Heterodimeric antibodies that bind ENPP3 and CD3

Abstract

The present invention relates to antibodies comprising novel antigen binding domains, as well as heterodimeric antibodies that bind to the extracellular nucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3).

Description

[0001] This application is a divisional application, and the corresponding parent application has an application number of 202080032762.3, a filing date of February 28, 2020, and an invention title of "Heterodimeric Antibodies Combining ENPP3 and CD3".

[0002] Priority Claim

[0003] This application claims the priority of U.S. Provisional Application No. 62 / 812,922 filed on March 1, 2019 and U.S. Provisional Application No. 62 / 929,687 filed on November 1, 2019, and these U.S. provisional applications are hereby incorporated by reference in their entireties. Background Art

[0004] Antibody-based therapeutic agents have been successfully used to treat various diseases, including cancer. An increasingly common approach being explored is the engineering of single immunoglobulin molecules that co-conjugate two different antigens. Such alternative antibody forms that conjugate two different antigens are commonly referred to as bispecific antibodies. Because of the considerable diversity of antibody variable regions (Fv) that allows for the generation of Fvs that recognize almost any molecule, a typical method for generating bispecific antibodies is to introduce new variable regions into the antibody.

[0005] A particularly useful method for bispecific antibodies is to engineer a first binding domain that conjugates CD3 and a second binding domain that conjugates an antigen associated with or upregulated on cancer cells, such that the bispecific antibody redirects CD3 + T cells to destroy cancer cells. It has previously been reported that ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) is highly expressed in renal cell carcinoma and is expressed at the lowest levels in normal tissues. In view of this, it is believed that anti-ENPP3 antibodies can be used, for example, to localize anti-tumor therapeutic agents (e.g., chemotherapeutic agents and T cells) to such ENPP3-expressing tumors. Novel bispecific antibodies against CD3 and ENPP3 are provided herein that are capable of localizing CD3 + effector T cells to tumors expressing ENPP3. Summary of the Invention

[0006] Accordingly, provided herein are ENPP3 antigen-binding domains and anti-ENPP3 antibodies (e.g., bispecific antibodies).

[0007] In one aspect, the present disclosure provides a composition comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the ENPP3 binding domain comprising variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) and variable light chain complementarity determining regions (vlCDR1-3) of any 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ). In some embodiments, the vhCDR1-3 and the vlCDR1-3 are selected from the vhCDR1-3 sequences and vlCDR1-3 sequences of the ENPP3 binding domains provided in Figure 12 , Figures 13A to 13B , and Figures 14A to 14I .

[0008] In another aspect, the present disclosure provides a composition comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the ENPP3 binding domain comprising a variable heavy chain domain and a variable light chain domain of any 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]H1.8L1.33, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0009] In another aspect, the present invention provides a composition comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain selected from 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0010] In another aspect, the present invention provides a nucleic acid composition comprising: a) a first nucleic acid encoding a variable heavy chain domain, the variable heavy chain domain comprising variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) of an ENPP3 binding domain; and b) a second nucleic acid encoding a variable light chain domain, the variable light chain domain comprising variable light chain complementarity determining regions 1-3 (vlCDR1-3) of an 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]H1.8L1.33, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ). In some embodiments, the vhCDR1-3 and the vlCDR1-3 are selected from the vhCDR1-3 sequences and vlCDR1-3 sequences provided in Figure 12 , Figures 13A to 13B , and Figures 14A to 14I .

[0011] 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]H1L1.77, 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0012] In some embodiments, the present invention provides an expression vector composition comprising: a) a first expression vector comprising the first nucleic acid; b) a second expression vector comprising the second nucleic acid. In other embodiments, the present invention provides a host cell comprising the expression vector composition.

[0013] In some embodiments, the present invention provides a method for preparing an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the method comprising culturing a host cell under conditions for expressing the ENPP3 binding domain and recovering the ENPP3 binding domain.

[0014] In another aspect, the present invention provides an anti-ENPP3 antibody, the anti-ENPP3 antibody comprising an extracellular nucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the ENPP3 binding domain comprising variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) and variable light chain complementarity determining regions (vlCDR1-3) of any of the following ENPP3 binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ). In some embodiments, vhCDR1-3 and vlCDR1-3 are selected from Figure 12 , Figures 13A to 13B , and Figures 14A to 14I vhCDR1-3 and vlCDR1-3 of any of the following ENPP3 binding domains in

[0015] In another aspect, the present invention provides an anti-ENPP3 antibody, the anti-ENPP3 antibody comprising an extracellular nucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the ENPP3 binding domain comprising the variable heavy chain domain and variable light chain domain of any 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0016] In another aspect, provided herein is an anti-ENPP3 antibody that comprises an extracellular nucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain selected from 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]H1L1.77, 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 Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0017] In some embodiments, the antibody comprises: a) a first monomer comprising a first antigen-binding domain and a first constant domain; and b) a second monomer comprising a second antigen-binding domain and a second constant domain, wherein either the first antigen-binding domain or the second antigen-binding domain is an ENPP3-binding domain. In other embodiments, the first antigen-binding domain and the second antigen-binding domain bind to different antigens. In other embodiments, the first antigen-binding domain is an ENPP3-binding domain and the second antigen-binding domain is a CD3-binding domain. In other embodiments, the CD3-binding domain comprises the 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 to 10F ). In other embodiments, the vhCDR1-3 and the vlCDR1-3 of the CD3-binding domain are selected from Figures 10A to 10F the vhCDR1-3 and vlCDR1-3 therein.

[0018] In some embodiments, the CD3-binding domain comprises the variable heavy chain domain and the variable light chain domain 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 to 10F ).

[0019] In some embodiments, the CD3-binding domain is an anti-CD3 scFv.

[0020] In some embodiments, wherein the first constant domain and the second constant domain each comprise CH2-CH3.

[0021] In some embodiments, the first constant domain and the second constant domain each comprise CH1-hinge-CH2-CH3.

[0022] In some embodiments, the first constant domain and the second constant domain are each a variant constant domain.

[0023] In some embodiments, the first monomer and the second monomer comprise a set of heterodimerization variants, and the set of heterodimerization variants is Figures 1A to 1E any one of the variants depicted in. In some embodiments, the set of heterodimerization variants comprises one variant from the following set of variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V.

[0024] In some embodiments, the first monomer and the second monomer each further comprise ablation variants. In other embodiments, the ablation variants are E233P / L234V / L235A / G236del / S267K.

[0025] In some embodiments, at least one of the first monomer or the second monomer further comprises a pI variant. In some embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D. In some embodiments, the scFv comprises a charged scFv linker.

[0026] In some embodiments, the present invention provides a nucleic acid composition comprising a nucleic acid encoding an anti-ENPP3. In some embodiments, the composition comprises nucleic acids encoding the first monomer and the second monomer. In some embodiments, the present invention provides an expression vector comprising the nucleic acid. In some embodiments, the present invention provides a host cell transformed with the expression vector.

[0027] In some embodiments, the present invention provides a method for preparing an anti-ENPP3 antibody according to any one of claims B1 to B21. The method comprises culturing the host cell according to claim B25 under conditions for expressing the anti-ENPP3 antibody, and recovering the anti-ENPP3 antibody. In some embodiments, the present invention provides a method for treating cancer, the method comprising administering the antibody to a patient in need thereof.

[0028] In another aspect, the present invention provides a heterodimeric antibody, the heterodimeric antibody comprising: a) a first monomer, the first monomer comprising: i) an anti-CD3 scFv, the anti-CD3 scFv comprising a first variable light chain domain, an scFv linker, and a first variable heavy chain domain; and ii) a first Fc domain, wherein the scFv is covalently attached to the N-terminus of the first Fc domain using a domain linker; b) a second monomer, the 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, the 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.

[0029] In some embodiments, the ENPP3 binding domain comprises the vhCDR1-3 and vlCDR1-3 of any of the following ENPP3 binding domains of the ENPP3 binding domain: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0030] In some embodiments, the vhCDR1-3 and vlCDR1-3 of the ENPP3 binding domain are selected from the vhCDR1-3 sequences and vlCDR1-3 sequences of the ENPP3 binding domains provided in Figure 12 , Figures 13A to 13B , and Figures 14A to 14I .

[0031] In some embodiments, the second heavy chain variable domain comprises a heavy chain variable domain, and the second light chain variable domain comprises a variable light chain domain of any 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, Ha16-1(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 Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0032] In some embodiments, the anti-CD3 scFv comprises 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 to 10F ).

[0033] In some embodiments, the vhCDR1-3 and vlCDR1-3 of the anti-CD3 scFv are selected from Figures 10A to 10F the vhCDR1-3 and vlCDR1-3 therein.

[0034] In some embodiments, the anti-CD3 scFv comprises a variable heavy chain domain and a variable light chain domain 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 to 10F ).

[0035] In some embodiments, the first variable light chain domain is covalently attached to the N-terminus of the first Fc domain using a domain linker.

[0036] In some embodiments, the first variable heavy chain domain is covalently attached to the N-terminus of the first Fc domain using a domain linker.

[0037] In some embodiments, the scFv linker is a charged scFv linker.

[0038] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0039] In some embodiments, the first monomer and the second monomer comprise a set of heterodimerization variants selected from any of the heterodimerization variants selected from Figures 1A to 1E . In some embodiments, the set of heterodimerization variants is selected from the following: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, wherein the numbering is according to EU numbering.

[0040] In some embodiments, the first monomer and the second monomer further comprise ablation variants. In some embodiments, the ablation variant is E233P / L234V / L235A / G236del / S267K, wherein the numbering is according to EU numbering.

[0041] In some embodiments, one of the first monomer or the second monomer comprises a pI variant.

[0042] In some embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D, wherein the numbering is according to EU numbering.

[0043] In some embodiments, the first monomer comprises the amino acid variants S364K / E357Q / E233P / L234V / L235A / G236del / S267K, wherein the second monomer comprises the amino acid variants L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and wherein the numbering is according to EU numbering.

[0044] In some embodiments, the scFv linker is a charged scFv linker having the amino acid sequence (GKPGS)4.

[0045] In some embodiments, the first monomer and the second monomer each further comprise the amino acid variant 428 / 434S.

[0046] In some embodiments, the heterodimeric antibody comprises the following heterodimeric antibodies: XENP24804, XENP26820, XENP28287, XENP28925, XENP29516, XENP30262, XENP26821, XENP29436, XENP28390, XENP29463, and XENP30263.

[0047] In another aspect, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising, from the N-terminus to the C-terminus, scFv-linker-CH2-CH3, wherein the scFv is an anti-CD3 scFV and CH2-CH3 is a first Fc domain; b) a second monomer comprising, from the N-terminus to the C-terminus, VH-CH1-hinge-CH2-CH3, wherein CH2-CH3 is a second Fc domain; and c) a light chain comprising VL-CL; wherein the first variant Fc domain comprises the amino acid variants S364K / E357Q, wherein the second variant Fc domain comprises the amino acid variants L368D / K370S, wherein the first variant Fc domain and the second variant Fc domain each comprise the amino acid variants E233P / L234V / L235A / G236del / S267K, wherein the hinge-CH2-CH3 of the second monomer comprises the amino acid variants N208D / Q295E / N384D / Q418E / N421D, wherein VH and VL form an ENPP3 binding domain, and the ENPP3 binding domain comprises the respective variable heavy chain domain and variable light chain domain of an ENPP3 binding domain selected from 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 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 Figure 12 , Figures 13A to 13B , and Figures 14A to 14I), wherein the anti-CD3 scFv comprises a variable heavy chain domain and a variable light chain domain of a CD3-binding domain selected from: 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 to 10F ), and wherein the numbering is according to EU numbering.

[0048] In some embodiments, the scFv comprises a charged scFv linker having the amino acid sequence (GKPGS)4.

[0049] In some embodiments, each of the first variant Fc domain and the second variant Fc domain further comprises the amino acid variant 428 / 434S, wherein the numbering is according to EU numbering.

[0050] In some embodiments, the present invention provides a nucleic acid composition comprising nucleic acids encoding a first monomer and a second monomer of an antibody and a light chain.

[0051] In some embodiments, the present invention provides an expression vector comprising a nucleic acid. In some embodiments, the present invention provides a host cell transformed with the expression vector.

[0052] In some embodiments, the present invention provides a method of treating ENPP3-related cancer, the method comprising administering to a patient in need thereof any one of the antibodies provided herein.

[0053] In another aspect, the present invention provides a heterodimeric antibody, the heterodimeric antibody comprising: a) a first monomer, the 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 an anti-CD3 scFV, linker1 and linker2 are a first domain linker and a second domain linker respectively, and CH2-CH3 is a first Fc domain; b) a second monomer, the second monomer comprising VH2-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH2 is a second variable heavy chain domain and CH2-CH3 is a second Fc domain; and c) a common light chain, the common light chain comprising 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.

[0054] In some embodiments, each of the first ENPP3 binding domain and the second ENPP3 binding domain comprises vhCDR1-3 and vlCDR1-3 of any of the following ENPP3 binding domains of the ENPP3 binding domain: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8L1.33, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0055] In some embodiments, the vhCDR1-3 and vlCDR1-3 of the first ENPP3 binding domain and the second ENPP3 binding domain are selected from the vhCDR1-3 and vlCDR1-3 provided in FIG. 14 and Figure 45 ).

[0056] In some embodiments, the first variable heavy chain domain and the second variable heavy chain domain each comprise a variable heavy chain domain of an ENPP3 binding domain, and the first variable light chain domain and the second variable light chain domain each comprise a variable light chain domain of an ENPP3 binding domain, wherein the ENPP3 binding domain is any ENPP3 binding domain among 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0057] In some embodiments, the scFv comprises vhCDR1-3 and vlCDR1-3 of any CD3 binding domain among 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 to 10F ).

[0058] In some embodiments, the vhCDR1-3 and vlCDR1-3 of the scFv are selected from Figures 10A to 10F the vhCDR1-3 and vlCDR1-3 in

[0059] In some embodiments, the scFv comprises a variable heavy chain domain and a variable light chain domain 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 to 10F ).

[0060] In some embodiments, the scFv comprises a scFv variable heavy chain domain, a scFv variable light chain domain, and a scFv linker connecting the scFv variable heavy chain domain and the scFv variable light chain domain.

[0061] In some embodiments, the scFv variable heavy chain domain is attached to the C-terminus of the CH1 of the first monomer using the first domain linker, and the scFv variable light chain domain is covalently attached to the N-terminus of the first Fc domain using the second domain linker.

[0062] In some embodiments, the scFv variable light chain domain is attached to the C-terminus of the CH1 of the first monomer using the first domain linker, and the scFv variable heavy chain domain is covalently attached to the N-terminus of the first Fc domain using the second domain linker.

[0063] In some embodiments, the scFv linker is a charged scFv linker.

[0064] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0065] In some embodiments, the first monomer and the second monomer comprise a set of heterodimerization variants selected from Figures 1A to 1E the group of heterodimerization variants depicted in

[0066] In some embodiments, the set of heterodimerization variants is selected 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 is according to EU numbering.

[0067] In some embodiments, the first monomer and the second monomer further comprise an ablation variant.

[0068] In some embodiments, the ablation variant is E233P / L234V / L235A / G236del / S267K, wherein the numbering is according to EU numbering.

[0069] In some embodiments, one of the first monomer or the second monomer further comprises a pI variant.

[0070] In some embodiments, the pI variant is N208D / Q295E / N384D / Q418E / N421D, wherein the numbering is according to EU numbering.

[0071] In some embodiments, the first variant Fc domain comprises the amino acid variants S364K / E357Q / E233P / L234V / L235A / G236del / S267K, wherein the second variant Fc domain comprises the amino acid variants L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, wherein the numbering is according to EU numbering.

[0072] In some embodiments, the scFv linker is a charged scFv linker having the amino acid sequence (GKPGS)4.

[0073] In some embodiments, each of the first variant Fc domain and the second variant Fc domain further comprises the amino acid variant 428 / 434S, wherein the numbering is according to EU numbering.

[0074] In some embodiments, the heterodimeric antibody comprises the following heterodimeric antibodies: XENP29437, XENP29520, XENP30264, XENP26822, XENP28438, XENP29438, XENP29467, XENP30469, XENP30470, XENP30819, XENP30821, XENP31148, XENP31149, XENP31150, XENP31419, and XENP31471.

[0075] In another aspect, the heterodimeric antibody comprises: a) a first monomer that, from the N-terminus to the C-terminus, comprises VH1-CH1-linker1-scFv-linker2-CH2-CH3, where the scFv is an anti-CD3 scFV and CH2-CH3 is a first Fc domain; b) a second monomer that, from the N-terminus to the C-terminus, comprises VH1-CH1-hinge-CH2-CH3, where CH2-CH3 is a second Fc domain; and c) a common light chain that comprises VL-CL; wherein the first variant Fc domain comprises the amino acid variants S364K / E357Q, wherein the second variant Fc domain comprises the amino acid variants L368D / K370S, wherein the first variant Fc domain and the second variant Fc domain each comprise the amino acid variants E233P / L234V / L235A / G236del / S267K, wherein the hinge-CH2-CH3 of the second monomer comprises the amino acid variants N208D / Q295E / N384D / Q418E / N421D, and wherein the VH and VL comprise variable heavy chain domains and variable light chain domains of an ENPP3-binding domain selected from: 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]H1L1.77, 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 Figure 12 , Figures 13A to 13B , and Figures 14A to 14I), wherein the anti-CD3 scFv comprises a variable heavy chain domain and a variable light chain domain of a CD3-binding domain selected from: 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 to 10F ), and wherein the numbering is according to EU numbering.

[0076] In some embodiments, the scFv comprises a charged scFv linker having the amino acid sequence (GKPGS)4.

[0077] In some embodiments, the first variant Fc domain and the second variant Fc domain each further comprise the amino acid variant 428 / 434S.

[0078] In some embodiments, the first monomer, the second monomer, and the common light chain of the antibody. In some embodiments, the invention provides an expression vector comprising a nucleic acid. In some embodiments, the invention provides a host cell transformed with the expression vector. In some embodiments, the invention provides a method of treating ENPP3-related cancer, the method comprising administering the antibody to a patient in need thereof.

[0079] In another aspect, the invention provides a heterodimeric antibody comprising the following heterodimeric antibodies: XENP24804, XENP26820, XENP28287, XENP28925, XENP29516, XENP30262, XENP26821, XENP29436, XENP28390, XENP29463, and XENP30263.

[0080] In another aspect, the present invention provides a heterodimeric antibody, said heterodimeric antibody comprising the following heterodimeric 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 a nucleic acid composition comprising a nucleic acid encoding a heterodimeric antibody. In some embodiments, the present invention provides an expression vector, said expression vector comprising a nucleic acid. In some embodiments, the present invention provides a host cell transformed with the expression vector.

[0081] In some embodiments, the method of the present invention provides a method for treating ENPP3-related cancer, said method comprising administering to a patient in need thereof any one of the heterodimeric antibodies provided herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figures 1A to 1E Depicts useful pairs of Fc heterodimerization variants (including skewed variants and pI variants). Some variants do not have a corresponding "monomer 2" variant; these variants are pI variants that can be used alone on either monomer.

[0083] Figure 2 Depicts a list of homotypic variant antibody constant regions and their corresponding substitutions. pI_(-) represents the lower pI variant, while pI_(+) represents the higher pI variant. These can optionally and independently be combined with other heterodimerization variants (and other variant types, as outlined herein) of the antibodies described herein.

[0084] Figure 3 Depicts useful ablation variants that ablate FcγR binding (sometimes referred to as "knockout" or "KO" variants). Generally, ablation variants are found on both monomers, although in some cases it may be present on only one monomer.

[0085] Figure 4 Depicts particularly useful embodiments of the "non-Fv" component of the antibodies described herein.

[0086] Figure 5 Depicts a plurality of charged scFv linkers as described herein for increasing or decreasing the pI of the heterodimeric bsAbs of the present invention that utilize one or more scFvs as components. The (+H) positive linker is particularly applicable herein, especially in combination with the anti-CD3 V L and VH Used in conjunction with the sequences. According to Whitlow et al., Protein Engineering 6(8):989-995 (1993), a single prior art scFv linker with a single charge is referred to as a "Whitlow". It should be noted that such linkers are used to reduce aggregation in scFv and enhance proteolytic stability in scFv. Such charged scFv linkers can be used in any of the inventive antibody forms disclosed herein, which include scFv (e.g., 1+1Fab-scFv-Fc form and 2+1Fab2-scFv-Fc form).

[0087] Figure 6 Depicts a number of exemplary domain linkers. In some embodiments, these linkers are used to connect single-chain Fv to the Fc chain. In some embodiments, these linkers can be combined. For example, the GGGGS linker can be combined with a "half-hinge" linker.

[0088] Figures 7A to 7DDepicts the sequences of the heavy chain backbones of several useful 1+1 Fab-scFv-Fc bispecific antibody forms based on human IgG1, without Fv sequences (e.g., scFv and VH on the Fab side). Backbone 1 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S skewed variant, C220S on the chain with the S364K / E357Q skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 2 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368D / K370S skewed variant, C220S on the chain with the S364K skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 3 is based on human IgG1 (356E / 358M allotype) and includes the S364K:L368E / K370S skewed variant, C220S on the chain with the S364K skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 4 is based on human IgG1 (356E / 358M allotype) and includes the D401K:K360E / Q362E / T411E skewed variant, C220S on the chain with the D401K skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the K360E / Q362E / T411E skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 5 is based on human IgG1 (356E / 358M allotype) and includes the S364K / E357Q:L368D / K370S skewed variant, C220S on the chain with the S364K / E357Q skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains.The backbone 6 is based on human IgG1 (356E / 358M allotype), and includes the S364K / E357Q:L368D / K370S skewed variant, C220S on the chain with the S364K / E357Q skewed variant, N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains, as well as the N297A variant on both chains. The backbone 7 is identical to the backbone 6 except that the mutation is N297S. The backbone 8 is based on human IgG4, and includes the S364K / E357Q:L368D / K370S skewed variant, N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the S228P (EU number, which is S241P in Kabat) variant on both chains, which ablates Fab arm exchange as known in the art. The backbone 9 is based on human IgG2, and includes the S364K / E357Q:L368D / K370S skewed variant, N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant. The backbone 10 is based on human IgG2, and includes the S364K / E357Q:L368D / K370S skewed variant, N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the S267K variant on both chains. The backbone 11 is identical to the backbone 1 except that the backbone 11 contains the M428L / N434S Xtend mutation. The backbone 12 is based on human IgG1 (356E / 358M allotype), and includes the S364K / E357Q:L368D / K370S skewed variant, C220S on the chain with the S364K / E357Q skewed variant and the P217R / P229R / N276K pI variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Included within each of these backbones are sequences that are 90%, 95%, 98% and 99% identical (as defined herein) to the recited sequences and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 additional amino acid substitutions (compared to the "parent" of the figure, as would be understood by one of ordinary skill in the art, compared to the parental human IgG1 (or IgG2 or IgG4, depending on the backbone), which already contains multiple amino acid modifications). That is, in addition to the skewed variants, pI variants and ablation variants contained in the backbones of this figure, the recited backbones may contain additional amino acid modifications (usually amino acid substitutions).

[0089] Figures 8A to 8CDepicts the sequences of the heavy chain backbones of several useful 2+1 Fab2-scFv-Fc bispecific antibody formats based on human IgG1, without the Fv sequences (e.g., scFv and VH on the Fab side). Backbone 1 is based on human IgG1 (356E / 358M allotype) and contains the S364K / E357Q:L368D / K370S skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 2 is based on human IgG1 (356E / 358M allotype) and contains the S364K:L368D / K370S skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 3 is based on human IgG1 (356E / 358M allotype) and contains the S364K:L368E / K370S skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368E / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 4 is based on human IgG1 (356E / 358M allotype) and contains the D401K:K360E / Q362E / T411E skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the K360E / Q362E / T411E skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Backbone 5 is based on human IgG1 (356D / 358L allotype) and contains the S364K / E357Q:L368D / K370S skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains.The backbone 6 is based on human IgG1 (356E / 358M allotype), and contains the S364K / E357Q:L368D / K370S skewed variant, the N208D / Q295E / N384D / Q418E / N421D pI variant on the chain with the L368D / K370S skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains, as well as the N297A variant on both chains. The backbone 7 is identical to the backbone 6 except that the mutation is N297S. The backbone 8 is identical to the backbone 1 except that the backbone 8 contains the M428L / N434S Xtend mutation. The backbone 9 is based on human IgG1 (356E / 358M allotype), and contains the S364K / E357Q:L368D / K370S skewed variant, the P217R / P229R / N276K pI variant on the chain with the S364K / E357Q skewed variant, and the E233P / L234V / L235A / G236del / S267K ablation variant on both chains. Included within each of these backbones are sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) to the recited sequences and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid substitutions (compared to the "parent" of the figure, as would be understood by one of ordinary skill in the art, compared to the parental human IgG1 (or IgG2 or IgG4, depending on the backbone), which already contains multiple amino acid modifications). That is, in addition to the skewed variants, pI variants, and ablation variants contained in the backbones of this figure, the recited backbones can contain additional amino acid modifications (generally amino acid substitutions).

[0090] Figure 9 Sequences of several useful constant light chain domain backbones based on human IgG1 are depicted, without Fv sequences (e.g., scFv or Fab). The constant light chain backbone sequences included herein are 90%, 95%, 98%, and 99% identical to the recited sequences (as defined herein), and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 additional amino acid modifications.

[0091] Figures 10A to 10F Sequences of exemplary anti-CD3 scFvs suitable for use in the bispecific antibodies described herein are depicted. The CDRs are underlined and 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 would be understood by one of ordinary skill in the art, this linker can be replaced by other linkers, including uncharged linkers or negatively charged linkers, some of which are depicted in Figure 5in), and the slashes indicate one or more boundaries of the variable domains. Additionally, the naming convention shows the orientation of the scFv from the N-terminus to the C-terminus. As noted herein and as is true for each sequence herein containing a CDR, as shown in Table 2, depending on the numbering used, the precise identification of the CDR positions may vary slightly, and thus herein includes not only the underlined CDRs but also the CDRs included within the VH and VL domains using other numbering systems. Further, for all sequences in the figures, these V H and V L sequences can be used in scFv form or Fab form.

[0092] Figures 11A to 11B depicts the antigen sequences of various antigens used in the antibodies described herein (including both human and cynomolgus monkey) to assist in the development of antigen-binding domains that bind to both and are amenable to clinical development.

[0093] Figure 12 depicts the variable heavy chain sequence and variable light chain sequence of an exemplary humanized ENPP3-binding domain designated AN1 herein, and the sequence of XENP28278, an anti-ENPP3 mAb based on AN1 and an IgG1 backbone having an E233P / L234V / L235A / G236del / S267K ablation variant. The CDRs are underlined, and the slashes indicate the boundaries between the variable regions and the constant domains. As noted herein and as is true for each sequence herein containing a CDR, as shown in Table 2, depending on the numbering used, the precise identification of the CDR positions may vary slightly, and thus herein includes not only the underlined CDRs but also the CDRs included within the VH and VL domains using other numbering systems. Further, for all sequences in the figures, these V H and V L sequences can be used in scFv form or Fab form.

[0094] Figures 13A to 13B depicts the variable heavy chain sequence and variable light chain sequence of an AN1 variant engineered to achieve improved purification and / or modulation of ENPP3-binding affinity and / or potency. The CDRs are underlined, and the slashes indicate the boundaries between the variable regions and the constant domains. As mentioned herein and as is true for each sequence containing a CDR, depending on the numbering method (as Figure 12 shown), the precise identification of the CDR positions may vary slightly, and thus herein includes not only the underlined CDRs but also the CDRs included within the VH and VL domains using other numbering systems. Further, as for all sequences in the figures, these V H and V LThe sequences can be used in the scFv form or the Fab form. In addition, each variable heavy chain domain depicted herein in the variable heavy chain structure can pair with any other αENPP3 variable light chain domain; and each variable light chain domain depicted herein in the variable light chain structure can pair with any other αENPP3 variable heavy chain domain.

[0095] Figures 14A to 14I The variable regions of additional ENPP3 antigen-binding domains that can be used for αENPP3xαCD3 antibodies are depicted. The CDRs are underlined. As mentioned herein and as is true for each sequence containing a CDR, depending on the numbering method used (as Figure 12 shown), the exact identification of the CDR positions may vary slightly, and thus this document includes not only the underlined CDRs but also the CDRs included within the VH and VL domains using other numbering systems. In addition, for all sequences in the figures, these V H and V L sequences can be used in the scFv form or the Fab form.

[0096] Figures 15A to 15B Several forms of the antibodies described herein are depicted. Figure 15A The "1+1 Fab-scFv-Fc" form is depicted, where the first arm contains an ENPP3-binding Fab and the second arm contains a CD3-binding scFv. Figure 30 B depicts the "2+1 Fab2-scFv-Fc" form, where the first arm contains an ENPP3-binding Fab and the second arm contains a Fab and a scFv, where the Fab binds ENPP3 and the scFv binds CD3.

[0097] Figure 16 The amino acid sequence of a control anti-RSVx anti-CD3 bispecific antibody (Fab-scFv-Fc) in the form of a bottle opener is depicted. The antibody is named first using the Fab variable region and second using the scFv variable region, separated by a dashed line. The CDRs are underlined, and the slashes indicate the boundaries of the variable regions. The scFv domain has a V H -scFv linker-V L orientation (N-terminus to C-terminus), but this can be reversed. Additionally, each sequence outlined herein can contain or not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0098] Figures 17A to 17CDepicts the sequence of an illustrative αENPP3 x αCD3 bsAb in the 1+1 Fab-scFv-Fc format and containing the H1.30_L1.47 anti-CD3 scFv (also known as CD3 High [VHVL]). The CDRs are underlined, and the slashes indicate the boundaries between the variable regions and other chain components (such as constant regions and domain linkers). It should be noted that the αENPP3 x αCD3 bsAb can utilize variable regions, Fc regions, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein can contain or not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0099] Figures 18A to 18C Depicts the sequence of an illustrative αENPP3 x αCD3 bsAb in the 1+1 Fab-scFv-Fc format and containing the H1.32_L1.47 anti-CD3 scFv (also known as CD3 High-Int Number 1 [VHVL]). The CDRs are underlined, and the slashes indicate the boundaries between the variable regions and other chain components (such as constant regions and domain linkers). It should be noted that the αENPP3 x αCD3 bsAb can utilize variable regions, Fc regions, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein can contain or not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0100] Figures 19A to 19C Depicts the sequence of an illustrative αENPP3 x αCD3 bsAb in the 2+1 Fab2-scFv-Fc format and containing the H1.30_L1.47 anti-CD3 scFv (also known as CD3 High [VHVL]). The CDRs are underlined, and the slashes indicate the boundaries between the variable regions and other chain components (such as constant regions and domain linkers). It should be noted that the αENPP3 x αCD3 bsAb can utilize variable regions, Fc regions, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein can contain or not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0101] Figures 20A to 20D Depicts the sequence of an illustrative αENPP3 x αCD3 bsAb in 2+1 Fab2-scFv-Fc format and containing the H1.32_L1.47 anti-CD3 scFv (also known as CD3 High-Int No. 1 [VHVL]). CDRs are underlined, and slashes indicate the boundaries between variable regions and other chain components (e.g., constant regions and domain linkers). It should be noted that the αENPP3 x αCD3 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0102] Figure 21 Depicts the sequence of an illustrative αENPP3 x αCD3 bsAb in 2+1 Fab2-scFv-Fc format and containing the L1.47_H1.30 anti-CD3 scFv (also known as CD3 High [VLVH]). CDRs are underlined, and slashes indicate the boundaries between variable regions and other chain components (e.g., constant regions and domain linkers). It should be noted that the αENPP3 x αCD3 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0103] Figures 22A to 22CDepicts the sequence of an illustrative αENPP3 xαCD3 bsAb in 2+1 Fab2-scFv-Fc format and containing the L1.47_H1.32 anti-CD3 scFv (also known as CD3 High-Int No. 1 [VHVL]). CDRs are underlined, and slashes indicate the boundaries between variable regions and other chain components (such as constant regions and domain linkers). It should be noted that the αENPP3 xαCD3 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0104] Figures 23A to 23E Depicts the sequence of an illustrative αENPP3 xαCD3 bsAb in 2+1 Fab2-scFv-Fc format and containing the L1.47_H1.89 anti-CD3 scFv (also known as CD3 High-Int No. 2 [VLVH]). CDRs are underlined, and slashes indicate the boundaries between variable regions and other chain components (such as constant regions and domain linkers). It should be noted that the αENPP3 xαCD3 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90%, 95%, 98%, and 99% identical (as defined herein) and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. Additionally, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably two Fc domains, which results in a longer serum half-life.

[0105] Figures 24A to 24B Depicts RTCC induction on CFSE-labeled KU812 cells, A) as indicated by a decrease in the number of KU812 cells and B) as shown by CFSE + KU812 cells, and after incubating CFSE-labeled KU812 with human PBMCs (effector-to-target ratio of 10:1) and αENPP3 xαCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours and staining with Zombie Aqua +As indicated by the percentage of KU812 cells. The controls used were the αRSV xαCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Overall, the data showed that the prototype αENPP3 xαCD3 bsAb dose-dependently induced redirected T cell cytotoxicity (RTCC) on KU812 cells; CD3 binding affinity correlated with RTCC potency (i.e., bsAbs with high CD3 induced RTCC more effectively than bsAbs with high-Int CD3 number 1); and bsAbs with the AN1 binding domain induced RTCC more effectively than bsAbs with the H16-7.8 binding domain.

[0106] Figures 25A to 25C Depicted activated CD4 + T cells, as indicated by: A) CD4 + T cell CD107a MFI, B) CD4 + T cell CD25 MFI, and C) CD4 + T cell CD69 MFI after incubation of CFSE-labeled KU812 with human PBMC (effector to target ratio of 10:1) and the αENPP3 xαCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours. The controls used were the αRSV xαCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Consistent with the RTCC data, the αENPP3 xαCD3 bsAb dose-dependently induced CD4 + T cell activation; CD3 binding affinity correlated with activation potency (i.e., bsAbs with high CD3 induced CD4 + T cell activation more effectively than bsAbs with high-Int CD3 number 1); and bsAbs with the AN1 binding domain induced CD4 + T cell activation more effectively than bsAbs with the H16-7.8 binding domain.

[0107] Figures 26A to 26C Depicted activated CD8 + T cells, as indicated by: A) CD8 + T cell CD107a MFI, B) CD8 +CD25 MFI on T cells, and C) CD8 after incubation of CFSE-labeled KU812 with human PBMC (effector-to-target ratio of 10:1) and αENPP3 x αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours + CD69 MFI on T cells. The controls used were αRSV x αCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Consistent with the RTCC data, αENPP3 x αCD3 bsAb induced CD8 + T cell activation in a dose-dependent manner; CD3 binding affinity was correlated with activation potency (i.e., bsAbs with high CD3 induced CD8 + T cell activation more effectively than bsAbs with high-CD3-Int number 1); and bsAbs with the AN1 binding domain induced CD8 + T cell activation more effectively than bsAbs with the H16-7.8 binding domain.

[0108] Figures 27A to 27B Depicted RTCC induction on CFSE-labeled RXF393 cells, A) as indicated by the decrease in the number of CFSE + RXF393 cells and B) as indicated by the percentage of CFSE + RXF393 cells stained with Zombie Aqua after incubation of CFSE-labeled RXF393 with human PBMC (effector-to-target ratio of 20:1) and αENPP3 x αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours. The controls used were αRSV x αCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Consistent with the data on KU812 cells, the data showed that the prototype αENPP3 x αCD3 bsAb induced redirected T cell cytotoxicity (RTCC) on RXF393 cells in a dose-dependent manner; CD3 binding affinity was correlated with RTCC potency (i.e., bsAbs with high CD3 induced RTCC more effectively than bsAbs with high-CD3-Int number 1); and bsAbs with the AN1 binding domain induced RTCC more effectively than bsAbs with the H16-7.8 binding domain.

[0109] Figures 28A to 28C Depicted activated CD4 + T cells, as indicated by: A) CD4 +MFI of CD107a on T cells, B) CD4 + MFI of CD25 on T cells, and C) CD4 + MFI of CD69 on T cells after incubation of CFSE-labeled RXF393 with human PBMC (effector-to-target cell ratio of 20:1) and αENPP3 x αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours. The controls used were αRSV x αCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Consistent with the RTCC data, αENPP3 x αCD3 bsAb induced CD4 + T cell activation in a dose-dependent manner; CD3 binding affinity was correlated with activation potency (i.e., bsAbs with high CD3 induced CD4 + T cell activation more effectively than bsAbs with high-CD3-Int number 1); and bsAbs with the AN1 binding domain induced CD4 + T cell activation more effectively than bsAbs with the H16-7.8 binding domain.

[0110] Figures 29A to 29C Depicted activated CD8 + T cells, as indicated by: A) CD8 + MFI of CD107a on T cells, B) CD8 + MFI of CD25 on T cells, and C) CD8 + MFI of CD69 on T cells after incubation of CFSE-labeled RXF393 with human PBMC (effector-to-target cell ratio of 20:1) and αENPP3 x αCD3 bispecific antibodies (XENP26820, XENP26821, XENP28287, and XENP28390) for 24 hours. The controls used were αRSV x αCD3 bispecific antibody (XENP13245), effector cells and target cells only, and target cells only. Consistent with the RTCC data, αENPP3 x αCD3 bsAb induced CD8 + T cell activation in a dose-dependent manner; CD3 binding affinity was correlated with activation potency (i.e., bsAbs with high CD3 induced CD8 + T cell activation more effectively than bsAbs with high-CD3-Int number 1); and bsAbs with the AN1 binding domain induced CD8 + T cell activation more effectively than bsAbs with the H16-7.8 binding domain.

[0111] Figure 30depicts A) a chromatogram showing the purified fraction 2 of XENP28287 (cation exchange chromatography after Protein A chromatography), and the purity and homogeneity of peaks B and BC (and pre-purified material) isolated from the cation exchange separation as depicted in Figure 30 A, as determined by B) analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS) and C) analytical cation exchange chromatography (aCIEX). Figure 30 B also depicts the molecular weight of the protein species in the peaks as determined by multi-angle light scattering.

[0112] Figure 31 depicts A) a chromatogram showing the purified fraction 2 of XENP28925 (cation exchange chromatography after Protein A chromatography), and the purity and homogeneity of peak B (and pre-purified material) isolated from the cation exchange separation as depicted in Figure 31 A, as determined by B) analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS) and C) analytical cation exchange chromatography (aCIEX). Figure 31 B also depicts the molecular weight of the protein species in the peaks as determined by multi-angle light scattering.

[0113] Figure 32 depicts A) a chromatogram showing the purified fraction 2 of XENP31149 (cation exchange chromatography after Protein A chromatography), and B) the identity of peaks A and B (and pre-purified material) isolated from the cation exchange separation as depicted in Figure XA, as determined by analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS).

[0114] Figure 33 depicts A) a chromatogram showing the purified fraction 2 of XENP31419 (cation exchange chromatography after Protein A chromatography), and B) the identity of peaks A and B (and pre-purified material) isolated from the cation exchange separation as depicted in Figure XA, as determined by analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS).

[0115] Figure 34 depicts CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, ENPP3 低)RTCC induction on cells, as indicated by the percentage of CFSE-stained cells with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibody XENP28925 for 18 hours. + Percentage of cells.

[0116] Figure 35 Depicts the binding of the affinity-engineered αENPP3 x αCD3 1+1 bsAb to ENPP3 高 KU812 cells.

[0117] Figure 36 Depicts RTCC induction on CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, ENPP3 低 ) cells, as indicated by the percentage of CFSE-stained cells with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibodies XENP28925 (WT high ENPP3 binding), XENP29516 (moderate ENPP3 binding), or XENP30262 (low ENPP3 binding) for 42 hours. Data show that compared to XENP28925, both XENP29516 and XENP30262 show substantially less efficient induction of RTCC on + RCC4 cells, where RTCC potency is correlated with binding potency as shown above. XENP29516 and XENP30262 also show less efficient induction of RTCC on 低 ENPP3 cells. 高 Less efficient induction of RTCC on cells.

[0118] Figures 37A to 37C Depicts the induction of A) IFNγ, B) IL-6, and C) TNFα release in human PBMC incubated with KU812 cells (effector-to-target ratio of 10:1) and the αENPP3xαCD3 bispecific antibodies XENP28925 (CD3 high) or XENP29436 (CD3 high - Int number 1) for 18 hours. Data indicate that compared to XENP28925, XENP29436 exhibits substantially less efficient induction of cytokine release.

[0119] Figure 38Depicts the induction of IFNγ release from human PBMCs incubated for 18 hours with RCC4 cells (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibodies XENP28925 (CD3 high) or XENP29436 (CD3 high - Int no. 1). Data show that in the presence of ENPP3 低 RCC4 cells, XENP29436 showed negligible induction of cytokine release compared to XENP28925.

[0120] Figure 39 Depicts the induction of RTCC on CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, RCC4 低 ) cells, as indicated by the percentage of CFSE + cells stained with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMCs (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibodies XENP28925 (CD3 high) or XENP29436 (CD3 high - Int no. 1) for 42 hours. Data indicate that XENP29436 showed substantially less effective induction of RTCC on ENPP3 低 cells compared to XENP28925; however, XENP29436 also showed reduced potency in inducing RTCC on ENPP3 高 cells.

[0121] Figure 40 Depicts the induction of IFNγ release from human PBMCs incubated with KU812 cells (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibodies XENP28925 (ENPP3 high; CD3 high), XENP29436 (ENPP3 high; CD3 high - Int no. 1), XENP29518 (ENPP3 moderate; CD3 high), XENP29463 (ENPP3 moderate; CD3 high - Int no. 1), XENP30262 (ENPP3 low; CD3 high) or XENP30263 (ENPP3 low; CD3 high - Int no. 1). Data show that reducing CD3 or ENPP3 binding potency reduced the induction of cytokine release. Notably, reducing both CD3 and ENPP3 binding potency further reduced the induction of cytokine release.

[0122] Figure 41 Depicts the induction of RTCC on CFSE-labeled KU812 (solid line, ENPP3 高) or CFSE-labeled RCC4 (dashed line, ENPP3 低 ) cells, as indicated by the percentage of CFSE + cells that were stained with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and the αENPP3 x αCD3 bispecific antibodies XENP28925 (WT high ENPP3 binding; CD3 high; monovalent ENPP3 binding), XENP29516 (moderate ENPP3 binding; CD3 high; monovalent ENPP3 binding), or XENP29520 (moderate ENPP3 binding; CD3 high; bivalent ENPP3 binding) for 42 hours. Data show that bivalent binding (with moderate ENPP3 binding) maintained reduced RTCC potency on ENPP3 低 cells but restored near the RTCC potency exhibited by XENP28925 on ENPP3 高 cells.

[0123] Figure 42 Depicts RTCC induction on CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, ENPP3 低 ) cells, as indicated by the percentage of CFSE + cells that were stained with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and the αENPP3 x 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; bivalent ENPP3 binding) for 42 hours. Data show that bivalent binding (with low ENPP3 binding) further reduced the RTCC potency on ENPP3 低 cells and restored some RTCC potency on ENPP3 高 cells.

[0124] Figure 43Depicts RTCC induction on CFSE-labeled KU812 as indicated by the percentage of CFSE-stained cells with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and αENPP3 x αCD3 bispecific antibodies XENP28925 (CD3 high; monovalent ENPP3 binding), XENP29437 (CD3 high; bivalent ENPP3 binding), XENP29436 (CD3 high - Int no. 1; monovalent ENPP3 binding), or XENP29438 (CD3 high - Int no. 1; bivalent ENPP3 binding) for 44 hours. Unexpectedly, XENP29438 failed to induce RTCC on KU812 cells. +

[0125] Figure 44 Depicts RTCC induction on CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, ENPP3 低 ) as indicated by the percentage of CFSE-stained cells with Zombie Aqua after incubation of CFSE-labeled target cells with human PBMC (effector-to-target ratio of 10:1) and αENPP3 x αCD3 bispecific antibodies XENP29437 (CD3 high VH / VL; bivalent ENPP3 binding), XENP30469 (CD3 high VL / VH; bivalent ENPP3 binding), XENP29428 (CD3 high - Int no. 1 VH / VL; bivalent ENPP3 binding), or XENP30470 (CD3 high - Int no. 2 VL / VH; bivalent ENPP3 binding) for 24 hours. Data show that swapping the orientation of the variable heavy chain domain and the variable light chain domain in the CD3 high - Int no. 1 scFv restored its activity in the context of the 2 + 1 Fab2 - scFv - Fc bsAb format (XENP29438 vs. XENP30470). Swapping the orientation of the variable heavy chain domain and the variable light chain domain in the CD3 high scFv led to a more modest improvement in RTCC potency in the context of the 2 + 1 Fab2 - scFv - Fc bsAb format (XENP29437 vs. XENP30469). +

[0126] Figure 45 Depicts RTCC induction on CFSE-labeled KU812 (solid line, ENPP3 高 ) or CFSE-labeled RCC4 (dashed line, ENPP3 低 ​​)RTCC induction on, as indicated by the percentage of CFSE-stained cells after incubation of CFSE-labeled target cells with human PBMC (effector-to-target cell ratio of 10:1) and αENPP3 x αCD3 bispecific antibodies XENP29520 (CD3 high [VH / VL]; bivalent ENPP3 moderate binding), XENP30819 (CD3 high-Int No. 1 [VL / VHL]; bivalent ENPP3 moderate binding), XENP31149 (CD3 high-Int No. 2 [VL / VHL]; bivalent ENPP3 moderate binding), XENP30264 (CD3 high [VH / VL]; bivalent ENPP3 low binding), XENP30821 (CD3 high-Int No. 1 [VL / VHL]; bivalent ENPP3 low binding) or XENP31150 (CD3 high-Int No. 2 [VL / VHL]; bivalent ENPP3 low binding) for 40 hours and staining with Zombie Aqua + as indicated by the percentage of cells.

[0127] Figure 46 Depicts the sequences of XENP16432, an anti-PD-1 mAb based on nivolumab and an Ig backbone with E233P / L234V / L235A / G236del / S267K ablation variants; and the sequence of XENP21461 (pembrolizumab).

[0128] Figure 47 Depicts the change in tumor volume over time (determined by caliper measurement) in NSG mice transplanted with KU812 and huPBMC administered with PBS, XENP16432 (bivalent anti-PD-1 mAb), or an illustrative αENPP3 x αCD3 2+1 bsAb (XENP30819, XENP30821 or XENP31419) used alone or in combination with XENP16432. Each of the αENPP3 x αCD3 bsAbs was able to enhance the allogeneic anti-tumor effect of T cells against KU812 cells and combined well with PD-1 blockade at low dose and / or higher dose treatments.

[0129] Figures 48A to 48C Depicts A) CD45 + lymphocytes, B) CD8 + T cells and C) CD4+ Expansion of T cells. In all cases, lymphocyte expansion was enhanced in combination with PD-1 blockade.

[0130] Figure 49 Depicts the change in tumor volume over time (determined by caliper measurement) in NSG mice transplanted with RXF-393 and huPBMCs dosed with A) PBS, B) XENP16432 (bivalent anti-PD-1 mAb), or an illustrative αENPP3 x αCD3 2+1 bsAb (XENP30819 or XENP31419) used alone or in combination with XENP16432. Each of the αENPP3 x αCD3 bsAbs was able to enhance the allogeneic anti-tumor effect of T cells against KU812 cells and combined well with PD-1 blockade at low, medium, and / or high dose treatments.

[0131] Figures 50A to 50N Depicts the change in tumor volume over time (determined by caliper measurement) in NSG mice transplanted with KU812 alone and huPBMCs dosed with A) PBS, B) XENP16432 (bivalent anti-PD-1 mAb), or an illustrative αENPP3 x αCD3 2+1 bsAb (XENP30819, XENP30821, or XENP31419) used alone or in combination with XENP16432. Each of the αENPP3 x αCD3 bsAbs was able to enhance the allogeneic anti-tumor effect of T cells against KU812 cells and combined well with PD-1 blockade at low and / or higher dose treatments.

[0132] Figures 51A to 51L Depicts the change in tumor volume over time (determined by caliper measurement) in NSG mice transplanted with RXF-393 alone and huPBMCs dosed with A) PBS, B) XENP16432 (bivalent anti-PD-1 mAb), or an illustrative αENPP3 x αCD3 2+1 bsAb (XENP30819 or XENP31419) used alone or in combination with XENP16432. Each of the αENPP3 x αCD3 bsAbs was able to enhance the allogeneic anti-tumor effect of T cells against KU812 cells and combined well with PD-1 blockade at low, medium, and / or high dose treatments.

[0133] Figures 52A to 52KDepicts several forms of the anti-ENPP3 x anti-CD3 bispecific antibody disclosed herein. The first is the "1+1Fab-scFv-Fc" form (also referred to as the "bottle opener" or "triple F" form), where the first antigen-binding domain is a Fab domain and the second antigen-binding domain is a scFv domain ( Figure 1A ). Additionally, "mAb-Fv", "mAb-scFv", "2+1Fab2-scFv-Fc" (also referred to as the "central scFv" or "central scFv" form), "central Fv", "single-arm central scFv", "single scFv-mAb", "scFv-mAb", "double scFv", "trifurcation", and non-heterodimeric bispecific forms are shown. Figure 49 The scFv domain depicted in Figure 49 can be variable heavy chain-(optional linker)-variable light chain or variable light chain-(optional linker)-variable heavy chain from the N-terminus to the C-terminus. Additionally, for the single-arm scFv-mAb, the scFv can be attached to the N-terminus of the heavy chain monomer or attached to 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 the disclosed ENPP3-binding domain or the CD3-binding domain can be included in the bispecific forms of Figure 52.

[0134] Figure 53 Provides schematic diagrams of the heterodimeric Fc proteins described herein, including 2:1Fab2-scFv-Fc, 1:1Fab=scFv-Fc, Y / Z-Fc (e.g., non-targeted interleukin-Fc), anti-X x Y / ZF (e.g., targeted interleukin-Fc)c, and single-arm Fc proteins.

[0135] Figure 54 Provides a structural model of the CH3-CH3 interface constructed using MOE based on Protein Data Bank entry 3AVE. A novel set of Fc substitutions enables a heterodimer yield of over 95% with little change in thermal stability.

[0136] Figure 55Depicts isotypic substitution for minimizing impact on tertiary structure. Engineered isoelectric point differences in the Fc region allow or facilitate direct purification of Fc heterodimers.

[0137] Figure 56 Depicts that hinge and CH2 substitutions abrogate FcγR binding.

[0138] Figures 57A to 57C Shows that the 2:1 Fab2-scFv-Fc format enables targeting of low-density tumor antigens on normal cells. Tuning TAA avidity and TAA / CD3 affinity enables selective cytotoxicity against cell lines mimicking cancerous and normal tissues (high / low antigen density). The tuned 2:1 bispecific antibody also reduces interference from soluble antigen and reduces cytokine release.

[0139] Figure 57A Shows that tuning FAP avidity and FAP / CD3 affinity enables selective cytotoxicity against cell lines mimicking cancerous and normal tissues (high / low antigen density). XENP23535 represents the tuned 1:1 format targeting FAP. XENP25967 represents the tuned 2:1 format targeting FAP.

[0140] Figure 57B Shows that tuning SSTR2 avidity and SSTR2 / CD3 affinity enables selective cytotoxicity against cell lines mimicking cancerous and normal tissues (high / low antigen density). XENP18087 represents the tuned 1:1 format targeting SSTR2. XENP30458 represents the tuned 2:1 format targeting SSTR2.

[0141] Figure 57C Shows that tuning ENPP3 avidity and ENPP3 / CD3 affinity enables selective cytotoxicity against cell lines mimicking cancerous and normal tissues (high / low antigen density). XENP28925 represents the tuned 1:1 format targeting ENPP3. XENP31149 represents the tuned 2:1 format targeting ENPP3.

[0142] Figure 58 Depicts the advantages of studying scale production of heterodimeric Fc proteins using the methods described herein. The method can be used for direct production of heterodimeric Fc proteins.

[0143] Figure 59 Shows the results of stable cell line development in clones with high titer and high heterodimer prevalence. The shake flask yield of the top clone is 1 - 2 g / L, with a heterodimer content of approximately 90%. Data was obtained after only the standard protein A purification step.

[0144] Figure 60Depicts the induction of RTCC on A549 cells transfected with SSTR2 (high, medium, and low density) by A) XENP18087 or B) XENP30458.

[0145] Figure 61 Depicts the reduction in the number of effector cells against A) target cells and the release of B) IL-6, C) TNFα, D) IFNγ, and E) IL-1β after incubation of CFSE-labeled SSTR2+COR-L279 target cells with human PBMC (effector:target ratio of 20:1) in the presence of XENP18087 or XENP30458 for 48 hours.

[0146] Figures 62A to 62D . The sequences of the illustrative 1:1-tuned form and 2:1-tuned form TAAx CD3 bispecifics described herein are shown. The anti-TTA (e.g., anti-FAP, anti-SSTR2, and anti-ENPP3) moieties such as variable regions, anti-CD3 moieties such as variable regions, constant / Fc regions, and linkers are shown. The linker is double-underlined (however, those skilled in the art will understand that the linker can be replaced by other linkers), and the slashes ( / ) indicate the boundaries between variable regions, constant regions / Fc regions, and linkers. The CDRs are underlined. In some embodiments, the 1:1 form of the TAAx CD3 bispecific is XENP23535, XENP18087, or XENP28925. In some embodiments, the 2:1 form of the TAAx CD3 bispecific is XENP25967, XENP30458, XENP31149.

[0147] Figure 63 Depicts the sequence of the SSTR2-binding domain [αSSTR2]_H1.24_L1.30. Detailed Description

[0148] I. Overview

[0149] Anti-bispecific antibodies that co-bind CD3 and tumor antigen targets are used to redirect T cells to attack and lyse targeted tumor cells. Examples include and DART forms, which monovalently bind CD3 and tumor antigen. While the approach of targeting CD3 has shown considerable promise, a common side effect of such therapies is the associated cytokine production, which often leads to toxic cytokine release syndrome. Since the anti-CD3 binding domain of the bispecific antibody binds to all T cells, the highly cytokine-producing CD4 T cell subset is recruited. In addition, the CD4 T cell subset includes regulatory T cells, and the recruitment and expansion of these regulatory T cells may lead to immunosuppression and have a negative impact on long-term tumor suppression. In addition, these forms do not contain an Fc domain and show a very short serum half-life in patients.

[0150] The present disclosure provides novel anti-CD3 x anti-ENPP3 (also referred to as anti-ENPP3 x anti-CD3, αCD3 x αENPP3, or αENPP3 x αCD3) heterodimeric bispecific antibodies and methods of using such antibodies to treat cancer. Specifically, the present disclosure provides various forms of anti-CD3, anti-ENPP3 bispecific antibodies, such as Figure 15A and Figure 15B those depicted in. These bispecific antibodies can be used to treat cancer, particularly those cancers with increased ENPP3 expression, such as renal cell carcinoma. Such antibodies are used to direct CD3+ effector T cells to ENPP3+ tumors, thereby allowing CD3+ effector T cells to attack and lyse ENPP3+ tumors.

[0151] In addition, in some embodiments, the present disclosure provides bispecific antibodies with different binding affinities for human CD3, which can alter or reduce the potential side effects of anti-CD3 therapy. That is, in some embodiments, the antibodies described herein provide antibody constructs that comprise an anti-CD3 antigen-binding domain that is a "strong" or "high-affinity" binder to CD3 (e.g., one example is the heavy and light chain variable domains described as H1.30_L1.47 (optionally including an appropriate charged linker)) and also binds to ENPP3. In other embodiments, the antibodies described herein provide antibody constructs that comprise an anti-CD3 antigen-binding domain that is a "weak" or "low-affinity" binder to CD3. Additional embodiments provide antibody constructs that comprise an anti-CD3 antigen-binding domain that has a moderate or "medium" affinity for CD3 that also binds to ENPP3. Although a very large number of anti-CD3 antigen-binding domains (ABDs) can be used, particularly useful embodiments use 6 different anti-CD3 ABDs, although they can be used in two scFv orientations as discussed herein. Affinity is typically measured using Biacore assays.

[0152] It should be understood that the "high, medium, low" anti-CD3 sequences provided herein can be used in a variety of heterodimerized forms, such as Figure 15A , Figure 15B and as shown. Generally, due to the potential side effects of T cell recruitment, exemplary embodiments utilize forms that bind CD3 only monovalently, such as Figure 15A and Figure 15B depicted in, and in the forms depicted herein, the CD3 ABD is an scFv as more fully described herein. In contrast, the subject bispecific antibodies can be monovalent (e.g., Figure 15A) or bivalent (e.g., Figure 15B ) to bind ENPP3.

[0153] Compositions containing an ENPP3 binding domain are provided herein, and such compositions include antibodies having such an ENPP binding domain (e.g., an ENPP3 x CD3 bispecific antibody). The subject antibodies containing such an ENPP3 binding domain advantageously elicit a range of different immune responses, depending on the particular ENPP3 binding domain used. For example, the subject antibodies exhibit differences in selectivity for cells having different ENPP3 expression, potency against ENPP3-expressing cells, ability to trigger cytokine release, and sensitivity to soluble ENPP3. Such ENPP3 binding domains and related antibodies can be used, for example, to treat ENPP3-related cancers.

[0154] Accordingly, in one aspect, heterodimeric antibodies that bind two different antigens are provided herein, e.g., the antibodies are "bispecific" in that the antibodies bind two different target antigens, typically ENPP3 and CD3 as described herein. These heterodimeric antibodies can bind monovalently to these target antigens (e.g., there is a single antigen-binding domain, such as a variable heavy chain and variable light chain domain pair) or bivalently (there are two antigen-binding domains that each independently bind to the antigen). In some embodiments, the heterodimeric antibodies provided herein contain one CD3 binding domain and one ENPP3 binding domain (e.g., the "1+1Fab-scFv-Fc" form of heterodimeric antibody described herein). In other embodiments, the heterodimeric antibodies provided herein include one CD3 binding domain and two ENPP3 binding domains (e.g., the "2+1Fab2-scFv-Fc" form of heterodimeric antibody described herein). The heterodimeric antibodies provided herein are based on the use of different monomers that contain amino acid substitutions that cause the heterodimer to "skew" formation relative to the homodimer, as more fully outlined below, and the different monomers bind to "pI variants" that allow for facile purification away from the homodimer, as similarly outlined below. The heterodimeric bispecific antibodies provided generally rely on the use of engineered or variant Fc domains that can self-assemble in production cells to produce the heterodimeric protein, as well as methods for producing and purifying such heterodimeric proteins.

[0155] II. Nomenclature

[0156] The antibodies provided herein are listed in several different forms. In some instances, each monomer in a particular antibody is given a unique "XENP" number, but as will be understood in the art, longer sequences may contain shorter XENP numbers. For example, the "scFv-Fc" monomer of a 1+1 Fab-scFv-Fc form antibody may have a first XENP number, while the scFv domain itself will have a different XENP number. Some molecules have three polypeptides, and thus the XENP numbers of the components are used as names. Thus, the molecule XENP29520 in the 2+1 Fab2-scFv-Fc form includes three sequences (see Figure 19A ): 1) the "Fab-Fc heavy chain" monomer; 2) the "Fab-scFv-Fc heavy chain" monomer; and 3) the "light chain" monomer or equivalent, but those skilled in the art will be able to readily identify these sequences by sequence alignment. These XENP numbers are in the sequence listing and identifiers and are used in the figures. Additionally, a molecule containing three components generates multiple sequence identifiers. For example, the list of Fabs includes the full heavy chain sequence, the variable heavy chain domain sequence and the three CDRs of the variable heavy chain domain sequence, the full light chain sequence, the variable light chain domain sequence and the three CDRs of the variable light chain domain sequence. The Fab-scFv-Fc monomer includes the full-length sequence, the variable heavy chain domain sequence, the 3 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). It should be noted that some molecules with an scFv domain herein use a single charged scFv linker (+H), but others may also be used. Additionally, the nomenclature for specific antigen-binding domains (e.g., ENPP3 and CD3-binding domains) uses the "Hx.xx_Ly.yy" class form, where the numbers serve as unique identifiers for specific variable chain sequences. Thus, the variable domain on the Fab side of the CD3-binding domain AN1[ENPP3]H1L1 (e.g., Figure 12 ) is "H1 L1", which indicates the combination of the variable heavy chain domain H1 and the light chain domain L1. In the case where these sequences are used as an scFv, the name "H1L1" indicates the variable heavy chain domain, H1 combined with the light domain L1 and in the VH-linker-VL orientation from the N-terminus to the C-terminus. Such a molecule with the same sequences of the heavy chain variable domain and the light chain variable domain but the sequences in reverse order (VL-linker-VH orientation from the N-terminus to the C-terminus) will be named "L1_H1.1". Similarly, different constructs may "mix and match" heavy and light chains, as will be apparent from the sequence listing and the figures.

[0157] III. Definitions

[0158] To more fully understand the present application, several definitions are set forth below. Such definitions are intended to cover grammatical equivalents.

[0159] "ENPP3" or "ectonucleotide pyrophosphatase / phosphodiesterase family member 3" (e.g., GenBank accession number NP005012.2) refers herein to a protein belonging to a family of extracellular enzymes involved in the hydrolysis of extracellular nucleotides. ENPP3 sequences are depicted, for example, Figure 11A and Figure 11B . ENPP3 is expressed in certain cancers, including renal cell carcinoma.

[0160] As used herein, "ablation" means a reduction or removal of activity. Thus, for example, "ablation of FcγR binding" means that the Fc region amino acid variant has less than 50% of the starting binding compared to the Fc region without the specific variant, preferably greater than 70% to 80% to 90% to 95% to 98% loss of activity, and generally, the activity is below the level of detectable binding in Biacore, SPR, or BLI assays. Particularly useful for ablation of FcγR binding are Figure 5 the variants shown, which variants are typically added to both monomers.

[0161] As used herein, "ADCC" or "antibody-dependent cell-mediated cytotoxicity" means a cell-mediated reaction in which non-specific cytotoxic cells expressing FcγR recognize a bound antibody on a target cell and subsequently cause lysis of the target cell. ADCC is associated with binding to FcγRIIIa; increased binding to FcγRIIIa results in increased ADCC activity.

[0162] As used herein, "ADCP" or antibody-dependent cell-mediated phagocytosis means a cell-mediated reaction in which non-specific phagocytic cells expressing FcγR recognize a bound antibody on a target cell and subsequently cause phagocytosis of the target cell.

[0163] As used herein, the term "antibody" is generally used. The antibodies described herein can take various forms as described herein, including conventional antibodies as well as antibody derivatives, fragments, and mimics.

[0164] Conventional immunoglobulin (Ig) antibodies are "Y"-shaped tetramers. Each tetramer is typically composed of two pairs of identical polypeptide chains, each pair having one "light chain" monomer (usually with a molecular weight of about 25 kDa) and one "heavy chain" monomer (usually with a molecular weight of about 50-70 kDa).

[0165] Other useful antibody forms include, but are not limited to, the 1+1 Fab-scFv-Fc form and the 2+1 Fab-scFv-Fc antibody form described herein, as well as "mAb-Fv", "mAb-scFv", "central-Fv", "single-arm scFv-mAb", "scFv-mAb", "bis-scFv", and "trifunctional" forms of antibodies, as Figure 49 shown.

[0166] An antibody heavy chain generally comprises a variable heavy chain (VH) domain, which contains vhCDR1-3; and an Fc domain, which contains CH2-CH3 monomers. In some embodiments, the antibody heavy chain comprises a hinge and CH1 domains. A conventional antibody heavy chain is a monomer organized from the N-terminus to the C-terminus as follows: VH-CH1-hinge-CH2-CH3. CH1-hinge-CH2-CH3 are collectively referred to as the antibody heavy chain "constant domain" or "constant region", and there are five different classes or "isotypes": IgA, IgD, IgG, IgE, and IgM. Thus, as used herein, "isotype" means any subclass of immunoglobulins defined according to the chemical and antigenic characteristics of their constant regions. It should be understood that therapeutic antibodies can also comprise hybrids of isotypes and / or subclasses. For example, as shown in U.S. Publication 2009 / 0163699, which is incorporated by reference herein, the antibodies described herein comprise the use of a human IgG1 / G2 hybrid.

[0167] In some embodiments, the antibodies provided herein include IgG isotype constant domains, which have several subclasses, including but not limited to IgG1, IgG2, IgG3, and IgG4. In the IgG subclasses of immunoglobulins, there are several immunoglobulin domains present in the heavy chain. As used herein, "immunoglobulin (Ig) domain" means an immunoglobulin region having a different tertiary structure. Of interest in the antibodies described herein are the heavy chain domains, including the constant heavy chain (CH) domains and the hinge domain. In the context of IgG antibodies, each IgG isotype has three CH regions. Thus, in the context of IgG, the "CH" domains are as follows: "CH1" refers to positions 118-220 according to the EU numbering as in Kabat. "CH2" refers to positions 237-340 according to the EU numbering as in Kabat, and "CH3" refers to positions 341-447 according to the EU numbering as in Kabat. As shown herein and discussed below, pI variants can be in one or more of the CH regions as well as the hinge region, as discussed below.

[0168] It should be noted that IgG1 has different allotypes with polymorphisms at 356 (D or E) and 358 (L or M). The sequences described herein use the 356D / 358M allotype, however other allotypes are included herein. That is, any sequence comprising an IgG1 Fc domain included herein may have 356E / 358L in place of the 356D / 358M allotype. It should be understood that therapeutic antibodies may also comprise hybrids of isotypes and / or subclasses. For example, as shown in U.S. Publication 2009 / 0163699, which is incorporated by reference, the antibodies of the invention in some embodiments comprise an IgG1 / IgG2 hybrid.

[0169] As used herein, "Fc" or "Fc region" or "Fc domain" means a polypeptide comprising the constant region of an antibody, in some cases not including all of the first constant region immunoglobulin domains (e.g., CH1) or a portion thereof, and in some cases optionally including all or a portion of the hinge. For IgG, the Fc domain includes the immunoglobulin domains CH2 and CH3 (Cγ2 and Cγ3), and optionally all or a portion of the hinge region between CH1 (Cγ1) and CH2 (Cγ2). Thus, in some cases, the Fc domain comprises CH2-CH3 and hinge-CH2-CH3 from the N-terminus to the C-terminus. In some embodiments, the Fc domain is an Fc domain from IgG1, IgG2, IgG3 or IgG4, with IgG1 hinge-CH2-CH3 and IgG4 hinge-CH2-CH3 being particularly useful in many embodiments. Additionally, in the case of the human IgG1 Fc domain, the hinge often contains a C220S amino acid substitution. Further, in the case of the human IgG4 Fc domain, the hinge often contains an S228P amino acid substitution. Although the boundaries of the Fc region can vary, the human IgG heavy chain Fc region is generally defined to include residues E216, C226 or A231 at its carboxyl terminus, numbered according to the EU index as in Kabat. In some embodiments, as described more fully below, amino acid modifications are made to the Fc region, such as altering binding to one or more FcγRs or to FcRn.

[0170] As used herein, "heavy chain constant region" means the CH1-hinge-CH2-CH3 portion of an antibody (or a fragment thereof) without the variable heavy chain domain; in the EU numbering of human IgG1, this is amino acids 118 to 447. As used herein, "heavy chain constant region fragment" means a heavy chain constant region that contains fewer amino acids from either or both the N-terminus and C-terminus but still retains the ability to dimerize with another heavy chain constant region.

[0171] Another type of Ig domain of the heavy chain is the hinge region. As used herein, "hinge" or "hinge region" or "antibody hinge region" or "hinge domain" means a flexible polypeptide comprising the amino acids between the first constant domain and the second constant domain of an antibody. Structurally, the IgG CH1 domain ends at EU position 215, and the IgG CH2 domain begins at residue EU position 231. Thus, for IgG herein, the antibody hinge is defined to include positions 216 (E216 in IgG1) to 230 (p230 in IgG1), where the numbering is according to the EU index as in Kabat. In some cases, a "hinge fragment" is used, which contains fewer amino acids at either or both of the N-terminus and C-terminus of the hinge domain. As mentioned herein, pI variants can also be made in the hinge region. Many antibodies herein have at least one cysteine at position 220 according to EU numbering (hinge region) that is replaced by serine. Typically, this modification is on the "scFv monomer" side of most of the sequences depicted herein, but it can also be on the "Fab monomer" side or on both to reduce disulfide formation. One or both of these cysteines are specifically included as substituted (C220S) in the sequences herein.

[0172] As will be appreciated by those skilled in the art, the exact numbering and positions of the heavy chain constant region domains may vary between different numbering systems. A useful comparison of the heavy chain constant regions according to EU and Kabat numbering is as follows, 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, which are incorporated herein by reference in their entirety.

[0173] Table 1

[0174] EU Number Kabat Number CH1 118-215 114-223 Hinge 216-230 226-243 CH2 231-340 244-360 CH3 341-447 361-478

[0175] An antibody light chain typically comprises two domains: a variable light chain domain (VL) that contains the light chain CDRs vlCDR1-3, and a constant light chain region (often referred to as CL or Cκ). An antibody light chain is typically organized from the N-terminus to the C-terminus as follows: VL-CL.

[0176] As used herein, an "antigen-binding domain" or "ABD" means a set of six complementarity determining regions (CDRs) that specifically bind to a target antigen as discussed herein (e.g., ENPP3 or CD3) when present as part of a polypeptide sequence. As is known in the art, these CDRs typically 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 CDR and vlCDR1, vlCDR2, and vlCDR3 vhCDR3 variable light chain CDR. The CDRs are present in the variable heavy chain domain (vhCDR1-3) and the variable light chain domain (vlCDR1-3). The variable heavy chain domain and the variable light chain domain from the Fv region.

[0177] The antibodies described herein provide a large number of different sets of CDRs. In this context, a "complete set of CDRs" contains three variable light chain and three variable heavy chain CDRs, e.g., vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. These can be part of a larger variable light chain domain or variable heavy chain domain, respectively. Additionally, as more fully outlined herein, when using heavy and light chains (e.g., when using Fab), the variable heavy chain domain and the variable light chain domain can be on separate polypeptide chains, or in the case of a scFv sequence on a single polypeptide chain.

[0178] As will be understood by those skilled in the art, the exact numbering and placement of CDRs can be different in different numbering systems. However, it should be understood that the disclosure of a variable heavy chain sequence and / or a variable light chain sequence includes the disclosure of the relevant (inherent) CDRs. Thus, the disclosure of each variable heavy region is the disclosure of the vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the disclosure of each variable light region is the disclosure of the vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3). A useful comparison of CDR numbering is as follows, see Lafranc et al., Dev. Comp. Immunol. 27(1):55-77 (2003):

[0179] Table 2

[0180]

[0181] Throughout this specification, when referring to residues in the variable domain (roughly residues 1-107 of the light chain variable region and residues 1-113 of the heavy chain variable region) and the EU numbering system for the Fc region, the Kabat numbering system is generally used (e.g., Kabat et al., supra (1991)).

[0182] CDRs contribute to the formation of the antigen-binding domain and the antigen-binding site of an antibody, or more specifically, the epitope-binding site. An "epitope" is a determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule called a paratope. An epitope is a group of molecules such as amino acids or sugar side chains and typically has specific structural and specific charge characteristics. A single antigen can have more than one epitope.

[0183] An epitope can include amino acid residues that directly participate in binding (also called the immunodominant component of the epitope) and other amino acid residues that do not directly participate in binding, such as amino acid residues that are effectively blocked by a specific antigen-binding peptide; in other words, the amino acid residues are within the coverage area of the specific antigen-binding peptide.

[0184] An epitope can be conformational or linear. Conformational epitopes are generated by amino acids that are spatially juxtaposed from different segments of a linear polypeptide chain. Linear epitopes are epitopes generated by adjacent amino acid residues in a polypeptide chain. The difference between conformational and non-conformational epitopes can be that binding is lost with the former but not the latter in the presence of a denaturing solvent.

[0185] An epitope typically includes at least 3, and more typically at least 5 or 8 - 10 amino acids in a unique spatial conformation. Antibodies that recognize the same epitope can be verified in a simple immunoassay, demonstrating the ability of one antibody to block the binding of another antibody to a target antigen, such as "capping". As outlined below, the present disclosure includes not only the antigen-binding domains and antibodies enumerated herein, but also antigen-binding domains and antibodies that competitively bind to epitopes that bind to the enumerated antigen-binding domains.

[0186] In some embodiments, the six CDRs of the antigen-binding domain are derived from the variable heavy chain domain and the variable light chain domain. In the "Fab" form, a set of 6 CDRs are derived from two different polypeptide sequences, the variable heavy chain domain (vh or VH; containing vhCDR1, vhCDR2, and vhCDR3) and the variable light chain domain (vl or VL; containing vlCDR1, vlCDR2, and vlCDR3), the C-terminus of the vh domain is linked to the N-terminus of the CH1 domain of the heavy chain and the C-terminus of the vl domain is linked to the N-terminus of the constant light chain (and thus forms the light chain). In the scFv form, the vh domain and the vl domain are typically bivalently linked into a single polypeptide sequence using a linker as outlined herein (the "scFv linker"), which can be (starting from the N-terminus) vh-linker-vl or vl-linker-vh, typically preferably the former (including optional domain linkers on each side, depending on the form used (e.g., according to Figure 1)). Typically, the C-terminus of the scFv domain is attached to the N-terminus of the hinge in a second monomer.

[0187] As used herein, "variable region" or "variable domain" means a region of an immunoglobulin that includes one or more Ig domains encoded substantially by any of the Vκ, Vλ, and / or VH genes that respectively make up the κ, λ, and heavy chain immunoglobulin gene loci and contains CDRs that confer antigen specificity. Thus, a "variable heavy chain domain" pairs with a "variable light domain" to form an antigen binding domain ("ABD"). Additionally, each variable domain includes three hypervariable regions ("complementary determining regions", "CDRs") (VHCDR1, VHCDR2, and VHCDR3 of the variable heavy chain domain and VLCDR1, VLCDR2, and VLCDR3 of the variable light domain) and four framework (FR) regions arranged in the following order from the amino terminus to the carboxy terminus: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. The hypervariable regions generally encompass the following amino acid residues: approximately amino acid residues 24-34 (LCDR1; "L" denotes light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the variable light chain region and approximately 31-35B (HCDR1; "H" denotes heavy chain), 50-65 (HCDR2), and 95-102 (HCDR3) in the variable heavy chain 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 hypervariable loops (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3) in the variable light chain region and 26-32 (HCDR1), 53-55 (HCDR2), and 96-101 (HCDR3) in the variable heavy chain region; Chothia and Lesk (1987) J. Mol. Biol. 196:901-917. Specific CDRs of the invention are described in Table 2.

[0188] As used herein, "Fab" or "Fab region" means a polypeptide that includes the VH, CH1, VL, and CL immunoglobulin domains, which domains are typically on two different polypeptide chains (e.g., VH-CH1 on one chain and VL-CL on another chain). Fab can refer to this region in isolation or in the context of the bispecific antibodies described herein. In the context of Fab, in addition to the CH1 and CL domains, Fab also includes the Fv region.

[0189] As used herein, "Fv" or "Fv fragment" means a polypeptide comprising the VL and VH domains of an ABD. The Fv region can be formatted as both Fab (as discussed above, typically two different polypeptides that also include constant regions as outlined above) and scFv, where the VL and VH domains are combined (typically with a linker as discussed herein) to form the scFv.

[0190] As used herein, "single-chain Fv" or "scFv" means a variable heavy chain domain that is covalently attached to a variable light chain domain using an scFv linker as discussed herein to form an scFv or scFv domain. The scFv domain can be in any orientation from N-terminus to C-terminus (VH-linker-VL or VL-linker-VH). In the sequences depicted in the sequence listing 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 from N-terminus to C-terminus, and L.Y_H.X is VL-linker-VH.

[0191] Some embodiments of the antibodies of the invention provided herein include at least one scFv domain, which, although not naturally occurring, typically comprises a variable heavy chain domain and a variable light domain joined together by an scFv linker. As outlined herein, although the scFv domain is typically oriented from N-terminus to C-terminus as VH-scFv linker-VL, this orientation can be reversed for any scFv domain (or those constructed using the vh and vl sequences of a Fab), to be VL-scFv linker-VH, where an optional linker (depending on the format) is located at one or both ends.

[0192] "Modification" as used herein means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence or a change to a moiety chemically linked to the protein. For example, a modification can be an altered carbohydrate or PEG structure attached to the protein. "Amino acid modification" as used herein means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence. For clarity, unless otherwise stated, amino acid modifications are always with respect to the amino acids encoded by DNA, e.g., the 20 amino acids with codons in DNA and RNA.

[0193] As used herein, "amino acid substitution" or "substitution" means replacing the amino acid at a specific position in a parental polypeptide sequence with a different amino acid. In particular, in some embodiments, the substitution is for an amino acid that is not naturally present at a particular position, not naturally present in a living organism or any living organisms. For example, substitution E272Y refers to a variant polypeptide, in this case an Fc variant, in which the glutamic acid at position 272 is replaced with tyrosine. For clarity, a protein that has been engineered to alter the nucleic acid coding sequence but not the starting amino acid (e.g., changing CGG (encoding arginine) to CGA (still encoding arginine) to increase host organism expression levels) is not an "amino acid substitution"; that is, although a new gene encoding the same protein is generated, if the protein has the same amino acid at the particular position where it starts, the protein is not an amino acid substitution.

[0194] As used herein, "amino acid insertion" or "insertion" means adding an amino acid sequence at a specific position in a parental polypeptide sequence. For example, -233E or 233E represents the insertion of glutamic acid after position 233 and before position 234. Additionally, -233ADE or A233ADE represents the insertion of AlaAspGlu after position 233 and before position 234.

[0195] As used herein, "amino acid deletion" or "deletion" means removing the amino acid sequence at a specific position in a parental polypeptide sequence. For example, E233- or E233#, E233() or E233del represents the deletion of the glutamic acid at position 233. Additionally, EDA233- or EDA233# represents the deletion of the sequence GluAspAla starting from position 233.

[0196] As used herein, "variant protein" or "protein variant" or "variant" means a protein that is different from a parental protein due to at least one amino acid modification. Compared to the parental protein, the protein variant has at least one amino acid modification, but not so many that the variant protein will not align with the parental protein using the alignment procedures described below. Generally, using alignment procedures such as BLAST described below, a variant protein (such as a variant Fc domain outlined herein) is typically at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the parental protein. As used herein, "variant" also refers to a specific amino acid modification that confers a particular function (e.g., "heterodimerization variant", "pI variant", "ablated variant", etc.).

[0197] As described below, in some embodiments, the parent polypeptide, e.g., the Fc parent polypeptide is a human wild-type sequence, such as a heavy constant domain or Fc region from IgG1, IgG2, IgG3, or IgG4, but a human sequence with a variant can also be used as a "parent polypeptide", such as an IgG1 / 2 hybrid of U.S. Publication 2006 / 0134105. The protein variant sequence herein will preferably have at least about 80% identity with the parent protein sequence, and most preferably at least about 90% identity, more preferably at least about 95% to 98% to 99% identity. Thus, as used herein, "antibody variant" or "variant antibody" means an antibody that is different from a parent antibody due to at least one amino acid modification, as used herein, "IgG variant" or "variant IgG" means an antibody that is different from a parent IgG (again, in many cases, from a human IgG sequence) due to at least one amino acid modification, and as used herein, "immunoglobulin variant" or "variant immunoglobulin" means an immunoglobulin sequence that is different from a parent immunoglobulin sequence due to at least one amino acid modification. As used herein, "Fc variant" or "variant Fc" means a protein that includes amino acid modifications in the Fc domain as compared to the Fc domain of human IgG1, IgG2 or IgG4.

[0198] As used herein, "Fc variant" or "variant Fc" means a protein comprising an amino acid modification in an Fc domain. The modification may be an addition, a deletion or a substitution. Fc variants are defined according to the amino acid modifications that constitute them. Thus, for example, N434S or 434S is an Fc variant having a serine substitution at position 434 relative to a parent Fc polypeptide, wherein the numbering is according to the EU index. Similarly, M428L / N434S defines an Fc variant having substitutions M428L and N434S relative to a parent Fc polypeptide. The identity of the WT amino acid may be non-specific, in which case the aforementioned variant is referred to as 428L / 434S. It should be noted that the order in which the substitution is provided is arbitrary, that is, for example, 428L / 434S is an Fc variant identical to 434S / 428L, etc. For all positions involving antibodies or derivatives thereof and fragments (e.g., Fc domains) discussed herein, unless otherwise noted, the amino acid position numbering is according to the EU index. The “EU index” or “EU index” as in Kabat or “EU numbering” scheme refers to the EU numbering of antibodies (Edelman et al., 1969, Proc. Natl. Acad. Sci. USA 63:78-85, which is hereby incorporated by reference in its entirety).

[0199] Typically, a variant Fc domain has at least about 80%, 85%, 90%, 95%, 97%, 98% or 99% identity with the corresponding parental human IgG Fc domain (using the identity algorithms discussed below, where in one embodiment the BLAST algorithm known in the art is utilized with default parameters). Alternatively, compared to the parental Fc domain, the variant Fc domain can 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. Alternatively, compared to the parental Fc domain, the variant Fc domain can have at most 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. Additionally, as discussed herein, the variant Fc domains described herein still retain the ability to form dimers with another Fc domain, as measured using known techniques described herein such as non-denaturing gel electrophoresis.

[0200] As used herein, "protein" means at least two covalently linked amino acids, which includes proteins, polypeptides, oligopeptides, and peptides. Additionally, the polypeptides that make up the antibodies described herein can include one or more synthetic derivatizations, glycosylations, polyethylene glycolations, cyclization conversions, cyclizations, linkers of other molecules, fusions with proteins or protein domains, and addition of peptide tags or labels at the side chains or termini.

[0201] As used herein, "residue" means a position in a protein and its associated amino acid entity. For example, asparagine 297 (also referred to as Asn297 or N297) is the residue at position 297 in human antibody IgG1.

[0202] As used herein, "IgG subclass modification" or "isotype modification" means an amino acid modification that converts an amino acid of one IgG isotype to the corresponding amino acid in a different aligned IgG isotype. For example, since IgG1 contains tyrosine at EU position 296 and IgG2 contains phenylalanine, the F296Y substitution in IgG2 is considered an IgG subclass modification.

[0203] As used herein, "non-naturally occurring modification" means an amino acid modification that is not an isotype. For example, since human IgG does not include serine at position 434, the substitution 434S in IgG1, IgG2, IgG3 or IgG4 (or hybrids thereof) is considered a non-naturally occurring modification.

[0204] As used herein, "amino acid" and "amino acid identity" mean one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0205] As used herein, "effector function" means a biochemical event resulting from the interaction of the Fc region of an antibody with an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

[0206] As used herein, "IgG Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an IgG antibody to form an Fc / Fc ligand complex. Fc ligands include, but are not limited to, FcγRI, FcγRII, FcγRIII, FcRn, C1q, C3, mannose-binding lectin, mannose receptor, staphylococcal protein A, streptococcal protein G, and viral FcγR. Fc ligands also include Fc receptor homologs (FcRH), which are a family of Fc receptors homologous to FcγR (Davis et al., 2002, Immunological Reviews 190:123-136, which is incorporated herein by reference in its entirety). Fc ligands can include undiscovered molecules that bind Fc. Specific IgG Fc ligands are FcRn and Fcγ receptors. As used herein, "Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an antibody to form an Fc / Fc ligand complex.

[0207] As used herein, "Fcγ receptor" or "FcγR" or "FcgammaR" means any member of the family of proteins that bind to the Fc region of IgG antibodies and are encoded by FcγR genes. In humans, this family includes, but is not limited to, FcγRI (CD64), including isotypes FcγRIa, FcγRIb, and FcγRIc; FcγRII (CD32), including isotypes FcγRIIa (including allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and FcγRIIc; and FcγRIII (CD16), including isotypes FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIIb-NA1 and FcγRIIIb-NA2) (Jefferis et al., 2002, Immunol Lett 82:57-65, which is incorporated herein by reference in its entirety), as well as any undiscovered human FcγR or FcγR isotype or allotype. FcγR can be 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 isotype or allotype.

[0208] As used herein, "FcRn" or "neonatal Fc receptor" means a protein that binds to the Fc region of IgG antibodies and is at least partially encoded by the FcRn gene. FcRn can be from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, functional FcRn protein includes two peptides, commonly referred to as the heavy chain and the light chain. The light chain is β-2 microglobulin and the heavy chain is encoded by the FcRn gene. Unless otherwise indicated herein, FcRn or FcRn protein refers to the complex of the FcRn heavy chain with β-2-microglobulin. A variety of FcRn variants are used to enhance binding to the FcRn receptor and, in some cases, increase serum half-life. "FcRn variants" enhance binding to the FcRn receptor, and suitable FcRn variants are shown below.

[0209] As used herein, "parent polypeptide" means the starting polypeptide that is subsequently modified to produce a variant. The parent polypeptide can 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 known commercially available recombinantly produced antibodies as outlined below. In this context, "parent Fc domain" will be relative to the recited variants; thus, comparing the "variant human IgG1 Fc domain" to the parent Fc domain of human IgG1, comparing the "variant human IgG4 Fc domain" to the parent Fc domain of human IgG4, etc.

[0210] As used herein, "position" means a position in a protein sequence. Positions can be numbered sequentially or numbered according to a defined format (such as the EU index used for antibody numbering).

[0211] As used herein, "target antigen" means a molecule that is specifically bound by an antigen-binding domain that includes the variable region of a given antibody.

[0212] The background of the monomers of the bispecific antibodies described herein by "strandedness" means similar to two DNA strands that "match", incorporating heterodimerization variants into each monomer to retain the ability to "match" to form a heterodimer. For example, if some pI variants are engineered into monomer A (such that the pI is higher), then spatial variants that can also be used as "charge pairs" will not interfere with the pI variants, e.g., placing charge variants that make the pI higher on the same "strand" or "monomer" to retain both functions. Similarly, for the "skewed" variants that occur in pairs as outlined more fully below, one of ordinary skill in the art considers the pI when determining which strand or monomer of the pair will receive which, such that the pI of the skewed variants used also maximizes the separation of the pIs.

[0213] As used herein, "target cell" means a cell that expresses a target antigen.

[0214] In the context of producing a bispecific antibody according to the present disclosure, "host cell" means a cell that contains exogenous nucleic acid encoding components of the bispecific antibody and is capable of expressing the bispecific antibody under suitable conditions. Suitable host cells are discussed below.

[0215] "Wild type or WT" as used herein means an amino acid sequence or nucleotide sequence found in nature, including allelic variants. A WT protein has an amino acid sequence or nucleotide sequence that has not been intentionally modified.

[0216] The present disclosure provides multiple antibody domains having sequence identity to human antibody domains. The sequence identity between two similar sequences (e.g., antibody variable domains) can be measured by algorithms such as the following algorithms: such as Smith, T.F. and Waterman, M.S. (1981) "Comparison Of Biosequences", Adv. Appl. Math. 2:482 [local homology algorithm]; Needleman, S.B. and Wunsch, C.D. (1970) "A General Method Applicable To The Search For Similarities In The Amino Acid Sequence Of Two Proteins", J. Mol. Biol. 48:443 [homologous alignment algorithm], Pearson, W.R. and Lipman, D.J. (1988) "Improved Tools For Biological Sequence Comparison", U.S.A. 85:2444 [similarity search method]; or Altschul, S.F. et al., (1990) "Basic Local Alignment Search Tool", J. Mol. Biol. 215:403-10, "BLAST" algorithm, see https: / / blast.ncbi.nlm.nih.gov / Blast.cgi When using any of the foregoing algorithms, default parameters (for window length, gap penalty, etc.) are used. In one embodiment, sequence identity is accomplished using the BLAST algorithm with default parameters.

[0217] The antibodies described herein are typically isolated or recombinant. When 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 expressing the polypeptide. Typically, an isolated polypeptide will be prepared by at least one purification step. An "isolated antibody" means an antibody that is substantially free of other antibodies having different antigen specificities. "Recombinant" means that an antibody is produced using recombinant nucleic acid techniques in an exogenous host cell, and the antibody can also be isolated.

[0218] "Specifically binds" or "binds specifically" or "is specific for" a particular antigen or epitope means binding that is measurably different from non-specific interactions. Specific binding can be measured, for example, by determining the binding of one molecule relative to the binding of a control molecule, which is typically a structurally similar molecule that does not have binding activity. For example, specific binding can be determined by competition with a control molecule that is similar to the target.

[0219] Specific binding to a particular antigen or epitope can be demonstrated, for example, by an antibody having a KD for the antigen or epitope of 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, or at least about 10 -10 M, at least about 10 -11 M, at least about 10 -12 M or greater, where KD refers to the dissociation rate of a particular antibody-antigen interaction. Typically, the KD of an antibody that specifically binds an antigen is 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold, or more than the KD of a control molecule relative to the antigen or epitope.

[0220] Furthermore, specific binding to a particular antigen or epitope can be manifested, for example, by an antibody having a KA or Ka for the antigen or epitope that is at least 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold, or more than that for an epitope relative to a control, where KA or Ka refers to the association rate of a particular antibody-antigen interaction. Binding affinity is typically measured using Biacore, SPR, or BLI assays.

[0221] IV. ENPP3 Binding Domain

[0222] In one aspect, the present disclosure provides ENPP3 antigen-binding domains (ABDs) and compositions comprising such ENPP3 antigen-binding domains (ABDs), including anti-ENPP3 antibodies. The subject antibodies comprising such ENPP3 antigen-binding domains (e.g., anti-ENPP3 x anti-CD3 bispecific antibodies) advantageously elicit a range of different immune responses (see Examples 5 and 6). Such ENPP3 binding domains and related antibodies can be used, for example, to treat ENPP3-related cancers.

[0223] As will be understood by those skilled in the art, suitable ENPP3 binding domains can include those as set forth in the Sequence Listing and Figure 12 ,Figures 13A to 13B and Figures 14A to 14I one of the sets of 6 CDRs depicted in Figures 14A to 14I or as underlined in the CDR, or in the case of using a different numbering scheme as described herein and shown in Table 2, using those identified by other alignments within the variable heavy (VH) chain domain and variable light domain (VL) sequences of the sequences depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I and other CDRs identified by alignment within the sequences depicted in Figures 14A to 14I (see Table 2). Suitable ENPP3 ABDs may also include the entire VH and VL sequences used as scFv or Fab domains as depicted in these sequences and the figures.

[0224] In one embodiment, the ENPP3 antigen-binding domain comprises the 6 CDRs of the ENPP3 ABD described herein, including the figures and sequence listing (i.e., vhCDR1-3 and vlCDR1-3). In an exemplary embodiment, the ENPP3 ABD is one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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, H16-9.65, Ha1-1(3,5)19 and Ha16-1.80( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ).

[0225] In addition to the parental CDR sets forming the ABD against ENPP3 disclosed in the figures and sequence listing, provided herein are ENPP3 ABD variants having CDRs that include at least one modification of the ENPP3 ABD CDRs disclosed herein. In one embodiment, compared to the 6 CDRs of the ENPP3 ABD depicted herein, including in the figures and sequence listing, the ENPP3 ABD comprises a set of 6 CDRs having 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications. In an illustrative embodiment, compared to the 6 CDRs of one ENPP3 ABD among the following ENPP3 ABDs, the ENPP3 ABD comprises a set of 6 CDRs having 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid modifications: 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ). In certain embodiments, the variant ENPP3 ABD is capable of binding to the ENPP3 antigen, as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., octet assay), the latter assay being particularly useful in many embodiments. In a particular embodiment, the ENPP3 ABD is capable of binding to the human ENPP3 antigen (see Example 5).

[0226] In one embodiment, the ENPP3 ABD comprises six CDRs having at least 90%, 95%, 97%, 98% or 99% identity to the six CDRs of the ENPP3 ABD as described herein, including the accompanying drawings and sequence listing. In an illustrative embodiment, the ENPP3 ABD comprises six CDRs having at least 90%, 95%, 97%, 98% or 99% identity to the six CDRs of one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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, H16-9.65, Ha1-1(3,5)19 and Ha16-1.80( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ). In certain embodiments, as determined by at least one of Biacore, surface plasmon resonance (SPR) and / or BLI (Biolayer Interferometry, e.g., octet assays), the ENPP3 ABD is capable of binding TIM-3, the latter assay being particularly useful in many embodiments. In a particular embodiment, the ENPP3 ABD is capable of binding to the human ENPP3 antigen (see Figure 2 ).

[0227] In another illustrative embodiment, the ENPP3 ABD includes any one of the ENPP3 ABDs described herein, including the accompanying drawings and sequence listing, the variable heavy chain (VH) domain and variable light chain (VL) domain of the ENPP3 ABD. In an illustrative embodiment, the ENPP3 ABD is one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8L1.33, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ).

[0228] In addition to the parental ENPP3 variable heavy and variable light domains disclosed herein, provided herein are ENPP3 ABDs comprising a variable heavy chain domain and / or a variable light domain, wherein the variable heavy chain domain and / or variable light domain are variants 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 compared to the VH and / or VL domain of the ENPP3 ABD described herein, including in the figures and sequence listing. In an illustrative embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes compared to the VH and / or VL domain of one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, 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( Figure 12 , Figures 13A to 13B , Figures 14A to 14I ). In certain embodiments, the ENPP3 ABD is capable of binding the ENPP3 as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, such as octet assays), with the latter assay being particularly useful in many embodiments. In a particular embodiment, the ENPP3 ABD is capable of binding the human ENPP3 antigen (see Example 5).

[0229] In one embodiment, the variant VH and / or VL domain is at least 90%, 95%, 97%, 98% or 99% identical to the VH and / or VL of the ENPP3 ABD as described herein, including in the figures and sequence listing. In an illustrative embodiment, the variant VH and / or VL domain is at least 90%, 95%, 97%, 98% or 99% identical to the VH and / or VL of one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ). In certain embodiments, the ENPP3 ABD is capable of binding the ENPP3 as determined by at least one of Biacore, surface plasmon resonance (SPR) and / or BLI (Biolayer Interferometry, such as octet assay), the latter assay being particularly useful in many embodiments. In a particular embodiment, the ENPP3 ABD is capable of binding the human ENPP3 antigen (see Example 5).

[0230] V. Antibodies

[0231] In one aspect, the present disclosure provides antibodies that bind ENPP3 (e.g., anti-ENPP3 antibodies). In certain embodiments, the antibodies bind to human ENPP3( Figure 11A ). The subject anti-ENPP3 antibodies include monospecific ENPP3 antibodies, as well as multispecific (e.g., bispecific) anti-ENPP3 antibodies. In certain embodiments, the anti-ENPP3 antibodies have a form of any of the antibody forms depicted in Figure 15A , Figure 15B and Figures 52A to 52K .

[0232] In some embodiments, the subject composition comprises an ENPP3 binding domain. In some embodiments, the composition comprises an antibody having an ENPP3 binding domain. The antibodies provided herein comprise one, two, three, four, five, and more than five ENPP3 binding domains. In certain embodiments, the ENPP3 binding domain comprises any one of the vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3 sequences of the ENPP3 binding domains depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In some embodiments, the ENPP3 binding domain comprises any one of the underlined vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3 sequences of the ENPP3 binding domains depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In some embodiments, the ENPP3 binding domain comprises the variable heavy chain domain and variable light domain of the ENPP3 binding domains depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . Figure 12 , Figures 13A to 13B and Figures 14A to 14I The ENPP3 binding domains described in include AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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.

[0233] In one aspect, the present disclosure provides bispecific antibodies that bind ENPP3 and CD3 in the various forms described below and are generally depicted inFigure 15A and Figure 15B . These bispecific heterodimeric antibodies include an ENPP3 binding domain. In certain embodiments, the ENPP3 binding domain comprises VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3 binding domain selected from the group consisting of those described in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In some embodiments, the ENPP3 binding domain comprises underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3 binding domain selected from those depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I .

[0234] These bispecific heterodimeric antibodies bind ENPP3 and CD3. Such antibodies comprise a CD3 binding domain and at least one ENPP3 binding domain. Any suitable ENPP3 binding domain can be included in an anti-ENPP3 X anti-CD3 bispecific antibody. In some embodiments, the anti-ENPP3 X anti-CD3 bispecific antibody comprises one, two, three, four, or more ENPP3 binding domains, including but not limited to Figure 12 , Figures 13A to 13B and Figures 14A to 14I described in. In certain embodiments, the anti-ENPP3 X anti-CD3 antibody comprises an ENPP3 binding domain comprising VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3 binding domain selected from the group consisting of those described in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In some embodiments, the anti-ENPP3 X anti-CD3 antibody comprises an ENPP3 binding domain comprising underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3 binding domain selected from the group consisting of those described in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In some embodiments, the anti-ENPP3 X anti-CD3 antibody comprises an ENPP3 binding domain comprising VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of an ENPP3 binding domain selected from the group consisting of those described in Figure 12 , Figures 13A to 13B and Figures 14A to 14IThe variable heavy chain domain and variable light chain domain of the ENPP3 binding domain of the group consisting of the ENPP3 binding domains depicted in []. In an exemplary embodiment, the anti-ENPP3 X anti-CD3 antibody comprises an anti-ENPP3 AN1[ENPP3]_H1L1 binding domain.

[0235] The anti-ENPP3 x anti-CD3 antibodies provided herein can comprise any suitable CD3 binding domain. In certain embodiments, the anti-ENPP3 X anti-CD3 antibody comprises a CD3 binding domain comprising a selection from the group consisting of Figures 10A to 10F the VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of the CD3 binding domains of the group consisting of those described in []. In some embodiments, the anti-ENPP3 X anti-CD3 antibody comprises a CD3 binding domain comprising a selection from the group consisting of Figures 10A to 10F the underlined VHCDR1, VHCDR2, VHCDR3, VLCDR1, VLCDR2, and VLCDR3 sequences of the CD3 binding domains of the group consisting of those described in []. In some embodiments, the anti-ENPP3 X anti-CD3 antibody comprises a CD3 binding domain comprising a selection from the group consisting of Figures 10A to 10F the variable heavy chain domain and variable light chain domain of the CD3 binding domains of the group consisting of the CD3 binding domains depicted in []. In some embodiments, the CD3 binding domain is selected from anti-CD3 H1.30_L1.47, anti-CD3 H1.32_L1.47; anti-CD3 H1.89_L1.48; anti-CD3 H1.90_L1.47; anti-CD3 H1.33_L1.47; and anti-CD3 H1.31_L1.47. As described herein, these anti-CD3 antigen binding domains (CD3-ABD) can be used in scFv form in either orientation (e.g., VH-scFv linker-VL or VL-scFv linker-VH from the N-terminus to the C-terminus).

[0236] The antibodies provided herein comprise different antibody domains. As described herein and known in the art, the antibodies described herein comprise different domains within the heavy and light chains, which domains can also be overlapping. These domains include, but are not limited to, the Fc domain, CH1 domain, CH2 domain, CH3 domain, hinge domain, heavy chain constant domain (CH1-hinge-Fc domain or CH1-hinge-CH2-CH3), variable heavy chain domain, variable light chain domain, light chain constant domain, FAb domain, and scFv domain.

[0237] As indicated herein, there are a number of suitable linkers (for use as domain linkers or scFv linkers) that can be used to covalently link the enumerated domains (e.g., scFv, Fab, Fc domain, etc.), including traditional peptide bonds produced by recombinant techniques. Figure 6 An exemplary linker for attaching the domains of the subject antibodies to each other is depicted in . In some embodiments, the linker peptide may primarily comprise the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should have a length sufficient to connect the two molecules so that they present the correct conformation relative to each other so that they retain the desired activity. In one embodiment, the linker has a length of about 1-50 amino acids, preferably a length of about 1-30 amino acids. In one embodiment, a linker with a length of 1-20 amino acids can be used, and in some embodiments, about 5 to about 10 amino acids can be used. Useful linkers include glycine-serine polymers (including, for example, (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer of at least one (and typically 3 to 4)), glycine-alanine polymers, alanine-serine polymers, and other flexible linkers, some of which are in Figure 5 and Figure 6 Alternatively, a variety of non-protein polymers including, but not limited to, polyethylene glycol (PEG), polypropylene glycol, polyalkylene oxide, or copolymers of polyethylene glycol and polypropylene glycol may be used as linkers.

[0238] Other linker sequences can include any sequence of any length of the CL / CH1 domain, but do not include all the residues of the CL / CH1 domain; for example, the first 5-12 amino acid residues of the CL / CH1 domain. The linker can be derived from an immunoglobulin light chain, such as Cκ or Cλ. The linker can be derived from an immunoglobulin heavy chain of any isotype, including, for example, Cγ1, Cγ2, Cγ3, Cγ4, Cα1, Cα2, Cδ, Cε and Cμ. The linker sequence can also be derived from other proteins, such as Ig-like proteins (e.g., TCR, FcR, KIR), hinge region source sequences, and other natural sequences from other proteins.

[0239] In some embodiments, the linker is a "domain linker" that is used to connect any two domains as outlined herein. For example, in Figure 15BThere can be a domain linker that connects the C-terminus of the CH1 domain of the Fab to the N-terminus of the scFv, and another optional domain linker that connects the C-terminus of the scFv to the CH2 domain (but in many embodiments, the hinge serves as this domain linker). While any suitable linker can be used, many embodiments utilize glycine-serine polymers as domain linkers, including for example (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer of at least one (and typically 3 to 4 to 5), and any peptide sequence that allows for recombinant ligation of the two domains and has sufficient length and flexibility such that each domain retains its biological function. In some cases, and noting the "chain type", as outlined below, charged domain linkers such as those used in some embodiments of scFv linkers can be used. Figure 6 Exemplary useful domain linkers are depicted.

[0240] Specifically referring to the domain linkers for attaching the scFv domain to the Fc domain in the "2+1" form, there are several domain linkers with special uses, including the "full hinge C220S variant", "flexible half hinge", "charged half hinge 1", and "charged half hinge 2" as shown in Figure 6 Figure.

[0241] In some embodiments, the linker is an "scFv linker" that is used to covalently attach the VH and VL domains as discussed herein. In many cases, the scFv linker is a charged scFv linker, and multiple such charged scFv linkers are shown in Figure 5 Figure. Thus, in some embodiments, the antibodies described herein also provide charged scFv linkers to facilitate pI separation between the first monomer and the second monomer. That is, by incorporating scFv linkers that are positively or negatively (or both, in the case of using the backbone of the scFv on different monomers) charged, the monomers containing the charged linkers are thereby allowed to change the pI without causing further changes in the Fc domain. These charged linkers can be substituted into any scFv containing a standard linker. Again, as will be understood by those skilled in the art, the charged scFv linker is used on the correct "chain" or monomer depending on the desired pI change. For example, as discussed herein, to form a heterodimeric antibody in the 1+1 Fab-scFv-Fc form, the original pI of the Fv region of each desired antigen-binding domain is calculated, and one is selected to form the scFv, and a positive or negative linker is selected based on the pI.

[0242] Charged domain linkers can also be used to increase the pI separation of the monomers of the antibodies described herein, and thus those included in Figure 5 Figure can be used in any embodiment that utilizes linkers herein.

[0243] Specifically, Figure 15A and 15B the form depicted is an antibody, commonly referred to as a "heterodimeric antibody", meaning that the protein has at least two related Fc sequences and at least two Fv regions that self-assemble into a heterodimeric Fc domain, whether as a Fab or as an scFv.

[0244] The provided ENPP3 binding domain can be included in any useful antibody form, including, for example, canonical immunoglobulins, as well as the 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv forms provided herein. Other useful antibody forms include, but are not limited to, "mAb-Fv", "mAb-scFv", "central-Fv", "single-arm scFv-mAb", "scFv-mAb", "dual scFv", and "trifunctional" forms of antibodies, as Figures 52A to 52K disclosed in

[0245] In some embodiments, the subject 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 an illustrative embodiment, the ENPP3 ABD comprises the variable heavy chain domain and variable light chain domain of one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I)。In some embodiments, ENPO3 ABD is one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, 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 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I )。

[0246] In an exemplary embodiment, the antibody is a bispecific antibody that comprises one or two ENPP3 ABDs, including any of the ENPP3 ABDs provided herein. Bispecific antibodies that comprise 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ). In an exemplary embodiment, the ENPP3 binding domain is a Fab. In some embodiments, such bispecific antibodies are heterodimeric bispecific antibodies that comprise any of the heterodimerization bias variants, pI variants, and / or ablation variants described herein.

[0247] A. Chimeric and Humanized Antibodies

[0248] In certain embodiments, the antibodies described herein comprise a heavy chain variable region from a particular germline heavy chain immunoglobulin gene and / or a light chain variable region from a particular germline light chain immunoglobulin gene. For example, such antibodies can comprise or consist of human antibodies that comprise a heavy or light chain variable region that is a “product” or “derived from” a particular germline sequence. A human antibody that is a “product” or “derived from” a human germline immunoglobulin sequence can be identified by comparing the amino acid sequence of the human antibody to the amino acid sequences of human germline immunoglobulins and selecting the human germline immunoglobulin sequence that is most similar in sequence to the sequence of the human antibody (i.e., greatest % identity) (using the methods outlined herein). Due to, for example, naturally occurring somatic mutations or intentionally introduced site-directed mutations, a human antibody that is a “product” or “derived from” a particular human germline immunoglobulin sequence can contain amino acid differences compared to the germline sequence. However, a humanized antibody typically has at least 90% amino acid sequence identity with the amino acid sequence encoded by a human germline immunoglobulin gene and, when compared to the germline immunoglobulin amino acid sequences of other species (e.g., murine germline sequences), contains amino acid residues that identify the antibody as being derived from a human sequence. In some cases, a humanized antibody can have at least 95%, 96%, 97%, 98%, or 99%, or even at least 96%, 97%, 98%, or 99% amino acid sequence identity with the amino acid sequence encoded by a germline immunoglobulin gene. Generally, a humanized antibody derived from a particular human germline sequence will show no more than 10 - 20 amino acid differences from the amino acid sequence encoded by a human germline immunoglobulin gene (before any skewing variants, pI variants, and ablation variants introduced herein; i.e., before the variants described herein are introduced, the number of variants is typically low). In some cases, compared to the amino acid sequence encoded by a germline immunoglobulin gene, a humanized antibody can show no more than 5 or even no more than 4, 3, 2, or 1 amino acid difference (again, before any skewing, pI, and ablation variants are introduced herein; i.e., before the variants described herein are introduced, the number of variants is typically low).

[0249] In one embodiment, the parental antibody has been affinity matured as is known in the art. Structure-based methods can be used for humanization and affinity maturation, such as those described in USSN 11 / 004,590. Selection-based methods can be used to humanize and / or affinity mature antibody variable regions, including but not limited to those described in the following references: Wu et al., 1999, Journal of Molecular Biology 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, which are incorporated herein by reference in their entirety. Other humanization methods can involve transplanting only portions of the CDRs, including but not limited to those described in the following references: USSN 09 / 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 herein by reference in their entirety.

[0250] B. Bispecific Antibodies

[0251] In an exemplary embodiment, the bispecific antibodies provided herein are heterodimeric bispecific antibodies comprising two variant Fc domain sequences. Such variant Fc domains contain amino acid modifications to facilitate self-assembly and / or purification of the heterodimeric antibody.

[0252] One problem in antibody technology is the desire for "bispecific" antibodies that bind two different antigens simultaneously, typically thereby allowing the different antigens to come into proximity and thus generate new functions and new therapies. Typically, these antibodies are produced by incorporating the genes for each heavy and light chain into a host cell. This typically results in the formation of the desired heterodimer (A-B) as well as two homodimers (A-A and B-B (excluding the light chain heterodimer problem)). However, the major obstacle in forming bispecific antibodies is the difficulty in biasing the formation of the desired heterodimeric antibody towards the formation of homodimers and / or the difficulty in purifying the heterodimeric antibody from the homodimeric antibodies.

[0253] There are multiple mechanisms that can be used to generate the heterodimeric antibodies of the present invention. Additionally, 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 generally collectively referred to as "heterodimerization variants". As discussed below, heterodimerization variants include "skew" variants (e.g., the "boss and hole" and "charge pair" variants described below) and "pI variants", which allow for the purification of heterodimers from homodimers. As generally described in U.S. Patent No. US 9,605,084, the entire content of which is incorporated herein by reference, and particularly as discussed below for heterodimerization variants, useful mechanisms for heterodimerization include the "knobs-into-holes" ("KIH") described in U.S. Patent US 9,605,084, the "electrostatic steering" or "charge pair" described in U.S. Patent US 9,605,084, the pI variants described in U.S. Patent US 9,605,084, and generally other Fc variants outlined in U.S. Patent US 9,605,084 and below.

[0254] Heterodimerization variants that can be used to form and purify the heterodimeric antibodies (e.g., bispecific antibodies) of the present invention are discussed further in detail below.

[0255] 1. Skew variants

[0256] In some embodiments, the heterodimeric antibody comprises a skew variant, which is one or more amino acid modifications in the first Fc domain (A) and / or the second Fc domain (B), and the one or more amino acid modifications facilitate the formation of an Fc heterodimer (an Fc dimer comprising the first Fc domain and the second Fc domain); (A - B) over Fc homodimers (Fc dimers comprising two in the first Fc domain or two in the second Fc domain; A - A or B - B). Suitable skew variants are included in Figure 29 of U.S. Patent Application Publication No. 2016 / 0355608, which is hereby incorporated by reference in its entirety and specifically for its disclosure of skew variants, and in Figures 1A to 1E and Figure 4 in.

[0257] A mechanism commonly known in the art as "the pestle and mortar" can also optionally use so-called amino acid engineering to create steric effects that favor heterodimer formation over homodimer formation; this is sometimes referred to as "the pestle and mortar" as described in the following documents: USSN 61 / 596,846; Ridgway et al., Protein Engineering 9(7):617 (1996); Atwell et al., Journal of Molecular Biology 1997 270:26; U.S. Patent No. 8,216,805, all of which are incorporated herein by reference in their entirety. The figure shows the number of "monomer A - monomer B" pairs that rely on "the pestle and mortar". Additionally, as described in Merchant et al., Nature Biotech. 16:677 (1998), these "pestle and mortar" mutations can be combined with disulfide bonds to bias heterodimerization.

[0258] Another mechanism for generating heterodimeric antibodies is sometimes referred to as "electrostatic steering", as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637 (2010), which is incorporated herein by reference in its entirety. This is sometimes referred to as "charge pairs" in this document. In this embodiment, electrostatics is used to bias formation towards heterodimerization. As will be understood by those skilled in the art, these charge pairs may also affect the pI and thus purification, and can therefore also be considered pI variants in some cases. However, since these are generated to enforce heterodimerization and not used as a purification tool, they are classified as "steric variants". These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (e.g., these are "monomer corresponding groups") and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0259] In some embodiments, the bias variants advantageously and simultaneously favor heterodimerization based on both the "pestle and mortar" mechanism and the "electrostatic manipulation" mechanism. In some embodiments, the heterodimeric antibody comprises one or more sets of such heterodimerization bias variants. These variants occur in "sets" of "pairs". That is, one set of pairs is incorporated into the first monomer and another set of pairs is incorporated into the second monomer. It should be noted that these sets do not necessarily manifest as "bump - into - hole" variants where there is a one - to - one correspondence between residues on one monomer and residues on the other monomer. That is, these paired sets can instead form an interface between the two monomers that encourages heterodimer formation and discourages homodimer formation, thereby allowing the percentage of heterodimers formed spontaneously under biological conditions to exceed 90%, rather than the expected 50% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B).Figure 4 Exemplary heterodimerizing “skewed” variants are depicted. In an illustrative embodiment, the heterodimeric antibody comprises 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 including bridging disulfide bonds, T366S / L368A / Y407V / Y349C:T366W / S354C) “skewed” variant amino acid substitution sets. In an exemplary embodiment, the heterodimeric antibody comprises the “S364K / E357Q:L368D / K370S” amino acid substitution set. In terms of nomenclature, the pair “S364K / E357Q:L368D / K370S” means that one monomer comprises an Fc domain having amino acid substitutions S364K and E357Q, and the other monomer comprises an Fc domain having amino acid substitutions L368D and K370S; as described above, the “chainness” of these pairs depends on the starting pI.

[0260] In some embodiments, the skewed variants provided herein can optionally and independently be combined, in one or both of the first and second Fc domains of the heterodimeric antibody, with any other modification—including but not limited to other skewed variants (see, e.g., FIG. 37 of U.S. Published Application No. 2012 / 0149876, which is incorporated herein by reference, particularly for its disclosure of skewed variants), pI variants, isotype variants, FcRn variants, ablation variants, etc. Additionally, a single modification can also be independently and optionally included in or excluded from the subject heterodimeric antibody.

[0261] Additional monomer A and monomer B variants can optionally and independently be combined in any amount with other variants (such as the pI variants outlined herein or the other steric variants shown in FIG. 37 of US2012 / 0149876, the figures and legends and SEQ ID NOs of which are hereby incorporated herein by reference).

[0262] In some embodiments, the steric variants outlined herein can optionally and independently be incorporated into one or both monomers with any pI variant (or other variant, such as an Fc variant, an FcRn variant, etc.), and can be independently and optionally included in or excluded from the proteins of the antibodies described herein.

[0263] A list of suitable skewed variants is found in Figures 1A to 1E whereFigure 4 Shows the specific uses of some pairs in many embodiments. In multiple embodiments, pairs that are particularly applicable include, but are not limited to, the following groups: S364K / E357Q:L368D / K370S; L368D / K370S:S364K; L368E / K370S:S364K; T411T / E360E / Q362E:D401K; L368D / K370S:S364K / E357L, and K370S:S364K / E357Q. In terms of nomenclature, the pair "S364K / E357Q:L368D / K370S" means that one of the monomers has the double variant group S364K / E357Q and the other has the double variant group L368D / K370S.

[0264] 2. pI (isoelectric point) variants for heterodimers

[0265] In some embodiments, the heterodimeric antibody comprises purification variants that advantageously allow the separation of the heterodimeric antibody (e.g., an anti-ENPP3 x anti-CD3 bispecific antibody) from the homodimeric protein.

[0266] There are several basic mechanisms that can lead to the easy purification of heterodimerizing antibodies. For example, modification of one or both of the heavy chain monomers A and B of the antibody such that each monomer has a different pI allows the isoelectric purification of the heterodimeric A-B antibody from the monomer A-A and B-B proteins. Alternatively, some scaffold forms, such as the "1+1Fab-scFv-Fc" and "2+1Fab2-scFv-Fc" forms, also allow separation based on size. As described herein, it is also possible to use skewing variants to "skew" the formation of the heterodimer relative to the homodimer. Thus, a combination of heterodimerizing skewing variants and pI variants is particularly useful in the heterodimeric antibodies provided herein.

[0267] In addition, as outlined more fully below, depending on the form of the heterodimeric antibody, pI variants and / or domain linkers contained within the constant region and / or Fc domain of the monomer can be used. In some embodiments, the heterodimeric antibody comprises additional modifications for alternative functions that can also produce pI changes, such as Fc, FcRn, and KO variants.

[0268] In some embodiments, the bispecific antibodies of the invention provided herein comprise at least one monomer having one or more modifications that alter the pI of the monomer (i.e., "pI variants"). Generally, as will be understood by those skilled in the art, there are two general classes of pI variants: variants that increase the pI of the protein (basic changes) and variants that decrease the pI of the protein (acidic changes). As described herein, all combinations of these variants can be made: one monomer can be wild-type or a variant that does not exhibit a pI significantly different from wild-type, and the other can be more basic or more acidic. Alternatively, each monomer can be altered, one being more basic and one being more acidic.

[0269] Depending on the format of the bispecific antibody, the pI variants can be included within the constant and / or Fc domain of the monomer, or a charged linker, domain linker, or scFv linker can be used. That is, an antibody format using one or more scFvs such as "1+1Fab-scFv-Fc" can include a charged scFv linker (positively or negatively charged) that gives an additional pI boost for purification purposes. As will be understood by those skilled in the art, some 1+1Fab-scFv-Fc formats can be used with only a charged scFv linker without additional pI adjustment, but the antibodies described herein do also provide pI variants and / or charged domain linkers on one or both of the monomers. Additionally, additional amino acid engineering for alternative functions can also confer pI changes, such as Fc, FcRn, and KO variants.

[0270] In the bispecific antibodies of the invention that use pI as a separation mechanism to allow purification of the heterodimeric protein, amino acid variants are introduced into one or both of the monomer polypeptides. That is, the pI of one of the monomers (referred to herein as "monomer A" for simplicity) can be engineered to be away from monomer B, or both monomer A and monomer B can be altered with an increase in the pI of monomer A and a decrease in the pI of monomer B. As outlined more fully below, the pI change of either or both monomers can be accomplished by: removing or adding charged residues (e.g., replacing a neutral amino acid with a positively or negatively charged amino acid residue, such as glycine to glutamate), changing a charged residue from positively or negatively charged to the opposite charge (e.g., aspartic acid to lysine), or changing a charged residue to a neutral residue (e.g., loss of charge; lysine to serine). Figure 3 and Figure 4 A variety of these variants are shown in

[0271] Thus, in some embodiments, the bispecific antibodies of the invention comprise amino acid modifications in the constant regions that change the isoelectric point (pI) of at least one (if not both) of the monomers of the dimeric protein by incorporating amino acid substitutions (“pI variants” or “pI substitutions”) into one or both of the monomers to form “pI antibodies”. As shown herein, separation of the heterodimer from the two homodimers can be achieved if the pI difference between the two monomers is as small as 0.1 pH unit, where 0.2, 0.3, 0.4, and 0.5 or higher can all be used for the antibodies described herein.

[0272] As will be appreciated by those skilled in the art, the number of pI variants to be incorporated on each or both monomers for 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, the starting pI of the scFv and Fab of interest. That is, to determine which monomer to engineer or in which “direction” (e.g., more positive or more negative), the Fv sequences of the two target antigens are calculated and a decision is made accordingly. As is known in the art, different Fvs will have different starting pIs for the antibodies described herein. Generally, as outlined herein, the pI is engineered such that the total pI difference between the individual monomers is at least about 0.1 log, and preferably 0.2 to 0.5 as outlined herein.

[0273] In cases where pI variants are used to effect heterodimerization, a more modular approach for designing and purifying bispecific proteins, including antibodies, is provided by using one or more constant regions of one or more heavy chains. Thus, in some embodiments, the heterodimerization variants (including skewed and pI heterodimerization variants) are not included in the variable regions such that each individual antibody has to be engineered. Additionally, in some embodiments, the likelihood that the pI variants will result in immunogenicity is significantly reduced by introducing pI variants from different IgG isotypes to change the pI without introducing significant immunogenicity. Thus, another problem to be solved is to elucidate low pI constant domains with high human sequence content, e.g., minimizing or avoiding non-human residues at any given position. Optionally or in addition to isotype substitutions, the likelihood that the pI variants will result in immunogenicity is significantly reduced by utilizing isosteric substitutions (e.g., Asn to Asp; and Gln to Glu).

[0274] As discussed below, negative effects that may occur in conjunction with this pI engineering are also increased serum half-life and increased FcRn binding. That is, as described in U.S. Published Application No. US2012 / 0028304 (incorporated herein by reference in its entirety), decreasing the pI of the antibody constant domains (including those found in antibody and Fc fusions) may result in the serum remaining in the body for a longer time. These pI variants with increased serum half-life also facilitate pI changes for purification.

[0275] In addition, it should be noted that pI variants provide additional benefits to the analytical and quality control processes of bispecific antibodies, as the ability to eliminate, minimize, and distinguish in the presence of homodimers is significant. Similarly, the ability to reliably test the reproducibility of heterodimer antibody production is important.

[0276] Generally, embodiments for a particular use rely on a group of variants that include skewed variants that, in combination with pI variants, facilitate heterodimerization formation as compared to homodimerization formation, where the pI variants increase the pI difference between the two monomers to facilitate purification of the heterodimer from the homodimer.

[0277] Exemplary combinations of pI variants are shown in U.S. Published Application No. 2016 / 0355608 Figure 4 and Figure 5 as well as Figure 30 and all such documents are incorporated herein by reference in their entirety, particularly the disclosures regarding pI variants. Preferred combinations of pI variants are shown in FIGS. 1 and Figure 2 and as outlined herein and shown in the figures, these changes are shown relative to IgG1, but all isotypes can be altered in this way, as well as isotype hybrids. In the case where the heavy chain constant domain is from IgG2-4, R133E and R133Q can also be used.

[0278] In one embodiment, a preferred combination of pI variants has one monomer (negative Fab side) that includes the 208D / 295E / 384D / 418E / 421D variant (N208D / Q295E / N384D / Q418E / N421D when relative to human IgG1) and a second monomer (positive scFv side) that includes a positively charged scFv linker, comprising (GKPGS)4 (SEQ ID NO: XX). However, as will be understood by those skilled in the art, the first monomer includes the CH1 domain, which includes position 208. Thus, in constructs that do not include the CH1 domain (e.g., for antibodies that do not utilize the CH1 domain on one of the domains), the preferred negative pI variant Fc group includes the 295E / 384D / 418E / 421D variation (Q295E / N384D / Q418E / N421D when relative to human IgG1).

[0279] Thus, in some embodiments, one monomer has a set of substitutions from Figure 2 and the other monomer has a charged linker (in the form of a charged scFv linker, since the monomer includes an scFv or a charged domain linker as indicated by the form, and the charged domain linker can be selected from those depicted in Figure 5 .

[0280] In some embodiments, modifications are made in the hinge of the Fc domain, including positions 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, and 230 based on EU numbering. Thus, pI mutations and in particular substitutions can be made in one or more of positions 216 - 230, where 1, 2, 3, 4, or 5 mutations are useful. Again, all possible combinations, either alone or in combination with other pI variants in other domains, are contemplated.

[0281] Specific substitutions useful for reducing the pI of the hinge domain include, but are not limited to, a deletion at position 221, a non - native valine or threonine at position 222, a deletion at position 223, a non - native glutamate at position 224, a deletion at position 225, a deletion at position 235, and a deletion or non - native alanine at position 236. In some cases, pI substitutions are made only in the hinge domain, and in other cases, these substitutions are added in any combination to other pI variants in other domains.

[0282] In some embodiments, mutations can be made in the CH2 region, including positions 233, 234, 235, 236, 274, 296, 300, 309, 320, 322, 326, 327, 334, and 339 based on EU numbering. It should be noted that changes in positions 233 to 236 can be made to increase effector function in the IgG2 backbone (as well as 327A). Again, all possible combinations can be made to these 14 positions; for example, variant Fc domains can include those having 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 CH2 pI substitutions.

[0283] Specific substitutions that can be used to lower the pI of the CH2 domain include, but are not limited to: non-natural glutamine or glutamate at position 274, non-natural phenylalanine at position 296, non-natural phenylalanine at position 300, non-natural valine at position 309, non-natural glutamate at position 320, non-natural glutamate at position 322, non-natural glutamate at position 326, non-natural glycine at position 327, non-natural glutamate at position 334, non-natural threonine at position 339, and all possible combinations within CH2 and with other domains.

[0284] In this embodiment, the modifications can independently and optionally be selected from positions 355, 359, 362, 384, 389, 392, 397, 418, 419, 444, and 447 (EU numbering) in the CH3 region. Specific substitutions that can be used to lower the pI of the CH3 domain include, but are not limited to: non-natural glutamine or glutamate at position 355, non-natural serine at position 384, non-natural asparagine or glutamate at position 392, non-natural methionine at position 397, non-natural glutamate at position 419, non-natural glutamate at position 359, non-natural glutamate at position 362, non-natural glutamate at position 389, non-natural glutamate at position 418, non-natural glutamate at position 444, and deletion or non-natural aspartic acid at position 447.

[0285] Generally, as will be understood by those skilled in the art, there are two general classes of pI variants: variants that increase the protein pI (basic changes) and variants that decrease the protein pI (acidic changes). As described herein, all combinations of these variants can be made: one monomer can be wild-type, or a variant that does not show a significantly different pI from wild-type, and the other can be more basic or more acidic. Alternatively, each monomer can be altered, one having a stronger basicity and one having a stronger acidity.

[0286] Preferred combinations of pI variants are shown in Figure 4 . As outlined herein and shown in the figures, these changes are presented relative to IgG1, but all isotypes can be altered in this manner, as well as isotype heterohybrids. In the case where the heavy chain constant domain is from IgG2-4, R133E and R133Q can also be used.

[0287] In one embodiment, for example in Figure 15A and Figure 15BIn one form, a preferred combination of pI variants has one monomer (negative Fab side) comprising the 208D / 295E / 384D / 418E / 421D variants (N208D / Q295E / N384D / Q418E / N421D relative to human IgG1) and a second monomer (positive scFv side) comprising a positively charged scFv linker, comprising (GKPGS)4 (SEQ ID NO: XXX). However, as will be understood by those skilled in the art, the first monomer comprises a CH1 domain, which comprises position 208. Thus, in constructs that do not contain the CH1 domain (e.g., for antibodies that do not utilize the CH1 domain on one of the domains, such as in the bis-scFv form or "single-arm" form, as depicted in Figure 52B , 52C or the form depicted in 52D), the preferred negative pI variant Fc group comprises the 295E / 384D / 418E / 421D variants (Q295E / N384D / Q418E / N421D relative to human IgG1).

[0288] Thus, in some embodiments, one monomer has a set of substitutions from Figure 4 and the other monomer has a charged linker (in the form of a charged scFv linker, since the monomer comprises an scFv or a charged domain linker as indicated by the form, and the charged domain linker can be selected from those depicted in Figure 5 .

[0289] 3. Isotype variants

[0290] In addition, many embodiments of the antibodies described herein rely on "introducing" pI amino acids from specific positions of one IgG isotype into another IgG isotype, thereby reducing or eliminating the possibility of introducing unwanted immunogenicity in the variant. A variety of these variants are shown in Figure 21 of U.S. Publication 2014 / 0370013, which is hereby incorporated by reference. That is, for a variety of reasons, IgG1 is a commonly used isotype for therapeutic antibodies, comprising high effector functions. However, the heavy chain constant region of IgG1 has a higher pI (8.10 relative to 7.31) than the heavy chain constant region of IgG2. By introducing IgG2 residues at specific positions into the IgG1 backbone, the pI of the resulting monomer is decreased (or increased) and additionally exhibits a longer serum half-life. For example, IgG1 has glycine at position 137 (pI of 5.97), and IgG2 has glutamate (pI of 3.22); introducing glutamate will affect the pI of the resulting protein. As described below, multiple amino acid substitutions are generally required to significantly affect the pI of the variant antibody. However, it should be noted that, as discussed below, homogeneous changes in the IgG2 molecule result in an extended serum half-life.

[0291] In other embodiments, non-isotypic amino acid changes are made to reduce the overall charge state of the resulting protein (e.g., by changing a higher pI amino acid to a lower pI amino acid), or to allow structural adaptation for stability, etc., as described further below.

[0292] Additionally, by pI engineering of the heavy and light chain constant domains, significant changes can be found in the individual monomers in the heterodimer. As discussed herein, having a pI difference of at least 0.5 between the two monomers can allow separation by ion exchange chromatography or isoelectric focusing or other methods sensitive to the isoelectric point.

[0293] 4. Calculate pI

[0294] The pI of each monomer can depend on the pI of the variant heavy chain constant domain and the pI of the total monomer, which includes the variant heavy chain constant domain and the fusion partner. Thus, in some embodiments, the pI change is calculated based on the variant heavy chain constant domain using the chart of FIG. 19 in US Publication 2014 / 0370013: Calculate the pI change based on the variant heavy chain constant domain. As discussed herein, which monomer to engineer is typically determined by the inherent pI of the Fv and the scaffold region. Alternatively, the pI of each monomer can be compared.

[0295] 5. pI Variants That Also Confer Better FcRn In Vivo Binding

[0296] In the case where the pI variant reduces the monomer pI, it can have the additional benefit of improving the in vivo serum residence time.

[0297] Although still under review, it is believed that the Fc region has a long half-life in vivo because binding to FcRn at pH 6 in endosomes sequesters the Fc (Ghetie and Ward, 1997 Immunology Today 18(12):592-598, which is hereby incorporated by reference in its entirety). The endosomal compartment then recycles the Fc to the cell surface. Once the compartment opens to the extracellular space, the higher pH (about 7.4) induces release of the Fc back into the blood. Dall'Acqua et al. showed that in mice, Fc mutant variants with enhanced 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, The Journal of Immunology 169:5171-5180, which is incorporated herein by reference in its entirety). It is thought that the increased affinity of FcRn for Fc at pH 7.4 prevents release of the Fc back into the blood. Thus, Fc mutants that increase the in vivo half-life of the Fc ideally enhance FcRn binding at lower pH while still allowing release of the Fc at higher pH. In the pH range of 6.0 - 7.4, the amino acid histidine changes its charge state. Thus, it is not surprising that His residues are found at important positions in the Fc / FcRn complex.

[0298] Recently, it has been proposed that antibodies with variable regions having a lower isoelectric point may also have a longer serum half-life (Igawa et al., 2010 Protein Engineering, Design & Selection (PEDS) 23(5):385-392, which is incorporated by reference). However, the mechanism is still poorly understood. Moreover, variable regions vary among antibodies. Constant region variants with a reduced pI and an extended half-life would provide a more modular approach to improving the pharmacokinetic properties of antibodies, as described herein.

[0299] C. Additional Fc Variants for Additional Functions

[0300] In addition to the heterodimerization variants discussed above, there are many useful Fc amino acid modifications that can be made for various reasons, including but not limited to altering binding to one or more FcγR receptors, altering binding to the FcRn receptor, etc., as discussed below.

[0301] Thus, the antibodies provided herein (heterodimers as well as homodimers) can include such amino acid modifications with or without the heterodimerization variants outlined herein (e.g., pI variants and spatial variants). Each set of variants can independently and optionally include any particular heterodimeric protein or exclude any particular heterodimeric protein.

[0302] 1. FcγR Variants

[0303] Thus, multiple useful Fc substitutions can be made to alter binding to one or more FcγR receptors. In certain embodiments, the antibodies of the invention comprise modifications that alter binding to one or more FcγR receptors (i.e., “FcγR variants”). Substitutions that result in increased binding as well as decreased binding can be useful. For example, it is known that increased binding to FcγRIIIa generally results in increased ADCC (antibody-dependent cell-mediated cytotoxicity; a cell-mediated reaction in which non-specific cytotoxic cells expressing FcγR recognize bound antibody on a target cell and subsequently cause lysis of the target cell). Similarly, in some cases, decreased binding to FcγRIIb, an inhibitory receptor, may also be advantageous. Amino acid substitutions useful for the antibodies described herein are those enumerated in U.S. Patent No. 8,188,321 (specifically Figure 41 ), U.S. Patent No. 8,084,582, and U.S. Published Application Nos. 20060235208 and 20070148170, all of which are hereby expressly incorporated by reference in their entirety, particularly with respect to the variants disclosed therein. Specific variants that can be 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.

[0304] In addition, as specifically disclosed in USSN 12 / 341,769, which is hereby incorporated by reference in its entirety, there are additional Fc substitutions useful for increasing binding to the FcRn receptor and extending 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 can be included in one or both Fc domains of the antibodies of the invention.

[0305] 2. Ablation Variants

[0306] Similarly, another class of functional variants are "FcγR ablation variants" or "Fc knock-out (FcKO or KO)" variants. In these embodiments, in some therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid additional mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific antibodies that bind monovalently to CD3, it is generally desirable to ablate FcγRIIIa binding to eliminate or significantly reduce ADCC activity. One of the Fc domains contains one or more Fcγ receptor ablation variants. These ablation variants are depicted in Figure 14 and each can be independently and optionally included or not included, and in a preferred aspect, ablation variants selected from the group consisting of: G236R / L328R, E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K, E233P / L234V / L235A / G236del / S239K / A327G, E233P / L234V / L235A / G236del / S267K / A327G, and E233P / L234V / L235A / G236del are utilized. It should be noted that the ablation variants mentioned herein ablate FcγR binding but generally do not ablate FcRn binding.

[0307] As is known in the art, the Fc domain of human IgG1 has the highest binding to Fcγ receptors and thus, when the constant domain (or Fc domain) in the backbone of the bispecific antibody is IgG1, ablation variants can be used. For example, alternatively or in addition to ablation variants in the IgG1 context, mutations at glycosylation position 297 (usually mutated to A or S) can significantly ablate binding to FcγRIIIa. Human IgG2 and IgG4 have a naturally reduced binding to Fcγ receptors and thus those backbones can be used with or without ablation variants.

[0308] D. Combinations of Bispecifics with Fc Variants

[0309] As will be understood by those skilled in the art, all of the described heterodimerization variants (including skewing and / or pI variants) can be optionally and independently combined in any manner as long as they retain their "chainness" or "monomer partitioning". In some embodiments, the bispecific antibodies provided herein include combinations of heterodimerization skewing variants, homotypic pI substitutions, and FcKO variants, as Figure 4 shown. Additionally, all of these variants can be combined into any of the heterodimerized forms in the heterodimerized form.

[0310] In the case of pI variants, while particularly useful embodiments are shown in the figures, other combinations can be generated according to the general rule of varying the pI difference between the two monomers to facilitate purification.

[0311] In addition, any heterodimerization variants, skews, and pIs are also independently and optionally combined with Fc ablation variants, Fc variants, FcRn variants, as generally outlined herein.

[0312] Exemplary combinations of variants included in some embodiments of antibodies in the heterodimeric 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fc forms are included in Figure 4 In certain embodiments, the antibody is an antibody in the heterodimeric 1+1Fab-scFv-Fc or 2+1Fab2-scFv-Fc form, as Figure 15A and Figure 15B shown.

[0313] E. Anti-ENPP3 x Anti-CD3 Bispecific Antibodies

[0314] In another aspect, the present disclosure provides anti-ENPP3 x anti-CD3 (also referred to herein as "αENPP3 x αCD3") bispecific antibodies. Such antibodies include at least one ENPP3 binding domain and at least one CD3 binding domain. In some embodiments, the bispecific αENPP3 x αCD3 provided herein selectively elicits an immune response at tumor sites expressing ENPP3.

[0315] It should be noted that, unless otherwise specified herein, the order of antigens listed by name does not confer structure; that is, an ENPP3 X CD3 1+1Fab-scFv-Fc antibody can have an scFv that binds to ENPP3 or CD3, but in some cases, the order specifies the structure as indicated.

[0316] As more fully outlined herein, these combinations of ABDs can be in various forms as outlined below, typically a combination where one ABD is in Fab form and the other is in scFv form. Exemplary forms for the bispecific antibodies provided herein include 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv forms (see, e.g., Figure 15A and Figure 15B ). Other useful antibody forms include, but are not limited to, "mAb-Fv", "mAb-scFv", "central-Fv", "single-arm scFv-mAb", "scFv-mAb", "dual scFv", and "trifunctional" forms of antibodies, as Figures 52A to 52K disclosed therein.

[0317] In addition, typically, one of the ABDs comprises an scFv having the orientation of VH-scFv linker-VL or VL-scFv linker-VH from the N-terminus to the C-terminus as outlined herein. Depending on the format, one or both of the other ABDs are typically Fabs that comprise a VH domain (usually as a component of the heavy chain) on one protein chain and a VL (usually as a component of the light chain) on another protein chain.

[0318] As will be understood by those skilled in the art, any set of 6 CDRs or VH and VL domains can be in the scFv format or the Fab format, which are then added to the heavy constant domain and the light constant domain, where the heavy constant domain comprises variants (contained within the CH1 domain as well as the Fc domain). The scFv sequences contained in the sequence listing utilize specific charged linkers, whereas as outlined herein, uncharged or other charged linkers can be used, including Figure 5 and Figure 6 those described in

[0319] In addition, as discussed above, the numbering used in the sequence listing to identify the CDRs is Kabat, however, different numberings can be used, which will change the amino acid sequences of the CDRs as shown in Table 2.

[0320] For all variable heavy and light domains listed herein, additional variants can be prepared. As outlined herein, in some embodiments, a set of 6 CDRs can have 0, 1, 2, 3, 4, or 5 amino acid modifications (where amino acid substitutions are particularly useful) and changes in the framework regions of the variable heavy and light domains, provided that the framework (excluding the CDRs) remains at least about 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in FIG. 1, the drawings and legends of which are incorporated herein by reference in their entirety. Thus, for example, the same CDRs as described herein can be combined with different framework sequences from germline sequences, provided that the framework regions remain at least 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in FIG. 1. Alternatively, the CDRs can have amino acid modifications (e.g., 1, 2, 3, 4, or 5 amino acid modifications in a set of CDRs (i.e., the CDRs can be modified provided that the total number of changes in the set of 6 CDRs is less than 6 amino acid modifications, where any combination of CDRs is altered; for example, there can be one change in vlCDR1, two changes in vhCDR2, no change in vhCDR3, etc.)) and have framework region changes, provided that the framework regions remain at least 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in FIG. 1.

[0321] The anti-ENPP3 x anti-CD3 bispecific antibody can include any suitable CD3 ABD, including those described herein (see, e.g., Figures 10A to 10F ). In some embodiments, the CD3 ABD of the anti-ENPP3 x anti-CD3 bispecific antibody includes the variable heavy chain domain and variable light chain domain of the CD3 ABDs provided herein, including Figures 10A to 10F and those described in the Sequence Listing. In some embodiments, the CD3 ABD comprises the variable heavy chain domain and variable light chain domain of any of the following CD3 ABDs: 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 to 10F)。In an exemplary embodiment, the CD3 ABD is one of the following CD3 ABDs: 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 to 10F ) or variants thereof.

[0322] The anti-ENPP3 x anti-CD3 bispecific antibody can comprise any suitable ENPP3 ABD, including those described herein (see, e.g., Figure 12 , Figures 13A to 13B and Figures 14A to 14I ). In some embodiments, the ENPP3 ABD of the anti-ENPP3 x anti-CD3 bispecific antibody comprises the variable heavy chain domain and variable light chain domain of the ENPP3 ABD provided herein, including those described in Figure 12 , Figures 13A to 13B and Figures 14A to 14I and those described in the sequence listing. In some embodiments, the ENPP3 ABD comprises the variable heavy chain domain and variable light chain domain of one of the following ENPP3 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I)。In an exemplary embodiment, the ENPP3 ABD is one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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, H16-9.65, Ha1-1(3,5)19 and Ha16-1.80( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ) or variants thereof.

[0323] F. Anti-SSTR2 x Anti-CD3 Bispecific Antibody

[0324] In another aspect, provided herein are anti-SSTR2 x anti-CD3 (also referred to herein as "αSSTR2 x αCD3") bispecific antibodies. Such antibodies include at least one SSTR2 binding domain and at least one CD3 binding domain. In some embodiments, the bispecific αSSTR2 x αCD3 provided herein selectively elicits an immune response at tumor sites expressing SSTR2.

[0325] Note that, unless otherwise indicated herein, the order of antigens listed by name does not confer structure; that is, an SSTR2 X CD3 1+1Fab-scFv-Fc antibody may have an scFv that binds to SSTR2 or CD3, but in some cases, the order specifies the structure as indicated.

[0326] As more fully outlined herein, these combinations of ABDs can take various forms as outlined below, typically a combination where one ABD is in Fab form and the other is in scFv form. Exemplary forms for the bispecific antibodies provided herein include 1+1Fab-scFv-Fc and 2+1Fab2-scFv-Fv forms (see, e.g., Figure 15A and Figure 15B)。Other useful antibody forms include, but are not limited to, "mAb-Fv", "mAb-scFv", "central-Fv", "single-arm scFv-mAb", "scFv-mAb", "dual scFv", and "trifunctional" forms of antibodies, such as Figures 52A to 52K disclosed in.

[0327] In addition, generally, one of the ABDs includes an scFv that is, as outlined herein, VH-scFv linker-VL or VL-scFv linker-VH in the orientation from the N-terminus to the C-terminus. Depending on the form, one or both of the other ABDs are typically Fabs that include a VH domain on one protein chain (usually as a component of the heavy chain) and a VL (usually as a component of the light chain) on another protein chain.

[0328] As will be understood by those skilled in the art, any set of 6 CDRs or VH and VL domains can be in the scFv form or the Fab form, which are then added to the heavy constant domain and the light constant domain, where the heavy constant domain includes variants (contained within the CH1 domain as well as the Fc domain). The scFv sequences contained in the sequence listing utilize specific charged linkers, and as outlined herein, uncharged or other charged linkers can be used, including Figure 5 and Figure 6 those described in.

[0329] In addition, as discussed above, the Kabat numbering used in the sequence listing to identify CDRs is used; however, different numberings can be used, which will change the amino acid sequences of the CDRs as shown in Table 2.

[0330] For all variable heavy and light domains listed herein, additional variants can be prepared. As outlined herein, in some embodiments, a set of 6 CDRs can have 0, 1, 2, 3, 4, or 5 amino acid modifications (where amino acid substitutions are particularly useful) and changes in the framework regions of the variable heavy and light domains, provided that the framework (excluding the CDRs) remains at least about 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in Figure 1, the drawings and legends of which are incorporated herein by reference in their entirety. Thus, for example, the same CDRs as described herein can be combined with different framework sequences from germline sequences, provided that the framework regions remain at least 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in Figure 1. Alternatively, the CDRs can have amino acid modifications (e.g., 1, 2, 3, 4, or 5 amino acid modifications in a set of CDRs (i.e., the CDRs can be modified as long as the total number of changes in the set of 6 CDRs is less than 6 amino acid modifications, where any combination of CDRs can be altered; for example, there can be one change in vlCDR1, two changes in vhCDR2, no change in vhCDR3, etc.)) and have framework region changes, provided that the framework regions remain at least 80%, 85%, or 90% identical to the germline sequences selected from those listed in U.S. Patent No. 7,657,380 in Figure 1.

[0331] The anti-SSTR2 x anti-CD3 bispecific antibody can include any suitable CD3 ABD, including those described herein (see, e.g., Figures 10A to 10F ). In some embodiments, the CD3 ABD of the anti-SSTR2 x anti-CD3 bispecific antibody includes the variable heavy chain domain and variable light chain domain of the CD3 ABDs provided herein, including Figures 10A to 10F and those described in the sequence listing. In some embodiments, the CD3 ABD comprises the variable heavy chain domain and variable light chain domain of any of the following CD3 ABDs: 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 to 10F)。In an exemplary embodiment, the CD3 ABD is one of the following CD3 ABDs: 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 to 10F ) or variants thereof.

[0332] The anti-SSTR2 x anti-CD3 bispecific antibody may comprise the variable heavy chain domain and variable light domain of [αSSTR2]H1.24_L1.30( Figure 63 ) or variants thereof.

[0333] G. Useful Forms of the Invention

[0334] As will be understood by those skilled in the art and discussed more fully below, the heterodimeric bispecific antibodies provided herein can take various configurations, generally depicted in FIG. 1. Some figures depict a "single-ended" configuration where one type of specificity is present on one "arm" of the molecule and a different specificity is present on the other "arm". Other figures depict a "double-ended" configuration where at least one type of specificity is present at the "top" of the molecule and one or more different specificities are present at the "bottom" of the molecule. Thus, in some embodiments, the antibodies described herein are novel immunoglobulin compositions that co-engage different first and second antigens.

[0335] As will be understood by those skilled in the art, the heterodimeric forms described herein can have different valencies and can be bispecific. That is, the heterodimeric antibodies described herein can be bivalent and bispecific, where one target tumor antigen (e.g., CD3) is bound by one binding domain and another target tumor antigen (e.g., ENPP3) is bound by a second binding domain. The heterodimeric antibody can also be trivalent and bispecific, where the first antigen is bound by two binding domains and the second antigen is bound by a second binding domain. As outlined herein, when CD3 is one of the target antigens, it is preferred that CD3 binds only monovalently to reduce potential side effects.

[0336] The antibodies described herein utilize a combination of an anti-CD3 antigen-binding domain and 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 and variable heavy chain domains, Fabs, and scFvs as depicted in any figure may be used. Similarly, any anti-ENPP3 antigen-binding domain may be used, whether or not the CDRs, variable light and variable heavy chain domains, Fabs, and scFvs as depicted in any figure (e.g., Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ) may be used, optionally and independently in any combination.

[0337] 1.1 + 1 Fab-scFv-Fc format

[0338] One heterodimeric scaffold found to be particularly useful for the antibodies described herein is the "1 + 1 Fab-scFv-Fc" or "bottle opener" format, as Figure 15A shown, 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 contains a single-chain Fv ("scFv", defined below) and an Fc domain. The scFv contains a variable heavy chain domain (VH1) and a variable light chain domain (VL1), where VH1 is attached to VL1 using a charged scFv linker (see, e.g., Figure 5 ). The scFv is attached to the heavy chain using a domain linker (see, e.g., Figure 6 ). The other heavy chain monomer is a "conventional" heavy chain (VH-CH1-hinge-CH2-CH3). The 1 + 1 Fab-scFv-Fc also contains a light chain that interacts with VH-CH1 to form a Fab. This structure is sometimes referred to herein as the "bottle opener" format due to its approximate visual similarity to a bottle opener. By using amino acid variants (e.g., the heterodimerization variants discussed above) in the constant regions (e.g., the Fc domain, CH1 domain, and / or hinge region) that promote heterodimer antibody formation as described more fully below, these two heavy chain monomers are brought together.

[0339] The "1 + 1 Fab-scFv-Fc" format of the present invention has several distinct advantages. As is known in the art, antibody analogs that rely on two scFv constructs often have stability and aggregation problems, which can be alleviated in the antibodies described herein by adding "conventional" heavy and light chain pairings. Additionally, in contrast to forms that rely on two heavy chains and two light chains, there is no problem of improper heavy and light chain pairing (e.g., heavy chain 1 paired with light chain 2, etc.).

[0340] Many of the embodiments outlined herein generally rely on a 1+1 Fab-scFv-Fc or "opener" form antibody comprising a first monomer, the first monomer comprising an scFv, the scFv comprising a variable heavy chain domain and a variable light chain domain covalently attached using an scFv linker (charged in many but not all cases), wherein the scFv is typically covalently attached to the N-terminus of a first Fc domain via a domain linker. The domain linker can be charged or uncharged and exogenous or endogenous (e.g., all or part of the native 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 Figure 6 the domain linkers in. The second monomer of the 1+1 Fab-scFv-Fc form or "opener" form is a heavy chain, and the composition further comprises a light chain.

[0341] Typically, in many preferred embodiments, the scFv is a domain that binds to CD3, and the Fab forms an ENPP3 binding domain. An exemplary anti-ENPP3 x anti-CD3 bispecific antibody in the 1+1 Fab-scFv-Fc form is depicted in Figure 15A . An exemplary anti-ENPP3 x anti-CD3 bispecific antibody in the 1+1 Fab-scFv-Fc form is depicted in Figures 17A to 17C and Figures 18A to 18C .

[0342] Additionally, the Fc domain of the antibodies described herein generally comprises skewed variants (e.g., a set of amino acid substitutions as shown in Figure 3 and Figure 9 , wherein particularly useful skewed variants are selected from the group consisting of: 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), optionally ablation variants (comprising those shown in Figure 3 ), optionally charged scFv linkers (comprising those shown in Figure 5 ), and the heavy chain comprises pI variants (comprising those shown in Figure 4 ).

[0343] In certain embodiments, the 1+1 Fab-scFv-Fc scaffold format comprises a first monomer comprising a scFv-domain linker-CH2-CH3 monomer; a second monomer comprising a first variable heavy chain domain-CH1-hinge-CH2-CH3 monomer; and a third monomer comprising 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 comprises a scFv variable heavy chain domain and a 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 linked using a scFv linker (charged, in many but not all cases). 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 moiety. Particularly useful ENPP3 and CD3 combinations for the 1+1 Fab-scFv-Fc ENPP3 xCD3 bispecific antibody format are disclosed in Figures 17A to 17C and Figures 18A to 18C and include: ENPP3 H16-1.93x CD3 H1.30 L1.47, ENPP3 H16-7.8x CD3 H1.30L1.47, ENPP3 AN1[ENPP3]H1L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1 x CD3H1.30L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33 x CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77 xCD3H.130L1.47. In some embodiments, the 1+1 Fab-scFv-Fc format comprises skewed variants, pI variants, and ablation variants. Thus, some embodiments comprise a 1+1 Fab-scFv-Fc format that includes: a) a first monomer (“scFv monomer”) that includes a charged scFv linker ( Figure 5The +H sequence is preferred in some embodiments), the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and the scFv that binds to CD3 as outlined herein; b) a second monomer (“Fab monomer”), the second monomer comprising the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and the variable heavy chain domain; and c) a light chain, the light chain comprising a variable light chain domain (VL) and a constant light chain domain (CL), where the numbering is according to EU numbering. The variable heavy chain domain and the variable light chain domain constitute the ENPP3 binding portion. CD3 binding domain sequences particularly useful 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, and those as Figures 10A to 10F shown. ENPP3 binding domain sequences particularly useful in these embodiments include, but are not limited to, AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8L1.33, 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I)。Particularly useful combinations of ENPP3 and CD3 sequences for antibodies in the 1+1 Fab2-scFv-Fc format include, for example, ENPP3 H16-1.93x CD3 H1.30 L1.47, ENPP3 H16-7.8x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33 x CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77 x CD3 H.130 L1.47.

[0344] Exemplary variable heavy and light domains of the scFv that binds CD3 are included in Figures 10A to 10F . Exemplary variable heavy and light domains of the Fv that binds ENPP3 are included in Figure 12 , Figures 13A to 13B and Figures 14A to 14I . In an exemplary embodiment, the ENPP3-binding domain of the 1+1 Fab-scFv-Fc ENPP3 x CD3 bispecific antibody comprises the VH and VL of one of the following ENPP3-binding domains: AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I)。In one embodiment, the CD3 binding domain of the 1+1 Fab-scFv-Fc ENPP3 x CD3 bispecific antibody comprises the VH and VL of 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 to 10F )。Particularly useful ENPP3 and CD3 combinations for the 1+1 Fab-scFv-Fc ENPP3 x CD3 bispecific antibody format are disclosed in Figures 17A to 17C and Figures 18A to 18C and include: ENPP3 H16-1.93 x CD3 H1.30 L1.47, ENPP3 H16-7.8 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33 x CD3 H1.30L1.47, and ENPP3 H1.8 L1.77 x CD3 H.130L1.47.

[0345] In some embodiments, the 1+1 Fab-scFv-Fc format comprises skewed variants, pI variants, ablation variants, and FcRn variants. Accordingly, some embodiments comprise a 1+1 Fab-scFv-Fc format that includes: a) a first monomer (“scFv monomer”) that comprises a charged scFv linker( Figure 6The +H sequence is preferred in some embodiments), the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and the scFv that binds to CD3 as outlined herein; b) a second monomer (“Fab monomer”), the second monomer comprising the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and the variable heavy chain domain; and c) a light chain, the light chain comprising a variable light chain domain (VL) and a constant light chain domain (CL), where the numbering is according to EU numbering. The variable heavy chain domain and the variable light domain constitute the ENPP3 binding domain. CD3 binding domain sequences particularly useful 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, and those as Figures 10A to 10F shown. ENPP3 binding domain sequences particularly useful in these embodiments include, but are not limited to, AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8L1.33, 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, Ha16-1(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 in Figure 12 , Figures 13A to 13B and Figures 14A to 14Idepicted. Particularly useful combinations of ENPP3 and CD3 sequences for 1+1 Fab2-scFv-Fc format antibodies include, for example, those disclosed in Figures 17A to 17C and Figures 18A to 18C and include: ENPP3 H16-1.93x CD3 H1.30 L1.47, ENPP3 H16-7.8x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1 x CD3 H1.30 L1.47, ENPP3 AN1[ENPP3]H1.8 L1.33 x CD3 H1.30 L1.47, and ENPP3 H1.8 L1.77 x CD3 H.130 L1.47.

[0346] Figures 7A to 7D Some exemplary Fc domain sequences useful in 1+1 Fab-scFv-Fc format antibodies are shown. Figures 7A to 7D The "monomer 1" sequence depicted in Figure 9 generally refers to the Fc domain of the "Fab-Fc heavy chain", while the "monomer 2" sequence refers to the Fc domain of the "scFv-Fc heavy chain". In addition,

[0347] In some embodiments, any VH and VL sequences depicted herein (including all VH and VL sequences depicted in the figures and sequence listing, including sequences against ENPP3) can be added to the Figures 7A to 7D opener backbone format as the "Fab side" using any anti-CD3 scFv sequence shown in the figures and sequence listing.

[0348] For opener backbone 1 from Figure 7A (optionally including the 428L / 434S variant), CD binding domain sequences particularly useful in these embodiments include, but are not limited to, CD3 binding domain 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 depicted in Figures 10A to 10F which are linked as the scFv side of the backbone shown in Figures 7A to 7D Particularly useful combinations of ENPP3 and CD3 sequences (optionally including the 428L / 434S variant) are disclosed in

[0349] Figures 17A to 17C and Figures 18A to 18C ​

[0350] 2.mAb-Fv

[0351] One heterodimeric scaffold particularly useful in the antibodies described herein is the mAb-Fv form. In this embodiment, the form relies on the use of an "extra" variable heavy chain domain attached to the C-terminus of one monomer and an "extra" variable light chain domain attached to the C-terminus of the other monomer, thus forming a third antigen-binding domain, wherein the Fab portions of the two monomers bind ENPP3 and the "extra" scFv domain binds CD3.

[0352] In this embodiment, the first monomer comprises a first heavy chain that comprises a first variable heavy chain domain and a first heavy chain constant domain that includes a first Fc domain, wherein a first variable light chain domain is covalently attached to the C-terminus of the first Fc domain using a domain linker (VH1-CH1-hinge-CH2-CH3-[optional linker]-VL2). The second monomer comprises a second variable heavy chain domain that belongs to a second heavy chain constant domain that includes a second Fc domain, and a third variable heavy chain domain that is covalently linked to the C-terminus of the second Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[optional linker]-VH2). The two C-terminally attached variable domains constitute an Fv that binds CD3 (since bivalent CD3 binding is less preferred). This embodiment further utilizes a light chain that comprises a variable light chain domain and a light chain constant domain, and the light chain associates with the heavy chain to form two identical Fabs that bind ENPP3. For many embodiments herein, these constructs include skewed variants, pI variants, ablation variants, additional Fc variants, etc. as desired and described herein.

[0353] The antibodies described herein provide the mAb-Fv form, wherein the CD3-binding domain sequence is as Figures 10A to 10F shown. The antibodies described herein provide the mAb-Fv form, wherein the ENPP3-binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0354] In addition, the Fc domain of the mAb-Fv form comprises skewed variants (e.g., Figure 3and a set of amino acid substitutions shown in Figure 8, and particularly suitable skewed variants are selected from the group consisting of: 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); optionally ablated variants (including those shown in Figure 3 ); optionally charged scFv linkers (including those shown in Figure 5 ); and heavy chains comprising pI variants (including those shown in Figure 2 ).

[0355] In some embodiments, the mAb-Fv form comprises skewed variants, pI variants, and ablated variants. Accordingly, some embodiments comprise an mAb-Fv form that includes: a) a first monomer that includes the skewed variant S364K / E357Q, the ablated variant E233P / L234V / L235A / G236del / S267K, and a first variable heavy chain domain and a second variable heavy chain domain, the first variable heavy chain domain and the first variable light domain of the light chain forming an Fv that binds to ENPP3; b) a second monomer that includes the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablated variant E233P / L234V / L235A / G236del / S267K, and a first variable heavy chain domain and a second variable light chain, the first variable heavy chain domain and the first variable light domain forming an Fv that binds to ENPP3 as outlined herein, the second variable light chain and the second variable heavy chain domain together forming an Fv (ABD) that binds to CD3; and c) a light chain that includes a first variable light domain and a constant light domain.

[0356] In some embodiments, the mAb-Fv format includes skewed variants, pI variants, ablation variants, and FcRn variants. Accordingly, some embodiments include an mAb-Fv format that includes: a) a first monomer that includes the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a first variable heavy chain domain and a second variable heavy chain domain, wherein the first variable heavy chain domain and the first variable light domain of the light chain form an Fv that binds to ENPP3; b) a second monomer that includes the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a first variable heavy chain domain and a second variable light chain, wherein the first variable heavy chain domain and the first variable light domain form an Fv that binds to ENPP3 as outlined herein, and the second variable light chain and the second variable heavy chain domain of the first monomer together form an Fv (ABD) that binds to CD3; and c) a light chain that includes a first variable light domain and a constant light domain.

[0357] 3. mAb-scFv

[0358] One heterodimeric scaffold particularly useful in the antibodies described herein is the mAb-scFv format. In this embodiment, the format relies on the attachment of an scFv to the C-terminus of one of the monomers to form a third antigen-binding domain, where the Fab portions of the two monomers bind to ENPP3 and the "extra" scFv domain binds to CD3. Accordingly, the first monomer includes a first heavy chain (including a variable heavy chain domain and a constant domain), with an scFv covalently attached to the C-terminus, which includes an scFv variable light chain domain, an scFv linker, and an scFv variable heavy chain domain, in either orientation. This embodiment further utilizes a light chain that includes a variable light domain and a light chain constant domain, which associates with the heavy chain to form two identical Fabs that bind to ENPP3. For many embodiments herein, these constructs include skewed variants, pI variants, ablation variants, additional Fc variants, etc., as desired and described herein.

[0359] The antibodies described herein provide an mAb-scFv format, where the CD-binding domain sequence is as Figures 10A to 10F shown, and the ENPP3-binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0360] In addition, the Fc domain in the mAb-scFv format contains skewed variants (e.g., a set of amino acid substitutions shown in Figure 1, and particularly suitable skewed variants are selected from the group consisting of: 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); optionally ablated variants (including those shown in Figure 3 ); optionally charged scFv linkers (including those shown in Figure 5 ); and heavy chains containing pI variants (including those shown in Figure 2 ).

[0361] In some embodiments, the mAb-scFv format contains skewed variants, pI variants, and ablated variants. Thus, some embodiments contain an mAb-scFv format that includes: a) a first monomer that includes the skewed variant S364K / E357Q, the ablated variant E233P / L234V / L235A / G236del / S267K, a variable heavy chain domain, and an scFv domain, the variable heavy chain domain and the variable light chain domain of a common light chain form an Fv that binds to ENPP3 as outlined herein, and the scFv domain binds to CD3; b) a second monomer that includes the skewed variant L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, the ablated variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light chain domain of a common light chain form an Fv that binds to ENPP3 as outlined herein; and c) a common light chain that includes a variable light domain and a constant light domain.

[0362] In some embodiments, the mAb-scFv format comprises skewed variants, pI variants, ablation variants, and FcRn variants. Thus, some embodiments include an mAb-scFv format that comprises: a) a first monomer that comprises the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy chain domain that together with the variable light chain domain of a common light chain forms an Fv that binds to ENPP3 as outlined herein and a ScFv domain that binds to CD3; b) a second monomer that comprises the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy chain domain that together with the variable light chain domain of a common light chain forms an Fv that binds to ENPP3 as outlined herein; and c) a common light chain that comprises a variable light chain domain and a light chain constant domain.

[0363] 4.2 + 1 Fab2-scFv-Fc format

[0364] One heterodimeric scaffold found to be particularly useful in the antibodies described herein is the "2 + 1 Fab2-scFv-Fc" format (also referred to in previous related documents as the "central-scFv format"), as Figure 15BAs shown, an exemplary combination having 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 to form a third antigen binding domain, wherein the Fab portions of two monomers bind to ENPP3 and the "extra" scFv domain binds to CD3. The scFv domain is inserted between the Fc domain and the CH1-Fv region of one of the monomers, thereby providing a third antigen binding domain. For example, an ENPP3 x CD3 bispecific antibody having a 2+1Fab2-scFv-Fc format is effective in inducing redirected T cell cytotoxicity in a cellular environment expressing low levels of ENPP3. Additionally, as shown in the examples, an ENPP3 x CD3 bispecific antibody having a 2+1Fab2-scFv-Fc format allows for "fine-tuning" of the immune response, as such antibodies exhibit a wide variety of different properties depending on the ENPP3 and / or CD3 binding domains used. For example, such antibodies exhibit differences in selectivity for cells having different ENPP3 expression, potency against ENPP3-expressing cells, ability to trigger cytokine release, and sensitivity to soluble ENPP3. These ENPP3 antibodies can be used, for example, to treat ENPP3-related cancers.

[0365] In this embodiment, one monomer includes a first heavy chain that includes a first variable heavy chain domain, a CH1 domain (and optionally a hinge), and an Fc domain, wherein the scFv includes an scFv variable light domain, an scFv linker, and an scFv variable heavy chain domain. The scFv is covalently linked 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 the reverse orientation of the scFv VH1-CH1-[optional linker]-VL2-scFv linker-VH2-[optional linker including hinge]-CH2-CH3). The optional linker can be any suitable peptide linker, including, for example Figure 6 the domain linkers included in. In some embodiments, the optional linker is a hinge or a fragment thereof. The other monomer is a standard Fab side (i.e., VH1-CH1-hinge-CH2-CH3). This embodiment further utilizes a light chain that includes a variable light chain domain and a light chain constant domain, and the light chain associates with the heavy chain to form two identical Fabs that bind ENPP3. For many of the embodiments herein, these constructs include skewed variants, pI variants, ablation variants, additional Fc variants, etc. as desired and described herein.

[0366] In one embodiment, the 2+1 Fab2-scFv-Fc format antibody comprises an scFv having a VH and a VL with a CD3 binding domain sequence as depicted in Figures 10A to 10F or in the Sequence Listing. In one embodiment, the 2+1 Fab2-scFv-Fc format antibody comprises two Fabs having a VH and a VL with an ENPP3 binding domain as in Figure 12 , Figures 13A to 13B and Figures 14A to 14I and as shown in the Sequence Listing. In an exemplary embodiment, the ENPP3 binding domain of the 2+1 Fab2-scFv-Fc ENPP3 x CD3 bispecific antibody comprises a VH and a VL of 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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ). In one embodiment, the CD3 binding domain of the 2+1 Fab2-scFv-Fc format antibody comprises a VH and a VL of 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 to 10F ). Particularly useful ENPP3 and CD3 combinations for the 2+1 Fab2-scFv-Fc format antibody are in Figures 19A to 19C ,Figures 20A to 20D , Figure 21 , Figures 22A to 22C , Figures 23A to 23E are disclosed in, and include: ENPP3 H1.8 L1 x CD3 H1.30 L1.47, ENPP3 H1.8 L1.33 x CD3 H1.30 L1.47, ENPP3 H1.8 L1.77 x CD3 H1.30 L1.47, ENPP3 H16-7.8 x CD3 H1.32 L1.47, ENPP3 AN[ENPP3]H1L1 x CD3 H1.32 L1.47, ENPP3 H1.8 L1 x CD3 H1.32 L1.47, ENPP3 H1.8 L1.33 x CD3 H1.32 L1.47, ENPP3 H1.8 L1 x CD3 L1.47 H1.30, ENPP3 H1.8 L1 x CD3 L1.47 H1.32, ENPP3 H1.8 L1.33 x CD3 L1.47 H1.32, ENPP3 H1.8 L1.77 x CD3 L1.47 H1.32, ENPP3 H1.8 L1 x CD3 L1.47 H1.89, ENPP3 H1.8 L1.33 x CD3 L1.47 H1.89, ENPP3 H1.8 L1.77 x CD3 L1.47 H1.89, ENPP3 H1.8 L1.33 x CD3 L1.47 H1.89, and ENPP3 H1.8 L1.77 x CD3 L1.47 H1.89.

[0367] In addition, the Fc domain in the 2+1Fab2-scFv-Fc form contains skewed variants (e.g., a set of amino acid substitutions shown in Figure 1, and particularly applicable skewed variants are selected from the group consisting of: 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); optionally ablated variants (including those shown in Figure 3 ); optionally charged scFv linkers (including those shown in Figure 5 ); and heavy chains containing pI variants (including those shown in Figure 2 ).

[0368] In some embodiments, the 2+1 Fab2-scFv-Fc format antibody comprises a skewed variant, a pI variant, and an ablation variant. Accordingly, some embodiments comprise the 2+1 Fab2-scFv-Fc format, which includes: a) a first monomer (Fab-scFv-Fc side), the first monomer comprising the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain that forms, with a variable light chain domain of a shared light chain, an Fv that binds to ENPP3 as outlined herein and an scFv domain that binds to CD3; b) a second monomer (Fab-Fc side), the second monomer comprising the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain that forms, with a variable light chain domain of a shared light chain, an Fv that binds to ENPP3 as outlined herein; and c) a light chain, the light chain comprising a variable light domain and a constant light domain, where the numbering is according to EU numbering. In some embodiments, the shared light chain and the variable heavy chain domains on each monomer form an ENPP3 binding domain. CD3 binding domain sequences particularly useful 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, and as Figures 10A to 10FThose shown. ENPP3 binding domain sequences particularly useful in these embodiments include, but are not limited to, AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, 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, as depicted in Figure 12 , Figures 13A to 13B and Figures 14A to 14I as depicted.

[0369] In some embodiments, the 2+1 Fab2-scFv-Fc format antibody comprises a skewed variant, a pI variant, an ablation variant, and an FcRn variant. Accordingly, some embodiments comprise the 2+1 Fab2-scFv-Fc format, which includes: a) a first monomer (Fab-scFv-Fc side), the first monomer comprising the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy domain, the variable heavy domain and the variable light domain of a shared light chain forming an Fv that binds to ENPP3 as outlined herein and an scFv domain that binds to CD3; b) a second monomer (Fab-Fc side), the second monomer comprising the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy domain, the variable heavy domain and the variable light domain of a shared light chain forming an Fv that binds to ENPP3 as outlined herein; and c) a light chain, the light chain comprising a variable light domain and a constant light domain, wherein the numbering is according to EU numbering. In some embodiments, the shared light chain and the variable heavy chain domains on each monomer form an ENPP3 binding domain. CD3 binding domain sequences particularly useful 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, and as Figures 10A to 10FThose shown. ENPP3 binding domain sequences particularly useful in these embodiments include, but are not limited to, 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, as in Figure 12 , Figures 13A to 13B and Figures 14A to 14I depicted.

[0370] Figures 8A to 8C Some exemplary Fc domain sequences are shown, which can be used in the 2+1Fab2-scFv-Fc format. Figures 8A to 8C The "monomer 1" sequence depicted in Figure 9 generally refers to the Fc domain of "Fab-Fc heavy chain", and the "monomer 2" sequence refers to the Fc domain of "Fab-scFv-Fc heavy chain". In addition,

[0371] 5. Central-Fv

[0372] One heterodimeric scaffold particularly useful in the antibodies described herein is the Central-Fv format. In this embodiment, the format relies on the use of an inserted Fv domain (i.e., the central Fv domain) to form an "extra" third antigen-binding domain, where the Fab portions of two monomers bind to ENPP3 and the "extra" central Fv domain binds to CD3. The "extra" central Fv domain is inserted between the Fc domain and the CH1-Fv region of the monomer, thus providing a third antigen-binding domain (i.e., the "extra" central Fv domain), where each monomer contains a component of the "extra" central Fv domain (i.e., one monomer includes the variable heavy chain domain and the other includes the variable light chain domain of the "extra" central Fv domain).

[0373] In this embodiment, a monomer comprises a first heavy chain that includes a first variable heavy chain domain, a CH1 domain, an Fc domain, and an additional variable light domain. The light chain domain is covalently linked between the C-terminus of the CH1 domain of the heavy chain constant domain and the N-terminus of the first Fc domain using a domain linker (VH1-CH1-[optional linker]-VL2-hinge-CH2-CH3). Another monomer comprises a first heavy chain that includes a first variable heavy chain domain, a CH1 domain, an Fc domain, and an additional variable heavy chain domain (VH1-CH1-[optional linker]-VH2-hinge-CH2-CH3). The light chain domain is covalently linked between the C-terminus of the CH1 domain of the heavy chain constant domain and the N-terminus of the first Fc domain using a domain linker.

[0374] This embodiment further utilizes a light chain that includes a variable light chain domain and a light chain constant domain, which associates with the heavy chain to form two identical Fabs that each bind ENPP3. For many embodiments herein, these constructs include skewed variants, pI variants, ablation variants, additional Fc variants, etc. as desired and described herein.

[0375] The antibodies described herein provide a central-Fv form, where the CD3 binding domain sequence is as Figures 10A to 10F shown, and the ENPP3 binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0376] 6. Single-arm central-scFv

[0377] One heterodimeric scaffold particularly useful for the antibodies described herein is the single-arm central-scFv form. In this embodiment, one monomer includes only the Fc domain, while the other monomer includes a Fab domain (first antigen-binding domain), an scFv domain (second antigen-binding domain), and an Fc domain, where the scFv domain is inserted between the Fc domains. In this form, the Fab portion binds to one receptor target, and the scFv binds to another receptor target. In this form, the Fab portion binds to ENPP3 and the scFv binds to CD3, or vice versa.

[0378] In this embodiment, a monomer comprises a first heavy chain, the first heavy chain comprising a first variable heavy chain domain, a CH1 domain, and an Fc domain, wherein the scFv comprises an scFv variable light domain, an scFv linker, and an scFv variable heavy chain domain. The scFv is covalently linked between the C-terminus of the CH1 domain of the heavy constant domain and the N-terminus of the first Fc domain using a domain linker in either orientation (VH1-CH1-[optional domain linker]-VH2-scFv linker-VL2-[optional domain linker]-CH2-CH3 or VH1-CH1-[optional domain linker]-VL2-scFv linker-VH2-[optional domain linker]-CH2-CH3). A second monomer comprises an Fc domain (CH2-CH3). This embodiment further utilizes a light chain comprising a variable light domain and a constant light domain, the light chain associating with the heavy chain to form a Fab.

[0379] For many embodiments herein, these constructs comprise skewed variants, pI variants, ablation variants, additional Fc variants, etc., as desired and as described herein.

[0380] The antibodies described herein provide a central-Fv form, wherein the CD3 binding domain sequence is as Figures 10A to 10F shown, and the ENPP3 binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0381] Additionally, the Fc domain of the single-arm central-scFv form typically comprises skewed variants (e.g., a set of amino acid substitutions as shown in FIG. 1, wherein particularly useful skewed variants are selected from the group consisting of: 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), optionally ablation variants (comprising those as shown in Figure 3 ), optionally charged scFv linkers (comprising those as shown in Figure 5 ), and the heavy chain comprises pI variants (comprising those as shown in Figure 2 ).

[0382] In some embodiments, the single-arm central-scFv format comprises skewed variants, pI variants, and ablation variants. Accordingly, some embodiments of the single-arm central-scFv format include: a) a first monomer comprising the skewed variants S364K / E357Q, the ablation variants E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain and an scFv domain, the variable heavy chain domain and the variable light domain of the light chain forming an Fv that binds to ENPP3 as outlined herein, the scFv domain binding to CD3; b) a second monomer comprising an Fc domain having the skewed variants L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, the ablation variants E233P / L234V / L235A / G236del / S267K; and c) a light chain comprising a variable light domain and a constant light domain.

[0383] In some embodiments, the single-arm central-scFv format comprises skewed variants, pI variants, ablation variants, and FcRn variants. Accordingly, some embodiments of the single-arm central-scFv format include: a) a first monomer comprising the skewed variants S364K / E357Q, the ablation variants E233P / L234V / L235A / G236del / S267K, the FcRn variants M428L / N434S, and a variable heavy chain domain and an scFv domain, the variable heavy chain domain and the variable light domain of the light chain forming an Fv that binds to ENPP3 as outlined herein, the scFv domain binding to CD3; b) a second monomer comprising an Fc domain having the skewed variants L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, the ablation variants E233P / L234V / L235A / G236del / S267K, and the FcRn variants M428L / N434S; and c) a light chain comprising a variable light domain and a constant light domain.

[0384] 7. Single-arm scFv-mAb

[0385] One heterodimeric scaffold particularly useful in the antibodies described herein is the single-arm scFv-mAb format. In this embodiment, one monomer contains only the Fc domain, while the other monomer uses an scFv domain attached to the N-terminus of the heavy chain, typically via the use of a linker: VH-scFv linker-VL-[optional domain linker]-CH1-hinge-CH2-CH3 or (in the reverse orientation) VL-scFv linker-VH-[optional domain linker]-CH1-hinge-CH2-CH3. In this form, the Fab portions each bind ENPP3 and the scFv binds CD3. This embodiment further utilizes a light chain comprising a variable light chain domain and a light chain constant domain, the light chain associating with the heavy chain to form a Fab. For many embodiments herein, these constructs contain skewed variants, pI variants, ablation variants, additional Fc variants, etc. as desired and described herein.

[0386] The antibodies described herein provide the single-arm scFv-mAb format, wherein the CD3 binding domain sequence is as Figures 10A to 10F shown, and wherein the ENPP3 binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0387] In addition, the Fc domain of the single-arm scFv-mAb format typically includes skewed variants (e.g., Figure 3 the set of amino acid substitutions shown in and 8, and particularly useful skewed variants are selected from the group consisting of: 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); optionally ablation variants (including Figure 3 those shown in); optionally charged scFv linkers (including Figure 5 those shown in); and heavy chains containing pI variants (including Figure 2 those shown in).

[0388] In some embodiments, the single-arm scFv-mAb format includes skewed variants, pI variants, and ablation variants. Accordingly, some embodiments of the single-arm scFv-mAb format include: a) a first monomer that includes the skewed variants S364K / E357Q, the ablation variants E233P / L234V / L235A / G236del / S267K, a variable heavy domain, and an scFv domain, the variable heavy domain and the variable light domain of the light chain forming an Fv that binds to ENPP3 as outlined herein, the scFv domain binding to CD3; b) a second monomer that includes an Fc domain having the skewed variants L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, and the ablation variants E233P / L234V / L235A / G236del / S267K; and c) a light chain that includes a variable light domain and a constant light domain.

[0389] In some embodiments, the single-arm scFv-mAb format includes skewed variants, pI variants, ablation variants, and FcRn variants. Accordingly, some embodiments of the single-arm scFv-mAb format include: a) a first monomer that includes the skewed variants S364K / E357Q, the ablation variants E233P / L234V / L235A / G236del / S267K, the FcRn variants M428L / N434S, a variable heavy domain, and an scFv domain, the variable heavy domain and the variable light domain of the light chain forming an Fv that binds to ENPP3 as outlined herein, the scFv domain binding to CD3; b) a second monomer that includes an Fc domain having the skewed variants L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, the ablation variants E233P / L234V / L235A / G236del / S267K, and the FcRn variants M428L / N434S; and c) a light chain that includes a variable light domain and a constant light domain.

[0390] 8. scFv-mAb

[0391] One heterodimeric scaffold particularly useful in the antibodies described herein is the mAb-scFv format. In this embodiment, the format relies on the attachment of an scFv to the N-terminus of one of the monomers to form a third antigen-binding domain, where the Fab portions of the two monomers bind to ENPP3 and the "extra" scFv domain binds to CD3.

[0392] In this embodiment, the first monomer comprises a first heavy chain (comprising a variable heavy chain domain and a constant domain), wherein the N-terminally covalently linked scFv comprises the scFv variable light domain, the scFv linker, and the scFv variable heavy chain domain in either orientation ((VH1 - scFv linker - VL1 - [optional domain linker] - VH2 - CH1 - hinge - CH2 - CH3) or (scFv in the opposite orientation) ((VL1 - scFv linker - VH1 - [optional domain linker] - VH2 - CH1 - hinge - CH2 - CH3)). This embodiment further utilizes a light chain comprising a variable light chain domain and a light chain constant domain, which associates with the heavy chain to form two identical Fabs that bind ENPP3. For many embodiments herein, these constructs comprise skewed variants, pI variants, ablation variants, additional Fc variants, etc. as desired and described herein.

[0393] The antibodies described herein provide an scFv - mAb format, wherein the CD3 binding domain sequence is as Figures 10A to 10F shown, and wherein the ENPP3 binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown.

[0394] Additionally, the Fc domain of the scFv - mAb format generally comprises skewed variants (e.g., a set of amino acid substitutions as shown in Figure 1, wherein particularly useful skewed variants are selected from the group consisting of: 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), optionally ablation variants (comprising those as shown in Figure 3 ), optionally charged scFv linkers (comprising those as shown in Figure 5 ), and the heavy chain comprises pI variants (comprising those as shown in Figure 2 ).

[0395] In some embodiments, the scFv-mAb format comprises skewed variants, pI variants, and ablation variants. Accordingly, some embodiments comprise an scFv-mAb format that includes: a) a first monomer that includes the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, a variable heavy chain domain, and an scFv domain, the variable heavy chain domain and the variable light domain of a common light chain forming an Fv that binds to ENPP3 as outlined herein, the scFv domain binding to CD3; b) a second monomer that includes the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light domain of a common light chain forming an Fv that binds to ENPP3 as outlined herein; and c) a common light chain that includes a variable light domain and a constant light domain.

[0396] In some embodiments, the scFv-mAb format comprises skewed variants, pI variants, ablation variants, and FcRn variants. Accordingly, some embodiments include an scFv-mAb format that comprises: a) a first monomer that comprises the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy chain domain (which together with the variable light chain domain of a common light chain forms an Fv that binds to ENPP3 as outlined herein and an scFv domain that binds to CD3); b) a second monomer that comprises the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a variable heavy chain domain (which together with the variable light chain domain of a common light chain forms an Fv that binds to ENPP3 as outlined herein); and c) a common light chain that comprises a variable light chain domain and a light chain constant domain.

[0397] 9. Dual scFv Format

[0398] The antibodies described herein also provide dual scFv formats known in the art. In this embodiment, the ENPP3 xCD3 heterodimeric bispecific antibody is composed of two scFv-Fc monomers (both monomers are in the form of (VH-scFv linker-VL-[optional domain linker]-CH2-CH3) or (VL-scFv linker-VH-[optional domain linker]-CH2-CH3), or one monomer in one orientation and the other monomer in the other orientation).

[0399] The antibodies described herein provide dual scFv formats, wherein the CD3 binding domain sequence is as Figures 10A to 10F shown, and wherein the ENPP3 binding domain sequence is as Figure 12 , Figures 13A to 13B and Figures 14A to 14I shown. In some embodiments, the dual scFv formats include skewed variants, pI variants, and ablation variants. Thus, some embodiments include dual scFv formats that include: a) a first monomer that includes the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and a first scFv that binds to CD3 or ENPP3; and b) a second monomer that includes the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant

[0400] E233P / L234V / L235A / G236del / S267K, and a second scFv that binds to CD3 or ENPP3. In some embodiments, the dual scFv formats include skewed variants, pI variants, ablation variants, and FcRn variants. In some embodiments, the dual scFv formats include skewed variants, pI variants, and ablation variants. Thus, some embodiments include dual scFv formats that include: a) a first monomer that includes the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a first scFv that binds to CD3 or ENPP3; and b) a second monomer that includes the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, the FcRn variant M428L / N434S, and a second scFv that binds to CD3 or ENPP3.

[0401] 10. Non-heterodimeric bispecific antibodies

[0402] As will be appreciated by those skilled in the art, the ENPP3 and CD3 Fv sequences outlined herein can also be used in both monospecific antibodies (eg, "conventional monoclonal antibodies") or non-heterodimeric bispecific formats.

[0403] Particularly useful CD3 binding domain sequences 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 ( Figures 10A to 10F ).

[0404] Particularly useful ENPP3 binding domain sequences include, but are not limited to, AN1[ENPP3]H1L1, AN1[ENPP3]H1L1.33, 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, Ha16-1. 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( Figure 12 , Figures 13A to 13B and Figures 14A to 14I ).

[0405] Suitable non-heterodimeric bispecific formats are known in the art and include many different formats as generally described in Spiess et al., Molecular Immunology (67): 95-106 (2015) and Kontermann, mAbs 4: 2, 182-197 (2012), both of which are expressly incorporated by reference, particularly with respect to the drawings, figure legends, and references to the formats therein.

[0406] 11. Trifunctional format

[0407] In some embodiments, the bispecific antibodies described herein are in a "trifunctional" format generally described in WO2015 / 184203, which is hereby incorporated by reference in its entirety, particularly for the figures, legends, definitions, and sequences (including the "K helix" and "E helix" sequences) of the "heterodimerization promoting domain" or "HPD". The trifunctional format relies on the use of two different HPDs that associate to form a heterodimeric structure as a component of the structure, see Figure 1K. In this embodiment, the trifunctional format includes "conventional" heavy and light chains (e.g., VH1-CH1-hinge-CH2-CH3 and VL1-CL), a third chain including a first "diabody-like binding domain" or VH2-(linker)-VL3-HPD1, and a fourth chain including a second VH3-(linker)-(linker)-VL2-HPD2. VH1 and VL1 form the first ABD, VH2 and VL2 form the second ABD, and VH3 and VL3 form the third ABD. In some cases, as shown in Figure 1K, the second and third ABDs bind to the same antigen, which is typically ENPP3 in this case, for example bivalently, while the first ABD binds to CD3 monovalently.

[0408] 12. Monospecific monoclonal antibodies

[0409] As will be appreciated by those skilled in the art, the novel Fv sequences outlined herein can also be used for both monospecific antibodies (e.g., "conventional monoclonal antibodies") or non-heterodimeric bispecific formats. Thus, in some embodiments, the antibodies described herein provide monoclonal (monospecific) antibodies that include the 6 CDRs and / or the vh and vl sequences from the figures, typically having an IgG1, IgG2, IgG3, or IgG4 constant region, where IgG1, IgG2, and IgG4 (including the IgG4 constant region with the S228P amino acid substitution) are particularly useful in some embodiments. That is, any sequence with the name "H_L" herein can be linked to the constant region of a human IgG1 antibody.

[0410] In some embodiments, the monospecific antibody is an ENPP3 monospecific antibody. In certain embodiments, the anti-ENPP3 monospecific antibody comprises any 6 CDRs of anti-ENPP3 antibodies selected from: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1 L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8L1.33, 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, Ha16-1(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( Figure 12 , Figures 13A to 13B , and Figures 14A to 14I ).

[0411] H. Antigen-binding domain

[0412] As discussed herein, the subject heterodimeric antibodies comprise two antigen-binding domains (ABDs), each binding to ENPP3 or CD3. As outlined herein, these heterodimeric antibodies can be bispecific and bivalent (e.g., in the form depicted in Figure 15A , where each antigen binds to a single ABD) or bispecific and trivalent (e.g., as depicted in Figure 15B , where one antigen binds to a single ABD and the other antigen binds to two).

[0413] Additionally, generally, one of the ABDs comprises an scFv in the orientation of VH-scFv linker-VL or VL-scFv linker-VH from the N-terminus to the C-terminus as outlined herein. Depending on the form, one or both of the other ABDs are typically Fabs that include a VH domain on one protein chain (usually as a component of the heavy chain) and a VL (usually as a component of the light chain) on another protein chain.

[0414] The present disclosure provides a variety of ABDs that bind to a variety of different checkpoint proteins, as outlined below. As will be appreciated by those skilled in the art, any set of 6 CDRs or VH and VL domains can be in scFv form or Fab form, which is then added to the heavy constant domain and the light constant domain, wherein the heavy constant domain includes a variant (contained in the CH1 domain and the Fc domain). The scFv sequences contained in the sequence table utilize specific charged joints, and as outlined herein, uncharged or other charged joints can be used, including those described in Figure 7.

[0415] Additionally, as discussed above, the numbering used in the sequence listing to identify the CDRs is Kabat, however, different numbering may be used which would alter the amino acid sequences of the CDRs as shown in Table 2.

[0416] For all variable heavy and light domains listed herein, additional variants can be prepared. As outlined herein, in some embodiments, a set of 6 CDRs can have 0, 1, 2, 3, 4 or 5 amino acid modifications (where amino acid substitutions are particularly useful) and changes in the framework regions of the variable heavy and light domains, as long as the framework (except for the CDRs) remains at least about 80%, 85% or 90% identical to the human germline sequence selected from those sequences listed in U.S. Patent No. 7,657,380 in Figure 1, the drawings and legends of which are incorporated herein by reference in their entirety. Thus, for example, the same CDRs as described herein can be combined with different framework sequences from human germline sequences, as long as the framework regions remain at least 80%, 85% or 90% identical to the human germline sequence selected from those sequences listed in U.S. Patent No. 7,657,380 in Figure 1. Alternatively, the CDRs can have amino acid modifications (e.g., 1, 2, 3, 4, or 5 amino acid modifications in a set of CDRs (that is, the CDRs can be modified as long as the total number of changes in a set of 6 CDRs is less than 6 amino acid modifications, wherein any combination of CDRs is altered; for example, there can be one change in VLCDR1, two changes in VHCDR2, no changes in VHCDR3, etc.)), and have framework region changes, as long as the framework regions maintain at least 80%, 85%, or 90% identity to human germline sequences selected from those sequences listed in Figure 1 of U.S. Patent No. 7,657,380.

[0417] 1. ENPP3 antigen binding domain

[0418] In some embodiments, one of the ABDs binds ENPP3. Figure 12 , Figures 13A to 13B and Figures 14A to 14IDepicted therein are suitable sets of 6 CDRs and / or VH and VL domains. In some embodiments, the heterodimeric antibody is an antibody in the form of 1+1 Fab-scFv-Fc or 2+1 Fab2-scFv-Fv (see, for example Figure 15A and Figure 15B ).

[0419] In one embodiment, the ENPP3 antigen-binding domain comprises the 6 CDRs of the ENPP3 ABD described herein, including the accompanying drawings and sequence listing (i.e., vhCDR1-3 and vlCDR1-3). In an illustrative embodiment, the ENPP3 ABD is one of the following ENPP3 ABDs: AN1[ENPP3]H1L1, AN1[ENPP3]H1 L1.33, AN1[ENPP3]H1L1.77, AN1[ENPP3]H1.8 L1, AN1[ENPP3]H1.8 L1.33, AN1[ENPP3]H1L1.77, 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, H16-9.65, Ha1-1(3,5)19 and Ha16-1.80( Figure 12 、 Figures 13A to 13B and Figures 14A to 14I ).

[0420] As will be understood by those skilled in the art, suitable ENPP3-binding domains can include a set of 6 CDRs as depicted in the accompanying drawings or as underlined, or in the case of a different numbering scheme as described herein and shown in Table 2, CDRs identified by other alignments within the VH and VL sequences of the sequences depicted in Figure 12 、 Figures 13A to 13B and Figures 14A to 14I . Suitable ABDs can also contain the entire VH and VL sequences as depicted as scFv or Fab in these sequences and the accompanying drawings. In many embodiments herein containing Fv to ENPP3, the Fab monomer binds to ENPP3.

[0421] In addition to the parental CDR sets forming ABD to ENPP3 disclosed in the figures and sequence listing, the present disclosure provides variant CDR sets. In one embodiment, a set of 6 CDRs can have 1, 2, 3, 4, or 5 amino acid changes from the parental CDRs, provided that the ENPP3 ABD is still capable of binding to the target antigen, as measured by at least one of the assays such as Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), with the latter assay being particularly useful in many embodiments.

[0422] In addition to the parental variable heavy chain domain and variable light domain forming ABD to ENPP3 disclosed herein, the present disclosure provides variant VH and VL domains. In one embodiment, each of the variant VH and VL domains can have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental VH and VL domains, provided that the ABD is still capable of binding to the target antigen, as measured by at least one of the assays such as Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), with the latter assay being particularly useful in many embodiments. In another embodiment, the variant VH and VL are at least 90%, 95%, 97%, 98%, or 99% identical to the corresponding parental VH or VL, provided that the ABD is still capable of binding to the target antigen, as measured by at least one of the assays such as Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), with the latter assay being particularly useful in many embodiments.

[0423] 2. CD3 antigen-binding domain

[0424] In some embodiments, one of the ABDs binds to CD3. Figures 10A to 10F And suitable sets of 6 CDRs and / or VH and VL domains and scFv sequences are depicted in the sequence listing. Particularly used CD3-binding domain sequences include but are not limited to anti-CD3 H1.30_L1.47, anti-CD3 H1.32, anti-CD3 L1.47, anti-CD3 H1.89_L1.47, anti-CD3 H1.90_L1.47, anti-CD3 H1.33_L1.47, anti-CD3 H1.31_L1.47, anti-CD3 L1.47_H1.30, anti-CD3 L1.47_H1.30, anti-CD3 L1.47_H1.32, anti-CD3 L1.47_H1.89, anti-CD3 L1.47_H1.90, anti-CD3 L1.47_H1.33, and anti-CD3 L1.47_H1.31, as Figures 10A to 10F shown.

[0425] As will be understood by those skilled in the art, suitable CD3 binding domains can include, for example, a set of 6 CDRs as depicted in Figures 10A to 10F or as underlined therein, or in the case of using a different numbering scheme as described herein and shown in Table 2, CDRs identified by other alignments within the VH and VL sequences of the sequences depicted in Figures 10A to 10F Suitable ABDs can also contain the entire VH and VL sequences as used for scFv or Fab as depicted in these sequences and the figures. In many embodiments herein containing Fv to CD3, the scFv monomer binds to CD3.

[0426] In addition to the parental CDR sets forming ABD to CD3 disclosed in the figures and sequence listing, the present disclosure provides variant CDR sets. In one embodiment, a set of 6 CDRs can have 1, 2, 3, 4, or 5 amino acid changes from the parental CDRs, provided that the CD3 ABD is still capable of binding to the target antigen, as measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), the latter assay being particularly useful in many embodiments.

[0427] In addition to the parental variable heavy chain domain and variable light domain forming ABD to CD3 disclosed herein, the present disclosure provides variant VH and VL domains. In one embodiment, each of the variant VH and VL domains can have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental VH and VL domains, provided that the ABD is still capable of binding to the target antigen, as measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), the latter assay being particularly useful in many embodiments. In another embodiment, the variant VH and VL are at least 90%, 95%, 97%, 98%, or 99% identical to the corresponding parental VH or VL, provided that the ABD is still capable of binding to the target antigen, as measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., Octet assay), the latter assay being particularly useful in many embodiments.

[0428] VI. SSTR2 Binding Domain

[0429] In one aspect, provided herein are somatostatin receptor 2 (SSTR2) antigen-binding domains (ABDs) and compositions comprising such SSTR2 antigen-binding domains (ABDs), including anti-SSTR2 antibodies.

[0430] Somatostatin receptors (SSTRs) belong to the superfamily of G protein-coupled receptors (GPCRs), each of which contains a single polypeptide chain composed of extracellular / intracellular domains and seven transmembrane domains. SSTRs are highly expressed in various cultured tumor cells and primary tumor tissues, including NETs (lung cancer, gastrointestinal cancer (GI), pancreatic cancer, pituitary cancer, medullary carcinoma, prostate cancer, pancreatic lung cancer, osteosarcoma, etc.) as well as non-NETs (breast cancer, lung cancer, colorectal cancer, ovarian cancer, cervical cancer, etc.) (Reubi., 2003, Endocr. Rev. 24:389-427; Volante et al., 2008, Mol. Cell. Endocrinol. 286:219-229; and Schulz et al., 2003, Gynecol. Oncol. 89:385-390). To date, five SSTR receptor subtypes have been identified (Patel et al., 1997, Trends Endocrinol. Metab. 8:398-405). In particular, SSTR2 is highly expressed at high concentrations on many tumor cells (Volante et al., 2008, Mol. Cell. Endocrinol. 286:219-229; and Reubi et al., 2003, Eur. J. Nucl. Med. Mol. Imaging 30:781-793), making it a candidate target antigen for bispecific antibody cancer targeting therapies. Given the high concentration of SSTR2 expressed on various tumors, it is believed that anti-SSTR2 antibodies can be used, for example, to localize anti-tumor therapeutic agents (e.g., chemotherapeutic agents and T cells) to such SSTR2-expressing tumors.

[0431] The subject antibodies (e.g., anti-SSTR2 x anti-CD3 bispecific antibodies) comprising the SSTR2 antigen-binding domains provided herein advantageously elicit a range of different immune responses. Such SSTR2-binding domains and related antibodies can be used, for example, to treat SSTR2-related cancers.

[0432] As will be understood by those skilled in the art, suitable SSTR2-binding domains can include a set of 6 CDRs as depicted in Figure 63 or as underlined, or in the case of using a different numbering scheme as described herein and shown in Table 2, using the ones in Figure 63Other CDRs identified by alignment within the VH and VL sequences of those sequences depicted. Suitable ABDs can also include the entire VH and VL sequences used as scFv or Fab as depicted in such sequences and the figures. In many embodiments herein containing Fv to SSTR2, the Fab monomer binds to SSTR2. In one embodiment, the SSTR2 antigen-binding domain comprises 6 CDRs of [αSSTR2]H1.24_L1.30 (i.e., vhCDR1-3 and vlCDR1-3)( Figure 63 ).

[0433] In addition to the parental CDR set forming the ABD against SSTR2 disclosed in the figures and sequence listing, SSTR2 ABD variants with CDRs are provided herein, said CDRs comprising at least one modification of the SSTR2 ABD CDRs disclosed herein. In one embodiment, compared to the 6 CDRs of the SSTR2 ABD depicted in this document, including the figures and sequence listing, the SSTR2 ABD comprises a set of 6 CDRs with 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 amino acid modifications. In an exemplary embodiment, compared to the 6 CDRs of [αSSTR2]H1.24_L1.30, the SSTR2 ABD comprises a set of 6 CDRs with 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 amino acid modifications( Figure 63 ). In certain embodiments, the variant SSTR2 ABD is capable of binding the SSTR2 antigen as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., octet assay), the latter assay being particularly useful in many embodiments. In a particular embodiment, the SSTR2 ABD is capable of binding the human SSTR2 antigen.

[0434] In one embodiment, the SSTR2 ABD comprises 6 CDRs having at least 90%, 95%, 97%, 98%, or 99% identity to the 6 CDRs of the SSTR2 ABD as described herein, including the figures and sequence listing. In an illustrative embodiment, the SSTR2 ABD comprises 6 CDRs having at least 90%, 95%, 97%, 98%, or 99% identity to the 6 CDRs of [αSSTR2]H1.24_L1.30( Figure 63 ). In certain embodiments, the SSTR2 ABD is capable of binding the SSTR2 antigen as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., octet assay), the latter assay being particularly useful in many embodiments. In a particular embodiment, the SSTR2 ABD is capable of binding the human SSTR2 antigen.

[0435] In another exemplary embodiment, the SSTR2 ABD comprises any one of the variable heavy chain (VH) domain and variable light chain (VL) domain of the ENPP3 ABD described herein, including the drawings and sequence listing. In an exemplary embodiment, the SSTR2 ABD is [αSSTR2]H1.24_L1.30( Figure 63 ).

[0436] In addition to the parental SSTR2 variable heavy and variable light domains disclosed herein, provided herein are SSTR2 ABDs comprising a variable heavy chain domain and / or a variable light chain domain, wherein the variable heavy chain domain and / or variable light chain domain is a variant of the SSTR2 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 compared to the VH and / or VL domains of the SSTR2 ABD described herein, including the drawings and sequence listing. 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 relative to the VH and / or VL domains of [αSSTR2]H1.24_L1.30( Figure 63 ). In certain embodiments, the SSTR2 ABD is capable of binding the SSTR2 as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., octet assay), with the latter assay being particularly useful in many embodiments. In a particular embodiment, the SSTR2 ABD is capable of binding to the human SSTR2 antigen.

[0437] In one embodiment, the variant VH and / or VL domain is at least 90%, 95%, 97%, 98%, or 99% identical to the VH and / or VL of the SSTR2 ABD as described herein, including in the drawings and sequence listing. In an exemplary embodiment, the variant VH and / or VL domain is at least 90%, 95%, 97%, 98%, or 99% identical to the VH and / or VL of [αSSTR2]H1.24_L1.30( Figure 63 ). In certain embodiments, the SSTR2 ABD is capable of binding the SSTR2 as determined by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (Biolayer Interferometry, e.g., octet assay), with the latter assay being particularly useful in many embodiments. In a particular embodiment, the SSTR2 ABD is capable of binding to the human SSTR2 antigen.

[0438] In some embodiments, the subject antibodies described herein comprise at least one SSTR2 binding domain. In certain embodiments, the antibody is a heterodimeric antibody. In some embodiments, the heterodimeric antibody is an antibody in the form of 1+1 Fab-scFv-Fc or 2+1 Fab2-scFv-Fv (see, for example, Figure 15A and Figure 15B ). Such heterodimeric antibodies can comprise any of the Fc variant amino acid substitutions provided herein independently or in combination (e.g., skewed, pI, and ablation variants, including those depicted in FIGS. 1 to Figure 4 ). Particularly useful skewed variants are selected from the group consisting of: 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), optionally ablation variants (including those shown in Figure 3 ), optionally charged scFv linkers (including those shown in Figure 5 ), and the heavy chain comprises pI variants (including those shown in Figure 2 ).

[0439] A. Useful Embodiments

[0440] Useful embodiments comprise the 1+1 Fab-scFv-Fc form, which includes: a) a first monomer (“scFv monomer”) that includes a charged scFv linker ( Figure 5 's +H sequence is preferred in some embodiments), the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and an scFv that binds to CD3 as outlined herein; b) a second monomer (“Fab monomer”) that includes the skewed variant L368D / K370S, the pI variants N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain; and c) a light chain that comprises a variable light chain domain (VL) and a constant light chain domain (CL), where the numbering is according to EU numbering. In some embodiments, the variable heavy chain domain and the variable light chain domain constitute the ENPP3 binding portion.

[0441] Other useful embodiments include the 1+1 Fab-scFv-Fc format, which includes: a) a first monomer (“scFv monomer”), the first monomer including a charged scFv linker ( Figure 5 The +H sequence of which is preferred in some embodiments), the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and an scFv that binds to CD3 as outlined herein; b) a second monomer (“Fab monomer”), the second monomer including the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain; and c) a light chain, the light chain including a variable light chain domain (VL) and a constant light chain domain (CL), where the numbering is according to EU numbering. In some embodiments, the variable heavy chain domain and the variable light chain domain constitute the SSTR2 binding portion.

[0442] Other useful embodiments include the 2+1 Fab2-scFv-Fc format, which includes: a) a first monomer (Fab-scFv-Fc side), the first monomer including the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light chain domain of a shared light chain constituting an Fv that binds to ENPP3 as outlined herein and an scFv domain that binds to CD3; b) a second monomer (Fab-Fc side), the second monomer including the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light chain domain of a shared light chain constituting an Fv that binds to ENPP3 as outlined herein; and c) a light chain, the light chain including a variable light domain and a constant light domain, where the numbering is according to EU numbering. In some embodiments, the shared light chain and the variable heavy chain domain on each monomer form the ENPP3 binding domain.

[0443] Other useful embodiments include the 2+1 Fab2-scFv-Fc format, which includes: a) a first monomer (Fab-scFv-Fc side), the first monomer including the skewed variant S364K / E357Q, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light chain domain of the common light chain constituting an Fv that binds to SSTR2 as outlined herein and an scFv domain that binds to CD3; b) a second monomer (Fab-Fc side), the second monomer including the skewed variant L368D / K370S, the pI variant N208D / Q295E / N384D / Q418E / N421D, the ablation variant E233P / L234V / L235A / G236del / S267K, and a variable heavy chain domain, the variable heavy chain domain and the variable light chain domain of the common light chain constituting an Fv that binds to SSTR2 as outlined herein; and c) a light chain, the light chain including a variable light domain and a constant light domain, wherein the numbering is according to EU numbering. In some embodiments, the common light chain and the variable heavy chain domains on each monomer form an SSTR2 binding domain (e.g., [αSSTR2]H1.24_L1.30( Figure 63 ))

[0444] Some useful embodiments include: XENP24804, XENP26820, XENP28287, XENP28925, XENP29516, XENP30262, XENP26821, XENP29436, XENP28390, XENP29463, and XENP30263

[0445] Other useful embodiments include: XENP29437, XENP29520, XENP30264, XENP26822, XENP28438, XENP29438, XENP29467, XENP30469, XENP30470, XENP30819, XENP30821, XENP31148, XENP31149, XENP31150, XENP31419, and XENP31471

[0446] Another useful embodiment is XENP30458

[0447] VII. Nucleic Acids of the Invention

[0448] The present disclosure also provides nucleic acid compositions encoding the anti-ENPP3 antibodies provided herein, the anti-ENPP3 antibodies including, but not limited to, anti-ENPP3 x anti-CD3 bispecific antibodies and ENPP3 monospecific antibodies

[0449] As will be understood by those skilled in the art, the nucleic acid compositions will depend on the form and scaffold of the heterodimeric protein. Thus, for example, when the form requires three amino acid sequences (such as the 1+1 Fab-scFv-Fc form (e.g., the first amino acid monomer comprises an Fc domain and an scFv, and the second amino acid monomer comprises a heavy chain and a light chain)), three nucleic acid sequences can be incorporated into one or more expression vectors for expression. Similarly, some forms (e.g., the dual scFv form disclosed in FIG. 1) require only two nucleic acids; likewise, these can be placed in one or two expression vectors.

[0450] As is known in the art, depending on the host cell used to generate the heterodimeric antibodies described herein, the nucleic acids encoding the components of the antibodies described herein can be incorporated into expression vectors known in the art. Generally, the nucleic acids are operably linked to any number of regulatory elements (promoters, origins of replication, selectable markers, ribosome binding sites, inducers, etc.). The expression vectors can be extrachromosomal or integrated vectors.

[0451] The nucleic acids and / or expression vectors of the antibodies described herein are then transformed into any number of different types of host cells well known in the art, including mammalian, bacterial, yeast, insect, and / or fungal cells, with mammalian cells (e.g., CHO cells) being used in many embodiments.

[0452] In some embodiments, if applicable depending on the form, the nucleic acids encoding each monomer and the optional nucleic acid encoding the light chain are generally each contained within a single expression vector under different or the same promoter control conditions. In embodiments particularly for the antibodies described herein, each of these two or three nucleic acids is contained on a different expression vector. As shown herein and in 62 / 025,931 (which is hereby incorporated by reference), different vector ratios can be utilized to drive heterodimer formation. That is, surprisingly, although the protein comprises a 1:1:2 ratio of first monomer:second monomer:light chain (in the case of many embodiments herein having three polypeptides including the heterodimeric antibody), these are not the ratios that give the best results.

[0453] The heterodimeric antibodies described herein are prepared by culturing host cells comprising one or more expression vectors well known in the art. Once produced, conventional antibody purification steps are performed, including an ion exchange chromatography step. As discussed herein, a difference in pI of at least 0.5 between the two monomers can allow separation by ion exchange chromatography or isoelectric focusing or other methods sensitive to the isoelectric point. That is, pI substitutions that involve altering the isoelectric point (pI) of each monomer such that each monomer has a different pI and the heterodimer also has a different pI facilitate isoelectric purification of "1+1Fab-scFv-Fc" and "2+1" heterodimers (e.g., anion exchange column, cation exchange column). These substitutions also aid in the determination and monitoring of any contaminating dual scFv-Fc and purified mAb homodimers (e.g., IEF gel, cIEF, and analytical IEX column).

[0454] VIII. Biological and biochemical functions of heterodimeric bispecific antibodies

[0455] Generally, the bispecific ENPP3 x CD3 antibodies described herein are administered to a patient suffering from cancer and the efficacy is evaluated in a variety of ways as described herein. Thus, while standard efficacy assays such as assessment of cancer burden, tumor size, presence or extent of metastasis, etc. can be performed, immuno-oncology therapies can also be evaluated based on immunological status assessments. This can be done in a variety of ways, including in vitro and in vivo analyses.

[0456] IX. Treatment

[0457] Once prepared, the compositions of the antibodies described herein can be used in many applications. ENPP3 is highly expressed in renal cell carcinoma, and thus, the heterodimeric compositions of the antibodies described herein can be used to treat such ENPP3-positive cancers.

[0458] X. Antibody compositions for in vivo administration

[0459] Antibody formulations for use with the antibodies described herein are prepared by mixing an antibody having the desired degree of purity with optional pharmaceutically acceptable carriers, excipients, or stabilizers (Remington’s Pharmaceutical Sciences, 16th edition, Osol, A. Ed.

[1980] ) for storage in the form of a lyophilized formulation or an aqueous solution. Acceptable carriers, excipients, or stabilizers are non-toxic to recipients at the dosages and concentrations employed, and include buffers such as phosphate, citrate, and other organic acids; antioxidants including ascorbic acid and methionine; preservatives (such as octadecyl dimethyl benzyl ammonium chloride; hexamethonium chloride; benzalkonium chloride, benzethonium chloride; phenol, butanol, or benzyl alcohol; alkyl parabens such as methylparaben or propylparaben; catechol; resorcinol; cyclohexanol; 3-pentanol; and m-cresol); low molecular weight (less than about 10 residues) polypeptides; proteins such as serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, histidine, arginine, or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrin; chelating agents such as EDTA; sugars such as sucrose, mannitol, trehalose, or sorbitol; salt-forming counterions such as sodium; metal complexes (e.g., zinc-protein complexes); and / or nonionic surfactants such as TWEEN TM , PLURONICS TM or polyethylene glycol (PEG).

[0460] XI. Mode of Administration

[0461] The antibodies and chemotherapeutic agents described herein are administered to a subject according to known methods, such as intravenous administration (as a bolus or by continuous infusion over a period of time).

[0462] XII. Mode of Treatment

[0463] In the methods described herein, a therapy is used to provide a positive treatment response to a disease or disorder. “Positive treatment response” is intended to mean an improvement in the disease or disorder and / or an improvement in the symptoms associated with the disease or disorder. For example, a positive treatment response will refer to one or more of the following improvements in a disease: (1) a decrease in the number of neoplastic cells; (2) an increase in neoplastic cell death; (3) inhibition of neoplastic cell survival; (5) inhibition (i.e., slowing to some extent and preferably stopping) of tumor growth; (6) an increase in patient survival rate; and (7) some alleviation of one or more symptoms associated with the disease or disorder.

[0464] A positive treatment response for any given disease or disorder can be determined by standardized response criteria specific to that disease or disorder. Screening techniques such as magnetic resonance imaging (MRI) scans, x-ray imaging, computed tomography (CT) scans, bone scan imaging, endoscopy, and tumor biopsy sampling, including bone marrow aspiration (BMA) and circulating tumor cell counting, can be used to estimate tumor response in terms of changes in tumor morphology (i.e., overall tumor burden, tumor size, etc.).

[0465] In addition to these positive treatment responses, a subject undergoing therapy may experience the beneficial effect of improvement in symptoms associated with the disease.

[0466] Treatment according to the present disclosure involves the use of a “therapeutically effective amount” of a drug. A “therapeutically effective amount” refers to an effective amount, in terms of dosage and over a period of time required, to achieve the desired therapeutic result.

[0467] The therapeutically effective amount can vary depending on factors such as disease state, age, gender, and individual body weight, as well as the ability of the drug to elicit the desired response in the individual. The therapeutically effective amount is also an amount where any toxic or detrimental effects of the antibody or antibody portion exceed the therapeutic beneficial effects.

[0468] The “therapeutically effective amount” for tumor treatment can also be measured by its ability to stabilize disease progression. The ability of a compound to inhibit cancer can be evaluated in an animal model system predictive of efficacy in human tumors.

[0469] Alternatively, this property of the composition can be evaluated by in vitro assays known to a skilled practitioner by examining the ability of the compound to inhibit tumor growth or induce apoptosis. A therapeutically effective amount of a therapeutic compound can reduce tumor size or otherwise alleviate the symptoms of the subject. One of ordinary skill in the art will be able to determine such amount based on factors such as the size of the subject, the severity of the subject's symptoms, and the particular composition or route of administration selected.

[0470] Adjust the dosage regimen to provide the optimal desired response (e.g., treatment response). For example, a single bolus can be administered, or several fractional doses can be administered over time, or the dose can be proportionally decreased or increased as indicated by the exigencies of the treatment situation. For convenience in administration and uniformity of dosage, parenteral compositions can be dispensed in dosage units. As used herein, a dosage unit refers to a physically discrete unit suitable as a single dose for a subject to be treated; each unit contains a predetermined quantity of the active compound associated with the required pharmaceutical carrier calculated to produce the desired therapeutic effect.

[0471] The specifications of the dosage unit forms of the present disclosure are subject to and directly depend on: (a) the unique properties of the active compound and the particular therapeutic effect to be achieved, and (b) the inherent limitations in the art due to individual sensitivity to such active compounds used in therapy.

[0472] The effective dosage and dosage regimen for the bispecific antibodies described herein depend on the disease or disorder to be treated and can be determined by those skilled in the art.

[0473] An exemplary non-limiting range of a therapeutically effective amount of the bispecific antibody in the antibodies described herein is about 0.1 - 100 mg / kg.

[0474] All cited references are hereby expressly incorporated by reference in their entirety.

[0475] Although the specific embodiments of the present disclosure have been described above for purposes of illustration, those skilled in the art will appreciate that many variations in the details can be made without departing from the invention described in the appended claims.

[0476] Examples

[0477] Examples are provided below to illustrate the antibodies described herein. These examples are not intended to limit the antibodies described herein to any particular application or theory of operation. For all constant region positions discussed in the antibodies described herein, numbering is according to the EU index as in Kabat (Kabat et al., 1991, Sequences of Proteins of Immunological Interest, 5th ed., United States Public Health Service, National Institutes of Health, Bethesda, which is hereby incorporated by reference in its entirety). Those skilled in the art of antibodies will understand that this convention consists of non-sequential numbering in specific regions of immunoglobulin sequences, which enables standardized reference to conserved positions in immunoglobulin families. Thus, the positions of any given immunoglobulin as defined by the EU index do not necessarily correspond to its sequential sequence.

[0478] General and specific scientific techniques are outlined in U.S. Publications 2015 / 0307629, 2014 / 0288275, and WO2014 / 145806, each of which is hereby expressly incorporated by reference in its entirety and particularly for the techniques outlined therein.

[0479] Example 1: Binding Domain

[0480] 1A: CD3 Binding Domain

[0481] The sequences of the CD3-binding domains with different CD3-binding affinities are shown in Figure 10.

[0482] 1B: ENPP3 Binding Domain

[0483] 1B(a): ENPP3 Binding Domain AN1

[0484] The variable region of the murine ENPP3-binding domain was humanized using codon content optimization (see, e.g., U.S. Patent No. 7,657,380, issued February 2, 2010). The sequence of the humanized ENPP3-binding domain, herein referred to as AN1, is as Figures 10A to 10F shown.

[0485] AN1 variants were engineered for improved purification (in the context of an αENPP3 x αCD3 bispecific antibody) and to modulate ENPP3-binding affinity / avidity. The sequences of illustrative such variants are depicted in Figure 13.

[0486] 1B(b): Additional ENPP3 Binding Domain

[0487] The sequences of additional ENPP3-binding domains that can be used in the αENPP3 x αCD3 bispecific antibodies described herein are depicted in Figure 14.

[0488] Example 2: Engineering and Producing αENPP3 x αCD3 Bispecific Antibody

[0489] Multiple forms of the αENPP3 x αCD3 bispecific antibody (bsAb) are envisioned, and illustrative forms are outlined below and in Figure 15.

[0490] One such form is the 1+1 Fab-scFv-Fc form, which comprises a single-chain Fv ("scFv") covalently attached to a first heterodimeric Fc domain, a variable heavy chain region (VH) covalently attached to a complementary second heterodimeric Fc domain, and a light chain (LC), which is transfected separately such that a Fab domain is formed together with the variable heavy chain domain.

[0491] Another form is the 2+1 Fab2-scFv-Fc form, which comprises a VH domain covalently attached to a CH1 domain, the CH1 domain covalently attached to an scFv, the scFv covalently attached to a first heterodimeric Fc domain (VH-CH1-scFv-Fc); a VH domain covalently attached to a complementary second heterodimeric Fc domain; and an LC, which is transfected separately such that a Fab domain is formed together with the VH domain.

[0492] The DNA coding strand of αENPP3 xαCD3 bsAb was generated by performing standard gene synthesis followed by isothermal cloning (Gibson assembly) or subcloning into the pTT5 expression vector containing fusion partners (such as the domain linker depicted in Figure 6 and / or the backbone depicted in Figures 7 to Figure 9 . The DNA was transfected into HEK293E cells for expression. The sequences of illustrative αENPP3 xαCD3 bsAbs in the 1+1 Fab-scFv-Fc form and 2+1 Fab2-scFv-Fc form (based on the binding domains described in Example 1) are depicted in Figures 17 to 23, respectively.

[0493] Example 3: αENPP3 x αCD3 bsAb Redirecting T Cells to Destroy ENPP3-Expressing Cells

[0494] The prototype αENPP3 xαCD3 bsAb in the 1+1 Fab-scFv-Fc form was engineered using the binding domains described in Example 1. Specifically, the sequences of XENP26820 (containing ENPP3 binding domain clone H16-7.8 and CD3 high scFv), XENP26821 (containing ENPP3 binding domain clone H16-7.8 and CD3 high-Int number 1 scFv), XENP28287 (containing ENPP3 binding domain clone AN1 and CD3 scFv), and XENP28390 (containing ENPP3 binding domain clone AN1 and CD3 high Int number 1 scFv) are depicted in Figures 17 and 18. XENP13245 (containing an RSV binding domain based on motavizumab and anti-CD3-high; the sequence is depicted in Figure 16 ) was used as a control.

[0495] The potential of the prototype αENPP3 xαCD3 bispecific antibody (bsAb) to redirect CD3 + effector T cells to destroy ENPP3-expressing cell lines was investigated. In the first experiment, KU812 (ENPP3 高 basophilic leukemia cell line) cells were incubated with human PBMC (effector-to-target cell ratio of 10:1) and the indicated concentrations of the test articles as shown above at 37 °C for 24 hours. After incubation, the cells were stained with Aqua Zombie dye at room temperature for 15 minutes. Then the cells were washed and stained with antibodies against cell surface markers and analyzed by flow cytometry. Two different methods were used to study the induction of redirected T cell cytotoxicity (RTCC): a) reduction in the number of CSFE+ target cells (the data of which is depicted in Figure 24A ), and b) Zombie Aqua staining of CSFE+ target cells (the data of which is depicted in Figure 24B ). CD4+ and CD8 + Activation and degranulation of T cells were also determined based on CD107a, CD25, and CD69 expression (the data are depicted in Figures 25 to 26).

[0496] In a second experiment, RXF393 (a clinically relevant renal cell carcinoma cell line expressing ENPP3) cells were incubated with human PBMC (effector-to-target cell ratio of 20:1) and the indicated concentrations of the prototype test articles shown above at 37 °C for 24 hours. After incubation, the cells were stained with Aqua Zombie dye for 15 minutes at room temperature. The cells were then washed and stained with antibodies against cell surface markers and analyzed by flow cytometry. As described above, two different methods were used to study the induction of RTCC: a) reduction in the number of CSFE+ target cells (the data are depicted in Figure 27A ), and b) Zombie Aqua staining of CSFE+ target cells (the data are depicted in Figure 27B ). CD4 + and CD8 + Activation and degranulation of T cells were also determined based on CD107a, CD25, and CD69 expression (the data are depicted in Figures 28 to 29).

[0497] Overall, the data show that the prototype αENPP3 x αCD3 bsAb dose-dependently induces RTC on ENPP3 cells; CD3 binding affinity correlates with RTCC potency (i.e., bsAbs with high CD3 are more effective in inducing RTCC than bsAbs with high-CD3-Int number 1); and bsAbs with an AN1-based binding domain are more effective in inducing RTCC than bsAbs with an H16-7.8-based binding domain. Consistent with the RTCC data, the αENPP3 x αCD3 bsAb dose-dependently induces activation of T cells; CD3 binding affinity correlates with activation potency (i.e., bsAbs with high CD3 are more effective in inducing T cell activation than bsAbs with high-CD3-Int number 1); and bsAbs with an AN1 binding domain are more effective in inducing T cell activation than bsAbs with an H16-7.8 binding domain.

[0498] Example 4: Increasing the Yield of αENPP3 x αCD3

[0499] Generally, bispecific antibodies are produced by transient transfection in HEK293E cells and purified by a two-step purification process including protein A chromatography (purification step 1) followed by ion exchange chromatography (purification step 2).

[0500] 4A: Engineering the AN1 variant to improve yield

[0501] 4A(a): Production of XENP28287 results in a homogeneous population containing aggregates and unpaired monomers

[0502] Purify XENP28287 from HEK293E supernatant as described above. Figure 30 A depicts a chromatogram showing purification fraction 2 of XENP28287 (cation exchange chromatography following protein A chromatography). The chromatogram shows the separation of two peaks (peak B and peak BC), and the identity, purity, and homogeneity of the peaks are further characterized by analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS) and analytical cation exchange chromatography (aCIEX) as generally described below.

[0503] Peaks B and BC isolated from purification fraction 2 of XENP28287 (and pre-purified material) were analyzed using aSEC-MALS to infer their component protein species. The analysis was performed on an Agilent 1200 high performance liquid chromatography (HPLC) system. Using a UV detection wavelength of 280 nM, samples were injected at a rate of 1.0 mL / min at 4 °C onto a Superdex TM 200 10 / 300GL column (GE Healthcare LifeSciences) for 25 minutes. On a T-rEX refractive index detector (Wyatt Technology, Santa Barbara, Cali.) and MALS was performed. Analysis was carried out using Agilent OpenLab Chromatography Data System (CDS) ChemStation Edition AICC.01.07 and ASTRA 6.1.7.15. Chromatograms depicting the SEC separation profiles of the pre-purified material, peak B, and peak BC, and the approximate MW of the component species determined by MALS are depicted in Figure 30 B. The profile shows that peak B contains a major species of approximately 126 kDa, which is consistent with the calculated molecular weight of the XENP28287 heterodimer (based on the amino acid sequence), but also contains a contaminant species of 75 kDa (presumably a monomer). Peak BC contains peaks with species of 308 kDa (presumably an aggregate), 121 kDa (XENP28287), and 82 kDa (contaminant monomer). Notably, the separation profile of the pre-purified material indicates that less than 85% of the material is the bispecific antibody heterodimer.

[0504] The peaks from purified fraction 2 were also analyzed using analytical CIEX to further assess the purity and homogeneity of peaks B and BC. The analysis was performed on an Agilent 1200 high performance liquid chromatography (HPLC) system. Samples were injected onto a Proteomix SCX-NP5 5 μM non-porous column (Sepax Technologies, Inc., Newark, Del.) at 1.0 mL / min using a 0 - 40% NaCl gradient in 20 mM MES, pH 6.0 buffer with a UV detection wavelength of 280 nM. The analysis was carried out using Agilent OpenLAB CDS ChemStation Edition AIC C.01.07 version. Figure 30 The chromatogram depicting the CIEX separation of peaks B and BC is shown in C. Notably, the aCIEX separation indicates that in the peak BC material, in addition to the main peak, there are many charge variants.

[0505] 4A(b): The AN1 VH variant H1.8 enables improved purification

[0506] Numerous AN1 variable heavy chain (VH) domains were designed with the aim of increasing the production of bispecific antibodies. A specific VH variant (H1.8; SEQ ID NO: XXX; also depicted in Figure 13) enabled improved separation of the bispecific antibody heterodimer from contaminant species. To illustrate this, XENP28925 (which contains an ENPP3 binding domain with the AN1 H1.8 VH variant; the sequence is depicted in Figure 17) was produced and purified from HEK293E supernatant as described above. Figure 31 A depicts the chromatogram showing purified fraction 2 of XENP28925 (cation exchange chromatography after protein A chromatography). The chromatogram shows the separation of a main peak (peak B), which was further characterized by aSEC-MALS and aCIEX as described above to determine identity, purity, and homogeneity.

[0507] Figure 31 B to Figure 31 C depicts the chromatograms showing the aSEC separation profiles of the pre-purified material and peak B (where the MW of the component species was determined by MALS) and the aCIEX separation profile of peak B. The profiles show that peak B contains a major species of approximately 128 kDa, which is consistent with the calculated molecular weight of the XENP28925 heterodimer (based on the amino acid sequence). Notably, the separation curve of the pre-purified material shows that greater than 97% of this material is the bispecific antibody heterodimer.

[0508] Overall, this indicates that the AN1 H1.8 VH variant enables an increase in the production of the αENPP3 xαCD3 heterodimer and an improvement in the separation of the heterodimer from contaminating species.

[0509] 4B: Engineering the backbone of the 2+1 Fab2-scFv-Fc bispecific format to improve yield

[0510] 4B(a): Production of XENP31419 results in a protein population skewed towards VH-Fc homodimers

[0511] XENP31149 (αENPP3 xαCD3 bsAb in 2+1Fab2-scFv-Fc form; sequence depicted in Figure 23) was purified from HEK293E supernatant as described above. Figure 32 A depicts a chromatogram showing the purification fraction 2 of XENP31149 (cation exchange chromatography after protein A chromatography). The chromatogram shows the separation of two peaks (peak A and peak B), which was further characterized by analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS) to obtain the identity, as generally described above.

[0512] Figure 32 The chromatogram of the aSEC separation curves of peak A and peak B and the MW of the component species determined by MALS are depicted in B. The spectrum shows that the main peak A contains a species with a molecular weight of 148.4 kDa, which is consistent with the calculated molecular weight of the VH-Fc homodimer, while the minor peak B contains a species with a molecular weight of 173.9 kDa, which is consistent with the calculated molecular weight of the XENP31149 heterodimer. Thus, XENP31149 was produced at a very low titer of 12.4 mg / L.

[0513] 4B(b): Engineering a full hinge in the Fab-scFv-Fc chain improves 2+1 Fab2-scFv-Fc heterodimer yield

[0514] Various methods for increasing the production of the 2+1Fab2-scFv-Fc heterodimer were investigated, including changing the linker between VH and scFv or between scFv and CH2 in the Fab-scFv-Fc chain. XENP31419 (sequence described in Figure 23) was engineered into a XENP31149 counterpart that has a full hinge (EPKSCDKTHTCPPCP; SEQ ID NO: XX) instead of a flexible half-hinge (GGGGSGGGGSKTHTCPPCP; SEQ ID NO: XX) between scFv and the CH2 region in the Fab-scFv-Fc chain. As described above, XENP31419 was produced and purified from HEK293E supernatant. Figure 33Panel A depicts a chromatogram showing the purified fraction 2 of XENP31419 (cation exchange chromatography following protein A chromatography). The chromatogram shows the separation of two peaks (minor peak A and major peak B), which was further characterized by analytical size exclusion chromatography with multi-angle light scattering (aSEC-MALS) to obtain the identity, as generally described above.

[0515] Figure 33 Panel B depicts a chromatogram of the aSEC separation profiles of peak A and peak B and the MW of the component species determined by MALS. The spectra show that minor peak A contains a species with a molecular weight of 152.2 kDa, which is consistent with the calculated molecular weight of the VH-Fc homodimer, while major peak B contains a species with a molecular weight of 180 kDa, which is consistent with the calculated molecular weight of the XENP31419 heterodimer. Thus, production yielded a significantly improved titer of 107.8 mg / L of the XENP31419 heterodimer.

[0516] Example 5: Tuning the αENPP3 xαCD3 bsAb to improve selectivity and therapeutic index

[0517] The following experiments generally use KU812 as ENPP3 高 Target cells (as ENPP3 + substitute for tumor cells) or RCC4 as ENPP3 低 target cells (as substitute for cells outside the tumor environment) were performed. The target cells were incubated with human PBMCs and the test article at the indicated effector-to-target ratio at 37 °C. After incubation, the cells were stained with AquaZombie dye for 15 minutes at room temperature. The cells were then washed and stained with antibodies against cell surface markers and analyzed by flow cytometry. The induction of RTCC was determined using Zombie Aqua staining of CSFE+ target cells; and the activation and degranulation of T cells were determined by the expression of CD107a, CD25, and CD69 on lymphocytes. It should also be noted that some of the data sets in the data set are from the same experiment because several engineering methods were explored simultaneously.

[0518] To study the on-target / off-tumor killing potential of the prototype 1+1 Fab-scFv-Fc bispecific antibody with high-affinity CD3 binding and high-affinity ENPP3 binding, KU812 cells and RCC4 cells were incubated with human PBMCs (effector-to-target ratio of 10:1) and the indicated concentrations of XENP28925 at 37 °C for 18 hours. Figure 34 The data depicted in Panel show t...

Claims

1. A composition, the composition comprising an ectonucleotide pyrophosphatase / phosphodiesterase family member 3 (ENPP3) binding domain, the ENPP3 binding domain consisting of a variable heavy chain domain and a variable light chain domain, wherein the variable heavy chain domain has variable heavy chain complementarity determining regions 1-3 (vhCDR1-3), and the variable light chain domain has variable light chain complementarity determining regions (vlCDR1-3), the vhCDR1-3 and vlCDR1-3 being selected from the group consisting of: a. vhCDR1 of SEQ ID NO:253, vhCDR2 of SEQ ID NO:254, vhCDR3 of SEQ ID NO:255, vlCDR1 of SEQ ID NO:257, vlCDR2 of SEQ ID NO:258, and vlCDR3 of SEQ ID NO:259; or b. vhCDR1 of SEQ ID NO:219, vhCDR2 of SEQ ID NO:220, vhCDR3 of SEQ ID NO:221, vlCDR1 of SEQ ID NO:223, vlCDR2 of SEQ ID NO:224, and vlCDR3 of SEQ ID NO:225; or c. vhCDR1 of SEQ ID NO:229, vhCDR2 of SEQ ID NO:230, vhCDR3 of SEQ ID NO:231, vlCDR1 of SEQ ID NO:233, vlCDR2 of SEQ ID NO:234, and vlCDR3 of SEQ ID NO:235; or d. vhCDR1 of SEQ ID NO:237, vhCDR2 of SEQ ID NO:238, vhCDR3 of SEQ ID NO:239, vlCDR1 of SEQ ID NO:241, vlCDR2 of SEQ ID NO:242, and vlCDR3 of SEQ ID NO:243; or e. vhCDR1 of SEQ ID NO:245, vhCDR2 of SEQ ID NO:246, vhCDR3 of SEQ ID NO:247, vlCDR1 of SEQ ID NO:249, vlCDR2 of SEQ ID NO:250, and vlCDR3 of SEQ ID NO:251; or f. vhCDR1 of SEQ ID NO:261, vhCDR2 of SEQ ID NO:262, vhCDR3 of SEQ ID NO:263, vlCDR1 of SEQ ID NO:265, vlCDR2 of SEQ ID NO:266, and vlCDR3 of SEQ ID NO:

267.

2. The composition according to claim 1, wherein the variable heavy chain domain and the variable light chain domain are selected from the group consisting of: a. SEQ ID NO:252 and SEQ ID NO:256; or b. SEQ ID NO:218 and SEQ ID NO:222; or c. SEQ ID NO:228 and SEQ ID NO:232; or d. SEQ ID NO:236 and SEQ ID NO:240; or e. SEQ ID NO:244 and SEQ ID NO:248; or f. SEQ ID NO:260 and SEQ ID NO:

264.

3. A composition, the composition comprising an anti-ENPP3 binding domain, the anti-ENPP3 binding domain comprising a variable heavy chain domain containing SEQ ID NO:252 and a variable light chain domain containing SEQ ID NO:

256.

4. A nucleic acid composition, the nucleic acid composition comprising: a. A first nucleic acid, the first nucleic acid encoding the variable heavy chain domain according to claim 1, 2 or 3; and b. A second nucleic acid, the second nucleic acid encoding the variable light chain domain according to claim 1, 2 or 3.

5. An expression vector composition, the expression vector composition comprising: a. A first expression vector, the first expression vector comprising the first nucleic acid according to claim 4; and b. A second expression vector, the second expression vector comprising the second nucleic acid according to claim 4.

6. A host cell, the host cell comprising the expression vector composition according to claim 5.

7. A method for preparing an ENPP3 binding domain, the method comprising culturing the host cell according to claim 6 under conditions for expressing the ENPP3 binding domain; and recovering the ENPP3 binding domain.

8. An anti-ENPP3 antibody, the anti-ENPP3 antibody comprising: a. A heavy chain, the heavy chain comprising VH-CH1-hinge-CH2-CH3; and b. A light chain, the light chain comprising VL-CL; wherein the VH and the VL are selected from the group consisting of: a. SEQ ID NO:252 and SEQ ID NO:256; or b. SEQ ID NO:218 and SEQ ID NO:222; or c. SEQ ID NO:228 and SEQ ID NO:232; or d. SEQ ID NO:236 and SEQ ID NO:240; or e. SEQ ID NO:244 and SEQ ID NO:248; or f. SEQ ID NO:260 and SEQ ID NO:

264.

9. A nucleic acid composition, the nucleic acid composition comprising: a. A first nucleic acid, the first nucleic acid encoding the heavy chain according to claim 8; and b. A second nucleic acid, the second nucleic acid encoding the light chain according to claim 8.

10. An expression vector composition, the expression vector composition comprising: a. A first expression vector, the first expression vector comprising the first nucleic acid according to claim 9; and b. A second expression vector, the second expression vector comprising the second nucleic acid according to claim 9.

11. A host cell comprising the expression vector composition according to claim 10.

12. A method for preparing an anti-ENPP3 antibody, the method comprising culturing the host cell according to claim 11 under conditions for expressing the anti-ENPP3 antibody; and recovering the anti-ENPP3 antibody.

13. A heterodimeric antibody, the heterodimeric antibody comprising: a) A first monomer, the first monomer comprising VH1-CH1-a first domain linker-scFv-a second domain linker-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable heavy chain domain, scFv is an anti-CD3 scFv, and CH2-CH3 is a first Fc domain; b) A second monomer, the second monomer comprising VH1-CH1-a hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein CH2-CH3 is a second Fc domain; and c) A common light chain, the common light chain comprising VL1-CL, wherein VL1 is a first variable light chain domain, and CL is a constant light chain domain; wherein the VH1 and the VL1 form an ENPP3 binding domain, the scFv comprises a second VH domain (VH2), an scFv linker, and a second VL domain (VL2), wherein the VH2 and the VL2 form a CD3 binding domain, and wherein the variable heavy chain complementarity determining regions 1-3 (vhCDR1-3) of the VH1 and the variable light chain complementarity determining regions (vlCDR1-3) of the VL1 are selected from the group consisting of: a. vhCDR1 of SEQ ID NO:253, vhCDR2 of SEQ ID NO:254, vhCDR3 of SEQ ID NO:255, vlCDR1 of SEQ ID NO:257, vlCDR2 of SEQ ID NO:258, and vlCDR3 of SEQ ID NO:259; or b. vhCDR1 of SEQ ID NO:219, vhCDR2 of SEQ ID NO:220, vhCDR3 of SEQ ID NO:221, vlCDR1 of SEQ ID NO:223, vlCDR2 of SEQ ID NO:224, and vlCDR3 of SEQ ID NO:225; or c. vhCDR1 of SEQ ID NO:237, vhCDR2 of SEQ ID NO:238, vhCDR3 of SEQ ID NO:239, vlCDR1 of SEQ ID NO:241, vlCDR2 of SEQ ID NO:242, and vlCDR3 of SEQ ID NO:243; or d. The vhCDR1 of SEQ ID NO:245, the vhCDR2 of SEQ ID NO:246, the vhCDR3 of SEQ ID NO:247, the vlCDR1 of SEQ ID NO:249, the vlCDR2 of SEQ ID NO:250, and the vlCDR3 of SEQ ID NO:251; or e. The vhCDR1 of SEQ ID NO:261, the vhCDR2 of SEQ ID NO:262, the vhCDR3 of SEQ ID NO:263, the vlCDR1 of SEQ ID NO:265, the vlCDR2 of SEQ ID NO:266, and the vlCDR3 of SEQ ID NO:

267.

14. The heterodimeric antibody according to claim 13, wherein said VH1 and VL1 are selected from the group consisting of: a. SEQ ID NO:252 and SEQ ID NO:256; or b. SEQ ID NO:218 and SEQ ID NO:222; or c. SEQ ID NO:228 and SEQ ID NO:232; or d. SEQ ID NO:236 and SEQ ID NO:240; or e. SEQ ID NO:244 and SEQ ID NO:248; or f. SEQ ID NO:260 and SEQ ID NO:

264.

15. The heterodimeric antibody according to any one of claims 13 and 14, wherein said scFv comprises VL2 - scFv linker - VH2 from the N - terminus to the C - terminus.

16. The heterodimeric antibody according to any one of claims 13 to 14, wherein said scFv comprises VH2 - scFv linker - VL2 from the N - terminus to the C - terminus.

17. The heterodimeric antibody according to any one of claims 13 to 14, wherein said first monomer comprises VH1 - CH1 - first domain linker - VL2 - scFv linker - VH2 - second domain linker - CH2 - CH3 from the N - terminus to the C - terminus.

18. The heterodimeric antibody according to any one of claims 13 to 14, wherein said first monomer comprises VH1 - CH1 - first domain linker - VH2 - scFv linker - VL2 - second domain linker - CH2 - CH3 from the N - terminus to the C - terminus.

19. The heterodimeric antibody according to any one of claims 13 to 18, wherein said VH2 and said VL2 are selected from the group consisting of: a. SEQ ID NO:93 and SEQ ID NO:97; or b. SEQ ID NO:103 and SEQ ID NO:107; or c. SEQ ID NO:113 and SEQ ID NO:117; or d. SEQ ID NO:123 and SEQ ID NO:127; or e. SEQ ID NO:133 and SEQ ID NO:137; or f. SEQ ID NO:143 and SEQ ID NO:

147.

20. The bispecific antibody according to any one of claims 13 to 19, wherein the scFv linker is a charged scFv linker.

21. The bispecific antibody according to claim 20, wherein the charged scFv linker comprises the sequence (GKPGS)4.

22. The bispecific antibody according to any one of claims 13 to 21, wherein the first monomer consists of the amino acid variants S364K / E357Q / E233P / L234V / L235A / G236del / S267K, wherein the second monomer consists of the amino acid variants L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and wherein the numbering is according to EU numbering.

23. The bispecific antibody according to any one of claims 13 to 21, wherein the first monomer consists of the amino acid variants S364K / E357Q / E233P / L234V / L235A / G236del / S267K / M428L / N434S, wherein the second monomer consists of the amino acid variants L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K / M428L / N434S, and wherein the numbering is according to EU numbering.

24. A bispecific antibody, the bispecific antibody comprising: a) a first monomer, the first monomer comprising: i) an anti-CD3 scFv, the anti-CD3 scFv comprising a first variable light chain domain (VL1), an scFv linker, and a first variable heavy chain domain (VH1); and ii) a first Fc domain, wherein the scFv is covalently attached to the N-terminus of the first Fc domain using a domain linker; b) a second monomer, the second monomer comprising a VH2-CH1-hinge-CH2-CH3 monomer from the N-terminus to the C-terminus, wherein VH2 is a second variable heavy chain domain, and CH2-CH3 is a second Fc domain; and c) a light chain, the light chain comprising VL2-CL, wherein VL2 is a second variable light chain domain, and CL is a constant light chain domain, wherein VH2 and VL2 form an ENPP3 binding domain, and wherein VH2 and VL2 are selected from the group consisting of: a. SEQ ID NO:252 and SEQ ID NO:256; or b. SEQ ID NO:218 and SEQ ID NO:222; or c. SEQ ID NO:228 and SEQ ID NO:232; or d. SEQ ID NO:236 and SEQ ID NO:240; or e. SEQ ID NO:244 and SEQ ID NO:248; or f. SEQ ID NO:260 and SEQ ID NO:

264.

25. The bispecific antibody according to claim 24, wherein the scFv comprises VL1 - scFv linker - VH1 from the N - terminus to the C - terminus.

26. The bispecific antibody according to claim 24, wherein the scFv comprises VH1 - scFv linker - VL1 from the N - terminus to the C - terminus.

27. The bispecific antibody according to any one of claims 24 to 26, wherein the VH1 and the VL1 are selected from the group consisting of: a. SEQ ID NO:93 and SEQ ID NO:97; or b. SEQ ID NO:103 and SEQ ID NO:107; or c. SEQ ID NO:113 and SEQ ID NO:117; or d. SEQ ID NO:123 and SEQ ID NO:127; or e. SEQ ID NO:133 and SEQ ID NO:137; or f. SEQ ID NO:143 and SEQ ID NO:

147.

28. The bispecific antibody according to any one of claims 24 to 27, wherein the scFv structure is a charged scFv linker.

29. The bispecific antibody according to claim 28, wherein the charged scFv linker has the sequence (GKPGS)4.

30. A nucleic acid composition, the nucleic acid composition comprising: a. A first nucleic acid encoding a first monomer according to any one of claims 13 - 29; b. A second nucleic acid encoding a second monomer according to any one of claims 13 - 29; and c. A third nucleic acid encoding a light chain according to any one of claims 13 - 29.

31. An expression vector composition, the expression vector composition comprising: a. A first expression vector comprising the first nucleic acid according to claim 30; b. A second expression vector comprising the second nucleic acid according to claim 30; and c. A third expression vector comprising the third nucleic acid according to claim 30.

32. A host cell comprising the expression vector composition according to claim 31.

33. A method for preparing a bispecific antibody, the method comprising culturing the host cell according to claim 32 under conditions for expressing the bispecific antibody, and recovering the bispecific antibody.

Images