Multivalent and bispecific antibody constructs and methods of use thereof

Abstract

The present disclosure relates to multivalent, e.g., trivalent, bispecific antibody constructs capable of binding to an antigen on a cytotoxic effector cell and a tumor-associated antigen (TAA) on a tumor cell. Pharmaceutical compositions comprising such antibody constructs, as well as methods for preparing and using such constructs and compositions, e.g., for the treatment of cancer, are also disclosed.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to and benefit of U.S. Provisional Application No. 63 / 465,483, filed May 10, 2023, U.S. Provisional Application No. 63 / 458,621, filed April 11, 2023, and U.S. Provisional Application No. 63 / 393,633, filed July 29, 2022, the entire contents of each of which are incorporated herein by reference for all purposes.

[0002] Sequence Listing This application contains a Sequence Listing that has been submitted electronically in XML format and is incorporated herein by reference in its entirety. A copy of this XML file, created on XXXX date, is named XXXX and is XXXX bytes in size.

[0003] Technical Field The present disclosure relates to multivalent and bispecific antibody constructs capable of targeting antigens on the surface of cytotoxic effector cells and tumor-associated antigens (TAA) located on tumor cells, as well as methods for producing such constructs and methods of use, for example, to treat cancer. [Background technology]

[0004] background Despite significant advances in cancer treatment over the past few decades, cancer continues to pose significant unmet medical needs. + They can be targeted to treat tumors, particularly solid tumors, as well as mesothelioma, ovarian cancer, lung cancer, colon cancer, and pancreatic cancer, and triple-negative breast cancer. Current standard and more recently developed treatments for these indications have shown some clinical progress, validating MSLNs as promising targets for cancer therapy, but these treatment modalities have reported significant systemic side effects and limited efficacy, primarily due to a narrow therapeutic window. Summary of the Invention

[0005] overview In various embodiments, the present disclosure relates to multivalent and bispecific antibody constructs capable of binding to antigens on the surface of cytotoxic effector cells and tumor-associated antigens (TAA) located on tumor cells. In some embodiments, the present disclosure relates to trivalent, bispecific antibody constructs capable of binding to cluster of differentiation 3 (CD3) on immune cells (e.g., T cells) and MSLN on tumor cells.

[0006] In various embodiments, the present disclosure relates to antibody constructs comprising a dimeric Fc domain, the dimeric Fc domain comprising: (i) a Fab domain capable of binding to an antigen on a cytotoxic effector cell; (ii) a first scFv domain and a second scFv domain, both of which are capable of binding mesothelin (MSLN); and (iii) a first Fc polypeptide and a second Fc polypeptide, wherein a) the Fab domain is linked to the N-terminus of the first Fc polypeptide; b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide; and c) the second scFv domain is linked to the N-terminus of the Fab domain. In some embodiments, at least one of the first scFv domain and the second scFv domain comprises a V HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. LIn various embodiments, the first scFv domain and the second scFv domain each comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L Includes the domain.

[0007] The present disclosure provides, in various embodiments, a VH1 antibody comprising: (i) a Fab domain capable of binding to an antigen on a cytotoxic effector cell; (ii) a VH1 antibody capable of binding to MSLN, the VH1 antibody comprising an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. Land (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and c) the second scFv domain is linked to the N-terminus of the Fab domain. In various embodiments, the second TAA to which the second scFv domain binds is also MSLN. Thus, in certain embodiments, the second scFv domain also comprises a V HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L Includes the domain.

[0008] In various embodiments of the antibody constructs disclosed herein, the cytotoxic effector cell is a T cell. In such embodiments, the antigen to which the Fab domain binds can be CD3. In various embodiments, the Fab domain is capable of binding to CD3 and comprises a V HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 128. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L Includes the domain.

[0009] In various embodiments of the antibody constructs disclosed herein, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103. H domain, and V comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 101. L In certain embodiments, the first scFv domain and the second scFv domain comprise a linker scFv By V L Domain is V H Such linkers are linked to domains. scFv is the amino acid sequence (G4S) n wherein n is 1, 2, 3, 4, or 5. In certain embodiments, n is 3. In other embodiments, n is 4.

[0010] In some embodiments of the antibody constructs disclosed herein, the first scFv domain (which may also be referred to herein as "scFv1"), the second scFv domain (which may also be referred to herein as "scFv2"), or both scFv domains, are arranged in the N-terminal to C-terminal direction as follows: V H -Linker scFv -V L In some embodiments, the first scFv domain and the second scFv domain may have a domain structure of, from N-terminal to C-terminal, V H -Linker scFv -V L In such embodiments, the first scFv domain and the second scFv domain of the antibody construct may each comprise or consist of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 150.

[0011] In other embodiments of the antibody constructs disclosed herein, the first scFv domain, the second scFv domain, or both scFv domains comprise, from N-terminal to C-terminal, a V L -Linker scFv -V H In various embodiments, the first scFv domain and the second scFv domain may have a domain structure of, from the N-terminus to the C-terminus, V L -Linker scFv -V H In such embodiments, the first scFv domain and the second scFv domain may each comprise or consist of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 149.

[0012] In various embodiments of the antibody constructs disclosed herein, the Fab domain comprises or consists of an amino acid sequence having, from N- to C-terminus, at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 105. H domain, and a C domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 106. H1 The Fab domain may further comprise or consist of a heavy chain H or a portion thereof, including a V domain, having, from the N-terminal to the C-terminal direction, an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:115. L domain, and a C domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 116. L the light chain L comprises a V domain of the heavy chain H; H -C H1In some embodiments, the light chain L of the antibody construct herein comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0013] In various embodiments of the antibody constructs disclosed herein, the antibody construct comprises, from N-terminal to C-terminal, scFv2-V H -C H1 wherein the C-terminus of the second scFv domain comprises V H In some embodiments, the linker scFv-Fab The C-terminus of the second scFv domain is V H The scFv2-linker is linked to the N-terminus of the domain. scFv-Fab -V H -C H1 Such a linker can be used to obtain a heavy chain domain structure of scFv-Fab is the amino acid sequence (G4S) n and n is 1, 2, 3, 4, or 5. In various embodiments, the first heavy chain H1 further comprises a first Fc polypeptide and a C of the Fab domain. H1 The C-terminus of the domain is linked to the N-terminus of the first Fc polypeptide, thereby forming the following domain structure of H1: scFv2-linker scFv-Fab -V H -C H1 In certain embodiments, the linker Fab-Fc By C H1 The domain is linked to the CH2 domain of the first Fc polypeptide. Fab-Fc may comprise or consist of an immunoglobulin (Ig) hinge region. In some embodiments, the Ig hinge region is an IgG hinge region. Such an IgG hinge region may be an IgG1 hinge region. In some embodiments, the linker Fab-Fccomprises or consists of an amino acid sequence having at least about 70%, 80%, 90%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 107. Thus, in certain embodiments, H1 can comprise or consist of the following domain structure: scFv2-linker scFv-Fab -V H -C H1 -Linker Fab-Fc -CH2-CH3.

[0014] In various embodiments of the antibody constructs disclosed herein, the dimeric Fc domain is a heterodimeric Fc domain. Such a heterodimeric Fc domain comprises a first polypeptide and a second Fc polypeptide. Such first and second Fc polypeptides are interchangeable unless otherwise specified and may be linked to either a Fab or a first scFv domain, as further described herein.

[0015] In certain embodiments, one of the first or second Fc polypeptides of the dimeric Fc domain is an IgG1-derived Fc polypeptide and comprises a first CH2 domain (e.g., a CH21 domain) and a first CH3 domain (e.g., a CH31 domain). Such a first CH3 domain may comprise one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH3 domain sequence. In some embodiments, the one or more amino acid substitutions are a set of amino acid substitutions selected from L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V (numbering of amino acid residues in the first Fc polypeptide is according to the EU numbering system). In certain embodiments, the first CH2 domain comprises one or more amino acid substitutions that reduce or abolish binding to an Fc receptor (e.g., an Fcγ receptor) compared to the corresponding wild-type IgG1 CH2 domain.

[0016] In various embodiments herein, trivalent and bispecific antibody constructs comprise a first heavy chain H1 comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100. In certain embodiments, such an H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100.

[0017] In various embodiments of the antibody constructs disclosed herein, the antibody construct further comprises, from N- to C-terminal, a second heavy chain H2 comprising the first scFv domain and the other of the first or second Fc polypeptide linked thereto (thereby forming the domain structure scFv1-CH2-CH3). In certain embodiments, the linker scFv-Fc The C-terminus of the first scFv domain is linked to the N-terminus of the second Fc polypeptide, e.g., the N-terminus of the second CH2 domain, via such a linker scFv-Fc may comprise or consist of an Ig hinge region. In some embodiments, the Ig hinge region is an IgG hinge region. Such an IgG hinge region may be an IgG1 hinge region. Thus, in some embodiments, the linker scFv-Fc comprises or consists of an amino acid sequence having at least about 70%, 80%, 90%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 111. Thus, in some embodiments, H2 can comprise or consist of the following domain structure: scFv1-linker scFv-Fc -CH2-CH3.

[0018] In various embodiments of the antibody constructs disclosed herein, the other of the first or second Fc polypeptide is also an IgG1-derived Fc polypeptide and comprises a second CH2 domain (e.g., a CH22 domain) and a second CH3 domain (e.g., a CH32 domain). In some embodiments, the second CH3 domain comprises one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH3 domain sequence. In some embodiments, the one or more amino acid substitutions are selected from T366L_K392M_T394W, T366L_K392L_T394W, T350V_T366L_K392L-_T394W, T350V_T366L_K392M_T394W, and T350V_T366L_N390R_K392M_T394W (numbering of amino acid residues in the second Fc polypeptide is according to the EU numbering system). Thus, in various embodiments of antibody constructs comprising heterodimeric Fc domains, one or more amino acid substitutions in the first CH3 domain and one or more amino acid substitutions in the second CH3 domain promote preferential pairing of heavy chains H1 and H2 and formation of heterodimeric Fc domains H1-H2 relative to the corresponding homodimeric Fc domains (e.g., dimers consisting of H1-H1 or H2-H2).Like the first CH2 domain of the first Fc polypeptide, the second CH2 domain may also contain one or more amino acid substitutions that reduce or abolish binding to Fc receptors (e.g., Fcγ receptors) relative to the corresponding wild-type IgG1 CH2 domain.

[0019] In various embodiments herein, the trivalent and bispecific antibody constructs comprise a second heavy chain, H2, comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 110. In certain embodiments, H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 110.

[0020] In various embodiments of the present disclosure, the antibody construct is bispecific for CD3 and MSLN and trivalent, i.e., monovalent for CD3 via the Fab domain and bivalent for MSLN via the first scFv domain and the second scFv domain.

[0021] In various embodiments of the antibody construct disclosed herein, the antibody construct comprises a first heavy chain (H1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100, a second heavy chain (H2) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110, and a light chain (L1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0022] In some embodiments, the antibody construct of the present disclosure comprises (i) a first heavy chain (H1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 100, (ii) a second heavy chain (H2) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 110, and (iii) a light chain (L1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 114.

[0023] In some embodiments, the antibody construct of the present disclosure comprises a first heavy chain (H1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 171, a second heavy chain (H2) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 172, and a light chain (L1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 114.

[0024] In other embodiments of the antibody constructs disclosed herein, the antibody construct comprises a first heavy chain (H1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 117, a second heavy chain (H2) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 119, and a light chain (L1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0025] In various embodiments of the trivalent and bispecific antibody constructs disclosed herein, such antibody constructs, when present at a concentration of about 10 picomoles per liter (pM) for about 72 hours in a cell population comprising immune cells expressing CD3 and tumor cells expressing about 500,000 MSLN per cell, are capable of one or more, two or more, or all of: A) inducing production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL of TNFα; B) inducing production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL of IFNγ; and / or C) inducing production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL of IL-2.

[0026] In various embodiments of the trivalent bispecific antibody constructs disclosed herein, such antibody constructs are capable of inducing at least 10-fold greater T cell-mediated cytotoxicity against MSLN-expressing tumor cells compared to a similar antibody construct comprising two anti-MSLN Fabs instead of the first and second anti-MSLN scFv domains.

[0027] In various embodiments of the trivalent and bispecific antibody constructs disclosed herein, such antibody constructs have a binding affinity for MSLN of about 0.7 nM, 0.8 nM, 0.9 nM, or about 1 nM as measured by surface plasmon resonance (SPR).

[0028] In certain embodiments of the trivalent and bispecific antibody constructs disclosed herein, such antibody constructs have a binding affinity for CD3 of about 30 nM, 40 nM, 50 nM, or about 60 nM as measured by SPR.

[0029] In certain embodiments of the trivalent bispecific antibody constructs disclosed herein, such antibody constructs are capable of inducing the production of inflammatory cytokines by cytotoxic effector cells in an MSLN-dependent manner, where the MSLN-dependent activation of cytotoxic effector cells is determined by measuring at least about a 20-fold, 50-fold, 100-fold, or at least about a 1000-fold decrease in cytokine production between a first cell population comprising first tumor cells and immune cells and a second cell population comprising second tumor cells and immune cells, where MSLN expression in the first tumor cells is at least about 3-fold, 4-fold, or 5-fold higher compared to MSLN expression in the second tumor cells.

[0030] In certain embodiments of the trivalent bispecific antibody constructs disclosed herein, when such antibody constructs are administered to a mammalian subject, the antibody constructs can reduce the volume of an MSLN-expressing tumor in the subject by at least about 5% for at least 20 days after administration to the subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg, when administered four times per week. In some embodiments, such reduction in tumor volume is at least about 7%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 50%, 60%, or at least about 70%. In some embodiments, the mammalian subject is a rodent or a non-human primate.

[0031] In some embodiments herein, the trivalent and bispecific antibody constructs of the present disclosure may have a stability of at least about 97%, 98%, or 99%, i.e., at least about 97%, 98%, or 99% of the intact antibody construct is measured using SEC after incubating the antibody construct in an aqueous solution or buffer system containing sucrose at 40° C. for 14 days.

[0032] In some embodiments, the present disclosure relates to pharmaceutical compositions comprising the trivalent and bispecific antibody constructs of the present disclosure and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.

[0033] Certain embodiments of the present disclosure relate to nucleic acid molecules or sets of nucleic acid molecules encoding one or more polypeptide chains of the antibody constructs described herein. In some embodiments, described herein are nucleic acid molecules or sets of nucleic acid molecules encoding one or more polypeptide chains of the bispecific and trivalent antibody constructs described herein, such as construct v32523 or v21812. In some embodiments, described herein are nucleic acid molecules or sets of nucleic acid molecules encoding the three polypeptide chains of bispecific and trivalent antibody construct v32523, the amino acid sequences of which are set forth in SEQ ID NOs: 100, 110, and 114, or in SEQ ID NOs: 171, 172, and 114, as further described herein.

[0034] Further described herein are vectors or sets of vectors that comprise a nucleic acid molecule or set of nucleic acid molecules that encode one or more polypeptide chains of the antibody constructs described herein.

[0035] Further described herein is a nucleic acid molecule or set of nucleic acid molecules that encode one or more polypeptide chains of the antibody constructs described herein, and / or a nucleic acid molecule or set of nucleic acid molecules that encode one or more polypeptide chains of the antibody constructs described herein. In some embodiments, the host cell may be a mammalian host cell.

[0036] Certain embodiments of the present disclosure relate to methods for producing antibody constructs described herein. In some embodiments, such methods relate to the production of trivalent, bispecific antibody constructs comprising two anti-MSLN scFv domains and one anti-CD3 Fab domain, as further described herein. In some embodiments, methods for producing such trivalent and bispecific antibody constructs include (a) obtaining a host cell culture comprising at least one host cell comprising one or more nucleic acid molecules encoding one or more polypeptide chains of the antibody construct, and (b) recovering the antibody construct from the host cell culture. In some embodiments, such methods may further include purifying the antibody construct, for example, using the chromatographic purification methods described herein.

[0037] In some embodiments, the amino acid sequence of one or more heavy chains of the antibody constructs herein may include an additional C-terminal lysine residue for production in a host cell. Thus, in some embodiments, the amino acid sequences set forth in SEQ ID NOs: 100 and 110 may further include a C-terminal lysine ("K") residue. Exemplary amino acid sequences of the two heavy chains H1 and H2 of antibody construct v32523, which include a C-terminal lysine residue, are set forth in SEQ ID NOs: 171 and 172, respectively. Such a C-terminal lysine residue may be enzymatically cleaved from one or more heavy chains after expression in a host cell. In some embodiments, a trivalent and bispecific antibody construct comprising the heavy chain amino acid sequences set forth in SEQ ID NOs: 171 and 172 and the light chain amino acid sequence set forth in SEQ ID NO: 114 may be referred to as v38490.

[0038] Further embodiments of the present disclosure relate to methods of eliciting an anti-tumor immune response in a cell population comprising immune cells expressing CD3 and tumor cells expressing MSLN, the method comprising contacting the cell population with an effective amount of a trivalent and bispecific antibody construct of the present disclosure. In some embodiments, the concentration of the antibody construct in the cell population is about 10 -2 Picomole (pM) to approximately 10 2When the dose is in the pM range (e.g., increasing), the anti-tumor immune response reduces viable tumor cells in the cell population by at least about 30%, 40%, 50%, or 60% in a dose-dependent manner, wherein the tumor cells express MSLNs at least about 15,000 MSLNs per cell, and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0039] Certain embodiments of the present disclosure relate to methods of inhibiting the proliferation of tumor cells expressing MSLN, comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of a trivalent or bispecific antibody construct of the present disclosure. In certain embodiments, the effective amount is at most about 10 -2 pM, 10 -1 Using an antibody construct concentration of 0.5 pM, or 1 pM, tumor cell proliferation is inhibited if an increase of about 5% in the count of viable tumor cells in the cell population is observed over at least about 5, 10, 20, or 48 hours, where the tumor cells express at least about 15,000 MSLNs per cell and the ratio of tumor cells to immune cells in the cell population is about 5:1. In some embodiments, tumor cell proliferation is inhibited if a decrease of at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, or at least about 50% in the count of viable tumor cells is measured over a given period of time.

[0040] Other embodiments of the present disclosure relate to methods for killing tumor cells expressing MSLN, comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of a trivalent or bispecific antibody construct of the present disclosure. In such embodiments, the concentration of the antibody construct is about 10 -2 pM~about 10 2 In the pM range (e.g., increasing), tumor cell killing is observed when a dose-dependent reduction of at least about 30%, 40%, 50%, or 60% of viable tumor cells in the cell population is measured, where the tumor cells express MSLNs at least about 15,000 MSLNs per cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0041] In certain embodiments of these methods, the immune cells comprise or consist of T cells. Such T cell populations may include one or more of the different types of T cells known in the art.

[0042] In certain embodiments of the methods described herein, trivalent and bispecific antibody constructs bind to CD3 on immune cells and MSLN on tumor cells. In various embodiments, such binding of both antigens on different cells can include simultaneous binding of CD3 and MSLN, which in some embodiments can establish a TCR-independent immune synapse capable of inducing anti-tumor cytotoxic activity of T cells against tumor cells.

[0043] In some embodiments, the present disclosure relates to a method for inhibiting the growth of an MSLN-expressing tumor or reducing the volume of such a tumor in a subject, the method comprising administering to the subject an effective amount of a trivalent bispecific antibody construct of the present disclosure to an MSLN-expressing tumor. In various embodiments, the antibody construct binds to CD3 on the subject's immune cells and to MSLN on the tumor cells, thereby inducing an anti-tumor immune response in the subject, thereby inhibiting tumor growth or reducing tumor volume in the subject. In some embodiments, the inhibition of tumor growth and / or reduction in tumor volume is induced, at least in part, by binding of the antibody construct to CD3 on the immune cells and MSLN on the tumor cells and formation of a TCR-independent artificial immune synapse within the tumor environment of the subject. As described herein, the immune cell can be a T cell. In various embodiments, and as further described herein, activation of the immune cell (e.g., a T cell) can induce the release of one or more cytokines, e.g., TNFα, IFNγ, and / or IL-2, from the activated immune cell. In various embodiments, tumor growth in a subject is inhibited for at least about 20 days, 30 days, or at least about 50 days using four weekly (Q7Dx4) administrations of an antibody construct to the subject (e.g., tumor growth of 5% or less over the course of the study). In certain embodiments, tumor volume is reduced by at least about 10%, 20%, 30%, 40%, or at least about 50% about 15 days after initiation of treatment. In such embodiments, about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg of the antibody construct is administered to the subject. The subject in the methods described herein may be a rodent, a non-human primate, or a human. In some embodiments, the antibody construct is administered intravenously.

[0044] In some embodiments, the present disclosure relates to a trivalent bispecific antibody construct according to any of the embodiments described herein for use in the treatment of cancer.

[0045] In some embodiments, the present disclosure relates to the use of an antibody construct according to any of the embodiments described herein in the manufacture of a medicament for treating cancer.

[0046] In some of these embodiments, the cancer is an MSLN-expressing cancer. [Brief explanation of the drawings]

[0047] Embodiments disclosed herein are illustrated by way of example, and not by way of limitation, in the accompanying drawings, which are merely illustrative and aid in understanding and are not intended as a definition of the limits of the antibody constructs, pharmaceutical compositions, and methods of the present disclosure.

[0048] [Figure 1]

[0023] Figures 1A and 1B show schematic diagrams of the geometry (e.g., relative orientation, relative spatial localization, and / or connectivity of binding domains, Fc domains, etc.) and format (e.g., antigen valency, type of binding domain (e.g., scFv, Fab, etc.), presence / absence / type of Fc domain, etc.) of antibody constructs according to certain embodiments of the present disclosure. Figure 1A shows a schematic diagram of the geometry and format of a trivalent bispecific antibody construct of the present disclosure capable of monovalent binding to CD3 (via one Fab domain) and bivalent binding to MSLN (via two scFv domains) according to various embodiments of the present disclosure (e.g., variants v21812, v32523, etc. described herein). Figure 1B shows a schematic representation of the geometry and format of a trivalent bispecific antibody construct capable of monovalent binding to CD3 (via one Fab domain) and bivalent binding to MSLN (via two Fab domains), for example, as used in the 2+1 Fab3 TCB benchmark construct v29191 or the Fab3 reference construct v21791. Figure 1C shows a schematic representation of the geometry and format of a bivalent bispecific antibody construct capable of monovalent binding to CD3 (e.g., via one scFv domain) and monovalent binding to MSLN (e.g., via one Fab domain), for example, as used in construct v21815 herein. FIG. 1D shows a schematic representation of the geometry and format of the trivalent trispecific MH6T-TriTAC benchmark polypeptide construct v31805, capable of monovalent binding to CD3 (e.g., via one scFv domain), monovalent binding to MSLN (e.g., via one VHH domain), and monovalent binding to human serum albumin (HSA or Alb, e.g., via one VHH domain). [Figure 2] Figure 2 shows the binding of the trivalent bispecific anti-(MSLN(scFv2) x CD3(Fab)) antibody construct v21812 to CD3 on CD3+ Jurkat cells as assessed by ELISA and compared to the bivalent bispecific construct v21815 (Figure 2A) and the trivalent bispecific anti-(MSLN x CD3) Fab3 antibody construct v21791 (Figure 2B). [Figure 3]Figure 3 shows the binding of the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812, the bivalent bispecific construct v21815, and the trivalent bispecific anti-(MSLN x CD3) Fab3 antibody construct v21791 to CD3 on CD4+ (Figure 3A) and CD8+ (Figure 3B) T cells as assessed by flow cytometry. [Figure 4] Figure 4A shows the binding of the bivalent, bispecific construct v21815 to MSLN on MSLN+H226 cells as measured by flow cytometry, with a measured K value of approximately 1.73 nM, and Figure 4B shows the binding of the trivalent, bispecific anti-(MSLN x CD3) antibody construct v21812 (i.e., comprising two anti-MSLN scFvs plus one anti-CD3 Fab) and Fab3 antibody construct v21791 to MSLN on MSLN+H226 cells as measured by flow cytometry, with measured K values of approximately 0.47 nM (v21812) and 0.74 nM (v21791), respectively. [Figure 5A] Figure 5A shows that the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812 and constructs v21815 and v21791 induced healthy donor-derived T cells to kill MSLN+ target tumor cells expressing different levels of MSLN, i.e., MCF-7 cells (approximately 500 MSLN per cell, Figure 5A). [Figure 5B] The trivalent, bispecific anti-(MSLN x CD3) antibody construct v21812, as well as constructs v21815 and v21791, induced healthy donor-derived T cells to kill MSLN+ target tumor cells expressing different levels of MSLN, i.e., BxPC3 (approximately 17,000 MSLN per cell, Figure 5B). These data show that construct v21812 exhibited the most potent killing of moderately and highly MSLN-expressing tumor cells, as shown in Figures 5B and 5C for BxPC3 and OVCAR-3 cells, respectively. [Figure 5C]The trivalent bispecific anti-(MSLN x CD3) antibody construct v21812 and constructs v21815 and v21791 induced healthy donor-derived T cells to kill MSLN+ target tumor cells expressing different levels of MSLN, i.e., OVCAR-3 (approximately 768,000 MSLN per cell, Figure 5C). These data show that construct v21812 exhibited the most potent killing of moderately and highly MSLN-expressing tumor cells, as shown in Figures 5B and 5C for BxPC3 and OVCAR-3 cells, respectively. [Figure 5D] These results show that the trivalent bispecific construct v21812 induced significantly higher IL-2 release from T cells when compared to both the triple Fab (Fab3, anti-CD3 Fab x anti-MSLN Fab2) construct v21791 and the 1 + 1 anti-CD3 scFv x anti-MSLN Fab construct v21815. Cytokine release was assessed by co-culturing OVCAR3 tumor cells with human pan T cells and treating the cell populations with each test article for 3 days. [Figure 6A] We show that the trivalent bispecific antibody construct v21812 suppressed tumor growth and further reduced tumor volume over the course of the experiment in OVCAR-3 tumor-bearing NOG mice engrafted with human PBMCs. The significant difference in tumor suppression exhibited by the antibody construct v21812 ((MSLN)-scFv2 x (CD3)-Fab, see Figure 1A) with a specially designed format and geometry compared to the corresponding trivalent bispecific Fab3 construct v21791 (see, e.g., Figure 1B), which contained identical anti-CD3 and anti-MSLN paratope sequences, was unexpected and surprising, but also demonstrates the superior in vivo performance of the scFv2 x Fab construct described herein compared to conventional constructs, as well as the impact that construct format (e.g., use of an anti-MSLN scFv domain versus a Fab domain) can have on construct properties. [Figure 6B]The results of a statistical analysis of experimental data obtained from in vivo studies are shown, further demonstrating that construct v21812 resulted in significantly lower tumor growth constants compared to Fab3 construct v21791, which contains the same anti-CD3 and anti-MSLN paratopes as v21812. In this study, a vehicle control was used as a negative control. [Figure 7A] The tumor cell-killing activity of the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812, as well as other MSLN-targeting constructs v29045 and v29048 bearing low-affinity anti-MSLN paratopes, is shown in three different cell lines with varying levels of MSLN expression and surface presentation: OVCAR-3 (Figure 7A, approximately 700,000 MSLNs per cell), H292 (Figure 7B, approximately 170,000 MSLNs per cell), and IGROV-1 (Figure 7C) (used as an MSLN-negative control cell line). The figure shows that the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812 exhibited the highest cell killing activity against OVCAR-3 cells, which highly express MSLN, in pan-T cell co-cultures compared to (i) two variants with low-affinity MSLN paratopes, v29045 and v29048, (ii) a trivalent bispecific Fab3 benchmark construct (2+1 Fab3TCB benchmark, v29191), and (iii) a negative control (v31926, an anti-CD3 / anti-HA construct). [Figure 7B]The tumor cell-killing activity of the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812, as well as other MSLN-targeting constructs v29045 and v29048 bearing low-affinity anti-MSLN paratopes, is shown in three different cell lines with varying levels of MSLN expression and surface presentation: OVCAR-3 (Figure 7A, approximately 700,000 MSLNs per cell), H292 (Figure 7B, approximately 170,000 MSLNs per cell), and IGROV-1 (Figure 7C) (used as an MSLN-negative control cell line). The figure shows that the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812 exhibited the highest cell killing activity against moderately MSLN-expressing H292 cells in pan-T cell co-cultures, similar to that of the anti-(MSLN x CD3) benchmark antibody construct MH6T-TriTAC (v31805, see e.g., Figure 1D for format), and compared to (i) two variants with low-affinity anti-MSLN paratopes, v29045 and v29048, (ii) the trivalent bispecific Fab3 benchmark construct (v29191), and (iii) the negative control (v31926). [Figure 7C] The figures show the tumor cell-killing activity of the trivalent bispecific anti-(MSLN x CD3) antibody construct v21812, as well as other MSLN-targeting constructs v29045 and v29048 bearing low-affinity anti-MSLN paratopes, in three different cell lines with varying levels of MSLN expression and surface presentation: OVCAR-3 (Figure 7A, approximately 700,000 MSLNs per cell), H292 (Figure 7B, approximately 170,000 MSLNs per cell), and IGROV-1 (Figure 7C) (used as an MSLN-negative control cell line). The figures show that all tested constructs did not significantly affect viable cell counts when tested in cocultures of the MSLN-negative tumor cell line IGROV-1 / pan-T cells. [Figure 8A]In particular, we show that the trivalent bispecific antibody construct v21812, and to a lesser extent the constructs v29045 and v29048 with low affinity anti-MSLN paratopes, induced production of the proinflammatory cytokine tumor necrosis factor alpha (TNFα) from T cells co-cultured with MSLN+OVCAR-3 cells for 48 hours (5:1 effector cell (E):tumor cell (T) ratio), and further when compared to the 2+1 Fab3 TCB benchmark construct v29191 (identical format but different paratope sequence compared to the Fab3 construct v21791). [Figure 8B] In particular, we show that the trivalent bispecific antibody construct v21812, and to a lesser extent the constructs v29045 and v29048 bearing low-affinity anti-MSLN paratopes, induced production of the proinflammatory cytokine interferon-gamma (IFNγ or IFNg) from T cells co-cultured with MSLN+OVCAR-3 cells for 48 hours (5:1 effector cell (E):tumor cell (T) ratio), and further when compared to the 2+1 Fab3 TCB benchmark construct v29191 (identical format but different paratope sequence compared to the Fab3 construct v21791). [Figure 8C] In particular, we show that the trivalent bispecific antibody construct v21812, and to a lesser extent the constructs v29045 and v29048 bearing low-affinity anti-MSLN paratopes, induced production of the pro-inflammatory cytokine interleukin-2 (IL-2) from T cells co-cultured with MSLN+OVCAR-3 cells for 48 hours (5:1 effector cell (E):tumor cell (T) ratio), and further when compared to the 2+1 Fab3 TCB benchmark construct v29191 (identical format but different paratope sequence compared to the Fab3 construct v21791). [Figure 9]This figure shows that the trivalent, bispecific anti-(MSLN x CD3) antibody construct v21812 of the present disclosure significantly suppressed tumor growth and reduced tumor volume over the course of the study in OVCAR-3 tumor-bearing NOG mice engrafted with human PBMCs. Construct v21812 suppressed tumor growth significantly more than other constructs tested, including the 2+1 Fab3TCB benchmark construct (v29191; see Figure 1B for construct format). ** indicates a p-value of <0.01, and *** indicates a p-value of <0.001. [Figure 10] Figure 1 shows the binding of the stability-improved (compared to v21812) trivalent bispecific antibody construct v32523 (identical domain sequence to v21812 except that both anti-MSLN scFv domains have a VL-linker-scFv-VH domain structure in the N- to C-terminal direction compared to the VH-linker-scFv-VL of v21812, and the linker-scFv has a (G4S)4 sequence in v32523 compared to the (G4S)3 sequence in v21812) to MSLN+OVCAR-3 cells as measured by flow cytometry, as well as compared to the benchmark control constructs v29191 and v31805. [Figure 11A] Figure 1 shows the binding of the trivalent bispecific antibody construct v32523, as well as the binding of the benchmark control constructs v29191 and v31805, to CD3+ human pan T cells in cell cultures containing CD4+ cells, as measured by flow cytometry. [Figure 11B] Figure 1 shows the binding of the trivalent bispecific antibody construct v32523, as well as the binding of the benchmark control constructs v29191 and v31805, to CD3+ human pan T cells in cell cultures containing CD8+ cells, as measured by flow cytometry. [Figure 11C] Figure 1 shows the binding of the trivalent bispecific antibody construct v32523, as well as the binding of the benchmark control constructs v29191 and v31805, to CD3+ human pan-T cells in cell cultures containing a mixture of CD4+ and CD8+ cells, as measured by flow cytometry. [Figure 12-1] Figure 1 shows MSLN-dependent T cell-mediated cytotoxicity induced by the trivalent bispecific antibody construct v32523 using cell lines with varying degrees of MSLN expression, compared to the benchmark construct v31805. The figure shows tumor cell killing induced by the antibody construct v32523 in MSLNhigh-expressing H292 cells compared to MSLNLlow-expressing OVTOKO and MCF7 cells. [Figure 12-2] Figure 1 shows MSLN-dependent T cell-mediated cytotoxicity induced by the trivalent bispecific antibody construct v32523 compared to the benchmark construct v31805 using cell lines with varying degrees of MSLN expression. The figure shows a direct comparison of tumor cell killing rates of the two tested constructs, v32523 and v31805, in MSLNhjgh-expressing H292 cells and MSLNLow-expressing A375 cells (*test article data shown as dashed squares). [Figure 12-3] Figure 1 shows MSLN-dependent T cell-mediated cytotoxicity induced by the trivalent bispecific antibody construct v32523 compared to the benchmark construct v31805 using cell lines with varying degrees of MSLN expression. (C) T cell-mediated cytotoxicity of the trivalent bispecific construct v32523* was also tested in the presence of H292 lung cancer cells (approximately 170,000 MSLNs per cell) and compared to a negative control. (D) T cell-mediated cytotoxicity of the trivalent bispecific construct v32523* was also tested in the presence of OVCAR8 ovarian cancer cells (approximately 85,000 MSLNs per cell) and compared to a negative control. [Figure 12-4] Figure 1 shows MSLN-dependent T cell-mediated cytotoxicity induced by the trivalent bispecific antibody construct v32523 compared to the benchmark construct v31805 using cell lines with varying degrees of MSLN expression. (E) T cell-mediated cytotoxicity of the trivalent bispecific construct v32523* was also tested in the presence of HCT-116 colon cancer cells (approximately 35,000 MSLNs per cell) and compared to a negative control. (F) T cell-mediated cytotoxicity of the trivalent bispecific construct v32523* was also tested in the presence of H2452 mesothelioma cells (approximately 11,000 MSLNs per cell) and compared to a negative control. [Figure 13A] Figure 1 shows that the trivalent bispecific antibody construct v32523 enabled healthy donor-derived T cells to kill MSLN+ tumor cells expressing different levels of MSLN, i.e., H292 (approximately 152,000 MSLN per cell). In all four tumor cell lines tested that exhibited moderate to high MSLN expression, the trivalent bispecific antibody construct v32523 of the present disclosure was at least as potent as, or even superior to, the benchmark control constructs v29191 (a 2+1 Fab3 construct) and v31805 (MH6T-TriTAC). [Figure 13B] These results show that the trivalent bispecific antibody construct v32523 enabled healthy donor-derived T cells to kill MSLN+ tumor cells expressing different levels of MSLN, i.e., HCT116 (approximately 35,000 MSLN per cell). In all four tumor cell lines tested that exhibited moderate to high MSLN expression, the trivalent bispecific antibody construct v32523 of the present disclosure was at least as potent as, or even superior to, the benchmark control constructs v29191 (a 2+1 Fab3 construct) and v31805 (MH6T-TriTAC). [Figure 13C] These results show that the trivalent bispecific antibody construct v32523 enabled healthy donor-derived T cells to kill MSLN+ tumor cells expressing different levels of MSLN, i.e., H2452 (approximately 11,000 MSLN per cell). In all four tumor cell lines tested that exhibited moderate to high MSLN expression, the trivalent bispecific antibody construct v32523 of the present disclosure was at least as potent as, or even superior to, the benchmark control constructs v29191 (a 2+1 Fab3 construct) and v31805 (MH6T-TriTAC). [Figure 13D]Figure 1 shows that the trivalent bispecific antibody construct v32523 enabled healthy donor-derived T cells to kill MSLN+ tumor cells expressing different levels of MSLN, i.e., A549 (approximately 11,000 MSLN per cell). In all four tumor cell lines tested that exhibited moderate to high MSLN expression, the trivalent bispecific antibody construct v32523 of the present disclosure was at least as potent as, or even superior to, the benchmark control constructs v29191 (a 2+1 Fab3 construct) and v31805 (MH6T-TriTAC). [Figure 14A] These data show that the trivalent bispecific antibody constructs v32523 and v31805 (MH6T-TriTAC), a benchmark control construct, induced TNFα production in T cells co-incubated with MSLN+H292 cells and MSLN+HCT116 cells (E:T ratio 5:1) for 72 hours, respectively. These data demonstrate that the trivalent bispecific antibody construct v32523 induced only minimal levels of both cytokines in the MSLN-low expressing cell line, HCT116, demonstrating the ability of the antibody constructs described herein to provide TAA (e.g., MSLN)-dependent T cell-mediated cytotoxicity. Such antigen-dependent properties may enable potent antitumor immune responses in an environment with high MSLN expression, while off-target tissues with no or low MSLN expression may not be exposed to the effects of T cell-mediated cytotoxicity, or at least only to a very low extent. [Figure 14B]The trivalent bispecific antibody constructs v32523 and v31805 (MH6T-TriTAC) benchmark control construct induced IFNγ production (Figures 14B and 14D) in T cells co-incubated with MSLN+H292 and MSLN+HCT116 cells (E:T ratio 5:1) for 72 hours, respectively. These data show that the trivalent bispecific antibody construct v32523 induced only minimal levels of both cytokines in the MSLN-low expressing cell line HCT116, demonstrating the ability of the antibody constructs described herein to provide TAA (e.g., MSLN)-dependent T cell-mediated cytotoxicity. Such antigen-dependent properties may enable potent antitumor immune responses in an environment with high MSLN expression, while off-target tissues with no or low MSLN expression may not be exposed to the effects of T cell-mediated cytotoxicity, or at least only to a very low extent. [Figure 14C] These data show that the trivalent bispecific antibody constructs v32523 and v31805 (MH6T-TriTAC), a benchmark control construct, induced TNFα production in T cells co-incubated with MSLN+H292 cells and MSLN+HCT116 cells (E:T ratio 5:1) for 72 hours, respectively. These data demonstrate that the trivalent bispecific antibody construct v32523 induced only minimal levels of both cytokines in the MSLN-low expressing cell line, HCT116, demonstrating the ability of the antibody constructs described herein to provide TAA (e.g., MSLN)-dependent T cell-mediated cytotoxicity. Such antigen-dependent properties may enable potent antitumor immune responses in an environment with high MSLN expression, while off-target tissues with no or low MSLN expression may not be exposed to the effects of T cell-mediated cytotoxicity, or at least only to a very low extent. [Figure 14D]These data show that the trivalent bispecific antibody constructs v32523 and v31805 (MH6T-TriTAC), a benchmark control construct, induced IFNγ production in T cells co-incubated with MSLN+H292 cells and MSLN+HCT116 cells (E:T ratio 5:1) for 72 hours, respectively. These data demonstrate that the trivalent bispecific antibody construct v32523 induced only minimal levels of both cytokines in the MSLN-low expressing cell line, HCT116, demonstrating the ability of the antibody constructs described herein to provide TAA (e.g., MSLN)-dependent T cell-mediated cytotoxicity. Such antigen-dependent properties may enable potent antitumor immune responses in an environment with high MSLN expression, while off-target tissues with no or low MSLN expression may not be exposed to the effects of T cell-mediated cytotoxicity, or at least only to a very low extent. [Figure 15] Figure 1 shows that when co-incubated with MSLN+OVCAR-3 cells (E:T ratio 10:1) for 72 hours, the trivalent bispecific antibody construct v32523 and the v31805 (MH6T-TriTAC) benchmark control construct induced T cell proliferation in a comparable and dose-dependent manner. [Figure 16A] 1 shows that the trivalent bispecific antibody construct v32523 of the present disclosure significantly suppressed tumor growth and reduced tumor volume at doses as low as 1 mg / kg or 3 mg / kg in OVCAR-3 tumor-bearing NOG mice (highly MSLN-expressing) engrafted with human PBMCs (arrows indicate the time points of administration). [Figure 16B] 1 shows that the trivalent bispecific antibody construct v32523 of the present disclosure significantly inhibited tumor growth and reduced tumor volume at doses as low as 1 mg / kg or 3 mg / kg in CT116 tumor-bearing NOG mice (moderate MSLN expression) engrafted with human PBMCs (arrows indicate administration time points). [Figure 17A] Figure 1 shows the pharmacokinetic profile and serum concentrations at various time points after injection with various doses of the trivalent bispecific antibody construct v32523 in OVCAR-3 tumor-bearing mice (at doses of 0.1, 1, and 3 mg / kg). [Figure 17B] Figure 1 shows the pharmacokinetic profile and serum concentrations at various time points after injection with various doses of the trivalent bispecific antibody construct v32523 in naive mice (at doses of 0.1, 0.5, 1, and 3 mg / kg). [Figure 17C] Figure 1 shows the pharmacokinetic profiles and serum concentrations at various time points following injection of various doses (i.e., 0.1, 1, and 3 mg / kg) of the trivalent bispecific antibody construct v32523 by PBMC donors A, B, and C. The lower limit of quantitation (LLOQ) was determined to be approximately 781 ng / mL. [Figure 18] This shows that the trivalent bispecific antibody construct v32523 of the present disclosure exhibited significantly greater antitumor activity at both 1 mg / kg and 3 mg / kg doses (intravenously) when compared to the benchmark control v31805 (MH6T-TriTAC) at various equivalent doses in HCT116 tumor-bearing mice engrafted with human PBMCs. Construct v31805 was dose-matched to v32523 by PK analysis to account for the different in vivo half-lives and allow for a head-to-head comparison of the two constructs. [Figure 19] Figure 1 shows that the trivalent bispecific antibody construct v32523 of the present disclosure exhibited significantly greater anti-tumor activity (intravenously, 3 mg / kg) at comparable dose levels compared to benchmark controls v31805 (MH6T-TriTAC) and v29191 (2+1 Fab3) in HCT116 tumor-bearing mice engrafted with human PBMCs. [Figure 20] 1 shows that the trivalent bispecific antibody construct v32523 of the present disclosure exhibited significantly greater anti-tumor activity (intravenous, 3 mg / kg) as measured by a significant reduction in tumor volume in OVCAR-3 tumor-bearing mice engrafted with human PBMCs, when compared to an equivalent dose of the benchmark control construct v31805. [Figure 21A]We show that the trivalent bispecific antibody construct v32523 induced MSLN-dependent cytokine production (IL-2) in either (i) the presence of T cells alone or (ii) the T cells were cocultured with MSLN+ tumor cells. Cytokine production was assessed by coculture of MSLN-expressing tumor cell lines with human pan-T cells and treating the cells with v32523 or a negative control for 3 days. [Figure 21B] We show that the trivalent bispecific antibody construct v32523 induced MSLN-dependent cytokine production (TNFα) in either (i) the presence of T cells alone or (ii) the T cells were cocultured with MSLN+ tumor cells. Cytokine production was assessed by coculture of MSLN-expressing tumor cell lines with human pan-T cells and treating the cells with v32523 or a negative control for 3 days. [Figure 21C] We show that the trivalent bispecific antibody construct v32523 induced T cell proliferation in either (i) the presence of T cells alone or (ii) the T cells co-cultured with MSLN+ tumor cells. T cell proliferation was assessed by co-culturing CFSE-labeled T cells with OVCAR3 cells and treating them with v32523 for 5 days. Proliferation was measured by flow cytometry. [Figure 22A] FIG. 1 shows that the trivalent bispecific construct v32523 exhibited higher antitumor activity when compared to the clinical benchmark construct v31805 and when tested in a high E:T ratio (5:1) environment (the same figure legend was used to identify the test conditions). [Figure 22B] FIG. 1 shows that the trivalent bispecific construct v32523 exhibited higher antitumor activity when compared to the clinical benchmark construct v31805 and when tested in a low E:T ratio (1:5) environment (the same figure legend was used to identify the test conditions). [Figure 23A] 1 shows serum levels of fibrinogen in cynomolgus monkeys after injection of 1 mg / kg, 10 mg / kg, and 30 mg / kg doses. [Figure 23B]1 shows the transient increase in IL-6 in cynomolgus monkeys after injection of 1 mg / kg, 10 mg / kg, and 30 mg / kg doses. [Figure 23C] Figure 1 shows the serum half-life of the trivalent bispecific antibody construct v38490 in cynomolgus monkeys following injection of 1 mg / kg, 10 mg / kg, and 30 mg / kg doses. [Figure 24A] Figures 24A-24B show that after incubation in A5Su buffer at 40°C for 14 days, the trivalent bispecific antibody construct v32523 remained monodisperse, as indicated by size exclusion chromatography (SEC) analysis, indicating that the purity of v32523 after incubation was at least about 99% as measured by SEC and contained less than about 1% other high molecular weight species (HMWS). Figures 24A-24B show that after incubation in A5Su buffer at 40°C for 14 days, the trivalent bispecific antibody construct v32523 remained monodisperse, as indicated by dynamic light scattering (DLS) analysis, indicating that the purity of v32523 after incubation was at least about 99% as measured by SEC and contained less than about 1% other high molecular weight species (HMWS). [Figure 24B] See legend to Figure 24A. DETAILED DESCRIPTION OF THE INVENTION

[0049] Detailed Description This disclosure is based on the MSLN +The present invention relates to multivalent and bispecific antibody constructs that may address unmet medical needs for improved treatment options for tumors. Specifically, the antibody constructs described herein may have fewer side effects than conventional constructs, allowing for a wider therapeutic window and the use of more effective doses. As further demonstrated herein, it has been unexpectedly discovered that such improved properties can be achieved, at least on a large scale, by improving the format of the antibody construct. In some embodiments, adjusting the format of the construct can induce more TAA (e.g., MSLN)-dependent T cell activity in vitro and / or in vivo, thereby providing more tumor-focused T cell-mediated cytotoxicity, while significantly reducing cytotoxic activity in non-tumor (e.g., MSLN-low or MSLN-negative) tissues.

[0050] As further described herein, MSLN + At least one of the remaining technical challenges in treating tumors is the MSLN + This can be seen as an imbalance between the cytotoxic activity of a molecule against tumor cells and its activity in tissues with low MSLN expression. Among other aspects involving enhanced anti-tumor activity, the antibody constructs of the present disclosure can target cytotoxic effector cells (e.g., CD3 T cells, etc.) in a more TAA (e.g., MSLN)-dependent manner, for example, when compared to conventional approaches. +Antibody constructs have been designed and manufactured to address this shortcoming by providing the ability to target TAA-specific immune cells (e.g., TAA-specific IgG1 and TAA-specific IgG2) to tumor cells. The in vitro and in vivo experiments described herein demonstrate the unexpected and superior performance (e.g., low TAA-independent activity, high tumor cell killing in a TAA-dependent manner, and in vivo tolerability) of multivalent (e.g., trivalent) bispecific antibody constructs over existing clinical benchmark constructs. Further disclosed herein are antibody constructs with specific formats and geometries, and demonstrate how the in vitro and in vivo performance (e.g., anti-tumor activity, TAA-specific tumor killing, etc.) of the constructs can depend on the format, e.g., by using an anti-TAA scFv domain instead of an anti-TAA Fab domain.

[0051] The present disclosure further describes pharmaceutical compositions comprising the multivalent, bispecific antibody constructs, e.g., trivalent, bispecific antibody constructs, of the present disclosure, as well as methods of producing and using such constructs and compositions for the treatment of TAA-positive tumors.

[0052] It should be understood that, in general, the positive recitation of a feature in one embodiment constitutes a basis for excluding that feature in an alternative embodiment. In particular, where a list of alternatives is presented for a given embodiment or claim, it should be understood that one or more alternatives may be deleted from the list, and that the shortened list may form an alternative embodiment, whether or not such alternative embodiment is specifically referenced.

[0053] It is further envisioned that any embodiment discussed herein can be implemented with respect to any antibody construct, method, use, or composition disclosed herein, and vice versa. Moreover, modifications of the specific embodiments described herein that will be apparent to those skilled in the art are intended to be included within the scope of the claims set forth herein.

[0054] definition Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.

[0055] The term "about," as used herein in reference to a numerical value or range, generally refers to ±10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, or ±1% of the recited or claimed numerical value or range, unless otherwise specified. In various embodiments, the term "about" refers to approximately a ±10% variation from a given value or range. In other embodiments, the term "about" refers to approximately a ±5% variation from a given value or range. In yet other embodiments, the term "about" refers to approximately a ±1% variation from a given value or range. It should be understood that, unless otherwise specified, such a variation is always included in any given value provided herein, whether or not it is specifically referred to.

[0056] The use of the words "a" or "an," when used herein in conjunction with the term "comprising," can mean "one," but is also consistent with the meaning of "one or more," "at least one," and "one or more than one."

[0057] As used herein, the terms "comprising," "having," "including," and "containing," and grammatical variations thereof, when used herein in connection with a construct, composition, use, or method, are inclusive, i.e., open-ended, and do not exclude additional, unrecited elements and / or method steps. The term "consisting essentially of," when used herein in connection with a construct, composition, use, or method, indicates that additional elements and / or method steps may be present, but that these additional features do not materially affect the manner in which the recited construct, composition, method, or use functions. The term "consisting of," when used herein in connection with a construct, composition, use, or method, excludes the presence of additional elements and / or method steps. An antibody construct, composition, use, or method described herein as including particular elements and / or steps may also, in certain embodiments, consist essentially of those elements and / or steps, and in other embodiments, consist of those elements and / or steps, whether or not those embodiments are specifically referred to.

[0058] The terms "subject" and "patient" are used interchangeably herein and generally refer to an animal in need of treatment. The animal in need of treatment can be a human or a non-human animal, e.g., a mammal, a bird, or a fish. In certain embodiments, the subject or patient is a mammal. In some embodiments, the subject is a human. In other embodiments, the subject is a rodent or a non-human primate.

[0059] An "effective amount" of an antibody construct described herein, or a pharmaceutical composition comprising such an antibody construct, with respect to a particular result to be achieved, is an amount sufficient to achieve the desired result. For example, when referring to the killing of cancer cells, an "effective amount" of an antibody construct or pharmaceutical composition refers to the amount of antibody construct or composition sufficient to produce a killing effect. In some embodiments, the achievement of the desired result (e.g., the killing of cancer cells) can be confirmed by measurement using one or more of the relevant methods described herein and / or methods known to those of skill in the art.

[0060] The terms "Fc region," "Fc," and "Fc domain" are used interchangeably herein to refer to the C-terminal region of an immunoglobulin heavy chain containing at least a portion of the constant region. In various embodiments, the Fc domain herein may be a dimer. Such a dimeric Fc domain may comprise a first Fc polypeptide and a second Fc polypeptide, and each Fc polypeptide may comprise a CH2 domain and / or a CH3 domain. Such a dimeric Fc may be either a homodimer, i.e., comprising a first and a second Fc polypeptide having identical amino acid sequences, or a heterodimer, i.e., comprising a first and a second Fc polypeptide having different amino acid sequences, e.g., sequences sharing about 95%, 96%, 97%, 98%, or 99% sequence identity. In some embodiments, the antibody constructs of the present disclosure comprise a homodimeric Fc domain. In still other embodiments, as further described herein, the antibody construct comprises a heterodimeric Fc domain, wherein at least one of the CH2 and / or CH3 domains of the first and second Fc polypeptides have amino acid sequences that share less than about 99%, less than 98%, or less than about 97% sequence identity.

[0061] The term "bispecific" as used herein with respect to an antibody construct refers to a biologically functional protein (e.g., an antibody construct described herein) that is "at least bispecific," i.e., comprises at least a first binding domain and a second binding domain, where such first and second binding domains are capable of specifically binding to two distinct epitopes, e.g., a first epitope and a second epitope. Such first and second epitopes may be located on the same antigen or on different antigens (e.g., a first epitope on CD3 and a second epitope on MSLN). Thus, in some embodiments, an antibody construct according to the present disclosure may comprise specificities for at least two different antigens, epitopes, or other targets. In certain embodiments, an antibody construct of the present disclosure is bispecific for two different epitopes, each epitope being located on a different antigen (e.g., CD3 and MSLN). The term "bispecific" as used herein in reference to antibody constructs also encompasses multispecific antibody constructs, such as trispecific antibody constructs (the latter comprising three binding domains, each binding to a distinct epitope or target), or constructs with more than three (e.g., four, five, etc.) specificities.

[0062] The term "trivalent," as used herein with respect to an antibody construct, refers to a biologically functional protein (e.g., an antibody construct described herein) that is "at least trivalent," i.e., it comprises three binding domains, e.g., at least a first binding domain, a second binding domain, and a third binding domain, where each of the first, second, and third binding domains can specifically bind to an epitope and / or antigen, e.g., CD3, MSLN, etc. The three binding domains can have specificity for three different epitopes or antigens, or two or more of the three binding domains have specificity for the same epitope or antigen. Thus, the valency, e.g., monovalent, bivalent, or trivalent, of an antibody construct herein describes the total number of antigen-binding domains of the antibody construct. Thus, the valency of an antibody construct should be at least equal to its specificity, i.e., a bispecific antibody construct should be at least bivalent. In embodiments herein where the antibody construct is trivalent and bispecific, two of the three binding domains of the construct are capable of bivalently binding to a first epitope or antigen (e.g., MSLN), and the third binding domain binds monovalently to a different epitope and / or antigen (e.g., CD3).

[0063] As used herein, the abbreviations "H," "H1," "H2," "L1," etc. are generally used as generic heavy and light chain identifiers, respectively. By way of example, "H1" and "H2" can broadly refer to the first and second heavy chains, respectively, of an antibody construct herein and are thus not intended to be limited in any way to a particular heavy chain amino acid (or polynucleotide) sequence. Rather, such chain identifiers may be used herein to describe and distinguish between two or more polypeptide chains of an antibody construct.

[0064] The term "amino acid modification," as used herein with respect to an amino acid sequence of a polypeptide, generally refers to an amino acid sequence of a polypeptide into which one or more amino acid substitutions, one or more amino acid insertions, and / or one or more amino acid deletions have been introduced, as compared to the corresponding unmodified (e.g., WT or reference) amino acid sequence of the polypeptide.

[0065] antibody construct In various embodiments, the present disclosure relates to multivalent and bispecific antibody constructs capable of targeting antigens on the surface of cytotoxic effector cells and tumor-associated antigens (TAA) located on tumor cells. In various embodiments, the antibody constructs of the present disclosure are trivalent and bispecific, comprising at least three binding domains. Such trivalent bispecific antibody constructs can monovalently bind to antigens on the surface of cytotoxic effector cells and can be bivalent for binding to TAA located on tumor cells. The two binding domains capable of bivalent binding to TAA can be directed against either the same TAA or two different TAAs. In various embodiments, both anti-TAA binding domains are directed against the same TAA. In some embodiments, both anti-TAA binding domains are directed against the same epitope of a TAA (e.g., MSLN).

[0066] In certain other embodiments, the trivalent bispecific antibody construct may be monovalently capable of binding to a TAA on the surface of a tumor cell and bivalent for binding an antigen on the surface of a cytotoxic effector cell.

[0067] In various embodiments, the trivalent bispecific antibody construct comprises (i) a first binding domain capable of monovalently binding to an antigen on a cytotoxic effector cell, and (ii) a second binding domain and a third binding domain capable of bivalently binding to a TAA on a tumor cell. In some embodiments, the antigen on the cytotoxic effector cell is CD3, the cytotoxic effector cell can be an immune cell such as a T cell, and the TAA is MSLN.

[0068] In some embodiments, the present disclosure relates to trivalent bispecific antibody constructs comprising (i) a first binding domain capable of binding to CD3 on the surface of a cytotoxic effector cell, and (ii) a second binding domain and a third binding domain, wherein at least one of the second and third binding domains is capable of binding to MSLN. In various embodiments, both the second and third binding domains are capable of binding to MSLN. Such constructs described herein, which are monovalent for CD3 and bivalent for MSLN, may have the ability to direct immune cells (e.g., T cells) to tumor cells and induce localized MSLN-dependent and immune cell-mediated anti-tumor immune responses. In certain embodiments, the antibody constructs of the present disclosure may not induce significant TAA-independent T cell or immune responses in the absence or low abundance of TAAs, thereby reducing off-target effects and enabling a wider therapeutic window compared to other existing approaches that are less TAA-dependent in their anti-tumor immune responses. The term "significant" with respect to a TAA-independent immune response generally refers to an immune response elicited by an antibody construct that is less than 10%, 5%, or 3% of the immune response elicited in an environment containing moderately or highly TAA-expressing cells or tissues, as further described herein.

[0069] To date, certain bispecific antibodies capable of targeting T cells to tumor cells have been identified and their effectiveness in cancer treatment tested. Blinatumomab is an example of a bispecific anti-CD3-CD19 antibody in a format called BiTE™ (Bispecific T-cell Engager), which has been identified and approved by the FDA for the treatment of B-cell disorders such as relapsed B-cell non-Hodgkin's lymphoma and chronic lymphocytic leukemia (see, e.g., Baeuerle et al (2009) Cancer Research 12:4941-4944). T cell enhancers directed against other tumor-associated target antigens have also been generated, and several have entered clinical trials, including AMG110 / MT110 EpCAM for lung, gastric, and colorectal cancer; AMG211 / MEDI565 CEA for gastrointestinal adenocarcinoma; and AMG 212 / BAY2010112 PSMA for prostate cancer (see, e.g., Suruadevara, CM et al., Oncoimmunology. 2015 Jun;4(6):e1008339). While these studies demonstrated some clinical efficacy, they were hampered by severe dose-limiting toxicities, primarily due to cytokine release syndrome (CRS). This narrowed the therapeutic window for these agents.

[0070] The T cell-binding antibody constructs of the present disclosure aim to address these shortcomings, for example, by reducing or rendering unmeasurable T cell activation when the targeted TAA, e.g., MSLN, is absent or low in abundance, thereby focusing their cytotoxic effect primarily on the tumor (micro)environment and are envisioned to provide not only higher anti-cancer activity when compared to conventional T cell-binding approaches, but also significantly reduce off-target effects, allowing for a wider therapeutic window.

[0071] In various embodiments, the T cell-binding trivalent bispecific antibody constructs of the present disclosure can bind to CD3 (monovalent) on T cells and MSLN (bivalent) on tumor cells, and upon binding to CD3 and MSLN can form a TCR-independent artificial immune synapse between the T cell and the tumor cell, which can activate the T cell and exert a cytotoxic effect on the tumor cell, resulting in tumor cell death, as further described herein.

[0072] A. Antibody Construct Formats and Geometries In certain embodiments, the present disclosure describes multivalent bispecific antibody constructs with formats and geometries specifically designed for improved in vivo tumor killing and tolerability characteristics (e.g., reduced off-target effects) compared to conventional constructs with different formats and / or geometries. In combination with the improved format, the antigen affinity of each binding domain present in the trivalent bispecific antibody constructs described herein has been rationally selected to provide potent anti-tumor activity in tissues that express the TAA, while limiting the activity of the construct in tissues with low TAA expression.

[0073] Generally, the format of an antibody construct described herein refers to the type(s) and number(s) of antibody construct domains (e.g., scFv, Fab, Fc, etc.) it contains, which further defines the valency of the construct and its specificity, i.e., bispecific, trispecific, etc., and bivalent, trivalent, etc., for its target (e.g., CD3, MSLN, etc.). As further described herein, the format of an antibody construct may also be designated as a combination of numbers indicating its valency and specificity for a particular target, including formats described as 1+1(a), 2+1(b), 2+2(c) (representing an antibody construct capable of binding, in the same order, (a) monovalently ("1") to a first antigen and monovalently ("1") to a second antigen, (b) bivalently ("2") to a first antigen and monovalently ("1") to a second antigen, and (c) bivalently ("2") to a first antigen and bivalently ("2") to a second antigen). The geometry of a construct herein is generally defined as the relative and three-dimensional orientation and arrangement of the various domains of an antibody construct. By way of example, an antibody construct comprising a Fab domain, an scFv domain, and a dimeric Fc domain may have different geometric geometries in which the domains are interconnected differently, resulting in different relative orientations. In one such embodiment, both the Fab domain and the scFv domain may each be linked to the N-terminus of the dimeric Fc domain (e.g., each to the N-terminus of the first or second Fc polypeptide). In another embodiment, one of the Fab domain and the scFv domain may be linked to the N-terminus of the Fc domain, and the other to the C-terminus of the dimeric Fc domain. In yet another embodiment, all three domains may be linked in tandem, e.g., the Fab domain may be linked to the N-terminus of the Fc domain, and the scFv domain may be linked to the N-terminus of the Fab domain.

[0074] In various embodiments, the present disclosure describes bispecific antibody constructs comprising (i) a first binding domain capable of binding to CD3, (ii) a second binding domain capable of binding to MSLN, and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. In some embodiments, the first binding domain capable of binding to CD3 comprises a first light chain variable domain (V L ) and the first light chain constant domain (C L a first heavy chain variable domain (V) paired with a light chain comprising H ) and heavy chain constant domain (C H1 In some embodiments, the second binding domain capable of binding to MSLN is a Fab domain comprising a heavy chain comprising a second heavy chain variable domain (V H ) and a second light chain variable domain (V L In some embodiments, the first binding domain is an scFv domain comprising a C H1 and the second binding domain is linked via the C-terminus of the second Fc polypeptide of the heterodimeric Fc domain to the N-terminus of the second heavy chain variable domain (V), depending on the relative orientation and domain structure of the scFv domains. H ) or the C-terminus of the second light chain variable domain (V L ) via either the C-terminus of

[0075] In some embodiments, the first binding domain is a Fab domain, either directly or via a linker. Fab-Fc Similarly, the second binding domain can be an scFv domain and can be linked to the second Fc polypeptide either directly or via a linker. scFv-Fc In various embodiments, the first binding domain and the second binding domain can be linked to the first and second Fc polypeptides, respectively, via a linker Fab-Fc and linker scFv-Fc In some embodiments, the linker Fab-Fc and linker scFv-Fcmay comprise or consist of different amino acid sequences, as further described herein. In other embodiments, the linker Fab-Fc and linker scFv-Fc comprise or consist of the same amino acid sequence.

[0076] In various embodiments of the present disclosure, antibody constructs can be bispecific and multivalent. In some embodiments, such constructs are bivalent and bispecific. In various other embodiments herein, antibody constructs of the present disclosure are trivalent and bispecific.

[0077] The trivalent bispecific antibody constructs of the present disclosure may further comprise a third binding domain. Such a third binding domain may be capable of binding to a different TAA compared to the second binding domain. In other embodiments, the third binding domain may bind to the same TAA as the second binding domain. In various embodiments, both the second binding domain and the third binding domain are capable of binding to MSLN. The third binding domain may be a Fab domain or a second scFv domain. In some embodiments, the third binding domain is a Fab domain. In various other embodiments, the third binding domain is a second scFv domain, resulting in an antibody construct comprising one Fab domain and two scFv domains, i.e., a first scFv domain (also referred to herein as "scFv1") and a second scFv domain (also referred to herein as "scFv2"), and such first and second scFv domains are capable of binding to MSLN.

[0078] In various embodiments, the trivalent bispecific antibody constructs of the present disclosure can be monovalent for CD3 and bivalent for MSLN. Such antibody constructs can comprise (i) a first binding domain capable of binding to CD3 on a cytotoxic effector cell, (ii) a second binding domain capable of binding to MSLN, (iii) a third binding domain capable of binding to MSLN, and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. In various embodiments, the first binding domain is a Fab domain, and the second and third binding domains are scFv domains, i.e., a first scFv domain and a second scFv domain, respectively. In some embodiments, the second and third binding domains are VF domains that share at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity with each other. H and V L Thus, in some embodiments, the second V of the second binding domain comprises or consists of the amino acid sequence H and V L domain (e.g., the first scFv domain, i.e., "scFv1") as well as the third V of the third binding domain H and V L The domains (e.g., the second scFv domain, i.e., "scFv2") comprise or consist of the same amino acid sequence, i.e., the same anti-MSLN paratope V H and V L In some embodiments, the second and third binding domains of the trivalent bispecific antibody construct of the present invention can bind to the same epitope on MSLN. However, in other embodiments, the trivalent bispecific antibody construct herein can comprise a second and third binding domain that can each bind to a different epitope on MSLN. In still other embodiments, the trivalent bispecific antibody construct can comprise a second and third binding domain that each comprise a different anti-MSLN paratope but can still bind to the same MSLN epitope.

[0079] In various embodiments of the present disclosure, the trivalent bispecific antibody construct comprises (i) a first binding domain capable of binding to CD3 on a cytotoxic effector cell (e.g., an immune cell such as a T cell), (ii) a second binding domain capable of binding to MSLN, (iii) a third binding domain capable of binding to MSLN, and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the first binding domain is linked to the first Fc polypeptide via a first linker, the second binding domain is linked to the second Fc polypeptide via a second linker, and the third binding domain is linked, either directly or via a third linker, to either (a) the first binding domain, (b) the second binding domain, or (c) the heterodimeric Fc domain.

[0080] In various embodiments, the third binding domain is linked to the first binding domain via a third linker.

[0081] In various embodiments of the trivalent bispecific antibody constructs of the present disclosure, the first binding domain capable of binding to CD3 is a first light chain variable domain (V L ) and the light chain constant domain (C L a first heavy chain variable domain (V) paired with a light chain (L1) comprising H ) and heavy chain constant domain (C H1 ), and the second and third binding domains capable of binding to MSLN are each an scFv domain, i.e., a first scFv domain and a second scFv domain. Thus, the second binding domain (e.g., the first scFv domain) is a Fab domain comprising a heavy chain comprising a second heavy chain variable domain (V H ) and a second light chain variable domain (V L ), and the third binding domain (e.g., the second scFv domain) may comprise a third heavy chain variable domain (V H ) and the third light chain variable domain (V L ).

[0082] In some embodiments of the trivalent bispecific antibody construct, the Fab domain is H1 a first linker (e.g., linker Fab-Fc The first scFv domain is linked via its V H Domain or its V L via the C-terminus of either domain (e.g., the V in the first scFv domain) H Sequence and V L Depending on the relative orientation of the sequences, a second linker (e.g., linker scFv-Fc Similarly, the second scFv domain can be linked via its V H Domain or its V L via the C-terminus of either domain (e.g., the V in the second scFv domain) H Sequence and V L Depending on the relative orientation of the sequences), a third linker (e.g., linker scFv-Fab ) Because a Fab domain comprises a heavy chain and a light chain, the second scFv domain can be linked to either the N-terminus of the heavy chain or the N-terminus of the light chain of the Fab domain. However, in various embodiments, the second scFv domain can be linked to the N-terminus of the heavy chain of the Fab domain, e.g., via the V H It is linked to the N-terminus of the domain.

[0083] In various embodiments, linkers (e.g., linkers) connecting the various domains of the antibody construct are used. Fab-Fc , linker scFv-Fc , linker scFv-Fab etc.) are all peptide (or peptidic) linkers, i.e., linkers comprising or consisting of an amino acid sequence of at least 2, 3, 4, 5, 10, 15, 20, 25, 30, 40, or more consecutive amino acid residues. In some embodiments, the linkers connecting the various binding domains to the dimeric Fc domain Fab-Fc and linker scFv-Fcis a peptide linker that may comprise or consist of an immunoglobulin hinge region, e.g., an IgG or IgG1 hinge region, as further described herein.

[0084] In some embodiments, the trivalent bispecific antibody constructs of the present disclosure may comprise or consist of two, three, or more polypeptide chains that can be expressed separately and then assembled (e.g., via heavy and / or light chain pairing) to form an antibody construct comprising its various domains. In various embodiments, the trivalent bispecific antibody constructs described herein may comprise or consist of three polypeptide chains: two heavy chains (e.g., designated H1 and H2), and one light chain (e.g., designated L1). In such embodiments, H1 has the domain structure (from N- to C-terminal direction): scFv2-V H -C H1 -CH2-CH3, and the second scFv domain (scFv2) may comprise V, -CH2-CH3, depending on their relative orientation within the scFv2. H-scFv2 or V L-scFv2 via either the C-terminus of the V of the Fab domain H In such an embodiment, H1 is linked to the N-terminus of the domain structure: L -V H ) scFv2 -V H -C H1 In other embodiments, H1 may comprise or consist of the domain structure: (V H -V L ) scFv2 -V H -C H1 Furthermore, H2 may comprise the domain structure (from N- to C-terminal direction): scFv1-CH2-CH3, and the first scFv domain may be V, ... H-scFv1 or V L-scFv1 In such embodiments, H2 can be linked to the N-terminus of CH2 via the C-terminus of either of the domain structures: (V L -VH ) scFv1 In other embodiments, H2 may comprise or consist of the domain structure: (V H -V L ) scFv1 The light chain L1 may comprise or consist of the domain structure: V L -C L (in the N-terminal to C-terminal direction) H -C H1 and L1 can form a Fab domain (i.e., a first binding domain) capable of binding to CD3. Formation of a Fab domain can involve covalent interactions (e.g., disulfide bonds), non-covalent interactions between amino acid residues of H1 and L1, or a combination of both. Furthermore, the CH2 and CH3 domains of H1 (e.g., corresponding to either part or all of a first Fc polypeptide) and the CH2 and CH3 domains of H2 (e.g., corresponding to either part or all of a second Fc polypeptide) can form a heterodimeric Fc domain, in which case the CH3 domain of H1 and the CH3 domain of H2 can each contain one or more asymmetric amino acid substitutions that promote the formation of heterodimeric Fc domains over the formation of homodimeric Fc domains, as further described herein.

[0085] Figure 1 shows schematic diagrams of the geometries (relative orientation and connectivity of binding domains) and formats (e.g., valency, presence or absence of an Fc domain, etc.) of various antibody constructs contemplated and described herein. Specifically, Figure 1A shows a schematic diagram of the geometry and format of a trivalent, bispecific antibody construct according to various embodiments of the present disclosure, which consists of three polypeptide chains, H1, H2, and L1, and is capable of binding to CD3 (monovalently via one Fab domain) and MSLN (bivalently via two scFv domains).

[0086] In various embodiments, the present disclosure describes antibody constructs comprising: (i) a Fab domain capable of binding to an antigen on a cytotoxic effector cell; (ii) a first scFv domain and a second scFv domain, both of which are capable of binding to MSLN; and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain.

[0087] In a further embodiment, the present disclosure provides a Fab antibody comprising: (i) a Fab domain capable of binding to an antigen on a cytotoxic effector cell; (ii) a first scFv domain and a second scFv domain, both of which are capable of binding to MSLN, and at least one of the first scFv domain and the second scFv domain comprising a heavy chain complementarity determining region 1 (HCDR1) sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and a light chain complementarity determining region 1 (LCDR1) sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain.

[0088] In certain embodiments, both the first scFv domain and the second scFv domain comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122, respectively. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L Includes the domain.

[0089] In some of these embodiments, the antigen on the cytotoxic effector cell is CD3. The cytotoxic effector cell can be an immune cell, for example, a T cell.

[0090] In various embodiments, the present disclosure provides a V Fab domain capable of binding to CD3 (the Fab domain comprises an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 128). H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L(ii) a first scFv domain and a second scFv domain (both of which are capable of binding to MSLN); and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain.

[0091] In various embodiments, the disclosure provides a VH1-binding domain comprising: (i) a Fab domain capable of binding to CD3 (the Fab domain comprises an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 128). H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L (ii) a first scFv domain and a second scFv domain (both of which are capable of binding to MSLN, and both scFv domains comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122); H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L(iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain.

[0092] In such embodiments, the antibody construct may be trivalent and bispecific.

[0093] In some embodiments, the trivalent bispecific antibody constructs described herein may comprise three polypeptide chains, two heavy chains (e.g., H1 and H2), and one light chain (e.g., L1). H1 has the domain structure (from N- to C-terminal): (V L -V H ) scFv2 -V H -C H1 -CH2-CH3, in which case "scFv2" refers to the second scFv domain capable of binding to MSLN. H2 has the domain structure (from N- to C-terminal): (V L -V H ) scFv1 L1 may comprise or consist of -CH2-CH3, in which case "scFv1" refers to the first scFv domain capable of binding to MSLN. L1 has the domain structure (from N- to C-terminal): V L -C L V of H1 H -C H1 and L1 V L -C Lcan pair (e.g., covalently) (e.g., via one or more disulfide bonds) to form a Fab domain capable of binding to CD3. The CH2 and CH3 domains of H1 that form at least in part the first Fc polypeptide, and the CH2 and CH3 domains of H2 that form at least in part the second Fc polypeptide, can form a heterodimeric Fc domain, and the CH3 domains of H1 and H2 can each comprise one or more asymmetric amino acid substitutions that promote formation of the heterodimeric Fc domain (e.g., formation of an H1-H1 or H2-H2 homodimeric Fc) compared to the respective homodimeric Fc domains (e.g., formation of an H1-H1 or H2-H2 homodimeric Fc).

[0094] In some embodiments, the trivalent bispecific antibody constructs described herein may comprise three polypeptide chains, two heavy chains (e.g., H1 and H2), and one light chain (e.g., L1). H1 has the domain structure (from N- to C-terminal): (V H -V L ) scFv2 -V H -C H1 -CH2-CH3, in which case "scFv2" refers to the second scFv domain capable of binding to MSLN. H2 has the domain structure (from N- to C-terminal): (V H -V L ) scFv1 L1 may comprise or consist of -CH2-CH3, in which case "scFv1" refers to the first scFv domain capable of binding to MSLN. L1 has the domain structure (from N- to C-terminal): V L -C L V of H1 H -C H1 and L1 V L -C Lcan pair (e.g., covalently) (e.g., via one or more disulfide bonds) to form a Fab domain capable of binding to CD3. The CH2 and CH3 domains of H1 that form at least in part the first Fc polypeptide, and the CH2 and CH3 domains of H2 that form at least in part the second Fc polypeptide, can form a heterodimeric Fc domain, and the CH3 domains of H1 and H2 can each comprise one or more asymmetric amino acid substitutions that promote formation of the heterodimeric Fc domain (e.g., formation of an H1-H1 or H2-H2 homodimeric Fc) compared to the respective homodimeric Fc domains (e.g., formation of an H1-H1 or H2-H2 homodimeric Fc).

[0095] Thus, in various embodiments, the antibody constructs of the present disclosure comprise or consist of three polypeptide chains, H1, H2, and L1, wherein (i) H1 has the domain structure (from N-terminal to C-terminal): (V L -V H ) scFv2 -V H -C H1 (ii) H2 comprises or consists of the domain structure (from the N-terminus to the C-terminus): (V L -V H ) scFv1 (iii) L1 comprises or consists of the domain structure (from the N-terminus to the C-terminus): V L -C L In this case, V of H1 H -C H1 and V of L1 L -C L form a Fab domain capable of binding to CD3, scFv1 and scFv2 are each capable of binding to MSLN, the first Fc polypeptide and the second Fc polypeptide form a heterodimeric Fc domain, and the CH3 domain of H1 and the CH3 domain of H2 each contain one or more asymmetric amino acid substitutions that promote formation of the heterodimeric Fc domain over the corresponding homodimeric Fc domain.

[0096] In various embodiments, the trivalent bispecific antibody constructs described herein may further comprise one or more linkers, as described herein, covalently linking two or more domains of the construct to one another. In such embodiments, the Fab domain may be linked to the first Fc polypeptide of the heterodimeric Fc domain via a first linker, the first scFv domain may be linked to the second Fc polypeptide of the heterodimeric Fc domain via a second linker, and the second scFv domain may be linked to the Fab domain via a third linker. The first linker linking the Fab domain to the Fc polypeptide is referred to herein as the linker Fab-Fc (i.e., referring to the linker that links the Fab domain to the Fc polypeptide). The second linker that links the scFv domain to the Fc polypeptide is referred to herein as the linker scFv-Fc Similarly, the third linker connecting the scFv domain to the Fab domain may be referred to herein as the linker scFv-Fab It can be referred to as.

[0097] Thus, in some embodiments, the first polypeptide chain H1 of the antibody constructs herein may comprise one or more peptide linkers. H1 a sequence (e.g., of a Fab domain) to the CH2 domain of the first Fc polypeptide via a first linker, e.g., linker Fab-Fc such a linker can be used to link the Fab domain to the first Fc polypeptide via Fab-Fc can be a peptide linker. In some embodiments, the linker Fab-Fc may comprise or consist of an immunoglobulin hinge region, e.g., an IgG or IgG1 hinge region. In some embodiments, the linker Fab-Fccomprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 107. Furthermore, in various embodiments, the second scFv domain (e.g., scFv2) is H or V L V of the Fab domain via either of the sequences H Linker in the sequence scFv-Fab The linker may be a linker. scFv-Fab can be a peptide linker. In various embodiments, the linker scFv-Fab is the amino acid sequence (Gly4Ser) n (or one-letter amino acid code: (G4S) n , SEQ ID NO: 132), where n is an integer from 1 to 5. In various embodiments, the linker scFv-Fab comprises or consists of the amino acid sequence (Gly4Ser) (SEQ ID NO: 104). In various embodiments, the first and / or second scFv domain (e.g., scFv1 or scFv2) itself comprises or consists of the amino acid sequence (Gly4Ser) (SEQ ID NO: 104). L Domain V H A fourth linker may be included to connect the domains. scFv is the linker scFv A linker may be referred to as a scFv can be a peptide linker. In some embodiments, the linker scFv is the amino acid sequence (Gly4Ser) n (SEQ ID NO: 132), where n is an integer from 1 to 5. In various embodiments, the linker scFv comprises or consists of the amino acid sequence (Gly4Ser)4 (SEQ ID NO: 102).

[0098] In some embodiments, the H1 polypeptide chain of the trivalent bispecific antibody construct herein has the domain structure (from N-terminal to C-terminal): (V L- Linker scFv -V H ) scFv2 -LinkerscFv-Fab -V H -C H1 -Linker Fab-Fc -CH2-CH3, where "scFv2" indicates the second scFv domain capable of binding to MSLN.

[0099] In various embodiments, the second polypeptide chain H2 of the antibody constructs herein may comprise one or more peptide linkers. In such embodiments, the first scFv1 domain is linked to its V H or V L a fifth linker (linker scFv-Fc (which may also be referred to as a linker) thereby linking the first scFv domain (scFv1) to the second Fc polypeptide. scFv-Fc can be a peptide linker. In some embodiments, the linker scFv-Fc may comprise or consist of an immunoglobulin hinge region, e.g., an IgG or IgG1 hinge region. In some embodiments, the linker scFv-Fc comprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 111 or SEQ ID NO: 112. Further, the first scFv domain (scFv1) comprises a V L Domain V H Linker that connects domains scFv1 (in which case V H and V L (The order of N-terminus to C-terminus of such linkers may vary.) scFv1 can be a peptide linker, and the linker present in scFv2 scFv (Linker scFv2 In some embodiments, the linker scFv1 is the amino acid sequence (Gly4Ser) n(SEQ ID NO: 132), where n is an integer from 1 to 5. In various embodiments, the linker scFv1 comprises or consists of the amino acid sequence (Gly4Ser)4 (SEQ ID NO: 102), wherein the first and second scFv domains are linked by the same linker scFv Includes:

[0100] In some embodiments, the H2 polypeptide chain of the trivalent bispecific antibody constructs herein has the domain structure (from N-terminal to C-terminal): (V L- Linker scFv -V H ) scFv1 -Linker scFv-Fc -CH2-CH3, where "scFv1" indicates the first scFv domain capable of binding to MSLN.

[0101] In various embodiments, the trivalent bispecific antibody constructs of the present disclosure comprise or consist of three polypeptide chains, which may be referred to as H1, H2, and L1, wherein (i) H1 has the domain structure (from N-terminal to C-terminal): (V L -Linker scFv -V H ) scFv2 -Linker scFv-Fab -V H -C H1 -Linker Fab-Fc (ii) H2 comprises or consists of the domain structure (from the N-terminus to the C-terminus): (V L -Linker scFv -V H ) scFv1 -Linker scFv-Fc (iii) L1 comprises or consists of the domain structure (from the N-terminus to the C-terminus): V L -C L In this case, V of H1 H -C H1 and V of L1 L -C Lform a Fab domain capable of binding to CD3, scFv1 and scFv2 are each capable of binding to MSLN, the first Fc polypeptide and the second Fc polypeptide form a heterodimeric Fc domain, and the CH3 domain of H1 and the CH3 domain of H2 each contain one or more asymmetric amino acid substitutions that promote heterodimeric Fc domain formation and H1-H2 pairing over the corresponding homodimeric Fc domains (e.g., H1-H1 or H2-H2).

[0102] In some embodiments, the present disclosure provides a VFv antibody comprising (i) a Fab domain capable of binding to CD3; (ii) a first scFv domain and a second scFv domain, both of which are capable of binding to MSLN, and at least one of the first and second scFv domains comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 101. L domain, V comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103; H and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain. In some embodiments, both the first and second scFv domains comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 101, respectively. L domain, and V comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103.H In such an embodiment, the first and second scFv domains, i.e., V of scFv1 and scFv2, L Domain and V H The domains may each comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequences set forth in SEQ ID NOs: 101 and 103, respectively. In such embodiments, the V of the first and second scFv domains, i.e., scFv1 and scFv2, L Domain and V H The domains may each comprise or consist of an amino acid sequence having at least about 95% sequence identity to the sequences set forth in SEQ ID NOs: 101 and 103, respectively. In such embodiments, the V of the first and second scFv domains, i.e., scFv1 and scFv2, L Domain and V H The domains may each comprise or consist of an amino acid sequence having at least about 97% sequence identity to the sequences set forth in SEQ ID NOs: 101 and 103, respectively. In such embodiments, the V of the first and second scFv domains, i.e., scFv1 and scFv2, L Domain and V H The domains may each comprise or consist of an amino acid sequence having at least about 99% sequence identity to the sequences set forth in SEQ ID NOs: 101 and 103, respectively. In yet other embodiments, the V of the first and second scFv domains, i.e., scFv1 and scFv2, L Domain and V H The domains may comprise or consist of the amino acid sequences set forth in SEQ ID NOs: 101 and 103, respectively.

[0103] In various embodiments herein, the trivalent, bispecific antibody construct comprises (i) a Fab domain capable of binding to CD3 (the Fab domain comprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 105). Ldomain, and a V domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 115. H (ii) a first scFv domain and a second scFv domain (both of which are capable of binding to mesothelin); and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain. In some embodiments, the V of the Fab domain L Domain and V H The V domains comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequences set forth in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V H The V domains comprise or consist of an amino acid sequence having at least about 95% sequence identity to the sequences set forth in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V H The V domains comprise or consist of an amino acid sequence having at least about 97% sequence identity to the sequences set forth in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V H In yet another embodiment, the V of the Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the sequences set forth in SEQ ID NOs: 105 and 115, respectively. L Domain and V H The domains comprise or consist of amino acid sequences having the sequences set forth in SEQ ID NOs: 105 and 115, respectively.

[0104] In various embodiments described herein, the trivalent bispecific antibody construct may comprise or consist of a set of three polypeptide chains, e.g., H1, H2, and L1, having the amino acid sequences set forth in (i) SEQ ID NOs: 100, 110, and 114, (ii) SEQ ID NOs: 171, 172, and 114, or (iii) SEQ ID NOs: 117, 119, and 114, or variant sequences thereof, e.g., sequences having at least about 80% sequence identity.

[0105] In embodiments in which the antibody construct comprises three polypeptide chains, H1, H2, and L1, (i) H1 may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100; (ii) H2 may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110; and (iii) L1 may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0106] In some embodiments, H1 comprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100, H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 110, and L1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114. In such embodiments, H1 may comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 100. In other embodiments, H1 comprises or consists of an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 100. In some embodiments, H1 comprises or consists of an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 100. In some embodiments, H1 comprises or consists of an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 100. In some embodiments, H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100.

[0107] In some embodiments, H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100, H2 comprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 110, and L1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114. In such embodiments, H2 may comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 110. In other embodiments, H2 comprises or consists of an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 110. In some embodiments, H2 comprises or consists of an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 110. In some embodiments, H2 comprises or consists of an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 110. In some embodiments, H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 110.

[0108] In some embodiments, H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100, H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 110, and L1 comprises or consists of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 114. In such embodiments, L1 may comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 114. In other embodiments, L1 comprises or consists of an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114.

[0109] Thus, in some embodiments, the present disclosure describes a trivalent bispecific antibody construct consisting of three polypeptide chains, H1, H2, and L1, where (i) H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100, (ii) H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 110, and (iii) L1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114. In some embodiments, such a trivalent bispecific antibody construct may be variant v32523.

[0110] In other embodiments, the present disclosure describes a trivalent bispecific antibody construct consisting of three polypeptide chains, H1, H2, and L1, where (i) H1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 117, (ii) H2 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 119, and (iii) L1 comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114. In some embodiments, such a trivalent bispecific antibody construct may be variant v21812.

[0111] In some embodiments, one or more heavy chains of the antibody constructs herein may further comprise a C-terminal lysine residue. Exemplary heavy chain sequences comprising such an additional C-terminal lysine residue are set forth in SEQ ID NOs: 171 and 172, which refer to the first and second heavy chains, respectively, of construct v32523, both of which further comprise a C-terminal lysine. As further described herein, such a C-terminal lysine residue may be cleaved (e.g., by enzymatic cleavage) after expression of the heavy chains of the antibody construct.

[0112] The antibody constructs of the present disclosure can simultaneously bind to CD3 on immune cells (e.g., T cells) and MSLN on tumor cells. Such simultaneous binding can target the anti-tumor activity of T cells to tumor cells and / or the tumor microenvironment. The simultaneous binding of CD3 and MSLN, when combined with the rationally selected antigen affinity of the constructs described herein, can provide TAA (e.g., MSLN)-dependent anti-tumor activity; i.e., the simultaneous binding of CD3 and MSLN can induce a tumor-killing effect in a local region, i.e., at the tumor site and / or in the tumor environment. Thus, the T cell-binding antibody constructs of the present disclosure can be used to treat MSLN-positive tumors while significantly reducing off-target effects following administration to a subject in need of treatment, when compared to existing T cell-binding constructs (e.g., the benchmark constructs described herein as v29191 (Roche) and v31805 (Harpoon)).

[0113] As further described herein, the antibody construct may comprise a heterodimeric Fc domain comprising a first Fc polypeptide comprising a first CH2 domain and a first CH3 domain, and a second Fc polypeptide comprising a second CH2 domain and a second CH3 domain, wherein at least one of the first and second Fc polypeptides may comprise one or more amino acid modifications in the first CH3 domain or the second CH3 domain, respectively, which may promote the formation of the heterodimeric Fc domain comprising the first and second Fc polypeptides over a corresponding homodimeric Fc domain comprising two copies of the first Fc polypeptide or the second Fc polypeptide. As further described herein, in some embodiments, the first and second Fc polypeptides each comprise at least one, two, three, four, or five amino acid modifications, which are modifications relative to the corresponding wild-type Fc polypeptide.

[0114] In various embodiments, the antibody constructs of the present disclosure may contain one or more amino acid modifications in the first CH2 domain of the first Fc polypeptide, the second CH2 domain of the second Fc polypeptide, or both CH2 domains. Such modifications of the CH2 domains can alter the interaction of the antibody construct with an Fc receptor, as further described herein. In various embodiments, one or more amino acid modifications to either one or both CH2 domains can reduce or "knock out" the endogenous Fc receptor function of the Fc domain. Thus, in various embodiments, either one or both CH2 domains of the antibody construct contain one or more amino acid substitutions that reduce or eliminate the interaction of the Fc domain with an Fc receptor, e.g., an Fcγ receptor (FcγR).

[0115] In some embodiments, as used herein, a VL1 antibody comprises: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell (the Fab domain comprises an HCDR1 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 128); H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2) (the first scFv domain and the second scFv domain are capable of binding mesothelin (MSLN)); and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0116] In some embodiments, at least one of the first scFv domain and the second scFv domain, or both scFv domains, comprises (i) a VFV comprising an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H domain, and (ii) a V domain comprising an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L Includes the domain.

[0117] In certain embodiments, as used herein, a Fab domain is a VH1 antibody that is capable of binding to CD3 on a cytotoxic effector cell (the Fab domain comprises an HCDR1 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising, or consisting of, the sequence set forth in SEQ ID NO: 128). H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2) (the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN), and the first scFv domain and the second scFv domain each comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122). H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0118] In yet another embodiment, the present disclosure describes a trivalent bispecific antibody construct comprising a first heavy chain (H1) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 100, a second heavy chain (H2) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 110, and a light chain (L1) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 114.

[0119] B. Domains of the antibody construct As further described herein, the antibody constructs of the present disclosure may comprise one or more antibody domains. In various embodiments, the antibody constructs comprise multiple (i.e., two or more) antibody domains. Such multiple antibody domains may include (i) one or more Fc domains (an Fc domain may comprise a first Fc polypeptide and a second Fc polypeptide), (ii) one or more Fab domains (a Fab domain may comprise a heavy chain variable domain (V H ) and heavy chain constant domain (C H1 ), and a light chain variable domain (V L ) and the light chain constant domain (C L )), and (iii) one or more scFv domains (an scFv domain may comprise a heavy chain variable domain (V H ) and the light chain variable domain (V L )) The various domains that the antibody construct may comprise are further described herein.

[0120] An immunoglobulin (Ig) structural unit generally consists of two pairs of polypeptide chains, each pair having one "light" chain (approximately 25 kilodaltons (kD)) and one "heavy" chain (approximately 50-70 kD). Light chains can be classified as either kappa or lambda. The "class" of an immunoglobulin refers to the type of constant domain possessed by its heavy chain. There are five major classes of antibodies: IgA, IgD, IgE, IgG, and IgM, several of which can be further divided into subclasses (isotypes), e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2. The heavy-chain constant domains corresponding to the different classes of immunoglobulins are called alpha (α), delta (δ), epsilon (ε), gamma (γ), and mu (μ), respectively.

[0121] In various embodiments, the antibody constructs described herein are based on IgG class immunoglobulins, e.g., IgG1, IgG2, IgG3, or IgG4 immunoglobulins. In some embodiments, the antibody constructs described herein are based on IgG1, IgG2, or IgG4 immunoglobulins. In particular embodiments, the antibody constructs described herein are based on IgG1 immunoglobulins. In the context of the present disclosure, when an antibody construct is based on a particular immunoglobulin isotype, it means that the antibody construct comprises either all or only a portion of the constant region (i.e., Fc domain) of the particular immunoglobulin isotype. It is understood that, according to certain embodiments of the present disclosure, the antibody construct may also comprise isotype and / or subclass hybrids.

[0122] In general, in an antibody, the N-terminal domain of each polypeptide chain is usually a variable region (e.g., V) of about 100 to 110 amino acids or more that is mainly involved in antigen recognition. H or V L ) defines the variable light chain (V L ) and variable heavy chain (V HThe terms (V) and (V) refer to these domains in the light and heavy chains, respectively. As described herein, in various embodiments, an antibody construct may comprise two or more variable domain sequences linked to each other in tandem and in a single polypeptide chain format, for example, as disclosed for a construct in which an scFv domain is linked to a Fab domain in a single chain configuration. Such constructs have the domain structure (V) in the N-terminal to C-terminal direction: H -V L ) / (V L -V H )] scFv -V H -C H1 (where " / " is synonymous with "or" herein).

[0123] Thus, in some embodiments, antibody constructs of the present disclosure derived from immunoglobulin molecules may comprise different Ig domains within their heavy and light chains. The heavy chain domain may comprise, for example, an Fc domain (or Fc region) comprising a CH2 domain and a CH3 domain, a hinge domain (or hinge region), and a variable heavy domain (V H ) and constant heavy chain domains (C H1 ), and the light chain domain may comprise a heavy chain Fab domain comprising a variable light chain domain (V L ) and the light chain constant domain (C L In some embodiments, according to a particular nomenclature, an "Fc domain" may include a CH2 and CH3 domain, and a hinge domain (or hinge region).

[0124] V of the antibody construct herein H Domain and V LEach domain has three loops whose sequences are hypervariable and form the antigen-binding site. Each of these loops is referred to as a "hypervariable region" or "HVR" or a "complementarity-determining region" or "CDR." The terms hypervariable region (HVR) and complementarity-determining region (CDR) are used interchangeably herein with respect to the portions of the variable region that form the antigen-binding site. With the exception of CDR1 of VH, CDRs generally comprise the amino acid residues that form the hypervariable loops. V H Domain and V L A domain consists of multiple relatively invariant stretches called framework regions (FRs), each about 15-30 amino acids long, separated by shorter CDRs, each typically about 5-15 amino acids long, but sometimes longer or shorter. H Domain and V L The three CDRs and four FRs constituting each domain are arranged as FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 from the N-terminus to the C-terminus.

[0125] Several different definitions and numbering conventions for the CDR regions of immunoglobulin molecules are commonly used, including those described by Kabat et al. (1983, Sequences of Proteins of Immunological Interest, NIH Publication No. 369-847, Bethesda, MD), Chothia et al. (1987, J Mol Biol, 196:901-917), as well as the IMGT, AbM, and Contact definitions. These different definitions include overlapping or subsets of amino acid residues when compared with each other. By way of example, CDR definitions according to Kabat, Chothia, IMGT, AbM, and Contact are set forth in Table 1 below.

[0126] Therefore, as will be readily appreciated by those skilled in the art, the exact numbering and arrangement of the CDRs may vary depending on the numbering system employed.H It will be understood that the disclosure herein of a variable light chain domain (VL) includes disclosure of the associated (unique) heavy chain CDRs (HCDRs) as defined by any known numbering system. Similarly, the disclosure herein of a variable light chain domain (VL) includes disclosure of the associated (unique) heavy chain CDRs (HCDRs) as defined by any known numbering system. Those skilled in the art will appreciate that a limited number of amino acid substitutions can be made in the CDR sequences or VLs of known antibodies without abolishing the ability of the antibody to bind its target. H Or V L It will be understood that candidate amino acid substitutions can be introduced into the sequence. Candidate amino acid substitutions can be identified by techniques such as computer modeling or alanine scanning, and the resulting variants can be analyzed by standard techniques to determine binding activity (e.g., expressed as binding affinity, e.g., measured EC 50 The antibody constructs described herein are tested for their affinity (represented as a value). As an example, in certain embodiments, the MSLN-binding domain(s) of the antibody constructs described herein may comprise a set of CDRs (i.e., heavy chain CDR1, CDR2, and CDR3, and light chain CDR1, CDR2, and CDR3) that have 90% or more, 95% or more, 98% or more, 99% or more, or 100% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 120-125, respectively, in which case the binding domain retains or substantially retains the ability to bind to MSLN. In this context, the term "substantially" refers to a change in binding affinity of less than about 5%, less than about 4%, less than about 3%, less than about 2%, or less than about 1% compared to the unmodified binding domain. [Table 1]

[0127] In some embodiments, the antibody constructs described herein comprise at least one immunoglobulin domain derived from a mammalian immunoglobulin, e.g., bovine immunoglobulin, human immunoglobulin, camelid immunoglobulin, rat immunoglobulin, or mouse immunoglobulin. In some embodiments, the antibody constructs herein may be chimeric constructs comprising two or more immunoglobulin domains, with at least one domain derived from a first mammalian immunoglobulin, e.g., a human immunoglobulin, and at least a second domain derived from a second mammalian immunoglobulin, e.g., a mouse or rat immunoglobulin. In other embodiments, the antibody constructs may be derived from immunoglobulins from different species; e.g., the antibody constructs may be chimerized or humanized. A "chimeric antibody construct" generally refers to an antibody comprising at least one variable domain derived from a rodent antibody (usually a mouse antibody) and at least one constant domain derived from a human antibody. A "humanized antibody construct" is a type of chimeric antibody that contains minimal sequences derived from a non-human antibody. In some embodiments, the antibody constructs herein may comprise at least one immunoglobulin constant domain derived from a human immunoglobulin, hi some embodiments, all domains of the antibody constructs described herein may be derived from (or be derived from) a human immunoglobulin.

[0128] In some embodiments, modifications (e.g., to the amino acid sequence) to one or more domains of the antibody construct can be made to further improve the properties and performance (e.g., antigen affinity, stability, pharmacokinetics, etc.) of the antibody construct, as further described elsewhere herein. For example, framework region (FR) residues of a human immunoglobulin can be replaced by corresponding non-human residues, or residues that are not present in either the recipient antibody or the donor antibody can be included in the humanized antibody. Generally, a humanized antibody variable domain or humanized antibody domain will contain all or substantially all of the hypervariable regions derived from a non-human immunoglobulin and all or substantially all of the FRs derived from a human immunoglobulin sequence. As further described herein, modifications in the Fc domain can allow the preferential pairing of Fc polypeptides to form heterodimeric Fc domains rather than homodimeric Fc domains.

[0129] In some embodiments, the present disclosure relates to antibody constructs that may have different valencies, e.g., bivalent, trivalent, tetravalent, or higher. Thus, in various embodiments, the antibody constructs herein comprise two or more, or three or more, antigen-binding domains, i.e., each is at least bivalent or at least trivalent. The antibody constructs herein may also comprise different valencies combined with specific antigen specificities. As an example, a bivalent, bispecific antibody construct herein may comprise two binding domains, e.g., a first binding domain (e.g., a Fab domain) and a second binding domain (e.g., an scFv domain). Each of these two binding domains may have unique binding specificities for an antigen (e.g., one domain for a first antigen (e.g., CD3) and the second domain for a second antigen (e.g., MSLN)). In various embodiments of the present disclosure, the antibody constructs may be bispecific and multivalent, and thus may comprise two or more binding domains, e.g., three binding domains. Each of the two or more binding domains may have a unique binding specificity for an antigen (the same epitope / antigen or different epitopes / antigens). In some embodiments, at least two of the two or more binding domains have binding specificity for two different antigens, e.g., CD3 and MSLN. In various embodiments, the antibody constructs herein may be trivalent and bispecific, i.e., comprise three binding domains, one of which has specificity for or is capable of binding to a first antigen (e.g., CD3) and the other two binding domains have specificity for or are capable of binding to a second antigen (e.g., MSLN).

[0130] In embodiments in which the antibody constructs herein comprise two binding domains that bind to the same target molecule (e.g., MSLN), the two binding domains may bind to the same epitope on the target molecule or to different epitopes on the target molecule. In some embodiments, the antibody constructs herein comprise two binding domains that bind to different epitopes on the target molecule (e.g., MSLN), which may be characterized as "biparatopic" binding. However, in other embodiments, the antibody constructs herein comprise two binding domains that bind to the same epitope on an antigen (e.g., MSLN). In such embodiments, the two binding domains may comprise heavy chain and light chain CDRs that share at least about 90%, 95%, 97%, 99%, or 100% amino acid sequence identity. In further embodiments, the V of two binding domains (e.g., first and second scFv domains) that bind to the same epitope on a target (e.g., MSLN) may be selected from the group consisting of: H and V L The domains may share at least about 90%, 95%, 97%, 99%, or 100% amino acid sequence identity, whereby in some of these embodiments, the complete amino acid sequences of the first scFv and the second scFv share at least about 90%, 95%, 97%, 99%, or 100% amino acid sequence identity.

[0131] As described herein, the antibody constructs of the present disclosure may comprise (i) one or more Fab domains, (ii) one or more scFv domains, and (iii) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. In various embodiments, the trivalent bispecific antibody constructs described herein comprise (i) a Fab domain capable of binding to a first antigen, (ii) a first scFv domain and a second scFv domain (both capable of binding to a second antigen), and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0132] Generally, as used herein, a "Fab domain" refers to the variable domain V of each of the light and heavy chains that contains the CDRs described herein. L and V H together with the constant domain of the light chain (C L ) and the first constant domain of the heavy chain (C H1 ). In some embodiments, the Fab domain can be a single-chain Fab. A single-chain Fab can be a Fab molecule in which a Fab light chain and a Fab constant heavy chain are connected by a peptide linker to form a single polypeptide chain. In such embodiments, typically, the C-terminus of the Fab light chain is connected to the N-terminus of the Fab heavy chain in a single-chain Fab molecule, although other formats are encompassed herein. However, in various embodiments herein, the Fab domain of an antibody construct is formed by two separately expressed polypeptide chains, i.e., a light chain and a heavy chain (or portions thereof). However, the heavy and light chain portions of a Fab domain can be interconnected by a covalent bond, such as a disulfide bond.

[0133] As used herein, an "scFv domain" generally refers to a heavy chain variable domain (V) in the format of a single polypeptide chain. H ) and the light chain variable domain (V L The scFv optionally comprises V H Domains and V L A peptide linker may be included between the domains, which may assist the scFv to form a functional structure for antigen binding. Thus, in various embodiments, the scFv domains herein may be linked via their C-terminus to a linker scFv By V H V linked to the N-terminus of the domain L domain, i.e., the scFv domain may comprise the domain structure: V L -Linker scFv -V H or alternatively, the scFv may have V H The C-terminus of scFv By V LThe domain structure: V H -Linker scFv -V L The structure may include a structure having the following structure:

[0134] In some embodiments, the antibody constructs described herein may further comprise another domain or portion that cannot be derived from an immunoglobulin molecule. Such a non-Ig domain may be referred to as a non-Ig portion. Such a non-Ig portion may be a detectable label (e.g., a radioactive or fluorescent label), a low molecular weight (<750 Da) drug molecule, another peptide (e.g., a signal peptide(s)), or a polypeptide molecule, or a combination thereof.

[0135] Fc domain As described herein, the antibody constructs of the present disclosure may comprise an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. In various embodiments, the Fc domain is a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the first Fc polypeptide and the second Fc polypeptide share about 90%, 95%, 97%, or about 99% amino acid sequence identity and each comprise one or more asymmetric amino acid substitutions that can promote preferential pairing of the Fc polypeptides to form the heterodimeric Fc domain relative to the formation of their respective homodimeric Fc domains.

[0136] As used herein, the term "Fc domain" or "Fc region" includes native (or wild-type) sequence Fc domains and variant Fc domains containing one or more amino acid modifications compared to the corresponding native or wild-type Fc domain. Unless otherwise specified herein, the numbering of amino acid residues in an Fc domain or constant region is according to the EU numbering system, also referred to as the EU index, as described, for example, in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD (1991). The "Fc polypeptide" of a dimeric (e.g., heterodimeric) Fc domain refers to one of the two polypeptide chains (e.g., the first and second Fc polypeptides) that form the dimeric (e.g., heterodimeric) Fc domain. In some embodiments, the Fc polypeptide can include a C-terminal constant region of an immunoglobulin heavy chain capable of stable self-association. In various embodiments, as further described herein, an Fc polypeptide (e.g., a first or second Fc polypeptide) comprises at least one of a CH2 domain sequence and / or a CH3 domain sequence. In certain embodiments, the Fc polypeptide of the antibody constructs described herein comprises a CH2 domain sequence and a CH3 domain sequence.

[0137] As disclosed herein, an antibody construct can comprise an Fc domain, and such an Fc domain can be a heterodimeric Fc domain. An Fc domain of an antibody construct, e.g., a heterodimeric Fc domain, comprises a first Fc polypeptide and a second Fc polypeptide, unless otherwise specified. Generally, each Fc polypeptide of a (e.g., heterodimeric) Fc domain can comprise a CH2 domain, a CH3 domain, or, as described in various embodiments herein, both a CH2 domain and a CH3 domain.

[0138] In certain embodiments, the antibody construct comprises an Fc domain based on a human IgG Fc domain. In some embodiments, the antibody construct comprises an Fc domain based on a human IgG1 Fc domain. In various embodiments, the antibody construct comprises two different Fc polypeptides, e.g., a heterodimeric IgG Fc domain comprising a first Fc polypeptide and a second polypeptide, wherein the first and second Fc polypeptides have different amino acid sequences, e.g., amino acid sequences that have about 90%, 95%, 97%, or 99% sequence identity when compared and aligned with each other, as further described herein for heterodimeric Fc domains. In some embodiments, the difference in the amino acid sequences of the first and second Fc polypeptides can be due to asymmetric amino acid substitutions that can be introduced into each Fc polypeptide chain to promote preferential pairing of heavy chains to form the heterodimeric Fc domain compared to the corresponding homodimeric Fc domain.

[0139] In various embodiments, the antibody constructs herein comprise an Fc domain that is a modified IgG Fc domain, wherein the CH3 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to the respective wild-type CH3 domain. In some embodiments, the antibody constructs herein comprise an Fc domain based on a modified IgG Fc domain, wherein the CH2 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to the respective wild-type CH2 domain. In some embodiments, the antibody constructs herein comprise an Fc domain based on a modified IgG Fc domain, wherein the CH3 domain and CH2 domain of at least one Fc polypeptide comprise one or more amino acid modifications compared to the respective wild-type CH3 and CH2 domains. In various embodiments, both Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in their CH3 domain. In some embodiments, both Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in their CH2 domain. In still other embodiments, both Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in both their CH2 domain and CH3 domain.

[0140] Modified Fc domain In some embodiments, the present disclosure relates to antibody constructs that may include heterodimeric immunoglobulin Fc domains comprising modified heterodimeric CH3 domains, wherein the modified heterodimeric CH3 domains comprise one or more asymmetric amino acid modifications, i.e., one or both of the first and second Fc polypeptides each comprise one or more amino acid modifications in their CH3 domain sequences compared to their respective wild-type sequences. As used herein, the term "asymmetric amino acid modification" generally refers to a modification in which an amino acid at a particular position on a first Fc polypeptide differs from the amino acid at the corresponding position on a second Fc polypeptide. These asymmetric amino acid modifications may include modification of only one of the two amino acids at the corresponding positions on each Fc polypeptide, or may include modification of both amino acids at the corresponding positions on each of the first and second Fc polypeptides. In various embodiments, an "asymmetric amino acid modification" is an asymmetric amino acid substitution.

[0141] In some embodiments, the antibody constructs herein comprise a heterodimeric Fc domain (i.e., a heterodimeric CH3 domain consisting of two CH3 domain sequences of a first and a second Fc polypeptide) comprising a modified CH3 domain, wherein the modified CH3 domain comprises one or more asymmetric amino acid modifications that promote the formation of the heterodimeric Fc domain (e.g., pairing of a first Fc polypeptide with a second Fc polypeptide) over the formation of a corresponding homodimeric Fc domain (e.g., pairing of a first Fc polypeptide with another first Fc polypeptide). Amino acid modifications that can be made to the CH3 domain of an Fc domain to promote the formation of heterodimeric Fc domains are known in the art, and include, for example, the techniques described in WO 96 / 027011 ("knobs-into-holes"), Gunasekaran et al., 2010, J Biol Chem, 285, 19637-46 ("electrostatic steering"), Davis et al., 2010, Prot Eng Des Sel, 23(4):195-202 (strand-exchange engineered domain (SEED) technology), and Labrijn et al., 2013, Proc Natl Acad Sci USA, 110(13):5145-50 (Fab arm exchange). Other examples include approaches that combine positive and negative design strategies resulting in asymmetrically modified stable Fc regions, as described in International Publication Nos. WO2012 / 058768 and WO2013 / 063702.

[0142] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain comprising a modified heterodimeric CH3 domain, wherein at least one, or both, of the Fc polypeptide chains comprises one or more amino acid modifications as described in International Publication No. WO2012 / 058768 or International Patent Publication No. WO2013 / 063702.

[0143] In some embodiments, the antibody constructs herein comprise a heterodimeric human IgG1 Fc domain with a modified CH3 domain. Table 2 herein shows full-length human IgG1 heavy chains (e.g., VH , C H1 The present invention provides an amino acid sequence of a human IgG1 Fc domain sequence (e.g., a sequence from which a first and / or second Fc polypeptide may be derived) identified in SEQ ID NO: 1, corresponding to amino acids 231 to 447 of the full-length human IgG1 heavy chain (including the hinge, hinge, CH2 domain, and CH3 domain). The CH2 domain is typically defined as comprising amino acids 231 to 340 of the full-length human IgG1 heavy chain, and the CH3 domain is typically defined as comprising amino acids 341 to 447 of the full-length human IgG1 heavy chain.

[0144] As described herein, an antibody construct can comprise a heterodimeric Fc domain having a modified CH3 domain comprising one or more asymmetric amino acid modifications that promote the formation of a heterodimeric Fc domain over the formation of a homodimeric Fc domain, wherein the modified CH3 domain comprises a first Fc polypeptide comprising amino acid modifications at positions F405 and Y407 relative to SEQ ID NO: 1, and a second Fc polypeptide comprising amino acid modifications at positions T366 and T394 relative to SEQ ID NO: 1. In various embodiments, the one or more amino acid modifications comprise one or more amino acid substitutions. Thus, in some embodiments, the amino acid modification at position F405 of the first Fc polypeptide of the modified CH3 domain is F405A, F405I, F405M, F405S, F405T, or F405V. In some embodiments, the amino acid modification at position Y407 of the first Fc polypeptide of the modified CH3 domain is Y407I or Y407V. In some embodiments, the amino acid modification at position T366 of the second Fc polypeptide of the modified CH3 domain is T366I, T366L, or T366M. In some embodiments, the amino acid modification at position T394 of the second Fc polypeptide of the modified CH3 domain is T394W. In some embodiments, the modified CH3 domain of the first Fc polypeptide further comprises an amino acid modification at position L351 relative to SEQ ID NO: 1. In some embodiments, the amino acid modification at position L351 of the first Fc polypeptide of the modified CH3 domain is L351Y. In some embodiments, the second Fc polypeptide of the modified CH3 domain further comprises an amino acid modification at position K392 relative to SEQ ID NO: 1. In some embodiments, the amino acid modification at position K392 of the second Fc polypeptide of the modified CH3 domain is K392F, K392L, or K392M. In some embodiments, one or both of the first and second Fc polypeptides of the modified CH3 domain further comprises the amino acid modification T350V. [Table 2]

[0145] In certain embodiments, the antibody constructs herein comprise heterodimeric Fc domains having modified CH3 domains comprising one or more asymmetric amino acid modifications that promote the formation of heterodimeric Fc domains over the formation of homodimeric Fc domains, wherein the modified CH3 domains comprise a first Fc polypeptide comprising the amino acid modifications F405A, F405I, F405M, F405S, F405T, or F405V together with the amino acid modification Y407I or Y407V relative to SEQ ID NO: 1, and a second Fc polypeptide comprising the amino acid modifications T366I, T366L, or T366M together with the amino acid modification T394W relative to SEQ ID NO: 1. In some embodiments, the first Fc polypeptide of the modified CH3 domain further comprises the amino acid modification L351Y. In some embodiments, the second Fc polypeptide of the modified CH3 domain further comprises the amino acid modification K392F, K392L, or K392M. In some embodiments, one or both of the first and second Fc polypeptides having a modified CH3 domain further comprise the amino acid modification T350V.

[0146] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain comprising a modified CH3 domain with a first Fc polypeptide comprising amino acid modifications at positions F405 and Y407, and optionally further comprising an amino acid modification at position L351, relative to SEQ ID NO: 1, and a second Fc polypeptide comprising amino acid modifications at positions T366 and T394, and optionally further comprising an amino acid modification at position K392, wherein the first Fc polypeptide comprises an amino acid modification at one of positions S400 or Q347 or and / or the second Fc polypeptide further comprises an amino acid modification at one or both of positions K360 or N390, wherein the amino acid modification at position S400 is S400E, S400D, S400R, or S400K, the amino acid modification at position Q347 is Q347R, Q347E, or Q347K, the amino acid modification at position K360 is K360D or K360E, and the amino acid modification at position N390 is N390R, N390K, or N390D.

[0147] In some embodiments, the antibody construct comprises a heterodimeric Fc domain comprising a modified CH3 domain comprising any one of variant 1, variant 2, variant 3, variant 4, or variant 5 modifications as shown in Table 2.

[0148] In various embodiments, antibody constructs of the present disclosure may comprise a heterodimeric Fc domain comprising a first Fc polypeptide (A) and a second Fc polypeptide (B), wherein the first Fc polypeptide (A) and the second Fc polypeptide (B) comprise an amino acid substitution in the CH3 domain described in variant #1 as shown in Table 2. In other embodiments, antibody constructs of the present disclosure may comprise a heterodimeric Fc domain comprising a first Fc polypeptide (A) and a second Fc polypeptide (B), wherein the first Fc polypeptide (A) and the second Fc polypeptide (B) comprise an amino acid substitution in the CH3 domain described in variant #2 as shown in Table 2. In some embodiments, antibody constructs of the present disclosure may comprise a heterodimeric Fc domain comprising a first Fc polypeptide (A) and a second Fc polypeptide (B), wherein the first Fc polypeptide (A) and the second Fc polypeptide (B) comprise an amino acid substitution in the CH3 domain described in variant #3 as shown in Table 2. In some embodiments, antibody constructs of the present disclosure may comprise a heterodimeric Fc domain comprising a first Fc polypeptide (A) and a second Fc polypeptide (B), wherein the first Fc polypeptide (A) and the second Fc polypeptide (B) comprise an amino acid substitution in the CH3 domain described in variant #4, as shown in Table 2. In yet other embodiments, antibody constructs of the present disclosure may comprise a heterodimeric Fc domain comprising a first Fc polypeptide (A) and a second Fc polypeptide (B), wherein the first Fc polypeptide (A) and the second Fc polypeptide (B) comprise an amino acid substitution in the CH3 domain described in variant #5, as shown in Table 2.

[0149] In certain embodiments, the CH3 domain of the first Fc polypeptide of the antibody construct herein has an amino acid sequence at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identical to the amino acid sequence set forth in SEQ ID NO: 109. In certain embodiments, the CH3 domain of the second Fc polypeptide of the antibody construct herein has an amino acid sequence at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identical to the amino acid sequence set forth in SEQ ID NO: 113. In some embodiments, the CH3 domain of the first Fc polypeptide of the antibody construct herein has the amino acid sequence set forth in SEQ ID NO: 109 and the second Fc polypeptide of the antibody construct herein has the amino acid sequence set forth in SEQ ID NO: 113.

[0150] It should be noted that the descriptors "first Fc polypeptide" and "second Fc polypeptide" in this section and throughout this disclosure may generally be used interchangeably to refer to both Fc polypeptides of an Fc domain, unless otherwise specified. Such descriptors are generally used herein to distinguish between the two Fc polypeptides of an Fc domain and are not intended to limit one particular Fc polypeptide to a particular set of asymmetric mutations.

[0151] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain based on an IgG Fc domain with a modified CH2 domain (i.e., a heterodimeric CH2 domain consisting of two CH2 domain sequences from each of the first and second Fc polypeptides). In some embodiments, the antibody constructs comprise an Fc domain based on an IgG Fc domain with a modified CH2 domain, wherein the modification(s) in the CH2 domain result in altered (e.g., reduced or abolished) binding to one or more Fc receptors (FcR), such as receptors of the FcγRI, FcγRII, and FcγRIII subclasses.

[0152] Several amino acid modifications to the CH2 domain of the first and / or second Fc polypeptide(s) of an Fc domain that selectively alter the affinity of such Fc domains for different Fcγ receptors are known in the art. Both amino acid modifications that increase binding and amino acid modifications that decrease binding may be useful in certain indications. For example, increasing the binding affinity of an Fc to FcγRIIIa (an activating receptor) can improve antibody-dependent cellular cytotoxicity (ADCC), thereby increasing target cell lysis. Reduced binding to FcγRIIb (an inhibitory receptor) may also be beneficial in some situations. In certain indications, reducing or eliminating ADCC and complement-dependent cytotoxicity (CDC) may be desirable. In such embodiments, modified CH2 domains containing amino acid modifications that increase binding to FcγRIIb or modified CH2 domains containing amino acid modifications that can reduce or eliminate binding of the Fc region to all Fcγ receptors ("knockout" variants) may be useful.

[0153] Among the amino acid modifications to the CH2 domain, non-limiting examples of amino acid modifications that alter binding to the Fc domain by Fcγ receptors include S298A / E333A / K334A and S298A / E333A / K334A / K326A (improved affinity for FcγRIIIa) (Lu, et al., 2011, J Immunol Methods, 365(1-2):132-41), F243L / R292P / Y300L / V305I / P396L (improved affinity for FcγRIIIa) (Stavenhagen, et al., 2007, Cancer Res, 67(18):8882-90), and F243L / R292P / Y300L / L235V / P396L (improved affinity for FcγRIIIa) (Nordstrom JL, et al. al., 2011, Breast Cancer Res, 13(6):R123), F243L (improved affinity for FcγRIIIa) (Stewart, et al., 2011, Protein Eng Des Sel., 24(9):671-8), S298A / E333A / K334A (improved affinity for FcγRIIIa) (Shields, et al., 2001, J Biol Chem, 276(9):6591-604), S239D / I332E / A330L and S239D / I332E (improved affinity for FcγRIIIa) (Lazar, et al., 2006, Proc Natl Acad Sci USA, 103(11):4005-10), and S239D / S267E and S267E / L328F (improved affinity for FcγRIIb) (Chu, et al., 2008, Mol Immunol, 45(15):3926-33). Additional modifications that affect Fc domain binding to Fcγ receptors are described in Therapeutic Antibody Engineering (Strohl & Strohl, Woodhead Publishing series in Biomedicine No 11, ISBN 1 907568 37 9, October 2012, page 283).

[0154] In various embodiments, the antibody constructs of the present disclosure comprise a heterodimeric Fc domain based on an IgG Fc domain with a modified CH2 domain that comprises one or more amino acid modifications that can reduce or abolish binding of the Fc domain to one or more or all Fcγ receptors (i.e., a "knockout" or "KO" variant).

[0155] Various publications describe strategies that have been used to engineer antibodies to generate "knockout" Fc variants (see, e.g., Strohl, 2009, Curr Opin Biotech 20:685-691, and Strohl & Strohl, "Antibody Fc engineering for optimal antibody performance" In Therapeutic Antibody Engineering, Cambridge: Woodhead Publishing, 2012, pp 225-249). These strategies include modifying glycosylation, using an IgG2 / IgG4 scaffold, or reducing effector function by introducing mutations in the hinge or CH2 domain of the Fc (see also U.S. Patent Publication No. 2011 / 0212087, International Publication No. WO2006 / 105338, U.S. Patent Publication No. 2012 / 0225058, U.S. Patent Publication No. 2012 / 0251531, and Strop et al., 2012, J. Mol. Biol., 420:204-219).

[0156] In some embodiments, the Fc domain of the antibody construct may include one or more known amino acid modifications to reduce FcgR and / or complement binding of the Fc domain. In some embodiments, such modifications may include those identified in Table 3. [Table 3]

[0157] Additional examples herein include Fc domains modified to include the amino acid modifications L235A / L236A / D265S, e.g., based on the sequence set forth in SEQ ID NO: 1. Additionally, asymmetric amino acid modifications in the CH2 domain that reduce Fc binding to all Fcγ receptors are described in International Publication No. WO2014 / 190441.

[0158] In certain embodiments, the CH2 domain of the first and second Fc polypeptides herein comprises or consists of an amino acid sequence having at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 108. In some embodiments, the CH2 domain of the first and / or second Fc polypeptides herein comprises or consists of the sequence set forth in SEQ ID NO: 108.

[0159] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain with native glycosylation modifications. As is known in the art, altering the glycosylation of Fc can increase or decrease effector function. For example, mutating the conserved asparagine residue at position 297 to alanine, glutamine, lysine, or histidine (i.e., N297A, Q, K, or H) results in a deglycosylated Fc that lacks all effector function (Bolt et al., 1993, Eur. J. Immunol., 23:403-411; Tao & Morrison, 1989, J. Immunol., 143:2595-2601). Conversely, removal of fucose from the oligosaccharide attached to heavy chain N297 has been shown to enhance ADCC by improving binding to FcγRIIIa (see, e.g., Shields et al., 2002, J. Biol. Chem., 277:26733-26740, and Niwa et al., 2005, J. Immunol. Methods, 306:151-160). Such low-fucose antibody constructs can be produced, for example, in knockout Chinese hamster ovary (CHO) cells lacking fucosyltransferase (FUT8) (Yamane-Ohnuki et al., 2004, Biotechnol. Bioeng., 87:614-622), in the variant CHO cell line Lec13, which has a reduced ability to attach fucose to the glycan attached to N297 (International Publication No. WO 03 / 035835), or in other cells that produce defucosylated antibodies (e.g., Li et al., 2006, Nat Biotechnol., 24:210-215; Shields et al., 2002, ibid., and Shinkawa et al., 2003, J. Biol. Chem., 278:3466-3473). Furthermore, International Publication No. WO2009 / 135181 describes the addition of a fucose analog to the culture medium during antibody production to inhibit the incorporation of fucose into the sugar chains on the antibody.

[0160] In certain embodiments, the trivalent bispecific antibody constructs of the present disclosure are capable of binding to fetal Fc receptors (eg, FcRn).

[0161] Linker As described herein, the antibody constructs of the present disclosure may include one or more linkers. In various embodiments, such one or more linkers are peptide (or peptidic) linkers comprising or consisting of an amino acid sequence of 1, 2, 3, 5, 10, 15, 20, 25, 30, or more consecutive amino acid residues. Such peptide linkers can link or connect two or more peptides or polypeptides to each other. Thus, in various embodiments, the linker herein is a first polypeptide sequence, e.g., a heavy chain constant domain (C H1 ) can be linked to a second polypeptide sequence, e.g., an Fc polypeptide. Thus, in some embodiments, a linker is used to link one domain of an antibody construct to another, e.g., a linker Fab-Fc For example, a Fab domain can be linked to an Fc domain via, for example, a linker scFv-Fab The scFv domain can be linked to the Fab domain via, for example, a linker scFv via V H Domain V L It can be concatenated to a domain, etc.

[0162] The linker may be, for example, a heavy chain variable domain (V H ) to, for example, a light chain variable domain (V L ), the linker may be of sufficient length to allow both domains to exert their biological functions. In addition to providing a spacer function, the linker (e.g., a peptide linker) herein may provide suitable flexibility or rigidity to properly orient one or more domains of the antibody construct, both within the antibody construct itself and between the antibody construct and its biological target(s) (e.g., T cell and / or cancer antigens).

[0163] Furthermore, the linkers (e.g., peptide linkers) herein can (i) support expression of full-length fusion proteins, e.g., full-length polypeptide chains H1, L1, H2, etc., of an antibody construct, and (ii) increase the stability of the purified protein both in vitro and in vivo after administration to a subject in need thereof, e.g., a human or rodent. One or more linkers used in the antibody constructs herein are generally non-immunogenic or have low immunogenicity in the mammalian subject to which the antibody construct may be administered. In certain embodiments, one or more linkers used in the antibody constructs herein may comprise part or all of a human immunoglobulin hinge region, a stalk region of a C-type lectin, a type II membrane protein family, or a combination thereof. In certain embodiments, one or more linkers used in the antibody constructs herein may comprise a hinge region of a human immunoglobulin, e.g., a hinge region of an IgG or IgG1, e.g., a linker Fab-Fc or linker scFv-Fc may include some or all of the above.

[0164] In certain embodiments, each linker of one or more linkers used in the antibody constructs herein can comprise or consist of an amino acid sequence having a length of from 2 to about 50 amino acids. In some embodiments, each linker of one or more linkers used in the antibody constructs herein can comprise or consist of an amino acid sequence having a length of from about 3 to about 40 amino acids, from about 10 to about 50 amino acids, from about 2 to about 40 amino acids, from about 5 to about 30 amino acids, from about 5 to about 25 amino acids, from about 4 to about 30 amino acids, from about 10 to about 30 amino acids, or from about 15 to about 25 amino acids. In some embodiments, one or more linkers of an antibody construct may each comprise an amino acid sequence comprising or consisting of 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, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 consecutive amino acids.

[0165] In certain embodiments, the linker of the antibody constructs herein has the amino acid sequence (EAAAK) n wherein n is an integer from 1 to 5 (SEQ ID NO: 133). In some embodiments, the linker comprises or consists of the sequence EAAAK (SEQ ID NO: 134). In some embodiments, the linker comprises or consists of the sequence EAAAKEAAAK (SEQ ID NO: 135). In some embodiments, the linker comprises a polyproline linker, e.g., having the amino acid sequence of PPP (SEQ ID NO: 136) or PPPP (SEQ ID NO: 137). In certain embodiments, the linker is a glycine (G)-proline (P) polypeptide linker, e.g., comprising or consisting of one or more of GPPPG (SEQ ID NO: 138), GGPPPGG (SEQ ID NO: 139), GPPPPG (SEQ ID NO: 140), or GGPPPPGG (SEQ ID NO: 141). In some embodiments, the linker herein is n Ser) m is a linker, and n and m are independently integers from 1 to 5 (SEQ ID NO: 142). In certain embodiments, the linker is (Gly3Ser) n (Gly4Ser)1 (SEQ ID NO: 143), (Gly3Ser)1 (Gly4Ser) n (SEQ ID NO: 144), or (Gly3Ser) n (Gly4Ser) n (SEQ ID NO: 145), wherein each n is independently an integer from 1 to 5. In certain embodiments, the linker herein is suitable for connecting two different domains of an antibody construct, for example, (G m S) n -GG (SEQ ID NO: 146), (SG n ) m (SEQ ID NO: 147), or (SEG n ) m (SEQ ID NO: 148), wherein m and n are independently integers of 0 to 20, 1 to 10, or 1 to 5.

[0166] Thus, in some embodiments, the antibody constructs described herein may contain one or more linkers, such as the amino acid sequence (Gly4Ser) n wherein n is an integer from 1 to 5 (SEQ ID NO: 132), 1 to 4, or 1 to 3. In such embodiments, n can be 1 (SEQ ID NO: 104). In other embodiments, n is 3 (SEQ ID NO: 118). In yet other embodiments, n is 4 (SEQ ID NO: 102).

[0167] In certain embodiments herein, the antibody construct comprises one or more linkers whose amino acid sequence is obtained, derived, or designed from the sequence of an antibody hinge region. In some embodiments, such linkers may have at least one cysteine capable of participating in or forming at least one disulfide bond under physiological conditions or other standard peptide conditions (e.g., peptide purification conditions, peptide storage conditions). In certain embodiments, linkers that correspond to or resemble immunoglobulin hinge peptides retain a cysteine corresponding to the hinge cysteine located toward the amino (i.e., N) terminus of the hinge. In further embodiments, the linker is derived from an IgG1 hinge and may be modified to remove any cysteine residues, or the linker is an IgG1 hinge with one or two cysteines corresponding to the hinge cysteines.

[0168] In certain embodiments, the linker of the antibody constructs described herein may comprise a "modified wild-type immunoglobulin hinge region" or "modified immunoglobulin hinge region." Such an altered hinge region may be a wild-type immunoglobulin hinge region that has (a) up to 30 percent amino acid changes (e.g., up to 25 percent, 20 percent, 15 percent, 10 percent, or 5 percent amino acid substitutions, insertions, and / or deletions), (b) a wild-type immunoglobulin hinge region that has up to 30 percent amino acid changes (e.g., up to 25 percent, 20 percent, 15 percent, 10 percent, or 5 percent amino acid substitutions, insertions, and / or deletions), or (c) a wild-type immunoglobulin hinge region that has up to 30 percent amino acid changes (e.g., up to 25 percent, 20 percent, 15 percent, 10 percent, or 5 percent amino acid substitutions, insertions, and / or deletions). (c) a portion of a wild-type immunoglobulin hinge region that is at least 10 amino acids in length (e.g., at least 12, 13, 14, or 15 amino acids), (d) a portion of a wild-type immunoglobulin hinge region that includes the core hinge region (a portion can be 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15, or at least 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 amino acids in length), or (d) a combination of any of (a)-(c). In certain embodiments, one or more cysteine residues in a wild-type immunoglobulin hinge region, such as an IgG1 hinge, including the upper and core regions, can be substituted with one or more other amino acid residues (e.g., one or more serine residues). Alternatively or additionally, the altered immunoglobulin hinge region may substitute proline residues of a wild-type immunoglobulin hinge region, such as an IgG1 hinge, including the upper and core regions, with another amino acid residue (e.g., serine residues).

[0169] Thus, in some embodiments, the antibody constructs of the present disclosure comprise a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 107. In some embodiments, the antibody constructs comprise a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 111. In some embodiments, the antibody constructs comprise a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 112.

[0170] In various embodiments, the antibody constructs of the present disclosure comprise a linker that links the Fab domain to the first Fc polypeptide. Fab-Fc (Linker Fab-Fc comprises or consists of the amino acid sequence set forth in SEQ ID NO: 107), and a linker which is another linker connecting the first scFv domain to the second Fc polypeptide. scFv-Fc (Linker scFv-Fc comprises or consists of the amino acid sequence set forth in SEQ ID NO: 111).

[0171] In other embodiments, the antibody construct of the disclosure comprises a linker, which is a linker connecting the Fab domain to the first Fc polypeptide. Fab-Fc (Linker Fab-Fc comprises or consists of the amino acid sequence set forth in SEQ ID NO: 107), and a linker which is another linker connecting the first scFv domain to the second Fc polypeptide. scFv-Fc (Linker scFv-Fc comprises or consists of the amino acid sequence set forth in SEQ ID NO: 112).

[0172] Binding domains for tumor-associated antigens As described herein, the antibody constructs of the present disclosure may comprise one or more binding domains capable of binding to one or more tumor-associated antigens (TAA). In various embodiments, the antibody constructs herein are trivalent and bispecific, comprising a first binding domain, a second binding domain, and a third binding domain, wherein one or more of such binding domains are capable of binding to a tumor-associated antigen (TAA). In various embodiments, two of such binding domains are capable of binding to a TAA. A TAA may be any antigenic substance expressed on the surface of a tumor cell.

[0173] Thus, in embodiments in which at least two binding domains of the antibody construct herein can bind to a TAA, both binding domains can bind to different TAAs. In various other embodiments, both anti-TAA binding domains can bind to the same TAA, i.e., the antibody construct is bivalent for that TAA. As further described herein, in such embodiments, both anti-TAA binding domains can bind to the same epitope of the TAA (e.g., bivalent binding of two TAA molecules (e.g., two MSLN molecules) to the same epitope is possible). In other embodiments, both anti-TAA binding domains may be capable of binding to two different epitopes of the same TAA.

[0174] As described herein, in various embodiments, the two anti-TAA binding domains of an antibody construct may be capable of binding to MSLN. In various embodiments, the anti-MSLN binding domains of such a construct are scFv domains and can bind to the same epitope of MSLN, either on the same MSLN protein or on two different MSLN proteins.

[0175] In some embodiments, the present disclosure relates to antibody constructs in which both anti-MSLN binding domains can have the same or similar (i.e., within about ±5%) binding affinity for MSLN, or both domains can have different binding affinities for MSLN. Such different binding affinities can be achieved, for example, by EC 50 When measured as a value, it may differ by at least about 2%, 5%, or 10%.

[0176] Thus, in various embodiments, at least one anti-MSLN binding domain of the construct herein may have a binding affinity to MSLN of about 0.5 nM, 0.6 nM, 0.7 nM, 0.8 nM, 0.9 nM, 1 nM, 2 nM, 3 nM, 4 nM, 5 nM, or about 10 nM. In some embodiments, both anti-MSLN binding domains may have a binding affinity to MSLN of about 0.5 nM, 0.6 nM, 0.7 nM, 0.8 nM, 0.9 nM, 1 nM, 2 nM, 3 nM, 4 nM, 5 nM, or about 10 nM. In some embodiments, at least one or both binding domain(s) to MSLN has an affinity to MSLN of about 3 nM. In some embodiments, at least one or both binding domain(s) for MSLN have an affinity for MSLN of about 0.5 nM to about 5 nM, about 1 nM to about 10 nM, about 1 nM to about 5 nM, or about 2 nM to about 4 nM. In some embodiments, both binding domains for MSLN have an affinity for MSLN of about 2 nM to about 4 nM.

[0177] As used throughout this disclosure, the term "binding affinity" in reference to an antibody construct or portion thereof (e.g., a binding domain) herein for binding to a particular antigen (e.g., CD3, MSLN, etc.) generally refers to the dissociation constant (K) measured using one or more binding assays known in the art, such as SPR, FACS, etc. D ) value. D (molar concentration "M", given in units of mM, μM, nM, pM, etc.) refers to the dissociation equilibrium constant of a particular antibody construct-antigen interaction. K DThere is an inverse relationship between the binding affinity and the K D The smaller the value, the higher (i.e., stronger) the affinity. Thus, the terms "higher affinity" or "stronger affinity" refer to a higher ability to form an interaction, and thus a higher K D Conversely, the terms "lower affinity" or "weaker affinity" refer to a lower ability to form interactions and thus a lower K D Unless otherwise indicated, the binding affinity of the antibody constructs herein to their targets, e.g., CD3, MSLN, etc., is measured by the EC 50 As used herein, the term "EC 50 " refers to the half maximal effective concentration, which includes the concentration of an antibody construct that elicits a response halfway between baseline and maximum after a specific exposure time. EC 50 The EC value essentially represents the concentration of the antibody construct at which 50% of its maximal effect is observed. 50 The value generally equals the concentration of an antibody construct of the present disclosure that gives half-maximal binding to cells expressing CD3 or a tumor-associated antigen, e.g., MSLN, as measured, e.g., by SPR or FACS binding assays.

[0178] As described herein, the two anti-MSLN binding domains of the antibody constructs herein can each comprise or consist of an scFv domain. In various embodiments, the two anti-MSLN binding domains each consist of an scFv domain, i.e., a first scFv domain and a second scFv domain, each of which has a V H Domain and V L Domain and V L The C-terminus of the domain is V H at the N-terminus of the domain, or V H The C-terminus of the domain is V Land a linker connecting the N-terminus of the domain. In various embodiments, the two anti-MSLN scFv domains comprise or consist of the same amino acid sequence. In such embodiments, the two anti-MSLN binding domains bind to the same epitope on MSLN and each comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L In these embodiments, both anti-MSLN scFv domains may have the same or similar binding affinity for MSLN.

[0179] In some embodiments, the CDRs of the anti-MSLN paratope comprise one or more amino acid modifications in one or more of the CDR sequences set forth in SEQ ID NOs: 120-125, wherein at least about 80%, 90%, or 95% of the binding affinity to MSLN is retained compared to the paratope without the amino acid modifications.

[0180] In some embodiments, the two anti-MSLN binding domains are scFv domains (e.g., first and second scFv domains) that comprise or consist of amino acid sequences shared with about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity.

[0181] In such embodiments, both anti-MSLN binding domains, i.e., the first scFv domain and the second scFv domain, comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 103. HIn certain embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 103. H In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 103. H In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 103. H In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 103. H In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 103. H It may include a domain.

[0182] In a further embodiment, both anti-MSLN binding domains, i.e., the first scFv domain and the second scFv domain, comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 102. L In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 102. L In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 102. LIn some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 102. L In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 102. L In some embodiments, both anti-MSLN binding domains may comprise a V domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 102. L It may include a domain.

[0183] Thus, in various embodiments, an antibody construct of the present disclosure comprises a first scFv domain and a second scFv domain, both capable of binding to the same epitope on MSLN, and wherein the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In some embodiments, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In some embodiments, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In some embodiments, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In some embodiments, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In some embodiments, the first scFv domain and the second scFv domain each comprise or consist of an amino acid sequence having about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 149. In various embodiments, the first scFv domain and the second scFv domain each comprise or consist of the amino acid sequence set forth in SEQ ID NO: 149.

[0184] In some embodiments, the first scFv domain and the second scFv domain (both capable of binding to MSLN) of the antibody construct herein each consist of the amino acid sequence set forth in SEQ ID NO: 149.

[0185] Binding domain for antigen on cytotoxic effector cells As further described herein, the antibody construct of the present disclosure may comprise at least one binding domain capable of binding to a molecule, e.g., a polypeptide, on the surface of a cytotoxic effector cell. Such a cytotoxic effector cell may be an immune cell. The immune cell may be a T cell, macrophage, dendritic cell, neutrophil, B cell, or NK cell. In various embodiments, the antibody construct of the present disclosure comprises at least one binding domain capable of binding to an antigen on a T cell. In various embodiments, such an antigen is CD3.

[0186] In some embodiments, the present disclosure relates to a trivalent, bispecific antibody construct comprising a first binding domain capable of binding to CD3, and a second and third binding domain, both capable of binding to MSLN. The antibody constructs described herein are also referred to as "T cell engagers," "TCEs," or "T cell engager molecules," which refer to their ability to bind to both CD3 on T cells and MSLN on tumor cells, thereby targeting T cell-mediated cytotoxic activity to the tumor environment, including cells expressing MSLN.

[0187] In various embodiments, the anti-CD3 binding domain of the antibody constructs herein has an EC 50 The affinity for CD3, given as a value, can be about 1 nM, 5 nM, 10 nM, 20 nM or less, or about 30 nM or less. In various embodiments, the anti-CD3 binding domains herein have an EC 50 The affinity for CD3, given as a value, can be about 20 nM to about 80 nM, about 30 nM to about 60 nM, or about 40 nM to about 50 nM. In some embodiments, the anti-CD3 binding domains herein have an EC 50In various embodiments, the antibody constructs herein have an EC50 of about 20 nM to about 40 nM, e.g., about 30 nM, for binding to CD3. 50 The antibody comprises an anti-CD3 binding domain having a value.

[0188] In various embodiments, the antibody constructs herein comprise a Fab domain capable of binding to CD3 on a T cell, and such a Fab domain is capable of binding to an EC 50 a VH1 / VH2 value of about 20 nM to about 40 nM, for example about 30 nM, and comprising an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 128; H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L Includes the domain.

[0189] In some embodiments, the CDRs of the anti-CD3 paratopes herein comprise one or more amino acid modifications in one or more of the CDR sequences set forth in SEQ ID NOs: 126-131, wherein at least about 80%, 90%, or 95% of the binding affinity to CD3 is retained compared to the paratope without the amino acid modifications.

[0190] In certain embodiments, as further described herein, the anti-CD3 Fab domain of the antibody construct comprises V H Domain and C H1 The heavy chain portion containing the V L Domain and C L It may comprise or consist of a light chain comprising a domain.

[0191] V of the anti-CD3 Fab domain of the antibody construct herein H The V domain may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. In some embodiments, the V of such a Fab domain H In some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. H In some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. H In some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. H In yet another embodiment, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. H The domain comprises or consists of the amino acid sequence set forth in SEQ ID NO:105.

[0192] V of the anti-CD3 Fab domain of the antibody construct herein L The V domain may be a portion of a light chain (e.g., L1) that pairs with the anti-CD3 domain of the heavy chain and may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. In some embodiments, the V domain of such a Fab domain may be a portion of a light chain (e.g., L1) that pairs with the anti-CD3 domain of the heavy chain and may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. L In some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. LIn some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. L In some embodiments, the V of such a Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. L The V domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. In various embodiments, the V of such a Fab domain L The domain comprises or consists of the amino acid sequence set forth in SEQ ID NO:115.

[0193] C of the anti-CD3 Fab domain of the antibody construct herein H1 The C domain may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. In some embodiments, the C domain of such a Fab domain may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1 In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1 In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1 In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1 In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1The domain comprises or consists of the amino acid sequence set forth in SEQ ID NO:106.

[0194] C of the anti-CD3 Fab domain of the antibody construct herein L The C domain may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. In some embodiments, the C domain of such a Fab domain may comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. L In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. L In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. L In some embodiments, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. L In yet another embodiment, the C of such a Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116. L The domain comprises or consists of the amino acid sequence set forth in SEQ ID NO:116.

[0195] Thus, in various embodiments, the antibody constructs herein comprise (i) a VV comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105. H (ii) a V domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115. L(iii) a C domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106. H1 domain, and (iv) an anti-CD3 Fab domain comprising a CL domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116.

[0196] In certain embodiments, the antibody constructs herein comprise (i) a V H (ii) a V domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 115; L (iii) a C domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 106; H1 domain, and (iv) an anti-CD3 Fab domain comprising a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO:116.

[0197] In various embodiments, as further described herein, the anti-CD3 Fab domain of the antibody construct comprises V H Domain and C H1 The portion of the heavy chain (e.g., H1) containing the V domain L In some embodiments, a heavy chain H1, a portion of which can form an anti-CD3 Fab domain, has the amino acid sequence set forth in SEQ ID NO: 100, and the corresponding light chain L1, which pairs with the anti-CD3 portion of the heavy chain to form the anti-CD3 Fab domain, has the amino acid sequence set forth in SEQ ID NO: 114.

[0198] C. Characterization of Antibody Constructs The trivalent, bispecific antibody constructs of the present disclosure may possess one or more properties that may make such antibody constructs superior when tested in vivo and / or in vitro and when compared to other known constructs currently used, for example, for the treatment of cancer, and with respect to anti-tumor activity and occurrence of side effects when administered to a subject in need thereof.

[0199] As described herein, the antibody constructs of the present disclosure may be capable of binding to CD3 on a cytotoxic effector cell (e.g., a T cell) and to MSLN located on a tumor cell. In some embodiments, such binding occurs simultaneously, e.g., the antibody construct can simultaneously bind to both CD3 on the cytotoxic effector cell and MSLN on the tumor cell. In various embodiments, the cytotoxic effector cell is an immune cell such as a T cell, and the tumor cell is an MSLN-expressing solid tumor cell.

[0200] In various embodiments, the antibody constructs described herein are engineered and designed to have a specific format (e.g., 2 (scFv) + 1 (Fab) for MSLN and CD3, respectively) and geometry that enables binding of MSLN driven by the avidity of two anti-MSLN scFv domains, each located on a different arm of the antibody construct. Furthermore, extensive characterization studies have demonstrated that when an anti-MSLN paratope with moderate, single-digit nanomolar affinity for MSLN (e.g., about 2 nM to about 4 nM) is used in combination with the antibody format, surprisingly potent T cell-mediated antitumor activity was measured in the presence of tumor cells with moderate to high MSLN expression, whereas reduced antitumor activity was observed in settings containing tumor cells with low MSLN expression. This superior TAA-dependent antitumor activity appears to be unique to the constructs described herein and was not observed with benchmark constructs tested with different formats and / or geometries. Furthermore, we found that combining such MSLN-binding properties with a single anti-CD3 binding domain with a moderate to low affinity for CD3 in the double-digit nanomolar range, e.g., about 20 nM to about 40 nM, provided the most favorable balance between T cell-mediated antitumor activity and selectivity for tumor cells that moderately or highly express MSLN compared with tissues that express MSLN poorly. Such properties may enhance immune cell-induced cytotoxicity, for example, at target sites where tumor cells expressing a TAA (e.g., MSLN) are present, compared with other organs or tissues where highly MSLN-expressing cells are absent or present in significantly reduced numbers. Thus, the trivalent bispecific antibody constructs described herein, comprising one anti-CD3 Fab binding domain and two anti-MSLN scFv binding domains, with both anti-MSLN scFv domains located in the N-terminal region of the antibody construct (e.g., at the N-terminus of the two heavy chains), may also target both CD3 and MSLN, but may provide superior in vivo anti-tumor activity and tolerability when compared to existing benchmark molecules with different molecular formats and / or geometries.

[0201] Thus, the antibody constructs of the present disclosure can be characterized as trivalent and bispecific T cell binding molecules that, upon binding to CD3 on immune cells and MSLN on tumor cells, can form a TCR-independent artificial immune synapse between immune cells, e.g., T cells, and tumor cells, which can activate the T cells and exert a cytotoxic effect on tumor cells, resulting in tumor cell lysis, as further described herein.

[0202] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains, when present in a cell population comprising CD3-expressing immune cells and MSLN-expressing tumor cells and at a concentration of about 10 pM, may be capable of one or more of: A) inducing production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL of TNFα; B) inducing production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL of IFNγ; and / or C) inducing production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL of IL-2. In some embodiments, the antibody construct comprises two or more of properties A, B, and C. In yet other embodiments, the antibody construct comprises all of properties A, B, and C. In some of these embodiments, the immune cells may include T cells and the tumor cells may express at least about 500,000 MSLNs per cell.

[0203] In various embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may have a binding affinity to MSLN of about 0.7 nM, 0.8 nM, 0.9 nM, or about 1 nM as measured by surface plasmon resonance (SPR). In some embodiments, such a trivalent bispecific antibody construct of the present disclosure may further have a binding affinity to CD3 of about 30 nM, 40 nM, 50 nM, or 60 nM, or between about 30 nM and about 60 nM.

[0204] Thus, in some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains, with the domain structure scFv2-Fab-CH21-CH31 in the first heavy chain and scFv1-CH22-CH32 in the second heavy chain, and comprising (i) an anti-MSLN paratope with the HCDRs and LCDRs set forth in SEQ ID NOs: 120-125, respectively, and an anti-CD3 paratope with the HCDRs and LCDRs set forth in SEQ ID NOs: 126-131, respectively, may have a binding affinity of about 0.7 nM, 0.8 nM, 0.9 nM, or about 1 nM to MSLN and a binding affinity of up to about 30 nM, 40 nM, 50 nM, or 60 nM to CD3, as measured by SPR. The relative orientation of the binding domains of such an antibody construct may be related to the binding affinity of CD3. + T cells and MSLNs + The constructs described herein can provide specific avidity-based binding to both MSLN-dependent T cell activation, as measured, for example, by cytokine release from activated T cells, and / or tumor cell killing activity based on MSLN expression by tumor cells. As further described and demonstrated herein, without being bound by any theory, the use of two anti-MSLN scFvs and one anti-CD3 Fab in the configuration described herein and shown, for example, in FIG. 1A , provides a more robust and efficient binding mechanism than other previously reported constructs, for example, a corresponding Fab comprising two anti-MSLN Fabs (instead of the two anti-MSLN scFv domains used in the constructs described herein) and one anti-CD3 Fab domain. 3This disclosure, including the experimental data provided herein, demonstrates the striking and surprising impact of the format and geometry of a construct, in combination with its antigen affinity, on its in vivo properties, for example, by directly comparing trivalent bispecific constructs comprising one anti-CD3 Fab domain and two anti-MSLN scFv domains (e.g., v21812, v32523, etc.) with other constructs comprising one anti-CD3 Fab domain and two anti-MSLN Fab domains (e.g., v29191, v21791), and, in some embodiments, by comparing constructs with different geometries but identical anti-CD3 / MSLN paratope sequences to demonstrate differences entirely attributable to format.

[0205] Thus, in some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can induce inflammatory cytokine production by T cells in an MSLN-dependent manner, where the MSLN-dependent activation of cytotoxic effector cells is determined by measuring at least about a 20-fold, 50-fold, 100-fold, or at least about a 1000-fold decrease in cytokine production between a first cell population comprising first tumor cells and immune cells and a second cell population comprising second tumor cells and immune cells, where MSLN expression in the first tumor cells is at least about 3-fold, 4-fold, or 5-fold higher compared to MSLN expression in the second tumor cells.

[0206] In further embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can induce at least 5-fold, 10-fold, 15-fold, or 20-fold greater T-cell-mediated cytotoxicity against tumor cells in in vitro tests when compared under the same experimental conditions to a similar antibody construct comprising two anti-MSLN Fab domains in place of the first and second anti-MSLN scFv domains of the construct disclosed herein.

[0207] MSLN + The cytotoxic activity of the antibody constructs described herein against cell lines was measured using half-maximal inhibitory concentrations (IC 50 ) As an example, such an IC 50 The value is MSLN + It can be used as a measure of the amount of antibody construct required to reduce the number of viable cells by 50%. IC of antibody construct 50 Several methods for determining IC are known in the art and are described herein, such as the use of cell viability assays and measurements to determine the concentration at which the count of viable tumor cells is reduced by 50%. Generally, a high IC 50 indicates that a larger amount of antibody construct is required to reduce the count of viable tumor cells by, for example, 50%, and therefore the ability of the antibody construct to induce antitumor cytotoxic activity is relatively low. Conversely, a low IC 50 indicates that less antibody construct is required to reduce the count of viable tumor cells by, for example, 50%, and therefore the ability of the antibody construct to induce anti-tumor cytotoxic activity is relatively high.

[0208] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains has an IC of about 1 pM to about 0.005 pM, about 1 pM to about 0.01 pM, about 0.5 pM to about 0.005 pM, about 0.1 pM to about 0.01 pM, or about 0.1 pM to about 0.05 pM for T cell-induced tumor cell killing. 50 In certain embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may have an IC of about 1 pM to about 0.01 pM or about 0.5 pM to about 0.01 pM for T cell-induced tumor cell killing. 50 In some embodiments, the tumor cells may have MSLNs, such as H292 or OVCAR3 cells, having an average number of MSLNs per cell of about 100,000 or more, as described herein. + It is a tumor cell.

[0209] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may be capable of either (i) inhibiting tumor growth (e.g., a 5% or less increase in tumor volume) and / or (ii) reducing tumor volume in a subject by at least about 5%, 10%, 15%, 20%, 25%, 30%, 40%, or at least about 50% (the tumor comprising at least about 100,000 MSLN-expressing cells per cell) when measured over a period of at least about 20 days after administration of the antibody construct to the subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg. In various embodiments, such trivalent bispecific antibody constructs may provide a Fab 3 A significantly greater reduction in tumor volume is achieved with a construct containing the format, i.e., two anti-MSLN Fabs and one anti-CD3 Fab. As further described herein, a difference is significant if the calculated p-value is <0.05, e.g., <0.005.

[0210] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can exhibit higher anti-tumor activity in the presence of soluble MSLN compared to a 1+1 format-based bispecific construct (e.g., a construct comprising one anti-CD3 (scFv or Fab) domain and one anti-MSLN (scFv or Fab) domain). Such an experimental scenario can mimic the in vivo release of MSLN from MSLN-expressing tumor cells, allowing MSLN to circulate in the serum of subjects bearing MSLN-expressing tumors. Thus, the released MSLN may be referred to as soluble MSLN, but may also comprise or consist of the extracellular portion of membrane-bound MSLN located on tumor cells. In some of these embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may exhibit greater anti-tumor activity compared to a 1+1 format-based construct when soluble MSLN is present at a concentration of about 50 ng / mL, 100 ng / mL, or about 150 ng / mL. In some embodiments, under such conditions, the anti-tumor activity of the trivalent bispecific construct may be about 10-fold, 20-fold, 50-fold, or about 100-fold greater than the anti-tumor activity of the corresponding 1+1 format-based construct.

[0211] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can induce cytokine release from immune cells (e.g., T cells) in an MSLN-dependent manner. In such embodiments, measurable cytokine release can be detected only when the construct is present in the presence of T cells and tumor cells that express MSLN, compared to when the construct is present in the presence of T cells and tumor cells that do not express MSLN. In further embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can induce T cell proliferation in an MSLN-dependent manner; for example, in the absence of MSLN-expressing tumor cells, no measurable T cell proliferation of T cells may be detected, but in the presence of MSLN-expressing tumor cells, at least about 40%, 50%, 60%, 70%, or at least about 80% T cell proliferation may be measured.

[0212] In some embodiments, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may be tolerated in cynomolgus monkeys at doses of approximately 1 mg / kg, 10 mg / kg, and 30 mg / kg. Following intravenous administration of the trivalent bispecific antibody construct v32523 to cynomolgus monkeys at doses of 1, 10, and 30 mg / kg, a transient increase in one or more biomarkers may be observed. Such one or more biomarkers may include one or more of IL-6, monocyte chemoattractant protein 1 (MCP-1), and granulocyte-macrophage colony-stimulating factor (GM-CSF). In certain embodiments, serum IL-6 concentrations may range from approximately 150 pg / mL to approximately 250 ng / mL approximately 24 hours after injection of 10 or 30 mg / kg of an antibody construct, e.g., v32523. Other effects of administration of the trivalent bispecific antibody construct may include, for example, increases in fibrinogen levels of approximately 38%, 68%, 166%, and 142% on day 4 after injection of 0.1, 1, 10, and 30 mg / kg of the antibody construct, respectively, compared to pre-administration levels.

[0213] In a further embodiment, a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may have an in vivo serum half-life of about 2.3, 2.6, or 3 days in cynomolgus monkeys.

[0214] In various embodiments herein, the trivalent bispecific antibody construct may be construct v32523 or v21812, as further described herein.

[0215] Sequence identity As described elsewhere in this disclosure, certain embodiments herein relate to isolated polypeptides or sets of isolated polypeptides of antibody constructs (e.g., polypeptide chains H1, H2, L1, etc., or portions thereof, e.g., domains), and polynucleotides or sets of polynucleotides that encode the antibody constructs described herein. Polynucleotides in this regard may encode all or a portion of an antibody construct, for example, one or more polypeptide chains (e.g., H1, H2, L1, etc.) of an antibody construct.

[0216] The terms "nucleic acid," "nucleic acid molecule," and "polynucleotide" are used interchangeably herein to refer to a polymeric form of nucleotides of any length, either deoxyribonucleotides or ribonucleotides, or analogs thereof. Non-limiting examples of polynucleotides include genes, gene fragments, messenger RNA (mRNA), cDNA, recombinant polynucleotides, plasmids, vectors, isolated DNA of any sequence, isolated RNA of any sequence, nucleic acid probes, and primers.

[0217] A polynucleotide that "encodes" a given polypeptide is one that is transcribed (in the case of DNA) and translated into a polypeptide (in the case of mRNA) in vivo when placed under the control of appropriate regulatory sequences. The boundaries of the coding sequence are determined by a start codon at the 5' (amino) terminus and a translation stop codon at the 3' (carboxy) terminus. A transcription termination sequence may be located 3' to the coding sequence.

[0218] In certain embodiments, the present disclosure relates to polynucleotide and / or polypeptide sequences that are identical or substantially identical to another polynucleotide and / or polypeptide sequence. The term "identical" with respect to two or more polynucleotide or polypeptide sequences refers to two or more sequences or subsequences that are identical, i.e., have the same sequence of nucleotide or amino acid monomers (i.e., 100% sequence identity), respectively. Polypeptide or polynucleotide sequences herein share "sequence identity" if they have a percentage or specified number of amino acid residues or nucleotides that are at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identical over a specified region when compared and aligned for maximum correspondence over a comparison window or designated region, as determined using one of the commonly used sequence comparison algorithms known to those of skill in the art or by manual alignment and visual inspection. This definition also refers to the complement of a test polynucleotide sequence. A particular sequence identity can exist over a region that is at least about 50 amino acids or nucleotides in length, or over a region that is about 75 to about 100 amino acids or nucleotides in length, or, if not specified, over the entire polypeptide or polynucleotide sequence. For sequence comparison, a test sequence is typically compared to a designated reference sequence. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters can be used, or alternative parameters can be designated. The sequence comparison algorithm then calculates the percent sequence identity for the test sequence(s) relative to the reference sequence based on the program parameters.

[0219] The term "comparison window," as used herein, refers to a segment of a sequence comprising about 20 to about 1,000 contiguous amino acid or nucleotide positions, e.g., about 50 to about 600, about 100 to about 300, or about 150 to about 200 contiguous amino acid or nucleotide positions, over which a test sequence can be compared with a reference sequence over the same number of contiguous positions after optimally aligning the two sequences. Longer segments (up to the full-length sequence) may also be used as the comparison window in certain embodiments. Methods for aligning sequences for comparison purposes are known to those of skill in the art. Optimal sequence alignment for comparison can be performed, for example, by the local homology algorithm of Smith & Waterman, 1970, Adv. Appl. Math., 2:482c, by the homology alignment algorithm of Needleman & Wunsch, 1970, J. Mol. Biol., 48:443, by the search for similarity method of Pearson & Lipman, 1988, Proc. Natl. Acad. Sci. USA, 85:2444, or by computer implementations of these algorithms (e.g., GAP, BESTFIT, FASTA, or TFASTA (Wisconsin Genetics Software Package, Genetics Computer Group, Madison, WI)), or by manual alignment and visual inspection (see, for example, Ausubel et al., Current Protocols in Molecular Biology, (1995 supplement), Cold Spring Harbor Laboratory Press). Examples of available algorithms suitable for determining percent sequence identity are the BLAST and BLAST 2.0 algorithms, which are described in Altschul et al., 1997, Nuc. Acids Res., 25:3389-3402, and Altschul et al., 1990, J. Mol. Biol., 215:403-410, respectively.Software for performing BLAST analyses is publicly available from the website of the National Center for Biotechnology Information (NCBI).

[0220] Certain embodiments described herein include variant sequences (e.g., variant V) that contain one or more amino acid modifications, e.g., one or more amino acid insertions, one or more amino acid deletions, and / or one or more amino acid substitutions, when compared to a reference sequence, e.g., a wild-type sequence. H The term "conservative substitution" refers to a variant sequence of a polypeptide (e.g., a polypeptide domain, a variant Fc polypeptide, etc.). In certain embodiments, the one or more amino acid modifications of the variant sequence include one or more amino acid substitutions when compared to a reference, such as a wild-type sequence. In such embodiments, the one or more amino acid substitutions are one or more non-conservative substitutions. In other embodiments, the one or more amino acid substitutions are one or more conservative substitutions. In general, a "conservative substitution," as used herein, is considered to be the replacement of one amino acid with another amino acid having similar physical, chemical, and / or structural properties. Common conservative substitutions are listed in column 1 of Table 4. One of skill in the art will understand that the primary factors determining what constitutes a conservative substitution are usually the size of the amino acid side chain and its physical / chemical properties, although in certain circumstances a given amino acid can be substituted with a wider range of amino acids than those listed in column 1 of Table 4. These additional amino acids tend to have similar properties to the amino acid being substituted but significantly different in size, or similar in size but significantly different in physical / chemical properties. This broad range of conservative substitutions is listed in column 2 of Table 4. [Table 4]

[0221] Those skilled in the art will be able to readily determine the most appropriate set of substituents to select, given the particular protein environment in which the amino acid substitutions are to be made.

[0222] Pharmaceutical Composition In some embodiments, the present disclosure relates to pharmaceutical compositions that can include one or more of the antibody constructs described herein. In various embodiments, the pharmaceutical compositions herein can further include pharmaceutically acceptable excipients, carriers, buffers, stabilizers, or other materials well known to those skilled in the art. Such materials are generally non-toxic and do not interfere with the efficacy of the active ingredient (i.e., the antibody construct). The precise nature of the carrier or other materials may depend on the route of administration. Thus, the pharmaceutical compositions herein can be formulated for a variety of uses and routes of administration, for example, oral, intravenous, cutaneous, subcutaneous, nasal, intramuscular, or intraperitoneal.

[0223] Oral pharmaceutical compositions can be in the form of tablets, capsules, powders, or liquids. Tablets can contain solid carriers such as gelatin or adjuvants. Liquid pharmaceutical compositions generally contain liquid carriers such as water, petroleum, animal or vegetable oils, mineral oil, or synthetic oil. Physiological saline, dextrose, or other sugar solutions, or glycols such as ethylene glycol, propylene glycol, or polyethylene glycol can be included.

[0224] For intravenous, cutaneous or subcutaneous injection, or injection into an affected area (e.g., a tumor site), the active ingredient (i.e., antibody construct) may be in the form of a parenterally acceptable aqueous solution that is pyrogen-free and has suitable pH, isotonicity, and stability. Those skilled in the art are well able to prepare suitable solutions using isotonic vehicles such as sodium chloride, Ringer's injection, lactated Ringer's injection, etc. Preservatives, stabilizers, buffers, antioxidants, and / or other additives may be included as necessary.

[0225] For antibody constructs described in the present disclosure to be administered to a subject, the administration is preferably in a "therapeutically effective amount" sufficient to provide efficacy in the individual, as further described herein. The actual amount administered, as well as the rate and duration of administration, may vary depending on the nature and severity of the disease (e.g., cancer) being treated. Prescribing treatment, e.g., determining dosage, etc., is the responsibility of a general practitioner or other physician, and typically takes into account the disorder being treated, the condition of the individual patient, the site of delivery, the method of administration, and other factors known to physicians. Examples of the techniques and protocols mentioned above are described in Remington's Pharmaceutical Sciences, 16th edition, Osol, A. (ed), 1980.

[0226] In some embodiments, the pharmaceutical composition may include a second active ingredient (eg, another protein or a small molecule) in addition to the antibody constructs described herein.

[0227] kit The present disclosure also describes kits comprising one or more of the antibody constructs described herein, or pharmaceutical compositions described herein comprising such antibody construct(s); such kits may further comprise instructions for use. Thus, in certain embodiments, the present disclosure describes kits comprising a vector for expressing an antibody construct described herein and instructions for use. In certain embodiments, the present disclosure describes kits comprising a host cell comprising a vector for expressing an antibody construct and instructions for use. In some embodiments, the present disclosure relates to kits comprising a purified antibody construct and instructions for use. The purified antibody construct may be lyophilized or provided in a dried form, such as a powder or granules, and the kit may further contain a solvent suitable for reconstituting the lyophilized or dried component(s).

[0228] The kit may further comprise a container and a label and / or package insert on or associated with the container. The label or package insert may include instructions customarily included in commercial packaging for therapeutic products, providing information or instructions regarding the indications, methods of use, dosage, administration, contraindications, and / or warnings for use of the therapeutic product (e.g., an antibody construct described herein). The label or package insert may further include a notice in a form prescribed by a government agency regulating the manufacture, use, or sale of pharmaceuticals or biological products, indicating approval by that agency for manufacture, use, or sale for administration to humans or animals. The container may hold a composition comprising an antibody construct of the present disclosure. In some embodiments, the container may have a sterile access port. For example, the container may be an intravenous infusion bag or a vial having a stopper pierceable by a hypodermic injection needle.

[0229] In addition to the container holding the composition comprising the antibody construct, the kit may further comprise one or more additional containers containing other components of the kit. For example, such a kit may include a pharmaceutically acceptable buffer (such as bacteriostatic water for injection (BWFI), phosphate-buffered saline, Ringer's solution, or dextrose solution), or other buffer or diluent.

[0230] Suitable containers may include, for example, bottles, vials, syringes, intravenous infusion bags, etc. The containers may be formed from a variety of materials, such as glass or plastic. If desired, one or more components of the kit (e.g., antibody constructs) can be lyophilized or provided in a dried form, such as a powder or granules, and the kit may further contain a suitable solvent for reconstituting the lyophilized or dried component(s).

[0231] The kits herein may further include other materials desirable from a commercial or user standpoint, including filters, needles, and syringes.

[0232] Methods for preparing antibody constructs In some embodiments, the present disclosure relates to methods for preparing the antibody constructs described herein. In various embodiments, the antibody constructs of the present disclosure can be produced using standard recombinant methods known in the art (e.g., U.S. Patent No. 4,816,567 and "Antibodies: A Laboratory Manual," 2 nd Edition, Ed. Greenfield, Cold Spring Harbor Laboratory Press, New York, 2014).

[0233] For recombinant production of the antibody constructs described herein, a polynucleotide or set of polynucleotides encoding the antibody construct can be generated and inserted into one or more vectors for further cloning and / or expression in a host cell. The polynucleotide(s) encoding the antibody construct can be generated by standard methods known in the art (e.g., Ausubel et al., Current Protocols in Molecular Biology, John Wiley & Sons, New York, 1994 & update, and "Antibodies: A Laboratory Manual," 2002). nd (See, for example, "The First Edition," Ed. Greenfield, Cold Spring Harbor Laboratory Press, New York, 2014). As will be appreciated by those of skill in the art, the number of polynucleotides required for expression of an antibody construct may depend on the format and / or geometry of the antibody construct, such as the number of polypeptide chains comprising the antibody construct. For example, if the antibody construct includes three polypeptide chains (e.g., H1, H2, and L1), three polynucleotides, each encoding one polypeptide chain, can be used. In embodiments in which two or more polynucleotides are used, such two or more polynucleotides can be incorporated into a single vector or multiple vectors (e.g., two or three separate vectors).

[0234] Generally, a polynucleotide or set of polynucleotides encoding an antibody construct herein can be incorporated into an expression vector for expression along with one or more regulatory elements, such as transcription elements, that can be used to efficiently transcribe the polynucleotide(s). Examples of such regulatory elements include, but are not limited to, promoters, enhancers, terminators, and polyadenylation signals. Those skilled in the art will understand that the choice of regulatory element can depend on the host cell selected for expression of the antibody construct polypeptide, and that such regulatory elements can be derived from a variety of sources, including bacterial, fungal, viral, mammalian, or insect genes. Optionally, the expression vector can further contain heterologous nucleic acid sequences that facilitate expression or purification of the expressed protein. Examples include, but are not limited to, signal peptides and affinity tags, such as metal affinity tags, histidine tags, avidin / streptavidin coding sequences, glutathione-S-transferase (GST) coding sequences, and biotin coding sequences. Expression vectors can be extrachromosomal or integrating vectors. Thus, in some embodiments, the amino acid sequence of the polypeptide chain of an antibody construct described herein to be expressed, e.g., chains H1, H2, L1, etc., may comprise a signal peptide sequence. Such signal peptide sequences may vary depending on the expression system and conditions used to produce the antibody construct. Exemplary signal peptide sequences may comprise the amino acid sequence METDTLLLWVLLLWVPGSTG (SEQ ID NO: 49) or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 50), e.g., H1, H2, etc., or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 50) or MGWSCIILFLVATATGVHS (SEQ ID NO: 51). In certain embodiments, one or more heavy chains (e.g., H1, H2, etc.) of an antibody construct described herein may comprise a C-terminal lysine residue after expression of the polypeptide chain inside a cell.In various embodiments, such C-terminal lysine residues may be enzymatically cleaved from the polypeptide chain prior to further processing (e.g., purification, formulation, etc.) and prior to use of the corresponding antibody construct, e.g., prior to administration of the construct to a subject in need thereof.

[0235] Certain embodiments for producing the antibody constructs of the present disclosure relate to vectors (such as expression vectors) containing one or more polynucleotides encoding at least a portion of the antibody constructs described herein. The polynucleotide(s) can be contained in a single vector or multiple vectors. In some embodiments, the polynucleotides are contained in a multicistronic vector. Expression vectors that can be used to express polynucleotides include, but are not limited to, pTT5 and pUC15 cells containing vectors encoding the antibody constructs.

[0236] Suitable host cells for cloning or expressing the polypeptides of the antibody construct include various prokaryotic or eukaryotic cells known in the art. Eukaryotic host cells include, for example, mammalian cells, plant cells, insect cells, and yeast cells (such as Saccharomyces or Pichia cells). Prokaryotic host cells include, for example, E. coli cells, A. salmonicida cells, or B. subtilis cells. In certain embodiments, antibody constructs can be produced in bacteria, particularly when glycosylation and Fc effector functions are not necessary or desirable for the intended purpose of the antibody construct, as described, for example, in U.S. Pat. Nos. 5,648,237, 5,789,199, and 5,840,523, and Charlton, Methods in Molecular Biology, Vol. 248, pp. 245-254, BKC Lo, ed., Humana Press, Totowa, NJ, 2003. Eukaryotic microbes, such as filamentous fungi or yeast, are suitable expression host cells in certain embodiments, particularly fungal and yeast strains in which the glycosylation pathway has been "humanized," resulting in the production of antibodies with partial or fully human glycosylation patterns (see, e.g., Gerngross, 2004, Nat. Biotech. 22:1409-1414, and Li et al., 2006, Nat. Biotech. 24:210-215).

[0237] In various embodiments, suitable host cells for expressing glycosylated antibody constructs are eukaryotic cells. For example, U.S. Patent Nos. 5,959,177, 6,040,498, 6,420,548, 7,125,978, and 6,417,429 describe the PLANTIBODIES™ technology for producing antibodies and portions thereof (e.g., scFv, Fab, etc.) in transgenic plants. Mammalian cell lines adapted to grow in suspension are particularly useful for expressing the antibody constructs described herein. Examples include SV40-transformed monkey kidney line CV1 (COS-7), human embryonic kidney (HEK) line 293 or 293 cells (see, e.g., Graham et al., 1977, J. Gen Virol., 36:59), baby hamster kidney cells (BHK), mouse Sertoli TM4 cells (see, e.g., Mather, 1980, Biol Reprod., 23:243-251); monkey kidney cells (CV1), African green monkey kidney cells (VERO-76), human cervical carcinoma (HeLa) cells, canine kidney cells (MDCK), buffalo rat liver cells (BRL3A), human lung cells (W138), human liver cells (HepG2), mouse mammary tumor (MMT060562), TRI cells (see, e.g., Mather et al., 1982, Annals of NY Acad. Sci, 383: 44-68), MRC5 cells, FS4 cells, Chinese hamster ovary (CHO) cells (DHFR - Examples of suitable mammalian host cell lines for antibody production include, but are not limited to, CHO cells; see Urlaub et al., 1980, Proc Natl Acad Sci USA, 77:4216), and myeloma cell lines (such as Y0, NS0, and Sp2 / 0). Exemplary mammalian host cell lines suitable for antibody production are reviewed in Yazaki & Wu, Methods in Molecular Biology, Vol. 248, pp. 255-268 (BKC Lo, ed., Humana Press, Totowa, NJ, 2003).

[0238] In certain embodiments, the host cells used to produce the antibody constructs are transient or stable higher eukaryotic cell lines, such as mammalian cell lines. In some embodiments, the host cells are mammalian HEK293T cells, CHO cells, HeLa cells, NS0 cells, or COS cells. In some embodiments, the host cells are stable cell lines that allow mature glycosylation of the antibody constructs.

[0239] Conventional methods for producing antibody constructs can be used to culture host cells containing expression vector(s) encoding the antibody construct. Alternatively, in some embodiments, host cells containing expression vector(s) encoding the antibody construct can be used therapeutically or prophylactically to deliver the antibody construct to a subject, or the polynucleotide or expression vector can be administered ex vivo to cells derived from a subject, which can then be returned to the subject's body.

[0240] In some embodiments, the host cell comprises a V nucleotide sequence of the binding domain of the antibody construct described herein. L and V H In some embodiments, the host cell comprises (e.g., has been transformed with) a vector comprising a polynucleotide encoding the full-length polypeptide chain of an antibody construct described herein, e.g., H1, H2, or L1 described herein. In another example, the host cell comprises (e.g., has been transformed with) a vector comprising a polynucleotide encoding the V of the binding domain. L a first vector comprising a polynucleotide encoding a corresponding binding domain V Hand a second vector comprising a polynucleotide encoding the vector. In various embodiments, the host cell is a eukaryote, such as a Chinese hamster ovary (CHO) cell, a human embryonic kidney (HEK) cell, or a lymphoid cell (e.g., a Y0, NS0, or Sp20 cell). In certain embodiments, the host cell is Expi293™ (Thermo Fisher, Waltham, MA). In certain embodiments, the host cell used herein is a CHO-S cell (National Research Council Canada) or a HEK293 cell.

[0241] Certain embodiments of the present disclosure relate to methods for producing an antibody construct, comprising culturing host cells into which one or more polynucleotides encoding the antibody construct or one or more expression vectors encoding the antibody construct have been introduced under conditions suitable for expression of the antibody construct. Such methods may further comprise recovering the antibody construct from the host cells (or from the culture medium of the host cells). In some embodiments, such methods may further comprise purifying the antibody construct.

[0242] Cell culture media that can be used include, but are not limited to, DMEM (Thermo Fisher, Waltham, MA), Opti-MEM™ (Thermo Fisher, Waltham, MA), Opti-MEM™ I Reduced Serum Medium (Thermo Fisher, Waltham, MA), RPMI-1640 medium, Expi293™ Expression Medium (Thermo Fisher, Waltham, MA), and FreeStyle CHO Expression Medium (Thermo Fisher Scientific, Waltham, MA). Cell culture media can be supplemented with serum, such as fetal bovine serum (FBS), amino acids, such as L-glutamine, antibiotics, such as penicillin and streptomycin, and / or antimycotics, such as amphotericin, or any other additives routinely used to support cell culture.

[0243] A. Purification of antibody constructs In various embodiments, the antibody constructs of the present disclosure are purified after expression. Proteins, such as the antibody constructs of the present disclosure, can be isolated or purified by various methods known to those skilled in the art (see, for example, Protein Purification: Principles and Practice, 3). rd (See, Ed., Scopes, Springer-Verlag, NY, 1994). Standard purification methods that can be used with the antibody constructs disclosed herein include chromatographic techniques, including ion exchange, hydrophobic interaction, affinity, sizing, or gel filtration, and reverse-phase chromatography, performed at atmospheric or elevated pressure using systems such as FPLC and HPLC. Additional purification methods include electrophoretic, immunological, precipitation, dialysis, and chromatofocusing techniques. Ultrafiltration and diafiltration techniques in combination with protein concentration can also be used. As is well known in the art, various natural proteins bind to the Fc domain and other structural elements of antibody constructs, and in some embodiments, these proteins can be used to purify antibody constructs. For example, bacterial proteins A and G can bind to the Fc domain of some antibody constructs. Similarly, bacterial protein L can bind to the Fab domain of some antibody constructs. Purification can also be enabled by specific fusion partners. For example, antibody constructs can be purified using glutathione resins when GST fusions are used, or Ni-labeled resins when His-tags are used. +2 Purification can be performed using affinity chromatography or, if a flag tag is used, using immobilized anti-flag antibodies. The degree of purification required can vary depending on the use of the antibody construct. Thus, in some embodiments, purification may not be necessary.

[0244] In certain embodiments, the antibody constructs of the present disclosure are substantially pure. The term "substantially pure" (or "substantially purified"), when used with reference to the antibody constructs described herein, refers to the antibody construct being substantially or essentially free from components that normally accompany or interact with the protein as found in its naturally occurring environment, e.g., in natural cells or, in the case of recombinantly produced antibody constructs, in host cells. In certain embodiments, a substantially pure antibody construct is an antibody construct that has been purified to contain less than about 30%, less than about 25%, less than about 20%, less than about 15%, less than about 10%, less than about 5%, or less than about 2% (by dry weight) of other contaminating protein species.

[0245] Assessment of the purity and / or homogeneity of antibody constructs can be performed by any method known in the art, including, but not limited to, non-reducing / reducing CE-SDS, non-reducing / reducing SDS-PAGE, ultra-high performance liquid chromatography-size exclusion chromatography (UPLC-SEC), high performance liquid chromatography (HPLC), mass spectrometry, multi-angle light scattering (MALS), and dynamic light scattering (DLS).

[0246] B. Post-translational Modifications In certain embodiments, the antibody constructs described herein may comprise one or more post-translational modifications. Such post-translational modifications may occur in vivo or may be performed in vitro after isolation of the antibody construct from a host cell.

[0247] Post-translational modifications may include various modifications known in the art (see, e.g., Proteins - Structure and Molecular Properties, 2nd Ed., TECreighton, W.H. Freeman and Company, New York, 1993; Post-Translational Covalent Modification of Proteins, B.C. Johnson, Ed., Academic Press, New York, pp. 1-12, 1983; Seifter et al., 1990, Meth. Enzymol., 182:626-646, and Rattan et al., 1992, Ann. N.Y. Acad. Sci., 663:48-62). In embodiments in which an antibody construct contains one or more post-translational modifications, the antibody construct may contain the same type of modification at one or more sites (e.g., amino acid residues) or may contain different modifications at different sites.

[0248] Examples of post-translational modifications may include glycosylation, acetylation, phosphorylation, amidation, derivatization with known protecting / blocking groups, formylation, oxidation, reduction, proteolytic cleavage or specific chemical cleavage (by cyanogen bromide, trypsin, chymotrypsin, papain, V8 protease or NaBH4).

[0249] Other examples of post-translational modifications may include, for example, the addition or removal of N- or O-linked glycans, chemical modification of N- or O-linked glycans, N- or C-terminal processing, attachment of chemical moieties to the amino acid backbone, and the addition or deletion of N-terminal methionine residues resulting from expression in prokaryotic host cells. Post-translational modifications may also include modification with detectable labels, such as enzymatic, fluorescent, isotopic, or affinity labels, to enable detection and isolation of the protein. Examples of suitable enzymatic labels include, but are not limited to, horseradish peroxidase, alkaline phosphatase, beta-galactosidase, and acetylcholinesterase. Examples of suitable prosthetic group complexes include, but are not limited to, streptavidin / biotin and avidin / biotin. Examples of suitable fluorescent substances include, but are not limited to, umbelliferone, fluorescein, fluorescein isothiocyanate, rhodamine, dichlorotriazinylamine fluorescein, dansyl chloride, and phycoerythrin. An example of a luminescent material is luminol, examples of bioluminescent materials include luciferase, luciferin and aequorin, and examples of suitable radioactive materials include iodine, carbon, sulfur, tritium, indium, technetium, thallium, gallium, palladium, molybdenum, xenon and fluorine.

[0250] Additional examples of post-translational modifications may include acylation, ADP-ribosylation, amidation, covalent attachment of flavin, covalent attachment of a heme moiety, covalent attachment of a nucleotide or nucleotide derivative, covalent attachment of a lipid or lipid derivative, covalent attachment of phosphatidylinositol, cross-linking, cyclization, disulfide bond formation, demethylation, formation of covalent cross-links, formation of cysteine, formation of pyroglutamate, gamma-carboxylation, GPI anchor formation, hydroxylation, iodination, methylation, myristylation, pegylation, prenylation, racemization, selenoylation, sulfation, transfer-RNA-mediated addition of amino acids to proteins such as arginylation, and ubiquitination.

[0251] How to use In some embodiments, the present disclosure relates to methods of using the antibody constructs of the present disclosure. In certain embodiments, the present disclosure describes methods of using the antibody constructs described herein to treat a disease or condition in a subject in need of treatment. Such methods may include administering the antibody construct or a pharmaceutical composition comprising the antibody construct to a subject in need thereof. In certain embodiments, the subject is a mammal. In certain embodiments, the subject is a human or a rodent.

[0252] In some embodiments, the present disclosure relates to a method of treating cancer in a subject in need thereof, the method comprising administering to the subject an antibody construct of the present disclosure, or a pharmaceutical composition comprising the antibody construct. Cancers that can be treated using the methods and antibody constructs disclosed herein include, but are not limited to, hematological tumors (including leukemia, myeloma, and lymphoma), carcinomas (including adenocarcinoma and squamous cell carcinoma), melanoma, and sarcoma. Carcinomas and sarcomas are often referred to as "solid tumors." In certain embodiments, the cancer is a solid tumor. In certain embodiments, the cancer is a leukemia. In certain embodiments, the cancer is a lymphoma.

[0253] When used in the methods described herein, the antibody constructs of the present disclosure can exert either a cytotoxic or cytostatic effect, and can result in one or more of the following: a reduction in tumor size in a tumor-bearing subject; a slowing or prevention of tumor growth; an increase in disease-free survival from the disappearance or removal of a tumor to its recurrence; prevention of initial or subsequent development of a tumor (e.g., metastasis); an increase in progression-free time; a reduction in one or more adverse symptoms associated with a tumor; an increase in overall survival; or a combination of the above.

[0254] The methods described herein may include administering an antibody construct to a subject in need thereof. The antibody construct may be administered to a subject by any suitable route of administration. As will be understood by one of skill in the art, the route and / or mode of administration may vary depending on the desired therapeutic outcome. In various embodiments, the antibody construct of the present disclosure may be administered by systemic administration or local administration. Local administration may be to the tumor site or tumor-draining lymph nodes. Generally, the antibody construct may be administered parenterally, for example, intravenously, intramuscularly, intradermally, intraperitoneally, subcutaneously, or spinally, for example, by injection or infusion.

[0255] Treatment (e.g., of cancer in a subject) can be achieved by administering a therapeutically effective amount of an antibody construct. As used herein, a "therapeutically effective amount" generally refers to an amount of an antibody construct described herein effective, at a dosage and for a period of time necessary to achieve a desired therapeutic result. A therapeutically effective amount may vary depending on factors such as the disease state, age, sex, and weight of the subject. A therapeutically effective amount is also one in which any potential toxic or detrimental effects of the antibody construct are outweighed by the therapeutically beneficial effects. A "sufficient amount" generally refers to an amount sufficient to produce a desired effect, e.g., an amount sufficient to generate an anti-tumor immune response against target (e.g., tumor) cells or tissues by using the trivalent bispecific antibody construct described herein to affect immune cells (e.g., T cells).

[0256] A skilled medical practitioner can determine appropriate dosages of the antibody constructs described herein. The selected dosage level may vary depending on various pharmacokinetic factors, including the activity (e.g., antigen affinity(ies)) of the particular antibody construct used, the route of administration, the time of administration, the rate of excretion of the construct, the duration of treatment, other drugs, compounds, and / or materials used in combination with the antibody construct, e.g., anti-cancer drugs, the age, sex, weight, condition, general health, and previous medical history of the subject being treated, and similar factors well known in the medical field.

[0257] In some embodiments, methods of treating a disease (e.g., cancer) in a subject include administering a second active ingredient (e.g., another protein or small molecule) in addition to the antibody constructs described herein. Such second active ingredient can be administered simultaneously or sequentially with the antibody construct, depending on the condition to be treated.

[0258] In some embodiments, the present disclosure relates to a method of eliciting an anti-tumor immune response in a cell population comprising immune cells expressing CD3 and tumor cells expressing MSLN, the method comprising contacting the cell population with an effective amount of an antibody construct of the present disclosure. In some embodiments, such an antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises a dimeric Fc domain comprising: (i) a Fab domain capable of binding to CD3 on cytotoxic effector cells, (ii) a first scFv domain and a second scFv domain, wherein the first scFv domain and the second scFv domain are capable of binding to MSLN, and (iii) a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain (e.g., the N-terminus of the heavy or light chain). In various embodiments, the concentration of the antibody construct in the cell population is about 10 -2 pM to about 10 2 When the dose is increased to pM, the anti-tumor immune response reduces viable tumor cells in the cell population by at least 30%, 40%, 50%, or 60% in a dose-dependent manner, where the tumor cells express MSLNs at least about 15,000 MSLNs per cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0259] In some embodiments, the present disclosure relates to a method of inhibiting the growth of tumor cells expressing MSLN, the method comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of an antibody construct of the present disclosure. In various embodiments, such an antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises a dimeric Fc domain comprising: (i) a Fab domain capable of binding to CD3 on cytotoxic effector cells; (ii) a first scFv domain and a second scFv domain, wherein the first scFv domain and the second scFv domain are capable of binding to MSLN; and (iii) a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain. As defined herein, up to about 10 -2 pM, 10 -1 Tumor cell proliferation is inhibited if an increase in the count of viable tumor cells of up to about 5%, 3%, or up to about 1% is observed in the cell population over a period of at least about 5, 10, 20, or 48 hours using an antibody construct concentration of 1 pM, or 1 pM, wherein tumor cell MSLN expression is at least about 15,000 MSLNs per cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0260] In some embodiments, the present disclosure relates to a method of killing tumor cells expressing MSLN, such a method comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of an antibody construct of the present disclosure. In various embodiments, such an antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises a dimeric Fc domain comprising: (i) a Fab domain capable of binding to CD3 on cytotoxic effector cells; (ii) a first scFv domain and a second scFv domain, wherein the first scFv domain and the second scFv domain are capable of binding to MSLN; and (iii) a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain. In various embodiments, the antibody construct is administered at a concentration of about 10 -2 pM to about 10 2 When the concentration of MSLN in the tumor cell population is increased to 100 pM, such tumor cell killing is observed as measured by a dose-dependent reduction of at least about 30%, 40%, 50%, or 60% of viable tumor cells in the cell population, where the tumor cells express at least about 15,000 MSLNs per cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0261] In any of the methods described herein, the immune cells comprise or consist of one or more types of T cells.

[0262] As described herein, in various embodiments, the antibody constructs described herein can be administered to a subject in need thereof, e.g., a subject with cancer, to modulate an immune response in the subject. The immune response that can be modulated using the antibody constructs of the present disclosure can be an anti-tumor immune response in the subject; for example, in various embodiments, such a modulated immune response can occur locally at the tumor site. Thus, in certain embodiments, the antibody constructs described herein can initiate and / or upregulate a local immune response, e.g., an anti-tumor response of the subject's immune system, to induce a localized cytotoxic effect against the tumor at the tumor site.

[0263] In various embodiments, antibody constructs described herein, e.g., trivalent bispecific antibody constructs capable of monovalent binding to CD3 (via one Fab domain) and bivalent binding to MSLN (via two scFv domains), may have a broader therapeutic window compared to comparable conventional molecules, e.g., v31805 (MH6T-TriTAC), thereby allowing for higher dose administration and potentially enhanced anti-tumor efficacy without inducing side effects and off-target effects. Such a broader therapeutic window may result from certain properties of the antibody constructs described herein, including a higher ratio of anti-tumor activity compared to cytokine induction, i.e., higher tumor cell killing activity can be achieved at lower cytokine induction levels.

[0264] In some embodiments, the present disclosure relates to a method of inhibiting the growth of an MSLN-expressing tumor and / or reducing tumor volume in a subject in need thereof, the method comprising administering to the subject an effective amount of an MSLN-expressing tumor antibody construct of the present disclosure. In various embodiments, such in vivo anti-tumor effects are elicited by an antibody construct that binds (e.g., simultaneously) to CD3 on immune cells and MSLN on tumor cells, and comprises (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, (ii) a first scFv domain and a second scFv domain (the first scFv domain and the second scFv domain are capable of binding to MSLN), and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is linked to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is linked to the N-terminus of the Fab domain.

[0265] In various embodiments, and as further described elsewhere herein, the antibody constructs simultaneously bind to CD3 on immune cells and MSLN on tumor cells, forming a TCR-independent artificial immune synapse within the tumor environment in the subject, thereby inducing an anti-tumor cytotoxic effect (mediated by the activated immune cells and directed against the tumor cells), thereby inducing suppression of tumor growth and / or reduction of tumor volume in the subject. The immune cells can be T cells.

[0266] In various embodiments, administering the antibody construct to a subject four times a week (Q7Dx4) can suppress tumor growth in the subject for at least 20, 30, or 50 days. Furthermore, in some embodiments, tumor volume can be reduced by at least 20%, 30%, 40%, or at least about 50% about 15 days after the start of treatment. Such therapeutic effects can be achieved when the antibody construct is administered to a subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg. The subject can be a rodent, a non-human primate, or a human.

[0267] In further embodiments, administration of a sufficient amount of an antibody construct to a subject in need thereof in connection with the methods described herein can activate or upregulate an immune response by providing one or more of the following: (i) modulation of T cell receptor signaling, (ii) modulation of T cell activation, (iii) modulation of proinflammatory cytokines, (iv) modulation of interferon gamma production by T cells, (v) modulation of T cell suppression, (vi) modulation of survival and / or differentiation of M2-type tumor-associated macrophages (TAMs) or myeloid-derived suppressor cells (MDSCs), and / or (vii) modulation of a cytotoxic or cytostatic effect on a cell, e.g., a cancer cell.

[0268] In some embodiments, the present disclosure relates to methods of modulating an immune response in a cell or subject using one or more of the trivalent bispecific antibody construct(s) of the present disclosure, where such modulation may include one or more of: (i) activation of immune cells; (ii) stimulation of T cell receptor signaling; (iii) stimulation of antibody-dependent cellular cytotoxicity (ADCC); (iv) T cell-dependent cytotoxicity (TDCC); (v) cell-dependent cytotoxicity (CDC); (vi) antibody-dependent cellular phagocytosis (ADCP); and combinations of the above. As described herein, in certain embodiments, the antibody construct activates T effector cells. In some embodiments, as shown herein, the antibody construct increases the production of one or more cytokines and / or signaling molecules, such as GM-CSF, TNF-α, MIP-1β, IFN-γ, IL-2, IL-12, IL-17, IL-21, and / or C-X-C motif ligand 13 (CXCL13), by effector T cells.

[0269] As described herein, in various embodiments, the antibody constructs of the present disclosure comprise an Fc domain comprising a first and a second Fc polypeptide, wherein one or more of the Fc polypeptides may comprise a modified CH2 domain (e.g., compared to the WT domain) that comprises one or more amino acid modifications that may result in reduced or eliminated binding of the Fc domain to one or more or all Fcγ receptors (also referred to herein as an Fc "knockout" or "KO" variant).

[0270] Experimental Method In some embodiments, the present disclosure relates to experimental methods for analyzing and / or detecting the antibody constructs of the present disclosure. Such methods can be used, for example, to evaluate the in vitro and / or in vivo properties of the antibody construct, such as its pharmacokinetic (PK) and pharmacodynamic (PD) properties. Other properties and characteristics of the antibody construct can be evaluated, such as its stability under certain conditions (e.g., temperature, pH, etc.), its solubility, or its behavior in the presence of certain other chemical components, such as other proteins or cells.

[0271] Specific binding of the antibody constructs described herein to an antigen (e.g., CD3, MSLN) can be measured, for example, by enzyme-linked immunosorbent assay (ELISA), surface plasmon resonance (SPR) technology (e.g., employing a BIAcore instrument) (Liljeblad et al., 2000, Glyco J, 17:323-329), or conventional binding assays (Heeley, 2002, Endocr Res, 28:217-229). In certain embodiments, specific binding is defined as the extent of binding to unrelated proteins being less than about 10%, about 5%, or about 2% of the binding to the target antigen (e.g., CD3, MSLN, etc.), as measured, for example, by SPR. In certain embodiments, the specific binding of the antibody constructs herein to a particular antigen or epitope has a dissociation constant (K) of <1 μM, e.g., <500 nM, <200 nM, <100 nM, <50 nM, <25 nM, or <5 nM. D In some embodiments, the specific binding of an antibody construct to a particular antigen or epitope is defined as about 10 -6 M~about 10 -10 M, about 10 -6 M~about 10 -9 M, about 10 -8 M~about 10 -10 M, or about 10 -7 M~about 10 -9 Dissociation constant of M (K D ) is defined by

[0272] Further experimental methods for characterizing and evaluating the trivalent bispecific antibody constructs of the present disclosure are described in Examples 1-20 herein.

[0273] Specific Embodiments of the Disclosure The present disclosure further describes antibody constructs, pharmaceutical compositions comprising such constructs, and methods of using such constructs according to any one or more of embodiments 1-124.

[0274] Embodiment 1. An antibody construct comprising a Fab domain capable of binding to an antigen on a cytotoxic effector cell, a first scFv domain and a second scFv domain capable of binding to mesothelin (MSLN), and a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0275] Embodiment 2. At least one of the first scFv domain and the second scFv domain comprises an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L 2. The antibody construct of embodiment 1, comprising a domain.

[0276] Embodiment 3. The first scFv domain and the second scFv domain each comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L 3. The antibody construct of embodiment 2, comprising a domain.

[0277] Embodiment 4. An antibody construct comprising a Fab domain capable of binding to an antigen on a cytotoxic effector cell, a V domain capable of binding to MSLN and comprising an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L 1. The antibody construct of claim 1, further comprising a first scFv domain comprising a Fab domain, a second scFv domain capable of binding to a second tumor-associated antigen (TAA), and a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0278] Embodiment 5. The antibody construct of embodiment 4, wherein the second TAA is also MSLN.

[0279] Embodiment 6. The second scFv domain comprises an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L 6. The antibody construct of any one of embodiments 4 to 5, comprising a domain.

[0280] Embodiment 7. The antibody construct of any one of embodiments 1 to 6, wherein the cytotoxic effector cell is a T cell.

[0281] Embodiment 8. The antibody construct of any one of embodiments 1 to 7, wherein the antigen to which the Fab domain binds is CD3.

[0282] Embodiment 9. The Fab domain comprises an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 128. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L 9. The antibody construct of any one of embodiments 1 to 8, comprising a domain.

[0283] Embodiment 10. The first scFv domain and the second scFv domain each comprise an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103. H domain, and V comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:101. L 10. The antibody construct of any one of embodiments 1 to 9, comprising a domain.

[0284] Embodiment 11. For the first scFv domain and the second scFv domain, the V L The domain is a linker scFv By V H 11. The antibody construct of embodiment 10, wherein the antibody construct is linked to a domain.

[0285] Embodiment 12. The linkerscFv However, the amino acid sequence (G4S) n and n is 1, 2, 3, 4, or 5 (SEQ ID NO: 132).

[0286] Embodiment 13. The first scFv domain and the second scFv domain are, in the N-terminal to C-terminal direction, V H -Linker scFv -V L or V L -Linker scFv -V H 13. The antibody construct according to any one of embodiments 10 to 12, having a domain structure as follows:

[0287] Embodiment 14. The antibody construct of any one of Embodiments 10 to 13, wherein each of the first scFv domain and the second scFv domain comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 149.

[0288] Embodiment 15. The Fab domain comprises an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 105. H domain, and a C comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 106. H1 15. The antibody construct of any one of embodiments 1 to 14, comprising a domain.

[0289] Embodiment 16. The Fab domain further comprises an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 115. L domain, and a C comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 116. La light chain (L1) comprising the V domain of the heavy chain, H -C H1 16. The antibody construct of embodiment 15, wherein the Fab domain is paired with a Fab domain.

[0290] Embodiment 17. The antibody construct of any one of embodiments 15-16, wherein the light chain (L1) comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0291] Embodiment 18. The antibody construct comprises, in N-terminal to C-terminal direction, scFv2-V H -C H1 and a C-terminus of the second scFv domain (scFv2) comprising the V H 18. The antibody construct of any one of embodiments 1 to 17, wherein the domain is linked to the N-terminus of the domain.

[0292] Embodiment 19. The C-terminus of the second scFv domain is H Linker at the N-terminus of the domain scFv-Fab 19. The antibody construct of embodiment 18, wherein the antibody construct is linked via

[0293] Embodiment 20. The linker scFv-Fab However, the amino acid sequence (G4S) n and n is 1, 2, 3, 4, or 5 (SEQ ID NO: 132).

[0294] Embodiment 21. The first heavy chain (H1) further comprises the first Fc polypeptide, and the C H1 The C-terminus of the domain is linked to the N-terminus of the first Fc polypeptide, thereby forming scFv2-V for H1. H -C H1 The antibody construct according to any one of embodiments 18 to 20, which forms a -CH2-CH3 domain structure.

[0295] Embodiment 22. The C H1 The domain is a linker Fab-Fc 22. The antibody construct of embodiment 21, wherein the first Fc polypeptide is linked to the first Fc polypeptide by:

[0296] Embodiment 23. The linker Fab-Fc 23. The antibody construct of embodiment 22, wherein said antibody construct comprises or consists of an Ig hinge region.

[0297] Embodiment 24 The antibody construct of embodiment 23, wherein the Ig hinge region is an IgG hinge region.

[0298] Embodiment 25. The antibody construct of embodiment 24, wherein the IgG hinge region is an IgG1 hinge region.

[0299] Embodiment 26. The linker Fab-Fc 26. An antibody construct according to any one of embodiments 22 to 25, comprising or consisting of an amino acid sequence having at least about 70%, 80%, 90%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 107.

[0300] Embodiment 27. The antibody construct of any one of embodiments 1 to 26, wherein the dimeric Fc domain is a heterodimeric Fc domain comprising the first Fc polypeptide and the second Fc polypeptide.

[0301] Embodiment 28. The antibody construct of any one of Embodiments 1 to 27, wherein the first Fc polypeptide is an IgG1-derived Fc polypeptide and comprises a first CH2 domain and a first CH3 domain.

[0302] Embodiment 29. The antibody construct of embodiment 28, wherein the first CH3 domain comprises one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH3 domain sequence.

[0303] Embodiment 30. The antibody construct of any one of embodiments 28-29, wherein the first CH2 domain comprises one or more amino acid substitutions that reduce or abolish binding to an Fc receptor compared to the corresponding wild-type IgG1 CH2 domain.

[0304] Embodiment 31. The antibody construct of any one of Embodiments 18 to 30, wherein the first heavy chain (H1) comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100.

[0305] Embodiment 32. A second heavy chain (H2) comprising, from N-terminal to C-terminal, the first scFv domain and the second Fc polypeptide linked thereto, thereby forming the domain structure scFv1-CH2-CH3. 32. The antibody construct of any one of embodiments 1 to 31, further comprising:

[0306] Embodiment 33. The C-terminus of the first scFv domain is linked to the N-terminus of the second Fc polypeptide via a linker scFv-Fc 33. The antibody construct of embodiment 32, wherein the antibody construct is linked via:

[0307] Embodiment 34. The linker scFv-Fc 34. The antibody construct of embodiment 33, wherein said antibody construct comprises or consists of an Ig hinge region.

[0308] Embodiment 35 The antibody construct of embodiment 34, wherein the Ig hinge region is an IgG hinge region.

[0309] Embodiment 36 The antibody construct of embodiment 35, wherein the IgG hinge region is an IgG1 hinge region.

[0310] Embodiment 37. The linker scFv-FcAn antibody construct according to any one of embodiments 33 to 36, comprising or consisting of an amino acid sequence having at least about 70%, 80%, 90%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 111.

[0311] Embodiment 38. The antibody construct of any one of Embodiments 32 to 37, wherein the second Fc polypeptide is also an IgG1-derived Fc polypeptide and comprises a second CH2 domain and a second CH3 domain.

[0312] Embodiment 39. The antibody construct of embodiment 38, wherein the second CH3 domain comprises one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH3 domain sequence.

[0313] Embodiment 40. The antibody construct of any one of embodiments 38-39, wherein the second CH2 domain comprises one or more amino acid substitutions that reduce or abolish binding to an Fc receptor compared to the corresponding wild-type IgG1 CH2 domain.

[0314] Embodiment 41. The first Fc polypeptide and the second Fc polypeptide each comprise a CH3 sequence, wherein one of the Fc polypeptides comprises an amino acid substitution in its CH3 sequence selected from L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V, and the other Fc polypeptide comprises an amino acid substitution in its CH3 sequence selected from T366L_K392M 41. The antibody construct of any one of embodiments 1 to 40, comprising an amino acid substitution selected from: T350V_T366L_K392L_T394W, T350V_T366L_K392L_T394W, T350V_T366L_K392M_T394W, and T350V_T366L_N390R_K392M_T394W, wherein the numbering of amino acid residues in said Fc polypeptide sequence is according to the EU numbering system.

[0315] Embodiment 42 The antibody construct of embodiment 41, wherein the amino acid substitutions in the CH3 sequence of the Fc polypeptide promote preferential pairing of the heavy chain H1 and H2 and formation of a heterodimeric Fc domain relative to formation of a homodimeric Fc domain.

[0316] Embodiment 43. The antibody construct of any one of Embodiments 32 to 42, wherein the second heavy chain (H2) comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110.

[0317] Embodiment 44. The antibody construct of any one of embodiments 1 to 43, wherein the antibody construct is bispecific for CD3 and MSLN, monovalent for CD3, and bivalent for MSLN.

[0318] Embodiment 45. The antibody construct of any one of embodiments 1 to 44, wherein the antibody construct comprises a first heavy chain (H1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100, a second heavy chain (H2) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110, and a light chain (L1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0319] Embodiment 46. The antibody construct of claim 45, wherein the first heavy chain (H1) comprises or consists of the amino acid sequence set forth in SEQ ID NO: 100.

[0320] Embodiment 47. The antibody construct of any one of Embodiments 45-46, wherein the second heavy chain (H2) comprises or consists of the amino acid sequence set forth in SEQ ID NO:110.

[0321] Embodiment 48. The antibody construct of any one of embodiments 45 to 47, wherein the light chain (L1) comprises or consists of the amino acid sequence set forth in SEQ ID NO: 114.

[0322] Embodiment 49. An antibody construct described in any one of embodiments 1 to 44, comprising: (i) a first heavy chain (H1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 171; (ii) a second heavy chain (H2) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 172; and (iii) a light chain (L1) comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 114.

[0323] Embodiment 50. The antibody construct of any one of embodiments 1 to 9, wherein the antibody construct comprises a first heavy chain (H1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 117, a second heavy chain (H2) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 119, and a light chain (L1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0324] Embodiment 51. The antibody construct of any one of embodiments 18 to 50, wherein one or more of the heavy chains H1 and H2 further comprise a C-terminal lysine residue.

[0325] Embodiment 52. The antibody construct of any one of embodiments 1 to 51, wherein when the antibody construct is present at a concentration of about 10 pM in a cell population comprising immune cells expressing CD3 and tumor cells expressing about 500,000 MSLN per cell, the antibody construct is capable of one or more, two or more, or all of the following: A) inducing production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL of TNFα; B) inducing production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL of IFNγ; or C) inducing production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL of IL-2.

[0326] Embodiment 53. The antibody construct of any one of embodiments 1 to 52, wherein the antibody construct is capable of inducing at least 10-fold greater T cell-mediated cytotoxicity against tumor cells compared to a similar antibody construct comprising two anti-MSLN Fabs instead of the first and second anti-MSLN scFv domains.

[0327] Embodiment 54. The antibody construct has a binding affinity (K) for MSLN of about 0.7 nM, 0.8 nM, 0.9 nM, 1 nM, or about 1.1 nM, as measured by SPR. D 54. The antibody construct of any one of embodiments 1 to 53, having the following structure:

[0328] Embodiment 55. The antibody construct has a binding affinity (K) for CD3 of about 30 nM, 40 nM, 50 nM, or about 60 nM as measured by SPR. D 55. The antibody construct of any one of embodiments 8 to 54, having the following structure:

[0329] Embodiment 56. The antibody construct of any one of embodiments 1 to 55, wherein the antibody construct is capable of inducing the production of inflammatory cytokines by the cytotoxic effector cells in an MSLN-dependent manner, wherein the MSLN-dependent activation of the cytotoxic effector cells is determined by measuring a decrease in cytokine production of at least about 20-fold, 50-fold, 100-fold, or at least about 1000-fold between a first cell population comprising first tumor cells and immune cells and a second cell population comprising second tumor cells and immune cells, and wherein MSLN expression in the first tumor cells is at least about 3-fold, 4-fold, or 5-fold higher than MSLN expression in the second tumor cells.

[0330] Embodiment 57. The antibody construct of any one of embodiments 1 to 56, wherein when administered to a subject, the antibody construct is capable of reducing the volume of an MSLN-expressing tumor in the subject by at least about 5% for at least 20 days after administration of the antibody construct to the subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg four times per week.

[0331] Embodiment 58. A pharmaceutical composition comprising the antibody construct of any one of Embodiments 1 to 57 and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.

[0332] Embodiment 59. A nucleic acid molecule or set of nucleic acid molecules encoding one or more, two or more, or three or more polypeptide chains forming the antibody construct of any one of embodiments 1 to 57.

[0333] Embodiment 60. A vector or set of vectors comprising the nucleic acid molecule or set of nucleic acid molecules of embodiment 59.

[0334] Embodiment 61. A cell comprising the nucleic acid molecule or set of nucleic acid molecules described in embodiment 59, or the vector or set of vectors described in embodiment 60.

[0335] Embodiment 62. A method for producing the antibody construct of any one of claims 1 to 57, comprising: (a) obtaining a host cell culture comprising at least one host cell comprising one or more nucleic acid molecules encoding one or more, two or more, three or more polypeptide chains forming said antibody construct; and (b) recovering said antibody construct from said host cell culture.

[0336] Embodiment 63 The method of embodiment 62, further comprising purifying the antibody construct after step (b).

[0337] Embodiment 64. A method for inducing an anti-tumor immune response in a cell population comprising immune cells expressing CD3 and tumor cells expressing MSLN, said method comprising contacting said cell population with an effective amount of the antibody construct of any one of embodiments 1 to 57.

[0338] Embodiment 65. The concentration of the antibody construct in the cell population is about 10 -2 pM to about 10 2 65. The method of embodiment 64, wherein when the anti-tumor immune response is increased to pM, the anti-tumor immune response reduces live tumor cells in the cell population by at least 30%, 40%, 50%, or 60% in a dose-dependent manner, the tumor cells express MSLNs at least about 15,000 MSLNs per cell, and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0339] Embodiment 66. A method for inhibiting the proliferation of tumor cells expressing MSLN, said method comprising contacting a cell population comprising said tumor cells and immune cells expressing CD3 with an effective amount of the antibody construct of any one of Embodiments 1 to 57.

[0340] Embodiment 67. Up to about 10 -2 pM, 10 -167. The method of embodiment 66, wherein tumor cell proliferation is inhibited when an increase in the count of viable tumor cells of up to about 5% is observed in the cell population over a period of at least about 5, 10, 20, or 48 hours using an antibody construct concentration of 1 pM, or 1 pM, and the tumor cells express at least about 15,000 MSLNs per cell, and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0341] Embodiment 68. A method of killing tumor cells that express MSLN, said method comprising contacting a cell population comprising said tumor cells and immune cells that express CD3 with an effective amount of the antibody construct of any one of embodiments 1 to 57.

[0342] Embodiment 69. The antibody construct has a concentration of about 10 -2 pM to about 10 2 69. The method of embodiment 68, wherein tumor cell killing is observed as measured by a dose-dependent reduction of at least about 30%, 40%, 50%, or 60% in viable tumor cells in the cell population when the MSLN expression level is increased to pM, the tumor cells express at least about 15,000 MSLNs per cell, and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0343] Embodiment 70. The method of any one of embodiments 64-69, wherein the immune cells comprise T cells.

[0344] Embodiment 71. The method of any one of embodiments 64 to 70, wherein the antibody construct binds to CD3 on immune cells and MSLN on tumor cells.

[0345] Embodiment 72. A method for inhibiting the growth of an MSLN-expressing tumor or reducing the volume of said tumor in a subject, said method comprising administering to said subject an effective amount of an antibody construct described in any one of embodiments 1 to 57.

[0346] Embodiment 73. The method of embodiment 72, wherein the antibody construct binds to CD3 on immune cells of the subject and MSLN on tumor cells, thereby eliciting an anti-tumor immune response in the subject, thereby inhibiting growth of the tumor or reducing the volume of the tumor in the subject.

[0347] Embodiment 74. The method of embodiment 73, wherein the inhibition of tumor growth or reduction in tumor volume is induced by simultaneous binding of the antibody construct to CD3 on immune cells and MSLN on tumor cells, and formation of a TCR-independent artificial immune synapse within the tumor environment of the subject.

[0348] Embodiment 75 The method of embodiment 74, wherein the immune cells are T cells.

[0349] Embodiment 76. The method of any one of embodiments 72-75, wherein tumor growth is inhibited for at least 20 days, 30 days, or 50 days using four weekly (Q7Dx4) administrations of the antibody construct to the subject.

[0350] Embodiment 77. The method of any one of embodiments 72-76, wherein the tumor volume is reduced by at least 20%, 30%, 40%, or at least about 50% about 15 days after initiation of treatment.

[0351] Embodiment 78 The method of any one of embodiments 72 to 77, wherein about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg of the antibody construct is administered to the subject.

[0352] Embodiment 79. The method of any one of embodiments 72-78, wherein the subject is a rodent, a non-human primate, or a human.

[0353] Embodiment 80. An antibody construct according to any one of embodiments 1 to 57 for use in the treatment of cancer.

[0354] Embodiment 81. Use of an antibody construct according to any one of embodiments 1 to 57 in the manufacture of a medicament for the treatment of cancer.

[0355] Embodiment 82. An antibody construct comprising: (i) a VH1 sequence comprising or consisting of an HCDR1 sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of an HCDR2 sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of an HCDR3 sequence set forth in SEQ ID NO: 128. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L 1. The antibody construct of claim 1, further comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, the Fab domain comprising a first scFv domain (scFv1) and a second scFv domain (scFv2) capable of binding to mesothelin (MSLN); and (ii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0356] Embodiment 83. At least one of the first scFv domain and the second scFv domain comprises (i) a VFv domain comprising an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H domain, and (ii) a V domain comprising an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L83. The antibody construct of embodiment 82, comprising a domain.

[0357] Embodiment 84. The first scFv domain and the second scFv domain each comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L 84. The antibody construct of embodiment 82 or embodiment 83, comprising a domain.

[0358] Embodiment 85. An antibody construct comprising: (i) a VH1 sequence comprising or consisting of an HCDR1 sequence set forth in SEQ ID NO: 126, an HCDR2 sequence comprising or consisting of an HCDR2 sequence set forth in SEQ ID NO: 127, and an HCDR3 sequence comprising or consisting of an HCDR3 sequence set forth in SEQ ID NO: 128. H a V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 129, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 130, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 131. L (ii) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, comprising a scFv domain; (iii) a first scFv domain (scFv1) and a second scFv domain (scFv2) capable of binding to mesothelin (MSLN), wherein the first scFv domain and the second scFv domain comprise an HCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 120, an HCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 121, and an HCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 122, respectively. Ha V domain, and an LCDR1 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 123, an LCDR2 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 124, and an LCDR3 sequence comprising or consisting of the sequence set forth in SEQ ID NO: 125. L and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein the Fab domain is linked to the N-terminus of the first Fc polypeptide, the first scFv domain is linked to the N-terminus of the second Fc polypeptide, and the second scFv domain is linked to the N-terminus of the Fab domain.

[0359] Embodiment 86. The first scFv domain or the second scFv domain comprises (i) a VFv comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103. H domain, and (ii) a V comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 101. L 86. An antibody construct according to any one of embodiments 82 to 85, comprising a domain.

[0360] Embodiment 87. The first scFv domain and the second scFv domain each comprise (i) an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 103. H domain, and (ii) a V comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 101. L 87. An antibody construct according to any one of embodiments 82 to 86, comprising a domain.

[0361] Embodiment 88. The first scFv domain and the second scFv domain each comprise (i) a V domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 103. H domain, and (ii) a V domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 101. L 88. An antibody construct according to any one of embodiments 82 to 87, comprising a domain.

[0362] Embodiment 89. The first scFv domain and the second scFv domain are L Domain V H Linker that connects domains scFv 89. The antibody construct of any one of embodiments 86 to 88, further comprising:

[0363] Embodiment 90. The linker scFv However, the amino acid sequence (G4S) n and n is 1, 2, 3, 4, or 5.

[0364] Embodiment 91. The first scFv domain and the second scFv domain are, in the N-terminal to C-terminal direction, V H -Linker scFv -V L or V L -Linker scFv -V H 91. The antibody construct of any one of embodiments 89 to 90, having a domain structure of:

[0365] Embodiment 92. The first scFv domain and the second scFv domain are, in the N-terminal to C-terminal direction, V H -Linker scFv -V L 92. The antibody construct of embodiment 91, having a domain structure of:

[0366] Embodiment 93. The first scFv domain and the second scFv domain are, in the N-terminal to C-terminal direction, V L -LinkerscFv -V H 92. The antibody construct of embodiment 91, having a domain structure of:

[0367] Embodiment 94. The Fab domain (i) comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 105. H (ii) a C domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 106. H1 (iii) a V domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 115. L domain, and (iv) a C domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 116. L 94. An antibody construct according to any one of embodiments 82 to 93, comprising a domain.

[0368] Embodiment 95. The antibody construct comprises, in N-terminal to C-terminal direction, scFv2-V H -C H1 and a C-terminus of the second scFv domain (scFv2) comprising the V of the Fab heavy chain. H 95. The antibody construct of any one of embodiments 82 to 94, wherein the domain is linked to the N-terminus of the domain.

[0369] Embodiment 96. The C-terminus of the scFv2 is H Linker at the N-terminus of the domain scFv-Fab 96. The antibody construct of embodiment 95, wherein the antibody construct is linked via

[0370] Embodiment 97. The linker scFv-Fab However, the amino acid sequence (G4S) nand n is 1, 2, 3, 4, or 5.

[0371] Embodiment 98. The H1 further comprises the first Fc polypeptide comprising a first CH2 sequence (CH21) and a first CH3 sequence (CH31), H1 The C-terminus of the scFv2-V domain is linked to the N-terminus of the first Fc polypeptide, thereby forming (from the N-terminus to the C-terminus) scFv2-V for H1. H -C H1 An antibody construct according to any one of embodiments 95 to 97, which forms a domain structure of -CH21-CH31.

[0372] Embodiment 99. The C H1 a linker domain to the CH2 sequence of the first Fc polypeptide; Fab-Fc 99. The antibody construct of embodiment 98, wherein the antibody construct is linked via

[0373] Embodiment 100. The linker Fab-Fc 100. The antibody construct of embodiment 99, wherein said antibody construct comprises or consists of an Ig hinge region.

[0374] Embodiment 101. The antibody construct of any one of embodiments 95 to 100, wherein H1 comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100.

[0375] Embodiment 102. A second heavy chain (H2) comprising, from N-terminal to C-terminal, the first scFv domain (scFv1) and, linked thereto, the second Fc polypeptide comprising a second CH2 sequence (CH22) and a second CH3 sequence (CH32), thereby forming the domain structure scFv1-CH22-CH32. 102. The antibody construct of any one of embodiments 82 to 101, further comprising:

[0376] Embodiment 103. The C-terminus of the scFv1 is linked to the N-terminus of the second Fc polypeptide via a linker scFv-Fc 103. The antibody construct of embodiment 102, wherein the antibody construct is linked via:

[0377] Embodiment 104. The linker scFv-Fc The antibody construct of embodiment 103, wherein said antibody construct comprises or consists of an Ig hinge region.

[0378] Embodiment 105. An antibody construct described in any one of embodiments 102 to 104, wherein H2 comprises or consists of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110.

[0379] Embodiment 106. The antibody construct of any one of Embodiments 82 to 105, wherein the first Fc polypeptide, the second Fc polypeptide, or both the first and second Fc polypeptides are IgG1-derived Fc polypeptides.

[0380] Embodiment 107. The antibody construct of any one of embodiments 82 to 106, wherein the dimeric Fc domain is a heterodimeric Fc domain and the amino acid sequence of the first Fc polypeptide differs from the amino acid sequence of the second Fc polypeptide by at least one amino acid residue.

[0381] Embodiment 108. The antibody construct of embodiment 107, wherein the first Fc polypeptide, the second Fc polypeptide, or both Fc polypeptides comprise one or more amino acid substitutions in the CH3 domain compared to the corresponding wild-type IgG1 CH3 domain sequence, and wherein the one or more amino acid substitutions promote preferential pairing of the first and second Fc polypeptides to form the heterodimeric Fc domain.

[0382] Embodiment 109. 109. The antibody construct of embodiment 108, wherein one Fc polypeptide comprises a set of amino acid substitutions selected from L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V, and the other Fc polypeptide comprises a set of amino acid substitutions selected from T366L_K392M_T394W, T366L_K392L_T394W, T350V_T366L_K392L_T394W, T350V_T366L_K392M_T394W, and T350V_T366L_N390R_K392M_T394W, and wherein the numbering of amino acid residues in said Fc polypeptides is according to the EU numbering system.

[0383] Embodiment 110. The antibody construct of any one of embodiments 98-109, wherein the first Fc polypeptide, the second Fc polypeptide, or both Fc polypeptides comprise one or more amino acid substitutions in the CH2 domain compared to the corresponding wild-type IgG1 CH2 domain sequence, and wherein the one or more amino acid substitutions reduce or abolish binding of the antibody construct to an Fcγ receptor.

[0384] Embodiment 111. The antibody construct of any one of Embodiments 82 to 110, wherein the antibody construct comprises a first heavy chain (H1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 100, a second heavy chain (H2) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 110, and a light chain (L1) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 114.

[0385] Embodiment 112. The antibody construct of any one of embodiments 95 to 111, wherein one or more of the heavy chains H1 and H2 further comprise a C-terminal lysine residue.

[0386] Embodiment 113. An antibody construct comprising a first heavy chain (H1) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 100, a second heavy chain (H2) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 110, and a light chain (L1) polypeptide comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 114.

[0387] Embodiment 114. The antibody construct of any one of embodiments 82 to 113, wherein when the antibody construct is present at a concentration of about 10 pM in a cell population comprising immune cells expressing CD3 and tumor cells expressing about 500,000 MSLN per cell, the antibody construct is capable of one or more, two or more, or all of the following: A) inducing production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL of TNFα; B) inducing production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL of IFNγ; or C) inducing production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL of IL-2.

[0388] Embodiment 115. A compound having a binding affinity (K) of about 0.7 nM, 0.8 nM, 0.9 nM, 1 nM, or about 1.1 nM to MSLN as measured by SPR. D 115. The antibody construct of any one of embodiments 82 to 114, having the following structure:

[0389] Embodiment 116. A compound having a binding affinity (K) for CD3 of about 30 nM, 40 nM, 50 nM, or about 60 nM, as measured by SPR. D 116. The antibody construct of any one of embodiments 82 to 115, having the following structure:

[0390] Embodiment 117. The antibody construct of any one of embodiments 82 to 116, wherein the antibody construct has a purity after incubation in a buffer solution at 40°C for 14 days of at least about 95%, 96%, 97%, 98%, or at least about 99% as measured by size exclusion chromatography relative to other high molecular weight species (HMWS) present in the buffer solution after the incubation period.

[0391] Embodiment 118. A pharmaceutical composition comprising the antibody construct of any one of embodiments 82 to 117 and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.

[0392] Embodiment 119. A nucleic acid molecule or set of nucleic acid molecules encoding one or more, two or more, or three or more polypeptide chains forming the antibody construct of any one of embodiments 82 to 117.

[0393] Embodiment 120. A vector or set of vectors comprising the nucleic acid molecule or set of nucleic acid molecules of embodiment 119.

[0394] Embodiment 121. A cell comprising the nucleic acid molecule or set of nucleic acid molecules described in embodiment 119, or the vector or set of vectors described in embodiment 120.

[0395] Embodiment 122. A method for producing an antibody construct described in any one of embodiments 82 to 117, comprising: (a) obtaining a host cell culture comprising at least one host cell that contains one or more nucleic acid molecules encoding one or more, two or more, three or more, or all of the polypeptide chains that form the antibody construct; and (b) recovering the antibody construct from the host cell culture.

[0396] Embodiment 123. An antibody construct according to any one of embodiments 82 to 117 for use in the treatment of cancer.

[0397] Embodiment 124. Use of an antibody construct according to any one of embodiments 82 to 117 in the manufacture of a medicament for the treatment of cancer. [Example]

[0398] The following examples are offered for illustrative purposes and are not intended to limit the scope of the present disclosure in any way.

[0399] The practice of the present disclosure may involve, and may employ, conventional methods of protein chemistry, biochemistry, recombinant DNA technology, and pharmacology, within the skill of the...

Claims

1. Antibody constructs, (i) V containing the HCDR1 sequence containing the sequence described in SEQ ID NO: 126, the HCDR2 sequence containing the sequence described in SEQ ID NO: 127, and the HCDR3 sequence containing the sequence described in SEQ ID NO: 128 Ｈ Domain, and V containing the LCDR1 sequence containing the sequence described in SEQ ID NO: 129, the LCDR2 sequence containing the sequence described in SEQ ID NO: 130, and the LCDR3 sequence containing the sequence described in SEQ ID NO:

131. Ｌ A Fab domain, including a domain, that can bind to CD3 on cytotoxic effector cells, (ii) A first scFv domain (scFv1) and a second scFv domain (scFv2) capable of binding to mesothelin (MSLN), wherein the first scFv domain and the second scFv domain each contain an HCDR1 sequence containing the sequence described in SEQ ID NO: 120, an HCDR2 sequence containing the sequence described in SEQ ID NO: 121, and an HCDR3 sequence containing the sequence described in SEQ ID NO:

122. Ｈ Domain, and V containing the LCDR1 sequence containing the sequence described in SEQ ID NO: 123, the LCDR2 sequence containing the sequence described in SEQ ID NO: 124, and the LCDR3 sequence containing the sequence described in SEQ ID NO:

125. Ｌ The first scFv domain and the second scFv domain, including the domain, and (iii) A dimerized Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. Includes, The Fab domain is ligated to the N-terminus of the first Fc polypeptide, The first scFv domain is ligated to the N-terminus of the second Fc polypeptide, The second scFv domain is ligated to the N-terminus of the Fab domain. The aforementioned antibody construct.

2. The first scFv domain and the second scFv domain each contain (i) the amino acid sequence described in SEQ ID NO:

103. Ｈ (ii) V containing the domain and the amino acid sequence described in SEQ ID NO: 101 Ｌ The antibody construct according to claim 1, comprising a domain.

3. The first scFv domain and the second scFv domain are the V Ｌ domain to the V Ｈ linker that links the domain to the V ｓｃＦｖ The antibody construct according to claim 1, further comprising.

4. The linker ｓｃＦｖ However, the amino acid sequence (G ４ S) ｎ The antibody construct according to claim 3, comprising or consisting of, where n is 1, 2, 3, 4, or 5.

5. The first scFv domain and the second scFv domain are V in the direction from the N-terminus to the C-terminus. Ｈ - Linker ｓｃＦｖ -V Ｌ , or V Ｌ - Linker ｓｃＦｖ -V Ｈ The antibody construct according to claim 3, having the domain structure.

6. The first scFv domain and the second scFv domain are V in the direction from the N-terminus to the C-terminus. Ｈ - Linker ｓｃＦｖ -V Ｌ The antibody construct according to claim 5, having the domain structure.

7. The first scFv domain and the second scFv domain are V in the direction from the N-terminus to the C-terminus. Ｌ - Linker ｓｃＦｖ -V Ｈ The antibody construct according to claim 5, having the domain structure.

8. The Fab domain (i) contains the sequence described in Sequence ID No. 105 Ｈ Domain, (ii) C containing the sequence described in sequence number 106 Ｈ１ Domain, (iii) V containing the sequence described in Sequence ID No. 115 Ｌ C containing the domain and the sequence described in (iv) Sequence ID 116. Ｌ The antibody construct according to claim 1, comprising a domain.

9. The antibody construct according to claim 1, wherein the first Fc polypeptide, the second Fc polypeptide, or both the first and second Fc polypeptides are IgG1-derived Fc polypeptides.

10. The antibody construct according to claim 1, wherein the dimeric Fc domain is a heterodimeric Fc domain, and the amino acid sequence of the first Fc polypeptide differs from the amino acid sequence of the second Fc polypeptide by at least one amino acid residue.

11. The antibody construct according to claim 10, wherein the first Fc polypeptide, the second Fc polypeptide, or both Fc polypeptides contain one or more amino acid substitutions in the CH3 domain compared to the corresponding wild-type IgG1 CH3 domain sequence, and the one or more amino acid substitutions promote preferential pairing of the first and second Fc polypeptides to form the heterodimeric Fc domain.

12. The antibody construct according to claim 11, wherein one Fc polypeptide comprises a set of amino acid substitutions selected from L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V, and the other Fc polypeptide comprises a set of amino acid substitutions selected from T366L_K392M_T394W, T366L_K392L_T394W, T350V_T366L_K392L_T394W, T350V_T366L_K392M_T394W, and T350V_T366L_N390R_K392M_T394W, and the numbering of amino acid residues in the first and second Fc polypeptides conforms to the EU numbering system.

13. The antibody construct according to claim 10, wherein the first Fc polypeptide, the second Fc polypeptide, or both Fc polypeptides contain one or more amino acid substitutions in the CH2 domain compared to the corresponding wild-type IgG1 CH2 domain sequence, the one or more amino acid substitutions reduce or eliminate the binding of the antibody construct to the Fcγ receptor, the one or more amino acid substitutions include the amino acid substitution L234A_L235A_D265S, and the numbering of amino acid residues in the first and second Fc polypeptides conforms to the EU numbering system.

14. The antibody construct according to claim 1, comprising a first heavy chain (H1) containing the sequence described in SEQ ID NO: 100, a second heavy chain (H2) containing the sequence described in SEQ ID NO: 110, and a light chain (L1) containing the sequence described in SEQ ID NO:

114.

15. The antibody construct according to claim 14, wherein one or more of the heavy chains H1 and H2 further comprises a C-terminal lysine residue.

16. A pharmaceutical composition comprising an antibody construct according to any one of claims 1 to 15, and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.

17. A nucleic acid molecule or set of nucleic acid molecules encoding one or more, two or more, or three or more polypeptide chains that form an antibody construct according to any one of claims 1 to 15.

18. A cell comprising a nucleic acid molecule or a set of nucleic acid molecules as described in claim 17.

19. A pharmaceutical product for the treatment of cancer, comprising an antibody construct according to any one of claims 1 to 15.

Images