Multivalent 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 antigens on cytotoxic effector cells and tumor associated antigens (TAAs) on tumor cells. Also disclosed are pharmaceutical compositions comprising such antibody constructs and methods of making and using such constructs and compositions, for example, for the treatment of cancer.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to and the benefit of U.S. Provisional Application No. 63 / 465,483 filed on May 10, 2023, U.S. Provisional Application No. 63 / 458,621 filed on April 11, 2023, and U.S. Provisional Application No. 63 / 393,633 filed on July 29, 2022, the entire contents of each of which are incorporated herein by reference for all purposes.

[0003] Sequence Listing

[0004] This application contains a sequence listing, which has been submitted electronically in XML format and is incorporated herein by reference in its entirety. A copy of the XML file was created on XXXX, is named XXXX, and is XXXX bytes in size. Technical Field

[0005] The present disclosure relates to multivalent bispecific antibody constructs capable of targeting antigens on the surface of cytotoxic effector cells and tumor-associated antigens (TAAs) located on tumor cells, and methods for making and using such constructs, e.g., for treating cancer. Background Art

[0006] Cancer still poses a major unmet medical need, despite considerable progress in cancer treatment over the past few decades. The TAA mesothelin (MSLN) can be targeted to treat MSLN + Tumors, particularly solid tumors, include those characterized by mesothelioma, ovarian, lung, colon and pancreatic cancer, and triple-negative breast cancer. Although current standards of care as well as recently developed therapies for these indications have shown some clinical progress, which validates MSLN as a promising target for cancer treatment, significant systemic side effects and limited efficacy have been reported, primarily due to the narrow therapeutic window of these modalities. Summary of the invention

[0007] In various embodiments, the present disclosure relates to multivalent bispecific antibody constructs that can bind to antigens on the surface of cytotoxic effector cells and tumor-associated antigens (TAAs) located on tumor cells. In some embodiments, the present disclosure relates to trivalent bispecific antibody constructs that can bind to cluster of differentiation 3 (CD3) on immune cells (e.g., T cells) and MSLN on tumor cells.

[0008] In various embodiments, the present disclosure relates to an antibody construct 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, wherein both the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN); and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and c) the second scFv domain is coupled 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 H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125. In various embodiments, the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0009] In various embodiments, the present disclosure relates to an antibody construct comprising: (i) a Fab domain capable of binding to an antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding to MSLN and comprising a V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; (iii) a second scFv domain capable of binding to a second tumor-associated antigen (TAA); and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and c) the second scFv domain is coupled to the N-terminus of the Fab domain. In various embodiments, the second TAA bound by the second scFv domain is also MSLN. Therefore, in certain embodiments, the second scFv domain also comprises V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0010] In various embodiments of the antibody constructs disclosed herein, the cytotoxic effector cell is a T cell. In such embodiments, the antigen bound by the Fab domain can be CD3. In various embodiments, the Fab domain is capable of binding CD3 and comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128; and V LA domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131.

[0011] In various embodiments of the antibody constructs disclosed herein, the first scFv domain and the second scFv domain each comprise a V H A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence shown in SEQ ID NO: 103; and V L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 101. In certain embodiments, for the first scFv domain and the second scFv domain, V L Linker scFv With V H Domain coupling. This linker scFv May contain amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of said amino acid sequence. In certain embodiments, n is 3. In other embodiments, n is 4.

[0012] 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 may have a V from N-terminus to C-terminus of H -Connectors scFv -V L In some embodiments, the first scFv domain and the second scFv domain have a V from the N-terminus to the C-terminus. H -Connectors 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 shown in SEQ ID NO: 150.

[0013] In other embodiments of the antibody constructs disclosed herein, the first scFv domain, the second scFv domain, or both scFv domains may have a V from N-terminus to C-terminus of L -Connectors scFv -V H In various embodiments, the first scFv domain and the second scFv domain have a V domain from the N-terminus to the C-terminus. L -Connectors 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 shown in SEQ ID NO: 149.

[0014] In various embodiments of the antibody constructs disclosed herein, the Fab domain comprises or consists of a heavy chain H or a portion thereof, which heavy chain H or a portion thereof comprises V from N-terminus to C-terminus. H A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 105, and C H1 A Fab domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 106. The Fab domain further comprises a light chain L, which comprises V from the N-terminus to the C-terminus. L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 115, and C L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 116, wherein the light chain L is identical to the V of the heavy chain H H -C H1 In 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 shown in SEQ ID NO: 114.

[0015] In various embodiments of the antibody constructs disclosed herein, the antibody construct comprises a first heavy chain H1 comprising, from N-terminus to C-terminus, scFv2-V H -C H1 , wherein the C-terminus of the second scFv domain is fused to the V H In some embodiments, the C-terminus of the second scFv domain is coupled via a linker scFv-Fab With V H N-terminal coupling of the domains to generate the heavy chain domain structure: scFv2-linker scFv-Fab -V H -C H1 This type of connector scFv-Fab May contain amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of the amino acid sequence. In various embodiments, the first heavy chain H1 further comprises a first Fc polypeptide, and wherein the C H1 The C-terminus of the domain is coupled to the N-terminus of the first Fc polypeptide, thereby forming the following H1 domain structure: scFv2-linker scFv-Fab -V H -C H1 -CH2-CH3. In certain embodiments, C H1 Linker Fab-Fc Coupled to the CH2 domain of the first Fc polypeptide. Fab-Fc The hinge region of the immunoglobulin (Ig) may include 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-Fc comprising or consisting of an amino acid sequence having at least about 70%, 80%, 90% or 100% sequence identity to the sequence shown in SEQ ID NO: 107. Thus, in certain embodiments, H1 may comprise the following domain structure: scFv2-linker scFv-Fab -V H -C H1 -Connectors Fab-Fc -CH2-CH3 or consists of the domain structure.

[0016] In various embodiments of the antibody constructs disclosed herein, the dimeric Fc domain is a heterodimeric Fc domain. This heterodimeric Fc domain comprises a first Fc polypeptide and a second Fc polypeptide. Unless otherwise indicated, this first Fc polypeptide and the second Fc polypeptide are interchangeable and can be coupled to a Fab domain or a first scFv domain, as further described herein.

[0017] In certain embodiments, one of the first Fc polypeptide or the second Fc polypeptide 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). Compared to the corresponding wild-type IgG1 CH3 domain sequence, this first CH3 domain may comprise one or more amino acid substitutions. In some embodiments, the one or more amino acid substitutions are selected from the following amino acid substitution sets: L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V, wherein the numbering of the amino acid residues in the first Fc polypeptide is according to the EU numbering system. In certain embodiments, compared to the corresponding wild-type IgG1 CH2 domain, the first CH2 domain comprises one or more amino acid substitutions that reduce or eliminate binding to an Fc receptor (e.g., an Fcγ-receptor).

[0018] In various embodiments herein, the trivalent bispecific 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 shown in SEQ ID NO: 100. In certain embodiments, such H1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 100.

[0019] In various embodiments of the antibody constructs disclosed herein, the antibody construct further comprises a second heavy chain H2, which comprises, from N-terminus to C-terminus, a first scFv domain coupled to the other of the first Fc polypeptide or the second Fc polypeptide, thereby forming a domain structure: scFv1-CH2-CH3. In certain embodiments, the C-terminus of the first scFv domain is coupled to the first scFv domain via a linker. scFv-Fc Coupled to the N-terminus of a second Fc polypeptide, such as the N-terminus of a second CH2 domain. scFv-Fc The linker 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 comprising or consisting of an amino acid sequence having at least about 70%, 80%, 90% or 100% sequence identity to the sequence shown in SEQ ID NO: 111. Thus, in some embodiments, H2 may comprise the following domain structure: scFv1-linker scFv-Fc -CH2-CH3 or consists of the domain structure.

[0020] In various embodiments of the antibody constructs disclosed herein, the other of the first Fc polypeptide or the 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, wherein the numbering of the 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 chain H1 and H2, and promote the formation of heterodimeric Fc domains H1-H2 compared to the formation of corresponding homodimeric Fc domains (e.g., domains consisting of H1-H1 or H2-H2). Similar to the first CH2 domain of the first Fc polypeptide, the second CH2 domain may comprise one or more amino acid substitutions that reduce or eliminate binding to an Fc receptor (e.g., an Fcγ-receptor) compared to a corresponding wild-type IgG1 CH2 domain.

[0021] In various embodiments herein, the trivalent bispecific antibody construct comprises 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 shown in SEQ ID NO: 110. In certain embodiments, H2 comprises or consists of the amino acid sequence shown in SEQ ID NO: 110.

[0022] In various embodiments of the present disclosure, the antibody construct is bispecific for CD3 and MSLN and is trivalent, ie, monovalent for CD3 via the Fab domain and bivalent for MSLN via the first and second scFv domains.

[0023] In various 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 shown 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 shown 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 shown in SEQ ID NO: 114.

[0024] 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 shown in SEQ ID NO: 100, (ii) a second heavy chain (H2) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 110, and (iii) a light chain (L1) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 114.

[0025] In some embodiments, the antibody construct of the present disclosure comprises a first heavy chain (H1) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 171, a second heavy chain (H2) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 172, and a light chain (L1) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 114.

[0026] 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 shown 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 shown 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 shown in SEQ ID NO:114.

[0027] In various embodiments of the trivalent, bispecific antibody constructs disclosed herein, such antibody constructs, when present at a concentration of about 10 picomoles / liter (pM) for about 72 hours in a cell population comprising immune cells expressing CD3 and tumor cells expressing about 500,000 MSLN / cell, are capable of achieving one or more, two or more, or all of the following: A) inducing the production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL TNFα; B) inducing the production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL IFNγ; and / or C) inducing the production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL IL-2.

[0028] In various embodiments of the trivalent bispecific antibody constructs disclosed herein, such antibody constructs are capable of inducing at least 10-fold higher T cell-mediated cytotoxicity against tumor cells expressing MSLN compared to a similar antibody construct comprising two anti-MSLN Fab domains instead of the first and second anti-MSLN scFv domains.

[0029] In various embodiments of the trivalent, bispecific antibody constructs disclosed herein, the binding affinity of such antibody constructs to MSLN as measured by surface plasmon resonance (SPR) is about 0.7 nM, 0.8 nM, 0.9 nM, or about 1 nM.

[0030] In certain embodiments of the trivalent, bispecific antibody constructs disclosed herein, the binding affinity of such antibody constructs for CD3 as measured by SPR is about 30 nM, 40 nM, 50 nM or about 60 nM.

[0031] In certain embodiments of the trivalent, bispecific antibody constructs disclosed herein, such antibody constructs are capable of inducing the production of proinflammatory cytokines by cytotoxic effector cells in an MSLN-dependent manner, wherein the MSLN-dependent activation of the cytotoxic effector cells is determined by measuring at least about 20-fold, 50-fold, 100-fold, or at least about 1000-fold reduction in cytokine production between a first cell population comprising a first tumor cell and an immune cell and a second cell population comprising a second tumor cell and an immune cell, wherein the MSLN expression of the first tumor cell is at least about 3, 4, or 5-fold higher than the MSLN expression of the second tumor cell.

[0032] In certain embodiments of the trivalent bispecific antibody constructs disclosed herein, when such antibody constructs are administered to a mammalian subject, the constructs are capable of reducing the volume of a tumor expressing MSLN in the subject by at least about 5% after the antibody constructs are administered to the subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg, and when administered 4 times per week. In some embodiments, the 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.

[0033] In some embodiments herein, the trivalent, bispecific antibody constructs of the present disclosure can have a stability of at least about 97%, 98% or 99%, i.e., at least about 97%, 98% or 99% intact antibody construct is measured after incubation of the antibody construct in an aqueous solution or buffer system comprising sucrose at 40° C. for 14 days and as measured using SEC.

[0034] In some embodiments, the present disclosure relates to a pharmaceutical composition comprising a trivalent, bispecific antibody construct of the present disclosure and a pharmaceutically acceptable carrier, excipient, diluent, or a combination thereof.

[0035] Certain embodiments of the present disclosure relate to a nucleic acid molecule or a set of nucleic acid molecules encoding one or more polypeptide chains of an antibody construct described herein. In some embodiments, a nucleic acid molecule or a set of nucleic acid molecules is described herein, encoding one or more polypeptide chains of a bispecific trivalent antibody construct (e.g., construct v32523 or v21812) described herein. In some embodiments, a nucleic acid molecule or a set of nucleic acid molecules is described herein, encoding three polypeptide chains of a bispecific trivalent antibody construct v32523, whose amino acid sequence is in SEQ ID NO: 100, 110 and 114, or in SEQ ID NO: 171, 172 and 114, as further described herein.

[0036] Further described herein is a vector or collection of vectors comprising a nucleic acid molecule or collection of nucleic acid molecules encoding one or more polypeptide chains of the antibody constructs described herein.

[0037] Further described herein is a host cell comprising a nucleic acid molecule or a collection of nucleic acid molecules encoding one or more polypeptide chains of an antibody construct described herein, and / or a vector or a collection of vectors comprising a nucleic acid molecule or a collection of nucleic acid molecules encoding one or more polypeptide chains of an antibody construct described herein. In some embodiments, the host cell can be a mammalian host cell.

[0038] Certain embodiments of the present disclosure relate to a method for producing an antibody construct as described herein. In some embodiments, this method relates to the production of a trivalent bispecific antibody construct comprising two anti-MSLN scFv domains and one anti-CD3 Fab domain, as further described herein. In some embodiments, a method for producing such a trivalent bispecific antibody construct includes (a) obtaining a host cell culture comprising at least one host cell, the at least one host cell comprising one or more nucleic acid molecules encoding one or more polypeptide chains of an antibody construct; and (b) recovering the antibody construct from the host cell culture. In some embodiments, this method may further include purifying the antibody construct using, for example, a chromatographic purification method as described herein.

[0039] In some embodiments, the amino acid sequence of one or more heavy chains of the antibody construct herein may include an additional C-terminal lysine residue for production in a host cell. Therefore, in some embodiments, the amino acid sequence shown in SEQ ID NO:100 and SEQ ID NO:110 may further include a C-terminal lysine ("K") residue. The exemplary amino acid sequences of the two heavy chains H1 and H2 of the antibody construct v32523 comprising the C-terminal lysine residue are shown in SEQ ID NO:171 and 172, respectively. After expression in a host cell, this C-terminal lysine residue can be enzymatically cleaved from one or more heavy chains. In some embodiments, a trivalent bispecific antibody construct comprising the heavy chain amino acid sequence shown in SEQ ID NO:171 and 172 and the light chain amino acid sequence shown in SEQ ID NO:114 may be referred to as v38490.

[0040] Additional embodiments of the present disclosure relate 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 a trivalent bispecific antibody construct of the present disclosure. In some embodiments, when the concentration of the antibody construct in the cell population ranges (e.g., increases) from about 10 -2 Picomolar (pM) to about 10 2 pM, the anti-tumor immune response provides a dose-dependent reduction of live tumor cells in the cell population by at least about 30%, 40%, 50% or 60%, wherein the MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0041] Certain embodiments of the present disclosure relate to a method of inhibiting the proliferation 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 a trivalent bispecific antibody construct of the present disclosure. In certain embodiments, when up to about 10 -2 pM, 10 -1 When an increase of up to about 5% in the count of live tumor cells in a cell population is observed at a concentration of pM or 1 pM of the antibody construct over a period of at least about 5, 10, 20 or 48 hours, the proliferation of tumor cells is inhibited, wherein the MSLN expression of the tumor cells is at least about 15,000 MSLN / cell, and the ratio of immune cells to tumor cells in the cell population is about 5: 1. In some embodiments, the proliferation of tumor cells is inhibited when the count of live tumor cells is reduced by at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, or at least about 50% over a given period of time.

[0042] Other embodiments of the present disclosure relate to a method of killing tumor cells expressing MSLN, the method comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of a trivalent bispecific antibody construct of the present disclosure. In such embodiments, when the concentration of the antibody construct ranges (e.g., increases) from about 10 -2 pM to about 10 2 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 at pM, wherein MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0043] In certain embodiments of these methods, the immune cells comprise or consist of T cells. Such T cell populations may comprise one or more different types of T cells known in the art.

[0044] In certain embodiments of the methods described herein, the trivalent bispecific antibody construct binds CD3 on immune cells and MSLN on tumor cells. In various embodiments, this binding of two antigens on different cells can include simultaneous binding to 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.

[0045] In some embodiments, the present disclosure relates to a method for inhibiting the growth of tumors expressing MSLN or reducing the volume of such tumors in a subject, the method comprising administering an effective amount of a trivalent bispecific antibody construct of the present disclosure to a subject. In various embodiments, the antibody construct engages CD3 on immune cells of the subject and MSLN on tumor cells, which triggers 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 the reduction of tumor volume are at least partially triggered by binding the antibody construct to CD3 on immune cells and MSLN on tumor cells in the tumor environment of the subject and forming a TCR-independent artificial immune synapse. As described herein, the immune cell can be a T cell. In various embodiments, and as further described herein, immune cell (e.g., T cell) activation can trigger the release of one or more cytokines from activated immune cells, such as TNFα, IFNγ and / or IL-2. In various embodiments, antibody constructs are administered to subjects 4 times a week (Q7Dx4), and the tumor growth of subjects is suppressed (for example, tumor growth does not exceed 5% during the study) for at least about 20 days, 30 days or at least about 50 days. In certain embodiments, about 15 days after the start of treatment, tumor volume is reduced by at least about 10%, 20%, 30%, 40% or at least about 50%. In such embodiments, antibody constructs of about 1mg / kg, 1.5mg / kg or about 3mg / kg are administered to subjects. The subject in the method described herein can be a rodent, a non-human primate or a human. In some embodiments, antibody constructs are administered intravenously.

[0046] In some embodiments, the disclosure relates to a trivalent, bispecific antibody construct according to any embodiment described herein for use in cancer treatment.

[0047] In some embodiments, the present disclosure relates to the use of an antibody construct according to any embodiment described herein in the manufacture of a medicament for cancer treatment.

[0048] In some of these embodiments, the cancer is a cancer that expresses MSLN. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Embodiments disclosed herein are illustrated by way of example and not limitation in the accompanying drawings. The description and drawings are for illustrative purposes only and to aid understanding and are not intended as definitions of the limits of the antibody constructs, pharmaceutical compositions and methods disclosed herein.

[0050] Figure 1 shows a schematic diagram of the geometry (e.g., relative orientation, relative spatial positioning and / or connectivity of binding domains, Fc domains, etc.) and format (e.g., antigen valency; type of binding domain, such as scFv, Fab, etc.; presence / absence / type of Fc domain, etc.) of antibody constructs according to certain embodiments of the present disclosure. Figure 1A Schematic diagrams showing the geometry and format of trivalent bispecific antibody constructs of the present disclosure capable of monovalently binding to CD3 (via one Fab domain) and divalently binding to MSLN (via two scFv domains) according to various embodiments of the present disclosure (e.g., as described herein with respect to variants v21812, v32523, etc.). Figure 1B Schematic diagram showing the geometry and format of a trivalent bispecific antibody construct capable of binding monovalently to CD3 (via one Fab domain) and divalently to MSLN (via two Fab domains), e.g., as in a 2+1 Fab 3 TCB benchmark construct, v29191 or Fab 3 Reference construct v21791 was used. Figure 1C Schematic diagram showing the geometry and format of a bivalent, bispecific antibody construct capable of monovalently binding to CD3 (e.g., via one scFv domain) and monovalently binding to MSLN (e.g., via one Fab domain), e.g., as used in construct v21815 herein. Figure 1D The ability to monovalently bind to CD3 (e.g., via a scFv domain), monovalently bind to MSLN (e.g., via a V HH domain) and binds monovalently to human serum albumin (HSA or Alb, e.g., via a V HH Schematic diagram of the geometry and format of the trivalent, trispecific MH6T-TriTAC benchmark polypeptide construct v31805 (Fig.

[0051] Figure 2A and Figure 2B Shown is the results of comparison with the bivalent bispecific construct v21815 ( Figure 2A ) and trivalent bispecific anti-(MSLNxCD3) Fab 3 Antibody construct v21791 ( Figure 2B ) were compared with the trivalent bispecific antibody (MSLN (scFv 2 )xCD3(Fab)) Antibody construct v21812 and CD3 + Binding of CD3 on Jurkat cells.

[0052] Figure 3A and Figure 3BThe trivalent bispecific anti-(MSLNxCD3) antibody construct v21812, the bivalent bispecific construct v21815 and the trivalent bispecific anti-(MSLNxCD3) Fab as assessed by flow cytometry are shown. 3 Antibody construct v21791 and CD4 + ( Figure 3A ) and CD8 + (FIG.3B) Binding of CD3 on T cells.

[0053] Figure 4A and Figure 4B The bivalent bispecific construct v21815 binds to MSLN as measured by flow cytometry. + Binding of MSLN on H226 cells ( Figure 4A ), measured K D The value was about 1.73 nM, and the trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 (ie, comprising 2 anti-MSLN scFv + 1 anti-CD3 Fab) and Fab as measured by flow cytometry 3 Antibody construct v21791 and MSLN + Binding of MSLN on H226 cells ( Figure 4B ), measured K D The values ​​were approximately 0.47 nM (v21812) and 0.74 nM (v21791), respectively.

[0054] Figures 5A-5D The trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 as well as constructs v21815 and v21791 are shown to direct T cells from healthy donors to kill MSLN+ target tumor cells expressing different levels of MSLN: MCF-7 cells (approximately 500 MSLN / cell, Figure 5A ), BxPC3 (about 17,000MSLN / cell, Figure 5B ) and OVCAR-3 (about 768,000MSLN / cell, Figure 5C ). The data showed that construct v21812 showed the most effective killing of tumor cells with medium and high MSLN expression, respectively. Figure 5B and Figure 5C As shown for BxPC3 and OVCAR-3 cells. Figure 5D It is shown that when the 3 ; Anti-CD3 Fab x Anti-MSLN Fab 2The trivalent bispecific construct v21812 induced significantly higher IL-2 release from T cells when compared to 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 the corresponding test articles for 3 days.

[0055] Fig. 6A The trivalent bispecific antibody construct v21812 was shown to inhibit tumor growth in OVCAR-3 tumor-bearing NOG mice transplanted with human PBMCs and even reduce tumor volume during the experiment. Antibody construct v21812 ((MSLN)-scFv with specially designed format and geometry 2 x(CD3)-Fab, see, Figure 1A ) and the corresponding trivalent bispecific Fab 3 Construct v21791 (see, e.g. Figure 1B ) (which contains the same anti-CD3 and anti-MSLN paratope sequences) compared to the significant difference in tumor inhibition was unexpected and surprising, not only demonstrating that the scFv described herein 2 The xFab construct outperformed the conventional construct in in vivo performance and also demonstrates the impact that the construct format (eg, use of an anti-MSLN scFv domain versus a Fab domain) can have on its properties. Figure 6B The results of statistical analysis of experimental data obtained from this in vivo study are shown, further demonstrating that the Fab containing the same anti-CD3 and anti-MSLN paratopes as v21812 3 Construct v21812 achieved a significantly lower tumor growth constant compared to construct v21791. Vehicle control was used as a negative control in this study.

[0056] Figures 7A-7C Shown are three different cell lines with different MSLN expression levels and surface presentation: OVCAR-3 ( Fig. 7A , about 700,000MSLN / cell), H292( Figure 7B , about 170,000MSLN / cell) and IGROV-1 ( Figure 7C ) (which was used as an MSLN negative control cell line), and the tumor cell killing activity of the trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 as well as other MSLN-targeting constructs v29045 and v29048 with lower affinity anti-MSLN paratopes. Fig. 7A Shown are (i) two variants v29045 and v29048 with lower affinity MSLN paratope, (ii) a trivalent bispecific Fab3 Benchmark construct (2+1Fab 3 The trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 showed the highest cell killing activity against high MSLN expressing OVCAR-3 cells in pan T cell co-cultures compared to (iii) TCB benchmark, v29191) and (iv) negative control (v31926, anti-CD3 / anti-HA construct). Figure 7B It was shown that the trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 exhibited the highest cell killing activity against moderate MSLN expressing H292 cells in pan T cell co-culture, which was comparable to the anti-(MSLNxCD3) benchmark antibody construct MH6T-TriTAC (v31805, see, e.g. Figure 1D The activity of the Fab was similar to that of the Fab 29045 and Fab 29048, and was further compared to (i) two variants with lower affinity anti-MSLN paratopes, v29045 and v29048, (ii) a trivalent bispecific Fab 3 The benchmark construct (v29191) and (iii) the negative control (v31926) were compared. Figure 7C It was shown that all tested constructs did not have any significant effect on viable cell counts when tested in the MSLN-negative tumor cell line IGROV-1 / pan T cell co-cultures.

[0057] Figures 8A-8C It was shown that in particular the trivalent bispecific antibody construct v21812 and to a lesser extent the constructs v29045 and v29048 with lower anti-MSLN affinity paratopes induced the production of the proinflammatory cytokine tumor necrosis factor alpha (TNFα, Fig. 8A ), interferon-γ (IFNγ or IFNg, Figure 8B ) and interleukin-2 (IL-2, Figure 8C ) and further cooperated with 2+1Fab 3 TCB benchmark construct v29191 (with Fab 3 Construct v21791 (same format but different paratope sequence) was compared.

[0058] Fig. 9 The trivalent bispecific anti-(MSLNxCD3) antibody construct v21812 of the present disclosure was shown to significantly inhibit tumor growth and reduce tumor volume during the study in OVCAR-3 tumor-bearing NOG mice transplanted with human PBMCs. Construct v21812 was superior to other tested constructs (including 2+1 Fab 3 TCB benchmark construct (v29191, see, Figure 1B The construct version of )) more significantly inhibited tumor growth. ** indicates p value < 0.01, *** indicates p value < 0.001.

[0059] Fig.10 The trivalent bispecific antibody construct v32523 (same domain sequence as v21812, however, with V in v21812) with improved stability (compared to v21812) is shown as measured by flow cytometry and compared to the benchmark control constructs v29191 and v31805. H -Connector scFv -V L In comparison, the two anti-MSLN scFv domains have a V L -Connectors scFv -V H The domain structure of v32523 is shown in Figure 2, and compared with (G4S)3 in v21812, the linker in v32523 is scFv Having (G4S)4 sequence) binds to MSLN+OVCAR-3 cells.

[0060] Figures 11A-11C As measured by flow cytometry, the + cell( Fig.11A )、CD8 + cell( Fig. 11B ) and CD4 + and CD8 + A mixture of cells ( Fig. 11C ) in cell culture with the trivalent bispecific antibody construct v32523 and the benchmark control constructs v29191 and v31805 and CD3 + Binding of human pan T cells.

[0061] Figures 12A-12F Shown is MSLN-dependent T cell-mediated cytotoxicity elicited by the trivalent bispecific antibody construct v32523 using cell lines with varying degrees of MSLN expression and compared to the benchmark construct v31805. Fig. 12A Shown with MSLN 低 Compared with OVTOKO and MCF7 cells, the expression of 高 The percentage of tumor cell killing induced by antibody construct v32523 in H292 cells expressing Fig. 12B Shown in MSLN 高 Expression of H292 cells and MSLN 低 A head-to-head comparison of tumor cell killing of the two tested constructs, v32523 and v31805, in A375 cells expressing . Fig. 12C , about 170,000MSLN / cell), OVCAR8 ovarian cancer cells ( Fig.12D , about 85,000MSLN / cell), HCT-116 colon cancer cells ( Fig.12E , about 35,000MSLN / cell) and H2452 mesothelioma cells ( Fig.12F The trivalent bispecific construct v32523* was tested for T cell-dependent cellular cytotoxicity in the presence of 50 μM HO (about 1 1,000 MSLN / cell) (*test article data are shown as square dotted lines) and compared to a negative control.

[0062] Figures 13A-13D The trivalent bispecific antibody construct v32523 is shown to direct T cells from healthy donors to kill MSLN+ tumor cells expressing different levels of MSLN: H292 (approximately 152,000 MSLN / cell, Fig.13A ), HCT116 (about 35,000MSLN / cell, Fig. 13B ), H2452 (about 11,000MSLN / cell, Fig. 13C ) and A549 (about 11,000MSLN / cell, Fig.13D The trivalent bispecific antibody construct v32523 of the present disclosure was at least as effective as the benchmark control construct v29191 (2+1 Fab 3 constructs) and v31805 (MH6T-TriTAC) were as effective or better than these constructs.

[0063] Figures 14A-14D Shown in the MSLN + H292 and MSLN + The trivalent bispecific antibody construct v32523 and the v31805 (MH6T-TriTAC) benchmark control construct induced TNFα in T cells co-incubated with HCT116 cells for 72 hours (E:T ratio of 5:1). Fig.14A and Fig. 14C ) and IFNγ( Fig. 14B and Fig.14D ) production. The data show that the trivalent bispecific antibody construct v32523 induces significantly lower amounts of two cytokines in the cell line HCT116 with lower MSLN expression, demonstrating that the antibody constructs described herein provide TAA (e.g., MSLN)-dependent T cell-mediated cytotoxicity properties. Such antigen-dependent properties can achieve a strong anti-tumor immune response in an environment with high MSLN expression, while off-target tissues with no or low MSLN expression may not be exposed to T cell-mediated cytotoxic effects, or at least to a much lower degree.

[0064] Fig.15 It is shown that the trivalent bispecific antibody construct v32523 as well as the v31805 (MH6T-TriTAC) benchmark control construct induced T cell proliferation in a comparable dose-dependent manner when incubated with MSLN+OVCAR-3 cells (E:T ratio of 10:1) for 72 hours.

[0065] Figures 16A-16B The results show that in OVCAR-3 tumor-bearing NOG mice transplanted with human PBMC (high MSLN expression, Fig.16A ) and HCT116 tumor-bearing NOG mice (moderate MSLN expression, Fig. 16B ) 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 (arrows indicate administration time points).

[0066] Figures 17A-17C In OVCAR-3 tumor-bearing mice ( Fig.17A , at doses of 0.1, 1, and 3 mg / kg) and naive mice ( Fig. 17B , at doses of 0.1, 0.5, 1 and 3 mg / kg) in the pharmacokinetic profiles and serum concentrations at different time points after injection of the trivalent bispecific antibody construct v32523 using different doses. Fig. 17C Pharmacokinetic profiles and serum concentrations at different time points after injection of different doses (ie, 0.1, 1 and 3 mg / kg) of the trivalent bispecific antibody construct v32523 are shown for PBMC donors A, B and C. The effective lower limit of quantitation (LLOQ) was determined to be approximately 781 ng / mL.

[0067] Fig.18 It is shown that the trivalent bispecific antibody construct v32523 of the present disclosure exhibited significantly greater anti-tumor activity at two doses of 1 mg / kg and 3 mg / kg (administered intravenously) in HCT116 tumor-bearing mice transplanted with human PBMCs when compared to various identical doses of the benchmark control v31805 (MH6T-TriTAC). The dose of construct v31805 was matched to that of v32523 by PK analysis to account for the different in vivo half-lives and to allow a head-to-head comparison of the two constructs.

[0068] Fig.19 The results are shown compared to the benchmark controls v31805 (MH6T-TriTAC) and v29191 (2+1 Fab 3) in HCT116 tumor-bearing mice transplanted with human PBMCs (intravenous, 3 mg / kg).

[0069] Fig. 20 It is shown that the trivalent bispecific antibody construct v32523 of the present disclosure exhibited significantly greater anti-tumor activity (iv, 3 mg / kg) as measured by a significant reduction in tumor volume in OVCAR-3 tumor-bearing mice transplanted with human PBMCs when compared to the same dose of the benchmark control construct v31805.

[0070] Figures 21A-21C The following conditions are shown: (i) T cells alone or (ii) with MSLN + In the presence of T cells co-cultured with tumor cells, the trivalent bispecific antibody construct v32523 induced MSLN-dependent cytokine production (IL-2( Fig.21A ) and TNFα( Fig. 21B )) and T cell proliferation ( Fig. 21C ). Cytokine production was assessed by co-culturing MSLN-expressing tumor cell lines with human pan T cells and treating those cells with v32523 or negative control for 3 days. T cell proliferation was assessed by co-culturing CFSE-labeled T cells with OVCAR3 cells and treating those cells with v32523 for 5 days. Proliferation was measured by flow cytometry.

[0071] Figures 22A-22B Shown is when compared to the clinical benchmark construct v31805 and when at high (5:1, Fig.22A ) and low (1:5, Fig. 22B )E:T ratio, the trivalent bispecific construct v32523 exhibited greater antitumor activity ( Figures 22A-22B Use the same legend to identify the test conditions).

[0072] Figures 23A-23C (i) Serum levels of fibrinogen in cynomolgus monkeys after injection of the trivalent bispecific antibody construct v38490 at 1 mg / kg, 10 mg / kg and 30 mg / kg doses ( Fig.23A ), (ii) transient increase of IL-6 ( Fig. 23B ) and (iii) serum half-life ( Fig.23C ).

[0073] Figures 24A-24B As shown by size exclusion chromatography (SEC, Fig.24A ) and dynamic light scattering (DLS, Fig. 24B) analysis, the trivalent bispecific antibody construct v32523 remained monodisperse after incubation in A5Su buffer at 40°C for 14 days, demonstrating that the purity of v32523 after incubation was at least about 99% as measured by SEC, with other high molecular weight species (HMWS) less than about 1%. DETAILED DESCRIPTION

[0074] The present disclosure relates to multivalent bispecific antibody constructs that can address the improvement of MSLN + Unmet medical needs of tumor treatment options. Specifically, compared with conventional constructs, the antibody constructs described herein can allow a wider therapeutic window due to smaller side effects, potentially enabling the use of more effective doses. As further demonstrated herein, it was unexpectedly found that such improved properties are achieved, at least to a large extent, by improving the format of the antibody construct. In some embodiments, the format of the regulatory construct can induce stronger TAA (e.g., MSLN)-dependent T cell activity in vitro and / or in vivo, which can provide more T cell-mediated cytotoxicity for tumors while significantly reducing cytotoxic activity in non-tumor (e.g., MSLN low or MSLN negative) tissues.

[0075] As further described herein, MSLN + At least one of the remaining technical problems in tumor therapy can be addressed by molecular targeting of MSLN + In other aspects including enhanced anti-tumor activity, the antibody constructs of the present disclosure have been designed and generated to enhance the ability of cytotoxic effector cells (e.g., CD3+) to target tumor cells in a more TAA (e.g., MSLN)-dependent manner (e.g., when compared to conventional methods). + The invention relates to the ability to guide immune cells, such as T cells, to tumor cells to address this shortcoming. 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 tolerance) of multivalent (e.g., trivalent) bispecific antibody constructs over existing clinical benchmark constructs. Antibody constructs with specific formats and geometric structures are further disclosed herein, and it is demonstrated how the in vitro and in vivo performance (e.g., anti-tumor activity, TAA-specific tumor killing, etc.) of the constructs can depend on their formats, for example, by using anti-TAA scFv domains instead of anti-TAA Fab domains, etc.

[0076] The present disclosure further describes pharmaceutical compositions comprising the multivalent bispecific antibody constructs (such as trivalent bispecific antibody constructs) of the present disclosure, as well as methods of making and using such constructs and compositions for TAA-positive tumor treatment.

[0077] Often, the positive recitation of a feature in one embodiment serves as a basis for excluding that feature in an alternative embodiment. In particular, where a list of options is presented for a given embodiment or claim, it should be understood that one or more options may be deleted from the list and the shortened list may form an alternative embodiment, whether or not such alternative embodiments are specifically mentioned.

[0078] It is also contemplated that any embodiment discussed herein can be implemented with respect to any antibody construct, method, use or composition disclosed herein, and vice versa. In addition, modifications of the specific embodiments described herein that are obvious to those skilled in the art are intended to be included within the scope of the claims described herein.

[0079] definition

[0080] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art.

[0081] Unless otherwise stated, as used herein, in the context of a numerical value or range, the term "about" generally refers to ±10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2% or ±1% of the numerical value or range listed or claimed. In various embodiments, the term "about" refers to a change of approximately ±10% from a given value or range. In other embodiments, the term "about" refers to a change of approximately ±5% from a given value or range. In other embodiments, the term "about" refers to a change of approximately ±1% from a given value or range. Unless otherwise stated, it should be understood that such changes are always included in any given value provided herein, whether specifically mentioned or not.

[0082] When used in conjunction with the term "comprising" in this document, the use of the word "a / an" may mean "one", but it is also consistent with the meaning of "one or more", "at least one", and "one or more than one".

[0083] As used herein, the terms "comprising", "having", "including" and "containing" and their grammatical variations are inclusive or open-ended when used in conjunction with constructs, compositions, uses or methods herein, and do not exclude additional, unlisted elements and / or method steps. When used in conjunction with constructs, compositions, uses or methods herein, the term "essentially consisting of..." indicates that additional features, elements and / or method steps may be present, but these additions do not substantially affect the manner in which the listed constructs, compositions, methods or uses function. The term "consisting of..." excludes the presence of additional elements and / or method steps when used in conjunction with constructs, compositions, uses or methods herein. Antibody constructs, compositions, uses or methods described herein as comprising certain elements and / or steps may also be essentially composed of those elements and / or steps in certain embodiments, and composed of those elements and / or steps in other embodiments, whether or not these embodiments are specifically mentioned.

[0084] 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, such as 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.

[0085] With respect to a specific result to be achieved, an "effective amount" of an antibody construct described herein or a pharmaceutical composition comprising such an antibody construct is an amount sufficient to achieve the desired result. For example, when it comes to killing cancer cells, an "effective amount" of an antibody construct or pharmaceutical composition refers to an amount of an antibody construct or composition sufficient to produce a killing effect. In some embodiments, the achievement of a desired result (e.g., cancer cell killing) can be demonstrated by measuring using one or more of the related methods described herein and / or those methods known to those skilled in the art.

[0086] The terms "Fc region", "Fc" and "Fc domain" are used interchangeably herein and refer to the C-terminal region of an immunoglobulin heavy chain that contains at least a portion of a constant region. In various embodiments, the Fc domain herein may be dimeric. Such a dimeric Fc domain may comprise a first Fc polypeptide and a second Fc polypeptide, wherein each Fc polypeptide may comprise a CH2 domain and / or a CH3 domain. Such a dimeric Fc may be homodimeric, i.e., comprising a first Fc polypeptide and a second Fc polypeptide having the same amino acid sequence, or heterodimeric, i.e., comprising a first Fc polypeptide and a second Fc polypeptide having different amino acid sequences (e.g., sequences having 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 other embodiments, and 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 Fc polypeptide and the second Fc polypeptide have an amino acid sequence that shares about 99% or less, 98% or less, or about 97% or less sequence identity.

[0087] As used herein, the term "bispecific" in the context of an antibody construct refers to a biological functional protein (e.g., an antibody construct as described herein) that "has at least bispecificity", that is, it comprises at least a first binding domain and a second binding domain, wherein such a first binding domain and a second binding domain can specifically bind to two different epitopes, such as a first epitope and a second epitope. Such a first epitope and a second epitope can be located on the same antigen or on different antigens, such as a first epitope on CD3 and a second epitope on MSLN. Therefore, in some embodiments, an antibody construct according to the present disclosure may include specificity for at least two different antigens, epitopes, or other targets. In certain embodiments, an antibody construct of the present disclosure has bispecificity for two different epitopes, wherein each epitope is located on different antigens (e.g., CD3 and MSLN). In the context of an antibody construct herein, the term "bispecific" also encompasses multispecific antibody constructs, such as trispecific antibody constructs, the latter including three binding domains, each binding domain binding to a different epitope or target, or having more than three (e.g., four, five, etc.) specific constructs.

[0088] As used herein, the term "trivalent" in the context of an antibody construct refers to a biologically functional protein (e.g., an antibody construct as 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, wherein each of the first binding domain, the second binding domain, and the third binding domain can specifically bind to an epitope and / or an antigen, e.g., CD3, MSLN, etc. The three binding domains can be specific to three different epitopes or antigens, or two or more of the three binding domains are specific to the same epitope or antigen. Therefore, in the context of an antibody construct herein, valence, e.g., monovalent, divalent, or trivalent, describes the total number of antigen binding domains of an antibody construct. Therefore, the valence of an antibody construct must be at least equal to its specificity, i.e., a bispecific antibody construct must be at least divalent. In embodiments where the antibody constructs herein are 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 monovalently engages a different epitope and / or antigen (e.g., CD3).

[0089] As used herein, the abbreviations "H", "H1", "H2", "L1", etc. are generally used as universal heavy chain and light chain identifiers, respectively. As an example, "H1" and "H2" herein can broadly refer to the first heavy chain and the second heavy chain of an antibody construct, respectively, and are thus not intended to be limited in any way to a specific heavy chain amino acid (or polynucleotide) sequence. Rather, such chain identifiers can be used herein to describe and distinguish two or more polypeptide chains of an antibody construct.

[0090] As used herein, the term "amino acid modification" in the context of an amino acid sequence of a polypeptide generally refers to an amino acid sequence of a polypeptide wherein one or more amino acid substitutions, one or more amino acid insertions, and / or one or more amino acid deletions have been introduced relative to the corresponding unmodified (e.g., WT or reference) amino acid sequence of the polypeptide.

[0091] Antibody Constructs

[0092] In various embodiments, the present disclosure relates to multivalent bispecific antibody constructs that can target 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 be monovalently bound to antigens on the surface of cytotoxic effector cells and bivalently bound to TAAs located on tumor cells. Two binding domains capable of bivalently binding TAAs can be directed to the same TAA or to two different TAAs. In various embodiments, two anti-TAA binding domains are directed to the same TAA. In some embodiments, two anti-TAA binding domains are directed to the same epitope (e.g., MSLN) on TAA.

[0093] In certain other embodiments, a trivalent, bispecific antibody construct may be capable of monovalently binding a TAA on the surface of a tumor cell and bivalently binding an antigen on the surface of a cytotoxic effector cell.

[0094] 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, wherein the cytotoxic effector cell can be an immune cell such as a T cell, and the TAA is MSLN.

[0095] In some embodiments, the present disclosure relates to a trivalent bispecific antibody construct 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 binding domain and the third binding domain is capable of binding to MSLN. In various embodiments, both the second binding domain and the third binding domain are capable of binding to MSLN. Such constructs described herein that are monovalent for CD3 and bivalent for MSLN may have the ability to guide immune cells (e.g., T cells) to tumor cells and induce local MSLN-dependent immune cell-mediated anti-tumor immune responses. In certain embodiments, the antibody constructs of the present disclosure may not induce significant TAA-independent T cells or immune responses in the absence or low abundance of TAAs, thereby having the ability to reduce off-target effects and achieve a wider therapeutic window when compared to other existing methods with less TAA dependence in anti-tumor immune responses. In the context of a TAA-independent immune response, the term "significant" generally refers to an immune response elicited by the antibody construct that is equal to or less than 10%, 5% or 3% of the immune response elicited in the context of cells or tissues containing moderate or high TAA expression, as further described herein.

[0096] Previously, certain bispecific antibodies that can direct T cells to tumor cells have been identified and tested for their efficacy in cancer treatment. TM Examples of bispecific anti-CD3-CD19 antibodies of the type described herein (bispecific T cell engagers) have been identified for use in the treatment of B cell diseases, such as relapsed B cell non-Hodgkin lymphoma and chronic lymphocytic leukemia (see, e.g., Baeuerle et al. (2009) Cancer Research 12:4941-4944) and have received FDA approval. T cell engagers against other tumor-associated target antigens have also been made, and several have entered clinical trials: e.g., AMG110 / MT110 EpCAM for lung, gastric, and colorectal cancers; AMG211 / MEDI565 CEA for gastrointestinal adenocarcinomas; and AMG 212 / BAY2010112 PSMA for prostate cancer (see, e.g., Suruadevara, CM et al., Oncoimmunology. 2015 Jun; 4(6):e1008339). Although these studies showed some clinical efficacy, they were hampered by severe dose-limiting toxicities, primarily caused by cytokine release syndrome (CRS). This resulted in a narrow therapeutic window for these agents.

[0097] The T cell engaging antibody constructs disclosed herein are intended to address these shortcomings, for example by providing reduced or even unmeasurable T cell activation in the absence or low abundance of the TAA to be targeted (e.g., MSLN), thereby focusing their cytotoxic effects primarily on the tumor (micro)environment, which is assumed to not only provide higher anti-cancer activity but also allow a wider therapeutic window with significantly reduced off-target effects when compared to conventional T cell engagement approaches.

[0098] In various embodiments, the T cell engaging trivalent bispecific antibody constructs of the present disclosure can be capable of binding CD3 on T cells (monovalent) and MSLN on tumor cells (bivalent), and can - when bound to CD3 and MSLN - form a TCR-independent artificial immune synapse between T cells and tumor cells. This can result in the T cells being activated and having a cytotoxic effect on the tumor cells, resulting in tumor cell killing, as further described herein.

[0099] A. Antibody Construct Format and Geometry

[0100] In certain embodiments, the present disclosure describes multivalent bispecific antibody constructs having formats and geometries specifically designed for improved in vivo tumor killing and tolerability properties (e.g., reduced off-target effects) compared to conventional constructs having different formats and / or geometries. In combination with the improved formats, the antigen affinity of each binding domain present in the trivalent bispecific antibody constructs described herein is rationally selected to provide strong anti-tumor activity in tissues expressing TAAs while limiting the activity of the constructs in tissues with low TAA expression.

[0101] Generally, the pattern of an antibody construct as described herein refers to the type and number of antibody construct domains (e.g., scFv, Fab, Fc, etc.) it contains, which further defines the valence of the construct to its target (e.g., CD3, MSLN, etc.) and its specificity, i.e., bispecific, trispecific, etc., as well as bivalent, trivalent, etc. As further described herein, the pattern of an antibody construct can also be expressed as a numerical combination indicating its valence and specificity to certain targets, including patterns described as 1+1 (a), 2+1 (b), 2+2 (c), etc., in the same order, referring to (a) an antibody construct capable of binding to a first antigen monovalently ("1") and binding to a second antigen monovalently ("1"), (b) binding to a first antigen divalently ("2") and binding to a second antigen monovalently ("1"), and (c) binding to a first antigen divalently ("2") and binding to a second antigen divalently ("2") The geometry of a construct herein is generally defined as the relative and three-dimensional orientation and positioning of the various domains of an antibody construct. As an example, the antibody construct comprising a Fab domain, a scFv domain and a dimeric Fc domain can have different geometric structures, wherein the domains are interconnected differently from each other, resulting in different relative orientations. In one such embodiment, the Fab domain and the scFv domain can each be coupled to the N-terminal of the dimeric Fc domain (e.g., each with the N-terminal of the first Fc polypeptide or the second Fc polypeptide). In another embodiment, one of the Fab domain and the scFv domain can be coupled to the N-terminal of the Fc domain, and the other is coupled to the C-terminal of the dimeric Fc domain. In yet another embodiment, all three domains can be coupled in series, for example, the Fab domain can be coupled to the N-terminal of the Fc domain, and the scFv domain can be coupled to the N-terminal of the Fab domain.

[0102] In various embodiments, the present disclosure describes a bispecific antibody construct 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 is a Fab domain comprising a heavy chain paired with a light chain, the heavy chain comprising a first heavy chain variable domain (V H ) and the heavy chain constant domain (C H1 ), the light chain comprising a first light chain variable domain (V L ) and the first light chain constant domain (C L In some embodiments, the second binding domain capable of binding to MSLN is a second heavy chain variable domain (V H ) and the second light chain variable domain (V L) of the scFv domain. In some embodiments, the first binding domain is via C H1 The C-terminus of the first Fc polypeptide of the heterodimeric Fc domain is coupled to the N-terminus of the first Fc polypeptide of the heterodimeric Fc domain, and the second binding domain is coupled to the second heavy chain variable domain (V H ) or the C-terminus of the second light chain variable domain (V L ) (depending on the relative orientation and domain structure of the scFv domains) is coupled to the N-terminus of the second Fc polypeptide of the heterodimeric Fc domain.

[0103] In some embodiments, the first binding domain is a Fab domain and can be directly or via a linker Fab-Fc Similarly, the second binding domain is a scFv domain and can be coupled directly or via a linker. scFv-Fc In various embodiments, the first binding domain and the second binding domain are coupled to the second Fc polypeptide via a linker. Fab-Fc and connectors scFv-Fc Coupled to the first Fc polypeptide and the second Fc polypeptide. Fab-Fc and connectors scFv-Fc In other embodiments, the linker comprises or consists of a different amino acid sequence, as further described herein. Fab-Fc and connectors scFv-Fc comprising or consisting of the same amino acid sequence.

[0104] In various embodiments of the present disclosure, the antibody construct can have bispecificity and be multivalent. In some embodiments, this construct is divalent and has bispecificity. In various other embodiments herein, the antibody construct of the present disclosure is trivalent and has bispecificity.

[0105] The trivalent bispecific antibody construct of the present disclosure may further include a third binding domain. Compared with the second binding domain, this third binding domain may be able to bind to different TAAs. In other embodiments, the third binding domain can bind to the same TAA as the second binding domain. In various embodiments, the second binding domain and the third binding domain are both able to bind to MSLN. The third binding domain may be a Fab domain, or it may be the second scFv domain. In some embodiments, the third binding domain is a Fab domain. In various other embodiments, the third binding domain is the second scFv domain, resulting in an antibody construct comprising a 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"), wherein such a first scFv domain and a second scFv domain can bind to MSLN.

[0106] In various embodiments, the trivalent bispecific antibody constructs of the present disclosure may be monovalent for CD3 and divalent for MSLN. Such antibody constructs may comprise (i) a first binding domain capable of binding to CD3 on cytotoxic effector cells, (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 binding domain and the third binding domain are each scFv domains, i.e., a first scFv domain and a second scFv domain. In some embodiments, the second binding domain and the third binding domain comprise V domains that each share at least about 80%, 90%, 95%, 97%, 99% or 100% sequence identity. H and V L Thus, in some embodiments, the second V binding domain of the second binding domain (e.g., the first scFv domain, or "scFv1") H and V L The third V domain of the second scFv domain and the third binding domain (e.g., the second scFv domain, or "scFv2"). H and V L The domains comprise or consist of the same amino acid sequence, i.e., comprise the same anti-MSLN paratope V H and V LSequence. In some embodiments, the second binding domain and the third binding domain of the trivalent bispecific antibody construct herein are capable of binding to the same epitope on MSLN. However, in other embodiments, the trivalent bispecific antibody construct herein may include a second binding domain and a third binding domain, each of which is capable of binding to a different epitope on MSLN. In other embodiments, the trivalent bispecific antibody construct may include a second binding domain and a third binding domain, each of which includes a different anti-MSLN paratope but can still bind to the same MSLN epitope.

[0107] In various embodiments of the present disclosure, a 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 coupled to the first Fc polypeptide via a first linker, the second binding domain is coupled to the second Fc polypeptide via a second linker, and the third binding domain is coupled directly or via a third linker to (a) the first binding domain, (b) the second binding domain, or (c) the heterodimeric Fc domain.

[0108] In various embodiments, the third binding domain is coupled to the first binding domain via a third linker.

[0109] In various embodiments of the trivalent bispecific antibody constructs of the present disclosure, the first binding domain capable of binding to CD3 is a Fab domain comprising a heavy chain paired with a light chain (L1), the heavy chain comprising a first heavy chain variable domain (V H ) and the heavy chain constant domain (C H1 ), the light chain comprising a first light chain variable domain (V L ) and the light chain constant domain (C L ), and the second binding domain and the third binding domain capable of binding to MSLN are each scFv domains, i.e., the first scFv domain and the second scFv domain. Therefore, the second binding domain (e.g., the first scFv domain) may comprise a second heavy chain variable domain (V H ) and the second light chain variable domain (V L ), and the third binding domain (e.g., a second scFv domain) may comprise a third heavy chain variable domain (V H ) and the third light chain variable domain (V L ).

[0110] In some embodiments of the trivalent bispecific antibody construct, the Fab domain is linked to the C H1 The C-terminus of the domain is connected via a first linker (e.g., linker Fab-Fc ) is coupled to the N-terminus of the first Fc polypeptide. The first scFv domain can be coupled to the first Fc polypeptide by its V H domain or its V L The C-terminus of the domain (e.g., depending on the V H and V L The relative orientation of the sequence) via a second linker (e.g., linker scFv-Fc ) is coupled to the N-terminus of the second Fc polypeptide. Similarly, the second scFv domain can be coupled to the second Fc polypeptide via its V H domain or its V L The C-terminus of the domain (e.g., depending on the V H and V L The relative orientation of the sequence) via a third linker (e.g., linker scFv-Fab ) is coupled to the N-terminus of the Fab domain. Since the Fab domain comprises a heavy chain and a light chain, the second scFv domain can be coupled to 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 is coupled to the N-terminus of the heavy chain of the Fab domain, for example, the V-terminus of the heavy chain. H N-terminal coupling of the domain.

[0111] In various embodiments, the linker (e.g., linker) connecting the various domains of the antibody construct Fab-Fc , Connectors scFv-Fc , Connectors scFv-Fab , etc.) are all peptide (or peptide) linkers, i.e., linkers that comprise or consist of an amino acid sequence of at least two, three, four, five, ten, fifteen, twenty, twenty-five, thirty, forty or more consecutive amino acid residues. In some embodiments, the linkers that couple each binding domain to the dimeric Fc domain Fab-Fc and connectors scFv-Fc is a peptide linker that may comprise or consist of an immunoglobulin hinge region, such as an IgG or IgG1 hinge region, as further described herein.

[0112] In some embodiments, the trivalent bispecific antibody constructs of the present disclosure may comprise or consist of two, three or more polypeptide chains, which may be expressed individually and then assembled (e.g., via heavy chain and / or light chain pairing) to form antibody constructs comprising their 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., referred to as H1 and H2) and one light chain (e.g., referred to as L1). In such embodiments, H1 may comprise a domain structure (from N-terminus to C-terminus): scFv2-V H -C H1 -CH2-CH3, wherein the second scFv domain (scFv2) is connected via V H-scFv2 or V L-scFv2 The C-terminus of the Fab domain is bound to the V H In such embodiments, H1 may comprise a domain structure: (V L -V H ) scFv2 -V H -C H1 -CH2-CH3 or consisting of said domain structure. In other embodiments, H1 may comprise a domain structure: (V H -V L ) scFv2 -V H -C H1 -CH2-CH3 or consisting of the domain structure. In addition, H2 may comprise a domain structure (from N-terminus to C-terminus): scFv1-CH2-CH3, wherein the first scFv domain may be connected via V H-scFv1 or V L-scFv1 The C-terminus of CH2 is coupled to the N-terminus of CH2 (depending on their relative orientation within scFv1). In such embodiments, H2 may comprise a domain structure: (V L -V H ) scFv1 -CH2-CH3 or consisting of said domain structure. In other embodiments, H2 may comprise a domain structure: (V H -V L ) scFv1 -CH2-CH3 or composed of the domain structure. The light chain L1 may comprise a domain structure (from N-terminus to C-terminus): V L -C L . V of H1 and L1 H -C H1A Fab domain (i.e., a first binding domain) capable of binding to CD3 may be formed. The Fab domain formation may include covalent interactions (e.g., disulfide bonds) and non-covalent interactions between the amino acid residues of H1 and L1, or a combination of the two. In addition, the CH2 and CH3 domains of H1 (e.g., corresponding to part or all of the first Fc polypeptide) and the CH2 and CH3 domains of H2 (e.g., corresponding to part or all of the second Fc polypeptide) may form a heterodimeric Fc domain, wherein the CH3 domain of H1 and the CH3 domain of H2 may each contain one or more asymmetric amino acid substitutions that promote heterodimeric Fc domain formation rather than homodimeric Fc domain formation, as further described herein.

[0113] Figure 1 provides a schematic diagram of the geometry (relative orientation and connectivity of binding domains) and format (e.g., valency, presence / absence of Fc domains, etc.) of various antibody constructs contemplated and described herein. Specifically, Figure 1A Schematic diagrams showing the geometry and format of trivalent bispecific antibody constructs according to various embodiments of the present disclosure, which consist of three polypeptide chains H1, H2 and L1 and are capable of binding to CD3 (monovalent, via one Fab domain) and MSLN (bivalent, via two scFv domains).

[0114] In various embodiments, the present disclosure describes an antibody construct, wherein the antibody construct comprises: (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, wherein both the first scFv domain and the second scFv domain 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 coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain.

[0115] In a further embodiment, the present disclosure describes an antibody construct 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, wherein both the first scFv domain and the second scFv domain are capable of binding to MSLN, and wherein at least one of the first scFv domain and the second scFv domain comprises a V HA domain comprising: a heavy chain complementary determining region 1 (HCDR1) sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and a V L A domain comprising: a light chain complementary determining region 1 (LCDR1) sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain.

[0116] In certain embodiments, both the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0117] In some of these embodiments, the antigen on the cytotoxic effector cell is CD3. The cytotoxic effector cell can be an immune cell, such as a T cell.

[0118] In various embodiments, the present disclosure describes an antibody construct comprising: (i) a Fab domain capable of binding to CD3, wherein the Fab domain comprises a V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain and a second scFv domain, wherein both the first scFv domain and the second scFv domain 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 coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain.

[0119] In various embodiments, the present disclosure describes an antibody construct comprising: (i) a Fab domain capable of binding to CD3, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain and a second scFv domain, wherein both the first scFv domain and the second scFv domain are capable of binding to MSLN, and the two scFv domains comprise V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V LA domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain.

[0120] In such embodiments, the antibody construct may be trivalent and bispecific.

[0121] 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 may comprise a domain structure (from N-terminus to C-terminus): (V L -V H ) scFv2 -V H -C H1 -CH2-CH3 or consisting of the domain structure, wherein "scFv2" represents the second scFv domain capable of binding to MSLN. H2 may comprise a domain structure (from N-terminus to C-terminus): (V L -V H ) scFv1 -CH2-CH3 or consisting of the domain structure, wherein "scFv1" represents the first scFv domain capable of binding to MSLN. L1 may comprise a domain structure (from N-terminus to C-terminus): V L -C L Or composed of the domain structure. H -C H1 and L1's V L -C LCan be (e.g., covalently) paired (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 (which—at least in part—form a first Fc polypeptide) and the CH2 and CH3 domains of H2 (which—at least in part—form a second Fc polypeptide) can form a heterodimeric Fc domain, wherein the CH3 domain of H1 and the CH3 domain of H2 can each comprise one or more asymmetric amino acid substitutions that promote heterodimeric Fc domain formation (e.g., H1-H1 or H2-H2 homodimeric Fc formation) compared to the corresponding homodimeric Fc domain (e.g., H1-H1 or H2-H2 homodimeric Fc formation).

[0122] 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 may comprise a domain structure (from N-terminus to C-terminus): (V H -V L ) scFv2 -V H -C H1 -CH2-CH3 or consisting of the domain structure, wherein "scFv2" represents the second scFv domain capable of binding to MSLN. H2 may comprise a domain structure (from N-terminus to C-terminus): (V H -V L ) scFv1 -CH2-CH3 or consisting of the domain structure, wherein "scFv1" represents the first scFv domain capable of binding to MSLN. L1 may comprise a domain structure (from N-terminus to C-terminus): V L -C L Or composed of the domain structure. H -C H1 and L1's V L -C L Can be (e.g., covalently) paired (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 (which—at least in part—form a first Fc polypeptide) and the CH2 and CH3 domains of H2 (which—at least in part—form a second Fc polypeptide) can form a heterodimeric Fc domain, wherein the CH3 domain of H1 and the CH3 domain of H2 can each comprise one or more asymmetric amino acid substitutions that promote heterodimeric Fc domain formation (e.g., H1-H1 or H2-H2 homodimeric Fc formation) compared to the corresponding homodimeric Fc domain (e.g., H1-H1 or H2-H2 homodimeric Fc formation).

[0123] Thus, in various embodiments, the antibody construct of the present disclosure may comprise or consist of three polypeptide chains H1, H2 and L1, wherein (i) H1 comprises a domain structure (from N-terminus to C-terminus): (V L -V H ) scFv2 -V H -C H1 -CH2-CH3 or consists of the said domain structure, (ii) H2 comprises the domain structure (from N-terminus to C-terminus): (V L -V H ) scFv1 -CH2-CH3 or consists of the domain structure, and (iii) L1 comprises the domain structure (from N-terminus to C-terminus): V L -C L or or consisting of said domain structure, wherein V H -C H1 and L1's V L -C L A Fab domain capable of binding to CD3 is formed, scFv1 and scFv2 are each capable of binding to MSLN, and the first Fc polypeptide and the second Fc polypeptide form a heterodimeric Fc domain, wherein the CH3 domain of H1 and the CH3 domain of H2 each contain one or more asymmetric amino acid substitutions that promote the formation of heterodimeric Fc domains rather than corresponding homodimeric Fc domains.

[0124] In various embodiments, the trivalent bispecific antibody constructs described herein may further comprise one or more linkers as described herein that covalently couple two or more domains of the construct to each other. In such embodiments, the Fab domain may be coupled to the first Fc polypeptide of the heterodimeric Fc domain via a first linker, the first scFv domain may be coupled to the second Fc polypeptide of the heterodimeric Fc domain via a second linker, and the second scFv domain may be coupled to the Fab domain via a third linker. The first linker coupling the Fab domain to the Fc polypeptide may be referred to herein as a linker Fab-Fc (i.e., represents the linker that couples the Fab domain to the Fc polypeptide). The second linker that couples the scFv domain to the Fc polypeptide may be referred to herein as the linker scFv-Fc Similarly, the third linker coupling the scFv domain to the Fab domain may be referred to herein as the linker scFv-Fab .

[0125] Thus, in some embodiments, the first polypeptide chain H1 of the antibody construct herein may comprise one or more peptide linkers. In such embodiments, (e.g., of the Fab domain) C H1The sequence can be connected via a first linker (e.g., Fab-Fc ) is coupled to the CH2 domain of the first Fc polypeptide, thereby coupling the Fab domain to the first Fc polypeptide. Fab-Fc It can be a peptide linker. In some embodiments, the linker Fab-Fc The linker may comprise or consist of an immunoglobulin hinge region, such as an IgG or IgG1 hinge region. Fab-Fc The invention also provides a method for preparing an scFv2 comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99% or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 107. In addition, in various embodiments, the second scFv domain (e.g., scFv2) can be formed by the V H or V L Sequence via adapter scFv-Fab V H Sequence coupling. Linkers scFv-Fab It can be a peptide linker. In various embodiments, the linker scFv-Fab Contains the amino acid sequence (Gly4Ser) n (or single letter amino acid code: (G4S) n , SEQ ID NO: 132), wherein n is an integer from 1 to 5, or consists of the amino acid sequence. In various embodiments, the linker scFv-Fab In various embodiments, the first scFv domain and / or the second scFv domain (e.g., scFv1 or scFv2) itself may comprise V L Domain and V H The fourth linker of the domain coupling. scFv Can be called a joint scFv . Connector scFv It can be a peptide linker. In some embodiments, the linker scFv May contain amino acid sequence (Gly4Ser) n (SEQ ID NO: 132), wherein n is an integer from 1 to 5, or consists of said amino acid sequence. In various embodiments, the linker scFv It comprises the amino acid sequence (Gly4Ser)4 (SEQ ID NO: 102) or consists of the amino acid sequence.

[0126] In some embodiments, the H1 polypeptide chain of the trivalent bispecific antibody construct herein may comprise a domain structure (from N-terminus to C-terminus): (V L -ConnectorsscFv -V H ) scFv2 -Connectors scFv-Fab -V H -C H1 -Connectors Fab-Fc -CH2-CH3 or consists of the domain structure, wherein "scFv2" represents the second scFv domain capable of binding to MSLN.

[0127] 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 may be linked to the first scFv2 domain by its V H or V L The domain is connected via a fifth linker (which may also be referred to as a linker scFv-Fc ) is coupled to the CH2 domain of the second Fc polypeptide, thereby coupling the first scFv domain (scFv1) to the second Fc polypeptide. scFv-Fc It can be a peptide linker. In some embodiments, the linker scFv-Fc The linker may comprise or consist of an immunoglobulin hinge region, such as an IgG or IgG1 hinge region. scFv-Fc comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99% or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 111 or SEQ ID NO: 112. In addition, the first scFv domain (scFv1) may comprise V L Domain and V H Domain-coupled linkers scFv1 (Where V H and V L The N-terminal to C-terminal sequence may be different). scFv1 It can be a peptide linker, and it can be combined with the linker present in scFv2 scFv (which may be referred to as a joint scFv2 ) are identical or different in sequence. In some embodiments, the linker scFv1 May contain amino acid sequence (Gly4Ser) n (SEQ ID NO: 132), wherein n is an integer from 1 to 5, or consists of said amino acid sequence. In various embodiments, the linker scFv1 comprising or consisting of the amino acid sequence (Gly4Ser)4 (SEQ ID NO: 102), in which case the first scFv domain and the second scFv domain comprise the same linker scFv .

[0128] In some embodiments, the H2 polypeptide chain of the trivalent bispecific antibody construct herein may comprise a domain structure (from N-terminus to C-terminus): (V L -Connectors scFv -V H ) scFv1 -Connectors scFv-Fc -CH2-CH3 or consists of the domain structure, wherein "scFv1" represents the first scFv domain capable of binding to MSLN.

[0129] In various embodiments, the trivalent bispecific antibody constructs of the present disclosure may comprise or consist of three polypeptide chains, which may be referred to as H1, H2, and L1, wherein (i) H1 comprises a domain structure (from N-terminus to C-terminus): (V L -Connectors scFv -V H ) scFv2 -Connectors scFv-Fab -V H -C H1 -Connectors Fab-Fc -CH2-CH3 or consists of the said domain structure, (ii) H2 comprises the domain structure (from N-terminus to C-terminus): (V L -Connectors scFv -V H ) scFv1 -Connectors scFv-Fc -CH2-CH3 or consists of the domain structure, and (iii) L1 comprises the domain structure (from N-terminus to C-terminus): V L -C L or or consisting of said domain structure, wherein V H -C H1 and L1's V L -C L A Fab domain capable of binding to CD3 is formed, scFv1 and scFv2 are each capable of binding to MSLN, and the first Fc polypeptide and the second Fc polypeptide form a heterodimeric Fc domain, wherein the CH3 domain of H1 and the CH3 domain of H2 each comprise one or more asymmetric amino acid substitutions that promote heterodimeric Fc domain formation and H1-H2 pairing rather than corresponding homodimeric Fc domains (e.g., resulting in H1-H1 or H2-H2 formation).

[0130] In some embodiments, the present disclosure relates to a trivalent bispecific antibody construct, which may comprise: (i) a Fab domain capable of binding to CD3; (ii) a first scFv domain and a second scFv domain, wherein both the first scFv domain and the second scFv domain are capable of binding to MSLN, and wherein at least one of the first scFv domain and the second scFv domain comprises a V L A 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 shown in SEQ ID NO: 101, and V H A 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 shown in SEQ ID NO: 103; and (iii) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, and wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain. In some embodiments, both the first scFv domain and the second scFv domain each comprise a V L A 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 shown in SEQ ID NO: 101; and V H A 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 shown in SEQ ID NO: 103. In such embodiments, the V of the first scFv domain and the second scFv domain (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 shown in SEQ ID NOs: 101 and 103, respectively. In some embodiments, the V of the first scFv domain and the second scFv domain (scFv1 and scFv2) are 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 shown in SEQ ID NOs: 101 and 103, respectively. In some embodiments, the V of the first scFv domain and the second scFv domain (scFv1 and scFv2) areL 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 shown in SEQ ID NOs: 101 and 103, respectively. In some embodiments, the V of the first scFv domain and the second scFv domain (scFv1 and scFv2) are 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 shown in SEQ ID NOs: 101 and 103, respectively. In other embodiments, the V of the first scFv domain and the second scFv domain (scFv1 and scFv2) are L Domain and V H The domains may comprise or consist of the amino acid sequences shown in SEQ ID NOs: 101 and 103, respectively.

[0131] In various embodiments herein, the trivalent bispecific antibody construct may comprise: (i) a Fab domain capable of binding to CD3, wherein the Fab domain comprises V L A 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 shown in SEQ ID NO: 105, and V H A 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 shown in SEQ ID NO: 115; (ii) a first scFv domain and a second scFv domain, wherein both the first scFv domain and the second scFv domain 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 coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain. In some embodiments, the V of the Fab domain L Domain and V H The Fab domains comprise or consist of an amino acid sequence having at least about 90% sequence identity to the sequences shown in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V HThe Fab domains comprise or consist of an amino acid sequence having at least about 95% sequence identity to the sequences shown in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V H The Fab domains comprise or consist of an amino acid sequence having at least about 97% sequence identity to the sequences shown in SEQ ID NOs: 105 and 115, respectively. In some embodiments, the V of the Fab domain L Domain and V H The Fab domains comprise or consist of amino acid sequences having at least about 99% sequence identity to the sequences shown in SEQ ID NOs: 105 and 115, respectively. In other embodiments, the V L Domain and V H The domains comprise or consist of the amino acid sequences shown in SEQ ID NOs: 105 and 115, respectively.

[0132] In various embodiments described herein, the trivalent bispecific antibody construct can comprise three sets of polypeptide chains, e.g., H1, H2, and L1, having or consisting of the amino acid sequences shown 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., those sequences having at least about 80% sequence identity).

[0133] In embodiments where 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 shown 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 shown 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 shown in SEQ ID NO: 114.

[0134] 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 shown in SEQ ID NO: 100, H2 comprises or consists of the amino acid sequence shown in SEQ ID NO: 110, and L1 comprises or consists of the amino acid sequence shown 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 shown 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 shown 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 shown 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 shown in SEQ ID NO: 100. In some embodiments, H1 comprises or consists of the amino acid sequence shown in SEQ ID NO:100.

[0135] In some embodiments, H1 comprises or consists of the amino acid sequence shown 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 shown in SEQ ID NO: 110, and L1 comprises or consists of the amino acid sequence shown 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 shown 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 shown 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 shown 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 shown in SEQ ID NO: 110. In some embodiments, H2 comprises or consists of the amino acid sequence shown in SEQ ID NO:110.

[0136] In some embodiments, H1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 100, H2 comprises or consists of the amino acid sequence shown 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 with the amino acid sequence shown 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 with the sequence shown in SEQ ID NO: 114. In other embodiments, L1 comprises or consists of an amino acid sequence having at least about 95% sequence identity with the sequence shown in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of an amino acid sequence having at least about 97% sequence identity with the sequence shown in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of an amino acid sequence having at least about 99% sequence identity with the sequence shown in SEQ ID NO: 114. In some embodiments, L1 comprises or consists of the amino acid sequence shown in SEQ ID NO:114.

[0137] Thus, in some embodiments, the present disclosure describes a trivalent bispecific antibody construct consisting of three polypeptide chains H1, H2, and L1, wherein (i) H1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 100, (ii) H2 comprises or consists of the amino acid sequence shown in SEQ ID NO: 110, and (iii) L1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 114. In some embodiments, this trivalent bispecific antibody construct can be variant v32523.

[0138] In other embodiments, the present disclosure describes a trivalent bispecific antibody construct consisting of three polypeptide chains H1, H2 and L1, wherein (i) H1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 117, (ii) H2 comprises or consists of the amino acid sequence shown in SEQ ID NO: 119, and (iii) L1 comprises or consists of the amino acid sequence shown in SEQ ID NO: 114. In some embodiments, this trivalent bispecific antibody construct can be variant v21812.

[0139] In some embodiments, one or more heavy chains in the antibody constructs herein may further comprise a C-terminal lysine residue. Exemplary heavy chain sequences containing such additional C-terminal lysine residues are shown in SEQ ID NOs: 171 and 172, and refer to the first heavy chain and the second heavy chain of construct v32523, respectively, wherein the two heavy chains further contain a C-terminal lysine. As further described herein, such C-terminal lysine residues can be cut (e.g., enzymatically cut) after expression of the antibody construct heavy chain.

[0140] The antibody constructs disclosed herein may be able to simultaneously bind CD3 on immune cells (e.g., T cells) and MSLN on tumor cells. This simultaneous binding can direct the anti-tumor activity of T cells to tumor cells and / or tumor microenvironment. The simultaneous binding of CD3 and MSLN, combined with the rationally selected antigen affinity of the constructs described herein, can provide TAA (e.g., MSLN)-dependent anti-tumor activity, that is, the simultaneous binding of CD3 and MSLN can trigger tumor killing in a local area (i.e., at the tumor site and / or in the tumor environment). Therefore, after administration to a subject in need, and when compared to existing T cell engager constructs (e.g., the benchmark constructs v29191 (Roche) and v31805 (Harpoon) described herein), the T cell engaging antibody constructs disclosed herein can be used to treat MSLN-positive tumors while causing significantly less off-target effects.

[0141] As further described herein, an 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 Fc polypeptide and the second Fc polypeptide may comprise one or more amino acid modifications in the first CH3 domain or the second CH3 domain, respectively, wherein such one or more amino acid modifications may promote the formation of a heterodimeric Fc domain comprising the first Fc polypeptide and the second Fc polypeptide, rather than 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 Fc polypeptide and the second Fc polypeptide each comprise at least one, two, three, four or five amino acid modifications, wherein such amino acid modifications are relative to the corresponding wild-type Fc polypeptide.

[0142] In various embodiments, the antibody constructs of the present disclosure may comprise 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. As further described herein, such CH2 domain modifications may alter the interaction of the antibody construct with an Fc receptor. In various embodiments, one or more amino acid modifications to one or both of the CH2 domains may reduce or "knock out" the endogenous Fc receptor function of the Fc domain. Thus, in various embodiments, one or both of the CH2 domains of the antibody constructs comprise one or more amino acid substitutions that reduce or weaken the interaction of the Fc domain with an Fc receptor, such as an Fc-γ receptor (FcγR).

[0143] In some embodiments, described herein is an antibody construct comprising (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN); and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: the Fab domain is coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0144] In some embodiments, at least one of the first scFv domain and the second scFv domain or both scFv domains comprise (i) V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and (ii) V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0145] In certain embodiments, described herein is an antibody construct comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN), and wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V LA domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: the Fab domain is coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0146] In a still further 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 shown in SEQ ID NO: 100; a second heavy chain (H2) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 110; and a light chain (L1) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 114.

[0147] B. Domain structures of antibody constructs

[0148] 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 (e.g., two or more) antibody domains. Such multiple antibody domains may comprise (i) one or more Fc domains, wherein the Fc domain may comprise a first Fc polypeptide and a second Fc polypeptide; (ii) one or more Fab domains, wherein the Fab domain may comprise a heavy chain portion, which comprises a heavy chain variable domain (V H ) and the heavy chain constant domain (C H1 ), and a light chain, which comprises a light chain variable domain (V L ) and the light chain constant domain (C L ); and (iii) one or more scFv domains, wherein the scFv domain may comprise a heavy chain variable domain (V H ) and the light chain variable domain (V L ). The various domains that an antibody construct may comprise are further described herein.

[0149] The immunoglobulin (Ig) structural unit is usually composed of two pairs of polypeptide chains, each pair having one "light" chain (about 25 kilodaltons (kD)) and one "heavy" chain (about 50-70 kD). Light chains can be classified as κ or λ. 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, and several of these classes 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 muon (μ), respectively.

[0150] In various embodiments, antibody constructs described herein are based on IgG class immunoglobulins, such as IgG1, IgG2, IgG3 or IgG4 immunoglobulins. In some embodiments, antibody constructs described herein are based on IgG1, IgG2 or IgG4 immunoglobulins. In certain embodiments, antibody constructs described herein are based on IgG1 immunoglobulins. In the context of the present disclosure, when antibody constructs are based on specific immunoglobulin isotypes, it refers to all or only part of the constant region (i.e., Fc domain) comprising a specific immunoglobulin isotype. It should be understood that, according to certain embodiments of the present disclosure, antibody constructs can also include a hybrid of isotypes and / or subclasses.

[0151] In general, the N-terminal domain of each polypeptide chain in an antibody generally defines a variable region (e.g., V L ... H or V L The term "variable light chain (V L )" and "variable heavy chain (V H )" refers to these domains in the light chain and heavy chain, respectively. As described herein, in various embodiments, an antibody construct may comprise two or more variable domain sequences coupled to each other in tandem and in a single polypeptide chain format, for example, as disclosed for a construct in which a scFv domain is coupled to a Fab domain in a single chain configuration. Such a construct may comprise a domain structure from N-terminus to C-terminus of: [(V H -V L ) / (V L -V H )] scFv -V H -C H1 (wherein “ / ” is equivalent to “or” in this specification).

[0152] Thus, in some embodiments, the antibody constructs derived from immunoglobulin molecules of the present disclosure may comprise different Ig domains within their heavy and light chains. The heavy chain domain may include an Fc domain (or Fc region) (e.g., comprising a CH2 domain and a CH3 domain), a hinge domain (or hinge region), and a variable heavy chain domain (V H ) and the constant heavy chain domain (C H1 ) of a heavy chain Fab domain, and the light chain domain may include a variable light chain domain (V L ) and the light chain constant domain (C L In some embodiments, and according to certain nomenclature, an "Fc domain" may include the CH2 and CH3 domains and a hinge domain (or hinge region).

[0153] Each V in the antibody constructs herein H and V L There are three loops in the domain that are hypervariable in sequence and form the antigen binding site. Each of these loops is called 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 to refer to the portion of the variable region that forms the antigen binding site. H In addition to CDR1 in a CDR, CDRs generally contain amino acid residues that form hypervariable loops. H and V L The domains consist of relatively invariant stretches called framework regions (FRs), which are between about 15 and 30 amino acids in length, separated by shorter CDRs, each of which is usually between about 5 and 15 amino acids in length, but can occasionally be longer or shorter. H and V L The three CDRs and four FRs of the domain are arranged from N-terminus to C-terminus as follows: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

[0154] Several different definitions and numbering conventions for CDR regions in 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), and IMGT, AbM and Contact definitions. These different definitions include superpositions or subsets of amino acid residues when compared to each other. For example, the CDR definitions according to Kabat, Chothia, IMGT, AbM and Contact are provided in Table 1 below.

[0155] Therefore, as will be apparent to one skilled in the art, the exact numbering and placement of the CDRs may differ based on the numbering system employed. However, it should be understood that reference is made herein to the heavy chain variable domain (V H ) includes disclosure of the associated (intrinsic) heavy chain CDRs (HCDRs) as defined by any known numbering system. Similarly, references herein to variable light domains (V L ) includes disclosure of the relevant (intrinsic) heavy chain CDRs (HCDRs) as defined by any known numbering system. Those skilled in the art will recognize that a limited number of amino acid substitutions can be introduced into the CDR sequences or V sequences of known antibodies. H Sequence or V L The antibody may be modified to modify the sequence without losing its ability to bind to its target. Candidate amino acid substitutions may be identified by computer modeling or by techniques such as alanine scanning, and the binding activity of the resulting variants (e.g., expressed as binding affinity, such as measured by EC 50 Values ​​are given) are tested by standard techniques. As an example, in certain embodiments, the MSLN binding domain of the antibody construct described herein comprises a CDR set (i.e., heavy chain CDR1, CDR2 and CDR3 and light chain CDR1, CDR2 and CDR3) having 90% or more, 95% or more, 98% or more, 99% or more, or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 120-125, respectively, wherein the binding domain retains or substantially retains the ability to bind to MSLN. In this context, the term "substantially" means that the change in binding affinity is less than about 5%, less than about 4%, less than about 3%, less than about 2%, or less than about 1% when compared to the unmodified binding domain.

[0156] Table 1: Common CDR definitions 1

[0157]

[0158] 1 For all definitions, either the Kabat or Chothia numbering systems may be used for HCDR2, HCDR3 and light chain CDRs, with the exception of Contact where Chothia numbering is used.

[0159] 2 Kabat numbering is used. The positions of the ends of the Kabat numbering scheme that distinguish Chothia and IMGT CDR-H1 loops vary depending on the length of the loop, as Kabat made insertions outside those CDR definitions at positions 35A and 35B. IMGT and Chothia CDR-H1 loops can be clearly defined using Chothia numbering. CDR-H1 definitions using Chothia numbering are: Kabat H31-H35, Chothia H26-H32, AbM H26-H35, IMGT H26-H33, Contact H30-H35.

[0160] In some embodiments, antibody constructs described herein include at least one immunoglobulin domain from mammalian immunoglobulin such as bovine immunoglobulin, human immunoglobulin, camel immunoglobulin, rat immunoglobulin or mouse immunoglobulin. In some embodiments, antibody constructs herein can be chimeric constructs comprising two or more immunoglobulin domains, wherein at least one domain is from a first mammalian immunoglobulin, such as human immunoglobulin, and at least the second domain is from a second mammalian immunoglobulin, such as mouse or rat immunoglobulin. In other embodiments, antibody constructs can be derived from immunoglobulins from different species, such as antibody constructs can be chimeric or humanized."Chimeric antibody constructs" refers to antibodies that generally include at least one variable domain from a rodent antibody (usually a mouse antibody) and at least one constant domain from a human antibody. "Humanized antibody constructs" are a class of chimeric antibodies containing the minimum sequence derived from a non-human antibody. In some embodiments, antibody constructs herein can include at least one immunoglobulin constant domain from a human immunoglobulin. In some embodiments, all domains of the antibody constructs described herein may be derived from (or derived from) human immunoglobulins.

[0161] In some embodiments, and as further described elsewhere herein, one or more domains of the antibody construct may be modified (e.g., to the amino acid sequence) to further improve the properties and performance (e.g., antigen affinity, stability, pharmacokinetics, etc.) of the antibody construct. For example, the framework region (FR) residues of a human immunoglobulin may be substituted with corresponding non-human residues, or the humanized antibody may contain residues that are not found in either the recipient antibody or the donor antibody. In general, the variable domains or humanized antibody domains in a humanized antibody contain all or nearly all of the hypervariable regions from a non-human immunoglobulin and all or nearly all of the FRs from a human immunoglobulin sequence. As further described herein, modifications in the Fc domain may achieve preferential pairing of Fc polypeptides to form heterodimeric Fc domains rather than homodimeric Fc domains.

[0162] In some embodiments, the disclosure relates to antibody constructs that can have different valences, for example, can be divalent, trivalent, tetravalent or have higher titer.Therefore, in various embodiments, the antibody construct herein comprises two or more, or three or more antigen binding domains, i.e., respectively at least divalent or at least trivalent.The antibody construct herein can also include different valences specifically bound to certain antigens.As an example, the bivalent bispecific antibody construct herein can include two binding domains, for example, a first binding domain (e.g., Fab domain) and a second binding domain (e.g., scFv domain).Such two binding domains can each have unique binding specificity to antigen, for example, a domain is for the first antigen (e.g., CD3), and the second domain is for the second antigen (e.g., MSLN).In various embodiments of the disclosure, the antibody construct can have bispecificity and be multivalent, so this antibody construct can include two or more binding domains, such as three binding domains.Each of the two or more binding domains can have unique binding specificity to antigen (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 construct herein can be trivalent and bispecific, i.e., the antibody comprises three binding domains, one of which is specific for or can bind to a first antigen (e.g., CD3), and the other two binding domains are specific for or can bind to a second antigen (e.g., MSLN).

[0163] In the embodiment where the antibody construct herein comprises two binding domains that bind to the same target molecule (e.g., MSLN), the two binding domains can bind to the same epitope on the target molecule, or they can bind to different epitopes on the target molecule. In some embodiments, the antibody construct herein comprises two binding domains that bind to different epitopes on the target molecule (e.g., MSLN), which can be characterized as "biparatope" binding. However, in other embodiments, the antibody construct herein comprises two binding domains that bind to the same epitope of an antigen (e.g., MSLN). In such embodiments, the two binding domains can comprise heavy and light chain CDRs that have at least about 90%, 95%, 97%, 99% or 100% amino acid sequence identity. In further embodiments, the V of the binding domains (e.g., the first scFv domain and the second scFv domain) that bind to the same epitope on the two binding targets (e.g., MSLN) is a double-paratope binding domain. H and V L The domains may share at least about 90%, 95%, 97%, 99% or 100% amino acid sequence identity, and thus 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.

[0164] 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 as 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.

[0165] Typically, a "Fab domain" as used herein contains the constant domain of the light chain (C L ) and the first constant domain of the heavy chain (C H1 ) and variable domains V on the light and heavy chains, respectively, containing the CDRs as described herein L and V H. In some embodiments, the Fab domain can be a single-chain Fab. A single-chain Fab can be a Fab molecule, wherein the Fab light chain and the Fab heavy constant 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 the single-chain Fab molecule, however, other types are also covered herein. However, in various embodiments herein, the Fab domain of the antibody construct is formed by two separately expressed polypeptide chains, i.e., a light chain and a heavy chain (or a portion thereof). However, the heavy chain and light chain portions of the Fab domain can be interconnected by covalent bonds (such as disulfide bonds).

[0166] As used herein, a "scFv domain" generally comprises a heavy chain variable domain (V H ) and the light chain variable domain (V L ). The scFv may optionally be in V H and V L The domains contain peptide linkers, which can help scFv form a functional structure for antigen binding. Therefore, in various embodiments, the scFv domains herein may include a peptide linker via its C-terminus. scFv With V H The N-terminal end of the domain is coupled to the V L A domain, i.e., a scFv domain may have a domain structure: V L -Connectors scFv -V H , or alternatively, the scFv may comprise a scFv With V L The N-terminal V H , i.e., with a domain structure: V H -Connectors scFv -V L .

[0167] In some embodiments, the antibody constructs described herein may further include another domain or portion that may not be derived from an immunoglobulin molecule. Such non-Ig domains may be referred to as non-Ig portions. Such non-Ig portions may be detectable markers (e.g., radiolabeled or fluorescent markers), low molecular weight (<750Da) drug molecules, another peptide (e.g., signal peptide) or polypeptide molecules, or combinations thereof.

[0168] Fc domain

[0169] 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, for example, each comprising one or more asymmetric amino acid substitutions that can promote preferential pairing of the Fc polypeptides to form a heterodimeric Fc domain compared to the corresponding homodimeric Fc domain formation.

[0170] As used herein, the term "Fc domain" or "Fc region" includes native (or wild-type) sequence Fc domains and variant Fc domains comprising one or more amino acid modifications relative to the corresponding native or wild-type Fc domain. Unless otherwise indicated 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 Edition, Public Health Service, National Institutes of Health, Bethesda, MD (1991). An "Fc polypeptide" of a dimeric (e.g., heterodimeric) Fc domain refers to one of the two polypeptide chains (e.g., a first Fc polypeptide and a second Fc polypeptide) that form a dimeric (e.g., heterodimeric) Fc domain. In some embodiments, the Fc polypeptide may include a C-terminal constant region of an immunoglobulin heavy chain that is capable of stabilizing self-association. In various embodiments, and as further described herein, the Fc polypeptide (e.g., a first Fc polypeptide or a 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 construct described herein comprises a CH2 domain sequence and a CH3 domain sequence.

[0171] As disclosed herein, the antibody construct may include an Fc domain, wherein the Fc domain may be a heterodimeric Fc domain. Unless otherwise indicated, the Fc domain of the antibody construct, such as a heterodimeric Fc domain, includes a first Fc polypeptide and a second Fc polypeptide. Typically, each Fc polypeptide of the (e.g., heterodimeric) Fc domain may include a CH2 domain, a CH3 domain, or both a CH2 domain and a CH3 domain as described in various embodiments herein.

[0172] 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 a heterodimeric IgG Fc domain containing two different Fc polypeptides (e.g., a first Fc polypeptide and a second polypeptide), wherein the first Fc polypeptide and the second Fc polypeptide have different amino acid sequences, such as amino acid sequences with about 90%, 95%, 97% or 99% sequence identity when compared and aligned to each other, for example, as further described herein about heterodimeric Fc domains. In some embodiments, the difference in the amino acid sequence of the first Fc polypeptide and the second Fc polypeptide can be attributed to asymmetric amino acid substitutions, which can be introduced into each Fc polypeptide chain to promote the preferential pairing of heavy chains to form heterodimeric Fc domains compared to corresponding homodimeric Fc domains.

[0173] In various embodiments, the antibody construct herein comprises an Fc domain as a modified IgG Fc domain, and wherein the CH3 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to the corresponding wild-type CH3 domain. In some embodiments, the antibody construct herein comprises 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 corresponding wild-type CH2 domain. In some embodiments, the antibody construct comprises an Fc domain based on a modified IgG Fc domain, wherein both the CH3 domain and the CH2 domain of at least one Fc polypeptide comprise one or more amino acid modifications compared to the corresponding wild-type CH3 domain and CH2 domain. In various embodiments, two Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in its CH3 domain. In some embodiments, two Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in its CH2 domain. In other embodiments, two Fc polypeptides of a heterodimeric Fc domain may comprise one or more amino acid modifications in its CH2 domain and CH3 domain.

[0174] Modified Fc domain

[0175] In some embodiments, the present disclosure relates to an antibody construct that may include a heterodimeric immunoglobulin Fc domain, the heterodimeric immunoglobulin Fc domain comprising a modified heterodimeric CH3 domain, wherein the modified heterodimeric CH3 domain comprises one or more asymmetric amino acid modifications, i.e., one or both of the first Fc polypeptide and the second Fc polypeptide each comprise one or more amino acid modifications in their CH3 domain sequence compared to the corresponding wild-type sequence. As used herein, the term "asymmetric amino acid modification" generally refers to a modification in which the amino acid at a specific position on the first Fc polypeptide is different from the amino acid at the corresponding position on the second Fc polypeptide. These asymmetric amino acid modifications may include modification of only one of the two amino acids at the corresponding position on each Fc polypeptide, or they may include modification of two amino acids at the corresponding position on each Fc polypeptide in the first Fc polypeptide and the second Fc polypeptide. In various embodiments, an "asymmetric amino acid modification" is an asymmetric amino acid substitution.

[0176] In some embodiments, the antibody constructs herein comprise a heterodimeric Fc domain comprising a modified CH3 domain (i.e., a heterodimeric CH3 domain consisting of two CH3 domain sequences of a first Fc polypeptide and a second Fc polypeptide), wherein the modified CH3 domain comprises one or more asymmetric amino acid modifications that promote heterodimeric Fc domain formation (e.g., a first Fc polypeptide is paired with a second Fc polypeptide) rather than corresponding homodimeric Fc domain formation (e.g., a first Fc polypeptide is paired with another first Fc polypeptide). Amino acid modifications that can be made to the CH3 domain of the Fc domain to promote heterodimeric Fc domain formation are known in the art and include, for example, those described in International Publication No. WO 96 / 027011 ("knobs into holes"); Gunasekaran et al., 2010, J Biol Chem, 285, 19637-46 ("electrostatic manipulation"); 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 methods of combining positive and negative design strategies to generate stable asymmetric modified Fc regions, such as those described in International Publication Nos. WO 2012 / 058768 and WO 2013 / 063702.

[0177] 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 comprise one or more amino acid modifications as described in International Publication No. WO 2012 / 058768 or International Patent Publication No. WO 2013 / 063702.

[0178] In some embodiments, the antibody constructs herein comprise a heterodimeric human IgG1 Fc domain with a modified CH3 domain. Table 2 herein provides the amino acid sequence of a human IgG1 Fc domain sequence (e.g., a sequence from which the first Fc polypeptide and / or the second Fc polypeptide may be derived), which corresponds to a full-length human IgG1 heavy chain (e.g., comprising V H , C H1 The CH2 domain is generally defined as comprising amino acids 231-340 of a full-length human IgG1 heavy chain, and the CH3 domain is generally defined as comprising amino acids 341-447 of a full-length human IgG1 heavy chain.

[0179] As described herein, the antibody construct may comprise a heterodimeric Fc domain having a modified CH3 domain comprising one or more asymmetric amino acid modifications that promote heterodimeric Fc domain formation rather than homodimeric Fc domain formation, 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 includes 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 includes 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 include an amino acid modification T350V.

[0180] Table 2: Exemplary human IgG1 Fc domain sequences and variants thereof

[0181]

[0182] *"A" corresponds to one of the first and second Fc polypeptide chains, and "B" corresponds to the other of the first and second Fc polypeptide chains of an antibody construct described herein.

[0183] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain having a modified CH3 domain comprising one or more asymmetric amino acid modifications that promote heterodimeric Fc domain formation rather than homodimeric Fc domain formation, wherein the modified CH3 domain comprises a first Fc polypeptide comprising amino acid modifications F405A, F405I, F405M, F405S, F405T or F405V and amino acid modifications Y407I or Y407V (relative to SEQ ID NO: 1) and a second Fc polypeptide comprising amino acid modifications T366I, T366L or T366M and amino acid modifications T394W (relative to SEQ ID NO: 1). In some embodiments, the first Fc polypeptide of the modified CH3 domain further comprises amino acid modification L351Y. In some embodiments, the second Fc polypeptide of the modified CH3 domain further comprises 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 comprises amino acid modification T350V.

[0184] In certain embodiments, the antibody constructs herein comprise a heterodimeric Fc domain comprising a modified CH3 domain having a first Fc polypeptide comprising amino acid modifications at positions F405 and Y407 and optionally further comprising an amino acid modification at position L351 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, as described above, and the first Fc polypeptide further comprises an amino acid modification at one or both of positions S400 or Q347, and / or the second Fc polypeptide further comprises an amino acid modification at one or both of positions K360 or N390, wherein relative to SEQ ID NO: 1, 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.

[0185] In some embodiments, the antibody construct comprises a heterodimeric Fc domain comprising a modified CH3 domain comprising a modification of any one of variant 1, variant 2, variant 3, variant 4, or variant 5 as shown in Table 2.

[0186] In various embodiments, the antibody construct 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 in their CH3 domains an amino acid substitution according to variant #1 as shown in Table 2. In other embodiments, the antibody construct 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 in their CH3 domains an amino acid substitution according to variant #2 as shown in Table 2. In some embodiments, the antibody construct 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 in their CH3 domains an amino acid substitution according to variant #3 as shown in Table 2. In some embodiments, the antibody construct 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 in their CH3 domains an amino acid substitution according to variant #4 as shown in Table 2. In other embodiments, the antibody construct 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 in their CH3 domains an amino acid substitution according to variant #5 as shown in Table 2.

[0187] In certain embodiments, the CH3 domain of the first Fc polypeptide of the antibody construct herein has an amino acid sequence that is at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identical to the amino acid sequence shown 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 that is at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identical to the amino acid sequence shown 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 shown in SEQ ID NO: 109, and the second Fc polypeptide of the antibody construct herein has the amino acid sequence shown in SEQ ID NO: 113.

[0188] It should be noted that unless otherwise indicated, "a first Fc polypeptide" and "a second Fc polypeptide" describing two Fc polypeptides of an Fc domain are generally used interchangeably throughout this section and this disclosure. Such descriptions are generally used herein to distinguish between two Fc polypeptides of an Fc domain, and are not intended to limit a particular Fc polypeptide to a certain set of asymmetric mutations.

[0189] 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 of a corresponding first Fc polypeptide and a second Fc polypeptide). In some embodiments, the antibody construct comprises an Fc domain based on an IgG Fc domain with a modified CH2 domain, wherein the modification of the CH2 domain results in an altered (e.g., reduced or attenuated) binding to one or more Fc receptors (FcRs) (such as receptors of the FcγRI, FcγRII, and FcγRIII subclasses).

[0190] Several amino acid modifications of the CH2 domain of the first Fc polypeptide and / or the second Fc polypeptide of the Fc domain that selectively changes the affinity of this Fc domain to different Fcγ receptors are known in the art. Both amino acid modifications that cause increased binding and amino acid modifications that cause reduced binding can be used for certain indications. For example, increasing the binding affinity of Fc to FcγRIIIa (activating receptor) can lead to increased antibody-dependent cell-mediated cytotoxicity (ADCC), which in turn can lead to increased target cell lysis. In some cases, the reduction of binding to FcγRIIb (inhibitory receptor) may also be beneficial. In certain indications, the reduction or elimination of ADCC and complement-mediated cytotoxicity (CDC) may be required. In such embodiments, it may be useful to include amino acid modifications that can lead to increased binding to FcγRIIb or reduce or eliminate the amino acid modifications of the binding of the Fc region to all Fcγ receptors.

[0191] Non-limiting examples of amino acid modifications to the CH2 domain that alter the binding of Fcγ receptors to the Fc domain include, but are not limited to, the following: S298A / E333A / K334A and S298A / E333A / K334A / K326A (increases affinity for FcγRIIIa) (Lu et al., 2011, J Immunol Methods, 365(1-2):132-41); F243L / R292P / Y300L / V305I / P396L (increases affinity for FcγRIIIa) (Stavenhagen et al., 2007, Cancer Res, 67(18):8882-90); F243L / R292P / Y300L / L235V / P396L (increases affinity for FcγRIIIa) (Nordstrom et al., 2007, Cancer Res, 67(18):8882-90); JL et al., 2011, Breast Cancer Res, 13(6):R123); F243L (increased affinity for FcγRIIIa) (Stewart et al., 2011, Protein Eng Des Sel., 24(9):671-8); S298A / E333A / K334A (increased affinity for FcγRIIIa) (Shields et al., 2001, J Biol Chem, 276(9):6591-604); S239D / I332E / A330L and S239D / I332E (increased affinity for FcγRIIIa) (Lazar et al., 2006, Proc Natl Acad Sci USA, 103 (11): 4005-10) and S239D / S267E and S267E / L328F (increase affinity for FcγRIIb) (Chu et al., 2008, Mol Immunol, 45 (15): 3926-33). Additional modifications that affect the binding of the Fc domain to the Fcγ receptor are described in Therapeutic Antibody Engineering (Strohl & Strohl, Woodhead Publishing series in Biomedicine No 11, ISBN 1 907568 37 9, October 2012, page 283).

[0192] 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, wherein the modified CH2 domain comprises one or more amino acid modifications that result in reduced or eliminated binding of the Fc domain to one or more, or all, Fcγ receptors (i.e., a "knockout" or "KO" variant).

[0193] 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 reducing effector function by modification of glycosylation, use of IgG2 / IgG4 scaffolds, or introducing mutations in the hinge or CH2 domain of Fc (see also, U.S. Patent Publication No. 2011 / 0212087, International Publication No. WO 2006 / 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).

[0194] In some embodiments, the Fc domain of the antibody construct may comprise 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.

[0195] Table 3: Modifications that reduce Fcγ receptor or complement binding

[0196]

[0197]

[0198] Additional examples herein include Fc domains engineered to include amino acid modifications L235A / L236A / D265S, for example based on the sequence shown 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. WO 2014 / 190441.

[0199] In certain embodiments, the CH2 domain of the first Fc polypeptide and the second Fc polypeptide 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 shown in SEQ ID NO: 108. In some embodiments, the CH2 domain of the first Fc polypeptide and / or the second Fc polypeptide herein comprises or consists of the sequence shown in SEQ ID NO: 108.

[0200] In certain embodiments, the antibody constructs herein comprise heterodimeric Fc domains in which native glycosylation has been modified. As is known in the art, the glycosylation of Fc can be modified to increase or decrease effector function. For example, mutation of the conserved asparagine residue at position 297 to alanine, glutamine, lysine or histidine (i.e., N297A, Q, K or H) produces an aglycosylated Fc lacking all effector functions (Bolt et al., 1993, Eur. J. Immunol., 23: 403-411; Tao & Morrison, 1989, J. Immunol., 143: 2595-2601). In contrast, removal of fucose from heavy chain N297-linked oligosaccharides has been shown to enhance ADCC based on improved 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 a variant CHO cell line Lec13 (International Publication No. WO 03 / 035835) with reduced ability to attach fucose to N297-linked carbohydrates, or in other cells that produce non-fucosylated antibodies (see, 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. WO 2009 / 135181 describes the addition of a fucose analog to a culture medium during antibody production to inhibit the incorporation of fucose into carbohydrates on the antibody.

[0201] In certain embodiments, the trivalent, bispecific antibody constructs of the present disclosure can bind to a neonatal Fc receptor (eg, FcRn).

[0202] Connectors

[0203] As described herein, the antibody constructs of the present disclosure may include one or more joints. In various embodiments, such one or more joints are amino acid sequences comprising 1, 2, 3, 5, 10, 15, 20, 25, 30 or more consecutive amino acid residues or peptide (or peptide) joints consisting of said amino acid sequences. Such peptide joints can couple or connect two or more peptides or polypeptides to each other. Therefore, in various embodiments, the joints herein can be a first polypeptide sequence (e.g., a heavy chain constant domain (C H1 )) is coupled to a second polypeptide sequence (e.g., an Fc polypeptide). Thus, in some embodiments, a linker can be used to couple one domain of an antibody construct to another domain, e.g., via a linker Fab-Fc Conjugating the Fab domain to the Fc domain, e.g. via a linker scFv-Fab Coupling the scFv domain to the Fab domain, e.g. via a linker scFv V H Domain and V L Domain coupling, etc.

[0204] In the linker, for example, the heavy chain variable domain (V H ) and, for example, the light chain variable domain (V L ) coupling, the linker can have sufficient length to allow the two domains to elicit their biological functions. In addition to providing a spacing function, the linker herein (e.g., a peptide linker) can provide flexibility or rigidity that is suitable for the correct orientation of one or more domains of the antibody construct within the antibody construct itself and between the antibody construct and its biological target (e.g., T cells and / or cancer antigens).

[0205] In addition, the linkers (e.g., peptide linkers) herein can support (i) the expression of full-length fusion proteins, such as the full-length polypeptide chains H1, L1, H2, etc. of the antibody construct, and (ii) provide increased stability of the purified protein in vitro and in vivo, for example, after administration to a subject in need thereof (such as a human or a rodent). The one or more linkers used in the antibody constructs herein are generally non-immunogenic or poorly immunogenic in a mammalian subject to which the antibody construct can be administered. In certain embodiments, the one or more linkers used in the antibody constructs herein may comprise part or all of a human immunoglobulin hinge region, a stem region of a C-type lectin, a type II membrane protein family, or a combination thereof. In certain embodiments, the one or more linkers used in the antibody constructs herein may comprise part or all of a human immunoglobulin hinge region (such as an IgG or IgG1 hinge region), such as a linker Fab-Fc or connectorscFv-Fc .

[0206] In certain embodiments, each joint in the one or more joints used in the antibody construct herein can comprise or consist of an amino acid sequence having a length of 2 to about 50 amino acids. In some embodiments, each joint in the one or more joints used in the antibody construct herein can comprise or consist of an amino acid sequence having a length of about 3 to about 40 amino acids, about 10 to about 50 amino acids, about 2 to about 40 amino acids, about 5 to about 30 amino acids, about 5 to about 25 amino acids, about 4 to about 30 amino acids, about 10 to about 30 amino acids, or about 15 to about 25 amino acids. In some embodiments, one or more linkers of the 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.

[0207] In certain embodiments, the linker of the antibody constructs herein comprises the amino acid sequence (EAAAK) n , wherein n is an integer from 1 to 5 (SEQ ID NO: 133), or consists of the amino acid sequence. In some embodiments, the linker comprises the sequence EAAAK (SEQ ID NO: 134) or consists of the sequence. In some embodiments, the linker comprises the sequence EAAAKEAAAK (SEQ ID NO: 135) or consists of the sequence. In some embodiments, the linker comprises a polyproline linker, such as an amino acid sequence having PPP (SEQ ID NO: 136) or PPPP (SEQ ID NO: 137). In certain embodiments, the linker is a glycine (G)-proline (P) polypeptide linker, such as one or more comprising GPPPG (SEQ ID NO: 138), GGPPPGG (SEQ ID NO: 139), GPPPG (SEQ ID NO: 140) or GGPPPPGG (SEQ ID NO: 141) or consists of one or more of the sequences. In some embodiments, the linker herein is (Gly n Ser) m A linker, wherein n and m are independently integers from 1 to 5 (SEQ ID NO: 142). In certain embodiments, the linker comprises (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) or consisting of the amino acid sequence, 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 and comprises a sequence containing a glycine-serine linker, such as, but not limited to, (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 from 0 to 20, from 1 to 10, or from 1 to 5.

[0208] Thus, in some embodiments, the antibody constructs described herein comprise one or more linkers, such as 1, 2, 3, 4, or 5 linkers, comprising the amino acid sequence (Gly4Ser) n , wherein n is an integer from 1 to 5 (SEQ ID NO: 132), from 1 to 4, or from 1 to 3, or consists of the amino acid sequence. In such embodiments, n may be 1 (SEQ ID NO: 104). In other embodiments, n is 3 (SEQ ID NO: 118). In other embodiments, n is 4 (SEQ ID NO: 102).

[0209] In certain embodiments herein, one or more linkers included in the antibody construct can be an amino acid sequence obtained, derived or designed from an antibody hinge region sequence. In some embodiments, this linker can have at least one cysteine ​​that can participate in or form at least one disulfide bond under physiological conditions or other standard peptide conditions (e.g., peptide purification conditions, conditions for peptide storage). In certain embodiments, a linker corresponding to or similar to an immunoglobulin hinge peptide retains a cysteine ​​corresponding to a hinge cysteine ​​arranged toward the amino (or N) end of the hinge. In a further embodiment, the linker is derived from an IgG1 hinge and can be modified to remove any cysteine ​​residues, or the linker is an IgG1 hinge with a cysteine ​​or two cysteines corresponding to a hinge cysteine.

[0210] In certain embodiments, the linker of the antibody constructs described herein may comprise an "altered wild-type immunoglobulin hinge region" or an "altered immunoglobulin hinge region." Such an altered hinge region can refer to (a) a wild-type immunoglobulin hinge region having up to 30% amino acid changes (e.g., up to 25%, 20%, 15%, 10% or 5% amino acid substitutions, insertions and / or deletions), (b) a portion of a wild-type immunoglobulin hinge region having up to 30% amino acid changes (e.g., up to 25%, 20%, 15%, 10% or 5% amino acid substitutions and / or deletions) having a length of at least 10 amino acids (e.g., at least 12, 13, 14 or 15 amino acids), (c) a portion of a wild-type immunoglobulin hinge region comprising a core hinge region (the portion can be 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 amino acids in length), or (d) any combination of (a)-(c). In certain embodiments, one or more cysteine ​​residues in a wild-type immunoglobulin hinge region (such as an IgG1 hinge comprising an upper region and a core region) may be substituted with one or more other amino acid residues (e.g., one or more serine residues). The altered immunoglobulin hinge region may alternatively or additionally have a proline residue in a wild-type immunoglobulin hinge region (such as an IgG1 hinge comprising an upper region and a core region) replaced with another amino acid residue (e.g., a serine residue).

[0211] 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 shown 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 shown 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 shown in SEQ ID NO: 112.

[0212] In various embodiments, the antibody constructs of the present disclosure comprise a linker coupling the Fab domain to the first Fc polypeptide, the linker Fab-Fc , where the connector Fab-Fc comprising or consisting of the amino acid sequence shown in SEQ ID NO: 107, and another linker coupling the first scFv domain to the second Fc polypeptide, the linkerscFv-Fc , where the connector scFv-Fc The invention relates to a polypeptide comprising the amino acid sequence shown in SEQ ID NO: 111 or consisting of the amino acid sequence.

[0213] In other embodiments, the antibody constructs of the present disclosure comprise a linker coupling the Fab domain to the first Fc polypeptide, the linker Fab-Fc , where the connector Fab-Fc comprising or consisting of the amino acid sequence shown in SEQ ID NO: 107, and another linker coupling the first scFv domain to the second Fc polypeptide, the linker scFv-Fc , where the connector scFv-Fc It comprises the amino acid sequence shown in SEQ ID NO: 112 or consists of the amino acid sequence.

[0214] Binding domains targeting tumor-associated antigens

[0215] As described herein, the antibody constructs disclosed herein may include 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 have bispecificity, and include a first binding domain, a second binding domain, and a third binding domain, wherein one or more of such binding domains can bind to tumor-associated antigens (TAA). In various embodiments, two of such binding domains can bind to TAA. TAA can be any antigenic substance expressed on the surface of tumor cells.

[0216] Therefore, in embodiments where at least two binding domains of the antibody construct herein are capable of binding to TAA, the two binding domains may bind to different TAAs. In various other embodiments, the two anti-TAA binding domains may bind to the same TAA, i.e., the antibody construct is divalent for the TAA. As further described herein, in such embodiments, the two anti-TAA binding domains may be capable of binding to the same epitope of TAA (e.g., capable of divalently binding to two TAA molecules on the same epitope (e.g., two MSLN molecules)). In other embodiments, the two anti-TAA binding domains may be capable of binding to two different epitopes of the same TAA.

[0217] As described herein, in various embodiments, the two anti-TAA binding domains of the antibody construct may be capable of binding to MSLN. In various embodiments, the anti-MSLN binding domains of such constructs are scFv domains and are capable of binding to the same MSL epitope on the same MSLN protein or on two different MSLN proteins.

[0218] In some embodiments, the present disclosure relates to antibody constructs in which two anti-MSLN binding domains can have the same or similar (i.e., within about ±5%) binding affinity to MSLN, or the two domains can have different binding affinities to MSLN. Such different binding affinities can differ by at least about 2%, 5% or 10%, for example, as EC 50 Value measured.

[0219] Thus, in various embodiments, at least one anti-MSLN binding domain of the constructs herein may have a binding affinity for 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, the binding affinity of two anti-MSLN binding domains for MSLN may be 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 two binding domains for MSLN have an affinity for MSLN of about 3 nM. In some embodiments, at least one or both binding domains to 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 to MSLN have an affinity for MSLN of about 2 nM to about 4 nM.

[0220] As used throughout this disclosure, the term "binding affinity" for binding to a certain antigen (e.g., CD3, MSLN, etc.) in the context of an antibody construct or portion thereof (e.g., a binding domain) herein broadly corresponds to the dissociation constant (K) measured using one or more binding assays known in the art (such as SPR, FACS, etc.). D ) value. The term “K D " (given in units such as molar "M", mM, μM, nM, pM, etc.) refers to the dissociation equilibrium constant for a particular antibody construct-antigen interaction. K D There is an inverse relationship between K and binding affinity, so K D The smaller the value, the higher the affinity, i.e. the stronger it is. Therefore, the term "higher affinity" or "stronger affinity" relates to a higher ability to form an interaction, and thus K D In contrast, the term "lower affinity" or "weaker affinity" relates to a lower ability to form an interaction, so that K D Unless otherwise stated, the binding affinity of an antibody construct herein for its target (e.g., CD3, MSLN, etc.) is expressed as the EC binding affinity of that target. 50As used herein, the term "EC 50 ” refers to the half maximal effective concentration, which includes the concentration of the antibody construct that induces a response halfway between baseline and maximum after a specified exposure time. EC 50 The value essentially represents the concentration of antibody construct at which 50% of the maximal effect is observed. Therefore, the EC 50 Values ​​are generally equal to the concentration of an antibody construct of the disclosure that produces half-maximal binding to cells expressing CD3 or a tumor-associated antigen (such as MSLN) as determined by, for example, an SPR or FACS binding assay.

[0221] 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 comprising a V H Domain, V L domain and V L The C-terminus of the domain is H N-terminal coupling of the domain or V H The C-terminus of the domain is L 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 may each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125. In these embodiments, the two anti-MSLN scFv domains can have the same or similar binding affinity to MSLN.

[0222] In some embodiments, the CDR of the anti-MSLN paratope comprises one or more amino acid modifications in one or more CDR sequences shown 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.

[0223] In some embodiments, the two anti-MSLN binding domains are scFv domains (e.g., a first scFv domain and a second scFv domain) that comprise or consist of amino acid sequences that share about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity.

[0224] In such embodiments, the two anti-MSLN binding domains, i.e., the first scFv domain and the second scFv domain may comprise a V H A domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99% or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 103. In certain embodiments, the two anti-MSLN binding domains may comprise V H A domain comprising or consisting of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 103. In some embodiments, the two anti-MSLN binding domains may comprise V H A domain comprising or consisting of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 103. In some embodiments, the two anti-MSLN binding domains may comprise V H A domain comprising or consisting of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 103. In some embodiments, the two anti-MSLN binding domains may comprise V H A domain comprising or consisting of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 103. In some embodiments, the two anti-MSLN binding domains may comprise V H A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 103.

[0225] In a further embodiment, the two anti-MSLN binding domains, namely the first scFv domain and the second scFv domain may comprise V LA domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 102. In some embodiments, the two anti-MSLN binding domains may comprise V L A domain comprising or consisting of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 102. In some embodiments, the two anti-MSLN binding domains may comprise V L A domain comprising or consisting of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 102. In some embodiments, the two anti-MSLN binding domains may comprise V L A domain comprising or consisting of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 102. In some embodiments, the two anti-MSLN binding domains may comprise V L A domain comprising or consisting of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 102. In some embodiments, the two anti-MSLN binding domains may comprise V L A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 102.

[0226] Thus, in various embodiments, the antibody constructs of the present disclosure comprise a first scFv domain and a second scFv domain, wherein both the first scFv domain and the second scFv domain are capable of binding to the same epitope on MSLN, and wherein each of the first scFv domain and the second scFv domain comprises or consists of an amino acid sequence having about 80%, 90%, 95%, 97%, 99% or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 149. In some embodiments, each of the first scFv domain and the second scFv domain comprises or consists of an amino acid sequence having about 90%, 95%, 97%, 99% or 100% sequence identity to the amino acid sequence shown in SEQ ID NO: 149. In some embodiments, each of the first scFv domain and the second scFv domain comprises or consists of an amino acid sequence having about 90% sequence identity to the amino acid sequence shown 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 shown 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 shown 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 shown 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 shown in SEQ ID NO: 149.

[0227] In some embodiments, the first scFv domain and the second scFv domain of the antibody construct herein, both of which are capable of binding to MSLN, each consist of the amino acid sequence shown in SEQ ID NO:149.

[0228] Binding domains directed against antigens on cytotoxic effector cells

[0229] As further described herein, the antibody constructs of the present disclosure may include at least one binding domain that can bind to a molecule (e.g., a polypeptide) on the surface of a cytotoxic effector cell. Such cytotoxic effector cells may be immune cells. Immune cells may be T cells, macrophages, dendritic cells, neutrophils, B cells, or NK cells. In various embodiments, the antibody constructs of the present disclosure include at least one binding domain that can bind to an antigen on a T cell. In various embodiments, such an antibody is CD3.

[0230] In some embodiments, the disclosure relates to a trivalent bispecific antibody construct comprising a first binding domain capable of binding to CD3 and a second binding domain and a third binding domain each capable of binding to MSLN. The antibody constructs described herein may also be referred to as "T cell engagers," "TCEs," or "T cell engager molecules," referring to the ability of the construct to bind both CD3 on T cells and MSLN on tumor cells, thereby directing T cell-mediated cytotoxic activity to the tumor environment containing cells expressing MSLN.

[0231] In various embodiments, the affinity of the anti-CD3 binding domain of the antibody constructs herein for CD3 (measured as the EC binding domain of CD3) is 50 In various embodiments, the affinity of the anti-CD3 binding domains herein for CD3 (given as EC for binding to CD3) may be no more than about 1 nM, 5 nM, 10 nM, 20 nM, or no more than about 30 nM. 50 In some embodiments, the affinity of the anti-CD3 binding domain herein for CD3 (given as EC for binding to CD3) is about 20 nM to about 80 nM, about 30 nM to about 60 nM, or about 40 nM to about 50 nM. 50 In various embodiments, the antibody constructs herein comprise an anti-CD3 binding domain that binds to the EC of CD3. 50 The value is about 20 nM to about 40 nM, such as about 30 nM.

[0232] In various embodiments, the antibody constructs herein comprise a Fab domain capable of binding to CD3 on T cells, wherein such Fab domain binds to the EC of CD3. 50 The value is about 20nM to about 40nM, such as about 30nM, and includes V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131.

[0233] 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.

[0234] In certain embodiments, and as further described herein, the anti-CD3 Fab domain of the antibody construct may comprise a V H Domain and C H1 The heavy chain portion of the domain and the V L Domain and C L The light chain of the domain may consist of the heavy chain portion and the light chain.

[0235] The V domain of the anti-CD3 Fab domain of the antibody construct herein H The 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 shown in SEQ ID NO: 105. In some embodiments, the V H The Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 105. In some embodiments, the V H The Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 105. In some embodiments, the V HThe Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 105. In some embodiments, the V H The Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 105. In other embodiments, the V H The domain comprises or consists of the amino acid sequence shown in SEQ ID NO:105.

[0236] The V domain of the anti-CD3 Fab domain of the antibody construct herein L The Fab domain may be a portion of a light chain (e.g., L1) that pairs with the anti-CD3 domain of a 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 shown in SEQ ID NO: 115. In some embodiments, the V domain of such a Fab domain is L The Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 115. In some embodiments, the V L The Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 115. In some embodiments, the V L The Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 115. In some embodiments, the V L The Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 115. In various embodiments, the V L The domain comprises or consists of the amino acid sequence shown in SEQ ID NO:115.

[0237] The anti-CD3 Fab domain of the antibody construct herein is C H1 The 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 shown in SEQ ID NO: 106. In some embodiments, the C of such Fab domain isH1 The Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 106. In some embodiments, the C H1 The Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 106. In some embodiments, the C H1 The Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 106. In some embodiments, the C H1 The Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 106. In some embodiments, the C H1 The domain comprises or consists of the amino acid sequence shown in SEQ ID NO:106.

[0238] The anti-CD3 Fab domain of the antibody construct herein is C L The 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 shown in SEQ ID NO: 116. In some embodiments, the C of such Fab domain is L The Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence shown in SEQ ID NO: 116. In some embodiments, the C L The Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence shown in SEQ ID NO: 116. In some embodiments, the C L The Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence shown in SEQ ID NO: 116. In some embodiments, the C L The Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence shown in SEQ ID NO: 116. In other embodiments, the C L The domain comprises or consists of the amino acid sequence shown in SEQ ID NO:116.

[0239] Thus, in various embodiments, the antibody constructs herein comprise an anti-CD3 Fab domain comprising (i) V H A 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 shown in SEQ ID NO: 105; (ii) V L A 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 shown in SEQ ID NO: 115; (iii) C H1 A 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 shown in SEQ ID NO:106; and (iv) 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 shown in SEQ ID NO:116.

[0240] In certain embodiments, the antibody constructs herein comprise an anti-CD3 Fab domain comprising (i) V H A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 105; (ii) V L A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 115; (iii) C H1 A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 106; and (iv) a CL domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 116.

[0241] In various embodiments, and as further described herein, when comprising V H Domain and C H1 A portion of the heavy chain (e.g., H1) of the domain is associated with a portion of the heavy chain (e.g., H1) comprising the V LWhen the light chain (L1) of the heavy chain and the CL domain are paired, an anti-CD3 Fab domain of the antibody construct can be formed. In some embodiments, the heavy chain H1, a portion of which can form the anti-CD3 Fab domain, has the amino acid sequence shown in SEQ ID NO: 100, and the corresponding light chain L1 that pairs with the anti-CD3 portion of the heavy chain to form the anti-CD3 Fab domain has the amino acid sequence shown in SEQ ID NO: 114.

[0242] C. Characteristics of Antibody Constructs

[0243] The trivalent, bispecific antibody constructs of the present disclosure may have 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 cancer treatment, and with respect to anti-tumor activity and the occurrence of side effects when the constructs are administered to a subject in need thereof.

[0244] As described herein, the antibody constructs of the present disclosure may be able to bind to CD3 on cytotoxic effector cells (e.g., T cells) and MSLN located on tumor cells. In some embodiments, this binding occurs simultaneously, for example, the antibody construct may bind to both CD3 on cytotoxic effector cells and MSLN on tumor cells simultaneously. In various embodiments, the cytotoxic effector cells are immune cells, such as T cells, and the tumor cells are solid tumor cells expressing MSLN.

[0245] In various embodiments, the antibody constructs described herein have been engineered and designed to have a certain format (e.g., 2 (scFv) + 1 (Fab) to MSLN and CD3, respectively) and geometry that allows affinity-driven MSLN engagement via two anti-MSLN scFv domains, each located on a different arm of the antibody construct. In addition, extensive evaluation studies have shown that anti-MSLN paratopes with moderate single-digit nanomolar affinity for MSLN (e.g., about 2nM to about 4nM) in combination with antibody formats surprisingly provide potent T cell-mediated anti-tumor activity in the presence of tumor cells with moderate to high MSLN expression, whereas lower anti-tumor activity was observed in the context of tumor cells with low MSLN expression. This superior TAA-dependent anti-tumor activity appears to be unique to the constructs described herein and was not observed in tested benchmark constructs with different formats and / or geometries. It was also found that combining this MSLN binding property with a monoclonal antibody CD3 binding domain having a moderate to low affinity for CD3 in the double-digit nanomolar range (e.g., about 20nM to about 40nM) showed the most favorable balance between T cell-mediated anti-tumor activity and selectivity for tumor cells with moderate or high MSLN expression rather than tissues with low MSLN expression. Such properties can provide enhanced cytotoxic effects triggered by immune cells at the target site, for example, where tumor cells expressing TAAs (e.g., MSLN) are present, and when compared to other organs or tissues where cells with high MSLN expression are absent or significantly less numerous. Therefore, the trivalent bispecific antibody constructs described herein comprising one anti-CD3 Fab binding domain and two anti-MSLN scFv binding domains in a configuration in which both anti-MSLN scFv domains are located in the N-terminal region of the antibody construct (e.g., at the N-termini of the two heavy chains) can provide excellent in vivo anti-tumor activity and tolerability when compared to existing benchmark molecules that can also target both CD3 and MSLN but have different molecular formats and / or geometric structures and different antigen affinities.

[0246] Therefore, the antibody constructs of the present disclosure can be considered as trivalent bispecific T cell engaging molecules that can form a TCR-independent artificial immune synapse between an immune cell (e.g., a T cell) and a tumor cell when binding to CD3 on an immune cell and MSLN on a tumor cell. This can result in the T cell being activated and producing a cytotoxic effect on the tumor cell, which can result in tumor cell lysis, as further described herein.

[0247] In some embodiments, the trivalent bispecific antibody constructs of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can be in a cell population comprising immune cells expressing CD3 and tumor cells expressing MSLN and when present at a concentration of about 10 pM, be able to achieve one or more of the following: A) induction of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL TNFα; B) induction of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL IFNγ; and / or C) induction of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL IL-2. In some embodiments, the antibody construct comprises two or more of properties A, B, and C. In other embodiments, the antibody construct comprises all properties A, B, and C. In some of these embodiments, the immune cells may comprise T cells and the tumor cells may express at least about 500,000 MSLN / cell.

[0248] In various embodiments, as measured by surface plasmon resonance (SPR), the 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. In some embodiments, the 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 about 30 nM to about 60 nM.

[0249] Thus, in some embodiments, the trivalent bispecific antibody construct of the present disclosure comprises an anti-CD3 Fab domain and two anti-MSLN scFv domains and has a domain structure scFv2-Fab-CH21-CH31 in the first heavy chain and scFv1-CH22-CH32 in the second heavy chain, and comprises (i) an anti-MSLN paratope having HCDRs and LCDRs as shown in SEQ ID NOs: 120-125, respectively, and an anti-CD3 paratope having HCDRs and LCDRs as shown in SEQ ID NOs: 126-131, respectively, and the trivalent bispecific antibody construct may have a binding affinity for MSLN of about 0.7 nM, 0.8 nM, 0.9 nM, or about 1 nM as measured by SPR, and a binding affinity for CD3 of up to about 30 nM, 40 nM, 50 nM, or 60 nM. The relative orientation of the binding domains of such an antibody construct may provide for binding to CD3. + T cells and MSLN +Specific affinity-based engagement of tumor cells, which can allow MSLN-dependent T cell activation, for example, as measured 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, and without being bound by any theory, when compared with other previously reported constructs, such as corresponding Fabs containing two anti-MSLN Fabs (rather than the two anti-MSLN scFv domains used in the constructs described herein) and one anti-CD3 Fab domain, 3 When comparing constructs, two anti-MSLN scFv and one anti-CD3 Fab in the configuration described herein (and as e.g. Figure 1A ) can provide a more effective and stronger MSLN-dependent anti-tumor response. The present disclosure (including the experimental data provided herein) demonstrates that the format and geometry of the construct combined with its antigen affinity can have a significant and surprising impact on its in vivo properties, for example, by head-to-head comparison of trivalent bispecific constructs, one of which comprises one anti-CD3 Fab domain and two anti-MSLN scFv domains (e.g., v21812, v32523, etc.), and the other comprises one anti-CD3 Fab domain and two anti-MSLN Fab domains (e.g., v29191, v21791), and in some embodiments, even constructs with different geometries but the same anti-CD3 / MSLN paratope sequences are compared to show differences driven entirely by format.

[0250] Thus, in some embodiments, the trivalent, bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains may be capable of inducing T cells to produce proinflammatory cytokines in an MSLN-dependent manner, wherein MSLN-dependent activation of cytotoxic effector cells is determined by measuring at least about 20-fold, 50-fold, 100-fold, or at least about 1000-fold reduction in cytokine production between a first cell population comprising a first tumor cell and an immune cell and a second cell population comprising a second tumor cell and an immune cell, wherein the MSLN expression of the first tumor cell is at least about 3, 4, or 5-fold higher than the MSLN expression of the second tumor cell.

[0251] In a further embodiment, the trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can be capable of inducing at least 5-fold, 10-fold, 15-fold or 20-fold higher T cell-mediated cytotoxicity against tumor cells in in vitro studies when compared to a similar antibody construct comprising two anti-MSLN Fab domains instead of the first anti-MSLN scFv domain and the second anti-MSLN scFv domain of the construct disclosed herein under the same experimental conditions.

[0252] The antibody constructs described herein are directed against MSLN + The cytotoxic activity of cell lines can be expressed as half maximal inhibitory concentration (IC 50 As an example, this IC 50 The value can be used to activate MSLN + A measure of how much antibody construct is needed to reduce cell number by 50%. Determine the IC of an antibody construct 50 Several methods are known in the art and described herein, such as using cell viability assays and measuring the concentration at which viable tumor cell counts are reduced by 50%. 50 This suggests that more antibody constructs are needed to reduce viable tumor cell counts by, for example, 50%, and therefore the ability of the antibody constructs to induce antitumor cytotoxic activity is relatively low. 50 It was shown that less antibody construct was needed to reduce the viable tumor cell count by, for example, 50%, and therefore the ability of the antibody construct to induce anti-tumor cytotoxic activity was relatively high.

[0253] In some embodiments, the IC of T cell-induced tumor cell killing of the trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains is 50 It can be 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. In certain embodiments, the IC of T cell-induced tumor cell killing of the trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains is 50 It can be about 1 pM to about 0.01 pM or about 0.5 pM to about 0.01 pM. In some embodiments, the tumor cells are MSLN with an average number equal to or greater than about 100,000 MSLN / cell. + Tumor cells, such as H292 or OVCAR3 cells, are as described herein.

[0254] In some embodiments, a trivalent, bispecific antibody construct comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains of the present disclosure may be able to (i) inhibit tumor growth (e.g., tumor volume increase ≤5%) and / or (ii) reduce tumor volume by at least about 5%, 10%, 15%, 20%, 25%, 30%, 40%, or at least about 50% in a subject when measured over a period of at least about 20 days after administration of the antibody construct to a subject at a dose of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg, wherein the tumor comprises cells expressing at least about 100,000 MSLN / cell. In various embodiments, the reduction in tumor volume achieved by such a trivalent, bispecific antibody construct is significantly greater than that achieved by a bispecific antibody construct comprising a Fab domain. 3 The constructs of the ANTI-MSLN Fab (ie, comprising two anti-MSLN Fabs and one anti-CD3 Fab) showed a reduction in tumor volume. As further described herein, a difference is significant when the calculated p value is < 0.05, such as < 0.005.

[0255] In some embodiments, when compared to a bispecific construct based on a 1+1 format (e.g., a construct comprising one anti-CD3 (scFv or Fab) domain and one anti-MSLN (scFv or Fab) domain), a trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can exhibit greater anti-tumor activity in the presence of soluble MSLN. This experimental scenario can simulate the shedding of MSLN from MSLN-expressing tumor cells in vivo, which can result in circulating MSLN in the serum of subjects with MSLN-expressing tumors. Therefore, the shed MSLN can be referred to as soluble MSLN and can comprise or consist of the extracellular portion of the membrane-bound MSLN located on the tumor cell. In some of these embodiments, when soluble MSLN is present at a concentration of about 50 ng / mL, 100 ng / mL or about 150 ng / mL, the trivalent bispecific antibody construct of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can exhibit greater anti-tumor activity compared to a construct based on a 1+1 format. In some embodiments, under such conditions, the anti-tumor activity of the trivalent bispecific construct can be about 10 times, 20 times, 50 times or about 100 times higher than the anti-tumor activity of the corresponding construct based on a 1+1 format.

[0256] In some embodiments, the trivalent bispecific antibody constructs of the present disclosure comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains can trigger cytokine release from immune cells (e.g., T cells) in an MSLN-dependent manner. In such embodiments, when compared with the constructs being present with T cells and tumor cells without MSLN expression, measurable cytokine release can be detected only when the constructs are present with T cells and tumor cells expressing MSLN. In further embodiments, the trivalent bispecific antibody constructs 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 tumor cells expressing MSLN, measurable T cell proliferation of T cells may not be detected, however, at least about 40%, 50%, 60%, 70% or at least about 80% T cell proliferation can be measured under conditions where tumor cells expressing MSLN are present.

[0257] In some embodiments, the trivalent bispecific antibody construct comprising an anti-CD3 Fab domain and two anti-MSLN scFv domains disclosed herein can be tolerated in cynomolgus monkeys at doses of about 1 mg / kg, 10 mg / kg, and 30 mg / kg. When the trivalent bispecific antibody construct v32523 is administered intravenously to cynomolgus monkeys at doses of 1, 10, and 30 mg / kg, a transient increase in one or more biomarkers can be observed. Such one or more biomarkers can include one or more of IL-6, monocyte chemotactic protein 1 (MCP-1), and granulocyte-macrophage colony stimulating factor (GM-CSF). In certain embodiments, about 24 hours after injection of 10 or 30 mg / kg antibody constructs (e.g., v32523), serum IL-6 concentrations can range from about 150 pg / mL to about 250 ng / mL. Other effects of administering the trivalent bispecific antibody constructs may include increases in fibrinogen levels, for example about 38%, 68%, 166% and 142% on day 4 after injection of 0.1, 1, 10 and 30 mg / kg antibody constructs, respectively, and compared to pre-dose levels.

[0258] In further embodiments, the 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 in cynomolgus monkeys of about 2.3, 2.6 or 3 days.

[0259] In various embodiments herein, the trivalent bispecific antibody construct can be construct v32523 or v21812, as further described herein.

[0260] Sequence identity

[0261] As described in other parts of the present disclosure, certain embodiments herein relate to an isolated polypeptide or isolated polypeptide set of an antibody construct (e.g., polypeptide chains H1, H2, L1, etc., or portions thereof, such as domains), and a polynucleotide or polynucleotide set encoding an antibody construct as described herein. In this context, a polynucleotide can encode all or part of an antibody construct, such as one or more polypeptide chains (e.g., H1, H2, L1, etc.) of an antibody construct.

[0262] The terms "nucleic acid", "nucleic acid molecule" and "polynucleotide" are used interchangeably herein and refer to a polymeric form of nucleotides (deoxyribonucleotides or ribonucleotides or their analogs) of any length. 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.

[0263] A polynucleotide that "encodes" a given polypeptide is one that is transcribed (in the case of DNA) or translated (in the case of mRNA) into the polypeptide 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' (carboxyl) terminus. A transcription termination sequence may be located 3' to the coding sequence.

[0264] In certain embodiments, the disclosure relates to polynucleotides and / or polypeptide sequences that are identical or substantially identical to another polynucleotide and / or polypeptide sequence.In the context of two or more polynucleotides or polypeptide sequences, the term "identical" refers to two or more identical sequences or subsequences, that is, respectively having the same nucleotide or amino acid monomer sequence (that is, 100% sequence identity).When sequences are compared and aligned to obtain the maximum consistency measured as one of the conventional sequence comparison algorithms known to those of ordinary skill in the art or by manual alignment and visual inspection on a comparison window or in a specified region, if these sequences have a certain percentage or a certain number of amino acid residues or nucleotides, respectively, i.e., at least about 80%, about 85%, about 90%, about 95%, about 97% or at least about 99% identity in a specified region, then the polypeptide or polynucleotide sequence herein has a total "sequence identity".This definition also refers to the component of the test polynucleotide sequence. Certain sequence identity can be present in a region of at least about 50 amino acids or nucleotides in length, or in a region of about 75 to about 100 amino acids or nucleotides in length, or, in the case of unspecified, in the entire sequence of a polynucleotide or polypeptide. For sequence comparison, a test sequence is usually compared with a specified reference sequence. When using a sequence comparison algorithm, the test sequence and the reference sequence are input into a computer, and if necessary, subsequence coordinates are specified, and sequence algorithm program parameters are specified. Default program parameters can be used, or alternative parameters can be specified. The sequence comparison algorithm then calculates the sequence identity percentage (%) of the test sequence relative to the reference sequence based on the program parameters.

[0265] The term "comparison window" as used herein refers to a sequence segment comprising continuous amino acid or nucleotide positions, which can be from about 20 to about 1000 continuous amino acid or nucleotide positions, for example from about 50 to about 600 or from about 100 to about 300 or from about 150 to about 200 continuous amino acid or nucleotide positions, over which the two sequences can be compared after the test sequence is optimally aligned with a reference sequence having the same number of continuous positions. In certain embodiments, longer segments up to and including the full-length sequence may also be used as a comparison window. Sequence alignment methods for comparison are known to those of ordinary skill in the art. Optimal alignment of sequences for comparison can be performed, for example, by the local homology algorithm of Smith & Waterman, 1970, Adv. Appl. Math., 2:482c; the homology alignment algorithm of Needleman & Wunsch, 1970, J. Mol. Biol., 48:443; the similarity search method of Pearson & Lipman, 1988, Proc. Natl. Acad. Sci. USA, 85:2444; or by computerized implementations of these algorithms (e.g., GAP, BESTFIT, FASTA or TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, Madison, WI), or by manual alignment and visual inspection (see, e.g., Ausubel et al., Current Protocols in Molecular Biology, (1995 supplement), Cold Spring Harbor Laboratory Press). Examples of available algorithms suitable for determining percentage of sequence identity are 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 analysis is publicly available through the website of the National Center for Biotechnology Information (NCBI).

[0266] Certain embodiments described herein relate to variant sequences (e.g., variant V HIn some embodiments, when compared to a reference sequence (such as a wild-type sequence), one or more amino acid modifications of the variant sequence include one or more amino acid substitutions. In such embodiments, 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, "conservative substitutions" as used herein are considered to be substitutions of one amino acid with another amino acid having similar physical, chemical and / or structural properties. Common conservative substitutions are listed in the first column of Table 4. It will be appreciated by those skilled in the art that the main factors determining the formation of conservative substitutions are usually the size of the amino acid side chain and its physical / chemical properties, but certain environments allow the substitution of a given amino acid with a wider range of amino acids than those listed in the first column of Table 4. These additional amino acids tend to have similar properties to the substituted amino acid, but vary more in size, or have similar sizes, but vary more in physical / chemical properties. This wider range of conservative substitutions is listed in the second column of Table 4.

[0267] Table 4: Conservative amino acid substitutions

[0268]

[0269] The skilled artisan can readily determine the most appropriate set of substitutions to select given the particular protein environment in which the amino acid substitution is to be made.

[0270] Pharmaceutical composition

[0271] In some embodiments, the disclosure relates to pharmaceutical compositions, which may include one or more antibody constructs described herein. In various embodiments, the pharmaceutical compositions herein may further include pharmaceutically acceptable excipients, carriers, buffers, stabilizers or other materials well known to those skilled in the art. Such materials are generally nontoxic and do not interfere with the efficacy of active ingredients (i.e., antibody constructs). The exact nature of carriers or other materials may depend on the route of administration. Therefore, the pharmaceutical compositions herein may be formulated for various uses and routes of administration, such as for oral, intravenous, skin, subcutaneous, nasal, intramuscular or intraperitoneal routes of administration.

[0272] Pharmaceutical compositions for oral administration can be in tablet, capsule, powder or liquid form. Tablets can contain solid carriers, such as gelatin or adjuvants. Liquid pharmaceutical compositions usually contain liquid carriers, such as water, petroleum, animal or vegetable oils, mineral oils or synthetic oils. Physiological saline solutions, dextrose or other sugar solutions or glycols, such as ethylene glycol, propylene glycol or polyethylene glycol can be included.

[0273] For intravenous, cutaneous or subcutaneous injection, or injection at the site of affliction (i.e., at the site of a tumor), the active ingredient (i.e., antibody construct) can be in the form of a parenteral acceptable aqueous solution, which is pyrogen-free and has a suitable pH, isotonicity and stability. Those skilled in the art are well able to prepare suitable solutions using, for example, isotonic vehicles such as sodium chloride injection, Ringer's injection, lactated Ringer's injection. Preservatives, stabilizers, buffers, antioxidants and / or other additives can be included as needed.

[0274] For antibody constructs according to the present disclosure administered to a subject, it is preferably administered with a "therapeutically effective amount" sufficient to show benefit to an individual, as further described herein. The actual amount administered, as well as the rate of administration and the time course, may depend on the nature and severity of the disease (e.g., cancer) being treated. Treatment prescriptions (e.g., decisions on dosage, etc.) are the responsibility of general practitioners and other doctors, and typically the condition to be treated, the condition of individual patients, the delivery site, the method of administration, and other factors known to practitioners will be considered. Examples of the above-mentioned techniques and protocols can be found in Remington's Pharmaceutical Sciences, 16th edition, Osol, A. (ed.), 1980.

[0275] In some embodiments, the pharmaceutical composition may comprise a second active ingredient (eg, another protein or small molecule) in addition to the antibody constructs described herein.

[0276] Reagent test kit

[0277] The disclosure also describes a medicine box, which includes one or more antibody constructs as described herein, or a pharmaceutical composition comprising such antibody constructs as described herein, wherein such a medicine box may further include instructions for use. Therefore, in certain embodiments, a medicine box is described herein, which includes a carrier and instructions for use for expressing an antibody construct described herein. In certain embodiments, a medicine box is described herein, which includes a host cell and instructions for use comprising a carrier for expressing an antibody construct. In some embodiments, the disclosure relates to a medicine box, which includes purified antibody constructs and instructions for use. Purified antibody constructs can be lyophilized or provided in dry form, such as powder or particles, and the medicine box may additionally contain a suitable solvent for reconstructing lyophilized or dried components.

[0278] The medicine box may also include a container and a label or package insert on or associated with the container. The label or package insert contains instructions typically included in the commercial packaging of the therapeutic product, which provides information or instructions about the indications, usage, dosage, administration, contraindications and / or warnings of the use of such therapeutic products (e.g., antibody constructs described herein). The label or package insert may also include a notification in the form of a government agency regulation for the manufacture, use or sale of a drug or biological product, which reflects the approval of the agency to manufacture, use or sale for human or animal administration. The container may hold a composition comprising the antibody construct of the present disclosure. In some embodiments, the container may have an aseptic access port. For example, the container may be an intravenous solution bag or a vial with a stopper that can be pierced by a hypodermic needle.

[0279] In addition to the container containing the composition comprising the antibody construct, the kit may further include one or more additional containers comprising other components of the kit. For example, a pharmaceutically acceptable buffer (such as bacteriostatic water for injection (BWFI), phosphate-buffered saline, Ringer's solution or dextrose solution) or other buffers or diluents may be included in such a kit.

[0280] Suitable containers may include, for example, bottles, vials, syringes, and intravenous solution bags, etc. The container may be formed of various materials such as glass or plastic. If appropriate, one or more components of the kit (e.g., antibody construct) may be lyophilized or provided in a dry form (such as a powder or granules), and the kit may additionally contain a suitable solvent for reconstitution of the lyophilized or dried components.

[0281] The kits herein may also include other materials desirable from a commercial and user perspective, such as filters, needles, and syringes.

[0282] Methods for preparing antibody constructs

[0283] In some embodiments, the disclosure relates to methods for preparing antibody constructs as described herein. In various embodiments, antibody constructs of the present disclosure can be produced using standard recombinant methods known in the art (see, e.g., U.S. Patent No. 4,816,567 and "Antibodies: A Laboratory Manual," 2nd edition, Greenfield, Cold Spring Harbor Laboratory Press, New York, 2014).

[0284] In order to recombinantly produce antibody constructs as described herein, polynucleotides or polynucleotide sets encoding antibody constructs can be produced, and inserted into one or more vectors for further cloning and / or expression in host cells. The polynucleotides encoding antibody constructs can be produced by standard methods known in the art (see, for example, Ausubel et al., Current Protocols in Molecular Biology, John Wiley & Sons, New York, 1994 & Update, and "Antibodies: A Laboratory Manual, " 2nd edition, Greenfield compiles, Cold Spring Harbor Laboratory Press, New York, 2014). As will be appreciated by those skilled in the art, the quantity of polynucleotides required for antibody construct expression may depend on the pattern and / or geometry of antibody constructs, including the quantity of polypeptide chains included by, for example, antibody constructs. For example, when antibody constructs include three polypeptide chains (for example, H1, H2 and L1), three kinds of polynucleotides encoding a kind of polypeptide chain can be used. In the embodiment using two or more polynucleotides, such two or more polynucleotides can be incorporated into a carrier or more than one carrier (for example, two or three independent carriers).

[0285] Typically, for expression, the polynucleotide or polynucleotide set encoding the antibody construct herein can be incorporated into an expression vector together with one or more regulatory elements (such as transcription elements that can be used for effective transcription of the polynucleotide). Examples of such regulatory elements include, but are not limited to, promoters, enhancers, terminators, and polyadenylation signals. It will be appreciated by those skilled in the art that the selection of regulatory elements may depend on the host cell selected for expressing the polypeptide of the antibody construct, and such regulatory elements may be derived from a variety of sources, including bacteria, fungi, viruses, mammals, or insect genes. The expression vector may optionally further contain a heterologous nucleic acid sequence that promotes the 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. The expression vector may be an extrachromosomal vector or an integration vector. Therefore, in some embodiments, the amino acid sequence of the polypeptide chain (e.g., chain H1, H2, L1, etc.) of the antibody construct expressed as described herein may include a signal peptide sequence. Such signal peptide sequences can vary depending on the expression system and conditions used to produce the antibody construct. Exemplary signal peptide sequences can include the amino acid sequence METDTLLLWVLLLWVPGSTG (SEQ ID NO: 49) or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 50) (e.g., for H1, H2, etc.), or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 50) or MGWSCIILFLVATATGVHS (SEQ ID NO: 51) (e.g., for L1, L2, etc.). In certain embodiments, after the intracellular polypeptide chain is expressed, one or more heavy chains (e.g., H1, H2, etc.) of the antibody constructs described herein may include a C-terminal lysine residue. In various embodiments, this C-terminal lysine residue can be enzymatically cleaved from the polypeptide chain before further processing (e.g., purification, formulation, etc.) and before using the corresponding antibody construct, such as before administering the construct to a subject in need thereof.

[0286] Certain embodiments of the present disclosure for producing antibody constructs relate to vectors (such as expression vectors) comprising one or more polynucleotides encoding at least a portion of the antibody constructs described herein. One or more polynucleotides may be contained in a single vector or more than one vector. In some embodiments, the polynucleotides are contained in a polycistronic vector. Expression vectors that can be used to express polynucleotides include, but are not limited to, pTT5 and pUC15 cells comprising vectors encoding antibody constructs.

[0287] Suitable host cells for cloning or expressing antibody construct polypeptides include various prokaryotic or eukaryotic cells as 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, Escherichia coli, Aeromonas salmonicida or Bacillus subtilis cells. In certain embodiments, antibody constructs can be produced in bacteria, particularly when glycosylation and Fc effector functions are not required or expected for the intended purpose of antibody constructs, such as, for example, U.S. Patent No. 5,648,237, No. 5,789,199 and No. 5,840,523 and Charlton, Methods in Molecular Biology, Vol. 248, pp. 245-254, BKCLo, Humana Press, Totowa, NJ, described in 2003. In certain embodiments, eukaryotic microorganisms such as filamentous fungi or yeast are suitable expression host cells, particularly fungi and yeast strains whose glycosylation pathways have been "humanized" to result in the production of antibodies with partially 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).

[0288] In various embodiments, suitable host cells for expressing glycosylated antibody constructs are eukaryotic cells. For example, U.S. Pat. Nos. 5,959,177, 6,040,498, 6,420,548, 7,125,978, and 6,417,429 describe PLANTIBODIES for producing antibodies and portions thereof (e.g., scFv, Fab, etc.) in transgenic plants. TMMammalian cell lines adapted for growth in suspension are particularly useful for expression of the antibody constructs described herein. Examples include, but are not limited to, monkey kidney CV1 line transformed by SV40 (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 (Hep G2), mouse mammary tumor (MMT 060562), TRI cells (see, e.g., Mather et al., 1982, Annals NY Acad Sci, 383:44-68), MRC 5 cells, FS4 cells, Chinese hamster ovary (CHO) cells (including DHFR - 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 (BKCLo, ed., Humana Press, Totowa, NJ, 2003).

[0289] In certain embodiments, the host cell used to produce the antibody constructs described herein is a transient or stable higher eukaryotic cell line, such as a mammalian cell line. In some embodiments, the host cell is a mammalian HEK293T, CHO, HeLa, NS0 or COS cell. In some embodiments, the host cell is a stable cell line that allows mature glycosylation of the antibody construct.

[0290] Host cells comprising one or more expression vectors encoding antibody constructs can be cultured using conventional methods to produce antibody constructs. Alternatively, in some embodiments, host cells comprising expression vectors encoding antibody constructs can be used therapeutically or prophylactically for delivering antibody constructs to subjects, or polynucleotides or expression vectors can be administered ex vivo to cells from subjects, and the cells are then returned to the body of the subject.

[0291] In some embodiments, the host cell comprises (e.g., has been transformed with) a vector comprising a V or V encoding a binding domain of an antibody construct described herein. L and V H In some embodiments, the host cell comprises a vector (e.g., has been transformed with a vector) comprising a polynucleotide encoding a full-length polypeptide chain of an antibody construct described herein (e.g., H1, H2, or L1 as described herein). In another example, the host cell comprises a first vector comprising a V encoding a binding domain. L and a second vector comprising a corresponding V encoding a binding domain. H In various embodiments, the host cell is eukaryotic, such as a Chinese hamster ovary (CHO) cell, a human embryonic kidney (HEK) cell, or a lymphoid cell (e.g., a Y0, NS0, Sp20 cell). In certain embodiments, the host cell is an Expi293 TM (Thermo Fisher, Waltham, MA). In certain embodiments, the host cells used herein are CHO-S cells (National Research Council Canada) or HEK293 cells.

[0292] Certain embodiments of the present disclosure relate to methods for preparing antibody constructs, the methods comprising culturing a host cell into which one or more polynucleotides encoding antibody constructs or one or more expression vectors encoding antibody constructs have been introduced under conditions suitable for expressing the antibody constructs. This method may further include recovering the antibody construct from the host cell (or from the host cell culture medium). In some embodiments, this method may further include purifying the antibody construct.

[0293] Cell culture media that can be used include, but are not limited to, DMEM (Thermo Fisher, Waltham, MA), Opti-MEM TM (Thermo Fisher, Waltham, MA), Opti-MEM TM I reduced serum medium (Thermo Fisher, Waltham, MA), RPMI-1640 medium, Expi293 TMExpression medium (Thermo Fisher, Waltham, MA) and FreeStyle CHO expression medium (Thermo Fisher Scientific, Waltham, MA). Cell culture media can be supplemented with serum (e.g., fetal bovine serum (FBS)), amino acids (e.g., L-glutamine), antibiotics (e.g., penicillin and streptomycin) and / or antifungal agents (e.g., amphotericin) or any other supplement routinely used to support cell culture.

[0294] A. Purification of Antibody Constructs

[0295] In various embodiments, the antibody constructs of the present disclosure are purified after expression. Proteins (such as antibody constructs of the present disclosure) can be separated or purified in a variety of ways known to those skilled in the art (see, for example, Protein Purification: Principles and Practice, 3rd edition, Scopes, Springer-Verlag, NY, 1994). Standard purification methods that can be used for antibody constructs disclosed herein include chromatographic techniques performed at atmospheric pressure or under high pressure using systems such as FPLC and HPLC, including ion exchange, hydrophobic interaction, affinity, molecular exclusion or gel filtration and reverse phase chromatography. Other purification methods include electrophoresis, immunity, precipitation, dialysis and chromatofocusing techniques. Ultrafiltration and diafiltration techniques combined with protein concentration can also be used. As is well known in the art, a variety of natural proteins bind to the Fc domains and other domain elements of antibody constructs, and in some embodiments, these proteins can be used for the purification of antibody constructs. For example, bacterial proteins A and G can be combined with the Fc domains of some antibody constructs. Similarly, bacterial protein L can be combined with the Fab domains of some antibody constructs. Purification can also be achieved by a specific fusion partner. For example, if a GST fusion is used, the antibody construct can be purified using glutathione resin, and if a His tag is used, Ni +2 The antibody construct may be purified by affinity chromatography or, if a Flag tag is used, by using an immobilized anti-Flag antibody. The necessary degree of purification may vary depending on the intended use of the antibody construct. Thus, in some embodiments, purification may not be required.

[0296] In certain embodiments, antibody constructs of the present disclosure are substantially pure. The term "substantially pure" (or "substantially purified"), when used in reference to antibody constructs described herein, refers to that the antibody construct is substantially or essentially free of components (such as initial cells, or host cells in the case of recombinantly produced antibody constructs) that are usually accompanied by the protein or interact with the protein as found in its naturally occurring environment. In certain embodiments, substantially pure antibody constructs are purified to have 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) other contaminating protein substances.

[0297] Assessment of antibody construct purity and / or homogeneity 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-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).

[0298] B. Post-translational modifications

[0299] In certain embodiments, the antibody constructs described herein may include one or more post-translational modifications. Such post-translational modifications may occur in vivo, or they may be performed in vitro after isolating the antibody construct from a host cell.

[0300] Post-translational modifications may include various modifications as known in the art (see, e.g., Proteins-Structure and Molecular Properties, 2nd Edition, TECreighton, WH Freeman and Company, New York, 1993; Post-Translational Covalent Modification of Proteins, BC 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. NY Acad. Sci., 663: 48-62). In those embodiments where the antibody construct comprises one or more post-translational modifications, the antibody construct may comprise the same type of modification at one or more sites (e.g., amino acid residues), or it may comprise different modifications at different sites.

[0301] Examples of post-translational modifications may include glycosylation, acetylation, phosphorylation, amidation, derivatization by known protecting / blocking groups, formylation, oxidation, reduction, proteolytic cleavage, or specific chemical cleavage by cyanogen bromide, trypsin, chymotrypsin, papain, V8 protease, or NaBH4.

[0302] Other examples of post-translational modifications may include, for example, the addition or removal of N- or O-linked carbohydrate chains, chemical modifications of N- or O-linked carbohydrate chains, the treatment of N- or C-terminal ends, the attachment of chemical moieties to the amino acid backbone, and the addition or deletion of N-terminal methionine residues produced by prokaryotic host cell expression. Post-translational modifications may also include modifications such as enzymatic, fluorescent, isotopic, or affinity labels with detectable markers to allow detection and separation of proteins. Examples of suitable enzyme 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 materials include, but are not limited to, umbelliferone, fluorescein, fluorescein isothiocyanate, rhodamine, dichlorotriazineamine 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, platinum, xenon and fluorine.

[0303] Additional examples of post-translational modifications may include acetylation, 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.

[0304] How to use

[0305] In some embodiments, the disclosure relates to methods of using the antibody constructs of the disclosure. In certain embodiments, methods of using the antibody constructs described herein are described herein for treating a disease or condition in a subject in need. This method may include administering an antibody construct or a pharmaceutical composition comprising an antibody construct to a subject in need. In certain embodiments, the subject is a mammal. In certain embodiments, the subject is a human or a rodent.

[0306] In some embodiments, the disclosure relates to a method for treating a cancer in a subject in need thereof, the method comprising administering to the subject an antibody construct of the disclosure or a pharmaceutical composition comprising the antibody construct. Cancers that can be treated using the methods and antibody constructs disclosed herein may include, but are not limited to, blood neoplasms (including leukemias, myelomas, and lymphomas), carcinomas (including adenocarcinomas and squamous cell carcinomas), melanomas, and sarcomas. Carcinomas and sarcomas are also often referred to as "solid tumors." In certain embodiments, cancer is a solid tumor. In certain embodiments, cancer is a leukemia. In certain embodiments, cancer is a lymphoma.

[0307] When used in the methods described herein, the antibody constructs of the present disclosure can exert a cytotoxic or cytostatic effect and can result in one or more of the following: a reduction in tumor size, a slowing or prevention of increase in tumor size, an increase in the disease-free survival time between the disappearance or removal of a tumor and its recurrence, prevention of the initial or subsequent occurrence of a tumor (e.g., metastasis), an increase in the time to progression, a reduction in one or more adverse symptoms associated with a tumor, or an increase in the overall survival time of a subject with a tumor, or a combination of the foregoing.

[0308] 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 appreciated by those skilled in the art, the route of administration and / or mode may vary depending on the desired treatment outcome. In various embodiments, the antibody construct of the present disclosure may be administered by systemic administration or topical administration. Topical administration may be administration at the tumor site or into a tumor draining lymph node. Typically, the antibody construct may be administered parenterally, for example, by intravenous, intramuscular, intradermal, intraperitoneal, subcutaneous or spinal administration, such as by injection or infusion.

[0309] The treatment of (e.g., a subject's cancer) 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 as described herein that effectively achieves the desired therapeutic outcome at the necessary dosage and within the necessary time period. The therapeutically effective amount may vary according to factors such as the subject's disease state, age, sex, and body weight. A therapeutically effective amount is also an amount in which the beneficial effects of treatment exceed any potential toxicity or deleterious effects of the antibody construct. A "sufficient amount" generally refers to an amount sufficient to produce the desired effect, such as an amount sufficient to produce an anti-tumor response to a target (e.g., tumor) cell or tissue, such as by engaging an immune cell (e.g., T cell) using a trivalent bispecific antibody construct as described herein.

[0310] A skilled medical practitioner can determine the appropriate dosage of the antibody constructs described herein. The selected dosage level may depend on various pharmacokinetic factors, including the activity of the specific antibody construct employed (e.g., antigen affinity), the route of administration, the time of administration, the excretion rate of the construct, the duration of treatment, other drugs, compounds and / or materials used in combination with the antibody construct, such as anticancer drugs, and 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.

[0311] In some embodiments, the method for treating a disease (e.g., cancer) of a subject comprises administering a second active ingredient (e.g., another protein or small molecule) in addition to the antibody constructs described herein. Such a second active ingredient can be administered simultaneously or sequentially with the antibody constructs according to the condition to be treated.

[0312] In some embodiments, the present disclosure relates to a method for 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, this antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises: (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 dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain (e.g., the N-terminus of the heavy chain or the light chain). In various embodiments, when the concentration of the antibody construct in the cell population is from about 10 -2 pM increased to about 10 2 pM, the anti-tumor immune response provides a dose-dependent reduction of at least 30%, 40%, 50% or 60% in viable tumor cells in the cell population, wherein the MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0313] In some embodiments, the present disclosure relates to a method for inhibiting the proliferation 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, this antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises: (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 dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain. As defined herein, when up to about 10 -2 pM, 10 -1 The proliferation of tumor cells is inhibited when an increase of up to about 5%, 3% or up to about 1% in the live tumor cell count in the cell population is observed over a period of at least about 5, 10, 20 or 48 hours at a pM or 1 pM concentration of the antibody construct, wherein the MSLN expression of the tumor cells is at least about 15,000 MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0314] In some embodiments, the present disclosure relates to a method for killing 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, this antibody construct binds to CD3 on immune cells and MSLN on tumor cells, and comprises: (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 dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain. In various embodiments, when the concentration of the antibody construct is from about 10 -2 pM increased to about 10 2 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 at pM, wherein MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0315] In any of the methods described herein, the immune cells include or consist of one or more types of T cells.

[0316] As described herein, in various embodiments, the antibody constructs described herein can be administered to a subject in need, such as a subject with cancer, to modulate the subject's immune response. The immune response that can be modulated using the antibody constructs disclosed herein can be an anti-tumor immune response of the subject, for example, in various embodiments, such a modulated immune response can occur locally at the tumor site. Therefore, in certain embodiments, the antibody constructs described herein can initiate and / or upregulate a local immune response, such as an anti-tumor response of the subject's immune system, to induce a local cytotoxic effect against a tumor at the tumor site.

[0317] In various embodiments, the antibody constructs described herein, such as trivalent bispecific antibody constructs capable of monovalently binding to CD3 (via one Fab domain) and divalently binding to MSLN (via two scFv domains) can have a wider therapeutic window than comparable conventional molecules (e.g., v31805 (MH6T-TriTAC)), which can allow for the administration of higher doses, potentially resulting in increased anti-tumor effects without causing side effects and off-target effects. This wider therapeutic window can be due to 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.

[0318] In some embodiments, the present disclosure relates to a method for inhibiting the growth of tumors expressing MSLN and / or reducing tumor volume in a subject in need thereof, the method comprising administering to the subject an effective amount of an antibody construct of the present disclosure. In various embodiments, this in vivo anti-tumor is triggered by an antibody construct that (e.g., simultaneously) binds to CD3 on immune cells and MSLN on tumor cells, and comprises: (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 dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the Fab domain.

[0319] In various embodiments, and as further described elsewhere herein, inhibition of tumor growth and / or reduction of tumor volume in a subject can be induced by causing the antibody construct to simultaneously bind to CD3 on an immune cell and MSLN on a tumor cell and form a TCR-independent artificial immune synapse within the tumor environment of the subject, thereby eliciting an anti-tumor cytotoxic effect mediated by the activated immune cell and directed against the tumor cell. The immune cell can be a T cell.

[0320] In various embodiments, antibody constructs are administered to subjects 4 times a week (Q7Dx4), and the tumor growth of subjects can be suppressed for at least 20 days, 30 days, or 50 days. In addition, in some embodiments, about 15 days after the start of treatment, tumor volume can be reduced by at least 20%, 30%, 40%, or at least about 50%. Such therapeutic effects can be achieved when antibody constructs are administered to subjects at a dosage of about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg. Subjects can be rodents, non-human primates, or people.

[0321] In further embodiments, and with respect to the methods described herein, administration of a sufficient amount of an antibody construct to a subject in need thereof can provide one or more of the following to activate or upregulate an immune response: (i) modulation of T cell receptor signaling, (ii) modulation of T cell activation, (iii) modulation of proinflammatory cytokines, (iv) modulation of interferon-γ production by T cells, (v) modulation of T cell suppression, (vi) modulation of survival and / or differentiation of M2 tumor-associated macrophages (TAMs) or myeloid-derived suppressor cells (MDSCs), and / or (vii) modulation of cytotoxic or cytostatic effects on cells (e.g., cancer cells).

[0322] In some embodiments, the present disclosure relates to methods of regulating the immune response of cells or subjects using one or more trivalent bispecific antibody constructs of the present disclosure, wherein such regulation may include one or more of the following: (i) immune cell activation, (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 thereof. As described herein, in certain embodiments, the antibody construct activates T effector cells. In some embodiments, and as demonstrated herein, the antibody construct increases the production of one or more cytokines and / or signaling molecules by T effector cells, such as GM-CSF, TNF-α, MIP-1β, IFN-γ, IL-2, IL-12, IL-17, IL-21, and / or CXC motif ligand 13 (CXCL13).

[0323] As described herein, in various embodiments, the antibody constructs of the present disclosure comprise an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein one or more of the Fc polypeptides may comprise a modified CH2 domain (e.g., compared to a WT domain) comprising one or more amino acid modifications that result in reduced or eliminated binding of the Fc domain to one or more, or all, Fcγ receptors (also referred to herein as Fc "knockout" or "KO" variants).

[0324] Experimental methods

[0325] In some embodiments, the disclosure relates to experimental methods for analyzing and / or detecting antibody constructs of the disclosure. Such methods can be used, for example, to assess the in vitro and / or in vivo properties of antibody constructs, such as their pharmacokinetic (PK) and pharmacodynamic (PD) properties. Other properties and features of antibody constructs can be evaluated, such as their stability under certain conditions (e.g., temperature, pH, etc.), their solubility, or their behavior in the presence of certain other chemical components (such as other proteins or cells).

[0326] The 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 (using, for example, a BIAcore instrument) (Liljeblad et al., 2000, Glyco J, 17: 323-329), or a traditional binding assay (Heeley, 2002, Endocr Res, 28: 217--229). In certain embodiments, specific binding is defined as a degree of binding to an unrelated protein that is less than about 10%, 5%, or less than about 2% of the binding to the target antigen (e.g., CD3, MSLN, etc.) as measured by, for example, SPR. In certain embodiments, the specific binding of the antibody constructs herein to a particular antigen or epitope is defined as a dissociation constant (K D )≤1 μM, for example, ≤500 nM, ≤200 nM, ≤100 nM, ≤50 nM, ≤25 nM or ≤5 nM. In some embodiments, the specific binding of an antibody construct to a particular antigen or epitope is defined as the dissociation constant (K D ) is about 10 -6 M to about 10 -10 M, about 10 -6 M to about 10 -9 M, about 10 -8 M to about 10 -10 M, or about 10 -7 M to about 10 -9 M.

[0327] Additional experimental methods to characterize and evaluate the trivalent, bispecific antibody constructs of the present disclosure are described in Examples 1-20 herein.

[0328] Certain embodiments of the present disclosure

[0329] According to any one or more of embodiments 1-124, the present disclosure further describes antibody constructs, pharmaceutical compositions comprising such constructs, and methods of using such constructs.

[0330] 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, wherein the first scFv domain and the second scFv domain are 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 coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0331] Embodiment 2. The antibody construct of embodiment 1, wherein at least one of the first scFv domain and the second scFv domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0332] Embodiment 3. The antibody construct of embodiment 2, wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0333] Embodiment 4. An antibody construct comprising: a Fab domain capable of binding to an antigen on a cytotoxic effector cell; a first scFv domain capable of binding to MSLN and comprising a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V LA domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; a second scFv domain, wherein the second scFv domain is 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 coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0334] Embodiment 5. The antibody construct of embodiment 4, wherein the second TAA is also MSLN.

[0335] Embodiment 6. An antibody construct as described in any one of Embodiments 4-5, wherein the second scFv domain comprises V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0336] Embodiment 7. An antibody construct as described in any one of embodiments 1-6, wherein the cytotoxic effector cells are T cells.

[0337] Embodiment 8. An antibody construct as described in any one of Embodiments 1-7, wherein the antigen bound by the Fab domain is CD3.

[0338] Embodiment 9. An antibody construct as described in any one of embodiments 1-8, wherein the Fab domain comprises V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131.

[0339] Embodiment 10. An antibody construct as described in any one of embodiments 1-9, wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence shown in SEQ ID NO: 103; and V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:101.

[0340] Embodiment 11. The antibody construct of embodiment 10, wherein for the first scFv domain and the second scFv domain, the V L Linker scFv With the V H Domain coupling.

[0341] Embodiment 12. The antibody construct of embodiment 11, wherein the linker scFv Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5 (SEQ ID NO: 132), or consists of said amino acid sequence.

[0342] Embodiment 13. An antibody construct as described in any one of Embodiments 10-12, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. H -Connectors scFv -V L or V L -Connectors scFv -V H domain structure.

[0343] Embodiment 14. An antibody construct as described in any of Embodiments 10-13, wherein 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 shown in SEQ ID NO: 149.

[0344] Embodiment 15. An antibody construct as described in any one of Embodiments 1-14, wherein the Fab domain comprises V H A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 105; and C H1 A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:106.

[0345] Embodiment 16. An antibody construct as described in Embodiment 15, wherein the Fab domain further comprises a light chain (L1), wherein the light chain comprises V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 115; and C L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 116, wherein the light chain (L1) is identical to the V H -C H1 The domains pair to form the Fab domain.

[0346] Embodiment 17. An antibody construct as described in any 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 shown in SEQ ID NO:114.

[0347] Embodiment 18. An antibody construct as described in any one of Embodiments 1-17, wherein the antibody construct comprises a first heavy chain (H1), the first heavy chain comprising scFv2-V from N-terminus to C-terminus. H -C H1 wherein the C-terminus of the second scFv domain (scFv2) is adjacent to the V H N-terminal coupling of the domain.

[0348] Embodiment 19. The antibody construct of embodiment 18, wherein the C-terminus of the second scFv domain is connected via a linker scFv-Fab With the V H N-terminal coupling of the domain.

[0349] Embodiment 20. The antibody construct of embodiment 19, wherein the linker scFv-Fab Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5 (SEQ ID NO: 132), or consists of said amino acid sequence.

[0350] Embodiment 21. An antibody construct as described in any one of embodiments 18-20, wherein the first heavy chain (H1) further comprises the first Fc polypeptide, and wherein the C H1 The C-terminus of the H1 domain is coupled to the N-terminus of the first Fc polypeptide, thereby forming the following H1 domain structure: scFv2-V H -C H1 -CH2-CH3.

[0351] Embodiment 22. The antibody construct of embodiment 21, wherein the C H1 Linker Fab-Fc coupled to the first Fc polypeptide.

[0352] Embodiment 23. The antibody construct of embodiment 22, wherein the linker Fab-Fc It comprises or consists of an Ig hinge region.

[0353] Embodiment 24. An antibody construct as described in Embodiment 23, wherein the Ig hinge region is an IgG hinge region.

[0354] Embodiment 25. An antibody construct as described in Embodiment 24, wherein the IgG hinge region is an IgG1 hinge region.

[0355] Embodiment 26. An antibody construct as described in any one of Embodiments 22-25, wherein the linker Fab-Fc Comprising or consisting of an amino acid sequence having at least about 70%, 80%, 90% or 100% sequence identity to the sequence shown in SEQ ID NO:107.

[0356] Embodiment 27. The antibody construct of any one of Embodiments 1-26, wherein the dimeric Fc domain is a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0357] Embodiment 28. The antibody construct of any one of Embodiments 1-27, wherein the first Fc polypeptide is an IgG1 derived Fc polypeptide and comprises a first CH2 domain and a first CH3 domain.

[0358] 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.

[0359] Embodiment 30. An antibody construct as described in any of embodiments 28-29, wherein the first CH2 domain comprises one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH2 domain, and the one or more amino acid substitutions reduce or eliminate binding to an Fc receptor.

[0360] Embodiment 31. An antibody construct as described in any of Embodiments 18-30, 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 shown in SEQ ID NO:100.

[0361] Embodiment 32. An antibody construct as described in any one of Embodiments 1-31, wherein the antibody construct further comprises a second heavy chain (H2), and the second heavy chain comprises the first scFv domain coupled to the second Fc polypeptide from the N-terminus to the C-terminus, thereby forming a domain structure: scFv1-CH2-CH3.

[0362] Embodiment 33. The antibody construct of embodiment 32, wherein the C-terminus of the first scFv domain is connected via a linker scFv-Fc Coupled to the N-terminus of the second Fc polypeptide.

[0363] Embodiment 34. The antibody construct of embodiment 33, wherein the linker scFv-Fc It comprises or consists of an Ig hinge region.

[0364] Embodiment 35. An antibody construct as described in Embodiment 34, wherein the Ig hinge region is an IgG hinge region.

[0365] Embodiment 36. An antibody construct as described in Embodiment 35, wherein the IgG hinge region is an IgG1 hinge region.

[0366] Embodiment 37. An antibody construct as described in any one of Embodiments 33-36, wherein the linker scFv-FcComprising or consisting of an amino acid sequence having at least about 70%, 80%, 90% or 100% sequence identity to the sequence shown in SEQ ID NO:111.

[0367] Embodiment 38. An antibody construct as described in any one of Embodiments 32-37, wherein the second Fc polypeptide is also an IgG1-derived Fc polypeptide and comprises a second CH2 domain and a second CH3 domain.

[0368] 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.

[0369] Embodiment 40. An antibody construct as described in any of Embodiments 38-39, wherein the second CH2 domain comprises one or more amino acid substitutions compared to the corresponding wild-type IgG1 CH2 domain, and the one or more amino acid substitutions reduce or eliminate binding to an Fc receptor.

[0370] Embodiment 41. An antibody construct as described in any of embodiments 1-40, wherein 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 the group consisting of: 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 the group consisting of: L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V, and T350V_L351Y_S400E_F405A_Y407V The Fc polypeptide comprises in its CH3 sequence an amino acid substitution selected from the group consisting of: 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 the Fc polypeptide sequence is according to the EU numbering system.

[0371] Embodiment 42. An antibody construct as described in Embodiment 41, wherein amino acid substitutions in the CH3 sequence of the Fc polypeptide promote preferential pairing of the heavy chains H1 and H2 and promote heterologous Fc domain formation compared to homodimeric Fc domain formation.

[0372] Embodiment 43. An antibody construct as described in any of Embodiments 32-42, 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 shown in SEQ ID NO:110.

[0373] Embodiment 44. The antibody construct of any one of Embodiments 1-43, wherein the antibody construct is bispecific for CD3 and MSLN, is monovalent for CD3 and bivalent for MSLN.

[0374] Embodiment 45. An antibody construct as described in any of embodiments 1-44, 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 shown 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 shown 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 shown in SEQ ID NO: 114.

[0375] Embodiment 46. The antibody construct of claim 45, wherein the first heavy chain (H1) comprises or consists of the amino acid sequence shown in SEQ ID NO: 100.

[0376] Embodiment 47. An antibody construct as described in any of Embodiments 45-46, wherein the second heavy chain (H2) comprises or consists of the amino acid sequence shown in SEQ ID NO: 110.

[0377] Embodiment 48. The antibody construct of any one of claims 45-47, wherein the light chain (L1) comprises or consists of the amino acid sequence shown in SEQ ID NO: 114.

[0378] Embodiment 49. An antibody construct as described in any of Embodiments 1-44, wherein the antibody construct comprises (i) a first heavy chain (H1) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 171, (ii) a second heavy chain (H2) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 172, and (iii) a light chain (L1) comprising or consisting of the amino acid sequence shown in SEQ ID NO: 114.

[0379] Embodiment 50. An antibody construct as described in any of embodiments 1-9, 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 shown 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 shown 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 shown in SEQ ID NO: 114.

[0380] Embodiment 51. The antibody construct of any one of embodiments 18-50, wherein one or more of said heavy chains H1 and H2 further comprises a C-terminal lysine residue.

[0381] Embodiment 52. An antibody construct as described in any of Embodiments 1-51, wherein the antibody construct, when 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 / cell, is capable of achieving one or more, two or more, or all of the following: A) inducing the production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL TNFα; B) inducing the production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL IFNγ; or C) inducing the production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL IL-2.

[0382] Embodiment 53. An antibody construct as described in any one of Embodiments 1-52, wherein the antibody construct is capable of inducing at least 10-fold higher T cell-mediated cytotoxicity against tumor cells compared to a similar antibody construct comprising two anti-MSLN Fab domains instead of the first and second anti-MSLN scFv domains.

[0383] Embodiment 54. The antibody construct of any one of embodiments 1-53, wherein the binding affinity (K) of the antibody construct to MSLN is measured by SPR. D ) is about 0.7 nM, 0.8 nM, 0.9 nM, 1 nM or about 1.1 nM.

[0384] Embodiment 55. An antibody construct as described in any one of embodiments 8-54, wherein the binding affinity (K) of the antibody construct to CD3 is measured by SPR. D ) is about 30 nM, 40 nM, 50 nM or about 60 nM.

[0385] Embodiment 56. An antibody construct as described in any of Embodiments 1-55, wherein the antibody construct is capable of inducing cytotoxic effector cells to produce proinflammatory cytokines in an MSLN-dependent manner, wherein the MSLN-dependent activation of the cytotoxic effector cells is determined by measuring a reduction in cytokine production by at least about 20-fold, 50-fold, 100-fold, or at least about 1000-fold between a first cell population comprising a first tumor cell and an immune cell and a second cell population comprising a second tumor cell and an immune cell, wherein the MSLN expression of the first tumor cell is at least about 3, 4, or 5 times higher than the MSLN expression of the second tumor cell.

[0386] Embodiment 57. An antibody construct as described in any of Embodiments 1-56, wherein when administered to a subject, the antibody construct is capable of reducing the volume of a MSLN-expressing tumor in the subject by at least about 5% over a period of 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 4 times per week.

[0387] Embodiment 58. A pharmaceutical composition comprising the antibody construct of any one of embodiments 1-57, and a pharmaceutically acceptable carrier, excipient, diluent, or a combination thereof.

[0388] Embodiment 59. A nucleic acid molecule or a collection 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 1-57.

[0389] Embodiment 60. A vector or a collection of vectors comprising the nucleic acid molecule or a collection of nucleic acid molecules as described in embodiment 59.

[0390] Embodiment 61. A cell comprising the nucleic acid molecule or a collection of nucleic acid molecules of embodiment 59, or the vector or a collection of vectors of embodiment 60.

[0391] Embodiment 62. A method for producing an antibody construct as described in any one of claims 1-57, the method comprising: (a) obtaining a host cell culture comprising at least one host cell, wherein the at least one host cell comprises one or more nucleic acid molecules encoding one or more, two or more, three or more polypeptide chains that form the antibody construct; and (b) recovering the antibody construct from the host cell culture.

[0392] Embodiment 63. A method as described in Embodiment 62, further comprising purifying the antibody construct after step (b).

[0393] Embodiment 64. 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 as described in any one of Embodiments 1-57.

[0394] Embodiment 65. A method as described in embodiment 64, wherein when the concentration of the antibody construct in the cell population is increased from about 10 -2 pM increased to about 10 2 pM, the anti-tumor immune response provides a dose-dependent reduction of at least 30%, 40%, 50% or 60% in viable tumor cells in the cell population, wherein the MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0395] Embodiment 66. A method of inhibiting proliferation 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 the antibody construct of any one of embodiments 1-57.

[0396] Embodiment 67. A method as described in embodiment 66, wherein when using up to about 10 -2 pM, 10 -1 The proliferation of tumor cells is inhibited when an increase of up to about 5% in the live tumor cell count in the cell population is observed over a period of at least about 5, 10, 20 or 48 hours at a pM or 1 pM concentration of the antibody construct, wherein the MSLN expression of the tumor cells is at least about 15,000 MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0397] Embodiment 68. A method of killing tumor cells expressing MSLN, the method comprising contacting a cell population comprising tumor cells and immune cells expressing CD3 with an effective amount of the antibody construct of any one of embodiments 1-57.

[0398] Embodiment 69. A method as described in embodiment 68, wherein when the concentration of the antibody construct is increased from about 10 -2 pM increased to about 10 2 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 at pM, wherein MSLN expression of the tumor cells is at least about 15,000MSLN / cell and the ratio of immune cells to tumor cells in the cell population is about 5:1.

[0399] Embodiment 70. A method as described in any of Embodiments 64-69, wherein the immune cells include T cells.

[0400] Embodiment 71. The method of any one of Embodiments 64-70, wherein the antibody construct binds to CD3 on immune cells and MSLN on tumor cells.

[0401] Embodiment 72. A method of inhibiting growth or reducing tumor volume of a tumor expressing MSLN in a subject, the method comprising administering to the subject an effective amount of the antibody construct of any one of embodiments 1-57.

[0402] Embodiment 73. A method as described in Embodiment 72, wherein the antibody construct binds to CD3 on the subject's immune cells and MSLN on tumor cells, which triggers an anti-tumor immune response in the subject, thereby inhibiting tumor growth or reducing tumor volume in the subject.

[0403] Embodiment 74. A method as described in Embodiment 73, wherein inhibition of tumor growth and / or reduction of tumor volume is triggered by causing the antibody construct to simultaneously bind to CD3 on immune cells and MSLN on tumor cells within the tumor environment of the subject and form a TCR-independent artificial immune synapse.

[0404] Embodiment 75. A method as described in Embodiment 74, wherein the immune cell is a T cell.

[0405] Embodiment 76. A method as described in any of Embodiments 72-75, wherein tumor growth is inhibited for at least 20 days, 30 days or 50 days using administration of the antibody construct to the subject 4 times a week (Q7Dx4).

[0406] Embodiment 77. The method of any one of Embodiments 72-76, wherein tumor volume is reduced by at least 20%, 30%, 40%, or at least about 50% about 15 days after initiation of treatment.

[0407] Embodiment 78. The method of any one of Embodiments 72-77, wherein about 1 mg / kg, 1.5 mg / kg, or about 3 mg / kg of the antibody construct is administered to the subject.

[0408] Embodiment 79. The method of any one of Embodiments 72-78, wherein the subject is a rodent, a non-human primate, or a human.

[0409] Embodiment 80. The antibody construct of any one of Embodiments 1-57, for use in cancer treatment.

[0410] Embodiment 81. Use of the antibody construct of any one of Embodiments 1-57 in the manufacture of a medicament for cancer treatment.

[0411] Embodiment 82. An antibody construct comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN); and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: the Fab domain is coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0412] Embodiment 83. The antibody construct of Embodiment 82, wherein at least one of the first scFv domain and the second scFv domain comprises (i) V HA domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and (ii) V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0413] Embodiment 84. The antibody construct of embodiment 82 or embodiment 83, wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122; and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125.

[0414] Embodiment 85. An antibody construct comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V LA domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN), and wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: the Fab domain is coupled to the N-terminus of the first Fc polypeptide, the first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and the second scFv domain is coupled to the N-terminus of the Fab domain.

[0415] Embodiment 86. An antibody construct as described in any one of Embodiments 82-85, wherein the first scFv domain or the second scFv domain comprises (i) V H A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 103; and (ii) V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:101.

[0416] Embodiment 87. The antibody construct of any one of Embodiments 82-86, wherein the first scFv domain and the second scFv domain each comprise (i) V HA domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 103; and (ii) V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:101.

[0417] Embodiment 88. An antibody construct as described in any one of Embodiments 82-87, wherein the first scFv domain and the second scFv domain each comprise (i) V H A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 103; and (ii) V L A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 101.

[0418] Embodiment 89. An antibody construct as described in any one of embodiments 86-88, wherein the first scFv domain and the second scFv domain further comprise the V L Domain and the V H Domain-coupled linkers scFv .

[0419] Embodiment 90. The antibody construct of embodiment 89, wherein the linker scFv Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of the amino acid sequence.

[0420] Embodiment 91. An antibody construct as described in any one of embodiments 89-90, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. H -Connectors scFv -V L or V L -Connectors scFv -V H domain structure.

[0421] Embodiment 92. The antibody construct of Embodiment 91, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. H -Connectors scFv -V L domain structure.

[0422] Embodiment 93. The antibody construct of Embodiment 91, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. L -Connectors scFv -V H domain structure.

[0423] Embodiment 94. An antibody construct as described in any one of Embodiments 82-93, wherein the Fab domain comprises (i) V H A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 105; (ii) C H1 A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 106; (iii) V L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 115; and (iv) C L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:116.

[0424] Embodiment 95. An antibody construct as described in any one of Embodiments 82-94, wherein the antibody construct comprises a first heavy chain (H1) comprising scFv2-V from N-terminus to C-terminus. H -C H1 wherein the C-terminus of the second scFv domain (scFv2) is adjacent to the V H N-terminal coupling of the domain.

[0425] Embodiment 96. The antibody construct of embodiment 95, wherein the C-terminus of scFv2 is connected via a linker scFv-Fab With the V H N-terminal coupling of the domain.

[0426] Embodiment 97. The antibody construct of embodiment 96, wherein the linker scFv-Fab Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of the amino acid sequence.

[0427] Embodiment 98. An antibody construct as described in any one of embodiments 95-97, wherein H1 further comprises the first Fc polypeptide, the first Fc polypeptide comprises a first CH2 sequence (CH21) and a first CH3 sequence (CH31), and wherein the CH H1 The C-terminus of the H1 domain is coupled to the N-terminus of the first Fc polypeptide, thereby forming the following H1 domain structure (from N-terminus to C-terminus): scFv2-V H -C H1 -CH21-CH31.

[0428] Embodiment 99. The antibody construct of embodiment 98, wherein the C H1 Linker Fab-Fc coupled to the CH2 sequence of the first Fc polypeptide.

[0429] Embodiment 100. The antibody construct of embodiment 99, wherein the linker Fab-Fc It comprises or consists of an Ig hinge region.

[0430] Embodiment 101. An antibody construct as described in any of Embodiments 95-100, 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 shown in SEQ ID NO:100.

[0431] Embodiment 102. An antibody construct as described in any of Embodiments 82-101, wherein the antibody construct further comprises a second heavy chain (H2), the second heavy chain comprising the first scFv domain (scFv1) from N-terminus to C-terminus, the first scFv domain being coupled to the second Fc polypeptide comprising a second CH2 sequence (CH22) and a second CH3 sequence (CH32), thereby forming a domain structure: scFv1-CH22-CH32.

[0432] Embodiment 103. The antibody construct of embodiment 102, wherein the C-terminus of scFv1 is connected via a linker scFv-Fc Coupled to the N-terminus of the second Fc polypeptide.

[0433] Embodiment 104. The antibody construct of embodiment 103, wherein the linker scFv-Fc It comprises or consists of an Ig hinge region.

[0434] Embodiment 105. An antibody construct as described in any of Embodiments 102-104, 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 shown in SEQ ID NO:110.

[0435] Embodiment 106. The antibody construct of any one of embodiments 82-105, wherein the first Fc polypeptide, the second Fc polypeptide, or both the first Fc polypeptide and the second Fc polypeptide are IgG1 derived Fc polypeptides.

[0436] Embodiment 107. An antibody construct as described in any of Embodiments 82-106, wherein the dimeric Fc domain is a heterodimeric Fc domain, wherein 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.

[0437] Embodiment 108. An antibody construct as described in 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, wherein the one or more amino acid substitutions promote preferential pairing of the first Fc polypeptide and the second Fc polypeptide to form a heterodimeric Fc domain.

[0438] Embodiment 109. An antibody construct as described in embodiment 108, wherein one Fc polypeptide comprises a set of amino acid substitutions selected from the group consisting of: 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 the group consisting of: 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 the Fc polypeptide is according to the EU numbering system.

[0439] Embodiment 110. An antibody construct as described in any 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 a corresponding wild-type IgG1 CH2 domain sequence, wherein the one or more amino acid substitutions reduce or eliminate binding of the antibody construct to an Fcγ-receptor.

[0440] Embodiment 111. An antibody construct as described in any of embodiments 82-110, 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 shown 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 shown 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 shown in SEQ ID NO: 114.

[0441] Embodiment 112. An antibody construct as described in any of Embodiments 95-111, wherein one or more of the heavy chains H1 and H2 further comprises a C-terminal lysine residue.

[0442] Embodiment 113. An antibody construct comprising: a first heavy chain (H1) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 100, a second heavy chain (H2) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 110, and a light chain (L1) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 114.

[0443] Embodiment 114. An antibody construct as described in any of Embodiments 82-113, wherein the antibody construct, when 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 / cell, is capable of achieving one or more, two or more, or all of the following: A) inducing the production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL TNFα; B) inducing the production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL IFNγ; or C) inducing the production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL IL-2.

[0444] Embodiment 115. The antibody construct of any one of embodiments 82-114, wherein the binding affinity (K) of the antibody construct to MSLN is measured by SPR. D ) is about 0.7 nM, 0.8 nM, 0.9 nM, 1 nM or about 1.1 nM.

[0445] Embodiment 116. An antibody construct as described in any one of Embodiments 82-115, wherein the binding affinity (K) of the antibody construct to CD3 is measured by SPR. D ) is about 30 nM, 40 nM, 50 nM or about 60 nM.

[0446] Embodiment 117. The antibody construct of any one of embodiments 82-116, wherein after incubation in buffer solution at 40°C for 14 days, the purity of the antibody construct is at least about 95%, 96%, 97%, 98% or at least about 99% as measured by size exclusion chromatography and relative to other high molecular weight species (HMWS) present in the buffer solution after the incubation period.

[0447] Embodiment 118. A pharmaceutical composition comprising the tri-specific fusion protein of any one of Embodiments 82-117, and a pharmaceutically acceptable carrier, excipient, diluent, or a combination thereof.

[0448] Embodiment 119. A nucleic acid molecule or a collection of nucleic acid molecules encoding one or more, two or more, or three or more polypeptide chains forming an antibody construct as described in any of embodiments 82-117.

[0449] Embodiment 120. A vector or a collection of vectors comprising the nucleic acid molecule or a collection of nucleic acid molecules as described in embodiment 119.

[0450] Embodiment 121. A cell comprising the nucleic acid molecule or a collection of nucleic acid molecules of embodiment 119, or the vector or a collection of vectors of embodiment 120.

[0451] Embodiment 122. A method for producing an antibody construct as described in any of embodiments 82-117, the method comprising: (a) obtaining a host cell culture comprising at least one host cell, wherein the at least one host cell comprises 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.

[0452] Embodiment 123. An antibody construct as described in any one of Embodiments 82-117, for use in cancer treatment.

[0453] Embodiment 124. Use of an antibody construct as described in any one of Embodiments 82-117 in the manufacture of a medicament for cancer treatment.

[0454] Example

[0455] The following examples are provided for illustrative purposes and are not intended to limit the scope of the present disclosure in any way.

[0456] The practice of the present disclosure may include and utilize conventional methods of protein chemistry, biochemistry, recombinant DNA technology and pharmacology within the technical scope of the art unless otherwise specified. Such techniques are fully explained in the literature. See, for example, TECreighton, Proteins: Structures and Molecular Properties (WH Freeman and Company, 1993); A Lehninger, Biochemistry (Worth Publishers, Inc., current edition); Sambrook, et al., Molecular Cloning: A Laboratory Manual (2nd edition, 1989); Methods In Enzymology (S. Colowick and N. Kaplan, Academic Press, Inc.); Remington's Pharmaceutical Sciences, 18th edition (Easton, Pennsylvania: Mack Publishing Company, 1990); Carey and Sundberg Advanced Organic Chemistry 3rd edition (Plenum Press) Volumes A and B (1992).

[0457] Example 1: Design and preparation of bivalent and trivalent MSLNxCD3 bispecific antibody constructs

[0458] As described below, several bivalent or trivalent bispecific MSLNxCD3 antibody constructs were generated. Antibody constructs and controls were prepared in different formats and geometries, such as Figures 1A-1D Such antibody constructs were prepared to examine the effects of antibody format and geometry as well as anti-MSLN paratope affinity on the efficacy of these T cell engager constructs in being able to target and kill tumor cells expressing MSLN.

[0459] Design of trivalent bispecific antibody constructs targeting MSLN and CD3

[0460] Initially, bispecific antibody constructs were prepared in a format in which the MSLN antigen binding domain was a scFv domain or a Fab domain and the CD3 antigen binding domain was a Fab domain or a scFv domain. These constructs (including controls) further comprised an IgG1 Fc domain unless otherwise indicated (see, e.g., Table 6). Bispecific antibody constructs comprising a human IgG1 heterodimeric Fc domain comprised a CH3 domain amino acid substitution set that promoted heterodimeric Fc domain formation. These amino acid substitution sets are referred to herein as (i) Het FcA, having amino acid substitutions T350V / L351Y / F405A / Y407V in the first Fc polypeptide chain (A) and (ii) Het FcB, having amino acid substitutions T350V / T366L / K392L / T394W in the second Fc polypeptide chain (B). Those antibody construct variants labeled "FcKO" in Table 6 comprise the following CH2 amino acid substitutions that knock out, e.g., measurably reduce or eliminate FcγR binding: L234A, L235A, and D265S. The amino acid residues of the Fc domain are identified according to the EU index. In addition, several antibody construct formats utilize heterodimeric Fab domains that use C H1 and C L The constructs were constructed using previously described mutations in the genomic domains that promote selective and preferential pairing of correct heavy and light chains (see, e.g., WO 2014 / 082179; WO 2015 / 181805; WO 2017 / 059551).

[0461] As described in WO2015109131, bispecific antibody constructs for CD3 and MSLN were designed, expressed, and characterized. In brief, genes encoding the heavy and light chains of the antibody constructs were constructed via gene synthesis using codons optimized for human / mammalian expression. The anti-CD3 scFv domain and Fab domain sequences were generated from the anti-CD3 monoclonal antibody C3E6 (see, e.g., U.S. Patent Publication No. 2019 / 0359712). The anti-MSLN scFv domain and Fab domain sequences were generated from the V domain of the mouse anti-MSLN SS antibody. H and V L Sequence generation (Chowdhury et al., 1998, PNAS, 95:669-674), including additional humanizing mutations (see, e.g., U.S. Pat. No. 9,388,222).

[0462] In certain bivalent bispecific antibody constructs of the present disclosure containing a C3E6 Fab domain (e.g., v21815 comprising an anti-CD3 Fab domain and an anti-MSLN scFv domain, see, e.g. Figure 1C), respectively, V H and V L Sequence similar to human IgG1 C H and C L In other bivalent constructs disclosed herein, the C3E6 scFv domain is fused to one chain of the heterodimeric Fc domain. In bivalent constructs containing SS Fab domains, SS V domains containing additional humanizing mutations (see, e.g., U.S. Pat. No. 9,388,222) are fused to one chain of the heterodimeric Fc domain. H and V L The domain sequence (Chowdhury et al., 1998, PNAS, 95:669-674) is similar to that of human IgG1 C H and C L Sequence fusion. In other bivalent constructs, the SS scFv domain is fused to one chain of the heterodimeric Fc domain (e.g., the first Fc polypeptide or the second Fc polypeptide). In all embodiments, the scFv domain is fused to one chain of the heterodimeric Fc domain, and the VH-CH domain of the Fab is fused to the second chain of the heterodimeric Fc domain.

[0463] In the trivalent bispecific antibody constructs of the present disclosure comprising two scFv domains targeting MSLN and one Fab domain targeting CD3 (e.g., v21812 and v32523), the anti-CD3 C3E6 Fab domains are synthesized by respectively placing V H and V L Sequence similar to human IgG1 C H and C L The V containing SS with additional humanized mutations (see, e.g., U.S. Pat. No. 9,388,222) was generated by fusion of the V containing SS. H and V L The anti-MSLN scFv domain of the sequence (Chowdhury et al., 1998, PNAS, 95:669-674) was fused to (i) the N-terminus of the heterodimeric Fc domain (ie, to the N-terminus of the first Fc polypeptide or the second Fc polypeptide) or (ii) the N-terminus of the anti-CD3C3E6 Fab domain.

[0464] In the trivalent bispecific antibody construct of the present disclosure comprising two Fab domains targeting MSLN and one Fab targeting CD3 (e.g., v21791 and 2+1 Fab 3In TCB), preferential pairing of light chains with specific heavy chains is used. This method is further described in WO2014082179, using WO2015181805 to design heavy chain-light chain heterodimer pairs (i.e., H1-L1 and H2-L2) that exhibit the selectivity and preferential pairing of the corresponding heavy and light chains to produce the desired heterodimers. The heterodimers are designed in pairs, including the C L domain and the C H1 When the immunoglobulin heavy chain is co-expressed with two immunoglobulin light chains, these modifications ensure that the MSLN light chain is bound to the MSLNV H -C H1 and the preferential pairing of CD3 light chain with CD3 VH-CH1. H and V L Sequence similar to human IgG1 C H and C L The anti-CD3 C3E6 Fab domain was generated by fusion of MSLN V H -C H1 The domain is attached to (i) the N-terminus of the heterodimeric Fc domain (i.e., to the N-terminus of the first Fc polypeptide or the second Fc polypeptide) or (ii) the anti-CD3 C3E6 Fab V H N-terminal fusion of the domain.

[0465] The heavy chain domain structures and formats (from N-terminus to C-terminus) used to generate certain antibody constructs of the present disclosure (e.g., trivalent bispecific constructs), as well as other constructs used to study the effects of overall construct format and geometry, are shown in Table 5 below.

[0466] Table 5: Heavy chain domain structures used in the antibody constructs herein

[0467] Heavy chain domain structure: An exemplary antibody construct that may comprise this chain format: <![CDATA[V H -V L -V H -C H1 -Hinge-CH2-CH3]]> v21812 <![CDATA[V H -C H1 -Hinge-CH2-CH3]]> <![CDATA[v18490、v21791、2+1Fab 3 TCB(v29191)]]> <![CDATA[V H -V L -Hinge-CH2-CH3]]> v21812 <![CDATA[V H -C H1 -V H -C H1 -Hinge-CH2-CH3]]> <![CDATA[v21791、v29191(2+1Fab 3 Benchmark)]]> <![CDATA[V L -V H -V H -C H1 -Hinge-CH2-CH3]]> v32523 <![CDATA[V L -V H -Hinge-CH2-CH3]]> v32523 <![CDATA[V HH -V HH -V H -V L ...

Claims

1. An antibody construct comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN), and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, in: The Fab domain is coupled to the N-terminus of the first Fc polypeptide, The first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and The second scFv domain is coupled to the N-terminus of the Fab domain.

2. The antibody construct of claim 1, wherein at least one of the first scFv domain and the second scFv domain comprises (i) V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and (ii) V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO:

125.

3. The antibody construct of claim 1 or claim 2, wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO:

125.

4. An antibody construct comprising: (i) a Fab domain capable of binding to CD3 on a cytotoxic effector cell, wherein the Fab domain comprises a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 126, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 127, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 128, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 129, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 130, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 131; (ii) a first scFv domain (scFv1) and a second scFv domain (scFv2), wherein the first scFv domain and the second scFv domain are capable of binding to mesothelin (MSLN), and wherein the first scFv domain and the second scFv domain each comprise a V H A domain comprising: a HCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 120, a HCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 121, and a HCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 122, and V L A domain comprising: a LCDR1 sequence comprising or consisting of the sequence shown in SEQ ID NO: 123, a LCDR2 sequence comprising or consisting of the sequence shown in SEQ ID NO: 124, and a LCDR3 sequence comprising or consisting of the sequence shown in SEQ ID NO: 125; and (iii) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, in: The Fab domain is coupled to the N-terminus of the first Fc polypeptide, The first scFv domain is coupled to the N-terminus of the second Fc polypeptide, and The second scFv domain is coupled to the N-terminus of the Fab domain.

5. The antibody construct of any one of claims 1 to 4, wherein the first scFv domain or the second scFv domain comprises (i) V H A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 103, and (ii) V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:

101.

6. The antibody construct of any one of claims 1 to 5, wherein the first scFv domain and the second scFv domain each comprise (i) V H A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 103, and (ii) V L A domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:

101.

7. The antibody construct of any one of claims 1 to 6, wherein the first scFv domain and the second scFv domain each comprise (i) V H A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO: 103, and (ii) V L A domain comprising or consisting of the amino acid sequence shown in SEQ ID NO:

101.

8. The antibody construct of any one of claims 5-7, wherein the first scFv domain and the second scFv domain further comprise L Domain and the V H Domain-coupled linkers scFv .

9. The antibody construct of claim 8, wherein the linker scFv Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of the amino acid sequence.

10. The antibody construct of any one of claims 8-9, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. H -Connectors scFv -V L or V L -Connectors scFv -V H domain structure.

11. The antibody construct of claim 10, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. H -Connectors scFv -V L The domain structure of the structure.

12. The antibody construct of claim 10, wherein the first scFv domain and the second scFv domain have a V from N-terminus to C-terminus. L -Connectors scFv -V H domain structure.

13. The antibody construct of any one of claims 1 to 12, wherein the Fab domain comprises (i) V H A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 105; (ii) C H1 A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 106; (iii) V L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO: 115; and (iv) C L A domain comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence shown in SEQ ID NO:

116.

14. The antibody construct of any one of claims 1 to 13, wherein the antibody construct comprises a first heavy chain (H1), the first heavy chain comprising scFv2-V from N-terminus to C-terminus. H -C H1 wherein the C-terminus of the second scFv domain (scFv2) is adjacent to the V H N-terminal coupling of the domain.

15. The antibody construct of claim 14, wherein the C-terminus of scFv2 is connected via a linker scFv-Fab With the V H The N-terminus of the domain is coupled.

16. The antibody construct of claim 15, wherein the linker scFv-Fab Contains amino acid sequence (G4S) n , wherein n is 1, 2, 3, 4 or 5, or consists of the amino acid sequence.

17. The antibody construct of any one of claims 14-16, wherein H1 further comprises the first Fc polypeptide, the first Fc polypeptide comprising a first CH2 sequence (CH21) and a first CH3 sequence (CH31), and wherein the CH H1 The C-terminus of the H1 domain is coupled to the N-terminus of the first Fc polypeptide, thereby forming the following H1 domain structure (from N-terminus to C-terminus): scFv2-V H -C H1 -CH21-CH31.

18. The antibody construct of claim 17, wherein the C H1 Linker Fab-Fc coupled to the CH2 sequence of the first Fc polypeptide.

19. The antibody construct of claim 18, wherein the linker Fab-Fc It comprises or consists of an Ig hinge region.

20. The antibody construct of any one of claims 14-19, 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 shown in SEQ ID NO:

100.

21. An antibody construct as described in any one of claims 1 to 20, wherein the antibody construct further comprises a second heavy chain (H2), the second heavy chain comprising, from N-terminus to C-terminus, the first scFv domain (scFv1) coupled to the second Fc polypeptide, the second Fc polypeptide comprising a second CH2 sequence (CH22) and a second CH3 sequence (CH32), thereby forming a domain structure: scFv1-CH22-CH32.

22. The antibody construct of claim 21, wherein the C-terminus of scFv1 is connected via a linker scFv-Fc coupled to the N-terminus of the second Fc polypeptide.

23. The antibody construct of claim 22, wherein the linker scFv-Fc It comprises or consists of an Ig hinge region.

24. The antibody construct of any one of claims 21-23, 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 shown in SEQ ID NO:

110.

25. The antibody construct of any one of claims 1-24, wherein the first Fc polypeptide, the second Fc polypeptide, or both the first Fc polypeptide and the second Fc polypeptide are IgGl derived Fc polypeptides.

26. The antibody construct of any one of claims 1-25, wherein the dimeric Fc domain is a heterodimeric Fc domain, wherein 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.

27. The antibody construct of claim 26, 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 a corresponding wild-type IgG1 CH3 domain sequence, wherein the one or more amino acid substitutions promote preferential pairing of the first Fc polypeptide and the second Fc polypeptide to form the heterodimeric Fc domain.

28. The antibody construct of claim 27, wherein one Fc polypeptide comprises a set of amino acid substitutions selected from the group consisting of 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 the group consisting of 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 the first Fc polypeptide and the second Fc polypeptide is according to the EU numbering system.

29. The antibody construct of any one of claims 17-28, 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 a corresponding wild-type IgG1 CH2 domain sequence, wherein the one or more amino acid substitutions reduce or eliminate binding of the antibody construct to an Fcγ-receptor.

30. An antibody construct as described in any one of claims 1 to 29, 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 shown 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 shown 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 shown in SEQ ID NO:

114.

31. The antibody construct of any one of claims 14-30, wherein one or more of the heavy chains H1 and H2 further comprises a C-terminal lysine residue.

32. An antibody construct comprising: a first heavy chain (H1) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 100, a second heavy chain (H2) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO: 110, and a light chain (L1) polypeptide comprising or consisting of the amino acid sequence shown in SEQ ID NO:

114.

33. The antibody construct of any one of claims 1-32, wherein the antibody construct, when 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 / cell, is capable of achieving one or more, two or more, or all of the following: A) induces production of at least about 20 pg / mL, 30 pg / mL, or 40 pg / mL TNFα; B) induces the production of at least about 500 pg / mL, 1000 pg / mL, or 2000 pg / mL IFNγ; or C) induces the production of at least about 50 pg / mL, 1000 pg / mL, or 150 pg / mL IL-2.

34. The antibody construct of any one of claims 1-33, wherein the binding affinity (K) of the antibody construct to MSLN is measured by SPR. D ) is about 0.7 nM, 0.8 nM, 0.9 nM, 1 nM or about 1.1 nM.

35. The antibody construct of any one of claims 1-34, wherein the binding affinity (K) of the antibody construct to CD3 is measured by SPR. D ) is about 30 nM, 40 nM, 50 nM or about 60 nM.

36. An antibody construct as described in any one of claims 1-35, wherein after incubation in buffer solution at 40°C for 14 days, the purity of the antibody construct is at least about 95%, 96%, 97%, 98% or at least about 99% as measured by size exclusion chromatography and relative to other high molecular weight species (HMWS) present in the buffer solution after the incubation period.

37. A pharmaceutical composition comprising the antibody construct of any one of claims 1-36, and a pharmaceutically acceptable carrier, excipient, diluent or a combination thereof.

38. A nucleic acid molecule or collection 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 claims 1-36.

39. A vector or a vector collection comprising the nucleic acid molecule or nucleic acid molecule collection of claim 38.

40. A cell comprising the nucleic acid molecule or collection of nucleic acid molecules according to claim 38, or the vector or collection of vectors according to claim 39.

41. A method for producing an antibody construct as described in any one of claims 1 to 36, the method 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, or all of the polypeptide chains forming the antibody construct; and (b) recovering the antibody construct from the host cell culture.

42. The antibody construct of any one of claims 1-36 for use in cancer therapy.

43. Use of an antibody construct as described in any one of claims 1 to 36 in the manufacture of a medicament for the treatment of cancer.

Images