Anti-CD28 composition

Abstract

This specification provides novel anti-CD28 × anti-STEAP1 antibodies and anti-CD28 × anti-CEACAM5 antibodies, as well as methods for using such antibodies for the treatment of related cancers.

Description

[Technical Field]

[0001] Claim of priority This application claims priority and interest in U.S. Provisional Applications No. 63 / 490,167 filed March 14, 2023, No. 63 / 496,374 filed April 14, 2023, No. 63 / 490,171 filed March 14, 2023, No. 63 / 495,989 filed April 13, 2023, and No. 63 / 578,592 filed August 24, 2023, the contents of which are incorporated herein by reference in their entirety. [Background technology]

[0002] Antibody-based therapies have been used to successfully treat a variety of diseases, including cancer. An increasingly widespread approach being explored is the manipulation of a single immunoglobulin molecule that co-engages with two different antigens. Such alternative antibody forms that engage with two different antigens are often referred to as bispecific antibodies. One particular approach for bispecific antibodies is that the bispecific antibody is CD3 + The goal is to design a first binding domain that engages with CD3 to redirect T cells and destroy cancer cells, and a second binding domain that engages with an antigen associated with cancer cells or an antigen upregulated on cancer cells.

[0003] However, TILs lose their cytotoxic capacity over time due to upregulation of inhibitory immune checkpoints. While checkpoint blockade has demonstrated increased clinical response rates compared to other treatment options, many patients still fail to achieve a response to checkpoint blockade. Engagement of costimulatory receptors on TILs could provide a positive signal that could overcome the negative signaling of immune checkpoints. Preclinical and clinical studies of agonist costimulatory receptor antibodies have indeed demonstrated that costimulatory receptor agonism can lead to superior antitumor responses by activating T cells to attack tumor cells.

[0004] Improving antitumor activity by specifically destroying tumor cells while minimizing peripheral toxicity is also important for cancer therapy. In this context, it is crucial that costimulatory signals are delivered only to T cells in the presence of target tumor cells. However, costimulatory receptor agonism with single-specific full-length antibodies is likely not discriminatory with respect to TILs, peripheral T cells, or autoantigen-reactive T cells that contribute to autoimmune toxicity.

[0005] Six-transmembrane prostatic epithelial antigen 1 (STEAP1) is a cell surface antigen highly expressed in certain cancers (e.g., prostate cancer, breast cancer, and bladder cancer). Carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) is a cell surface glycoprotein and a member of the carcinoembryonic (CEA) antigen family. CEACAM5 is overexpressed in many tumor types, including gastrointestinal, breast, and lung tumors. Because these antigens are overexpressed on a variety of tumors, STEAP1 and CEACAM5 are promising candidates for the development of targeted cancer therapies.

[0006] While immunotherapies targeting STEAP1 and CEACAM5 have been attempted, novel immune response-enhancing compositions are still needed to treat STEAP1-related and CEACAM5-related cancers. [Overview of the project]

[0007] This specification provides novel anti-CD28 × anti-STEAP1 antibodies and anti-CD28 × anti-CEACAM5 antibodies, as well as methods for using such antibodies for the treatment of related cancers (e.g., STEAP1-related cancer or CEACAM5-related cancer). The target anti-CD28 × anti-STEAP1 antibody can agonist-conjugate to CD28 costimulatory molecules on T cells and STEAP1 on tumor cells. The target anti-CD28 × anti-CEACAM5 antibody can agonist-conjugate to CD28 costimulatory molecules on T cells and CEACAM5 on tumor cells. Thus, such antibodies selectively enhance antitumor activity at tumor sites while minimizing peripheral toxicity. The target antibodies provided herein are particularly useful in combination with other anticancer therapies, including, for example, bispecific antibodies for the treatment of STEAP1-related cancer and CEACAM5-related cancer.

[0008] In a first aspect, the herein provides an anti-CD28 × anti-STEAP1 heterodimer antibody comprising a) a first monomer, b) a second monomer, and c) a light chain. The first monomer comprises i) a single-stranded variable fragment (scFv) and ii) a first Fc domain, the scFv being covalently bound to the N-terminus of the first Fc domain using a domain linker. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable heavy domain and CH2-CH3 is the second Fc domain. The light chain comprises VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). VH1 and VL1 together form a first antigen-binding domain (ABD), while VH2 and VL2 together form a second ABD. Furthermore, one of the first and second ABDs is a CD28-binding domain, and the other is a 6-transmembrane prostatic epithelial antigen 1 (STEAP1)-binding domain.

[0009] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0010] In exemplary embodiments, the first ABD is a STEAP1 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: a) VH and VL of either of the STEAP1 binding domains of Figure 22 or 39, or a variant thereof; or b) VH of Figure 31 and VL of Figure 32, or a variant thereof. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL of either of the CD28 binding domains of Figures 15, 18, and 21, or a variant thereof; or 2) (i) VH of either of the CD28 binding domains of Figure 15 or 16 and (ii) VL of either of the CD28 binding domains of Figures 15, 17, or 82, or a variant thereof.

[0011] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0012] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0013] In certain embodiments, the first and second Fc domains each comprise one or more attenuation variants. In an exemplary embodiment, the one or more attenuation variants comprise E233P / L234V / L235A / G236del / S267K, and the numbering follows the EU numbering.

[0014] In some embodiments, one of the first or second monomers further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises the pI variants N208D / Q295E / N384D / Q418E / N421D, and the numbering follows the EU numbering.

[0015] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises the amino acid variants E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain comprises the amino acid variants E233P / L234V / L235A / G236del / S267K / S364K / E357Q, and the numbering follows the EU numbering. In an exemplary embodiment, the first and second variant Fc domains each further comprise the amino acid variants 428L / 434S.

[0016] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0017] In another aspect, provided herein is an anti-CD28×anti-STEAP1 heterodimeric antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises, from the N-terminus to the C-terminus, VH1-CH1-the first domain linker-scFv-the second domain linker-CH2-CH3, wherein VH1 is the first variable heavy domain and CH2-CH3 is the first Fc domain. The second monomer comprises, from the N-terminus to the C-terminus, VH1-CH1-hinge-CH2-CH3, wherein CH2-CH3 is the second Fc domain. The first light chain comprises, from the N-terminus to the C-terminus, VL1-CL, wherein VL1 is the first variable light domain and CL is the constant light domain. The second light chain comprises, from the N-terminus to the C-terminus, VL1-CL, wherein VL1 is the first variable light domain and CL is the constant light domain. The scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and VH2 and VL2 form a second ABD. Further, the first ABD is a STEAP1-binding domain and the second ABD is a CD28-binding domain, or the first ABD is a CD28-binding domain and the second ABD is a STEAP1-binding domain.

[0018] In some embodiments, the scFv comprises, from the N-terminus to the C-terminus, VH2-scFv linker-VL2. In some embodiments, the scFv comprises, from the N-terminus to the C-terminus, VL2-scFv linker-VH2.

[0019] In exemplary embodiments, the first ABD is a STEAP1 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: a) VH and VL of either of the STEAP1 binding domains of Figure 22 or 39, or a variant thereof; or b) VH of Figure 31 and VL of Figure 32, or a variant thereof. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL of either of the CD28 binding domains of Figures 15, 18, and 21, or a variant thereof; or 2) (i) VH of either of the CD28 binding domains of Figure 15 or 16 and (ii) VL of either of the CD28 binding domains of Figures 15, 17, or 82, or a variant thereof.

[0020] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0021] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0022] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0023] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0024] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0025] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0026] In another embodiment, provided herein is an anti-CD28 × anti-STEAP1 heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-hinge-CH2-CH3-domain linker-scFv from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the second Fc domain. The first light chain comprises VL1-CL from N-terminus to C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The second light chain contains VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv contains a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, the first ABD is a STEAP1-binding domain and the second ABD is a CD28-binding domain, or the first ABD is a CD28-binding domain and the second ABD is a STEAP1-binding domain.

[0027] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0028] In exemplary embodiments, the first ABD is a STEAP1 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: a) VH and VL of either of the STEAP1 binding domains of Figure 22 or 39, or a variant thereof; or b) VH of Figure 31 and VL of Figure 32, or a variant thereof. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL of either of the CD28 binding domains of Figures 15, 18, and 21, or a variant thereof; or 2) (i) VH of either of the CD28 binding domains of Figure 15 or 16 and (ii) VL of either of the CD28 binding domains of Figures 15, 17, or 82, or a variant thereof.

[0029] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0030] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0031] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0032] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0033] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0034] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0035] In another embodiment, provided herein is an anti-CD28 × anti-STEAP1 heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where each VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain. The second monomer comprises scFv-domain linker-CH2-CH3 from N-terminus to C-terminus, where VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain. The first light chain comprises VL1-CL from N-terminus to C-terminus, where VL1 is a first variable light domain and CL is a constant light domain. The second light chain contains VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv contains a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), and VH2 and VL2 form a second ABD. Furthermore, the first ABD is a STEAP1 binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a STEAP1 binding domain. In some embodiments, scFv contains VH2-scFv linker-VL2 from the N-terminus to the C-terminus. In some embodiments, scFv includes a VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

[0036] In exemplary embodiments, the first ABD is a STEAP1 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: a) VH and VL of either of the STEAP1 binding domains of Figure 22 or 39, or a variant thereof; or b) VH of Figure 31 and VL of Figure 32, or a variant thereof. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL of either of the CD28 binding domains of Figures 15, 18, and 21, or a variant thereof; or 2) (i) VH of either of the CD28 binding domains of Figure 15 or 16 and (ii) VL of either of the CD28 binding domains of Figures 15, 17, or 82, or a variant thereof.

[0037] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0038] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0039] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0040] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0041] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0042] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0043] In another embodiment, provided herein is a bispecific antibody comprising a STEAP1-binding domain comprising a)i) a first variable heavy domain (VH1) and ii) a first variable light domain (VL1), and an anti-CD28-binding domain comprising b)i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2). In some embodiments, VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1-binding domain of Figure 22 or 39 or a variant thereof, or b) VH of Figure 31 and VL of Figure 32 or a variant thereof. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 15, 18, and 21; or 2) (i) VH or a variant thereof in Figure 15 or 16 and (ii) VL or a variant thereof in Figures 15, 17, or 82.

[0044] In some embodiments of the bispecific antibody, the first Fc domain and the second Fc domain are variant Fc domains, respectively.

[0045] In some embodiments of the bispecific antibody, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0046] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0047] In some embodiments, one of the first or second monomers further comprises one or more pI variants. In some embodiments, the pI variants are N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0048] Furthermore, this specification also provides nucleic acid compositions comprising nucleic acids encoding antibodies described herein, expression vector compositions comprising such nucleic acids, host cells for producing antibodies comprising the expression vector compositions, and methods for producing antibodies.

[0049] In another embodiment, the Specified herein provides a method for treating STEAP1-associated cancer in a patient requiring treatment for STEAP1-associated cancer, the method comprising administering to the patient an anti-CD28 × anti-STEAP1 bispecific antibody as described herein.

[0050] In another embodiment, the foregoing provides a method for treating STEAP1-related cancer in a patient requiring treatment for STEAP1-related cancer, the method comprising administering to the patient an anti-CD28 × anti-STEAP1 bispecific antibody and an anti-CD3 × anti-STEAP1 bispecific antibody as described herein.

[0051] In another embodiment, the composition provided herein comprises a STEAP1 antigen-binding domain (ABD). In some embodiments, STEAP1 comprises a) a variable heavy domain having vhCDR1-3 of either the STEAP1 binding domain variable heavy domain of Figure 22 or Figure 31, and b) a variable light domain having vlCDR1-3 of either the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

[0052] In some embodiments, the STEAP1 ABD includes a) a variable weight domain having at least 85% sequence identity to the STEAP1 binding domain variable weight domain of Figure 22 or Figure 31, and b) a variable light domain having at least 85% sequence identity to the STEAP1 binding domain variable light domain of Figure 22 or Figure 32. In some embodiments, the variable weight domain has at least 90% sequence identity to the STEAP1 binding domain variable weight domain of Figure 22 or Figure 31, and the variable light domain has at least 90% sequence identity to the STEAP1 binding domain variable light domain of Figure 22 or Figure 32. In some embodiments, the variable weight domain has at least 95% sequence identity to the STEAP1 binding domain variable weight domain of Figure 22 or Figure 31, and the variable light domain has at least 95% sequence identity to the STEAP1 binding domain variable light domain of Figure 22 or Figure 32. In some embodiments, the variable weight domain has at least 99% sequence identity to the STEAP1 binding domain variable weight domain of Figure 22 or Figure 31, and the variable light domain has at least 99% sequence identity to the STEAP1 binding domain variable light domain of Figure 22 or Figure 32. In some embodiments, the variable weight domain has the amino acid sequence of the STEAP1 binding domain variable weight domain of Figure 22 or Figure 31, and the variable light domain has the amino acid sequence of the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

[0053] In some embodiments, the STEAP1 ABD includes STEAP1 binding domain variable heavy domains and STEAP1 binding domain variable light domains selected from: H1 and L1, H1.1 and L1, H1.2 and L1, H1.3 and L1, H1.4 and L1, H1.5 and L1, H1.6 and L1, H1.7 and L1, H1.8 and L1, H1.9 and L1, H1.10 and L1, H1.11 and L1, H1.12 and L1, H1.13 and L1, H1.14 and L1, H1.15 and L1, H1.16 and L1, H1.17 and L1, H1.18 and L1, H1.19 and L 1, H1.20 and L1, H1.21 and L1, H1.22 and L1, H1.23 and L1, H1.24 and L1, H1.25 and L1, H1.26 and L1, H1.27 and L1, H1.28 and L1, H1.29 and L1, H1.30 and L1, H1.31 and L1, H1.32 and L1, H1.33 and L1, H1.34 and L1, H1.35 and L1, H1.36 and L1, H1.37 and L1, H1.38 and L1, H1 and L1.1, H1 and L1.2, H1 and L1.3, H0 and L0, H1.21 and L1.3, H1.34 and and L1.3, H1.39 and L1, H1.40 and L1, H1.41 and L1, H1.42 and L1, H1.43 and L1, H1.44 and L1, H1.45 and L1, H1.46 and L1, H1.47 and L1, H1.48 and L1, H1.49 and L1, H1.50 and L1, H1.51 and L1, H1.52 and L1, H1.53 and L1, H1.54 and L1, H1.55 and L1, H1.56 and L1, H1.57 and L1, H1.58 and L1, H1.59 and L1, H1.60 and L1, H1.61 and L1, H1.62 and L 1. H1.63 and L1, H1.64 and L1, H1.65 and L1, H1.66 and L1, H1.67 and L1, H1.68 and L1, H1.69 and L1, H1.70 and L1, H1.71 and L1, H1.72 and L1, H1.73 and L1, H1.74 and L1, H1.75 and L1, H1.76 and L1, H1.77 and L1, H1.78 and L1, H1.79 and L1, H1.80 and L1, H1.81 and L1, H1.82 and L1, H1.83 and L1, H1.84 and L1, H1.85 and L1, H1.86 and L1, H1.87 and L1, H1.88 and L1, H1.89 and L1, H1.90 and L1, H1.91 and L1, H1.92 and L1, H1.93 and L1, H1.94 and L1, H1.95 and L1, H1.96 and L1, H1.97 and L1, H1.98 and L1, H1.99 and L1, H1.100 and L1, H1.101 and L1, H1.102 and L1, H1.103 and L1, H1.104 and L1, H1.105 and L1, H1.106 and L1, H1.107 and L1, H1.108 and L1, H1.109 and L1, H1.110 and L1, H 1.111 and L1, H1.112 and L1, H1.113 and L1, H1.114 and L1, H1.115 and L1, H1.116 and L1, H1.117 and L1, H1.118 and L1, H1.119 and L1, H1.120 and L1, H1.121 and L1, H1.122 and L1, H1.123 and L1, H1.124 and L1, H1.125 and L1, H1.126 and L1, H1.127 and L1, H1.128 and L1, H1.129 and L1, H1.130 and L1, H1.131 and L1, H1.132 and L1, H1.133 and L 1. H1.134 and L1, H1.135 and L1, H1.136 and L1, H1.137 and L1, H1.138 and L1, H1.139 and L1, H1.140 and L1, H1.141 and L1, H1.142 and L1, H1.143 and L1, H1.144 and L1, H1.145 and L1, H1.146 and L1, H1.147 and L1, H1.148 and L1, H1.149 and L1, H1.150 and L1, H1.151 and L1, H1.152 and L1, H1.153 and L1, H1.154 and L1, H1.155 and L1, H1.156 and L1, H1.157 and L1, H1.158 and L1, H1.159 and L1, H1.160 and L1, H1.161 and L1, H1.162 and L1, H1.163 and L1, H1.164 and L1, H1.165 and L1, H1.166 and L1, H1.167 and L1, H1.168 and L1, H1.169 and L1, H1.170 and L1, H1.171 and L1, H1.172 and L1, H1.173 and L1, H1.174 and L1, H1.175 and L1, H1.176 and L1, H1.177 and L1, H1.178 and L1, H1.179 and L1, H1.180 and L1, H1.181 and L1, H1.182 and L1, H1.183 and L1, H1.184 and L1, H1.185 and L1, H1.186 and L1, H1.187 and L1, H1.188 and L1, H1.189 and L1, H1 and L1.4, H1 and L1.5, H1 and L1.6, H1 and L1.7, H1 and L1.8, H1 and L1.9, H1 and L1.10, H1 and L1.11, H1 and L1.12, H1 and L1.13, H1 and L1.14, H1 and L1.15, H1 and L1.16, H1 and L1.1 7, H1 and L1.18, H1 and L1.19, H1 and L1.20, H1 and L1.21, H1 and L1.22, H1 and L1.23, H1 and L1.24, H1 and L1.25, H1 and L1.26, H1 and L1.27, H1 and L1.28, H1 and L1.29, H1 and L1.30, H1 and L1.31, H1 and L1.32, H1 and L1.33, H1 and L1.34, H1 and L1.35, H1 and L1.36, H1 and L1.37, H1 and L1.38, H1 and L1.39, H1 and L1.40, H1 and L1.41, H1 and L1.4 2. H1 and L1.43, H1 and L1.44, H1 and L1.45, H1 and L1.46, H1 and L1.47, H1 and L1.48, H1 and L1.49, H1 and L1.50, H1 and L1.51, H1 and L1.52, H1 and L1.53, H1 and L1.54, H1 and L1.55, H1 and L1.56, H1 and L1.57, H1 and L1.58, H1 and L1.59, H1 and L1.60, H1 and L1.61, H1 and L1.62, H1 and L1.63, H1 and L1.64, H1 and L1.65, H1 and L1.66, H1 and L1.6 7, H1 and L1.68, H1 and L1.69, H1 and L1.70, H1 and L1.71, H1 and L1.72, H1 and L1.73, H1 and L1.74, H1 and L1.75, H1 and L1.76, H1 and L1.77, H1 and L1.78, H1 and L1.79, H1 and L1.80, H1 and L1.81, H1 and L1.82, H1 and L1.83, H1 and L1.84, H1 and L1.85, H1.42 and L1.47, H1.42 and L1.56, H1.190 and L1.3, H1.191 and L1.3, H1.192 and L1.3, H1.195 and L1.3, H1.196 and L1.3, H1.197 and L1.3, H1.198 and L1.3, H1.199 and L1.3, H1.200 and L1.3, H1.201 and L1.3, H1.202 and L1.3, H1.203 and L1.3, H1.204 and L1.3, H1.205 and L1.3, H1.206 and L1.3, H1.207 and L1.3, H1.208 and L1.3, H1.209 and L1.3, H1.210 and L1.3, H1.211 and L1.3, H1.212 and L1.3, H1.213 and L1.3, H1.214 and L1.3, H1.215 and L1.3, H1.216 and L1.3, H1.217 and L1.3, H1.218 and L1.3, H1.219 and L1.3, H1.220 and L1.3, H1.221 and L1.3, H1.222 and L1.3, H1.223 and L1.3, H1.224 and L1.3, H1.225 and L1.3, H1.226 and L1.3, H1.227 and L1.3, H1.228 and L1.3, H1.21 and L1.92, H1.21 and L1.93, H1.21 and L1.94, H1.21 and L1.95, H1.21 and L1 .96, H1.21 and L1.97, H1.21 and L1.98, H1.21 and L1.99, H1.21 and L1.100, H1.21 and L1.101, H1.21 and L1.102, H1.21 and L1.103, H1.21 and L1.104, H1.21 and L1.105, H1.21 and L1.106, H1.21 and L1.107, H1.21 and L1.108, H1.21 and L1.109, H1.21 and L1.110, H1.21 and L1.111, H1.21 and L1.112, H1.21 and L1.113, H1.21 and L1.1 14, H1.21 and L1.115, H1.21 and L1.116, H1.21 and L1.117, H1.21 and L1.118, H1.21 and L1.119, H1.195 and L1.92, H1.195 and L1.94, H1.195 and L1.95, H1.195 and L1.104, H1.195 and L1.105, H1.195 and L1.109, H1.196 and L1.92, H1.196 and L1.94, H1.196 and L1.95, H1.196 and L1.104, H1.196 and L1.105, H1.196 and L1.109, H1.205 and L1.92, H1.205 and L1.94, H1.205 and L1.95, H1.205 and L1.104, H1.205 and L1.105, H1.205 and L1.109, H1.193 and L1.91, H1.194 and L1.86, H1.194 and L1.87, H1.194 and L1.88, H1.194 and L1.89, or H1.194 and L1.90, or H1.2320 and L1.100, or H1.230 and L1.115 (Figures 22, 31, 32, and 39).

[0054] In one embodiment, the herein provides an anti-CD28 × anti-CEACAM5 heterodimer antibody comprising a) a first monomer, b) a second monomer, and c) a light chain. The first monomer comprises i) a single-stranded variable fragment (scFv) and ii) a first Fc domain, the scFv being covalently bound to the N-terminus of the first Fc domain using a domain linker. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable heavy domain and CH2-CH3 is the second Fc domain. The light chain comprises VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). VH1 and VL1 together form a first antigen-binding domain (ABD), and VH2 and VL2 together form a second ABD. Furthermore, one of the first and second ABDs is a CD28-binding domain, and the other is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5)-binding domain.

[0055] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0056] In exemplary embodiments, the first ABD is a CEACAM5 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from the VH and VL or variants thereof of any of the CEACAM5 binding domains shown in Figures 41 and 45 below. In some embodiments, VH2 and VL2 are selected from one of the following: 1) the VH and VL or variants thereof of any of the CD28 binding domains shown in Figures 15, 18, 21 and 65, or 2) (i) the VH or variant thereof of Figure 15 or 16 and (ii) the VL or variant thereof of Figures 15, 17 or 82.

[0057] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0058] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0059] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0060] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0061] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0062] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0063] In another embodiment, provided herein is an anti-CD28 × anti-CEACAM5 heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-first domain linker-scFv-second domain linker-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable heavy domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where CH2-CH3 is the second Fc domain. The first light chain comprises VL1-CL from N-terminus to C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The second light chain contains VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv contains a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

[0064] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0065] In exemplary embodiments, the first ABD is a CEACAM5 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from: 1) VH and VL or variants thereof of any of the CEACAM5 binding domains in Figures 41 and 45. In some embodiments, VH2 and VL2 are selected from: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 15, 18, 21 and 65, or 2) (i) VH or a variant thereof in Figure 15 or 16 and (ii) VL or a variant thereof in Figures 15, 17 or 82.

[0066] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0067] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0068] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0069] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0070] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0071] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0072] In another embodiment, provided herein is an anti-CD28 × anti-CEACAM5 heterodimer antibody comprising a) a first monomer, b) a second monomer, and c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-hinge-CH2-CH3-domain linker-scFv from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the second Fc domain. The first light chain comprises VL1-CL from N-terminus to C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The second light chain contains VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv contains a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

[0073] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0074] In exemplary embodiments, the first ABD is a CEACAM5 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from: 1) VH and VL or variants thereof of any of the CEACAM5 binding domains in Figures 41 and 45. In some embodiments, VH2 and VL2 are selected from: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 15, 18, 21 and 65, or 2) (i) VH or a variant thereof in Figure 15 or 16 and (ii) VL or a variant thereof in Figures 15, 17 or 82.

[0075] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0076] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0077] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0078] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0079] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0080] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0081] In another embodiment, provided herein is a heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where each VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain. The second monomer comprises scFv-domain linker-CH2-CH3 from N-terminus to C-terminus, where VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain. The first common light chain comprises VL1-CL from N-terminus to C-terminus, where VL1 is a first variable light domain and CL is a constant light domain. The second light chain contains VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv contains a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

[0082] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0083] In exemplary embodiments, the first ABD is a CEACAM5 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from: 1) VH and VL or variants thereof of any of the CEACAM5 binding domains in Figures 41 and 45. In some embodiments, VH2 and VL2 are selected from: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 15, 18, 21 and 65, or 2) (i) VH or a variant thereof in Figure 15 or 16 and (ii) VL or a variant thereof in Figures 15, 17 or 82.

[0084] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0085] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0086] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0087] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0088] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0089] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS.

[0090] In another embodiment, provided herein is a bispecific antibody comprising a CEACAM5-binding domain comprising a)i) a first variable heavy domain (VH1) and ii) a first variable light domain (VL1), and an anti-CD28-binding domain comprising b)i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2). In some embodiments, VH1 and VL1 are selected from: 1) VH and VL or variants thereof of any of the CEACAM5-binding domains in Figures 41 and 45. In some embodiments, VH2 and VL2 are selected from: 1) VH and VL or variants thereof of any of the CD28-binding domains in Figures 15, 18, 21, and 65, or 2) (i) VH or a variant thereof in Figure 15 or 16 and (ii) VL or a variant thereof in Figures 15, 17, or 82.

[0091] In some embodiments of the bispecific antibody, the first Fc domain and the second Fc domain are variant Fc domains, respectively.

[0092] In some embodiments of the bispecific antibody, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0093] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0094] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0095] Furthermore, this specification also provides nucleic acid compositions comprising nucleic acids encoding antibodies described herein, expression vector compositions comprising such nucleic acids, host cells for producing antibodies comprising the expression vector compositions, and methods for producing antibodies.

[0096] In another embodiment, provided herein is a method for treating CEACAM5 cancer in a patient requiring treatment for CEACAM5 cancer, the method comprising administering to the patient an anti-CD28 × anti-CEACAM5 bispecific antibody as described herein.

[0097] In another embodiment, provided herein is a method for treating CEACAM5-related cancer in a patient requiring treatment for CEACAM5-related cancer, the method comprising administering to the patient an anti-CD28 × anti-CEACAM5 bispecific antibody and an anti-CD3 × anti-CEACAM5 bispecific antibody as described herein.

[0098] In another embodiment, provided herein is a CEACAM5 antigen-binding domain composition comprising a) a variable heavy domain and b) a variable light domain. The variable heavy domain comprises vhCDR1-3 of any of the CEACAM5-binding domain variable heavy domains shown in Figure 45. The variable light domain comprises vlCDR1-3 of any of the CEACAM5-binding domain variable light domains shown in Figure 45. In another embodiment, provided herein is a CEACAM5 antigen-binding domain composition comprising a) a variable heavy domain and b) a variable light domain. The variable heavy domain comprises at least 85% sequence identity to the CEACAM5-binding domain variable heavy domain shown in Figure 45. The variable light domain comprises at least 85% sequence identity to the CEACAM5-binding domain variable light domain shown in Figure 45.

[0099] In some embodiments, the variable heavy domain has at least 90% sequence identity with the CEACAM5 binding domain variable heavy domain in Figure 45, and the variable light domain has at least 90% sequence identity with the CEACAM5 binding domain variable light domain in Figure 45.

[0100] In some embodiments, the variable heavy domain has at least 95% sequence identity with the CEACAM5 binding domain variable heavy domain in Figure 45, and the variable light domain has at least 95% sequence identity with the CEACAM5 binding domain variable light domain in Figure 45.

[0101] In some embodiments, the variable heavy domain has at least 99% sequence identity with the CEACAM5 binding domain variable heavy domain in Figure 45, and the variable light domain has at least 99% sequence identity with the CEACAM5 binding domain variable light domain in Figure 45.

[0102] In some embodiments, the variable heavy domain has the amino acid sequence of the CEACAM5 binding domain variable heavy domain shown in Figure 45, and the variable light domain has the amino acid sequence of the CEACAM5 binding domain variable light domain shown in Figure 45.

[0103] In another embodiment, provided herein are CD28 antigen-binding domain compositions comprising a) a variable heavy domain and b) a variable light domain. The variable heavy domain comprises vhCDR1-3 of any of the CD28-binding domain variable heavy domains shown in Figure 65. The variable light domain comprises vlCDR1-3 of any of the CD28-binding domain variable light domains shown in Figure 65.

[0104] In another embodiment, provided herein is a CD28 antigen-binding domain composition comprising a) a variable heavy domain and b) a variable light domain. The variable heavy domain contains at least 85% sequence identity to the CD28-binding domain variable heavy domain of Figure 65. The variable light domain contains at least 85% sequence identity to the CD28-binding domain variable light domain of Figure 65.

[0105] In some embodiments, the variable weight domain has at least 90% sequence identity with the CD28-binding domain variable weight domain in Figure 65, and the variable light domain has at least 90% sequence identity with the CD28-binding domain variable light domain in Figure 65.

[0106] In some embodiments, the variable heavy domain has at least 95% sequence identity with the CD28-binding domain variable heavy domain in Figure 65, and the variable light domain has at least 95% sequence identity with the CD28-binding domain variable light domain in Figure 65.

[0107] In some embodiments, the variable weight domain has at least 99% sequence identity with the CD28-binding domain variable weight domain in Figure 65, and the variable light domain has at least 99% sequence identity with the CD28-binding domain variable light domain in Figure 65.

[0108] In some embodiments, the variable heavy domain has the amino acid sequence of the CD28-binding domain variable heavy domain shown in Figure 65, and the variable light domain has the amino acid sequence of the CD28-binding domain variable light domain shown in Figure 65.

[0109] In some embodiments, the CD28 binding domain is selected from the following: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, and CD28.3 in Figure 65. [CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD28]_H1L 1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD 28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, T N228[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.7 1[SS], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[S S], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS]. [Brief explanation of the drawing]

[0110] [Figure 1A] The sequences of human and mouse CD28 are shown. [Figure 1B] The sequence of CD28 from a cynomolgus monkey is shown. Such CD28 is useful for developing cross-reactive CD28 antigen-binding domains to facilitate clinical development. [Figure 2A] The sequences of human, mouse, and cynomolgus monkey STEAP1 are shown. Such STEAP1 is useful for developing cross-reactive STEAP1 antigen-binding domains to facilitate clinical development. [Figure 2B]The sequences of human, mouse, and cynomolgus monkey CEACAM5 are shown. Such CEACAM5 sequences are useful for developing cross-reactive CEACAM5 antigen-binding domains to facilitate clinical development. [Figure 3A] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3B] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3C] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3D] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3E] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3F] Useful pairs of heterodimerization variant sets (including skew and pI variants) are shown. Figure 3F shows variants in which there is no variant corresponding to "monomer 2". Such variants are pI variants and can be used alone for either monomer of a bispecific antibody (e.g., STEAP1×CD28 bsAb or CEACAM5×CD28 bsAb), or may be included in the non-scFv side in a form that utilizes scFv as a component, for example, and a suitable charged scFv linker can be used for the second monomer that utilizes scFv as the CD28 binding domain. Preferred charged linkers are shown in Figure 6. The heterodimer yield (%) and CH3 Tm (°C) of preferred Fc heterodimerization variants have been previously described (see, for example, Figure 8 of U.S. Patent Application No. 2019 / 0248898). [Figure 4]A list of isosteric variant antibody constant regions and their respective substitutions is shown. pI_(-) indicates a lower pI variant, while pI_(+) indicates a higher pI variant. These variants can be combined arbitrarily and independently with other variants, including the heterodimerized variants outlined herein. [Figure 5] The document exhibits useful attenuation variants (also referred to as "knockout" or "KO" variants) that reduce FcγR binding. In some embodiments, such attenuation variants are present in the Fc domains of both monomers of the target antibody described herein. In other embodiments, the attenuation variant is present in only one variant's Fc domain. [Figure 6] As described herein, numerous charged scFv linkers are used to increase or decrease the pI of heterodimer bispecific antibodies (e.g., STEAP1×CD28 and CEACAM5×CD28 bsAb) that utilize one or more scFv components. (+H) positive linkers are particularly used herein, especially in conjunction with the anti-CD28 VL and VH sequences shown herein. A single-charged, single-prior-art scFv linker is referred to as "Whitlow" from Whitlow et al., Protein Engineering 6(8):989-995 (1993). It should be noted that this linker was used to reduce aggregation and improve proteolytic stability in scFv. Such charged scFv linkers may be used in any of the target antibody forms disclosed herein that include scFv (e.g., 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc forms). [Figure 7] Numerous exemplary domain linkers are shown. In some embodiments, these linkers are used to link a single-stranded Fv to an Fc chain. In some embodiments, these linkers can be combined in any orientation. For example, the GGGGS linker can be combined with a “bottom half-hinge” linker at the N-terminus or C-terminus. [Figure 8] This invention presents a bispecific antibody platform particularly useful for STEAP1×CD28 and CEACAM5×CD28 bsAb bsAb. While the platform is presented in relation to the 1+1 Fab-scFv-Fc format, it can also be applied to the use of other bispecific antibody formats. [Figure 9A] Based on human IgG1, the sequences of several useful heterodimer TEAP1×CD28 bsAb and CEACAM5×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 9B] Based on human IgG1, the sequences of several useful heterodimer TEAP1×CD28 bsAb and CEACAM5×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 9C] Based on human IgG1, the sequences of several useful heterodimer TEAP1×CD28 bsAb and CEACAM5×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 9D]Based on human IgG1, the sequences of several useful heterodimer TEAP1×CD28 bsAb and CEACAM5×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. Heterodimer Fc skeleton 1, based on human IgG1 (356E / 358M allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the L368E / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the K360E / Q362E / T411E scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the D401K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains.The heterodimer Fc skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the N297A variant that removes glycosylation on both chains. The heterodimer Fc skeleton 7 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K decrement variant and the N297S variant that removes glycosylation on both chains. The heterodimer Fc skeleton 8 is based on human IgG4 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variants on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the S228P (according to EU numbering, S241P in Kabat) variant which reduces Fab arm exchange (as known in the art) on both chains. The heterodimer Fc skeleton 9 is based on human IgG2 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variants on the first heterodimer Fc chain, and the S364K / E357Q scuba riant on the second heterodimer Fc chain.The heterodimer Fc skeleton 10 is based on human IgG2 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the S267K attenuation variant on both chains. The heterodimer Fc skeleton 11 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434S Xtend variant on both chains. The heterodimer Fc skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba rianto on the first heterodimer Fc chain, the S364K / E357Q scuba rianto and P217R / P229R / N276K pI variant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 13, based on human IgG1 (356D / 358L allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and M428L / N434S Xtend variant on both chains. The heterodimer Fc skeleton 14, based on human IgG1 (356E / 358M allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434A Xtend variant on both chains.The heterodimer Fc skeleton 15 is based on human IgG1 (356D / 358L allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434A Xtend variant on both chains. Sequences that are 90, 95, 98, and 99% identical (as defined herein) to the described sequences, and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more additional amino acid substitutions (as understood by those skilled in the art, compared to parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), compared to the “parent” in the figure which already contains numerous amino acid modifications) are included in each of these skeletons. That is, the described skeletons may contain additional amino acid modifications (usually amino acid substitutions) in addition to or as alternatives to the skew, pI, and attenuation variants contained within the skeleton in this figure. Furthermore, the skeletons shown herein may include deletions of C-terminal glycine (K446_) and / or lysine (K447_). C-terminal glycine and / or lysine deletions may be intentionally manipulated to reduce heterogeneity or in relation to a predetermined bispecific form such as the mAb-scFv form. Furthermore, C-terminal glycine and / or lysine deletions may occur naturally, for example, during production and storage. [Figure 10] Exemplary sequences of the heterodimer TEAP1×CD28 and CEACAM5×CD28 bsAb skeletons for use in the 2+1 mAb-scFv configuration are shown. The configurations shown here are based on the heterodimer Fc skeleton 1 shown in Figure 9, but further including G446_ on monomer 1(-) and G446_ / K447_ on monomer 2(+). It should be noted that any of the additional skeletons shown in Figure 9 can be applied to use in the 2+1 mAb-scFv configuration, whether or not they include K447_ on one or both strands. It should be noted that these sequences may further include the M428L / N434S variant. [Figure 11] The sequence of "CH1" used in the embodiment of STEAP1×CD28 bsAb is shown. [Figure 12] The arrangement of "hinge" used in the embodiments of STEAP1×CD28 and CEACAM5×CD28 bsAb is shown. [Figure 13] This shows the constant domains of the congeneral light chains used in STEAP1×CD28 and CEACAM5×CD28 bsAb, which utilize the Fab-binding domain. [Figure 14A] The present invention presents a bispecific form. It presents a "1+1 Fab-scFv-Fc" form having a first Fab arm that binds to a first antigen and a second scFv arm that binds to a second antigen. The 1+1 Fab-scFv-Fc form comprises a first monomer containing a first heavy chain variable region (VH1) covalently bonded (optionally via a linker) to the N-terminus of a first heterodimer Fc skeleton, a second monomer containing a single-stranded Fv covalently bonded (optionally via a linker) to the N-terminus of a second corresponding heterodimer Fc skeleton, and a third monomer containing a light chain variable region covalently bonded to a light chain constant domain (the light chain variable region is complementary to VH1). [Figure 14B] The present invention presents a bispecific form. It presents a "2+1 Fab2-scFv-Fc" form having a first Fab arm and a second Fab-scFv arm, where Fab binds to the first antigen and scFv binds to the second antigen. The 2+1 Fab2-scFv-Fc form comprises a first monomer containing a first heavy chain variable region (VH1) covalently bonded (optionally via a linker) to the N-terminus of a first heterodimer Fc skeleton, a second monomer containing VH1 covalently bonded (optionally via a linker) to a single-stranded Fv covalently bonded (optionally via a linker) to the N-terminus of a second corresponding heterodimer Fc skeleton, and a third monomer containing a light chain variable region covalently bonded to a light chain constant domain (the light chain variable region is complementary to VH1). [Figure 14C]The present invention presents a bispecific form. It presents a "1+1 common light chain" or "1+1 CLC" form having a first Fc containing a first Fab arm that binds to a first antigen and a second Fc containing a second Fab arm that binds to a second antigen. The 1+1 CLC form comprises a first monomer containing VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. [Figure 14D] The present invention presents a bispecific form. It presents a "2+1 common light chain" or "2+1 CLC" form having a first Fc containing two Fab arms that bind to a first antigen and a second Fc containing one Fab arm that binds to a second antigen. The 2+1 CLC form comprises a first monomer containing VH1-CH1-hinge-VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with the first and second VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. [Figure 14E] The present invention presents a bispecific form. It presents a "2+1 mAb-scFv" form having a first Fc containing an N-terminal Fab arm that binds to a first antigen, and a second Fc containing an N-terminal Fab arm that binds to a first antigen and a C-terminal scFv that binds to a second antigen. The 2+1 mAb-scFv form comprises a first monomer containing VH1-CH1-hinge-CH2-CH3, a second monomer containing VH1-CH1-hinge-CH2-CH3-scFv, and a third monomer containing VL-CL. VL pairs with the first and second VH1 to form a binding domain having binding specificity for the first antigen. [Figure 14F]The present invention presents a bispecific form. It presents a "2+1 Fab2-Fc×scFv-Fc" form having a first Fc domain containing two N-terminal Fab arms that bind to a first antigen and a second Fc domain containing an N-terminal scFv that binds to a second antigen. The 2+1 Fab2-Fc×scFv-Fc form comprises a first monomer containing VH1-CH1-VH2-CH1-hinge-CH2-CH3, a second monomer containing scFv-domain linker-CH2-CH3, and third and fourth monomers containing VL-CL. VL pairs with VH1 and VH2 to form a binding domain having binding specificity to the first antigen. [Figure 14G] This shows the dual-specific form of the present invention. It is a dual scFv, an additional dual-specific form. [Figure 14H] The present invention presents a dual-specific form. An additional dual-specific form is a 1-arm scFv-mAb. [Figure 14I] The present invention presents a bispecific form. An additional bispecific form is scFv-mAb. [Figure 14J] This presents a bispecific form of the present invention. It is a bispecific mAb of an additional bispecific form. [Figure 14K] The present invention presents a dual-specific form. An additional dual-specific form is a 1-arm central-scFv. [Figure 14L] The present invention presents a bispecific form. An additional bispecific form is mAb-Fv. [Figure 14M] The present invention presents a bispecific form. The additional bispecific form is the central-Fv. [Figure 14N] This presents a dual-specific form of the present invention. It is a trident of additional dual-specific forms. [Figure 15]The variable heavy chain and variable light chain sequences of 1A7, an exemplary phage-derived CD28-binding domain, are shown, as well as the sequence of XENP28428, an anti-CD28 mAb based on the IgG1 scaffold with 1A7 and the E233P / L234V / L235A / G236del / S267K attenuation variant. CDRs are indicated by underlines, and slashes indicate the boundary(s) between the variable region and the constant domain. As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 16A] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 16B] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 16C] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 16D] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 16E] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 16F] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. It should be noted that the variable weight domain can pair with any of the other variable light domains shown in Figures 15, 17, and 18, including sequence numbers XXX-YYY. [Figure 17A] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17B] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17C]The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17D] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17E] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17F] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17G] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17H] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 17I] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. It should be noted that the variable heavy domain can pair with any of the other variable light domains shown in Figures 15, 16, and 18, including sequence numbers XXX-YYY. [Figure 18A] The sequence of an exemplary affinity-optimized 1A7VH / VL pair is shown. [Figure 18B] The sequence of an exemplary affinity-optimized 1A7VH / VL pair is shown. [Figure 18C] Exemplary affinity-optimized 1A7VH / VL pair sequences are shown. It should be noted that these pairs can be formalized as Fab or scFv. Furthermore, in scFv form, these pairs can be formalized as VH VL orientation or VLVH orientation. [Figure 19A] The consensus framework region (FR) and complementarity determination region (CDR) (in Kabat) of the variable weight and variable light domain variants of anti-CD28 clone 1A7 are shown. [Figure 19B] The consensus framework region (FR) and complementarity determination region (CDR) (in Kabat) of the variable weight and variable light domain variants of anti-CD28 clone 1A7 are shown. [Figure 20]This shows the binding affinity in relation to the exemplary affinity operation 1A7VH / VL pair and scFv (in relation to the 1+1 Fab-scFv-FcbsAb form). [Figure 21A] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21B] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21C] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21D] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21E] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21F] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21G] The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. [Figure 21H]The variable heavy chain and variable light chain sequences of the additional CD28-binding domains used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 22] The variable heavy chain and variable light chain sequences of the additional STEAP1 binding domain used in STEAP1×CD28 and CEACAM5×CD28 bsAb of the present invention are shown. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 23] Figures A and B demonstrate, A) classical T cell / APC interaction, and B) reproduction of classical T cell / APC interaction by combining CD3 bispecific antibodies and CD28 bispecific antibodies. In classical T cell / APC interaction, there is a first signal (signal 1) provided by TCR reactivity with peptide-MHC and a second signal (signal 2) provided by crosslinking of CD28 by CD80 / CD86 expressed on APCs, which together fully activate T cells. In contrast, treatment with CD3 bispecificity provides only the first signal. The CD28 signal may be provided by CD28 bispecificity with the intention of promoting activation and proliferation via CD28 costimulation. In some embodiments, TAA1 and TAA2 may be different antigens. In some embodiments, TAA1 and TAA2 may be the same antigen but may have different epitopes. In some embodiments, TAA1 and TAA2 may be the same antigen and have the same epitope. [Figure 24A] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24B] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24C] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24D] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24E] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24F] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24G] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24H] The sequence of STEAP1×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 24I]The sequence of an exemplary STEAP1×CD28 bsAb in the 1+1 Fab-scFv-Fc format is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that STEAP1×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 25A] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 25B] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 25C] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 25D] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 25E] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 25F]The following is a sequence of an exemplary anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. The CDR is indicated by an underline, the scFv linker by a double underline (in the sequence, the scFv linker is the positively charged scFv(GKPGS)4 linker (SEQ ID NO: XXX), but as will be understood to those skilled in the art, this linker can be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 6), and a slash indicates the boundary(s) of a variable domain. In addition, the nomenclature indicates the orientation of the scFv from the N-terminus to the C-terminus. As is true for all sequences described herein that contain a CDR, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2; therefore, this specification includes not only underlined CDRs but also CDRs contained within the VH and VL domains using other numbering systems. Furthermore, for all sequences shown in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 26] This study demonstrates the induction of IL-2 secretion by combining STEAP1×CD28 bsAb XENP44562 with a constant dose of 1 μg / ml of exemplary EpCAM×CD3 bsAb. Purified T cells were incubated with 22Rv1-STEAP1 tumor cells (1:10 effector:target (E:T) ratio) and the test substance for 24 hours, and the cell supernatant was assayed for IL-2 by MSD. [Figure 27A] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27B] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27C] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27D] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27E]An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27F] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27G] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27H] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27I] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27J] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27K] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27L] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27M] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27N] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 27O] An example of STEAP1×CD28 bsAb in the 2+1 mAb-scFv-Fc format is shown. [Figure 28] An example of the 2+1 Fab2-Fc×scFv-Fc format, STEAP1×CD28, is shown. [Figure 29] XENP46214, a STEAP1×CD28 bsAb in the 1+1 Fab-scFv-Fc form, exhibits the ability to conditionally enhance T cell activation in the presence of the STEAP1 antigen when used in combination with an exemplary B7H3×CD3 engager. [Figure 30A] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30B] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30C] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30D] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30E] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30F] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30G] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30H] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30I] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30J] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30K] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30L] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30M] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30N] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30O] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30P] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30Q]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30R] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30S] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30T] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30U] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30V] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30W] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30X] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30Y] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30Z] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AA] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AB] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AC] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AD] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AE] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AF]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AG] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AH] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AI] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AJ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AK] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AL] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AM] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AN] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AO] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AP] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AQ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AR] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AS] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AT] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AU]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AV] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AW] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AX] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AY] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30AZ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BA] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BB] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BC] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BD] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BE] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BF] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BG] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BH] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BI] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BJ]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BK] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BL] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BM] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BN] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BO] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BP] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BQ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BR] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BS] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BT] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BU] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BV] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BW] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BX] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BY]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30BZ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CA] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CB] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CC] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CD] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CE] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CF] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CG] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CH] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CI] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CJ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CK] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CL] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CM] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CN]The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CO] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CP] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CQ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CR] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CS] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CT] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CU] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CV] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CW] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CX] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CY] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. [Figure 30CZ] The amino acid sequence of the divalent mAb form STEAP-A[STEAP1] variant is shown. CDRs are underlined, and slashes indicate the boundary(s) of the variable region. [Figure 31A] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31B] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31C] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31D] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31E] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31F] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31G] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31H] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31I] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31J] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31K] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31L] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31M] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31N] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31O] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31P] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31Q] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31R]This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31S] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31T] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 31U] This shows an operational variant of the variable weight domain of STEAP-A[STEAP-1]Fv. [Figure 32A] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32B] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32C] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32D] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32E] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32F] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32G] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32H] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32I] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 32J] This shows an operational variant of the variable light domain of STEAP-A[STEAP-1]Fv. [Figure 33]The results of experiments treating co-cultures of T cells and 22Rv1-STEAP1 cancer cells (1:1 E:T ratio) with 0.5 μg / mL exemplary B7H3×CD3 bsAb alone or in combination with 1 μg / mL STEAP1×CD28 bsAb are shown. Cell killing was tracked and analyzed over 6 days via Incucyte. As shown, STEAP1×CD28 bsAb significantly enhanced target cell killing when combined with B7H3×CD3. [Figure 34A] The C4-2B cell binding data obtained from the second round of affinity manipulation is shown, sorted by maximum MFI. [Figure 34B] The C4-2B cell binding data obtained from the second round of affinity manipulation is shown, sorted by maximum MFI. [Figure 34C] The C4-2B cell binding data obtained from the second round of affinity manipulation is shown, sorted by maximum MFI. [Figure 34D] The C4-2B cell binding data obtained from the second round of affinity manipulation is shown, sorted by maximum MFI. [Figure 35A] The results of experiments in which co-cultures of T cells and 22Rv1 cells (1:1 E:T ratio) were treated with 1 μg / mL of STEAP1×CD28 bsAb or exemplary B7H3×CD28 bsAb in combination with A) 10 μg / mL of exemplary PSMA×CD3 bsAb or B) 10 μg / mL of exemplary STEAP1×CD3 bsAb are shown. [Figure 35B] The results of experiments in which co-cultures of T cells and 22Rv1 cells (1:1 E:T ratio) were treated with 1 μg / mL of STEAP1×CD28 bsAb or exemplary B7H3×CD28 bsAb in combination with A) 10 μg / mL of exemplary PSMA×CD3 bsAb or B) 10 μg / mL of exemplary STEAP1×CD3 bsAb are shown. Cell killing was tracked and analyzed over time via Incucyte. As shown, STEAP1×CD28 bsAb induced significantly superior tumor cell killing compared to exemplary B7H3×CD28 bsAb. [Figure 36]The 2+1 mAb-scFv configuration shows strong binding of STEAP1×CD28 bsAb and more potent induction of IL-2. Furthermore, the 1+1 Fab-scFv-Fc configuration shows higher antigen-binding ability of STEAP1×CD28 bsAb. [Figure 37] This demonstrates the induction of IL2 secretion from 22RV1-STEAP1 tumor cells (1:1 effector:target ratio) and purified T cells incubated with 1 ug / ml of specified exemplary CD3 bispecific and escalating doses of STEAP1×CD28 bsAb. [Figure 38] T cell proliferation, as indicated by the number of CD4+ T cells, is shown after incubation of purified T cells with 22Rv1 tumor cells (1:1 effector:target ratio) and exemplary A) B7H3×CD3 bsAb and B) PSMA×CD3 bsAb at 1 μg / ml, as well as specified STEAP1×CD28 bsAb at escalating doses. [Figure 39] The variable weight, variable light weight, and CDR sequences of the STEAP1 binding domains [STEAP-A]_H1.230_L1.100 and [STEAP-A]_H1.230_L1.115 are shown. [Figure 40] An example of the 2+1 central-scFv format STEAP1×CD28 bsAb is shown. [Figure 41A] The following shows exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain used in the CEACAM5×CD28 bsAb of the present invention. [Figure 41B] The following shows exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain used in the CEACAM5×CD28 bsAb of the present invention. [Figure 41C] The following shows exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain used in the CEACAM5×CD28 bsAb of the present invention. [Figure 41D] The following shows exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain used in the CEACAM5×CD28 bsAb of the present invention. [Figure 41E]Exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain utilized in the CEACAM5×CD28 bsAb of the present invention are shown. [Figure 41F] Exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain utilized in the CEACAM5×CD28 bsAb of the present invention are shown. [Figure 41G] Exemplary variable heavy chain and variable light chain sequences of the CEACAM5 binding domain utilized in the CEACAM5×CD28 bsAb of the present invention are shown. As applicable to all sequences herein described and containing CDRs, the identification of CDR positions can be determined according to the numbering scheme shown in Table 2. Further, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 42A] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42B] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42C] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42D] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42E] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42F] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42G] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42H] Sequences of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format are shown. [Figure 42I]Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42J] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42K] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42L] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42M] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42N] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42O] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42P] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42Q] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [[ID=二十五]] [Figure 42R] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42S] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42T] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42U] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42V] Shows the sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format. [Figure 42W] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42X] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42Y] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42Z] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42AA] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42BB] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42CC] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42DD] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42EE] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42FF] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42GG] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42HH] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42II] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42JJ] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42KK] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42LL] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42MM] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. <{ [Figure 42NN] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 4200] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42PP] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42QQ] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42RR] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42SS] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42TT] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42UU] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42VV] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42WW] The sequence of an exemplary CEACAM5×CD28 bsAb in 1+1 Fab-scFv-Fc format is shown. [Figure 42XX]The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42YY] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42ZZ] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42AAA] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42BBB] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42CCC] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42DDD] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42EEE] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42FFF] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42GGG] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42HHH] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42III] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42JJJ] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42KKK] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42LLL] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42MM] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42 NNN] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 4200] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42PPP] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42QQQ] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42RRR] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42SSS] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42TTT] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42UUU] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42VVV] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42WWW] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42XXX] The sequence of CEACAM5×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 42YYY]An exemplary sequence of CEACAM5×CD28 bsAb in the 1+1 Fab-scFv-Fc format is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that CEACAM5×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 43] This shows the induction of IL2 secretion from purified T cells incubated with A)MKN45, B)HPAF-II, C)LS1034, D)LS174T, E)Lovo, or F)HT-29 tumor cells (specified effector:target (E:T) ratio) and 1 μg / ml of specified CEACAM5×CD28 bsAb and exemplary B7H3×CD3 bsAb in escalating doses. [Figure 44] This shows the induction of IL2 secretion from purified T cells incubated with A) LS1034, B) HPAF-II, C) Lovo, D) LS174T, E) SNU-C1, or F) HT-29 tumor cells, as well as exemplary B7H3×CD3 bsAb at 1 μg / ml and specified CEACAM5×CD28 bsAb at escalating doses. [Figure 45A] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45B]This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45C] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45D] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45E] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45F] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45G] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45H] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45I] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45J] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45K] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45L] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45M] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45N] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45O] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45P] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45Q]This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 45R] This shows the humanized CEACAM5 binding domain CEA-A, exemplary affinity-manipulated VH and VL variants, and the consensus framework region (FR) and complementarity determination region (CDR) of the humanized and affinity-manipulated variants. [Figure 46A] This document shows a novel 1A7 VH sequence manipulated to pair with the IGKV1-39 germline sequence (although it may also pair with any 1A7 VL sequence), as well as its consensus framework region (FR) and complementarity determination region (CDR). [Figure 46B] This document shows a novel 1A7 VH sequence manipulated to pair with the IGKV1-39 germline sequence (although it may also pair with any 1A7 VL sequence), as well as its consensus framework region (FR) and complementarity determination region (CDR). [Figure 46C] This document shows a novel 1A7 VH sequence manipulated to pair with the IGKV1-39 germline sequence (although it may also pair with any 1A7 VL sequence), as well as its consensus framework region (FR) and complementarity determination region (CDR). [Figure 46D] This document shows a novel 1A7 VH sequence manipulated to pair with the IGKV1-39 germline sequence (although it may also pair with any 1A7 VL sequence), as well as its consensus framework region (FR) and complementarity determination region (CDR). [Figure 47] This study demonstrates the affinity of a novel 1A7 variant paired with IGKV1-39 human germline VL to human CD28. [Figure 48A] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48B] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48C]The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48D] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48E] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48F] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48G] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48H] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48I] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 mAb-scFv format is shown. [Figure 48J] The sequence of an exemplary CEA×CD28 bsAb in 2+1 mAb-scFv format is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that CEA×CD28 bsAb can be 90, 95, 98, and 99% identical (as defined herein) and / or may utilize variable, Fc, and constant domain sequences containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 49A] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49B]The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49C] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49D] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49E] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49F] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49G] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49H] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49I] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49J] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49K] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49L] The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc format is shown. [Figure 49M]The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-scFv-Fc form is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that CEA×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 50A] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50B] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50C] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50D] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50E] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50F] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50G] The sequence of CEA×CD28 bsAb, an example of the 2+1 Fab2-Fc×scFv-Fc format, is shown. [Figure 50H]The sequence of an exemplary CEA×CD28 bsAb in the 2+1 Fab2-Fc×scFv-Fc form is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that CEA×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 51] A-D show induced T cell activation (A) indicated by CD25 expression on CD4 T cells and B) CD8 T cells and T cell proliferation (C) indicated by CD4 and D) CD8 T cell counts) after incubation of purified T cells with SNU-C1 tumor cells (1:1 effector:target ratio) and exemplary B7H3×CD3 bsAb at 1 μg / ml and escalating doses of XENP44338. [Figure 52] Figures A-E show the induction of IL2 secretion from purified T cells incubated with A) HPAF-II (high CEA), B) MKN45 (high CEA), C) LS1074, D) LS174T (medium-low CEA), and E) HT-29 (low CEA) tumor cells (1:1 effector:target ratio) as well as with exemplary B7H3×CD3 bsAb at 1 μg / ml and specified CEA×CD28 bsAb with an elevating CD28 affinity ladder. [Figure 53]Figures A-E show the induction of IL2 secretion from purified T cells incubated with specified effector:target ratios of tumor cells (A) HPAF-II (high CEA), B) SNU-C1 (high CEA), C) LS1034 (medium CEA), D) LS513 (medium-low CEA), and E) LS174T (medium-low CEA) as well as exemplary B7H3×CD3 bsAb at 1 μg / ml and specified CEA×CD28 bsAb in escalating doses in the form of 1+1 Fab-scFv-Fc, 2+1 mAb-scFv, or 2+1 Fab2-scFv-Fc. [Figure 54] The EC50 of IL2 secretion induction from purified T cells incubated with LS174T (medium-low CEA) tumor cells (1:1 effector:target ratio) and exemplary B7H3×CD3 bsAb at 1 μg / ml and escalating doses of specified CEA×CD28 bsAb in the form of 1+1 Fab-scFv-Fc or 2+1 Fab2-Fc×scFv-Fc, along with escalating concentrations of soluble CEA. The data show that while the 2+1 form exhibits lower efficacy and potency, it demonstrates superior tolerance to soluble CEA, as indicated by the large amount of sCEA required to shift the EC50 from baseline. [Figure 55] Figures A-C illustrate mechanism-based modeling of trimer formation (i.e., bispecific antibody, T cell, and tumor cell trimers) for bsAb having a) a 0.2 nM CEA-binding domain, b) a 2 nM CEA-binding domain, and c) a 15 nM CEA-binding domain, as well as for soluble CEA at increasing concentrations. [Figure 56] This demonstrates the induction of IL2 secretion from HPAF-II tumor cells (1:1 effector:target ratio) and purified T cells incubated with 1 μg / ml exemplary B7H3×CD3 bsAb and specified CEA×CD28 bsAb in escalating doses, each containing CEA-A H0L0 (approx. 0.7–2 nM), CEA-A H1L1 (approx. 5–8 nM), or CEA-A H1L2 (57–69 nM), with or without 1 μg / ml soluble CEA. [Figure 57]This demonstrates the induction of IL2 secretion from purified T cells incubated with HPAF-II (high CEA) tumor cells (1:1 effector:target ratio) and specified CEA×CD28 bsAb at escalating doses (with or without 1 μg / ml sCEA) containing exemplary B7H3×CD3 bsAb and affinity-modified CEA-A H1L1 variants. [Figure 58] This demonstrates the induction of IL2 secretion from purified T cells incubated with HPAF-II (high CEA) tumor cells (1:1 effector:target ratio) and specified CEA×CD28 bsAb at escalating doses (with or without 1 μg / ml sCEA) containing exemplary B7H3×CD3 bsAb and affinity-modified CEA-A H1L2 variants. [Figure 59] This demonstrates the induction of IL2 secretion from purified T cells incubated with HPAF-II (high CEA) tumor cells (1:1 effector:target ratio) and specified CEA×CD28 bsAb at escalating doses, including exemplary B7H3×CD3 bsAb and affinity-modified CEA-A H1L1 and H1L2 variants, at a dose of 1 μg / ml. [Figure 60] This demonstrates the induction of IL2 secretion from purified T cells incubated with HPAF-II (high CEA) tumor cells (1:1 effector:target ratio) and specified CEA×CD28 bsAb at escalating doses (with or without 1 μg / ml sCEA) containing exemplary B7H3×CD3 bsAb and affinity-modified CEA-A H1L1 and H1L2 variants. [Figure 61] A and B show the induction of IL2 secretion at A) 6 hours or B) 24 hours from purified T cells incubated with specified doses of CEA×CD28 bsAb in an escalating dose with MKN-45 (high CEA) or Lovo (low CEA) cells (1:1 effector:target ratio). [Figure 62] This shows the induction of IL2 secretion from LS1034 tumor cells (effector:target ratio of 10:1) and purified T cells incubated with or without specified doses of CEA×CD28 bsAb and 20nMPD-1mAb. [Figure 63]This shows increased thermal stability of disulfide-stabilized 1A7 scFv. [Figure 64] The CD28 affinity of additional 1A7 scFv variants, with or without disulfide stabilization (indicated as [SS]), is shown, obtained from different experiments in association with different CD28 multispecific antibodies. All scFvs shown below were performed using scFv linker (GKPGS) 4, but any scFv linker (including that in Figure 5) may also be used. Furthermore, the scFv may be in either the VH-scFv linker-VL or VL-scFv linker-VH direction. [Figure 65A] An exemplary 1A7-based scFv is shown, in which cysteine ​​was manipulated for disulfide stabilization. [Figure 65B] An exemplary 1A7-based scFv is shown, in which cysteine ​​was manipulated for disulfide stabilization. [Figure 65C] An exemplary 1A7-based scFv is shown, in which cysteine ​​was manipulated for disulfide stabilization. [Figure 65D] An exemplary 1A7-based scFv is shown, in which cysteine ​​was manipulated for disulfide stabilization. [Figure 65E] An exemplary 1A7-based scFv is shown, in which cysteine ​​is manipulated for disulfide stabilization. All scFvs shown below use the scFv linker (GKPGS) 4, but any scFv linker (including that shown in Figure 5) may be used. Furthermore, the scFv can be in either the VH-scFv linker-VL or VL-scFv linker-VH direction; however, only the former is illustrated. [Figure 66] This demonstrates the induction of IL2 secretion from HPAF-II tumor cells (1:1 effector:target ratio) and purified T cells incubated with 1 μg / ml exemplary B7H3×CD3 bsAb and specified CEA×CD28 bsAb in escalating doses, each containing or not containing disulfide-stabilized CD28 scFv. [Figure 67]This study demonstrates the manipulation of A431-β2M-deficient cells that stably express CEACAM5 to express CMV-derived (HLA-A2 / NLVa) or melanoma-derived (HLA-A2 / G209; negative control) pMHC antigen complexes. The cells were co-cultured with PBMCs derived from HLA-A2+ and CMV-serologically positive donors and treated with titration doses of CEACAM5 × CD28 XENP45906. [Figure 68] This shows the induction of IL2 from purified T cells incubated with MKN45 (high CEA), HPAF-II (high CEA), LS1034 (medium CEA), LS174T (medium CEA), Lovo (medium CEA), and HT29 (low CEA) cells (1:1 effector:target ratio), as well as with exemplary B7H3×CD3 bsAb at 1 μg / ml and escalating doses of CEA×CD28 XENP44915. [Figure 69] A and B show A) CD8 T cell activation (indicated by CD25 expression) and B) induction of tumor cell killing after incubation of purified T cells with HPAF-II (high CEA) cells and exemplary B7H3×CD3 bsAb at 0.1 or 1 μg / ml and escalating doses of CEA×CD28 XENP44915. The data demonstrate that CEA×CD28 synergistically interacts with high and low concentrations of T cell activators. [Figure 70] Figures A-F show the induction of IL2 secretion from purified T cells incubated with A)MKN45, B)HPAF-II, C)LS1034, D)LS174T, E)Lovo, and F)HT-29 cells (1:1 effector:target ratio) and specified CEA×CD28 bsAb in escalating doses containing 1 μg / ml exemplary B7H3×CD3 bsAb and CD28 scFv with or without disulfide stabilization and CEA-binding domains with different affinity. [Figure 71]Figures A-D show the induction of IL2 secretion from purified T cells incubated with A) MKN-45, B) LS174T, C) Lovo, and D) HT-29 tumor cells (1:1 effector:target ratio) and exemplary B7H3×CD3 bsAb and specified CEA×CD28 bsAb in escalating doses in the form of 1+1 Fab-scFv-Fc or 2+1 Fab2-Fc×scFv-Fc. [Figure 72] Figures A and B show the induction of IL2 secretion from specified tumor cells (1:1 effector:target ratio) and purified T cells incubated with exemplary B7H3×CD3 bsAb at 1 μg / ml and escalating doses of A)XENP45906 or B)XENP46322. [Figure 73] This demonstrates the induction of A) LS-174T or B) Lovo cancer cell killing by purified T cells (1:1 effector:target ratio) incubated with 1 μg / ml CEACAM5×CD28 bsAb XENP45906 or XENP46322 and exemplary B7H3×CD3 bsAb in escalating doses. [Figure 74] This shows the induction of HPAF-II cancer cell killing by purified T cells (1:1 effector:target ratio) incubated with CEACAM5×CD28 bsAb A)XENP45906 or B)XENP46322 and exemplary doses of EpCAM×CD3 bsAb. [Figure 75] This shows induction of IL2 secretion by A431 HLA-A2 / NLV CEA knock-in cells with CEACAM5 antigen densities of A) 250K, B) 100K, and C) 34K, as well as by purified CMV-reactive T cells (10:1 effector:target ratio) incubated with escalating doses of CEACAM5 × CD28 bsAb XENP45906 or XENP46322 alone or with anti-PD1. [Figure 76] This shows the induction of IL2 secretion by purified T cells (1:1 effector:target ratio) incubated with 1 μg of HPAF-II and escalating doses of CEACAM5×CD28 bsAb A)XENP45906 or B)XENP46322 with 1 μg of B7H3×CD3 and various concentrations of soluble CEA, with or without. C) Particle changes. [Figure 77] This shows the time course of tumor volume (determined by caliber measurement) in HPAF-II and huPBMC-transplanted NSG-DKO mice administered with A) XENP46322 or B) XENP45906 along with exemplary EpCAM×CD3 bsAb. [Figure 78] The following shows the number of human CD45+, human CD4+, and human CD8+ cells at day 21 in HPAF-II and huPBMC-transplanted NSG-DKO mice administered with XENP46322 or XENP45906 along with exemplary EpCAM×CD3 bsAb. [Figure 79] This shows an Octet sensorgram at 25°C demonstrating the binding of CEACAM5×CD28 bsAb containing CEA-A_H0L0, CEA-AH1L1, or CEA-A_H1L2 to CEACAM5 and CEACAM8. [Figure 80] This shows the binding of CEACAM5×CD28 bsAb to CEACAM5 and CEACAM8, with further operational variants of CEA-A_H1.7_L2.15 determined by Octet at 25°C. [Figure 81] A) Induction of IL2 secretion by purified T cells in the presence of HPAF-II or B) LS174T cancer cells, and with or without titrated doses of CEACAM5×CD28 and 1 μg / ml B7H3×CD3 bsAb and 1 μg / ml soluble CEA, manipulated to reduce CEACAM8 cross-reactivity. [Figure 82] The sequences of the IGKV1-39 human germline (CDRs are underlined) are shown. These can be paired with any of the 1A7s shown herein, but the preferred pair is shown in Figure 47. [Modes for carrying out the invention]

[0111] I. Overview Six-transmembrane prostatic epithelial antigen 1 (STEAP1) is a cell surface antigen that functions in maintaining intracellular iron homeostasis. Carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) is a cell surface glycoprotein and a member of the carcinoembryonic (CEA) antigen family. Because STEAP1 and CEACAM5 are overexpressed on various tumors, these tumor antigens are candidates for targeted therapy development.

[0112] T cell activation in cancer treatment is a widely studied topic. T cells require multiple signals for complete activation and differentiation. As shown in Figure 23A, signal 1, facilitated by the recognition of peptide-MHC (pMHC) complexes by the T cell receptor (TCR), is absolutely essential for T cell activation. Signal 2, which acts synergistically with and amplifies signal 1, is typically provided by the interaction of CD28 itself with CD80 and CD86, which are ligands for CD28. CD28 engagement alone is typically inactive, but when combined with signal 1 activation, it promotes additional activation, survival, and proliferation signals (including IL-2 secretion). Since CD80 and CD86 are spontaneously expressed only by professional antigen-presenting cells (APCs), the degree of CD28 costimulation in the tumor environment can vary considerably. Therefore, the present invention relates to a novel class of tumor-targeted anti-CD28 × anti-STEAP1 bispecific antibodies, wherein the CD80 / CD86 engagement of CD28 mimics the CD80 / CD86 engagement of CD28, thereby providing an artificial source of signal 2 (Figure 23B). Notably, signal 1 may be provided by the innate TCR:pMHC recognition of tumor cells, or by a combination of CD28 bispecificity and CD3 bispecificity that can mimic signal 1 (e.g., anti-CD3 × anti-STEAP1).

[0113] Accordingly, provided herein are novel anti-CD28×anti-STEAP1 (also referred to as "αCD28×αSTEAP1," and sometimes simply "CD28×STEAP1") and anti-CD28×anti-CEACAM5 (also referred to as "αCD28×αCEACAM5," and sometimes simply "CD28×CEACAM5") bispecific antibodies, as well as methods for using such antibodies for the treatment of cancer. In many cases, these bispecific antibodies are heterodimers. The αCD28×αSTEAP1 antibody of interest can agonist bind to CD28 costimulatory molecules on T cells and target STEAP1 on STEAP1-expressing tumor cells. Similarly, the αCD28×αCEACAM5 antibody of interest can agonist bind to CD28 costimulatory molecules on T cells and target CEACAM5 on CEACAM5-expressing tumor cells. Therefore, such antibodies selectively enhance antitumor activity at tumor sites expressing the target antigen while minimizing peripheral toxicity.

[0114] The antibodies provided herein are particularly useful for enhancing antitumor activity, whether used alone as monotherapy or in combination with other anticancer therapies as more fully described herein.

[0115] Accordingly, in one embodiment, what is provided herein is a heterodimer antibody that binds to two different antigens, for example, the antibody is "bispecific" in that it binds to two different target antigens, generally CD28 and STEAP1 or CD28 and CEACAM5, as described below. These heterodimer antibodies can bind to each of the target antigens in either a monovalent (e.g., having a single antigen-binding domain, e.g., a pair of variable heavy and variable light domains) or a bivalent (having two antigen-binding domains, each independently binding to the antigen). In some embodiments, the heterodimer antibody provided herein comprises a) one CD28-binding domain and b) one STEAP1 or CEACAM5-binding domain (e.g., a heterodimer antibody of the "1+1 Fab-scFv-Fc" form described herein, and thus bispecific and bivalent). In other embodiments, the heterodimer antibodies provided herein contain a) one CD28-binding domain and b) two STEAP1 or CEACAM5-binding domains (e.g., heterodimer antibodies of the “2+1 Fab2-scFv-Fc”, “2+1 mAb-scFv”, and “2+1 Fab2-Fc×scFv-Fc” forms described herein, which are thus bispecific but trivalent because they contain three antigen-binding domains (ABD)). The heterodimer antibodies provided herein are based on the use of different monomers containing amino acid substitutions (i.e., “scuba riant”) that “bias” towards heterodimer formation rather than homodimer formation, as will be more thoroughly outlined below. In some embodiments, the heterodimer antibodies are also combined with a purified variant (e.g., “pI variant”) that allows for the simple purification of heterodimers from homodimers, as will be similarly outlined below. The heterodimer bispecific antibodies provided typically rely on the use of engineered or variant Fc domains that can be self-constructed in generative cells to generate heterodimer proteins, and on methods for generating and purifying such heterodimer proteins.

[0116] II. Nomenclature The nomenclature for specific antigen-binding domains (e.g., STEAP1 and CD28-binding domains) uses the format "Hx.xx_Ly.yy", where the numbers are unique identifiers for specific variable-chain sequences. For example, the CD28-binding domain "1A7[CD28]_H1_L1" (Figure 15) contains a variable heavy domain, H1, and a variable light domain, L1. When these sequences are used as scFv, the name "H1_L1" indicates that the binding domain contains a variable heavy domain "H1" combined with a variable light domain "L1", with an N-to-C-terminal VH-linker-VL orientation. A molecule having the same sequences of heavy and light variable domains but in the reverse order (N-to-C-terminal VL-linker-VH orientation) would be designated "L1_H1". Similarly, different constructs can "mix and match" heavy and light chains, as is evident from the sequence listings and figures.

[0117] III. Definition To ensure that this application can be fully understood, some definitions are provided below. Such definitions are intended to encompass grammatical equivalents.

[0118] In this specification, “CD28,” “differentiation antigen group 28,” and “Tp44” (e.g., Genebank accession numbers NP_001230006 (human), NP_001230007 (human), NP_006130 (human), and NP_031668 (mouse)) refer to B7 receptors expressed on T cells that provide costimulatory signals necessary for T cell activation and survival. T cell stimulation via CD28, in addition to the T cell receptor (TCR), provides a potent signal for the production of various interleukins. CD28 is a receptor for the CD80 (B7.1) and CD86 (B7.2) proteins. CD28 contains an intercellular domain with a YMNM motif important for the recruitment of SH2 domain-containing proteins, particularly PI3K. CD28 also contains two proline-rich motifs that can bind SH3-containing proteins. An exemplary CD28 sequence is shown in Figure 1. Unless otherwise stated, references to CD28 refer to human CD28 sequences.

[0119] As used herein, “STEAP1,” “PRSS24,” “6-transmembrane prostatic epithelial antigen 1,” or “STEAP family member 1” refers to a metalloproteinase that functions in maintaining iron and copper homeostasis and comprises six transmembrane domains, three extracellular loops, and two intracellular loops. STEAP1 is expressed in certain cancers, including prostate cancer, and is overexpressed in malignant prostate tissue compared to normal prostate tissue. An exemplary STEAP1 sequence is shown in Figure 2.

[0120] In this specification, “CEACAM5,” “carcinoembryonic antigen-associated cell adhesion molecule 5,” “CEA cell adhesion molecule 5,” “CD66e,” or “differentiation antigen group 66e,” or “CEA” are members of the carcinoembryonic antigen (CEA) family, which are cell surface glycoproteins involved in cell adhesion. The CEACAM5 sequence is shown, for example, in Figure 2. CECAM is overexpressed in various cancers, including gastrointestinal, respiratory, and genitourinary cancers, as well as breast cancer.

[0121] As used herein, “reduction” means a decrease or removal of activity. Therefore, for example, “reduction of FcγR binding” means that the Fc region amino acid variant has less than 50% of the initial binding compared to an Fc region without the particular variant, with a loss of activity exceeding 70-80-90-95-98% being preferred, and typically below the binding level detectable in Biacore, SPR, or BLI assays. Those specifically used in FcγR binding reduction are shown in Figure 5, and these are typically added to both monomers.

[0122] "ADCC," or antibody-dependent cell-mediated cytotoxicity, as used herein, refers to a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize a bound antibody on a target cell, subsequently causing lysis of the target cell. ADCC correlates with binding to FcγRIIIa; increased binding to FcγRIIIa results in increased ADCC activity.

[0123] When used herein, "ADCP" or "antibody-dependent cell-mediated phagocytosis" refers to a cell-mediated response in which nonspecific phagocytic cells expressing FcγR recognize bound antibodies on target cells, and subsequently cause phagocytosis of the target cells.

[0124] As used herein, the term “antibody” is used in a general sense. The antibodies provided herein may take many forms as described herein, including conventional antibodies and antibody derivatives, fragments, and mimetic compounds as described herein.

[0125] Conventional immunoglobulin (Ig) antibodies are "Y"-shaped tetramers. Each tetramer typically consists of two identical polypeptide chain pairs, each pair containing one "light chain" monomer (typically with a molecular weight of about 25 kDa) and one "heavy chain" monomer (typically with a molecular weight of about 50–70 kDa).

[0126] Other useful antibody formats include, but are not limited to, the “1+1 Fab-scFv-Fc,” “2+1 Fab2-scFv-Fc,” “2+1 mAb-scFv,” and “2+1 Fab2-FcxscFv-Fc” formats provided herein (see, for example, Figure 14). Additional useful antibody formats include, but are not limited to, the “1+1 common light chain,” “2+1 common light chain,” “mAb-Fv,” “mAb-scFv,” “central-Fv,” “1-arm scFv-mAb,” “scFv-mAb,” “double scFv,” and “trident” format antibodies (Figure 14). See also US20180127501A1 (which is incorporated herein by reference, particularly the relevant parts concerning antibody formats) (see, for example, Figure 2 of US20180127501A1).

[0127] Antibody heavy chains typically include a variable weight (VH) domain containing vhCDR1-3 and an Fc domain containing CH2-CH3 monomers. In some embodiments, antibody heavy chains also include hinge and CH1 domains. Conventional antibody heavy chains are monomers organized from N to C-terminus: VH-CH1-hinge-CH2-CH3. The CH1-hinge-CH2-CH3 are collectively referred to as the "constant domain" or "constant region" of the heavy chain of antibodies in which five different categories or "isotypes" exist: IgA, IgD, IgG, IgE, and IgM.

[0128] In some embodiments, the antibodies provided herein include a constant IgG isotype domain, which has several subclasses, including but not limited to IgG1, IgG2, IgG3, and IgG4. In the IgG subclasses of immunoglobulins, several immunoglobulin domains are present in the heavy chain. As used herein, “immunoglobulin (Ig) domain” means a region of immunoglobulin having a distinctive tertiary structure. Heavy chain domains, including constant heavy (CH) domains and hinge domains, are of interest in the present invention. In relation to IgG antibodies, each IgG isotype has three CH domains. Thus, in relation to IgG, the “CH” domains are as follows: “CH1” points to positions 118-215 according to the EU index in Kabat; “Hinge” points to positions 216-230 according to the EU index in Kabat; “CH2” points to positions 231-340 according to the EU index in Kabat; and “CH3” points to positions 341-447 according to the EU index in Kabat. As shown in Figure 1, the exact numbering and arrangement of heavy chain domains may differ between different numbering systems. As shown herein and described below, pI variants may be in one or more CH regions and in the hinge regions discussed below.

[0129] It should be noted that IgG1 has different allotypes with polymorphisms at 356(D or E) and 358(L or M). The sequences shown herein use the 356E / 358M allotype, but other allotypes are included herein. That is, any sequence containing the IgG1 Fc domain included herein may have 356D / 358L, which replaces the 356E / 358M allotype. It should also be understood that therapeutic antibodies may also contain isotype and / or subclass hybrids. For example, as shown in U.S. Publication 2009 / 0163699 (incorporated by reference), the antibody in some embodiments contains a human IgG1 / G2 hybrid.

[0130] When used herein, “Fc,” “Fc region,” or “Fc domain” refers to a polypeptide containing the constant region of an antibody, in some examples all or part of the first constant region immunoglobulin domain (e.g., CH1), and in some cases optionally all or part of the hinge. In the case of IgG, the Fc domain includes the immunoglobulin domains CH2 and CH3 (Cγ2 and Cγ3), and optionally all or part of the hinge region between CH1 (Cγ1) and CH2 (Cγ2). Thus, in some cases, the Fc domain includes CH2-CH3 and hinge-CH2-CH3 from N to C-terminus. In some embodiments, the Fc domain is from IgG1, IgG2, IgG3, or IgG4, with IgG1 hinge-CH2-CH3 and IgG4 hinge-CH2-CH3 being particularly utilized in many embodiments. Also, in the case of the human IgG1 Fc domain, the hinge may include the C220S amino acid substitution. Furthermore, in the case of the human IgG4 Fc domain, the hinge may include the S228P amino acid substitution. While the boundaries of the Fc region can vary, the human IgG heavy chain Fc region is typically defined as containing residues E216, C226, or A231 at its carboxyl terminus, and the numbering follows the EU index in Kabat. In some embodiments, amino acid modifications are made to the Fc region to alter binding to one or more FcγR or FcRn, as more fully described below.

[0131] In this specification, “heavy chain constant region” means the CH1-hinge-CH2-CH3 portion of an antibody (or fragment thereof), excluding the variable heavy domain; in the EU numbering of human IgG1, this corresponds to amino acids 118-447. In this specification, “heavy chain constant region fragment” means a heavy chain constant region containing fewer amino acids from either or both of the N and C-terminuses, but still retaining the ability to form dimers with another heavy chain constant region.

[0132] Another type of heavy chain domain is the hinge region. In this specification, “hinge,” “hinge region,” “antibody hinge region,” or “hinge domain” refers to a flexible polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain terminates at EU position 215, and the IgG CH2 domain begins at residue EU position 231. Thus, in the case of IgG, the antibody hinge is defined herein as containing positions 216 (E216 in IgG1) to 230 (P230 in IgG1), with numbering following the EU index in Kabat. In some cases, “hinge fragments” containing fewer amino acids at either or both of the N and C-terminuses of the hinge domain are used. As described herein, pI variants can also be made from hinge regions. Many of the antibodies herein have at least one cysteine ​​at position 220, following the EU numbering (hinge region), replaced with serine. Typically, this modification is on the “scFv monomer” side (when the 1+1 or 2+1 form is used) for most of the sequences shown herein, but it can also be on the “Fab monomer” side, or both, to reduce disulfide formation. One or both of these replaced cysteines (C220S) are specifically included in the sequences shown herein.

[0133] As will be understood by those skilled in the art, the precise numbering and arrangement of heavy chain constant region domains (i.e., CH1, hinge, CH2, and CH3 domains) may differ between different numbering systems. A useful comparison of heavy constant region numbering according to EU and Kabat is provided below; please refer to Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85 and Kabat et al., 1991, Sequences of Proteins of Immunological Interest, 5th Ed., United States Public Health Service, National Institutes of Health, Bethesda (incorporated by reference throughout).

[0134] [Table 1]

[0135] Antibody light chains typically contain two domains: a variable light domain (VL) containing light chain CDR vlCDR1-3, and a constant light chain region (often referred to as CL or Cκ). Antibody light chains are typically sequenced VL-CL from the N-terminus to the C-terminus.

[0136] In this specification, “antigen-binding domain” or “ABD” means a set of six complementarity-determining regions (CDRs) that, when present as part of a polypeptide sequence, specifically bind to the target antigens discussed herein (e.g., STEAP1, CEACAM5, or CD28). As is known in the art, these CDRs generally exist as a first set of variable weight CDRs (vhCDR or VHCDR) and a second set of variable light CDRs (vlCDR or VLCDR), each containing three CDRs: vhCDR1, vhCDR2, vhCDR3 variable weight CDRs and vlCDR1, vlCDR2, vlCDR3, and vhCDR3 variable light CDRs. The CDRs reside in variable weight domains (vhCDR1-3) and variable light domains (vlCDR1-3). The variable weight domains and variable light domains form the Fv region.

[0137] The present invention provides a number of different CDR sets. In this case, a “complete CDR set” includes three variable light CDRs and three variable heavy CDRs, for example, vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. These may each be part of a larger variable light or variable heavy domain. Furthermore, as will be more fully outlined herein, the variable heavy and variable light domains may be on separate polypeptide chains when heavy and light chains are used (e.g., when Fab is used), or on a single polypeptide chain in the case of scFv sequences.

[0138] As will be understood by those skilled in the art, the exact numbering and arrangement of CDRs may differ between different numbering systems. However, it should be understood that the disclosure of variable weight and / or variable light sequences includes the disclosure of the associated (intrinsic) CDRs. Thus, the disclosure of each variable weight region is a disclosure of vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the disclosure of each variable light region is a disclosure of vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3). A useful comparison of CDR numbering is as follows: See Lafranc et al., Dev.Comp.Immunol.27(1):55-77(2003):

[0139] [Table 2]

[0140] Throughout this specification, the Kabat numbering system is generally used when referring to residues in the variable domain (approximately residues 1-107 in the light chain variable region and residues 1-113 in the heavy chain variable region), while the EU numbering system is used for the Fc region (e.g., Kabat et al. (cited above) (1991)).

[0141] CDRs contribute to the antigen-binding domain and antigen binding of antibodies, or more specifically, the formation of epitope-binding sites. An "epitope" refers to a determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule, also known as a paratope. Epitopes are groups of molecules, such as amino acids or sugar side chains, and typically possess specific structural and electrical characteristics. A single antigen may have multiple epitopes.

[0142] An epitope may include amino acid residues directly involved in binding (also known as the immunodominant component of the epitope) and other amino acid residues not directly involved in binding, such as amino acid residues that are effectively blocked by a specific antigen-binding peptide (i.e., those amino acid residues are within the footprint of the specific antigen-binding peptide).

[0143] Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are generated by adjacent amino acid residues in a polypeptide chain. Conformational and non-conformational epitopes can be distinguished in that binding to the former, rather than the latter, is lost in the presence of a denaturing solvent.

[0144] An epitope typically contains at least three, more commonly five or eight to ten, amino acids in its own spatial conformation. Antibodies that recognize the same epitope can be identified in a simple immunoassay, e.g., "binning," in which one antibody demonstrates its ability to block the binding of another antibody to a target antigen. As outlined below, the present invention includes not only the antigen-binding domains and antibodies listed herein, but also those that compete for binding to the epitopes bound by the listed antigen-binding domains.

[0145] In some embodiments, the six CDRs of the antigen-binding domain are provided by variable heavy and variable light domains. In the "Fab" form, the set of six CDRs consists of two distinct polypeptide sequences: a variable heavy domain (vh or VH; containing vhCDR1, vhCDR2, and vhCDR3) and a variable light domain (vl or VL; containing vlCDR1, vlCDR2, and vlCDR3), where the C-terminus of the vh domain is bound to the N-terminus of the CH1 domain of the heavy chain, and the C-terminus of the vl domain is bound to the N-terminus of the constant light domain (thus forming the light chain). In the scFv form, the vh and vl domains are covalently linked to a single polypeptide sequence, typically via the use of a linker outlined herein ("scFv linker"), which can be either vh-linker-vl or vl-linker-vh (starting from the N-terminus), with the former being the preferred configuration (each side may include an arbitrary domain linker, depending on the form used). Typically, the C-terminus of the scFv domain is linked to the N-terminus of all or part of the hinge in the second monomer.

[0146] When used herein, “variable region” or “variable domain” means a region of immunoglobulin that includes one or more Ig domains substantially encoded by any of the Vκ, Vλ, and / or VH genes constituting kappa, lambda, and each heavy chain immunoglobulin locus, and contains a CDR that confers antigen specificity. Thus, a “variable heavy domain” pairs with a “variable light domain” to form an antigen-binding domain (“ABD”). Each variable domain also includes three hypervariable regions (“complementarity-determining regions,” “CDRs”) (vhCDR1, vhCDR2, and vhCDR3 in the case of a variable heavy domain and vlCDR1, vlCDR2, and vlCDR3 in the case of a variable light domain) and four framework (FR) regions, arranged from the amino terminus to the carboxyl terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

[0147] "Fab" or "Fab region," as used herein, typically refers to an antibody region on two different polypeptide chains containing VH, CH1, VL, and CL immunoglobulin domains (e.g., VH-CH1 on one chain and VL-CL on the other). Fab may refer to this isolated region or to this region in relation to the bispecific antibody of the present invention. In relation to Fab, Fab includes the Fv region in addition to the CH1 and CL domains.

[0148] When used herein, “Fv,” “Fv fragment,” or “Fv region” means an antibody region containing the VL and VH domains. The Fv region can be formalized as both Fab (two distinct polypeptides, including the constant region typically outlined above, as discussed above) and single-chain Fv (scFv) (where the vl and vh domains are contained in a single peptide conjugated to a linker, typically as discussed herein).

[0149] In this specification, “single-stranded Fv” or “scFv” means a variable weight domain covalently linked to a variable light domain using an scFv linker, typically as discussed herein, to form an scFv or scFv domain. The scFv domain can be oriented either N-to-C-terminus (vh-linker-vl or vl-linker-vh). In sequences shown in the sequence listings and figures, the order of the vh and vl domains is included in the name. For example, H.X_L.Y means that the N-to-C-terminus is vh-linker-vl, and L.Y_H.X means vl-linker-vh.

[0150] Some embodiments of the target antibodies provided herein include at least one scFv domain, which typically contains a variable heavy domain and a variable light domain that are not naturally present but are linked together by an scFv linker. As outlined herein, scFv domains are typically oriented from N to C-terminus as VH-scFv linker-VL, whereas this can be reversed to VL-scFv linker-VH for any scFv domain (or one constructed using vh and vl sequences from Fab) using any linker at one or both ends, depending on the form.

[0151] In this specification, “modification” or “variant” means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence or an alteration to a site chemically linked to a protein. For example, a modification may be an altered carbohydrate or PEG structure linked to a protein. In this specification, “amino acid modification” means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence. For clarity, unless otherwise stated, amino acid modification always refers to amino acids encoded by DNA, e.g., the 20 amino acids that have codons in DNA and RNA.

[0152] In this specification, “amino acid substitution” or “substitution” means the replacement of an amino acid at a specific position in the parent polypeptide sequence with a different amino acid. In particular, in some embodiments, a substitution is a substitution of an amino acid at a specific position that is not found in nature and does not exist naturally in any organism, including within a living organism. For example, substitution E272Y refers to a variant polypeptide, in this case the Fc variant, in which glutamic acid at position 272 is replaced with tyrosine. For clarity, a protein that is manipulated to alter the nucleic acid coding sequence but not alter the starting amino acid (e.g., replacing CGG (encoding arginine) with CGA (still encoding arginine) to increase the host organism’s expression level) is not an “amino acid substitution”; that is, even though it is the generation of a new gene encoding the same protein, if the protein has the same amino acid at the specific position where it starts, it is not an amino acid substitution.

[0153] As used herein, "amino acid insertion" or "insertion" means the addition of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, -233E or 233E specifies the insertion of glutamic acid after position 233 and before position 234. Also, -233ADE or A233ADE specifies the insertion of AlaAspGlu after position 233 and before position 234.

[0154] "Amino acid deletion" or "deletion," as used herein, means the removal of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, E233- or E233#, E233() or E233del specify the deletion of glutamic acid at position 233. Also, EDA233- or EDA233# specifies the deletion of the sequence GluAspAla that begins at position 233.

[0155] "Variant protein," "protein variant," or "variant" as used herein means a protein that differs from that of the parent protein by at least one amino acid modification. A protein variant has at least one amino acid modification compared to the parent protein, but not so many that the variant protein cannot be aligned with the parent protein using an alignment program such as those described below. Typically, a variant protein (e.g., a variant Fc domain outlined herein) is at least 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identical to the parent protein using an alignment program such as BLAST, as described below.

[0156] When used herein, "variant" also refers to a specific amino acid modification that confers a particular function (e.g., "heterodimerated variant," "pI variant," "decay variant," etc.).

[0157] As described below, in some embodiments, the parent polypeptide, e.g., the Fc parent polypeptide, is a multiple constant domain or Fc region from a human wild-type sequence, e.g., IgG1, IgG2, IgG3, or IgG4. However, a human sequence having a variant may also function as a “parent polypeptide,” e.g., the IgG1 / 2 hybrid described in U.S. Publication 2006 / 0134105. The protein variant sequences herein preferably have at least about 80% identity with the parent protein sequence, most preferably at least about 90% identity, and more preferably at least about 95-98-99% identity. Therefore, as used herein, “antibody variant” or “variant antibody” means an antibody that differs from the parent antibody by at least one amino acid modification; “IgG variant” or “variant IgG” means an antibody that differs from the parent IgG (which, in this case, is also often derived from a human IgG sequence) by at least one amino acid modification; and “immunoglobulin variant” or “variant immunoglobulin” means an immunoglobulin sequence that differs from the parent immunoglobulin sequence by at least one amino acid modification. “Fc variant” or “variant Fc” means a protein that contains amino acid modifications in its Fc domain compared to the Fc domain of human IgG1, IgG2, or IgG4.

[0158] "Fc variant" or "variant Fc," as used herein, means a protein containing amino acid modifications in the Fc domain. Modifications may be additions, deletions, or substitutions. Fc variants are defined according to the amino acid modifications that constitute them. For example, N434S or 434S is an Fc variant having a serine substitution at position 434 compared to the parent Fc polypeptide, and the numbering follows the EU index. Similarly, M428L / N434S is an Fc variant having the substitutions M428L and N434S compared to the parent Fc polypeptide. WT amino acid identity may not be specified, in which case the aforementioned variant is referred to as 428L / 434S. It should be noted that the order in which substitutions are provided is arbitrary (i.e., for example, 428L / 434S is the same Fc variant as 434S / 428L). For all positions discussed herein relating to antibodies or their derivatives and fragments (e.g., Fc domains), unless otherwise noted, amino acid numbering follows the EU index. The “EU index,” “EU index in Kabat,” or “EU numbering” scheme refers to the numbering of EU antibodies (Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85 (incorporated herein by reference)). Modifications may be additions, deletions, or substitutions.

[0159] Typically, a variant Fc domain has at least approximately 80, 85, 90, 95, 97, 98, or 99 percent identity with the corresponding parental human IgG Fc domain (using the identity algorithms discussed below (one embodiment utilizes the BLAST algorithm as known in the art using default parameters)). Alternatively, a variant Fc domain may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 or more amino acid modifications compared to the parental Fc domain. Alternatively, the variant Fc domain may have up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications compared to the parent Fc domain. Furthermore, as discussed herein, the variant Fc domains described herein still retain the ability to form dimers with other Fc domains when measured using known techniques described herein, such as non-denaturing gel electrophoresis.

[0160] As used herein, "protein" means at least two covalently linked amino acids, and includes proteins, polypeptides, oligopeptides, and peptides. Furthermore, polypeptides constituting the antibodies of the present invention may include synthetic derivatization, glycosylation, PEGylation, cyclic substitution, cyclization, linking to other molecules, fusion to proteins or protein domains, and addition of peptide tags or labels to one or more side chains or terminals.

[0161] As used herein, "residue" means the position in a protein and its associated amino acid identity. For example, asparagine 297 (also referred to as Asn297 or N297) is the residue at position 297 in the human antibody IgG1.

[0162] As used herein, "IgG subclass modification" or "isotype modification" refers to an amino acid modification that converts a certain amino acid in one IgG isotype to a corresponding amino acid in a differently aligned IgG isotype. For example, since IgG1 contains tyrosine at EU position 296 and IgG2 contains phenylalanine, the F296Y substitution in IgG2 is considered an IgG subclass modification.

[0163] As used herein, “modifications not found in nature” means non-isomorphic amino acid modifications. For example, since none of the human IgGs contain serine at position 434, the substitution 434S in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a modification not found in nature.

[0164] As used herein, "amino acid" and "amino acid identity" mean one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0165] As used herein, "effector function" means a biochemical event resulting from the interaction between an antibody Fc region and an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

[0166] As used herein, "IgG Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an IgG antibody to form an Fc / Fc ligand complex. Fc ligands include, but are not limited to, FcγRI, FcγRII, FcγRIII, FcRn, C1q, C3, mannan-binding lectins, mannose receptors, staphylococcal protein A, streptococcal protein G, and viral FcγR. Fc ligands also include Fc receptor homologs (FcRH), a family of Fc receptors homologous to FcγR (Davis et al., 2002, Immunological Reviews 190:123-136 (integrated as a whole by reference)). Fc ligands may include undiscovered molecules that bind Fc. Specific IgG Fc ligands are FcRn and Fc gamma receptors. As used herein, "Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an antibody in order to form an Fc / Fc ligand complex.

[0167] When used herein, “Fc gamma receptor,” “FcγR,” or “Fc gamma R” means any member of the family of proteins that ligate to the Fc region of an IgG antibody and are encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI(CD64) including isoforms FcγRIa, FcγRIb, and FcγRIc; FcγRII(CD32) including isoform FcγRIIa (including allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and FcγRIIc; and FcγRIII(CD16) including isoforms FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIb-NA1 and FcγRIIb-NA2) (Jefferis et al., 2002, Immunol Lett 82:57-65 (incorporated by reference as a whole)), as well as any undiscovered human FcγR or FcγR isoform or allotype. FcγR can originate from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. Mouse FcγR includes, but is not limited to, FcγRI(CD64), FcγRII(CD32), FcγRIII(CD16), and FcγRIII-2(CD16-2), as well as any undiscovered mouse FcγR or FcγR isoform or allotype.

[0168] When used herein, “FcRn” or “fetal Fc receptor” means a protein that conjugates the Fc region of an IgG antibody and is at least partially encoded by the FcRn gene. FcRn may be from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, a functional FcRn protein comprises two polypeptides, often referred to as a heavy chain and a light chain. The light chain is beta-2-microglobulin, and the heavy chain is encoded by the FcRn gene. Unless otherwise described herein, FcRn or FcRn protein refers to the complex of the FcRn heavy chain and beta-2-microglobulin. Various FcRn variants are used to increase binding to the FcRn receptor and, in some cases, to increase the serum half-life. “FcRn variant” is an amino acid modification that contributes to increased binding to the FcRn receptor, and preferred FcRn variants are shown below.

[0169] Where used herein, “parent polypeptide” means the initiating polypeptide that is later modified to produce a variant. The parent polypeptide may be a naturally occurring polypeptide, or a variant or engineered version of a naturally occurring polypeptide. Thus, where used herein, “parent immunoglobulin” means an unmodified immunoglobulin polypeptide that is modified to produce a variant, and “parent antibody” means an unmodified antibody that is modified to produce a variant antibody. It should be noted that “parent antibody” includes known commercially available recombinantly produced antibodies, as outlined below. In this context, “parent Fc domain” is relative to the described variant; thus, “variant human IgG1 Fc domain” is compared to the parent Fc domain of human IgG1, “variant human IgG4 Fc domain” is compared to the parent Fc domain of human IgG4, and so on.

[0170] As used herein, "position" means a location in the sequence of a protein. Positions may be numbered sequentially or in an established format, for example, according to the EU index for numbering antibody domains (e.g., CH1, CH2, CH3, or hinge domains).

[0171] As used herein, "target antigen" means a molecule that is specifically bound by an antigen-binding domain, which includes a variable region of a given antibody.

[0172] In the context of the monomers of the heterodimer antibodies of the present invention as used herein, “strandiness” means maintaining the ability to “pair” to form a heterodimer by incorporating the heterodimerizing variant into each monomer, similar to how two strands of DNA “pair.” For example, when manipulating monomer A with a certain pI variant (e.g., to increase the pI), a stereovariant that is a “charge pair” and can be similarly utilized does not interfere with the pI variant, and for example, a charge variant that increases the pI will be placed on the same “strand” or “monomer” to maintain the functionality of both. Similarly, with regard to “skew” variants provided in pairs of sets, as will be more thoroughly outlined below, those skilled in the art will consider the pI when determining which strand or monomer one set of the pair will be placed on, so that pI separation is also maximized using the pI of the skew.

[0173] As used herein, "target cell" means a cell that expresses a target antigen.

[0174] In relation to generating the bispecific antibody according to the present invention, "host cell" means a cell that contains exogenous nucleic acids encoding the components of the bispecific antibody and is capable of expressing the bispecific antibody under suitable conditions. Suitable host cells are discussed below.

[0175] In this specification, “wild-type,” “WT,” or “wild-type” means a naturally occurring amino acid or nucleotide sequence, including allele variations. WT proteins have an amino acid or nucleotide sequence that is not intentionally modified.

[0176] This specification provides numerous antibody domains (e.g., Fc domains) that have sequence identity with human antibody domains. Sequence identity between two similar sequences (e.g., antibody variable domains) can be determined using the following methods: Smith, TF & Waterman, MS (1981) “Comparison Of Biosequences,” Adv.Appl.Math.2:482 [Local homology algorithm]; Needleman, SB & Wunsch, CD (1970) “A General Method Applicable To The Search For Similarities In The Amino Acid Sequence Of Two Proteins,” J.Mol.Biol.48:443 [Homologous alignment algorithm]; Pearson, WR & Lipman, DJ (1988) “Improved Tools For Biological Sequence Comparison,” Proc.Natl.Acad.Sci.(USA)85:2444 [Similarity search method]; or Altschul, SF et al, (1990) “Basic Local Alignment Search Tool,” Sequence identity can be measured by algorithms such as the "BLAST" algorithm (see https: / / blast.ncbi.nlm.nih.gov / Blast.cgi), as described in J.Mol.Biol.215:403-10. When using any of the aforementioned algorithms, default parameters (regarding window length, gap penalty, etc.) are used. In one embodiment, sequence identity is performed using the BLAST algorithm with default parameters.

[0177] The antibodies of the present invention are typically isolated or recombinant. "Isolated", when used to describe the various polypeptides disclosed herein, means a polypeptide that has been identified, separated and / or recovered from the cells or cell culture in which it was expressed. Usually, an isolated polypeptide will be prepared by at least one purification step. An "isolated antibody" refers to an antibody that is substantially free of other antibodies having different antigen specificities. "Recombinant" means that the antibodies are produced using recombinant nucleic acid techniques in a foreign host cell, and they can also be isolated.

[0178] "Specific binding" or "specifically binds to" or "specific for" a particular antigen or epitope means a binding that is clearly distinct from non-specific interactions. Specific binding can be measured, for example, by determining the binding of a molecule as compared to the binding of a control molecule, which is usually a molecule of similar structure that does not have binding activity. For example, specific binding can be determined by competition with a control molecule similar to the target.

[0179] Specific binding to a particular antigen or epitope is at least about 10 -4 M, at least about 10 -5 M, at least about 10 -6 M, at least about 10 -7 M, at least about 10 -8 M, at least about 10 -9 M, alternatively at least about 10 -10 M, at least about 10 -11 M, at least about 10 -12 M, or can be shown by an antibody having a KD for an antigen or epitope greater than that, where KD refers to the dissociation rate of a particular antibody-antigen interaction. Typically, an antibody that specifically binds an antigen will have a KD that is 20, 50, 100, 500, 1000, 5,0OO, 10,000 times higher or more than that for the control molecule with respect to the antigen or epitope.

[0180] Furthermore, specific binding to a particular antigen or epitope can be demonstrated by antibodies with a KA or Ka constant for the antigen or epitope that is at least 20, 50, 100, 500, 1,000, 5,000, or more than 10,000 times compared to the control, where KA or Ka refers to the binding constant of a particular antibody-antigen interaction. Binding affinity is typically measured using Biacore, SPR, or BLI assays.

[0181] IV. Anti-CD28 x anti-STEAP1 antibody In one embodiment, a novel anti-CD28 × anti-STEAP1 antibody is provided herein. In some embodiments, the anti-CD28 × anti-STEAP1 antibody described herein can agonist-conjugate to the CD28 costimulatory molecule on T cells and to STEAP1 on tumor cells. Such antibodies selectively enhance antitumor activity at STEAP1-related tumor sites while minimizing peripheral toxicity. The target antibodies provided herein are particularly useful in combination with other anticancer therapies, for example, including bispecific antibodies for the treatment of STEAP1-related cancers.

[0182] The anti-CD28 × anti-STEAP1 antibody is polyvalent and comprises at least two antigen-binding domains (ABDs), where at least one antigen-binding domain is a CD28-binding domain and at least one antigen-binding domain is a STEAP1-binding domain. The anti-CD28 × anti-STEAP1 antibody of interest may contain any suitable CD28-binding domains and STEAP1-binding domains, including, for example, the CD28-binding domains and STEAP1-binding domains provided herein.

[0183] The antigen-binding domains provided herein generally include variable heavy domains (VH) having VH-CDR1, VH-CDR-2, and VH-CDR-3, and variable light domains (VL) having VL-CDR1, VL-CDR-2, and VL-CDR-3.

[0184] Furthermore, as discussed above, the numbering used to identify CDRs in sequence listings and figures is Kabat, but as shown in Table 2, different numbering can be used, which changes the amino acid sequence of the CDR.

[0185] Further variants can be created for all of the variable weight domains and variable light domains listed herein. As outlined herein, in some embodiments, a set of six CDRs may have 0, 1, 2, 3, 4, or 5 amino acid modifications (including amino acid substitutions used for specific applications), in addition to variations in the framework regions of the variable weight domains and variable light domains, provided that the framework (excluding the CDRs) retains at least approximately 80, 85, 90, 95, or 99% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380 (the figure and legend are incorporated in their entirety by reference). Thus, for example, identical CDRs described herein can be combined with different framework sequences of human germline sequences, provided that the framework regions retain at least 80, 85, or 90% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380. Alternatively, a CDR may have amino acid modifications (for example, one, two, three, four, or five amino acid modifications in a set of CDRs (i.e., a CDR may be modified as long as the total number of changes in a set of six CDRs is less than six amino acid modifications, and any combination of CDRs to be modified is arbitrary, for example, one change in vlCDR1, two changes in vhCDR2, and zero changes in vhCDR3)) as well as changes in the framework region, provided that the framework region maintains at least 80, 85, or 95-99% identity with respect to a human germline sequence selected from those enumerated in Figure 1 of U.S. Patent No. 7,657,380).

[0186] As will be understood by those skilled in the art, any set of six CDR or VH and VL domains may be in scFv or Fab form, which is then appended to the heavy and light chain constant domains, the heavy chain constant domain containing variants (including within the Fc domain in addition to within the CH1 domain).

[0187] In addition, in embodiments in which the target antibody contains scFv, the scFv may be oriented from the N-terminus to the C-terminus of a VH-scFv linker-VL or a VL-scFv linker-VH. In some forms, one or more ABDs are generally Fabs containing a VH domain on one protein chain (generally as a component of the heavy chain) and a VL domain on another protein chain (generally as a component of the light chain). Exemplary scFv linkers used in the target antibody are shown in Figure 6.

[0188] In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a bispecific antibody. In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a bivalent antibody. In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a trivalent antibody. In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a bispecific bivalent antibody. In some embodiments, the anti-CD28 × anti-STEAP1 antibody contains one CD28-binding domain and one STEAP1-binding domain. In exemplary embodiments, the anti-CD28 × anti-STEAP1 antibody is a bispecific trivalent antibody. In some embodiments, the anti-CD28 × anti-STEAP1 antibody contains one CD28-binding domain and two STEAP1-binding domains.

[0189] The anti-CD28 × anti-STEAP1 antibodies provided herein may be any useful form, including, for example, standard immunoglobulins, as well as the “1+1 Fab-scFv-Fc,” “2+1 mAb-scFv,” “2+1 Fab2-scFv-Fc,” and “2+1 Fab2-Fc × scFv-Fc” forms described herein (Figure 14). Additional useful forms include, but are not limited to, the “mAb-Fv,” “central-Fv,” “1-arm scFv-mAb,” “scFv-mAb,” “double scFv,” and “trident” forms provided herein (e.g., Figure 14). See also US20180127501A1 (which is incorporated herein by reference, particularly the relevant parts concerning antibody forms) (e.g., see Figure 2). In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a heterodimer bispecific antibody containing a variant Fc domain having one of the heterodimerized scuba riant, pI variant, and / or attenuation variant described herein. See, for example, Figure 8.

[0190] Unless otherwise specified herein, the order of names in the antigen list does not confer structure. That is, in the anti-STEAP1 × anti-CD28 1+1 Fab-scFv-Fc antibody, scFv may bind to either STEAP1 or CD28. However, in some cases, the order indicates the structure.

[0191] The anti-CD28 × anti-STEAP1 antibodies provided herein further comprise different antibody domains. As described herein and known in the art, the antibodies described herein comprise different domains within the heavy and light chains, which may overlap. These domains include, but are not limited to, Fc domains, CH1 domains, CH2 domains, CH3 domains, hinge domains, heavy constant domains (CH1-hinge-Fc domain or CH1-hinge-CH2-CH3), variable heavy domains, variable light domains, light constant domains, Fab domains, and scFv domains.

[0192] As shown herein, there are numerous suitable linkers that can be used to covalently link described domains (e.g., scFv, Fab, Fc domains, VH domains, VL domains, etc.) containing conventional peptide bonds, generated by recombinant technology (for use as either domain linkers or scFv linkers). An exemplary linker for linking domains of a target antibody is shown in Figure 7. In some embodiments, the linker peptide may primarily contain the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should have a length appropriate for linking two molecules in such a way that they assume the correct conformation to each other so that they retain the desired activity. In one embodiment, the linker is about 1 to 50 amino acids long, preferably about 1 to 30 amino acids long. In one embodiment, a linker of 1 to 20 amino acids long may be used, and about 5 to about 10 amino acids are utilized in some embodiments. Useful linkers include, for example, glycine-serine polymers, glycine-alanine polymers, alanine-serine polymers, and other flexible linkers, including (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n (wherein n is an integer of at least 1 (and generally 3 to 4)). Alternatively, a variety of non-proteinaceous polymers, including but not limited to polyethylene glycol (PEG), polypropylene glycol, polyoxyalkylene, or copolymers of polyethylene glycol and polypropylene glycol, may be used as linkers.

[0193] Other linker sequences may include any sequence of any length from the CL / CH1 domain, but not all residues of the CL / CH1 domain, for example, the first 5-12 amino acid residues of the CL / CH1 domain. Linkers may originate from immunoglobulin light chains, e.g., Cκ or Cλ. Linkers may originate from immunoglobulin heavy chains of any isotype, including, for example, Cγ1, Cγ2, Cγ3, Cγ4, Cα1, Cα2, Cδ, Cε, and Cμ. Linker sequences may also originate from other proteins, e.g., Ig-like proteins (e.g., TCR, FcR, KIR), hinge region-derived sequences, and other native sequences from other proteins.

[0194] In some embodiments, the linker is a “domain linker” used to ligate any two domains outlined herein. For example, in the 2+1 Fab2-scFv-Fc form, there may be a domain linker that ligates the C-terminus of the CH1 domain of Fab to the N-terminus of scFv, and another arbitrary domain linker that ligates the C-terminus of scFv to the CH2 domain (although in many embodiments, a hinge is used as this domain linker). Any suitable linker may be used, but many embodiments utilize an arbitrary peptide sequence that enables recombination of two domains having sufficient length and flexibility to allow each domain to retain its biological function, in addition to a glycine-serine polymer as a domain linker, including, for example, (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n (wherein n is an integer of at least 1 (and generally 3-4-5)). In some cases, charged domain linkers, as used in some embodiments of the scFv linker, may be used, with regard to the “strandiness” outlined below. An example of a useful domain linker is shown in Figure 7.

[0195] In some embodiments, the linker is an scFv linker used to covalently link the VH and VL domains, as discussed herein. Often, the scFv linker is a charged scFv linker, some of which are shown in Figure 6. Provided herein, so as therein, are charged scFv linkers for facilitating separation in pI between a first monomer and a second monomer. That is, by incorporating either a positive or negative charged scFv linker (or both; in the case of a skeleton using scFv on different monomers), monomers containing a charged linker can alter pI without causing further alteration to the Fc domain. These charged linkers can be substituted within any scFv containing a standard linker. In this case as well, as will be understood by those skilled in the art, the charged scFv linker is used on the correct “chain” or monomer according to the desired alteration of pI. For example, to produce a 1+1 Fab-scFv-Fc heterodimer antibody as discussed herein, the original pI of the Fv region for each of the desired antigen-binding domains is calculated, one is selected to produce the scFv, and a positive or negative linker is selected depending on the pI. Charged domain linkers can also be used to increase the pI separation of monomers according to the present invention, and thus those included in Figures 6 and 7 can be used in any embodiment of this specification in which a linker is utilized.

[0196] An exemplary anti-CD28×anti-STEAP1 antibody is shown, for example, in Figure 24. During cell culture generation of the anti-CD28×anti-STEAP1 antibodies provided herein, a C-terminal lysine residue or a C-terminal lysine and glycine residue may be cleaved from the heavy chain monomer, thereby resulting in a C-terminus "cut-off" variant. See, for example, Jiang et al., Journal of Pharmaceutical Sciences 105:2066-2072 (2016). Thus, in some embodiments provided herein, the anti-CD28×anti-STEAP1 antibody is one variant of the anti-CD28×anti-STEAP1 antibody and includes a deletion of a C-terminal lysine (-K) terminus or a C-terminal lysine and glycine (-GK) residue in one or both Fc domains of the anti-CD28×anti-STEAP1 antibodies described herein. In some embodiments, the deletion is G446del and / or K447del (EU numbering).

[0197] The form of the anti-CD28 × anti-STEAP1 antibody is described in further detail below.

[0198] A. CD28 binding domain The anti-CD28 × anti-STEAP1 antibodies provided herein include at least one CD28-binding domain. The anti-CD28 × anti-STEAP1 antibodies provided herein may include any preferred CD28-binding domain. In exemplary embodiments, the CD28-binding domain is an agonist CD28 ABD that advantageously provides T-cell costimulatory activity.

[0199] As will be understood by those skilled in the art, a suitable CD28-binding domain may comprise a set of six CDRs, either as the underlined CDRs shown in the figure, or as CDRs identified using other alignments within the variable weight (VH) domain and variable light domain (VL) sequences of the CD28-binding domain shown in Figures 15, 18, and 21, when different numbering schemes are used as described herein and shown in Table 2. Additional VH and VL sequences of exemplary CD28-binding domains that can be used in the target antibody are shown in Figures 16 and 17. A suitable CD28 ABD may also include these sequences and the entire VH and VL sequences shown in the figure, used as scFv or Fab.

[0200] In one embodiment, the CD28 antigen-binding domain includes, but is not limited to, those shown in Figures 15, 18, 21, and 65, any six CDRs of any of the CD28-binding domains described herein (i.e., vhCDR1-3 and vlCDR1-3). In some embodiments, the CD28 ABD that binds to human CD28 is one of the following CD28 ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD2 8]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 2 81VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28] _H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN228[CD28]_H4L2, 1A7 [CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H 1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). In some embodiments, the CD28 ABD includes a VH / VL pair selected from VH and VL shown in Figures 15, 16, 17, and 82.

[0201] In addition to the parent CDR sets disclosed in the Figures and Sequence Listings that form the ABD against CD28, variant CD28 ABDs are provided herein that have CDRs containing at least one modification of the CD28 ABD CDRs disclosed herein (e.g., Figures 15, 18, 21, and 65, and the Sequence Listings). In one embodiment, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody contains a set of six CDRs having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, compared to the six CDRs of the CD28 ABDs described herein, including the Figures and Sequence Listings.In an exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody includes a set of six CDRs having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, compared to one of the following six CD28 ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28 ]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341 VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0 , m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1 , TN228[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[S S], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). In certain embodiments, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody is capable of binding to the CD28 antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments). In certain embodiments, the CD28 ABD can bind to the human CD28 antigen (see Figure 1).

[0202] In some embodiments, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody comprises six CDRs that are at least 90, 95, 97, 98, or 99% identical to six CDRs of the CD28 ABD described herein, including in the figures and sequence listings. In exemplary embodiments, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody comprises six CDRs that are at least 90, 95, 97, 98, or 99% identical to six CDRs of one of the following CD28 ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1 .71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD2 8]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[ CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN2 28[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65).In certain embodiments, CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0203] In another exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-STEAP1 antibody comprises one variable weight (VH) domain and a variable light (VL) domain of any of the CD28 ABDs described herein, including in the figures and sequence listings. In the exemplary embodiment, the CD28 ABD is one of the following CD28 ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD2 8]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 2 81VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28] _H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN228[CD28]_H4L2, 1A7 [CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H 1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). In some embodiments, the CD28 ABD includes pairs of VH and VL selected from VH and VL shown in Figures 15, 16, 17, and 82.

[0204] In some embodiments, the anti-CD28 × anti-STEAP1 antibody comprises a CD28 ABD including a variable heavy domain and / or variable light domain, which are variants of the CD28 ABD VH and VL domains disclosed herein. In one embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of the CD28 ABD described herein, including figures and sequence listings. In exemplary embodiments, the variant VH domain and / or VL domain is the following CD28 One of the ABD domains has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14 _L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[C D28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9. 3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN 228[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS] , 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). In some embodiments, the variation is within the VH domain shown in Figures 15, 18, 21, and 65.In some embodiments, the alteration is located within the VL domain as shown in Figures 15, 18, 21, and 65. In some embodiments, the alteration is located within the VH and VL domains as shown in Figures 15, 18, 21, and 65. In some embodiments, one or more amino acid alterations are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid alterations are located within one or more CDRs. In certain embodiments, the CD28 ABD can bind to CD28 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, the CD28 ABD can bind to the human CD28 antigen (see Figure 1).

[0205] In one embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of the CD28 ABDs described herein, including in the figures and sequence listings. In an exemplary embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of one of the following CD28 ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD28]_H 1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD2 8]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN228[ CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A 7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). In some embodiments, the CD28 ABD contains VH that is at least 90, 95, 97, 98, or 99% identical to the VH domains shown in Figures 15-18, 21, and 65. In some embodiments, the CD28 ABD includes a VL that is at least 90, 95, 97, 98, or 99% identical to the VL domain shown in Figures 15-18 and 21.In some embodiments, the CD28 ABD includes VH and VL domains that are at least 90, 95, 97, 98, or 99% identical to the VH and VL domains shown in Figures 15-18, 21, and 65. In certain embodiments, the CD28 ABD is capable of binding to CD28 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments). In certain embodiments, the CD28 ABD is capable of binding to the human CD28 antigen (see Figure 1).

[0206] In some embodiments, the CD28-binding domain of the target anti-CD28 × anti-STEAP1 antibody includes VH, which contains one of the sequences VHCDR1-3 or HFR1-4 shown in Figure 19A. In some embodiments, the CD28-binding domain includes VL, which contains one of the sequences VLCDR1-3 or LFR1-4 shown in Figure 19B.

[0207] In some embodiments, the anti-CD28 × anti-STEAP1 antibody comprises a CD28-binding domain including VH and VL selected from the following: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, wherein VL has the CD28-binding domain shown in Figures 15, 18, 21, and 65, or variants thereof. VL having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the riant, respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figure 15 or 16, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figure 15, 17, or 82, respectively.

[0208] In some embodiments, the anti-CD28 × anti-STEAP1 antibody comprises a CD28-binding domain including VH and VL selected from the following: (i) VH having the amino acid sequence of any VH or a variant thereof of the CD28-binding domain shown in Figures 15, 18, 21, and 65; and (ii) VL having the amino acid sequence of any VL or a variant thereof of the CD28-binding domain shown in Figures 15, 18, 21, and 65; or (i) VH having the amino acid sequence of any VH or a variant thereof shown in Figure 15 or 16; and (ii) VL having the amino acid sequence of any VL or a variant thereof shown in Figures 15, 17, or 82.

[0209] B.STEAP1 binding domain In one embodiment, what is provided herein is a STEAP1 antigen-binding domain (ABD) composition. In embodiments, the STEAP1 ABD comprises a set of six CDRs, either as underlined CDRs as shown in the sequence listing and in Figures 22 or 31 and 32, or as CDRs identified using other alignments within the variable weight (VH) domain of those shown in Figures 22, 31, or 39 and the variable light domain (VL) sequence of those shown in Figures 22, 32, or 39 and the sequence listing (see Table 2), when different numbering schemes are used as described herein and as shown in Table 2. A preferred STEAP1 ABD may also include these sequences and the entire VH and VL sequences shown in the figures, used as scFv or Fab domains. In embodiments, the STEAP1 ABD comprises the variable weight (VH) domain or a variant thereof and / or the variable light domain or a variant thereof of any of the STEAP1 binding domains disclosed herein (e.g., Figures 22 and 39). Such STEAP1-binding domains are useful, for example, for inclusion in anti-CD28 × anti-STEAP1 antibodies provided herein.

[0210] In exemplary embodiments, STEAP1 ABD comprises a variable weight (VH) domain and a variable light (VL) domain from any one of the STEAP1 ABDs described herein, including figures and sequence listings. In exemplary embodiments, STEAP1 ABD is the STEAP1 ABD shown in Figure 22 or 39. In exemplary embodiments, STEAP1 ABD comprises a pair of VH and VL selected from those shown in Figures 22 and 39. In exemplary embodiments, STEAP1 ABD comprises a VH selected from those shown in Figure 31 and a VL selected from those shown in Figure 32.

[0211] STEAP1 ABDs are provided herein that include variable weight domains and / or variable light domains which are variants of the STEAP1 ABD VH and VL domains disclosed herein, in addition to the parent STEAP1 variable weight domains and variable light domains disclosed herein. In one embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of STEAP1 ABD described herein, including figures and sequence listings. In exemplary embodiments, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of STEAP1 ABD shown in Figure 22 or 39. In some embodiments, the changes are located within the VH domain shown in Figure 22 or Figure 31. In some embodiments, the changes are located within the VL domain shown in Figure 22 or Figure 32. In some embodiments, the changes are located within the VH and VL domains shown in Figure 22. In some embodiments, the changes are located within the VH domain shown in Figure 31 and the VL domain shown in Figure 32. In some embodiments, one or more amino acid changes are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid changes are located within one or more of the vhCDR1-3 and / or vlCDR1-3. In some embodiments, the STEAP1 variant includes vhCDR1-3 and / or vlCDR1-3 of either the STEAP1 ABD in Figures 22 and 39. In some embodiments, the STEAP1 variant includes vhCDR1-3 of either the VH domain in Figure 31 and vlCDR1-3 of either the VL domain in Figure 32. In certain embodiments, STEAP1 ABD can be bound to STEAP1 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments).In certain embodiments, STEAP1 ABD can bind to the human STEAP1 antigen (Figure 2).

[0212] In one embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of STEAP1 ABD described herein, including figures and sequence listings. In an exemplary embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of STEAP1 ABD shown in Figure 22. In some embodiments, STEAP1 ABD includes a VH that is at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figure 22 or Figure 31. In some embodiments, STEAP1 ABD includes a VL that is at least 90, 95, 97, 98, or 99% identical to the VL domain shown in Figure 22 or Figure 32. In some embodiments, STEAP1 ABD includes VH and VL that are at least 90, 95, 97, 98, or 99% identical to the VH and VL domains shown in Figure 22. In some embodiments, STEAP1 ABD includes VH and VL that are at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figure 31 and the VL domain shown in Figure 32. In some embodiments, STEAP1 ABD includes any six CDRs (vhCDR1-3 and vlCDR1-3) of the STEAP1 ABD in Figure 22. In some embodiments, STEAP1 ABD includes vhCDR1-3 of any VH domain shown in Figure 31 and vlCDR1-3 of any VL domain shown in Figure 32. In certain embodiments, STEAP1 can bind to STEAP1 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments). In certain embodiments, STEAP1 ABD can bind to human STEAP1 antigen (Figure 2).

[0213] The anti-CD28 × anti-STEAP1 antibodies provided herein contain at least one STEAP1-binding domain. A target antibody containing such a STEAP1 antigen-binding domain (e.g., an anti-STEAP1 × anti-CD3 bispecific antibody) favorably targets cells expressing higher levels of STEAP1 than cells expressing STEAP1 levels (e.g., normal cells).

[0214] In some embodiments, the STEAP1 ABD of the anti-CD28 × anti-STEAP1 antibody comprises a set of six CDRs, either as underlined CDRs as shown in the sequence listing and in Figure 22 or Figures 31 and 32, or as variable weight (VH) domain sequences as shown in Figures 22 and 31, variable light domain (VL) sequences as shown in Figures 22 and 32, and as CDRs identified using other alignments in the sequence listing (see Table 2), when different numbering schemes are used as described herein and as shown in Table 2.

[0215] In exemplary embodiments, the STEAP1 ABD of the anti-CD28 × anti-STEAP1 antibody comprises one of the variable weight (VH) domains and a variable light (VL) domain of the STEAP1 ABDs described herein, including figures and sequence listings. In exemplary embodiments, the STEAP1 ABD of the anti-CD28 × anti-STEAP1 antibody is the STEAP1 ABD shown in Figure 22 or 39. In exemplary embodiments, the STEAP1 ABD of the anti-CD28 × anti-STEAP1 antibody comprises one of the variable weight (VH) domains shown in Figure 31 and one of the variable light (VL) domains shown in Figure 32.

[0216] STEAP1 ABD is provided herein that includes a variable weight domain and / or variable light domain that is a variant of the STEAP1 ABD VH and VL domains disclosed herein, in addition to the parent STEAP1 variable weight domain and variable light domain disclosed herein. In one embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of STEAP1 ABD described herein, including figures and sequence listings. In an exemplary embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of STEAP1 ABD shown in Figure 22 or 39. In exemplary embodiments, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH domain shown in Figure 31 and / or the VL domain shown in Figure 32. In some embodiments, the changes are located within the VH domain shown in Figure 22 or Figure 31. In some embodiments, the changes are located within the VL domain shown in Figure 22 or Figure 32. In some embodiments, the changes are located within the VH and VL domains shown in Figure 22. In some embodiments, the changes are located within the VH domain shown in Figure 31 and the VL domain shown in Figure 32. In some embodiments, one or more amino acid changes are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid changes are located within one or more of the vhCDR1-3 and / or vlCDR1-3. In some embodiments, the STEAP1 variant includes vhCDR1-3 and / or vlCDR1-3 of either the STEAP1 ABD shown in Figure 22 or 39. In some embodiments, the STEAP1 variant includes vhCDR1-3 of either the VH domain shown in Figure 31 and vlCDR1-3 of either the VL domain shown in Figure 32.In certain embodiments, the STEAP1 ABD of anti-CD28 × anti-STEAP1 antibody can bind to STEAP1 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly utilized in many embodiments). In certain embodiments, the STEAP1 ABD can bind to the human STEAP1 antigen (Figure 2).

[0217] In one embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of STEAP1 ABD described herein, including in the Figures and Sequence Listings. In an exemplary embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of STEAP1 ABD shown in Figure 22 or 39. In an exemplary embodiment, the variant VH domain and / or VL domain are at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figure 31 and / or the VL domain shown in Figure 32. In some embodiments, STEAP1 ABD includes a VH that is at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figure 22 or Figure 31. In some embodiments, STEAP1 ABD includes a VL that is at least 90, 95, 97, 98, or 99% identical to the VL domain shown in Figure 22 or Figure 32. In some embodiments, STEAP1 ABD includes VH and VL that are at least 90, 95, 97, 98, or 99% identical to the VH and VL domains shown in Figure 22 or 39. In some embodiments, STEAP1 ABD includes VH and VL that are at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figure 31 and the VL domain shown in Figure 32. In some embodiments, STEAP1 ABD includes any six CDRs (vhCDR1-3 and vlCDR1-3) of the STEAP1 ABD in Figure 22. In some embodiments, STEAP1 ABD includes vhCDR1-3 of any VH domain shown in Figure 31 and vlCDR1-3 of any VL domain shown in Figure 32. In certain embodiments, the STEAP1 ABD of an anti-CD28 × anti-STEAP1 antibody can bind to STEAP1 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly utilized in many embodiments).In certain embodiments, STEAP1 ABD can bind to the human STEAP1 antigen (Figure 2).

[0218] In some embodiments, the anti-CD28 × anti-STEAP1 antibody is a bivalent antibody (e.g., a 1+1 Fab-scFv-Fc antibody) containing one STEAP1 binding domain. In other embodiments, the anti-CD28 × anti-STEAP1 antibody is a trivalent antibody (e.g., a 2+1 mAb-scFv, 2+1 Fab2-scFv-Fc, and 2+1 Fab2-FcxscFv-Fc antibody) containing two STEAP1 binding domains.

[0219] C. Chimeric antibodies and humanized antibodies In certain embodiments, the target antibodies provided herein include heavy chain variable regions from specific germline heavy chain immunoglobulin genes and / or light chain variable regions from specific germline light chain immunoglobulin genes. For example, such antibodies include, or may consist of, human antibodies containing heavy chain or light chain variable regions that are "products of" or "derived from" a specific germline sequence. Human antibodies that are "products of" or "derived from" a human germline immunoglobulin sequence can be identified, for example, by comparing the amino acid sequence of the human antibody with the amino acid sequence of the human germline immunoglobulin and selecting the human germline immunoglobulin sequence that is closest in sequence to (i.e., maximum % identity) the sequence of the human antibody (using the methods outlined herein). Human antibodies that are "products of" or "derived from" a specific human germline immunoglobulin sequence may contain amino acid differences when compared to the germline sequence, for example, by the intentional introduction of naturally occurring somatic mutations or site-directed mutations. However, humanized antibodies are typically at least 90% identical to the amino acid sequence encoded by human germline immunoglobulin genes and contain amino acid residues that identify the antibody as derived from a human sequence when compared to germline immunoglobulin amino acid sequences of other species (e.g., mouse germline sequences). In given cases, a humanized antibody may be at least 95, 96, 97, 98, or 99% identical in amino acid sequence to the amino acid sequence encoded by a germline immunoglobulin gene, or even at least 96%, 97%, 98, or 99%. Typically, a humanized antibody derived from a particular human germline sequence will show a difference of 10 to 20 amino acids or less from the amino acid sequence encoded by the human germline immunoglobulin gene (before the introduction of any skew, pI, and attenuation variants as used herein (i.e., the number of variants is usually small before the introduction of the variants of the present invention)).In certain cases, the humanized antibody may exhibit a difference of 5 or fewer amino acids, or further 4, 3, 2, or 1 or fewer, from the amino acid sequence encoded by the germline immunoglobulin gene (in this case as well, before the introduction of any skew, pI, and attenuation variants as specified herein (i.e., the number of variants is usually small before the introduction of the variants of the present invention)).

[0220] In one embodiment, the parent antibody is affinity-matured as known in the art. Structure-based methods may be used for humanization and affinity maturation, for example, as described in USSN 11 / 004,590. Humanization and / or affinity maturation of the antibody variable region may be performed using selection-based methods, including, but not limited to, those described in Wu et al., 1999, J.Mol.Biol.294:151-162; Baca et al., 1997, J.Biol.Chem.272(16):10678-10684; Rosok et al., 1996, J.Biol.Chem.271(37):22611-22618; Rader et al., 1998, Proc.Natl.Acad.Sci.USA 95:8910-8915; Krauss et al., 2003, Protein Engineering 16(10):753-759 (all incorporated by reference). Other humanization methods, including but not limited to those described in USSN09 / 810,510; Tan et al.,2002,J.Immunol.169:1119-1125;De Pascalis et al.,2002,J.Immunol.169:3076-3084 (all incorporated by reference), may involve grafting only a portion of the CDR.

[0221] D. Heterodimer antibodies In exemplary embodiments, the anti-CD28×anti-STEAP1 antibody provided herein is a heterodimer bispecific antibody comprising two variant Fc domain sequences. Such variant Fc domains include amino acid modifications to facilitate the self-construction and / or purification of the heterodimer antibody.

[0222] An ongoing challenge in antibody technology is the desire for "bispecific" antibodies that bind simultaneously to two different antigens, typically by bringing the different antigens into proximity and resulting in novel functionalities and therapies. These antibodies are usually produced by incorporating genes for each heavy and light chain into the host cell. This typically results in the formation of the desired heterodimer (AB), as well as two homodimers (AA and BB (without the light chain heterodimer problem)). However, a major obstacle in the formation of bispecific antibodies is the bias towards the formation of the desired heterodimer antibody relative to the formation of the homodimer, and / or the difficulty in purifying the heterodimer antibody from the homodimer.

[0223] Numerous mechanisms exist that can be used to generate target heterodimer antibodies. Furthermore, as will be understood by those skilled in the art, these different mechanisms can be combined to ensure high heterodimerization. Amino acid modifications that facilitate the generation and purification of heterodimers are usually collectively referred to as “heterodimerizing variants.” As will be discussed below, heterodimerizing variants include “skew” variants (e.g., the “knob and hole” and “charge pair” variants described below) as well as “pI variants” that enable the purification of heterodimers from homodimers. Useful mechanisms for heterodimerization, as described in U.S. Patent No. US9,605,084 (which in whole and in particular is incorporated herein by reference with respect to the discussion of heterodimerization variants below), include “knobs and holes” ("KIH") as described in U.S. Patent No. US9,605,084, “electrostatic steering” or “charge pairs” as described in U.S. Patent No. US9,605,084, pI variants as described in U.S. Patent No. US9,605,084, and additional general Fc variants outlined in U.S. Patent No. US9,605,084 and below.

[0224] Heterodimizing variants useful for the formation and purification of target heterodimer antibodies (e.g., bispecific antibodies) are discussed in more detail below.

[0225] 1. Scubarian In some embodiments, heterodimer antibodies include scubarians, which are one or more amino acid modifications in the first Fc domain (A) and / or the second Fc domain (B) that prioritize the formation of an Fc heterodimer (an Fc dimer (AB) containing the first and second Fc domains) rather than an Fc homodimer (an Fc dimer (AA or BB) containing two first Fc domains or two second Fc domains). Preferred scubarians are shown in Figure 29, and in Figures 3 and 8, of U.S. Publication No. 2016 / 0355608 (which is incorporated herein by reference in whole and in particular with respect to its disclosure of scubarians).

[0226] One particular type of scubarian is usually referred to in the art as “knob and hole” when referring to amino acid manipulations that produce steric effects that favor heterodimerization and not homodimerization, as described in USSN 61 / 596,846, Ridgway et al., Protein Engineering 9(7):617(1996); Atwell et al., J.Mol.Biol.1997 270:26; U.S. Patent No. 8,216,805 (all of these are incorporated herein by reference in whole and in particular with respect to the disclosure of “knob and hole” mutations). This is sometimes referred to herein as “steric variant.” The figure identifies numerous “monomer A-monomer B” pairs that depend on “knob and hole.” Furthermore, as described by Merchant et al., Nature Biotech. 16:677 (1998), these “knob and hole” mutations can be combined with disulfide bonds to further favor the formation of Fc heterodimers.

[0227] Another method used in the formation of heterodimers is sometimes referred to as “electrostatic steering,” as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637 (2010) (the whole of which is incorporated herein by reference). This is sometimes referred to herein as “charge pairing.” In this embodiment, electrostatics is used to bias towards the formation of heterodimers. As those skilled in the art will understand, these can also affect pI and thus purification, and thus in some cases can be considered pI variants. However, these are classified as “scuba riants” because they were generated to force heterodimerization and were not used as purification tools. These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (for example, these are "corresponding monomer sets") and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0228] In some embodiments, scubarians favorably and simultaneously prioritize heterodimerization based on both a “knob and hole” mechanism and an “electrostatic steering” mechanism. In some embodiments, a heterodimer antibody contains one or more sets of such heterodimerized scubarians. These variants are provided in “pairs” of “sets,” that is, one set of the pair is incorporated into a first monomer, and the other set of the pair is incorporated into a second monomer. It should be noted that these sets do not necessarily behave as “knob in hole” variants having a one-to-one correspondence between residues on one monomer and residues on the other monomer. In other words, pairs of these sets can rather form interfaces between the two monomers, thereby promoting heterodimerization and suppressing homodimerization, making it possible to raise the proportion of heterodimers spontaneously formed under biological conditions to over 90% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B) rather than the expected 50%. Exemplary heterodimerizing "skewed" variants are shown in Figures 3 and 8. Such "skewed" variants include, but are not limited to, S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q (EU numbering).

[0229] In exemplary embodiments, the heterodimer antibody includes the S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q; or T366S / L368A / Y407V:T366W (optionally containing a cross-linked disulfide, T366S / L368A / Y407V / Y349C:T366W / S354C) "skew" variant amino acid substitution set (EU numbering). In exemplary embodiments, the heterodimer antibody contains the "S364K / E357Q:L368D / K370S" amino acid substitution set. From a nomenclature standpoint, the pair "S364K / E357Q:L368D / K370S" means that one monomer contains an Fc domain with amino acid substitutions S364K and E357Q, and the other monomer contains an Fc domain with amino acid substitutions L368D and K370S; as described above, the "strandiness" of these pairs depends on the initiation pI.

[0230] In some embodiments, the scuba riants provided herein may, optionally and independently, be incorporated into one or both of the first and second Fc domains of a heterodimer antibody along with any other modifications, including but not limited to other scuba riants (see, for example, Figure 37 of U.S. Publication No. 2012 / 0149876 (which is incorporated herein by reference in particular for its disclosure of scuba riants)), pI variants, isomorphic variants, FcRn variants, attenuation variants, and the like. Furthermore, individual modifications may also, independently and optionally, be included in or excluded from the heterodimer antibody in question.

[0231] In some embodiments, the scuba rians outlined herein may be optionally and independently incorporated into one or both heavy chain monomers together with any pI variant (or other variants, e.g., Fc variant, FcRn variant, etc.), and may be independently and optionally included in or excluded from the heterodimer antibody of interest.

[0232] 2. Purified variant In some embodiments, the heterodimer antibody includes a purified variant that advantageously enables the separation of the heterodimer protein (e.g., anti-CD28 × anti-STEAP1 bispecific antibody) from the homodimer protein.

[0233] Several fundamental mechanisms exist that can facilitate the purification of heterodimer antibodies. For example, modifications to one or both of antibody heavy chain monomers A and B, such that each monomer has a different pI, enable isoelectric purification of heterodimer AB antibodies from monomer AA and BB proteins. Alternatively, several skeletal configurations, such as the "1+1 Fab-scFv-Fc" configuration and the "2+1 Fab2-scFv-Fc" configuration, enable size-based separation. As mentioned above, it is also possible to "distort" heterodimer formation relative to homodimers using scuba riants. Therefore, combinations of heterodimerizing scuba riants and purification variants are particularly useful for the heterodimer antibodies provided herein.

[0234] Furthermore, as will be outlined more thoroughly below, depending on the form of the heterodimeric antibody, purified variants contained in the constant region and / or Fc domain and / or domain linker of the monomer may be used. In some embodiments, the heterodimeric antibody includes additional modifications for alternative functionality, such as Fc, FcRn, and KO variants, which may also produce pI changes.

[0235] In some embodiments, the heterodimer antibodies provided herein include at least one monomer having one or more modifications (i.e., “pI variants”) that alter the pI of the monomer. Generally, as understood by those skilled in the art, there are two general categories of pI variants: those that increase the pI of the protein (basic changes) and those that decrease the pI of the protein (acidic changes). As described herein, all combinations of these variants are possible: one monomer may be wild-type, or a variant that does not exhibit a significantly different pI from the wild-type, while the other may be either more basic or more acidic. Alternatively, each monomer may be modified to be one more basic and the other more acidic.

[0236] Depending on the form of the heterodimer antibody, the pI variant may be contained within the constant and / or Fc domain of the monomer, or a charged linker (either a domain linker or an scFv linker) may be used. That is, antibody forms utilizing scFv(s), e.g., the "1+1 Fab-scFv-Fc" form, may include a charged scFv linker (either positive or negative) to provide a further increase in pI for purification purposes. As will be understood by those skilled in the art, some 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc forms are useful simply with a charged scFv linker and without additional pI adjustment, but the present invention also provides pI variants located in one or both monomers and / or in the charged domain linker. Furthermore, additional amino acid manipulation for alternative functionality may also confer pI changes, e.g., Fc, FcRn, and KO variants.

[0237] In target heterodimer antibodies that utilize pI as a separation mechanism to enable the purification of heterodimer proteins, amino acid variants are introduced into one or both monomeric polypeptides. That is, the pI of one monomer (referred herein as "monomer A" for simplicity) may be manipulated separately from monomer B, or both monomer A and B changes may be altered by increasing pI for monomer A and decreasing pI for monomer B. As will be more thoroughly outlined below, pI changes in either or both monomers may be achieved by removing or adding charged residues (e.g., a neutral amino acid is replaced by a positively or negatively charged amino acid residue (e.g., glycine to glutamic acid)), changing a charged residue from positive or negative to the opposite charge (e.g., aspartic acid to lysine), or changing a charged residue to a neutral residue (e.g., loss of charge; lysine to serine). Numerous of these variants are shown in Figures 3 and 4.

[0238] Therefore, in some embodiments, the target heterodimer antibody includes amino acid modifications in the constant region that alter the isoelectric point (pI) of at least one (if not both) monomer of the dimer protein in order to form a “pI antibody” by incorporating an amino acid substitution (“pI variant” or “pI substitution”) in one or both monomers. As shown herein, if the pIs of the two monomers differ by only 0.1 pH units (0.2, 0.3, 0.4, and 0.5 or more are all utilized in the present invention), separation of heterodimers from two homodimers can be achieved.

[0239] As will be understood by those skilled in the art, the number of pI variants present in each or both monomers(s) to obtain good separation will depend in part on the starting pIs of the components, e.g., in the form 1+1 Fab-scFv-Fc, 2+1 Fab2-scFv-Fc, 1+1 CLC, and 2+1 CLC, on the scFv(1+1 Fab-scFv-Fc, 2+1 Fab2-scFv-Fc) and Fab(s)(s) of the target(s). That is, the Fv sequences of the two target antigens are calculated and then the decision is made to determine which monomers to operate on or in which "direction" (e.g., more positive or more negative). As is known in the art, different Fvs will have different starting pIs to utilize in the present invention. Typically, as outlined herein, the pIs are operated to result in a total pI difference of at least about 0.1 log (preferably 0.2 to 0.5 as outlined herein) between each monomer.

[0240] When pI variants are used to achieve heterodimerization, a more modular approach is provided for designing and purifying antibody-containing bispecific proteins by utilizing the constant region(s) of the heavy chain(s). Thus, in some embodiments, heterodimerizing variants (including skew and pI heterodimerizing variants) are not included in the variable region, so that each individual antibody must be manipulated. Also, in some embodiments, the potential for immunogenicity due to pI variants is significantly reduced by introducing pI variants from different IgG isotypes so that pI can be altered without introducing significant immunogenicity. Thus, an additional problem to be solved is the elucidation of low pI constant domains with a high human sequence content, e.g., minimizing or avoiding non-human residues at any particular position. As an alternative to or in addition to isomorphic substitutions, the potential for immunogenicity due to pI variants is significantly reduced by utilizing isosteric substitutions (e.g., Asn to Asp; and Gln to Glu).

[0241] As discussed below, the secondary benefits that may arise from this pI manipulation are also the extension of serum half-life and increased FcRn binding. Specifically, as described in U.S. Publication No. US2012 / 0028304 (which is incorporated in its entirety by reference), reducing the pI of antibody constant domains (including those found in antibodies and Fc fusions) can result in longer serum retention in vivo. These pI variants for increased serum half-life also facilitate pI alteration for purification.

[0242] Furthermore, it should be noted that pI variants provide additional benefits for the analysis and quality control processes of bispecific antibodies, as their ability to exclude, minimize, and distinguish homodimers is important. Similarly, the ability to reliably test the reproducibility of heterodimeric antibody production is important.

[0243] Typically, the most commonly used embodiments rely on a set of variants, combining a scuba riant that promotes heterodimerization rather than homodimerization with a pI variant that increases the pI difference between the two monomers, making it easier to purify the heterodimer from the homodimer.

[0244] Exemplary combinations of pI variants are shown in Figures 3 and 4, and Figure 30, of U.S. Publication No. 2016 / 0355608 (all of which are incorporated herein by reference in whole and in particular for the disclosure of pI variants). Preferred combinations of pI variants are shown in Figures 3 and 4. As outlined herein and shown in the figures, these variations are shown relative to IgG1, but all isotypes and isotype hybrids can be modified in the same way. R133E and R133Q can also be used when the heavy chain constant domain is derived from IgG2-4.

[0245] In one embodiment, a preferred combination of pI variants has one monomer (negative Fab side) containing the has208D / 295E / 384D / 418E / 421D variant (N208D / Q295E / N384D / Q418E / N421D compared to human IgG1) and a second monomer (positive scFv side) containing a positively charged scFv linker containing (GKPGS)4 (SEQ ID NO: XXX). However, as will be understood by those skilled in the art, the first monomer contains a CH1 domain including position 208. Therefore, in constructs that do not contain the CH1 domain (for example, antibodies that do not utilize the CH1 domain in one of their domains), the preferred negative pI variant Fc set includes the 295E / 384D / 418E / 421D variants (Q295E / N384D / Q418E / N421D compared to human IgG1).

[0246] Therefore, in some embodiments, one monomer has the set of substitutions shown in Figure 8, and the other monomer has a charged linker (either a charged scFv linker (since the monomer contains scFv) or a charged domain linker (which can be selected from those shown in Figure 6, as the form indicates)).

[0247] In some embodiments, modifications are made at the hinge of the Fc domain, including positions 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, and 230, based on EU numbering. Thus, pI mutations and especially substitutions may occur at one or more of positions 216-230, with 1, 2, 3, 4, or 5 mutations being utilized. In this case as well, all possible combinations are intended alone or in conjunction with other pI variants in other domains.

[0248] Specific substitutions used to reduce the pI of the hinge domain include, but are not limited to, deletions at position 221, non-native valine or threonine at position 222, deletion at position 223, non-native glutamate at position 224, deletion at position 225, deletion at position 235, and deletion or non-native alanine at position 236. In some cases, only pI substitutions occur in the hinge domain, while in other cases, these substitutions are added in any combination to other pI variants in other domains.

[0249] In some embodiments, mutations may occur in the CH2 region, including positions 233, 234, 235, 236, 274, 296, 300, 309, 320, 322, 326, 327, 334, and 339, based on EU numbering. It should be noted that changes at positions 233–236 may occur (along with 327A) to increase effector function in the IgG2 backbone. In this case as well, all possible combinations of these 14 positions may occur; for example, antibodies provided herein may include variant Fc domains having 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 CH2 pI substitutions.

[0250] Specific substitutions used to reduce the pI of the CH2 domain include, but are not limited to, non-native glutamine or glutamate at position 274, non-native phenylalanine at position 296, non-native phenylalanine at position 300, non-native valine at position 309, non-native glutamate at position 320, non-native glutamate at position 322, non-native glutamate at position 326, non-native glycine at position 327, non-native glutamate at position 334, non-native threonine at position 339, and all possible combinations within the CH2 and other domains.

[0251] In this embodiment, modifications may be independently and optionally selected from positions 355, 359, 362, 384, 389, 392, 397, 418, 419, 444, and 447 (EU numbering) of the CH3 domain. Specific substitutions used to reduce the pI of the CH3 domain include, but are not limited to, non-native glutamine or glutamate at position 355, non-native serine at position 384, non-native asparagine or glutamate at position 392, non-native methionine at position 397, non-native glutamate at position 419, non-native glutamate at position 359, non-native glutamate at position 362, non-native glutamate at position 389, non-native glutamate at position 418, non-native glutamate at position 444, and deletion or non-native aspartate at position 447.

[0252] In some embodiments, anti-CD28 × anti-STEAP1 antibodies include an amino acid substitution in one of their Fc domains that reduces binding to protein A. Such purified variants produce heterodimers with asymmetric binding to protein A, which can be used to separate the heterodimers from the homodimer population by a pH gradient. Exemplary purified amino acid substitutions that reduce binding to protein A include, but are not limited to, H435R and Y436F (IgG1 CH3 domain, EU numbering). See, for example, US2010331527 (which is incorporated by reference in whole and in particular with regard to relevant disclosures concerning Fc domain modifications for reducing protein A binding).

[0253] 3. Isomorphic Variants Furthermore, many embodiments of the target heterodimer antibodies rely on the “introduction” of pI amino acids at specific positions from one IgG isotype to another, thus reducing or eliminating the possibility of unwanted immunogenicity being introduced into the variant. Numerous of these are shown in Figure 21 of U.S. Publication 2014 / 0370013 (incorporated herein by reference). That is, IgG1 is a common isotype for therapeutic antibodies for a variety of reasons, including high effector function. However, the polyconstant region of IgG1 has a higher pI than that of IgG2 (8.10 vs. 7.31). By introducing IgG2 residues at specific positions into the IgG1 backbone, the pI of the resulting monomer is reduced (or increased), additionally exhibiting a longer serum half-life. For example, IgG1 has glycine at position 137 (pI 5.97), while IgG2 has glutamic acid (pI 3.22); introducing glutamic acid would affect the pI of the resulting protein. As described below, numerous amino acid substitutions are usually required to significantly affect the pI of variant antibodies. However, it should be noted that even changes in the IgG2 molecule can lead to an increase in serum half-life, as discussed below.

[0254] In other embodiments, as described further below, non-isotypic amino acid changes are made to reduce the overall charge state of the resulting protein (for example, by changing higher pI amino acids to lower pI amino acids) or to allow structural modifications for stability or other reasons.

[0255] Furthermore, significant changes can be observed in each monomer of the heterodimer by manipulating the pI of both the heavy and light steady domains. As discussed herein, by making the pI of the two monomers differ by at least 0.5, separation by ion-exchange chromatography, isoelectric focusing, or other isoelectric-sensitive methods may be possible.

[0256] 4. Calculation of pI The pI of each monomer of the antibodies provided herein may depend on the pI of the variant heavy chain constant domain as well as the pI of the total monomer including the variant heavy chain constant domain and its fusion partner. Therefore, in some embodiments, the pI change is calculated based on the variant heavy chain constant domain using the chart in Figure 19 of U.S. Publication 2014 / 0370013. As discussed herein, which monomer to manipulate is typically determined by the Fv and the pI inherent in the skeletal region. Alternatively, the pI of each monomer may be compared.

[0257] 5. pI variants that also confer better FcRn in vivo binding. If a pI variant reduces monomeric pI, the pI variant may have the added benefit of improving serum retention in vivo.

[0258] Although still under investigation, the Fc region is thought to have a longer half-life in vivo because binding to FcRn at pH 6 in endosomes sequesters Fc (Ghetie and Ward, 1997 Immunol Today. 18(12):592-598 (integrated by reference)). The endosomal compartment then recycles Fc to the cell surface. When the compartment opens to the extracellular space, a higher pH of approximately 7.4 induces the release of Fc back into the bloodstream. In mice, Dall' Acqua et al. showed that Fc mutants with increased FcRn binding at pH 6 and pH 7.4 actually had lower serum concentrations and the same half-life as wild-type Fc (Dall' Acqua et al. 2002, J.Immunol. 169:5171-5180 (integrated by reference)). The increased affinity of Fc for FcRn at pH 7.4 is thought to hinder the release and return of Fc into the bloodstream. Therefore, Fc mutations that would increase the half-life of Fc in vivo would ideally increase FcRn binding at lower pH levels while still allowing Fc release at higher pH levels. The amino acid histidine changes its charge state in the pH range of 6.0–7.4. Therefore, it is not surprising to find His residues at important positions in the Fc / FcRn complex.

[0259] Recently, it has been suggested that antibodies with variable regions having lower isoelectric points may also have longer serum half-lives (Igawa et al., 2010 PEDS.23(5):385-392 (incorporated by reference)). However, the mechanism is still not well understood. Moreover, the variable region differs from antibody to antibody. Constant region variants with reduced pI and extended half-lives would offer a more modular approach to improving the pharmacokinetic properties of the antibodies described herein.

[0260] E. Additional Fc variants for additional functionality In addition to the heterodimerization variants discussed above, there are numerous useful Fc amino acid modifications that can be performed for a variety of reasons, including, but not limited to, altering binding to one or more FcγR receptors and modified binding to FcRn receptors, as discussed below.

[0261] Accordingly, the antibodies (heterodimers and homodimers) provided herein may include such amino acid modifications with or without the heterodimerizing variants (e.g., pI variants and stereovariants) outlined herein. Each set of variants may be independently and optionally included in or excluded from any particular heterodimer protein.

[0262] 1. FcγR and FcRn variants Therefore, there are numerous useful Fc substitutions that can be performed to alter the binding of one or more FcγR receptors. In a given embodiment, the target antibody contains modifications (i.e., "FcγR variants") that alter the binding of one or more FcγR receptors. Substitutions resulting in increased and decreased binding may be useful. For example, increased binding to FcγRIIIa is known to typically result in increased ADCC (antibody-dependent cell-mediated cytotoxicity; a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize the bound antibody on target cells, subsequently causing lysis of the target cells). Similarly, decreased binding to FcγRIIb (an inhibitory receptor) may be equally beneficial in several situations. The amino acid substitutions used in the target antibodies include those listed in U.S. Patent Nos. 8,188,321 (particularly Figure 41) and 8,084,582, and U.S. Publication Nos. 20060235208 and 20070148170 (all of which are expressly incorporated herein by reference with respect to the variants disclosed therein that affect them in whole and in particular to Fcγ receptor binding). Specific variants used include, but are not limited to, 236A, 239D, 239E, 332E, 332D, 239D / 332E, 267D, 267E, 328F, 267E / 328F, 236A / 332E, 239D / 332E / 330Y, 239D, 332E / 330L, 243A, 243L, 264A, 264V, and 299T. Such modifications may be present in one or both Fc domains of the target antibody.

[0263] In some embodiments, the target antibody includes one or more Fc modifications that increase the serum half-life. These include, but are not limited to, 434S, 434A, 428L, 308F, 259I, 428L / 434S, 259I / 308F, 436I / 428L, 436I or V / 434S, 436V / 428L, 259I / 308F / 428L, and M252Y / S254T / T256E, as specifically disclosed in USSN12 / 341,769 (the entirety of which is incorporated herein by reference), Fc substitutions utilized for increased binding to the FcRn receptor and increased serum half-life. Such modifications may be contained in one or both Fc domains of the target antibody.

[0264] 2. Decline variant In some embodiments, heterodimer antibodies include one or more modifications that reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) in order to avoid additional mechanisms of action. Such modifications are referred to as “FcγR reduction variants” or “Fc knockout (FcKO or KO)” variants. In these embodiments, for some therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) in order to avoid additional mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific antibodies that monovalently bind CD28, it is usually desirable to reduce FcγRIIIa binding in order to eliminate or significantly reduce ADCC activity. In some embodiments of the target antibody described herein, at least one of the Fc domains contains one or more Fcγ receptor attenuation variants. In some embodiments of the target antibody described herein, both of the Fc domains contain one or more Fcγ receptor attenuation variants. These attenuation variants are shown in Figure 5, each of which may be included or excluded independently and optionally. A preferred embodiment utilizes attenuation variants selected from the group consisting of L234A / L235A / D265S, G236R / L328R, E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K, E233P / L234V / L235A / G236del / S239K / A327G, E233P / L234V / L235A / G236del / S267K / A327G, and E233P / L234V / L235A / G236del. It should be noted that the diminishing variants referred to herein diminish FcγR binding but do not typically diminish FcRn binding.

[0265] As is known in the art, the Fc domain of human IgG1 has the highest binding to the Fcγ receptor, and therefore the reduced-binding variant can be used when the constant domain (or Fc domain) in the heterodimeric antibody backbone is IgG1. Alternatives to, or in addition to, the reduced-binding variant in the IgG1 background, mutations at glycosylation site 297 (usually to A or S) can significantly reduce binding to FcγRIIIa, for example. Human IgG2 and IgG4 have naturally reduced binding to the Fcγ receptor, and therefore their backbones can be used with or without the reduced-binding variant.

[0266] F. Combination of heterodimer and Fc variant As will be understood by those skilled in the art, all of the listed heterodimerized variants (including skewed and / or purified variants) can be combined in any way, independently and arbitrarily, as long as they retain their "strandiness" or "monomer partitioning." Furthermore, all of these variants can be combined with any of the heterodimerized forms.

[0267] While the most commonly used embodiments for the pI variant are shown in the figure, other combinations can be generated according to fundamental rules for modifying the pI difference between the two monomers to facilitate purification.

[0268] Furthermore, both heterodimerized variants (skewed and purified variants) can be independently and optionally combined with Fc-decaying variants, Fc variants, and FcRn variants, as generally outlined herein.

[0269] Exemplary combinations of variants included in some embodiments of heterodimer antibodies in the form of 1+1 Fab-scFv-Fc, 2+1 mAb-Fc, 2+1 Fab2-scFv-Fc, and 2+1 Fab2-Fc×scFv-Fc are shown in Figure 8. In some embodiments, the heterodimer antibodies include the variant combinations shown in Figure 8.

[0270] G. Useful antibody formats As will be understood by those skilled in the art and will be discussed more thoroughly below, the heterodimer bispecific antibodies provided herein can take on several different configurations, as schematically shown in Figure 14.

[0271] As will be understood by those skilled in the art, the heterodimer form of the present invention may have different valencies and may be bispecific. That is, the heterodimer antibody of the present invention may be bivalent and bispecific, or trivalent and bispecific, where the first antigen is bound by two binding domains and the second antigen is bound by a second binding domain. As outlined herein, when CD28 is one of the target antigens, it is preferable that CD28 binds only in a monovalent state.

[0272] The present invention utilizes a CD28-binding domain in combination with a STEAP1-binding domain. As will be understood by those skilled in the art, any combination of anti-CD28 CDRs, anti-CD28 variable light domains and variable heavy domains, Fab and scFv or their variants shown in any of the figures (see in particular Figures 15, 18, 21 and 65) may be used. Similarly, regardless of which of the CDRs, variable light domains and variable heavy domains, Fab and scFv or their variants shown in any of the figures (see, for example, Figures 22, 31 and 32) may be used, any of the STEAP1 antigen-binding domains may be used and may be combined in any combination, either independently or arbitrarily.

[0273] 1.1+1 Fab-scFv-Fc format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-STEAP1 antibodies provided herein is the “1+1 Fab-scFv-Fc” or “bottle opener” form, as shown in Figure 14A. The 1+1 Fab-scFv-Fc form antibody contains a first monomer which is a “standard” heavy chain (VH1-CH1-hinge-CH2-CH3), where VH1 is the first variable heavy domain and CH2-CH3 is the first Fc domain. The 1+1 Fab-scFv-Fc also contains a light chain which includes a first variable light domain VL1 and a constant light domain CL. The light chain interacts with VH1-CH1 of the first monomer to form a first antigen-binding domain which is Fab. The second monomer of the antibody contains a second binding domain which is a single-chain Fv (defined hereafter as “scFv”) and a second Fc domain. The scFv comprises a second variable heavy domain (VH2) and a second variable light domain (VL2), where VH2 is bound to VL2 using a potentially charged scFv linker (see, e.g., Figure 6). The scFv is bound to the heavy chain using a domain linker (see, e.g., Figure 7). The two monomers are brought about by the use of amino acid variants (e.g., heterodimerizing variants discussed above) in constant regions (e.g., Fc domain, CH1 domain, and / or hinge region) that promote the formation of a heterodimer antibody, as will be described more thoroughly below. This structure is roughly visually similar to a bottle opener and is therefore sometimes referred to herein as the “bottle opener” form. In some embodiments, the 1+1 Fab-scFv-Fc form antibody is a bivalent antibody.

[0274] The "1+1 Fab-scFv-Fc" format offers several unique advantages. As is known in the art, antibody analogs that rely on two scFv constructs often have stability and aggregation problems, which can be mitigated in this invention by the addition of "standard" heavy-chain and light-chain pairings. Furthermore, in contrast to formats that rely on two heavy chains and two light chains, there is no problem with incorrect pairing of heavy and light chains (e.g., heavy 1 pairing with light 2).

[0275] In some embodiments of the 1+1 Fab-scFv-Fc antibody, one of the first or second antigen-binding domains is a CD28-binding domain, and the other binding domain is a STEAP1-binding domain. In some embodiments of 1+1 Fab-scFv-Fc, it is an scFv that binds to CD28 and a Fab that binds to STEAP1. An exemplary anti-CD28 × anti-STEAP1 bispecific antibody of the 1+1 Fab-scFv-Fc form is shown in Figure 24.

[0276] In some embodiments, the first and second Fc domains of a 1+1 Fab-scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 8). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0277] In some embodiments, the variant Fc domain includes a degraded variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes the degraded variant E233P / L234V / L235A / G236del / S267K, numbered according to EU numbering.

[0278] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants (including those shown in Figures 3 and 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0279] In exemplary embodiments, the 1+1 Fab-scFv-Fc form antibody contains the amino acid modification combination shown in Figure 8. In such embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0280] In some embodiments, the scFv of the 1+1 Fab-scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 1+1 Fab-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0281] In the exemplary embodiment of the 1+1 Fab-scFv-Fc form antibody, the first Fc domain contains the heterodimerized scuba riant L368D / K370S, the second Fc domain contains the heterodimerized scuba riant S364K / E357Q, each of the first and second Fc domains contains the diminished variant E233P / L234V / L235A / G236del / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 1+1 Fab-scFv-Fc form antibody provided herein contains the (GKPGS)4-charged scFv linker. In some embodiments, the 1+1 Fab-scFv-Fc format antibody provided herein includes the FcRn variant M428L / N434S, and the numbering follows EU numbering. In some embodiments, the scFv of the 1+1 Fab-scFv-Fc format antibody provided herein includes a charged scFv linker (including one shown in Figure 6).

[0282] The target 1+1 Fab-scFv-Fc antibody may contain any suitable CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28-binding domain is one of the following CD28-binding domains or a variant thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1 .71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD2 8]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[ CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN2 28[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). Additional VH and VL sequences of exemplary CD28-binding domains that can be used in the target 1+1 Fab-scFv-Fc type antibody are shown in Figures 15, 16, 17, and 82.

[0283] In some embodiments of the 1+1 Fab-scFv-Fc form, the anti-CD28 ABD has VH and VL domains selected from the following: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof. VL having the amino acid sequences of R2 and vlCDR3, respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variants shown in Figure 15 or 16, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variants shown in Figure 15, 17, or 82, respectively.

[0284] In some embodiments of the 1+1 Fab-scFv-Fc form, the anti-CD28 ABD has VH and VL domains selected from: (i) VH having the amino acid sequence of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; and (ii) VL having the amino acid sequence of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; or (i) VH having the amino acid sequence of the VH or a variant thereof shown in Figure 15 or 16; and (ii) VL having the amino acid sequence of the VL or a variant thereof shown in Figures 15, 17, or 82.

[0285] The target 1+1 Fab-scFv-Fc antibody may contain any preferred STEAP1 binding domain, including any of the STEAP1 binding domains provided herein. In some embodiments, the STEAP1 binding domain is one of the following STEAP1 binding domains or a variant thereof: H1 and L1, H1.1 and L1, H1.2 and L1, H1.3 and L1, H1.4 and L1, H1.5 and L1, H1.6 and L1, H1.7 and L1, H1.8 and L1, H1.9 and L1, H1.10 and L1, H1.11 and L1, H1.12 and L1, H1.13 and L1, H1.14 and L1, H1.15 and L1, H1.16 and L1, H1.17 and L 1. H1.18 and L1, H1.19 and L1, H1.20 and L1, H1.21 and L1, H1.22 and L1, H1.23 and L1, H1.24 and L1, H1.25 and L1, H1.26 and L1, H1.27 and L1, H1.28 and L1, H1.29 and L1, H1.30 and L1, H1.31 and L1, H1.32 and L1, H1.33 and L1, H1.34 and L1, H1.35 and L1, H1.36 and L1, H1.37 and L1, H1.38 and L1, H1 and L1.1, H1 and L1. 2. H1 and L1.3, H0 and L0, H1.21 and L1.3, H1.34 and L1.3, H1.39 and L1, H1.40 and L1, H1.41 and L1, H1.42 and L1, H1.43 and L1, H1.44 and L1, H1.45 and L1, H1.46 and L1, H1.47 and L1, H1.48 and L1, H1.49 and L1, H1.50 and L1, H1.51 and L1, H1.52 and L1, H1.53 and L1, H1.54 and L1, H1.55 and L1, H1.56 and L1, H1.57 and L1, H1.58 and L1, H1.59 and L1, H1.60 and L1, H1.61 and L1, H1.62 and L1, H1.63 and L1, H1.64 and L1, H1.65 and L1, H1.66 and L1, H1.67 and L1, H1.68 and L1, H1.69 and L1, H1.70 and L1, H1.71 and L1, H1.72 and L1, H1.73 and L1, H1.74 and L1, H1.75 and L1, H1.76 and L1, H1.77 and L1, H1.78 and L1, H1.79 and L1, H1.80 and L1, H1.81 and L1, H1.82 and L1, H1.83 and L1, H1.84 and L1, H1.85 and L1, H1.86 and L1, H1.87 and L1, H1.88 and L1, H1.89 and L1, H1.90 and L1, H1.91 and L1, H1.92 and L1, H1.93 and L1, H1.94 and L1, H1.95 and L1, H1.96 and L1, H1.97 and L1, H1.98 and L1, H1.99 and L1, H1.100 and L1, H1.101 and L1, H1.102 and L1, H1.103 and L1, H1.104 and L1, H1.105 and L1, H1.106 and L1, H1.107 and L1, H1.108 and L1, H1.109 and L1, H1.110 and L1, H1.111 and L1, H1.112 and L1, H1.113 and L1, H1.114 and L1, H1.115 and L1, H1.116 and L1, H1.117 and L1, H1.118 and L1, H1.119 and L1, H1.120 and L1, H1.121 and L1, H1.122 and L1, H1.123 and L1, H1.124 and L1, H1.125 and L1, H1.126 and L1, H 1.127 and L1, H1.128 and L1, H1.129 and L1, H1.130 and L1, H1.131 and L1, H1.132 and L1, H1.133 and L1, H1.134 and L1, H1.135 and L1, H1.136 and L1, H1.137 and L1, H1.138 and L1, H1.139 and L1, H1.140 and L1, H1.141 and L1, H1.142 and L1, H1.143 and L1, H1.144 and L1, H1.145 and L1, H1.146 and L1, H1.147 and L1, H1.148 and L1, H1.149 and and L1, H1.150 and L1, H1.151 and L1, H1.152 and L1, H1.153 and L1, H1.154 and L1, H1.155 and L1, H1.156 and L1, H1.157 and L1, H1.158 and L1, H1.159 and L1, H1.160 and L1, H1.161 and L1, H1.162 and L1, H1.163 and L1, H1.164 and L1, H1.165 and L1, H1.166 and L1, H1.167 and L1, H1.168 and L1, H1.169 and L1, H1.170 and L1, H1.171 and L1, H1.172 and L1, H1.173 and L1, H1.174 and L1, H1.175 and L1, H1.176 and L1, H1.177 and L1, H1.178 and L1, H1.179 and L1, H1.180 and L1, H1.181 and L1, H1.182 and L1, H1.183 and L1, H1.184 and L1, H1.185 and L1, H1.186 and L1, H1.187 and L1, H1.188 and L1, H1.189 and L1, H1 and L1.4, H1 and L1.5, H1 and L1.6, H1 and L1.7, H1 and L1.8, H1 and L1.9, H1 and L1.10, H1 and L1.11, H1 and L1.12, H1 and L1.13, H1 and L1.14, H1 and L1.15, H1 and L1.16, H1 and L1.17, H1 and L1.18, H1 and L1.19, H1 and L1.20, H1 and L1.21, H1 and L1.22, H1 and L1.23, H1 and L1.24, H1 and L1.25, H1 and L1.26, H1 and L1.27, H1 and L1.28, H1 and L1.29, H1 and L1.30, H1 and L1.31, H1 and L1.32, H1 and L1.33, H1 and L1.34 , H1 and L1.35, H1 and L1.36, H1 and L1.37, H1 and L1.38, H1 and L1.39, H1 and L1.40, H1 and L1.41, H1 and L1.42, H1 and L1.43, H1 and L1.44, H1 and L1.45, H1 and L1.46, H1 and L1.47, H1 and L1.48, H1 and L1.49, H1 and L1.50, H1 and L1.51, H1 and L1.52, H1 and L1.53, H1 and L1.54, H1 and L1.55, H1 and L1.56, H1 and L1.57, H1 and L1.58, H1 and L1.5 9, H1 and L1.60, H1 and L1.61, H1 and L1.62, H1 and L1.63, H1 and L1.64, H1 and L1.65, H1 and L1.66, H1 and L1.67, H1 and L1.68, H1 and L1.69, H1 and L1.70, H1 and L1.71, H1 and L1.72, H1 and L1.73, H1 and L1.74, H1 and L1.75, H1 and L1.76, H1 and L1.77, H1 and L1.78, H1 and L1.79, H1 and L1.80, H1 and L1.81, H1 and L1.82, H1 and L1.83, H1 and L1.84, H1 and L1.85, H1.42 and L1.47, H1.42 and L1.56, H1.190 and L1.3, H1.191 and L1.3, H1.192 and L1.3, H1.195 and L1.3, H1.196 and L1.3, H1.197 and L1.3, H1.198 and L1.3, H1.199 and L1.3, H1.200 and L1.3, H1.201 and L1.3, H1.202 and L1.3, H1.203 and L1.3, H1.204 and L1.3, H1.205 and L1.3, H1.206 and L1.3, H1.207 and L1.3, H 1.208 and L1.3, H1.209 and L1.3, H1.210 and L1.3, H1.211 and L1.3, H1.212 and L1.3, H1.213 and L1.3, H1.214 and L1.3, H1.215 and L1.3, H1.216 and L1.3, H1.217 and L1.3, H1.218 and L1.3, H1.219 and L1.3, H1.220 and L1.3, H1.221 and L1.3, H1.222 and L1.3, H1.223 and L1.3, H1.224 and L1.3, H1.225 and L1.3, H1.226 and L1.3, H1. 227 and L1.3, H1.228 and L1.3, H1.21 and L1.92, H1.21 and L1.93, H1.21 and L1.94, H1.21 and L1.95, H1.21 and L1.96, H1.21 and L1.97, H1.21 and L1.98, H1.21 and L1.99, H1.21 and L1.100, H1.21 and L1.101, H1.21 and L1.102, H1.21 and L1.103, H1.21 and L1.104, H1.21 and L1.105, H1.21 and L1.106, H1.21 and L1.107, H1.21 and L1 .108, H1.21 and L1.109, H1.21 and L1.110, H1.21 and L1.111, H1.21 and L1.112, H1.21 and L1.113, H1.21 and L1.114, H1.21 and L1.115, H1.21 and L1.116, H1.21 and L1.117, H1.21 and L1.118, H1.21 and L1.119, H1.195 and L1.92, H1.195 and L1.94, H1.195 and L1.95, H1.195 and L1.104, H1.195 and L1.105, H1.195 and L1.109, H1.196 and L1.92, H1.196 and L1.94, H1.196 and L1.95, H1.196 and L1.104, H1.196 and L1.105, H1.196 and L1.109, H1.205 and L1.92, H1.205 and L1.94, H1.205 and L1.95, H1.205 and L1.104, H1.205 and L1.105 , H1.205 and L1.109, H1.193 and L1.91, H1.194 and L1.86, H1.194 and L1.87, H1.194 and L1.88, H1.194 and L1.89, or H1.194 and L1.90, or H1.2320 and L1.100, or H1.230 and L1.115 (Figures 22, 31, 32, and 39). Additional VH and VL sequences of exemplary STEAP1 binding domains that can be used in the target 1+1 Fab-scFv-Fc type antibody are shown in Figures 22, 31, and 32.

[0286] In some embodiments of the 1+1 Fab-scFv-Fc form, the anti-STEAP1 ABD has VH and VL domains selected from: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the STEAP1 binding domains shown in Figure 22 or its variants; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and VL having the amino acid sequence of vlCDR3 respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variants shown in Figure 22 or Figure 31, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variants shown in Figure 22 or Figure 32, respectively.

[0287] In some embodiments of the 1+1 Fab-scFv-Fc form, the STEAP1 ABD has VH and VL domains selected from: (i) a VH having the amino acid sequence of any VH or any variant thereof of the STEAP1 binding domain shown in Figure 22, and (ii) a VL having the amino acid sequence of any VL of the STEAP1 binding domain shown in Figure 22; or (i) a VH having the amino acid sequence of any VH or any variant thereof shown in Figure 22 or Figure 31, and (ii) a VL having the amino acid sequence of any VL or any variant thereof shown in Figure 22 or Figure 32.

[0288] Figure 9 shows several exemplary Fc domain sequences useful in 1+1 Fab-scFv-Fc type antibodies. The “monomer 1” sequence shown in Figure 9 typically refers to the Fc domain of the “Fab-Fc heavy chain,” and the “monomer 2” sequence refers to the Fc domain of the “scFv-Fc heavy chain.” Figures 11 and 12 also provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the first or second monomer of the 1+1 Fab-scFv-Fc form. Furthermore, Figure 13 provides useful CL sequences that can be used in this form.

[0289] 2.2+1 mAb-scFv format One heterodimeric antibody form particularly used in the target bispecific anti-CD28 × anti-STEAP1 antibody is the 2+1 mAb-scFv form shown in Figure 14E. This antibody form contains three antigen-binding domains: two Fab moieties and an scFv bound to one C-terminus of the heavy chain. In some embodiments of this form, each Fab moiety binds to STEAP1 (in this case, human STEAP1), and the "additional" scFv domain binds to CD28. In other words, this mAb-scFv form is a trivalent antibody.

[0290] In these embodiments, the first chain or monomer comprises VH1-CH1-hinge-CH2-CH3 from N to the C-terminus, and the second monomer comprises VH1-CH1-hinge-CH2-CH3-domain linker-scFv domain from N to the C-terminus, where the scFv domain comprises a second VH(VH2), a second VL(VL2), and an scFv linker. For all scFv domains in this specification, the scFv domain may be VH2-scFv linker-VL2 or VL2-scFv linker-VH2 in either orientation from N to the C-terminus. Therefore, the second monomer may contain, from N to C-terminus, VH1-CH1-hinge-CH2-CH3-domain linker-VH2-scFv linker-VL2 or VH1-CH1-hinge-CH2-CH3-domain linker-VL2-scFv linker-VH2. The composition also contains a light chain VL1-CL. In embodiments, this form contains two identical light chains (VL1-CL). In this form, VH1 is a first variable weight (VH) domain, VL1 is a first variable light (VL) domain, VH2 is a second variable weight domain, and VL2 is a second variable light domain. In some embodiments, the form comprises two identical common light chains (VL1-CL, each VL1-CL), each of which associates with one of the first and second monomer VH1-CH1 to form two identical Fabs (i.e., the first antigen-binding domain). In some embodiments, scFv is the second antigen-binding domain. In some embodiments, the first ABD binds to human STEAP1, and the second ABD binds to human CD28.

[0291] In some embodiments, the first and second Fc domains of a 2+1 mAb-scFv antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 8). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0292] In some embodiments, the variant Fc domain includes a degraded variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes the degraded variant E233P / L234V / L235A / G236del / S267K, numbered according to EU numbering.

[0293] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants (including those shown in Figures 3 and 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0294] In some embodiments, the scFv of the 2+1 mAb-scFv antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 mAb-scFv antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0295] In exemplary embodiments, the 2+1 mAb-scFv form antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236del / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 mAb-scFv antibody provided herein includes a (GKPGS) 4-charged scFv linker. In some embodiments, the 2+1 mAb-scFv antibody provided herein includes the FcRn variant M428L / N434S, with numbering according to EU numbering.

[0296] In some embodiments, the scFv of the second monomer of a 2+1 Fab2-scFv-Fc antibody is CD28-conjugated, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a binding domain that binds to STEAP1. A 2+1 mAb-scFv antibody may include any preferred STEAP1-binding domain and CD28-binding domain, including the STEAP1-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 15, 18, and 21).

[0297] The target 2+1 mAb-scFv antibody may contain any suitable CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28-binding domain is one of the following CD28-binding domains or a variant thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1 .71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD2 8]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[ CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN2 28[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65). Additional VH and VL sequences of exemplary CD28-binding domains that can be used in the target 2+1 mAb-scFv format antibodies are shown in Figures 15, 16, 17, and 82.

[0298] In some embodiments of the 2+1 mAb-scFv form, the anti-CD28 ABD has VH and VL domains selected from the following: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof. VL having the amino acid sequences of R2 and vlCDR3, respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variants shown in Figure 15 or 16, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variants shown in Figure 15, 17, or 82, respectively.

[0299] In some embodiments of the 2+1 mAb-scFv form, the anti-CD28 ABD has VH and VL domains selected from: (i) VH having the amino acid sequence of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; and (ii) VL having the amino acid sequence of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; or (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 15 or 16; and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figures 15, 17, or 82.

[0300] The target 2+1 mAb-scFv format may include any preferred STEAP1 binding domain, including any of the STEAP1 binding domains provided herein. In some embodiments, the STEAP1 binding domain is one of the following STEAP1 binding domains or a variant thereof: H1 and L1, H1.1 and L1, H1.2 and L1, H1.3 and L1, H1.4 and L1, H1.5 and L1, H1.6 and L1, H1.7 and L1, H1.8 and L1, H1.9 and L1, H1.10 and L1, H1.11 and L1, H1.12 and L1, H1.13 and L1, H1.14 and L1, H1.15 and L1, H1.16 and L1, H1.17 and L1, H 1.18 and L1, H1.19 and L1, H1.20 and L1, H1.21 and L1, H1.22 and L1, H1.23 and L1, H1.24 and L1, H1.25 and L1, H1.26 and L1, H1.27 and L1, H1.28 and L1, H1.29 and L1, H1.30 and L1, H1.31 and L1, H1.32 and L1, H1.33 and L1, H1.34 and L1, H1.35 and L1, H1.36 and L1, H1.37 and L1, H1.38 and L1, H1 and L1.1, H1 and L1.2, H1 and and L1.3, H0 and L0, H1.21 and L1.3, H1.34 and L1.3, H1.39 and L1, H1.40 and L1, H1.41 and L1, H1.42 and L1, H1.43 and L1, H1.44 and L1, H1.45 and L1, H1.46 and L1, H1.47 and L1, H1.48 and L1, H1.49 and L1, H1.50 and L1, H1.51 and L1, H1.52 and L1, H1.53 and L1, H1.54 and L1, H1.55 and L1, H1.56 and L1, H1.57 and L1, H1. 58 and L1, H1.59 and L1, H1.60 and L1, H1.61 and L1, H1.62 and L1, H1.63 and L1, H1.64 and L1, H1.65 and L1, H1.66 and L1, H1.67 and L1, H1.68 and L1, H1.69 and L1, H1.70 and L1, H1.71 and L1, H1.72 and L1, H1.73 and L1, H1.74 and L1, H1.75 and L1, H1.76 and L1, H1.77 and L1, H1.78 and L1, H1.79 and L1, H1.80 and L1, H1.81 and L1, H1.82 and L1, H1.83 and L1, H1.84 and L1, H1.85 and L1, H1.86 and L1, H1.87 and L1, H1.88 and L1, H1.89 and L1, H1.90 and L1, H1.91 and L1, H1.92 and L1, H1.93 and L1, H1.94 and L1, H1.95 and L1, H1.96 and L1, H1.97 and L1, H1.98 and L1, H1.99 and L1, H1.100 and L1, H1.101 and L1, H1.102 and L1, H1.103 and L1, H1.104 and L1, H1.10 5 and L1, H1.106 and L1, H1.107 and L1, H1.108 and L1, H1.109 and L1, H1.110 and L1, H1.111 and L1, H1.112 and L1, H1.113 and L1, H1.114 and L1, H1.115 and L1, H1.116 and L1, H1.117 and L1, H1.118 and L1, H1.119 and L1, H1.120 and L1, H1.121 and L1, H1.122 and L1, H1.123 and L1, H1.124 and L1, H1.125 and L1, H1.126 and L1, H1.127 and L1, H 1.128 and L1, H1.129 and L1, H1.130 and L1, H1.131 and L1, H1.132 and L1, H1.133 and L1, H1.134 and L1, H1.135 and L1, H1.136 and L1, H1.137 and L1, H1.138 and L1, H1.139 and L1, H1.140 and L1, H1.141 and L1, H1.142 and L1, H1.143 and L1, H1.144 and L1, H1.145 and L1, H1.146 and L1, H1.147 and L1, H1.148 and L1, H1.149 and L1, H1.150 and and L1, H1.151 and L1, H1.152 and L1, H1.153 and L1, H1.154 and L1, H1.155 and L1, H1.156 and L1, H1.157 and L1, H1.158 and L1, H1.159 and L1, H1.160 and L1, H1.161 and L1, H1.162 and L1, H1.163 and L1, H1.164 and L1, H1.165 and L1, H1.166 and L1, H1.167 and L1, H1.168 and L1, H1.169 and L1, H1.170 and L1, H1.171 and L1, H1.172 and L1, H1.173 and L1, H1.174 and L1, H1.175 and L1, H1.176 and L1, H1.177 and L1, H1.178 and L1, H1.179 and L1, H1.180 and L1, H1.181 and L1, H1.182 and L1, H1.183 and L1, H1.184 and L1, H1.185 and L1, H1.186 and L1, H1.187 and L1, H1.188 and L1, H1.189 and L1, H1 and L1.4, H1 and L1.5, H1 and L1.6, H1 and L1.7, H1 and L1.8, H1 and L1.9, H1 and L1.10, H1 and L1.11, H1 and L1.12, H1 and L1.13, H1 and L1.14, H1 and L1.15, H1 and L1.16, H1 and L1.17, H1 and L1.18, H1 and L1.19, H1 and L1.20, H1 and L1.21, H1 and L1.22, H1 and L1.23, H1 and L1.24, H1 and L1.25, H1 and L1.26, H1 and L1.27, H1 and L1.28, H1 and L1.29, H1 and L1.30, H1 and L1.31, H1 and L1.32, H1 and L1.33, H1 and L1.34, H1 and L1.35, H1 and L1.36, H1 and L1.37, H1 and L1.38, H1 and L1.39, H1 and L1.40, H1 and L1.41, H1 and L1.42, H1 and L1.43, H1 and L1.44, H1 and L1.45, H1 and L1.46, H1 and L1.47, H1 and L1.48, H1 and L1.49, H1 and L1.50, H1 and L1.51, H1 and L1.52, H1 and L1.53, H1 and L1.54, H1 and L1.55, H1 and L1.56, H1 and L1.57, H1 and L1.58, H1 and L1.59, H1 and L1.60, H1 and L1.61, H1 and L1.62, H1 and L1.63, H1 and L1.64, H1 and L1.65, H1 and L1.66, H1 and L1.67, H1 and L1.68, H1 and L1.69, H1 and L1.70, H1 and L1.71, H1 and L1.72, H1 and L1.73, H1 and L1.74, H1 and L1.75, H1 and L1.76, H1 and L1.77, H1 and L1.78, H1 and L1.79, H1 and L1.80, H1 and L1.81, H1 and L1.82, H1 and L1.83, H1 and L1.84, H1 and L1.85, H1.42 and L1.47, H1.42 and L1.56, H1.190 and L1.3, H1.191 and L1.3, H1.192 and L1.3, H1.195 and L1.3, H1.196 and L1.3, H1.197 and L1.3, H1.198 and L1.3, H1.199 and L1.3, H1.200 and L1.3, H1.201 and L1.3, H1.202 and L1.3, H1.203 and L1.3, H1.204 and L1.3, H1.205 and L1.3, H1.206 and L1.3, H1.207 and L1.3, H1.208 and L1.3, H1.209 and L1.3, H1.210 and L1.3, H1.211 and L1.3, H1.212 and L1.3, H1.213 and L1.3, H1.214 and L1.3, H1.215 and L1.3, H1.216 and L1.3, H1.217 and L1.3, H1.218 and L1.3, H1.219 and L1.3, H1.220 and L1.3, H1.221 and L1.3, H1.222 and L1.3, H1.223 and L1.3, H1.224 and L1.3, H1.225 and L1.3, H1.226 and L1.3, H1.227 and L1.3, H1.228 and L 1.3, H1.21 and L1.92, H1.21 and L1.93, H1.21 and L1.94, H1.21 and L1.95, H1.21 and L1.96, H1.21 and L1.97, H1.21 and L1.98, H1.21 and L1.99, H1.21 and L1.100, H1.21 and L1.101, H1.21 and L1.102, H1.21 and L1.103, H1.21 and L1.104, H1.21 and L1.105, H1.21 and L1.106, H1.21 and L1.107, H1.21 and L1.108, H1.21 and L1.109, H1 .21 and L1.110, H1.21 and L1.111, H1.21 and L1.112, H1.21 and L1.113, H1.21 and L1.114, H1.21 and L1.115, H1.21 and L1.116, H1.21 and L1.117, H1.21 and L1.118, H1.21 and L1.119, H1.195 and L1.92, H1.195 and L1.94, H1.195 and L1.95, H1.195 and L1.104, H1.195 and L1.105, H1.195 and L1.109, H1.196 and L1.92, H1.196 and L1.94, H1.196 and L1.95, H1.196 and L1.104, H1.196 and L1.105, H1.196 and L1.109, H1.205 and L1.92, H1.205 and L1.94, H1.205 and L1.95, H1.205 and L1.104, H1.205 and L1.105, H1.205 and L1 .109, H1.193 and L1.91, H1.194 and L1.86, H1.194 and L1.87, H1.194 and L1.88, H1.194 and L1.89, or H1.194 and L1.90, or H1.2320 and L1.100, or H1.230 and L1.115 (Figures 22, 31, 32, and 39). Additional VH and VL sequences of exemplary STEAP1 binding domains that can be used in the target 2+1 mAb-scFv format antibody are shown in Figures 22, 31, and 32.

[0301] In some embodiments of the 2+1 mAb-scFv format, the anti-STEAP1 ABD has VH and VL domains selected from the following: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the STEAP1 binding domains shown in Figure 22 or its variants; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and VL having the amino acid sequence of vlCDR3 respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variants shown in Figure 22 or Figure 31, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variants shown in Figure 22 or Figure 32, respectively.

[0302] In some embodiments of the 2+1 mAb-scFv format, the STEAP1 ABD has VH and VL domains selected from: (i) VH having the amino acid sequence of any of the STEAP1 binding domains shown in Figure 22 or a variant thereof; and (ii) VL having the amino acid sequence of any of the STEAP1 binding domains shown in Figure 22 or a variant thereof; or (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 22 or Figure 31; and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figure 22 or Figure 32.

[0303] Figure 10 shows several exemplary Fc domain sequences useful in 2+1 mAb-scFv format antibodies. Figures 11 and 12 also provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the first or second monomer of the 2+1 mAb-scFv format. Furthermore, Figure 13 provides useful CL sequences that may be used in this format.

[0304] 3.2+1 Fab2-scFv-Fc format One heterodimeric antibody form particularly utilized in the anti-CD28 × anti-STEAP1 antibodies provided herein is the 2+1 Fab2-scFv-Fc form (also referred to as the “central-scFv form”), shown in Figure 14B. This antibody form includes three antigen-binding domains: two Fab moieties and an scFv inserted between the VH-CH1 and CH2-CH3 regions of one of the monomers. In some embodiments of this form, each Fab moiety binds to STEAP1, and the “additional” scFv domain binds to CD28. In some embodiments, the 2+1 Fab2-scFv-Fc form antibody is a trivalent antibody.

[0305] In some embodiments of the 2+1 Fab2-scFv-Fc form, the first monomer includes a standard heavy chain (i.e., VH1-CH1-hinge-CH2-CH3), where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer includes another first variable weight domain (VH1), a CH1 domain (and an optional hinge), a second Fc domain, and an scFv containing an scFv variable light domain (VL2), an scFv linker, and an scFv variable weight domain (VH2). The scFv is covalently linked between the C-terminus of the CH1 domain of the second monomer and the N-terminus of the second Fc domain using an arbitrary domain linker (VH1-CH1-[arbitrary linker]-VH2-scFv linker-VH2-[arbitrary linker]-CH2-CH3, or the opposite orientation for scFv: VH1-CH1-[arbitrary linker]-VL2-scFv linker-VH2-[arbitrary linker]-CH2-CH3). The arbitrary linker can be any suitable peptide linker, including, for example, the domain linker shown in Figure 7. This embodiment further utilizes a common light chain comprising a variable light domain (VL1) and a constant light domain (CL). In some embodiments, the form comprises two identical common light chains (VL1-CL, respectively), each of which associates with one of the VH1-CH1 domains of the first and second monomers to form two identical Fabs. In some embodiments, each identical Fab is coupled to STEAP1, and scFv is coupled to CD28. As with many of the embodiments herein, these constructs may include, as desired and as described herein, scuba riants, pI variants, attenuation variants, additional Fc variants, etc.

[0306] In some embodiments, the second and second Fc domains of the 2+1 Fab2-scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 8). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0307] In some embodiments, the variant Fc domain includes a degraded variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes the degraded variant E233P / L234V / L235A / G236del / S267K, numbered according to EU numbering.

[0308] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants (including those shown in Figures 3 and 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0309] In some embodiments, the scFv of the 2+1 Fab2-scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0310] In exemplary embodiments, the 2+1 Fab2-scFv-Fc form antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236del / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 Fab2-scFv-Fc antibody provided herein includes a (GKPGS)4-charged scFv linker. In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, with numbering according to EU numbering.

[0311] In some embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0312] In some embodiments, the scFv of the second monomer of the 2+1 Fab2-scFv-Fc antibody is a CD28-binding domain, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a binding domain that binds to STEAP1. The 2+1 Fab2-scFv-Fc antibody of interest may contain any suitable CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28-binding domain is one of the following CD28-binding domains or a variant thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1 .71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD2 8]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[ CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN2 28[CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[SS], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS], 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS] (Figures 15, 18, 21, and 65).

[0313] In some embodiments of the 2+1 Fab2-scFv-Fc form, the anti-CD28 ABD has VH and VL domains selected from the following: (i) VH comprising vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof; and (ii) VL comprising vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vhCDR2, and vhCDR3 of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65, or variants thereof. VL having the amino acid sequences of R2 and vlCDR3, respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variants shown in Figure 15 or 16, respectively, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variants shown in Figure 15, 17, or 82, respectively.

[0314] In some embodiments of the 2+1 Fab2-scFv-Fc form, the anti-CD28 ABD has VH and VL domains selected from: (i) VH having the amino acid sequence of any of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; and (ii) VL having the amino acid sequence of the CD28-binding domains shown in Figures 15, 18, 21, and 65 or a variant thereof; or (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 15 or 16; and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figures 15, 17, or 82.

[0315] In some embodiments, the VH1 of the first and second monomers and the VL1 of the common light chain of the 2+1 Fab2-scFv-Fc antibody form a binding domain that binds to STEAP1, respectively. The 2+1 Fab2-scFv-Fc of interest may contain any preferred STEAP1 binding domain, including any of the STEAP1 binding domains provided herein. In some embodiments, the STEAP1 binding domain is one of the following STEAP1 binding domains or a variant thereof: H1 and L1, H1.1 and L1, H1.2 and L1, H1.3 and L1, H1.4 and L1, H1.5 and L1, H1.6 and L1, H1.7 and L1, H1.8 and L1, H1.9 and L1, H1.10 and L1, H1.11 and L1, H1.12 and L1, H1.13 and L1, H1.14 and L1, H1.15 and L 1. H1.16 and L1, H1.17 and L1, H1.18 and L1, H1.19 and L1, H1.20 and L1, H1.21 and L1, H1.22 and L1, H1.23 and L1, H1.24 and L1, H1.25 and L1, H1.26 and L1, H1.27 and L1, H1.28 and L1, H1.29 and L1, H1.30 and L1, H1.31 and L1, H1.32 and L1, H1.33 and L1, H1.34 and L1, H1.35 and L1, H1.36 and and L1, H1.37 and L1, H1.38 and L1, H1 and L1.1, H1 and L1.2, H1 and L1.3, H0 and L0, H1.21 and L1.3, H1.34 and L1.3, H1.39 and L1, H1.40 and L1, H1.41 and L1, H1.42 and L1, H1.43 and L1, H1.44 and L1, H1.45 and L1, H1.46 and L1, H1.47 and L1, H1.48 and L1, H1.49 and L1, H1.50 and L1, H1.51 and L1, H1.52 and L1, H1.53 and L1, H1.54 and L1, H1.55 and L1, H1.56 and L1, H1.57 and L1, H1.58 and L1, H1.59 and L1, H1.60 and L1, H1.61 and L1, H1.62 and L1, H1.63 and L1, H1.64 and L1, H1.65 and L1, H1.66 and L1, H1.67 and L1, H1.68 and L1, H1.69 and L1, H1.70 and L1, H1.71 and L1, H1.72 and L1, H1.73 and L1, H1.74 and L1, H1.75 and L1, H1.76 and L1, H1.77 and L1, H1.78 and L1, H1.79 and L1, H1.80 and L1, H1.81 and L1, H1.82 and L1, H1.83 and L1, H1.84 and L1, H1.85 and L1, H1.86 and L1, H1.87 and L1, H1.88 and L1, H1.89 and L1, H1.90 and L1, H1.91 and L1, H1.92 and L1, H1.93 and L1, H1.94 and L1, H1.95 and L1, H1.96 and L1, H 1.97 and L1, H1.98 and L1, H1.99 and L1, H1.100 and L1, H1.101 and L1, H1.102 and L1, H1.103 and L1, H1.104 and L1, H1.105 and L1, H1.106 and L1, H1.107 and L1, H1.108 and L1, H1.109 and L1, H1.110 and L1, H1.111 and L1, H1.112 and L1, H1.113 and L1, H1.114 and L1, H1.115 and L1, H1.116 and L1, H1.117 and L1, H1.118 and L1, H1.119 and L1, H1.120 and L1, H1.121 and L1, H1.122 and L1, H1.123 and L1, H1.124 and L1, H1.125 and L1, H1.126 and L1, H1.127 and L1, H1.128 and L1, H1.129 and L1, H1.130 and L1, H1.131 and L1, H1.132 and L1, H1.133 and L1, H1.134 and L1, H1.135 and L1, H1.136 and L1, H1.137 and L1, H1.138 and L1, H1.139 and L1, H1.140 and L1, H1.141 and L1, H1.142 and and L1, H1.143 and L1, H1.144 and L1, H1.145 and L1, H1.146 and L1, H1.147 and L1, H1.148 and L1, H1.149 and L1, H1.150 and L1, H1.151 and L1, H1.152 and L1, H1.153 and L1, H1.154 and L1, H1.155 and L1, H1.156 and L1, H1.157 and L1, H1.158 and L1, H1.159 and L1, H1.160 and L1, H1.161 and L1, H1.162 and L1, H1.163 and L1, H1.164 and L1, H1.165 and L1, H1.166 and L1, H1.167 and L1, H1.168 and L1, H1.169 and L1, H1.170 and L1, H1.171 and L1, H1.172 and L1, H1.173 and L1, H1.174 and L1, H1.175 and L1, H1.176 and L1, H1.177 and L1, H1.178 and L1, H1.179 and L1, H1.180 and L1, H1.181 and L1, H1.182 and L1, H1.183 and L1, H1.184 and L1, H1.185 and L1, H1.186 and L1, H1.187 and L 1, H1.188 and L1, H1.189 and L1, H1 and L1.4, H1 and L1.5, H1 and L1.6, H1 and L1.7, H1 and L1.8, H1 and L1.9, H1 and L1.10, H1 and L1.11, H1 and L1.12, H1 and L1.13, H1 and L1.14, H1 and L1.15, H1 and L1.16, H1 and L1.17, H1 and L1.18, H1 and L1.19, H1 and L1.20, H1 and L1.21, H1 and L1.22, H1 and L1.23, H1 and L1.24, H1 and L1.25, H1 and L1.26, H1 and L1.27, H1 and L1.28, H1 and L1.29, H1 and L1.30, H1 and L1.31, H1 and L1.32, H1 and L1.33, H1 and L1.34, H1 and L1.35, H1 and L1.36, H1 and L1.37, H1 and L1.38, H1 and L1.39, H1 and L1.40, H1 and L1.41, H1 and L1.42, H1 and L1.43, H1 and L1.44, H1 and L1.45, H1 and L1.46, H1 and L1.47, H1 and L1.48, H1 and L1.49, H1 and L1.50, H1 and L1.5 1, H1 and L1.52, H1 and L1.53, H1 and L1.54, H1 and L1.55, H1 and L1.56, H1 and L1.57, H1 and L1.58, H1 and L1.59, H1 and L1.60, H1 and L1.61, H1 and L1.62, H1 and L1.63, H1 and L1.64, H1 and L1.65, H1 and L1.66, H1 and L1.67, H1 and L1.68, H1 and L1.69, H1 and L1.70, H1 and L1.71, H1 and L1.72, H1 and L1.73, H1 and L1.74, H1 and L1.75, H1 and L1.76, H1 and L1.77, H1 and L1.78, H1 and L1.79, H1 and L1.80, H1 and L1.81, H1 and L1.82, H1 and L1.83, H1 and L1.84, H1 and L1.85, H1.42 and L1.47, H1.42 and L1.56, H1.190 and L1.3, H1.191 and L1.3, H1.192 and L1.3, H1.195 and L1.3, H1.196 and L1.3, H1.197 and L1.3, H1.198 and L1.3, H1.199 and L1.3, H1.200 and L1.3, H1.201 and L1.3, H 1.202 and L1.3, H1.203 and L1.3, H1.204 and L1.3, H1.205 and L1.3, H1.206 and L1.3, H1.207 and L1.3, H1.208 and L1.3, H1.209 and L1.3, H1.210 and L1.3, H1.211 and L1.3, H1.212 and L1.3, H1.213 and L1.3, H1.214 and L1.3, H1.215 and L1.3, H1.216 and L1.3, H1.217 and L1.3, H1.218 and L1.3, H1.219 and L1.3, H1.220 and L1.3, H1.2 21 and L1.3, H1.222 and L1.3, H1.223 and L1.3, H1.224 and L1.3, H1.225 and L1.3, H1.226 and L1.3, H1.227 and L1.3, H1.228 and L1.3, H1.21 and L1.92, H1.21 and L1.93, H1.21 and L1.94, H1.21 and L1.95, H1.21 and L1.96, H1.21 and L1.97, H1.21 and L1.98, H1.21 and L1.99, H1.21 and L1.100, H1.21 and L1.101, H1.21 and L1.102, H1.2 1 and L1.103, H1.21 and L1.104, H1.21 and L1.105, H1.21 and L1.106, H1.21 and L1.107, H1.21 and L1.108, H1.21 and L1.109, H1.21 and L1.110, H1.21 and L1.111, H1.21 and L1.112, H1.21 and L1.113, H1.21 and L1.114, H1.21 and L1.115, H1.21 and L1.116, H1.21 and L1.117, H1.21 and L1.118, H1.21 and L1.119, H1.195 and L1.92, H1.195 and L1.94, H1.195 and L1.95, H1.195 and L1.104, H1.195 and L1.105, H1.195 and L1.109, H1.196 and L1.92, H1.196 and L1.94, H1.196 and L1.95, H1.196 and L1.104, H1.196 and L1.105, H1.196 and L1.109, H1.205 and L1.92, H1.205 and L1.94, H1.205 and L1.95, H1.205 and L1.104, H1.205 and L1.105, H1.205 and L1.109, H1.193 and L1.91, H1.194 and L1.86, H1.194 and L1.87, H1.194 and L1.88, H1.194 and L1.89, or H1.194 and L1.90, or H1.2320 and L1.100, or H1.230 and L1.115 (Figures 22, 31, 32, and 39). Additional VH and VL sequences of exemplary STEAP1 binding domains that can be used in the target 2+1 Fab2-scFv-Fc type antibody are shown in Figures 22, 31, and 32.

[0316] In some embodiments of the 2+1 Fab2-scFv-Fc form, the anti-STEAP1 ABD each comprises a VH and a VL domain selected from the following: (i) a VH comprising vhCDR1, vhCDR2, and vhCDR3, wherein the VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any of the STEAP1 binding domains shown in Figure 22 or variants thereof; and (ii) a VL comprising vlCDR1, vlCDR2, and vlCDR3, wherein the VL has the amino acid sequences of vlCDR1, vlCDR2, and vhCDR3 of any of the STEAP1 binding domains shown in Figure 22 or variants thereof. VL having the amino acid sequences of vhCDR1 and vhCDR3 respectively; or (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of VH or its variant shown in Figure 22 or Figure 31, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of VL or its variant shown in Figure 22 or Figure 32.

[0317] In some embodiments of the 2+1 Fab2-scFv-Fc form, each STEAP1 ABD comprises a VH and a VL domain selected from: (i) a VH having the amino acid sequence of any VH or any variant thereof of the STEAP1 binding domain shown in Figure 22; and (ii) a VL having the amino acid sequence of any VL or any variant thereof of the STEAP1 binding domain shown in Figure 22; or (i) a VH having the amino acid sequence of any VH or any variant thereof shown in Figure 22 or Figure 31; and (ii) a VL having the amino acid sequence of any VL or any variant thereof shown in Figure 22 or Figure 32.

[0318] Figure 9 shows several exemplary Fc domain sequences useful in 2+1 Fab2-scFv-Fc type antibodies. Figures 11 and 12 also provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the second or third monomer of the 2+1 Fab2-scFv-Fc form. Furthermore, Figure 13 provides useful CL sequences that may be used in this form.

[0319] An exemplary anti-CD28 × anti-STEAP1 bispecific antibody in the 2+1 Fab2-scFv-Fc format is shown in Figure 40.

[0320] 4.2+1 Fab2-Fc×scFv-Fc format One heterodimer antibody configuration particularly utilized in the anti-CD28 × anti-STEAP1 antibodies provided herein is the 2+1 Fab2-Fc × scFv-Fc configuration (also known as the “stacked bottle opener”) shown in Figure 14F. This configuration comprises a first monomer, a second monomer, and a common light chain. The first monomer contains VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first variant Fc domain. The domain linker can be any useful domain linker (see, for example, Figure 7). The second monomer contains a single-stranded Fv ("scFv") covalently linked to the second variant Fc domain by the domain linker (scFv-domain linker-CH2-CH3). The common light chain comprises VL1-CL, where VL1 is the first variable light domain. The second monomer scFv comprises a second variable heavy domain (VH2) linked to the second variable light domain (VL2) by an scFv linker. In some embodiments, this form comprises two identical common light chains (VL1-CL). In this embodiment, the two VH1-CH1 of the first monomer each interact with the common light chain to form two identical Fabs, which are the first antigen-binding domains, and VH2 and VL2 form the second antigen-binding domain. In some embodiments, each of the first antigen-binding domains is a STEAP1-binding domain, and the second antigen-binding domain is a CD28-binding domain. As with many of the embodiments herein, these constructs may include scuba riants, pI variants, attenuation variants, additional Fc variants, etc., as desired and as described herein.

[0321] In some embodiments, the second and second Fc domains of the 2+1 Fab2-Fc×scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 8). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0322] In some embodiments, the variant Fc domain includes a degraded variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes the degraded variant E233P / L234V / L235A / G236del / S267K, numbered according to EU numbering.

[0323] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants (including those shown in Figures 3 and 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0324] In some embodiments, the 2+1 Fab2-Fc×scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0325] In exemplary embodiments, the 2+1 Fab2-Fc×scFv-Fc antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236del / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 Fab2-Fc×scFv-Fc antibody provided herein includes a (GKPGS) 4-charged scFv linker (SEQ ID NO: XX). In some embodiments, the 2+1 Fab2-FcxscFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, with numbering according to EU numbering.

[0326] In some embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0327] In some embodiments, the scFv of the second monomer of the 2+1 Fab2-Fc×scFv-Fc antibody is a CD28-binding domain, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a binding domain that binds to STEAP1. The 2+1 Fab2-scFv-Fc antibody of interest may contain any preferred CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28-binding domain is one of the following CD28-binding domains or a variant thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1 .71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD2 8]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L...

Claims

1. A heterodimer antibody, a) The first monomer, i) Single-stranded variable fragments (scFv); and ii) The first monomer comprising the first Fc domain, wherein the scFv is covalently bonded to the N-terminus of the first Fc domain using a domain linker; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; and c) A light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 and VL1 together form a first antigen-binding domain (ABD), and the VH2 and VL2 together form a second ABD. The heterodimer antibody wherein one of the first ABD and the second ABD is a CD28-binding domain, and the other of the first ABD and the second ABD is a 6-transmembrane prostatic epithelial antigen 1 (STEAP1)-binding domain.

2. The heterodimer antibody according to claim 1, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

3. The heterodimer antibody according to claim 1, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

4. The heterodimer antibody according to any one of claims 1 to 3, wherein the first ABD is the STEAP1 binding domain and the second ABD is the CD28 binding domain.

5. The heterodimer antibody according to claim 4, wherein VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1 binding domain in Figure 22 or 39, or a variant thereof, b) VH in Figure 31 and VL in Figure 32, or their variants.

6. The heterodimer antibody according to claim 4 or 5, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, and 21, or their variants, (2) VH or its variant in Figure 15 or 16 and (ii) VL or its variant in Figure 15, 17, or 82.

7. A heterodimer antibody according to any one of claims 1 to 6, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

8. The heterodimer antibody according to claim 7, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

9. The heterodimer antibody according to claim 8, wherein the first and second Fc domains comprise the heterodimerized scuba riant S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

10. A heterodimer antibody according to any one of claims 7 to 9, wherein each of the first and second Fc domains comprises one or more attenuation variants.

11. The heterodimer antibody according to claim 10, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

12. A heterodimer antibody according to any one of claims 7 to 11, wherein one of the first or second monomers further comprises one or more pI variants.

13. The heterodimer antibody according to claim 12, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

14. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 7 to 13, wherein the numbering follows EU numbering.

15. The heterodimer antibody according to any one of claims 8 to 14, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

16. The heterodimer antibody according to any one of claims 1 to 15, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

17. A heterodimer antibody, a) A first monomer comprising VH1-CH1-first domain linker-scFv-second domain linker-CH2-CH3 from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a six-transmembrane prostatic epithelial antigen 1 (STEAP1) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a STEAP1 binding domain.

18. The heterodimer antibody according to claim 17, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

19. The heterodimer antibody according to claim 17, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

20. The heterodimer antibody according to any one of claims 17 to 19, wherein the first ABD is the STEAP1 binding domain and the second ABD is the CD28 binding domain.

21. The heterodimer antibody according to claim 20, wherein VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1 binding domain in Figure 22 or 39, or a variant thereof, b) VH in Figure 31 and VL in Figure 32, or their variants.

22. The heterodimer antibody according to claim 20 or 21, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, and 21, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

23. The heterodimer antibody according to any one of claims 17 to 22, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

24. The heterodimer antibody according to claim 23, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

25. The heterodimer antibody according to claim 24, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

26. A heterodimer antibody according to any one of claims 23 to 25, wherein each of the first and second Fc domains comprises one or more attenuation variants.

27. The heterodimer antibody according to claim 26, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

28. A heterodimer antibody according to any one of claims 23 to 27, wherein one of the first or second monomers further comprises one or more pI variants.

29. The heterodimer antibody according to claim 28, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

30. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 23 to 29, wherein the numbering follows EU numbering.

31. The heterodimer antibody according to any one of claims 24 to 30, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

32. The heterodimer antibody according to any one of claims 17 to 31, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

33. A heterodimer antibody, a) A first monomer comprising VH1-CH1-hinge-CH2-CH3-domain linker-scFv from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a six-transmembrane prostatic epithelial antigen 1 (STEAP1) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a STEAP1 binding domain.

34. The heterodimer antibody according to claim 33, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

35. The heterodimer antibody according to claim 33, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

36. The heterodimer antibody according to any one of claims 33 to 35, wherein the first ABD is the STEAP1 binding domain and the second ABD is the CD28 binding domain.

37. The heterodimer antibody according to claim 36, wherein VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1 binding domain in Figure 22 or 39, or a variant thereof, b) VH in Figure 31 and VL in Figure 32, or their variants.

38. The heterodimer antibody according to claim 36 or 37, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, and 21, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

39. The heterodimer antibody according to any one of claims 33 to 38, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

40. The heterodimer antibody according to claim 39, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

41. The heterodimer antibody according to claim 40, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

42. A heterodimer antibody according to any one of claims 39 to 41, wherein each of the first and second Fc domains comprises one or more attenuation variants.

43. The heterodimer antibody according to claim 42, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

44. A heterodimer antibody according to any one of claims 39 to 43, wherein one of the first or second monomers further comprises one or more pI variants.

45. The heterodimer antibody according to claim 44, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

46. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 39 to 45, wherein the numbering follows EU numbering.

47. The heterodimer antibody according to any one of claims 40 to 46, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

48. The heterodimer antibody according to any one of claims 33 to 47, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

49. A heterodimer antibody, a) A first monomer comprising VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, Each of the VH1 is a first variable weight domain, and CH2-CH3 is a first Fc domain, in the first monomer; b) The second monomer comprising scFv-domain linker-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a six-transmembrane prostatic epithelial antigen 1 (STEAP1) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a STEAP1 binding domain.

50. The heterodimer antibody according to claim 49, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

51. The heterodimer antibody according to claim 49, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

52. The heterodimer antibody according to any one of claims 49 to 51, wherein the first ABD is the STEAP1 binding domain and the second ABD is the CD28 binding domain.

53. The heterodimer antibody according to claim 52, wherein VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1 binding domain in Figure 22 or 39, or a variant thereof, b) VH in Figure 31 and VL in Figure 32, or their variants.

54. The heterodimer antibody according to claim 52 or 53, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, and 21, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

55. A heterodimer antibody according to any one of claims 49 to 54, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

56. The heterodimer antibody according to claim 55, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

57. The heterodimer antibody according to claim 56, wherein the first and second Fc domains comprise the heterodimerized scuba riant S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

58. A heterodimer antibody according to any one of claims 55 to 57, wherein each of the first and second Fc domains comprises one or more attenuation variants.

59. The heterodimer antibody according to claim 58, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

60. A heterodimer antibody according to any one of claims 55 to 59, wherein one of the first or second monomers further comprises one or more pI variants.

61. The heterodimer antibody according to claim 60, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

62. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 55 to 61, wherein the numbering follows EU numbering.

63. The heterodimer antibody according to any one of claims 56 to 62, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

64. The heterodimer antibody according to any one of claims 49 to 63, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

65. a) i) a first variable heavy domain (VH1) and ii) a first variable light domain (VL1), comprising a 6-transmembrane prostatic epithelial antigen 1 (STEAP1) binding domain, b) A bispecific antibody comprising an anti-CD28 binding domain containing i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2).

66. The bispecific antibody according to claim 65, wherein VH1 and VL1 are selected from one of the following: a) VH and VL of either the STEAP1 binding domain in Figure 22 or 39, or a variant thereof, b) VH in Figure 31 and VL in Figure 32, or their variants.

67. The bispecific antibody according to claim 68 or 69, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, and 21, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

68. The bispecific antibody according to any one of claims 65 to 67, wherein the bispecific antibody further comprises a first Fc domain and a second Fc domain.

69. The bispecific antibody according to claim 68, wherein the first and second Fc domains comprise a set of heterodimerized scuba ariants selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

70. The bispecific antibody according to claim 69, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

71. A bispecific antibody according to any one of claims 68 to 70, wherein each of the first and second Fc domains comprises one or more attenuation variants.

72. The bispecific antibody according to claim 71, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

73. The bispecific antibody according to any one of claims 68 to 72, wherein one of the first or second monomers further comprises one or more pI variants.

74. The bispecific antibody according to claim 73, wherein the pI variant is N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

75. a) A first nucleic acid encoding a first monomer according to any one of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64; b) A second nucleic acid encoding a second monomer according to any one of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64; and c) A nucleic acid composition comprising a third nucleic acid encoding a light chain as described in any of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64.

76. a) A first expression vector comprising the first nucleic acid according to claim 75; b) A second expression vector comprising the second nucleic acid described in claim 75; and c) An expression vector composition comprising a third expression vector each containing the third nucleic acid described in claim 75.

77. A host cell comprising the expression vector according to claim 76.

78. A method for producing a heterodimer antibody according to any one of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64, comprising culturing a host cell according to claim F3 under conditions in which the heterodimer antibody is expressed, and recovering the heterodimer antibody.

79. A method for treating STEAP-associated cancer in a patient requiring treatment for STEAP-associated cancer, comprising administering to the patient a heterodimer antibody according to any one of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64.

80. A method for treating STEAP-associated cancer in a patient requiring treatment for STEAP-associated cancer, comprising administering to the patient a heterodimer antibody and an anti-CD3 × anti-STEAP1 bispecific antibody according to any one of claims 1 to 16, 17 to 32, 33 to 38, or 49 to 64.

81. A method for treating STEAP-related cancer in a patient requiring treatment for STEAP-related cancer, comprising administering a bispecific antibody according to any one of claims 65 to 74 to the patient.

82. A method for treating STEAP-related cancer in a patient requiring treatment for STEAP-related cancer, comprising administering to the patient a bispecific antibody and an anti-CD3 × anti-STEAP1 bispecific antibody according to any one of claims 65 to 74.

83. A composition comprising a STEAP1 antigen-binding domain, wherein the STEAP1-binding domain is a) A variable weight domain having vhCDR1-3 of either the STEAP1 binding domain variable weight domain in Figure 22 or Figure 31; and b) The composition comprising a variable light domain having vlCDR1 to vlCDR1 of either the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

84. A composition comprising a STEAP1 antigen-binding domain, wherein the STEAP1-binding domain is a) A variable weight domain having at least 85% sequence identity with the STEAP1 binding domain variable weight domain in Figure 22 or Figure 31; and b) The composition comprising a variable light domain having at least 85% sequence identity with the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

85. The composition according to claim 84, wherein the variable heavy domain has at least 90% sequence identity with the STEAP1 binding domain variable heavy domain of Figure 22 or Figure 31, and the variable light domain has at least 90% sequence identity with the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

86. The composition according to claim 84, wherein the variable heavy domain has at least 95% sequence identity with the STEAP1 binding domain variable heavy domain of Figure 22 or Figure 31, and the variable light domain has at least 95% sequence identity with the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

87. The composition according to claim 84, wherein the variable heavy domain has at least 99% sequence identity with the STEAP1 binding domain variable heavy domain of Figure 22 or Figure 31, and the variable light domain has at least 99% sequence identity with the STEAP1 binding domain variable light domain of Figure 22 or Figure 32.

88. The composition according to claim 84, wherein the variable heavy domain has the amino acid sequence of the STEAP1 binding domain variable heavy domain shown in Figure 22 or Figure 31, and the variable light domain has the amino acid sequence of the STEAP1 binding domain variable light domain shown in Figure 22 or Figure 32.

89. The aforementioned STEAP1 binding domain, The following STEAP1 binding domain variable weight domains and STEAP1 binding domain variable light domains are selected: H1 and L1, H1.1 and L1, H1.2 and L1, H1.3 and L1, H1.4 and L1, H1.5 and L1, H1.6 and L1, H1.7 and L1, H1.8 and L1, H1.9 and L1, H1.10 and L1, H1.11 and L1, H1.12 and L1, H1.13 and L1, H1.14 and L1, H1.15 and L1, H1.16 and L1, H1.17 and L1, H1.18 and L1, H1.19 and L1, H1.20 and L1, H 1.21 and L1, H1.22 and L1, H1.23 and L1, H1.24 and L1, H1.25 and L1, H1.26 and L1, H1.27 and L1, H1.28 and L1, H1.29 and L1, H1.30 and L1, H1.31 and L1, H1.32 and L1, H1.33 and L1, H1.34 and L1, H1.35 and L1, H1.36 and L1, H1.37 and L1, H1.38 and L1, H1 and L1.1, H1 and L1.2, H1 and L1.3, H0 and L0, H1.21 and L1.3, H1.34 and L1.3, H1.39 and L1, H1 . 40 and L1, H1.41 and L1, H1.42 and L1, H1.43 and L1, H1.44 and L1, H1.45 and L1, H1.46 and L1, H1.47 and L1, H1.48 and L1, H1.49 and L1, H1.50 and L1, H1.51 and L1, H1.52 and L1, H1.53 and L1, H1.54 and L1, H1.55 and L1, H1.56 and L1, H1.57 and L1, H1.58 and L1, H1.59 and L1, H1.60 and L1, H1.61 and L1, H1.62 and L1, H1.63 and L1, H1.64 and L1, H 1.65 and L1, H1.66 and L1, H1.67 and L1, H1.68 and L1, H1.69 and L1, H1.70 and L1, H1.71 and L1, H1.72 and L1, H1.73 and L1, H1.74 and L1, H1.75 and L1, H1.76 and L1, H1.77 and L1, H1.78 and L1, H1.79 and L1, H1.80 and L1, H1.81 and L1, H1.82 and L1, H1.83 and L1, H1.84 and L1, H1.85 and L1, H1.86 and L1, H1.87 and L1, H1.88 and L1, H1.89 and L1,H1.90 and L1, H1.91 and L1, H1.92 and L1, H1.93 and L1, H1.94 and L1, H1.95 and L1, H1.96 and L1, H1.97 and L1, H1.98 and L1, H1.99 and L1, H1.100 and L1, H1.101 and L1, H1.102 and L1, H1.103 and L1, H1.104 and L1, H1.105 and L1, H1.106 and L1, H1.107 and L1, H1.108 and L1, H1.109 and L1, H1.110 and L1, H1.111 and L1, H1.112 and L1, H1.1 13 and L1, H1.114 and L1, H1.115 and L1, H1.116 and L1, H1.117 and L1, H1.118 and L1, H1.119 and L1, H1.120 and L1, H1.121 and L1, H1.122 and L1, H1.123 and L1, H1.124 and L1, H1.125 and L1, H1.126 and L1, H1.127 and L1, H1.128 and L1, H1.129 and L1, H1.130 and L1, H1.131 and L1, H1.132 and L1, H1.133 and L1, H1.134 and L1, H1.135 and L1 , H1.136 and L1, H1.137 and L1, H1.138 and L1, H1.139 and L1, H1.140 and L1, H1.141 and L1, H1.142 and L1, H1.143 and L1, H1.144 and L1, H1.145 and L1, H1.146 and L1, H1.147 and L1, H1.148 and L1, H1.149 and L1, H1.150 and L1, H1.151 and L1, H1.152 and L1, H1.153 and L1, H1.154 and L1, H1.155 and L1, H1.156 and L1, H1.157 and L1, H1.15 8 and L1, H1.159 and L1, H1.160 and L1, H1.161 and L1, H1.162 and L1, H1.163 and L1, H1.164 and L1, H1.165 and L1, H1.166 and L1, H1.167 and L1, H1.168 and L1, H1.169 and L1, H1.170 and L1, H1.171 and L1, H1.172 and L1, H1.173 and L1, H1.174 and L1, H1.175 and L1, H1.176 and L1, H1.177 and L1, H1.178 and L1, H1.179 and L1, H1.180 and L1,H1.181 and L1, H1.182 and L1, H1.183 and L1, H1.184 and L1, H1.185 and L1, H1.186 and L1, H1.187 and L1, H1.188 and L1, H1.189 and L1, H1 and L1.4, H1 and L1.5, H1 and L1.6, H1 and L1.7, H1 and L1.8, H1 and L1.9, H1 and L1.10, H1 and L1.11, H1 and L1.12, H1 and L1.13, H1 and L1.14, H1 and L1.15, H1 and L1.16, H1 and L1.17, H1 and L1.18, H1 and L1.19, H1 and L1.20, H1 and L1.21, H1 and L1.22, H1 and L1.23, H1 and L1.24, H1 and L1.25, H1 and L1.26, H1 and L1.27, H1 and L1.28, H1 and L1.29, H1 and L1.30, H1 and L1.31, H1 and L1.32, H1 and L1.33, H1 and L1.34, H1 and L1.35, H1 and L1.36, H1 and L1.37, H1 and L1.38, H1 and L1.39, H1 and L1.40, H1 and L1.41, H1 and L1.42, H1 and L1.43, H1 and L1.44, H1 and L1.45, H1 and L1.46, H1 and L1.47, H1 and L1.48, H1 and L1.49, H1 and L1.50, H1 and L1.51, H1 and L1.52, H1 and L1.53, H1 and L1.54, H1 and L1.55, H1 and L1.56, H1 and L1.57, H1 and L1.58, H1 and L1.59, H1 and L1.60, H1 and L1.61, H1 and L1.62, H1 and L1.63, H1 and L1.64, H1 and L1.65, H1 and L1.66, H1 and L1.67, H1 and L1.68, H1 and L1.69, H1 and L1.70, H1 and L1.71, H1 and L1.72, H1 and L1.73, H1 and L1.74, H1 and L1.75, H1 and L1.76, H1 and L1.77, H1 and L1.78, H1 and L1.79, H1 and L1.80, H1 and L1.81, H1 and L1.82, H1 and L1.83, H1 and L1.84, H1 and L1.85, H1.42 and L1.47, H1.42 and L1.56, H1.190 and L1.3, H1.191 and L1.3, H1.192 and L1.3, H1.195 and L1.3,H1.196 and L1.3, H1.197 and L1.3, H1.198 and L1.3, H1.199 and L1.3, H1.200 and L1.3, H1.201 and L1.3, H1.202 and L1.3, H1.203 and L1.3, H1.204 and L1.3, H1.205 and L1.3, H1.206 and L1.3, H1.207 and L1.3, H1.208 and L1.3, H1.209 and L1.3, H1.210 and L1.3, H1.211 and L1.3, H1.212 and L1.3, H1.213 and L1.3, H1.214 and L1.3, H1. 215 and L1.3, H1.216 and L1.3, H1.217 and L1.3, H1.218 and L1.3, H1.219 and L1.3, H1.220 and L1.3, H1.221 and L1.3, H1.222 and L1.3, H1.223 and L1.3, H1.224 and L1.3, H1.225 and L1.3, H1.226 and L1.3, H1.227 and L1.3, H1.228 and L1.3, H1.21 and L1.92, H1.21 and L1.93, H1.21 and L1.94, H1.21 and L1.95, H1.21 and L1.96, H1.21 and and L1.97, H1.21 and L1.98, H1.21 and L1.99, H1.21 and L1.100, H1.21 and L1.101, H1.21 and L1.102, H1.21 and L1.103, H1.21 and L1.104, H1.21 and L1.105, H1.21 and L1.106, H1.21 and L1.107, H1.21 and L1.108, H1.21 and L1.109, H1.21 and L1.110, H1.21 and L1.111, H1.21 and L1.112, H1.21 and L1.113, H1.21 and L1.114, H1.21 and L1.115, H1.21 and L1.116, H1.21 and L1.117, H1.21 and L1.118, H1.21 and L1.119, H1.195 and L1.92, H1.195 and L1.94, H1.195 and L1.95, H1.195 and L1.104, H1.195 and L1.105, H1.195 and L1.109, H1.196 and L1.92, H1.196 and L1.94, H1.196 and L1.95, H1.196 and L1.104, H1.196 and L1.105, H1.196 and L1.109, H1.205 and L1.92,The composition according to claim 83, comprising H1.205 and L1.94, H1.205 and L1.95, H1.205 and L1.104, H1.205 and L1.105, H1.205 and L1.109, H1.193 and L1.91, H1.194 and L1.86, H1.194 and L1.87, H1.194 and L1.88, H1.194 and L1.89, or H1.194 and L1.90, or H1.2320 and L1.100, or H1.230 and L1.115 (Figures 22, 31, 32, and 39).

90. A heterodimer antibody, a) The first monomer, i) Single-stranded variable fragments (scFv); and ii) The first monomer comprising the first Fc domain, wherein the scFv is covalently bonded to the N-terminus of the first Fc domain using a domain linker; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; and c) A light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 and VL1 together form a first antigen-binding domain (ABD), and the VH2 and VL2 together form a second ABD. The heterodimer antibody wherein one of the first ABD and the second ABD is a CD28-binding domain, and the other of the first ABD and the second ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5)-binding domain.

91. The heterodimer antibody according to claim 90, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

92. The heterodimer antibody according to claim 90, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

93. The heterodimer antibody according to any one of claims 90 to 92, wherein the first ABD is the CEACAM5 binding domain and the second ABD is the CD28 binding domain.

94. VH1 and VL1 are a) VH and VL of any of the CEACAM5 binding domains in Figure 41 or their variants, b) The heterodimer antibody according to claim 93, which is VH and VL or a variant thereof as shown in Figure 45.

95. The heterodimer antibody according to claim 93 or 94, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, 21, and 65, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

96. A heterodimer antibody according to any one of claims 90 to 95, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

97. The heterodimer antibody according to claim 96, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

98. The heterodimer antibody according to claim 97, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

99. A heterodimer antibody according to any one of claims 96 to 98, wherein each of the first and second Fc domains comprises one or more attenuation variants.

100. The heterodimer antibody according to claim 99, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

101. A heterodimer antibody according to any one of claims 96 to 100, wherein one of the first or second monomers further comprises one or more pI variants.

102. The heterodimer antibody according to claim 101, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

103. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 96 to 102, wherein the numbering follows EU numbering.

104. The heterodimer antibody according to any one of claims 97 to 103, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

106. The heterodimer antibody according to any one of claims 90 to 104, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

107. A heterodimer antibody, a) A first monomer comprising VH1-CH1-first domain linker-scFv-second domain linker-CH2-CH3 from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

108. The heterodimer antibody according to claim 107, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

109. The heterodimer antibody according to claim 107, wherein the scFv comprises a VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

110. The heterodimer antibody according to any one of claims 107 to 109, wherein the first ABD is the CEACAM5 binding domain and the second ABD is the CD28 binding domain.

111. The heterodimer antibody according to claim 110, wherein VH1 and VL1 are selected from the following: a) VH and VL of any of the CEACAM5 binding domains in Figure 41 or their variants, b) VH and VL or their variants in Figure 45.

112. The heterodimer antibody according to claim 110 or 111, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, 21, and 65, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

113. A heterodimer antibody according to any one of claims 107 to 112, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

114. The heterodimer antibody according to claim 113, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

115. The heterodimer antibody according to claim 114, wherein the first and second Fc domains comprise the heterodimerized scuba riant S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

116. A heterodimer antibody according to any one of claims 113 to 115, wherein each of the first and second Fc domains comprises one or more attenuation variants.

117. The heterodimer antibody according to claim 116, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

118. A heterodimer antibody according to any one of claims 113 to 117, wherein one of the first or second monomers further comprises one or more pI variants.

119. The heterodimer antibody according to claim 118, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

120. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 113 to 119, wherein the numbering follows EU numbering.

121. The heterodimer antibody according to any one of claims 114 to 120, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

122. The heterodimer antibody according to any one of claims 107 to 121, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

123. A heterodimer antibody, a) A first monomer comprising VH1-CH1-hinge-CH2-CH3-domain linker-scFv from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

124. The heterodimer antibody according to claim 123, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

125. The heterodimer antibody according to claim 123, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

126. The heterodimer antibody according to any one of claims 123 to 125, wherein the first ABD is the CEACAM5 binding domain and the second ABD is the CD28 binding domain.

127. The heterodimer antibody according to claim 126, wherein VH1 and VL1 are selected from the following: a) VH and VL of any of the CEACAM5 binding domains in Figure 41 or their variants, b) VH and VL or their variants in Figure 45.

128. The heterodimer antibody according to claim 126 or 127, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, 21, and 65, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

129. A heterodimer antibody according to any one of claims 123 to 128, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

130. The heterodimer antibody according to claim 129, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

131. The heterodimer antibody according to claim 130, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

132. A heterodimer antibody according to any one of claims 129 to 131, wherein each of the first and second Fc domains comprises one or more attenuation variants.

133. The heterodimer antibody according to claim 132, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

134. A heterodimer antibody according to any one of claims 129 to 133, wherein one of the first or second monomers further comprises one or more pI variants.

135. The heterodimer antibody according to claim 134, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

136. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 129 to 135, wherein the numbering follows EU numbering.

137. The heterodimer antibody according to any one of claims 130 to 136, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

138. The heterodimer antibody according to any one of claims 123 to 137, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

139. A heterodimer antibody, a) A first monomer comprising VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, Each of the VH1 is a first variable weight domain, and CH2-CH3 is a first Fc domain, in the first monomer; b) The second monomer comprising scFv-domain linker-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein the first ABD is a carcinoembryonic antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain and the second ABD is a CD28 binding domain, or the first ABD is a CD28 binding domain and the second ABD is a CEACAM5 binding domain.

140. The heterodimer antibody according to claim 139, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

141. The heterodimer antibody according to claim 139, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

142. The heterodimer antibody according to any one of claims 139 to 141, wherein the first ABD is the CEACAM5 binding domain and the second ABD is the CD28 binding domain.

143. The heterodimer antibody according to claim 142, wherein VH1 and VL1 are selected from the following: a) VH and VL of any of the CEACAM5 binding domains in Figure 41 or their variants, b) VH and VL or their variants in Figure 45.

144. The heterodimer antibody according to claim 142 or 143, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, 21, and 65, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

145. A heterodimer antibody according to any one of claims 139 to 144, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

146. The heterodimer antibody according to claim 145, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

147. The heterodimer antibody according to claim 146, wherein the first and second Fc domains comprise the heterodimerized scuba riant S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

148. A heterodimer antibody according to any one of claims 145 to 147, wherein each of the first and second Fc domains comprises one or more attenuation variants.

149. The heterodimer antibody according to claim 148, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

150. A heterodimer antibody according to any one of claims 145 to 149, wherein one of the first or second monomers further comprises one or more pI variants.

151. The heterodimer antibody according to claim 150, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

152. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 145 to 151, wherein the numbering follows EU numbering.

153. A heterodimer antibody according to any one of claims 146 to 152, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

154. The heterodimer antibody according to any one of claims 139 to 153, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS.

155. a) i) a first variable heavy domain (VH1) and ii) a first variable light domain (VL1) comprising a fetal antigen-associated cell adhesion molecule 5 (CEACAM5) binding domain, b) A bispecific antibody comprising an anti-CD28 binding domain containing i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2).

156. The bispecific antibody according to claim 155, wherein VH1 and VL1 are selected from the following: a) VH and VL of any of the CEACAM5 binding domains in Figure 41 or their variants, b) VH and VL or their variants in Figure 45.

157. The bispecific antibody according to claim 158 or 159, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 15, 18, 21, and 65, or their variants, (2) (i) VH or a variant thereof as shown in Figure 15 or 16 and (ii) VL or a variant thereof as shown in Figure 15, 17, or 82.

158. The bispecific antibody according to any one of claims 155 to 157, wherein the bispecific antibody further comprises a first Fc domain and a second Fc domain.

159. The bispecific antibody according to claim 158, wherein the first and second Fc domains comprise a set of heterodimerized scuba ariants selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

160. The bispecific antibody according to claim 159, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

161. A bispecific antibody according to any one of claims 158 to 160, wherein each of the first and second Fc domains comprises one or more attenuation variants.

162. The bispecific antibody according to claim 161, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

163. A bispecific antibody according to any one of claims 158 to 162, wherein one of the first or second monomers further comprises one or more pI variants.

164. The bispecific antibody according to claim 163, wherein the pI variant is N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

165. a) A first nucleic acid encoding a first monomer according to any one of claims 90 to 106, 107 to 122, 123 to 138 and 139 to 154; b) A second nucleic acid encoding a second monomer according to any one of claims 90 to 106, 107 to 122, 123 to 138 and 139 to 154; and c) A nucleic acid composition comprising a third nucleic acid encoding a light chain as described in any of claims 90 to 106, 107 to 122, 123 to 138 and 139 to 154.

166. a) A first expression vector comprising the first nucleic acid described in claim 165; b) A second expression vector comprising the second nucleic acid described in claim 165; and c) An expression vector composition comprising a third expression vector each containing the third nucleic acid described in claim 165.

167. A host cell comprising the expression vector composition according to claim 166.

168. A method for producing a heterodimer antibody according to any one of claims 90 to 106, 107 to 122, 123 to 138 and 139 to 154, comprising culturing a host cell according to claim 167 under conditions in which the heterodimer antibody is expressed, and recovering the heterodimer antibody.

169. A method for treating CEACAM5-related cancer in a patient requiring treatment for CEACAM5-related cancer, comprising administering to the patient a heterodimer antibody according to any one of claims 190-106, 107-122, 123-138 and 139-154.

170. A method for treating CEACAM5-related cancer in a patient requiring treatment for CEACAM5-related cancer, comprising administering to the patient a heterodimer antibody and an anti-CD3 × anti-CEACAM5 bispecific antibody according to any one of claims 90 to 106, 107 to 122, 123 to 138 and 139 to 154.

171. A method for treating CEACAM5-related cancer in a patient requiring treatment for CEACAM5-related cancer, comprising administering a bispecific antibody according to any one of claims 155 to 168 to the patient.

172. A method for treating CEACAM5-related cancer in a patient requiring treatment for CEACAM5-related cancer, comprising administering to the patient a bispecific antibody and an anti-CD3 × anti-CEACAM5 bispecific antibody according to any one of claims 155 to 168.

173. A composition comprising a CEACAM5 antigen-binding domain, wherein the CEACAM5-binding domain is a) A variable weight domain having vhCDR1-3 of any of the CEACAM5 binding domain variable weight domains in Figure 45; and b) The composition comprising a variable light domain having vlCDR1 to 3 of any of the CEACAM5 binding domain variable light domains of Figure 45.

174. A composition comprising a CEACAM5 antigen-binding domain, wherein the CEACAM5-binding domain is a) A variable weight domain having at least 85% sequence identity with the CEACAM5 binding domain variable weight domain in Figure 45; and b) The composition comprising a variable light domain having at least 85% sequence identity with respect to the CEACAM5 binding domain variable light domain of Figure 45.

175. The composition according to claim 174, wherein the variable heavy domain has at least 90% sequence identity with the CEACAM5 binding domain variable heavy domain of Figure 45, and the variable light domain has at least 90% sequence identity with the CEACAM5 binding domain variable light domain of Figure 45.

176. The composition according to claim 174, wherein the variable heavy domain has at least 95% sequence identity with the CEACAM5 binding domain variable heavy domain of Figure 45, and the variable light domain has at least 95% sequence identity with the CEACAM5 binding domain variable light domain of Figure 45.

177. The composition according to claim 174, wherein the variable heavy domain has at least 99% sequence identity with the CEACAM5 binding domain variable heavy domain of Figure 45, and the variable light domain has at least 99% sequence identity with the CEACAM5 binding domain variable light domain of Figure 45.

178. The composition according to claim 174, wherein the variable heavy domain has the amino acid sequence of the CEACAM5 binding domain variable heavy domain shown in Figure 45, and the variable light domain has the amino acid sequence of the CEACAM5 binding domain variable light domain shown in Figure 45.

179. A composition comprising a CD28 antigen-binding domain, wherein the CD28-binding domain is a) A variable heavy domain having vhCDR1-3 of any of the CD28 binding domain variable heavy domains in Figure 65; and b) The composition comprising a variable light domain having vlCDR1 to 3 of any of the CD28 binding domain variable light domains of Figure 65.

180. A composition comprising a CD28 antigen-binding domain, wherein the CD28-binding domain is a) A variable weight domain having at least 85% sequence identity with the CD28 binding domain variable weight domain in Figure 65; and b) The composition comprising a variable light domain having at least 85% sequence identity with respect to the CD28-binding domain variable light domain of Figure 65.

181. The composition according to claim 180, wherein the variable heavy domain has at least 90% sequence identity with the CD28-binding domain variable heavy domain of Figure 65, and the variable light domain has at least 90% sequence identity with the CD28-binding domain variable light domain of Figure 65.

182. The composition according to claim 180, wherein the variable heavy domain has at least 95% sequence identity with the CD28-binding domain variable heavy domain of Figure 65, and the variable light domain has at least 95% sequence identity with the CD28-binding domain variable light domain of Figure 65.

183. The composition according to claim 180, wherein the variable heavy domain has at least 99% sequence identity with the CD28-binding domain variable heavy domain of Figure 65, and the variable light domain has at least 99% sequence identity with the CD28-binding domain variable light domain of Figure 65.

184. The composition according to claim 180, wherein the variable heavy domain has the amino acid sequence of the CD28 binding domain variable heavy domain shown in Figure 65, and the variable light domain has the amino acid sequence of the CD28 binding domain variable light domain shown in Figure 65.

185. The composition according to claim 180, wherein the CD28 binding domain is selected from the following: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3 [CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28] _H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, TN228[ CD28]_H4L2, 1A7[CD28]_H1.1_L1[SS], 1A7[CD28]_H1_L1.71[SS], 1A7[CD28]_H1.1_L1.71[S- S], 1A7[CD28]_H1.14_L1[SS], 1A7[CD28]_H1.14_L1.71[SS], 1A7[CD28]_H1.129_L1.71[SS] , 1A7[CD28]_H1.106_L1.71[SS], 1A7[CD28]_H1.129_L1[SS], and 1A7[CD28]_H1.106_L1[SS].

Images