Heterodimer antibodies that bind to ENPP3 and CD3

Abstract

To provide heterodimeric antibodies that bind ENPP3 and CD3.SOLUTION: The present invention is directed to antibodies, including novel antigen binding domains and heterodimeric antibodies, that bind Ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3).SELECTED DRAWING: Figure 15A

Description

[Technical Field]

[0001] Priority Claim This application claims priority to U.S. Provisional Application No. 62 / 812,922, filed on 1 March 2019, and U.S. Provisional Application No. 62 / 929,687, filed on 1 November 2019, both of which are incorporated herein by reference in their entirety. [Background technology]

[0002] Antibody-based therapies have been successfully used to treat a variety of diseases, including cancer. An increasingly popular avenue being explored is the engineering of a single immunoglobulin molecule that simultaneously binds to two different antigens. Such alternative antibody formats that bind to two different antigens are often referred to as bispecific antibodies. Due to the considerable diversity of antibody variable regions (Fv), it is possible to produce Fv that recognize virtually any molecule, so a typical approach to producing bispecific antibodies is to introduce a novel variable region into the antibody.

[0003] A particularly useful approach with bispecific antibodies involves manipulating the first binding domain that binds to CD3 and the second binding domain that binds to an antigen associated with or upregulated by cancer cells, thereby enabling the bispecific antibody to bind to CD3 + The goal is to redirect T cells to destroy cancer cells. Ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) has been previously reported to be highly expressed in renal cell carcinoma and minimally expressed in normal tissues. Considering this, anti-ENPP3 antibodies are thought to be useful, for example, for localizing antitumor agents (e.g., chemotherapeutic agents and T cells) to such ENPP3-expressing tumors. CD3 + Novel bispecific antibodies against CD3 and ENPP3 that can localize effector T cells to ENPP3-expressing tumors are provided herein. [Overview of the project]

[0004] Therefore, ENPP3 antigen-binding domains and anti-ENPP3 antibodies are provided herein (e.g., bispecific antibodies).

[0005] In one embodiment, the following ENPP3 binding domains are included: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Compositions comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain containing any of the following variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) and variable light chain complementarity determining regions (vlCDR1-3) are provided herein. In some embodiments, vhCDR1-3 and vlCDR1-3 are selected from the vhCDR1-3 and vlCDR1-3 sequences of ENPP3 binding domains provided in Figures 12, 13A-13B, and 14A-14I.

[0006] In another embodiment, the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Compositions comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain containing any of the following variable heavy chain domains and variable light chain domains: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0007] In another embodiment, the present invention relates to the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 The present invention provides a composition containing an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain selected from 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha16-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H16-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0008] In another embodiment, the present invention provides a nucleic acid composition comprising: a) a first nucleic acid encoding a variable heavy chain domain including variable heavy chain complementarity determination regions 1-3 (vhCDR1-3) of the ENPP3 binding domain; and b) a second nucleic acid encoding a variable light chain domain including variable light chain complementarity determination regions 1-3 (vlCDR1-3) of the ENPP3 binding domain, wherein the ENPP3 binding domain is one of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 One of 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). In some embodiments, vhCDR1-3 and vlCDR1-3 are selected from the vhCDR1-3 and vlCDR1-3 sequences provided in Figures 12, 13A-13B, and 14A-14I.

[0009] In another aspect, the present invention provides a nucleic acid composition comprising: a) a first nucleic acid encoding a variable heavy chain domain comprising a variable heavy chain domain of an ENPP3 binding domain, and b) a second nucleic acid encoding a variable light chain domain comprising a variable light chain domain of an ENPP3 binding domain, wherein the ENPP3 binding domain is any one of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (FIGS. 12, 13A-13B, and 14A-14I).

[0010] In some embodiments, the present invention provides an expression vector composition comprising: a) a first expression vector comprising a first nucleic acid, and b) a second expression vector comprising a second nucleic acid. In further embodiments, the present invention provides a host cell comprising the expression vector composition.

[0011] In some embodiments, the present invention provides a method for producing an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, comprising culturing a host cell under conditions in which the ENPP3 binding domain is expressed and recovering the ENPP3 binding domain.

[0012] In one aspect, the present invention provides an anti-ENPP3 antibody comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, wherein the ENPP3 binding domain is one of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I) and comprises any of the variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) and variable light chain complementarity determining regions (vlCDR1-3). In some embodiments, vhCDR1-3 and vlCDR1-3 are selected from the vhCDR1-3 and vlCDR1-3 of any of the following ENPP3 binding domains of Figures 12, 13A-13B, and 14A-14I.

[0013] In another embodiment, the present invention provides an anti-ENPP3 antibody comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, wherein the ENPP3 binding domain is one of the following: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 It contains any of the following variable heavy and light chain domains: 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0014] In another embodiment, the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Anti-ENPP3 antibodies containing an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain selected from any one of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0015] In some embodiments, the antibody comprises a) a first monomer containing a first antigen-binding domain and a first constant domain, and b) a second monomer containing a second antigen-binding domain and a second constant domain, wherein either the first or second antigen-binding domain is an ENPP3-binding domain. In further embodiments, the first and second antigen-binding domains bind to different antigens. In further embodiments, the first antigen-binding domain is an ENPP3-binding domain, and the second antigen-binding domain is a CD3-binding domain. In further embodiments, the CD3 binding domain includes vhCDR1-3 and vlCDR1-3 of any of the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F). In further embodiments, the CD3-binding domains vhCDR1-3 and vlCDR1-3 are selected from vhCDR1-3 and vlCDR1-3 shown in Figures 10A-10F.

[0016] In some embodiments, the CD3 binding domain includes one of the following variable heavy chain domains and variable light chain domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).

[0017] In some embodiments, the CD3-binding domain is anti-CD3scFv.

[0018] In some embodiments, the first and second constant domains each include CH2-CH3.

[0019] In some embodiments, the first and second steady-state domains each include CH1-hinge-CH2-CH3.

[0020] In some embodiments, the first and second steady-state domains are each variant steady-state domains.

[0021] In some embodiments, the first and second monomers comprise a set of heterodimerization variants, which are any one of the variants shown in Figures 1A to 1E. In some embodiments, the set of heterodimerization variants comprises one of the following sets of variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V.

[0022] In some embodiments, the first and second monomers each further include a removal variant. In further embodiments, the removal variant is E233P / L234V / L235A / G236del / S267K.

[0023] In some embodiments, at least one of the first or second monomers further comprises a pI variant. In further embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D. In some embodiments, the scFv comprises a charged scFv linker.

[0024] In some embodiments, the present invention provides a nucleic acid composition comprising nucleic acid encoding anti-ENPP3. In some embodiments, the composition comprises nucleic acids encoding first and second monomers. In some embodiments, the present invention provides an expression vector comprising nucleic acid. In further embodiments, the present invention provides host cells transformed with the expression vector.

[0025] In some embodiments, the present invention provides a method for producing an anti-ENPP3 antibody according to any one of claims B1 to B21. This method comprises culturing host cells according to claim B25 under conditions in which the anti-ENPP3 antibody is expressed, and recovering the anti-ENPP3 antibody. In some embodiments, the present invention provides a method for treating cancer, comprising administering the antibody to a patient in need.

[0026] In another aspect, the present invention provides a heterodimer antibody comprising: a) i) an anti-CD3scFv comprising a first variable light chain domain, an scFv linker, and a first variable heavy chain domain; ii) a first Fc domain, wherein the scFv is covalently bonded to the N-terminus of the first Fc domain using a domain linker; b) a second monomer comprising a VH2-CH1-hinge-CH2-CH3 monomer, wherein VH is a second variable heavy chain domain and CH2-CH3 is a second Fc domain; and c) a light chain comprising a second variable light chain domain, wherein the second variable heavy chain domain and the second variable light chain domain form an ENPP3 binding domain.

[0027] In some embodiments, the ENPP3 binding domains are the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Includes vhCDR1-3 and vlCDR1-3 of any of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0028] In some embodiments, the vhCDR1-3 and vlCDR1-3 of the ENPP3 binding domains are selected from the vhCDR1-3 and vlCDR1-3 sequences of the ENPP3 binding domains provided in Figures 12, 13A-13B, and 14A-14I.

[0029] In some embodiments, the second heavily variable domain is the following ENPP3 binding domain: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 The second lightly variable domain includes one of the following heavily variable domains: 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I), and the second lightly variable domain is one of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8L1, AN1[ENPP3]H1.8L1.33, AN1[ENPP3]H1L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16- It includes any of the following variable light domains: 1.67, Ha16-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha16-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha16-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H16-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0030] In further embodiments, the anti-CD3 binding scFv includes vhCDR1-3 and vlCDR1-3 of any of the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).

[0031] In some embodiments, the anti-CD3scFv vhCDR1-3 and vlCDR1-3 are selected from vhCDR1-3 and vlCDR1-3 shown in Figures 10A-10F.

[0032] In some embodiments, the anti-CD3scFv includes any of the following variable heavy chain domains and variable light chain domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).

[0033] In some embodiments, the first variable light chain domain is covalently bonded to the N-terminus of the first Fc domain using a domain linker.

[0034] In some embodiments, the first variable heavy chain domain is covalently bonded to the N-terminus of the first Fc domain using a domain linker.

[0035] In some embodiments, the scFv linker is a charged scFv linker.

[0036] In some embodiments, the first and second Fc domains are variant Fc domains.

[0037] In some embodiments, the first and second monomers include a set of heterodimerized variants selected from either the heterodimers or variants shown in Figures 1A-1E. In some embodiments, the selected set of heterodimerized variants derives from the following:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, where the numbering follows EU numbering.

[0038] In some embodiments, the first and second monomers further comprise a removal variant. In some embodiments, the removal variant is E233P / L234V / L235A / G236del / S267K, where the numbering follows EU numbering.

[0039] In some embodiments, one of the first or second monomers includes a pI variant.

[0040] In some embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D, where the numbering follows EU numbering.

[0041] In some embodiments, the first monomer comprises the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, and the second monomer comprises the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, where the numbering follows EU numbering.

[0042] In some embodiments, the scFv linker is a charged scFv linker having the amino acid sequence (GKPGS)4.

[0043] In some embodiments, the first and second monomers each further contain the amino acid variant 428 / 434S.

[0044] In some embodiments, the heterodimer antibodies include the following heterodimer antibodies: XENP24804, XENP26820, XENP28287, XENP28925, XENP29516, XENP30262, XENP26821, XENP29436, XENP28390, XENP29463, and XENP30263.

[0045] In another embodiment, the present invention provides a heterodimer antibody comprising: a) a first monomer comprising scFv-linker-CH2-CH3 from the N-terminus to the C-terminus, wherein scFv is anti-CD3scFV and CH2-CH3 is a first Fc domain; b) a second monomer comprising VH-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein CH2-CH3 is a second Fc domain; and c) a light chain comprising VL-CL, wherein the first variant Fc domain is amine The first variant contains the amino acid variant S364K / E357Q, the second variant Fc domain contains the amino acid variant L368D / K370S, the first and second variant Fc domains each contain the amino acid variant E233P / L234V / L235A / G236del / S267K, the second monomer hinge-CH2-CH3 contains the amino acid variant N208D / Q295E / N384D / Q418E / N421D, and VH and VL each contain AN1[ENPP3]H1L1 and AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I) form an ENPP3 binding domain containing a variable heavy chain domain and a variable light chain domain, and anti-CD3scFv is H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.This includes variable heavy and light chain domains of CD3-binding domains selected from 30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F), where numbering follows EU numbering.

[0046] In some embodiments, the scFv includes a charged scFv linker having the amino acid sequence (GKPGS)4.

[0047] In some embodiments, the first and second variant Fc domains each further comprise amino acid variants 428 / 434S, where the numbering follows EU numbering.

[0048] In some embodiments, the present invention provides nucleic acid compositions comprising first and second monomers and nucleic acids encoding the light chain of an antibody.

[0049] In some embodiments, the present invention provides an expression vector comprising nucleic acid. In some embodiments, the present invention provides a host cell transformed with the expression vector.

[0050] In some embodiments, the present invention provides a method for treating ENPP3-associated cancer, comprising administering to a patient in need of it one of the antibodies provided herein.

[0051] In another embodiment, the present invention relates to a) a first monomer comprising VH1-CH1-linker1-scFv-linker2-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable heavy chain domain, scFv is anti-CD3scFV, linker1 and linker2 are a first domain linker and a second domain linker, respectively, and CH2-CH3 is a first Fc domain, and b) from the N-terminus to the C-terminus The present invention provides a heterodimer antibody comprising a second monomer containing VH2-CH1-hinge-CH2-CH3, wherein VH2 is a second variable heavy chain domain and CH2-CH3 is a second Fc domain, and a common light chain containing a variable light chain domain, wherein the first variable heavy chain domain and the variable light chain domain form a first ENPP3 binding domain, and the second variable heavy chain domain and the variable light chain domain form a second ENPP3 binding domain.

[0052] In some embodiments, the first and second ENPP3 binding domains are, respectively, the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Includes vhCDR1-3 and vlCDR1-3 of any of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).

[0053] In some embodiments, the first and second ENPP3 binding domains vhCDR1-3 and vlCDR1-3 are selected from vhCDR1-3 and vlCDR1-3 provided in Figures 14 and 45.

[0054] In some embodiments, the first and second variable heavy chain domains each include a variable heavy chain domain of an ENPP3 binding domain, and the first and second variable light chain domains each include a variable light chain domain of an ENPP3 binding domain, and the ENPP3 binding domain is one of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 It is one of the following: 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A~13B, and 14A~14I).

[0055] In some embodiments, scFv includes vhCDR1-3 and vlCDR1-3 of any of the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).

[0056] In some embodiments, the vhCDR1-3 and vlCDR1-3 of scFv are selected from the vhCDR1-3 and vlCDR1-3 shown in Figures 10A-10F.

[0057] In some embodiments, the scFv includes one of the following variable heavy chain domains and variable light chain domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).

[0058] In some embodiments, the scFv includes a variable heavy chain domain, a variable light chain domain, and an scFv linker connecting the variable heavy chain domain and the variable light chain domain.

[0059] In some embodiments, the scFv variable heavy chain domain is bonded to the C-terminus of CH1 of the first monomer using a first domain linker, and the scFv variable light chain domain is covalently bonded to the N-terminus of the first Fc domain using a second Fc domain linker.

[0060] In some embodiments, the scFv variable light chain domain is covalently bonded to the C-terminus of CH1 of the first monomer using a first domain linker, and the scFv variable heavy chain domain is covalently bonded to the N-terminus of the first Fc domain using a second Fc domain linker.

[0061] In some embodiments, the scFv linker is a charged scFv linker.

[0062] In some embodiments, the first and second Fc domains are variant Fc domains.

[0063] In some embodiments, the first and second monomers include a set of heterodimerized variants selected from those shown in Figures 1A to 1E.

[0064] In some embodiments, the selected set of heterodimerization variants derive from the following:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, where the numbering follows EU numbering.

[0065] In some embodiments, the first and second monomers further include removal variants.

[0066] In a further embodiment, the elimination variant is E233P / L234V / L235A / G236del / S267K, where the numbering follows the EU numbering system.

[0067] In some embodiments, one of the first or second monomers further comprises a pI variant.

[0068] In some embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D, where the numbering follows EU numbering.

[0069] In some embodiments, the first variant Fc domain comprises the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, and the second variant Fc domain comprises the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, where the numbering follows EU numbering.

[0070] In some embodiments, the scFv linker is a charged scFv linker having the amino acid sequence (GKPGS)4.

[0071] In some embodiments, the first and second variant Fc domains each further comprise amino acid variants 428 / 434S, where the numbering follows EU numbering.

[0072] In some embodiments, the heterodimer antibodies include the following heterodimer antibodies: XENP29437, XENP29520, XENP30264, XENP26822, XENP28438, XENP29438, XENP29467, XENP30469, XENP30470, XENP30819, XENP30821, XENP31148, XENP31149, XENP31150, XENP31419, and XENP31471.

[0073] In another embodiment, the heterodimer antibody comprises: a) a first monomer comprising VH1-CH1-linker1-scFv-linker2-CH2-CH3 from the N-terminus to the C-terminus, where scFv is anti-CD3scFV and CH2-CH3 is a first Fc domain; b) a second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where CH2-CH3 is a second Fc domain; and c) a common light chain comprising VL-CL, and the first variant Fc domain The main component contains the amino acid variant S364K / E357Q, the second variant Fc domain contains the amino acid variant L368D / K370S, the first and second variant Fc domains each contain the amino acid variant E233P / L234V / L235A / G236del / S267K, the hinge-CH2-CH3 of the second monomer contains the amino acid variant N208D / Q295E / N384D / Q418E / N421D, and the VH and VL are AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 The ENPP3 binding domains selected from 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I) contain variable heavy and light chain domains, and the anti-CD3scFv is H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.This includes variable heavy and light chain domains of CD3-binding domains selected from 30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F), where numbering follows EU numbering.

[0074] In some embodiments, the scFv includes a charged scFv linker having the amino acid sequence (GKPGS)4.

[0075] In some embodiments, the first and second variant Fc domains each further comprise the amino acid variant 428 / 434S.

[0076] In some embodiments, the antibody comprises first and second monomers and a common light chain. In some embodiments, the present invention provides an expression vector comprising nucleic acid. In some embodiments, the present invention provides host cells transformed with the expression vector. In some embodiments, the present invention provides treatment for ENPP3-associated cancer, comprising administering an antibody to a patient in need thereof.

[0077] In another embodiment, the present invention provides heterodimer antibodies comprising the following heterodimer antibodies: XENP24804, XENP26820, XENP28287, XENP28925, XENP29516, XENP30262, XENP26821, XENP29436, XENP28390, XENP29463, and XENP30263.

[0078] In another embodiment, the present invention provides heterodimer antibodies comprising the following heterodimer antibodies: XENP29437, XENP29520, XENP30264, XENP26822, XENP28438, XENP29438, XENP29467, XENP30469, XENP30470, XENP30819, XENP30821, XENP31148, XENP31149, XENP31150, XENP31419, and XENP31471. In some embodiments, the present invention provides nucleic acid compositions comprising nucleic acids encoding heterodimer antibodies. In some embodiments, the present invention provides expression vectors comprising nucleic acids. In further embodiments, the present invention provides host cells transformed with an expression vector.

[0079] In some embodiments, the present invention provides a method for treating ENPP3-associated cancer, comprising administering to a patient in need of it one of the heterodimeric antibodies provided herein. [Brief explanation of the drawing]

[0080] [Figure 1A] This shows useful pairs of Fc heterodimerization variants (including scuba dimers and PI variants). Some variants lack a corresponding "monomer 2" variant; these are pI variants that can be used with any monomer alone. [Figure 1B] This shows useful pairs of Fc heterodimerization variants (including scuba dimers and PI variants). Some variants lack a corresponding "monomer 2" variant; these are pI variants that can be used with any monomer alone. [Figure 1C] This shows useful pairs of Fc heterodimerization variants (including scuba dimers and PI variants). Some variants lack a corresponding "monomer 2" variant; these are pI variants that can be used with any monomer alone. [Figure 1D]This shows useful pairs of Fc heterodimerization variants (including scuba dimers and PI variants). Some variants lack a corresponding "monomer 2" variant; these are pI variants that can be used with any monomer alone. [Figure 1E] This shows useful pairs of Fc heterodimerization variants (including scuba dimers and PI variants). Some variants lack a corresponding "monomer 2" variant; these are pI variants that can be used with any monomer alone. [Figure 2] This section lists the constant regions of equivalent variant antibodies and their respective substitutions. pI_(-) indicates a low pI variant, and pI_(+) indicates a high pI variant. These can be optionally and independently combined with other heterodimerized variants of the antibodies described herein (as with other variant types, as outlined herein). [Figure 3] This exhibits a useful removal variant (often called a "knockout" or "KO" variant) in which the FcγR bond is removed. Generally, removal variants are found in both monomers, but in some cases they may be present in only one of the monomers. [Figure 4] Particularly useful embodiments of the “non-Fv” components of the antibodies described herein are shown. [Figure 5]As described herein, several charged scFv linkers are shown that are used to increase or decrease the pI of a heterodimer bsAb of a subject utilizing one or more scFvs as components. (+H) positive linkers are used herein, particularly in anti-CD3VL and VH sequences as shown herein. A single prior art scFv linker with a single charge 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 in scFvs and enhance proteolytic stability. Such charged scFv linkers can be used in any of the antibody formats of the subject disclosed herein that include scFvs (e.g., 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fc formats). [Figure 6] Several exemplary domain linkers are shown. In some embodiments, these linkers find applications in linking single-chain Fv to Fc chains. In some embodiments, these linkers can be combined. For example, the GGGGS linker can be combined with a “half-hinge” linker. [Figure 7A]Without including the Fv sequence (e.g., VH for scFv and Fab sides), the following sequences of several useful 1+1 Fab-scFv-Fc bispecific antibody format heavy chain backbones are shown based on human IgG1. Backbone 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the chain with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, C220S on the strand with the S364K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba ant, C220S on the strand with the S364K scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368E / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. The skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, C220S on the strand with the D401K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the strand with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba rianto, a chain-like C220S with the S364K / E357Q scuba rianto, a chain-like N208D / Q295E / N384D / Q418E / N421D pI variant with the L368D / K370S scuba rianto, an E233P / L234V / L235A / G236del / S267K removal variant on both chains, and a chain-like N297A variant on both chains. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is based on human IgG4 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the S228P (EU numbering, which is S241P in Kabat) variant on both strands that removes the Fab arm exchange as known in the art. Skeleton 9 is based on human IgG2 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant. Skeleton 10 is based on human IgG2 and contains the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368D / K370S scuba riant, and the S267K variant on both strands. Skeleton 11 is identical to skeleton 1 except that it contains the M428L / N434S Xtend mutation.The skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q:L368D / K370S scuba riant, C220S, and S364K / E357Q scuba riant, as well as the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 7B]Without including the Fv sequence (e.g., VH for scFv and Fab sides), the following sequences of several useful 1+1 Fab-scFv-Fc bispecific antibody format heavy chain backbones are shown based on human IgG1. Backbone 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the chain with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, C220S on the strand with the S364K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba ant, C220S on the strand with the S364K scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368E / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. The skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, C220S on the strand with the D401K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the strand with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba rianto, a chain-like C220S with the S364K / E357Q scuba rianto, a chain-like N208D / Q295E / N384D / Q418E / N421D pI variant with the L368D / K370S scuba rianto, an E233P / L234V / L235A / G236del / S267K removal variant on both chains, and a chain-like N297A variant on both chains. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is based on human IgG4 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the S228P (EU numbering, which is S241P in Kabat) variant on both strands that removes the Fab arm exchange as known in the art. Skeleton 9 is based on human IgG2 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant. Skeleton 10 is based on human IgG2 and contains the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368D / K370S scuba riant, and the S267K variant on both strands. Skeleton 11 is identical to skeleton 1 except that it contains the M428L / N434S Xtend mutation.The skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q:L368D / K370S scuba riant, C220S, and S364K / E357Q scuba riant, as well as the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 7C]Without including the Fv sequence (e.g., VH for scFv and Fab sides), the following sequences of several useful 1+1 Fab-scFv-Fc bispecific antibody format heavy chain backbones are shown based on human IgG1. Backbone 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the chain with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, C220S on the strand with the S364K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba ant, C220S on the strand with the S364K scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368E / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. The skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, C220S on the strand with the D401K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the strand with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba rianto, a chain-like C220S with the S364K / E357Q scuba rianto, a chain-like N208D / Q295E / N384D / Q418E / N421D pI variant with the L368D / K370S scuba rianto, an E233P / L234V / L235A / G236del / S267K removal variant on both chains, and a chain-like N297A variant on both chains. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is based on human IgG4 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the S228P (EU numbering, which is S241P in Kabat) variant on both strands that removes the Fab arm exchange as known in the art. Skeleton 9 is based on human IgG2 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant. Skeleton 10 is based on human IgG2 and contains the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368D / K370S scuba riant, and the S267K variant on both strands. Skeleton 11 is identical to skeleton 1 except that it contains the M428L / N434S Xtend mutation.The skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q:L368D / K370S scuba riant, C220S, and S364K / E357Q scuba riant, as well as the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 7D]Without including the Fv sequence (e.g., VH for scFv and Fab sides), the following sequences of several useful 1+1 Fab-scFv-Fc bispecific antibody format heavy chain backbones are shown based on human IgG1. Backbone 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the chain with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, C220S on the strand with the S364K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba ant, C220S on the strand with the S364K scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368E / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. The skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, C220S on the strand with the D401K scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, C220S on the strand with the S364K / E357Q scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba rianto, a chain-like C220S with the S364K / E357Q scuba rianto, a chain-like N208D / Q295E / N384D / Q418E / N421D pI variant with the L368D / K370S scuba rianto, an E233P / L234V / L235A / G236del / S267K removal variant on both chains, and a chain-like N297A variant on both chains. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is based on human IgG4 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the S228P (EU numbering, which is S241P in Kabat) variant on both strands that removes the Fab arm exchange as known in the art. Skeleton 9 is based on human IgG2 and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant. Skeleton 10 is based on human IgG2 and contains the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368D / K370S scuba riant, and the S267K variant on both strands. Skeleton 11 is identical to skeleton 1 except that it contains the M428L / N434S Xtend mutation.The skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q:L368D / K370S scuba riant, C220S, and S364K / E357Q scuba riant, as well as the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 8A]Without including the Fv sequence (e.g., VH for scFv and Fab sides), we present sequences of several useful 2+Fab2-scFv-Fc bispecific antibody format heavy chain skeletons based on human IgG1. Skeleton 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368E / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, the E233P / L234V / L235A / G236del / S267K removal variant on both strands, and the N297A variant on both strands. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is identical to skeleton 1 except that it includes the M428L / N434S Xtend mutation. The skeleton 9 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 8B]Without including the Fv sequence (e.g., VH for scFv and Fab sides), we present sequences of several useful 2+Fab2-scFv-Fc bispecific antibody format heavy chain skeletons based on human IgG1. Skeleton 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368E / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, the E233P / L234V / L235A / G236del / S267K removal variant on both strands, and the N297A variant on both strands. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is identical to skeleton 1 except that it includes the M428L / N434S Xtend mutation. The skeleton 9 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 8C]Without including the Fv sequence (e.g., VH for scFv and Fab sides), we present sequences of several useful 2+Fab2-scFv-Fc bispecific antibody format heavy chain skeletons based on human IgG1. Skeleton 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both chains. Skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the L368E / K370S scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand containing the K360E / Q362E / T411E scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the S364K / E357Q:L368D / K370S scuba ant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand having the L368D / K370S scuba ant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands.Skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the strand with the L368D / K370S scuba riant, the E233P / L234V / L235A / G236del / S267K removal variant on both strands, and the N297A variant on both strands. Skeleton 7 is identical to 6 except that the mutation is N297S. Skeleton 8 is identical to skeleton 1 except that it includes the M428L / N434S Xtend mutation. The skeleton 9 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S scuba riant, the P217R / P229R / N276K pI variant on the strand having the S364K / E357Q scuba riant, and the E233P / L234V / L235A / G236del / S267K removal variant on both strands. Each of these skeletons contains a sequence that is 90, 95, 98, and 99% identical (as defined herein) to the enumerated sequences and / or includes 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the “parent” in the figure, which already includes some amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), as will be understood by those skilled in the art. In other words, the listed skeletons may include additional amino acid modifications (generally amino acid substitutions) in addition to the scuba riants, pI variants, and removal variants contained within the skeletons in this diagram. [Figure 9] Based on human IgG1, without including the Fv sequence (e.g., scFv or Fab), several useful constant light chain domain skeleton sequences are shown. Constant light chain skeleton sequences that are 90, 95, 98, and 99% identical to the enumerated sequences (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid modifications are included herein. [Figure 10A]This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 10B] This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 10C] This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 10D] This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 10E] This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 10F]This specification shows exemplary anti-CD3scFv sequences suitable for use with the bispecific antibodies described herein. The CDR is underlined, the scFv linker is double underlined (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, some of which are uncharged or negatively charged, as shown in Figure 5), and a slash indicates the boundary of a variable domain. Furthermore, the nomenclature convention indicates the direction of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR position may differ slightly depending on the numbering used as shown in Table 2, and therefore, not only the underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 11A] The antigen sequences of several antigens used in the antibodies described herein, including both human and cynomolgus monkey antigens, are shown to facilitate the development of antigen-binding domains that bind to both antigens for easier clinical development. [Figure 11B] The antigen sequences of several antigens used in the antibodies described herein, including both human and cynomolgus monkey antigens, are shown to facilitate the development of antigen-binding domains that bind to both antigens for easier clinical development. [Figure 12]This specification shows the variable heavy and variable light chain sequences of an exemplary humanized ENPP3-binding domain, referred to herein as AN1, as well as the sequence of XENP28278, an anti-ENPP3 mAb based on the AN1 and IgG1 skeletons, having the E233P / L234V / L235A / G236del / S267K removal variant. CDRs are underlined, and slashes indicate the boundary between the variable region and the constant domain. As is the case with all sequences described herein and including CDRs herein, the precise identification of the CDR location may differ slightly depending on the numbering used, as shown in Table 2, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figures, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 13A] This document shows the variable heavy chain and variable light chain sequences of AN1 variants designed for improved purification and / or regulation of ENPP3 binding affinity and / or potency. CDRs are underlined, and slashes indicate the boundary between the variable region and the constant domain. As is true for all sequences described herein and including CDRs herein, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. In addition, each of the variable heavy chain domains shown herein can be paired with any other αENPP3 variable light chain domain, and each of the variable light chain domains shown herein can be paired with any other αENPP3 variable heavy chain domain. [Figure 13B]This document shows the variable heavy chain and variable light chain sequences of AN1 variants designed for improved purification and / or regulation of ENPP3 binding affinity and / or potency. CDRs are underlined, and slashes indicate the boundary between the variable region and the constant domain. As is true for all sequences described herein and including CDRs herein, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. In addition, each of the variable heavy chain domains shown herein can be paired with any other αENPP3 variable light chain domain, and each of the variable light chain domains shown herein can be paired with any other αENPP3 variable heavy chain domain. [Figure 14A] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14B]This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14C] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14D] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14E]This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14F] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14G] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14H]This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 14I] This shows the variable regions of additional ENPP3 antigen-binding domains that can be used with αENPP3 × αCD3 antibodies. CDRs are underlined. As is the case with all sequences described herein and including CDRs herein, the precise identification of CDR locations may vary slightly depending on the numbering used, as shown in Figure 12, and therefore, not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv format or Fab format. [Figure 15A] Several formats of the antibodies described herein are shown. Figure 15A shows the "1+1 Fab-scFv-Fc" format, where the first arm contains ENPP3-conjugated Fab and the second arm contains CD3-conjugated scFv. Figure 30B shows the "2+1 Fab2-scFv-Fc" format, where the first arm contains ENPP3-conjugated Fab and the second arm contains Fab and scFv, where Fab is conjugated to ENPP3 and scFv is conjugated to CD3. [Figure 15B]Several formats of the antibodies described herein are shown. Figure 15A shows the "1+1 Fab-scFv-Fc" format, where the first arm contains ENPP3-conjugated Fab and the second arm contains CD3-conjugated scFv. Figure 30B shows the "2+1 Fab2-scFv-Fc" format, where the first arm contains ENPP3-conjugated Fab and the second arm contains Fab and scFv, where Fab is conjugated to ENPP3 and scFv is conjugated to CD3. [Figure 16] The amino acid sequences of control anti-RSV × high CD3 bispecific antibodies in bottle-opener format (Fab-scFv-Fc) are shown. The antibodies are named using the Fab variable region 1 and scFv variable region 2, separated by dashes. CDRs are underlined, and slashes indicate the boundaries of the variable regions. The scFv domain has a VH-scFv linker-VL orientation (from N-terminus to C-terminus), but this can be reversed. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 17A] The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may be available as a variable region, Fc region, and constant domain sequence that is 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 17B]The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may be available as a variable region, Fc region, and constant domain sequence that is 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 17C] The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may be available as a variable region, Fc region, and constant domain sequence that is 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 18A]The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 18B] The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 18C]The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 1+1Fab-scFv-Fc format and containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]), is shown. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 19A] The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 2+1Fab2-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 19B]The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 2+1Fab2-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 19C] The sequence of an exemplary αENPP3×αCD3 bsAb, shown in 2+1Fab2-scFv-Fc format and containing H1.30_L1.47 anti-CD3scFv (also known as CD3 High [VHVL]), is shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 20A]The following sequences of exemplary αENPP3×αCD3 bsAb containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]) are shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 20B] The following sequences of exemplary αENPP3×αCD3 bsAb containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]) are shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 20C]The following sequences of exemplary αENPP3×αCD3 bsAb containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]) are shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 20D] The following sequences of exemplary αENPP3×αCD3 bsAb containing H1.32_L1.47 anti-CD3scFv (also known as CD3 High-Int#1[VHVL]) are shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 21]The sequence of an exemplary αENPP3×αCD3 bsAb is shown in 2+1Fab2-scFv-Fc format and contains L1.47_H1.30 anti-CD3scFv (also known as CD3 High [VLVH]). The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 22A] The following sequences of exemplary αENPP3×αCD3 bsAb, containing L1.47_H1.32 anti-CD3 scFv (also known as CD3 High-Int#1[VLVH]) in 2+1Fab2-scFv-Fc format, are shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 22B]The following sequences of exemplary αENPP3×αCD3 bsAb, containing L1.47_H1.32 anti-CD3 scFv (also known as CD3 High-Int#1[VLVH]) in 2+1Fab2-scFv-Fc format, are shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 22C] The following sequences of exemplary αENPP3×αCD3 bsAb, containing L1.47_H1.32 anti-CD3 scFv (also known as CD3 High-Int#1[VLVH]) in 2+1Fab2-scFv-Fc format, are shown. The CDR is underlined, and slashes indicate the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have variable regions, Fc regions, and constant domains that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 23A]The sequence of an exemplary αENPP3×αCD3 bsAb containing L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int#2[VLVH]) is shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 23B] The sequence of an exemplary αENPP3×αCD3 bsAb containing L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int#2[VLVH]) is shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 23C]The sequence of an exemplary αENPP3×αCD3 bsAb containing L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int#2[VLVH]) is shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 23D] The sequence of an exemplary αENPP3×αCD3 bsAb containing L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int#2[VLVH]) is shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 23E]The sequence of an exemplary αENPP3×αCD3 bsAb containing L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int#2[VLVH]) is shown in 2+1Fab2-scFv-Fc format. The CDR is underlined, and a slash indicates the boundary between the variable region and other chain components (e.g., the constant region and domain linker). Note that αENPP3×αCD3bsAb may have a variable region, Fc region, and constant domain that are 90, 95, 98, and 99% identical (as defined herein), as well as containing 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Furthermore, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, which results in a longer half-life in serum. [Figure 24A] The images show the induction of RTCC in CFSE-labeled KU812 cells after 24-hour incubation with human PBMCs of CFSE-labeled KU812 (effector-to-target cell ratio of 10:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390), stained with Zombie Aqua. This is indicated by A) a decrease in the number of CFSE+KU812 cells and B) a decrease in the percentage of CFSE+KU812 cells. The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. In summary, the data show that the prototype αENPP3×αCD3bsAb induced dose-dependent redirected T cell cytotoxicity (RTCC) on KU812 cells, that CD3 binding affinity correlated with RTCC potency (i.e., bsAbs containing CD3 High induced RTCC more strongly than bsAbs containing CD3 High-Int#1), and that bsAbs with an AN1 binding domain induced RTCC more strongly than bsAbs with an H16-7.8 binding domain. [Figure 24B]The images show the induction of RTCC in CFSE-labeled KU812 cells after 24-hour incubation with human PBMCs of CFSE-labeled KU812 (effector-to-target cell ratio of 10:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390), stained with Zombie Aqua. This is indicated by A) a decrease in the number of CFSE+KU812 cells and B) a decrease in the percentage of CFSE+KU812 cells. The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. In summary, the data show that the prototype αENPP3×αCD3bsAb induced dose-dependent redirected T cell cytotoxicity (RTCC) on KU812 cells, that CD3 binding affinity correlated with RTCC potency (i.e., bsAbs containing CD3 High induced RTCC more strongly than bsAbs containing CD3 High-Int#1), and that bsAbs with an AN1 binding domain induced RTCC more strongly than bsAbs with an H16-7.8 binding domain. [Figure 25A]The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled KU812 (effector-to-target cell ratio of 10:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD4+ T cell activation more potently than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD4+ T cell activation more potently than bsAbs with an H16-7.8 binding domain. [Figure 25B] The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled KU812 (effector-to-target cell ratio of 10:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD4+ T cell activation more potently than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD4+ T cell activation more potently than bsAbs with an H16-7.8 binding domain. [Figure 25C] The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled KU812 (effector-to-target cell ratio of 10:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD4+ T cell activation more potently than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD4+ T cell activation more potently than bsAbs with an H16-7.8 binding domain. [Figure 26A] The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with CFSE-labeled KU812 human PBMCs (10:1 effector-to-target cell ratio) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD8+ T cell activation more strongly than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAbs with an H16-7.8 binding domain. [Figure 26B] The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with CFSE-labeled KU812 human PBMCs (10:1 effector-to-target cell ratio) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD8+ T cell activation more strongly than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAbs with an H16-7.8 binding domain. [Figure 26C]The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with CFSE-labeled KU812 human PBMCs (10:1 effector-to-target cell ratio) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAbs containing CD3 High induced CD8+ T cell activation more strongly than bsAbs containing CD3 High-Int#1), and bsAbs with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAbs with an H16-7.8 binding domain. [Figure 27A]The images show the induction of RTCC in CFSE-labeled RXF393 cells after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390), stained with Zombie Aqua. This is indicated by A) a decrease in the number of CFSE+RXF393 cells and B) a decrease in the percentage of CFSE+RXF393 cells. The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with the data for KU812 cells, the data show that the prototype αENPP3×αCD3bsAb induced dose-dependent redirected T cell cytotoxicity (RTCC) on RXF393 cells, that CD3 binding affinity correlated with RTCC potency (i.e., bsAbs containing CD3 High induced RTCC more strongly than bsAbs containing CD3 High-Int#1), and that bsAbs with an AN1 binding domain induced RTCC more strongly than bsAbs with an H16-7.8 binding domain. [Figure 27B]The images show the induction of RTCC in CFSE-labeled RXF393 cells after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390), stained with Zombie Aqua. This is indicated by A) a decrease in the number of CFSE+RXF393 cells and B) a decrease in the percentage of CFSE+RXF393 cells. The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with the data for KU812 cells, the data show that the prototype αENPP3×αCD3bsAb induced dose-dependent redirected T cell cytotoxicity (RTCC) on RXF393 cells, that CD3 binding affinity correlated with RTCC potency (i.e., bsAbs containing CD3 High induced RTCC more strongly than bsAbs containing CD3 High-Int#1), and that bsAbs with an AN1 binding domain induced RTCC more strongly than bsAbs with an H16-7.8 binding domain. [Figure 28A]The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3 High induced CD4+ T cell activation more strongly than bsAb containing CD3 High-Int#1), and bsAb with an AN1 binding domain induced CD4+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 28B] The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3 High induced CD4+ T cell activation more strongly than bsAb containing CD3 High-Int#1), and bsAb with an AN1 binding domain induced CD4+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 28C] The images show CD4+ T cell activation as indicated by A) CD107a MFI on CD4+ T cells, B) CD25 MFI on CD4+ T cells, and C) CD69 MFI on CD4+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD4+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3 High induced CD4+ T cell activation more strongly than bsAb containing CD3 High-Int#1), and bsAb with an AN1 binding domain induced CD4+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 29A] The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3High induced CD8+ T cell activation more strongly than bsAb containing CD3High-Int#1), and bsAb with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 29B] The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3High induced CD8+ T cell activation more strongly than bsAb containing CD3High-Int#1), and bsAb with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 29C]The images show activated CD8+ T cells as indicated by A) CD107a MFI on CD8+ T cells, B) CD25 MFI on CD8+ T cells, and C) CD69 MFI on CD8+ T cells, after 24-hour incubation with human PBMCs of CFSE-labeled RXF393 (effector-to-target cell ratio of 20:1) and αENPP3 × αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390). The controls used were αRSV × αCD3 bispecific antibody (XENP13245), effector and target cells only, and target cells only. Consistent with RTCC data, αENPP3×αCD3bsAb dose-dependently induced CD8+ T cell activation, with CD3 binding affinity correlating with activating potency (i.e., bsAb containing CD3High induced CD8+ T cell activation more strongly than bsAb containing CD3High-Int#1), and bsAb with an AN1 binding domain induced CD8+ T cell activation more strongly than bsAb with an H16-7.8 binding domain. [Figure 30] A) Part 2 of the purification of XENP28287 (cation exchange chromatography following protein A chromatography), and chromatograms showing the purity and homogeneity of peaks B and BC isolated from the cation exchange separation shown in Figure 30A (as well as pre-purified material), B) chromatograms from analytical size exclusion chromatography using multi-angle light scattering (aSEC-MALS) and C) chromatograms from analytical cation exchange chromatography (aCIEX). Figure 30B also shows the molecular weight of the protein species of the peaks determined by multi-angle light scattering. [Figure 31]A) Part 2 of the purification of XENP28925 (cation exchange chromatography following protein A chromatography), and a chromatogram showing the purity and homogeneity of peak B isolated from the cation exchange separation shown in Figure 31A (as well as the pre-purified material), B) chromatograms from analytical size exclusion chromatography using multi-angle light scattering (aSEC-MALS) and C) chromatograms from analytical cation exchange chromatography (aCIEX). Figure 31B also shows the molecular weight of the protein species of the peak determined by multi-angle light scattering. [Figure 32] A) A chromatogram showing part 2 of the purification of XENP31149 (cation exchange chromatography following protein A chromatography), and B) A chromatogram showing the identity of peaks A and B isolated from the cation exchange separation shown in Figure XA (as well as material pre-purified by analytical size exclusion chromatography using multi-angle light scattering (aSEC-MALS)). [Figure 33] A) A chromatogram showing part 2 of the purification of XENP31419 (cation exchange chromatography following protein A chromatography), and B) A chromatogram showing the identity of peaks A and B isolated from the cation exchange separation shown in Figure XA (as well as material pre-purified by analytical size exclusion chromatography using multi-angle light scattering (aSEC-MALS)). [Figure 34] This shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) cells, indicated by the percentage of CFSE+ cells, after 18 hours of incubation with human PBMCs (effector-to-target cell ratio) of CFSE-labeled target cells (10:1 effector-to-target cell ratio) and the αENPP3×αCD3 bispecific antibody XENP28925, stained with Zombie Aqua. [Figure 35] This shows the binding of affinity-modified αENPP3×αCD31+1bsAb to ENPP3-high KU812 cells. [Figure 36]The data shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) cells, indicated by the percentage of CFSE+ cells stained with Zombie Aqua, after 42-hour incubation with human PBMCs (effector-to-target cell ratio of 10:1) and the αENPP3×αCD3 bispecific antibodies XENP28925 (WT, high ENPP3 binding), XENP29516 (intermediate ENPP3 binding), or XENP30262 (low ENPP3 binding). The data indicate that both XENP29516 and XENP30262 were substantially less potent in inducing RTCC in ENPP3-low RCC4 cells compared to XENP28925, and that RTCC potency correlates with binding potency as described above. XENP29516 and XENP30262 also showed that RTCC induction in ENPP3-high cells was not as potent. [Figure 37A] This shows the induction of A) IFNγ, B) IL-6, and C) TNFα release by KU812 cells (10:1 effector-to-target cell ratio) and human PBMCs incubated for 18 hours with the αENPP3 × αCD3 bispecific antibody XENP28925 (CD3 High) or XENP29436 (CD3 High-Int#1). The data indicate that XENP29436 has substantially weaker resistance to cytokine release induction compared to XENP28925. [Figure 37B] This shows the induction of A) IFNγ, B) IL-6, and C) TNFα release by KU812 cells (10:1 effector-to-target cell ratio) and human PBMCs incubated for 18 hours with the αENPP3 × αCD3 bispecific antibody XENP28925 (CD3 High) or XENP29436 (CD3 High-Int#1). The data indicate that XENP29436 has substantially weaker resistance to cytokine release induction compared to XENP28925. [Figure 37C]This shows the induction of A) IFNγ, B) IL-6, and C) TNFα release by KU812 cells (10:1 effector-to-target cell ratio) and human PBMCs incubated for 18 hours with the αENPP3 × αCD3 bispecific antibody XENP28925 (CD3 High) or XENP29436 (CD3 High-Int#1). The data indicate that XENP29436 has substantially weaker resistance to cytokine release induction compared to XENP28925. [Figure 38] This study demonstrates the induction of IFNγ release in RCC4 cells (with a 10:1 effector-to-target cell ratio) and human PBMCs incubated for 18 hours with the αENPP3 × αCD3 bispecific antibody XENP28925 (CD3 High) or XENP29436 (CD3 High-Int#1). The data show that XENP29436 demonstrated a very slight induction of cytokine release compared to XENP28925 in the presence of ENPP3-low RCC4 cells. [Figure 39] The images show the induction of RTCC in CFSE-labeled KU812 (solid line, ENPP3 high) or CFSE-labeled RCC4 (dashed line, RCC4 low) cells, indicated by the percentage of CFSE+ cells, after 42-hour incubation with human PBMCs (effector-to-target cell ratio) of CFSE-labeled target cells (10:1 effector-to-target cell ratio) and the αENPP3×αCD3 bispecific antibody XENP28925 (CD3 High) or XENP29436 (CD3 High-Int#1), stained with Zombie Aqua. The data demonstrate that XENP29436 is substantially less potent in inducing RTCC in ENPP3 low cells compared to XENP28925, but also demonstrates reduced potency in inducing RTCC in ENPP3 high cells. [Figure 40]This shows the induction of IFNγ release by human PBMCs incubated with KU812 cells (10:1 effector-to-target cell ratio) and the αENPP3×αCD3 bispecific antibodies XENP28925 (ENPP3 High; CD3 High), XENP29436 (ENPP3 High; CD3 High-Int#1), XENP29518 (ENPP3 Intermediate; CD3 High), XENP29463 (ENPP3 Intermediate; CD3 High-Int#1), XENP30262 (ENPP3 Low; CD3 High), or XENP30263 (ENPP3 Low; CD3 High-Int#1). The data indicate that reducing the binding affinity of CD3 or ENPP3 reduces the induction of cytokine release. In particular, reducing the binding affinity of both CD3 and ENPP3 further reduces the induction of cytokine release. [Figure 41] The image shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) cells, indicated by the percentage of CFSE+ cells, after 42-hour incubation with human PBMCs (10:1 effector-to-target cell ratio) of CFSE-labeled target cells (10:1 effector-to-target cell ratio) and αENPP3×αCD3 bispecific antibodies XENP28925 (WT, high ENPP3 binding; CD3 High; monovalent ENPP3 binding), XENP29516 (intermediate ENPP3 binding; CD3 High; monovalent ENPP3 binding), or XENP29520 (intermediate ENPP3 binding; CD3 High; bivalent ENPP3 binding), stained with Zombie Aqua. The data show that the divalent binding (with intermediate ENPP3 binding) maintained reduced RTCC efficacy in ENPP3-low cells, but restored RTCC efficacy in ENPP3-high cells, similar to that demonstrated by XENP28925. [Figure 42]The data shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) cells, indicated by the percentage of CFSE+ cells stained with Zombie Aqua, after 42-hour incubation with human PBMCs (10:1 effector-to-target cell ratio) and αENPP3×αCD3 bispecific antibodies XENP28925 (WT, high ENPP3 binding; CD3 High; monovalent ENPP3 binding), XENP30262 (low ENPP3 binding; CD3 High; monovalent ENPP3 binding), or XENP30264 (low ENPP3 binding; CD3 High; divalent ENPP3 binding). The data indicate that divalent binding (with low ENPP3 binding) further reduced RTCC efficacy in low ENPP3 cells and restored some RTCC efficacy in high ENPP3 cells. [Figure 43] This shows the induction of RTCC in CFSE-labeled KU812 cells, indicated by the percentage of CFSE+ cells stained with Zombie Aqua, after 44-hour incubation with human PBMCs (effector-to-target cell ratio of 10:1) and αENPP3×αCD3 bispecific antibodies XENP28925 (CD3 High; monovalent ENPP3-bound), XENP29437 (CD3 High; divalent ENPP3-bound), XENP29436 (CD3 High-Int#1; monovalent ENPP3-bound), or XENP29438 (CD3 High-Int#1; divalent ENPP3-bound). Unexpectedly, XENP29438 failed to induce RTCC in KU812 cells. [Figure 44]The image shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) after 24-hour incubation with human PBMCs (effector-to-target cell ratio) of CFSE-labeled target cells (10:1 effector-to-target cell ratio) and αENPP3×αCD3 bispecific antibodies XENP294377 (CD3 High VH / VL; bivalent ENPP3 conjugated), XENP30469 (CD3 High VL / VH; bivalent ENPP3 conjugated), XENP29428 (CD3 High-Int#1 VH / VL; bivalent ENPP3 conjugated), indicated by the percentage of CFSE+ cells stained with Zombie Aqua. The data showed that swapping the orientations of the variable heavy and light domains of CD3 High-Int#1scFv restored its activity in relation to the 2+1Fab2-scFv-Fc bsAb format (XENP29438 vs. XENP30470). Swapping the orientations of the variable heavy and light domains of CD3 High scFv allowed for a much milder improvement in RTCC potency in relation to the 2+1Fab2-scFv-Fc bsAb format (XENP29437 vs. XENP30469). [Figure 45]Human PBMCs of CFSE-labeled target cells (10:1 effector-to-target cell ratio) and αENPP3 × αCD3 bispecific antibodies XENP29520 (CD3 High [VH / VL]; divalent ENPP3 intermediate binding), XENP30819 (CD3 High-Int#1 [VL / VHL]; divalent ENPP3 intermediate binding), XENP31149 (CD3 High-Int#2 [VL / VHL]; divalent ENPP3 intermediate binding), XENP30264 (CD3 High [VH / VL]; divalent ENPP3 low binding), XENP30821 (CD3 High-Int#1 [VL / VHL]; divalent ENPP3 low binding), or XENP31150 (CD3 The image shows the induction of RTCC in CFSE-labeled KU812 (solid line, high ENPP3) or CFSE-labeled RCC4 (dashed line, low ENPP3) after 40 hours of incubation with High-Int#2[VL / VHL] (low binding of divalent ENPP3), indicated by the percentage of CFSE+ cells stained with Zombie Aqua. [Figure 46] The sequence of XENP16432, an anti-PD-1 mAb based on nivolumab and an IgG1 scaffold with an E233P / L234V / L235A / G236del / S267K removal variant, and XENP21461 (pembrolizumab) is shown. [Figure 47] The images show the time course of tumor volume (determined by caliper measurements) in NSG mice transplanted with KU812 and huPBMCs, administered with PBS, XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) alone or in combination with XENP16432. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low and / or high doses and was well combined with PD-1 blockers. [Figure 48A]The images show the proliferation of A) CD45+ lymphocytes, B) CD8+ T cells, and C) CD4+ T cells in the blood of NSG mice transplanted with KU812 and huPBMCs up to day 14, administered with PBS, XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1bsAb (XENP30819, XENP30821, or XENP31419) alone or in combination with XENP16432. In all cases, lymphocyte proliferation was promoted when combined with a PD-1 blocker. [Figure 48B] The images show the proliferation of A) CD45+ lymphocytes, B) CD8+ T cells, and C) CD4+ T cells in the blood of NSG mice transplanted with KU812 and huPBMCs up to day 14, administered with PBS, XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1bsAb (XENP30819, XENP30821, or XENP31419) alone or in combination with XENP16432. In all cases, lymphocyte proliferation was promoted when combined with a PD-1 blocker. [Figure 48C] The images show the proliferation of A) CD45+ lymphocytes, B) CD8+ T cells, and C) CD4+ T cells in the blood of NSG mice transplanted with KU812 and huPBMCs up to day 14, administered with PBS, XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1bsAb (XENP30819, XENP30821, or XENP31419) alone or in combination with XENP16432. In all cases, lymphocyte proliferation was promoted when combined with a PD-1 blocker. [Figure 49]The images show the time course of tumor volume (determined by caliper measurements) in NSG mice transplanted with RXF-393 and huPBMCs, administered with PBS, XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 50A] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50B] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50C]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50D] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50E] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50F]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50G] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50H] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50I]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50J] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50K] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50L]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50M] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 50N] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) administered alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with these cells are shown. Each αENPP3×αCD3 bsAb, at low and / or high doses, was able to enhance the allogeneic antitumor effect of T cells against KU812 cells and was well combined with PD-1 blockers. [Figure 51A]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51B] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51C] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51D]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51E] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51F] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51G]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51H] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51I] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51J]A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51K] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 51L] A) PBS, B) XENP16432 (a divalent anti-PD-1 mAb), or exemplary αENPP3×αCD3 2+1 bsAb (XENP30819, or XENP31419) alone or in combination with XENP16432, are shown. The changes in tumor volume over time (determined by caliper measurements) in NSG mice transplanted with individual RXF-393 and huPBMCs are shown. Each αENPP3×αCD3 bsAb was able to enhance the allogeneic antitumor effect of T cells against KU812 cells at low, intermediate, and / or high doses and was well combined with PD-1 blockers. [Figure 52A]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52B]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52C]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52D]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52E]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52F]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52G]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52H]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52I]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52J]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 52K]Several formats for use with the anti-ENPP3 × anti-CD3 bispecific antibodies disclosed herein are shown. The first is the “1+1Fab-scFv-Fc” format (also called the “bottle opener” or “triple F” format), which has a first antigen-binding domain, which is the Fab domain, and a second anti-light source-binding domain, which is the scFv domain (Figure 1A). Furthermore, “mAb-Fv”, “mAb-scFv”, “2+1Fab2-scFv-Fc” (also called the “central scFv” or “central-scFv” format), “central-Fv”, “one-arm central-scFv”, “one-scFv-mAb”, “scFv-mAb”, “dual scFv”, “Trident”, and non-heterodimeric bispecific formats are all shown. The scFv domain shown in Figure 49 may be either a variable heavy chain-(any linker)-variable light chain or a variable light chain-(any linker)-variable heavy chain, from the N-terminus to the C-terminus. Furthermore, in the case of a one-arm scFv-mAb, the scFv can bind to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. In certain embodiments, “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the ENPP3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the CD3 binding domain. In some embodiments, “Anti-antigen 1” in Figure 52 refers to the CD3 binding domain, and “Anti-antigen 2” in Figure 52 refers to the ENPP3 binding domain. Either of the disclosed ENPP3 binding domains and CD3 binding domains can be included in the bispecificity format of Figure 52. [Figure 53]This document provides schematic diagrams of heterodimeric Fc proteins described herein, including 2:1 Fab2-scFv-Fc, 1:1 Fab=scFv-Fc, Y / Z-Fc (e.g., non-target interleukin-Fc), anti-X×Y / ZF (e.g., target interleukin-Fc)c, and one-arm Fc proteins. [Figure 54] Based on entry 3AVE in the Protein Data Bank, we provide a structural model of the CH3-CH3 interface constructed using MOE. A novel set of Fc substitutions can achieve heterodimer yields exceeding 95% with minimal change in thermal stability. [Figure 55] This shows isosteric substitutions used to minimize the impact on the tertiary structure. Differences in the isoelectric point of the Fc region allow for or facilitate the direct purification of Fc heterodimers. [Figure 56] This demonstrates that hinges and CH2 substitutions invalidate the FcγR bond. [Figure 57A] Figures 57A-C show that the 2:1 Fab2-scFv-Fc format enables targeting of tumor antigens with low density on normal cells. By adjusting the valence of TAA and the affinity of TAA / CD3, selective cytotoxicity of cell lines mimicking cancer and normal tissues is possible (high / low antigen density). The adjusted 2:1 bispecific antibody reduces interference by soluble antigens and also reduces cytokine release. Figure 57A shows that by adjusting the valence of FAP and the affinity of FAP / CD3, selective cytotoxicity of cell lines mimicking cancer and normal tissues is possible (high / low antigen density). XENP23535 represents an adjusted 1:1 format targeting FAP. XENP25967 represents an adjusted 2:1 format targeting FAP. [Figure 57B]We demonstrate that selective cytotoxicity (high / low antigen density) of cell lines mimicking cancer and normal tissues is possible by adjusting the valence of SSTR2 and the affinity of SSTR2 / CD3. XENP18087 represents a modified 1:1 format that targets SSTR2. XENP30458 represents a modified 2:1 format that targets SSTR2. [Figure 57C] We demonstrate that selective cytotoxicity (high / low antigen density) of cell lines mimicking cancer and normal tissues is possible by adjusting the valence of ENPP3 and the affinity of ENPP3 / CD3. XENP28925 represents a modified 1:1 format that targets ENPP3. XENP31149 represents a modified 2:1 format that targets ENPP3. [Figure 58] The advantages of using the method described herein for the study-scale production of heterodimeric Fc proteins are demonstrated. This method is useful for the direct production of heterodimeric Fc proteins. [Figure 59] The establishment of stable cell lines results in clones with high titer and a high proportion of heterodimers. The top clones had yields of 1–2 g / L in shaking flasks and a heterodimer content of approximately 90%. The data were obtained after only a standard protein A purification process. [Figure 60] A) Induction of RTCC in A549 cells transfected with SSTR2 (high density, medium density, low density) using XENP18087 or B) XENP30458. [Figure 61] This study shows A) a decrease in the number of target cells and B) the release of IL-6, C) TNFα, D) IFNγ, and E) IL-1β by effector cells after 48 hours of incubation of CFSE-labeled SSTR2+[COR-L279] target cells with human PBMCs (effector:target ratio of 20:1) in the presence of XENP18087 or XENP30458. [Figure 62A]Sequences of exemplary 1:1 and 2:1 tuned TAA×CD3 bispecificity sequences described herein. Anti-TTA (e.g., anti-FAP, anti-SSTR2, and anti-ENPP3) components such as a variable region, an anti-CD3 component (variable region, constant / Fc region), and a linker are shown. Linkers are double-underlined (although linkers can be replaced with other linkers, as will be understood by those skilled in the art), and slashes ( / ) indicate the boundary between the variable region, constant / Fc region, and linker. CDRs are underlined. In some embodiments, the 1:1 format TAA×CD3 bispecificity sequences are XENP23535, XENP18087, or XENP28925. In some embodiments, the 2:1 format TAA×CD3 bispecificity sequences are XENP25967, XENP30458, and XENP31149. [Figure 62B] Sequences of exemplary 1:1 and 2:1 tuned TAA×CD3 bispecificity sequences described herein. Anti-TTA (e.g., anti-FAP, anti-SSTR2, and anti-ENPP3) components such as a variable region, an anti-CD3 component (variable region, constant / Fc region), and a linker are shown. Linkers are double-underlined (although linkers can be replaced with other linkers, as will be understood by those skilled in the art), and slashes ( / ) indicate the boundary between the variable region, constant / Fc region, and linker. CDRs are underlined. In some embodiments, the 1:1 format TAA×CD3 bispecificity sequences are XENP23535, XENP18087, or XENP28925. In some embodiments, the 2:1 format TAA×CD3 bispecificity sequences are XENP25967, XENP30458, and XENP31149. [Figure 62C]Sequences of exemplary 1:1 and 2:1 tuned TAA×CD3 bispecificity sequences described herein. Anti-TTA (e.g., anti-FAP, anti-SSTR2, and anti-ENPP3) components such as a variable region, an anti-CD3 component (variable region, constant / Fc region), and a linker are shown. Linkers are double-underlined (although linkers can be replaced with other linkers, as will be understood by those skilled in the art), and slashes ( / ) indicate the boundary between the variable region, constant / Fc region, and linker. CDRs are underlined. In some embodiments, the 1:1 format TAA×CD3 bispecificity sequences are XENP23535, XENP18087, or XENP28925. In some embodiments, the 2:1 format TAA×CD3 bispecificity sequences are XENP25967, XENP30458, and XENP31149. [Figure 62D] Sequences of exemplary 1:1 and 2:1 tuned TAA×CD3 bispecificity sequences described herein. Anti-TTA (e.g., anti-FAP, anti-SSTR2, and anti-ENPP3) components such as a variable region, an anti-CD3 component (variable region, constant / Fc region), and a linker are shown. Linkers are double-underlined (although linkers can be replaced with other linkers, as will be understood by those skilled in the art), and slashes ( / ) indicate the boundary between the variable region, constant / Fc region, and linker. CDRs are underlined. In some embodiments, the 1:1 format TAA×CD3 bispecificity sequences are XENP23535, XENP18087, or XENP28925. In some embodiments, the 2:1 format TAA×CD3 bispecificity sequences are XENP25967, XENP30458, and XENP31149. [Figure 63] The sequence for the SSTR2 binding domain [αSSTR2]_H1.24_L1.30 is shown. [Modes for carrying out the invention]

[0081] I. Overview Anti-bispecific antibodies that co-link CD3 to tumor antigen targets are used to redirect T cells to attack and lyse targeted tumor cells. Examples include the Bite® and DART formats, which monovalently link CD3 to tumor antigens. While CD3-targeted approaches show considerable promise, a common side effect of such therapies is the production of associated cytokines, often leading to toxic cytokine release syndrome. Because the anti-CD3 binding domain of bispecific antibodies links to all T cells, a highly cytokine-producing CD4 T cell subset is recruited. Furthermore, this CD4 T cell subset includes regulatory T cells, and their recruitment and proliferation can lead to immunosuppression and negatively impact long-term tumor suppression. In addition, these formats do not contain an Fc domain, indicating a very short serum half-life in patients.

[0082] Novel anti-CD3 × anti-ENPP3 (also known as anti-ENPP3 × anti-CD3, αCD3 × αENPP3, or αENPP3 × αCD3) heterodimer bispecific antibodies and methods of using such antibodies for the treatment of cancer are provided herein. In particular, anti-CD3, anti-ENPP3 bispecific antibodies in various formats, such as those shown in Figures 15A and 15B, are provided herein. These bispecific antibodies are useful for the treatment of cancers, especially cancers with increased ENPP3 expression, such as renal cell carcinoma. Such antibodies are used to direct CD3+ effector T cells to ENPP3+ tumors, thereby enabling CD3+ effector T cells to attack and lyse ENPP3+ tumors.

[0083] Furthermore, in some embodiments, this disclosure provides bispecific antibodies having different binding affinities to human CD3 that can modify or reduce the potential side effects of anti-CD3 therapy. Specifically, in some embodiments, the antibodies described herein provide an antibody construct comprising an anti-CD3 antigen-binding domain that is a “strong” or “high affinity” conjugate to CD3 (e.g., a heavy chain and light variable domain indicated as H1.30_L1.47 (optionally, including a charged linker as needed)) and also binds to ENPP3. In other embodiments, the antibodies described herein provide an antibody construct comprising an anti-CD3 antigen-binding domain that is a “light” or “low affinity” conjugate to CD3. Additional embodiments provide an antibody construct comprising an anti-CD3 antigen-binding domain having an intermediate or “medium” affinity to CD3 that also binds to ENPP3. While a very large number of anti-CD3 antigen-binding domains (ABDs) can be used, particularly useful embodiments utilize six different anti-CD3 ABDs, which can be used in two scFv directions as described herein. Affinity is generally measured using a Biacore assay.

[0084] It should be understood that the “high, medium, and low” anti-CD3 sequences provided herein can be used in various heterodimerization formats, as shown in Figures 15A and 15B. Generally, due to the potential side effects of T cell recruitment, exemplary embodiments utilize a format that binds to CD3 only in a monovalent state, as shown in Figures 15A and 15B, where the format shown herein is CD3 ABD, which is scFv, as fully described herein. In contrast, the bispecific antibodies of interest can bind to ENPP3 in either a monovalent (e.g., Figure 15A) or bivalent (e.g., Figure 15B) state.

[0085] Compositions comprising an ENPP3 binding domain, including an antibody having such an ENPP3 binding domain (e.g., an ENPP3 × CD3 bispecific antibody), are provided herein. A target antibody comprising such an ENPP3 binding domain favorably induces a variety of different immune responses depending on the specific ENPP3 binding domain used. For example, the target antibody exhibits differences in selectivity for cells with different ENPP3 expression, efficacy against ENPP3-expressing cells, ability to induce cytokine release, and sensitivity to soluble ENPP3. Such ENPP3 binding domains and associated antibodies are used, for example, in the treatment of ENPP3-related cancers.

[0086] Accordingly, in one embodiment, heterodimer antibodies that bind to two different antigens are provided herein, for example, the antibody is “bispecific” in that it binds to two different target antigens described herein, generally ENPP3 and CD3. These heterodimer antibodies can bind to these target antigens in either a monovalent (e.g., having a single antigen-binding domain such as a pair of variable heavy-chain and variable light-chain domains) or a bivalent (having two antigen-binding domains, each independently binding to the antigen). In some embodiments, the heterodimer antibodies provided herein include one CD3-binding domain and one ENPP3-binding domain (e.g., a heterodimer antibody in the “1+1Fab-scFv-Fc” format described herein). In other embodiments, the heterodimer antibodies provided herein include one CD3-binding domain and two ENPP3-binding domains (e.g., a heterodimer antibody in the “2+1Fab2-scFv-Fc” format described herein). The heterodimer antibodies provided herein are based on the use of different monomers containing amino acid substitutions that “distort” heterodimer formation on homodimers, as will be outlined more fully below, and are linked to “pI variants” that enable the simple purification of heterodimers detached from homodimers, as will also be outlined below. The heterodimer bispecific antibodies provided generally depend on the use of an engineered or variant Fc domain that can self-assemble in producing cells to produce heterodimer proteins, and on methods for generating and purifying such heterodimer proteins.

[0087] II. Nomenclature The antibodies provided herein are listed in several different formats. In some cases, each monomer of a particular antibody is given a unique "XENP" number, although longer sequences may contain shorter ones, as can be dissociated in the art. For example, the "scFv-Fc" monomer in a 1+1 Fab-scFv-Fc format antibody may have a first XENP number, but the scFv domain itself may have a different XENP number. Since some molecules have three polypeptides, the XENP number is used as a name along with the constituents. Thus, the molecule XENP29520 in 2+1 Fab2-scFv-Fc format contains three sequences (see Figure 19A): 1) "Fab-Fc heavy chain" monomer; 2) "Fab-scFv-Fc heavy chain" monomer; and 3) "light chain" monomer or equivalent, which can be readily identified by those skilled in the art through sequence alignment. These XENP numbers are found in the sequence listings and identifiers and are used in the figures. Furthermore, a single molecule containing three components generates multiple sequence identifiers. For example, the Fab list contains the complete heavy chain sequence, a variable heavy chain domain sequence, and three CDRs of the variable heavy chain domain sequence; the complete light chain sequence, a variable light chain domain sequence, and three CDRs of the variable light chain domain sequence. The Fab-scFv-Fc monomer contains the full-length sequence, a variable heavy chain domain sequence, three heavy chain CDR sequences, and an scFv sequence (including the scFv variable heavy chain domain sequence, the scFv variable light chain domain sequence, and the scFv linker). Note that in this specification, among molecules having some scFv domains, a single charged scFv linker (+H) is used, but others may be used. In addition, the nomenclature for specific antigen-binding domains (e.g., ENPP3 and CD3-binding domains) uses the format of type "Hx.xx_Ly.yy", where the numbers are unique identifiers for specific variable chain sequences. Therefore, the variable domain on the Fab side of the CD3-binding domain AN1[ENPP3]H1L1 (for example, Figure 12) is "H1L1," which indicates that the variable heavy chain domain H1 is combined with the light chain domain L1.When these sequences are used as scFv, the name "H1L1" indicates that the variable heavy-chain domain H1 is combined with the light-chain domain L1, in the VH-linker-VL direction from N to C-terminus. A molecule having the same sequences of heavy and light variable domains but in the reverse order (VL-linker-VH direction, from N-terminus to C-terminus) would be named "L1_H1.1". Similarly, different constructs can "mix and adapt" the heavy and light chains, as is evident from the sequence listings and figures.

[0088] III. Definition To facilitate a more complete understanding of this application, several definitions are provided below. Such definitions are intended to encompass grammatical equivalents.

[0089] In this specification, “ENPP3” or “Ectonucleotide pyrophosphatase / phosphodiesterase family member 3” (e.g., Genebank deposit number NP005012.2) refers to a protein belonging to a series of extrinsic enzymes that ligate to the hydrolysis of extracellular nucleotides. The ENPP3 sequence is shown, for example, in Figures 11A and 11B. ENPP3 is expressed in certain cancers, including renal cell carcinoma.

[0090] In this specification, “removal” means a reduction or elimination of activity. Therefore, for example, “cleaving the FcγR bond” means that the Fc region amino acid variant has less than 50% of the initiation bond compared to an Fc region that does not contain the particular variant, and preferably the activity is lost by more than 70-80-90-95-98%, and generally the activity is below a detectable binding level in Biacore, SPR, or BLI assays. Those particularly used in cleaving the FcγR bond are shown in Figure 5, and these are generally attached to both monomers.

[0091] As used herein, “ADCC” or “antibody-dependent cell-mediated cytotoxicity” refers to a cell-mediated response in which nonspecific cytotoxic cells expressing FcγR recognize a bound antibody on target cells, subsequently causing lysis of the target cells. ADCC correlates with binding to FcγRIIIa, and increased binding to FcγRIIIa results in increased ADCC activity.

[0092] As used herein, "ADCP" or "antibody-dependent cell-mediated phagocytosis" refers to a cell-mediated response in which nonspecific phagocytic cells expressing FcγR recognize a bound antibody on a target cell, subsequently triggering phagocytosis of the target cell.

[0093] As used herein, the term "antibody" is used in general. The antibodies described herein may take several forms as described herein, including conventional antibodies and antibody derivatives, fragments, and mimics as described herein.

[0094] 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).

[0095] Other useful antibody formats include, but are not limited to, the 1+1 Fab-scFv-Fc format and 2+1 Fab-scFv-Fc antibody formats described herein, as shown in Figure 49, as well as antibodies in the “mAb-Fv”, “mAb-scFv”, “Central-Fv”, “One-arm scFv-mAb”, “scFv-mAb”, “Dual scFv”, and “Trident” formats.

[0096] Antibody heavy chains typically comprise a variable heavy chain (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 the N-terminus to the C-terminus: VH-CH1-hinge-CH2-CH3. CH1-hinge-CH2-CH3 is collectively referred to as the heavy chain "constant domain" or "constant region," although there are five distinct categories or "isotypes" of antibodies: IgA, IgD, IgG, IgE, and IgM. Thus, as used herein, "isotype" means any subclass of immunoglobulin defined by the chemical and antigenic characteristics of their constant regions. It should be understood that therapeutic antibodies may also include hybrids of isotypes and / or subclasses. For example, as shown in U.S. Patent Application Publication No. 2009 / 0163699 incorporated by reference, the antibodies described herein include the use of human IgG1 / G2 hybrids.

[0097] In some embodiments, the antibodies provided herein include a constant domain of an IgG isotype having several subclasses, including but not limited to IgG1, IgG2, IgG3, and IgG4. The IgG subclasses of immunoglobulins have several immunoglobulin domains in the heavy chain. “Immunoglobulin (Ig) domain” as used herein means a region of immunoglobulin having a distinctly different tertiary structure. The heavy chain domain, including the constant heavy chain CH domain and hinge domain, is of interest in the antibodies described herein. In relation to IgG antibodies, each IgG isotype has three CH domains. Thus, the “CH” domains in relation to IgG are as follows: “CH1” points to positions 118–220 according to the same EU index as Kabat; “CH2” points to positions 237–340 according to the same EU index as Kabat; and “CH3” points to positions 341–447 according to the same EU index as Kabat. As shown herein and described below, pI variants may reside in one or more of the CH domains and, as discussed below, the hinge domains.

[0098] It should be noted that IgG1 has different allotypes with polymorphisms at 356(D or E) and 358(L or M). While the sequences illustrated herein use the 356D / 358M allotype, other allotypes are included herein. That is, any sequence containing an IgG1 Fc domain included herein may have 356E / 358L instead of the 356D / 358M allotype. It should be understood that therapeutic antibodies may also include isotype and / or subclass hybrids. For example, as shown in U.S. Patent Application Publication No. 2009 / 0163699, incorporated by reference, this antibody includes IgG1 / IgG2 hybrids in some embodiments.

[0099] As used herein, “Fc,” “Fc region,” or “Fc domain” means a polypeptide including the constant region of an antibody, excluding in some cases all or part of the first constant region immunoglobulin domain (e.g., CH1), and in some cases optionally including 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 the N-terminus to the 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 finding specific uses in many embodiments. In addition, in the case of the human IgG1 Fc domain, the hinge frequently includes the C220S amino acid substitution. Furthermore, in the case of the human IgG4 Fc domain, the hinge frequently contains the S228P amino acid substitution. Although the boundaries of the Fc region may differ, the human IgG heavy chain Fc region is typically defined as having residues E216, C226, or A231 at its carboxyl terminus, the numbering here following the EU index similar to that of Kabat. In some embodiments, amino acid modifications are made to the Fc region to alter the linkage to one or more FcγR or FcRn, for example, as will be described in more detail below.

[0100] In this specification, “heavy chain constant region” means the CH1-hinge-CH2-CH3 portion of an antibody (or its fragment) excluding the variable heavy chain domain, which in the EU numbering of human IgG1 corresponds to amino acids 118-447. In this specification, “heavy chain constant region fragment” means a heavy chain constant region with fewer amino acids from either or both of the N-terminus and C-terminus, but which still retains the ability to form dimers with another heavy chain constant region.

[0101] Another type of heavy chain Ig domain is the hinge region. In this specification, “hinge,” “hinge region,” “antibody hinge region,” or “hinge domain” refers to a mobile polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain ends at EU position 215, and the IgG CH2 domain begins at residue EU position 231. Therefore, for IgG, the antibody hinge is defined herein as containing positions 216 (E216 in IgG1) to 230 (P230 in IgG1), numbered by EU index as in Kabat. In some cases, a “hinge fragment” is used, which contains fewer amino acids at either the N-terminus or C-terminus or both of the hinge domain. As described herein, pI variants can also be constructed in the hinge region. Many of the antibodies herein have at least one cysteine ​​at position 220, by EU numbering (hinge region) replaced with serine. Generally, this modification is located on the "scFv monomer" side for most of the sequences shown herein, but it can also be located on the "Fab monomer" side, or both, to reduce disulfide formation. Specifically included in the sequences herein are those in which one or both of these cysteines are replaced (C220S).

[0102] As those skilled in the art will understand, the precise numbering and arrangement of multiple constant region domains may differ depending on the numbering system. A useful comparison of EU and Kabat numbering is provided below; see 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 (in its entirety by reference). [Table 1]

[0103] An antibody light chain generally consists of two domains: a variable light chain domain (VL) containing light chain CDR vlCDR1-3, and a constant light chain region (often referred to as CL or Cκ). The antibody light chain is usually organized from the N-terminus to the C-terminus: VL-CL.

[0104] 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 a target antigen (e.g., ENPP3 or CD3) as considered herein. As is known in the art, these CDRs generally exist as a first set of variable heavy chain CDRs (vhCDR or VHCDR) and a second set of variable light chain CDRs (vlCDR or VLCDR), each containing three CDRs: vhCDR1, vhCDR2, vhCDR3 variable heavy chain CDRs and vlCDR1, vlCDR2, vlCDR3 vhCDR3 variable light chain CDRs. The CDRs reside in the variable heavy chain domains (vhCDR1-3) and the variable light chain domains (vlCDR1-3). Variable heavy chain domains and variable light chain domains from the Fv region.

[0105] The antibodies described herein provide a number of different CDR sets. In this case, a “complete CDR set” includes three variable light chain CDRs and three variable heavy chain CDRs, e.g., vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. These may each be part of a larger variable light chain or variable heavy chain domain. Furthermore, as fully outlined herein, the variable heavy and variable light chain domains may be on separate polypeptide chains when the heavy and light chains are used (e.g., when Fab is used), or on a single polypeptide chain in the case of scFv sequences.

[0106] As those skilled in the art will understand, the exact numbering and arrangement of CDRs may differ depending on the numbering system. However, it should be understood that the disclosure of variable heavy chain and / or variable light chain sequences includes the disclosure of the relevant (unique) CDRs. Thus, the disclosure of each variable heavy 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 given below; see Lafranc et al., Dev. Comp. Immunol. 27(1):55-77 (2003). [Table 2]

[0107] Throughout this specification, the Kabat numbering system is generally used when referring to residues within 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 for the Fc region (see, e.g., Kabat et al., above (1991)).

[0108] CDRs contribute to the formation of antigen binding, or more specifically, the formation of antigen-binding domains and antibody epitope-binding sites. An "epitope" refers to a determinant that interacts with a specific antigen-binding site within 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 charge properties. A single antigen may have two or more epitopes.

[0109] An epitope may include amino acid residues directly involved in binding (also called the immunodominant components of the epitope) and other amino acid residues not directly involved in binding, such as amino acid residues that are effectively blocked by specific antigen-binding peptides, in other words, amino acid residues that lie within the footprint of the specific antigen-binding peptide.

[0110] Epitopes can be either conformational or linear. Conformational epitopes are generated by the spatial juxtaposition of amino acids from different segments of a linear polypeptide chain. Linear epitopes are generated by adjacent amino acid residues within a polypeptide chain. Conformational and non-conformational epitopes can be distinguished in that binding to the former is lost in the presence of a denaturing solvent, while binding to the latter is not.

[0111] An epitope typically contains at least three, more commonly, at least five, or eight to ten amino acids within its unique spatial structure. Antibodies that recognize the same epitope can be validated in a simple immunoassay demonstrating the ability of one antibody to block the binding of another antibody to its target antigen, e.g., "binning." As outlined below, this disclosure includes not only the enumerated antigen-binding domains and antibodies herein, but also those that compete for binding to epitopes bound by the enumerated antigen-binding domains.

[0112] In some embodiments, the six CDRs of the antigen-binding domain are contributed by variable heavy chain and variable light chain domains. In the "Fab" format, the set of six CDRs consists of two different polypeptide sequences, namely the variable heavy chain domain (vh or V). H , including vhCDR1, vhCDR2 and vhCDR3) and variable light chain domains (vl or V LThe vh and vl domains (including vlCDR1, vlCDR2, and vlCDR3) contribute, of which the C-terminus of the vh domain binds to the N-terminus of the CH1 domain of the heavy chain, and the C-terminus of the vl domain binds to the N-terminus of the constant light chain domain (thus forming the light chain). In the scFv format, the vh and vl domains are covalently linked to a single polypeptide sequence by the use of a linker ("scFv linker"), as generally outlined herein, which may be either vh-linker-vl or vl-linker-vh (starting from the N-terminus), with the former being generally preferred (including optional domain linkers on both sides, depending on the format used (e.g., from Figure 1)). Generally, the C-terminus of the scFv domain binds to the N-terminus of the hinge of the second monomer.

[0113] As used herein, “variable region” or “variable domain” means a region of immunoglobulin containing one or more Ig domains substantially encoded by any of the Vκ, Vλ, and / or VH genes that constitute the kappa, lambda, and heavy chain immunoglobulin loci, respectively, and containing a CDR that confers antigen specificity. Thus, a “variable heavy chain domain” pairs with a “variable light chain domain” to form an antigen-binding domain (“ABD”). Furthermore, each variable domain contains three hypervariable regions (“complementarity-determining regions,” “CDR”) (VHCDR1, VHCDR2, and VHCDR3 in the variable heavy chain domain; VLCDR1, VLCDR2, and VLCDR3 in the variable light chain domain) and four framework (FR) regions, arranged in the following order from the amino terminus to the carboxyl terminus: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. The hypervariable region generally consists of amino acid residues from approximately 24-34 (LCDR1, where "L" represents the light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the light chain variable region, and approximately 31-35B (HCDR1, where "H" represents the heavy chain), 50-65 (HCDR2), and 95-102 (HCDR3) in the heavy chain variable region; Kabat et al., SEQUENCES OF PROTEINS OF IMMUNOLOGICAL INTEREST, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991) and / or those residues that form the hypervariable loop (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3) in the light chain variable region and 26-32 (HCDR1), 53-55 (HCDR2), and 96-101 (HCDR3) in the heavy chain variable region; including Chothia and Lesk (1987) J.Mol.Biol.196:901-917. Specific CDRs of the present invention are listed in Table 2.

[0114] As used herein, “Fab” or “Fab region” generally refers to a polypeptide comprising two distinct polypeptide chains (e.g., VH-CH1 on one chain and VL-CL on the other), VH, CH1, VL, and CL immunoglobulin domains. Fab may refer to this region alone or to this region in relation to the bispecific antibodies described herein. In relation to Fab, Fab includes the Fv region in addition to the CH1 and CL domains.

[0115] As used herein, “Fv,” “Fv fragment,” or “Fv region” means a polypeptide containing the VL and VH domains of an ABD. An Fv region can be formatted as both a Fab (two different polypeptides, including the constant region outlined above) and an scFv, with the VL and VH domains being combined (generally with the linker discussed herein) to form an scFv.

[0116] In this specification, "single-stranded Fv" or "scFv" generally refers to a variable heavy-chain domain covalently bonded to a variable light-chain domain, which forms an scFv or scFv domain using the scFv linker discussed herein. The scFv domain may be oriented in either direction from the N-terminus to the C-terminus (VH-linker-VL or VL-linker-VH). In the sequences shown in the sequence listings and figures, the order of the VH and VL domains is indicated in the name. For example, H.X_L.Y means VH-linker-VL and L.Y_H.X means VL-linker-VH from the N-terminus to the C-terminus.

[0117] Embodiments of the antibodies provided herein include at least one scFv domain, which, although not naturally occurring, generally include a variable heavy-chain domain and a variable light-chain domain linked together by an scFv linker. As outlined herein, the scFv domain is generally oriented from the N-terminus to the C-terminus as VH-scFv linker-VL, but this can be reversed to VL-scFv linker-VH with respect to either the scFv domain (or a domain constructed using Fab-derived VH and VL sequences) by using any linker at one or both ends, depending on the format.

[0118] In this specification, “modification” means the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence, or an alteration of a portion chemically linked to a protein. For example, the modification may be an altered carbohydrate or PEG structure bound to a protein. In this specification, “amino acid modification” means the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence. For clarity, unless otherwise stated, amino acid modification always refers to amino acids encoded by DNA, such as the 20 amino acids that have codons in DNA and RNA.

[0119] In this specification, “amino acid substitution” or “substitution” means replacing an amino acid at a specific position in a parent polypeptide sequence with a different amino acid. In particular, in some embodiments, the substitution is for an amino acid that does not naturally exist at a particular position (either not naturally present in an organism or not present in any 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. To clarify, a protein that has been manipulated to alter the nucleic acid coding sequence but not the starting amino acid (e.g., replacing CGG (coding arginine) with CGA (still coding for arginine) to increase the expression level in a host organism) is not an “amino acid substitution.” That is, despite the creation of a new gene encoding the same protein, if the protein has the same amino acid at the specific position where it is initiated, it is not an amino acid substitution.

[0120] As used herein, “amino acid insertion” or “insertion” means the addition of an amino acid sequence to a specific position in the parent polypeptide sequence. For example, -233E or 233E indicates an insertion of glutamic acid after position 233 and before position 234. Furthermore, -233ADE or A233ADE indicates an insertion of AlaAspGlu after position 233 and before position 234.

[0121] As used herein, “amino acid deletion” or “deletion” 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 indicate a deletion of glutamic acid at position 233. Furthermore, EDA233- or EDA233# indicates a deletion of the sequence GluAspAla beginning at position 233.

[0122] As used herein, “variant protein,” “protein variant,” or “variant” means a protein that differs from that of the parent protein based on 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 align with the parent protein using an alignment program such as those described below. Generally, the variant proteins outlined herein (e.g., variant Fc domains) are generally 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. As used herein, “variant” also refers to a specific amino acid modification that confers a particular function (e.g., “heterodimerized variant,” “pI variant,” “removal variant,” etc.).

[0123] As described below, in some embodiments, the parent polypeptide, for example, the Fc parent polypeptide, is a human wild-type sequence such as a heavy chain constant domain or Fc region derived from IgG1, IgG2, IgG3, or IgG4. However, a human sequence having a variant can also function as a “parent polypeptide,” and may include, for example, the IgG1 / 2 hybrid described in U.S. Publication No. 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. Thus, as used herein, “antibody variant” or “variant antibody” means an antibody that is different from the parent antibody based on at least one amino acid modification, “IgG variant” or “variant IgG” means an antibody that is different from the parent IgG (which, again, is often derived from human IgG) based on at least one amino acid modification, and “immunoglobulin variant” or “variant immunoglobulin” means an immunoglobulin sequence that is different from that of the parent immunoglobulin sequence based on at least one amino acid modification. As used herein, "Fc variant" or "variant Fc" refers to a protein that contains amino acid modifications compared to the Fc domain of human IgG1, IgG2, or IgG4.

[0124] As used herein, “Fc variant” or “variant Fc” means a protein that contains amino acid modifications in its Fc domain. These 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 substituted serine at position 434 relative to the parent Fc polypeptide, and the numbering is by EU index. Similarly, M428L / N434S defines an Fc variant having substitutions M428L and N434S relative to the parent Fc polypeptide. The identity of the WT amino acids 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 in this invention relating to antibodies or derivatives and their fragments (e.g., Fc domains), unless otherwise specified, the amino acid position numbering is by EU index. The “EU index,” “Kabat-like EU index,” or “EU numbering” scheme refers to the EU antibody numbering (Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85, thus fully incorporated by reference).

[0125] Generally, a variant Fc domain has at least approximately 80, 85, 90, 95, 97, 98, or 99 percent identity with respect to the corresponding parental human IgGFc domain (using the identity algorithm discussed below (in one embodiment, the BLAST algorithm known in the art) with 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 amino acid modifications relative 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, which can be measured using known techniques described herein, such as non-denaturing gel electrophoresis.

[0126] In this specification, “protein” means at least two covalently bonded amino acids, including proteins, polypeptides, oligopeptides, and peptides. Furthermore, polypeptides used to create antibodies described herein may include synthetic derivatization of one or more side chains or terminals, glycosylation, PEGylation, circular permutation, cyclization, linking to other molecules, fusion to proteins or protein domains, and addition of peptide tags or labels.

[0127] As used herein, "residue" refers to a position in a protein and its associated amino acid feature. For example, asparagine 297 (also known as Asn297 or N297) is the residue at position 297 in the human antibody IgG1.

[0128] As used herein, "IgG subclass modification" or "isotype modification" refers to an amino acid modification that converts one amino acid in one IgG isotype to a corresponding amino acid in a different, compatible 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.

[0129] As used herein, “modifications not occurring naturally” means non-isotypic 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 occurring naturally.

[0130] As used herein, “amino acid” and “amino acid identity” mean one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0131] As used herein, “effector function” refers to 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.

[0132] As used herein, “IgG Fc ligand” means any biologically derived molecule, preferably a polypeptide, 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 (incorporated whole by reference)). Fc ligands may include undiscovered molecules that bind to Fc. Specific IgG Fc ligands are FcRn and Fc gamma receptors. As used herein, "Fc ligand" means any biologically derived molecule, preferably a polypeptide, that binds to the Fc region of an antibody to form an Fc / Fc ligand complex.

[0133] As used herein, “Fc gamma receptor,” “FcγR,” or “Fc gamma R” means any member of the protein family that binds to the Fc region of an IgG antibody and is encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI(CD64), which contains the isoforms FcγRIa, FcγRIb, and FcγRIc; FcγRII(CD32), which contains the isoforms FcγRIIa (including the allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and FcγRIIc; and FcγRIII(CD16), which contains the isoforms FcγRIIIa (including the allotypes V158 and F158), and FcγRIIIb (including the allotypes FcγRIIb-NA1 and FcγRIIb-NA2) (the whole is incorporated by reference Jefferis et al., 2002, Immunol Lett 82:57-65), 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.

[0134] As used herein, “FcRn” or “neonatal Fc receptor” means a protein that binds to the Fc region of an IgG antibody and is at least partially encoded by the FcRn gene. FcRn may be derived from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, a functional FcRn protein contains two polypeptides, often referred to as a heavy chain and a light chain. The light chain is β-2-microglobulin, and the heavy chain is encoded by the FcRn gene. Unless otherwise stated herein, FcRn or FcRn protein refers to the complex of the FcRn heavy chain and β-2-microglobulin. Various FcRn variants can be used to increase binding to the FcRn receptor and, in some cases, to increase the serum half-life. “FcRn variant” means one that increases binding to FcRn, and preferred FcRn variants are listed below.

[0135] As used herein, “parent polypeptide” means a starting polypeptide that is subsequently 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, as used herein, “parent immunoglobulin” means an unmodified immunoglobulin polypeptide that is modified to produce a variant, and as used herein, “parent antibody” means an unmodified antibody that is modified to produce a variant antibody. It should be noted that “parent antibody” includes antibodies produced by known commercially available recombinants, as outlined below. In such relationships, the “parent Fc domain” is related to the listed variants, and therefore, “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.

[0136] As used herein, “position” means a location within a protein sequence. Positions may be numbered sequentially or according to established formats, such as the EU index for antibody numbering.

[0137] As used herein, "target antigen" means a molecule that is specifically bound by an antibody-binding domain, which includes a variable region of a given antibody.

[0138] In relation to the monomers of the heterodimer antibodies described herein, “strandedness” means incorporating heterodimerizing mutations into each monomer so as to “match” double-stranded DNA, as to retain the ability to “match” and form heterodimers. For example, if several pI variants are manipulated into monomer A (e.g., to increase the pI), then steric mutations, which are “charge pairs” that can also be utilized, do not interfere with the pI variant, and for example, the charge mutations that increase the pI are placed on the same “strand” or “monomer” and retain both functions. Similarly, for “skew” variants that form sets of pairs, as outlined in more detail below, those skilled in the art will consider the pI when determining which strand or monomer one set of pairs goes into, and therefore pI separation is also maximized using the pI of the skew.

[0139] As used herein, "target cell" means a cell that expresses a target antigen.

[0140] In the context of producing bispecific antibodies using the antibodies described herein, "host cell" means a cell that contains exogenous nucleic acids encoding the components of the bispecific antibody and that can express the bispecific antibody under suitable conditions. Suitable host cells are described below.

[0141] In this specification, “wild-type” or “WT” means a naturally occurring amino acid or nucleotide sequence, including allelic mutations. WT proteins have an amino acid or nucleotide sequence that is not intentionally modified.

[0142] This specification provides several antigen-binding domains that have sequence identity with human antibody domains. Sequence identity between two similar sequences (such as 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 [homology alignment algorithm], Pearson, WR & Lipman, DJ (1988) “Improved Tools For Biological Sequence Comparison,” Proc.Natl.Acad.Sci.(USA)85:2444 [search for similarity method], or Altschul, SF et al, (1990) “Basic Local Alignment Search Sequence identity can be measured by algorithms such as Tool, "J.Mol.Biol.215:403-10, the "BLAST" algorithm (see https: / / blast.ncbi.nlm.nih.gov / Blast.cgi). When using any of the aforementioned algorithms, default parameters (window length, gap penalty, etc.) are used. In one embodiment, sequence identity is measured using the BLAST algorithm with default parameters.

[0143] The antibodies described herein are generally isolated or recombinant. As used to describe the various polypeptides disclosed herein, "isolated" means a polypeptide that has been identified and separated and / or recovered from a cell or cell culture in which it was expressed. Typically, an isolated polypeptide will be prepared by at least one purification step. An "isolated antibody" refers to an antibody that substantially lacks other antibodies having different antigen specificities. "Recombinant" means that an antibody is produced using recombinant nucleic acid techniques in a foreign host cell, and such antibodies can also be isolated.

[0144] "Specific binding" to a particular antigen or epitope, or "specifically binds to" or "is specific for" a particular antigen or epitope means a binding that is measurably different from non-specific interactions. Specific binding can be measured, for example, by determining the binding of a molecule compared to the binding of a control molecule that is a molecule of similar structure that generally does not have binding activity. For example, specific binding can be determined by competition with a control molecule similar to the target.

[0145] Specific binding to a particular antigen or epitope can be, for example, 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 greater KD, 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-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold, or greater than that for a control molecule compared to the antigen or epitope.

[0146] Furthermore, specific binding to a particular antigen or epitope can be demonstrated, for example, by an antibody whose KA or Ka for the antigen or epitope is at least 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold, or more than 20-fold, 50-fold, 10,000-fold, or greater than that of a control, where KA or Ka refers to the association rate of a particular antibody-antigen interaction. Binding affinity is generally measured using Biacore, SPR, or BLI assays.

[0147] IV.ENPP3 binding domain In one embodiment, compositions comprising an ENPP3 antigen-binding domain (ABD) and an anti-ENPP3 antibody are provided herein. A target antibody comprising such an ENPP3 antigen-binding domain (e.g., an anti-ENPP3 × anti-CD3 bispecific antibody) favorably induces a variety of different immune responses (see Examples 5 and 6). Such ENPP3-binding domains and associated antibodies are used, for example, in the treatment of ENPP3-related cancers.

[0148] As will be understood by those skilled in the art, preferred ENPP3 binding domains may include a set of six CDRs as illustrated in the sequence listing and Figures 12, 13A–13B, and 14A–14I, identified using other alignments within the variable heavy chain (VH) domain and variable light chain (VL) domain sequences, as illustrated in the sequence listing and Figures 12, 13A–13B, and 14A–14I, if underlined or if a different numbering scheme is used as shown in Table 2. Preferred ENPP3 ABDs may also include these sequences and the entire sequences of VH and VL as shown in the figures, which may be used as scFv or Fab domains.

[0149] In one embodiment, the ENPP3 antigen-binding domain comprises six CDRs of the ENPP3 ABD described herein (i.e., vhCDR1-3 and vlCDR1-3), including the figure and sequence list. In exemplary embodiments, the ENPP3 ABD includes the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha16-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It is one of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A~13B, and 14A~14I).

[0150] Provided herein are variant ENPP3 ABDS having a CDR comprising at least one modification of the ENPP3 ABD CDR disclosed herein, in addition to the parent CDR set disclosed in Figure and Sequence Listing that forms the ABD for ENPP3. In one embodiment, the ENPP3 ABD comprises a set of six CDRs having amino acid modifications 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, compared to the six CDRs of the ENPP3 ABD described herein, including Figure and Sequence Listing. In exemplary embodiments, ENPP3 ABD is as follows: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Compared to one of the six CDRs among 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I), this set includes six CDRs with 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications. In certain embodiments, the variant ENPP3 ABD can bind to the ENPP3 antigen when measured by at least one of the Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., octet assay) assays, the latter finding specific applications in many embodiments. In certain embodiments, the ENPP3 ABD can bind to the human ENPP3 antigen (see Example 5).

[0151] In one embodiment, the ENPP3 ABD includes six CDRs that are at least 90, 95, 97, 98, or 99% identical to the six CDRs of the ENPP3 ABD described herein, including figures and sequence listings. In exemplary embodiments, ENPP3 ABD is as follows: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 Includes six CDRs that are at least 90, 95, 97, 98, or 99% identical to one of six CDRs from among 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). In certain embodiments, ENPP3 ABD can bind to the ENPP3 antigen when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., octet assay), the latter finding specific applications in many embodiments. In certain embodiments, ENPP3 ABD can bind to the human ENPP3 antigen (see Figure 2).

[0152] In another exemplary embodiment, the ENPP3 ABD comprises one of the variable heavy chain (VH) domains and a variable light chain (VL) domain from the ENPP3 ABDs described herein, including the figures and sequence listings. In exemplary embodiments, the ENPP3 ABD includes the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha16-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It is one of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A~13B, and 14A~14I).

[0153] In addition to the parent ENPP3 variable heavy chain domain and variable light chain domain disclosed herein, ENPP3 ABDs are provided herein that include a variable heavy chain domain and / or a variable light chain domain which is a variant of the ENPP3 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 ENPP3 ABD described herein, including figures and sequence listings. In exemplary embodiments, the variant VH domain and / or VL domain are the following ENPP3 ABD:AN1[ENPP3]H1L1,AN1[ENPP3]H1 L1.33,AN1[ENPP3]H1 L1.77,AN1[ENPP3]H1.8 L1,AN1[ENPP3]H1.8 L1.33,AN1[ENPP3]H1 L1.77,H16-7.213,H16-9.69,H16-1.52,Ha16-1(1)23,H16-9.44,H16-1.67,Ha1 6-1(3,5)36,H16-1.86,H16-9.10,H16-9.33,H16-1.68,Ha16-1(1)1,Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I) have 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid changes from one of the six VH and / or VL domains. In certain embodiments, ENPP3 ABD can bind to ENPP3 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., octet assay), the latter finding specific applications in many embodiments. In certain embodiments, ENPP3 ABD can bind to the human ENPP3 antigen (see Example 5).

[0154] 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 ENPP3 ABD described herein, including in the figures and sequence listings. In exemplary embodiments, variant VH and / or VL domains are the following ENPP3 ABD:AN1[ENPP3]H1L1,AN1[ENPP3]H1 L1.33,AN1[ENPP3]H1 L1.77,AN1[ENPP3]H1.8 L1,AN1[ENPP3]H1.8 L1.33,AN1[ENPP3]H1 L1.77,H16-7.213,H16-9.69,H16-1.52,Ha16-1(1)23,H16-9.44,H16-1.67,Ha1 6-1(3,5)36,H16-1.86,H16-9.10,H16-9.33,H16-1.68,Ha16-1(1)1,Ha1 It is at least 90, 95, 97, 98, or 99% identical to one of the VH and / or VL samples from among 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha16-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H16-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). In certain embodiments, ENPP3 ABD can bind to ENPP3 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., octet assay), the latter finding specific applications in many embodiments. In certain embodiments, ENPP3 ABD can bind to the human ENPP3 antigen (see Example 5).

[0155] V. Antibodies In one embodiment, an antibody that binds to ENPP3 (e.g., an anti-ENPP3 antibody) is provided herein. In a particular embodiment, the antibody binds to human ENPP3 (Figure 11A). The anti-ENPP3 antibodies of interest include monospecific ENPP3 antibodies and multispecific (e.g., bispecific) anti-ENPP3 antibodies. In a particular embodiment, the anti-ENPP3 antibody has a format according to one of the antibody formats shown in Figures 15A, 15B, and 52A-52K.

[0156] In some embodiments, the composition in question includes an ENPP3-binding domain. In some embodiments, the composition includes an antibody having an ENPP3-binding domain. The antibodies provided herein include one, two, three, four, and five or more ENPP3-binding domains. In certain embodiments, the ENPP3-binding domain includes one of the vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3 sequences of ENPP3-binding domains selected from those shown in Figures 12, 13A-13B, and 14A-14I. In some embodiments, the ENPP3-binding domain includes the underlined vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3 sequences of ENPP3-binding domains selected from those shown in Figures 12, 13A-13B, and 14A-14I. In some embodiments, the ENPP3 binding domain includes a variable heavy chain domain and a variable light chain domain of the ENPP3 binding domain selected from those shown in Figures 12, 13A-13B, and 14A-14I. The ENPP3 binding domains shown in Figures 12, 13A-13B, and 14A-14I are AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8L1, AN1[ENPP3]H1.8L1.33, AN1[ENPP3]H1L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha16-1 Includes (3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha16-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha16-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H16-9.65, Ha1-1(3,5)19, and Ha16-1.80.

[0157] In one embodiment, bispecific antibodies that bind to ENPP3 and CD3 are provided herein in various formats, as outlined below and generally shown in Figures 15A and 15B. These bispecific heterodimer antibodies include an ENPP3-binding domain. In certain embodiments, the ENPP3-binding domain includes vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3 sequences of an ENPP3-binding domain selected from the group shown in Figures 12, 13A-13B, and 14A-14I. In some embodiments, the ENPP3-binding domain includes underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3-binding domain selected from those shown in Figures 12, 13A-13B, and 14A-14I.

[0158] These bispecific heterodimer antibodies bind to ENPP3 and CD3. Such antibodies include a CD3-binding domain and at least one ENPP3-binding domain. Any suitable ENPP3-binding domain can be included in the anti-ENPP3X anti-CD3 bispecific antibody. In some embodiments, the anti-ENPP3X anti-CD3 bispecific antibody includes, but is not limited to, the one shown in Figures 12, 13A-13B, and 14A-14I, and includes one, two, three, four or more ENPP3-binding domains. In certain embodiments, the anti-ENPP3X anti-CD3 antibody includes an ENPP3-binding domain containing the VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of ENPP3-binding domains selected from the group consisting of those shown in Figures 12, 13A-13B, and 14A-14I. In some embodiments, the anti-ENPP3X anti-CD3 antibody includes an ENPP3-binding domain selected from the group consisting of those shown in Figures 12, 13A-13B, and 14A-14I, comprising the underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences. In some embodiments, the anti-ENPP3X anti-CD3 antibody includes an ENPP3-binding domain comprising a variable heavy chain domain and a variable light chain domain selected from the group consisting of those shown in Figures 12, 13A-13B, and 14A-14I. In exemplary embodiments, the anti-ENPP3X anti-CD3 antibody includes an anti-ENPP3 AN1[ENPP3]_H1L1-binding domain.

[0159] The anti-ENPP3X anti-CD3 antibodies provided herein may contain any suitable CD3-binding domain. In certain embodiments, the anti-ENPP3X anti-CD3 antibody includes a CD3-binding domain comprising the VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of a CD3-binding domain selected from the group shown in Figures 10A-F. In some embodiments, the anti-ENPP3X anti-CD3 antibody includes a CD3-binding domain comprising the underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of a CD3-binding domain selected from the group shown in Figures 10A-10F. In some embodiments, the anti-ENPP3X anti-CD3 antibody includes a CD3-binding domain comprising a variable heavy chain domain and a variable light chain domain of a CD3-binding domain selected from the group shown in Figures 10A-10F. In some embodiments, the CD3 binding domain is selected from anti-CD3 H1.30_L1.47, anti-CD3 H1.32_L1.47; anti-CD3H1.89_L1.48; anti-CD3H1.90_L1.47; anti-CD3H1.33_L1.47; and anti-CD3H1.31_L1.47. As outlined herein, these anti-CD3 antigen-binding domains (CD3-ABD) can be used in scFv format in either direction (e.g., from N-terminus to C-terminus, VH-scFv linker-VL or VL-scFv linker-VH).

[0160] The antibodies provided herein contain different antibody domains. As described herein and known in the art, the heterodimer antibodies described herein contain different domains within the heavy and light chains, which may also overlap. These domains include, but are not limited to, Fc domains, CH1 domains, CH2 domains, CH3 domains, hinge domains, heavy chain constant domains (CH1-hinge-Fc domain or CH1-hinge-CH2-CH3), variable heavy chain domains, variable light chain domains, light chain constant domains, Fab domains, and scFv domains.

[0161] As shown herein, several suitable linkers (used as either domain linkers or scFv linkers) can be used to covalently bond (including conventional peptide bonds generated by recombination techniques) to the listed domains (e.g., scFv, Fab, Fc domains, etc.). An exemplary linker for attaching the domains of the antibody of interest to each other is shown in Figure 6. In some embodiments, the linker peptide may primarily consist of the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should be long enough to join the two molecules in such a way that they assume the correct conformations relative 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, linkers of 1 to 20 amino acids long may be used, and in some embodiments, about 5 to about 10 amino acids are used. Useful linkers include, for example, glycine-serine polymers, glycine-alanine polymers, alanine-serine polymers, and other mobile linkers, including (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n (where n is an integer of at least 1 (and generally 3-4)), some of which are shown in Figures 5 and 6. 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 useful as linkers.

[0162] Other linker sequences may contain any sequence of a CL / CH1 domain of any length, but may not contain 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, such as Ig-like proteins (e.g., TCR, FcR, KIR), hinge region-derived sequences, and other native sequences from other proteins.

[0163] In some embodiments, the linker is a “domain linker” used to link any two domains outlined herein together. For example, in Figure 15B, there may be a domain linker that links the C-terminus of the CH1 domain of Fab to the N-terminus of scFv, and another arbitrary domain linker that links the C-terminus of scFv to the CH2 domain (although in many embodiments, a hinge is used as this domain linker). Any suitable linker can be used, but many embodiments utilize glycine-serine polymers as domain linkers, for example, including (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer of at least 1 (generally 3-4-5), as well as any peptide sequence that allows for the recombination of two domains having sufficient length and flexibility so that each domain retains its biological function. In some cases, charged domain linkers can be used, as used in some embodiments of scFv linkers, paying attention to “strandedness” outlined below. Exemplary useful domain linkers are shown in Figure 6.

[0164] Specifically referring to the domain linkers used to connect scFv domains to Fc domains in "2+1" format, there are several domain linkers that find specific uses, including "Full Hinge C220S Variant," "Flex Half Hinge," "Charged Half Hinge 1," and "Charged Half Hinge 2," as shown in Figure 6.

[0165] In some embodiments, the linker is an “scFv linker” used to covalently bond the VH and VL domains discussed herein. In many cases, the scFv linker is a charged scFv linker, some of which are shown in Figure 5. Thus, in some embodiments, the antibodies described herein further provide a charged scFv linker to facilitate the separation of pI between the first monomer and the second monomer. That is, by incorporating either a positive or negative charged scFv linker (or both in the case of a scaffold using scFvs on different monomers), this makes it possible to alter the pI of the monomer containing the charged linker without further altering the Fc domain. These charged linkers can be substituted in any scFv containing a standard linker. Also, 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 change in pI. For example, as discussed herein, in order to produce a 1+1Fab-scFv-Fc format heterodimer antibody, the original pI of the Fv region for each of the desired antigen-linking domains is calculated, one is selected to produce the scFv, and depending on the pI, either a positive or negative linker is selected.

[0166] Charged domain linkers can also be used to increase the pI separation of monomers of the antibodies described herein, and thus those shown in Figure 5 can be used in any embodiment of this specification in which a linker is utilized.

[0167] In particular, the format shown in Figures 15A and 15B is an antibody commonly referred to as a "heterodimal antibody," meaning that the protein has at least two associated Fc sequences that self-assemble into a heterodimer Fc domain, and at least two Fv regions, whether as Fab or scFv.

[0168] The provided ENPP3-binding domains can be incorporated, for example, into standard immunoglobulins, as well as into any useful antibody format, including the 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv formats provided herein. Other useful antibody formats include, but are not limited to, antibodies in the “mAb-Fv”, “mAb-scFv”, “Central-Fv”, “One-arm scFv-mAb”, “scFv-mAb”, “Dual scFv”, and “Trident” formats, as shown in Figures 52A–52K.

[0169] In some embodiments, the target antibody comprises one or more of the ENPP3 ABDs provided herein. In some embodiments, the antibody comprises one ENPP3 ABD. In other embodiments, the antibody comprises two ENPP3 ABDs. In exemplary embodiments, the ENPP3 ABD includes the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha16-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It includes one variable heavy chain domain and one variable light chain domain from among 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).In exemplary embodiments, the ENPO3 ABD includes the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It is one of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A~13B, and 14A~14I).

[0170] In an exemplary embodiment, the antibody is a bispecific antibody comprising one or two ENPP3 ABDs, including any of the ENPP3 ABDs provided herein. Examples of bispecific antibodies comprising such ENPP3 ABDs include, for example, 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc bispecific format antibodies. In an exemplary embodiment, the ENPP3 ABD is one of the following B7H3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). In an exemplary embodiment, the ENPP3 binding domain is a Fab. In some embodiments, such bispecific antibodies are heterodimeric bispecific antibodies comprising any of the heterodimerization skewer variants, pI variants, and / or removal variants described herein.

[0171] A. Chimeric and Humanized Antibodies In certain embodiments, the antibodies described herein include heavy chain variable regions derived from specific germline heavy chain immunoglobulin genes and / or light chain variable regions derived from specific germline light chain immunoglobulin genes. For example, such antibodies may include or consist of human antibodies that include 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 as such by comparing the amino acid sequence of the human antibody with the amino acid sequence of a human germline immunoglobulin and selecting the human germline immunoglobulin that is the sequence closest to the sequence of the human antibody (i.e., the highest identity %) (using the methods outlined herein). Human antibodies that are "products" of or "derived from" a specific human germline immunoglobulin sequence may include amino acid differences compared to the germline sequence, for example, due to naturally occurring somatic mutations or the intentional introduction of site-directed mutations. However, humanized antibodies are typically at least 90% identical in amino acid sequence 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 certain 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% identical. Typically, a humanized antibody derived from a particular human germline sequence shows no more than 10-20 amino acid differences from the amino acid sequence encoded by a human germline immunoglobulin gene (the number of variants is generally small before the introduction of any skew, pI, and exclusion mutations described herein, i.e., before the introduction of the variants described herein).In certain cases, a humanized antibody may differ from the amino acid sequence encoded by a germline immunoglobulin gene by no more than 5 amino acids, or even by no more than 4, 3, 2, or 1 amino acid (similarly, prior to the introduction of any skews, pI, and removal mutations herein, i.e., prior to the introduction of the variants described herein, the number of variants is generally low).

[0172] In one embodiment, the parent antibody is affinity matured as is known in the art. Structure-based methods can be used for humanization and affinity maturation, as described, for example, in USSN11 / 004,590. Antibody variable regions may be humanized and / or affinity matured using selection-based methods, including, but not limited to, the methods 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 of which are incorporated by reference in their entirety. Other humanization methods may include transplanting only a portion of the CDRs, including, but not limited to, the methods 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 of which are incorporated by reference in their entirety.

[0173] B. Heterodimeric Antibodies In an exemplary embodiment, the bispecific antibodies provided herein are heterodimeric bispecific antibodies that include two variant Fc domain sequences. Such variant Fc domains include amino acid modifications to facilitate self-assembly and / or purification of the heterodimeric antibody.

[0174] A continuing challenge in antibody technology is the demand for "bispecific" antibodies that bind simultaneously to two different antigens, generally bringing different antigens into proximity and potentially leading to new functions and therapeutic approaches. These antibodies are typically created by incorporating the genes for each heavy and light chain into the host cell. This generally results in the formation of not only two homodimers (AA and BB (excluding the issue of light chain heterodimers)) but also the desired heterodimer (AB). However, the main obstacle in bispecific antibody formation is the difficulty in skewing heterodimer formation over homodimer formation and / or purifying heterodimeric antibodies away from homodimers.

[0175] Many mechanisms exist that can be used to generate the target heterodimer antibody. 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 collectively called “heterodimerizing variants.” As will be discussed below, heterodimerizing variants include “skew” variants (e.g., the “knob-and-hole” and “charge pair” variants discussed below), as well as “pI variants” that enable the purification of heterodimers from homodimers. As is generally described in U.S. Patent No. 9,605,084, which is incorporated herein by reference in its entirety, and more specifically, regarding the discussion of heterodimerization variants, useful mechanisms for heterodimerization include “knob and hole” (“KIH”) as described in U.S. Patent No. 9,605,084, “electrostatic steering” or “charge pair” as described in U.S. Patent No. 9,605,084, the pI variant as described in U.S. Patent No. 9,605,084, and further general Fc variants outlined in U.S. Patent No. 9,605,084 and below.

[0176] The following section discusses in more detail the heterodimerization variants useful for the formation and purification of target heterodimer antibodies (e.g., bispecific antibodies).

[0177] 1. Scubarian In some embodiments, the heterodimer antibody comprises a scubarian, which is one or more amino acid modifications in the first Fc domain (A) and / or the second Fc domain (B), and the scubarian promotes the formation of an Fc dimer (AB) containing both the first and second Fc domains over an Fc homodimer (AA or BB). A suitable scubarian is shown in Figure 29 of U.S. Patent Application Publication No. 2016 / 0355608 and is incorporated herein by reference, in whole, and in particular, with regard to the disclosure of scubarians, as well as in Figures 1A–1E and 4.

[0178] One mechanism, commonly referred to in the art as "knobs and hole," refers to amino acid engineering that produces steric effects favoring heterodimerization and unfavoring homodimerization, and may be used arbitrarily; this is often referred to as "knobs and hole" and is described in USSN61 / 596,846, Ridgway et al., Protein Engineering 9(7):617(1996); Atwell et al., J.Mol.Biol.1997 270:26; USSN8,216,805, all of which are incorporated herein by reference in their entirety. These figures identify several "monomer A-monomer B" pairs that depend on "knobs and hole." Furthermore, as described in Merchant et al., Nature Biotech.16:677(1998), these "knobs and hole" mutations can be combined with disulfide bonding to skew formation for heterodimerization.

[0179] Further mechanisms used for the formation of heterodimers, as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637(2010), are all incorporated herein by reference and are sometimes referred to as “electrostatic steering.” This is sometimes referred to herein as “charge pairing.” In this embodiment, electrostatics is used to skew the formation toward heterodimerization. As will be understood by those skilled in the art, these can also affect pI, i.e., purification, and therefore in some cases can be considered pI variants. However, since these are generated to force heterodimerization and are not used as a means of purification, they are classified as “steric variants.” These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (for example, these are "monomer correspondence sets"), and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0180] In some embodiments, scubarians favorably and simultaneously promote heterodimerization based on both “knob-and-hole” and “electrostatic steering” mechanisms. In some embodiments, heterodimerizing antibodies comprise one or more sets of such heterodimerizing scubarians. These variants form “pairs” of “sets,” i.e., a pair from one set is incorporated into the first monomer, and a pair from the other set is incorporated into the second monomer. It should be noted that these sets do not necessarily behave as “knob-and-hole” variants, but rather have a one-to-one correspondence between residues of one monomer and residues of the other monomer. That is, these pairs of sets may instead form an interface between two monomers that promotes heterodimerization and not homodimerization, and the percentage of heterodimers spontaneously formed under biological conditions may exceed 90%, rather than the expected 50% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B). An exemplary heterodimerized "skew" variant is shown in Figure 4. In exemplary embodiments, the heterodimerized antibody comprises 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. In exemplary embodiments, the heterodimer antibody contains the "S364K / E357Q:L368D / K370S" amino acid substitution set. In terms of nomenclature, the "S364K / E357Q:L368D / K370S" pair 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 "twist" of these pairs depends on the initial pI.

[0181] In some embodiments, the scuba riants provided herein are incorporated independently into one or both of the first and second Fc domains of the IL-15-Fc fusion protein, along with other modifications to one or both of the first and second Fc domains of the heterodimer antibody, including but not limited to other scuba riants (e.g., Figure 37 of U.S. Patent Application Publication 2012 / 0149876, in particular, as incorporated herein by reference for the disclosure of scuba riants), pI variants, isotype variants, FcRn variants, elimination variants, and the like. Furthermore, individual modifications can also be independently and optionally included in or excluded from the heterodimer antibody.

[0182] Further variants of monomer A and monomer B can be combined in any quantity, selectively and independently, with other variants such as the pI variant outlined herein or other stereovariants shown in Figure 37 of US2012 / 0149876, whose figures, descriptions, and sequence numbers are all expressly incorporated herein by reference.

[0183] In some embodiments, the steric mutations outlined herein may optionally and independently incorporate any pI mutation (or other mutations such as Fc mutations, FcRn mutations, etc.) into one or both monomers, and may independently and optionally be included in or excluded from the antibody proteins described herein.

[0184] A list of preferred scuba antlers can be seen in Figures 1A-1E, and Figure 4 shows some pairs of particularly useful ones in many embodiments. Particularly used in many embodiments are pairs of sets including, but are not limited to, S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L and K370S:S364K / E357Q. Regarding nomenclature, the pair "S364K / E357Q:L368D / K370S" means that one monomer has the double variant set S364K / E357Q, and the other has the double variant set L368D / K370S.

[0185] 2. pI (isoelectric point) variants of heterodimers In some embodiments, the heterodimer antibody includes a purified variant that favorably enables the separation of the heterodimer antibody (e.g., an anti-ENPP3 × anti-CD3 bispecific antibody) from a homodimer protein.

[0186] Several fundamental mechanisms exist that can facilitate the purification of heterodimer antibodies. For example, modification of one or both of antibody heavy chain monomers A and B so that each monomer has a different pI allows for isoelectric focusing of heterodimer AB antibodies from monomer AA and BB proteins. Alternatively, some scaffold formats, such as the "1+1Fab-scFv-Fc" format and the "2+1Fab2-scFv-Fc" format, also allow for separation based on size. As mentioned above, it is also possible to "skew" the formation of heterodimers on homodimers using scuba riants. Therefore, combinations of heterodimerizing scuba riants and pI variants find specific applications in the heterodimer antibodies provided herein.

[0187] Furthermore, as will be outlined more fully below, depending on the format of the heterodimeric antibody, pI variants and / or domain linkers contained within the constant region and / or Fc domain of the monomer may be used. In some embodiments, the heterodimeric antibody may include additional modifications for alternative functions, such as Fc, FcRn, and KO variants, which can also produce pI changes.

[0188] In some embodiments, the heterodimer antibodies of interest provided herein include at least one monomer having one or more modifications that alter the pI of the monomer (i.e., a “pI variant”). Generally, as will be understood by those skilled in the art, pI variants fall into two common categories: those that increase the pI of a protein (basic changes) and those that decrease the pI of a 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 changed, one to be more basic and the other to be more acidic.

[0189] Depending on the heterodimer antibody format, the pI mutation may be contained within the constant and / or Fc domain of the monomer, or a charged linker, domain linker, or scFv linker may be used. That is, antibody formats utilizing scFv, such as "1+1Fab-scFv-Fc," may include a charged scFv linker (either positive or negative) to provide an additional pI boost for purification purposes. As will be understood by those skilled in the art, the antibodies described herein also provide pI variants present in one or both monomers, and / or charged domain linkers, although some 1+1Fab-scFv-Fc formats are useful with only a charged scFv linker and without additional pI adjustment. In addition, further amino acid manipulation for alternative functionality may also confer pI changes such as Fc, FcRn, and KO variants.

[0190] In heterodimeric antibodies that utilize pI as a separation mechanism to enable the purification of heterodimeric proteins, amino acid variants are introduced into one or both monomeric polypeptides. That is, the pI of one monomer (referred to herein as "monomer A" for simplicity) can be manipulated to move away from monomer B, or the changes in both monomers can be altered by increasing the pI of monomer A and decreasing the pI of monomer B. As will be fully outlined below, changes in the pI of either or both monomers can be carried out by removing or adding charged residues (e.g., substituting a neutral amino acid with a positively or negatively charged amino acid residue, e.g., glycine to glutamic acid), by changing charged residues from positive or negative to the opposite charge (e.g., aspartic acid to lysine), or by changing charged residues to neutral residues (e.g., loss of charge, lysine to serine). Some of these variants are shown in Figures 3 and 4.

[0191] Therefore, in some embodiments, the heterodimer antibody of interest includes amino acid modifications in a constant region that alter the isoelectric point (pI) of at least one, if not both, of the monomers of the dimer protein by incorporating an amino acid substitution ("pI variant" or "pI substitution") into one or both monomers. As shown herein, the separation of a heterodimer from two homodimers can be achieved when the pIs of the two monomers differ by a small amount of 0.1 pH units, and differences of 0.2, 0.3, 0.4, and 0.5 or more are all used in the antibodies described herein.

[0192] As will be understood by those skilled in the art, the number of pI mutations to be included in each monomer or both monomers to obtain good separation will depend in part on the starting pI of the components, e.g., in the 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc formats, on the starting pI of the scFv and Fab of interest. That is, to determine which monomer to manipulate or which “direction” (e.g., more positive or more negative), the Fv sequences of the two target antigens are calculated and a determination is made therefrom. As is known in the art, different Fvs will have different starting pIs utilized in the antibodies described herein. Generally, as outlined herein, the pI is manipulated to provide at least about 0.1 log total pI difference per monomer, with 0.2-0.5 being preferred as outlined herein.

[0193] By using the constant region of the heavy chain, a more modular approach for designing and purifying bispecific proteins containing antibodies is provided when pI variants are used to achieve heterodimers. Thus, in some embodiments, heterodimerization mutations (including skews and pI heterodimerization variants) are not included in the variable region, and thus each individual antibody has to be manipulated. In addition, in some embodiments, the potential immunogenicity resulting from pI variants is significantly reduced by transferring pI variants from different IgG isotypes such that the pI changes without introducing significant immunogenicity. Thus, a further problem to be solved is the elucidation of low pI constant domains with high human sequence content, e.g., minimizing or avoiding non-human residues at any given position. Instead of or in addition to isotype substitution, the potential immunogenicity resulting from pI variants is significantly reduced by utilizing isoelectronic substitutions (e.g., Asn to Asp and Gln to Glu).

[0194] As will be discussed below, the secondary benefits that can occur with this pI engineering are also the extension of serum half-life and increased FcRn ligation. Specifically, as described in U.S. Patent Application Publication US2012 / 0028304 (which is incorporated herein by reference in its entirety), 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 extending serum half-life also facilitate pI changes for purification.

[0195] Furthermore, the pI variant exhibits remarkable ability to exclude, minimize, and distinguish homodimers when present, thus providing further benefits to the analysis and quality control processes of bispecific antibodies. Similarly, the ability to reliably test the reproducibility of heterodimeric antibody production is important.

[0196] Generally, the embodiments used in particular depend on a set of variants containing a scuba riant, which, in combination with a pI variant that increases the pI difference between two monomers, promotes heterodimerization in preference to homodimerization, thereby facilitating the removal of homodimers and the purification of heterodimers.

[0197] Exemplary combinations of pI variants are shown in Figures 4 and 5 and 30 of U.S. Patent Application Publication No. 2016 / 0355608, all of which are incorporated herein by reference, in general and in particular, with respect to the disclosure of pI variants. Preferred combinations of pI variants are shown in Figures 1 and 2. As outlined herein and shown in the figures, these variations are shown in comparison to IgG1, but all isotypes can be modified in this way, as can isotype hybrids. R133E and R133Q may also be used when the heavy chain constant domains are derived from IgG2-4.

[0198] In one embodiment, a preferred combination of pI variants comprises one monomer (negative Fab side) containing the 208D / 295E / 384D / 418E / 421D variant (N208D / Q295E / N384D / Q418E / N421D when compared to human IgG1) and a second monomer (positive scFv side) containing a positively charged scFv linker containing (GKPGS)4 (SEQ ID NO: XX). 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 a CH1 domain (e.g., in the case of an antibody that does not utilize a CH1 domain in one of its domains), a preferred negative pI mutation Fc set contains the 295E / 384D / 418E / 421D mutation (Q295E / N384D / Q418E / N421D when compared to human IgG1).

[0199] Therefore, in some embodiments, one monomer has a set of substitutions from Figure 2, and the other monomer has a charged linker (either in the format of a charged scFv linker, as the monomer includes an scFv or a charged domain linker, as shown in the format, which can be selected from those shown in Figure 5).

[0200] In some embodiments, modifications occur 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 variants, particularly substitutions, can occur at one or more positions 216–230, having one, two, three, four, or five mutations that find use. Similarly, all possible combinations are intended to be alone or with other pI variants in other domains.

[0201] Specific substitutions used to reduce the pI of a hinge domain include, but are not limited to, deletions at position 221, unnatural valine or threonine at position 222, deletion at position 223, unnatural glutamate at position 224, deletion at position 225, deletion at position 235, and deletion or unnatural alanine at position 236. In some cases, only pI substitutions are performed in the hinge domain, while in other examples, these substitutions are added in any combination to other pI variants in other domains.

[0202] In some embodiments, mutations can be introduced within 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 233–236 may be made to increase effector function (along with 327A) in the IgG2 backbone. Similarly, all possible combinations of these 14 positions can be made, for example, = may include a variant Fc domain with pI substitutions of CH2 at positions 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10.

[0203] Specific substitutions used to reduce the pI of the CH2 domain include, but are not limited to, unnatural glutamine or glutamate at position 274, unnatural phenylalanine at position 296, unnatural phenylalanine at position 300, unnatural valine at position 309, unnatural glutamate at position 320, unnatural glutamate at position 322, unnatural glutamate at position 326, unnatural glycine at position 327, natural glutamate at position 334, unnatural threonine at position 339, and all possible combinations within CH2 and with other domains.

[0204] In this embodiment, modifications can be selected independently and optionally 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, unnatural glutamine or glutamate at position 355, unnatural serine at position 384, unnatural asparagine or glutamate at position 392, unnatural methionine at position 397, unnatural glutamate at position 419, unnatural glutamate at position 359, unnatural glutamate at position 362, unnatural glutamate at position 389, unnatural glutamate at position 418, unnatural glutamate at position 444, and deletion or unnatural aspartate at position 447.

[0205] Generally, as will be understood by those skilled in the art, pI variants fall into two common categories: those that increase the pI of a protein (basic changes) and those that decrease the pI of a 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 change, one to be more basic and the other to be more acidic.

[0206] Preferred combinations of pI variants are shown in Figure 4. While these variations are shown in comparison to IgG1, as outlined herein and shown in the figure, all isotypes can be modified in this way, as can isotype hybrids. R133E and R133Q can also be used when the heavy chain constant domains are derived from IgG2-4.

[0207] In one embodiment, for example in the format of Figures 15A and 15B, a preferred combination of pI variants comprises one monomer (negative Fab side) containing the 208D / 295E / 384D / 418E / 421D variant (N208D / Q295E / N384D / Q418E / N421D for human IgG1) and a second monomer (positive scFv side) containing a positively charged scFv linker containing (GKPGS)4 (sequence number 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 a CH1 domain (for example, in the case of heterodimer antibodies that do not utilize a CH1 domain in one of their domains, such as in the dual scFv or "one-arm" format shown in Figures 42B, C, or D), the preferred negative pI variant Fc set includes the 295E / 384D / 418E / 421D variants (Q295E / N384D / Q418E / N421D for human IgG1).

[0208] Therefore, in some embodiments, one monomer has a set of substitutions from Figure 4, and the other monomer has a charged linker (either in the format of a charged scFv linker, as the monomer includes an scFv or a charged domain linker, as shown in the format, which can be selected from those shown in Figure 5).

[0209] 3. Isotype Variants Furthermore, many embodiments of the antibodies described herein rely on the “transfer” of pI amino acids at specific positions from one IgG isotype to another, thus reducing or eliminating the possibility of introducing undesirable immunogenicity into the variant. Some of these are shown in Figure 21 of U.S. Patent Application Publication 2014 / 0370013, which is incorporated herein by reference. Specifically, IgG1 is a common isotype for therapeutic antibodies for various reasons, including its 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 into the IgG1 backbone at specific positions, the pI of the resulting monomer is reduced (or increased) and exhibits a longer serum half-life. For example, IgG1 has glycine at position 137 (pI 5.97), while IgG2 has glutamic acid (pI 3.22), and the transfer of glutamic acid affects the pI of the resulting protein. As described below, several amino acid substitutions are generally required to have a significant effect on 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 will be discussed below.

[0210] In other embodiments, non-isotype amino acid changes are performed (for example, by changing high-pI amino acids to low-pI amino acids) to reduce the overall charge state of the resulting protein or to allow for structural adjustments for stability, etc., as will be described in more detail below.

[0211] In addition, 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, a difference of at least 0.5 in the pI of the two monomers may allow separation by ion exchange chromatography, isoelectric focusing, or other isoelectric-sensitive methods.

[0212] 4. Calculate pI The pI of each monomer depends on the pI of the variant heavy chain steady domain and the pI of the total monomer, and may include the variant heavy chain steady domain and fusion partners. Therefore, in some embodiments, the pI change is calculated based on the variant heavy chain steady domain using the chart in Figure 19 of U.S. Patent Application Publication 2014 / 0370013. As discussed herein, which monomer to work with is generally determined by the Fv and the intrinsic pI of the scaffold. Alternatively, the pI of each monomer can be compared.

[0213] 5. pI variants that also confer better in vivo binding of FcRn If pI variants reduce the pI of the monomer, they may have the additional benefit of improving serum retention in vivo.

[0214] Although still under investigation, it is thought that the Fc region has a longer half-life in vivo because Fc is sequestered upon binding to FcRn at pH 6 within the endosome (integrated whole by reference, Ghetie and Ward, 1997 Immunol Today. 18(12):592-598). 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 into the bloodstream. In mice, Dall'Acqua et al. showed that Fc variants with increased FcRn binding at pH 6 and pH 7.4 actually had reduced serum concentrations and the same half-life as wild-type Fc (Dall'Acqua et al. 2002, J.Immunol. 169:5171-5180, integrated whole by reference). The increased affinity of Fc to FcRn at pH 7.4 is thought to hinder the release of Fc into the bloodstream. Therefore, Fc mutations that increase the half-life of Fc in vivo 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. Thus, it is not surprising to find His residues at key positions in the Fc / FcRn complex.

[0215] 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, all incorporated herein by reference). However, the mechanism is still not well understood. Furthermore, the variable region differs from antibody to antibody. As described herein, constant-region variants with reduced pI and extended half-lives would provide a more modular approach to improving the pharmacokinetic properties of antibodies.

[0216] C. Additional Fc variants for additional functionality In addition to the heterodimerized variants outlined above, there are several useful Fc amino acid modifications that can occur for various reasons, including altering binding to one or more FcγR receptors and altered binding to FcRn, as outlined below.

[0217] Accordingly, the antibodies (heterodimers and homodimers) provided herein may include, or may not include, the heterodimerizing variants (e.g., pI variants and stereovariants) outlined herein, or such amino acid modifications. Each set of variants may be independently and optionally included in or excluded from a particular heterodimer protein.

[0218] 1. FcγR variant Several useful Fc substitutions exist that can be performed to alter binding to one or more FcγR receptors. In certain embodiments, the antibody in question includes modifications that alter binding to one or more FcγR receptors (i.e., "FcγR variants"). Substitutions that result in increased or decreased binding may be useful. For example, increased binding to FcγRIIIa is generally known to result in increased ADCC (antibody-dependent cell-mediated cytotoxicity, i.e., a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize a linked antibody on target cells, subsequently causing lysis of the target cells). Similarly, decreased binding to FcγRIIb (an inhibitory receptor) may also be beneficial under certain circumstances. The amino acid substitutions used in the antibodies described herein include those enumerated in U.S. Patent No. 8,188,321 (particularly Figure 41) and No. 8,084,582, and U.S. Patent Publication Applications 20060235208 and 20070148170, all of which are incorporated herein by reference in their entirety, particularly expressly with respect to the variants disclosed therein. 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.

[0219] Furthermore, as specifically disclosed in USSN12 / 341,769, which is incorporated entirely herein by reference, there are additional Fc substitutions used to increase binding to the FcRn receptor and increase serum half-life, including but not limited to 434S, 434A, 428L, 308F, 259I, 428L / 434S, 259I / 308F, 436I / 428L, 436I or V / 434S, 436V / 428L, and 259I / 308F / 428L. Such modifications may be contained in one or both Fc domains of the antibody in question.

[0220] 2. Removal Variant Similarly, another category of functional variants is “FcγR elimination 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 Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid additional mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific antibodies that monovalently bind to CD3, it is generally desirable to eliminate FcγRIIIa binding to eliminate or significantly reduce ADCC activity, and one of the Fc domains comprises one or more Fcγ receptor elimination variants. These removal variants are shown in Figure 14 and can be independently and optionally included or excluded in a preferred embodiment using removal variants selected from the group consisting of G236R / L328R, E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K / A327G, E233P / L234V / L235A / G236del / S267K / A327G, and E233P / L234V / L235A / G236del. Note that the removal variants referred to herein remove FcγR bonds but not FcRn bonds.

[0221] As is known in the art, the Fc domain of human IgG1 has the highest binding to the Fcγ receptor, and therefore, if the constant domain (or Fc domain) of the heterodimeric antibody backbone is IgG1, the elimination variant can be used. Alternatively, or in addition to the elimination variant of the IgG1 background, mutations at glycosylation site 297 (generally to A or S) can significantly eliminate binding to FcγRIIIa, for example. Human IgG2 and IgG4 have naturally reduced binding to the Fcγ receptor, so these backbones can be used with or without the elimination variant.

[0222] D. Combinations of heterodimers and Fc variants As those skilled in the art will understand, all listed heterodimerized variants (including scuba riants and / or pI variants) can be combined at will and independently, as long as their "strandedness" or "monomer splitting" is preserved. In some embodiments, the heterodimerized antibodies provided herein include combinations of heterodimerized scuba riants, equimolecular-volume pI substitutions, and FcKO variants, as shown in Figure 4. Furthermore, all of these variants can be combined in any of the heterodimerized formats.

[0223] In the case of pI variants, the embodiments used in particular are shown in the figure, while other combinations can be generated by following the basic rule of changing the pI difference between the two monomers to facilitate purification.

[0224] Furthermore, any of the heterodimerization variants, skew, and pI can be independently and optionally combined with the Fc removal variant, the Fc variant, and the FcRn variant, as generally outlined herein.

[0225] Figure 4 shows exemplary combinations of variants included in several embodiments of heterodimer 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fc format antibodies. In certain embodiments, the antibody is a heterodimer 1+1Fab-scFv-Fc or 2+1Fab2-scFv-Fc format antibody, as shown in Figures 15A and 15B.

[0226] E. Anti-ENPP3 x anti-CD3 bispecific antibody In another embodiment, a bispecific anti-ENPP3 × anti-CD3 (also referred to herein as "αENPP3 × αCD3") antibody is provided herein. Such an antibody comprises at least one ENPP3-binding domain and at least one CD3-binding domain. In some embodiments, the bispecific αENPP3 × αCD3 selectively provided an immune response at tumor sites expressing ENPP3, as described herein.

[0227] Please note that unless otherwise indicated herein, the order of the antigen list in the names does not give a structure; that is, the ENPP3XCD3 1+1Fab-scFv-Fc antibody can bind scFv to ENPP3 or CD3, but in some cases the structure is identified in the order indicated.

[0228] As outlined in more detail herein, these combinations of ABDs can take various forms, generally in which one ABD is in Fab format and the other is in scFv format, as outlined below. Exemplary formats used with the bispecific antibodies provided herein include the 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv formats (see, for example, Figures 15A and 15B). Other useful antibody formats include, but are not limited to, antibodies in “mAb-Fv”, “mAb-scFv”, “Central-Fv”, “One-arm scFv-mAb”, “scFv-mAb”, “Dual scFv”, and “Trident” formats, as shown in Figures 52A–52K.

[0229] Furthermore, generally speaking, one of the ABDs contains scFv as outlined herein, in an N-terminus to C-terminus direction of VH-scFv linker-VL or VL-scFv linker-VH. According to the format, one or both of the other 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 the other protein chain (generally as a component of the light chain).

[0230] As will be understood by those skilled in the art, any set of six CDR or VH and VL domains can be in scFv or Fab format, which is then appended to the heavy and light chain constant domains, where the heavy chain constant domains include mutations (including within the CH1 and Fc domains). While the scFv sequences included in the sequence listing utilize a specific charged linker, uncharged or other charged linkers can be used, including those shown in Figures 5 and 6, as outlined herein.

[0231] Furthermore, as mentioned above, the numbering used in sequence listings for CDR identification is Kabat, but different numbering can be used, which will alter the amino acid sequence of the CDR as shown in Table 2.

[0232] Further variants can be created for all variable heavy and light chain domains described herein. As outlined herein, in some embodiments, a set of six CDRs may have 0, 1, 2, 3, 4, or 5 amino acid modifications (particularly by the amino acid substitutions used), and variations in the framework region of the variable heavy and light chain domains may be present, insofar as this framework (excluding the CDRs) retains at least about 80, 85, or 90% identity with human germline sequences selected from those listed in Figure 1 of U.S. Patent No. 7,657,380 (which are incorporated herein by reference in their entirety). Thus, for example, the same CDRs described herein can be combined with different framework sequences derived from human germline sequences, insofar as the framework region retains at least 80, 85, or 90% identity with human germline sequences 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, a set of CDRs may have 1, 2, 3, 4, or 5 amino acid modifications (i.e., any combination of CDRs may be modified, e.g., one modification in vlCDR1, two modifications in vhCDR2, and no modification in vhCDR3)), and similarly, a modification of the framework region may have a framework region, provided that the framework region maintains at least 80, 85, or 90% identity with respect to a human germline sequence selected from those enumerated in Figure 1 of U.S. Patent No. 7,657,380).

[0233] The anti-ENPP3 × anti-CD3 bispecific antibody may contain any suitable CD3 ABD, including those described herein (see, for example, Figures 10A–10F). In some embodiments, the CD3 ABD of the anti-ENPP3 × anti-CD3 bispecific antibody includes a variable heavy chain domain and a variable light chain domain of the CD3 ABDs provided herein, including those described in Figures 10A–10F and the sequence listings. In some embodiments, the CD3 ABD includes one variable heavy-chain domain and a variable light-chain domain from the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F). In exemplary embodiments, the CD3 ABD is one of the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F) or one of these variants. The anti-ENPP3 × anti-CD3 bispecific antibody may include any suitable ENPP3 ABD, including those described herein (see, for example, Figures 12, 13A-13B, and 14A-14I). In some embodiments, the ENPP3 ABD of the anti-ENPP3 × anti-CD3 bispecific antibody includes a variable heavy chain domain and a variable light chain domain of the ENPP3 ABDs provided herein, including those described in Figures 12, 13A-13B, and 14A-14I and the sequence listings. In some embodiments, ENPP3 ABD is the following ENPP3 ABD: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It includes one variable heavy chain domain and one variable light chain domain from among 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I).In exemplary embodiments, ENPP3 ABD is as follows: ENPP3 ABD: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 It is one of the following: 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A~13B, and 14A~14I).

[0234] F. Anti-SSTR2 x anti-CD3 bispecific antibody In another embodiment, a bispecific anti-SSTR2 × anti-CD3 (also referred to herein as "αSSTR2 × αCD3") antibody is provided herein. Such an antibody comprises at least one SStR2-binding domain and at least one CD3-binding domain. In some embodiments, the bispecific αSSTR2 × αCD3 has provided a selective immune response in tumor sites expressing SSTR2, as described herein.

[0235] Please note that, unless otherwise indicated herein, the order of the antigen list in the names does not give a structure; that is, the SSTR2XCD3 1+1Fab-scFv-Fc antibody can bind scFv to SSTR2 or CD3, but in some cases the structure is identified in the order indicated.

[0236] As outlined in more detail herein, these combinations of ABDs can take various forms, generally in which one ABD is in Fab format and the other is in scFv format, as outlined below. Exemplary formats used with the bispecific antibodies provided herein include the 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv formats (see, for example, Figures 15A and 15B). Other useful antibody formats include, but are not limited to, antibodies in “mAb-Fv”, “mAb-scFv”, “Central-Fv”, “One-arm scFv-mAb”, “scFv-mAb”, “Dual scFv”, and “Trident” formats, as shown in Figures 52A–52K.

[0237] Furthermore, generally speaking, one of the ABDs contains scFv as outlined herein, in an N-terminus to C-terminus direction of VH-scFv linker-VL or VL-scFv linker-VH. According to the format, one or both of the other 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 the other protein chain (generally as a component of the light chain).

[0238] As will be understood by those skilled in the art, any set of six CDR or VH and VL domains can be in scFv or Fab format, which is then appended to the heavy and light chain constant domains, where the heavy chain constant domains include mutations (including within the CH1 and Fc domains). While the scFv sequences included in the sequence listing utilize a specific charged linker, uncharged or other charged linkers can be used, including those shown in Figures 5 and 6, as outlined herein.

[0239] Furthermore, as mentioned above, the numbering used in sequence listings for CDR identification is Kabat, but different numbering can be used, which will alter the amino acid sequence of the CDR as shown in Table 2.

[0240] Further variants can be created for all variable heavy and light chain domains described herein. As outlined herein, in some embodiments, a set of six CDRs may have 0, 1, 2, 3, 4, or 5 amino acid modifications (particularly by the amino acid substitutions used), and variations in the framework region of the variable heavy and light chain domains may be present, insofar as this framework (excluding the CDRs) retains at least about 80, 85, or 90% identity with human germline sequences selected from those listed in Figure 1 of U.S. Patent No. 7,657,380 (which are incorporated herein by reference in their entirety). Thus, for example, the same CDRs described herein can be combined with different framework sequences derived from human germline sequences, insofar as the framework region retains at least 80, 85, or 90% identity with human germline sequences 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, a set of CDRs may have 1, 2, 3, 4, or 5 amino acid modifications (i.e., any combination of CDRs may be modified, e.g., one modification in vlCDR1, two modifications in vhCDR2, and no modification in vhCDR3)), and similarly, a modification of the framework region may have a framework region, provided that the framework region maintains at least 80, 85, or 90% identity with respect to a human germline sequence selected from those enumerated in Figure 1 of U.S. Patent No. 7,657,380).

[0241] The anti-SSTR2 × anti-CD3 bispecific antibody may contain any suitable CD3 ABD, including those described herein (see, for example, Figures 10A–10F). In some embodiments, the CD3 ABD of the anti-SSTR2 × anti-CD3 bispecific antibody includes a variable heavy chain domain and a variable light chain domain of the CD3 ABDs provided herein, including those described in Figures 10A–10F and the sequence listings. In some embodiments, the CD3 ABD includes one variable heavy-chain domain and a variable light-chain domain from the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F). In exemplary embodiments, the CD3 ABD is one of the following CD3 binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F) or one of these variants. Anti-SSTR2 × anti-CD3 bispecific antibodies may contain the variable heavy chain domain and variable light chain domain of [αSSTR2]H1.24_L1.30 (Figure 63), or variants thereof.

[0242] G. Useful formats of the present invention As will be understood by those skilled in the art and will be described in more detail below, the bispecific heterodimer antibodies provided herein can take on a wide variety of configurations, generally as shown in Figure 1. Some figures show a “single-ended” configuration, where one “arm” of the molecule has one type of specificity and the other “arm” has a different specificity. Other figures show a “dual-ended” configuration, where the “top” of the molecule has at least one type of specificity and the “bottom” of the molecule has one or more different specificities. Accordingly, some embodiments address novel immunoglobulin compositions that co-binding with different first and second antigens.

[0243] As those skilled in the art will understand, the heterodimer format of the present invention may have different valencies and may be bispecific. That is, the heterodimer antibody of the antibody described herein may be bivalent and bispecific, with one target tumor antigen (e.g., CD3) bound by one binding domain and the other target tumor antigen (e.g., ENPP3) bound by a second binding domain. The heterodimer antibody may also be trivalent and bispecific, with the first antigen bound by two binding domains and the second antigen bound by a second binding domain. As outlined herein, when CD3 is one of the target antigens, it is preferable that CD3 binds only monovalently to reduce potential side effects.

[0244] The antibodies described herein utilize an anti-CD3 antigen-binding domain in combination with an anti-ENPP3-binding domain. As will be understood by those skilled in the art, any collection of anti-CD3 CDRs, anti-CD3 variable light chain domains and variable heavy chain domains, Fabs, and scFvs may be used, as shown in any of the figures. Similarly, any of the anti-ENPP3 antigen-binding domains may be used, and any of the CDRs, variable light chain domains and variable heavy chain domains, Fabs, and scFvs may be used in any combination, either independently or arbitrarily, as shown in any of the figures (e.g., Figures 12, 13A-13B, and 14A-14I).

[0245] 1.1+1Fab-scFv-Fc format One heterodimer scaffold for which specific applications are found in the antibodies described herein is a “1+1 Fab-scFv-Fc” or “bottle opener” format, as shown in Figure 15A, with an exemplary combination of a CD3-binding domain and a tumor target antigen (ENPP3)-binding domain. In this embodiment, one heavy chain monomer of the antibody comprises a single Fv (defined below as “scFv”) and an Fc domain. The scFv comprises a variable heavy chain domain (VH1) and a variable light chain domain (VL1), where VH1 is conjugated to VL1 using a facilitable scFv linker (see, for example, Figure 5). The scFv is conjugated to the heavy chain using a domain linker (see, for example, Figure 6). The other heavy chain monomer is a “normal” heavy chain (VH-CH1-hinge-CH2-CH3). The 1+1 Fab-scFv-Fc also includes a light chain that interacts with VH-CH1 to form a Fab. Because this structure is visually roughly similar to a bottle opener, it may be referred to herein as the “bottle opener” format. Two heavy chain monomers are joined together by the use of amino acid variants (e.g., the heterodimerizing variants described above) in constant regions (e.g., the Fc domain, CH1 domain, and / or hinge region) that promote the formation of a heterodimer antibody, as described in more detail below.

[0246] The current "1+1 Fab-scFv-Fc" format has several clear 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 the antibodies described herein by the addition of "normal" heavy-chain and light-chain pairing. Furthermore, in contrast to formats that rely on two heavy chains and two light chains, there is no problem of incorrect pairing of heavy-chain and light-chain (e.g., pairing of heavy chain 1 and light chain 2).

[0247] Many of the embodiments outlined herein generally rely on a 1+1 Fab-scFv-Fc or "bottle opener" format antibody containing a first monomer containing scFv, which includes variable heavy and variable light chain domains covalently linked using an scFv linker (often charged, but not always), where the scFv is typically covalently bound to the first Fc domain via a domain linker. The domain linker may be charged or uncharged, and may be exogenous or endogenous (e.g., all or part of a natural hinge domain). Any suitable linker can be used to attach the scFv to the N-terminus of the first Fc domain. In some embodiments, the domain linker is selected from the domain linkers shown in Figure 6. The second monomer in the 1+1 Fab-scFv-Fc or "bottle opener" format is a heavy chain, and the composition further comprises a light chain.

[0248] Generally, in many preferred embodiments, scFv is the domain that binds to CD3, and Fab forms the ENPP3 binding domain. An exemplary anti-ENPP3 × anti-CD3 bispecific antibody in the 1+1 Fab-scFv-Fc format is shown in Figure 15A. Exemplary anti-ENPP3 × anti-CD3 bispecific antibodies in the 1+1 Fab-scFv-Fc format are shown in Figures 17A-17C and 18A-18C.

[0249] Furthermore, the Fc domain of the antibodies described herein is generally a scuba rian (a set of amino acid substitutions, e.g., as shown in Figures 3 and 9), with particularly useful scuba rians being 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 (selected from the group), optionally including a removal variant (including the one shown in Figure 3), optionally including a charged scFv linker (including the one shown in Figure 5), and the heavy chain including a pI variant (including the one shown in Figure 4).

[0250] In certain embodiments, the 1+1 Fab-scFv-Fc scaffold format includes a first monomer containing an scFv domain linker-CH2-CH3 monomer, a second monomer containing a first variable heavy-chain domain-CH1-hinge-CH2-CH3, and a third monomer containing a first variable light-chain domain. In some embodiments, the CH2-CH3 of the first monomer is a first variant Fc domain, and the CH2-CH3 of the second monomer is a second variant Fc domain. In some embodiments, the scFv includes an scFv variable heavy-chain domain and an scFv variable light-chain domain that form a CD3 binding moiety. In certain embodiments, the scFv variable heavy-chain domain and the scFv variable light-chain domain are covalently bound using an scFv linker (which is often, though not always, charged). See, for example, Figure 5. In some embodiments, the first variable heavy-chain domain and the first variable light-chain domain form an ENPP3 binding domain. Particularly useful ENPP3 and CD3 combinations for use in the 1+1 Fab-scFv-Fc ENPP3×CD3 bispecific antibody format are disclosed in Figures 17A-17C and 18A-18C, including ENPP3 H16-1.93×CD3 H1.30 L1.47, ENPP3 H16-7.8×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, and ENPP3 AN1[ENPP3]H1.8 L1.33×CD2 H1.30 L1.47, and ENPP3 H1.8 This includes L1.77×CD3 H.130 L1.47. In some embodiments, the 1+1 Fab-scFv-Fc format includes a scuba variant, a pI variant, and a elimination variant.Therefore, some embodiments are in a 1+1 Fab-scFv-Fc format, comprising: a) a first monomer ("scFv monomer") comprising a charged scFv linker (preferably the +H sequence in Figure 5 in some embodiments), a scuba riant S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, and an scFv bound to CD3 as outlined herein; and b) a scuba riant L368D / K370S, The 1+1 Fab-scFv-Fc format comprises a second monomer ("Fab monomer") containing a variable heavy chain domain that constitutes an Fv binding to a second antigen outlined herein, together with the pI variant N208D / Q295E / N384D / Q418E / N421D, the elimination variant E233P / L234V / L235A / G236del / S267K, and a variable light chain domain, and a light chain containing a variable light chain domain (VL) and a constant light chain domain (CL), wherein the numbering follows EU numbering. The variable heavy chain domain and the variable light chain domain constitute the ENPP3 binding portion. CD3-binding domain sequences that find specific uses in these embodiments include, but are not limited to, H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31, as well as those shown in Figures 10A to 10F.Examples of ENPP3-binding domain sequences with specific uses in these embodiments include AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, and Ha1 Examples include, but are not limited to, 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). Particularly useful ENPP3 and CD3 sequence combinations for use with 1+1Fab2-scFv-Fc format antibodies include, for example, ENPP3 H16-1.93×CD3 H1.30 L1.47, ENPP3 H16-7.8×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33×CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77×CD3 H.130 L1.47.

[0251] Exemplary variable heavy and light chain domains of scFv that bind to CD3 are included in Figures 10A–10F. Exemplary variable heavy and light chain domains of Fv that bind to ENPP3 are included in Figures 12, 13A–13B, and 14A–14I. In exemplary embodiments, the ENPP3 binding domains of the 1+1Fab-scFv-Fc ENPP3×CD3 bispecific antibody are as follows: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 Includes one VH and VL from among 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, Ha1 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (Figures 12, 13A-13B, and 14A-14I). In one embodiment, the CD3-binding domain of the ENPP3×CD3 bispecific antibody of 1+1Fab-scFv-Fc includes one VH and VL from the following CD3-binding domains: H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31 (Figures 10A-10F).Particularly useful ENPP3 and CD3 combinations for use in the 1+1 Fab-scFv-Fc ENPP3×CD3 bispecific antibody format are disclosed in Figures 17A-17C and 18A-18C, and include ENPP3 H16-1.93×CD3 H1.30 L1.47, ENPP3 H16-7.8×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77×CD3 H.130 L1.47.

[0252] In some embodiments, the 1+1Fab-scFv-Fc format includes a scuba riant, a pI variant, a removal variant, and an FcRn variant. Thus, some embodiments of the 1+1Fab-scFv-Fc format include a) a first monomer ("scFv monomer") comprising a charged scFv linker (preferably the +H sequence in Figure 6 in some embodiments), a scuba riant S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and an scFv bound to CD3 as outlined herein, and b) The 1+1 Fab-scFv-Fc format includes a second monomer ("Fab monomer") containing the new variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the removal variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy chain domain, and a light chain containing a variable light chain domain (VL) and a constant light chain domain (CL), where the numbering follows EU numbering. The variable heavy chain domain and the variable light chain domain constitute the ENPP3 binding domain. CD3-binding domain sequences that find specific uses in these embodiments include, but are not limited to, H1.30_L1.47, H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, H1.31_L1.47, L1.47_H1.30, L1.47_H1.30, L1.47_H1.32, L1.47_H1.89, L1.47_H1.90, L1.47_H1.33, and L1.47_H1.31, as well as those shown in Figures 10A to 10F.Examples of ENPP3-binding domain sequences with specific uses in these embodiments include AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1 L1.77, H16-7.213, H16-9.69, H16-1.52, Ha16-1(1)23, H16-9.44, H16-1.67, Ha1 6-1(3,5)36, H16-1.86, H16-9.10, H16-9.33, H16-1.68, Ha16-1(1)1, and Ha1 Examples include, but are not limited to, 6-1(3,5)18, Ha16-1(2,4)4, Ha16-1(3,5)56, H16-7.8, H16-1.93, Ha1 6-1(3,5)27.1, H16-1.61, H16-1(3,5)5, ​​H16-7.200, H16-1(3,5)42, H1 6-9.65, Ha1-1(3,5)19, and Ha16-1.80 (as shown in Figures 12, 13A-13B, and 14A-14I). Combinations of ENPP3 and CD3 sequences that are particularly useful for use with antibodies in the 1+1 Fab-scFv-Fc format are disclosed, for example, in Figures 17A-17C and 18A-18C, and include ENPP3 H16-1.93×CD3 H1.30 L1.47, ENPP3 H16-7.8×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1×CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33×CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77×CD3 H.130 L1.47.

[0253] Figures 7A–7D show several exemplary Fc domain sequences useful in 1+1 Fab-scFv-Fc format antibodies. The "monomer 1" sequence shown in Figures 7A–7D 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." Furthermore, Figure 9 shows useful CL sequences that can be used in this format.

[0254] In some embodiments, any of the VH and VL sequences shown herein (including all VH and VL sequences shown in the figures and sequence listings, including those directed to ENPP3) can be added as a "Fab side" to the bottle opener skeleton format of Figures 7A–7D using any of the anti-CD3scFv sequences shown in the figures and sequence listings.

[0255] In the case of bottle opener skeleton 1 from Figure 7A (including optionally the 428L / 434S variant), CD-binding domain sequences that find specific applications in these embodiments include, but are not limited to, the CD3-binding domains anti-CD3 H1.30_L1.47, anti-CD3 H1.32_L1.47, anti-CD3 H1.89_L1.47, anti-CD3 H1.90_L1.47, anti-CD3 H1.33_L1.47, and anti-CD3 H1.31_L1.47, as well as those shown in Figures 10A-10F, which are attached as the scFv side of the skeletons shown in Figures 7A-7D.

[0256] Particularly useful combinations of ENPP3 and CD3 sequences for use (including optionally 428L / 434S variants) are disclosed in Figures 17A-17C and 18A-18C.

[0257] 2.mAb-Fv One heterodimer scaffold particularly used in the antibodies described herein is the mAb-Fv format. In this embodiment, the format relies on the use of C-terminal binding of an "additional" variable heavy chain domain to one monomer and C-terminal binding of an "additional" variable light chain domain to the other monomer, thereby forming a third antigen-binding domain, where the Fab portions of the two monomers bind to ENPP3 and the "additional" scFv domain binds to CD3.

[0258] In this embodiment, the first monomer comprises a first heavy chain comprising a first variable heavy chain domain and a first steady heavy chain domain comprising a first Fc domain, and having a first variable light chain domain covalently bonded to the C-terminus of the first Fc domain using a domain linker (VH1-CH1-hinge-CH2-CH3-[arbitrary linker]-VL2). The second monomer comprises a second variable heavy chain domain of a second steady heavy chain domain containing a second Fc domain, and a third variable heavy chain domain covalently bonded to the C-terminus of the second Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[optional linker]-VH2). The variable domains bonded to the two C-terminuses constitute an Fv that binds to CD3 (as having a divalent CD3 bond is less desirable). This embodiment further utilizes a common light chain containing a variable light chain domain and a steady light chain domain, which associate with the heavy chain to form two identical Fabs that link to ENPP3. With respect to many of the embodiments herein, these constructs include scuba riants, pI variants, elimination variants, additional Fc variants, etc., as desired and described herein.

[0259] The antibodies described herein provide mAb-Fv formats in which the CD3 binding domain sequence is shown in Figures 10A-10F. The antibodies described herein provide mAb-Fv formats in which the ENPP3 binding domain sequence is shown in Figures 12, 13A-13B, and 14A-14I.

[0260] Furthermore, the Fc domain in the mAb-Fv format is a scuba rian (a set of amino acid substitutions, e.g., as shown in Figures 3 and 8, with particularly useful scuba rians being S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E35 The group consists of 7L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes a removal variant (including the one shown in Figure 3), optionally includes a charged scFv linker (including the one shown in Figure 5), and the heavy chain includes a pI variant (including the one shown in Figure 2).

[0261] In some embodiments, the mAb-Fv format includes a scuba riant, a pI variant, and a removal variant. Thus, some embodiments of the mAb-Fv format include a) a first monomer comprising a scuba riant S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, and a first variable heavy chain domain and a second variable heavy chain domain constituting an Fv that binds to ENPP3 together with a first variable light chain domain of the light chain, and b) a scuba riant L368D / K370S, a pI variant N208D / Q295E / N384D / Q4 The mAb-Fv format includes a second monomer comprising 18E / N421D, the removal variant E233P / L234V / L235A / G236del / S267K, and a first variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with a first variable light chain domain, and a second variable light chain that, together with a second variable heavy chain domain, forms an Fv(ABD) that binds to CD3; and a light chain comprising a first variable light chain domain and a constant light chain domain.

[0262] In some embodiments, the mAb-Fv format includes a scuba riant, a pI variant, a removal variant, and an FcRn variant. Therefore, some embodiments of the mAb-Fv format include a) a first monomer comprising a scuba riant S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and a first variable heavy chain domain and a second variable heavy chain domain that constitute an Fv that binds to the antigen together with a first variable light chain domain of the light chain; and b) a scuba riant L368D / K370S, a pI variant N208D / Q295E / N384D / Q418 The mAb-Fv format includes a second monomer comprising E / N421D, the removal variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a first variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with the first variable light chain domain, and a second variable light chain that, together with the second variable heavy chain domain of the first monomer, forms an Fv(ABD) that binds to CD3; and c) a light chain comprising the first variable light chain domain and a constant light chain domain.

[0263] 3.mAb-scFv One heterodimer scaffold particularly used in the antibodies described herein is the mAb-scFv format. In this embodiment, the format relies on the use of C-terminal binding of scFv to one of the monomers, thereby forming a third antigen-binding domain, with the Fab portions of the two monomers binding to ENPP3 and the “additional” scFv domain binding to CD3. Thus, the first monomer comprises a first heavy chain (including a variable heavy chain domain and a constant domain) and has a C-terminally covalently bound scFv (VH1-CH1-hinge-CH2-CH3-[optional linker]-VH2-scFv linker-VL2 or VH1-CH1-hinge-CH2-CH3-[optional linker]-VL2-scFv linker-VH2). This embodiment further utilizes a common light chain including variable light chain domains and constant light chain domains, which associate with the heavy chain to form two identical Fabs that link to ENPP3. With respect to many of the embodiments herein, these constructs include scuba riants, pI variants, elimination variants, additional Fc variants, etc., as desired and described herein.

[0264] The antibodies described herein provide an mAb-scFv format in which the CD-binding domain sequence is shown in Figures 10A-10F and the ENPP3-binding domain sequence is shown in Figures 12, 13A-13B, and 14A-14I.

[0265] Furthermore, the Fc domain of the central scFv format is a scuba rian (for example, a set of amino acid substitutions as shown in Figure 1, with particularly useful scuba rians being S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K). The group consists of / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes a removal variant (including those shown in Figure 3), optionally includes a charged scFv linker (including those shown in Figure 5), and the heavy chain includes a pI variant (including those shown in Figure 2).

[0266] In some embodiments, the mAb-scFv format includes a scubarian, a pI variant, and a removal variant. Thus, some embodiments of the mAb-scFv format include a) a first monomer comprising a scubarian S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with the variable light chain domain of the common light chain, and an scFv domain that binds to CD3, and b) a scubarian L368D / The mAb-scFv format comprises a second monomer comprising K370S, pI variants N208D / Q295E / N384D / Q418E / N421D, elimination variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with the variable light chain domain of the common light chain, and a common light chain comprising a variable light chain domain and a constant light chain domain.

[0267] In some embodiments, the mAb-scFv format includes a scubarian, a pI variant, a removal variant, and an FcRn variant. Thus, some embodiments of the mAb-scFv format include a) a first monomer comprising a scubarian S364K / E357Q, a removal variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and a variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with the variable light chain domain of the common light chain, and an scFv domain that binds to CD3, and b) a scubarian L368D / The mAb-scFv format comprises a second monomer comprising K370S, pI variants N208D / Q295E / N384D / Q418E / N421D, removal variants E233P / L234V / L235A / G236del / S267K, FcRn variant M428L / N434S, and a variable heavy chain domain constituting an Fv that binds to ENPP3 as outlined herein together with the variable light chain domain of the common light chain, and a common light chain comprising a variable light chain domain and a constant light chain domain.

[0268] 4.2+1Fab2-scFv-Fc format One heterodimer scaffold for which specific applications are found in the antibodies described herein is the “2+1 Fab2-scFv-Fc” format (also referred to as the “central-scFv format” in previous related filings), shown in Figure 15B, which is an exemplary combination of a CD3-binding domain and two tumor target antigen (ENPP3)-binding domains. In this embodiment, the format relies on the use of an inserted scFv domain, thereby forming a third antigen-binding domain, where the Fab portions of the two monomers bind to ENPP3, and the “additional” scFv domain binds to CD3. The scFv domain is inserted between the Fc domain of one of the monomers and the CH1-Fv region, thereby providing a third antigen-binding domain. As described, ENPP3 × CD3 bispecific antibodies having the 2+1Fab2-scFv-Fc format are potent for inducing redirected T cell cytotoxicity in cellular environments expressing low levels of ENPP3. Furthermore, as illustrated in the example, ENPP3 × CD3 bispecific antibodies with a 2+1Fab2-scFv-Fc format exhibit a wide variety of different properties depending on the ENPP3 and / or CD3 binding domains used by such antibodies, thus enabling "fine-tuning" of the immune response. For example, such antibodies exhibit differences in selectivity for cells with different ENPP3 expression, efficacy against ENPP3-expressing cells, ability to induce cytokine release, and sensitivity to soluble ENPP3. These ENPP3 antibodies are used, for example, in the treatment of ENPP3-related cancers.

[0269] In this embodiment, one monomer comprises a first heavy chain including a first variable heavy chain domain, a CH1 domain (and an optional hinge), and an Fc domain, and has an scFv including an scFv variable light chain domain, an scFv linker, and an scFv variable heavy chain domain. The scFv is covalently bonded between the C-terminus of the CH1 domain of the heavy chain constant domain and the N-terminus of the first Fc domain using an optional domain linker (VH1-CH1-[optional linker]-VH2-scFv linker-VL2-[optional linker including hinge]-CH2-CH3, or VH1-CH1-[optional linker]-VL2-scFv linker-VH2-[optional linker including hinge]-CH2-CH3, in the opposite direction to the scFv). The optional linker ...

Claims

1. A pharmaceutical composition for treating ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3)-related cancers in patients in need, wherein the pharmaceutical composition comprises an anti-CD3 × anti-ENPP3 antibody, and the antibody is a. The first monomer containing sequence number 531, b. A second monomer containing Sequence ID No. 532, and c. Including a light chain containing sequence number 533, Pharmaceutical composition.

2. The pharmaceutical composition according to claim 1, wherein the pharmaceutical composition further comprises an anti-PD1 antibody.

Images