Anti-CD137 antibodies and methods of use

Abstract

The present disclosure provides antibodies, or antigen-binding fragments thereof, that specifically bind to human CD137, multispecific antibodies, and antigen-binding fragments thereof, that specifically bind to human GPC3 and CD137, pharmaceutical compositions comprising said antibodies, and the use of the antibodies, multispecific antibodies, or compositions for the treatment of diseases such as cancer.

Description

Technical Field

[0001] This disclosure relates to antibodies that specifically bind to human CD137 (TNF receptor superfamily member 9 (TNFRSF9)), bispecific antibodies or antigen-binding fragments thereof that bind to human CD137, compositions comprising said antibodies, and methods or uses for treating cancer. Background Art

[0002] Glypican-3 (GPC3) belongs to the heparan sulfate proteoglycan (HSPG) family and comprises a 60 - 70 kD core protein that is linked to the cell membrane surface via a glycosylphosphatidylinositol anchor (GPI). The carboxyl terminus of GPC3 is modified with heparan sulfate side chains (Filmus J et al., J. Clin. Inv. [Journal of Clinical Investigation] 2001; 108: 497 - 501).

[0003] The specific expression of GPC3 in tumor cells has received extensive attention. GPC3 is expressed in hepatocellular carcinoma (HCC), which is the most common type of liver cancer. Notably, its expression is not detected in non-malignant tissues. Overexpression of GPC3 has also been reported in hepatoblastoma, lung squamous cell carcinoma (LSCC), and other cancers. Therefore, GPC3 is suitable as a tumor antigen for targeted therapy. (Li N et al., Trends Cancer. [Cancer Trends] 2018; 4: 741 - 54; Ho M, et al., Eur J Cancer. [European Journal of Cancer] 2011; 47: 333 - 8; Moek et al., Am. J. Pathol. [American Journal of Pathology] 2018; 188(9): 1973 - 1981).

[0004] CD137 (also known as TNFRSF9 / 41BB) is a co-stimulatory molecule belonging to the TNFRSF family. It was discovered by screening for T cell factors in murine helper and cytotoxic cells stimulated with concanavalin A. It was identified as an inducible gene in 1989 that is expressed on antigen-primed T cells but not on resting T cells (Kwon et al., Proc. Natl. Acad. Sci. USA 1989; 86:1963–1967). In addition, it is known to be expressed in the following cells: dendritic cells (DC), natural killer cells (NK) (Vinay et al., Mol. Cancer Ther. 2012; 11:1062–1070), activated CD4+ and CD8+ T lymphocytes, eosinophils, natural killer T cells (NKT), and mast cells (Kwon et al., 1989 ibid.; Vinay D., Int. J. Hematol. 2006; 83:23–28). CD137 maintains and enhances immune effector functions by inducing Th1 cytokine production (Bartkowiak et al., Front Oncol. 2015; 5:117; Shuford et al., J Exp Med. 1997; 186:47-55). Upon binding to its sole ligand (CD137L, 4-1BBL or TNFSF9), CD137 signaling activates via the NF-κB pathway leading to increased expression of prosurvival molecules (Wang et al., Immunol Rev. 2009; 229:192-215).

[0005] The anti-CD137 antibody urelumab (BMS-663513), which binds to CRD I of CD137, and utomilumab (PF-05082566), which binds to CRD III and IV of CD137, have shown potential as cancer therapeutics due to their ability to activate cytotoxic T cells and increase the production of interferon gamma (IFN-γ). The mechanism by which these antibodies regress tumors is their enhancement of the cancer immune cell response. In particular, anti-CD137 antibodies stimulate and activate effector T lymphocytes (e.g., by stimulating CD8+ T lymphocytes to produce INFγ), and enhance the production of NKT and APC (e.g., macrophages).

[0006] Urelumab showed promising results in preclinical experiments and early clinical studies (Sznol et al., Clin. Oncol. [Clinical Oncology] 2008; 26 (Suppl 15)). However, in later studies, urelumab exhibited hepatotoxicity, leading to the suspension of the development of this antibody until February 2012 (Segal et al., Clin. Cancer Res. [Clinical Cancer Research] 2017; 23:1929–1936). The hepatotoxicity was mainly due to the S100A4 protein secreted by tumor and stromal cells, and the study that restricted the dose of urelumab to 8 mg or 0.1 mg / kg every 3 weeks per patient restored the interest in this antibody (Segal et al., Clin. Cancer Res. [Clinical Cancer Research] 2017; 23:1929–1936).

[0007] Compared with urelumab, utomilumab showed a better safety profile, and preliminary studies showed no hepatotoxicity or other dose-limiting factors (Segal et al., J. Clin. Oncol. [Journal of Clinical Oncology] 2014; 32 (Suppl 15)). The reported results of the phase I trial of utomilumab as a single agent showed a good safety profile (Segal et al., Clin. Cancer Res. [Clinical Cancer Research] 2018; 24:1816–1823). It is speculated that the difference between these two antibodies is due to their different binding sites on the CD137 receptor.

[0008] There is no approved therapeutic antibody against CD137, and the medical need for therapeutic agents targeting CD137 remains unmet. In addition, anti-TAAxCD137 bispecific antibodies that recruit immune cells to cancers expressing tumor-associated antigens (TAAs) can be used to treat cancers. SUMMARY OF THE DISCLOSURE

[0009] This disclosure contains antibodies and antigen-binding fragments that specifically bind to human CD137. In addition, the CD137 VHH domain fragments disclosed herein can be used to construct bispecific antibodies together with other forms (such as TAAs, immune checkpoints, or immune stimulatory factors). The CD137 antibodies alone or in combination with other forms can potentially be used to treat or prevent cancer, autoimmune diseases, or infectious diseases.

[0010] In addition, this disclosure relates to bispecific anti-GPC3xCD137 antibodies and their antigen-binding fragments.

[0011] This disclosure encompasses the following examples.

[0012] Example 1: An antibody or its antigen-binding fragment that specifically binds to human CD137, wherein: (1) The antibody or its antigen-binding fragment specifically binds to the following epitope, which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35); (2) The antibody or its antigen-binding fragment specifically binds to the following epitope, which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97 and Gly98 of human CD137 (SEQ ID NO: 35); or (3) The antibody or its antigen-binding fragment specifically binds to a human CD137 dimer, which comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or its antigen-binding fragment specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or its antigen-binding fragment specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97 and Gly98; and / or the antibody or its antigen-binding fragment binds to the human CD137 dimer and promotes human CD137 aggregation.

[0013] Example 2. An antibody or its antigen-binding fragment that specifically binds to human CD137, the antibody or its antigen-binding fragment comprising: (i) A heavy chain variable region (VH) that comprises (a) HCDR1 (heavy chain complementarity-determining region 1) of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) A heavy chain variable region that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO: 3.

[0014] Example 3. The antibody or its antigen-binding fragment according to any one of Examples 1-2, the antibody or its antigen-binding fragment comprising: (i) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 4.

[0015] Example 4. The antibody or antigen-binding fragment thereof according to Example 3, wherein one, two, three, four, five, six, seven, eight, nine or ten amino acids have been inserted, deleted or substituted in SEQ ID NO: 17, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15 or SEQ ID NO: 4.

[0016] Example 5. The antibody or antigen-binding fragment thereof according to any one of the preceding examples, wherein the antibody or antigen-binding fragment comprises: (i) A variable heavy chain (VH) that comprises SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises SEQ ID NO: 4.

[0017] Example 6. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), Fab fragment, Fab' fragment or F(ab’)2 fragment.

[0018] Example 7. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof comprises a heavy chain constant region of IgG1, IgG2, IgG3 or IgG4 subclass and / or a light chain constant region of κ or λ type.

[0019] Example 8. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof has antibody-dependent cell cytotoxicity (ADCC) or complement-dependent cytotoxicity (CDC).

[0020] Example 9. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof has reduced glycosylation or is non-glycosylated or is hypofucosylated.

[0021] Example 10. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof comprises an increased bisecting GlcNac structure.

[0022] Example 11. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:19 or SEQ ID NO:53.

[0023] Example 12. The antibody or antigen-binding fragment thereof according to any one of the foregoing examples, wherein the antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function and / or an extended half-life, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:20.

[0024] Example 13. A pharmaceutical composition, which comprises the antibody or antigen-binding fragment thereof according to any one of the foregoing examples, and a pharmaceutically acceptable carrier.

[0025] Example 14. A method for treating cancer, which comprises administering to a patient in need a therapeutically effective amount of the antibody or antigen-binding fragment thereof according to any one of Examples 1-12, or the pharmaceutical composition according to Example 13.

[0026] Example 15. The method according to Example 14, wherein the cancer is gastric cancer, colon cancer, pancreatic cancer, breast cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, small cell lung cancer, non-small cell lung cancer, ovarian cancer, skin cancer, mesothelioma, lymphoma, leukemia, myeloma, and sarcoma.

[0027] Example 16. The method according to any one of Examples 14-15, wherein the antibody or its antigen-binding fragment is administered in combination with another therapeutic agent.

[0028] Example 17. The method according to Example 16, wherein the therapeutic agent is an anti-PD-1 antibody.

[0029] Example 18. The method according to Example 17, wherein the anti-PD1 antibody is tislelizumab.

[0030] Example 19: A multispecific antibody or its antigen-binding fragment, the multispecific antibody or its antigen-binding fragment comprising at least a first antigen-binding domain that specifically binds to a human tumor-associated antigen (TAA), and at least a second antigen-binding domain that specifically binds to human CD137, wherein the second antigen-binding domain is: (1) An antibody or its antigen-binding fragment that specifically binds to the following epitope, the epitope comprising amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35), or consisting thereof; (2) An antibody or its antigen-binding fragment that specifically binds to the following epitope, the epitope comprising amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35), or consisting thereof; or (3) An antibody or antigen-binding fragment thereof that specifically binds to a human CD137 dimer, wherein the human CD137 dimer comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), the epitope comprising or consisting of the amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), the epitope comprising or consisting of the amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98; and / or the antibody or antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

[0031] Example 20. A multispecific antibody or antigen-binding fragment thereof, the multispecific antibody or antigen-binding fragment thereof comprising at least a first antigen-binding domain that specifically binds to a human tumor-associated antigen (TAA) and at least a second antigen-binding domain that specifically binds to human CD137, wherein the second antigen-binding domain comprises: (i) A heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) A heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO: 3.

[0032] Example 21. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-20, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 17; (ii) A heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 11; (iii) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 13; (iv) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 15; or (v) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 4.

[0033] Example 22. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19 - 21, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises SEQ ID NO: 17; (ii) A heavy chain variable region (VH) that comprises SEQ ID NO: 11; (iii) A heavy chain variable region (VH) that comprises SEQ ID NO: 13; (iv) A heavy chain variable region (VH) that comprises SEQ ID NO: 15; or (v) A heavy chain variable region (VH) that comprises SEQ ID NO: 4.

[0034] Example 23. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19 - 22, wherein the TAA is GPC3.

[0035] Example 24. A multispecific antibody or antigen-binding fragment thereof that comprises a first antigen-binding domain that specifically binds to human phosphatidylinositol proteoglycan 3 (GPC3) and a second antigen-binding domain that specifically binds to human CD137.

[0036] Example 25: The multispecific antibody or antigen-binding fragment thereof according to Example 24, wherein the second antigen-binding domain that specifically binds to human CD137 is: (1) An antibody or antigen-binding fragment thereof that specifically binds to the following epitope, which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35); (2) An antibody or antigen-binding fragment thereof that specifically binds to the following epitope, which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35); or (3) An antibody or antigen-binding fragment thereof that specifically binds to a human CD137 dimer, which comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or antigen-binding fragment specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or antigen-binding fragment specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98; and / or the antibody or antigen-binding fragment binds to the human CD137 dimer and promotes human CD137 aggregation.

[0037] Example 26. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 24 - 25, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO: 3.

[0038] Example 27. The multispecific antibody or antigen-binding fragment thereof according to Example 26, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 4.

[0039] Example 28. The multispecific antibody or antigen-binding fragment thereof according to Example 27, wherein one, two, three, four, five, six, seven, eight, nine, or ten amino acids have been inserted, deleted, or substituted in SEQ ID NO: 17, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15, or SEQ ID NO: 4.

[0040] Example 29. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 24-28, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A variable heavy chain (VH) that comprises SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises SEQ ID NO: 4.

[0041] Example 30. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-29, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: A heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:45, (b) HCDR2 of SEQ ID NO:46, and (c) HCDR3 of SEQ ID NO:47; and A light-chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO:48, (e) LCDR2 of SEQ ID NO:49, and (f) LCDR3 of SEQ ID NO:50.

[0042] Example 31. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-30, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: A heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO:41, and a light-chain variable region (VL) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO:43.

[0043] Example 32. The multispecific antibody or antigen-binding fragment thereof according to Example 31, wherein one, two, three, four, five, six, seven, eight, nine or ten amino acids have been inserted, deleted or substituted in SEQ ID NO:41 or SEQ ID NO:43.

[0044] Example 33. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-32, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: a heavy-chain variable region (VH) that comprises SEQ ID NO:41, and a light-chain variable region (VL) that comprises SEQ ID NO:43.

[0045] Example 34. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-33, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises A heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:45, (b) HCDR2 of SEQ ID NO:46, and (c) HCDR3 of SEQ ID NO:47; and A light chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO: 48, (e) LCDR2 of SEQ ID NO: 49, and (f) LCDR3 of SEQ ID NO: 50, and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO: 3.

[0046] Example 35. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19 - 34, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: A heavy chain variable region (VH) that comprises SEQ ID NO: 41, and a light chain variable region (VL) that comprises SEQ ID NO: 43; and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises SEQ ID NO: 4; (ii) A heavy chain variable region (VH) that comprises SEQ ID NO: 11; (iii) A heavy chain variable region (VH) that comprises SEQ ID NO: 13; (iv) A heavy chain variable region (VH) that comprises SEQ ID NO: 15; or (v) A heavy chain variable region (VH) that comprises SEQ ID NO: 17.

[0047] Example 36. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19 - 35, wherein the multispecific antibody or antigen-binding fragment is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), Fab fragment, Fab' fragment, or F(ab’)2 fragment.

[0048] Example 37. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-36, wherein the first antigen-binding domain that specifically binds to human GPC3 is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), single-domain antibody, Fab fragment, Fab' fragment or F(ab')2 fragment, and the second antigen-binding domain that specifically binds to human CD137 is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), single-domain antibody, Fab fragment, Fab' fragment or F(ab')2 fragment.

[0049] Example 38. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-37, wherein the multispecific antibody or antigen-binding fragment thereof is a bispecific antibody.

[0050] Example 39. The multispecific antibody or antigen-binding fragment thereof according to Example 38, wherein the bispecific antibody is in 2+2 form.

[0051] Example 40. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-39, wherein the multispecific antibody or antigen-binding fragment thereof contains a linker of SEQ ID NO: 60 to SEQ ID NO: 101.

[0052] Example 41. The multispecific antibody or antigen-binding fragment thereof according to Example 40, wherein the linker is SEQ ID NO: 62.

[0053] Example 42. The multispecific antibody or antigen-binding fragment thereof according to Example 40, wherein the linker is SEQ ID NO: 67.

[0054] Example 43. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-42, wherein the multispecific antibody or antigen-binding fragment contains a heavy chain constant region of IgG1, IgG2, IgG3 or IgG4 subclass and / or a light chain constant region of κ or λ type, and wherein the heavy chain constant region contains CH1 and / or Fc domain.

[0055] Example 44. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-43, wherein the multispecific antibody or antigen-binding fragment thereof has antibody-dependent cell cytotoxicity (ADCC) or complement-dependent cell cytotoxicity (CDC).

[0056] Example 45. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-44, wherein the multispecific antibody or antigen-binding fragment thereof has reduced glycosylation or no glycosylation or is hypofucosylated.

[0057] Example 46. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-45, wherein the multispecific antibody or antigen-binding fragment thereof comprises an increased bisecting GlcNac structure.

[0058] Example 47. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-46, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:53.

[0059] Example 48. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-47, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function and / or an extended half-life, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:20.

[0060] Example 49. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-48, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG4 Fc.

[0061] Example 50. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-49, wherein: a) The heavy chain variable region (VH), CH1 domain, the Fc domain of the first antigen-binding domain that specifically binds to human GPC3, and the heavy chain variable region (VH) of the second antigen-binding domain that specifically binds to human CD137 are arranged in the first polypeptide in the N-terminal to C-terminal direction; Optionally, the C-terminus of the Fc domain is linked to the N-terminus of the heavy chain variable region (VH) of the second antigen-binding domain through a linker; and b) The light chain variable region (VL) and the first light chain constant region of the first antigen-binding domain that specifically binds to human GPC3 are arranged in the second polypeptide in the N-terminal to C-terminal direction.

[0062] Example 51. The multispecific antibody or antigen-binding fragment thereof according to any one of Examples 19-50, wherein the multispecific antibody or antigen-binding fragment comprises (i) The first polypeptide of SEQ ID NO:25 and the second polypeptide of SEQ ID NO:23; (ii) The first polypeptide of SEQ ID NO:21 and the second polypeptide of SEQ ID NO:23; (iii) The first polypeptide of SEQ ID NO:33 and the second polypeptide of SEQ ID NO:23; (iv) The first polypeptide of SEQ ID NO:27 and the second polypeptide of SEQ ID NO:23; (v) The first polypeptide of SEQ ID NO:29 and the second polypeptide of SEQ ID NO:23; or (vi) The first polypeptide of SEQ ID NO:31 and the second polypeptide of SEQ ID NO:23.

[0063] Example 52. A pharmaceutical composition comprising a multispecific antibody or an antigen-binding fragment thereof according to any one of Examples 19 - 51, and a pharmaceutically acceptable carrier.

[0064] Example 53. A method for treating cancer, the method comprising administering to a patient in need thereof a therapeutically effective amount of a multispecific antibody or an antigen-binding fragment thereof according to any one of Examples 19 - 51, or a pharmaceutical composition according to Example 52.

[0065] Example 54. The method according to Example 53, wherein the cancer is an advanced or metastatic solid tumor.

[0066] Example 55. The method according to any one of Examples 53 - 54, wherein the cancer expresses GPC3.

[0067] Example 56. The method according to any one of Examples 53 - 55, wherein the cancer is liver cancer, lung cancer, gastric cancer, germ cell tumor, thyroid cancer, pancreatic cancer, ovarian cancer, skin cancer, kidney cancer, esophageal cancer, atypical teratoid rhabdoid tumor of the brain or undifferentiated synovial sarcoma.

[0068] Example 57. The method according to Example 56, wherein the liver cancer is hepatoblastoma or hepatocellular carcinoma (HCC).

[0069] Example 58. The method according to Example 56, wherein the lung cancer is non-small cell lung cancer (NSCLC) or small cell lung cancer (SCLC).

[0070] Example 59. The method according to Example 58, wherein the non-small cell lung cancer is squamous non-small cell lung cancer.

[0071] Example 60. The method according to Example 58, wherein the non-small cell lung cancer is GPC3+ squamous non-small cell lung cancer.

[0072] Example 61. The method according to Example 56, wherein the gastric cancer is alpha-fetoprotein+(AFP+) gastric cancer.

[0073] Example 62. The method according to Example 56, wherein the renal cancer is nephroblastoma.

[0074] Example 63. The method according to Example 56, wherein the esophageal cancer is esophageal squamous cell carcinoma.

[0075] Example 64. The method according to Example 56, wherein the esophageal cancer is GPC3+ esophageal squamous cell carcinoma.

[0076] Example 65. The method according to Example 56, wherein the germ cell tumor is yolk sac tumor or non-dysgerminoma.

[0077] Example 66. The method according to any one of Examples 53-65, wherein the multispecific antibody or its antigen-binding fragment or pharmaceutical composition is administered in combination with another therapeutic agent.

[0078] Example 67. The method according to Example 66, wherein the therapeutic agent is an anti-PD1 or anti-PDL1 antibody.

[0079] Example 68. The method according to Example 67, wherein the anti-PD1 antibody is tislelizumab.

[0080] Example 69. An isolated nucleic acid encoding an antibody, multispecific antibody or antigen-binding fragment thereof according to any one of Examples 1-12 and 19-51.

[0081] Example 70. A vector comprising the nucleic acid according to Example 69.

[0082] Example 71. A host cell comprising the nucleic acid according to Example 69 or the vector according to Example 70.

[0083] Example 72. A process for producing a multispecific antibody or its antigen-binding fragment, the process comprising culturing the host cell according to Example 71 and recovering the antibody or its antigen-binding fragment from the culture.

[0084] In some embodiments, the present disclosure provides an anti-CD137 antibody or an antigen-binding fragment thereof, which shows specific binding and high affinity for human CD137 and cynomolgus monkey CD137, shows superior overall biophysical properties, such as Tm or Tagg, shows superior pharmacokinetics, and / or is capable of binding to the CD137 dimer and promoting CD137 aggregation.

[0085] In some embodiments, the present disclosure provides an anti-CD137 antibody or an antigen-binding fragment thereof having at least one or more of the following characteristics: (1) Shows specific binding and high affinity for human CD137 and cynomolgus monkey CD137; (2) Has superior pharmacokinetics; (3) Has superior overall biophysical properties, such as Tm or Tagg, and / or has superior overall stability; (4) Is a humanized antibody with a low immunogenicity risk in humans; and (5) Is capable of binding to the CD137 dimer and promoting CD137 aggregation.

[0086] In some embodiments, the present disclosure provides an anti-CD137 antibody or an antigen-binding fragment thereof, which is a humanized antibody with a low immunogenicity risk in humans, while maintaining specific binding and high affinity for human CD137 and cynomolgus monkey CD137 and showing superior overall biophysical properties and / or stability. In some embodiments, the present disclosure provides an anti-CD137 antibody or an antigen-binding fragment thereof, which shows specific binding and high affinity for human CD137 and cynomolgus monkey CD137 and / or is capable of binding to the CD137 dimer and promoting CD137 aggregation (e.g., through CDR residues).

[0087] In some embodiments, the present disclosure provides a GPC3xCD137 bispecific antibody or an antigen-binding fragment thereof having at least one or more of the following characteristics: (1) Shows specific binding and high affinity for human CD137 and cynomolgus monkey CD137; (2) Has specific binding and high affinity for human GPC3 and cynomolgus monkey GPC3, and shows high affinity for a wide range of GPC expression (low to high expression); (3) Induces T cell activation in a GPC3-dependent manner, including cytokine release (e.g., IFN-γ or IL-2) and T cell killing activity, and reduces T cell activation or T cell killing activity in the absence of cells expressing GPC3; (4) Induce T cell activation and effective T cell killing activity against a wide range of cells expressing GPC3 (low, medium, high expression); (5) Effectively inhibit tumor growth when administered alone; (7) Induce a synergistic effect (e.g., tumor growth inhibition and / or tumor-free rate) when administered together with an anti-PD-1 antibody; (8) Have excellent pharmacokinetics; (9) Have excellent overall biophysical properties, such as Tm or Tagg, and / or stability; and (10) Be able to bind to CD137 dimer and promote CD137 aggregation. Brief Description of the Drawings

[0088] Figure 1 It was demonstrated by ELISA that compared with urelumab (BMS-663513), BGA-9612 binds to huCD137 and partially competes with 20 μg / ml CD137L.

[0089] Figure 2 Showed a comparison of the FACS binding affinity of BGA-9612 with other humanized VHHs in HuT78 cells overexpressing human CD137.

[0090] Figure 3 Is a schematic design of the GPC3xCD137 multi-specific antibody form targeting tumors.

[0091] Figure 4A Showed a binding assay of BE-774 with Hut78 / CD137 cells overexpressing CD137 by flow cytometry, demonstrating the binding of BE-774 to native huCD137 expressed on the cell surface. Figure 4B Showed a binding assay of BE-774 with HepG2 cells expressing GPC3 by flow cytometry, demonstrating the binding of BE-774 to native huGPC3 expressed on the cell surface.

[0092] Figures 5A-5B Proved that BE-774 and BE-653 induce T cell activation when co-cultured with GPC3-positive tumor cells HepG2. Figure 5A is a schematic diagram of activating CD137 (41BB) by co-stimulating huPBMC with BE-774 or BE-653 and the hepatocellular carcinoma (HCC) cell line expressing OS8. Figure 5B showed that BE-774 and BE-653 induced dose-dependent cytokine release in PBMC co-cultured with HepG2 cells, but not in PBMC co-cultured with GPC3-negative cells.

[0093] Figures 6A-6B It was demonstrated that BE-774 and BE-653 enhanced the T cell killing activity against GPC3-positive tumor cells HepG2. Figure 6A is a schematic diagram of activating CD137 (4-1BB) by co-stimulating huPBMC with BE-774 or BE-653 in combination with an EpCAM / CD3 bispecific T cell engager (BiTE) that provides the first signal for T cell activation. Figure 6B shows that BE-774 and BE-653 enhanced the T cell killing activity against cells expressing GPC3 in a dose-dependent manner, but did not enhance the killing activity against GPC3-negative cells.

[0094] Figure 7 The PK profiles of BE-933 and BE-774 in cynomolgus monkeys were shown.

[0095] Figure 8 The PK profiles of BE-933 and BE-774 in the hFcRn mouse model were shown.

[0096] Figures 9A-9B The binding of BE-915 to human CD137 overexpressed on Hut78 (Figure 9B) and human GPC3 expressed on HepG2 (Figure 9A) was shown.

[0097] Figure 10 The binding specificity of BE-915 to CD137 and other TNFRSF members was shown.

[0098] Figures 11A-11B The cross-competitive binding of BE-915 to human CD137 with CD137L was shown. CD137L blocked the binding of BE-915 to CD137 expressed on HuT78 (Figure 11A). BE-915 blocked the binding of CD137L to CD137 expressed on HuT78 (Figure 11B).

[0099] Figures 12A-12C The release of IL-2 and IFN-γ from human PBMC induced by BE-915 was shown. Figure 12A is a schematic diagram of activating CD137 by co-stimulating huPBMC with BE-915 and a hepatocellular carcinoma (HCC) cell line expressing OS8. Figures 12B-12C show that BE-915 induced dose-dependent cytokine release in PBMC in a GPC3-expression-dependent manner. PBMC from two donors were tested.

[0100] Figures 13A-13CShows the T cell killing activity of BE-915 induced in human PBMC. Figure 13A is a schematic diagram of activating CD137 by co-stimulating huPBMC with BE-915 in combination with an EpCAM / CD3 bispecific T cell engager (BiTE) that provides the first signal for T cell activation. Figures 13B-C show that BE-915 dose-dependently enhances the T cell killing activity against cells expressing GPC3, but not against GPC3-negative cells. PBMC from two donors were tested.

[0101] Figure 14 Shows the pharmacokinetic characteristics of BE-915 in cynomolgus monkeys after i.v. infusion (5 mg / kg, N = 2).

[0102] Figure 15 Shows the efficacy of BE-915 monotherapy in the MC38 / hGPC3 model of humanized CD137 knock-in mice.

[0103] Figure 16 Shows the efficacy of the combination of BE-915 and anti-PD-1 antibody in the LL / 2 / hGPC3 model of humanized CD137 knock-in mice.

[0104] Figure 17 Shows a schematic diagram of the partial competitive binding of VHH (BGA-2524) to CD137L and CD137. The crystal structure of VHH (BGA-2524) / CD137 was superimposed with the CD137L / CD137 complex (PDB: 6MGP) through the CD137 CRD1 and CRD2 domains. CD137, CD137L, and VHH (BGA-2524) are colored black, white, and gray, respectively.

[0105] Figure 18 Shows the binding of VHH (BGA-2524) to the CD137 dimer. Crystal structure analysis shows that VHH (BGA-2524) has the ability to bind to the CD137 dimer to promote CD137 aggregation. Each monomer of the CD137 dimer is shown in white or gray, and VHH (BGA-2524) on the surface is shown as a black cartoon (left). The epitope of VHH (BGA-2524) is shown in black on the surface of the CD137 dimer (right, with BGA-2524 removed).

[0106] Figure 19Atomic interactions on the binding surface of the VHH(BGA-2524) / CD137 complex are shown. The binding interface between VHH(BGA-2524) and CD137 recognizes certain key residues of VHH(BGA-2524) (para residues, underlined amino acids) and CD137 (epitope residues). Each monomer of the CD137 dimer is shown in white or gray cartoons, covered with a transparent surface, and the CRD1, CRD2, and CRD3 domains are represented by lines, respectively. Para residues are shown in black lines, and amino acids are underlined (most of the framework regions are removed). DETAILED DESCRIPTION

[0107] The present disclosure provides anti-CD137 antibodies and antigen-binding fragments thereof, as well as multispecific antibodies or antigen-binding fragments thereof that recognize CD137 as an antigen and recognize at least one tumor-associated antigen (TAA) as another antigen. The present disclosure also provides anti-GPC3xCD137 multispecific antibodies and antigen-binding fragments thereof. In addition, the present disclosure provides antibodies with desired pharmacokinetic properties, desired biophysical properties, and other desired properties, and can therefore be used to reduce the possibility of cancer or treat cancer. The present disclosure further provides pharmaceutical compositions comprising antibodies and methods for preparing and using such pharmaceutical compositions for preventing and treating cancer and related disorders. I. Anti-GPC3 Antibodies

[0108] The present disclosure provides antibodies or antigen-binding fragments thereof that specifically bind to human GPC3. In one embodiment, the anti-GPC3 antibody or antigen-binding fragment thereof is present at 1×10 -6 M to 1×10 -10 The binding affinity (K D ) specifically binds to human GPC3. In another embodiment, the anti-GPC3 antibody or antigen-binding fragment thereof is present at about 1×10 -6 M, about 1×10 -7 M, about 1×10 - 8 M, about 1×10 -9 M or about 1×10 -10 The binding affinity (K D ) binds to human GPC3.

[0109] In one embodiment, the anti-GPC3 antibody or antigen-binding fragment thereof comprises: a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:45, (b) HCDR2 of SEQ ID NO:46, and (c) HCDR3 of SEQ ID NO:47; and a light chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO:48, (e) LCDR2 of SEQ ID NO:49, and (f) LCDR3 of SEQ ID NO:50, according to Kabat numbering.

[0110] In another embodiment, the anti-GPC3 antibody or antigen-binding fragment thereof comprises: HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:41; and LCDR1, LCDR2, and LCDR3 from the light chain variable region (VL) shown in SEQ ID NO:43.

[0111] In another embodiment, the anti-GPC3 antibody or antigen-binding fragment thereof further comprises no more than one, two, three, four, or five amino acid deletions, insertions, or substitutions in the CDRs, preferably the amino acid substitutions are conservative amino acid substitutions, while maintaining binding specificity and affinity.

[0112] In another embodiment, the anti-GPC3 antibody or its antigen-binding fragment comprises a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:41, and a light chain variable region (VL) that comprises an amino acid sequence that is at least 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:43. In another embodiment, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 amino acids have been inserted, deleted, or substituted (optionally conservative amino acid substitutions) in SEQ ID NO:41 or SEQ ID NO:43. In another embodiment, such variations are in the framework regions of the variable regions. In another embodiment, the anti-GPC3 antibody or its antigen-binding fragment having such variations retains binding specificity and affinity.

[0113] In another embodiment, the anti-GPC3 antibody or its antigen-binding fragment comprises a heavy chain variable region (VH) comprising SEQ ID NO:41 and a light chain variable region (VL) comprising SEQ ID NO:43.

[0114] In another embodiment, the anti-human GPC3 antibody or its antigen-binding fragment exhibits cross-species binding activity to cynomolgus monkey GPC3. II. Anti-CD137 Antibody Table 1: Sequences of Anti-CD137 Antibodies

[0115] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137. The antibody or antigen-binding fragment of the present disclosure includes, but is not limited to, an antibody or an antigen-binding fragment thereof produced as described below.

[0116] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or antigen-binding fragment thereof specifically binds to the following epitope, which comprises amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35), consists essentially of or consists of the same, and optionally determines the epitope by X-ray diffraction.

[0117] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or antigen-binding fragment thereof specifically binds to the following epitope, which comprises amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35), consists essentially of or consists of the same, and optionally determines the epitope by X-ray diffraction.

[0118] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or antigen-binding fragment thereof specifically binds to a human CD137 dimer, which comprises a first human CD137 monomer and a second human CD137 monomer, or consists of the same; wherein the antibody or antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), which comprises amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67, consists essentially of or consists of the same; and the antibody or antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), which comprises amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98, consists essentially of or consists of the same, optionally determines the epitope by X-ray diffraction, and optionally the antibody or antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

[0119] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to an epitope comprising one or more amino acid residues selected from the group consisting of Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, and Gln67 of human CD137 (SEQ ID NO: 35).

[0120] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to an epitope comprising one or more amino acid residues selected from the group consisting of Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35).

[0121] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to a human CD137 dimer comprising or consisting of a first human CD137 monomer and a second human CD137 monomer, wherein the antibody or the antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35) comprising one or more amino acid residues selected from the group consisting of Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67, and the antibody or the antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35) comprising one or more amino acid residues selected from the group consisting of Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98, optionally the antibody or the antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

[0122] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to an epitope comprising one or more amino acid residues selected from the group consisting of Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67, Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35).

[0123] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to the following epitope, which is composed of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, and Gln67 of human CD137 (SEQ ID NO:35).

[0124] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof specifically binds to the following epitope, which is composed of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO:35).

[0125] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137 dimer, wherein the human CD137 dimer comprises or consists of a first human CD137 monomer and a second human CD137 monomer, and wherein the antibody or the antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 (SEQ ID NO:35) monomer, which is composed of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, and Gln67, and the antibody or the antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO:35), which is composed of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98, and wherein the antibody or the antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

[0126] The present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or the antigen-binding fragment thereof comprises a paratope containing one or more amino acid residues selected from the group consisting of Asn31, Tyr32, Ala33, Trp52, Ser53, Tyr55, His57, Leu98, Lys99, Tyr100, Pro101, Thr104, Thr106, Tyr109 (in the natural order of the sequence) or Asn31, Tyr32, Ala33, Trp52, Ser54, Tyr56, His58, Leu96, Lys97, Tyr98, Pro99, Thr100B, Thr100D, Tyr102 (Kabat nomenclature).

[0127] In one embodiment, an antibody or antigen-binding fragment thereof that specifically binds to human CD137 binds to the lateral surface of the CD137 CRD2 domain (e.g., of a human) mainly through CDR residues (e.g., Asn31, Tyr32, Ala33, Trp52, Ser54, Tyr56, His58, Leu96, Lys97, Tyr98, Pro99, Thr100B, Thr100D, Tyr102 (Kabat nomenclature) of a human CD137 VHH).

[0128] In some embodiments, the epitope of human CD137 to which an antibody or antigen-binding fragment thereof that specifically binds to the human CD137 disclosed herein binds is determined by X-ray diffraction.

[0129] The present disclosure provides an antibody or antigen-binding fragment that specifically binds to human CD137, wherein the antibody or antibody fragment (e.g., antigen-binding fragment) comprises a VH domain having an amino acid sequence of SEQ ID NO:4, SEQ ID NO:8, SEQ ID NO:6, SEQ ID NO:11, SEQ ID NO:13, SEQ ID NO:15, or SEQ ID NO:17 (Table 1). The present disclosure also provides an antibody or antigen-binding fragment that specifically binds to human CD137, wherein the antibody or antigen-binding fragment comprises an HCDR having an amino acid sequence of any one of the HCDRs listed in Table 1. In one aspect, the present disclosure provides an antibody or antigen-binding fragment that specifically binds to human CD137, wherein the antibody comprises (or alternatively, consists of) one, two, three, or more HCDRs having an amino acid sequence of any one of the HCDRs listed in Table 1.

[0130] In one embodiment, the antibody or antigen-binding fragment thereof comprises one or more complementarity-determining regions (CDRs) that comprise an amino acid sequence selected from SEQ ID NO:1, SEQ ID NO:2, SEQ ID NO:3, SEQ ID NO:10; or an amino acid sequence selected from SEQ ID NO:54, SEQ ID NO:55, SEQ ID NO:56; or an amino acid sequence selected from SEQ ID NO:57, SEQ ID NO:58, SEQ ID NO:59.

[0131] In one embodiment, the anti-CD137 antibody or antigen-binding fragment thereof comprises: (i) HCDR1 (heavy chain complementarity-determining region 1), HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:4; (ii) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:8; (iii) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:6; (iv) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:11; (v) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:13; (vi) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:15; or (vii) HCDR1, HCDR2, and HCDR3 from the heavy chain variable region (VH) shown in SEQ ID NO:17.

[0132] In one embodiment, the anti-CD137 antibody or antigen-binding fragment thereof comprises: (i) a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:2, and (c) HCDR3 of SEQ ID NO:3; or (ii) a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:10, and (c) HCDR3 of SEQ ID NO:3, according to Kabat numbering.

[0133] Other antibodies or antigen-binding fragments thereof of the present disclosure include amino acids that have been altered but have at least 60%, 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% percent identity with the CDR regions disclosed in Table 1 in the CDR regions. In some aspects, it includes amino acid alterations (insertions, deletions, or substitutions, optionally conservative amino acid substitutions) where no more than 1, 2, 3, 4, or 5 amino acids are altered in the CDR regions when compared to the CDR regions depicted in the sequences described in Table 1, while maintaining binding specificity and affinity.

[0134] Other antibodies of the present disclosure include those in which the amino acids or the nucleic acids encoding the amino acids have been altered, but have at least 60%, 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% percent identity with the sequences described in Table 1, optionally with the corresponding sequences of the CDRs unchanged. In some aspects, it includes alterations in the amino acid sequence in which when compared to the variable regions depicted in the sequences described in Table 1, no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 amino acids are altered in the variable region (e.g., the framework region of the variable region), while retaining the therapeutic activity / binding specificity / affinity, optionally with the corresponding sequences of the CDRs unchanged. In some aspects, it includes alterations in the amino acid sequence in which when compared to the variable regions depicted in the sequences described in Table 1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 amino acids are altered in the variable region (e.g., the framework region of the variable region), such as insertions, deletions, or substitutions (optionally conservative amino acid substitutions), while retaining the therapeutic activity / binding specificity / affinity, optionally with the corresponding sequences of the CDRs unchanged.

[0135] In some embodiments, the present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, the antibody or the antigen-binding fragment thereof comprising a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:2, and (c) HCDR3 of SEQ ID NO:3; wherein the amino acids F37, Y47, G49, and I94 (Kabat numbering) in the framework region are retained.

[0136] In some embodiments, the present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, the antibody or the antigen-binding fragment thereof comprising a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:10, and (c) HCDR3 of SEQ ID NO:3; wherein the amino acids F37, Y47, G49, and I94 (Kabat numbering) in the framework region are retained.

[0137] In some embodiments, the present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or antigen-binding fragment thereof comprises a VH domain having the amino acid sequence described in Table 1 or a variant thereof, wherein HCDR1, HCDR2, and HCDR3 are unchanged, and the amino acids F37, Y47, G49, and I94 (Kabat numbering) in the framework region are retained.

[0138] In some embodiments, the present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137, wherein the antibody or antigen-binding fragment thereof comprises a heavy chain variable region that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:4, SEQ ID NO:8, SEQ ID NO:6, SEQ ID NO:11, SEQ ID NO:13, SEQ ID NO:15, or SEQ ID NO:17, wherein HCDR1, HCDR2, and HCDR3 are unchanged, and the amino acids F37, Y47, G49, and I94 (Kabat numbering) in the framework region are retained.

[0139] In another embodiment, the present disclosure provides an antibody or an antigen-binding fragment thereof that specifically binds to human CD137 with a binding affinity (K -6 from 1×10 -10 M to 1×10 D M). In another embodiment, the anti-CD137 antibody or antigen-binding fragment thereof binds to human CD137 with a binding affinity (K -6 of approximately 1×10 -7 M, approximately 1×10 -8 M, approximately 1×10 -9 M, approximately 1×10 -10 M, or approximately 1×10 D M).

[0140] The present disclosure also provides nucleic acid sequences encoding the VH and full-length heavy chains of an antibody that specifically binds to human CD137. Such nucleic acid sequences can be optimized for expression in mammalian cells.

[0141] The present disclosure also provides antibodies and antigen-binding fragments thereof that bind the same epitope as the anti-CD137 antibodies described in Table 1. Thus, other antibodies and antigen-binding fragments thereof can be identified based on their ability to cross-compete with other antibodies in a binding assay (e.g., competitively inhibit their binding in a statistically significant manner). Testing the ability of an antibody to inhibit the binding of an antibody and antigen-binding fragment thereof of the present disclosure to CD137 demonstrates that the test antibody can compete with the antibody or its antigen-binding fragment for binding to CD137. Without being bound by any one theory, such an antibody can bind to the same or a related (e.g., structurally similar or spatially proximal) epitope on CD137 as the competing antibody or its antigen-binding fragment. In some aspects, an antibody that binds the same epitope on CD137 as the antibody or antigen-binding fragment thereof of the present disclosure is a human or humanized monoclonal antibody. Such human or humanized monoclonal antibodies can be prepared and isolated as described herein.

[0142] In some embodiments, the anti-CD137 antibody comprises at least one antigen-binding site and at least one variable region. In some embodiments, the anti-CD137 antibody comprises an antigen-binding fragment from a CD137 antibody described herein. In some embodiments, the anti-CD137 antibody is isolated or recombinant. In some embodiments, the anti-CD137 antibody also encompasses multispecific antibodies that target CD137 as at least one arm and one or more other antigens as one or more additional arms. III. Anti-CD137 Multispecific Antibodies

[0143] In one embodiment, the anti-CD137 antibodies disclosed herein can be used to construct multispecific antibodies in combination with other forms (e.g., human tumor-associated antigens (TAAs), immune checkpoint molecules, or immunostimulatory factors).

[0144] In one embodiment, the anti-CD137 antibodies disclosed herein can be incorporated into an anti-CD137 x TAA multispecific antibody, where TAA is an antibody or fragment thereof directed against any human tumor-associated antigen. The antibody molecule is a multispecific antibody molecule, e.g., it comprises multiple antigen-binding domains, where at least one antigen-binding domain sequence specifically binds to a human TAA as a first antigen / epitope and a second antigen-binding domain sequence specifically binds to human CD137 as a second antigen / epitope. In one embodiment, the multispecific antibody comprises a third, fourth, or fifth antigen-binding domain. In one embodiment, the multispecific antibody is a bispecific, trispecific, or tetra-specific antibody. In each instance, the multispecific antibody comprises at least one anti-TAA antigen-binding domain and at least one anti-CD137 antigen-binding domain.

[0145] In one embodiment, the multispecific antibody is a bispecific antibody. As used herein, a bispecific antibody specifically binds only two antigens. The bispecific antibody comprises a first antigen-binding domain that specifically binds to a TAA and a second antigen-binding domain that specifically binds to human CD137. This includes bispecific antibodies that comprise a heavy-chain variable domain and a light-chain variable domain that specifically bind to a TAA and a heavy-chain variable domain that specifically binds to human CD137. In some embodiments, the bispecific antibody comprises an antigen-binding fragment, wherein the antigen-binding fragment can be a Fab, F(ab’)2, Fv, single-chain Fv (scFv), or single-domain antibody.

[0146] In some embodiments, the second antigen-binding domain that specifically binds to human CD137 comprises the anti-CD137 antibodies disclosed in Section II.

[0147] In one embodiment, the multispecific antibodies disclosed herein bind to human TAA and / or human CD137 with a binding affinity (K -6 d) of 1×10 -10 M to 1×10 D M. In another embodiment, the multispecific antibodies disclosed herein bind to human TAA and / or human CD137 with a binding affinity (K -6 d) of approximately 1×10 -7 M, approximately 1×10 -8 M, approximately 1×10 -9 M, approximately 1×10 -10 M, or approximately 1×10 D M.

[0148] In one embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the first antigen-binding domain specifically binds to a human TAA, and the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:2, (c) HCDR3 of SEQ ID NO:3; or (ii) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:10, (c) HCDR3 of SEQ ID NO:3.

[0149] In another embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the first antigen-binding domain specifically binds to a human TAA, and the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 17; (ii) a heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 11; (iii) a heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 13; (iv) a heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 15; or (v) a heavy-chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 4.

[0150] In another embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the first antigen-binding domain specifically binds to a human TAA, and the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises SEQ ID NO: 4; (ii) a heavy-chain variable region (VH) that comprises SEQ ID NO: 11; (iii) a heavy-chain variable region (VH) that comprises SEQ ID NO: 13; (iv) a heavy-chain variable region (VH) that comprises SEQ ID NO: 15; or (v) a heavy-chain variable region (VH) that comprises SEQ ID NO: 17.

[0151] In one embodiment, the TAA is human GPC3. The first antigen-binding domain that specifically binds to human GPC3 comprises the anti-GPC3 antibodies disclosed in Section I.

[0152] The present disclosure provides multivalent antibodies (e.g., tetravalent antibodies) having at least two antigen-binding domains, which can be readily produced by recombinant expression of nucleic acids encoding antibody polypeptide chains. The multivalent antibodies herein comprise three to eight, but preferably four, antigen-binding domains that specifically bind to at least two antigens. IV. Anti-GPC3xCD137 Multispecific Antibodies

[0153] In one embodiment, the anti-GPC3 and anti-CD137 antibodies disclosed herein can be incorporated into an anti-GPC3xCD137 multispecific antibody. The antibody molecule is a multispecific antibody molecule that, for example, comprises multiple antigen-binding domains, wherein at least one antigen-binding domain sequence specifically binds to GPC3 as a first epitope / antigen, while a second antigen-binding domain sequence specifically binds to CD137 as a second epitope / antigen. In one embodiment, the multispecific antibody comprises a third, fourth, or fifth antigen-binding domain. In one embodiment, the multispecific antibody is a bispecific antibody, trispecific antibody, or tetravalent antibody. In each instance, the multispecific antibody comprises at least one anti-GPC3 antigen-binding domain and at least one anti-CD137 antigen-binding domain.

[0154] In one embodiment, the multispecific antibody is a bispecific antibody. As used herein, a bispecific antibody specifically binds only two antigens. The bispecific antibody comprises a first antigen-binding domain that specifically binds to human GPC3 and a second antigen-binding domain that specifically binds to human CD137. This includes bispecific antibodies that comprise a heavy-chain variable domain and a light-chain variable domain that specifically bind to human GPC3 as a first epitope / antigen and a heavy-chain variable domain that specifically binds to human CD137 as a second epitope / antigen. In some embodiments, the bispecific antibody comprises an antigen-binding fragment, wherein the antigen-binding fragment can be a Fab, F(ab’)2, Fv, single-chain Fv (scFv), or single-domain antibody.

[0155] The present disclosure provides a multispecific antibody or an antigen-binding fragment thereof that comprises a first antigen-binding domain that specifically binds to human glypican-3 (GPC3) and a second antigen-binding domain that specifically binds to human CD137.

[0156] The first antigen-binding domain that specifically binds to human glypican-3 (GPC3) includes the anti-GPC3 antibodies described in Section I. The second antigen-binding domain that specifically binds to human CD137 includes the anti-CD137 antibodies disclosed in Section II.

[0157] In one embodiment, the multispecific antibody of the present disclosure binds to human GPC3 and / or human CD137 with a binding affinity (K -6 M to 1×10 -10 ) of 1×10 D In another embodiment, the multispecific antibody of the present disclosure binds to human GPC3 and / or human CD137 with a binding affinity of approximately 1×10 - 6M, about 1×10 -7 M, about 1×10 -8 M, about 1×10 -9 M or about 1×10 -10 M of the binding affinity (K D ) binds to human GPC3 and / or human CD137.

[0158] In one embodiment, the multispecific antibody of the present disclosure specifically binds to human GPC3 and exhibits high affinity for both human GPC3 and monkey GPC3. In another embodiment, the multispecific antibody of the present disclosure specifically binds to human CD137. In another embodiment, the multispecific antibody of the present disclosure exhibits high affinity for both human CD137 and monkey CD137.

[0159] In one embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 45, (b) HCDR2 of SEQ ID NO: 46, (c) HCDR3 of SEQ ID NO: 47; and a light-chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO: 48, (e) LCDR2 of SEQ ID NO: 49, (f) LCDR3 of SEQ ID NO: 50, according to Kabat numbering; and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, (c) HCDR3 of SEQ ID NO: 3; or (ii) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, (c) HCDR3 of SEQ ID NO: 3, according to Kabat numbering.

[0160] In another embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: a heavy-chain variable region (VH) comprising SEQ ID NO: 41, and a light-chain variable region (VL) comprising SEQ ID NO: 43; and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises SEQ ID NO: 4; (ii) a heavy-chain variable region (VH) that comprises SEQ ID NO: 11; (iii) a heavy-chain variable region (VH) that comprises SEQ ID NO: 13; (iv) a heavy-chain variable region (VH) that comprises SEQ ID NO: 15; or (v) a heavy-chain variable region (VH) that comprises SEQ ID NO: 17.

[0161] In another embodiment, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, wherein the multispecific antibody or antigen-binding fragment is (i) BE-933, which comprises a first polypeptide of SEQ ID NO: 27 and a second polypeptide of SEQ ID NO: 23; (ii) BE-774, which comprises a first polypeptide of SEQ ID NO: 25 and a second polypeptide of SEQ ID NO: 23; (iii) BE-653, which comprises a first polypeptide of SEQ ID NO: 29 and a second polypeptide of SEQ ID NO: 23; (iv) BE-915, which comprises a first polypeptide of SEQ ID NO: 21 and a second polypeptide of SEQ ID NO: 23; (v) BE-647, which comprises a first polypeptide of SEQ IDNO: 31 and a second polypeptide of SEQ ID NO: 23; or (vi) BE-621, which comprises a first polypeptide of SEQ ID NO: 33 and a second polypeptide of SEQ ID NO: 23.

[0162] Other multispecific antibodies or antigen-binding fragments thereof disclosed herein include those in which the amino acids or the nucleic acids encoding the amino acids have been altered, but have at least 60%, 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% percent identity to the sequences described herein (e.g., the CDRs are not altered). In some aspects, it includes alterations of the amino acid sequence in which when compared to the variable regions described herein, no more than 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 amino acids are altered in the variable region (e.g., framework region), while retaining the therapeutic activity / binding specificity / affinity. V. Others Form and module ratio

[0163] In one embodiment, the anti-CD137 antibodies disclosed herein can be used to construct multispecific antibodies together with other forms (e.g., TAAs, immune checkpoint or immune stimulatory factors). TAAs (e.g., GPC3) are used as examples for the forms and ratios described below. The description of GPC3 in the following examples can also be applied to other TAAs.

[0164] The multispecific antibodies disclosed herein can be in different forms. In one embodiment, the multispecific antibodies disclosed herein have the forms disclosed below, including: (1) Form A provides a symmetric IgG-like multispecific molecule with a Fab×VH configuration. The anti-huCD137 VH domain antibody is fused to the C-terminus of the Fc (CH3 domain) of the anti-GPC3 antibody, with a linker therebetween, as Figure 3 shown; (2) Form B also provides a symmetric IgG-like multispecific molecule with a Fab×VH configuration. The anti-huCD137 VH domain antibody is fused to the C-terminus of the light chain (Cκ) of the anti-GPC3 antibody, with a linker therebetween; (3) Form C provides a symmetric VH antibody-like multispecific molecule with a Fab×VH configuration. The Fab region of the anti-GPC3 antibody is fused to the N-terminus of the VH of the anti-huCD137 VH domain Ab, with a linker therebetween; and (4) Form D also provides a symmetric IgG-like multispecific molecule with a Fab×VH configuration. The anti-huCD137 VH domain antibody is fused to the N-terminus of the heavy chain (VH) of the anti-GPC3 antibody, with a linker therebetween. In one embodiment, the multispecific antibody is Form A, as Figure 3 shown.

[0165] The multispecific antibodies of the present disclosure can be constructed with different module ratios such as 1:1. In one embodiment, an inert Fc can be used for the multispecific antibody, and the Azymetric TM platform from Zymeworks can be used to assemble the Fab×VH configuration, in which the ZW1 mutations (A chain: T350V / L351Y / F405A / Y407V; B chain: T350V / T366L / K392L / T394W) can be introduced into the CH3 domain of the heavy chain to allow efficient heterodimer formation (Von Kreudenstein et al., (2013) Mabs [Monoclonal Antibodies] 5(5):646-54, incorporated by reference in its entirety). In one aspect, a specific ratio activates CD137 in a GPC3-dependent manner and does not activate CD137 in the absence of GPC3.

[0166] In one embodiment, the multispecific antibody or its antigen-binding fragment comprises: a) a first polypeptide that, from the N-terminus to the C-terminus, comprises: a first heavy-chain variable region (such as a first heavy-chain variable region); a CH1 domain, an Fc domain, and a second heavy-chain variable region (such as a second heavy-chain variable region); optionally, the C-terminus of the Fc domain is linked to the N-terminus of the second heavy-chain variable region by a linker; and b) a second polypeptide that, from the N-terminus to the C-terminus, comprises: a first light-chain variable region (such as a first light-chain variable region); and a first light-chain constant region; wherein the first heavy-chain variable region and the first light-chain variable region form a first antigen-binding domain that specifically binds to human GPC3, and the second heavy-chain variable region forms a second antigen-binding domain that specifically binds to human CD137. In another embodiment, the multispecific antibody or its antigen-binding fragment comprises two first polypeptides and two second polypeptides. Linker

[0167] It should also be understood that the domains and / or regions of the polypeptide chains of the bispecific antibody can be separated by linkers of various lengths. In some embodiments, the antigen-binding domains are separated from each other by a linker region from CL, CH1, hinge, CH2, CH3, or the entire Fc. For example, VL1-CL-(linker)VH2-CH1. Such linker regions can comprise randomly sorted amino acids or a restricted set of amino acids. Such linker regions can be flexible or rigid (see, for example, US2009 / 0155275, incorporated by reference in its entirety).

[0168] Multispecific antibodies have been constructed by gene fusion of two single-chain Fv (scFv) or Fab fragments, with or without a flexible linker (Mallender et al., J. Biol. Chem. 1994; 269: 199-206; Mack et al., Proc. Natl. Acad. Sci. USA 1995; 92: 7021-5; Zapata et al., Protein Eng. 1995; 8: 1057-62), via dimerization devices such as leucine zippers (Kostelny et al., J. Immunol. 1992; 148: 1547-53; de Kruif et al., J. Biol. Chem. 1996; 271: 7630-4) and Ig C / CH1 domains (Muller et al., FEBS Lett. 1998; 422: 259-64); by diabodies (Holliger et al., Proc. Nat. Acad. Sci. USA 1993; 90: 6444-8; Zhu et al., Bio / Technology (NY) 1996; 14: 192-6); Fab-scFv fusions (Schoonjans et al., J. Immunol. 2000; 165: 7050-7); and minibody forms (Pack et al., Biochemistry 1992; 31: 1579-84; Pack et al., Bio / Technology 1993; 11: 1271-7). Each reference mentioned in this paragraph is incorporated by reference in its entirety.

[0169] The multispecific antibodies disclosed herein comprise a linker region of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or more amino acid residues between one or more of their antigen-binding domains, CL domain, CH1 domain, hinge region, CH2 domain, CH3 domain or Fc region. In some embodiments, the amino acids glycine and serine are included in the linker region. In another embodiment, the linker can be GS (SEQ ID NO:97), GGS (SEQ ID NO:98), GSG (SEQ ID NO:99), SGG (SEQ ID NO:100), GGG (SEQ ID NO:101), GGGS (SEQ ID NO:60), SGGG (SEQ ID NO:61), GGGGS (SEQ ID NO:62), GGGGSGS (SEQ ID NO:63), GGGGSGS (SEQ ID NO:64), GGGGSGGS (SEQ ID NO:65), GGGGSGGGGS (SEQ ID NO:66), GGGGSGGGGSGGGGS (SEQ ID NO:67), AKTTPKLEEGEFSEAR (SEQ ID NO:68), AKTTPKLEEGEFSEARV (SEQ ID NO:69), AKTTPKLGG (SEQ ID NO:70), SAKTTPKLGG (SEQ ID NO:71), AKTTPKLEEGEFSEARV (SEQ ID NO:72), SAKTTP (SEQ ID NO:73), SAKTTPKLGG (SEQ IDNO:74), RADAAP (SEQ ID NO:75), RADAAPTVS (SEQ ID NO:76), RADAAAAGGPGS (SEQ ID NO:77), RADAAAA(G4S)4 (SEQ ID NO:78), SAKTTP (SEQ ID NO:79), SAKTTPKLGG (SEQ ID NO:80), SAKTTPKLEEGEFSEARV (SEQ ID NO:81), ADAAP (SEQ ID NO:82), ADAAPTVSIFPP (SEQ ID NO:83), TVAAP (SEQ ID NO:84), TVAAPSVFIFPP (SEQ ID NO:85), QPKAAP (SEQ ID NO:86), QPKAAPSVTLFPP (SEQ ID NO:87), AKTTPP (SEQ IDNO:88), AKTTPPSVTPLAP (SEQ ID NO:89), AKTTAP (SEQ ID NO:90), AKTTAPSVYPLAP (SEQ ID NO:91), ASTKGP (SEQ ID NO:92), ASTKGPSVFPLAP (SEQ ID NO:93), GENKVEYAPALMALS (SEQ ID NO:94), GPAKELTPLKEAKVS (SEQ ID NO:95), and GHEAAAVMQVQYPAS (SEQ ID NO:96) or any combination thereof (see WO 2007 / 024715, which is incorporated by reference in its entirety). Dimerization-specific amino acids

[0170] In one embodiment, the multivalent antibody comprises at least one dimerization-specific amino acid alteration. The dimerization-specific amino acid alteration results in a "protrude-into-hole" interaction and increases the assembly of the correct multivalent antibody. The dimerization-specific amino acids can be within the CH1 domain or the CL domain or a combination thereof. The dimerization-specific amino acids for pairing the CH1 domain with other CH1 domains (CH1-CH1) and the CL domain with other CL domains (CL-CL) can be found at least in the disclosures of WO 2014082179, the WO 2015181805 family, and WO 2017059551, all of which are incorporated by reference in their entirety. The dimerization-specific amino acids can also be within the Fc domain and can be combined with the dimerization-specific amino acids within the CH1 or CL domain. In one embodiment, the present disclosure provides a bispecific antibody comprising at least one pair of dimerization-specific amino acids. Alterations in the Fc region

[0171] The Fc region, if present, can be a wild-type Fc region of the IgG1, IgG2, IgG3, or IgG4 subclass.

[0172] In one embodiment, the antibody, multispecific antibody, or antigen-binding fragment thereof comprises an Fc domain of IgG1 or IgG4 with reduced effector function. In another embodiment, the Fc domain comprises the amino acid sequence of SEQ ID NO:53 or SEQ ID NO:19. In another embodiment, the IgG1 Fc comprises the mutations E233P, L234A, L235A, G236del, and P329A.

[0173] In one embodiment, the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain with an extended half-life. In another embodiment, the multispecific antibody or antigen-binding fragment thereof comprises the Fc domain of IgG1, wherein the YTE mutation (M252Y / S254T / T256E, EU numbering, as described in US7658921 and incorporated herein by reference in its entirety) located in the CH2 of the IgG Fc region is introduced.

[0174] In one embodiment, the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain with reduced effector function and an extended half-life. In another embodiment, the Fc domain comprises the amino acid sequence of SEQ ID NO:20.

[0175] In another embodiment, the antibodies of the present disclosure have strong Fc-mediated effector function, and the antibodies mediate antibody-dependent cell cytotoxicity (ADCC) against target cells expressing TAA (e.g., GPC3).

[0176] In other aspects, the Fc region is altered by substituting at least one amino acid residue with a different amino acid residue to alter the effector function of the antibody. For example, one or more amino acids can be substituted with different amino acid residues such that the antibody has an altered affinity for effector ligands while retaining the antigen-binding ability of the parental antibody. The effector ligands with altered affinity can be, for example, Fc receptors or the C1 component of complement. This method is described, for example, in U.S. Patent Nos. 5,624,821 and 5,648,260 to Winter et al., both of which are incorporated herein by reference in their entirety.

[0177] In another aspect, one or more amino acid residues can be substituted with one or more different amino acid residues such that the antibody has altered C1q binding and / or reduced or eliminated complement-dependent cytotoxicity (CDC). This method is described, for example, in U.S. Patent No. 6,194,551 to Idusogie et al., which is incorporated herein by reference in its entirety.

[0178] In another aspect, one or more amino acid residues are altered to change the ability of the antibody to fix complement. This method is described, for example, in Published WO 94 / 29351 to Bodmer et al., which is incorporated herein by reference in its entirety. In certain aspects, one or more amino acids of the antibodies or antigen-binding fragments of the present disclosure are replaced with one or more allotypic amino acid residues of the IgG1 subclass and κ isotype. Allotypic amino acid residues also include, but are not limited to, the heavy chain constant regions of the IgG1, IgG2, and IgG3 subclasses and the light chain constant region of the κ isotype, as described by Jefferis et al., Mabs. [Monoclonal Antibodies] 2009; 1:332-338, which is incorporated herein by reference in its entirety.

[0179] In another aspect, the Fc region is modified by modifying one or more amino acids to increase the ability of the antibody to mediate antibody-dependent cell cytotoxicity (ADCC) and / or to increase the affinity of the antibody for Fcγ receptors. This method is described, for example, in the publication WO 00 / 42072 by Presta, which is incorporated herein by reference in its entirety. In addition, the binding sites for FcγRI, FcγRII, FcγRIII, and FcRn on human IgG1 have been mapped, and variants with improved binding have been described (see Shields et al., J. Biol. Chem. [Journal of Biological Chemistry] 2001; 276:6591-6604), which is incorporated herein by reference in its entirety.

[0180] In another aspect, the glycosylation of the multispecific antibody is modified. For example, aglycosylated antibodies can be prepared (i.e., the antibody lacks or has reduced glycosylation). For example, the glycosylation can be altered to increase the affinity of the antibody for the "antigen". Such carbohydrate modifications can be achieved, for example, by altering one or more glycosylation sites within the antibody sequence. For example, one or more amino acid substitutions can be made that result in the elimination of one or more variable region framework glycosylation sites, thereby eliminating glycosylation at that site. Such aglycosylation can increase the affinity of the antibody for the antigen. This method is described, for example, in U.S. Patent Nos. 5,714,350 and 6,350,861 by Co et al., both of which are incorporated herein by reference in their entirety.

[0181] Additionally or alternatively, antibodies with altered glycosylation patterns can be prepared, such as hypofucosylated antibodies with a reduced amount of fucosyl residues or antibodies with an increased bisecting GlcNac structure. Such altered glycosylation patterns have been shown to increase the ADCC ability of antibodies. Such carbohydrate modifications can be achieved, for example, by expressing the antibody in host cells with an altered glycosylation pathway. Cells with an altered glycosylation pathway have been described in the art and can be used as host cells in which to express a recombinant antibody to produce an antibody with altered glycosylation. For example, EP 1,176,195 to Hang et al. (incorporated herein by reference in its entirety) describes a cell line with a functionally disrupted FUT8 gene that encodes fucosyltransferase such that antibodies expressed in such a cell line exhibit hypofucosylation. The disclosure of Presta, WO 03 / 035835 (incorporated herein by reference in its entirety), describes the variant CHO cell line Lecl3 cells, which have a reduced ability to attach fucose to Asn(297)-linked carbohydrates, also resulting in hypofucosylation of antibodies expressed in this host cell (see also Shields et al., J. Biol. Chem. [Journal of Biological Chemistry] 2002; 277:26733-26740, incorporated herein by reference in its entirety). WO 99 / 54342 to Umana et al. (incorporated herein by reference in its entirety) describes a cell line engineered to express a glycosyltransferase that modifies glycoproteins (e.g., β(1,4)-N-acetylglucosaminyltransferase III (GnTIII)) such that antibodies expressed in the engineered cell line exhibit an increased bisecting GlcNac structure, which results in an increase in the ADCC activity of the antibody (see also Umana et al., Nat. Biotech. [Nature Biotechnology] 1999; 17:176-180, incorporated herein by reference in its entirety).

[0182] On the other hand, if a reduction in ADCC is desired, many previous reports have shown that the human antibody subclass IgG4 has only modest ADCC and little or no CDC effector function (Moore G L, et al., MAbs. 2010;2:181-189, incorporated herein by reference in its entirety). However, native IgG4 has been found to be less stable under stress conditions such as in acidic buffer or at elevated temperatures (Angal, S. Mol Immunol. 1993;30:105-108; Dall'Acqua, W. et al., 1998 Biochemistry, 37:9266-9273; Aalberse et al., Immunol. 2002;105:9-19, all incorporated herein by reference in their entirety). Reduced ADCC can be achieved by operably linking the antibody to an IgG4 Fc engineered with a combination of alterations that reduce FcγR binding or C1q binding activity, thereby reducing or eliminating ADCC and CDC effector functions. Considering the physicochemical properties of antibodies as biopharmaceuticals, one of the less desirable intrinsic properties of IgG4 is that its two heavy chains dynamically separate in solution to form half-antibodies, which results in the generation of bispecific antibodies in vivo through a process called "Fab arm exchange" (Van der Neut Kolfschoten M, et al., Science. 2007;317:1554-157, incorporated herein by reference in its entirety). Mutation of serine at position 228 (EU numbering system) to proline has been shown to inhibit the separation of IgG4 heavy chains (Angal, S. Mol Immunol. 1993;30:105-108, incorporated herein by reference in its entirety; Aalberse et al., Immunol. 2002;105:9-19, incorporated herein by reference in its entirety).It has been reported that some amino acid residues in the hinge and γFc regions affect the interaction of antibodies with Fcγ receptors (Chappel S M, et al., Proc. Natl. Acad. Sci. USA [Proceedings of the National Academy of Sciences of the United States of America] 1991; 88:9036 - 9040; Mukherjee, J., et al., FASEB J [The FASEB Journal] 1995; 9:115 - 119; Armour, K. L., et al., Eur J Immunol [European Journal of Immunology] 1999; 29:2613 - 2624; Clynes, R. A., et al., 2000 Nature Medicine, 6:443 - 446; Arnold J. N., Annu Rev Immunol [Annual Review of Immunology] 2007; 25:21 - 50, each incorporated by reference in its entirety). In addition, some rare IgG4 isotypes in the human population can also confer different physicochemical properties (Brusco, A., et al., Eur J Immunogenet [European Journal of Immunogenetics] 1998; 25:349 - 55; Aalberse et al., Immunol [Immunology] 2002; 105:9 - 19, each incorporated by reference in its entirety). To generate multispecific antibodies with low ADCC and CDC but good stability, the hinge region and Fc region of human IgG4 can be modified and many changes can be introduced. These modified IgG4 Fc molecules can be found in SEQ ID NOs: 83 - 88 of U.S. Patent No. 8,735,553 to Li et al., which is incorporated by reference in its entirety.

[0183] In another embodiment, the antibodies of the present disclosure comprise an Fc domain of human IgG4 having an S228P and / or R409K substitution (according to the EU numbering system). Antibody production

[0184] Antibodies and their antigen - binding fragments can be produced by any method known in the art, including but not limited to recombinant expression of antibody tetramers, chemical synthesis, and enzymatic digestion, while full - length monoclonal antibodies can be obtained by, for example, hybridoma or recombinant production. Recombinant expression can be from any suitable host cell known in the art, such as mammalian host cells, bacterial host cells, yeast host cells, insect host cells, etc.

[0185] The present disclosure further provides polynucleotides encoding the antibodies described herein, such as polynucleotides encoding a heavy or light chain variable region or segment comprising the complementarity determining regions described herein. In some aspects, the polynucleotide encoding the heavy chain variable region or the light chain variable region has at least 85%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% nucleic acid sequence identity to a polynucleotide selected from the group consisting of SEQ ID NO:5, SEQ ID NO:9, SEQ ID NO:7, SEQ ID NO:12, SEQ ID NO:14, SEQ ID NO:16, SEQ ID NO:18, SEQ ID NO:42 or SEQ ID NO:44.

[0186] The polynucleotides of the present disclosure can encode the variable region sequences of anti-TAA (such as GPC3) x CD137 antibodies. They can also encode the variable and constant regions of the antibody. Some polynucleotide sequences encode polypeptides comprising the variable regions of the heavy and light chains of exemplary anti-TAA (such as GPC3) x CD137 antibodies.

[0187] The present disclosure further provides polynucleotides encoding the anti-GPC3 x CD137 antibodies described herein. In some aspects, the polynucleotide encoding the first polypeptide or the second polypeptide of the anti-GPC3 x CD137 antibody has at least 85%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% nucleic acid sequence identity to a polynucleotide selected from the group consisting of SEQ ID NO:22, SEQ ID NO:26, SEQ ID NO:28, SEQ ID NO:30, SEQ ID NO:32, SEQ ID NO:34 or SEQ ID NO:24. In some embodiments, the polynucleotides described herein can be codon-optimized for expression in a host cell (such as, a eukaryotic cell, more particularly, a mammalian cell (such as, a CHO cell)).

[0188] The present disclosure also provides expression vectors and host cells for producing the antibodies herein (e.g., anti-CD137 antibodies and anti-TAA (e.g., GPC3)xCD137 antibodies). The choice of expression vector depends on the intended host cell of the expression vector. Generally, the expression vector contains a promoter and other regulatory sequences (e.g., enhancers) operably linked to a polynucleotide encoding an antibody chain or antigen-binding fragment. In some aspects, an inducible promoter is used to prevent expression of the inserted sequence except under the control of inducing conditions. Inducible promoters include, for example, the arabinose, lacZ, metallothionein promoter, or heat shock promoter. The culture of the transformed organism can be expanded under non-inducing conditions and without biasing the population towards coding sequences for which the host cell better tolerates their expression products. In addition to the promoter, other regulatory elements may be required or desirable for efficient expression of the antibody or antigen-binding fragment. These elements typically include the ATG start codon and adjacent ribosome-binding site or other sequences. Furthermore, the expression efficiency can be increased by including an enhancer suitable for the cell system in use (see, e.g., Scharf et al., Results Probl. Cell Differ. [Results and Problems in Cell Differentiation] 1994; 20:125; and Bittner et al., Meth. Enzymol. [Methods in Enzymology] 1987; 153:516, both incorporated by reference in their entireties). For example, the SV40 enhancer or CMV enhancer can be used to increase expression in mammalian host cells.

[0189] Host cells for carrying and expressing antibody chains can be prokaryotic or eukaryotic cells. Escherichia coli is a prokaryotic host that can be used for cloning and expressing the polynucleotides of the present disclosure. Other suitable microbial hosts include bacilli, such as Bacillus subtilis, and other enterobacteriaceae, such as Salmonella, Serratia, and various Pseudomonas species. In these prokaryotic hosts, expression vectors can also be prepared, which typically contain expression control sequences (e.g., origin of replication) compatible with the host cell. In addition, there will be any number of well-known promoters, such as the lactose promoter system, the tryptophan (trp) promoter system, the β-lactamase promoter system, or the promoter system from bacteriophage λ. The promoter typically optionally controls expression with an operon sequence and has ribosome binding site sequences, etc., for initiating and completing transcription and translation. Other microorganisms such as yeast can also be used for expressing antibodies. A combination of insect cells and baculovirus vectors can also be used. In other aspects, mammalian host cells are used for expressing and producing the antibodies of the present disclosure. For example, they can be hybridoma cell lines expressing endogenous immunoglobulin genes or mammalian cell lines carrying exogenous expression vectors. These include any normally dying or normal or abnormally immortalized animal or human cells. For example, several suitable host cell lines capable of secreting intact immunoglobulins have been developed, including CHO cell lines, various COS cell lines, HEK 293 cells, myeloma cell lines, transformed B cells, and hybridomas. The use of mammalian tissue cell cultures for expressing polypeptides is generally discussed, for example, in Winnacker, From Genes to Clones, VCH Publishers, New York, N.Y., 1987, which is incorporated herein by reference in its entirety. Expression vectors for mammalian host cells can include expression control sequences, such as origin of replication, promoter, and enhancer (see, e.g., Queen et al., Immunol. Rev. 1986; 89:49-68, which is incorporated herein by reference in its entirety), as well as necessary processing information sites, such as ribosome binding sites, RNA splicing sites, polyadenylation sites, and transcription terminator sequences. These expression vectors usually contain promoters derived from mammalian genes or mammalian viruses. Suitable promoters can be constitutive, cell type-specific, stage-specific, and / or regulatable or adjustable.Useful promoters include, but are not limited to, the metallothionein promoter, the constitutive adenovirus major late promoter, the dexamethasone-inducible MMTV promoter, the SV40 promoter, the MRPpolIII promoter, the constitutive MPSV promoter, the tetracycline-inducible CMV promoter (such as the human immediate-early CMV promoter), the constitutive CMV promoter, and promoter-enhancer combinations known in the art. Generation of bispecific antibodies

[0190] The current standard for engineering heterodimeric antibody Fc domains is the knobs-into-holes (KiH) design, which introduces mutations at the core CH3 domain interface. The resulting heterodimer has a reduced CH3 melting temperature (69 °C or lower). In contrast, the Zymeworks Azymetric TM platform (supra) heterodimeric Fc design has a thermal stability of 81.5 °C, which is comparable to that of the wild-type CH3 domain. Pharmaceutical compositions

[0191] Also provided are compositions comprising the antibodies or antigen-binding fragments thereof herein or polynucleotides comprising sequences encoding the antibodies or antigen-binding fragments herein, including pharmaceutical formulations. In certain embodiments, the composition comprises one or more antibodies or antigen-binding fragments herein, or one or more polynucleotides comprising sequences encoding one or more antibodies or antigen-binding fragments herein. These compositions may also comprise suitable carriers, such as pharmaceutically acceptable excipients well known in the art, including buffers.

[0192] The compositions disclosed herein can be in a variety of forms. These include, for example, liquid, semi-solid, and solid dosage forms, such as liquid solutions (e.g., injectable and infusion solutions), dispersions or suspensions, liposomes, and suppositories. The appropriate form depends on the intended mode of administration and therapeutic application. A typical suitable composition is in the form of an injectable or infusion solution. A suitable mode of administration is parenteral (e.g., intravenous, subcutaneous, intraperitoneal, intramuscular). In some embodiments, the antibody is administered by intravenous infusion or injection. In certain embodiments, the antibody is administered by intramuscular or subcutaneous injection. Detection and diagnostic methods

[0193] The antibodies or antigen-binding fragments disclosed herein can be used in a variety of applications, including, but not limited to, methods for detecting CD137 or GPC3. In one aspect, the antibody or antigen-binding fragment can be used to detect the presence of CD137 or GPC3 in a biological sample. As used herein, the term "detect" includes quantitative or qualitative detection. In certain aspects, the biological sample includes cells or tissues. In other aspects, such tissues include normal and / or cancerous tissues that express CD137 or GPC3 at higher levels relative to other tissues.

[0194] In one aspect, the present disclosure provides a method for detecting the presence of CD137 or GPC3 in a biological sample. In certain aspects, the method includes contacting the biological sample with an antibody herein under conditions that permit binding of the antibody to the antigen, and detecting whether a complex is formed between the antibody and the antigen. The biological sample can include, but is not limited to, urine, tissue, sputum, or blood samples.

[0195] Also included is a method for diagnosing a disorder associated with GPC3 expression. In certain aspects, the method includes contacting test cells with an anti-GPC3xCD137 antibody; determining the expression level (quantitative or qualitative) of GPC3 expressed by the test cells by detecting the binding of the anti-GPC3xCD137 antibody to the GPC3 polypeptide; and comparing the expression level of the test cells with the expression level of GPC3 in control cells (e.g., normal cells of the same tissue origin as the test cells or cells that do not express GPC3), wherein a higher level of GPC3 expression in the test cells compared to the control cells indicates the presence of a disorder associated with GPC3 expression. VI. Therapeutic Methods Anti-CD137 Antibody

[0196] The antibodies or antigen-binding fragments of the present disclosure can be used in a variety of applications, including but not limited to methods for treating CD137-related disorders or diseases. In one aspect, the CD137-related disorder or disease is cancer. In the case of a CD137xTAA bispecific antibody, the cancer can be TAA-specific, and CD137 is used to recruit immune cells to tumors expressing the TAA.

[0197] In one aspect, the present disclosure provides a method for treating cancer. In certain aspects, the method includes administering to a patient in need a therapeutically effective amount of an anti-CD137 antibody or antigen-binding fragment or a bispecific antibody containing CD137, or a pharmaceutical composition thereof. In another aspect, the present disclosure provides an anti-CD137 antibody or antigen-binding fragment or bispecific antibody, or a pharmaceutical composition, for use in treating cancer. In another aspect, the present disclosure provides the use of an anti-CD137 antibody or antigen-binding fragment, bispecific antibody or its antigen-binding fragment or pharmaceutical composition in the manufacture of a medicament for treating cancer.

[0198] The cancer can include, but is not limited to, gastric cancer, colon cancer, pancreatic cancer, breast cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, small cell lung cancer, non-small cell lung cancer, ovarian cancer, skin cancer, mesothelioma, lymphoma, leukemia, myeloma, and sarcoma. Anti-GPC3xCD137 Bispecific Antibody

[0199] The antibodies or antigen-binding fragments disclosed herein can be used in a variety of applications, including but not limited to methods for treating GPC3-related disorders or diseases. In one aspect, the GPC3-related disorder or disease is cancer.

[0200] In one aspect, the present disclosure provides a method for treating cancer. In certain aspects, the method comprises administering to a patient in need thereof a therapeutically effective amount of an anti-GPC3xCD137 antibody or antigen-binding fragment, or a pharmaceutical composition thereof. In another aspect, the present disclosure provides a multispecific antibody or an antigen-binding fragment thereof, or a pharmaceutical composition, for use in treating cancer. In another aspect, the present disclosure provides the use of a multispecific antibody or an antigen-binding fragment thereof or a pharmaceutical composition in the manufacture of a medicament for treating cancer.

[0201] In one embodiment, the cancer expresses GPC3. In one embodiment, the cancer is an advanced or metastatic solid tumor.

[0202] The cancer can include but is not limited to any one or more of liver cancer, lung cancer, gastric cancer, germ cell tumor, thyroid cancer, pancreatic cancer, ovarian cancer, skin cancer, kidney cancer (such as nephroblastoma), esophageal cancer, atypical teratoid rhabdoid tumor of the brain, or undifferentiated synovial sarcoma. In one embodiment, the liver cancer is hepatoblastoma or hepatocellular carcinoma (HCC). In another embodiment, the lung cancer is non-small cell lung cancer (NSCLC) or small cell lung cancer (SCLC). In another embodiment, the non-small cell lung cancer is squamous non-small cell lung cancer. In another embodiment, the non-small cell lung cancer is GPC3+ squamous non-small cell lung cancer. In another embodiment, the gastric cancer is alpha-fetoprotein positive (AFP+) gastric cancer. In another embodiment, the kidney cancer is nephroblastoma. In another embodiment, the esophageal cancer is esophageal squamous cell carcinoma. In another embodiment, the esophageal cancer is GPC3+ esophageal squamous cell carcinoma. In another embodiment, the germ cell tumor is yolk sac tumor or non-dysgerminoma. Other

[0203] The antibodies or antigen-binding fragments disclosed herein can be administered by any suitable means, including parenterally, intratracheally, and intranasally, and if desired for local treatment, intralesional administration. Parenteral infusion includes intramuscular, intravenous, intraarterial, intraperitoneal, or subcutaneous administration. Administration can be by any suitable route, such as by injection, such as intravenous or subcutaneous injection, which depends in part on whether the administration is short-term or long-term. A variety of dosing regimens are contemplated herein, including but not limited to single administration or multiple administrations at different time points, bolus administration, and pulsed infusion.

[0204] The antibodies or antigen-binding fragments of the present disclosure can be formulated, administered, and dosed in a manner consistent with good medical practice. Factors to consider in this regard include the particular disorder being treated, the particular mammal being treated, the clinical condition of the individual patient, the cause of the disorder, the site of delivery of the agent, the method of administration, the dosing regimen, and other factors known to the medical practitioner. The antibody need not but optionally can be formulated with one or more agents currently used to prevent or treat the disorder being studied. The effective amount of such other agents will depend on the amount of antibody present in the formulation, the type of disorder or treatment, and the other factors discussed above.

[0205] For the prevention or treatment of a disease, a suitable dosage of the antibodies or antigen-binding fragments of the present disclosure will depend on the type of disease to be treated, the type of antibody, the severity and course of the disease, whether the antibody is being administered for prophylactic or therapeutic purposes, previous therapy, the clinical history of the patient and response to the antibody, and the judgment of the attending physician. VII. Combination Therapies

[0206] In one aspect, the anti-CD137 antibodies of the present disclosure or multispecific antibodies containing anti-CD137, e.g., anti-CD137xTAA antibodies or anti-GPC3xCD137 antibodies, can be used in combination with other therapeutic agents.

[0207] Other therapeutic agents include, for example, other immune checkpoint antibodies. Such immune checkpoint antibodies can include anti-PD1 antibodies. Anti-PD1 antibodies can include, but are not limited to, tislelizumab, Pembrolizumab, or Nivolumab. Tislelizumab is disclosed in US 8,735,553. Pembrolizumab (previously known as MK-3475) is disclosed in US 8,354,509 and US 8,900,587 and is a humanized IgG4-K immunoglobulin that targets the PD1 receptor and inhibits the binding of the PD1 receptor ligands PD-L1 and PD-L2. Nivolumab (as disclosed by Bristol-Meyers Squibb) is a fully human IgG4-K monoclonal antibody. Nivolumab (clone 5C4) is disclosed in US Patent No. US 8,008,449 and WO 2006 / 121168.

[0208] Other immune checkpoint antibodies that can be combined with the anti-CD137 antibodies of the present disclosure or multispecific antibodies containing anti-CD137 can include anti-TIGIT antibodies. Such anti-TIGIT antibodies can include, but are not limited to, the anti-TIGIT antibodies disclosed in WO 2019 / 129261.

[0209] In one embodiment, the present disclosure provides the use of a combination of an anti-CD137 antibody of the present disclosure or a multispecific antibody comprising anti-CD137 (e.g., an anti-CD137xTAA antibody or an anti-GPC3xCD137 antibody) and an anti-PD-1 antibody (e.g., tislelizumab or the other anti-PD-1 antibodies described above) in the manufacture of a medicament for treating cancer (e.g., the cancers described above). In another embodiment, the present disclosure provides a combination of an anti-CD137 antibody of the present disclosure or a multispecific antibody comprising anti-CD137 (e.g., an anti-CD137xTAA antibody or an anti-GPC3xCD137 antibody) and an anti-PD-1 antibody (e.g., tislelizumab or the other anti-PD-1 antibodies described above) for use in treating cancer (e.g., the cancers described above).

[0210] Combination therapy can refer to and include any one of the following: - Simultaneously administering such a combination therapy to a patient in need of treatment, wherein such components are formulated together into a single dosage form that substantially simultaneously releases such components to the patient, - Substantially simultaneously administering such a combination to a patient in need of treatment, wherein such components are formulated separately into separate dosage forms for administration by the patient at substantially the same time, whereby such components are released to the patient at substantially the same time, - Sequentially administering such a combination therapy to a patient in need of treatment, wherein such components are formulated separately into separate dosage forms for administration by the patient at successive times, with a significant time interval between each administration, whereby such components are released to the patient at substantially different times; and - Sequentially administering such a combination to a patient in need of treatment, wherein such components are formulated together into a single dosage form that releases such components in a controlled manner, whereby they are released to the patient simultaneously, continuously, and / or overlappingly at the same and / or different times, and wherein each portion can be administered by the same or different routes. Definitions

[0211] Unless specifically defined elsewhere in this document, all other technical and scientific terms used herein have the meaning commonly understood by a person of ordinary skill in the art.

[0212] As used herein, including in the appended claims, unless the context clearly dictates otherwise, the singular forms of "a", "an", and "the" include their corresponding plural referents.

[0213] Unless the context clearly dictates otherwise, the term "or" means the term "and / or" and can be used interchangeably with the term "and / or".

[0214] As used herein, the term "anticancer agent" refers to any agent that can be used to treat a proliferative disorder (such as cancer), including but not limited to cytotoxic agents, chemotherapeutic agents, radiotherapy and radiotherapeutic agents, targeted anticancer agents, and immunotherapeutic agents.

[0215] The term "CD137" or "TNFRSF9", "ILA" or "41BB" or "4-1BB" refers to a co-stimulatory molecule belonging to the TNFRSF family. The nucleic acid sequence of human CD137 is shown in SEQ ID NO:36, based on GenBank sequence accession number: NM_001561.4. The amino acid sequence of human CD137 is SEQ ID NO:35.

[0216] The term "glypican 3" (GPC3) is also known as DGSX, GTR2-2, MXR7, OCI-5, SDYS, SGB, SGBS, SGBS1. The amino acid sequence of human GPC3 (SEQ ID NO:51) can also be found in the NCBI reference sequence: NP_004475.1. The nucleic acid sequence of human GPC3 is shown in SEQ ID NO:52.

[0217] As used herein, the terms "administration / administering" and "treating / treatment", when applied to an animal, a human, an experimental subject, a cell, a tissue, an organ, or a biological fluid, mean that an exogenous drug, therapeutic agent, diagnostic agent, or composition contacts the animal, human, subject, cell, tissue, organ, or biological fluid. The treatment of cells encompasses the contact of a reagent with the cells and the contact of a reagent with a fluid, where the fluid contacts the cells. The terms "administration" and "treatment" also mean in vitro and ex vivo treatment, for example, treating cells with a reagent, a diagnostic agent, a binding compound, or by another cell. The term "subject" herein includes any organism, preferably an animal, more preferably a mammal (such as a rat, a mouse, a dog, a cat, a rabbit), and most preferably a human. In one aspect, treating any disease or disorder means improving the disease or disorder (i.e., slowing or preventing or reducing the development of the disease or at least one of its clinical symptoms). In another aspect, "treating / treatment" means alleviating or improving at least one physical parameter, including those that may not be distinguishable to the patient. In yet another aspect, "treating / treatment" means modulating a disease or disorder physically (e.g., stabilization of discernible symptoms), physiologically (e.g., stabilization of physical parameters), or both. In yet another aspect, "treating / treatment" means preventing or delaying the onset or development or progression of a disease or disorder.

[0218] In the context of this disclosure, the term "subject" is a mammal, e.g., a primate, preferably a higher primate, such as a human (e.g., a patient suffering from or at risk of developing a disorder described herein).

[0219] As used herein, the term "affinity" refers to the strength of the interaction between an antibody and an antigen. Within an antigen, the variable regions of the antibody interact with the antigen at multiple sites through non-covalent forces. Generally, the more interactions, the stronger the affinity.

[0220] As used herein, the term "antibody" refers to a polypeptide of the immunoglobulin family that can bind to a corresponding antigen non-covalently, reversibly, and in a specific manner. For example, a naturally occurring IgG antibody is a tetramer comprising at least two heavy (H) chains and two light (L) chains that are interconnected by disulfide bonds. Each heavy chain is composed of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. The heavy chain constant region is composed of three domains CH1, CH2, and CH3. Each light chain is composed of a light chain variable region (abbreviated herein as VL or Vκ) and a light chain constant region. The light chain constant region is composed of one domain CL. The VH and VL regions can be further subdivided into hypervariable regions, called complementarity determining regions (CDRs), which are interspersed with more conserved regions, called framework regions (FRs). Each VH and VL is composed of three CDRs and four framework regions (FRs) arranged in the following order from the amino terminus to the carboxyl terminus: FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4. The variable regions of the heavy and light chains contain binding domains that interact with the antigen. The constant region of the antibody can mediate the binding of the immunoglobulin to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component of the classical complement system (Clq).

[0221] The term "antibody" includes, but is not limited to, monoclonal antibodies, human antibodies, humanized antibodies, chimeric antibodies, and anti-idiotypic (anti-Id) antibodies, human engineered antibodies, single-chain antibodies (scFv), single-domain antibodies, Fab fragments, Fab' fragments, or F(ab')2 fragments. Antibodies can be of any isotype / class (e.g., IgG, IgE, IgM, IgD, IgA, and IgY) or subclass (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2). In addition, antibodies include their derivatives, e.g., by direct or indirect linkage or complex formation with another agent, such as other drugs. The term "antibody" as used herein encompasses monospecific antibodies, bispecific antibodies, and multispecific antibodies.

[0222] The term "chimeric antibody" means a molecule composed of domains from different species, i.e., fusing the variable domains of an antibody from one host species (such as mouse, rabbit, llama, etc.) with the constant domains of an antibody from a different species (such as human).

[0223] In some embodiments, the anti-GPC3 antibody comprises at least one antigen-binding site and at least one variable region. In some embodiments, the anti-GPC3 antibody comprises an antigen-binding fragment from the GPC3 antibodies described herein. In some embodiments, the anti-GPC3 antibody is isolated or recombinant.

[0224] In some embodiments, the anti-CD137 antibody comprises at least one antigen-binding site and at least one variable region. In some embodiments, the anti-CD137 antibody comprises an antigen-binding fragment from the CD137 antibodies described herein. In some embodiments, the anti-CD137 antibody is isolated or recombinant.

[0225] The term "monoclonal antibody" or "mAb" or "Mab" as used herein refers to a population of antibodies that are substantially homogeneous, i.e., the antibody molecules comprised in the population are identical in amino acid sequence except for possible naturally occurring mutations that may be present in minor amounts. In contrast, conventional (polyclonal) antibody preparations typically include a multiplicity of different antibodies that have different amino acid sequences in their variable domains, particularly their complementarity determining regions (CDRs), and which are usually specific for different epitopes. The modifier "monoclonal" indicates the character of the antibody obtained from a substantially homogeneous population of antibodies and should not be construed as requiring that the antibody be made by any particular method. Monoclonal antibodies (mAbs) can be obtained by methods known to those of skill in the art. See, e.g., Kohler et al., Nature 1975; 256:495-497; U.S. Patent No. 4,376,110; Ausubel et al., CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, 1992; Harlow et al., ANTIBODIES: A LABORATORY MANUAL, Cold spring Harbor Laboratory, 1988; and Colligan et al., CURRENT PROTOCOLS IN IMMUNOLOGY, 1993. The antibodies disclosed herein can be of any immunoglobulin class (including IgG, IgM, IgD, IgE, IgA), and any subclass thereof (e.g., IgG1, IgG2, IgG3, IgG4). Hybridomas producing monoclonal antibodies can be cultured in vitro or in vivo. High titer monoclonal antibodies can be obtained in in vivo production, where cells from a single hybridoma are injected intraperitoneally into a mouse, e.g., a naïve Balb / c mouse, to produce ascites fluid containing a high concentration of the desired antibody. Monoclonal antibodies of isotype IgM or IgG can be purified from such ascites fluid, or from culture supernatants, using column chromatography methods well known to those of skill in the art.

[0226] Typically, the basic antibody structural unit comprises a tetramer. Each tetramer includes two pairs of identical polypeptide chains, each pair having one "light chain" (about 25 kDa) and one "heavy chain" (about 50 - 70 kDa). The amino-terminal portion of each chain includes a variable region of about 100 to 110 or more amino acids that is primarily responsible for antigen recognition. The carboxyl-terminal portion of the heavy chain can be defined as a constant region that is primarily responsible for effector functions. Typically, human light chains are classified as κ and λ light chains. In addition, human heavy chains are typically classified as α, δ, ε, γ, or μ, and the isotypes of the antibodies are defined as IgA, IgD, IgE, IgG, and IgM, respectively. Within the light and heavy chains, the variable and constant regions are joined by a "J" region of about 12 or more amino acids, and the heavy chain also includes a "D" region of about 10 or more amino acids.

[0227] The variable regions of each light chain / heavy chain (VL / VH) pair form the antibody binding site. Thus, generally speaking, an intact antibody has two binding sites. Except in bifunctional or bispecific antibodies, the two binding sites are usually identical in the primary sequence.

[0228] Typically, the variable domains of the heavy and light chains contain three hypervariable regions, also called "complementary determining regions (CDRs)", which are located between relatively conserved framework regions (FRs). The CDRs are typically aligned by the framework regions such that specific epitopes can be bound. Generally, from the N-terminus to the C-terminus, both the light and heavy chain variable domains contain FR-1 (or FR1), CDR-1 (or CDR1), FR-2 (FR2), CDR-2 (CDR2), FR-3 (or FR3), CDR-3 (CDR3), and FR-4 (or FR4). The positions of the CDRs and framework regions can be determined using a variety of definitions well known in the art, e.g., Kabat, Chothia, AbM, and IMGT (see, e.g., Johnson et al., Nucleic Acids Res. [Nucleic Acids Research] 2001; 29:205-206; Chothia and Lesk, J. Mol. Biol. [Journal of Molecular Biology] 1987; 196:901-917; Chothia et al., Nature. [Nature] 1989; 342:877-883; Chothia et al., J. Mol. Biol. [Journal of Molecular Biology] 1992; 227:799-817; Al-Lazikani et al., J. Mol. Biol. [Journal of Molecular Biology] 1997; 273:927-748; ImMunoGenTics (IMGT) numbering (Lefranc, M.-P., The Immunologist. [The Immunologist] 1999; 7, 132-136; Lefranc, M.-P. et al., Dev. Comp. Immunol. [Developmental and Comparative Immunology] 27, 55-77 (2003) ("IMGT" numbering scheme)).The definition of antigen-binding sites is also described in the following references: Ruiz et al., Nucleic Acids Res. 28:219-221 (2000); and Lefranc, M.P., Nucleic Acids Res., 29:207-209 (2001); MacCallum et al. J. Mol. Biol. 262:732-745 (1996); and Martin et al., Proc. Natl. Acad. Sci. USA 86:9268-9272 (1989); Martin et al., Methods Enzymol. 203:121-153 (1991); and Rees et al., in Sternberg M.J.E. (ed.), Protein Structure Prediction, Oxford University Press, Oxford, 141-172 (1996). For example, according to Kabat, the CDR amino acid residues in the heavy-chain variable domain (VH) are numbered 31-35 (HCDR1), 50-65 (HCDR2), and 95-102 (HCDR3); and the CDR amino acid residues in the light-chain variable domain (VL) are numbered 24-34 (LCDR1), 50-56 (LCDR2), and 89-97 (LCDR3). According to Chothia, the CDR amino acid numbers in VH are 26-32 (HCDR1), 52-56 (HCDR2), and 95-102 (HCDR3); and the amino acid residues in VL are numbered 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3). By combining the CDR definitions of Kabat and Chothia, the CDR consists of amino acid residues 26-35 (HCDR1), 50-65 (HCDR2), and 95-102 (HCDR3) in human VH and amino acid residues 24-34 (LCDR1), 50-56 (LCDR2), and 89-97 (LCDR3) in human VL. According to IMGT, the CDR amino acid residues in VH are numbered approximately 26-35 (HCDR1), 51-57 (HCDR2), and 93-102 (HCDR3), and the CDR amino acid residues in VL are numbered approximately 27-32 (LCDR1), 50-52 (LCDR2), and 89-97 (LCDR3) (numbering according to Kabat). According to IMGT, the CDR regions of antibodies can be determined using the program IMGT / DomainGap Align.

[0229] The term "hypervariable region" refers to the amino acid residues in an antibody that are responsible for antigen binding. The hypervariable regions contain amino acid residues from "CDRs" (e.g., LCDR1, LCDR2, and LCDR3 in the light chain variable domain and HCDR1, HCDR2, and HCDR3 in the heavy chain variable domain). See Kabat et al., (1991) Sequences of Proteins of Immunological Interest, 5th ed. Public Health Service, National Institutes of Health, Bethesda, Maryland (defining the CDR regions of antibodies by sequence); see also Chothia and Lesk, J. Mol. Biol. 1987; 196:901-917 (defining the CDR regions of antibodies by structure). The term "framework" or "FR" residues means those variable domain residues other than the hypervariable region residues defined herein as CDR residues.

[0230] Unless otherwise specified, "antigen-binding fragment" refers to an antigen-binding fragment of an antibody, i.e., an antibody fragment that retains the ability to specifically bind an antigen with the same antigen specificity as the full-length antibody, e.g., a fragment that retains one or more CDR regions. Examples of antigen-binding fragments include, but are not limited to, Fab, Fab', F(ab')2, and Fv fragments; diabodies; linear antibodies; single-chain antibody molecules (e.g., single-chain Fv (ScFv)); nanobodies; and multispecific antibodies formed from antibody fragments.

[0231] As used herein, an antibody "specifically binds" to a target protein if the antibody exhibits preferential binding to the target protein as compared to other proteins, but this specificity does not require absolute binding specificity. Antibody "specific binding" or "selective binding" as used in the context of describing the interaction between an antigen (e.g., a protein) and an antibody or an antigen-binding antibody fragment refers to a binding reaction that determines the presence of the antigen in a heterogeneous population of proteins and other biological agents (e.g., in a biological sample, blood, serum, plasma, or tissue sample). Thus, under certain specified immunoassay conditions, an antibody or its antigen-binding fragment specifically binds to a particular antigen at least two-fold above background levels and does not specifically bind to other antigens present in the sample in significant amounts. In one aspect, under the specified immunoassay conditions, the specific binding of an antibody or its antigen-binding fragment to a particular antigen is at least ten (10)-fold above background binding levels and does not specifically bind to other antigens present in the sample in significant amounts.

[0232] As used herein, "antigen-binding domain" means the portion of an antibody that specifically binds to an antigen. In some embodiments, it comprises at least six CDRs and specifically binds to an epitope (or three CDRs, in the case of a single-domain antibody). The "antigen-binding fragment" of a multispecific antibody (e.g., a bispecific antibody) comprises a first antigen-binding domain that specifically binds to a first epitope and a second antigen-binding domain that specifically binds to a second epitope. A multispecific antibody can be a bispecific antibody, a trispecific antibody, a tetra-specific antibody, etc., having antigen-binding domains for each specific epitope. A multispecific antibody can be a multivalent antibody (e.g., a bispecific tetravalent antibody), which comprises multiple antigen-binding domains, such as 2, 3, 4, or more antigen-binding domains that specifically bind to a first epitope and 2, 3, 4, or more antigen-binding domains that specifically bind to a second epitope.

[0233] The term "human antibody" as used herein means an antibody that contains only human immunoglobulin protein sequences. If produced in a mouse, mouse cell, or hybridoma derived from a mouse cell, a human antibody can contain murine carbohydrate chains. Similarly, a "mouse antibody" or "rat antibody" means an antibody that contains only mouse or rat immunoglobulin protein sequences, respectively.

[0234] The term "humanized" or "humanized antibody" means an antibody form that contains sequences from non-human (e.g., murine, rabbit, camelid, etc.) antibodies as well as human antibodies. Such an antibody contains a minimal sequence derived from a non-human immunoglobulin. Typically, a humanized antibody will comprise substantially all of at least one, and typically two, variable domains, all or substantially all of whose hypervariable loops correspond to those of the non-human immunoglobulin, and all or substantially all of the FRs are those of human immunoglobulin sequences. A humanized antibody will also optionally comprise at least a portion of the immunoglobulin constant region (Fc), typically at least a portion of a human immunoglobulin. When it is necessary to distinguish a humanized antibody from the parental rodent antibody, the prefix "hum", "hu", "Hu", or "h" is added to the antibody clone name. The humanized form of a rodent / camelid antibody typically contains the same CDR sequences as the parental rodent antibody, but may include certain amino acid substitutions to increase affinity, increase the stability of the humanized antibody, remove post-translational modifications, or for other reasons.

[0235] The term "corresponding germline sequence" refers to a nucleic acid sequence encoding an amino acid sequence or subsequence of a human variable region that has the highest determined amino acid sequence identity to a reference variable region amino acid sequence or subsequence compared to all other known variable region amino acid sequences encoded by the human germline immunoglobulin variable region sequences. The corresponding germline sequence can also refer to a human variable region amino acid sequence or subsequence that has the highest amino acid sequence identity to a reference variable region amino acid sequence or subsequence compared to all other evaluated variable region amino acid sequences. The corresponding germline sequence can be only a framework region, only a complementarity determining region, a framework and a complementarity determining region, a variable segment (as defined above), or other combinations of sequences or subsequences containing the variable region. Sequence identity can be determined using the methods described herein, such as by aligning two sequences using BLAST, ALIGN, or another alignment algorithm known in the art. The corresponding germline nucleic acid or amino acid sequence can have at least about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% sequence identity to the reference variable region nucleic acid or amino acid sequence. Additionally, if the antibody contains a constant region, the constant region is also derived from such a human sequence, such as a germline sequence, or a mutated form of a germline sequence or an antibody containing a consensus framework sequence derived from the analysis of human framework sequences, as described, for example, in Knappik et al., J. Mol. Biol. [Journal of Molecular Biology] 296:57 - 86, 2000.

[0236] The "2 + 2 format" means that in a bispecific antibody that targets two different antigens or two different epitopes, such an antibody contains two first antigen - binding domains that specifically bind to a first antigen or a first epitope, and two second antibody - binding domains that specifically bind to a second antigen or a second epitope.

[0237] The term "equilibrium dissociation constant (K D , M)" refers to the dissociation rate constant (kd, time -1 ) divided by the association rate constant (ka, time -1 , M -l ). The equilibrium dissociation constant can be measured using any known method in the art. The antibodies of the present disclosure will generally have an equilibrium dissociation constant of less than about 10 -7 or 10 -8 M, such as less than about 10 -9 M or 10 -10 M, and in some aspects, less than about 10 -11 M, 10 -12 M or 10 -13 M.

[0238] The term "cancer" or "tumor" in this text has the broadest meaning as understood in the art and refers to a physiological disorder in mammals typically characterized by unregulated cell growth. In the context of this disclosure, cancer is not limited to a particular type or location.

[0239] In the context of this disclosure, when referring to an amino acid sequence, the term "conservative substitution" means replacing an original amino acid with a new amino acid that substantially does not change the chemical, physical, and / or functional properties of the antibody or fragment, such as its binding affinity to GPC3 or CD137. In particular, common conservative substitutions of amino acids are well known in the art.

[0240] As used herein, the term "knob-into-hole" technology refers to the introduction of amino acids in vitro or in vivo by introducing a spatial protrusion (knob) into one polypeptide and a pocket or cavity (hole) into another polypeptide at the interface where they interact, thereby guiding the pairing of the two polypeptides together. For example, knob-into-hole has been introduced into the Fc:Fc binding interface, C L :C H I interface, or V H / V L interface (see, for example, US2011 / 0287009, US2007 / 0178552, WO 96 / 027011, WO98 / 050431, and Zhu et al., Protein Science. 1997; 6:781-788). In some embodiments, knob-into-hole ensures the correct pairing of two different heavy chains during the production of a multispecific antibody. For example, a multispecific antibody having knob-into-hole amino acids in its Fc region may also contain a single variable domain linked to each Fc region, or further contain different heavy chain variable domains paired with similar or different light chain variable domains. Knob-into-hole technology can also be used in the VH or VL regions to ensure correct pairing.

[0241] As used herein in the context of "knob-into-hole" technology, the term "knob" refers to an amino acid change that introduces a protrusion (knob) into a polypeptide at the interface where the polypeptide interacts with another polypeptide. In some embodiments, the other polypeptide has a hole mutation.

[0242] As used herein in the context of "knob-into-hole", the term "hole" refers to an amino acid change that introduces a pocket or cavity (hole) into a polypeptide at the interface where the polypeptide interacts with another polypeptide. In some embodiments, the other polypeptide has a knob mutation.

[0243] Examples of algorithms suitable for determining percent sequence identity and sequence similarity are the BLAST algorithms, which are described in Altschul et al., Nuc. Acids Res. [Nucleic Acids Research] 25:3389-3402, 1977; and Altschul et al., J. Mol. Biol. [Journal of Molecular Biology] 1990; 215:403-410. Software for performing BLAST analyses is publicly available through the National Center for Biotechnology Information. This algorithm involves first identifying high-scoring sequence pairs (HSPs) by identifying short word lengths W in the query sequence that, when aligned with the same word length in a database sequence, match or satisfy some positive-valued threshold score T. T is referred to as the neighborhood word score threshold. These initial neighborhood word hits serve as seeds for a search to find longer HSPs that contain them. The word hits are extended in both directions along each sequence until the cumulative alignment score can no longer increase. For nucleotide sequences, parameters M (reward score for a pair of matching residues; always >0) and N (penalty score for a mismatch residue; always <0) are used to calculate the cumulative score. For amino acid sequences, a scoring matrix is used to calculate the cumulative score. Extension of the word hits in each direction is stopped when: the cumulative alignment score drops from its maximum achieved value by an amount X; the cumulative score tends to zero or lower due to the accumulation of one or more negative-scoring residue alignments; or the end of either sequence is reached. The BLAST algorithm parameters W, T, and X determine the sensitivity and speed of the alignment. The BLASTN program (for nucleotide sequences) defaults to a word length (W) of 11, an expectation value (E) of 10, M = 5, N = -4, and the comparison of both strands. For amino acid sequences, the BLAST program defaults to a word length of 3, an expectation value (E) of 10, and the BLOSUM62 scoring matrix (see, Henikoff and Henikoff, Proc. Natl. Acad. Sci. USA [Proceedings of the National Academy of Sciences of the United States of America] 1989; 89:10915) alignment (B) 50, expectation value (E) 10, M = 5, N = -4, and the comparison of both strands.

[0244] The BLAST algorithm also performs a statistical analysis of the similarity between two sequences (see, e.g., Karlin and Altschul, Proc. Natl. Acad. Sci. USA 90:5873-5787, 1993). One measure of similarity provided by the BLAST algorithm is the minimum sum probability (P(N)), which gives an indication of the probability that a match occurs by chance between two nucleotide or amino acid sequences. For example, if the minimum sum probability in a comparison of a test nucleic acid with a reference nucleic acid is less than about 0.2, more preferably less than about 0.01, and most preferably less than about 0.001, then the nucleic acid is considered similar to the reference sequence.

[0245] The percent identity between two amino acid sequences can also be determined using the algorithm of E. Meyers and W. Miller, Comput. Appl. Biosci., 1988; 4:11-17, incorporated into the ALIGN algorithm (version 2.0), using a PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4. In addition, the percent identity between two amino acid sequences can be determined using the algorithm of Needleman and Wunsch, J. Mol. Biol. 48:444-453 (1970), which has been incorporated into the GAP program in the GCG software package, using a BLOSUM62 matrix or a PAM250 matrix, a gap weight of 16, 14, 12, 10, 8, 6, or 4, and a length weight of 1, 2, 3, 4, 5, or 6.

[0246] The term "nucleic acid" is used interchangeably herein with the term "polynucleotide" and refers to deoxyribonucleotides or ribonucleotides and polymers thereof in single-stranded or double-stranded form. The term encompasses nucleic acids containing known nucleotide analogs or modified backbone residues or linkages, which are synthetic, naturally occurring, and non-naturally occurring, having similar binding properties as the reference nucleic acid, and which are metabolized in a manner similar to the reference nucleotides. Examples of such analogs include, but are not limited to, phosphorothioates, phosphoramidates, methylphosphonates, chiral methylphosphonates, 2-O-methyl ribonucleotides, and peptide-nucleic acids (PNAs).

[0247] In the context of nucleic acids, the term "operably linked" refers to the functional relationship between two or more polynucleotide (e.g., DNA) segments. Typically, it refers to the functional relationship between a transcriptional regulatory sequence and a transcriptional sequence. For example, a promoter or enhancer sequence is operably linked to a coding sequence if it stimulates or regulates the transcription of the coding sequence in a suitable host cell or other expression system. Generally, a promoter transcriptional regulatory sequence operably linked to a transcriptional sequence is physically contiguous with the transcriptional sequence, i.e., they are cis-acting. However, some transcriptional regulatory sequences (such as enhancers) do not need to be physically contiguous or adjacent to the coding sequence whose transcription they enhance.

[0248] As used herein, the term "pharmaceutically acceptable excipient" includes any and all solvents, dispersion media, isotonic agents, absorption delaying agents, and the like that are physiologically compatible. The excipients may be suitable for intravenous, intramuscular, subcutaneous, parenteral, rectal, spinal, or epidermal administration (e.g., by injection or infusion).

[0249] The term "therapeutically effective amount" as used herein refers to the amount of an antibody sufficient to affect the treatment of a disease, or at least one clinical symptom of a disease or disorder, when administered to a subject to treat the disease, disorder, or symptom. The "therapeutically effective amount" can vary with the antibody, the disease, the disorder, and / or the symptom of the disease or disorder, the severity of the disease, disorder, and / or symptom of the disease or disorder, the age of the subject to be treated, and / or the weight of the subject to be treated. The appropriate amount in any given case will be apparent to those skilled in the art or can be determined by routine experimentation. In the case of combination therapy, the "therapeutically effective amount" refers to the total amount of the combination of agents used to effectively treat the disease, disorder, or condition.

[0250] The term "combination therapy" refers to the administration of two or more therapeutic agents to treat a therapeutic condition or disorder described in the present disclosure. Such administration encompasses co-administering these therapeutic agents in a substantially simultaneous manner. Such administration also encompasses co-administering in multiple containers or in separate containers for each active ingredient (e.g., capsules, powders, and liquids). The powder and / or liquid can be reconstituted or diluted to the desired dose before administration. In addition, such administration also encompasses using each type of therapeutic agent in a sequential manner at approximately the same time or at different times. In either case, the treatment regimen will provide a beneficial effect of the drug combination in treating the condition or disorder described herein.

[0251] As used herein, the phrase "in combination with" means administering the antibody herein to a subject either simultaneously with, before, or after the administration of an additional therapeutic agent. In certain embodiments, the antibody herein is administered as a co-formulation with an additional therapeutic agent. equivalent

[0252] It should be understood that although the present invention has been described in conjunction with the detailed description, the foregoing description is intended to illustrate and not limit the scope of the present invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

[0253] It should be understood that one, some, any, or all of the features of the various embodiments described herein can be combined to form additional embodiments of the present disclosure. These and other aspects of the present disclosure will become apparent to those skilled in the art. Examples Example 1. Generation of a single-chain anti-huCD137 VHH antibody CD137 recombinant protein for phage activity and binding assays

[0254] To discover VHH antibodies against human CD137, several recombinant proteins were designed and expressed for phage panning and screening. The cDNA coding region of full-length human CD137 (huCD137) was aligned based on the CD137 GenBank sequence (accession number: NM_001561.4, this gene was obtained from Sinobio, catalog number: HG10041-M, herein incorporated by reference as SEQ ID NO:36). The human CD137 ligand (TNFSF9) was aligned based on the CD137 ligand GenBank sequence (accession number: NM_003811.3, this gene was obtained from Sinobio, catalog number: HG15693-G). Briefly, the coding regions of the extracellular domain (ECD) consisting of amino acids (AA) 24 - 183 of huCD137 (SEQ ID NO:35) and the ECD consisting of AA 71 - 254 of huCD137 ligand (SEQ ID NO:37) were PCR amplified respectively. The coding region of mIgG2a Fc (SEQ ID NO:39) was PCR amplified and then conjugated with the ECD of human CD137 or the ECD of human CD137 ligand by overlap PCR to prepare mIgG2a Fc fusion proteins. Then the PCR products were cloned into a pcDNA3.1-based expression vector (Invitrogen, Carlsbad, California, USA), thus generating two recombinant mIgG2a Fc fusion protein expression plasmids, human CD137 ECD-mIgG2a, human CD137 ligand ECD-mIgG2a. Alternatively, the coding region of the ECD consisting of AA 24 - 183 of huCD137 (SEQ ID NO:35) was also cloned into a pcDNA3.1-based expression vector (Invitrogen, Carlsbad, California, USA), and its C-terminus was fused with a 6xHis tag, thus generating human CD137-ECD-his. To produce the recombinant fusion proteins, the plasmids were transiently transfected into a HEK293-based mammalian cell expression system (internally developed) and cultured in a CO2 incubator equipped with a rotary shaker for 5 - 7 days. The supernatant containing the recombinant proteins was collected and centrifuged to clarify. The recombinant proteins were purified by a protein A column (catalog number: 17127901, General Life Sciences) or Ni-NTA agarose (catalog number: R90115, Invitrogen). All the recombinant proteins were dialyzed with phosphate buffered saline (PBS) and stored in small aliquots in an -80 °C freezer. Llama Immunization and Phage Library Construction

[0255] One llama was immunized with human CD137 ECD-mIgG2a. Two weeks after the fourth immunization, llama PBMCs were collected using standard techniques for RNA extraction (Chomczynski, et al., Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Analytical Biochem. 1987; 162(1):156-159).

[0256] A phage library was constructed by reverse transcription and spliced overlap extension PCR. The PCR products were double digested with NcoI / NotI and ligated into the phagemid vector pCANTAB-5E. The library was then transformed into Escherichia coli TG1 bacteria and verified by random clone DNA Sanger sequencing (analysis of >96 clones). After the rescue step using KM13 helper phage, the phages were purified by precipitation twice directly from the culture supernatant with PEG / NaCl. After transformation into Escherichia coli bacteria, a library of size >10 7 was obtained. Phage display panning and screening

[0257] Phage display selections were performed by phage display using standard protocols (Silacci et al., (2005) Proteomics, 5, 2340-50; Zhao et al., (2014) PLoS One, 9, e111339). Briefly, 10 μg / ml immobilized human CD137 ECD-mIgG2a (catalog number 470319, Thermo Fisher Scientific) in immunotubes was used in rounds 1 and 2. Hut78 / huCD137 cells were used for selection in round 3. Immunotubes were blocked with PBS solution containing 5% milk powder (w / v) supplemented with 1% Tween 20 (MPBST) for 1 hour. After washing with PBST (PBS buffer supplemented with 0.05% Tween 20), 5 × 10 12 (round 1) or 5 × 10 11Phages (in the second round) were depleted with human CD40 ECD-mIgG2a in MPBST for 1 hour and then incubated with the antigen for 1 hour. For the third round of selection, cell panning was performed using Hut78 / huCD137 cells and HEK293 (ATCC, CRL-1573) cells as depletion cells. After washing with PBST, the bound phages were eluted with 100 mM triethylamine (Sigma-Aldrich). The eluted phages were used to infect mid-log Escherichia coli TG1 bacteria and plated on TYE agar plates supplemented with 2% glucose and 100 μg / ml ampicillin. After three rounds of selection, single colonies were picked and supernatants containing phages were prepared using standard protocols. Phage ELISA was used to screen for anti-huCD137 VHH antibodies.

[0258] For phage ELISA, Maxisorp immunoplates were coated with the antigen and blocked with PBS buffer containing 5% milk powder (w / v). Phage supernatants were blocked with MPBST for 30 minutes and then added to the wells of the ELISA plate for 1 hour. After washing with PBST, bound phages were detected using HRP-conjugated anti-M13 antibody (GE Healthcare) and 3,3',5,5'-tetramethylbenzidine substrate (Catalog No.: 00-4201-56, eBioscience, USA). Cells expressing CD137 (10 5 cells / well) were incubated with ELISA-positive phage supernatants and then bound with AlexaFluro-647-labeled anti-M13 antibody (GE Healthcare). Cell fluorescence was quantified using a flow cytometer (Guava easyCyte TM 8HT, Merck-Millipore, USA). Expression and purification of Fc-fused VHH antibodies

[0259] Then, using an in-house developed expression vector, the anti-huCD137 VHH antibody was constructed into the form of a human Fc-fused VHH antibody (VHH-Fc). The VHH domain antibody was fused to the N-terminus of human Fc with an intervening G4S linker (SEQ ID NO: 62). The Fc null form of human IgG1 (inert Fc without binding to FcγR, SEQ ID NO: 19) was used. Expression and preparation of the Fc-fused VHH antibody were achieved by transfection into 293G cells and purification using a protein A column (Catalog No. 17543802, GE Life Sciences). The purified antibody was concentrated to 0.5 - 5 mg / mL in PBS and stored in aliquots at -80 °C in a refrigerator. Example 2. Characterization of purified anti - huCD137 VHH antibody

[0260] For antigen ELISA, Maxisorp immunoplates were coated with the antigen and blocked with 3% BSA (w / v) in PBS buffer (blocking buffer). Monoclonal VHH antibodies were blocked with the blocking buffer for 30 minutes and then added to the wells of the ELISA plate for 1 hour. After washing with PBST, bound antibodies were detected using HRP - conjugated anti - human IgG antibody (Sigma, A0170) and 3,3',5,5' - tetramethylbenzidine substrate (Catalog No.: 00 - 4201 - 56, Invitrogen, USA). Ligand competition was also applied to the ELISA assay. The ELISA analysis and ligand competition results of a representative clone BGA - 9612 are shown in Figure 1 which. The results showed that BGA - 9612 (SEQ ID NO: 1 - 5) could bind to human CD137 with good affinity but not to mouse CD137. The binding of human CD137 to BGA - 9612 could be reduced by competition with the huCD137 ligand. Example 3. Humanization of anti - human CD137 VHH BGA - 9612

[0261] For the humanization of BGA - 9612, sequences highly homologous to the cDNA sequence of the variable region of BGA - 9612 were searched in the germline human immunoglobulin gene databases in the IMGT and NCBI websites. Human IGVH genes that are present at high frequency in the human antibody repertoire (Glanville 2009 PNAS [Proceedings of the National Academy of Sciences of the United States of America] 106:20216 - 20221) and highly homologous to BGA - 9612 were selected as the humanization templates.

[0262] Humanization was performed by CDR grafting (Methods in Molecular Biology, Vol 248: Antibody Engineering, Methods and Protocols [Molecular Biology Methods, Volume 248: Antibody Engineering, Methods and Protocols], Humana Press), and the humanized VHH from BGA - 9612 was engineered into the Fc - VHH form using an in - house developed expression vector. Using an in - house developed expression vector containing an Fc variant of human IgG1 (SEQ ID NO: 19), the humanized VHH from BGA - 9612 was fused to the C - terminus of Fc in the Fc - VHH form with a G4S linker (SEQ ID NO: 62), and these vectors have easily adaptable sub - cloning sites. By transfecting the construct into ExpiCHO TMExpression and preparation of the humanized VHH from BGA-9612 were achieved in cells and purified using a Protein A column. The purified antibody was concentrated to 0.5 - 5 mg / mL in PBS and stored in aliquots at -80 °C in a refrigerator. Framework exchange

[0263] In the first round of humanization, mutations from camelid to human amino acid residues in the framework region were guided by the modeled 3D structure, and camelid framework residues that were structurally important for maintaining the canonical structure of the CDR were retained in the first round of humanization, including amino acid residues R27, F37, E44, R45, Q46, Y47, G49, V78, I94, and Q103 (Kabat numbering). Specifically, the CDRs (SEQ ID NO: 1 - 3) of BGA-9612 VHH were transplanted into the framework of the human germline variable gene IGVH3-23, where several camelid framework residues were retained, and BGA-6582 (SEQ ID NO: 8 - 9) was generated. The binding affinities of BGA-9612 and BGA-6582 by SPR are shown in Table 2.

[0264] In the examples, Kabat numbering and Kabat definitions were used for the sequences of the CDR and VH / VL. The Fc sequence was numbered using EU numbering. Table 2. Binding affinities of BGA-9612 and BGA-6582 by SPR VHH <![CDATA[k on (M -1 s -1 )]]> <![CDATA[k off (s -1 )]]> <![CDATA[K D (M)]]> BGA-9612 2.07E+04 4.19E-04 2.03E-08 BGA-6582 2.12E+04 5.88E-04 2.77E-08

[0265] Based on BGA-6582, we performed several single mutations (converting the camelid residues retained in the framework region to the corresponding human germline residues), as well as combinations of single mutations, as shown in Table 3. All humanization mutations were performed using primers containing mutations at specific positions and a site-directed mutagenesis kit (Catalog No. FM111-02, TransGen Biotech Co., Ltd., Beijing, China). The desired mutations were verified by sequencing analysis. These further humanized VHHs from BGA-6582 were tested in SPR binding assays, as shown in Table 3. BGA-3726 (SEQ ID NO: 1 - 3 and 6 - 7) including amino acid Q46E based on BGA-6582 was selected for further engineering. Amino acid residues F37, Y47, G49, and I94 in the framework region are crucial for binding to CD137 (Kabat numbering). Table 3. Binding affinities of variants with single amino acid mutations based on BGA-6582 Improved biophysical properties

[0266] The humanized VHH BGA-3726 did not show overall biophysical properties (e.g., Tm or Tagg) superior to those of the camelid VHH BGA-9612 (data not shown). Therefore, BGA-3726 was further engineered by introducing mutations in the CDR and framework regions to improve the biophysical properties for human therapeutic use.

[0267] The amino acid N73 in framework region 3 (FR3) of BGA-3726 was identified as a deamination hot spot. To mitigate the risk of post-translational modification (PTM), the subsequent amino acid S74 was mutated to alanine (reverted to the camelid amino acid residue).

[0268] In summary, based on BGA-3726, the following engineered versions of humanized VHHs were derived from the mutation process described above: (1) BGA-3544, containing the amino acids N64K and N65G in HCDR2 and L5V, S82bN, A84P, and L108Q in the framework (Kabat numbering) (SEQ ID NO:1, 10, 3, and 11-12), (2) BGA-7031, containing L5V, S82bN, A84P, and L108Q in the framework (Kabat numbering) (SEQ ID NO:1-3 and 15-16), (3) BGA-9502, containing the amino acids N64K and N65G in HCDR2 and L5V, S74A, S82bN, A84P, and L108Q in the framework (Kabat numbering) (SEQ ID NO:13-14), (4) BGA-2524, containing L5V, S74A, S82bN, A84P, and L108Q in the framework (Kabat numbering) (SEQ ID NO:17-18).

[0269] For affinity determination, the antibodies were surface captured with anti-human Fc and used for affinity determination based on surface plasmon resonance (SPR) technology. The results of the binding characteristics of the SPR assay for anti-CD137 antibodies are summarized in Table 4. BGA-3544 and BGA-7031 had similar binding affinities, with dissociation constants of 17.6 nM and 11.7 nM, respectively, which were comparable to the dissociation constant of BGA-9612 (15.2 nM). BGA-6582 also had a binding affinity comparable to that of BGA-9612. Table 4. Comparison of the binding affinities of BGA-9612 and humanized VHHs by SPR

[0270] The biophysical properties of chimeric and humanized Fc-VHHs were tested. The biophysical properties tested included melting temperature by DSC, aggregation temperature by SLS266, hydrophobicity by HIC-HPLC, and self-association tendency by AC-SINS (described in detail below). BGA-9612 showed the best thermal stability by Tm and Tagg, and good colloidal stability by AC-SINS. As shown in Table 5, the humanized VHHs BGA-3544 and BGA-7031 showed overall biophysical properties comparable to BGA-9612 (chimeric). In addition, BGA-3544 and BGA-7031 showed improved Tm compared to BGA-9612 (chimeric). Table 5. Summary of the Biophysical Properties of Anti-CD137 VHHs

[0271] The melting temperature (Tm) was determined using a high-throughput MicroCal TM VP-Capillary DSC (Malvern Instruments, Northampton, MA). Thermal profiles of each protein (350 μL, 0.5 mg / mL) from 20 °C to 100 °C were obtained using a scan rate of 60 °C / hr. The thermal profile of the individual buffer was subtracted from each protein sample. The results obtained showed the values of the transition temperature midpoint (Tm) and calorimetric enthalpy (ΔH) of the sample.

[0272] The aggregation temperature Tagg (°C) represents the colloidal stability of the sample and was obtained by monitoring the onset of aggregation using UNCLE TM (Unchained lab, Pleasanton, CA) by SLS266. The sample was loaded into the Uni and the temperature was raised from 15 °C to 95 °C. The backward-reflected optics cannot detect the near-UV light scattering of protein aggregates, so only non-scattered light reaches the detector. Therefore, the decrease in backward-reflected light is a direct measure of aggregation in the sample.

[0273] To determine the hydrophobicity of a given VHH using HPLCe2695 (Waters Corporation, Milford, MA), the protein sample was diluted in mobile phase A (50 mM sodium phosphate, 1.5 M ammonium sulfate, pH 7.0), followed by loading onto a MAbPac equilibrated in mobile phase A TMPre-filtered on a HIC-10 column (ThermoFisher Scientific, Waltham, MA). Samples were eluted using a reverse gradient from mobile phase A to mobile phase B (50 mM sodium phosphate, pH 7.0). After elution, the A280 nm fractions were recorded over time, and then the data was exported and analyzed using Empower TM software. The retention time of each sample was compared to a reference and was a property of VHH hydrophobicity, where longer elution times were associated with higher hydrophobicity.

[0274] The AC-SINS assay measures protein self-interaction by capturing VHH on the surface of gold colloids that exhibit surface resonance oscillations at visible light frequencies. When immobilized antibodies self-interact, the colloids aggregate, changing the oscillation frequency to absorb longer wavelengths. Gold nanoparticles were incubated with an 80 / 20 (v / v) mixture of capture antibody / non-capture antibody. The coated gold nanoparticles were then concentrated 10x into PBS. Samples were pre-diluted in PBS at a concentration of 50 μg / ml. 10 μL of 10X concentrated AuNP was incubated with 100 μL of sample in the dark at room temperature for 2 hours in a 384-well plate. Then the absorbance spectra of each well were read from 510 to 570 nm using a BMG ClarioStar TM (BMG Labtech, Offenburg, Germany). A red shift in the maximum absorption peak and its intensity indicated the self-interaction propensity of the tested VHH sample. Example 4. Binding activity of humanized VHH to native CD137

[0275] To evaluate the binding activity of anti-CD137 antibodies to native CD137 on live cells, HuT78 cells were engineered to overexpress human CD137. Live HuT78 / CD137 cells were seeded in 96-well plates and incubated with a series of diluted anti-CD137 antibodies. Goat anti-human IgG was used as a secondary antibody to detect antibody binding to the cell surface. The EC 50 value for dose-dependent binding to human native CD137 was determined by fitting the dose-response data to a four-parameter logistic model using GraphPad Prism TM . The data is shown in Figure 2 and Table 6. The humanized VHH retained sub-nanomolar binding affinity for native CD137. Table 6. Binding of chimeric and humanized VHH to HuT78 / CD137 VHH HuT78 / CD137 Binding EC50 (nM) BGA-9612 0.85 BGA-3544 0.85 BGA-7031 0.88 Example 5. Anti-GPC3xCD137 multispecific antibody

[0276] Agonistic anti - huCD137 antibodies have shown toxicity in the clinical setting, which may indicate that systemic FcγR cross - linking is not optimal for CD137 activation. The aim was to specifically achieve effective CD137 stimulation at the tumor site without systemic CD137 activation in a wide range of cancers. To overcome the dependence on FcγR cross - linking, we generated anti - GPC3xCD137 multispecific antibodies with the following characteristics, as Figure 3 shown in. This specific construct includes an IgG fusion - like multispecific antibody form with a module ratio of 2:2, a bivalent F(ab')2 fragment that binds human GPC3, a VH domain fragment fused to the C - terminus of CH3 (which binds huCD137), and an Fc - null form of huIgG1 (IgG1mf Fc, SEQ ID NO:53), which does not have FcγR binding but retains FcRn binding. The amino acid and DNA sequences of the GPC3 antibody are shown in Table 7. Table 7. Amino acid and DNA sequences of the GPC3 antibody

[0277] To further increase the in - vivo half - life of the anti - GPC3xCD137 multispecific antibody, the YTE mutation (M252Y / S254T / T256E, EU numbering) located in CH2 of the IgG Fc region was introduced into IgG1mf Fc to generate IgG1mf Fc - YTE (SEQ ID NO:20).

[0278] The following anti - GPC3xCD137 multispecific antibodies were generated.

[0279] BE - 933: The GPC3 antibody (SEQ ID No:41 - 50) was combined with the CD137 VHH BGA - 9612 (SEQ ID NO:1 - 5) to generate the construct BE - 933 (VL is SEQ ID NO:23 - 24, VH is SEQ ID NO:27 - 28).

[0280] BE - 774: The YTE mutation in Fc was introduced into BE - 933 to generate BE - 774 (VL is SEQ ID NO:23 - 24, VH is SEQ ID NO:25 - 26).

[0281] BE-653: Combine the GPC3 antibody (SEQ ID No: 41 - 50) with the CD137 VHH BGA-7031 (SEQ ID No: 1 - 3 and 15 - 16) to generate the construct BE-653 (VL is SEQ ID NO: 23 - 24, VH is SEQ ID NO: 29 - 30).

[0282] BE-915: Combine the GPC3 antibody (SEQ ID No: 41 - 50) with the CD137 VHH BGA-2524 (SEQ ID No: 1 - 3 and 17 - 18) and the YTE mutation in Fc to generate the construct BE-915 (VL is SEQ ID NO: 23 - 24, VH is SEQ IDNO: 21 - 22).

[0283] BE-647: Combine the GPC3 antibody (SEQ ID No: 41 - 50) with the CD137 VHH BGA-9502 (SEQ ID No: 1, 10, 3, 13 - 14) to generate the construct BE-647 (VL is SEQ ID NO: 23 - 24, VH is SEQ ID NO: 31 - 32).

[0284] BE-621: Combine the GPC3 antibody (SEQ ID No: 41 - 50) with the CD137 VHH BGA-9502 (SEQ ID No: 1, 10, 3, 13 - 14) and the YTE mutation in Fc to generate the construct BE-621 (VL is SEQ ID NO: 23 - 24, VH is SEQID NO: 33 - 34). Example 6. Target Binding Activity of Anti-GPC3xCD137 Bispecific Antibodies

[0285] Using BIAcore TM T-200 (General Life Sciences) was used to characterize the binding kinetics of BE-774 by SPR assay. Briefly, the anti-κ antibody was immobilized on an activated CM5 biosensor chip (Catalog No.: BR100839, General Life Sciences). BE-774 was passed over the chip surface and captured by the anti-κ antibody. Then, serial dilutions (6.0 nM to 2150 nM) of human CD137 ECD-mIgG2a or huGPC3-His were passed over the chip surface, and the changes in surface plasmon resonance signals were analyzed using a one-to-one Langmuir binding model (BIA evaluation software, General Life Sciences) to calculate the association rate (k on ) and dissociation rate (k off ). The equilibrium dissociation constant (K D ) was calculated as the ratio k off / k on The results showed that BE-774 exhibited binding to huCD137 and huGPC3 as shown in Table 8 below. To evaluate the binding activity of the BE-774 bispecific antibody to native huCD137 on live cells, Hut78 cells were transfected to overexpress human CD137. Live cells expressing Hut78 / huCD137 were seeded in 96-well plates and incubated with serial dilutions of BE-774. Goat anti-human IgG was used as the secondary antibody to detect the binding of the antibody to the cell surface. The EC 50 value for dose-dependent binding to human native CD137 was determined by fitting the dose-response data to a four-parameter logistic model using GraphPad Prism TM . As shown in Figure 4A , BE-774 exhibited specific binding to native CD137 on live cells in a dose-responsive manner with an EC50 of 0.7 nM. To evaluate the binding to huGPC3 on live cells, HepG2 cells expressing huGPC3 were seeded in 96-well plates and incubated with serial dilutions of BE-774. Goat anti-human IgG was used as the secondary antibody to detect the binding of the antibody to the cell surface. The EC 50 value for dose-dependent binding to human native GPC3 was determined by fitting the dose-response data to a four-parameter logistic model using GraphPad Prism TM . As shown in Figure 4B , BE-774 exhibited specific binding to native GPC3 on live cells in a dose-responsive manner with an EC50 of 0.9 nM. Table 8. SPR Affinity of BE-774 for huGPC3 and huCD137 Example 7. Anti-GPC3xCD137 Induces T Cell Activation in Coculture with GPC3-Positive Tumor Cells

[0286] The functional activity of the anti-GPC3xCD137 bispecific antibody was evaluated in an in vitro coculture experiment using human peripheral blood mononuclear cells (PBMC) and a hepatocellular carcinoma (HCC) cell line expressing OS8. OS8 is a single-chain variable fragment (scFv) of the anti-human CD3 antibody OKT3 fused to the C-terminal domain (113 - 220 aa) of murine CD8α, which includes a hinge domain, a transmembrane domain, and a cytoplasmic domain. When expressed on target cells, OS8 can provide signal 1 for T cell activation (Figure 5A). GPC3-high-expressing HepG2 cells were selected to evaluate the functional activity of the GPC3xCD137 bispecific antibody, while SK-HEP-1 (GPC3-negative) was used as a negative control cell line.

[0287] Frozen human PBMC (Ausbian) was thawed in RPMI 1640 medium and incubated overnight at 37 °C. Target cells expressing OS8 were seeded into 384-well plates and allowed to attach for 16 hours. The next day, PBMC were added to the 384-well plates at an effector-to-target cell ratio (E:T) of 2:1. The co-cultured cells were then treated with serial dilutions of BE-774 or BE-653 at 37 °C for 48 hours. Culture supernatants were collected for subsequent measurement of IFN-γ and IL-2 concentrations by a TR-FRET-based method (Degorce, et al. Current chemical genomics. 2009, 3:22) as described in the manufacturer's manual (Cisbio). Results showed that anti-GPC3xCD137 bispecific antibodies (including BE-774 and BE-653) induced dose-dependent cytokine release in PBMC co-cultured with HepG2 cells but not GPC3-negative cells (Figure 5B). PBMC from two donors were tested and the results are shown in Figure 5B. Example 8. Anti-GPC3xCD137 enhances T cell killing activity against GPC3-positive tumor cells

[0288] The xCELLigence TM RTCA MP instrument (Agilent Technologies) was used in co-culture experiments to evaluate the ability of anti-GPC3xCD137 bispecific antibodies to induce T cell killing activity by impedance measurement. Frozen human PBMC (Ausbian) was thawed in RPMI 1640 medium and incubated overnight at 37 °C. Target cells were seeded into 96-well E-plates (Agilent Technologies) and allowed to attach for 16 hours. The next day, PBMC were added to the 96-well E-plates at an effector-to-target cell ratio (E:T) of 5:1. The co-cultured cells were then treated with a combination of serial dilutions of BE-774 or BE-653 and an EpCAM / CD3 bispecific T cell engager (BiTE) that provides signal 1 for T cell activation (Figure 6A). The experiment was allowed to proceed for 4 days and electrode impedance was measured by viable adherent target cells. Results showed that both BE-774 and BE-653 dose-dependently enhanced T cell killing activity against HepG2 cells expressing GPC3 but not against GPC3-negative SK-OV-3 cells (Figure 6B). Example 9. Pharmacokinetics of BE-774 and BE-933 in cynomolgus monkeys

[0289] Blood samples were collected from cynomolgus monkeys at 0, 0.00694 (10 minutes), 0.0417 (1 hour), 0.167 (4 hours), 0.333 (8 hours), 1, 4, 7, 10, 14, 21, and 28 days after intravenous administration of 5 mg / kg of BE-774 or BE-933, and then centrifuged (4 °C, 3500 × g, 2 min) to separate the serum. The concentrations of BE-774 or BE-933 were measured by an in-house developed ELISA. Briefly, a labeled GPC3 protein (Catalog No.: C414, Novoprotein Co., Ltd., China) was used as the capture reagent, and a biotinylated CD137 antigen (Catalog No.: 41B-H82E6, ACRO Co., China) was used as the detection reagent for the measurement of BE-774 or BE-933. The pharmacokinetic (PK) characteristics of BE-774 and BE-933 at the 5 mg / kg dose level are shown in Figure 7 . The PK parameters of BE-774 and BE-933 at the 5 mg / kg dose level are shown in Table 9. Compared with BE-933, BE-774 with the YTE mutation in Fc showed significantly improved pharmacokinetic characteristics. Table 9. PK parameters of BE-774 and BE-933 in cynomolgus monkeys Example 10. Pharmacokinetics of BE-774 and BE-933 in the hFcRn mouse model

[0290] Blood samples were collected from hFcRn mice at 0, 0.0833 (2 hours), 1, 3, 7, 10, 14, 21, and 28 days after intravenous administration of 3 mg / kg of BE-774 or BE-933, and then centrifuged (4 °C, 3500 × g, 2 min) to separate the serum. The concentrations of BE-774 or BE-933 were measured by an in-house developed ELISA. Briefly, a labeled GPC3 protein (Catalog No.: C414, Novoprotein Co., Ltd., China) was used as the capture reagent, and a biotinylated CD137 antigen (Catalog No.: 41B-H82E6, ACRO Co., China) was used as the detection reagent for the measurement of BE-774 or BE-933. The pharmacokinetic (PK) characteristics of BE-774 and BE-933 at the 3 mg / kg dose level are shown in Figure 8 . The PK parameters of BE-774 and BE-933 at the 3 mg / kg dose level are shown in Table 10. BE-933 had favorable pharmacokinetic characteristics in hFcRn mice, and compared with BE-933, BE-774 with the YTE mutation in Fc showed significantly improved pharmacokinetic characteristics. Table 10. PK Parameters of BE-774 and BE-933 in hFcRn Mice Example 11. SPR Binding of BE-915 to CD137 and GPC3

[0291] The binding kinetics of the antibody was measured using surface plasmon resonance (SPR). The association rate constant (K a ) and dissociation rate constant (K d ) of the antibody with CD137 and GPC3 recombinant proteins were measured by SPR, and then the affinity constant (K D ) was determined. The results showed that BE-915 had good binding affinities for both human CD137 and human GPC3 (Tables 11 and 12).

[0292] To test the binding specificity of BE-915 to CD137 from different species, human CD137 (Catalog No.: 41B-H5256, Acrobio, China) and cynomolgus monkey CD137 ECD (SEQ ID NO:102) were used as bait proteins for SPR binding studies. BE-915 showed high binding affinity for human CD137, with a K D of approximately 3.6 nM, which was similar to its binding affinity for cynomolgus monkey CD137 (K D of approximately 4.4 nM), as shown in Table 11. This also indicated that the anti-CD137 antibody BGA-2524 used in BE-915 had a great binding affinity for both human CD137 and cynomolgus monkey CD137.

[0293] To test the binding specificity of BE-915 to GPC3 from different species, human GPC3 (Catalog No.: 10088-H08H, Sino Biological, China) and cynomolgus monkey GPC3 (Catalog No.: GP3-C5225, Acrobio, China) were used as bait proteins for SPR binding studies. BE-915 showed binding affinity for human GPC3 (K D of approximately 2.4 nM) and for cynomolgus monkey GPC3 (K D of approximately 0.53 nM), as shown in Table 12. Table 11. SPR Binding of BE-915 to Human and Cynomolgus Monkey CD137 Antigen Antibody <![CDATA[K a (1 / Ms)]]> <![CDATA[K d (1 / s)]]> <![CDATA[K D (M)]]> Human CD137 BE-915 1.2e+5 4.2e-4 3.6e-9 Cynomolgus Macaque CD137 BE-915 1.3e+5 5.5e-4 4.4e-9 Table 12. SPR Binding of BE-915 to Human and Cynomolgus Monkey GPC3 Antigen Antibody <![CDATA[K a (1 / Ms)]]> <![CDATA[K d (1 / s)]]> <![CDATA[K D (M)]]> Human GPC3 BE-915 5.8e+4 1.4e-4 2.4e-9 Cynomolgus Macaque GPC3 BE-915 2.1e+5 1.1e-4 5.3e-10 Example 12. Binding of BE-915 to native human CD137 and native human GPC3

[0294] To verify the binding of BE-915 to native human CD137 and native human GPC3 expressed on cells, HuT78 cells overexpressing human CD137 (HuT78 / CD137) and HepG2 cells expressing native human GPC3 were used to evaluate the binding of BE-915, respectively. Fluorescence-activated cell sorting (FACS) showed that BE-915 had strong binding activity to human CD137 and human GPC3 in a dose-dependent manner, with EC50 values of 0.69 nM (Figure 9B) and 1.09 nM (Figure 9A), respectively, as shown in Figure 9. The isotype control huIgG (Catalog No.: 02-7102, Thermo, USA) had no binding activity to HuT78 / CD137 or HepG2 cells. Example 13. Binding specificity of BE-915 to other TNFRSF members

[0295] To test the binding specificity of BE-915 to other TNFRSF members, ELISA was performed by coating TNFRSF4 / OX40 (Catalog No.: OXO-H5252, Acrobio, China), TNFRSF7 / CD27 (Catalog No.: CD7-H5257, Acrobio, China), TNFRSF14 / HVEM (Catalog No.: CD7-H522b, Acrobio, China), TNFRSF8 / CD30 (Catalog No.: HVM-H5258, Acrobio, China), and human CD137 (Catalog No.: 41B-H5256, Acrobio, China) on plates and then binding with BE-915. As Figure 10 shown, BE-915 specifically bound to human CD137 but not to other TNFRSF members. This also indicated that the anti-CD137 antibody BGA-2524 used in BE-915 specifically bound to human CD137. Example 14. Competitive binding of BE-915 with human CD137L to human CD137

[0296] To accurately evaluate the blockade of CD137-CD137L binding at the cellular level, a cell-based blockade assay using HuT78 / CD137 was established. The competitive blockade of CD137L on the interaction between CD137 and BE-915 was measured by detecting the binding of BE-915 (starting from a concentration of 15 μg / mL and then serially diluted 3-fold) to human CD137 expressed on HuT78 in the presence of 10 μg / mL, 1 μg / mL, 0.1 μg / mL, or 0 μg / mL CD137L (Catalog No.: 41L-H52D4, Acrobio, China). The binding signal was detected by using a secondary antibody anti-hFc 647 (Catalog No.: 109-605-098, Jackson, USA). huIgG (Catalog No.: 02-7102, Thermo, USA) was used as an isotype control (Figure 11A). As the CD137L concentration increased, the binding of BE-915 to CD137 decreased (Figure 11A). The competitive blockade of BE-915 on the interaction between CD137 and CD137L was measured by detecting the binding of CD137L (starting from a concentration of 15 μg / mL and then serially diluted 3-fold) to CD137 expressed on HuT78 in the presence of 1 μg / mL, 0.1 μg / mL, or 0 μg / mL BE-915. The binding signal was detected by using a secondary antibody anti-his647 (Catalog No.: A01802, Genscript, China). huIgG (Catalog No.: 02-7102, Thermo, USA) was used as an isotype control (Figure 11B). As the BE-915 concentration increased, the binding of CD137L to CD137 decreased (Figure 11B). These data indicate that BE-915 cross-competes with CD137L for binding to human CD137. This also indicates that the anti-CD137 antibody BGA-2524 used in BE-915 cross-competes with CD137L for binding to human CD137. Example 15. BE-915 induces activation of T cells co-cultured with GPC3-positive tumor cells

[0297] The functional activity of the GPC3 x CD137 bispecific antibody BE-915 was evaluated in an in vitro co-culture experiment using human peripheral blood mononuclear cells (PBMC) and a hepatocellular carcinoma (HCC) cell line expressing OS8 (Figure 12A). Three HCC cell lines, HepG2, Huh7, and Hep3B, with high to low GPC3 expression based on FACS analysis (Figure 12B) were selected to evaluate the effect of GPC3 levels on the functional activity of BE-915. SK-HEP-1, which does not express GPC3, was used as a negative control cell line.

[0298] Thawed frozen human PBMCs (OriBiotech) in RPMI 1640 medium and incubated overnight at 37 °C. Target cells expressing OS8 were seeded into 384-well plates and allowed to attach for 16 h. The next day, PBMCs were added to the 384-well plates at an effector-to-target cell ratio (E:T) of 2:1. Then, the co-cultured cells were treated with serial dilutions of BE-915 at 37 °C for 48 h. Culture supernatants were collected for subsequent measurement of IFN-γ and IL-2 concentrations by a TR-FRET-based method (Degorce, et al. Current Chemical Genomics. 2009, 3:22) as described in the manufacturer's manual (Cisbio). Results showed that BE-915 induced dose-dependent cytokine release in PBMCs from two independent donors co-cultured with GPC3-expressing cells but not GPC3-negative cells (Figure 12C). Example 16. Enhancement of PBMC-based cell killing by BE-915 in co-culture with GPC3-positive tumor cells

[0299] In co-culture experiments, the T cell killing activity modulated by BE-915 was evaluated by impedance measurement using the xCELLigence RTCA MP instrument (Agilent Technologies). Thawed frozen human PBMCs (OriBiotech) in RPMI 1640 medium and incubated overnight at 37 °C. Target cells were seeded into 96-well E-plates (Agilent Technologies) and allowed to attach for 16 h. The next day, PBMCs were added to the 96-well E-plates at an effector-to-target cell ratio (E:T) of 5:1. Then, the co-cultured cells were treated with a combination of serial dilutions of BE-915 and an EpCAM / CD3 bispecific T cell engager (BiTE) that provides signal 1 for T cell activation (Figure 13A). Three HCC cell lines, HepG2, Huh7, and Hep3B, with high to low GPC3 expression based on FACS analysis (Figure 13B) were selected to evaluate the effect of GPC3 levels on the functional activity of BE-915. SK-OV-3, which does not express GPC3, was used as a negative control cell line.

[0300] The experiment was allowed to proceed for 4 days, and the electrode impedance was measured by viable adherent target cells. Consistent with the cytokine production assay, BE-915 dose-dependently enhanced the killing activity of T cells against GPC3-expressing cells but not against GPC3-negative cells (Figure 13C). PBMCs from two donors were used in this experiment. Example 17. Pharmacokinetic characteristics of BE-915 in cynomolgus monkeys

[0301] Blood samples were collected from cynomolgus monkeys at 0, 0.167 h, 1 h, 4 h, 8 h, 1, 3, 6, 9, 13, 20, and 27 days after intravenous infusion of 5 mg / kg of BE-915, and then centrifuged (4 °C, 3000×g, 15 min) to separate the serum. The concentration of BE-915 was measured by an in-house developed ELISA ligand-binding method. Briefly, a labeled GPC3 antigen (Catalog No.: C414, Novoprotein Scientific Inc., China) was used as the capture reagent, and a biotinylated CD137 antigen (Catalog No.: 41B-H82E6, ACRO Biosystems, China) was used as the detection reagent for BE-915. The obtained pharmacokinetic characteristics and parameters are shown respectively in Figure 14 and Table 13. In the 5 mg / kg dosing group, BE-915 was below the lower limit of quantification (0.0391 μg / mL) on day 13 after dosing. Anti-drug antibodies (ADA) were detected in the serum in the 5 mg / kg dosing group starting from day 9, indicating a potential impact on the pharmacokinetic curve. PK parameters were calculated after removing the concentration values at the time points. The clearance rate of BE-915 was 10.4 mL / day / kg, and the antibody-like half-life was 3.2 days. Table 13. Pharmacokinetic parameters of BE-915 in cynomolgus monkeys after i.v. infusion Note: A non-compartmental model was used to calculate the pharmacokinetic parameters Example 18. Efficacy of BE-915 monotherapy in the MC38 / hGPC3 model of humanized CD137 knock-in mice

[0302] The in vivo efficacy of BE-915 was examined in the MC38 / hGPC3 murine colorectal cancer model of humanized CD137 knock-in mice. MC38 / hGPC3 cells were subcutaneously implanted into the right flank of recipient mice. Seven days after cell inoculation, the mice were randomly divided into 4 groups according to tumor volume. BE-915 was administered intraperitoneally on day 1 and once a week for 18 days. BE-915 (0.1, 0.5, and 3.0 mg / kg, once a week) effectively inhibited tumor growth. The tumor volume was significantly reduced at the study endpoint (D18). In addition, on day 18, the tumor-free rates in the 0.1, 0.5, and 3.0 mg / kg groups were 0%, 0%, and 10% respectively ( Figure 15 and Table 14). Throughout the study, no significant effect on animal body weight was observed in any treatment group. Table 14. Efficacy of BE-915 in the MC38 / hGPC3 syngeneic tumor model of humanized CD137 knock-in mice Abbreviations: hGPC3, human glypican 3; n, number of animals; NA, not applicable; QW, once a week; SEM, standard deviation of the mean; TGI, tumor growth inhibition. Note: The TGI rate was calculated according to the following formula: %TGI = [1 - (Tt of treatment - T0 of treatment) / (Tt of vehicle - T0 of vehicle)] × 100%. Tt of treatment = mean tumor volume of the treatment group on day t; T0 of treatment = mean tumor volume of the treatment group on day 0; Tt of vehicle = mean tumor volume of the vehicle group on day t; and T0 of vehicle = mean tumor volume of the vehicle group on day 0. Example 19. Efficacy of the combination of BE-915 and anti-PD-1 antibody in the LL / 2 / hGPC3 model of humanized 4-1BB knock-in mice

[0303] The anti-tumor activity of the combination of BE-915 and anti-mouse PD-1 antibody was investigated in the LL / 2 / hGPC3 syngeneic model (lung cancer) of humanized CD137 knock-in mice. LL / 2 / hGPC3 cells were implanted into mice. Seven days after cell inoculation, the mice were randomly divided into 4 groups according to tumor volume. Mice treated with the combination of BE-915 (10.0 mg / kg, once a week) and anti-mouse PD-1 antibody (10.0 mg / kg, once a week) showed synergistic tumor growth inhibition. On day 13, the tumor growth inhibition rate in the combination group was 74.7%, significantly higher than that in the groups treated with BE-915 alone (34.1%) or anti-PD-1 (38.0%) ( Figure 16 and Table 15). During the entire study period, no significant effect on the body weight of animals in any treatment group was observed. Table 15. Anti-tumor effects of BE-915 and anti-mouse PD-1 antibody in the LL / 2 / hGPC3 syngeneic model of humanized CD137 knock-in mice Abbreviations: hGPC3, human glypican 3; n, number of animals; NA, not applicable; QW, once a week; SEM, standard deviation of the mean; TGI, tumor growth inhibition. Note: The TGI rate was calculated according to the following formula: %TGI = [1 - (Tt of treatment - T0 of treatment) / (Tt of vehicle - T0 of vehicle)] × 100%. Tt of treatment = mean tumor volume of the treatment group on day t; T0 of treatment = mean tumor volume of the treatment group on day 0; Tt of vehicle = mean tumor volume of the vehicle group on day t; and T0 of vehicle = mean tumor volume of the vehicle group on day 0. Example 20. Biophysical properties of BE-915

[0304] The biophysical properties of BE-774 (using camelid CD137 VHH BGA-9612) and BE-915 (using humanized CD137 VHH BGA-2524) were tested. The biophysical properties tested included melting temperature, aggregation temperature, hydrophobicity by HIC-HPLC, and self-association tendency by AC-SINS (described in detail below). BE-915 showed the best thermal stability by Tm and Tagg, as well as good colloidal stability by AC-SINS. As shown in Table 16, BE-915 showed overall biophysical properties comparable to BE-774. This also indicates that the humanized CD137 VHH BGA-2524 used in BE-915 has overall biophysical properties comparable to the camelid CD137 VHH BGA-9612 used in BE-774. Table 16. Summary of biophysical properties of BE-915 and BE-774

[0305] The melting temperature (Tm) and aggregation temperature Tagg (°C) were determined by UNCLE TM (Unchained lab, Pleasanton, CA), an instrument that simultaneously measures intrinsic fluorescence and static light scattering. During the measurement, 9 μL of a 1 mg / mL protein sample in PBS buffer was loaded into a cuvette; the sample was held at 20 °C for 120 s and then heated to 95 °C at a rate of 0.3 °C / min. After excitation at 266 nm, fluorescence and static light scattering (at 266 nm) were collected.

[0306] To determine the hydrophobicity of a given antibody using an HPLC system, a 50 μg sample (1 mg / ml) was diluted with mobile phase A solution (1.5 M ammonium sulfate, 50 mM sodium phosphate, pH 7.0) before analysis to achieve a final ammonium sulfate concentration of approximately 1 M. Using a MABPac HIC-10 column, a linear gradient of mobile phase A and mobile phase B solution (50 mM sodium phosphate, pH 7.0) was run at a flow rate of 0.5 mg / min for 29 minutes. Peak retention times were monitored at A280 absorbance.

[0307] AC-SINS assays measure protein self-interaction by capturing antibodies on the surface of gold colloids, which exhibit surface resonance oscillations at visible light frequencies. When the immobilized antibodies self-interact, the colloids aggregate, changing the oscillation frequency to absorb longer wavelengths. Gold nanoparticles are incubated with an 80 / 20 (v / v) mixture of capture antibody / non-capture antibody. The coated gold nanoparticles are then centrifuged and resuspended in PBS. The samples are diluted to 0.05 mg / ml (in conjugation buffer), and 45 μl of each dilution is loaded onto a 384-well plate. Then 5 μl of the previously prepared gold nanoparticles is added to each well of the plate containing mAb and buffer controls. The plate is then covered with an aluminum lid, incubated at room temperature for 2 hours, and centrifuged rapidly at 3000 rpm before reading the absorption spectrum of each well from 450 to 650 nm using a plate reader. The spectrum of each sample is recorded, and the red shift of the maximum absorption peak compared to the buffer is analyzed. The red shift and its intensity indicate the self-interaction propensity of the tested mAb sample.

[0308] The amino acid and DNA sequences of the anti-CD137 VHH and anti-GPC3xCD137 bispecific antibodies are shown in Table 17 below. Table 17. Amino acid and DNA sequences of the anti-CD137 VHH and anti-GPC3xCD137 bispecific antibodies Example 21. Structural and Functional CD137 Epitope Mapping

[0309] To better understand how the anti-CD137 single-domain antibody arm can have high affinity for CD137 and be a potent agonist of CD137 / CD137L interaction, the crystal structure of the VHH (BGA-2524) complex with CD137 was determined. A. CD137 and VHH (BGA-2524) Expression, Purification, and Crystallization

[0310] The extracellular domain of human CD137 containing a partial three CRDs (CRD1-3; amino acids 24-105 of SEQ ID NO:35 (human CD137-full length)) was expressed in HEK293G cells. The cDNA encoding CD137 was cloned into the pMAX vector with an N-terminal secretion sequence and a C-terminal TEV cleavage site followed by an Fc tag. The culture supernatant containing the secreted CD137-Fc fusion protein was mixed with Mab Select Sure TM resin (GE Healthcare LifeSciences) at 4 °C for 3 hours. The protein was washed with a buffer containing 20 mM Tris-HCl pH 8.0, 150 mM NaCl, then eluted with 50 mM acetic acid (pH adjusted to 3.5 with 5 M NaOH), and finally neutralized with 1 / 10 CV 1.0 M Tris-HCl pH 8.0. The eluted protein was mixed with TEV protease (10:1 molar ratio) and dialyzed overnight at 4 °C against a buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl). The mixture was loaded onto a Ni-NTA column (Qiagen) and Mab Select Sure TM resin to remove TEV protease and the Fc tag, and then a HiLoad 16 / 600 Superdex TMThe 75pg column (GE Healthcare Life Sciences) further purified the effluent by size exclusion chromatography in buffer (20 mM Tris pH 8.0, 100 mM NaCl).

[0311] The DNA sequence encoding VHH (BGA-2524) was cloned into the PET21a vector with an N-terminal HIS-MBP tag followed by a TEV protease cleavage site. Protein expression in Shuffle T7 was induced with 1 mM IPTG at 18 °C for 16 h when OD600 was 0.6 - 1.0. Cells were harvested by centrifugation at 7,000 g for 10 min. The cell pellet was resuspended in lysis buffer (50 mM Tris-HCl pH 8.0, 300 mM NaCl) and lysed by sonication on ice. The lysate was then centrifuged at 48,000 g for 30 min at 4 °C. The supernatant was mixed with Talon resin and batch-incubated at 4 °C for 3 h. The resin was washed with lysis buffer containing 5 mM imidazole and the protein was eluted with lysis buffer containing an additional 100 mM imidazole. The eluate was mixed with TEV protease (10:1 molar ratio) and dialyzed overnight at 4 °C against buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl). The mixture was loaded onto a Talon column to remove TEV protease and the HIS-MBP tag, and then used HiLoad 16 / 600 Superdex TM The 75pg column (GE Healthcare Life Sciences) further purified the effluent by size exclusion chromatography in buffer (20 mM Tris pH 8.0, 100 mM NaCl).

[0312] Excess purified CD137 was mixed with purified VHH (BGA-2524) (1.2:1 molar ratio) to generate the CD137 / VHH (BGA-2524) complex. Then, using Superdex TM The 75Increase 10 / 300 column (GE Healthcare Life Sciences) further purified the complex by gel filtration in buffer (20 mM Tris pH 8.0, 100 mM NaCl). The CD137 / VHH (BGA-2524) complex (10 mg / ml) was crystallized in 18% PEG 4000, 0.1 M Tris pH 8.7, 0.2 M Li2SO4. Crystals cryoprotected with 20% PEG 4000, 0.1 M Tris pH 8.7, 0.2 M Li2SO4, 10% glycerol were flash-frozen in liquid nitrogen. X-ray diffraction data were collected at the beamline BL02U1 of the Shanghai Synchrotron Radiation Facility (Shanghai, China). B. Data collection and structure determination

[0313] X-ray diffraction data were collected at beamline BL02U1 of the Shanghai Synchrotron Radiation Facility (Shanghai, China) under cryogenic cooling conditions of 100 Kelvin. The diffraction images were processed using the integrated data processing software XDS (Kabsch, W., Xds. Acta Crystallogr D Biol Crystallogr [Acta Crystallographica Section D: Biological Crystallography], 2010.66 (Pt 2): pp. 125 - 32). The structures of human CD137 (PDB: 6MGP) and an internal VHH model were used as search models. The initial solution was found by the molecular replacement program PHASER (McCoy, A.J., et al., Phaser crystallographic software. [Phaser Crystallography Software] J Appl Crystallogr [Journal of Applied Crystallography], 2007.40 (Pt 4): pp. 658 - 674). Then, the model was built manually and iteratively using the program COOT (Emsley, P. and K. Cowtan, Coot: model-building tools for molecular graphics [Coot: Molecular Graphics Model Building Tools]. Acta Crystallogr D Biol Crystallogr [Acta Crystallographica Section D: Biological Crystallography], 2004.60 (Pt 12 Pt 1): pp. 2126 - 32) and refined using PHENIX (Adams, P.D., et al., PHENIX: a comprehensive Python-based system for macromolecular structure solution [PHENIX: A Comprehensive Python-Based System for Macromolecular Structure Determination]. Acta Crystallogr D Biol Crystallogr [Acta Crystallographica Section D: Biological Crystallography], 2010.66 (Pt 2): pp. 213 - 21). The final model was refined to acceptable R and R-free values and Ramachandran statistics (calculated by Molprobity). The data processing and refinement statistics can be found in Table 18. Table 18. Data collection and refinement statistics a Values in parentheses are for the highest resolution shell. b Calculated from approximately 5% of the reflections set aside during refinement cr.m.s.d., root mean square deviation C. Structure of VHH (BGA-2524) that binds to human CD137 dimer

[0314] VHH (BGA-2524) complexed with CD137 crystallized in the P21 space group, with two complexes in the asymmetric unit and diffracted to The structure of VHH (BGA-2524) that binds to human CD137 indicates that the binding part of VHH (BGA-2524) to CD137L is spatially connected in part ( Figure 17 ). The buried surface area between VHH (BGA-2524) and CD137 is approximately The interactions of VHH (BGA-2524) cluster around the CD137 CRD2 domain. VHH (BGA-2524) binds mainly to the lateral surface of the CD137 CRD2 domain through CDR residues. All CDRs of VHH (BGA-2524) are involved in the binding of CD137 dimer, especially CDR3 contributes the greatest potential. CDR1 and CDR3 bind to the two monomers of the CD137 dimer, while CDR2 binds only to one monomer of the CD137 dimer. The curved CDR3 loop always covers the hydrophobic patch of the VHH framework, which may result in better biophysical properties ( Figure 18)。The CDR1 Tyr32 of VHH (BGA-2524) contacts one monomer of the CD137 dimer at residue Gly98, while the CDR1 Asn31 and Ala33 of VHH (BGA-2524) contact the other monomer of the CD137 dimer at residues Ile64 and Gln67. The CDR2 Trp52, Ser53, Tyr55, His57 of VHH (BGA-2524) contact only one monomer of the CD137 dimer at residues Asp38, Pro49, Pro50, Asn51, Ile64. The CDR3 residues Leu98, Thr104, Thr106 and Tyr109 of VHH (BGA-2524) contact one monomer of the CD137 dimer at residues Ser55, Ala56, Arg75, Glu85 and Ala97, while the CDR3 residues Leu98, Lys99, Tyr100 and Pro101 of VHH (BGA-2524) contact the other monomer of the CD137 dimer at residues Phe36, Pro49, Thr61, Cys62, Asp63 and Ile64. VHH (BGA-2524) interacts with CD137 using a combination of hydrogen bonds and salt bridges as well as hydrophobic interactions. For example, the CDR2 residue His57 of VHH (BGA-2524) forms two salt bridges with the CD137 residue Asp38. The CDR3 residue Lys99 of VHH (BGA-2524) forms two salt bridges with the CD137 residue Asp63. The residues Tyr32, Ser53, His57, Leu98, Lys99, Pro101 and Thr106 of VHH (BGA-2524) form one hydrogen bond each with the CD137 residues Gly98, Asn51, Asp38, Ile64, Asp63, Thr61 and Ser55 respectively (i.e., Tyr32-Gly98, Ser53-Asn51, His57-Asp38, Leu98-Ile64, Lys99-Asp63, Pro101-Thr61 and Thr106-Ser55). The residue Trp52 of VHH (BGA-2524) forms two hydrogen bonds with the CD137 residues Pro50 and Asn51. The residue Tyr109 of VHH (BGA-2524) forms two hydrogen bonds with the CD137 residues Arg75 and Glu85. The residue Tyr100 of VHH (BGA-2524) forms two hydrogen bonds with the CD137 residue Cys62( Figure 19 )。

[0315] Based on the crystal structure of the VHH (BGA-2524) / CD137 complex, the CD137 residues contacted by VHH (BGA-2524) (i.e., the epitope residues of CD137 bound by VHH (BGA-2524)) and the VHH (BGA-2524) residues contacted by CD137 (i.e., the counter-residues of VHH (BGA-2524) contacted by CD137) were determined. Tables 19 and 20 below show the residues of CD137 and VHH (BGA-2524) that they contact, as evaluated using the contact distance strictness, which is the point with the highest van der Waals (non-polar) interaction force. Table 19. Epitope Residues of CD137 and Their Corresponding Counter-Residues of VHH (BGA-2524) Table 20. Counter-Residues of VHH (BGA-2524) and Their Corresponding CD137 Epitope Residues

Claims

1. An antibody or antigen-binding fragment thereof that specifically binds to human CD137, wherein (1) the antibody or antigen-binding fragment thereof specifically binds to the following epitope, which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35); (2) the antibody or antigen-binding fragment thereof specifically binds to the following epitope, which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97 and Gly98 of human CD137 (SEQ ID NO: 35); or (3) the antibody or antigen-binding fragment thereof specifically binds to a human CD137 dimer, which comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), which comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97 and Gly98; and / or the antibody or antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

2. An antibody or antigen-binding fragment thereof that specifically binds to human CD137, the antibody or antigen-binding fragment thereof comprising: (i) a heavy chain variable region (VH) that comprises (a) HCDR1 (heavy chain complementarity-determining region 1) of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) a heavy chain variable region that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO:

3.

3. The antibody or antigen-binding fragment thereof according to any one of claims 1-2, the antibody or antigen-binding fragment thereof comprising: (i) a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:

4.

4. The antibody or antigen-binding fragment thereof according to claim 3, wherein one, two, three, four, five, six, seven, eight, nine, or ten amino acids have been inserted, deleted, or substituted in SEQ ID NO: 17, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15, or SEQ ID NO:

4.

5. The antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment comprises: (i) A variable heavy chain (VH) that comprises SEQ ID NO: 17; (ii) A variable heavy chain (VH) that comprises SEQ ID NO: 11; (iii) A variable heavy chain (VH) that comprises SEQ ID NO: 13; (iv) A variable heavy chain (VH) that comprises SEQ ID NO: 15; or (v) A variable heavy chain (VH) that comprises SEQ ID NO:

4.

6. The antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), Fab fragment, Fab' fragment, or F(ab’)2 fragment.

7. The antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment comprises a heavy chain constant region of the IgG1, IgG2, IgG3, or IgG4 subclass and / or a light chain constant region of the κ or λ type.

8. The antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment has antibody-dependent cell cytotoxicity (ADCC) or complement-dependent cell cytotoxicity (CDC).

9. An antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment thereof has reduced glycosylation or is non-glycosylated or is hypofucosylated.

10. An antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment thereof comprises an increased bisecting GlcNac structure.

11. An antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO: 19 or SEQ ID NO:

53.

12. An antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function and / or an extended half-life, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:

20.

13. A pharmaceutical composition comprising an antibody or antigen-binding fragment thereof according to any one of the preceding claims, and a pharmaceutically acceptable carrier.

14. A method of treating cancer, the method comprising administering to a patient in need thereof a therapeutically effective amount of an antibody or antigen-binding fragment thereof according to any one of claims 1-12, or a pharmaceutical composition according to claim 13.

15. The method according to claim 14, wherein the cancer is gastric cancer, colon cancer, pancreatic cancer, breast cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, small cell lung cancer, non-small cell lung cancer, ovarian cancer, skin cancer, mesothelioma, lymphoma, leukemia, myeloma and sarcoma.

16. The method according to any one of claims 14-15, wherein the antibody or antigen-binding fragment thereof is administered in combination with another therapeutic agent.

17. The method according to claim 16, wherein the therapeutic agent is an anti-PD-1 antibody.

18. The method according to claim 17, wherein the anti-PD1 antibody is tislelizumab.

19. A multispecific antibody or antigen-binding fragment thereof, the multispecific antibody or antigen-binding fragment thereof comprising at least a first antigen-binding domain that specifically binds to a human tumor-associated antigen (TAA), and at least a second antigen-binding domain that specifically binds to human CD137, wherein the second antigen-binding domain is (1) an antibody or antigen-binding fragment thereof that specifically binds to the following epitope, the epitope comprising the amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35), or consisting of the same; (2) An antibody or antigen-binding fragment thereof that specifically binds to the following epitope, which epitope comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35); or (3) An antibody or antigen-binding fragment thereof that specifically binds to a human CD137 dimer, which human CD137 dimer comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO: 35), which epitope comprises or consists of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO: 35), which epitope comprises or consists of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98; and / or the antibody or antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

20. A multispecific antibody or antigen-binding fragment thereof, which multispecific antibody or antigen-binding fragment thereof comprises at least a first antigen-binding domain that specifically binds to a human tumor-associated antigen (TAA) and at least a second antigen-binding domain that specifically binds to human CD137, wherein the second antigen-binding domain comprises: (i) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 2, and (c) HCDR3 of SEQ ID NO: 3; or (ii) A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO: 1, (b) HCDR2 of SEQ ID NO: 10, and (c) HCDR3 of SEQ ID NO:

3.

21. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-20, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 17; (ii) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 11; (iii) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 13; (iv) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO: 15; or (v) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:

4.

22. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-21, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises SEQ ID NO: 17; (ii) A heavy chain variable region (VH) that comprises SEQ ID NO: 11; (iii) A heavy chain variable region (VH) that comprises SEQ ID NO: 13; (iv) A heavy chain variable region (VH) that comprises SEQ ID NO: 15; or (v) A heavy chain variable region (VH) that comprises SEQ ID NO:

4.

23. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-22, wherein the TAA is GPC3.

24. A multispecific antibody or antigen-binding fragment thereof that comprises a first antigen-binding domain that specifically binds to human phosphatidylinositol proteoglycan 3 (GPC3) and a second antigen-binding domain that specifically binds to human CD137.

25. The multispecific antibody or antigen-binding fragment thereof according to claim 24, wherein the second antigen-binding domain that specifically binds to human CD137 is (1) An antibody or antigen-binding fragment thereof that specifically binds to the following epitope, which comprises the amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67 of human CD137 (SEQ ID NO: 35), or consists of the same; (2) An antibody or antigen-binding fragment thereof that specifically binds to the following epitope, which comprises the amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98 of human CD137 (SEQ ID NO: 35), or consists of the same; or (3) An antibody or antigen-binding fragment thereof that specifically binds to a human CD137 dimer, which human CD137 dimer comprises or consists of a first human CD137 monomer and a second human CD137 monomer; wherein the antibody or antigen-binding fragment thereof specifically binds to an epitope of the first human CD137 monomer (SEQ ID NO:35), the epitope comprising or consisting of amino acid residues Phe36, Asp38, Pro49, Pro50, Asn51, Thr61, Cys62, Asp63, Ile64, Gln67; and the antibody or antigen-binding fragment thereof specifically binds to an epitope of the second human CD137 monomer (SEQ ID NO:35), the epitope comprising or consisting of amino acid residues Ser55, Ala56, Arg75, Glu85, Ala97, and Gly98; and / or the antibody or antigen-binding fragment thereof binds to the human CD137 dimer and promotes human CD137 aggregation.

26. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 24-25, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:2, and (c) HCDR3 of SEQ ID NO:3; or (ii) a heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:10, and (c) HCDR3 of SEQ ID NO:

3.

27. The multispecific antibody or antigen-binding fragment thereof according to claim 26, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:17; (ii) a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:11; (iii) a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:13; (iv) a heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:15; or (v) A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:

4.

28. The multispecific antibody or antigen-binding fragment thereof according to claim 27, wherein one, two, three, four, five, six, seven, eight, nine, or ten amino acids have been inserted, deleted, or substituted in SEQ ID NO:17, SEQ ID NO:11, SEQ ID NO:13, SEQ ID NO:15, or SEQ ID NO:

4.

29. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 24-28, wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) A heavy chain variable region (VH) that comprises SEQ ID NO:17; (ii) A heavy chain variable region (VH) that comprises SEQ ID NO:11; (iii) A heavy chain variable region (VH) that comprises SEQ ID NO:13; (iv) A heavy chain variable region (VH) that comprises SEQ ID NO:15; or (v) A heavy chain variable region (VH) that comprises SEQ ID NO:

4.

30. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-29, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: A heavy chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:45, (b) HCDR2 of SEQ ID NO:46, and (c) HCDR3 of SEQ ID NO:47; and A light chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO:48, (e) LCDR2 of SEQ ID NO:49, and (f) LCDR3 of SEQ ID NO:

50.

31. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-30, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: A heavy chain variable region (VH) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:41, and a light chain variable region (VL) that comprises an amino acid sequence that is at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to SEQ ID NO:

43.

32. The multispecific antibody or antigen-binding fragment thereof according to claim 31, wherein one, two, three, four, five, six, seven, eight, nine, or ten amino acids have been inserted, deleted, or substituted in SEQ ID NO:41 or SEQ ID NO:

43.

33. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-32, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: a heavy-chain variable region (VH) comprising SEQ ID NO:41, and a light-chain variable region (VL) comprising SEQ ID NO:

43.

34. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-33, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:45, (b) HCDR2 of SEQ ID NO:46, and (c) HCDR3 of SEQ ID NO:47; and a light-chain variable region (VL) that comprises (d) LCDR1 of SEQ ID NO:48, (e) LCDR2 of SEQ ID NO:49, and (f) LCDR3 of SEQ ID NO:50, and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:2, and (c) HCDR3 of SEQ ID NO:3; or (ii) a heavy-chain variable region (VH) that comprises (a) HCDR1 of SEQ ID NO:1, (b) HCDR2 of SEQ ID NO:10, and (c) HCDR3 of SEQ ID NO:

3.

35. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-34, wherein the first antigen-binding domain that specifically binds to human GPC3 comprises: a heavy-chain variable region (VH) comprising SEQ ID NO:41, and a light-chain variable region (VL) comprising SEQ ID NO:43; and wherein the second antigen-binding domain that specifically binds to human CD137 comprises: (i) a heavy-chain variable region (VH) that comprises SEQ ID NO:4; (ii) a heavy-chain variable region (VH) that comprises SEQ ID NO:11; (iii) a heavy-chain variable region (VH) that comprises SEQ ID NO:13; (iv) a heavy-chain variable region (VH) that comprises SEQ ID NO:15; or (v) a heavy-chain variable region (VH) that comprises SEQ ID NO:

17.

36. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-35, wherein the multispecific antibody or antigen-binding fragment thereof is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), Fab fragment, Fab' fragment or F(ab')2 fragment.

37. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-36, wherein the first antigen-binding domain that specifically binds to human GPC3 is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), single-domain antibody, Fab fragment, Fab' fragment or F(ab')2 fragment, and the second antigen-binding domain that specifically binds to human CD137 is a monoclonal antibody, chimeric antibody, humanized antibody, human engineered antibody, single-chain antibody (scFv), single-domain antibody, Fab fragment, Fab' fragment or F(ab')2 fragment.

38. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-37, wherein the multispecific antibody or antigen-binding fragment thereof is a bispecific antibody.

39. The multispecific antibody or antigen-binding fragment thereof according to claim 38, wherein the bispecific antibody is in the 2+2 form.

40. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-39, wherein the multispecific antibody or antigen-binding fragment thereof contains a linker of SEQ ID NOs: 60 to 101.

41. The multispecific antibody or antigen-binding fragment thereof according to claim 40, wherein the linker is SEQ ID NO:

62.

42. The multispecific antibody or antigen-binding fragment thereof according to claim 40, wherein the linker is SEQ ID NO:

67.

43. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-42, wherein the multispecific antibody or antigen-binding fragment contains a heavy chain constant region of IgG1, IgG2, IgG3 or IgG4 subclass and / or a light chain constant region of κ or λ type, and wherein the heavy chain constant region contains CH1 and / or Fc domain.

44. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-43, wherein the multispecific antibody or antigen-binding fragment has antibody-dependent cell cytotoxicity (ADCC) or complement-dependent cell cytotoxicity (CDC).

45. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-44, wherein the multispecific antibody or antigen-binding fragment has reduced glycosylation or is non-glycosylated or is hypofucosylated.

46. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-45, wherein the multispecific antibody or antigen-binding fragment contains an increased bisecting GlcNac structure.

47. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-46, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:

53.

48. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-47, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG1 Fc with reduced effector function and / or an extended half-life, optionally the Fc domain comprises the amino acid sequence of SEQ ID NO:

20.

49. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-48, wherein the multispecific antibody or antigen-binding fragment thereof comprises an Fc domain, and wherein the Fc domain is an IgG4 Fc.

50. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-49, wherein: a) the heavy chain variable region (VH), CH1 domain, the Fc domain of the first antigen-binding domain that specifically binds to human GPC3, and the heavy chain variable region (VH) of the second antigen-binding domain that specifically binds to human CD137 are arranged in the first polypeptide in the N-terminal to C-terminal direction; optionally, the C-terminus of the Fc domain is linked to the N-terminus of the heavy chain variable region (VH) of the second antigen-binding domain through a linker; and b) the light chain variable region (VL) and the first light chain constant region of the first antigen-binding domain that specifically binds to human GPC3 are arranged in the second polypeptide in the N-terminal to C-terminal direction.

51. The multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-50, wherein the multispecific antibody or antigen-binding fragment comprises (i) the first polypeptide of SEQ ID NO:25 and the second polypeptide of SEQ ID NO:23; (ii) the first polypeptide of SEQ ID NO:21 and the second polypeptide of SEQ ID NO:23; (iii) the first polypeptide of SEQ ID NO:33 and the second polypeptide of SEQ ID NO:23; (iv) the first polypeptide of SEQ ID NO:27 and the second polypeptide of SEQ ID NO:23; (v) the first polypeptide of SEQ ID NO:29 and the second polypeptide of SEQ ID NO:23; or (vi) the first polypeptide of SEQ ID NO:31 and the second polypeptide of SEQ ID NO:

23.

52. A pharmaceutical composition comprising the multispecific antibody or antigen-binding fragment thereof according to any one of claims 19-51, and a pharmaceutically acceptable carrier.

53. A method for treating cancer, the method comprising administering to a patient in need thereof a therapeutically effective amount of a multispecific antibody or an antigen-binding fragment thereof according to any one of claims 19-51, or a pharmaceutical composition according to claim 52.

54. The method according to claim 53, wherein the cancer is an advanced or metastatic solid tumor.

55. The method according to any one of claims 53-54, wherein the cancer expresses GPC3.

56. The method according to any one of claims 53-55, wherein the cancer is liver cancer, lung cancer, gastric cancer, germ cell tumor, thyroid cancer, pancreatic cancer, ovarian cancer, skin cancer, kidney cancer, esophageal cancer, atypical teratoid rhabdoid tumor of the brain or undifferentiated synovial sarcoma.

57. The method according to claim 56, wherein the liver cancer is hepatoblastoma or hepatocellular carcinoma (HCC).

58. The method according to claim 56, wherein the lung cancer is non-small cell lung cancer (NSCLC) or small cell lung cancer (SCLC).

59. The method according to claim 58, wherein the non-small cell lung cancer is squamous non-small cell lung cancer.

60. The method according to claim 58, wherein the non-small cell lung cancer is GPC3+ squamous non-small cell lung cancer.

61. The method according to claim 56, wherein the gastric cancer is alpha-fetoprotein+(AFP+) gastric cancer.

62. The method according to claim 56, wherein the kidney cancer is nephroblastoma.

63. The method according to claim 56, wherein the esophageal cancer is esophageal squamous cell carcinoma.

64. The method according to claim 56, wherein the esophageal cancer is GPC3+ esophageal squamous cell carcinoma.

65. The method according to claim 56, wherein the germ cell tumor is yolk sac tumor or non-dysgerminoma.

66. The method according to any one of claims 53-65, wherein the multispecific antibody or an antigen-binding fragment thereof or the pharmaceutical composition is administered in combination with another therapeutic agent.

67. The method according to claim 66, wherein the therapeutic agent is an anti-PD1 or anti-PDL1 antibody.

68. The method according to claim 67, wherein the anti-PD1 antibody is tislelizumab.

69. An isolated nucleic acid encoding an antibody, multispecific antibody or an antigen-binding fragment thereof according to any one of claims 1-12 and 19-51.

70. A vector comprising the nucleic acid according to claim 69.

71. A host cell comprising the nucleic acid according to claim 69 or the vector according to claim 70.

72. A process for producing a multispecific antibody or an antigen-binding fragment thereof, the process comprising culturing the host cell according to claim 71 and recovering the antibody or an antigen-binding fragment thereof from the culture.

Images