduced costs and increased efficiency. In order that this disclosure may be more readily understood, certain technical and scientific terms are specifically defined below. Unless specifically defined elsewhere in this document, all other technical and scientific terms used herein have the meanings commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0237] The following abbreviations are used throughout this disclosure: mAb or Mab or MAb- Monoclonal Antibody. CDR--Complementarity determining region in an immunoglobulin variable region. VH or VH-immunoglobulin heavy chain variable region. VL or VL- immunoglobulin light chain variable region. Fc or Fc region - the constant region of an immunoglobulin heavy chain comprising the CH2 and CH3 domains and part of the hinge region. Fab or Fab fragment - a monovalent antigen-binding fragment of an antibody consisting of the VH, CH1 and VL, CL domains. FR - antibody framework region, immunoglobulin variable region excluding CDR regions

[0238] As used herein, the term "PD-L1" includes variants, isoforms, homologs, orthologs, and paralogs. For example, an antibody specific for human PD-L1 protein may, in certain cases, cross-react with PD-L1 protein from species other than human. In other embodiments, an antibody specific for human PD-L1 protein may be specific for human PD-L1 protein, but may exhibit species or other type cross-reactivity, or may cross-react with PD-L1 from certain other species, but not all other species (e.g., cross-react with monkey PD-L1, but not with mouse PD-L1). The term "human PD-L1" refers to human sequence PD-L1, such as the complete amino acid sequence of human PD-L1 having NCBI Accession Number NP_054862. PD-L1 is a member of the B7 protein family and shares approximately 20% amino acid sequence identity with B7.1 and B7.2. Human PD-L1 shares 70% and 93% amino acid sequence identity with the mouse and cynomolgus monkey PD-L1 orthologues, respectively.

[0239] As used herein, the terms "PD-1", "PD1", "programmed cell death protein 1", "CD279", and "cluster of differentiation 279" (e.g., Genebank Accession No. NP_005009 (human)) refer to a type I membrane protein that is a member of the extended CD28 / CTLA-4 family of T cell regulators. PD1 contains an extracellular IgV domain followed by a transmembrane region and an intracellular tail PD1 is expressed on the surface of activated T cells, B cells, and macrophages.

[0240] The term "CD137" refers to 4-1BB or TNFRSF9 (TNF receptor superfamily member 9), a member of the TNF receptor superfamily (TNFRSF) and a costimulatory molecule expressed after activation of immune cells, both innate and adaptive immune cells. As used herein, 4-1BB may be derived from a mammal, for example, Homo sapiens (human) (NCBI accession number NP_001552). As described herein, the term CD137 includes variants, isoforms, homologs, orthologs, and paralogs. For example, an antibody specific for human CD137 protein may cross-react with CD137 protein from species other than human in certain cases. In other embodiments, an antibody specific for human CD137 protein may be completely specific for human CD-137 protein, but may exhibit species or other type cross-reactivity, or may cross-react with CD137 from certain other species, but not all other species (e.g., cross-react with monkey CD137, but not with mouse 4-1BB). The term "cyno CD137" refers to cynomolgus monkey CD137, such as the complete amino acid sequence having NCBI accession number XP_005544945.1. The term "mouse CD137" refers to the mouse sequence 4-1BB, such as the complete amino acid sequence of mouse 4-1BB, having NCBI accession number NP_035742.1. The human CD137 sequence in the present disclosure may differ from human CD137 in NCBI accession number NP_001552, for example, by having conserved mutations or mutations in non-conserved regions, but CD137 in the present disclosure has substantially the same biological function as human CD137 in NCBI accession number NP_001552.

[0241] As used herein, the term "transforming growth factor beta", "TGF-beta", or "TGFβ" can refer to any TGF-beta protein, including, but not limited to, TGF-beta 1, TGF-beta 2, and TGF-beta 3, including natural TGF-beta proteins and synthetic proteins, including variants and mimetics. TGF-beta proteins are members of a superfamily of structurally similar regulatory proteins, including, but not limited to, mammalian TGF-beta-1, 2, and 3, inhibins, activins, and bone morphogenetic proteins. Mature TGF-beta typically exists as a homodimer, such as a dimeric mature TGF-beta molecule comprising two covalently associated TGF-beta molecules.

[0242] As used herein, the term "TGFβ receptor II" ("TGFβRII") refers to a polypeptide having a wild-type human TGFβ receptor type 2 isoform A or isoform B sequence, or a portion thereof that binds TGFβ (e.g., the amino acid sequence of NCBI Reference Sequence (RefSeq) Accession No. NP_001020018 or NP_003233.4, respectively), such as SEQ ID NO: 74, or a sequence substantially identical to the amino acid sequence of SEQ ID NO: 74. TGFβRII can retain at least 0.1%, 0.5%, 1%, 5%, 10%, 25%, 35%, 50%, 75%, 90%, 95%, or 99% of the TGFβ binding activity of the wild-type sequence. The expressed TGFβRII polypeptide lacks a signal sequence.

[0243] The term "antibody" herein is used in the broadest sense and encompasses a variety of antibody structures, including, but not limited to, monoclonal antibodies, polyclonal antibodies, bispecific antibodies, and trispecific antibodies.

[0244] The term "antibody scaffold module" herein refers to a Y-shaped antibody having two heavy chains and two light chains. The two heavy chains are linked to each other by disulfide bonds, and each heavy chain is linked to a light chain by a disulfide bond. An antibody scaffold may have one or more binding modules attached to one or more of its heavy and / or light chains. The antibody binding scaffold includes two Fab and an Fc portion having two constant region sequences.

[0245] An exemplary antibody, such as an IgG, comprises two heavy chains and two light chains. Each heavy chain is composed of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. Each light chain is composed of a light chain variable region (abbreviated herein as VL) and a light chain constant region. The VH and VL regions can be further subdivided into regions of hypervariability called complementarity determining regions (CDRs), interspersed with more conserved regions called framework regions (FRs). Each VH and VL is composed of three CDRs and four FRs, arranged from amino terminus to carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4.

[0246] The term "monoclonal antibody" or "mAb" as used herein refers to an antibody obtained from a population of substantially homogeneous antibodies, e.g., the individual antibodies comprising the population are identical and / or bind the same epitope, except for possible variant antibodies, e.g., including naturally occurring mutations or arising during production and / or storage of the monoclonal antibody preparation. In contrast to polyclonal antibody preparations, which typically include different antibodies directed against different determinants (epitopes), each monoclonal antibody of a monoclonal antibody preparation is directed against a single determinant on an antigen. Thus, the modifier "monoclonal" indicates the character of the antibody as being obtained from a substantially homogeneous population of antibodies and should not be construed as requiring production of the antibody by any method. For example, monoclonal antibodies used in accordance with the present disclosure may be produced by a variety of techniques, including, but not limited to, hybridoma methods, recombinant DNA methods, phage display methods, and methods utilizing transgenic animals containing all or a portion of the human immunoglobulin loci, such methods and other exemplary methods for producing monoclonal antibodies described herein.

[0247] The term "chimeric antibody" refers to a recombinant antibody in which a portion of the heavy and / or light chain is identical or homologous to the corresponding sequence in an antibody derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain is identical or homologous to the corresponding sequence in an antibody derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, so long as they exhibit the desired biological activity. Complementarity determining region (CDR) grafting may also be performed to alter certain properties of the antibody molecule, including affinity or specificity. Typically, the variable domains are obtained from an antibody from a laboratory animal, such as a rodent (parent antibody), and the constant domain sequences are obtained from a human antibody, such that the resulting chimeric antibody is capable of directing effector functions in a human subject and is less likely to provoke an adverse immune response than the parent (e.g., murine) antibody from which it is derived.

[0248] A "human antibody" is an antibody having an amino acid sequence corresponding to that of an antibody produced by a human and / or has been produced using any of the techniques for producing human antibodies known to those of skill in the art. This definition of a human antibody specifically excludes humanized antibodies that contain non-human antigen-binding residues. Human antibodies can be produced using a variety of techniques known in the art, including those described in Cole et al, Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, p. 77 (1985); Boerner et al, J. Immunol, 147(I):86-95 (1991). See also van Dijk and van de Winkel, Curr. Opin. Pharmacol, 5:368-74 (2001). Human antibodies can be produced by transgenic animals that have been engineered to produce such antibodies in response to antigenic challenge, but in which the endogenous locus has been disabled, e.g., immunized HuMab mice (see, e.g., Nils Lonberg et al., 1994, Nature 2000, 14:131-135, 2000, for HuMab mice). 368:856-859, WO98 / 24884, WO94 / 25585, WO93 / 1227, WO92 / 22645, WO92 / 03918, and WO01 / 09187), Xenomice (see, e.g., U.S. Pat. Nos. 6,075,181 and 6,150,584 for XENOMOUSE™ technology), or Trianni mice (see, e.g., WO2013 / 063391, WO2017 / 035252, and WO2017 / 136734) with the target antigen.

[0249] The term "humanized antibody" refers to an antibody that has been engineered to contain one or more human framework regions in the variable regions together with non-human (e.g., mouse, rat, or hamster) complementarity determining regions (CDRs) of the heavy and / or light chains. In certain embodiments, a humanized antibody contains sequences that are completely human except for the CDR regions. Humanized antibodies are typically less immunogenic in humans than non-humanized antibodies, and thus provide therapeutic benefits in certain situations. Those of skill in the art will be aware of humanized antibodies as well as appropriate techniques for their production. For example, Hwang,WYK,et al.,Methods 36:35,2005;Queen et al.,Proc.Natl.Acad.Sci.USA,86:10029-10033,1989;Jones et al.,Nature,321:522-25,1986;Riechmann et al. al.,Nature,332:323-27,1988;Verhoeyen et al.,Science,239:1534-36,1988;Orlandi et al. See, e.g., et al., Proc. Natl. Acad. Sci. USA, 86:3833-37, 1989; U.S. Patent Nos. 5,225,539; 5,530,101; 5,585,089; 5,693,761; 5,693,762; 6,180,370; and Selick et al., WO 90 / 07861, each of which is incorporated herein by reference in its entirety.

[0250] The "class" of an antibody refers to the type of constant domain or constant region carried by its heavy chain. There are five major classes of antibodies: IgA, IgD, IgE, IgG, and IgM, several of which can be further divided into subclasses (isotypes), e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2. The heavy chain constant domains that correspond to the different classes of immunoglobulins are called α, δ, ε, γ, and μ, respectively.

[0251] The term "antigen-binding domain" (or simply "binding domain") of an antibody or similar terms refers to one or more fragments of an antibody that retain the ability to specifically bind to an antigen complex. Examples of binding fragments encompassed within the term "antigen-binding portion" of an antibody include: (i) a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH1 domains; (ii) a F(ab')2 fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CH domains; (iv) a Fv fragment consisting of the VL and VH domains of a single arm of an antibody; (v) a dAb fragment consisting of the VH domain (Ward et al., (1989) Nature 341:544-546); (vi) an isolated complementarity determining region (CDR), and (vii) a combination of two or more isolated CDRs, optionally linked by a synthetic linker.

[0252] "Complementarity determining region" or "CDR", as the term is used herein, refers to short polypeptide sequences within the variable regions of both heavy and light chain polypeptides that are primarily responsible for mediating specific antigen recognition. There are three CDRs (called CDR1, CDR2, and CDR3) within each VL and each VH. Unless otherwise noted herein, CDRs and framework regions are annotated according to the Kabat numbering scheme (Kabat EA, et al., 1991, Sequences of proteins of Immunological interest, In: NIH Publication No. 91-3242, USDepartment of Health and Human Services, Bethesda, Md).

[0253] In other embodiments, the CDRs of an antibody can be determined according to MacCallum RM et al., (1996) J Mol Biol 262:732-745, which is incorporated herein by reference in its entirety. In other embodiments, the CDRs of an antibody can be determined according to the AbM numbering scheme, which refers to the AbM hypervariable regions, which represent a compromise between the Kabat CDRs and the Chothia structural loops, and are used by Oxford Molecular's AbM antibody modeling software (Oxford Molecular Group, Inc.), which is incorporated herein by reference in its entirety. CDRs can also be determined according to the AbM numbering scheme, which refers to the AbM hypervariable regions, which represent a compromise between the Kabat CDRs and the Chothia structural loops, and are used by Oxford Molecular's AbM antibody modeling software (Oxford Molecular Group, Inc.), which is incorporated herein by reference in its entirety. thWhile HVLs can be defined by sequence comparison in the Ed. Public Health Service, National Institutes of Health, Bethesda, Md., HVLs are structurally defined according to the three-dimensional structure of the variable domain as described by Chothia and Lesk, 1987, J. Mol. Biol. 196:901-917. If these two methods result in slightly different identifications of the CDRs, the structural definition is preferred. As defined by Kabat, in the light chain variable domain, CDR-L1 is located at about residues 24-34, CDR-L2 at about residues 50-56, and CDR-L3 at about residues 89-97; in the heavy chain variable domain, CDR-H1 is located at about residues 31-35, CDR-H2 at about residues 50-65, and CDR-H3 at about residues 95-102. IMGT and NORTH provide alternative definitions of CDRs (see Lefranc MP. Unique database numbering system for immunogenetic analysis. Immunol Today (1997) 18:509; and North B, Lehmann A, Dunbrack RLJ. A new clustering of antibody CDR loop conformations. J Mol Biol. (2011) 406:228-56). In addition, CDRs can be defined according to Chemical Computing Group (CCG) numbering (Almagro et al., Proteins 2011; 79:3050-3066 and Maier et al, Proteins 2014; 82:1599-1610). CDR1, CDR2, CDR3 thus define the specific, intrinsic and functional properties for a given antibody.

[0254] The "variable domains" (V domains) of antibodies mediate binding and confer antigen specificity of a particular antibody. However, variability is not evenly distributed across the 110 amino acid span of the variable domain. Instead, the V region consists of relatively invariant stretches called framework regions (FR) of 15-30 amino acids separated by shorter regions of extreme variability, referred to herein as "hypervariable regions" or CDRs, each 9-12 amino acids in length. As will be appreciated by those of skill in the art, the exact numbering and arrangement of the CDRs may vary among different numbering systems. However, it should be understood that the disclosure of a variable heavy and / or variable light chain sequence includes the disclosure of the associated CDRs. Thus, the disclosure of each variable heavy chain region is a disclosure of the vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the disclosure of each variable light chain region is a disclosure of the vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3).

[0255] An "Fv" consists of a dimer of one heavy- and one light-chain variable domain in tight, non-covalent association. The folding of these two domains gives rise to six hypervariable loops (three loops each from the H and L chain) that contribute amino acid residues for antigen binding and confer antigen-binding specificity to the antibody.

[0256] A "single-chain variable fragment" or "scFv" is a fragment of an immunoglobulin heavy chain (V H ) and light chain (V L sFv refers to a fusion protein of the variable region of V. For a review of sFv, see Pluckthun in The Pharmacology of Monoclonal Antibodies, vol. 113, Rosenburg and Moore eds., Springer-Verlag, New York, pp. 269-315 (1994). In some embodiments, the regions are connected with a short linker peptide of 10 to about 25 amino acids. The linker can be rich in glycines for flexibility and in serine or threonines for solubility, but V H N-terminus of V Lor vice versa. The protein retains the specificity of the original immunoglobulin despite the removal of the constant region and the introduction of a linker. Disulfide-stabilized scFvs can be engineered by introducing paired cysteines by mutating specific residues in VH or VL. These residues are at the interface of VH and VL. See reference Weatherill, EE et al. Towards a universal disulphide stabilized single chain Fv format: importance of interchain disulphide bond location and vL-vH orientation. Protein Eng Des Sel 25, 321-329. NovaRock used VH44-VL100 in the examples.

[0257] "Framework" or "Framework Region" or "FR" refers to variable domain residues other than hypervariable region (HVR) residues. The FR of a variable domain generally consists of four FR domains: FR1, FR2, FR3, and FR4.

[0258] A "human consensus framework" is a framework that represents the most commonly occurring amino acid residues in a selection of human immunoglobulin VL or VH framework sequences. Generally, the selection of human immunoglobulin VL or VH sequences is from a subgroup of variable domain sequences. Generally, the subgroup of sequences is a subgroup in Kabat et al., Sequences of Proteins of Immunological Interest, Fifth Edition, NIH Publication 91-3242, Bethesda Md. (1991), Vols. 1-3. In one embodiment, for VL, the subgroup is subgroup kappa I, as in Kabat et al. (supra). In one embodiment, for VH, the subgroup is subgroup III, as in Kabat et al. (supra).

[0259] The "hinge region" is generally defined as a stretch of 216-238 (EU numbering) or 226-251 (Kabat numbering) of human IgG1. The hinge can be further divided into three distinct regions, the upper, middle (e.g., core), and lower hinge.

[0260] The terms "Fc region" and "constant region" are used herein to define a C-terminal region of an immunoglobulin heavy chain that comprises at least a portion of the constant region. This term includes native sequence Fc regions and variant Fc regions. In one embodiment, a human IgG heavy chain Fc region extends from Cys226, or from Pro230, to the carboxyl-terminus of the heavy chain. However, the C-terminal lysine (Lys447) of the Fc region may or may not be present. Unless otherwise specified herein, numbering of amino acid residues in an Fc region or constant region is according to the EU numbering system, also referred to as the EU index, as described in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991).

[0261] The term "effector function" derived from the interaction of an antibody Fc region with a specific Fc receptor includes, but is not limited to, Clq binding, complement dependent cytotoxicity (CDC), FcyR-mediated effector functions such as antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent cell-mediated phagocytosis (ADCP), and down-regulation of cell surface receptors. Such effector functions generally require an Fc region in combination with an antigen-binding domain (e.g., an antibody variable domain).

[0262] The term "T-cell dependent cytotoxicity" (TDCC) describes a sequence of events where molecules simultaneously bind to tumor cells, associate with cytotoxic T cells, and redirect cell lysis by bringing the T cells and the target cells in close proximity. Both ADCC and TDCC activation result in the killing of the target cells. However, there are some key differences between these two distinct types of cytotoxicity. The ADCC effect is mediated through Fcγ receptors expressed on natural killer (NK) cells, which bind to the constant regions of antibodies attached on the surface of the target cells. The TDCC effect is mediated through the association of cytotoxic T cells in close proximity to the target cells.

[0263] The term "Fc receptor" or "FcR" describes an antibody receptor that binds to the Fc region of immunoglobulins involved in antigen recognition located on the membrane of certain immune cells including B lymphocytes, natural killer cells, macrophages, neutrophils, and mast cells. Fc receptors that recognize the Fc portion of IgG are called Fc gamma receptors (FcγR). The FcγR family includes allelic variants and alternatively spliced ​​forms of these receptors. Based on differences in structure, function, and affinity for IgG binding, FcγRs are classified into three major groups: FcγRI, FcγRII (FcγRIIa and FcγRIIb), and FcγRIII (FcγRIIIa and FcγRIIIb). Among them, FcγRI (CD64), FcγRIIa (CD32a), and FcγRIIIa (CD16a) are activating receptors that contain a signaling motif, an immunoreceptor tyrosine-based activation motif (ITAM), in the gamma subunit of FcγRI and FcγRIIIa or in the cytoplasmic tail of FcγRIIa. After binding of antigen-antibody complexes, activating Fcγ receptors (human: FcγRI, FcγRIIA, FcγRIIC, FcγRIIIA, FcγRIIIB and mouse: FcγRI, FcγRIII, FcγRIV) induce immune effector functions. In contrast, FcγRIIb (CD32b) is an inhibitory receptor. Cross-linking of FcγRIIb leads to phosphorylation of immunoreceptor tyrosine-based inhibitory motifs (ITIMs) and inhibitory signaling transduction (Patel et al. Front Immunol. 2019;10:223.).

[0264] The term "Fc silencing" refers to an Fc region that has been engineered to minimize / eliminate binding activity with Fc γR and complement, leading to silencing or elimination of Fc-mediated effector functions. Strategies for engineering Fc include modifying Fc glycosylation, using hybrids of IgG subclasses, or introducing one or more mutations in the hinge and / or CH2 regions. Effector function silencing of Fc and the important residues for each mutation are known in the art, for example, Strohl, WR and Strohl LM, "Antibody Fc engineering for optimal antibody performance" In Therapeutic Antibody Engineering, Cambridge: Woodhead Publishing (2012), pp 242, International Patent Application Nos. WO2017008169A1 and WO2021055669.

[0265] Specific, non-limiting examples of sites that can be engineered to silence human IgG1 Fc include L234, L235, G237, D265, N297, P329, P331, all in EU numbering.

[0266] As used herein, the term "T regulatory cells" or "Treg" refers to cells of the immune system that have a regulatory role by suppressing / inhibiting the proliferation, activation, and cytotoxicity of other immune cells, such as CD8 positive (CD8+) effector T cells. Regulatory T cells (Treg) are characterized by the expression of the master transcription factor forkhead box P3 (Foxp3). There are two major subsets of Treg cells: "natural" Treg (nTreg) cells, which arise in the thymus, and "induced" Treg (iTreg) cells, which arise in the periphery from CD4+Foxp3- conventional T cells. Natural Tregs are characterized as expressing both the CD4 T cell coreceptor and CD25, a component of the IL-2 receptor. Tregs are thus CD4+CD25+. Expression of the nuclear transcription factor forkhead box P3 (FoxP3) is a defining characteristic that determines the development and function of natural Tregs. Treg cells exert their suppressive effects through multiple modes of action, including secretion of inhibitory cytokines (e.g., IL-10, TGFβ, IL-35), modulation of dendritic cell function / maturation, expression of immunoregulatory surface molecules (e.g., CTLA-4, LAG-3), or suppression by cytolysis (e.g., granzyme A and / or B mediated).

[0267] As used herein, the term "bispecific" refers to a binding protein comprising an antibody scaffold module and a first link module, where the modules are derived from an antibody and / or receptor protein with binding specificities for two different antigens. In one embodiment, the antibody scaffold module has binding specificity for PD-L1 and the first link module has binding specificity for any other antigen, e.g., for a cell surface protein, receptor, receptor subunit, tissue specific antigen, tumor microenvironment regulator, cytokine, etc. In another embodiment, the antibody scaffold module has binding specificity for CD137 and the first link module has binding specificity for any other antigen, e.g., for a cell surface protein, receptor, receptor subunit, tissue specific antigen, tumor microenvironment regulator, cytokine, etc.

[0268] As used herein, the term "trispecific" refers to a binding protein comprising an antibody scaffold module and a first link module and a second link module, where the modules are derived from an antibody and / or receptor protein with binding specificities for three different antigens. In one embodiment, the antibody scaffold module has binding specificity for PD-L1, and the first and second link modules have binding specificity for any other antigen (apart from the antigens of the other link modules), e.g., cell surface costimulatory receptors (including, but not limited to, CD137), receptors, receptor subunits, tissue specific antigens, tumor microenvironment regulators, cytokines, etc. In one embodiment, the antibody scaffold module has binding specificity for CD137, and the first and second link modules have binding specificity for any other antigen (apart from the antigens of the other link modules), e.g., cell surface costimulatory receptors (including, but not limited to, PD-L1), receptors, receptor subunits, tissue specific antigens, tumor microenvironment regulators, cytokines, etc.

[0269] With respect to the binding of an antibody to a target molecule, the term "specific binding" or "specifically binds to" or "specific for" a particular polypeptide or epitope on a particular polypeptide target means binding that is measurably different from non-specific interactions. Specific binding can be measured, for example, by determining the binding of a molecule compared to the binding of a control molecule. For example, specific binding can be determined by competition with a control molecule that is similar to the target, for example, an excess of unlabeled target. In this case, specific binding is indicated if the binding of the labeled target to the probe is competitively inhibited by an excess of unlabeled target. As used herein, the term "specific binding" or "specifically binds to" or "specific for" to a particular polypeptide or epitope on a particular polypeptide target can be exhibited by a molecule having a Kd of, for example, 10-4M or less, alternatively 10-5M or less, alternatively 10-6M or less, alternatively 10-7M or less, alternatively 10-8M or less, alternatively 10-9M or less, alternatively 10-10M or less, alternatively 10-11M or less, alternatively 10-12M or less, or in the range of 10-4M to 10-6M, or 10-6M to 10-10M, or 10-7M to 10-9M. As will be appreciated by those skilled in the art, affinity and KD values ​​are inversely related. A high affinity for an antigen is measured by a low KD value. In one embodiment, the term "specific binding" refers to binding whereby a molecule binds to a particular polypeptide or an epitope on a particular polypeptide without substantially binding to any other polypeptides or polypeptide epitopes. As used herein, the terms "specific binding," "specifically binds," and "selectively binds" refer to antibody binding to an epitope of CD137, PDL1, and / or TGFb.

[0270] The term "affinity" as used herein refers to the strength of binding of an antibody to an epitope. The affinity of an antibody is given by the dissociation constant Kd, defined as [Ab] x [Ag] / [Ab-Ag], where [Ab-Ag] is the molar concentration of the antibody-antigen complex, [Ab] is the molar concentration of unbound antibody, and [Ag] is the molar concentration of unbound antigen. The affinity constant Ka is defined by 1 / Kd. Methods for determining the affinity of mAbs are described in Harlow, et al., Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1988), Coligan et al., eds., Current Protocols in Immunology, Greene Publishing Assoc. and Wiley Interscience, NY, (1992, 1993), and Muller, Meth. Enzymol. 92:589-601 (1983), which references are incorporated herein by reference in their entireties. One standard method well known in the art for determining the affinity of mAbs is the use of surface plasmon resonance (SPR) screening (e.g., by analysis using a BIAcore™ SPR analyzer).

[0271] "Epitope" is a term of the art that denotes the site or site of interaction between an antibody and its antigen. As described by (Janeway, C, Jr., P. Travers, et al. (2001). Immunobiology: the immune system in health and disease. Part II, Sections 3-8. New York, Garland Publishing, Inc.): "Antibodies generally recognize only small regions on the surface of macromolecules, such as proteins...[a particular epitope] is likely to be composed of amino acids from different parts of the [antigen] polypeptide chain that are brought together by protein folding. This type of antigenic determinant is known as a conformational or discontinuous epitope because the structure recognized is composed of segments of the protein that are discontinuous in the antigen's amino acid sequence but that are brought together in a three-dimensional structure. In contrast, an epitope that is composed of a single segment of a polypeptide chain is referred to as a continuous or linear epitope" (Janeway, C. Jr., P. Travers, et al. (2001). Immunobiology: the immune system in health and disease. Part II, Sections 3-8. New York, Garland Publishing, Inc.) Publishing, Inc.).

[0272] The term "KD", as used herein, is intended to refer to the dissociation constant of a particular antibody-antigen interaction. It is calculated by the following formula: Koff / Kon=KD

[0273] The term “IC 50 " as used herein is intended to refer to the effective concentration of a binding protein disclosed herein required to neutralize 50% of the biological activity of the antigen to which it binds.

[0274] EC for drugs and their specific activities (e.g., binding to cells, inhibition of enzyme activity, activation or inhibition of immune cells) 50" refers to the effective concentration of an agent that produces 50% of its maximum response or effect for such activity. The EC for a drug and a particular activity 100 refers to the effective concentration of an agent that produces substantially its maximal response for such activity.

[0275] As used herein, the term "T cell exhaustion" is defined as the impaired ability of T cells to proliferate and secrete cytokines caused by prolonged antigen stimulation-induced overexpression of immune checkpoint receptors such as PD-1, CTLA-4, T cell Ig and mucin domain-containing (TIM)-3, and lymphocyte activation gene 3.

[0276] As used herein, the term "costimulatory receptor" on T cells refers to cell surface molecules that can positively induce signal transduction to fully activate T cells with TCR signaling and cytokine stimulation. Co-signaling pathways play a critical role in priming and activating T cells, as well as in regulating T cell differentiation, effector function, and survival. Co-stimulatory receptors are generally categorized into two groups: Ig receptor superfamily (IgSF) and TNF receptor superfamily (TNFRSF).

[0277] The term "binding module" refers to any substance that binds to PD-L1, CD137, TGFβ, or any other target, which may enhance the specific activity of the binding protein of the invention in comparison to the scaffold module itself. Non-limiting examples of binding modules include anticalins, lipebodies, monobodies, scFv, Fab, scFab, affibodies, finomers, DARPins, nanobodies, peptide aptamers, and nucleic acid aptamers.

[0278] The term "Fab" or "antigen-binding fragment" refers to two identical fragments of an antibody typically prepared by enzymatic digestion that confer binding specificity to the antibody. Papain digestion of an antibody produces two identical Fab fragments consisting of the entire light (L) chain along with the variable region domain of the heavy (H) chain (VH) and the first constant domain of one heavy chain (CH1). Pepsin treatment of an antibody yields a single large F(ab)2 fragment that roughly corresponds to two disulfide-linked Fab fragments that have bivalent antigen-binding activity and are still capable of cross-linking antigen. Fab fragments differ from Fab' fragments by having additional residues at the carboxy terminus of the CH1 domain including one or more cysteines from the antibody hinge region.

[0279] The term "linker" refers to at least one atom that forms a covalent bond between two chemical entities. The term "linker" may refer to at least one atom that forms a covalent bond between a scaffold module and another covalent bond to a binding module. If the scaffold module and the binding module are linked only through peptide bonds, the linker is referred to as a "peptide linker". Otherwise, the linker is referred to as a "chemical linker". Furthermore, "flexible peptide linkers" contain mostly small, non-polar or polar amino acids, whereas "rigid peptide linkers" contain alpha-helix-forming sequences and / or are rich in proline residues (Chen et al., 2013. Adv Drug Deliv Rev. 65(10):1357-1369).

[0280] The term "scaffold module" refers to a protein that contains two antigen binding sites and can act as a support structure for one or more binding modules. A binding module can be attached to a scaffold module through a linker and / or a binding module can be incorporated into any loop region present in the scaffold module.

[0281] It is noted that as used herein in the specification and the appended claims, the singular forms include plural references unless the context clearly dictates otherwise.

[0282] T cell activation & exhaustion T cell activation The T cell receptor (TCR) on a T cell binds to antigens presented by MHC complexes on the surface of antigen-presenting cells (APCs). TCR / MHC / antigen binding triggers the initial activation of the T cell. T cell signaling through the T cell antigen receptor (TCR) / CD3 complex triggers an array of signals that activate multiple effector pathways.

[0283] T cell activation is regulated, positively and negatively, by both signals generated by the TCR / CD3 complex and signals emanating from other cell surface receptors and / or delivered by soluble mediators to ensure that T cells respond to appropriate ligands for an appropriate period of time. The central idea of ​​T cell activation is that signaling through the TCR alone leads to a state of anergy (a state of hyporesponsiveness of T cells to a particular antigen that can be induced by the lack of costimulation).

[0284] Several surface receptors have been described that are capable of providing costimulation for T cells, including CD28, CD30, CD5, CD2, ICOS, OX40, and 4-1BB (CD137). CD28 is expressed during the induction of an immune response and promotes the expression of several other costimulatory molecules, including ICOS, OX40, and CD137.

[0285] TCR-generated regulatory signals are also complemented by signals from the ligation of other coreceptors, such as cytotoxic T-lymphocyte antigen-4 (CTLA-4) and programmed death-1 (PD-1, also known as CD279), both of which function to limit TCR-triggered T cell proliferation and activation. In light of their function as negative regulators, CTLA-4 and PD-1 are described as immune checkpoints. It is widely known that ligation of PD-1 and the PD-L1 ligand (also known as B7-H1 or CD274), expressed on activated T cells, activates critical immune checkpoints, leading to T cell tolerance, dysfunction and T cell exhaustion.

[0286] Prolonged exposure to tumor antigen peptide / MHC class I complexes in the presence of inflammatory cytokines induces a distinct phenotype in T cells present in solid tumors, characterized by a progressive loss of effector function, high and persistent expression of inhibitory immune checkpoint receptors, poor proliferative capacity (Pauken KE and Wherry EJ.,Trends Immunol 2015;36(4):265-276(2015),Wherry EJ and Kurachi M.,Nat Rev Immunol 2015 15(8):486-499(2015), metabolic dysregulation, poor memory recall, and distinct transcriptional and epigenetic programs (McLane,L et al.,Ann.Rev.Immunol.37:457(2019).

[0287] T cell exhaustion Dysfunctional or chronically stimulated T cells are distinct lineages found in mouse models and humans that arise upon repeated TCR stimulation in cancer (Zajac AJ et al, J. Exp. Med 188, 2205-2213 (1998), Bakhli MY Cytokine, 71, 339-347 (2011), Wherry and Kurachi Nat. Rev. Immunol 15, 486-499 (2015). Prolonged exposure to tumor antigen peptide / MHC class I complexes in the presence of inflammatory cytokines induces distinct phenotypes in T cells present in solid tumors, with progressive loss of effector function, high and persistent expression of inhibitory immune checkpoint receptors, poor proliferative capacity (Pauken KE and Wherry EJ., Trends Immunol 2015;36(4):265-276 (2015), Wherry EJ and Kurachi M., Nat Rev Immunol 2015 15(8):486-499(2015), and are characterized by metabolic dysregulation, poor memory recall, and distinct transcriptional and epigenetic programs (McLane, L et al., Ann. Rev. Immunol. 37:457(2019).

[0288] Negative regulatory signals resulting from activation of immune suppressive checkpoints are believed to be the main mechanism for effector T cell dysfunction in the TME. Tumor-antigen specific responses in cancer research and the presence of concurrent disease have identified PD-1 as a marker for exhaustion. Studies have shown that checkpoint blockade can at least partially reactivate T cell function (Fuertes Marraco SA, et al Front Immunol 6,310(2015). Blockade of PD-1 / PD-L1 interaction has been shown to partially restore T cell function.

[0289] PD-1 / PD-L1 pathway The PD-1 / PD-L1 pathway has been extensively examined (Salmaninejad et al., 2019; Han et al., 2020; Makuku et al., 2021). PD-1 is a 55-kDa transmembrane protein with 288 amino acids that includes an IgV-like extracellular domain followed by a transmembrane region and an intracellular tail with two phosphorylation sites for TCR signaling and regulation (Ishida, Agata, Shibahara, & Honjo). PD-L1 is a type 1 membrane-passing 33-kDa glycoprotein with 290 amino acids with Ig-like and IgC-like domains in its extracellular region (Freeman et al., 2000). PD-1 and PD-L1 are immune checkpoint proteins.

[0290] PD-1 is expressed on various activated immune cells, such as activated T cells, B cells and natural killer (NK), activated monocytes, dendritic cells (DC), macrophages, and immature Langerhans cells. PD-1 is also upregulated on the surface of T cells constantly exposed to antigens and is one of the markers of exhausted T cells (Ahmadzadeh et al., 2009). Transcription factors including NFAT, NOTCH, FOXO1, and IRF9 induce transcription of PD-1 expression (Staron et al., 2014). Cytokines such as IL-2, IL-21, IL-15, IL-7, and type 1 IFN can enhance PD-1 expression. IL-6 and IL-12 increase PD-1 expression in splenic CD8 T cells using signal transducer and activator of transcription 3 (STAT3) and STAT4, respectively (Salmaninejad et al., 2019).

[0291] PD-L1 is constitutively expressed by antigen-presenting cells (APCs) such as macrophages, B cells, DCs, and some epithelial cells, especially under inflammatory conditions (Sharpe et al., 2007). The presence of PD-L1 in peripheral tissues is necessary to prevent autoimmune damage. PD-L1 is also expressed by several types of tumor cells, such as non-small cell lung cancer (NSCLC), hematological malignancies, and virus-infected cells, as an "adaptive immune mechanism" to evade antitumor responses (Ohaegbulam et al., 2015). Several transcription factors have been found to be involved in the transcriptional upregulation of PD-L1 in cancer cells, such as hypoxia-inducible factor a (HIF-1a), STAT3, and NF-kB (Chen et al., 2015). In addition, cytokines produced by infiltrating immune cells, such as IL-4, IL-10, TNFa, IFNg, and growth stem cell factor, as well as bacterial LPS and VEGF, upregulate the expression of the PD-L1 gene (Ji et al., 2015).

[0292] The physiological role of PD-1 is the inhibition of functional T cells and the development of T regulatory (Treg) cells. Treg and co-inhibitory immune checkpoints such as PD-1 / PD-L1 act as fail-safes to prevent aberrant and chronic activation of the immune system and immune reactivity against self-antigens. However, it is well known that the PD-1 / PD-L1 pathway is co-selected by cancer cells or tumor-associated antigen-presenting cells as a means to avoid anti-tumor T cell responses and promote tumor immune evasion (Hargadon et al., 2018). Tumors can escape from host immune surveillance by expressing PD-L1. The interaction between PD-1 and PD-L1 leads to the downregulation of T cells as well as their apoptosis and elimination from the tumor microenvironment, thus allowing cancer cells to escape the immune response (Iwai et al., 2017).

[0293] Targeting PD-1 / PD-L1 PD-1 is a member of the B7 / CD28 family of receptors that share a common structure: an immunoglobulin-like extracellular domain, a transmembrane domain, and an intracellular domain that contains immunoreceptor tyrosine-based inhibitory and signaling motifs. When PD-1 is engaged by its ligand, PD-L1, such interaction leads to the recruitment of the SRC homologous phosphatases SHP1 and SHP2, thereby transmitting signals into the cell. PD-1 is primarily expressed on activated T cells, the primary cytotoxic effectors of the adaptive immune response, and the signals it transmits serve to attenuate T cell-mediated immune responses. Studies have shown that by blocking the inhibitory effects of PD-1, effective immune responses against tumor cells can be elicited. PD-1 / PD-L1 checkpoint inhibitors, which block this interaction, increase immune cell proliferation and enhance the effectiveness of the body's natural antitumor surveillance system. Understanding the mechanisms of cancer evasion of immune checkpoints is critical to approaches to personalizing the delivery of immunotherapies.

[0294] Two separate signals are required to fully activate T cells that keep their cytotoxic activity in check. The first arises from the interaction between the T cell receptor and major histocompatibility complex-presented antigen epitopes on the surface of antigen-presenting cells (APCs). The second arises from the association of costimulatory receptor-ligand pairs on the T cell surface and the APC surface. PD-1 belongs to a group of co-inhibitory receptors that regulate T cell activity through ligand binding, causing T cells to fall into a state known as exhaustion, in which they are unable to proliferate or perform their effector functions.

[0295] Blockade of the PD-1 / PD-L1 pathway has been shown to reactivate / restore the function of exhausted tumor-infiltrating T lymphocytes. For example, treatment with anti-PD-1 / PD-L1 and anti-CTLA-4 immune checkpoint inhibitors has been reported to reactivate dysfunctional TILs and enhance their antitumor effects (Wherry and Kurachi, 2015; Zarour, 2016; Miller et al., 2019).

[0296] CD137 costimulatory pathway CD137, also called 4-1BB, or TNFRSF9 (TNF receptor superfamily member 9), is a 30 kDa transmembrane glycoprotein of the tumor necrosis factor (TNF) receptor superfamily (TNFRSF), originally identified as an inducible co-stimulatory receptor expressed on activated T cells. Current understanding of 4-1BB indicates that expression is generally activation-dependent and encompasses a broad range of immune cell subsets, including activated T cells, activated natural killer (NK) and natural killer T (NKT) cells, regulatory T cells, dendritic cells (DCs), including follicular DCs, stimulated mast cells, differentiated myeloid cells, monocytes, neutrophils, and eosinophils. 4-1BB expression has also been demonstrated on tumor vasculature and atherosclerotic endothelium. The stimulatory ligand for CD137 (CD137L) is expressed on activated antigen-presenting cells (APCs), bone marrow progenitor cells, and hematopoietic stem cells (Wang et al., Immunological Reviews, 2009, 229, 192-215).

[0297] CD137 is undetectable on the surface of naive T cells. Upon stimulatory TCR signaling, CD137 expression induces and peaks 2-3 days after stimulation, then declines after 3 days. CD137 is expressed as both a monomer and a dimer on the cell surface of activated T cells. Based on homology to other members in the TNFRSF family, ligand binding induces receptor trimerization, resulting in receptor activation. Some members of the TNFRSF may exist in a soluble form after cleavage of the extracellular domain from the cell surface. Soluble 4-1BB and soluble 4-1BBL have been detected in the serum of some patients with autoimmune diseases and cancer (Wang et al., Immunological Reviews, 2009, 229, 192-215).

[0298] CD137 is a member of the TNF receptor (TNFR) superfamily with no known intrinsic enzymatic activity in its cytoplasmic domain. It relies on the TNFR-associated factor (TRAF) family of adaptor proteins to assemble the CD137 signalosome to transmit signals into cells. Upon CD137 activation by binding of CD137L trimers or by cross-linking with agonistic monoclonal antibodies, TRAF1, TRAF2, and TRAF3 are readily recruited to the cytoplasmic domain of CD137 as homo- and / or heterotrimers, possibly with different arrangements, to initiate the assembly of the CD137 signalosome, which leads to downstream activation of NF-κB and mitogen-activated protein (MAP) kinase cascades, including ERK, JNK, and p38 MAP kinase (Bartkowiak et al. Clin. Cancer Res. 2018, 24, 1138-1151).

[0299] CD137 plays a key role in maintaining effective T cell immune responses and in generating immunological memory. The expression profile of CD137 and its unique ability to enhance robust effector responses in multiple subsets of lymphocytes relevant for tumor immunity make it a uniquely attractive target for immunotherapy.

[0300] Multiple studies in mice and human T cells indicate that CD137 promotes cell proliferation, survival, and cytokine production. CD137 agonistic antibodies have been shown to significantly enhance cytolytic T lymphocyte responses. Agonistic CD137 antibodies, either as monotherapy or in combination with other treatments, such as checkpoint inhibitors, have provided evidence of antitumor benefit in prophylactic and therapeutic settings. Stimulation of CD137 has been shown to result in sustained antitumor protective T cell memory responses in vivo (Fisher et al, Cancer Immunol Immunother 2012, 61, 1721-1733). More recently, CD137 agonistic antibodies have been shown to increase expression of the cell adhesion molecules ICAM-1, VCAM-1, and E-selectin on tumor vasculature, resulting in increased T cell migration into the tumor microenvironment (Palazon et al, Cancer Research, 2011, 71(3), 8001-811).

[0301] However, the puzzling observation that both CD137- / - mice and agnostic CD137 antibodies exhibit enhanced antitumor activity indicates that mere activation of CD137 signaling cannot fully explain its antitumor effect. Several studies have reported CD137 signaling upon binding to CD137L. Recently, accumulating evidence indicates that CD137L / CD137 interaction results in bidirectional signaling. As CD137L is mainly expressed on activated APCs, such as DCs, reverse signaling from CD137L has been reported to affect DC development and function (Kwon, Immune Network, 2015, 15(3), 121-124). In 2017, a report demonstrated reverse signaling by CD137 ligand (CD137L) in antigen-presenting dendritic cells (DCs) in tumors, explaining these paradoxical results. Specifically, CD137L reverse signaling suppressed intratumoral differentiation of IL12-producing CD103+ DCs and type 1 tumor-associated macrophages (TAMs), which play an important role in generating IFNγ-producing CD8+ cytotoxic T lymphocytes. Notably, CD137L blockade increased the levels of IL12 and IFNγ, which further promoted intratumoral differentiation of IFNγ-producing CD8+ T cells, IL12-producing CD103+ DCs, and type 1 TAMs in tumors. Thus, by activating CD137 signaling in T cells while blocking CD137L reverse signaling in DCs, the antitumor activity of the C137 pathway should be fully elicited (Kang et al., Cancer Research, 2017, 77(21), 5989-6000).

[0302] High doses of CD137 agonist antibodies in naive and tumor-bearing mice have been reported to induce T cell infiltration into the liver and elevation of aspartate aminotransferase (AST) and alanine aminotransferase (ALT), indicators of liver inflammation or injury. Based on clinical trials evaluating the efficacy of an anti-CD137 antibody (Urelumab, Bristol Myers), activation of CD137 activity can induce therapeutic antitumor activity. However, hepatotoxicity limits its clinical development. Clinical evaluation of a second anti-CD137 (Utomilumab, Pfizer) demonstrated an acceptable safety profile, but limited clinical efficacy. To date, there are no approved therapeutic antibodies directed against CD137.

[0303] Bispecific targeting the PD / PD-L1 checkpoint and CD137 Since costimulatory receptors play a key role in regulating the effector functions of T cells, agonism of costimulatory pathways can improve checkpoint blockade efficacy and lead to sustained antitumor responses. Bispecific molecules designed to target the PD-1 / PD-L1 pathway and T cell costimulatory molecules can derepress checkpoints and simultaneously costimulate T cells to provide efficient induction of antitumor immunity. Depending on the aspect of the molecular design, T cell activation can occur through trans-binding and be dependent on PD-L1 binding, thereby effectively restricting immune activity to the tumor microenvironment. Suitable costimulatory targets on T cells include, but are not limited to, CD137, OX40, CD28, CD27, CD226, GITR, ICOS, TNFRSF25, LIGHT (TNFSF14), TIM-1, and LFA-1.

[0304] Previous preclinical experiments and clinical data clearly support that stimulating the costimulatory pathway may be a potentially effective strategy to reactivate T cell responses in cancer, especially when used in combination with other immune activation strategies. Binding proteins that are bispecific for PD-L1 and costimulatory molecules on T cells mediate simultaneous binding to PD-L1-expressing antigen-presenting cells (APCs) or tumor cells and activated T cells, and promote the activation of tumor-specific T cells, such as tumor-infiltrating T cells and CD8 + This should result in conditional activation of cytolytic T cells and reduce the on-target off-tumor toxicity of agonistic anti-CD137 antibodies.

[0305] TGF-β The TGFβ superfamily is a large group of structurally related proteins that includes TGFβ, nodal, activin, lefty, bone morphogenetic proteins, and growth and differentiation factors. TGFβ signaling is transduced through Smad and non-Smad pathways. TGFβ is a pleiotropic cytokine with crucial functions in mediating immune suppression and evasion of immune surveillance in the TME. TGFβ is aberrantly produced by tumors and promotes cancer progression primarily by suppressing both the innate and adaptive immune systems. As a negative regulator of antitumor immunity, TGF-β impairs the efficacy of anti-PD-1 / PD-L1 and induces drug resistance.

[0306] It has been reported that only about 10-30% of patients with most carcinomas achieve an objective response when treated with anti-PD-1 / PD-L1 monotherapy, even in clinical trials that selectively enroll only patients whose pretreatment tumor specimens express PD-L1 (Lipson, EJ et al, Semin. Oncol. 42(4):587 (2015)). One possible explanation for this lack of response is the immunosuppressive nature of the tumor microenvironment due to the presence of immunosuppressive cells (e.g., regulatory T cells (Tregs)) and other factors (e.g., cytokines and metabolic pathways) that function to inhibit T cell priming or suppress effector T cell function. The efficacy of anti-PD-1 / PD-L1 therapy has been reported to be limited in TMEs characterized by hyperactive TGFβ signaling (Tauriello DVF, et al, Nature, 554:538-543 (2018)).

[0307] The effects of TGFβ signaling are mediated by three TGFβ ligands (TGF-β1-3), all of which exist as homodimers through TGFβ type 1 (TGFβR1) and type 2 receptors TGFβR2. In the tumor microenvironment, there are numerous cell context-dependent factors that are closely involved in the balance of TGFβ signaling (Liu S, et al., Mol Med Rep 17:699-704 (2018)). TGFβ has been shown to be a key factor for suppressing anti-tumor immune responses, however, the precise roles of T cell and tumor-derived TGF-β are still poorly understood.

[0308] Tregs, monocytic myeloid-derived suppressor cells (MDSCs), alternatively polarized macrophages (M2 phenotype), and their associated soluble factors are well-recognized inhibitory mechanisms that can suppress antitumor immunity. MDSCs have been reported to be the main source of TGFβ in the tumor microenvironment.

[0309] Tregs are commonly found in solid tumors and can promote immunosuppression through several mechanisms, including competing with effector cells for activating cytokines and secreting immunosuppressive cytokines. For example, Tregs contribute to the levels of transforming growth factor-beta (TGFβ) in the tumor microenvironment, which is an immunosuppressant that subverts both adaptive immune priming and effector responses.

[0310] Designing therapeutic agents that target tumor microenvironment regulators is a recognized strategy used to optimize tumor immunotherapy. For example, the interaction of TGFβ and immune cells has been identified as a key regulatory axis in the tumor microenvironment and cancer progression. TGFβ regulates the function of multiple immune cells by inducing the differentiation of Tregs, reducing the cytotoxicity of T cells and natural killer (NK) cells, limiting tumor infiltration of immune cells, and suppressing antigen presentation by dendritic cells (DCs) (Yi, M et al. J. Hematol. Oncol. 14(1):27(2021). The main strategies for inhibition of the TGF-β signaling pathway are to design therapeutic agents that interfere with the binding of TGF-β to its receptor, block intracellular signaling, or disrupt expression using antisense oligonucleotides.

[0311] TGF-β as a regulator of the tumor microenvironment The efficacy of anti-PD-1 / PD-L1 immunotherapy is limited in TMEs characterized by hyperactive TGFβ signaling. Data from preclinical studies support the premise that blocking or antagonizing TGFβ could potentially disrupt some of the factors that contribute to the immunosuppressive effects of TGFβ. A study by Budhu et al. found that the antitumor activity of T cells was inhibited by tumor-resident regulatory T (Treg) cells, and that immunosuppression was dependent on the presence of TGFβ on the surface of Treg cells. Blocking antibodies against TGF-β reversed immune suppression and boosted antitumor responses (S. Budhu et al, Sci. Signal. 10, eaak9702 (2017)). Other published data suggest that TGFβ limits antitumor immune responses by blocking T cell infiltration into tumors. Studies performed in mouse models of urothelial carcinoma showed that combination treatment with antibodies targeting TGF-β signaling and PD-L1 increased the ability of T cells to enter tumors, induced tumor regression, and enhanced antitumor immune responses (Mariathasan, S et al, Nature 554, 544-548 (2018)).

[0312] Tauriello, DVF et al. studied mice expressing four mutations associated with colorectal cancer and found that metastasis in the mice was characterized by reduced T cell infiltration and active TGFβ in the stroma. Inhibition of the PD-1-PD-L1 immune checkpoint elicited a limited response in the model system, but inhibition of TGFβ unleashed a strong and sustained cytotoxic T cell response that prevented metastasis. In mice with advanced hepatic metastatic disease, blockade of TGFβ signaling sensitized tumors to anti-PD-1 / PD-L1 treatment and led to increased survival (Tauriello, Nature 554, 544-548 (2018)).

[0313] M7824 (Bintrafusp alfa) is a bifunctional fusion protein under development by Merck KGaA, Darmstadt, Germany, and GlaxoSmithKline. M7824 contains a flexible (Gly) residue at the N-terminus of the soluble extracellular domain (136 amino acids) of TGF-βRII. 4 Ser) 4 It contains VH and VL sequences derived from the humanized IgG1 monoclonal antibody avelumab genetically fused via a Gly linker, which functions as a cytokine trap for all three TGF-β (TGF-β1-3) ligand isoforms.

[0314] Several preclinical studies have demonstrated that vintrafusp alfa (1) can prevent or reverse TGF-β-induced epithelial-mesenchymal transition in human cancer cells; this change in tumor cell plasticity renders human tumor cells more susceptible to immune-mediated attack as well as to some chemotherapeutic agents; (2) alters the phenotype of natural killer cells and T cells, thus enhancing their cytolytic capacity against tumor cells; (3) mediates enhanced lysis of human tumor cells via antibody-dependent cell-mediated cytotoxicity mechanisms; (4) reduces the suppressive activity of Treg cells; (5) mediates antitumor activity in multiple preclinical models; and (6) enhances antitumor activity in combination with radiation, chemotherapy, and several other immunotherapeutic agents designed to simultaneously block the PD-L1 and TGF beta pathways (Lind H, et al, J Immunother Cancer, Feb;8(1):e000433(2020) shows that treatment with anti-PD-L1 / TGFβ trap (Bintrafsp) induces synergistic antitumor effects that are reportedly superior to those of single-agent anti-PD-L1 or TGFβ trap proteins (US9,676,863).

[0315] The observed synergy is due to the simultaneous blockade of the interaction between PD-1 on immune cells and PD-L1 on tumor cells and the neutralization of TGFβ in the tumor microenvironment (US9,676,863). The synergy is presumably based on the simultaneous blockade of two major immune evasion mechanisms: TGFβ depletion is achieved by antagonizing TGFβ cytokine levels in the tumor microenvironment (as a result of anti-PD-L1 targeting of tumor cells) and by the destruction of TGFβ through PD-L1 receptor-mediated endocytosis (US9,676,863).

[0316] PD-L1 binding proteins targeting the PD / PD-L1 checkpoint and TGFβ It has been reported that dual blockade of PD-1 / PD-L1 and TGFβ has synergistic antitumor activity (Chen,X et al. Int. J. Cancer 143:2561(2018) and Yi,M et al. J. Hematol. Oncol.14(1):27(2021). Considering that the immunosuppressive effects of PD-1 / PD-L1 axis and TGFβ are complementary and independent, it is reasonable to design PD-L1 binding proteins that can block TGFβ signaling to enhance the efficacy of anti-PD-L1 immunotherapy (Yi,M et al. J. Hematol. Oncol.14(1):27(2021)). In addition, bi- or tri-specific binding proteins that bind PD-L1 and contain TGF-β pathway antagonists may provide an effective option for patients resistant to PD-1 / PD-L1 monotherapy (Kim et al.,J Hematol Oncol,14:55(2021)).

[0317] To enhance the efficacy of anti-PD-L1 immunotherapy, binding proteins are provided that can simultaneously block the PD-1 / PD-L1 axis and the TGFβ signaling pathway. In some embodiments, the binding proteins can also bind to CD137 and provide a costimulatory signal that promotes T cell responsiveness.

[0318] PD-L1 monospecific In some embodiments, the disclosed PD-L1 monospecifics (1923Ab2 and 1923Ab3) are specific for (e.g., specifically bind to) human PD-L1. These binding proteins and fragments thereof are characterized by a unique set of CDR sequences, specificity for PD-L1, and exhibit potent inhibitory activity of PD-1 / PD-L1 signaling. More specifically, in one aspect, the disclosure relates to binding proteins that bind to human PD-L1 and their use as monotherapy or in combination with other anti-cancer agents to modulate PD-L1-mediated activity of cells localized to the tumor microenvironment. In alternative embodiments, the disclosed binding proteins that bind to PD-L1 can also be used as subunits of bispecifics and trispecifics designed to disinhibit / release effector T cells from PD-1 / PD-L1 checkpoint inhibition.

[0319] In alternative aspects, the disclosure relates to the use of the disclosed binding proteins that bind PD-L1 to design bispecifics or trispecifics that bind PD-L1, and their use to deactivate the PD-L1 / PD-L1 checkpoint and promote T cell activation.

[0320] In an exemplary embodiment, the invention provides a binding protein that binds to PD-L1, and comprises an antibody scaffold module that comprises: (a) a heavy chain variable region sequence that comprises the CDR sequences for 1923Ab2 in Table 1; and (b) a light chain variable region sequence that comprises the CDR sequences for 1923Ab2 in Table 2.

[0321] In an exemplary embodiment, the invention provides a binding protein that binds to PD-L1, and comprises an antibody scaffold module that comprises: (a) a heavy chain variable region sequence that comprises the CDR sequences for 1923Ab3 in Table 1; and (b) a light chain variable region sequence that comprises the CDR sequences for 1923Ab3 in Table 2. [Table 1] [Table 2]

[0322] In some embodiments, binding proteins that bind PD-L1 may be monoclonal, chimeric, bispecific, or trispecific, with a heavy chain region sequence comprising a VH (e.g., SEQ ID NOs: 1 and 3) and a VL (e.g., SEQ ID NOs: 2 and 4), and the constant region may be IgG1, IgG2, IgG3, or IgG4. In a further embodiment, the antibodies of the invention are human IgG1 type with Fc silencing mutations (L234A L235A or N297A).

[0323] In some embodiments, it is advantageous for the disclosed anti-PD-L1 antibodies to be Fc-engineered.

[0324] In yet a further embodiment, the invention provides a binding protein that binds to PD-L1 comprising a heavy chain sequence and a light chain sequence, wherein the heavy chain sequence has at least 85% sequence identity to SEQ ID NO: 42 or 45, and the light chain sequence has at least 85% sequence identity to SEQ ID NO: 40.

[0325] In one embodiment, the disclosed antibodies are capable of binding to human or cynomolgus PD-L1 and blocking the interaction between human PD-L1 and the PD1 receptor.

[0326] In one embodiment, the binding protein is 5×10 -9 With a KD of M or less, preferably 2×10 -9 M or less, and even more preferably 1×10 -9 Binds to human PD-L1 with a KD of M or less.

[0327] In a further embodiment, the invention relates to a binding protein that binds to human PD-L1, or a fragment thereof, which cross-competes for binding to PD-L1 with the antibody according to the invention (Tecentriq), as described herein.

[0328] In some embodiments, the binding proteins that bind to PD-L1 or a fragment thereof exhibit one or more of the following structural and functional characteristics, alone or in combination: (a) are specific for human PD-L1, (b) cross-react with cynomolgus PD-L1, (c) disrupt the binding of PD-L1 to PD-1, or (d) ablate a T cell PD-L1-mediated checkpoint inhibitory signal.

[0329] Disruption of the PD-L1 / PD-1 interaction and de-inhibition of activation checkpoints by the disclosed binding proteins that bind to PD-L1 were examined using multiple in vitro assays. The biological activity of the anti-PD-L1 moieties was determined by their ability to disrupt the interaction between PD-1 and PD-L1, restoring TCR signaling using a PD-1 / PD-L1 blocking assay.

[0330] CD137 monospecific In some embodiments, the disclosed CD137 monospecifics (1923Ab4, 1923Ab5, and 1923Ab6) are specific for (e.g., specifically bind to) human CD137. These binding proteins and fragments thereof are characterized by a unique set of CDR sequences, specificity for CD137, and are useful in cancer immunotherapy as monotherapy or in combination with other anti-cancer agents. As demonstrated herein, binding proteins that bind CD137 can also be used as subunits of bispecifics and trispecifics designed to reactivate effector T cells released from PD1 / PD-L1 checkpoint inhibition. More specifically, the present disclosure relates to binding proteins that bind human CD137 and their use to modulate CD137-mediated activity of cells localized to the tumor microenvironment.

[0331] In an exemplary embodiment, the invention provides a binding protein that binds to CD137, comprising: (a) an antibody scaffold module comprising: a heavy chain variable region sequence comprising the CDR sequences for 1923Ab4 in Table 3; and a light chain variable region sequence comprising the CDR sequences for 1923Ab4 in Table 4.

[0332] In an exemplary embodiment, the invention provides a binding protein that binds to CD137, comprising: (a) an antibody scaffold module comprising: a heavy chain variable region sequence comprising the CDR sequences for 1923Ab5 in Table 3; and a light chain variable region sequence comprising the CDR sequences for 1923Ab5 in Table 4.

[0333] In an exemplary embodiment, the invention provides a binding protein that binds to CD137, comprising: (a) an antibody scaffold module comprising: a heavy chain variable region sequence comprising the CDR sequences for 1923Ab6 in Table 3; and a light chain variable region sequence comprising the CDR sequences for 1923Ab6 in Table 4. [Table 3] [Table 4]

[0334] In some embodiments, the binding proteins that bind CD137 may be monoclonal, chimeric, bispecific or trispecific with heavy chain region sequences comprising VH (e.g., SEQ ID NOs: 16, 18, and 20) and VL (e.g., SEQ ID NOs: 17, 19, and 21), and the constant region may be IgG1, IgG2, IgG3 or IgG4. In further embodiments, the antibodies of the invention are human IgG1 type with Fc silencing mutations (L234A L235A or N297A).

[0335] In some embodiments, it is advantageous for the disclosed anti-CD137 antibodies to be Fc-engineered.

[0336] In yet a further embodiment, the invention provides a binding protein that binds to CD137 comprising a heavy chain sequence and a light chain sequence, wherein the heavy chain sequence has at least 85% sequence identity with the heavy chain sequence: SEQ ID NO: 75: and the light chain sequence has at least 85% sequence identity with the light chain sequence: SEQ ID NO: 76.

[0337] Applicants set out to discover binding proteins that bind to CD137 that demonstrate a desirable profile to overcome on-target off-tumor toxicity and immunosuppressive settings for better immunotherapy. The disclosed binding proteins that bind to CD137 may be particularly beneficial for tumor microenvironments enriched in exhausted T cells or regulatory T cells that contribute to anti-PD-1 / PD-L1 resistance.

[0338] In some embodiments, the binding proteins and fragments thereof that bind CD137 exhibit one or more of the following structural and functional characteristics, alone or in combination: (a) specific for human CD137; (b) cross-reacts with cynomolgus monkey CD137; (c) disrupting (e.g., reducing or preventing) human CD137L binding to CD137; (d) exhibits fast on and fast off properties for CD137; (e) has cross-linking-dependent agonistic activity against CD137 signaling; or (f) Activates T cells in a cross-linking dependent manner.

[0339] In one embodiment, the disclosed binding proteins activate CD137 signaling in the presence of a cross-linking agent. In a T cell activation assay using primary PBMCs, they enhanced anti-CD3 stimulated IFN-gamma release in a cross-linking dependent manner.

[0340] In some embodiments, the disclosed binding proteins advantageously bind to both human CD137 and cynomolgus monkey CD137. Cross-reactivity with CD137 expressed on cells in cynomolgus monkeys (e.g., Macaca fascicularis) is advantageous because it allows for animal testing of the antibody molecule without the need to use surrogate antibodies. The disclosed binding proteins that bind to CD137, 1923Ab4, 1923Ab5, and 1923Ab6, all bind CD137 from cynomolgus monkeys with significant affinity. Although the disclosed antibodies do not bind to mouse CD137, humanized CD137 mice are available and can be used to test the efficacy of the disclosed antibodies in preclinical mouse tumor models in vivo.

[0341] T cell activation by the disclosed anti-CD137 antibodies was examined using multiple in vitro assays. The biological activity of the anti-CD137 moiety was determined by its ability to induce crosslinking-dependent CD137 signaling using assay cells overexpressing CD137 and carrying an NFκB luciferase reporter. Crosslinking anti-CD137 also enhances anti-CD3-stimulated IFNγ release in PBMCs.

[0342] In some embodiments, the binding proteins disclosed herein may contain one or more conservative amino acid substitutions. One of skill in the art will recognize that a conservative amino acid substitution is the replacement of one amino acid with another amino acid that has similar structural or chemical properties, such as a similar side chain. Exemplary conservative substitutions are described in the art, for example, in Watson et al., Molecular Biology of the Gene, The Benjamin / Cummings Publication Company, 4th Ed. (1987).

[0343] "Conservative modifications" refer to amino acid modifications that do not significantly affect or alter the binding characteristics of the binding protein containing the amino acid sequence. Conservative modifications include amino acid substitutions, additions, and deletions. A conservative substitution is one in which an amino acid is replaced with an amino acid residue having a similar side chain. Families of amino acid residues having similar side chains have been well defined, including amino acids with acidic side chains (e.g., aspartic acid, glutamic acid), basic side chains (e.g., lysine, arginine, histidine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine), uncharged polar side chains (e.g., glycine, asparagine, glutamine, cysteine, serine, threonine, tyrosine, tryptophan), aromatic side chains (e.g., phenylalanine, tryptophan, histidine, tyrosine), aliphatic side chains (e.g., glycine, alanine, valine, leucine, isoleucine, serine, threonine), amides (e.g., asparagine, glutamine), beta-branched side chains (e.g., threonine, valine, isoleucine), and sulfur-containing side chains (cysteine, methionine). Additionally, any naturally occurring residues in the polypeptide may also be substituted with alanine as previously described for alanine scanning mutagenesis (MacLennan et al. (1998) Acta Physiol Scand Suppl 643:55-67; ​​Sasaki et al. (1998) Adv Biophys 35:1-24). Amino acid substitutions into the binding proteins disclosed herein may be made by known methods, such as PCR mutagenesis (U.S. Patent No. 4,683,195).

[0344] In some embodiments, the binding protein comprises a variable heavy chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 16, 18, or 20. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable heavy chain sequence of SEQ ID NO: 16, 18, or 20. In still further embodiments, the binding protein comprises a variable heavy chain sequence of SEQ ID NO: 16, 18, or 20 and has one or more conservative amino acid substitutions in the heavy chain variable sequence, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 16, 18, or 20 (based on the Kabat numbering system).

[0345] In certain embodiments, the binding protein comprises a variable heavy chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to a binding protein heavy chain variable region sequence set forth in SEQ ID NO: 16, 18, or 20, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising a variable heavy chain sequence set forth in SEQ ID NO: 16, 18, or 20 and a variable light chain sequence set forth in SEQ ID NO: 17, 19, or 21.

[0346] In some embodiments, the binding protein comprises a variable light chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 17, 19, or 21. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable light chain sequence of SEQ ID NO: 17, 19, or 21. In still further embodiments, the binding protein comprises the variable light chain sequence of SEQ ID NO: 17, 19, or 21 and has one or more conservative amino acid substitutions, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions in the light chain variable sequence. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 17, 19, or 21 (based on the Kabat numbering system).

[0347] In certain embodiments, the binding protein comprises a variable light chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to a binding protein light chain variable region sequence set forth in SEQ ID NO: 17, 19, or 21, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising a variable heavy chain sequence set forth in SEQ ID NO: 16, 18, or 29 and a variable light chain sequence set forth in SEQ ID NO: 17, 19, or 21.

[0348] PD-L1 / CD137 bispecific In certain embodiments, the present disclosure provides a PD-L1 / CD137 bispecific comprising a PD-L1 antibody scaffold module derived from an anti-PD-L1 antibody and a CD137 first binding module derived from an anti-CD137 antibody, where the bispecific stimulates CD137 and activates T cells in a PD-L1 dependent manner.

[0349] In some embodiments, a PD-L1 / CD137 bispecific comprises a heavy chain disclosed in Table 5. For example, a PD-L1 / CD137 bispecific may comprise a HC with a set of CDR sequences derived from the VH of 1923Ab3, and a modified human IgG1 Fc region, with or without amino acid modifications to reduce or ablate Fc receptor function. In some embodiments, the HC may further comprise a binding protein that binds CD137 disclosed herein, for example, an scFv with a set of CDR sequences derived from the VH and VL regions of 1923Ab4. In some embodiments, the scFv may be attached to the HC by a linker positioned at the N- or C-terminus of the HC.

[0350] In some embodiments, a PD-L1 / CD137 bispecific comprises a light chain disclosed in Table 5. For example, a PD-L1 / CD137 bispecific may comprise a LC with a set of CDR sequences derived from the VL of 1923Ab3. In one embodiment, the LC may further comprise a binding protein that binds CD137 disclosed herein, for example an scFv with a set of CDR sequences derived from the VH and VL regions of 1923Ab4. In some embodiments, the scFv may be attached to the LC by a linker positioned at the N- or C-terminus of the LC. [Table 5]

[0351] In some embodiments, PD-L1 / CD137 bispecific 1923Ab8 (Figure 13B) comprises two Fabs from 1923Ab3 that bind PD-L1, a human IgG1 Fc with L234A L235A mutations, and two scFv fragments derived from 1923Ab4 (VH precedes VL) attached to the C-terminus of the Ab3 heavy chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab8. Figure 15 provides a more detailed description of the subcomponents of the disclosed binding protein. The amino acid sequences of the heavy and light chains are provided in SEQ ID NO:38 and SEQ ID NO:40, respectively.

[0352] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab11 (Figure 13E) comprises two Fabs from 1923Ab3 that bind PD-L1, a human IgG1 Fc with L234A L235A mutations, and two scFv fragments derived from 1923Ab4 (VL precedes VH) attached to the C-terminus of the Ab3 heavy chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab11. Figure 15 provides a more detailed description of the subcomponents. The amino acid sequences of the heavy and light chains are provided in SEQ ID NO:44 and SEQ ID NO:40, respectively.

[0353] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab12 (Figure 13F) comprises two Fabs from 1923Ab3 that bind PD-L1, a human IgG1 Fc with L234A L235A mutations, and two disulfide bond stabilized scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the N-terminus of the Ab3 light chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab12. Figure 15 provides a more detailed description of the subcomponents. The amino acid sequences of the heavy and light chains are provided in SEQ ID NO:45 and SEQ ID NO:46, respectively.

[0354] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab13 (Figure 13G) comprises two Fabs from 1923Ab3 that bind PD-L1, a human IgG1 Fc with L234A L235A mutations, and two disulfide bond stabilized scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the N-terminus of the Ab3 heavy chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab13. Figure 15 provides a more detailed description of the subcomponents. The amino acid sequences of the heavy and light chains are provided in SEQ ID NO:47 and SEQ ID NO:40, respectively.

[0355] In some embodiments, the PD-L1 / CD137 bispecific 1923Ab18 (Figure 13J) comprises two Fabs from 1923Ab3 that bind PD-L1, a human IgG1 Fc with L234A L235A mutations, and two disulfide bond stabilized scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the Ab3 heavy chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab18. Figure 15 provides a more detailed description of the subcomponents. The amino acid sequences of the heavy and light chains are provided in SEQ ID NO:50 and SEQ ID NO:40, respectively.

[0356] The biological activity of the anti-PD-L1 moiety was determined by its ability to disrupt the interaction between PD-1 and PD-L1, restoring TCR signaling using a PD-1 / PD-L1 blocking assay.The biological activity of the anti-CD137 moiety was determined by its ability to induce crosslinking-dependent CD137 signaling using assay cells overexpressing CD137 and carrying an NFκB luciferase reporter.

[0357] In some embodiments, the PD-L1 / CD137 bispecific exhibits one or more of the following characteristics, alone or in combination: (a) is specific for human PD-L1 and binds to human CD137; (b) cross-reacts with cynomolgus monkey PD-L1 and CD137; (c) disrupting the interaction of PD-1 and PD-L1; (d) disrupting (e.g., reducing or preventing) human CD137L binding to CD137; (e) exhibits fast-on and fast-off properties for CD137; (f) derepressing T cell PD-L1-mediated checkpoint inhibitory signals; (g) has PD-L1-dependent agonistic activity on CD137 signaling; (h) activates T cells in a PD-L1-dependent manner; (i) killing PD-L1-expressing tumor cells by activating CD8 T cells; (j) Demonstrate anti-tumor efficacy in human PD-L1 and CD137 knock-in using the MC38-hPD-L1 syngeneic tumor model; (k) increasing CD8+ T cells in the tumor microenvironment; (l) reducing the percentage of Treg cells in the tumor microenvironment; and (m) Reducing on-target toxicity outside the tumor microenvironment.

[0358] In certain embodiments, the bispecifics have the ability to enhance immune cell proliferation, survival, cytolytic activity of CD8 T cells and secretion of cytokines. In certain embodiments, the disclosed PD-L1 / CD137 bispecifics have been shown to activate human T cells in a CMV recall assay with greater potency than monospecific antibody combinations.

[0359] In some embodiments, the binding proteins described herein have a characteristic selected from the group consisting of: disrupting the interaction of PD-1 and PD-L1, ablating T cell PD-L1 mediated checkpoint inhibitory signals, having PD-L1 dependent agonistic activity on CD137 signaling, activating T cells in a PD-L1 dependent manner, killing PD-L1 expressing tumor cells by activating CD8 T cells, demonstrating anti-tumor efficacy in human PD-L1 and CD137 knock-in using the MC38-hPD-L1 syngeneic tumor model, increasing CD8+ / T cells in the tumor microenvironment, and decreasing the percentage of Treg cells in the tumor.

[0360] In certain embodiments, the bispecifics have the ability to bind to CD137 and PD-L1 expressing cells. In another embodiment, the binding of the bispecifics to PD-L1 disrupts its interaction with PD-1, thereby resulting in the restoration of TCR signaling in Jurkat T cells. Also, PD-L1 binding bispecifics activate CD137 signaling, whereas monospecific anti-PD-L1 antibodies alone do not induce CD137 signaling. Furthermore, they enhance anti-CD3 stimulated IFNγ release in PBMCs. More specifically, the bispecifics redirect CD8+ to kill PD-L1 expressing tumor cells in vitro. In one embodiment, the disclosed bispecifics showed strong anti-tumor efficacy using human PD-L1 expressing MC38 mouse tumors in human CD137 and human PD-L1 double knock-in mice.

[0361] In some embodiments, the binding protein comprises a variable heavy chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 38, 44, 45, 47, or 50. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable heavy chain sequence of SEQ ID NO: 38, 44, 45, 47, or 50. In yet further embodiments, the binding protein comprises a variable heavy chain sequence of SEQ ID NO: 38, 44, 45, 47, or 50 and has one or more conservative amino acid substitutions in the heavy chain variable sequence, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 38, 44, 45, 47, or 50 (based on the Kabat numbering system).

[0362] In certain embodiments, the binding protein comprises a variable heavy chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to a binding protein heavy chain variable region sequence set forth in SEQ ID NO: 38, 44, 45, 47, or 50, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising a variable heavy chain sequence set forth in SEQ ID NO: 38, 44, 45, 47, or 50 and a variable light chain sequence set forth in SEQ ID NO: 40 or 46.

[0363] In some embodiments, the binding protein comprises a variable light chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 40 or 46. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable light chain sequence of SEQ ID NO: 40 or 46. In yet further embodiments, the binding protein comprises a variable light chain sequence of SEQ ID NO: 40 or 46 and has one or more conservative amino acid substitutions in the light chain variable sequence, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 40 or 46 (based on the Kabat numbering system).

[0364] In certain embodiments, the binding protein comprises a variable light chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the binding protein light chain variable region sequence set forth in SEQ ID NO: 40 or 46, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising a variable heavy chain sequence set forth in SEQ ID NO: 38, 44, 45, 47, or 50 and a variable light chain sequence set forth in SEQ ID NO: 40 or 46.

[0365] PD-L1 / TGFβ bispecific In an alternative embodiment, the disclosure provides a PD-L1 / TGFβ bispecific comprising a PD-L1 antibody scaffold module derived from an anti-PD-L1 antibody and a tumor microenvironment regulator derived from the ECD of human TGFβRII, where the bispecific deactivates the PD-1 / PD-L1 checkpoint and neutralizes the immunosuppressive effects of TGFβ in the tumor microenvironment.

[0366] In another embodiment, the disclosure provides a PD-L1 / TGFβ bispecific comprising a PD-L1 antibody scaffold module derived from an anti-PD-L1 antibody and a TGFβ first binding module derived from the ECD of human TGFβ receptor type II, where the bispecific is capable of binding to both PD-L1 and TGFβ, and neutralizing the biological activity of TGFβ in the TME.

[0367] In some embodiments, the PD-L1 / TGFβ bispecific 1923Ab20 (Figure 13L) comprises two Fabs from 1923Ab3 that bind to PD-L1, a human IgG1 Fc with L234A L235A mutations, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the 1923Ab3 heavy chain. Figure 14(A) provides a description of the heavy and light chains of 1923Ab20. Figure 15 provides a more detailed description of the subcomponents. Table 6 provides the amino acid sequences of the heavy and light chains, SEQ ID NO:51 and SEQ ID NO:40, respectively. [Table 6]

[0368] The biological activity of the anti-PD-L1 moiety was determined by its ability to disrupt the interaction between PD-1 and PD-L1, restoring TCR signaling using a PD-1 / PD-L1 blocking assay. The ability of the extracellular domain (ECD) of TGFβRII to neutralize the biological activity of human TGFβ was determined by its ability to block TGFβ-induced signaling cascades using SBE (SMAD-binding element) reporter cells.

[0369] In some embodiments, the bispecifics exhibit one or more of the following characteristics, alone or in combination: (a) are specific for human PD-L1 and bind to human TGFβ; (b) disrupt the interaction of PD-1 and PD-L1; (c) de-inhibit T cell PD-L1-mediated checkpoint inhibition signals; (d) bind to and neutralize the biological activity of human TGFβ; (e) have reduced toxicity outside the tumor microenvironment; (f) increase CD8+ T cells in the tumor microenvironment; or (g) reduce the percentage of Treg cells in the tumor microenvironment.

[0370] In certain embodiments, the disclosed PD-L1 / TGFβ bispecifics have the ability to enhance immune cell proliferation, survival, CD8 T cell cytolytic activity, and cytokine secretion. In certain embodiments, the disclosed PD-L1 / TGFβ bispecifics have been shown to activate human T cells in a CMV recall assay with greater potency than monospecific antibody combinations.

[0371] In some embodiments, the binding proteins described herein have a characteristic selected from the group consisting of: disrupting the interaction of PD-1 and PD-L1, ablating T cell PD-L1 mediated checkpoint inhibitory signals, activating T cells in a PD-L1 dependent manner, killing PD-L1 expressing tumor cells by activating CD8 T cells, increasing CD8+ T cells in the tumor microenvironment, and decreasing the percentage of Treg cells in the tumor.

[0372] In some embodiments, the PD-L1 / TGFβ bispecifics exhibit one or more of the following functional characteristics, alone or in combination: disrupting the interaction of PD-1 and PD-L1, ablating T cell PD-L1-mediated checkpoint inhibition signals, and inhibiting TGFβ signaling.

[0373] In certain embodiments, the disclosed bispecifics in combination with urelumab-NR have been shown to activate human T cells in a CMV recall assay.

[0374] In one embodiment, the disclosed antibodies comprising the extracellular domain (ECD) of TGFβRII block TGFβ-induced signaling in HEK cells carrying an SBE reporter gene.

[0375] In a particular embodiment, the bispecific is capable of binding to PD-L1 expressing cells, hi another embodiment, binding of the bispecific to PD-L1 disrupts its interaction with PD-1, thereby resulting in restoration of TCR signaling in Jurkat T cells.

[0376] In some embodiments, the binding protein comprises a variable heavy chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 51. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable heavy chain sequence of SEQ ID NO: 51. In yet further embodiments, the binding protein comprises the variable heavy chain sequence of SEQ ID NO: 51 and has one or more conservative amino acid substitutions, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions in the heavy chain variable sequence. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 51 (based on the Kabat numbering system).

[0377] In certain embodiments, the binding protein comprises a variable heavy chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the binding protein heavy chain variable region sequence set forth in SEQ ID NO:51, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising the variable heavy chain sequence set forth in SEQ ID NO:51 and the variable light chain sequence set forth in SEQ ID NO:40.

[0378] In some embodiments, the binding protein comprises a variable light chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 40. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable light chain sequence of SEQ ID NO: 40. In yet further embodiments, the binding protein comprises the variable light chain sequence of SEQ ID NO: 40 and has one or more conservative amino acid substitutions, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions in the light chain variable sequence. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 40 (based on the Kabat numbering system).

[0379] In certain embodiments, the binding protein comprises a variable light chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the binding protein light chain variable region sequence set forth in SEQ ID NO:40, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising the variable heavy chain sequence set forth in SEQ ID NO:51 and the variable light chain sequence set forth in SEQ ID NO:40.

[0380] PD-L1 / TGFβ / CD137 trispecific In a specific embodiment, the binding proteins of the disclosure are trispecific and are constructed in the form of recombinant proteins that comprise an antibody scaffold module (derived from an antibody) that binds to PD-L1, a first binding module that comprises a TGF beta receptor II binding protein, and a second binding module (derived from an antibody) that binds to CD137.

[0381] In certain embodiments, the disclosure provides a trispecific comprising a PD-L1-binding antibody scaffold module derived from an anti-PD-L1 antibody, a first binding module for TGFβ derived from the ECD of TGFβ receptor type 2, and a second binding module for CD137 derived from an anti-CD137 antibody, where the trispecific binds to PD-L1, CD137, and depletes TGFβ from the local microenvironment, thereby activating T cells in a PD-L1-dependent manner. [Table 7]

[0382] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab7 (Figure 13A) comprises two Fabs from 1923Ab3 that bind to PD-L1, a human IgG1 Fc with L234A L235A mutations, two scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the 1923Ab3 heavy chain, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the 1923Ab3 light chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab7. Figure 15 provides a more detailed description of the subcomponents of the trispecific. Table 7 provides the amino acid sequences of the heavy and light chains, SEQ ID NO:38 and SEQ ID NO:39, respectively.

[0383] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab9 (Figure 13C) comprises two Fabs from 1923Ab3 that bind to PD-L1, a heterodimeric human IgG1 Fc with L234A L235A and knob-in-hole (KiH) mutations, one scFv fragment (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the "knob" heavy chain, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the 1923Ab3 light chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab9. Figure 15 provides a more detailed description of the subcomponents of the trispecific. Table 7 provides the amino acid sequences of heavy chain 1 (knob) (SEQ ID NO: 41), heavy chain 2 (hole) (SEQ ID NO: 42), and light chain (SEQ ID NO: 39).

[0384] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab10 (Figure 13D) comprises two Fabs from 1923Ab3 that bind to PD-L1, a heterodimeric human IgG1 Fc with L234A L235A and knob-in-hole (KiH) mutations, one scFv fragment (VL precedes VH) derived from 1923Ab4 attached to the C-terminus of the "knob" heavy chain, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the Ab3 light chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab10. Figure 15 provides a more detailed description of the subcomponents of the trispecific. Table 7 provides the amino acid sequences of heavy chain 1 (knob) (SEQ ID NO: 43), heavy chain 2 (hole) (SEQ ID NO: 42), and light chain (SEQ ID NO: 39).

[0385] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab17 (Figure 13I) comprises two Fabs from 1923Ab3 that bind to PD-L1, a human IgG1 Fc with L234A L235A mutations, two disulfide bond stabilized scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the Ab3 heavy chain, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the Ab3 light chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab17. Figure 15 provides a more detailed description of the subcomponents of the trispecific. Table 7 provides the amino acid sequences of the heavy and light chains, SEQ ID NO:50 and SEQ ID NO:39, respectively.

[0386] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific 1923Ab19 (Figure 13K) comprises two Fabs from 1923Ab3 that bind to PD-L1, a human IgG1 Fc with L234A L235A mutations, two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the 1923Ab3 light chain, and two disulfide bond stabilized scFv fragments (VH precedes VL) derived from 1923Ab4 attached to the C-terminus of the 1923Ab3 heavy chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab19. Figure 15 provides a more detailed description of the subcomponents of the trispecific. Table 7 provides the amino acid sequences of the heavy and light chains, SEQ ID NO:51 and SEQ ID NO:52, respectively.

[0387] To facilitate heterodimerization of recombinants based on a natural IgG-like structural scaffold, the knob-into-hole (KiH) technique has been widely applied, which involves engineering the CH3 domain to create either a "knob" or a "hole" in each heavy chain to facilitate heterodimerization. In some embodiments, binding proteins that specifically bind to PD-L1 are characterized by an asymmetric design. An example of a "knob" mutation includes T366W in the CH3 domain, and an example of a "hole" mutation includes T366S, L368A, and Y407V in the CH3 domain. In one embodiment, a stabilizing disulfide bond can be introduced by an additional S354C mutation in the "knob" and an additional Y349C mutation in the "hole". All residue numbers are in EU numbering.

[0388] The biological activity of the PD-L1 / TGFβ / CD137 trispecifics was determined by the corresponding signaling events regulated by PD-L1, TGFβ, or CD137. In some embodiments, the anti-PD-L1 moiety was determined by its ability to disrupt the interaction between PD-1 and PD-L1 using a PD-1 / PD-L1 blocking assay. The biological activity of the anti-CD137 moiety was determined by its ability to induce crosslinking-dependent CD137 signaling using assay cells overexpressing CD137 and carrying an NFκB luciferase reporter. The biological activity of the anti-TGF-β moiety was determined by its ability to block TGFβ-induced signaling cascades using SBE (SMAD binding element) reporter cells.

[0389] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific exhibits one or more of the following functional characteristics, alone or in combination: (a) is capable of binding to human PD-L1, CD137 and TGFβ; (b) cross-reacts with cynomolgus monkey PD-L1 and CD137; (c) disrupting (e.g., reducing or preventing) the interaction of PD-1 and PD-L1; (d) disrupting (e.g., reducing or preventing) human CD137L binding to CD137; (e) exhibits fast-on and fast-off properties for CD137; (f) derepressing T cell PD-L1-mediated checkpoint inhibitory signals; (g) inhibiting TGFβ signaling and neutralizing its biological activity; (h) has PD-L1-dependent agonistic activity on CD137 signaling; (pi) activates T cells in a PD-L1-dependent manner; and (j) Killing PD-L1-expressing tumor cells by activating CD8 T cells.

[0390] In some embodiments, the PD-L1 / TGFβ / CD137 trispecific exhibits one or more of the following functional characteristics, alone or in combination: disrupts the interaction of PD-1 and PD-L1, ablates T cell PD-L1-mediated checkpoint inhibition signals, inhibits TGFβ signaling, has PD-L1-dependent agonist activity on CD137 signaling, activates T cells in a PD-L1-dependent manner, and kills PD-L1-expressing tumor cells by activating CD8 T cells.

[0391] In certain embodiments, the trispecifics have the ability to enhance immune cell proliferation, survival, cytolytically active CD8 T cells, and cytokine secretion. In certain embodiments, the disclosed trispecifics have been shown to activate human T cells in a CMV recall assay with greater potency than combinations of monospecific antibodies.

[0392] In one embodiment, the trispecific blocks TGFβ-induced signaling in HEK cells carrying an SBE reporter gene.

[0393] In a particular embodiment, the trispecific has the ability to bind to CD137 and PD-L1 expressing cells. In another embodiment, the binding of the trispecific to PD-L1 disrupts its interaction with PD-1, thereby resulting in the restoration of TCR signaling in Jurkat T cells. PD-L1 binding bispecifics also activate CD137 signaling. Furthermore, they enhance anti-CD3 stimulated IFNγ release in PBMCs. More specifically, the trispecific redirects CD8+ to kill PD-L1 expressing tumor cells in vitro.

[0394] In some embodiments, the binding protein comprises a variable heavy chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 38, 41, 42, 43, 50, or 51. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable heavy chain sequence of SEQ ID NO: 38, 41, 42, 43, 50, or 51. In yet further embodiments, the binding protein comprises the variable heavy chain sequence of SEQ ID NO: 38, 41, 42, 43, 50, or 51 and has one or more conservative amino acid substitutions in the heavy chain variable sequence, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 38, 41, 42, 43, 50, or 51 (based on the Kabat numbering system).

[0395] In certain embodiments, the binding protein comprises a variable heavy chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to a binding protein heavy chain variable region sequence set forth in SEQ ID NO: 38, 41, 42, 43, 50, or 51, and contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising the variable heavy chain sequence set forth in SEQ ID NO: 38, 41, 42, 43, 50, or 51 and the variable light chain sequence set forth in SEQ ID NO: 39 or 52.

[0396] In some embodiments, the binding protein comprises a variable light chain sequence comprising an amino acid sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 39 or 52. In other embodiments, the binding protein retains the binding and / or functional activity of a binding protein comprising the variable light chain sequence of SEQ ID NO: 39 or 52. In yet further embodiments, the binding protein comprises the variable light chain sequence of SEQ ID NO: 39 or 52 and has one or more conservative amino acid substitutions in the light chain variable sequence, e.g., 1, 2, 3, 4, 5, 1-2, 1-3, 1-4, or 1-5 conservative amino acid substitutions. In still further embodiments, the one or more conservative amino acid substitutions fall within one or more framework regions of SEQ ID NO: 39 or 52 (based on the Kabat numbering system).

[0397] In certain embodiments, the binding protein comprises a variable light chain sequence with at least about 95%, about 96%, about 97%, about 98%, or about 99% sequence identity to the binding protein light chain variable region sequence set forth in SEQ ID NO: 39 or 52, contains one or more conservative amino acid substitutions in the framework regions (based on the Kabat numbering system), and retains the binding activity and / or functional activity of a binding protein comprising a variable heavy chain sequence set forth in SEQ ID NO: 38, 41, 42, 43, 50, or 51 and a variable light chain sequence set forth in SEQ ID NO: 39 or 52.

[0398] CD137 / TGFβ / PD-L1 trispecific In certain embodiments, the binding proteins of the disclosure are trispecific and constructed in the form of recombinant proteins comprising an antibody scaffold module (derived from an antibody) that binds to CD137, a first binding module that comprises a TGFβ receptor II binding protein, and a second binding module (derived from an antibody) that binds to PD-L1.

[0399] In certain embodiments, the present disclosure provides a trispecific comprising a CD137-binding antibody scaffold module derived from an anti-CD137 antibody, a first binding module for TGFβ derived from the ECD of TGFβ receptor type 2, and a second binding module for PD-L1 derived from an anti-PD-L1 antibody, where the trispecific binds to CD137, PD-L1, and depletes TGFβ from the local microenvironment, thereby activating T cells in a PD-L1-dependent manner. [Table 8]

[0400] In some embodiments, the CD137 / TGFβRII / PD-L1 trispecific 1923Ab16 (Figure 13H) comprises two Fabs from 1923Ab4 that bind to CD137, a human IgG1 Fc with L234A L235A mutations, two scFv fragments (VL precedes VH) derived from 1923Ab3 attached to the C-terminus of the 1923Ab4 heavy chain, and two polypeptides encoding the extracellular domain of TGFβRII attached to the C-terminus of the Ab4 light chain. Figure 14(B) provides a description of the heavy and light chains of 1923Ab16. Figure 15 provides a more detailed description of the subcomponents. Table 8 provides the amino acid sequences of the heavy and light chains, SEQ ID NO:48 and SEQ ID NO:49, respectively.

[0401] Methods for Producing Binding Proteins The binding proteins disclosed herein may be made by any method known in the art. For example, a recipient may be immunized with soluble recombinant human PD-L1 and / or CD137 protein, or fragments or peptides thereof conjugated to a carrier protein. Any suitable method of immunization may be used. Such methods may include the use of adjuvants, other immune stimulants, repeated booster immunizations, and one or more immunization routes.

[0402] Any suitable source of human PD-L1 and / or CD137 may be used as the immunogen for the generation of non-human or human anti-PD-L1 and / or CD137 antibodies in the compositions and methods disclosed herein.

[0403] Different forms of PD-L1 and / or CD137 antigens may be used to generate antibodies that are sufficient to generate biologically active antibodies. Thus, the inducing PD-L1 and / or CD137 antigen may be a single epitope, multiple epitopes, or the entire protein alone or in combination with one or more immunogenicity enhancing agents. In some embodiments, the inducing antigen is an isolated soluble full-length protein, or a soluble protein that comprises less than the full-length sequence (e.g., immunizing with a peptide that comprises a particular portion or epitope of PD-L1 and / or CD137). As used herein, the term "portion" refers to the minimum number of amino acids or nucleic acids that constitute an immunogenic epitope of the antigen of interest, as appropriate. Any genetic vector suitable for transformation of the cells of interest may be used, including, but not limited to, adenoviral vectors, plasmids, and non-viral vectors, such as cationic lipids.

[0404] It may be desirable to prepare monoclonal antibodies (mAbs) from a variety of mammalian hosts, e.g., mice, rodents, primates, humans, etc. Descriptions of techniques for preparing such monoclonal antibodies can be found, for example, in Sties et al. (eds.) BASIC AND CLINICAL IMMUNOLOGY (4 thed.) Lance Medical Publication, Los Altos, CA, and references cited therein; Harlow and Lane (1988) ANTIBODIES: A LABORATORY MANUAL CSH Press; Goding (1986) MONOCLONAL ANTIBODIES: PRINCIPLES AND PRACTICE (2 nd (ed.) Academic Press, New York, NY. Typically, spleen cells from animals immunized with a desired antigen are immortalized, commonly by fusion with myeloma cells. See Kohler and Milstein (196) Eur. J. Immunol. 6:511-519. Alternative methods of immortalization include transformation with Epstein-Barr virus, oncogenes, or retroviruses, or other methods known in the art. See, e.g., Doyle et al. (eds. 1994 and periodic supple supples) CELL and TISSUE CULTURE: LABORATORY PROCEDURES, John Wiley and Sons, New York, NY. Colonies arising from a single immortalized cell are screened for production of antibodies of the desired specificity and affinity for the antigen, and the yield of monoclonal antibodies produced by such cells can be enhanced by a variety of techniques, including injection into the peritoneal cavity of a vertebrate host. Alternatively, DNA sequences encoding monoclonal antibodies or antigen-binding fragments thereof may be isolated by screening a DNA library from human B cells, for example, following the general protocols outlined by Huse et al., (1989) Science 246:1275-1281. Antibodies may thus be obtained by a variety of techniques familiar to the skilled worker in the art.

[0405] Other suitable techniques include the selection of libraries of antibodies in phage, yeast, viral or similar vectors. See, for example, Huse et al. (supra); Ward et al., (1989) Nature 341:544-546. The polypeptides and antibodies disclosed herein can be used with or without modification, including chimeric or humanized antibodies. Frequently, the polypeptides and antibodies are labeled by covalently or non-covalently linking a substance which provides a detectable signal. A wide variety of labels and conjugation techniques are known and have been reported extensively in both the scientific and patent literature. Suitable labels include radionuclides, enzymes, substrates, cofactors, inhibitors, fluorescent moieties, chemiluminescent moieties, magnetic particles, and the like. Patents which teach the use of such labels include U.S. Patent Nos. 3,817,837; 3,850,752; 3,996,345; 4,277,437; 4,275,149; and 4,366,241. Recombinant immunoglobulins may also be produced, see Cabilly, U.S. Patent No. 4,816,567; and Queen et al. (1989) Proc. Nat'l Acad. Sci. USA 86:10029-10023; or made in transgenic mice. See Nils Lonberg et al., (1994), Nature 368:856-859; and Mendez et al. (1997) Nature Genetics 15:146-156; TRANSGENIC ANIMALS AND METHODS OF USE (WO2012 / 62118), Medarex, Trianni, Abgenix, Ablexis, OminiAb, Harbour and other technologies.

[0406] In some embodiments, the ability of the produced antibodies to bind PD-L1 and / or CD137 can be assessed using standard binding assays such as surface plasmon resonance (SPR), ELISA, Western blot, immunofluorescence, flow cytometry analysis, chemotaxis assays, and cell migration assays, etc. In some aspects, the produced antibodies can also be assessed for their ability to inhibit PD-L1 and / or block PD-L1 and / or activate CD137 receptor signaling and thereby activate CD137.

[0407] The antibody composition prepared from the cells can be purified using, for example, hydroxylapatite chromatography, gel electrophoresis, dialysis, and affinity chromatography, with affinity chromatography being a typical purification technique. The suitability of protein A as an affinity ligand depends on the species and isotype of any immunoglobulin Fc domain present in the antibody. Protein A can be used to purify antibodies based on human gamma 1, gamma 2, or gamma 4 heavy chains (see, e.g., Lindmark et al., 1983 J. Immunol. Meth. 62:1-13). Protein G is recommended for all mouse isotypes and for human gamma 3 (see, e.g., Guss et al., 1986 EMBO J. 5:1567-1575). The matrix to which the affinity ligand is attached is most often agarose, but other matrices are available. Mechanically stable matrices such as controlled pore glass or poly(styrenedivinyl)benzene allow for faster flow rates and shorter processing times than can be achieved with agarose. If the antibody contains a CH3 domain, Bakerbond ABX™ resin (JT Baker, Phillipsburg, NJ) is useful for purification. Other techniques for protein purification, such as fractionation on ion exchange columns, ethanol precipitation, reverse-phase HPLC, chromatography on silica, chromatography on heparin SEPHAROSE™ chromatography on anion or cation exchange resins (such as polyaspartic acid columns), chromatofocusing, SDS-PAGE, and ammonium sulfate precipitation, are also available depending on the antibody to be recovered.

[0408] Following any preliminary purification steps, the mixture containing the antibody of interest and contaminants may be subjected to low pH hydrophobic interaction chromatography, typically performed at low salt concentrations (e.g., about 0-0.25 M salt) and using an elution buffer at a pH between about 2.5-4.5.

[0409] Also included are nucleic acids that hybridize under low, medium, and high stringency conditions, as defined herein, to all or a portion (e.g., a portion encoding a variable region) of the nucleotide sequence represented by the isolated polynucleotide sequence encoding an antibody or antibody fragment of the disclosure. The hybridizing portion of the hybridizing nucleic acid is typically at least 15 (e.g., 20, 25, 30, or 50) nucleotides in length. The hybridizing portion of the hybridizing nucleic acid is at least 80%, e.g., at least 90%, at least 95%, or at least 98% identical to a portion or all of the sequence of a nucleic acid encoding an anti-PD-L1 and / or CD137 polypeptide (e.g., a heavy or light chain variable region), or its complement. Hybridizing nucleic acids of the type described herein can be used, for example, as cloning probes, primers, e.g., PCR primers, or diagnostic probes.

[0410] Polynucleotides, Vectors, and Cells Other embodiments include isolated polynucleotides comprising sequences encoding the binding proteins or fragments thereof disclosed herein, vectors and cells comprising the polynucleotides, and recombinant techniques for the production of the disclosed binding proteins. The isolated polynucleotides can encode any desired form of the binding protein, including, for example, full length monoclonal antibodies, Fab, Fab', F(ab')2, and Fv fragments, diabodies, linear antibodies, single chain antibody molecules, miniantibodies.

[0411] Some embodiments include an isolated polynucleotide comprising a sequence encoding a heavy chain variable region of a binding protein or fragment thereof having the amino acid sequence of any of SEQ ID NOs: 1, 3, 16, 18, and 20. Some embodiments include an isolated polynucleotide comprising a sequence encoding a light chain variable region of a binding protein or fragment thereof having the amino acid sequence of any of SEQ ID NOs: 2, 4, 17, 19, and 21.

[0412] In one embodiment, the isolated polynucleotide sequence encodes a binding protein or fragment thereof having a heavy chain variable region and a light chain variable region comprising the following amino acid sequences: (a) a heavy chain variable region sequence comprising SEQ ID NO:1 and a light chain variable region sequence comprising SEQ ID NO:2; (b) a heavy chain variable region sequence comprising SEQ ID NO:3 and a light chain variable region sequence comprising SEQ ID NO:4; (c) a heavy chain variable region sequence comprising SEQ ID NO:16 and a light chain variable region sequence comprising SEQ ID NO:17; (d) a heavy chain variable region sequence comprising SEQ ID NO:18 and a light chain variable region sequence comprising SEQ ID NO:19; or (e) a heavy chain variable region sequence comprising SEQ ID NO:20 and a light chain variable region sequence comprising SEQ ID NO:21.

[0413] In another embodiment, the isolated polynucleotide sequence encodes a binding protein or fragment thereof having a heavy chain variable region and a light chain variable region comprising the following amino acid sequences: (a) a heavy chain variable region sequence that is 90%, 95%, or 99% identical to SEQ ID NO:1 and a light chain variable region sequence that is 90%, 95%, or 99% identical to SEQ ID NO:2; (b) a heavy chain variable region sequence that is 90%, 95%, or 99% identical to SEQ ID NO:3 and a light chain variable region sequence that is 90%, 95%, or 99% identical to SEQ ID NO:4; (c) (d) a heavy chain variable region sequence that is 90%, 95% or 99% identical to SEQ ID NO:18 and a light chain variable region sequence that is 90%, 95% or 99% identical to SEQ ID NO:19; or (e) a heavy chain variable region sequence that is 90%, 95% or 99% identical to SEQ ID NO:20 and a light chain variable region sequence that is 90%, 95% or 99% identical to SEQ ID NO:21.

[0414] A polynucleotide comprising a sequence encoding a binding protein or fragment thereof disclosed herein can be fused to one or more regulatory or control sequences known in the art and can be included in a suitable expression vector or cell known in the art. Each of the polynucleotide molecules encoding a heavy or light chain variable domain can be fused independently to a polynucleotide sequence encoding a constant domain, such as a human constant domain, to allow for the production of an intact antibody. Alternatively, the polynucleotides, or portions thereof, can be fused together to provide a template for the production of a single chain antibody.

[0415] For recombinant production, a polynucleotide encoding the binding protein or a fragment thereof is inserted into a replicable vector for cloning (amplification of the DNA) or for expression. Many suitable vectors for expressing the binding protein or a fragment thereof are available. Vector components generally include, but are not limited to, one or more of the following: a signal sequence, an origin of replication, one or more marker genes, an enhancer element, a promoter, and a transcription termination sequence.

[0416] The binding proteins or fragments thereof disclosed herein can also be produced as fusion polypeptides, in which the binding protein is fused to a heterologous polypeptide, such as a signal sequence or other polypeptide having a specific cleavage site at the amino terminus of the mature protein or polypeptide. The heterologous signal sequence selected is typically one that is recognized and processed by the cell (i.e., cleaved by a signal peptidase). For prokaryotic cells, the signal sequence can be replaced by a prokaryotic signal sequence. The signal sequence can be, for example, alkaline phosphatase, penicillinase, lipoprotein, heat-stable enterotoxin II leaders, and the like. For yeast secretion, the native signal sequence can be replaced by a leader sequence derived, for example, from yeast invertase alpha factor (including Saccharomyces and Kluyveromyces alpha-factor leaders), acid phosphatase, C. albicans glucoamylase, or signals described in WO 90 / 13646. In mammalian cells, mammalian signal sequences as well as viral secretory leaders, such as the herpes simplex gD signal, can be used. The DNA for such precursor region is ligated in reading frame to DNA encoding the binding protein or fragment thereof.

[0417] Expression and cloning vectors contain a nucleic acid sequence that allows the vector to replicate in one or more selected cells. Generally, in cloning vectors, this sequence allows the vector to replicate independently of the host chromosomal DNA and includes an origin of replication or an autonomously replicating sequence. Such sequences are well known for a variety of bacteria, yeast, and viruses. The origin of replication from the plasmid pBR322 is suitable for most gram-negative bacteria, the 2μ plasmid origin is suitable for yeast, and various viral origins (SV40, polyoma, adenovirus, VSV, and BPV) are useful for cloning vectors in mammalian cells. Generally, an origin of replication is not required for mammalian expression vectors (the SV40 origin may typically be used because it contains the early promoter).

[0418] Expression and cloning vectors may contain a gene encoding a selectable marker to facilitate identification of expression. Typical selectable marker genes confer resistance to antibiotics or other toxins, such as ampicillin, neomycin, methotrexate, or tetracycline, or alternatively, complement auxotrophic deficiency, or in other alternatives, encode a protein that supplies a particular nutrient not present in complex media, such as the gene encoding D-alanine racemase for Bacillus.

[0419] cell culture The cells used to produce the binding proteins or fragments thereof disclosed herein may be cultured in a variety of media. Commercially available media such as Ham's F10 (Sigma), Minimum Essential Medium ((MEM), Sigma), RPMI-1640 (Sigma), FreeStyle™ (Cibco), and Dulbecco's Modified Eagle's Medium ((DMEM), Sigma) are suitable for culturing the host cells. Any of these or other media may be supplemented, if necessary, with hormones and / or other growth factors (such as insulin, transferrin, or epidermal growth factor), salts (such as sodium chloride, calcium, magnesium, and phosphate), buffers (such as HEPES), nucleotides (such as adenosine and thymidine), antibiotics (such as gentamicin), trace elements (such as inorganic compounds usually present at final concentrations in the micromolar or lower range), and glucose or an equivalent energy source. Any other necessary supplements may also be included at appropriate concentrations that would be known to those skilled in the art. Culture conditions, such as temperature, pH, and the like, will include those previously used with the cell selected for expression and will be apparent to one of skill in the art.

[0420] Non-therapeutic Use The binding proteins described herein are useful as affinity purification agents. In this process, the binding proteins are immobilized on a solid phase, such as a protein A resin, using methods well known in the art. The immobilized binding proteins are contacted with a sample containing the PD-L1, TGFβ, and / or CD137 protein (or fragments thereof) to be purified, and the support is then washed with a suitable solvent that removes substantially all of the material in the sample except for the PD-L1, TGFβ, and / or CD137 protein that is bound to the immobilized binding protein. Finally, the support is washed with another suitable solvent that releases the PD-L1, TGFβ, and / or CD137 protein from the binding protein.

[0421] The binding proteins disclosed herein are also useful in diagnostic assays for detecting and / or quantifying PD-L1, TGFβ, and / or CD137 proteins, e.g., for detecting PD-L1, TGFβ, and / or CD137 expression in specific cells, tissues, or serum. The binding proteins can be used diagnostically, e.g., to monitor disease onset or progression as part of a clinical testing procedure, e.g., to determine the effectiveness of a given therapeutic and / or prophylactic regimen. Detection can be facilitated by conjugating the binding protein to a detectable substance. Examples of detectable substances include various enzymes, prosthetic groups, fluorescent materials, luminescent materials, bioluminescent materials, radioactive materials, various positron emitting metals using positron emission tomography, and non-radioactive paramagnetic metal ions. See, e.g., U.S. Pat. No. 4,741,900 for metal ions that can be conjugated to the binding proteins for use as diagnostics according to the present disclosure.

[0422] The binding proteins can be used in methods for diagnosing a PD-L1, TGFβ, and / or CD137-associated disorder (e.g., a disorder characterized by aberrant expression of PD-L1, TGFβ, and / or CD137) or for determining whether a subject has an increased risk of developing a PD-L1, TGFβ, and / or CD137-associated disorder. Such methods include contacting a biological sample from a subject with a binding protein disclosed herein and detecting binding of the molecule to PD-L1, TGFβ, and / or CD137. By "biological sample" is intended any biological sample obtained from an individual, a cell line, tissue culture, or other source of cells potentially expressing PD-L1, TGFβ, and / or CD137. Methods for obtaining tissue biopsies and body fluids from mammals are well known in the art.

[0423] In some embodiments, the methods may further include comparing the levels of PD-L1, TGFβ, and / or CD137 in the patient sample to a control sample (e.g., a subject without a PD-L1, TGFβ, and / or CD137-associated disorder) to determine whether the patient has, or is at risk for developing, a PD-L1, TGFβ, and / or CD137-associated disorder.

[0424] In some embodiments, it may be advantageous to label the binding protein with a detectable moiety, for example, for diagnostic purposes. Numerous detectable labels are available, including radioisotopes, fluorescent labels, enzyme substrate labels, and the like. The label may be indirectly conjugated to the binding protein using a variety of known techniques. For example, the binding protein may be conjugated with biotin, and any of the three broad categories of labels mentioned above may be conjugated with avidin, or vice versa. Biotin selectively binds to avidin, and thus the label may be conjugated to the binding protein in this indirect manner. Alternatively, to achieve indirect conjugation of the label to the binding protein, the binding protein may be conjugated with a small hapten (such as digoxin), and one of the different types of labels mentioned above is conjugated with an anti-hapten antibody (e.g., anti-digoxin antibody). In this way, indirect conjugation of the label to the binding protein may be achieved.

[0425] Exemplary radioisotope labels include: 35 S, 14 C. 125 I, 3 H, and 131I. The binding protein can be labeled with a radioisotope using techniques described, for example, in Current Protocols in Immunology, Volumes 1 and 2, 1991, Coligen et al., Ed. Wiley-Interscience, New York, NY, Pubs. Radioactivity can be measured, for example, by scintillation counting.

[0426] Exemplary fluorescent labels include those derived from rare earth chelates (europium chelates) or are available as fluorescein and its derivatives, rhodamine and its derivatives, dansyl, Lissamine, phycoerythrin, and Texas Red. Fluorescent labels can be conjugated to binding proteins via known techniques, such as those disclosed in Current Protocols in Immunology. Fluorescence can be quantified using a fluorometer.

[0427] There are a variety of well-characterized enzyme-substrate labels known in the art (see, e.g., U.S. Pat. No. 4,275,149). The enzyme generally catalyzes a chemical change in a chromogenic substrate that can be measured using a variety of techniques. For example, the change can be a color change in the substrate that can be measured spectrophotometrically. Alternatively, the enzyme can change the fluorescence or chemiluminescence of the substrate. Techniques for quantifying the change in fluorescence are described above. The chemiluminescent substrate becomes electronically excited by a chemical reaction and can then emit light that can be measured using, for example, a chemiluminometer, or can donate energy to a fluorescent acceptor.

[0428] Examples of enzyme labels include luciferases, such as firefly luciferase and bacterial luciferase (U.S. Pat. No. 4,737,456), luciferin, 2,3-dihydrophthalazinedione, malate dehydrogenase, urease, peroxidases, such as horseradish peroxidase (HRPO), alkaline phosphatase, β-galactosidase, glucoamylase, lysozyme, sugar oxidases (such as glucose oxidase, galactose oxidase, and glucose-6-phosphate dehydrogenase), heterocyclic oxidases (such as uricase and xanthine oxidase), lactoperoxidase, microperoxidase, and the like. Techniques for conjugating enzymes to binding proteins are described, for example, in O'Sullivan et al., 1981, Methods for the Preparation of Enzyme-Antibody Conjugates for use in Enzyme Immunoassay, in Methods in Enzym. (J. Langone & H. Van Vunakis, eds.), Academic press, NY, 73:147-166.

[0429] Examples of enzyme-substrate combinations include, for example: horseradish peroxidase (HRPO) with hydrogen peroxidase as a substrate, in which hydrogen peroxidase oxidizes a dye precursor, such as orthophenylenediamine (OPD) or 3,3,5,5-tetramethylbenzidine hydrochloride (TMB); alkaline phosphatase (AP) with paranitrophenyl phosphate as a chromogenic substrate; and β-D-galactosidase (β-D-Gal) with a chromogenic substrate, such as p-nitrophenyl-β-D-galactosidase or the fluorogenic substrate 4-methylumbelliferyl-β-D-galactosidase.

[0430] In another embodiment, the binding proteins disclosed herein are used unlabeled and detected with a labeled antibody that binds to the binding protein.

[0431] The binding proteins described herein may be used in any known assay method, such as competitive binding assays, direct and indirect sandwich assays, and immunoprecipitation assays, see, e.g., Zola, Monoclonal Antibodies: A Manual of Techniques, pp. 147-158 (CRC Press, Inc. 1987).

[0432] The binding proteins disclosed herein can be used to inhibit the binding of PD-L1, TGFβ, and / or CD137 to their respective receptors. Such methods include administering a binding protein disclosed herein to a cell (e.g., a mammalian cell) or cellular environment, thereby inhibiting receptor-mediated signaling. These methods can be performed in vitro or in vivo. By "cellular environment" is intended the tissue, medium, or extracellular matrix surrounding the cell.

[0433] Compositions and methods of treatment The disclosure also provides compositions, including, for example, pharmaceutical compositions, that include the binding proteins disclosed herein. Such compositions have numerous therapeutic uses for the treatment, prevention, or amelioration of a disease or disorder, such as cancer.

[0434] The present disclosure also provides methods for the treatment or prevention of cancer comprising administering to a subject in need thereof a composition or formulation comprising the binding proteins disclosed herein, and, optionally, another immune-based treatment.

[0435] The disclosed binding proteins are also useful in methods of treatment of cancer, either alone (eg, as monotherapy) or in combination with other immunotherapeutic agents and / or chemotherapy.

[0436] The binding proteins can be administered alone or in combination with other compositions useful for treating immune-mediated inflammatory disorders or autoimmune diseases.

[0437] In some aspects, compositions, e.g., pharmaceutical compositions, are provided that include one or more of the binding proteins disclosed herein. Pharmaceutical compositions can be formulated using pharma- ceutical acceptable carriers or diluents and any other known adjuvants and excipients according to conventional techniques, such as those disclosed in Remington: The Science and Practice of Pharmacy, 19th Edition, Gennaro, Ed., Mack Publishing Co., Easton, Pa., 1995.

[0438] Typically, the composition for administration by injection is a solution in a sterile isotonic aqueous buffer. If necessary, the pharmaceutical agent may also include a solubilizing agent and a local anesthetic, such as lignocaine, to ease pain at the site of injection. Generally, the ingredients are supplied separately or mixed in unit dosage form as a dry lyophilized powder or water-free concentrate in a sealed container, such as an ampoule or sachet indicating the quantity of active agent. When the pharmaceutical agent is administered by injection, it may be dispensed in an infusion bottle containing sterile pharmaceutical grade water or saline. When the pharmaceutical agent is administered by injection, an ampoule of sterile water for injection or saline may be provided so that the ingredients can be mixed prior to administration.

[0439] As used herein, "pharmaceutical acceptable carriers" include any and all solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic and absorption delaying agents, and the like that are physiologically compatible. Preferably, the carrier is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal or epidermal administration (e.g., by injection or infusion). Depending on the route of administration, the active compounds, i.e., antibodies, bispecific and multispecific molecules, may be coated in a material to protect the compound from the action of acids and other natural conditions that may inactivate the compound.

[0440] The composition can be administered by various methods known in the art. As will be appreciated by those skilled in the art, the route and / or mode of administration varies depending on the desired results. The active compound can be prepared with a carrier that protects the compound against rapid release, such as controlled release formulations, including implants, transdermal patches, and microencapsulated delivery systems. Biodegradable, biocompatible polymers can be used, such as ethylene vinyl acetate, polyanhydrides, polyglycolic acid, collagen, polyorthoesters, and polylactic acid. Methods for the preparation of such formulations are generally known to those skilled in the art. For example, see Sustained and Controlled Release Drug Delivery Systems, JR Robinson, ed., Marcel Dekker, Inc., New York, 1978.

[0441] The dosage level of the active ingredient in the pharmaceutical composition can be varied to obtain an amount of the active ingredient effective to achieve the desired therapeutic response for a particular subject, composition, and mode of administration without being toxic to the subject. The selected dosage level will depend on various pharmacokinetic factors, including the activity of the particular composition used, the route of administration, the time of administration, the rate of excretion of the particular compound used, the duration of treatment, other drugs, compounds and / or materials used in combination with the particular composition used, the age, sex, weight, condition, general health and previous medical history of the patient being treated, and similar factors well known in the medical arts.

[0442] The pharmaceutical compositions described herein may be administered in an effective amount. "Effective amount" refers to an amount that achieves the desired response or the desired effect, either alone or together with further doses. In the case of the treatment of a particular disease or condition, the desired response preferably relates to the inhibition of the course of the disease. This includes slowing the progression of the disease, and in particular, halting or reversing the progression of the disease.

[0443] In some aspects, the compositions described herein are administered to a patient, e.g., in vivo, to treat or prevent various disorders, such as those described herein. Preferred patients include human patients having a disorder that can be corrected or ameliorated by administering a binding protein disclosed herein.

[0444] In some aspects, conventional viral and non-viral based gene transfer methods can be used to introduce nucleic acids encoding the antibodies or derivatives thereof described herein in mammalian cells or target tissues. Such methods can be used to administer nucleic acids encoding antibodies to cells in vitro. In some embodiments, nucleic acids encoding antibodies or derivatives thereof are administered for in vivo or ex vivo gene therapy uses. In other embodiments, gene delivery techniques are used to test the activity of antibodies in cell-based or animal models. Non-viral vector delivery systems include DNA plasmids, naked nucleic acids, and nucleic acids complexed with delivery vehicles such as liposomes. Viral vector delivery systems include DNA and RNA viruses, which have either episomal or integrated genomes after delivery to cells. Such methods are well known in the art.

[0445] Methods of non-viral delivery of nucleic acids encoding engineered polypeptides of the present disclosure include lipofection, microinjection, biolistics, virosomes, liposomes, immunoliposomes, polycation or lipid:nucleic acid conjugates, naked DNA, artificial virions, and drug-enhanced uptake of DNA. Lipofection methods and lipofection reagents are well known in the art (e.g., Transfectam™ and Lipofectin™). Cationic and neutral lipids that are suitable for efficient receptor-recognition lipofection of polynucleotides include the lipids of Felgner, WO91 / 17424, WO91 / 16024. Delivery can be to cells (ex vivo administration) or target tissues (in vivo administration). Preparation of lipid:nucleic acid complexes, including targeted liposomes, such as immunolipid complexes, is well known to those of skill in the art.

[0446] The use of RNA or DNA virus-based systems for delivery of nucleic acids encoding the antibodies described herein takes advantage of the highly evolved process for targeting viruses to specific cells in the body and transporting the viral payload to the nucleus. Viral vectors can be administered directly to the patient (in vivo) or they can be used to treat cells in vitro and the modified cells are administered to the patient (ex vivo). Conventional virus-based systems for delivery of the polypeptides of the present disclosure can include retroviral, lentiviral, adenoviral, adeno-associated and herpes simplex virus vectors for gene transfer. Viral vectors are currently the most efficient and versatile method of gene transfer in target cells and tissues. Integration in the host genome is possible with retroviral, lentiviral and adeno-associated virus gene transfer methods, often resulting in long-term expression of the inserted transgene. In addition, high transduction efficiency has been observed in many different cell types and target tissues. All patents and publications identified are expressly incorporated herein by reference for the purpose of describing and disclosing, for example, the methodologies described in such publications that may be used in connection with the present disclosure. These publications are provided solely for their disclosure prior to the filing date of this application. Nothing in this regard should be construed as an admission that the inventors are not entitled to antedate such disclosure by virtue of prior invention or for any other reason. All statements as to dates or representations as to the contents of these documents are based on the information available to the applicant and do not constitute any admission as to the accuracy of the dates or contents of these documents. EXAMPLES

[0447] Common methods Methods for protein purification, including immunoprecipitation, chromatography, and electrophoresis, are described. See, e.g., Coligan et al. (2000) Current Protocols in Protein Science, Vol. 1, John Wiley and Sons, Inc., New York. Chemical analysis, chemical modifications, post-translational modifications, production of fusion proteins, and glycosylation of proteins are described. See, e.g., Coligan et al. (2000) Current Protocols in Protein Science, Vol. 2, John Wiley and Sons, Inc., New York; Ausubel et al. (2001) Current Protocols in Molecular Biology, Vol. 3, John Wiley and Sons, Inc., NY, NY, pp. 16.0.5-16.22.17; Sigma-Aldrich, Co. (2001) Products for Life Science Research, St. Louis, Mo.; pp. 45-89; Amersham Pharmacia Biotech (2001) BioDirectory, Piscataway, NJ, pp. 384-391. Production, purification, and fragmentation of polyclonal and monoclonal antibodies are described. Coligan et al. (2001) Current Protocols in Immunology, Vol. 1, John Wiley and Sons, Inc., New York; Harlow and Lane (1999) Using Antibodies, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; Harlow and Lane, supra.

[0448] Hybridoma or cell culture supernatants containing anti-PD-L1 or anti-CD137 antibodies were purified via HiTrap Protein G columns (GE, Cat. No. 17040401) according to the manufacturer's procedure. Briefly, the supernatants were equilibrated with DPBS (Gibco, Cat. No. 14190-136) for 5 CV and loaded via syringe / infusion pump (Legato 200, KDS) at ambient temperature and 3 min residence time. The column was washed with 5 CV of DPBS and elution was performed with 4 CV of pH 2.8 elution buffer (Fisher Scientific, Cat. No. PI21004). The eluate was fractionated and fractions were neutralized with 1 M Tris-HCL, pH 8.5 (Fisher Scientific, Cat. No. 50-843-270) and assayed by A280 (Dropsensiti96, Trinean). Peak fractions were pooled and buffer exchanged into DPBS. Centrifugal filters (EMD Millipore, Cat. No. UFC803024) were equilibrated in DPBS for 2 minutes at 4,000×g. Purified samples were loaded, DPBS was added, and samples were spun at 4,000×g for 5-10 minutes until the total DPBS volume reached ≧6 DV. The final pool was analyzed by A280.

[0449] Standard methods in molecular biology are described. See, for example, Maniatis et al. (1982) Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; Sambrook and Russell (2001) Molecular Cloning, 3rd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY; Wu (1993) Recombinant DNA, Vol. 217, Academic Press, San Diego, Calif. Standard methods can also be found in Ausbel et al. (2001) Current Protocols in Molecular Biology, Vols. 1-4, John Wiley and Sons, Inc. New York, NY, which describes cloning and DNA mutagenesis in bacterial cells (Vol. 1), cloning in mammalian cells and yeast (Vol. 2), glycoconjugates and protein expression (Vol. 3), and bioinformatics (Vol. 4).

[0450] Stable cell lines expressing human PD-L1 and CD137 were generated by transfecting selected host cells (i.e., CHO-K1 or HEK293) with pcDNA3.1-based plasmids expressing Homo sapiens target proteins using electroporation or lipid-based transfection. Integrator cells were selected using geneticin or puromycin. After 7-10 days of antibiotic selection, stable clones were isolated by FACS using labeled antibodies or serial dilution. After expansion, stable clones were further confirmed for target protein expression by flow cytometry. Mouse and cynomolgus targets were each transiently expressed in HEK293T cells using lipid-based transfection.

[0451] The sequences of the heavy and light chain variable regions for the hybridoma clones were determined as described below. Total RNA was purified by dilution from 1-2×10 6 The hybridoma cells were extracted from the 1000-kDa hybridoma. CDNA was generated by performing a 5'RACE reaction using the SMARTer RACE 5' / 3' kit from Takara (Mountainview, CA, USA). PCR was performed using Q5 High-Fidelity DNA Polymerase from NEB (Ipsowich, MA, USA) to amplify the variable regions from the heavy and light chains using the Takra Universal Primer Mix in combination with gene-specific primers for the 3' mouse constant regions of the appropriate immunoglobulins. The amplified variable regions for the heavy and light chains were run on a 2% agarose gel, the appropriate bands were excised, and then gel purified using the Mini Elute Gel Extraction Kit from Qiagen. The purified PCR products were cloned using the Zero Blunt PCR Cloning Kit from Invitrogen (Carlsbad, CA, USA), transformed into Stella Competent E.Coli cells from Tarora, and plated on LB Agar+50ug / ml kanamycin plates. Direct colony Sanger sequencing was performed by GeneWiz (South Plainfield, NJ, USA). The resulting polynucleotide sequences were analyzed using IMGT V-QUEST to identify productive rearrangements and to analyze the translated protein sequences. CDR determination was based on Kabat numbering.

[0452] The selected VH or VL chains were PCR amplified and cloned into pcDNA3.4-based expression vectors, which carry the constant regions from human IgG1 (Uniprot P01857) or human kappa light chain (UniProt P01834). Paired heavy and light chain expression plasmids were transfected into Expi293 cells (Thermo Fisher Scientific) according to the supplier's Expi293 expression system protocol. Five days after transfection, culture supernatants were collected by centrifugation. The expressed antibodies were purified by one-step affinity purification using a Protein A column and buffer exchanged into PBS pH 7.2.

[0453] Methods for flow cytometry are available, including the Fluorescence Activated Cell Sorting Detection System (FACS®). See, e.g., Owens et al. (1994) Flow Cytometry Principles for Clinical Laboratory Practice, John Wiley and Sons, Hoboken, NJ; Givan (2001) Flow Cytometry, 2nd ed.; Wiley-Liss, Hoboken, NJ; Shapiro (2003) Practical Flow Cytometry, John Wiley and Sons, Hoboken, NJ. Suitable fluorescent reagents are available for modifying nucleic acids, including, for example, nucleic acid primers and probes, polypeptides, and antibodies for use as diagnostic reagents. Molecular Probes (2003) Catalogue, Molecular Probes, Inc., Eugene, Oreg.; Sigma-Aldrich (2003) Catalogue, St. Louis, Mo.

[0454] Standard techniques are available for characterizing ligand / receptor interactions. See, for example, Coligan et al. (2001) Current Protocols in Immunology, Vol. 4, John Wiley, Inc., New York. Standard methods of antibody function characterization suitable for characterizing antibodies with specific mechanisms of action are well known to those of skill in the art.

[0455] For example, software packages and databases are available for determining antigenic fragments, leader sequences, protein folding, functional domains, CDR annotations, glycosylation sites, and sequence alignments.

[0456] An autologous anti-PD-L1 antibody based on the anti-PD-L1 antibody (avelumab), referred to herein as "avelumab-NR" (PC1), was prepared based on the publicly available information published in US 10,759,856 (VH SEQ ID NO:24 and VL SEQ ID NO:25 therein). The PC1 antibody was used to establish functional assays used to evaluate and characterize the anti-PD-L1 specific antibodies disclosed herein.

[0457] An autologous anti-CD137 antibody based on the anti-CD137 antibody (Urelumab), referred to herein as "Urelumab-NR" (PC2), was prepared based on the publicly available information published in US 7,288,638 (VH SEQ ID NO: 3 and VL SEQ ID NO: 6 therein). A second autologous CD137-reactive antibody (Utomilumab), referred to herein as "Utomilumab-NR" (PC3), was prepared based on the publicly available information published in US 8,337,850 (VH SEQ ID NO: 43 and VL SEQ ID NO: 45 therein). The PC2 and PC3 antibodies were used to confirm CD137 expression in the cell lines used in the examples and to establish binding and functional assays used to evaluate and characterize the anti-CD137-specific antibodies disclosed herein. Examples 18 and 22 in the present application demonstrate that the disclosed bispecifics and trispecifics induced stronger CD137 signaling and T cell activation compared to urelumab-NR and utomilumab-NR.

[0458] The reference sequences utilized herein are illustrated in Table 9. [Table 9]

[0459] Example 1: Generation of a binding protein that binds to PD-L1 Fully human anti-human PD-L1 antibodies were generated by immunizing human Ig transgenic mice, Trianni mice, which express human antibody VH and VL genes (see, e.g., WO2013 / 063391, TRIANNI® mice).

[0460] immunization - TRIANNI mice, as described above, were immunized by injection with recombinant human PD-L1 protein via intraperitoneal (IP), subcutaneous (SC), tail base or footpad injection.

[0461] Immune responses were monitored by retro-orbital bleeds. Plasma was screened by ELISA, flow cytometry (FACS), or imaging (as described below). Mice with sufficient anti-PD-L1 titers were used for fusions. Mice were boosted intraperitoneally, at the base of the tail, in the footpad, or intravenously with immunogen before sacrifice and removal of the spleen and lymph nodes.

[0462] Selection of mice producing anti-PD-L1 antibodies - To select mice producing antibodies that bind to PD-L1, sera from immunized mice are screened for binding to human PD-L1 protein by ELISA, FACS, or imaging in cells expressing PD-L1 protein (HEK293T cells transfected with the PD-L1 gene) and not control cells that do not express PD-L1 (HEK293T cells).

[0463] For ELISA, briefly, ELISA plates coated with recombinant human PD-L1 (AcroBiosystems, Catalog No: PD1-H5229) were incubated with dilutions of sera from immunized mice for 1 hour at room temperature, the assay plates were washed, and specific antibody binding was detected with HRP-labeled anti-mouse IgG antibody (Jackson ImmunoResearch, Catalog No: 109-035-088) after 1 hour incubation at room temperature, washed, and followed by incubation with ABTS substrate (Moss, Catalog No: ABTS-1000) for 30 minutes at room temperature. Plates were read using an ELISA plate reader (Biotek).

[0464] For FACS, briefly, PD-L1-HEK293T cells or parental HEK293T cells were incubated with dilutions of serum from immunized mice for 2 hours at 4°C. Cells were fixed with 2% PFA (Alfa Aesar, Cat. No. J61899) for 15 minutes at 4°C and then washed. Specific antibody binding was detected using Alexa 647-labeled goat anti-mouse IgG antibody (ThemoFisher Scientific, Cat. No. A-21445) after 1 hour incubation at 4°C. Flow cytometry analysis was performed on a flow cytometry instrument (Intellicyte, IQue plus, Sartorius).

[0465] Mouse sera were also tested by imaging. Briefly, PD-L1-HEK293T cells were incubated with dilutions of sera from immunized mice. Cells were washed, fixed with paraformaldehyde, washed, and specific antibody binding was detected with secondary Alexa488 goat anti-mouse antibody and Hoechst (Invitrogen). Plates were scanned and analyzed on an imager (Cytation 5, Biotek).

[0466] Generation of hybridomas producing antibodies against PD-L1-To generate hybridomas producing the human antibodies of the invention, splenocytes and lymph node cells were isolated from immunized mice and fused to a suitable immortalized cell line, such as a mouse myeloma cell line. The resulting hybridomas were screened for production of antigen-specific antibodies. For example, single cell suspensions of splenocytes, lymph node cells from immunized mice were fused to an equal number of Sp2 / 0 non-secreting mouse IgG myeloma cells (ATCC, CRL 1581) by electrofusion. Cells were seeded into flat-bottom 96-well tissue culture plates, followed by incubation in selection medium (HAT medium) for about one week, then switched to hybridoma culture medium. Approximately 10-14 days after cell seeding, supernatants from individual wells were screened by ELISA, imaging, or FACS as described above. Antibody-secreting hybridomas were transferred to 24-well plates, screened again, and if still positive for anti-PD-L1, positive hybridomas were subcloned by sorting using a single cell sorter. The subclones were screened again by ELISA, imaging or FACS as described above. Stable subclones were then cultured in vitro to generate small amounts of antibody for purification and characterization.

[0467] Example 2: Binding specificity of mAbs that bind PD-L1 to human, mouse, and cynomolgus PD-L1 proteins The binding specificity of the disclosed anti-PD-L1 antibodies (1923Ab2 and 1923Ab3) to PD-L1 proteins of different species was assessed by ELISA. Briefly, human recombinant PD-L1 protein (Acro Biosystems, Catalog No.: PD1-H5229), cynomolgus monkey PD-L1 protein (Acro Biosystems, Catalog No.: PD1-C52H4), and mouse PD-L1 protein (Acro Biosystems, Catalog No.: PD1-M5220) were directly coated onto an ELISA plate. 1923Ab2 and 1923Ab3 were then added to the plate and incubated with peroxidase AffiniPure F(ab') 2Fragmented goat anti-mouse IgG, Fcγ fragment specific (Jackson ImmunoResearch, Cat. No. 115-036-071) or Peroxidase AffiniPure F(ab') 2 Detection followed with fragmented goat anti-human IgG, Fcγ fragment specific (Jackson ImmunoResearch, Cat# 109-036-098). After addition of ABTS substrate (Moss, Cat#: ABTS-1000), ELISA plates were read using a Synergy Neo2 multimode reader (Bioteck, SN#: 180213F).

[0468] Figures 2A&B show the binding activity of the disclosed anti-PD-L1 antibodies to human and cynomolgus monkey PD-L1 proteins, but not to mouse PD-L1, in a dose-dependent manner; isotype controls mIgG or hIgG1 do not bind to PD-L1 of any species. The ELISA binding EC50 values ​​of the disclosed anti-PD-L1 antibodies to human and cynomolgus monkey PD-L1 proteins are provided in Figure 2A and Figure 2B.

[0469] Example 3: Binding affinity of recombinant anti-PD-L1 to human PD-L1 Recombinant anti-PD-L1 antibodies were expressed and purified from Expi293 using human IgG1 or human IgG1 variant constant regions. The binding affinity of the anti-PD-L1 antibodies was assessed using an immunofluorescence imaging assay.

[0470] The cell binding affinity of anti-PD-L1 antibodies was tested on HEK293T-PD-L1 cells. Cells were seeded in complete medium containing DMEM with 10% FBS and then incubated at 37℃ overnight. Anti-PD-L1 antibodies were serially diluted and added to the assay plate and incubated at 4℃ for 2 hours, then cells were fixed at room temperature for 15 minutes. Fixed cells were washed three times with PBS and then stained with Alexa Fluor® 488 goat anti-human IgG (H+L) secondary antibody (Invitrogen, Cat. No. A-11013) for detection at room temperature for 1 hour. Binding signals were assessed by imaging the cells and quantifying the fluorescence intensity using Cytation (Biotek, VT).

[0471] The binding results showed that 1923Ab3 exhibited strong binding to PD-L1 expressed on the cell surface. The binding affinity of 1923Ab3 is similar to that of the reference antibody atezolizumab (Roche). The binding EC50 of 1923Ab3 and atezolizumab to human PD-L1 was determined to be 0.18 nM and 0.21 nM, respectively (see Figure 3).

[0472] Example 4: Effect of PD-L1 antibodies on PD-1 / PD-L1 interaction The effect of PD-L1 antibodies on the interaction of PD-L1 and PD-LPD-1 was determined by a PD-1 / PD-L1 inhibition bioassay developed by Promega (Madison, USA). This assay is a luciferase cell-based assay consisting of two genetically engineered cell lines. The PD-1 effector cells are Jurkat T cells expressing human PD-1 on the cell surface and a stably integrated luciferase reporter driven by the NFAT response element (NFAT-RE), and the artificial APC cells are CHO-K1 cells expressing cell surface human PD-L1 and an engineered cell surface protein designed to activate the cognate TCR in an antigen-independent manner. When these two cell types are co-cultured, the PD-1 / PD-L1 interaction inhibits TCR signaling, resulting in a decrease in the luminescent signal. The addition of an antibody that blocks the PD-1 / PD-L1 interaction removes the inhibitory signal, leading to activation of the TCR and increased luminescence.

[0473] Artificial APC CHO-K1 cells (Promega, Catalog No. J109A) were cultured using Ham's F-12K medium (ThermoFisher, Catalog No. 21127022) containing 10% heat-inactivated fetal bovine serum (Sigma, Catalog No. 17H165) according to the manufacturer's protocol. These artificial APC cells were seeded into 384-well white TC-treated plates (Corning, Catalog No. 3570). Plates were incubated at 37°C, 5% CO 2 The cells were incubated at 37°C for 16 hours. The supernatant was removed and serially diluted antibodies were added. PD-1 effector cells (Promega, Cat#: J115A) were cultured using RPMI-1640 (ThermoFisher, Cat#: 11875-085) containing 10% heat-inactivated fetal bovine serum according to the manufacturer's protocol. The effector cells were added to white 384-well plates containing the artificial APC cells and antibodies. The plates were incubated at 37°C, 5% CO 2Plates were incubated at RT for 6 hours. After equilibration to room temperature, One-Glo Luciferase Reagent (Promega, Cat# E6130) was added to each well. Plates were then incubated at room temperature for 5 minutes and luminescence was measured using a Synergy Neo2 plate reader (Biotek). Data was analyzed with GraphPad Prism software.

[0474] Figure 4 shows that both 1923Ab2 and 1923Ab3 antibodies efficiently inhibited PD1 / PD-L1 interaction, resulting in the restoration of luminescent signal in the assay. The IC50 of 1923Ab2, 1923Ab3 and atezolizumab for PD1 / PD-L1 blockade was determined to be 0.90 nM, 0.43 nM, and 0.29-0.31 nM, respectively.

[0475] Example 5: Generation of a binding protein that binds to CD137 Fully human anti-human CD137 antibodies were generated by immunizing human Ig transgenic mice, Trianni mice expressing human antibody VH and VL genes (see, e.g., WO2013 / 063391, TRIANNI® mice).

[0476] Immunization-TRIANNI mice, as described above, were immunized by injection with an immunogen that contained HEK293 cells stably transfected with the human CD137 gene and recombinant human CD137 ECD protein.TRIANNI mice were immunized via intraperitoneal (IP), subcutaneous (SC), tail base or footpad injection.

[0477] Immune responses were monitored by retro-orbital bleeding. Plasma was screened by ELISA, flow cytometry (FACS), or imaging (as described below). Mice with sufficient anti-CD137 titers were used for fusions. Mice were boosted intraperitoneally with immunogen at the base of the tail or in the footpad before sacrifice and removal of the spleen and lymph nodes.

[0478] Selection of mice producing anti-CD137 antibodies - To select mice producing antibodies that bind to CD137, sera from immunized mice are screened for binding to human CD137 protein by ELISA, FACS, or imaging in cells expressing CD137 protein (HEK293T transfected with the CD137 gene) and not control cells that do not express CD137 (HEK293T cells).

[0479] For ELISA, briefly, recombinant human CD137 (R&D, Cat. No. 9220-4B) coated ELISA plates were incubated with dilutions of serum from immunized mice for 1 hour at room temperature, the assay plates were washed, and specific antibody binding was detected with HRP-labeled anti-mouse IgG antibody (Jackson ImmunoResearch, Cat. No. 109-035-088) after 1 hour incubation at room temperature, washed, and followed by incubation with ABTS substrate (Moss, Cat. No. ABTS-1000) for 30 minutes at room temperature. Plates were read using an ELISA plate reader (Biotek).

[0480] For FACS, briefly, CD137-HEK293T cells or parental HEK293T cells were incubated with dilutions of serum from immunized mice for 2 hours at 4°C. Cells were fixed with 2% PFA (Alfa Aesar, Cat. No. J61899) for 15 minutes at 4°C and then washed. Specific antibody binding was detected with Alexa 647-labeled goat anti-mouse IgG antibody (ThemoFisher Scientific, Cat. No. A-21445) after 1 hour of incubation at 4°C. Flow cytometry analysis was performed on a flow cytometry instrument (Intellicyte, IQue plus, Sartorius).

[0481] Mouse sera were also tested by imaging. Briefly, CD137-HEK293T cells were incubated with dilutions of sera from immunized mice. Cells were washed, fixed with paraformaldehyde, washed, and specific antibody binding was detected with secondary Alexa488 goat anti-mouse antibody and Hoechst (Invitrogen). Plates were scanned and analyzed on an imaging device (Cytation 5, Biotek).

[0482] Generation of hybridomas producing antibodies against CD137 -To generate hybridomas producing human antibodies of the invention, splenocytes and lymph node cells were isolated from immunized mice and fused to a suitable immortalized cell line, such as a mouse myeloma cell line. The resulting hybridomas were screened for production of antigen-specific antibodies. For example, single cell suspensions of splenocytes, lymph node cells from immunized mice were fused to an equal number of Sp2 / 0 non-secreting mouse IgG myeloma cells (ATCC, CRL 1581) by electrofusion. Cells were seeded into flat-bottom 96-well tissue culture plates, followed by incubation in selection medium (HAT medium) for about one week, then switched to hybridoma culture medium. Approximately 10-14 days after cell seeding, supernatants from individual wells were screened by imaging or FACS as described above. Antibody-secreting hybridomas were transferred to 24-well plates, screened again, and if still positive for anti-CD137, positive hybridomas were subcloned by sorting using a single cell sorter. Subclones were screened again by imaging or FACS as described above. Stable subclones were then cultured in vitro to generate small amounts of antibody for purification and characterization.

[0483] Example 6: Binding affinity of recombinant anti-CD137 antibodies to human, mouse and cynomolgus CD137 Analysis of binding affinity of anti-CD137 mAbs using immunofluorescence imaging assay was performed using HEK293T cells stably transfected with human, mouse or cynomolgus CD137 expression constructs. These cell lines express species-specific forms of CD137 protein on the cell surface. Cells were seeded in complete medium containing DMEM with 10% FBS and then incubated overnight at 37°C. Cells were stained with serial dilutions of anti-CD137 for 2 hours at 4°C, and then cells were fixed for 15 minutes at room temperature. Fixed cells were washed three times with PBS and then stained with Alexa Fluor® 488 goat anti-human IgG (H+L) secondary antibody for detection (Invitrogen, Cat. No. A-11013) for 1 hour at room temperature. Binding signals were evaluated by imaging cells and quantifying fluorescence intensity using Biotek Cytation.

[0484] The results show that 1923Ab4 (FIG. 5A), 1923Ab5 (FIG. 5B), and 1923Ab6 (FIG. 5C) antibodies efficiently bound to human and cynomolgus CD137 on the cell surface. However, none of them showed binding to mouse CD137. The binding EC50 of 1923Ab4, 1923Ab5, and 1923Ab6 to human CD137 was determined as 0.29 nM, 1.62 nM, and 10.5 nM, respectively. The binding EC50 of 1923Ab4, 1923Ab5, and 1923Ab6 to cynomolgus CD137 was determined as 0.22 nM, 2.77 nM, and 4.23 nM, respectively.

[0485] Experiments were also performed to compare the binding affinity of 1923Ab4, 1923Ab5, and 1923Ab6 antibodies with reference antibodies, including PC2 Urelumab-NR (BMS) and PC3 Utomilumab-NR (Pfizer). 1923Ab4 antibody showed similar binding affinity to human CD137 as Urelumab-NR and Utomilumab-NR (Figure 6). 1923Ab5 and 1923Ab6 antibodies showed weaker binding compared to the reference antibodies.

[0486] Example 7: CD137 Ligand Competition To evaluate the ability of the disclosed anti-CD137 antibodies to block CD137 ligand binding to CD137, the disclosed antibodies 1923Ab4, 1923Ab5, and 1923Ab6, as well as two reference antibodies PC2 Urelumab-NR and PC3 Utomilumab-NR, were tested by biolayer interferometry (Gator Bio, CA). Urelumab-NR and Utomilumab-NR are reported to be nonligand and ligand-blocking antibodies, respectively.

[0487] Briefly, streptavidin probes (Probe Life, Catalog No. PL168-1600002) were first added to a 96-well plate containing assay buffer (PBS with 0.02% Tween20 and 0.05% sodium azide) for 30 seconds (baseline step). The probes were then added to a 96-well plate containing CD137L His-Avi-Tag protein (BPS Bioscience, Catalog No. 100238; 10ug / ml) for 180 seconds (filling step, to capture biotin-CD137L), followed by a 30-second baseline step. The CD137L-loaded probes were then bound to CD137 protein (R&D, Catalog No. 9220-4B, 10ug / ml) for 180 seconds, followed by association with the disclosed or reference antibodies at a concentration of 10ug / ml for 180 seconds.

[0488] The data was processed using software provided by the manufacturer (Gator Bio, CA). 1923Ab5, 1923Ab6, and Urelumab-NR bound to the CD137 / CD137L complex added on the probe. However, 1923Ab4 and Utomilumab-NR did not show binding to the CD137 / CD137L complex added on the probe. These results implied that 1923Ab4 and Utomilumab-NR had ligand blocking activity. 1923Ab5, 1923Ab6, and Urelumab-NR bound to regions of CD137 that are not located in the ligand binding site (Figure 7).

[0489] Example 8: Cross-linking dependent agonist activity of anti-CD137 antibodies in NFκB luciferase reporter assay The agonist activity of the antibodies was evaluated using an NFκB luciferase reporter assay. 293T cells stably transfected with a human CD137 expression plasmid and an NFκB luciferase reporter plasmid were used to measure the agonist activity of anti-CD137 antibodies. These reporter cells (HEK-CD137 reporter cells) were stimulated with either an anti-CD137 antibody or a mixture containing the tested anti-CD137 antibody and a 3:1 ratio of a bridging antibody (anti-human Fcγ fragment specific (Jackson ImmunoResearch Lab, Cat. No.: 109-005-098) and incubated at 37°C, 5% CO 2 Plates were incubated at RT for 16 hours. ONE-Glo™ Luciferase Reagent (Promega, Cat#: E6130) was added and plates were incubated at room temperature for 10 minutes. Luminescence signals were measured using a Synergy Neo2 plate reader (Biotek) and data was analyzed using GraphPad Prism.

[0490] PC3 Utomirumab-NR was used as a control antibody, which has been reported to be a cross-linking dependent agonist antibody for CD137 signaling. As shown in FIG. 8, compared to the isotype control, all tested antibodies showed very minimal agonist activity in the absence of cross-linking antibody. However, all tested antibodies strongly activated NFκB luciferase gene expression, except for the isotype control cross-linked by anti-human Fcγ. The results demonstrate that the agonist activity of 1923Ab4, 1923Ab5, and 1923Ab6 antibodies is cross-linking dependent.

[0491] Example 9: Cross-linking-dependent agonist activity of anti-CD137 antibodies in human primary T cell activation assays The agonist activity of the antibodies was further confirmed in a T cell activation assay. Human PBMCs were prepared from healthy donors. These human PBMCs were cultured at 1×10 in RPMI1640 medium supplemented with 10% FBS and 0.5ug / ml of mouse anti-hCD3 clone OKT3 (Biolegend, Cat. No. 317325). 6 The plates were cultured at a density of 100 cells / mL. T cells were stimulated by adding either anti-CD137 antibody or a mixture containing the tested anti-CD137 antibody and a 3:1 ratio of cross-linking antibody. The plates were incubated at 37°C in 5% CO 2 After 72 hours of incubation, the supernatants were used to measure secreted IFNγ by AlphaLISA (PerkinElmer, Cat. No. AL217C / F) using the protocol according to the manufacturer's instructions.

[0492] PC3 Utomirumab-NR was used as a positive control antibody, which has been reported to be a cross-linking-dependent agonist antibody for T cell activation. As shown in FIG. 9, compared to the isotype control, PBMCs treated with the disclosed antibodies did not increase IFNγ production in the absence of cross-linking antibody. However, all tested antibodies, except for the isotype control, cross-linked by anti-human Fcγ strongly stimulated IFNγ production. This result demonstrates that 1923Ab4, 1923Ab5, and 1923Ab6 antibodies activated T cells in a cross-linking-dependent manner.

[0493] Example 10: Identification of the binding epitope region of 1923Ab4 on human CD137 The extracellular region of CD137 contains four cysteine-rich domains (CRD1-4) that are conserved among species. To identify which CRD domains are required for binding to the 1923Ab4 antibody, human / mouse hybrid CD137 expression constructs were generated by replacing individual human CRD domains with their mouse counterparts (Figure 10A). Example 6 shows that 1923Ab4 binds to human CD137 on the cell surface, but not to mouse CD137.

[0494] Analysis of anti-CD137 1923Ab4 binding to HEK293T cells transiently transfected with either human CD137 (WT) or human / mouse hybrid CD137 expression construct (msCRD1-4) was performed by flow cytometry-based binding assay. Cells were stained with 1923Ab4 for 2 hours at 4°C, followed by fixing the cells for 15 minutes at room temperature. Fixed cells were washed three times with PBS, followed by staining with Alexa Fluor® 488 goat anti-human IgG antibody (Invitrogen, Cat. No. A-11013) for detection for 1 hour at room temperature. Binding signals were assessed by quantifying the fluorescence intensity using iQue Screener PLUS (Sartorius, MI).

[0495] Both PC2 Urelumab-NR and PC3 Utomirumab-NR bind to human CD137, but not to mouse CD137. The binding of PC2 and PC3 to human CD137 was through the CRD1 domain and the CRD3 / 4 domain, respectively. In this binding experiment, PC2 lost binding to CD137 only when the human CRD1 domain was replaced with the mouse CRD1 domain, whereas PC3 lost binding to CD137 when either the human CRD3 or CRD4 domain was replaced with the mouse counterpart (Figure 10B-F). The disclosed antibody 1923Ab4 shows reduced binding to cells transfected with msCRD2, msCRD3, and msCRD4 expression constructs (Figure 10B-F). This result suggests that 1923Ab4 binds to human CD137 through the CRD2, CRD3, and CRD4 regions.

[0496] In Figure 11A, sequence alignment of the human and mouse CRD4 regions revealed five distinct differences. Additional expression constructs (M1-M5) were generated by changing the human amino acid sequence to that of the mouse. For example, the human "CF" sequence in the M1 region was mutated to the mouse "SL" sequence by site-directed mutagenesis.

[0497] Analysis of anti-CD137 1923Ab4 binding to HEK293T cells transiently transfected with either human CD137 (WT) or mutant CD137 expression constructs (M1-M5) was performed by flow cytometry-based binding assay. Cells were stained with 1923Ab4 for 2 hours at 4°C, followed by fixing the cells for 15 minutes at room temperature. Fixed cells were washed three times with PBS, followed by staining for 1 hour at room temperature with Alexa Fluor® 488 goat anti-human IgG antibody (Invitrogen, Cat. No. A-11013) for detection. Binding signals were assessed by quantifying the fluorescence intensity using iQue Screener PLUS (Sartorius, MI).

[0498] In Figure 11D, PC2 urelumab-NR was used as an expression control and its binding epitope was mapped to the CRD1 domain (Chin, SM et al, Nat Commun, 2018 Nov 8;9(1):4679). Changing the human amino acid sequence of "KRGI" (SEQ ID NO:81) to the mouse amino acid sequence of "NGTGV" (SEQ ID NO:77) in the M2 region of the CRD4 domain of CD137 significantly reduced its binding to the disclosed antibody 1923Ab4. Mutagenesis on M1, M3, M4, and M5 did not change the binding activity of 1923Ab4 to CD137 protein expressed on the cell surface (Figures 11B, 11C, &11E-G).

[0499] Example 11: Epitope mapping of 1923Ab4 on human CD137 using HDX mass spectrometry Using domain swapping and mutagenesis studies, 1923Ab4 was shown to bind to human CD137 through the CRD2, CRD3 and CRD4 domains (Example 10). To further identify the binding site of 1923Ab4 to human CD137, hydrogen deuterium exchange (HDX) mass spectrometry was used. The region of human CD137 bound by the antibody, which is defined as the epitope, is protected from hydrogen / deuterium exchange. The mass difference of the digested peptides was revealed by LC-MS.

[0500] Recombinant CD137 was first incubated in deuterium oxide, alone or in complex with 1923Ab4 antibody. Deuterium exchange was performed for 0, 15, 60, 600, or 3600 seconds at 20°C. The exchange reaction was quenched by low pH and the protein was digested with pepsin / prolyl endopeptidase / XIII. Deuterium levels in digested peptides of CD137 were monitored from mass shifts on LC-MS.

[0501] Recombinant CD137 showed a significant reduction in deuterium uptake upon binding to 1923Ab4 antibodies with sequences AA46-51 (KGVFRT; SEQ ID NO: 78), AA65-90 (CTPGFHCLGAGCSMCEQDCKQGQELT; SEQ ID NO: 79) and AA104-107 (DQKR; SEQ ID NO: 80) (regions depicted as shaded bars in FIG. 12). This data is consistent with the domain swapping experiments illustrated in Example 10. Recombinant CD137 showed a significant reduction in deuterium uptake upon binding to 1923Ab4 antibodies with sequences AA46-51 (KGVFRT; SEQ ID NO: 78), AA65-90 (CTPGFHCLGAGCSMCEQDCKQGQELT; SEQ ID NO: 79) and AA104-107 (DQKR; SEQ ID NO: 80). The epitope mapping results, depicted as shaded bars, are summarized in FIG. 12. This data is consistent with the domain swapping experiments illustrated in Example 10.

[0502] Example 12: Preparation of scFv that bind to PD-L1 or CD137 ScFvs that bind to PD-L1 with the structure (N)-VL-linker-VH-(C) or (N)-VH-linker-VL-(C) were prepared using the variable regions of a fully human monoclonal antibody against PD-L1 shown in Figure 1A.

[0503] An scFv that binds to CD137 with the structure (N)-VL-linker-VH-(C) or (N)-VH-linker-VL-(C) was prepared using the variable regions of a fully human monoclonal antibody against CD137 shown in Figures 1B and 1C, in which the amino acid residue "G" at position 44 of the heavy chain variable region may be replaced with "C", and the amino acid residue "G" at position 100 of the light chain variable region may be replaced with "C". Such an amino acid substitution of "G" to "C" in the scFv may improve the stability of the scFv as one target-specific portion of a bispecific or trispecific antibody.

[0504] Example 13: Molecular design and production of PD-L1 / CD137 bispecifics As a representative example of a binding protein that binds to PD-L1, a symmetric bispecific (PD-L1 x CD137) was prepared characterized by the molecular format depicted in Figure 13J, containing the subunits / components summarized in Figures 13 and 14: 1923Ab18 1. Heavy chain Components: SEQ ID NO:50 comprising the heavy chain of an anti-PD-L1 antibody, a linker and an anti-CD137 scFv (VH-VL with CC) (N→C); and 2. Light chain : SEQ ID NO: 40 comprising the anti-PD-L1 antibody light chain.

[0505] DNA segment 1 (sequence number 50) having a polynucleotide sequence encoding the heavy chain component of 1923Ab18 was inserted into the expression vector, and DNA segment 2 (sequence number 40) having a polynucleotide sequence encoding the light chain component of 1923Ab18 was inserted into the expression vector.

[0506] The constructed expression vector was transiently expressed in Expi293 cells (ThermoFisher) and incubated for 1 h at 4 °C for 2 h. 2 The cells were cultured in Expi293 expression medium for 5 days at 37°C in an incubator. The bispecific antibodies were purified from cell culture supernatants by recombinant Protein A affinity chromatography (Hitrap Mabselect SuRe, GE) and, if necessary, a second step purification by ion exchange or gel filtration chromatography. SDS-PAGE (BiRad), size exclusion HPLC (Agilent, 1100 series) analysis using a SE-HPLC column (TOSO, G3000SWXL) and CE-SDS (SCIEX, PA800 Plus) were performed to detect and confirm the size and purity of the bispecific antibodies. The purified proteins were buffer exchanged into the desired buffer and concentrated by ultrafiltration using an Amicon Ultra 15 30K device and the protein concentration was estimated using DropSense (Unchained Lab). Transient transfection could be used in a two-vector system or with a one-vector system containing both heavy and light chain components in one single vector. Alternatively, bispecific antibodies can be purified from the supernatant of a stable CHO expressing cell line.

[0507] Example 14: Molecular design and production of asymmetric PD-L1 / TGFβ / CD137 trispecifics As a representative example of an asymmetric binding protein that binds to PD-L1, a trispecific (PD-L1×CD137×TGFβRII) was prepared that is characterized by the molecular format depicted in FIG. 13C, containing the subunits / components summarized in FIGS. 13 and 14: 1923Ab9 1. Heavy Chain (Knob) Components: SEQ ID NO:41 comprising the heavy chain of an anti-PD-L1 antibody, a linker and an anti-CD137 scFv (VH-VL with CC) (N→C); 2. Heavy Chain (Hole): SEQ ID NO:42, which comprises the heavy chain of an anti-PD-L1 antibody component (N→C); and 3. Light chain: SEQ ID NO: 39 containing the anti-PD-L1 antibody light chain.

[0508] DNA segment 1 (SEQ ID NO: 41) having a polynucleotide sequence encoding the heavy chain (knob) component of 1923Ab9 was inserted into the expression vector, DNA segment 2 (SEQ ID NO: 42) having a polynucleotide sequence encoding the heavy chain (hole) component of 1923Ab9 was inserted into the expression vector, and DNA segment 3 (SEQ ID NO: 39) having a polynucleotide sequence encoding the light chain of 1923Ab9 was inserted into the expression vector.

[0509] The constructed expression vector was transiently expressed in Expi293 cells (ThermoFisher) and incubated for 1 h at 4 °C for 2 h. 2 The cells were cultured in Expi293 expression medium for 5 days at 37°C in an incubator. The bispecific antibodies were purified from cell culture supernatants by recombinant Protein A affinity chromatography (Hitrap Mabselect SuRe, GE) and, if necessary, a second step purification by ion exchange or gel filtration chromatography. SDS-PAGE (BioRad), size exclusion HPLC (Agilent, 1100 series) analysis using a SE-HPLC column (TOSO, G3000SWXL) and CE-SDS (SCIEX, PA800 Plus) were performed to detect and confirm the size and purity of the trispecific antibodies. The purified proteins were buffer exchanged into the desired buffer and concentrated by ultrafiltration using an Amicon Ultra 15 30K device and the protein concentration was estimated using DropSense (Unchained Lab). Transient transfection could be used in a 3-vector system or with a 1-vector system containing both heavy and light chain components in one single vector. Alternatively, bispecific antibodies can be purified from the supernatant of a stable CHO expressing cell line.

[0510] Example 15: Molecular design and production of PD-L1 / TGFβ / CD137 trispecifics As a representative example of a symmetric binding protein that binds to PD-L1, a trispecific (PD-L1×CD137×TGFβRII, 1923Ab17) was prepared characterized by the molecular format depicted in FIG. 13I, containing the subunits / components summarized in FIGS. 13 and 14: 1923Ab17 1. Heavy chain SEQ ID NO: 50 comprising the component, the heavy chain of an anti-PD-L1 antibody, a linker and an anti-CD137 scFv (VH-VL with CC) (N→C); and 2. Light chain : SEQ ID NO:39 comprising the component, the light chain of an anti-PD-L1 antibody, a linker, and TGFβRII ECD.

[0511] A DNA segment (1) (SEQ ID NO:50) having a polynucleotide sequence encoding the heavy chain component of 1923Ab17 was inserted into an expression vector, and a DNA segment (2) (SEQ ID NO:39) having a polynucleotide sequence encoding the light chain component of 1923Ab17 was inserted into an expression vector.

[0512] The constructed expression vector was transiently expressed in Expi293 cells (ThermoFisher) and incubated for 1 h at 4 °C for 2 h. 2The cells were cultured in Expi293 expression medium for 5 days at 37°C in an incubator. The trispecific antibodies were purified from cell culture supernatants by recombinant Protein A affinity chromatography (Hitrap Mabselect SuRe, GE) and, if necessary, a second step purification by ion exchange or gel filtration chromatography. SDS-PAGE (BioRad), size exclusion HPLC (Agilent, 1100 series) analysis using a SE-HPLC column (TOSO, G3000SWXL) and CE-SDS (SCIEX, PA800 Plus) were performed to detect and confirm the size and purity of the trispecific antibodies. The purified proteins were buffer exchanged into the desired buffer and concentrated by ultrafiltration using an Amicon Ultra 15 30K device and the protein concentration was estimated using DropSense (Unchained Lab). Transient transfection could be used in a two-vector system or with a one-vector system containing both heavy and light chain components in one single vector. Alternatively, bispecific antibodies can be purified from the supernatant of a stable CHO expressing cell line.

[0513] Example 16: Characterization of bispecific and trispecific binding to PD-L1: Binding to CD137 Bispecific and trispecific antibodies were generated, produced, and purified as described in Examples 13 to 15. To examine the binding activity of these antibodies to CD137, immunofluorescence imaging assays were performed using HEK293T cells stably transfected with a human CD137 expression construct. This cell line expressed human CD137 protein on the cell surface. The cells were seeded in complete medium containing DMEM with 10% FBS and then incubated overnight at 37°C. The cells were stained with these antibodies for 2 hours at 4°C, and then the cells were fixed for 15 minutes at room temperature. The fixed cells were washed three times with PBS, and then stained with Alexa Fluor® 488 goat anti-human IgG (H+L) secondary antibody for detection (Invitrogen, Cat. No. A-11013) for 1 hour at room temperature. The binding signal was evaluated by imaging the cells and quantifying the fluorescence intensity using Cytation 5 (Biotek, VT).

[0514] In Figure 16A, all of the disclosed bispecific and trispecific antibodies, including 1923Ab7, 1923Ab8, 1923Ab9, 1923Ab10 and 1923Ab11, bound similarly to human CD137 on the cell surface compared to 1923Ab4, indicating that the ScFv format of anti-CD137 retained binding activity to CD137.

[0515] The inventors also generated trispecific antibodies using 1923Ab3 as an ScFv format. In Figure 16B, both 1923Ab7 and 1923Ab16 bound similarly to human PD-L1 on the cell surface as measured by flow cytometry. This result indicates that both 1923Ab3 and 1923Ab4 can be used as ScFv formats in the disclosed bispecific and trispecific antibodies.

[0516] Example 17: Characterization of bispecifics and trispecifics binding to PD-L1:CD137 signaling These antibodies were further evaluated by measuring their agonistic activity on CD137 signaling using an NFκB luciferase reporter assay. 293T cells stably transfected with a human CD137 expression plasmid and an NFκB luciferase reporter plasmid were used as reporter cells. These reporter cells were co-cultured with 293T cells expressing PD-L1 and stimulated with antibodies and incubated at 37°C with 5% CO. 2 The plates were incubated at RT for 16 hours. ONE-Glo™ Luciferase Reagent (Promega, Cat. No. E6130) was added and the plates were incubated at room temperature for 10 minutes. Luminescence signals were measured using a Synergy Neo2 plate reader (Biotek, VT) and data was analyzed using GraphPad Prism. Activation of CD137 signaling resulted in an increase in luminescence signal.

[0517] In Figure 17A, all of the disclosed antibodies, including 1923Ab7, 1923Ab8, 1923Ab9, 1923Ab10 and 1923Ab11, activated CD137 signaling compared to the isotype control antibody. 1923Ab7, 1923Ab8 and 1923Ab11 showed stronger agonist activity compared to 1923Ab9 and 1923Ab10 (Figure 17A). The results demonstrate that bivalent binding of anti-CD137 induced stronger anti-CD137 signaling than monovalent binding of anti-CD137.

[0518] We also generated a trispecific antibody using 1923Ab3 as an ScFv format. The agonistic activity of these antibodies against CD137 signaling was also compared using Jurkat T CD137 reporter cells expressing recombinant CD137 and NF B luciferase reporter. Briefly, Jurkat T NFkB reporter cell line was used to measure the activity of CD137 signaling, and PD-L1-expressing HEK293T cells were used as target cells to deliver PD-L1. In Figure 17B, both 1923Ab7 and 1923Ab16 showed similar agonistic activity with CD137 signaling measured by Jurkat T CD137 reporter cells. 1923Ab16 effectively binds to human PD-L1 and activates CD137 signaling.

[0519] Example 18: Characterization of bispecifics and trispecifics - PD-L1 dependent activation of CD137 signaling The disclosed antibodies were evaluated for their ability to induce PD-L1-dependent CD137 agonism. Figure 18 demonstrates the ability of 1923Ab7, 1923Ab8, 1923Ab17, and 1923Ab18 to induce CD137 signaling using Jurkat T CD137 reporter cells in the presence (18A) or absence (18B) of target cells. Briefly, CD137-expressing Jurkat T NFkB reporter cell lines were used to measure the activity of CD137 signaling, and PD-L1-expressing HEK293T cells were used as target cells to provide PD-L1. When reporter cells were co-cultured with target cells, the disclosed antibodies, including 1923Ab7, 1923Ab8, 1923Ab17, 1923Ab18, and urelumab-NR, demonstrated activation of CD137 signaling (Figure 18A). However, none of the disclosed antibodies induced CD137 signaling in the absence of target cells, except for Urelumab-NR (Figure 18B). Urelumab is a cross-linking independent antibody deve...

Claims

1. An anti-CD137 antibody or antigen-binding fragment thereof, (i) a heavy chain variable region comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:22, and CDR3: SEQ ID NO:23; and a light chain variable region comprising CDR1: SEQ ID NO:24, CDR2: SEQ ID NO:25, and CDR3: SEQ ID NO:26; (ii) a heavy chain variable region comprising CDR1: SEQ ID NO: 27, CDR2: SEQ ID NO: 28, and CDR3: SEQ ID NO: 29; and a light chain variable region comprising CDR1: SEQ ID NO: 30, CDR2: SEQ ID NO: 9, and CDR3: SEQ ID NO: 31; or (iii) a heavy chain variable region comprising CDR1: SEQ ID NO: 32, CDR2: SEQ ID NO: 33, and CDR3: SEQ ID NO: 34; and a light chain variable region comprising CDR1: SEQ ID NO: 35, CDR2: SEQ ID NO: 36, and CDR3: SEQ ID NO: 37; An anti-CD137 antibody or antigen-binding fragment thereof. (i) a heavy chain variable region comprising SEQ ID NO: 16 and a light chain variable region comprising SEQ ID NO: 17; (ii) a heavy chain variable region comprising SEQ ID NO: 18 and a light chain variable region comprising SEQ ID NO: 19; or (iii) An anti-CD137 antibody or antigen-binding fragment thereof according to claim 1, comprising a heavy chain variable region comprising SEQ ID NO: 20 and a light chain variable region comprising SEQ ID NO:

21.

3. An anti-CD137 antibody or antigen-binding fragment thereof described in claim 1, wherein the binding protein or binding fragment thereof comprises a heavy chain constant region.

4. An anti-CD137 antibody or its antigen-binding fragment described in claim 3, wherein the constant region comprises SEQ ID NO: 60, SEQ ID NO: 61, SEQ ID NO: 62, SEQ ID NO: 63, or SEQ ID NO:

64.

5. An anti-CD137 antibody described in claim 1, comprising a heavy chain comprising sequence number 75 and a light chain comprising sequence number 76.

6. An anti-CD137 antibody or antigen-binding fragment thereof described in claim 1, comprising a light chain constant region comprising sequence number 65 or sequence number 66.

7. An anti-CD137 antibody described in claim 1, which is a multispecific antibody, preferably a bispecific or trispecific antibody.

8. The antigen-binding fragment of claim 1, which is a Fab, Fab', F(ab)2, F(ab')2, Fd, Fv, scFv, single-chain antibody, miniantibody, or diabody.

9. An anti-PD-L1 antibody or antigen-binding fragment thereof, (i) a heavy chain variable region comprising CDR1: SEQ ID NO:5, CDR2: SEQ ID NO:6, and CDR3: SEQ ID NO:7; and a light chain variable region comprising CDR1: SEQ ID NO:8, CDR2: SEQ ID NO:9, and CDR3: SEQ ID NO:10; or (ii) a heavy chain variable region comprising CDR1: SEQ ID NO: 11, CDR2: SEQ ID NO: 12, and CDR3: SEQ ID NO: 13; and a light chain variable region comprising CDR1: SEQ ID NO: 14, CDR2: SEQ ID NO: 9, and CDR3: SEQ ID NO: 15; An anti-PD-L1 antibody or an antigen-binding fragment thereof.

10. A nucleic acid composition comprising one or more nucleic acids encoding the antibody or antigen-binding fragment thereof described in claim 1 or 9.

11. An expression vector composition comprising one or more expression vectors containing the nucleic acid composition described in claim 10.

12. A host cell comprising the nucleic acid composition described in claim 10.

13. A host cell comprising the expression vector composition described in claim 11.

14. A method for producing the anti-CD137 antibody or its antigen-binding fragment described in claim 1, comprising culturing the host cell described in claim 12.

15. A pharmaceutical composition comprising the anti-CD137 antibody or antigen-binding fragment thereof described in claim 1 and a pharmaceutically acceptable carrier.

Images