Heterodimeric antibodies binding cd3 and tumor antigen

Abstract

The present invention relates to heterodimeric antibodies binding CD3 and a tumor antigen. The heterodimeric antibodies comprise: a) a first monomer comprising a first heavy chain, b) a second monomer comprising a second heavy chain and b) a common light chain; the heterodimeric antibodies bind to a target tumor antigen and human CD3.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority to U.S. Provisional Patent Application No. 62 / 159,111, filed May 8, 2015; U.S. Provisional Patent Application No. 62 / 251,005, filed November 4, 2015; U.S. Provisional Patent Application No. 62 / 250,971, filed November 4, 2015; USSN 14 / 952,714, filed November 11, 2015; and PCT / US2015 / 062772, filed November 25, 2015, the entire contents of which are incorporated herein by reference, especially the drawings, descriptions, and claims. Background of the Invention

[0003] Antibody-based therapies have been successfully used to treat a variety of diseases, including cancer and autoimmune / inflammatory conditions. However, improvements to these drugs, particularly enhancing their clinical efficacy, are still needed. One approach being explored is to modify other and neoantigen binding sites into antibody-based drugs, enabling a single immunoglobulin molecule to co-bind two different antigens. This non-natural or altered antibody form that binds to two different antigens is often referred to as a bispecific antibody. Due to the significant diversity of antibody variable regions (Fv), theoretically capable of generating Fvs that recognize any molecule, a common approach to bispecific antibody generation is to introduce new variable regions into the antibody.

[0004] Various methods of altering antibody forms for bispecific targeting have been explored (Chames and Baty, 2009, mAbs 1[6]:1-9; Holliger and Hudson, 2005, Nature Biotechnology 23[9]:1126-1136; Kontermann, mAbs 4(2):182(2012), all of which are incorporated herein by reference). Initially, bispecific antibodies were prepared by fusing two cell lines that each produced a single monoclonal antibody (Milstein et al., 1983, Nature 305:537-540). Although the resulting hybridomas or tetraploids did produce bispecific antibodies, they were only a small population and required extensive purification to isolate the desired antibody. The engineered solution was to prepare bispecific antibodies using antibody fragments. Because such fragments lack the complex quaternary structure of full-length antibodies, variable light and heavy chains can be linked in a single genetic construct. Many different forms of antibody fragments have been produced, including biantibodies, single-chain biantibodies, tandem scFv and Fab2 bispecific antibodies (Chames and Baty, 2009, mAbs 1[6]: 1-9; Holliger and Hudson, 2005, Nature Biotechnology 23[9]: 1126-1136; expressly incorporated herein by reference). While these forms can be expressed at high levels in bacteria and may have a favorable permeation benefit due to their small size, they are rapidly cleared in vivo and may present manufacturing difficulties related to their production and stability. The main reason for these drawbacks is that antibody fragments generally lack antibody constant regions with relevant functions, including larger size, high stability, and binding to various Fc receptors and ligands that maintain a long half-life in serum (i.e., neonatal Fc receptor FcRn) or binding sites used for purification (i.e., protein A and protein G).

[0005] Attempts have begun to address the shortcomings of fragment-based bispecific antibodies by modifying dual binding into a full-length antibody-like form (Wu et al., 2007, Nature Biotechnology 25

[11] : 1290-1297; USSN12 / 477,711; Michaelson et al., 2009, mAbs 1[2]: 128-141; PCT / US2008 / 074693; Zuo et al., 2000, Protein Engineering 13[5]: 361-367; USSN09 / 865,198; Shen et al., 2006, J Biol Chem 281

[16] : 10706-10714; Lu et al., 2005, J Biol Chem 280

[20] : 19665-19672; PCT / US2005 / 025472; expressly incorporated herein by reference). These forms overcome some of the obstacles of bispecific antibodies, mainly because they contain an Fc region. A significant drawback of these forms is that, because they construct a new antigen-binding site at the top of the homodimeric constant chain, the binding to the new antigen is always divalent.

[0006] For many antigens that have attracted attention as common targets in therapeutic bispecific antibody forms, it is desirable for binding to be monovalent rather than bivalent. For many immune receptors, cell activation is achieved through cross-linking of monovalent binding interactions. Cross-linking mechanisms are typically mediated by antibody / antigen immune complexes or by effector cell binding to target cells. For example, low-affinity Fcγ receptors (FcγRs) such as FcγRIIa, FcγRIIb, and FcγRIIIa bind monovalently to the Fc region of antibodies. Monovalent binding does not activate cells expressing these FcγRs; however, after immune complexation or cell-to-cell contact, the receptors are cross-linked and aggregate on the cell surface, leading to activation. For receptors responsible for mediating cell killing, such as FcγRIIIa on natural killer (NK) cells, receptor cross-linking and cell activation occur when effector cells bind to target cells in a high-affinity form (Bowles and Weiner, 2005, J Immunol Methods 304:88-99, incorporated herein by reference). Similarly, on B cells, the inhibitory receptor FcγRIIb downregulates B cell activation only when it binds to the cell surface B cell receptor (BCR) into an immune complex, a mechanism mediated by immune complexes of soluble IgG and the same antigen recognized by the BCR (Heyman 2003, Immunol Lett 88[2]:157-161; Smith and Clatworthy, 2010, Nature Reviews Immunology 10:328-343; explicitly incorporated by reference). As another example, CD3 activation of T cells occurs only when its associated T cell receptor (TCR) binds to the MHC of the antigen-presenting cell loaded with the antigen at a high-affinity cell-cell synapse (Kuhns et al., 2006, Immunity 24:133-139). In fact, the nonspecific bivalent cross-linking of CD3 using anti-CD3 antibodies causes cytokine storms and toxicity (Perruche et al., 2009, J Immunol 183[2]:953-61; Chatenoud and Bluestone, 2007, Nature Reviews Immunology 7:622–632; explicitly incorporated by reference). Therefore, for practical clinical applications, the preferred mode of CD3 co-binding for redirecting the killing of target cells is monovalent binding, which leads to activation only when bound to the co-binding target.

[0007] CD38, also known as cyclic ADP-ribolytic enzyme, is a type II transmembrane glycoprotein with a long C-terminal extracellular domain and a short N-terminal cytoplasmic domain. In hematopoietic cells, various functional effects are attributed to CD38-mediated signaling, including lymphocyte proliferation, cytokine release, regulation of B cell and bone marrow cell development and survival, and induction of dendritic cell maturation. CD38 is upregulated in many hematopoietic malignancies and cell lines derived from various hematopoietic malignancies, including non-Hodgkin lymphoma (NHL), Burkitt lymphoma (BL), multiple myeloma (MM), B chronic lymphocytic leukemia (B-CLL), B and T acute lymphoblastic leukemia (ALL), T-cell lymphoma (TCL), acute myeloid leukemia (AML), hairy cell leukemia (HCL), Hodgkin lymphoma (HL), and chronic myeloid leukemia (CML). On the other hand, most primitive pluripotent stem cells in the hematopoietic system are CD38-positive. Despite recent discoveries and advancements in anticancer drugs, many cancers involving CD38-expressing tumors still have poor prognoses. Therefore, better treatments for these cancers are needed.

[0008] B-cell antigen CD19 (CD19, also known as B-cell surface antigen B4, Leu-12) is a pan-B-cell surface marker expressed from early pre-B cell development to terminal differentiation into plasma cells. CD19 promotes the proliferation and survival of mature B cells. It complexes with CD21 on the cell surface. It also binds to CD81 and Leu-13 and enhances B-cell receptor (BCR) signaling. Together with BCR, CD19 regulates intrinsic and antigen receptor-induced signaling thresholds that are crucial for B-cell clonal expansion and humoral immunity. In cooperation with CD21, it links the adaptive and innate immune systems. Upon activation, CD19 undergoes cytoplasmic tail phosphorylation, leading to Src family kinase binding and PI-3 kinase recruitment. It is a promising target for lymphoblastic cancer immunotherapy because it is also expressed on the vast majority of NHL cells and in some leukemias.

[0009] Many antibodies or antibody-drug conjugates targeting CD19 have been evaluated in preclinical studies or clinical trials for cancer treatment. These anti-CD19 antibodies or antibody-drug conjugates include, but are not limited to, MT-103 (a single-chain bispecific CD19 / CD3 antibody; Hoffman et al., 2005 Int J Cancer 115: 98-104; Schlereth et al., 2006 Cancer Immunol Immunother 55: 503-514), CD19 / CD16 biantibody (Schlenzka et al., 2004 Anti-cancer Drugs 15: 915-919; Kipriyanov et al., 2002 J Immunol 169: 137-144), BU12-saponin (Flavell et al., 1995 Br J Cancer 72: 1373-1379), and anti-CD19-idarubicin (Rowland et al., 1993 Cancer Immunol Immunother 55: 503-514); all of which are explicitly incorporated herein by reference.

[0010] CD123, also known as interleukin-3 receptor α (IL-3Rα), is expressed on dendritic cells, monocytes, eosinophils, and basophils. Most myeloid cells (CD13+, CD14+, CD33+, low CD15) and some CD19+ cells, as well as directed hematopoietic stem cells / progenitor cells, also constitutively express CD123. It is not present on CD3+ cells.

[0011] Therefore, although bispecific antibodies derived from antibody fragments are limited by biophysical and pharmacokinetic barriers, those constructed in full-length antibody-like forms suffer from the disadvantage of multivalent binding to co-target antigens in the absence of the primary target antigen, leading to nonspecific activation and potentially toxic effects. This invention addresses this problem by introducing novel bispecific antibodies targeting CD3 and CD38. Summary of the Invention

[0012] Therefore, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising: i) a first heavy chain comprising: 1) a first variable heavy chain domain; 2) a first constant heavy chain comprising a first Fc domain; 3) an scFv comprising an scFv variable light chain domain, an scFv linker, and an scFv variable heavy chain domain; wherein the scFv is covalently linked to the C-terminus of the Fc domain by a domain linker; b) a second monomer comprising a second heavy chain comprising a second variable heavy chain domain and a second constant heavy chain, the second constant heavy chain comprising a second Fc domain; and c) a common light chain comprising a variable light chain domain and a constant light chain domain.

[0013] On the other hand, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising: i) a first heavy chain comprising: 1) a first variable heavy chain domain; 2) a first constant heavy chain domain comprising a first Fc domain; and 3) a first variable light chain domain, wherein the first variable light chain domain is covalently linked to the C-terminus of the first Fc domain by a domain linker; b) a second monomer comprising: i) a second variable heavy chain domain; ii) a second constant heavy chain domain comprising a second Fc domain; and iii) a third variable heavy chain domain, wherein the second variable heavy chain domain is covalently linked to the C-terminus of the second Fc domain by a domain linker; and c) a common light chain comprising the variable light chain domain and the constant light chain domain.

[0014] In another aspect, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising: i) a first heavy chain comprising: 1) a first variable heavy chain domain; 2) a first constant heavy chain comprising a first CH1 domain and a first Fc domain; 3) an scFv comprising an scFv variable light chain domain, an scFv linker, and an scFv variable heavy chain domain; wherein the scFv is covalently linked between the C-terminus of the CH1 domain and the N-terminus of the first Fc domain by a domain linker; b) a second monomer comprising a second heavy chain comprising a second variable heavy chain domain and a second constant heavy chain, the second constant heavy chain comprising a second Fc domain; and c) a common light chain comprising a variable light chain domain and a constant light chain domain.

[0015] On the other hand, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising: i) a first heavy chain comprising: 1) a first variable heavy chain domain; 2) a first constant heavy chain domain comprising a first Fc domain; and 3) a first variable light chain domain, wherein the second variable light chain domain is covalently linked by a domain linker between the C-terminus of the CH1 domain of the first constant heavy chain domain and the N-terminus of the first Fc domain; b) a second monomer comprising: i) a second variable heavy chain domain; ii) a second constant heavy chain domain comprising a second Fc domain; and iii) a third variable heavy chain domain, wherein the second variable heavy chain domain is covalently linked by a domain linker to the C-terminus of the second Fc domain; and c) a common light chain comprising the variable light chain domain and the constant light chain domain.

[0016] In another aspect, the present invention provides a heterodimeric antibody comprising: a) a first monomer comprising: i) a first heavy chain comprising: 1) a first variable heavy chain domain; 2) a first constant heavy chain comprising a first CH1 domain and a first Fc domain; 3) an scFv comprising an scFv variable light chain domain, an scFv linker, and an scFv variable heavy chain domain; wherein the scFv is covalently linked between the C-terminus of the CH1 domain and the N-terminus of the first Fc domain by a domain linker; b) a second monomer comprising a second Fc domain; and c) a light chain comprising a variable light chain domain and a constant light chain domain.

[0017] In some aspects, the first and second Fc domains have combined amino acid substitutions selected from: S364K / E357Q: L368D / K370S; L368D / K370S: S364K; L368E / K370S: S364K; T411T / E360E / Q362E: D401K; L368D / K370S: S364K / E357L and K370S: S364K / E357Q. Furthermore, the variable heavy chain domain and the variable light chain domain bind to a first target tumor antigen (TTA), and the scFv binds to a second TTA or human CD3. In some embodiments, the TTA is selected from CD19, CD20, and CD123.

[0018] In another aspect, the present invention provides an anti-CD3 antigen-binding domain, the sequences of which, such as the CDR and / or variable domains and / or scFv, are shown in the figures for H1.32_L1.47, H1.89_L1.47, H1.90_L1.47, H1.33_L1.47, and H1.31_L1.47. The present invention also provides nucleic acid compositions, expression vector compositions, and host cells.

[0019] In another aspect, the present invention also provides a heterodimeric antibody comprising a) a first monomer comprising: i) a first Fc domain; ii) an anti-CD3scFv comprising a scFv variable light chain domain, a scFv linker, and a scFv variable heavy chain domain; wherein the scFv is covalently linked to the N-terminus of the Fc domain by a domain linker; b) a second monomer comprising a heavy chain comprising: i) a heavy chain variable domain; and ii) a heavy chain constant domain comprising a second Fc domain; and c) a light chain comprising a variable light chain domain and a variable light chain constant domain; wherein the anti-CD3 scFv is selected from anti-CD3 H1.32_L1.47, anti-CD3H1.89_L1.47, anti-CD3 H1.90_L1.47, and anti-CD3 H1.33_L1.47 (SEQ ID NO:XX). Heavy chain variable domains and light chain variable domains combine to form TTAs (including but not limited to CD19, CD20, CD38 and CD123).

[0020] On the other hand, the present invention provides an anti-CD20 antibody binding domain comprising: a) a variable light chain domain comprising v1CDR1 having the sequence RASWSVSYIH (SEQ ID NO: XX), v1CDR2 having the sequence ATSNLAS (SEQ ID NO: XX), and v1CDR3 having the sequence QQWTHNPPT (SEQ ID NO: XX); and b) a variable heavy chain domain comprising vhCDR1 having the sequence SYNMH (SEQ ID NO: XX), vhCDR2 having the sequence AIYPGNGATSYSQKFQG (SEQ ID NO: XX), and vhCDR3 having the sequence SYYMGGDWYFDV (SEQ ID NO: XX). In some embodiments, the anti-CD20 antibody binding domain has the sequence C2B8 H1.202_L1.113.

[0021] On the other hand, the present invention provides an anti-CD20 antibody binding domain comprising: a) a variable light chain domain comprising v1CDR1 having the sequence RASSSVSYIH (SEQ ID NO: XX), v1CDR2 having the sequence ATSNLAS (SEQ ID NO: XX), and v1CDR3 having the sequence QQWTSNPPT (SEQ ID NO: XX); and b) a variable heavy chain domain comprising vhCDR1 having the sequence SYNMH (SEQ ID NO: XX), vhCDR2 having the sequence AIYPGNGDTSYNQKFQG (SEQ ID NO: XX), and vhCDR3 having the sequence STYYGGDWYFNV (SEQ ID NO: XX).

[0022] In some embodiments, the anti-CD20 antibody binding domain has a C2B8_H1L1 sequence.

[0023] In another aspect, the present invention provides a heterodimeric antibody comprising a) a first monomer comprising: i) a first Fc domain; ii) an anti-CD3 scFv comprising a scFv variable light chain domain, a scFv linker, and a scFv variable heavy chain domain; wherein the scFv is covalently linked to the N-terminus of the Fc domain by a domain linker; b) a second monomer comprising a heavy chain comprising: i) a heavy chain variable domain; and ii) a heavy chain constant domain comprising a second Fc domain; and c) a light chain comprising a variable light chain domain and a variable light chain constant domain; wherein the variable heavy chain and the light chain form a C2B8 H1.202_L1.113 or C2B8_H1L1 binding domain.

[0024] In another aspect, the present invention provides a heterodimeric antibody comprising a) a first monomer comprising: i) a first Fc domain; ii) an anti-CD3 scFv comprising a scFv variable light chain domain, a scFv linker, and a scFv variable heavy chain domain; wherein the scFv is covalently linked to the N-terminus of the Fc domain by a domain linker; b) a second monomer comprising a heavy chain comprising: i) a heavy chain variable domain; and ii) a heavy chain constant domain comprising a second Fc domain; and c) a light chain comprising a variable light chain domain and a variable light chain constant domain. In this embodiment, the variable domain binds to CD123 and may have the sequence 7G3_H1.109_L1.47.

[0025] In addition, the present invention provides heterodimeric antibodies selected from the following: XENP15049, XENP15051, XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP15635, XENP15636, XENP15638, XENP15639, and XENP13677. XENP14388, XENP14389, XENP14390, XENP14391, XENP14392, XENP14393, XENP16366, XENP16367, XENP16368, XENP16369, XENP16370, XENP16371, XENP16372, XENP16373, XENP16375, XENP16376, XENP16377, XENP14045, and XENP13928. This invention provides methods for preparing these proteins and using them to treat patients, and also provides nucleic acids, expression vectors, and host cells.

[0026] Furthermore, the present invention provides a heterodimeric antibody having six CDRs (vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3) of a variable region selected from one of the antigen-binding domains of heterodimeric antibodies described below: XENP15049, XENP15051, XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP15635, XENP15636, XENP15637, XENP15638, XENP15639 ... ENP15636, XENP15638, XENP15639, XENP13677, XENP14388, XENP14389, XENP14390, XENP14391, XENP14392, XENP14393, XENP16366, XENP16367, XENP16368, XENP16369, XENP16370, XENP16371, XENP16372, XENP16373, XENP16375, XENP16376, XENP16377, XENP14045, and XENP13928. This invention provides methods for preparing these proteins and using them to treat patients, and also provides nucleic acids, expression vectors, and host cells.

[0027] Furthermore, the present invention provides a heterodimeric antibody having two sets of CDRs. The first set contains six CDRs (vhCDR1, vhCDR2, vhCDR3, vlCDR1, vlCDR2, and vlCDR3) derived from the variable region of one of the antigen-binding domains of the following heterodimeric antibodies. The second set contains the variable region of the second antigen-binding domain of the other heterodimeric antibody. The heterodimeric antibody is selected from: XENP15049, XENP15051, XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, ...632, XENP15049, XENP15051, XENP15051, XENP15632, XENP15049, XENP15051, XENP15051, XENP15052, XENP15049, XENP15051, XENP15051, XENP15052, XENP15049, XENP15051, XENP15052, XENP15049, XENP15051, XENP15052, XENP15053, XENP15053, XEN P15633, XENP15634, XENP15635, XENP15636, XENP15638, XENP15639, ENP16366,XENP16367,XENP16368,XENP16369,XENP16370,XENP16371,XENP16372,XENP16373, This invention provides methods for preparing these proteins and using them to treat patients, and also provides nucleic acids, expression vectors, and host cells.

[0028] Furthermore, the present invention provides a heterodimeric antibody having two sets of VH and VL domains, the first set being a variable region derived from one of the antigen-binding domains of the following heterodimeric antibodies, and the second set being a variable region derived from the second antigen-binding domain of another heterodimeric antibody, wherein the heterodimeric antibody is selected from: XENP15049, XENP15051, XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP1 XENP15635, XENP15636, XENP15638, XENP15639, XENP13677, XENP14388, XENP14389, XENP14390, XENP14391, XENP14392, XENP14393, XENP16366, XENP16367, XENP16368, XENP16369, XENP16370, XENP16371, XENP16372, XENP16373, XENP16375, XENP16376, XENP16377, XENP14045, and XENP13928. This invention provides methods for preparing these proteins and using them to treat patients, and also provides nucleic acids, expression vectors, and host cells. Attached Figure Description

[0029] Figure 1 A, 1B, and 1C illustrate various forms of the invention. Two styles of "opener" forms are shown, one with an anti-CD3 antigen-binding domain containing scFv and an anti-TTA antigen-binding domain containing Fab, and the other the opposite. mAb-Fv, mAb-scFv, center-scFv, and center-Fv forms are all shown in the figures. Although anti-CD3 is shown as the scFv, as described herein, the Fv sequences can be interchanged. That is, the anti-CD3 and anti-TTA domains of mAb-Fv, mAb-scFv, center-scFv, and center-Fv are interchangeable. Additionally, a "single-arm" form where one monomer contains only the Fc domain is shown, with both a single-arm center scFv and a single-arm center Fv. A double scFv form is also shown.

[0030] Figure 2Sequences of the "High CD3" anti-CD3-H1.30_L1.47 construct are shown, including the variable heavy chain domain and variable light chain domain (CDRs are underlined), each single-listed v1CDR and vhCDR, and the scFv construct with a charged linker (doubly underlined). This charged linker can be replaced with an uncharged linker or a different other charged linker as needed, and the same applies to all sequences in the figures.

[0031] Figure 3 Sequences of the "High-Medium #1" anti-CD3_H1.32_L1.47 construct are shown, including the variable heavy chain domain and variable light chain domain (CDRs are underlined), each single-listed v1CDR and vhCDR, and the scFv construct with a charged linker (doubly underlined). This charged linker can be replaced with an uncharged linker or a different other charged linker as needed, and the same applies to all sequences in the figures.

[0032] Figure 4 Sequences of the "High-Medium #2" anti-CD3_H1.89_L1.47 construct are shown, including the variable heavy chain domain and variable light chain domain (CDRs are underlined), each single-listed v1CDR and vhCDR, and the scFv construct with a charged linker (doubly underlined). This charged linker can be replaced with an uncharged linker or a different other charged linker as needed, and the same applies to all sequences in the figures.

[0033] Figure 5 Sequences of the "High-Medium #3" anti-CD3_H1.90_L1.47 construct are shown, including the variable heavy chain domain and variable light chain domain (CDRs are underlined), each single-listed v1CDR and vhCDR, and the scFv construct with a charged linker (doubly underlined). This charged linker can be replaced with an uncharged linker or a different other charged linker as needed, and the same applies to all sequences in the figures.

[0034] Figure 6 Sequences of the "Medium" anti-CD3_H1.90_L1.47 construct are shown, including the variable heavy chain domain and variable light chain domain (CDRs are underlined), each single-listed v1CDR and vhCDR, and the scFv construct with a charged linker (doubly underlined). This charged linker can be replaced with an uncharged linker or a different other charged linker as needed, and the same applies to all sequences in the figures.

[0035] Figure 7The sequences of the "low" anti-CD3_H1.31_L1.47 construct are shown, including the variable heavy chain domain and the variable light chain domain (CDRs are underlined), each single-column v1CDR and vhCDR, and the scFv construct with a charged connector (double underlined). This charged connector can be replaced with an uncharged connector or other different charged connectors as needed, as is the case for all sequences in the figures.

[0036] Figure 8 Show the sequences of the high CD38:OKT10_H1.77_L1.24 construct, including the variable heavy chain domain and the variable light chain domain (CDRs are underlined), each single column v1CDR and vhCDR, and the scFv construct with a charged connector (double underlined).

[0037] Figure 9 The display shows the sequences of the CD38:OKT10_H1L1.24 construct, including the variable heavy chain domain and the variable light chain domain (CDRs are underlined), each single column v1CDR and vhCDR, and the scFv construct with a charge connector (double underlined).

[0038] Figure 10 Show the sequences of the low CD38:OKT10_H1L1 construct, including the variable heavy chain domain and the variable light chain domain (CDRs are underlined), each single column v1CDR and vhCDR, and the scFv construct with a charged connector (double underlined).

[0039] Figure 11 Display the sequence of XENP15331.

[0040] Figure 12 Display the sequence of XENP13243.

[0041] Figure 13 Display the sequence of XENP14702.

[0042] Figure 14 Display the sequence of XENP15426.

[0043] Figure 15 Display the sequence of XENP14701.

[0044] Figure 16 Display the sequence of XENP14703.

[0045] Figure 17 Display the sequence of XENP13243.

[0046] Figure 18 Display the sequence of XENP18967.

[0047] Figure 19 Display the sequence of XENP18971.

[0048] Figure 20 Display the sequence of XENP18969.

[0049] Figure 21 Display the sequence of XENP18970.

[0050] Figure 22 Display the sequence of XENP18972.

[0051] Figure 23 Display the sequence of XENP18973.

[0052] Figure 24 Display the sequence of XENP15055.

[0053] Figure 25 Display the sequence of XENP13544.

[0054] Figure 26 Display the sequence of XENP13694.

[0055] Figure 27 Display the sequence of human CD3ε.

[0056] Figure 28 The full-length human CD38 protein (SEQ ID NO: 130) and its extracellular domain (ECD; SEQ ID NO: 131) are shown.

[0057] Figures 29A-29E This demonstrates useful combinations of paired heterodimer variants (including biased variants and pI variants). Figure 29E Some variants do not have a corresponding "monomer 2" variant; these are standalone pI variants that can be used with either monomer or located on the Fab side of the "opener," for example, an scFv connector with an appropriate charge can be used on a second monomer employing scFv as the second antigen-binding domain. Suitable charged connectors are shown in Figure 33.

[0058] Figure 30 This diagram displays a series of isosteric variant antibody constant regions and corresponding substitutions. pI_(-) indicates a low pI variant, while pI_(+) indicates a high pI variant. These can optionally and independently be combined with other heterodimerizing variants of the present invention (and other variant types as described herein).

[0059] Figure 31 This demonstrates the elimination variants (sometimes referred to as "knockout" or "KO" variants) that are useful for eliminating FcγR binding.

[0060] Figure 32This invention demonstrates two particularly useful embodiments.

[0061] Figure 33 shows various charged scFv linkers used to enhance or reduce the pI of heterodimeric antibodies using single or multiple scFvs as components. The (+H) positively charged linker is particularly useful here, especially in conjunction with the shown anti-CD3 vl and vh sequences. A single-charged prior art scFv linker, known as “Whitlow,” is described in Whitlow et al., Protein Engineering 6(8): 989-995 (1993). It should be noted that this linker is used to reduce scFv aggregation and enhance proteolytic stability.

[0062] Figure 34 This paper presents a series of modified heterodimers-biased Fc variants, along with their heterodimer yields (as determined by HPLC-CIEX) and thermal stability (as determined by DSC). Thermal stability not determined is indicated by "nd".

[0063] Figure 35 The expression yield of bispecific antibody after affinity purification of protein A.

[0064] Figure 36 Chromatogram of cation exchange purification.

[0065] Figure 37 Redirected T cell cytotoxicity assay: 24-hour incubation with 10kbpM RPMI 8226 cells and 400kbpM T cells. The test sample was anti-CD38×anti-CD3 bispecific antibody. LDH detection was used.

[0066] Figure 38 Redirected T cell cytotoxicity assay: 24-hour incubation with 10kbpM RPMI 8226 cells and 500kbpM human PBMCs. The test sample was an anti-CD38×anti-CD3 bispecific antibody. LDH detection was used.

[0067] Figure 39 Display the sequence of XENP14419.

[0068] Figure 40 Display the sequence of XENP14420.

[0069] Figure 41 Display the sequence of XENP14421.

[0070] Figure 42 Display the sequence of XENP14422.

[0071] Figure 43 Display the sequence of XENP14423.

[0072] Figure 44 Redirected T cell cytotoxicity assay: 96-hour incubation with 40kbpM RPMI8226 cells and 400kbpM human PBMCs. The test sample was anti-CD38 × anti-CD3 Fab-scFv-Fc. Flow cytometry was used to detect the disappearance of CD38+ cells.

[0073] Figure 45 :right Figure 1 Further analysis of the redirected T cell cytotoxicity assay is shown. The first row shows the mean fluorescence intensity (MFI) of the activation marker CD69 on CD4+ and CD8+ T cells as measured by flow cytometry. The second row shows the percentage of Ki-67+ CD4+ and CD8+ T cells as an indicator of cell proliferation. The third row shows the mean intracellular fluorescence intensity (MFI) of the granzyme B inhibitor PI-9 on CD4+ and CD8+ T cells as measured by flow cytometry.

[0074] Figure 46 The mouse assay was designed to detect the antitumor activity of the anti-CD38×anti-CD3 Fab-scFv-Fc bispecific antibody.

[0075] Figure 47 : The relationship between measured tumor size and time and treatment.

[0076] Figure 48 : Bioluminescence image (day 10).

[0077] Figure 49 CD38 in macaques after a single dose of the specified test sample + Cell clearance.

[0078] Figure 50 T cell activation in macaques was measured by CD69 mean fluorescence intensity (MFI), with color coding as follows: Figure 49 As shown.

[0079] Figure 51 Serum IL-6 levels after a single dose of the specified test sample.

[0080] Figure 52 Display the sequence of XENP15427.

[0081] Figure 53 Display the sequence of XENP15428.

[0082] Figure 54 Display the sequence of XENP15429.

[0083] Figure 55 Display the sequence of XENP15430.

[0084] Figure 56 Display the sequence of XENP15431.

[0085] Figure 57 Display the sequence of XENP15432.

[0086] Figure 58 Display the sequence of XENP15433.

[0087] Figure 59 Display the sequence of XENP15434.

[0088] Figure 60 Display the sequence of XENP15435.

[0089] Figure 61 Display the sequence of XENP15436.

[0090] Figure 62 Display the sequence of XENP15437.

[0091] Figure 63 Display the sequence of XENP15438.

[0092] Figure 64 This demonstrates binding affinity in the Biacore assay.

[0093] Figure 65 The experiment demonstrated the use of different ratios of light chains, Fab-Fc, and scFv-Fc to stabilize the heterodimer purity during aggregate formation.

[0094] Figure 66 Clearance of human IgM and IgG2 from anti-CD38×anti-CD3 bispecific antibodies in the huPBMC mouse model.

[0095] Figure 67 shows the stability-optimized humanized anti-CD3 variant scFv. Substitutions are shown compared to the H1_L1.4 scFv sequence. Amino acid numbers are Kabat numbers.

[0096] Figure 68: Amino acid sequences of stability-optimized humanized anti-CD3 variant scFv. CDRs are underlined. For each heavy / light chain combination, four sequences are listed: (i) scFv with a C-terminal 6×His tag, (ii) scFv only, (iii) VH only, and (iv) VL only.

[0097] Figure 69Redirected T cell cytotoxicity assay: 24-hour incubation with 10kJ RPMI 8226 cells and 500kJ PBMCs. The test sample was anti-CD38 (OKT10_H1L1, OKT10_H1.77_L1.24) × anti-CD3Fab-scFv-Fc. LDH assay was used.

[0098] Figure 70 huPBL-SCID Ig-clearance assay. The test substance was administered 8 days post-PBMC transplantation at a dose of 0.03, 0.3, or 3 mg / kg via intraperitoneal administration. Blood samples were collected 14 days post-PBMC transplantation, and serum was obtained for the determination of human IgM and IgG2.

[0099] Figure 71 Display the sequence of XENP15049.

[0100] Figure 72 Display the sequence of XENP15051.

[0101] Figure 73 Display the sequence of XENP15050.

[0102] Figure 74 Display the sequence of XENP13676.

[0103] Figure 75 Display the sequence of XENP14696.

[0104] Figure 76 Display the sequence of XENP15629.

[0105] Figure 77 Display the sequence of XENP15053.

[0106] Figure 78 Display the sequence of XENP15630.

[0107] Figure 79 Display the sequence of XENP15631.

[0108] Figure 80 Display the sequence of XENP15632.

[0109] Figure 81 Display the sequence of XENP15633.

[0110] Figure 82 Display the sequence of XENP15634.

[0111] Figure 83 Display the sequence of XENP15635.

[0112] Figure 84Display the sequence of XENP15636.

[0113] Figure 85 Display the sequence of XENP15638.

[0114] Figure 86 Display the sequence of XENP15639.

[0115] Figure 87 Display the sequence of XENP13677.

[0116] Figure 88 Display the sequence of XENP14388.

[0117] Figure 89 Display the sequence of XENP14389.

[0118] Figure 90 Display the sequence of XENP14390.

[0119] Figure 91 Display the sequence of XENP14391.

[0120] Figure 92 Display the sequence of XENP14392.

[0121] Figure 93 Display the sequence of XENP14393.

[0122] Figure 94 Display the sequence of XENP16366.

[0123] Figure 95 Display the sequence of XENP16367.

[0124] Figure 96 Display the sequence of XENP16368.

[0125] Figure 97 Display the sequence of XENP16369.

[0126] Figure 98 Display the sequence of XENP16370.

[0127] Figure 99 Display the sequence of XENP16371.

[0128] Figure 100 Display the sequence of XENP16372.

[0129] Figure 101 Display the sequence of XENP16373.

[0130] Figure 102 Display the sequence of XENP16374.

[0131] Figure 103 Display the sequence of XENP16375.

[0132] Figure 104 The sequence of XENP16376 is shown. The CDR, vh, and vl sequences of the anti-CD20 Fab arm are shown below. Figure 121 As shown in the image.

[0133] Figure 105 Display the sequence of XENP16377.

[0134] Figure 106 The sequences of CD20 and CD123 antigens are displayed.

[0135] Figure 107 Surface plasmon resonance determination of CD3 affinity. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. Human CD3δε-Fc (Sino Biological) was covalently bound to the chip surface. The test sample passed at 3.125, 12.5, 50, and 200 nM.

[0136] Figure 108 Surface plasmon resonance determination of CD3 affinity. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. Macaque CD3δε-Fc (Sinochem Biotechnology Co., Ltd.) was covalently bound to the chip surface. The test sample passed at 3.125, 12.5, 50, and 200 nM.

[0137] Figure 109 Surface plasmon resonance determination of CD3 affinity. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. Human CD3δε-Fc (Sino Biological) was covalently bound to the chip surface. The test sample passed at 31.25, 125, 500, and 2000 nM.

[0138] Figure 110 Surface plasmon resonance determination of CD3 affinity. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. Macaque CD3δε-Fc (Sinochem Biotechnology Co., Ltd.) was covalently bound to the chip surface. The test sample passed at 31.25, 125, 500, and 2000 nM.

[0139] Figure 111Surface plasmon resonance determination of CD3 affinity. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. Macaque CD3δε-Fc (Sinochem Biotechnology Co., Ltd.) was covalently bound to the chip surface. The test sample passed at 31.25, 125, 500, and 2000 nM.

[0140] Figure 112 Redirected T cell cytotoxicity assay: 24-hour incubation with 10kJ Ramos cells and 250kJ PBMCs. The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3 Fab-scFv-Fc. LDH assay was used.

[0141] Figure 113 Redirected T cell cytotoxicity assay: 24-hour incubation with 20kJE Jeko cells and 200kPBMCs (CD19-cleared). The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3Fab-scFv-Fc. Flow cytometry was used for detection, specifically detecting CD19. + The disappearance of cells.

[0142] Figure 114 : Figure 113 IL-6 was produced 24 hours after the experiment described in the paper.

[0143] Figure 115: Redirected T cell cytotoxicity assay, 5-hour incubation, 20kJ Jeko cells, 500kPBMCs (CD19-cleared). The test sample was anti-CD20 (C2B8_H1L1) × anti-CD3 Fab-scFv-Fc. Flow cytometry was used for detection, specifically for CD19. + The disappearance of cells.

[0144] Figure 116: Redirected T cell cytotoxicity assay, 24-hour incubation, 20kJE Jeko cells, 500kPBMCs (CD19-cleared). The test sample was anti-CD20 (C2B8_H1.202_L1.113) × anti-CD3Fab-scFv-Fc. Flow cytometry was used for detection, specifically for CD19. + The disappearance of cells.

[0145] Figure 117 : Figure 113 IL-6 was produced 24 hours after the experiment described in the paper.

[0146] Figure 118Redirected T cell cytotoxicity assay: 24-hour incubation with 10kJ RPMI 8226 cells and 500kJ PBMCs. The test sample was anti-CD38 (OKT10_H1L1, OKT10_H1.77_L1.24) × anti-CD3Fab-scFv-Fc. LDH assay was used.

[0147] Figure 119 huPBL-SCID Ig-clearance assay. The test sample was administered at 5 mg / kg on days 1 and 8 post-PBMC transplantation. The route of administration was intraperitoneal. Blood samples were collected 14 days post-PBMC transplantation, serum was processed, and human IgM and IgG2 were measured.

[0148] Figure 120 huPBL-SCID Ig-clearance assay. Eight days after PBMC implantation, 0.03, 0.3, or 3 mg / kg of the test sample was administered intraperitoneally. Blood samples were collected 14 days post-PBMC transplantation, serum was obtained, and human IgM and IgG2 were measured.

[0149] Figure 121 The sequence shown is CD20 C2B8_H1.202_L1.113. Although the live connector shown is (+H), other live or non-live connectors may also be used, such as those shown in Figure 33.

[0150] Figure 122 The sequence shown is low CD20 C2B8_H1L1. Although the live connector shown is (+H), other live or non-live connectors may also be used, such as those shown in Figure 33.

[0151] Figure 123 The sequence number CD123 7G3_H1.109_L1.57 is shown. Although the live connector is shown as (+H), other live or non-live connectors may also be used, such as those shown in Figure 33.

[0152] Figure 124The matrices in the diagram represent possible combinations of the present invention. "A" indicates that the reference CD3 sequence CDR can be combined with the CDR on the right-hand side of the TTA. That is, the vhCDR from the variable heavy chain CD3 H1.30 sequence and the vlCDR from the variable light chain of the CD3 L1.57 sequence can be combined with the vhCDR from the CD38 OKT10 H1.77 sequence and the vlCDR from the OKT10L1.24 sequence. "B" indicates that the CDR from the CD3 construct can be combined with the variable heavy chain and variable light chain domains of the TTA. That is, the vhCDR from the variable heavy chain CD3 H1.30 sequence and the vlCDR from the variable light chain of the CD3 L1.57 sequence can be combined with the variable heavy chain domains CD38 OKT10 H1.77 and OKT10L1.24 sequences. "C" indicates the opposite, i.e., the variable heavy chain and variable light chain domains of the CD3 sequence are used in conjunction with the CDR of the TTA. "D": Combination of variable heavy chain and variable light chain in each. "E": CD3 scFv combined with TTA CDR; "F": CD3 scFv combined with the variable heavy chain and variable light chain domains of the TTA antigen-binding domain. These combinations can be made into bottle opener forms, such as the various backbone patterns in Figure 162, or into other forms, such as... Figure 1 The mAb-Fv, mAb-scFv, center-scFv, center-Fv, or double-scFv forms include Figure 131 and 132 The pattern shown is the backbone. However, generally speaking, forms containing divalent CD3 binding are not recommended. That is, "A" (CD3CDR×TTA CDR) can be added to the opener sequence (including those in Figure 162 or various different heterodimerized variants) or added to... Figure 132 The mAb-scFv trunk, center-scFv, mAb-Fv form, or center-Fv form are used.

[0153] Figure 125 Schematic diagram of anti-CD123×anti-CD3 Fab-scFv-Fc bispecific antibody.

[0154] Figure 126 The table shows that 7G3_H1L1 is a modified variant with improved affinity and stability.

[0155] Figure 127 The table shows the properties of the final affinity and stability optimized 7G3 humanized variant.

[0156] Figure 128 The binding of XENP14045 (anti-CD123×anti-CD3) bispecific antibody to the CD123-positive AML cell line KG-1a.

[0157] Figure 129 XENP14045 kills KG-1a cells via redirected T cell cytotoxicity (RTCC).

[0158] Figure 130 XENP14045 was used to induce T cell "serial killing" of target cells in KG-1a cells by different effector cell to target cell ratios (E:T).

[0159] Figure 131 Serum drug concentration in (IV)C57BL / 6 mice after intravenous administration of 2 mg / kg XENP14045; half-life of the bispecific antibody is 6.2 days.

[0160] Figure 132 The killing of CD123+ basophils and plasmacytoid dendritic cells (PDCs) in the blood of rhesus monkeys after receiving a single intravenous (IV) dose of 0.01, 0.1, or 1 mg / kg XENP14045.

[0161] Figure 133 The killing of CD123+ basophils and plasmacytoid dendritic cells (PDCs) in the bone marrow of rhesus monkeys after receiving a single intravenous (IV) dose of 0.01, 0.1, or 1 mg / kg XENP14045.

[0162] Figure 134 T cell redistribution in macaques after receiving a single intravenous (IV) dose of XENP14045.

[0163] Figure 135 T cells in macaques were induced to CD69 after receiving a single intravenous (IV) dose of XENP14045.

[0164] Figure 136A-136C The sequence of this invention. The CDR area is underlined.

[0165] Figure 137 The purity of heterodimers during stable aggregate formation was measured using different ratios of light chains, Fab-Fc, and scFv-Fc (top table); the purity of heterodimers of F2 aggregates under different conditions (bottom table).

[0166] Figure 138 SEC results show high purity of XENP14045 cell line material after two-step purification.

[0167] Figure 139 T cells kill CD123+ cells.

[0168] Figure 140The mechanism of bispecific antibodies that aggregate cytotoxic T cells to kill AML stem cells and blast cells.

[0169] Figure 141 Effective production of XENP14045 bispecific antibody.

[0170] Figure 142 The XENP14045 bispecific antibody binds to human AML, with a KD of 8.1 nM with human CD3.

[0171] Figure 143 The XENP14045 bispecific antibody exhibits cross-reactivity with primate cells, and its KD with rhesus monkey (Cyno) CD3 is 5.7 nM.

[0172] Figure 144 Anti-CD123 × anti-CD3 killing human AML cell lines.

[0173] Figure 145 Anti-CD123 × anti-CD3 killing human AML cell lines.

[0174] Figure 146 : The long half-life of bispecific antibodies in mice.

[0175] Figure 147 : Single dose to monkeys.

[0176] Figure 148 Clearance of CD123+ cells from monkey blood basophils. Flow cytometry analysis showed the basophil phylum to be: CD20-CD16+CD14-CD4-CD8-FceR1+.

[0177] Figure 149 Bone marrow basophil clearance was performed using the same phylogenetic setup.

[0178] Figure 150 In monkeys, repeated administration cleared CD123+ cells.

[0179] Figure 151 In monkeys, the clearance of CD123+ cells. Flow cytometry showed that the basophil phylum was: CD20-CD16+CD14-CD4-CD8-FceR1+. Flow cytometry showed that the plasmacytoid dendritic cell phylum was: CD20-CD16-CD14-Cd4-CD8-CD303+.

[0180] Figure 152 : Clearance from the bone marrow of monkeys. (Same as above) Figure 151 .

[0181] Figure 153 CD123+ cell clearance is associated with T cell redistribution and activation; Figure 153T cell redistribution.

[0182] Figure 154 CD123+ cell clearance is associated with T cell redistribution and activation; Figure 154 T cell activation.

[0183] Figure 155 CD123+ cell clearance is associated with T cell redistribution and activation; Figure 155 Cytokine release.

[0184] Figures 156A-156D Materials related to the difficulty of humanizing anti-CD123 mouse sequences, as described in Example 3. Figure 156A-C Humanization showed decreased affinity (mainly vH). 13760 is the Fab of the H0L0 initiating mouse antibody, 13763 is the first humanized candidate vH, and both the heavy and light chain Fabs of 13761 have been humanized. Figure 156D Humanization reduces RTCC efficiency by approximately 10 times.

[0185] Figure 157 The results of the first round of humanization (“Library 1”) are shown, which produced 108 variants, including LDA, targeted and reversion substitution, which underwent affinity screening in the form of Fab on the Biacore CD123 chip, while the neutral and high affinity variants underwent stability screening on DSF.

[0186] Figure 158A and 158B The Tm level was elevated, as described in Example 3.

[0187] Figure 159A and 159B The results show the transformation of each Fab into a bottle opener form, using anti-CD3 scFv and the developed Fab. Figure 159A The combined experiment showed that Figure 159B Display RTCC test.

[0188] Figure 160A-160E The results of the "second round" of humanization are shown, as described in Example 3. It should be noted that XENP13967 is equivalent to XENP14045 on the CD123 side, but differs in CD3 scFv, as shown in their sequences.

[0189] Figure 161 The second round of Tm testing is shown in Example 3.

[0190] Figure 162A-162DSeveral useful opener-type trunk sequences are shown, but Fv sequences (such as scFv and vh and vl on the Fab side) are not shown. As will be apparent to those skilled in the art and as described below, these sequences can be paired with any vh / vl herein, such that one monomer contains scFv (optionally containing a charged scFv connector) and another monomer contains Fab sequences (e.g., vh coupled to the “Fab-side heavy chain” and vl coupled to the “constant light chain”). The scFv can be CD3-resistant or TTA-resistant, and the Fab is resistant to the other. That is, the various Fv sequences for CD3, CD123, CD38, CD19, or CD20 herein can be arbitrarily combined and included in the various trunk patterns shown in FIG162.

[0191] It should be noted that these bottle opener backbone sequences can be used Figure 1 The central-scFv form in B, wherein a second Fab (vh-CH1 and vl-constant light chain) is added to the N-terminus of the scFv on the "opener side" to bind the same antigen as the first Fab.

[0192] Figure 163 The sequence of the mAb-scFv backbone of this invention is shown, with the addition of the Fv sequence of this invention. As will be apparent to those skilled in the art and as described below, these sequences can be conjugated with any vh / vl pair herein, such that one monomer contains both Fab and scFv (optionally containing a charged scFv connector) while the other monomer contains a Fab sequence (e.g., a vh conjugated to the “Fab-side heavy chain” and a vl conjugated to the “constant light chain”). Monomer 1 is the negative pI side of Fab-scFv and contains heterodimer variants L368D / K370S, isoelectric pI variants N208D / Q295E / N384D / Q418E / N421D, and elimination variants E233P / L234V / L235A / G236del / S267K, (all relative to IgG1). Monomer 2 is the positive pI side of scFv and has the heterodimer variant 364K / E357Q. However, it can also be replaced by other biased variant pairs, such as [S364K / E357Q:L368D / K370S], [L368D / K370S:S364K], [L368E / K370S:S364K], [T411T / E360E / Q362E:D401K], [L368D / K370S:S364K / E357L] and [K370S:S364K / E357Q]. Invention Details

[0193] I. Definition

[0194] To provide a more complete understanding of this application, several definitions are given below. These definitions include grammatical synonyms and equivalents.

[0195] In this text, "elimination" means a reduction or removal of activity. Therefore, for example, "elimination of FcγR binding" means that the binding of an Fc region amino acid variant is less than 50% of the initial binding compared to the Fc region without the specific variant, preferably a loss of 70-80-90-95-98% or more of activity, and overall, the activity is lower than the level detectable in a biacore assay. Those that are particularly useful for eliminating FcγR binding are those... Figure 16 .

[0196] As used in this article, "ADCC" or "antibody-dependent cell-mediated cytotoxicity" refers to the cell-mediated reaction described below: non-specific cytotoxic cells expressing FcγR recognize antibodies bound to target cells, subsequently causing target cell lysis. ADCC is associated with FcγRIIIa binding; increased FcγRIIIa binding leads to enhanced ADCC activity.

[0197] The term “ADCP” or antibody-dependent cell-mediated phagocytosis used in this article refers to the cell-mediated response described below: nonspecific cytotoxic cells expressing FcγR recognize antibodies bound to target cells, subsequently inducing phagocytosis of the target cells.

[0198] As used herein, “modification” refers to the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence, or changes in chemically linked structural units on a protein. For example, modification can be a change in the structure of a sugar or PEG bonded to a protein. “Amino acid modification” as used herein refers to the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence. For clarity, unless otherwise stated, amino acid modifications always refer to the DNA-encoded amino acids, for example, the 20 amino acids that have both DNA and RNA codons.

[0199] In this document, "amino acid substitution" or "replacement" means replacing an amino acid at a specific position in the parental polypeptide sequence with a different amino acid. Specifically, in some embodiments, substitution refers to an amino acid that is not naturally present at a certain position, either in that organism or in any organism. For example, substitution of E272Y refers to a variant polypeptide, in this case the Fc variant, in which glutamic acid at position 272 is replaced by tyrosine. For clarity, proteins whose nucleic acid coding sequence has been modified without changing the initial amino acid (e.g., changing CGG (encoding arginine) to CGA (still encoding arginine) to improve host expression levels) are not considered "amino acid substitution"; that is, although a new gene encoding the same protein is generated, if the specific position of the protein still contains the original, identical amino acid, it is not an amino acid substitution.

[0200] As used in this article, "amino acid insertion" or "insertion" refers to the addition of an amino acid sequence at a specific position in the parental polypeptide sequence. For example, -233E or 233E means the insertion of glutamic acid after position 233 and before position 234. Similarly, -233ADE or A233ADE means the insertion of AlaAspGlu after position 233 and before position 234.

[0201] As used in this article, "amino acid deletion" or "deletion" refers to the removal of an amino acid sequence at a specific position in the parental polypeptide sequence. For example, E233-, E233#, or E233() indicates the deletion of glutamic acid at position 233. Additionally, EDA233- or EDA233# indicates the deletion of the GluAspAla sequence at position 233.

[0202] As used herein, "variant protein," "protein variant," or "variant" refers to a protein that differs from a parent protein by at least one amino acid modification. A protein variant can refer to the protein itself, a composition containing the protein, or the amino acid sequence encoding the protein. Preferably, the protein variant has at least one amino acid modification compared to the parent protein, for example, about 1 to about 70 amino acid modifications compared to the parent, more preferably about 1 to about 5 amino acid modifications. As described below, in some embodiments, the parent polypeptide, such as an Fc parent polypeptide, is a human wild-type sequence, such as from the Fc region of IgG1, IgG2, IgG3, or IgG4; however, human sequences having variants can also be used as "parent polypeptides," for example, Figure 19The IgG1 / 2 hybrid is described herein. The protein variant sequence is preferably at least about 80% identical to the parental protein sequence, preferably at least about 90%, and more preferably at least about 95-98-99%. Variant protein may refer to the variant protein itself, a composition containing the protein variant, or the DNA sequence encoding the variant protein. Therefore, as used herein, “antibody variant” or “variant antibody” means an antibody that differs from the parental antibody by at least one amino acid modification; “IgG variant” or “variant IgG” means an antibody that differs from the parental IgG (again, in many cases, the human IgG sequence) by at least one amino acid modification; and “immunoglobulin variant” or “variant immunoglobulin” means an immunoglobulin sequence that differs from the parental immunoglobulin sequence by at least one amino acid modification. “Fc variant” or “variant Fc” means a protein containing an amino acid modification within the Fc domain. The Fc variants of the present invention are described according to the amino acid modifications they contain. Therefore, for example, N434S or 434S is an Fc variant with a substituted serine residue at position 434 compared to the parental Fc peptide, where the coding is according to the EU index. Similarly, M428L / N434S describes an Fc variant with substitutions of M428L and N434S compared to the parental Fc peptide. Wild-type (WT) amino acids may not be specified, for example, variants described as 428L / 434S. It should be noted that the order of substitutions described is arbitrary, that is, for example, 428L / 434S and M428L / N434S are the same Fc variant, and so on. All positions related to antibodies in this invention, unless otherwise stated, are numbered according to the EU index. The EU index in the EU index or Kabat or EU coding scheme refers to the EU antibody coding system (Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85, incorporated herein by reference in its entirety). Modifications can be additions, deletions, or substitutions. Substitutions may include natural amino acids, and sometimes synthetic amino acids.Examples include U.S. Patent Nos. 6,586,207; WO 98 / 48032; WO 03 / 073238; US2004-0214988A1; WO 05 / 35727A2; WO 05 / 74524A2; JWChin et al., (2002), Journal of the American Chemical Society 124:9026-9027; JWChin and PGSchultz, (2002), ChemBioChem 11:1135-1137; JWChin et al., (2002), PICAS United States of America 99:11020-11024; and L.Wang and PGSchultz, (2002), Chem. 1-10, all of which are incorporated herein by reference.

[0203] As used herein, “protein” refers to at least two covalently bonded amino acids, including proteins, polypeptides, oligopeptides, and peptides. A peptide group may comprise a native amino acid and a peptide bond or a synthetic peptide-like structure, i.e., an “analyte,” such as a peptide-like substance (see Simon et al., PNAS USA 89(20):9367(1992), incorporated herein by reference in its entirety). Amino acids may be native or synthetic (e.g., amino acids not encoded by DNA); as will be understood by those skilled in the art. For example, homophenylalanine, citrulline, ornithine, and leucine are considered synthetic amino acids in this invention, and amino acids with both D- and L- (R or S) conformations may be used. Variations of the invention may include modifications using synthetic amino acids, which may be incorporated using techniques developed by, for example, Schultz et al., including but not limited to, the methods described in: Cropp and Shultz, 2004, Trends Genet. 20(12):625-30; Anderson et al., 2004, Proc Natl Acad Sci USA 101(2):7566-71; Zhang et al., 2003, 303(5656):371-3; and Chin et al., 2003, Science 301(5635):964-7, all of which are incorporated herein by reference in their entirety. Additionally, peptides may include one or more side chains, synthetic derivatization of one or more ends, glycosylation, PEGylation, cyclic arrangement, cyclization, linkers to other molecules, fusion with proteins or protein domains, and the addition of peptide tags or labels.

[0204] As used in this article, “residue” refers to a position in a protein and the amino acid at that position. For example, asparagine 297 (also known as Asn297 or N297) refers to the residue at position 297 in human antibody IgG1.

[0205] As used herein, “Fab” or “Fab region” refers to a polypeptide containing the VH, CH1, VL, and CL immunoglobulin domains. Fab can refer to the isolated region or the region contained within a full-length antibody, antibody fragment, or Fab fusion protein. As used herein, “Fv” or “Fv fragment” or “Fv region” refers to a polypeptide containing the VL and VH domains of a single antibody. Those skilled in the art will recognize that these generally consist of two chains.

[0206] The terms "IgG subclass modification" or "isotype modification" used in this article refer to amino acid modifications that change an amino acid in one IgG isotype to the corresponding amino acid in another parallel IgG isotype. For example, since IgG1 has tyrosine at position 296 in the EU and IgG2 has phenylalanine, the F296Y substitution in IgG2 is considered an IgG subclass modification.

[0207] As used in this article, “non-natural modification” refers to a non-isotype amino acid modification. For example, since serine is not present at position 434 in all IgGs, the substitution of 434S in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a non-natural modification.

[0208] The terms “amino acid” and “specific amino acid” used in this article refer to one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0209] As used in this article, "effective function" refers to a biochemical event resulting from the interaction between the antibody's Fc region and an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

[0210] As used herein, “IgG Fc ligand” refers to such molecules, preferably peptides, from various organisms that bind to the Fc region of an IgG antibody to form an Fc / Fc ligand complex. Fc ligands include, but are not limited to, FcγRI, FcγRII, FcγRIII, FcRn, C1q, C3, mannan-binding lectins, mannose receptors, staphylococcal protein A, streptococcal protein G, and viral FcγR. Fc ligands also include Fc receptor homologs (FcRH), a family of Fc receptors homologous to FcγR (Davis et al., 2002, Immunological Reviews 190:123-136, incorporated herein by reference in full). Fc ligands include molecules that have not yet been identified that bind Fc. Specific IgG Fc ligands are FcRn and Fcγ receptors. As used herein, “Fc ligand” refers to such molecules, preferably peptides, from various organisms that bind to the Fc region of an antibody to form an Fc / Fc ligand complex.

[0211] As used in this article, “Fcγ receptor,” “FcγR,” or “FcqammaR” refers to a member of a protein family that binds to the Fc region of IgG antibodies and is encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI (CD64), including subtypes FcγRIa, FcγRIb, and FcγRIc; FcγRII (CD32), including subtypes FcγRIIa (including allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and FcγRIIc; and FcγRIII (CD16), including subtypes FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIb-NA1 and FcγRIIb-NA2) (Jefferis et al., 2002, Immunol Lett 82:57-65, incorporated herein by reference in its entirety), and also includes undiscovered human FcγR or FcγR subtypes or allotypes. FcγR can be derived 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 undiscovered mouse FcγR or FcγR subtypes or allotypes.

[0212] As used herein, the term "FcRn" or "neonatal Fc receptor" refers to a protein that binds to the Fc region of an IgG antibody and is at least partially encoded by the FcRn gene. FcRn can be derived from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, functional FcRn proteins comprise two polypeptides, 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 stated herein, FcRn or FcRn protein refers to the complex of the FcRn heavy chain and β-2-microglobulin. Various FcRn variants are used to enhance binding to the FcRn receptor and sometimes to prolong serum half-life, see [link to relevant documentation]. Figure 83 And its explanation.

[0213] The term "parental polypeptide" as used herein refers to an initial polypeptide that has been modified to form a variant. A parental polypeptide can be a natural polypeptide, or a variant or modified version of a natural polypeptide. A parental polypeptide can refer to the polypeptide itself, a composition containing a parental polypeptide, or the amino acid sequence encoding a parental polypeptide. Therefore, the term "parental immunoglobulin" as used herein refers to an unmodified immunoglobulin polypeptide that has been modified to form a variant, and the term "parental antibody" as used herein refers to an unmodified antibody that has been modified to form a variant antibody. It should be noted that "parental antibody" includes known commercially available, recombinant antibodies, as described below.

[0214] As used herein, “Fc”, “Fc region”, or “Fc domain” refers to a polypeptide containing the portion of the antibody constant region excluding the first constant region, the immunoglobulin domain, and sometimes excludes a portion of the hinge region. Therefore, Fc refers to the latter two constant regions of the immunoglobulin domains of IgA, IgD, and IgG, the latter three constant regions of the immunoglobulin domains of IgE and IgM, and the flexible hinge region at the N-terminus of these domains. For IgA and IgM, Fc may contain the J chain. For IgG, the Fc region includes the lower portion of the hinge region between the immunoglobulin domains Cγ2 and Cγ3 (Cγ2 and Cγ3) and between Cγ1 (Cγ1) and Cγ2 (Cγ2). Although the boundaries of the Fc region are variable, the human IgG heavy chain Fc region is generally defined as the region containing residues C226 or P230 to the carboxyl terminus, numbered according to Kabat’s EU index. In some implementations, as described below, the Fc region is modified with amino acids, for example, thereby altering its binding to one or more FcγR receptors or FcRn receptors.

[0215] In this article, "heavy chain constant region" refers to the CH1-hinge-CH2-CH3 portion of the antibody.

[0216] As used herein, "Fc fusion protein" or "immunoadhesin" refers to a protein containing an Fc region that is generally linked (optionally via a linker, as described herein) to another protein, such as a binding portion for a target protein, as described herein. In some cases, one monomer of a heterodimeric antibody contains an antibody heavy chain (or includes scFV or also includes a light chain) and the other monomer is an Fc fusion body containing a variable Fc domain and a ligand. In some embodiments, these "half-antibody-half-fusion proteins" are referred to as "fusion bodies."

[0217] The term "position" as used in this article refers to a location within a protein sequence. Positions can be numbered sequentially or according to existing formats, such as the EU index used for antibody numbering.

[0218] As used in this article, "target antigen" refers to a molecule that specifically binds to the variable region of a given antibody. Target antigens can be proteins, sugars, lipids, or other compounds. Several suitable target antigens are described below.

[0219] Regarding the monomers of the heterodimeric antibodies of this invention, "chain-like" means that, similar to "matching" DNA double strands, heterodimeric variants are introduced into each monomer to retain the ability to "match" each other to form heterodimers. For example, if certain pI variants are modified to monomer A (e.g., to increase pI), stereotypes of "charge pairs" can also be used without interfering with the pI variants; for example, charge variants that increase pI are placed on the same "chain" or "monomer" to retain both functional groups. Similarly, for the "skew" variants appearing in the pairing combinations below, those skilled in the art will consider pI when deciding which chain or monomer to include a pair in, thereby maximizing pI separation by utilizing the pI of the skew variant.

[0220] The term "target cell" as used in this article refers to a cell that expresses the target antigen.

[0221] The “variable region” used in this article refers to the region of an immunoglobulin as described below, which contains one or more Ig domains encoded by VH genes that are essentially composed of V.κ, V.λ and / or the κ, λ and heavy chain immunoglobulin loci, respectively.

[0222] In this article, "wild-type or WT" refers to an amino acid or nucleotide sequence found in nature, including allelic variations. WT proteins have an amino acid or nucleotide sequence that has not yet been specifically modified.

[0223] The antibodies of this invention are generally isolated or recombinant. When used to describe the various polypeptides disclosed herein, “isolated” means that the polypeptide has been isolated and / or recovered and identified from the cells or cell cultures in which it is expressed. Typically, isolated polypeptides undergo at least one purification step in their preparation. “Isolated antibody” refers to an antibody that is substantially free of other antibodies with different antigen specificities. “Recombinant” means that the antibody is generated in exogenous host cells using recombinant nucleic acid technology.

[0224] "Specific binding," "specific binding to," or "specific targeting" of a specific antigen or epitope indicates that the binding is measurably distinct from non-specific interactions. For example, specific binding can be determined by measuring the binding of a molecule and comparing it to the binding of a control molecule, which is typically a structurally similar molecule but without binding activity. For instance, specific binding can be determined by competition with a control molecule similar to the target.

[0225] Specific binding to a particular antigen or epitope can manifest as the K-axis of the antibody against the antigen or epitope. D 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 higher, where K D This refers to the dissociation rate of a specific antibody-antigen interaction. Generally speaking, the Kc of an antibody that specifically binds to a certain antigen is... D It is 20, 50, 100, 500, 1000, 5,000, 10,000 or more times higher than the antigen or epitope-related control molecule.

[0226] Furthermore, specific binding to a particular antigen or epitope can also manifest as, for example, the K-axis of an antibody against a certain antigen or epitope. A or K a It was at least 20, 50, 100, 500, 1000, 5000, 10000 or more times higher than the control, where K A or K a This refers to the binding rate of a specific antibody-antigen interaction. Binding affinity is generally determined using the Biacore assay.

[0227] II. Overview

[0228] Bispecific antibodies that co-bind to CD3 and tumor antigen targets have been designed to redirect T cells to attack and lyse targeted tumor cells. Examples include BiTE and DART types, which monovalently bind to both CD3 and tumor antigens. While CD3-targeting strategies have shown considerable promise, a common side effect of these therapies is associated cytokine production, often leading to harmful cytokine release syndrome. Because the anti-CD3 binding domain of bispecific antibodies binds to all T cells, it recruits a high-cytokine-producing CD4 T cell subset. Furthermore, the CD4 T cell subset includes regulatory T cells, whose recruitment and expansion can lead to immunosuppression and negatively impact long-term tumor suppression. Additionally, these forms of bispecific antibodies lack an Fc domain and have a very short serum half-life in patients.

[0229] While CD3-targeting strategies have shown considerable promise, a common side effect of these therapies is associated cytokine production, often leading to harmful cytokine release syndrome. This is because the anti-CD3 binding domain of bispecific antibodies binds to all T cells, recruiting a high-cytokine-producing CD4 T cell subset. Furthermore, this CD4 T cell subset includes regulatory T cells, whose recruitment and expansion can lead to immunosuppression and negatively impact long-term tumor suppression. A feasible way to reduce cytokine production and potentially decrease CD4 T cell activation is by reducing the affinity of the anti-CD3 domain for CD3.

[0230] Therefore, in some embodiments, the antibody constructs provided by the present invention have an anti-CD3 antigen-binding domain that is a “strong” or “high-affinity” binder to CD3 (e.g., variable heavy and light chain domains of H1.30-L1.47 (optionally including suitable charged linkers)) and also bind CD38. In other embodiments, the antibody constructs provided by the present invention have an anti-CD3 antigen-binding domain that is a “weak” or “low-affinity” binder to CD3. In still other embodiments, the antibody constructs have an anti-CD3 antigen-binding domain that has a moderate or “moderate” affinity for CD3 and also bind CD38. Affinity is generally determined using the Biacore assay.

[0231] It should be understood that the “high, medium, and low” anti-CD3 sequences of this invention can be used for various forms of heterodimerization. While “bottle opener” type heterodimers are predominantly used herein, these variable heavy chain and variable light chain sequences, as well as scFv sequences (and Fab sequences comprising variable heavy chain and variable light chain sequences), can be used for other forms, such as those described in WO Publication No. 2014 / 145806. Figure 2 The accompanying drawings, illustrations, and descriptions of the drawings are incorporated herein by reference.

[0232] Therefore, this invention provides heterodimeric antibodies capable of binding to two different antigens, for example, antibodies that are "bispecific," meaning they bind to two different target antigens, typically tumor antigens (TTAs) as described below. These heterodimeric antibodies can bind to these target antigens monovalently (e.g., with only one antigen-binding domain, such as a variable heavy chain domain and a variable light chain domain pair) or bivalently (with two antigen-binding domains, each binding an antigen independently). The heterodimeric antibodies of this invention are based on the use of different monomers containing amino acid substitutions (as described below) that are "biased" in forming heterodimers compared to homodimers, combined with "pI variants" that facilitate the purification and separation of the heterodimer from the homodimer (also described below). For the heterodimeric bispecific antibodies of this invention, this invention primarily relies on the use of modified Fc domains or variant Fc domains that can self-assemble into heterodimeric proteins in producing cells, and this invention is also based on methods for generating and purifying such heterodimeric proteins.

[0233] III. Antibodies

[0234] This invention relates to the generation of bispecific antibodies that bind to two different antigens, such as CD3, and target tumor antigens such as CD19, CD20, CD38, and CD123, and are therapeutic antibodies in general. The term "antibody" is used in a broad sense. Antibodies that can be used in this invention can take many forms as described herein, including conventional antibodies as well as antibody derivatives, fragments, and mimics, as described herein.

[0235] Conventional antibody structural units generally include tetramers. Each tetramer typically consists of two identical pairs of polypeptide chains, each pair having one "light chain" (generally with a molecular weight of about 25 kDa) and one "heavy chain" (generally with a molecular weight of about 50-70 kDa). Human light chains are classified as κ and λ light chains. This invention relates to IgG classes, which have several subclasses, including but not limited to IgG1, IgG2, IgG3, and IgG4. It should be noted that IgG1 has multiple allotypes, exhibiting polymorphism at positions 356 (D or E) and 358 (L or M). The sequence shown here uses the 356D / 358M type, but other allotypes are also within the scope of this application. That is, any sequence containing the IgG1 Fc domain herein can be 356E / 358L, and not just 356D / 358M.

[0236] Furthermore, in many sequences described herein, at least cysteine ​​at position 220 is replaced by serine; typically, for most sequences described herein, this is located on the "scFv monomer" side, but it can also be located on the "Fab monomer" side, or on both sides, in order to reduce disulfide bond formation. Specific sequences described herein contain one or both of the aforementioned cysteine ​​substitutions (C220S).

[0237] Therefore, as used herein, “isotype” refers to the various subclasses of immunoglobulins determined by the chemical and antigenic characteristics of constant regions. It should be understood that therapeutic antibodies may also comprise hybrids of isotypes and / or subclasses. As described in U.S. Publication No. 2009 / 0163699, incorporated herein by reference, this invention includes pI modification of IgG1 / G2 hybrids.

[0238] The amino-terminal portion of each chain includes a variable region of approximately 100-110 or more amino acids, which is primarily responsible for antigen recognition and is generally referred to in this art and herein as the "Fv domain" or "Fv region". Within the variable region, three loops in the V domains of both the heavy and light chains together form the antigen-binding site. Each loop is called a complementarity-determining region (hereinafter referred to as "CDR"), where the amino acid sequence changes most significantly. "Variable" refers to the wide sequence variation between antibodies in certain segments of the variable region. The variation within the variable region is not uniformly distributed. In fact, the V region consists of a relatively stable 15-30 amino acid extension called the frame region (FR) and highly variable short regions of 9-15 amino acids or longer distributed between the frame regions, called "hypervariates".

[0239] Each VH and VL consists of three hypervariable regions (“complementary determinant regions”, “CDRs”) and four FRs, arranged in the following order from the amino terminus to the carboxyl terminus: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

[0240] Hypervariable regions generally include approximately amino acid residues 24-34 (LCDR1; "L" indicates light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the light chain variable region, and approximately amino acid residues 31-35B (HCDR1; "H" indicates heavy chain), 50-65 (HCDR2), and 95-102 (HCDR3) in the heavy chain variable region; Kabat et al., *SEQUENCES OF PROTEINS OF IMMUNOLOGICAL INTEREST*, 5th edition, Public Health Service, National Institutes of Health, Bethesda, Maryland. Health, Bethesda, Md. (1991) and / or those residues forming hypervariable rings (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3) in the light chain variable region and residues 26-32 (HCDR1), 53-55 (HCDR2), and 96-101 (HCDR3) in the heavy chain variable region; Chothia and Lesk (1987) J. Mol. Biol. 196: 901-917. Specific CDRs of the present invention are described below.

[0241] As those skilled in the art will know, the exact numbering and location of CDRs may differ across numbering systems. However, it must be understood that revealing the variable heavy chain and / or variable light chain sequence reveals the associated CDR. Thus, revealing the variable region of each heavy chain reveals each vhCDR (e.g., vhCDR1, vhCDR2, and vhCDR3), and revealing the variable region of each light chain reveals each vlCDR (e.g., vlCDR1, vlCDR2, and vlCDR3).

[0242] In this specification, when referring to residues in variable domains, the Kabat numbering system is generally used (roughly, light chain variable region residues 1-107 and heavy chain variable region residues 1-113), and the Fc region is numbered using the EU numbering system (e.g., Kabat et al., ibid. (1991)).

[0243] This invention provides a large number of different CDR combinations. In this case, a “full CDR set” comprises three variable light chain CDRs and three variable heavy chain CDRs, for example, vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. This can be localized to larger variable light chain and variable heavy chain domains, respectively. Furthermore, as discussed in more detail herein, the variable heavy chain and variable light chain domains can reside in separate polypeptide chains, when using heavy and light chains (e.g., when using Fab), or on the same polypeptide chain, such as an scFv sequence.

[0244] CDRs (corresponding antigen-binding sites), or more specifically, epitope-binding sites, are involved in the formation of antigen-binding sites. An epitope is a determinant that interacts with a specific antigen-binding site ("antibody determinant") within the variable region of an antibody molecule. Epitopes are composed of molecules such as amino acids or sugar side chains and typically possess specific structural and charge characteristics. An antigen may have more than one epitope.

[0245] Epitopes may contain amino acid residues that directly participate in binding (also known as the immunodeterminant component of the epitope) and other amino acid residues that do not participate in binding, such as amino acid residues that are effectively blocked by the specific antigen-binding peptide; in other words, these amino acid residues are included within the footprint of the specific antigen-binding peptide.

[0246] Epitopes can be conformational or linear. Conformational epitopes are formed by the spatial arrangement of amino acids in different segments of a linear polypeptide chain. Linear epitopes are formed by adjacent amino acid residues in a polypeptide chain. The difference between conformational and non-conformational epitopes is that binding to conformational epitopes is lost in the presence of denaturing solvents, while binding to non-conformational epitopes is not.

[0247] Epitopes typically consist of at least three, or more commonly at least five or eight to ten amino acids, in a unique spatial conformation. Antibodies that recognize the same epitope can be identified by simple immunoassays, such as binning, which demonstrate the ability of one antibody to block the binding of another antibody to a target antigen.

[0248] The carboxyl terminus of each chain is defined as the constant region primarily responsible for effector function. Kabat et al. compiled the primary sequences of numerous variable regions of heavy and light chains. Based on the degree of sequence conservation, each primary sequence was classified into CDRs and framework regions and listed (see Sequences of Proteins of Immunological Interest, 5th Edition, NIH Publication No. 91-3242, E.A. Kabat et al., included by full citation).

[0249] In the IgG subclass of immunoglobulins, multiple immunoglobulin domains exist within the heavy chain. In this document, "immunoglobulin (Ig) domain" refers to a region of immunoglobulin possessing a unique tertiary structure. This invention is of interest to the heavy chain domain, including the constant heavy chain (CH) domain and the hinge domain. For IgG antibodies, each IgG isotype has three CH regions. Therefore, the CH domains of IgG are as follows: according to the Kabat EU index, "CH1" refers to positions 118-220; "CH2" refers to positions 237-340; and "CH3" refers to positions 341-447. As described herein, pI variants may be present in one or more CH regions as well as the hinge region, as discussed later.

[0250] It should be noted that the sequence here starts at CH1, position 118; it does not include the variable region unless otherwise stated. For example, the first amino acid of SEQ ID NO:2, although indicated as position "1" in the sequence listing, corresponds to position 118 in the CH1 region according to the EU number.

[0251] Another type of heavy chain Ig domain is the hinge region. As used herein, “hinge,” “hinge region,” “antibody hinge region,” or “immunoglobulin hinge region” refers to a flexible polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain terminates at EU position 220, and the IgG CH2 domain begins at EU position 237. Therefore, for IgG, the antibody hinge is defined herein as including positions 221 (D221 in IgG1) through 236 (G236 in IgG1), where the numbering is based on the Kabat EU index. In some embodiments, for example, with respect to the Fc region, a lower hinge portion is included; “lower hinge portion” generally refers to position 226 or 230. As described herein, pI variants can also be fabricated in the hinge region.

[0252] Light chains generally consist of two structural domains: the light chain variable structural domain (containing the light chain CDR and forming the Fv region together with the heavy chain variable structural domain) and the light chain constant region (usually called CL or Cκ).

[0253] Another area that is considered as a replacement is the Fc region.

[0254] Therefore, this invention provides different antibody domains. As described herein and known in the art, the heavy and light chains of the heterodimeric antibodies of this invention contain different domains, which may also overlap. These domains include, but are not limited to, Fc domains, CH1 domains, CH2 domains, CH3 domains, hinge domains, heavy chain constant domains (CH1-hinge-Fc domains or CH1-hinge-CH2-CH3), variable heavy chain domains, variable light chain domains, light chain constant domains, FAb domains, and scFv domains.

[0255] Therefore, the “Fc domain” includes the -CH2-CH3 domain and an optional hinge domain. In the embodiments described herein, when the scFv is connected to the Fc domain, the C-end of the scFv construct is connected to the hinge of the Fc domain, for example, typically to the EPKS sequence at the beginning of the hinge. The heavy chain includes a variable heavy chain domain and a constant domain, which includes the CH1-optional hinge-Fc domain, the Fc domain containing CH2-CH3. The light chain includes a light chain variable domain and a light chain constant domain. The scFv includes a variable heavy chain, an scFv connector, and a variable light chain domain. In most constructs and sequences described herein, the C-end of the variable light chain is connected to the N-end of the scFv connector, and the C-end of the scFv connector is connected to the N-end of the variable heavy chain (N-vh-connector-vl-C), which can also be interchanged (N-vl-connector-vh-C).

[0256] Some embodiments of the present invention include at least one scFv domain, which is not natural and generally includes a variable heavy chain domain and a variable light chain domain linked together by an scFv linker. As shown herein, a variety of suitable scFv linkers can be employed, including conventional peptide bonds generated by recombination techniques.

[0257] The linker peptide may primarily comprise the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should be of sufficient length to ensure that the two molecules are correctly conformated together to retain their respective activities. In one embodiment, the linker is about 1-50 amino acids long, preferably about 1-30 amino acids long. In one embodiment, linkers of 1-20 amino acids in length can be used, and in some embodiments, linkers of about 5 to about 10 amino acids are used. Available 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 1 (generally 3-4), glycine-alanine polymers, alanine-serine polymers, and other flexible linkers. Alternatively, a variety of non-protein polymers can be used as linkers, including but not limited to polyethylene glycol (PEG), polypropylene glycol, polyoxyethylene, or copolymers of polyethylene glycol and polypropylene glycol.

[0258] Other linker sequences may include sequences of various lengths within the CL / CH1 domain, but not all residues of the CL / CH1 domain; for example, the first 5-12 amino acid residues of the CL / CH1 domain. Linkers may be derived from immunoglobulin light chains, such as Cκ or Cλ. Linkers may be derived from various isotypes of immunoglobulin heavy chains, such as Cγ1, Cγ2, Cγ3, Cγ4, Cα1, Cα2, Cδ, Cε, and Cμ. Linker sequences may also originate from other proteins, such as Ig-like proteins (e.g., TCR, FcR, KIR), sequences derived from the hinge region, and other natural sequences from other proteins.

[0259] In some embodiments, the linker is a “domain linker” used to connect any two domains described herein. While various suitable linkers can be used, many embodiments employ glycine-serine polymers, including, for example, (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer of at least 1 (generally 3 to 4 to 5), and various peptide sequences of sufficient length and flexibility to recombine the two domains while preserving their respective biological functions. In some cases, where a “chain-like” configuration is also noted, charged domain linkers, such as those used in some embodiments of the scFv linker, can be used, as described later.

[0260] In some embodiments, the scFv connector is a charged scFv connector, and Figure 33 shows several such connectors. Therefore, the present invention also provides charged scFv connectors to facilitate the separation of pI between the first and second monomers. That is, by incorporating charged scFv connectors, whether positively or negatively charged (or both, if the scFv framework is used on different monomers), the monomer containing the charged connector can change the pI without making other changes in the Fc domain. These charged connectors can be replaced by various standard scFv-containing connectors. Similarly, those skilled in the art will appreciate that the charged scFv connector is used on the correct “chain” or monomer according to the desired pI. For example, as described herein, to prepare a heterodimeric antibody in the form of a triF, the original pI of the Fv region of each desired antigen-binding domain is calculated, and one of them is selected to prepare the scFv, with a positively or negatively charged connector chosen based on the pI.

[0261] Charged structural domain connectors can also be used to enhance pI separation of the monomers of the present invention; therefore, those shown in Figure 33 can be used in various embodiments of the connectors used herein.

[0262] In some embodiments, the antibody is a full-length antibody. As used herein, “full-length antibody” refers to the structure constituting the native biological form of the antibody, including variable and constant regions, including one or more modifications described herein, particularly in the Fc domain, to form heterodimers or to purify and separate heterodimers from homodimers. Full-length antibodies generally include Fab and Fc domains and may also contain additional antigen-binding domains, such as scFv, as shown in the figure.

[0263] In one embodiment, the antibody is an antibody fragment, provided that the fragment contains at least one constant domain that can be modified (e.g., pI-modified) to produce a heterodimer. Other antibody fragments that can be used include fragments containing one or more pI-modified CH1, CH2, CH3, hinge, and CL domains of the present invention. For example, an Fc fusion is a fusion of an Fc region (CH2 and CH3, which may have a hinge region) with another protein. Several Fc fusions are known in the art and can be improved by adding heterodimerizing variants of the present invention. Here, using various combinations of heterodimerizing variants described herein, antibody fusions can be prepared comprising CH1; CH1, CH2, and CH3; CH2; CH3; CH2 and CH3; CH1 and CH3, wherein any or all of them can also optionally have a hinge region.

[0264] Specifically, Figure 1 The form shown is an antibody, often called a "heterodimolecular antibody," which means that the protein has at least two associated Fc sequences that self-assemble into a heterodimeric Fc domain.

[0265] Chimeric antibodies and humanized antibodies

[0266] In some implementations, the antibody may be a mixture of different species, such as chimeric antibodies and / or humanized antibodies. Generally, "chimeric antibody" and "humanized antibody" refer to antibodies that combine regions from more than one species. For example, a "chimeric antibody" typically contains a variable region from a mouse (and sometimes a rat) and a constant region from a human. A "humanized antibody" generally refers to a non-human antibody in which the framework region of the variable domain has been replaced with a sequence from a human antibody. Generally, in a humanized antibody, the entire antibody, except for the CDR, is encoded by human polynucleotides and is identical to the antibody except for the CDR. The CDR, partially or entirely encoded by non-human biological nucleic acids, is transplanted into the β-sheet framework of the variable region of the human antibody to form the antibody; the specificity of the antibody is determined by the transplanted CDR. Examples of this type of antibody production can be found in, for instance, WO 92 / 11018, Jones, 1986, Nature 321:522-525, and Verhoeyen et al., 1988, Science 239:1534-1536, all of which are incorporated herein by reference. It is often necessary to "reverse mutate" the framework region residues of selected receptors to the corresponding donor residues, thereby restoring the affinity lost in the initial transplantation construct (US 5530101, US5585089, US 5693761, US 5693762, US 6180370, US 5859205, US 5821337, US 6054297, US 6407213, all of which are incorporated herein by reference in full). Humanized antibodies may also selectively contain at least a portion of the immunoglobulin constant region, generally human immunoglobulins, and therefore usually contain the human Fc region. Humanization can also be achieved using mice with genetically modified immune systems. (Roque et al., 2004, Biotechnol. Prog. 20:639-654, cited in full hereby.) Various well-known techniques and methods for the humanization and modification of non-human antibodies exist in this field (see Tsurushita and Vasquez, 2004, Humanization of Monoclonal Antibodies, Molecular Biology of B Cells, 533-545, Elsevier Science (USA), and all other references cited herein, all of which are included in full hereby.Humanization methods include, but are not limited to: Jones et al., 1986, Nature 321:522-525; Riechmann et al., 1988; Nature 332:323-329; Verhoeyen et al., 1988, Science, 239:1534-1536; Queen et al., 1989, Proc.Natl.Acad.Sci.USA 86:10029-33; He et al., 1998, J.Immunol.160:1029-1035; Carter et al., 1992, Proc.Natl.Acad.Sci.USA 89:4285-9; Presta et al., 1997, Cancer Res.57(20):4593-9; Gorman et al., 1991, Proc.Natl.Acad.Sci.USA The methods described in O'Connor et al., 1998, Protein Eng 11:321-8, are all incorporated herein by reference in their entirety. Humanization or other methods for reducing the immunogenicity of the variable region of nonhuman antibodies may include surface remodeling, as described in Roguska et al., 1994, Proc. Natl. Acad. Sci. USA 91:969-973, which are incorporated herein by reference in their entirety.

[0267] In some embodiments, the antibodies of the present invention comprise a heavy chain variable region from a germline heavy chain immunoglobulin gene and / or a light chain variable region from a germline light chain immunoglobulin gene. For example, such antibodies may comprise or be composed of human antibodies described below, wherein the heavy chain or light chain variable region comprised in the human antibody is a "product" of or "derived from" a germline sequence. Human antibodies belonging to a human germline immunoglobulin sequence "product" or "derived from" a human germline immunoglobulin sequence can be identified as follows: by comparing the amino acid sequence of the human antibody with the amino acid sequence of a human germline immunoglobulin, the human germline immunoglobulin sequence that is sequence-closest (i.e., has the highest percentage similarity) to the sequence of the human antibody is selected. Human antibodies belonging to a human germline immunoglobulin sequence "product" or "derived from" a human germline immunoglobulin sequence may have amino acid differences from the germline sequence, possibly due to, for example, naturally occurring somatic mutations or intentionally introduced site-directed mutations. However, the amino acid sequence of a humanized antibody is generally at least 90% identical to the amino acid sequence encoded by the human germline immunoglobulin gene and contains amino acid residues indicating that the antibody is derived from a human sequence when compared with the amino acid sequences of germline immunoglobulins from other species (such as mouse germline sequences). In some cases, the amino acid sequence of a humanized antibody may be 95%, 96%, 97%, 98%, or 99% identical to the amino acid sequence encoded by the germline immunoglobulin gene, or even at least 96%, 97%, 98%, or 99%. Typically, a humanized antibody derived from a human germline sequence differs from the amino acid sequence encoded by the human germline immunoglobulin gene by no more than 10-20 amino acids (before the introduction of the various biases, pIs, and elimination variants described herein; i.e., the number of variants is generally low before the introduction of the variants of the present invention). In some cases, the amino acid sequence encoded by the humanized antibody and the human germline immunoglobulin gene may differ by no more than 5, or even no more than 4, 3, 2 or 1 amino acid (again before the introduction of the various biased, pI and elimination variants described herein; that is, the number of variants is generally low before the introduction of the variants of the present invention).

[0268] In one embodiment, the parent antibody is affinity-ripened, as is known in the art. Humanization and affinity ripening can be performed using structure-based methods, such as those described in USSN 11 / 004,590. Selective approaches can be used to humanize and / or ripen the antibody variable region, including but not limited to: Wu et al., 1999, J.Mol.Biol.294:151-162; Baca et al., 1997, J.Biol.Chem.272(16):10678-10684; Rosok et al., 1996, J.Biol.Chem.271(37):22611-22618; Rader et al., 1998, Proc.Natl.Acad.Sci.USA 95:8910-8915; Krauss et al., 2003, Protein Engineering 16(10):753-759, all of which are incorporated herein by reference in their entirety. Other humanization methods may include porting only a portion of the CDR, including but not limited to the methods described in 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.

[0269] IV. Heterodimeric antibodies

[0270] Therefore, in some embodiments, the present invention provides a heterodimeric antibody that relies on employing two different heavy chain variant Fc sequences, which self-assemble into a heterodimeric Fc domain and a heterodimeric antibody.

[0271] This invention relates to novel constructs of heterodimeric antibodies that allow binding to more than one antigen or ligand, for example, allowing bispecific binding. The heterodimeric antibody constructs are based on the self-assembly property of the Fc domains of the two heavy chains of the antibody; for example, two “monomers” self-assemble into a “dimer.” As described below, heterodimeric antibodies are prepared by altering the amino acid sequences of the monomers. Therefore, this invention relates to the creation of heterodimeric antibodies that rely on amino acid variants within different constant regions on each chain to promote heterodimer formation and / or simplify the purification and separation of heterodimers and homodimers, enabling these heterodimeric antibodies to co-bind multiple antigens in various ways.

[0272] Therefore, this invention provides bispecific antibodies. A current problem in the field of antibody technology is the need for “bispecific” antibodies that simultaneously bind to two different antigens, thereby bringing the different antigens closer together and generating novel functions and therapies. Generally, these antibodies are prepared by introducing genes containing the heavy and light chains into host cells. This typically forms the desired heterodimer (AB), as well as two homodimers (AA and BB (excluding the issue of light chain heterodimers)). However, a major obstacle to forming bispecific antibodies is the difficulty in purifying and separating the heterodimer antibody from the homodimer antibody, and / or the difficulty in forming a biased proportion of heterodimers relative to homodimers.

[0273] There are various mechanisms that can be used to generate the heterodimers of this invention. Furthermore, those skilled in the art will recognize that these mechanisms can be combined to ensure high heterodimerization. Therefore, the amino acid variants that generate the heterodimers are referred to as "heterodimerizing variants." As described below, heterodimerizing variants may include stereovariates (e.g., the "sphere and pore" or "biased" variants and the "charge pair" variants described below) and "pI variants," which enable the purification and separation of homodimers and heterodimers. As described in WO2014 / 145806, which is incorporated herein by reference in its entirety, and specifically as described below under “Heterodimerization Variants,” the mechanisms available for heterodimerization include “sphere-and-pore” (“KIH”; sometimes referred to herein as the “biased” variant (see WO2014 / 145806), “electrostatic manipulation” or “charge pairing” as described in WO2014 / 145806, the pI variant as described in WO2014 / 145806, and other Fc variants as described in WO2014 / 145806 and below.

[0274] In this invention, several basic mechanisms facilitate the purification of heterodimeric antibodies: one relies on the use of pI variants, giving each monomer a different pI, thereby allowing isoelectric purification of AA, AB, and BB dimer proteins. Alternatively, partial structural forms, such as the "triF" form, allow size-based separation. As described later, it is also possible to form more heterodimers relative to homodimers. Therefore, combinations of stereodimerizing variants and pI or charge-pair variants are particularly useful in this invention.

[0275] Generally, particularly useful embodiments of the present invention rely on combined variants, including biased variants and pI variants, whereby the biased variants promote a superior heterodimerization above homodimerization, and the pI variants increase the pI difference between the two monomers.

[0276] Furthermore, as described later, depending on the form of the heterodimeric antibody, the pI variant can be contained either within the constant and / or Fc domain of the monomer or within a charged linker, employing either a domain linker or an scFv linker. That is, a scFv-based framework, such as the tri-F form, can include a charged (positive or negative) scFv linker, which optimizes purification through pI. Those skilled in the art will recognize that some tri-F forms are usable simply by having a charged scFv linker, without requiring further pI modifications, although the present invention does provide pI variants and / or charged domain linkers present on one or both of the monomers. Additionally, amino acid modifications to other functional groups can also result in pI variations, such as Fc, FcRn, and KO variants.

[0277] In this invention, which utilizes pI as a separation mechanism to purify heterodimeric proteins, amino acid variants can be introduced into one or both of the monomeric polypeptides; that is, the pI of one of the monomers (referred to as "monomer A") can be modified to be different from that of monomer B, or both monomers A and B can be modified, wherein the pI of monomer A is increased and the pI of monomer B is decreased. As described below, the pI modification of one or both monomers can be performed as follows: removal or addition of charged residues (e.g., substitution of neutral amino acids with positively or negatively charged amino acid residues, e.g., replacement of glycine with glutamic acid), conversion of positively charged residues to negatively charged residues or vice versa (e.g., conversion of aspartic acid to lysine), or conversion of charged residues to neutral residues (e.g., charge loss; conversion of lysine to serine). Several such variants are shown in the accompanying figures.

[0278] Therefore, this embodiment of the invention allows for sufficient pI variation in at least one of the monomers to separate the heterodimer from the homodimer. Those skilled in the art will recognize, and as described below, that this can be done by using a constant region of the "wild-type" heavy chain and a variant region whose pI has been increased or decreased (wt A-+B or wt A--B), or by increasing one region while decreasing another (A+-B- or A-B+).

[0279] Therefore, in general, one aspect of this invention is the amino acid variant in the constant region of the antibody, which aims to alter the isoelectric point (pI) of at least one (if not both) of the dimer protein monomers to form a "pI antibody," by introducing an amino acid substitution ("pI variant" or "pI substitution") into one or both monomers. As shown herein, a heterodimer can be separated from two homodimers if the pI difference between the two monomers is only 0.1 pH units; pI differences of 0.2, 0.3, 0.4, and 0.5 pH units or greater are all employed in this invention.

[0280] Those skilled in the art will recognize that the number of pI variants introduced on each or both monomers to achieve good separation depends in part on the initial pI of the components, taking the three-F form as an example, i.e., the initial pI of the target scFv and Fab. That is, to determine which monomer to modify or the modification "direction" (e.g., more positive or more negative charge), the Fc sequences of the two target antigens are calculated and the decision is made accordingly. As is known in the art, different Fvs have different initial pIs, and this invention utilizes this. Generally, as described herein, the pIs are modified so that the total pI difference between the monomers is at least about 0.1 log, preferably 0.2-0.5.

[0281] Furthermore, those skilled in the art will recognize, and as illustrated herein, that in some embodiments, heterodimers and homodimers can be separated based on size. For example... Figure 1 As shown, there are several forms that allow for the separation of heterodimers and homodimers based on size.

[0282] In cases where pI variants are used to achieve heterodimerization, a more modular approach to designing and purifying bispecific proteins (including antibodies) is provided by utilizing the constant region of the heavy chain. Therefore, in some embodiments, the heterodimerizing variant (including biased and purified heterodimerizing variants) is not located in the variable region, meaning each antibody must be modified. Additionally, in some embodiments, the pI is altered by introducing a pI variant with another IgG isotype without introducing significant immunogenicity, thereby reducing the likelihood of the pI variant causing immunogenicity. Therefore, another problem to be addressed is elucidating low pIs with high human sequence content constant domains, for example, minimizing or avoiding non-human residues at any particular position.

[0283] The potential side benefits of this pI modification include prolonged serum half-life and enhanced FcRn binding. That is, as described in USSN 13 / 194,904 (included herein by reference in its entirety), reducing the pI of constant antibody domains (including those in the antibody-Fc fusion complex) can result in longer in vivo serum retention. Such pI variants with prolonged serum half-life are also beneficial for pI modifications intended for purification purposes.

[0284] Additionally, it should be noted that pI variants of heterodimerized antibodies offer further advantages for the analysis and quality control of bispecific antibodies, namely, significant elimination, minimization, or differentiation capabilities in the presence of homodimers. Similarly, the ability to reliably test the reproducibility of heterodimer antibody production is important.

[0285] Heterodimer variants

[0286] The present invention provides heterodimeric proteins, including various forms of heterodimeric antibodies, which employ heterodimeric variants to allow heterodimer formation and / or purification and separation from homodimers.

[0287] There are several suitable combinations of paired heterodimerization-biased variants. These variants appear as "paired" combinations, where one pair incorporates the first monomer and the other incorporates the second. It should be noted that these combinations are not necessarily like the "sphere and pore" variants, where residues on one monomer correspond one-to-one with residues on the other; rather, these paired combinations form an interface between the two monomers, promoting heterodimer formation while inhibiting homodimer formation. Consequently, the percentage of spontaneously formed heterodimers under biological conditions exceeds 90%, rather than the previously assumed 50% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B).

[0288] Stereomorphs

[0289] In some implementations, the formation of heterodimers can be promoted by adding stereovariates. That is, by altering the amino acids in each heavy chain, different heavy chains are more likely to combine to form heterodimer structures rather than homodimers with the same Fc amino acid sequence. Suitable stereovariates are shown in Figure 29.

[0290] Another mechanism commonly referred to in the art as "knobs and holes" can be employed here, referring to amino acid modifications that produce stereochemical effects that favor heterodimer formation but discourage homodimer formation; this is sometimes referred to as "knobs and holes," as described in USSN 61 / 596,846, Ridgway et al., Protein Engineering 9(7):617(1996); Atwell et al., J. Mol. Biol. 1997 270:26; and US Patent No. 8,216,805, all of which are incorporated herein by reference in their entirety. The figures illustrate various "monomer A-monomer B" pairings that depend on "knobs and holes." In addition, as described in Merchant et al., Nature Biotech. 16:677(1998), these "knobs and holes" mutations can combine with disulfide bonds to selectively form heterodimers.

[0291] Another mechanism for generating heterodimers is sometimes referred to as “electrostatic manipulation,” as described by Gunasekaran et al., J. Biol. Chem. 285(25):19637(2010), which is incorporated herein by reference in its entirety. This is sometimes referred to herein as a “charge pair.” In this embodiment, electrostatics are used to favor the formation of heterodimers. Those skilled in the art will recognize that these may also affect the pI, and consequently, purification, and thus may be considered pI variants in certain cases. However, since these are manufactured to drive dimerization rather than for use as purification tools, they are classified as “stereovariants.” These variants include, but are not limited to: D221E / P228E / L368E paired with D221R / P228R / K409R (e.g., these are “monomer-corresponding combinations”) and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0292] Monomer variant A and monomer variant B may optionally and independently be used in combination with any number of other variants, such as pI variants or those specified in US 2012 / 0149876. Figure 37 Other stereovariates shown, along with the accompanying figures and descriptions, and SEQ ID NO are incorporated herein by reference.

[0293] In some embodiments, the stereovariates described herein may optionally and independently be incorporated into one or two monomers along with various pI variants (or other variants such as Fc variants, FcRn variants, etc.), and may be independently and optionally included or excluded from the proteins of the present invention.

[0294] Figure 29 shows suitable biased variants. Figure 34 This section illustrates some pairings that are particularly useful in certain embodiments. Particularly useful pairings in various embodiments include, but are not limited to: S364K / E357Q:L368D / K370S; L368D / K370S:S364K; L368E / K370S:S364K; T411T / E360E / Q362E:D401K; L368D / K370S:S364K / E357L and K370S:S364K / E357Q. In naming, the pairing "S364K / E357Q:L368D / K370S" indicates that one monomer has the dual variant combination S364K / E357Q while the other has the dual variant combination L368D / K370S.

[0295] pI (isoelectric point) variants of heterodimers

[0296] Generally, those skilled in the art will recognize that there are two types of pI variants: those that increase protein pI (basic alteration) and those that decrease protein pI (acidic alteration). As described herein, these variants can be combined arbitrarily: one monomer can be wild-type, or a variant with a pI not significantly different from wild-type, and the other can be more basic or more acidic. Alternatively, each monomer can be altered, with one becoming more basic and the other becoming more acidic.

[0297] Preferred combinations of pI variants are shown in [link to pI variants]. Figure 30 As described herein and illustrated in the figures, these changes pertain to IgG1, but all isotypes and isotype hybrids may be modified accordingly. When the heavy chain constant domain originates from IgG2 through 4, R133E and R133Q may also be used.

[0298] In one embodiment, for example in a bottle opener form, one of the monomers (negatively charged Fab side) of a preferred pI variant combination comprises the 208D / 295E / 384D / 418E / 421D variant (i.e., N208D / Q295E / N384D / Q418E / N421D relative to human IgG1), and the other monomer (positively charged scFv side) comprises a positively charged scFv linker containing (GKPGS)4. However, as those skilled in the art will know, the first monomer has a CH1 domain containing position 208. Therefore, in constructs without a CH1 domain (e.g., heterodimeric Fc fusion proteins that do not employ a CH1 domain in one of their domains, such as a dual scFv form), the preferred negatively charged pI variant Fc group comprises the 295E / 384D / 418E / 421D variant (i.e., Q295E / N384D / Q418E / N421D relative to human IgG1).

[0299] Antibody heterodimer light chain variant

[0300] If the heterodimer is antibody-based, for example, at least one of the monomers contains a light chain in addition to the heavy chain domain, pI variants can also be created in the light chain. Amino acid substitutions used to reduce the light chain pI include, but are not limited to: K126E, K126Q, K145E, K145Q, N152D, S156E, K169E, S202E, K207E, and the addition of the peptide DEDE to the C-terminus of the light chain. Such modifications based on a constant λ light chain include one or more substitutions at R108Q, Q124E, K126Q, N138D, K145T, and Q199E. Alternatively, the pI of the light chain can also be increased.

[0301] Isotype variant

[0302] Furthermore, many embodiments of the present invention are based on “introducing” a pI amino acid at a specific position of one IgG isotype into another isotype, thereby reducing or eliminating the possibility of introducing undesirable immunogenicity into the variant. See U.S. Publication No. 2014 / 0370013 for this purpose. Figure 21 The above is incorporated herein by reference. That is, IgG1 is a commonly used isotype of therapeutic antibodies for various reasons, including high-efficiency function. However, the pI of the IgG1 heavy chain constant region is higher than that of the IgG2 heavy chain constant region (8.10 vs. 7.31). By introducing specific IgG2 residues into the IgG1 backbone, the resulting monomer exhibits a lower (or higher) pI and a longer serum half-life. For example, IgG1 has glycine at position 137 (pI 5.97), while IgG2 has glutamate (pI 3.22); the introduction of 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 variant antibodies. However, it should be noted that, as described below, changes in the IgG2 molecule can also prolong the serum half-life.

[0303] In other embodiments, different types of amino acids are changed to reduce the overall charge state of the resulting protein (e.g., replacing high pI amino acids with low pI amino acids) or to make structural adjustments for stability, as described below.

[0304] Furthermore, significant changes can be observed in the monomers of the heterodimer by modifying the pI of the heavy chain constant region and the light chain constant region. As described herein, making the pI of the two monomers differ by at least 0.5 allows for separation by ion exchange chromatography, isoelectric focusing, or other methods sensitive to the isoelectric point.

[0305] Calculate pI

[0306] The pI of each monomer may depend on the pI of the heavy chain constant domain of the variant and the pI of the entire monomer, including the variable heavy chain constant domain and the fusion chaperone. Therefore, in some embodiments, the pI of U.S. Publication No. 2014 / 0370013 is used. Figure 19 The charts in the paper calculate the change in pI based on the variable heavy chain constant structural domain. As described in this paper, which monomer is modified is generally determined by the inherent pI of the Fv and the frame region. Alternatively, the pI of each monomer can be compared.

[0307] It also provides pI variants that bind better to FcRn in vivo.

[0308] In cases where pI variants reduce monomeric pI, they also additionally improve serum retention in vivo.

[0309] Although still under validation, the Fc region is thought to have a long in vivo half-life because the binding of FcRn in the endosome at pH 6 isolates Fc (Ghetie and Ward, 1997 Immunol Today. 18(12):592-598, incorporated herein by reference in full). The endosome compartment then recycles Fc to the cell surface. Once the compartment opens to the extracellular space, higher pH (~7.4) induces the release of Fc back into the bloodstream. In mice, Dall'Acqua et al. showed that Fc mutants with enhanced FcRn binding at pH 6 and pH 7.4 actually had lower serum concentrations and the same half-life as wild-type Fc (Dall'Acqua et al., 2002, J. Immunol. 169:5171-5180, incorporated herein by reference in full). The increased affinity of Fc for FcRn at pH 7.4 is thought to inhibit the release of Fc back into the bloodstream. Therefore, ideally, an Fc mutation that enhances the in vivo half-life of Fc would strengthen FcRn binding at low pH while still allowing Fc release at high pH. The charge state of the amino acid histidine changes within a pH range of 6.0–7.4. Therefore, it is unsurprising that His residues are found at key positions in the Fc / FcRn complex.

[0310] It has recently been proposed that antibodies with low isoelectric point variable regions can also have longer serum half-lives (Igawa et al., 2010 PEDS.23(5):385-392, included in this article by reference in full). However, the mechanism is still poorly understood. Furthermore, the variable regions of antibodies vary considerably. Constant regions with low pI and long half-lives could provide a more modular approach to improving the pharmacokinetic properties of antibodies, as described in this article.

[0311] Other Fc variants for other functions

[0312] Besides pI amino acid variants, a variety of other useful Fc amino acid modifications can be made for a variety of reasons, including but not limited to altering the binding to one or more FcγR receptors, altering the binding to FcRn receptors, etc.

[0313] Therefore, the proteins of the present invention may include amino acid modifications, including heterodimerization variants as described herein, which include pI variants and stereovariates. Various combinatorial variants may be included independently, optionally, or not included in various specific heterodimer proteins.

[0314] FcγR variant

[0315] Therefore, a variety of usable Fc substitutions can be manufactured to alter the binding to one or more of the FcγR receptors. Useful substitutions include those that enhance and reduce binding. For example, it is known that enhanced binding to FcγRIIIa generally enhances ADCC (antibody-dependent cell-mediated cytotoxicity; a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize antibodies bound to target cells and then cause target cell lysis). Similarly, in some cases, reduced binding to FcγRIIb (inhibitory receptor) is beneficial. Useful amino acid substitutions in this invention include USSN 11 / 124,620 (especially...) Figure 41 Those listed in 11 / 174,287, 11 / 396,495, and 11 / 538,406 are all incorporated herein by reference in their entirety, especially their variants. Specific useful variants 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.

[0316] In addition, other Fc substitutions are used to enhance binding to the FcRn receptor and prolong serum half-life, as detailed in USSN 12 / 341,769, the full text of which is incorporated herein by reference, 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.

[0317] Eliminate variants

[0318] Similarly, another class of functional variants are “FcγR elimination variants” or “Fc knockout (FcKO or KO)” variants. In these embodiments, for certain therapeutic applications, it is necessary to reduce or eliminate the normal binding of the Fc domain to one, more, or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid other mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific antibodies that monovalently bind to CD3, it is often necessary to eliminate FcγRIIIa binding to eliminate or significantly reduce ADCC activity, where one of the Fc domains contains one or more Fcγ receptor elimination variants. These elimination variants are described in […]. Figure 31Each of these can be independently and optionally included or excluded, with the following elimination variants preferred: G236R / L328R, E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K, E233P / L234V / L235A / G236del / S239K / A327G, E233P / L234V / L235A / G236del / S267K / A327G, and E233P / L234V / L235A / G236del. It should be noted that the elimination variants cited herein eliminate FcγR binding but generally do not eliminate FcRn binding.

[0319] Combination of heterodimer and Fc variant

[0320] Those skilled in the art will recognize that all the mentioned heterodimerization variants (including biased variants and / or pI variants) can be optionally and independently combined in any manner, as long as they retain their “chain-like” or “monomer-like” configuration. Furthermore, all these variants can be incorporated into any heterodimerization form.

[0321] Regarding pI variants, although several particularly useful embodiments are shown in the figures, other embodiments can be formed based on the fundamental principle of changing the pI difference between the two monomers to facilitate purification.

[0322] In addition, any one of the heterodimerization variant, bias, and pI can be independently or optionally combined with the Fc elimination variant, Fc variant, or FcRn variant, as generally described herein.

[0323] Useful antibody form of the present invention

[0324] As will be apparent to those skilled in the art, and as described below, the heterodimeric fusion protein of the present invention can take on various conformations, generally as follows: Figure 1 As shown. Some figures depict a "single-terminal" configuration, where one "arm" of the molecule has one specificity and the other "arm" has a different specificity. Other figures depict a "double-terminal" configuration, where the "top" of the molecule has at least one specificity and the "bottom" of the molecule has one or more other specificities. Therefore, the present invention relates to novel immunoglobulin compositions that co-bind different first and second antigens.

[0325] Those skilled in the art will recognize that the heterodimeric form of the present invention can have different valence states in addition to being bispecific. That is, the heterodimeric antibody of the present invention can be bivalently bispecific, where one target tumor antigen (e.g., CD3) is bound by one binding domain and another target tumor antigen (e.g., CD20, CD38, CD123, etc.) is bound by a second binding domain. The heterodimeric antibody can also be trivalently bispecific, where the first antigen is bound by both binding domains and the second antigen is bound by the second binding domain. As described herein, when CD3 is one of the target antigens, it is preferable that CD3 is bound only monovalently to reduce potential side effects.

[0326] This invention combines an anti-CD3 antigen-binding domain with an anti-target tumor antigen (TTA) antigen-binding domain. Those skilled in the art will recognize that the various figures (see attached figures for details) can be used to illustrate this. Figure 2-7 The anti-CD3 CDR, anti-CD3 variable light chain domain and variable heavy chain domain, Fab and scFv shown in Figure 68) can be used. Similarly, various anti-TTA antigen-binding domains, such as anti-CD38, anti-CD20, anti-CD19 and anti-CD123 antigen-binding domains, can be used, and the CDR, variable light chain domain and variable heavy chain domain, Fab and scFv shown in the respective figures can be used, optionally and independently in any combination.

[0327] Bottle opener type

[0328] One of the particularly useful heterodimer frameworks in this invention is Figure 1 The “three-F” or “bottle opener” architecture shown in A, A, and B. In this embodiment, one heavy chain of the antibody contains a single-chain Fv (“scFv” as described herein) while the other heavy chain is a “conventional” Fab form containing variable heavy and light chains. This structure is sometimes referred to herein as a “three-F” type (scFv-FAb-Fc) or “bottle opener” type because it looks quite similar to a bottle opener (see [link to documentation]). Figure 1 The two chains are joined together by amino acid variants in constant regions (e.g., Fc domain, CH1 domain, and / or hinge region), which promote the formation of heterodimeric antibodies, as described below.

[0329] The “three-F” configuration offers numerous advantages. As is known in the art, antibody analogs relying on two scFv constructs typically suffer from stability and aggregation issues, which can be mitigated in this invention by adding a “conventional” heavy-light chain pairing. Furthermore, unlike configurations relying on two heavy chains and two light chains, there is no issue of heavy-light chain mismatch (e.g., heavy chain 1 paired with light chain 2, etc.).

[0330] Many of the embodiments described herein are based on a bottle opener form, in which the first monomer comprises scFv, which includes variable heavy chain domains and variable light chain domains, covalently bonded by scFv joints (charged in many cases, but not all), wherein the scFv is typically covalently bonded to the N-terminus of the first Fc domain via domain joints (charged or uncharged, as described herein). The second monomer of the bottle opener form is a heavy chain, and the composition also includes a light chain.

[0331] Generally, in many preferred embodiments, scFv is a Fab-bound TTA that combines CD3 domains with both heavy and light chains.

[0332] In addition, the Fc domain of the present invention generally includes biased variants (e.g., as shown in Figure 29 and...). Figure 34 The combined amino acid substitutions shown, particularly useful preferred variants are selected from: S364K / E357Q:L368D / K370S; L368D / K370S:S364K; L368E / K370S:S364K; T411T / E360E / Q362E:D401K; L368D / K370S:S364K / E357L and K370S:S364K / E357Q), and optional elimination variants (including...). Figure 31 Those shown), optional live scFv connectors (including those shown in Figure 33), heavy chains containing pI variants (including Figure 30 (Those shown).

[0333] In some implementations, any vh and vl sequences herein (including all vh and vl sequences in the figures, including those for CD20, CD38, and CD123) can be incorporated into the opener backbone pattern shown in Figure 162 as the "Fab side," employing any anti-CD3 scFv sequence shown in each figure. Specifically, the anti-CD3 sequences used in these implementations are: anti-CD3H1.30_L1.47, anti-CD3 H1.32_L1.47, anti-CD3 H1.89_L1.47, anti-CD3 H1.90_L1.47, anti-CD3H1.33_L1.47, and anti-CD3H1.31_L1.47, as the scFv side of the backbone pattern in Figure 162.

[0334] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 The bottle opener form shown in Figure 68 includes any combination with the various trunk patterns in Figure 162. Furthermore, Figures 2 to 7 The anti-CD3 vh and vl sequences shown in Figure 68 can be used as Fab sides.

[0335] This invention provides a bottle opener form having a CD38 antigen-binding domain, wherein the anti-CD38 sequence is shown in the figure, including... Figure 8-10 As described above, each VH and VL anti-CD38 sequence can be either Fab-side or scFv-side and can be linked to form one of the antigen-binding domains in the opener pattern, including those in Figure 162. When the anti-CD38 sequence is Fab-side, any anti-CD3 scFv sequence in the figure can be used, specifically including 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, to become the scFv side of the backbone shown in Figure 162.

[0336] This invention provides a bottle opener pattern with a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure. As described above, each vh and vl anti-CD20 sequence can be either Fab-side or scFv-side, and can be connected to form one of the antigen-binding domains of the bottle opener pattern, including those bottle opener patterns in Figure 162. When the anti-CD20 sequence is Fab-side, any anti-CD3 scFv sequence in the figure can be used, specifically including 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, becoming the scFv side of the backbone shown in Figure 162.

[0337] This invention provides a bottle opener pattern with a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure. As described above, each vh and vl anti-CD123 sequence can be either Fab-side or scFv-side, and can be connected to form one of the antigen-binding domains of the bottle opener pattern, including those bottle opener patterns in Figure 162. When the anti-CD123 sequence is Fab-side, any anti-CD3 scFv sequence in the figure can be used, specifically including 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, becoming the scFv side of the backbone shown in Figure 162.

[0338] mAb-Fv type

[0339] One of the particularly useful heterodimer frameworks in this invention is Figure 1The mAb-Fv form is shown. In this embodiment, the form attaches an "additional" variable heavy chain domain to the C-terminus of one of the monomers and an "additional" variable light chain domain to the C-terminus of the other monomer, thereby forming a third antigen-binding domain, wherein the Fab portions of both monomers bind TTA and the "additional" scFv domain binds CD3.

[0340] In this embodiment, the first monomer includes a first heavy chain comprising a first variable heavy chain structural domain and a first constant heavy chain structural domain containing a first Fc structural domain. The first variable light chain structural domain is covalently joined to the C-end of the first Fc structural domain via a structural domain joint (vh1-CH1-hinge-CH2-CH3-[optional joint]-vl2). The second monomer includes a second variable heavy chain structural domain and a second constant heavy chain structural domain containing a second Fc structural domain, and further includes a third variable heavy chain structural domain (vj1-CH1-hinge-CH2-CH3-[optional joint]-vh2) covalently joined to the C-end of the second Fc structural domain via a structural domain joint. The two C-end joined variable structural domains form a CD3-bonded scFv (since forming a divalent CD3 bond is not recommended). This embodiment also employs a common light chain comprising a variable light chain structural domain and a constant light chain structural domain, which associates with the heavy chain to form two identical Fabs bonded to the TTA. As with many embodiments of the present invention, these constructs may include biased variants, pI variants, elimination variants, other Fc variants, etc., as required herein.

[0341] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the mAb-Fv type shown in Figure 68.

[0342] This invention provides an anti-CD38 sequence, such as Figure 8-10 The mAb-Fv form shown.

[0343] The present invention provides a mAb-Fv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0344] The present invention provides a mAb-Fv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0345] The present invention provides a mAb-Fv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure.

[0346] This invention provides a method comprising Figure 31 The mAb-Fv form of the elimination variant is shown.

[0347] The present invention provides mAb-Fv forms that include the biased variants shown in Figures 29 and 34.

[0348] mAb-scFv

[0349] One of the particularly useful heterodimer frameworks in this invention is Figure 1 The mAb-scFv form is shown. In this embodiment, such a form is based on the C-terminal binding of scFv to one of the monomers, thereby forming a third antigen-binding domain, wherein the Fab portions of both monomers bind TTA while the “additional” scFv domain binds CD3. Thus, the first monomer contains a first heavy chain (containing a variable heavy chain domain and a constant domain), and the C-terminal covalently bound scFv contains a scFv variable light chain domain, a scFv connector, and a scFv variable heavy chain domain, in either order (vh1-CH1-hinge-CH2-CH3-[optional connector]-vh2-scFv connector-vl2, or vh1-CH1-hinge-CH2-CH3-[optional connector]-vl2-scFv connector-vh2). This embodiment also employs a common light chain containing a variable light chain domain and a constant light chain domain, which associates with the heavy chain to form two identical Fabs that bind TTA. As with many embodiments of the present invention, these constructs may include biased variants, pI variants, elimination variants, other Fc variants, etc., as required herein.

[0350] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the mAb-Fv form shown in Figure 68.

[0351] This invention provides an anti-CD38 sequence, such as Figure 8-10 The mAb-Fv form shown.

[0352] The present invention provides a mAb-Fv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0353] The present invention provides a mAb-Fv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0354] The present invention provides a mAb-Fv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure.

[0355] This invention provides a method comprising Figure 31 The mAb-Fv form of the elimination variant is shown.

[0356] The present invention provides mAb-Fv forms that include the biased variants shown in Figures 29 and 34.

[0357] Center scFv

[0358] One of the particularly useful heterodimer frameworks in this invention is Figure 1 The central-scFv form is shown. This form employs an inserted scFv domain to form a third antigen-binding domain, where the Fab portions of both monomers bind TTA while the "additional" scFv domain binds CD3. The scFv domain is inserted between the Fc domain and the CH1-Fv region of one of the monomers, thereby providing the third antigen-binding domain.

[0359] In this embodiment, one of the monomers includes a first heavy chain comprising a first variable heavy chain structural domain, a CH1 structural domain (and an optional hinge), and an Fc structural domain, and also includes an scFv comprising a scFv variable light chain structural domain, an scFv connector, and a scFv variable heavy chain structural domain. The scFv is covalently joined between the C-end of the CH1 structural domain of the heavy chain constant structural domain and the N-end of the first Fc structural domain by an optional structural domain connector (vh1-CH1-[optional connector]-vh2-scFv connector-vl2-[optional connector including hinge]-CH2-CH3, or the reverse for the scFv, vh1-CH1-[optional connector]-vl2-scFv connector-vh2-[optional connector including hinge]-CH2-CH3). The other monomer is the standard Fab side. This embodiment also employs a common light chain comprising a variable light chain structural domain and a constant light chain structural domain, associated with the heavy chain to form two identical Fabs with a combined TTA. As with many embodiments of the present invention, these constructs may include biased variants, pI variants, elimination variants, other Fc variants, etc., as required herein.

[0360] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the center-scFv form shown in Figure 68.

[0361] This invention provides an anti-CD38 sequence, such as Figure 8-10 The center-scFv form shown.

[0362] The present invention provides a central-scFv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0363] The present invention provides a central-scFv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0364] The present invention provides a central-scFv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in v.

[0365] This invention provides a method comprising Figure 31 The center-scFv form of the elimination variant is shown.

[0366] The present invention provides a center-scFv form that includes the biased variants shown in Figures 29 and 34.

[0367] Center-Fv type

[0368] One of the particularly useful heterodimer frameworks in this invention is Figure 1 The central-Fv form is shown. This form employs an inserted scFv domain, thereby forming a third antigen-binding domain, in which the Fab portions of the two monomers bind TTA while the "additional" scFv domain binds CD3. The scFv domain is inserted between the CH1-Fv region and the Fc domain of the monomer, thereby providing the third antigen-binding domain, wherein each monomer contains one of the components of scFv (e.g., one monomer contains a variable heavy chain domain and the other contains a variable light chain domain).

[0369] In this embodiment, one monomer includes a first heavy chain comprising a first variable heavy chain structural domain, a CH1 structural domain, an Fc structural domain, and an additional variable light chain structural domain. The additional light chain structural domain is covalently joined between the C-end of the heavy chain constant structural domain CH1 and the N-end of the first Fc structural domain by a structural domain joint (vh1-CH1-[optional joint]-vl2-hinge-CH2-CH3). The other monomer includes a first heavy chain comprising a first variable heavy chain structural domain, a CH1 structural domain, an Fc structural domain, and an additional variable heavy chain structural domain (vh1-CH1-[optional joint]-vh2-hinge-CH2-CH3). The additional light chain structural domain is covalently joined between the C-end of the heavy chain constant structural domain CH1 and the N-end of the first Fc structural domain by a structural domain joint.

[0370] This embodiment also employs a common light chain containing both a variable light chain structure domain and a constant light chain structure domain, which associates with the heavy chain to form two identical Fabs that bind the TTA. As with many embodiments of the invention, these constructs may include biased variants, pI variants, elimination variants, other Fc variants, etc., as desired, as described herein.

[0371] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the center-Fv form shown in Figure 68.

[0372] This invention provides an anti-CD38 sequence, such as Figure 8-10 The center-Fv form shown.

[0373] The present invention provides a central-Fv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0374] The present invention provides a central-Fv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0375] The present invention provides a central-Fv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure.

[0376] This invention provides a method comprising Figure 31 The center-Fv form of the elimination variant is shown.

[0377] The present invention provides a center-Fv form comprising the biased variants shown in Figures 29 and 34.

[0378] Single-arm center - scFv

[0379] One of the particularly useful heterodimer frameworks in this invention is Figure 1 The single-arm center-scFv configuration is shown. In this embodiment, one monomer contains only the Fc domain, while the other monomer incorporates an inserted scFv domain to form a second antigen-binding domain. In this configuration, the Fab portion binds to TTA while the scFv binds to CD3, or vice versa. The scFv domain is inserted between the CH1-Fv region and the Fc domain of one of the monomers.

[0380] In this embodiment, one of the monomers includes a first heavy chain comprising a first variable heavy chain domain, a CH1 domain, and an Fc domain, and also includes an scFv comprising an scFv variable light chain domain, an scFv connector, and an scFv variable heavy chain domain. The scFv is covalently bonded between the C-terminus of the heavy chain constant CH1 domain and the N-terminus of the first Fc domain by a domain connector. The second monomer includes an Fc domain. This embodiment also employs a light chain comprising a variable light chain domain and a constant light chain domain, which associates with the heavy chain to form a Fab. As with many embodiments of the invention, these constructs may include biased variants, pI variants, elimination variants, other Fc variants, etc., as desired, as described herein.

[0381] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the single-arm center-scFv form shown in Figure 68.

[0382] This invention provides an anti-CD38 sequence, such as Figure 8-10 The single-arm center-scFv form shown.

[0383] The present invention provides a single-arm center-scFv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0384] The present invention provides a single-arm center-scFv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0385] The present invention provides a single-arm center-scFv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure.

[0386] This invention provides a method comprising Figure 31 The eliminated variant is shown as the single-arm center-scFv form.

[0387] The present invention provides a single-arm center-scFv form comprising the bias variants shown in Figures 29 and 34.

[0388] Double SCFV type

[0389] This invention also provides, as is known in the art. Figure 1 The double scFv form shown.

[0390] This invention provides an anti-CD3 scFv sequence, such as Figures 2 to 7 And the double scFv form shown in Figure 68.

[0391] This invention provides an anti-CD38 sequence, such as Figure 8-10 The double scFv form shown.

[0392] The present invention provides a dual scFv form having a CD20 antigen-binding domain, wherein the anti-CD20 sequence is shown in the figure.

[0393] The present invention provides a dual scFv form having a CD19 antigen-binding domain, wherein the anti-CD19 sequence is shown in the figure.

[0394] The present invention provides a dual scFv form having a CD123 antigen-binding domain, wherein the anti-CD123 sequence is shown in the figure.

[0395] This invention provides a method comprising Figure 31 The double scFv form of the elimination variant is shown.

[0396] The present invention provides a dual scFv form that includes the biased variants shown in Figures 29 and 34.

[0397] The present invention provides a dual scFv form containing pI variants and / or charged scFv connectors (typically, one of the monomers contains Q295E / N384D / Q418E / N481D while the other monomer has a positively charged scFv connector, or both monomers contain scFv connectors with opposite charges).

[0398] target antigen

[0399] The bispecific antibody of the present invention has two distinct antigen-binding domains: one that binds to CD3 (usually monovalently) and the other that binds to a target tumor antigen (sometimes referred to herein as "TTA"). Suitable target tumor antigens include, but are not limited to: CD20, CD38, CD123; ROR1, ROR2, BCMA; PSMA; SSTR2; SSTR5, CD19, FLT3, CD33, PSCA, ADAM17, CEA, Her2, EGFR, EGFR-vIII, CD30, FORR1, GD-2, CA-IX, Trop-2, CD70, CD38, mesothelin, EphA2, CD22, CD79b, GPNMB, CD56, CD138, CD52, CD74, CD30, CD123, RON, ERBB2, and EGFR.

[0400] The “triple-F” type is particularly well-suited for targeting two (or more) different antigens. (As described herein, this targeting can be any combination of monovalent and bivalent binding, depending on the antibody form.) Therefore, immunoglobulins are preferably co-bound to two target antigens in this study. The specificities of each monomer can be selected from the list provided herein. Other useful bispecific antibody forms employing the anti-CD3 binding domain are listed below. Figure 1 .

[0401] The heterodimeric antibodies described herein are particularly suitable for the monovalent binding of beneficial or essential co-target pairs of target antigens. These antigens can be, for example, immune receptors activated after immune complexation. Cellular activation of many immune receptors occurs solely through cross-linking, typically achieved by antibody / antigen immune complexes or through binding to effector cell targets. For some immune receptors, such as the CD3 signaling receptor on T cells, activation after co-target binding is crucial because nonspecific cross-linking in the clinical setting can trigger cytokine storms and toxicity. Therapeutoriously, by using the immunoglobulins described herein to monovalently bind such antigens, this activation occurs only in the microenvironment of the primary target antigen, where cross-linking should occur. The ability to target two different antigens in different valence states is a novel and useful aspect of this invention. Examples of therapeutically preferred or necessary monovalent co-binding target antigens include, but are not limited to, immune-activating receptors such as CD3, FcγR, toll-like receptors (TLRs) such as TLR4 and TLR9, cytokines, chemokines, cytokine receptors, and chemokine receptors. In many implementations, one of the antigen-binding sites binds to CD3, and in some implementations, this is a monomer containing scFv.

[0402] Theoretically, the immunoglobulins described in this article can target any antigen, including but not limited to proteins, subunits, domains, motifs, and / or epitopes belonging to the following target antigen list, including soluble factors such as cytokines and membrane-bound factors, including transmembrane receptors: 17-IA, 4-1BB, 4Dc, 6-keto-PGF1a, 8-iso-PGF2a, 8-oxo-dG, A1 adenosine receptor, A33, ACE, ACE-2, activin, activin A, activin AB, activin B, activin C, activin RIA, activin RIA ALK-2, and activin RIB. ALK-4, activator RIIA, activator RIIB, ADAM, ADAM10, ADAM12, ADAM15, ADAM17 / TACE, ADAM8, ADAM9, ADAMTS, ADAMTS4, ADAMTS5, addressins, aFGF, ALCAM, ALK, ALK-1, ALK-7, α-1-antitrypsin, α-V / β-1 antagonist, ANG, Ang, APAF-1, APE, APJ, APP, APRIL, AR, ARC ART, Artesunate, Anti-Id, Aspartic, Atrial Natriuretic Factor, av / b3 Integrin, Axl, b2M, B7-1, B7-2, B7-H, B-Lymphocyte Stimulator (BlyS), BACE, BACE-1, Bad, BAFF, BAFF-R, Bag-1, BAK, Bax, BCA-1, BCAM, Bcl, BCMA, BDNF, β-ECGF, bFGF, BID, Bik, BIM, BLC, BL-CAM, BLK, BMP, BMP-2, BMP-2a BMP-3 osteoblast, BMP-4BMP-2b, BMP-5, BMP-6Vgr-1, BMP-7(OP-1), BMP-8(BMP-8a, OP-2), BMPR, BMPR-IA(ALK-3), BMPR-IB(ALK-6), BRK-2, RPK-1, BMPR-II(BRK-3), BMP, β-NGF, BOK, bufotalin, bone-derived neurotrophic factor, BPDE, BPDE-DNA, BTC, complement factor 3 (C3), C3a, C4, C5, C 5a, C10, CA125, CAD-8, calcitonin, cAMP, carcinoembryonic antigen (CEA), cancer-associated antigen, cathepsin A, cathepsin B, cathepsin C / DPPI, cathepsin D, cathepsin E, cathepsin H, cathepsin L, cathepsin O, cathepsin S, cathepsin V, cathepsin X / Z / P, CBL, CCI, CCK2, CCL, CCL1, CCL11, CCL12, CCL13, CCL14, CCL15, CCL16, CCL17, CCL18CCL19, CCL2, CCL20, CCL21, CCL22, CCL23, CCL24, CCL25, CCL26, CCL27, CCL28, CCL3, CCL4, CCL5, CCL6, CCL7, CCL8, CCL9 / 10, CCR, CCR1, CCR10, CCR10, CCR2, CCR3, CCR4, CCR5, CCR6, CCR7, CCR8, CCR9, CD1, CD2, CD3, CD3E, CD4, CD5, CD6, CD7, CD8, CD10, CD11a, CD11b, CD11c, CD13, CD14, CD15, CD16, CD18, CD19, CD20, CD21, CD22, CD23, CD25, CD27L, CD28, CD29, CD30, CD30L, CD32, CD33 (p67 protein), CD34, CD38, CD40, CD40L, CD44, CD45, CD46, CD49a, CD52, CD54, CD55, CD56, CD61, CD64, CD66e, CD74, CD80 (B7-1), CD89, CD95, CD123, CD137, CD138, CD140a, CD146, CD147, CD148, CD152, CD164, CEACAM5, CFTR, cGMP, CINC, Clostridium botulinum toxin, Clostridium perfringens toxin, CKb8-1, CLC, CMV, CMV UL, CNTF, CNTN-1, COX, C-Ret, CRG-2, CT-1, CTACK, CTGF, CTLA-4, CX3CL1, CX3CR1, CXCL, CXCL1, CXCL2, CXCL3, CXCL4, CXCL5, CXCL6, CXCL7, CXCL8, CXCL9, CXCL10, CXCL11, CXCL12, CXCL13, CXCL14, CXCL15, CXCL16, CXCR, CXCR1, CXCR2, CXCR3, CXCR4, CXCR5, CXCR6, cytokeratin tumor-associated antigen, DAN, DCC, DcR3, DC-SIGN, decay accelerator, des(1-3)-IGF-I (brain IGF-1), Dhh, digoxigenin, DNAM-1, DNase, Dpp, DPPIV / CD26, Dtk, ECAD, EDA, EDA-A1, EDA-A2, EDAR, EGF, EGFR (ErbB-1), EMA, EMMPRIN, ENA, endothelin receptor, enkephalinase, eNOS, Eot, eosinophil chemokine 1, EpCAM, liver glycoside B2 / EphB4, EPO, ERCC, E-selectin, ET-1, factor IIaFactor VII, Factor VIIIc, Factor IX, Fibroblast Activating Protein (FAP), Fas, FcR1, FEN-1, Ferritin, FGF, FGF-19, FGF-2, FGF3, FGF-8, FGFR, FGFR-3, Fibrin, FL, FLIP, Flt-3, Flt-4, Follicle-Stimulating Hormone, Fractal Chemokines, FZD1, FZD2, FZD3, FZD4, FZD5, FZD6, FZD7, FZD8, FZD9, FZD10, G250, Gas 6. GCP-2, GCSF, GD2, GD3, GDF, GDF-1, GDF-3 (Vgr-2), GDF-5 (BMP-14, CDMP-1), GDF-6 (BMP-13, CDMP-2), GDF-7 (BMP-12, CDMP-3), GDF-8 (myosin), GDF-9, GDF-15 (MIC-1), GDNF, GDNF, GFAP, GFRa-1, GFR-α1, GFR-α2, GFR-α3, GITR, glucagon, Glut 4. Glycoprotein IIb / IIIa (GP IIb / IIIa), GM-CSF, gp130, gp72, GRO, growth hormone releasing factor, hapten (NP-cap or NIP-cap), HB-EGF, HCC, HCMV gB envelope glycoprotein, HCMV gH envelope glycoprotein, HCMV UL, Hematopoietic Growth Factor (HGF), Hep B gp120, Heparinase, Her2, Her2 / neu (ErbB-2), Her3 (ErbB-3), Her4 (ErbB-4), Herpes Simplex Virus (HSV) gB glycoprotein, HSV gD glycoprotein, HGFA, High Molecular Weight Melanoma-Associated Antigen (HMW-MAA), HIV gp120, HIV IIIB gp120V3 loop, HLA, HLA-DR, HM1.24, HMFG PEM, HRG, Hrk, human cardiac myosin, human cytomegalovirus (HCMV), human growth hormone (HGH), HVEM, I-309, IAP, ICAM, ICAM-1, ICAM-3, ICE, ICOS, IFNg, Ig, IgA receptor, IgE, IGF, IGF-binding protein, IGF-1R, IGFBP, IGF-I, IGF-II, IL, IL-1, IL-1R, IL-2, IL-2R, IL-4, IL-4R, IL-5, IL-5R, IL-6, IL-6R, IL-8, IL-9, IL-10, IL-12, IL-13, IL-15, IL-18, IL-18R, IL-23, interferon (INF)-α, INF-β, INF-γ, inhibin, iNOS, insulin A-chain, insulin B-chainInsulin-like growth factor 1, integrin α2, integrin α3, integrin α4, integrin α4 / β1, integrin α4 / β7, integrin α5 (αV), integrin α5 / β1, integrin α5 / β3, integrin α6, integrin β1, integrin β2, interferon γ, IP-10, I-TAC, JE, kallikrein 2, kallikrein 5, kallikrein 6, kallikrein 11, kallikrein 12, kallikrein 14, kallikrein 15, kinin Kallikrein L1, kallikrein L2, kallikrein L3, kallikrein L4, KC, KDR, keratinocyte growth factor (KGF), laminin 5, LAMP, LAP, LAP (TGF-1), latent TGF-1, latent TGF-βb1, LBP, LDGF, LECT2, Lefty, Lewis-Y antigen, Lewis-Y related antigen, LFA-1, LFA-3, Lfo, LIF, LIGHT, lipoprotein, LIX, LK N, Lptn, L-selectin, LT-a, LT-b, LTB4, LTBP-1, lung surfactant, luteinizing hormone, lymphotoxin B receptor, Mac-1, MAdCAM, MAG, MAP2, MARC, MCAM, MCAM, MCK-2, MCP, M-CSF, MDC, Mer, METALLOPROTEASES, MGDF receptor, MGMT, MHC (HLA-DR), MIF, MIG, MIP, MIP-1-α, M K, MMAC1, MMP, MMP-1, MMP-10, MMP-11, MMP-12, MMP-13, MMP-14, MMP-15, MMP-2, MMP-24, MMP-3, MMP-7, MMP-8, MMP-9, MPIF, Mpo, MSK, MSP, Mucin (Muc1), MUC18, Muellerian natriuretic substance, Mug, Musk, NAIP, NAP, NCAD, N-cadherin, NCA 90, NCAM, NCAM, Enkephalin, Neurotrophin-3, -4, or -6, Neuroturin, Nerve Growth Factor (NGF), NGFR, NGF-β, nNOS, NO, NOS, Npn, NRG-3, NT, NTN, OB, OGG1, OPG, OPN, OSM, OX40L, OX40R, p150, p95, PADPr, Parathyroid Hormone, PARC, PARP, PBR, PBSF, PCAD, P-cadherin, PCNA, PDGF, PDGF, PDK-1, PECAM, PEM, PF4, PGE, PGF, PGI2, PGJ2, PIN, PLA2, Placental Alkaline Phosphatase (PLAP), PlGF, PLP, PP14, ProinsulinRelaxin proton, protein C, PS, PSA, PSCA, prostate-specific membrane antigen (PSMA), PTEN, PTHrp, Ptk, PTN, R51, RANK, RANKL, RANTES, RANTES, relaxin A-chain, relaxin B-chain, renin, respiratory syncytial virus (RSV) F, RSV Fgp, Ret, Rheumatoid Factor, RLIP76, RPA2, RSK, S100, SCF / KL, SDF-1, SERINE, Serum Albumin, sFRP-3, Shh, SIGIRR, SK-1, SLAM, SLPI, SMAC, SMDF, SMOH, SOD, SPARC, Stat, STEAP, STEAP-II, TACE, TACI, TAG-72 (Tumor-associated Glycoprotein-72), TARC, TCA-3, T-cell receptors (e.g., T-cell receptor α / β), TdT, TECK, TEM1, TEM5, TEM7, TEM8, TERT, Testis PLAP-like alkaline phosphatase, TfR, TGF, TGF-α, TGF-β, TGF-β panspecific, TGF-βRI (ALK-5), TGF-βRII, TGF-βRIIb, TGF-βRIII, TGF-β1, TGF-β2, TGF-β3, TGF-β4, TGF-β5, thrombin, thymic Ck-1, thyroid-stimulating hormone, Tie, TIMP, TIQ, tissue factor, TMEFF2, Tmpo, TMPRSS2, TNF, TNF-α, TNF-αβ, TNF-β2, TNFc, TNF-RI, TNF-RII, TNFRSF10A (TRAIL) R1 Apo-2, DR4), TNFRSF10B (TRAIL R2 DR5, KILLER, TRICK-2A, TRICK-B), TNFRSF10C (TRAIL R3 DcR1, LIT, TRID), TNFRSF10D (TRAIL R4 DcR2, TRUNDD), TNFRSF11A (RANK ODF R, TRANCE R), TNFRSF11B (OPG OCIF, TR1), TNFRSF12 (TWEAK R FN14), TNFRSF13B (TACI), TNFRSF13C (BAFFR), TNFRSF14 (HVEM ATAR, HveA, LIGHT R, TR2), TNFRSF16 (NGFR p75NTR), TNFRSF17 (BCMA), TNFRSF18 (GITR AITR), TNFRSF19 (TROY TAJ, TRADE), TNFRSF19L (RELT), TNFRSF1A (TNF RICD120a, p55-60)TNFRSF1B (TNF RII CD120b, p75-80), TNFRSF26 (TNFRH3), TNFRSF3 (LTbRTNF RIII, TNFC R), TNFRSF4 (OX40 ACT35, TXGP1 R), TNFRSF5 (CD40 p50), TNFRSF6 (FasApo-1, APT1, CD95), TNFRSF6B (DcR3M68, TR6), TNFRSF7 (CD27), TNFRSF8 (CD30), TNFRSF9 (4-1BB CD137, ILA), TNFRSF21 (DR6), TNFRSF22 (DcTRAIL R2 TNFRH2), TNFRST23 (DcTRAIL R1TNFRH1), TNFRSF25 (DR3 Apo-3, LARD, TR-3, TRAMP, WSL-1), TNFSF10 (TRAIL Apo-2 ligand, TL2), TNFSF11 (TRANCE / RANK ligand ODF, OPG ligand), TNFSF12 (TWEAK Apo-3 ligand, DR3 ligand), TNFSF13 (APRIL TALL2), TNFSF13B (BAFF BLYS, TALL1, THANK, TNFSF20), TNFSF14 (LIGHTHVEM ligand, LTg), TNFSF15 (TL1A / VEGI), TNFSF18 (GITR ligand AITR ligand, TL6), TNFSF1A (TNF-α myonein, DIF, TNFSF2), TNFSF1B (TNF-β LTa, TNFSF1), TNFSF3 (LTb) TNFC, p33), TNFSF4 (OX40 ligand gp34, TXGP1), TNFSF5 (CD40 ligand CD154, gp39, HIGM1, IMD3, TRAP), TNFSF6 (Fas ligand Apo-1 ligand, APT1 ligand), TNFSF7 (CD27 ligand CD70), TNFSF8 (CD30 ligand CD153), TNFSF9 (4-1BB ligand CD137 ligand), TP-1, t-PA, Tpo, TRAIL, TRAIL R, TRAIL-R1, TRAIL-R2, TRANCE, transfer receptor, TRF, Trk, TROP-2, TSG, TSLP, tumor-associated antigen CA 125, expression Lewis Tumor-associated antigens of Y-related sugars, TWEAK, TXB2, Ung, uPAR, uPAR-1, urokinase, VCAM, VCAM-1, VECAD, VE-cadherin, VE-cadherin-2, VEFGR-1 (flt-1), VEGF, VEGFR,VEGFR-3 (flt-4), VEGI, VIM, viral antigens, VLA, VLA-1, VLA-4, VNR integrin, von Willebrand factor, WIF-1, WNT1, WNT2, WNT2B / 13, WNT3, WNT3A, WNT4, WNT5A, WNT5B, WNT6, WNT7A, WNT7B, WNT8A, WNT8B, WNT9A, WNT9A, WNT9B, WNT10A, WNT10B, WNT11, WNT16, XCL1, XCL2, XCR1, XCR1, XEDAR, XIAP, XPD, and receptors for hormones and growth factors.

[0403] Exemplary antigens that can be specifically targeted by the immunoglobulins of this invention include, but are not limited to: CD20, CD19, Her2, EGFR, EpCAM, CD3, FcγRIIIa (CD16), FcγRIIa (CD32a), FcγRIIb (CD32b), FcγRI (CD64), Toll-like receptors (TLRs) such as TLR4 and TLR9, cytokines such as IL-2, IL-5, IL-13, IL-12, IL-23, and TNFα, cytokine receptors such as IL-2R, chemokines, chemokine receptors, and growth factors such as VEGF and HGF. To form the bispecific antibodies of this invention, antibodies against any combination of these antigens can be manufactured; that is, each of these antigens may optionally and independently be included or excluded by the bispecific antibodies of this invention.

[0404] A particularly preferred combination of bispecific antibodies is an antigen-binding domain targeting CD3 and an antigen-binding domain selected from domains binding CD19, CD20, CD38, and CD123, as shown in the figure.

[0405] The nucleic acid of the present invention

[0406] This invention also provides nucleic acid compositions encoding the bispecific antibodies of this invention. Those skilled in the art will recognize that the nucleic acid composition will depend on the form and framework of the heterodimeric protein. Therefore, for example, when the antibody form requires three amino acid sequences, such as the three-F form (e.g., the first amino acid monomer contains an Fc domain and scFv, and the second amino acid monomer contains a heavy chain and a light chain), the three nucleic acid sequences can be incorporated into one or more expression vectors for expression. Similarly, some forms (e.g.) Figure 1 The dual scFv form (in which only two nucleic acids are required) requires only two nucleic acids; similarly, they can be placed in one or two expression vectors.

[0407] As is known in the art, nucleic acids encoding the components of the present invention can be incorporated into an expression vector according to host cells used for producing the heterodimeric antibodies of the present invention, as is known in the art. The nucleic acids are generally operatively linked to any number of regulatory elements (promoters, origins of replication, selectivity markers, ribosome binding sites, inducers, etc.). The expression vector can be an extrachromosomal vector or an integration vector.

[0408] The nucleic acids and / or expression vectors of the present invention can then be transformed into any number of different types of host cells, including mammalian, bacterial, yeast, insect and / or fungal cells, as is well known in the art. Mammalian cells (e.g., CHO cells) are particularly useful in many embodiments.

[0409] In some embodiments, the individual nucleic acids encoding each monomer and optionally the nucleic acid encoding the light chain (depending on the form of the antibody) are contained in a single expression vector, typically under the control of different or the same promoter. In embodiments particularly useful in this invention, the two or three nucleic acids are contained in separate expression vectors. As described herein and as in 62 / 025,931 (incorporated herein by reference), different vector ratios can be used to drive heterodimer formation. That is, surprisingly, although the protein contains a 1:1:2 ratio of first monomer: second monomer: light chain (the heterodimeric antibody in many embodiments herein contains three polypeptides), these are not optimal ratios.

[0410] The heterodimeric antibodies of the present invention are prepared by culturing host cells containing an expression vector, as is well known in the art. After antibody generation, conventional antibody purification steps are performed, including ion-exchange chromatography. As described herein, ensuring that the pIs of the two monomers differ by at least 0.5 allows separation by ion-exchange chromatography, isoelectric focusing, or other methods sensitive to the isoelectric point. That is, pI substitutions that alter the isoelectric point (pI) of each monomer result in different pIs for each monomer and also different pIs for the heterodimer, thereby facilitating isoelectric purification of the “triF” heterodimer (e.g., anion-exchange columns, cation-exchange columns). These substitutions also facilitate the determination and monitoring of various contaminating bis-scFv-Fc and mAb homodimers after purification (e.g., IEF gels, cIEF, and analytical IEX columns).

[0411] treat

[0412] The compositions of this invention have a variety of applications. CD20, CD38, and CD123 are dysregulated in many hematopoietic malignancies and various hematopoietic malignancy cell lines. Therefore, the heterodimer antibodies of this invention can be used to treat cancers including, but not limited to, all B-cell lymphomas and leukemias, including, but not limited to, non-Hodgkin lymphoma (NHL), Burkitt lymphoma (BL), multiple myeloma (MM), B-type chronic lymphocytic leukemia (B-CLL), B-type and T-type acute lymphoblastic leukemia (ALL), T-cell lymphoma (TCL), acute myeloid leukemia (AML), hairy cell leukemia (HCL), Hodgkin lymphoma (HL), chronic lymphocytic leukemia (CLL), non-Hodgkin lymphoma, and chronic myeloid leukemia (CML).

[0413] Therefore, the heterodimeric compositions of the present invention can be used to treat these cancers.

[0414] Antibody compositions for in vivo administration

[0415] The antibody formulations used according to the present invention are prepared and preserved by mixing an antibody of the desired purity with an optional pharmaceutically acceptable carrier, excipient, or stabilizer (Remington's Pharmaceutical Sciences, 16th edition, Osol, A. ed.,

[1980] ) to form a lyophilized formulation or an aqueous solution. Acceptable carriers, excipients, or stabilizers that are non-toxic to the recipient at the doses and concentrations used include: buffers such as phosphates, citrates, and other organic acid buffers; antioxidants, including ascorbic acid and methionine; preservatives (such as octadecyl dimethyl benzyl ammonium chloride; hexachlorocyclohexane quaternary ammonium chloride; benzalkonium chloride, benzyl chloride; phenol, butyl or benzyl alcohol; alkyl p-hydroxybenzoate, such as methyl or propyl p-hydroxybenzoate; catechol; resorcinol; cyclohexanol; 3-pentanol; and m-cresol); low molecular weight (less than about 10 residues) Polypeptides (based on hydroxyl groups); 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 sugars, including glucose, mannose, or dextrin; chelating agents, such as EDTA; sugars, such as sucrose, mannitol, trehalose, or sorbitol; counterions that form salts, such as sodium; metal complexes (such as Zn-protein complexes); and / or nonionic surfactants, such as Tween. TM ), PLURONICS TM ) or polyethylene glycol (PEG).

[0416] Depending on the specific indications required for treatment, the formulations described herein may also contain more than one active compound, preferably those with complementary activities that do not interfere with each other. For example, it may be necessary to provide antibodies with additional specificity. Alternatively, or in addition to these, the composition may contain cytotoxic agents, cytokines, growth inhibitors, and / or small molecule antagonists. These molecules are preferably present in the composition at an effective amount for the target effect.

[0417] Active ingredients can also be encapsulated, for example through coagulation techniques or interfacial polymerization, such as hydroxymethyl cellulose or gelatin microcapsules and poly(methyl methacrylate) microcapsules, or encapsulated in colloidal drug delivery systems (e.g., liposomes, albumin microspheres, microemulsions, nanoparticles, and nanocapsules) or macroemulsions. These techniques are described in Remington Pharmaceutical Sciences, 16th edition, Osol, A. (ed., 1980).

[0418] Formulations intended for internal administration should be sterile or nearly sterile. This can be easily achieved through filtration using sterile membrane filters.

[0419] Sustained-release formulations can be prepared. Suitable examples of sustained-release formulations include antibodies contained in a semi-permeable matrix of a solid hydrophobic polymer, wherein the matrix is ​​in the form of a molded article, such as a membrane or microcapsule. Examples of sustained-release matrices include: polyesters, hydrogels (e.g., poly(2-hydroxyethyl-methacrylate) or poly(vinyl alcohol)), polylactide (US Patent No. 3,773,919), copolymers of L-glutamic acid and γ-ethyl-L-glutamic acid, non-degradable ethylene-vinyl acetate, and degradable lactic acid-glycolic acid copolymers such as LUPRON DEPOT. TM (Injectable microspheres composed of lactic-glycolic acid copolymer and leuprolide acetate) and poly-D-(-)-3-hydroxybutyric acid. Polymers such as ethylene-vinyl acetate and lactic-glycolic acid can release molecules for more than 100 days, while some hydrogels release proteins for a shorter time.

[0420] Encapsulated antibodies can persist in vivo for extended periods. However, exposure to a humid environment (37°C) can cause denaturation or aggregation, leading to decreased biological activity and potential alterations in the immunoprototype. Stabilization can be achieved through well-designed strategies, depending on the underlying mechanisms. For example, if aggregation is found to occur via sigma-disulfide interconversion to form SS bonds, stabilization can be achieved by modifying thiol residues, lyophilizing with acidic solutions, controlling moisture content, using appropriate additives, and developing specific polymer matrix compositions.

[0421] Dosage mode

[0422] The antibodies and chemotherapeutic agents of the present invention are administered to the treated subject according to known methods, such as intravenous administration by bolus injection or continuous infusion over a period of time, via intramuscular, intraperitoneal, intraspinal, subcutaneous, intra-articular, intrasynovial, intrathecal, oral, local, or inhalation routes. Intravenous or subcutaneous antibody administration is preferred.

[0423] Treatment mode

[0424] In the method of the present invention, the treatment provides a positive therapeutic effect with respect to a disease or condition. A “positive therapeutic effect” refers to improvement of the disease or condition and / or improvement of symptoms associated with the disease or condition. For example, a positive therapeutic effect would refer to one or more of the following improvements: (1) a reduction in the number of tumor cells; (2) an increase in tumor cell death; (3) inhibition of tumor cell survival; (5) inhibition of tumor growth (i.e., slowing it to a certain extent, preferably stopping it); (6) an increase in patient survival; and (7) a certain degree of relief of one or more symptoms associated with the disease or condition.

[0425] The effectiveness of treatment in various specific diseases or conditions can be evaluated using standardized efficacy criteria specific to that disease or condition. Tumor efficacy can be assessed based on changes in tumor morphology (i.e., total tumor burden, tumor size, etc.) using techniques such as magnetic resonance imaging (MRI), X-ray imaging, computed tomography (CT) scans, bone scans, endoscopy, and tumor biopsy sampling, including screening techniques such as bone marrow aspiration (BMA) and circulating tumor cell counting.

[0426] In addition to these positive treatment effects, patients receiving treatment will also experience improvement in disease-related symptoms.

[0427] Improvement in disease can be characterized as complete remission. "Complete remission" means the absence of clinically detectable disease, normalization of previously abnormal radiographic findings, and (in the case of myeloma) normalization of bone marrow and cerebrospinal fluid (CSF) or abnormal monoclonal proteins.

[0428] Following treatment according to the method of the invention, this effect can last for at least 4 to 8 weeks, sometimes 6 to 8 weeks. Alternatively, disease improvement can be classified as partial remission. "Partial remission" means a reduction of at least about 50% in all measurable tumor burden (i.e., the number of malignant cells in the subject, or the measured tumor mass volume or the amount of aberrant monoclonal proteins), with no new lesions, lasting for 4 to 8 weeks or 6 to 8 weeks.

[0429] The treatment of this invention includes a "therapeuticly effective amount" of the drug used. "Therapeuticly effective amount" refers to the amount that, according to a certain dosage and necessary duration, can achieve the desired therapeutic effect.

[0430] Therapeutic effective doses vary depending on factors such as an individual's disease state, age, sex, weight, and the drug's ability to produce the desired effect in that individual. Therapeutic effective doses are also the amount of antibody or antibody fraction whose beneficial therapeutic effect outweighs its toxic or harmful effects.

[0431] The "therapeutic effective dose" of cancer treatment can also be measured by its ability to stabilize disease progression. The ability of compounds to inhibit cancer is evaluated in animal model systems that can predict efficacy in human tumors.

[0432] Alternatively, this property of the composition can be evaluated by testing the compound's ability to inhibit cell growth or induce apoptosis, such as using in vitro assays known to those skilled in the art. A therapeutically effective amount of the therapeutic compound can reduce tumor size or alleviate symptoms in the subject. Those skilled in the art can determine the therapeutically effective amount based on factors such as the subject's physical condition, the severity of the subject's symptoms, and the specific composition or chosen route of administration.

[0433] The dosing regimen can be adjusted to provide the optimal desired effect (e.g., therapeutic effect). For example, it can be a single bolus injection, divided doses over time, or the dosage can be proportionally reduced or increased according to the urgency of the treatment situation. Parenteral compositions can be formulated into unit dosage forms for ease of administration and uniform dosage. In this context, a unit dosage form refers to a physically discrete unit used as a single dose for the subject of treatment, each unit containing a predetermined amount of the active compound and the desired drug delivery vehicle, which is calculated to produce the desired therapeutic effect.

[0434] The specifications of the unit dosage form in this invention depend on or are directly dependent on (a) the unique properties of the active compound and the specific therapeutic effect to be achieved, and (b) the inherent limitations in the art regarding the response of such active compound to individual sensitivities.

[0435] The effective dosage and administration regimen of the bispecific antibody of the present invention depend on the disease or condition to be treated and can be determined by those skilled in the art.

[0436] Exemplary, and non-limiting, ranges for the therapeutically effective amount of the bispecific antibody used in this invention are about 0.1-100 mg / kg, such as about 0.1-50 mg / kg, such as about 0.1-20 mg / kg, such as about 0.1-10 mg / kg, such as about 0.5, about 0.3, about 1, or about 3 mg / kg. In another embodiment, the dosage of the antibody is 1 mg / kg or higher, such as 1 to 20 mg / kg, such as 5 to 20 mg / kg, such as 8 mg / kg.

[0437] A medical professional of ordinary skill in the art can readily determine and prescribe the required effective amount of a pharmaceutical composition. For example, for a particular drug in a pharmaceutical composition, a physician or veterinarian may start with a level below what is required for the desired therapeutic effect and then gradually increase the dosage until the desired effect is achieved.

[0438] In one embodiment, the bispecific antibody is administered by weekly infusion at a dose of 10 to 500 mg / kg (e.g., 200 to 400 mg / kg). Such administration can be repeated, for example, 1 to 8 times, or 3 to 5 times. Administration can be a continuous infusion over 2 to 24 hours (e.g., 2 to 12 hours).

[0439] In one implementation method, if it is necessary to reduce side effects (including toxicity), the bispecific antibody is administered via slow, continuous infusion over a prolonged period, such as more than 24 hours.

[0440] In one embodiment, the bispecific antibody is administered at a weekly dose of 250 mg to 2000 mg (e.g., 300 mg, 500 mg, 700 mg, 1000 mg, 1500 mg, or 2000 mg), up to eight times, such as four to six times. Administration can be a continuous infusion over 2 to 24 hours (e.g., 2 to 12 hours). This regimen can be repeated once or multiple times as needed, for example, after 6 or 12 months. The dosage of the compound of the invention can be determined or adjusted by measuring the amount of the compound in the blood after administration, such measurement may be, for example, taking a biological sample and using an anti-idiotype antibody targeting the antigen-binding region of the bispecific antibody.

[0441] In another embodiment, the bispecific antibody is administered weekly for 2 to 12 weeks, for example 3 to 10 weeks, or for example 4 to 8 weeks.

[0442] In one implementation method, the bispecific antibody is administered as maintenance therapy, for example, once a week for 6 months or longer.

[0443] In one implementation, the bispecific antibody administration regimen includes a single bispecific antibody infusion followed by an infusion of a bispecific antibody conjugated to a radioisotope. This regimen can be repeated, for example, after 7 to 9 days.

[0444] As a non-limiting example, the treatment according to the invention may be to provide antibodies at a daily dose of about 0.1-100 mg / kg, for example, 0.5, 0.9, 1.0, 1.1, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 40, 45, 50, 60, 70, 80, 90, or 100 mg / kg daily, on the 1st, 2nd, 3rd, 4th, 5th, 6th, 7th, 8th, or 9th day after the start of treatment. At least one day of 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, or 40 days, or at least one week of weeks 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20, or any combination thereof, using a single dose or multiple doses at intervals of 24, 12, 8, 6, 4, or 2 hours, or any combination thereof.

[0445] In some embodiments, the bispecific antibody molecule is used in combination with one or more other therapeutic agents, such as chemotherapeutic agents. Non-limiting examples of DNA-damaging chemotherapeutic agents include topoisomerase I inhibitors (e.g., irinotecan, topotecan, camptothecin and its analogues or metabolites, and doxorubicin); topoisomerase II inhibitors (e.g., etoposide, teniposide, and daunorubicin); alkylating agents (e.g., melphalan, chlorambucil, busulfan, thiotepa, ifosfamide, carmustine, lomustine, ulinastatin, streptozotocin, desmethylpyrazine, methotrexate, mitomycin C, and cyclophosphamide); DNA intercalating agents (e.g., cisplatin, oxaliplatin, and carboplatin); DNA intercalating agents and free radical generators such as bleomycin; and nucleoside mimics (e.g., 5-fluorouracil, capecitabine, gemcitabine, fludarabine, cytarabine, mercaptopurine, thioguanine, pentostatin, and hydroxyurea).

[0446] Chemotherapy agents that disrupt cell replication include: paclitaxel, docetaxel, and related analogs; vincristine, vinblastine, and related analogs; thalidomide, lenalidomide, and related analogs (e.g., CC-5013 and CC-4047); protein tyrosine kinase inhibitors (e.g., imatinib mesylate and gefitinib); proteasome inhibitors (e.g., bortezomib); NF-κB inhibitors, including IκB kinase inhibitors; antibodies that bind to proteins overexpressed in cancer and thereby downregulate cell replication (e.g., trastuzumab, rituximab, cetuximab, and bevacizumab); and other inhibitors of proteins or enzymes known to be upregulated, overexpressed, or activated in cancer, the inhibition of which downregulates cell replication.

[0447] In some embodiments, the antibody of the present invention can be (Bortezomib) can be used before, during, or after treatment.

[0448] All cited references are explicitly included in this paper through citation.

[0449] While several specific embodiments of the present invention have been described above for illustrative purposes, those skilled in the art will recognize that many changes can be made to the details, all of which fall within the scope of the present invention as defined in the claims. Example

[0450] The invention is illustrated below by way of examples. These examples are not intended to limit the invention to any particular application or principle. All constant region locations described in this invention are numbered according to Kabat's EU index (Kabat et al., 1991, Sequences of Proteins of Immunological Interest, 5th ed., Public Health Service, National Institutes of Health, Bethesda, incorporated herein by reference in its entirety). Those skilled in the art of antibodies will know that this convention provides a standardized reference for conserved locations within the immunoglobulin family by using non-sequential numbering of specific regions of the immunoglobulin sequence. Therefore, the location of any given immunoglobulin as defined by the EU index does not necessarily correspond to its sequential sequence.

[0451] U.S. Publications 2015 / 0307629, 2014 / 0288275, and WO2014 / 145806 contain general and specific technologies, all of which are incorporated herein by reference, and the technologies contained therein are also included.

[0452] Example

[0453] Example 1: Transformation

[0454] Production of bispecific antibodies

[0455] A schematic diagram of the anti-CD38×anti-CD3 bispecific antibody is shown below. Figure 1 The amino acid sequences of various modified forms of anti-CD38×anti-CD3 bispecific antibodies are listed below. Figures 39 to 43The DNA encoding the three strands required for bispecific antibody expression was produced via gene synthesis (Blue Heron Biotechnology, Bosell, Washington) and subcloned into the expression vector pTT5 using standard molecular biology techniques. Substitution was introduced using site-directed mutagenesis (QuikChange, Stratagene, Cedar Creek, Texas) or other gene synthesis and subcloning methods. The DNA was transfected into HEK293E cells for expression, and the resulting protein was purified from the supernatant using protein A affinity chromatography (GE Healthcare) and cation exchange chromatography. The yield of protein A affinity purification was as follows: Figure 35 As shown. Cation exchange chromatography purification was performed using a HiTrap SP HP column (GE Healthcare), 50 mM MES, pH 6.0 wash / equilibration buffer and 50 mM MES elution buffer, with a linear gradient of pH 6.0 + 1 M NaCl (see...). Figure 36 (Chromatogram).

[0456] Redirecting T cell cytotoxicity

[0457] In vitro identification of anti-CD38×anti-CD3 bispecific antibodies against CD38 + Redirected T-cell cytotoxicity (RTCC) of the RPMI8266 myeloma cell line. 10kJ RPMI8266 cells were incubated with 500kJ human PBMCs for 24 hours. RTCC was measured by LDH fluorescence (see [link to LDH assay]). Figure 37 ).

[0458] Example 2

[0459] Redirecting T cell cytotoxicity

[0460] In vitro identification of the redirected T-cell cytotoxicity (RTCC) of the anti-CD38×anti-CD3 Fab-scFv-Fc bispecific antibody against the CD38+RPMI8266 myeloma cell line. 40kJ RPMI8266 cells were incubated with 400kJ human PBMCs for 96 hours. RTCC was measured by flow cytometry. Figure 44 As shown. The expression of CD69, Ki-67, and PI-9 in CD4+ and CD8+ T cells was also identified by flow cytometry, such as... Figure 45 As shown.

[0461] Mouse models of anti-tumor activity

[0462] On day -23, five NODscidγ (NSG) mice in each of the four groups (five mice per group, totaling four groups) were injected intravenously with 5 × 10⁶ N / A of the drug. 6 RPMI8226TrS tumor cells (multiple myeloma, expressing luciferase). On day 0, 10 × 10⁶ cells were implanted intraperitoneally in mice.6 Personal PBMCs. Following PBMC implantation on day 0, the test drug was administered weekly (on days 0 and 7) via intraperitoneal injection at the following dosage levels. Figure 4 As shown. Figure 46 The experimental design is summarized. An in vivo imaging system was used. Total flux was measured in each mouse to monitor tumor growth. Both XmAb13551 and XmAb15426 showed significant antitumor activity (see [link to relevant documentation]). Figure 47 and Figure 48 ).

[0463] Macaque experiment

[0464] Rhesus monkeys were administered a single dose of anti-CD38×anti-CD3 bispecific antibody. The experiment also included an anti-RSV×anti-CD3 bispecific antibody control. Dosage levels were: 20 μg / kg XmAb13551 (n=2), 0.5 mg / kg XmAb15426 (n=3), 3 mg / kg XmAb14702 (n=3), or 3 mg / kg XmAb13245 (anti-RSV×anti-CD3 control, n=3) (3 independent experiments). The anti-CD38×anti-CD3 bispecific antibody rapidly depleted CD38+ cells in peripheral blood (see...). Figure 49 According to CD69 expression assays, anti-CD38×anti-CD3 bispecific antibodies led to T cell activation (see...). Figure 50 Serum IL-6 levels were also measured (see...). Figure 51 Compared with XmAb13551, XmAb15426 was found to have a prolonged CD38+ cell clearance duration, lower T cell activation, and lower IL-6 production levels.

[0465] XmAb15426 and XmAb14702 were tested at single doses of 0.5 mg / kg and 3 mg / kg. Both antibodies were well tolerated at these high doses, consistent with moderate serum IL6 levels in the treated monkeys. Furthermore, XmAb15426 with moderate CD3 affinity at 0.5 mg / kg was more effective at clearing CD38+ cells than the previously high-affinity XmAb13551 administered at 2, 5, or 20 μg / kg. The clearance effect of XmAb15426 lasted longer than that of the previous highest dose of XmAb13551 (7 days for XmAb15426 vs. 2 days for XmAb13551). Notably, despite the greater target cell clearance efficacy of XmAb15426, T cell activation (CD69, CD25, and PD1 induction) was significantly lower, even in monkeys treated with XmAb15426 at a dose 25 times higher than the 20 μg / kg XmAb13551 group. XmAb14702, with its very low CD3 affinity, had virtually no effect on CD38+ cell and T cell activation.

[0466] These results demonstrate that modulating T cell activation by reducing CD3 affinity is a feasible approach to improving the therapeutic window of T cell-binding bispecific antibodies. This strategy improves tolerability and allows overcoming antigen sink clearance of targets such as CD38 at high doses, thereby expanding the antigenome suitable for targeted T cell immunotherapy. We have demonstrated that XmAb15426 effectively eliminates CD38+ cells by reducing CD3 affinity, while minimizing the CRS effect visible at comparable doses of its high-affinity counterpart, XmAb13551.

[0467] Example 3

[0468] CD123's CDR Development

[0469] The starting point for the CDR of the humanized CD123 antibody Fab is the variable and light chain region of the 7G3 mouse antibody, named "7G3H0L0", derived from ATCC HB-12009. However, the initial humanization (H1_L1, sequence shown in Figure 136) resulted in a severe decrease in affinity (5 to 10-fold affinity, see Figure 136). Figure 156B and C This decrease in affinity is primarily due to the humanization of the heavy chain, as seen in the H1_L0 construct (e.g., the first humanized heavy chain versus the mouse light chain), while the H1_L1 construct exhibits a full 10-fold reduction. Consistent with this is a 10-fold decrease in RTCC (redirected T cell cytotoxicity), as seen in... Figure 156D The experiment targeted KG1a cells that express CD123.

[0470] Therefore, two rounds of affinity / stabilization optimization were performed. The first round (see "Library 1") Figure 157 108 variants were generated, including LDA, targeted and reversion substitutions, which were then subjected to affinity screening on the CD123 chip in Fab form (humanized variable heavy chain domain fused with CH1 of human IgG1), and neutral and high affinity variants were then subjected to stability screening on DSF.

[0471] As shown in Figure 158, the Tm of the original H1L1 variant is higher than that of the original H0L0. Figure 158 also shows the results of other parent variants of H1L1.

[0472] The first-round variant was then constructed into a bottle opener form, as described later, using an anti-CD3 scFv and the developed Fab, and KG-1a binding tests and RTCC experiments were conducted, as shown in Figure 159. Although the first round of optimization improved the affinity and efficacy of the variant, further optimization is still needed.

[0473] In the second round (“Round 2”), as shown in Figure 160, binding affinity was restored to the H1L1 mouse level, and RTCC activity was also restored. The optimal variant, XENP14045, exhibited improved affinity higher than both the first humanized sequence and the parent mouse antibody, representing a +21-fold increase over the first humanized sequence (H1L1) and a 2-fold increase over the parent mouse antibody (7G3, H0L0). It should be noted that the CD123 side of XENP13967 is identical to that of XENP14045, but the CD3 scFv differs, as shown in the sequence.

[0474] The second round of optimization also improved stability, as shown by the measured Tm. Figure 161 The Tm test results showed that XENP13967 (and its corresponding counterpart XENP14045) had a +5C improvement over the original chimeric antibody (such as the variable heavy chain and light chain mouse sequences) and a +4C improvement over the H1L1 variant. (13967 / 14045 has 11 substitutions compared to the original H1L1 sequence). Furthermore, the potential deamide site (-NS motif) of the light chain CDR1 was eliminated in the second round.

[0475] Example 4

[0476] CD20's CDR Development

[0477] Binding affinity and potency of two anti-CD20 Fab antibodies in the context of CD20×CD3 bispecific antibodies were investigated. Both XENP13677 and XENP13676 are based on rituximab. The 13677 variant was significantly more potent than the 13676 variant, and its CD20 affinity was close to that of the parent rituximab antibody. In vivo performance of the two bispecific antibodies was compared in rhesus monkeys. However, due to the higher potency of the 13677 variant, the dosage was 0.03 mg / kg, 10 times lower than the lower potency of the 13676 (0.3 mg / kg). At these doses, both antibodies significantly depleted monkey B cells. However, surprisingly, the significantly more potent 13677 showed faster B cell recovery at the lower dose. On the other hand, the IL6 release induced by both antibodies was roughly the same. As such, the low-affinity variant 13676 unexpectedly exhibited a more suitable therapeutic profile, inducing more durable B-cell clearance while maintaining similar IL6 levels.

[0478] Examples of embodiments of the present invention include:

[0479] 1. A heterodimeric antibody comprising:

[0480] a) The first monomer, which comprises:

[0481] i) The first chain, which comprises:

[0482] 1) First variable heavy chain structural domain;

[0483] 2) The first constant heavy chain containing the first Fc structural domain;

[0484] 3) An scFv comprising a variable light chain structural domain, an scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded to the C-terminus of the Fc structural domain by a structural domain connector;

[0485] b) A second monomer comprising a second heavy chain, the second heavy chain comprising a second variable heavy chain domain and a second constant heavy chain, the second constant heavy chain comprising a second Fc domain; and

[0486] c) A common light chain containing both variable light chain structural domains and constant light chain structural domains;

[0487] The first and second Fc domains have amino acid substitutions selected from the following combinations: S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, and K370S:S364K / E357Q, and wherein the first variable heavy chain domain and the variable light chain domain bind to a first target tumor antigen (TTA), the second variable heavy chain domain and the variable light chain domain bind to the first TTA, and the scFv binds to human CD3 (SEQ ID NO:XX).

[0488] 2. The heterodimeric antibody as described in Embodiment 1, wherein the scFv has a polypeptide sequence selected from the group consisting of: SEQ ID NO:XX (scFv 13551), SEQ ID NO:XX (scFv 15426), SEQ ID NO:XX (scFv 13423) and SEQ ID NO:XX (scFv 14702).

[0489] 3. The heterodimeric antibody as described in Embodiment 1 or 2, wherein the first variable heavy chain domain and the variable light chain domain are bound to TTAs selected from the following: CD19, CD20 and CD123.

[0490] 4. A heterodimeric antibody comprising:

[0491] a) The first monomer, which comprises:

[0492] i) The first chain, which comprises:

[0493] 1) First variable heavy chain structural domain;

[0494] 2) A first constant heavy-chain structural domain containing a first Fc structural domain; and

[0495] 3) A first variable light chain structural domain, wherein the first variable light chain structural domain is covalently joined to the C-end of the first Fc structural domain by a structural domain connector;

[0496] b) The second monomer, which comprises:

[0497] i) Second variable heavy chain structural domain;

[0498] ii) A second constant heavy chain structure containing a second Fc structure domain; and

[0499] iii) A third variable heavy chain structural domain, wherein the second variable heavy chain structural domain is covalently joined to the C-end of the second Fc structural domain by a structural domain joint;

[0500] c) A common light chain containing both variable light chain structural domains and constant light chain structural domains;

[0501] The first and second Fc domains have amino acid substitutions selected from the following combinations: S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, and K370S:S364K / E357Q, wherein the first variable heavy chain domain and the variable light chain domain are bound to a first TTA, the second variable heavy chain domain and the variable light chain domain are bound to the TTA, and the second variable light chain domain and the third variable heavy chain domain are bound to CD3.

[0502] 5. The heterodimeric antibody as described in Embodiment 4, wherein the scFv has a polypeptide sequence selected from the group consisting of: SEQ ID NO:XX (scFv 13551), SEQ ID NO:XX (scFv 15426), SEQ ID NO:XX (scFv 13423) and SEQ ID NO:XX (scFv 14702).

[0503] 6. The heterodimeric antibody as described in embodiment 4 or 5, wherein the first variable heavy chain domain and the variable light chain domain are bound to TTAs selected from the following: CD19, CD20 and CD123.

[0504] 7. A heterodimeric antibody comprising:

[0505] a) The first monomer, which comprises:

[0506] i) The first chain, which comprises:

[0507] 1) First variable heavy chain structural domain;

[0508] 2) A first constant heavy chain containing a first CH1 domain and a first Fc domain;

[0509] 3) An scFv comprising a variable light chain structural domain, an scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded between the C-end of the CH1 structural domain and the N-end of the first Fc structural domain by a structural domain connector.

[0510] b) A second monomer comprising a second heavy chain, the second heavy chain comprising a second variable heavy chain domain and a second constant heavy chain, the second constant heavy chain comprising a second Fc domain; and

[0511] c) A common light chain containing both variable light chain structural domains and constant light chain structural domains;

[0512] The first and second Fc domains have amino acid substitutions selected from the following combinations: S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, and K370S:S364K / E357Q, wherein the first variable heavy chain domain and the variable light chain domain bind to a first TTA, the second variable heavy chain domain and the variable light chain domain bind to the TTA, and the scFv binds to human CD3.

[0513] 8. The heterodimeric antibody as described in Embodiment 7, wherein the scFv has a polypeptide sequence selected from the group consisting of: SEQ ID NO:XX (scFv 13551), SEQ ID NO:XX (scFv 15426), SEQ ID NO:XX (scFv 13423) and SEQ ID NO:XX (scFv 14702).

[0514] 9. The heterodimeric antibody as described in embodiment 7 or 8, wherein the first variable heavy chain domain and the variable light chain domain are bound to TTAs selected from the following: CD19, CD20 and CD123.

[0515] 10. A heterodimeric antibody comprising:

[0516] a) The first monomer, which comprises:

[0517] i) The first chain, which comprises:

[0518] 1) First variable heavy chain structural domain;

[0519] 2) A first constant heavy-chain structural domain containing a first Fc structural domain; and

[0520] 3) A first variable light chain structural domain, wherein the second variable light chain structural domain is covalently joined by a structural domain joint between the C-end of the CH1 structural domain of the first constant heavy chain structural domain and the N-end of the first Fc structural domain.

[0521] b) The second monomer, which contains

[0522] i) Second variable heavy chain structural domain;

[0523] ii) A second constant heavy chain structure containing a second Fc structure domain; and

[0524] iii) A third variable heavy chain structural domain, wherein the second variable heavy chain structural domain is covalently joined to the C-end of the second Fc structural domain by a structural domain joint;

[0525] c) A common light chain containing both variable light chain structural domains and constant light chain structural domains;

[0526] The first and second Fc domains have amino acid substitutions selected from the following combinations: S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, and K370S:S364K / E357Q, wherein the first variable heavy chain domain and the variable light chain domain bind to a first TTA, the second variable heavy chain domain and the variable light chain domain bind to the TTA, and the second variable light chain domain and the third variable heavy chain domain bind to human CD3.

[0527] 11. The heterodimeric antibody as described in Embodiment 10, wherein the scFv has a polypeptide sequence selected from the group consisting of: SEQ ID NO:XX (scFv 13551), SEQ ID NO:XX (scFv 15426), SEQ ID NO:XX (scFv 13423) and SEQ ID NO:XX (scFv 14702).

[0528] 12. The heterodimeric antibody as described in embodiment 10 or 11, wherein the first variable heavy chain domain and the variable light chain domain are bound to TTAs selected from the group consisting of CD19, CD20 and CD123.

[0529] 13. A heterodimeric antibody comprising:

[0530] a) The first monomer, which comprises:

[0531] i) The first chain, which comprises:

[0532] 1) First variable heavy chain structural domain;

[0533] 2) A first constant heavy chain containing a first CH1 domain and a first Fc domain;

[0534] 3) An scFv comprising a variable light chain structural domain, an scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded between the C-end of the CH1 structural domain and the N-end of the first Fc structural domain by a structural domain connector.

[0535] b) A second monomer containing a second Fc domain; and

[0536] c) Light chains containing variable light chain structural domains and constant light chain structural domains;

[0537] The first and second Fc domains have amino acid substitutions selected from the following combinations: S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, and K370S:S364K / E357Q, wherein the first variable heavy chain domain and the variable light chain domain bind to the first antigen, and the scFv binds to the second antigen.

[0538] 14. The heterodimeric antibody as described in Embodiment 13, wherein the scFv has a polypeptide sequence selected from the group consisting of: SEQ ID NO:XX (scFv 13551), SEQ ID NO:XX (scFv 15426), SEQ ID NO:XX (scFv 13423) and SEQ ID NO:XX (scFv 14702).

[0539] 15. The heterodimeric antibody as described in embodiment 13 or 14, wherein the first variable heavy chain domain and the variable light chain domain are bound to TTAs selected from the group consisting of CD19, CD20 and CD123.

[0540] 16. An anti-CD3 antibody binding domain, comprising:

[0541] a) A variable light chain domain comprising vlCDR1 with the sequence GSSTGAVTTSNYAN (SEQ ID NO:XX), vlCDR2 with the sequence GTNKRAP (SEQ ID NO:XX), and vlCDR3 with the sequence ALWYSNHWV (SEQ ID NO:XX); and

[0542] b) A variable heavy chain domain comprising vhCDR1 with sequence TYAMN (SEQ ID NO:XX), vhCDR2 with sequence RIRSKANNYATYYADSVKG (SEQ ID NO:XX), and vhCDR3 with sequence HGNFGDSYVSWFAY (SEQ ID NO:XX).

[0543] 17. The anti-CD3 antibody binding domain as described in Embodiment 16, wherein the binding domain is scFv.

[0544] 18. The anti-CD3 antibody binding domain as described in embodiment 16 or 17, wherein the variable light chain domain has sequence L1.47 (SEQ ID NO:XX) and the variable heavy chain domain has sequence H1.32 (SEQ ID NO:XX).

[0545] 19. The anti-CD3 antibody binding domain as described in Embodiment 18, wherein the scFv has the sequence H1.32_L1.47 (SEQ ID NO:XX).

[0546] 20. A nucleic acid composition encoding scFv as described in Embodiment 19.

[0547] 21. An expression vector comprising the nucleic acid composition as described in Embodiment 20.

[0548] 22. A host cell comprising the expression vector described in embodiment 21.

[0549] 23. An anti-CD3 antibody binding domain, comprising:

[0550] a) A variable light chain domain comprising vlCDR1 with the sequence GSSTGAVTTSNYAN (SEQ ID NO:XX), vlCDR2 with the sequence GTNKRAP (SEQ ID NO:XX), and vlCDR3 with the sequence ALWYSNHWV (SEQ ID NO:XX); and

[0551] b) A variable heavy chain domain comprising vhCDR1 with sequence TYAMN (SEQ ID NO:XX), vhCDR2 with sequence RIRSKYNNYATYYADSVKG (SEQ ID NO:XX), and vhCDR3 with sequence HGNFGDEYVSWFAY (SEQ ID NO:XX).

[0552] 24. The anti-CD3 antibody binding domain as described in Embodiment 23, wherein the binding domain is scFv.

[0553] 25. The anti-CD3 antibody binding domain as described in embodiment 23 or 24, wherein the variable light chain domain has sequence L1.47 (SEQ ID NO:XX) and the variable heavy chain domain has sequence H1.89 (SEQ ID NO:XX).

[0554] 26. The anti-CD3 antibody binding domain as described in Embodiment 23, wherein the scFv has the sequence H1.89_L1.47 (SEQ ID NO:XX).

[0555] 27. A nucleic acid composition encoding scFv as described in Embodiment 26.

[0556] 28. An expression vector comprising the nucleic acid composition as described in Embodiment 27.

[0557] 29. A host cell comprising the expression vector described in embodiment 28.

[0558] 30. An anti-CD3 antibody binding domain, comprising:

[0559] a) A variable light chain domain comprising vlCDR1 with the sequence GSSTGAVTTSNYAN (SEQ ID NO:XX), vlCDR2 with the sequence GTNKRAP (SEQ ID NO:XX), and vlCDR3 with the sequence ALWYSNHWV (SEQ ID NO:XX); and

[0560] b) A variable heavy chain domain comprising vhCDR1 with sequence TYAMN (SEQ ID NO:XX), vhCDR2 with sequence RIRSKYNNYATYYADSVKG (SEQ ID NO:XX), and vhCDR3 with sequence HGNFGDPYVSWFAY (SEQ ID NO:XX).

[0561] 31. The anti-CD3 antibody binding domain as described in embodiment 30, wherein the binding domain is scFv.

[0562] 32. The anti-CD3 antibody binding domain as described in embodiment 30 or 31, wherein the variable light chain domain has sequence L1.47 (SEQ ID NO:XX) and the variable heavy chain domain has sequence H1.90 (SEQ ID NO:XX).

[0563] 33. The anti-CD3 antibody binding domain as described in embodiment 30, wherein the scFv has the sequence H1.90_L1.47 (SEQ ID NO:XX).

[0564] 34. A nucleic acid composition encoding scFv as described in Embodiment 33.

[0565] 35. An expression vector comprising the nucleic acid composition as described in embodiment 34.

[0566] 36. A host cell comprising the expression vector described in embodiment 35.

[0567] 37. An anti-CD3 antibody binding domain, comprising:

[0568] a) A variable light chain domain comprising vlCDR1 with the sequence GSSTGAVTTSNYAN (SEQ ID NO:XX), vlCDR2 with the sequence GTNKRAP (SEQ ID NO:XX), and vlCDR3 with the sequence ALWYSNHWV (SEQ ID NO:XX); and

[0569] b) A variable heavy chain domain comprising vhCDR1 with sequence TYAMN (SEQ ID NO:XX), vhCDR2 with sequence RIRSKYNNYATYYADSVKG (SEQ ID NO:XX), and vhCDR3 with sequence HGNFGDSYVSWFDY (SEQ ID NO:XX).

[0570] 38. The anti-CD3 antibody binding domain as described in embodiment 37, wherein the binding domain is scFv.

[0571] 39. The anti-CD3 antibody binding domain as described in embodiment 37 or 38, wherein the variable light chain domain has sequence L1.47 (SEQ ID NO:XX) and the variable heavy chain domain has sequence H1.33 (SEQ ID NO:XX).

[0572] 40. The anti-CD3 antibody binding domain as described in Embodiment 38, wherein the scFv has the sequence H1.33_L1.47 (SEQ ID NO:XX).

[0573] 41. A nucleic acid composition encoding scFv as described in Embodiment 38.

[0574] 42. An expression vector comprising the nucleic acid composition as described in embodiment 41.

[0575] 43. A host cell comprising the expression vector described in embodiment 42.

[0576] 44. An anti-CD3 antibody binding domain, comprising:

[0577] a) A variable light chain domain comprising vlCDR1 with the sequence GSSTGAVTTSNYAN (SEQ ID NO:XX), vlCDR2 with the sequence GTNKRAP (SEQ ID NO:XX), and vlCDR3 with the sequence ALWYSNHWV (SEQ ID NO:XX); and

[0578] b) A variable heavy chain domain comprising vhCDR1 with sequence TYAMS (SEQ ID NO:XX), vhCDR2 with sequence RIRSKYNNYATYYADSVKG (SEQ ID NO:XX), and vhCDR3 with sequence HGNFGDSYVSWFAY (SEQ ID NO:XX).

[0579] 45. The anti-CD3 antibody binding domain as described in embodiment 44, wherein the binding domain is scFv.

[0580] 46. ​​The anti-CD3 antibody binding domain as described in embodiment 44 or 45, wherein the variable light chain domain has sequence L1.47 (SEQ ID NO:XX) and the variable heavy chain domain has sequence H1.31 (SEQ ID NO:XX).

[0581] 47. The anti-CD3 antibody binding domain as described in embodiment 46, wherein the scFv has the sequence H1.31_L1.47 (SEQ ID NO:XX).

[0582] 48. A nucleic acid composition encoding scFv as described in Embodiment 47.

[0583] 49. An expression vector comprising the nucleic acid composition as described in embodiment 48.

[0584] 50. A host cell comprising the expression vector described in embodiment 49.

[0585] 51. A heterodimeric antibody comprising:

[0586] a) The first monomer, which comprises:

[0587] i) The first Fc structural domain;

[0588] ii) An anti-CD3 scFv comprising a scFv variable light chain structural domain, a scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded to the N-end of the Fc structural domain by a structural domain connector;

[0589] b) The second monomer, comprising a heavy chain, includes:

[0590] i) Heavy chain variable structural domain; and

[0591] ii) Heavy-chain constant structural domains containing a second Fc structural domain; and

[0592] c) Light chains containing variable light chain structural domains and variable light chain constant structural domains;

[0593] The anti-CD3 scFv is selected from: anti-CD3 H1.32_L1.47 (SEQ ID NO:XX), anti-CD3H1.89_L1.47 (SEQ ID NO:XX), anti-CD3 H1.90_L1.47 (SEQ ID NO:XX) and anti-CD3 H1.33_L1.47 (SEQ ID NO:XX), and the heavy chain variable domain and the light chain variable domain are combined with TTA.

[0594] 52. The heterodimeric antibody as described in embodiment 51, wherein the TTA is selected from CD19, CD20 and CD123.

[0595] 53. An anti-CD20 antibody binding domain, comprising:

[0596] a) A variable light chain domain comprising vlCDR1 with the sequence RASWSVSYIH (SEQ ID NO:XX), vlCDR2 with the sequence ATSNLAS (SEQ ID NO:XX), and vlCDR3 with the sequence QQWTHNPPT (SEQ ID NO:XX); and

[0597] b) A variable heavy chain domain comprising vhCDR1 with sequence SYNMH (SEQ ID NO:XX), vhCDR2 with sequence AIYPGNGATSYSQKFQG (SEQ ID NO:XX), and vhCDR3 with sequence SYYMGGDWYFDV (SEQ ID NO:XX).

[0598] 54. The anti-CD20 antibody binding domain as described in embodiment 53, wherein the variable light chain domain has the sequence C2B8 L1.113 (SEQ ID NO:XX) and the variable heavy chain domain has the sequence C2B8 H1.202 (SEQ ID NO:XX).

[0599] 55. A nucleic acid composition encoding a binding domain as described in Embodiment 53.

[0600] 56. An expression vector comprising the nucleic acid composition as described in embodiment 55.

[0601] 57. A host cell comprising the expression vector described in embodiment 56.

[0602] 58. An anti-CD20 antibody binding domain, comprising:

[0603] a) A variable light chain domain comprising vlCDR1 with sequence RASSSVSYIH (SEQ ID NO:XX), vlCDR2 with sequence ATSNLAS (SEQ ID NO:XX), and vlCDR3 with sequence QQWTSNPPT (SEQ ID NO:XX); and

[0604] b) A variable heavy chain domain comprising vhCDR1 with sequence SYNMH (SEQ ID NO:XX), vhCDR2 with sequence AIYPGNGDTSYNQKFQG (SEQ ID NO:XX), and vhCDR3 with sequence STYYGGDWYFNV (SEQ ID NO:XX).

[0605] 59. The anti-CD20 antibody binding domain as described in embodiment 58, wherein the variable light chain domain has the sequence C2B8 L1 (SEQ ID NO:XX) and the variable heavy chain domain has the sequence C2B8 H1 (SEQ ID NO:XX).

[0606] 60. A nucleic acid composition encoding a binding domain as described in Embodiment 58.

[0607] 61. An expression vector comprising the nucleic acid composition as described in embodiment 60.

[0608] 62. A host cell comprising the expression vector described in embodiment 61.

[0609] 63. A heterodimeric antibody comprising:

[0610] a) The first monomer, which comprises:

[0611] i) The first Fc structural domain;

[0612] ii) An anti-CD3 scFv comprising a scFv variable light chain structural domain, a scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded to the N-end of the Fc structural domain by a structural domain connector;

[0613] b) The second monomer, comprising a heavy chain, includes:

[0614] i) Heavy chain variable structural domain; and

[0615] ii) Heavy-chain constant structural domains containing a second Fc structural domain; and

[0616] c) A light chain comprising a variable light chain structural domain and a variable light chain constant structural domain; wherein the variable light chain structural domain comprises v1CDR1 having the sequence RASSSVSYIH (SEQ ID NO: XX), v1CDR2 having the sequence ATSNLAS (SEQ ID NO: XX), and v1CDR3 having the sequence QQWTSNPPT (SEQ ID NO: XX), and the variable heavy chain structural domain comprises vhCDR1 having the sequence SYNMH (SEQ ID NO: XX), vhCDR2 having the sequence AIYPGNGDTSYNQKFQG (SEQ ID NO: XX), and vhCDR3 having the sequence STYYGGDWYFNV (SEQ ID NO: XX).

[0617] 64. A heterodimeric antibody comprising:

[0618] a) The first monomer, which comprises:

[0619] i) The first Fc structural domain;

[0620] ii) An anti-CD3 scFv comprising a scFv variable light chain structural domain, a scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded to the N-end of the Fc structural domain by a structural domain connector;

[0621] b) The second monomer, comprising a heavy chain, includes:

[0622] i) Heavy chain variable structural domain; and

[0623] ii) Heavy-chain constant structural domains containing a second Fc structural domain; and

[0624] c) Light chains containing variable light chain structural domains and variable light chain constant structural domains;

[0625] The variable light chain domain comprises v1CDR1 with the sequence RASSSVSYIH (SEQ ID NO: XX), v1CDR2 with the sequence ATSNLAS (SEQ ID NO: XX), and v1CDR3 with the sequence QQWTSNPPT (SEQ ID NO: XX), and the variable heavy chain domain comprises vhCDR1 with the sequence SYNMH (SEQ ID NO: XX), vhCDR2 with the sequence AIYPGNGDTSYNQKFQG (SEQ ID NO: XX), and vhCDR3 with the sequence STYYGGDWYFNV (SEQ ID NO: XX).

[0626] 65. A heterodimeric antibody comprising:

[0627] a) The first monomer, which comprises:

[0628] i) The first Fc structural domain;

[0629] ii) An anti-CD3 scFv comprising a scFv variable light chain structural domain, a scFv connector, and a scFv variable heavy chain structural domain; wherein the scFv is covalently bonded to the N-end of the Fc structural domain by a structural domain connector;

[0630] b) The second monomer, comprising a heavy chain, includes:

[0631] i) Heavy chain variable structural domain; and

[0632] ii) Heavy-chain constant structural domains containing a second Fc structural domain; and

[0633] c) A light chain comprising a variable light chain structural domain and a variable light chain constant structural domain; wherein the variable light chain structural domain comprises v1CDR1 having the sequence KSSQSLLNTGNQKNYLT (SEQ ID NO: XX), v1CDR2 having the sequence WASTRES (SEQ ID NO: XX), and v1CDR3 having the sequence QNDYSYPYT (SEQ ID NO: XX), and the variable heavy chain structural domain comprises vhCDR1 having the sequence DYYMK (SEQ ID NO: XX), vhCDR2 having the sequence DIIPSNGATFYNQKFKG (SEQ ID NO: XX), and vhCDR3 having the sequence SHLLRASWFAY (SEQ ID NO: XX).

[0634] 66. A heterodimeric antibody selected from: XENP15049, XENP15051; XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP15635, XENP15636, XENP15638, XENP15639, XE NP13677, XENP14388, XENP14389, XENP14390, XENP14391, XENP14392, ENP16368, XENP16369, XENP16370, XENP16371, XENP16372, XENP16373, XENP16375, XENP16376, and

[0635] 67. A nucleic acid composition comprising three nucleic acids encoding a heterodimeric antibody selected from the following: XENP15049, XENP15051; XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP15635, XENP15636, XENP15638, XENP15639 ... P15639, XENP13677, XENP14388, XENP14389, XENP14390, XENP14391, 367, XENP16368, XENP16369, XENP16370, XENP16371, XENP16372, XENP16373,XENP16375,

[0636] 68. An expression vector composition comprising three expression vectors, each containing a nucleic acid such that the three expression vectors encode a heterodimeric antibody selected from the following: XENP15049, XENP15051; XENP15050, XENP13676, XENP14696, XENP15629, XENP15053, XENP15630, XENP15631, XENP15632, XENP15633, XENP15634, XENP15635, XENP15636. XENP15638, XENP15639, XENP13677, XENP14388, XENP14389, XENP14390, XENP16367, XENP16368, XENP16369, XENP16370, XENP16371, XENP16372,XENP16373,

[0637] 69. A host cell comprising the nucleic acid composition as described in Embodiment 67.

[0638] 70. A host cell comprising the expression vector composition of embodiment 68.

[0639] 71. A method for preparing a heterodimeric antibody as described in Embodiment 66, comprising culturing host cells as described in Embodiment 69 or 70 under conditions in which the antibody is expressed, and recovering the antibody.

[0640] 72. A method of treating cancer, comprising administering a heterodimeric antibody as described in embodiment 66 to a patient in need of such treatment.

Claims

1. An anti-CD3 single-chain Fv (scFv) polypeptide, wherein the scFv polypeptide comprises, from the N-terminus to the C-terminus, a variable light chain domain (VL), an scFv linker, and a variable heavy chain domain (VH), wherein, The variable light chain structural domain and the variable heavy chain structural domain are selected from: a) A variable light chain structural domain comprising vlCDR1 shown in SEQ ID NO:15, vlCDR2 shown in SEQ ID NO:16, and vlCDR3 shown in SEQ ID NO:17, and a variable heavy chain structural domain comprising vhCDR1 shown in SEQ ID NO:11, vhCDR2 shown in SEQ ID NO:12, and vhCDR3 shown in SEQ ID NO:13; b) A variable light chain structural domain comprising vlCDR1 shown in SEQ ID NO:24, vlCDR2 shown in SEQ ID NO:25 and vlCDR3 shown in SEQ ID NO:26, and a variable heavy chain structural domain comprising vhCDR1 shown in SEQ ID NO:20, vhCDR2 shown in SEQ ID NO:21 and vhCDR3 shown in SEQ ID NO:22; c) A variable light chain structural domain comprising vlCDR1 shown in SEQ ID NO:33, vlCDR2 shown in SEQ ID NO:34, and vlCDR3 shown in SEQ ID NO:35, and a variable heavy chain structural domain comprising vhCDR1 shown in SEQ ID NO:29, vhCDR2 shown in SEQ ID NO:30, and vhCDR3 shown in SEQ ID NO:31; d) A variable light chain domain comprising vlCDR1 shown in SEQ ID NO:42, vlCDR2 shown in SEQ ID NO:43, and vlCDR3 shown in SEQ ID NO:44, and a variable heavy chain domain comprising vhCDR1 shown in SEQ ID NO:38, vhCDR2 shown in SEQ ID NO:39, and vhCDR3 shown in SEQ ID NO:40; e) A variable light chain domain comprising vlCDR1 shown in SEQ ID NO:51, vlCDR2 shown in SEQ ID NO:52, and vlCDR3 shown in SEQ ID NO:53, and a variable heavy chain domain comprising vhCDR1 shown in SEQ ID NO:47, vhCDR2 shown in SEQ ID NO:48, and vhCDR3 shown in SEQ ID NO:49; and f) A variable light chain structural domain comprising vlCDR1 shown in SEQ ID NO:6, vlCDR2 shown in SEQ ID NO:7, and vlCDR3 shown in SEQ ID NO:8; and a variable heavy chain structural domain comprising vhCDR1 shown in SEQ ID NO:2, vhCDR2 shown in SEQ ID NO:3, and vhCDR3 shown in SEQ ID NO:

4.

2. The polypeptide of claim 1, wherein VH and VL are selected from: a) VL having the amino acid sequence SEQ ID NO:14 and VH having the amino acid sequence SEQ ID NO:10; b) VL having the amino acid sequence SEQ ID NO:23 and VH having the amino acid sequence SEQ ID NO:19; c) VL having the amino acid sequence SEQ ID NO:32 and VH having the amino acid sequence SEQ ID NO:28; d) VL having the amino acid sequence SEQ ID NO:41 and VH having the amino acid sequence SEQ ID NO:37; e) VL having the amino acid sequence SEQ ID NO:50 and VH having the amino acid sequence SEQ ID NO:46; and f) VL having the amino acid sequence SEQ ID NO:5 and VH having the amino acid sequence SEQ ID NO:

1.

3. The polypeptide of claim 1 or 2, wherein the scFv adapter comprises SEQ ID NO:

142.

4. A nucleic acid encoding a polypeptide according to any one of claims 1 to 3.

5. An expression vector comprising the nucleic acid of claim 4.

6. A host cell comprising the expression vector of claim 5.

7. A method for producing a polypeptide according to any one of claims 1 to 3, comprising culturing a host cell according to claim 6 under conditions in which the polypeptide is expressed, and harvesting the polypeptide.

Images