Trispecific antibody constructs targeting DLL3 and methods of use thereof

Abstract

The present disclosure describes costimulatory trivalent and trispecific antibody constructs capable of binding two different antigens on one or more cytotoxic effector cell(s) and a TAA (such as DLL3) on a tumor cell. Pharmaceutical compositions comprising such antibody constructs and methods of preparing and using such constructs and compositions, e.g., for the treatment of cancer, are also disclosed.

Description

TRISPECIFIC ANTIBODY CONSTRUCTS TARGETING DLL3 AND METHODS OF USE THEREOFCROSS-REFERENCE

[0001] This application claims priority to and the benefit of U.S. Provisional Application No. 63 / 730,303, filed December 10, 2024, U.S. Provisional Application No. 63 / 685,886, filed August 22, 2024, U.S. Provisional Application No. 63 / 575,343, filed April 5, 2024, and U.S. Provisional Application No. 63 / 560,940, filed March 4, 2024, the entire contents of which are incorporated by reference herein for all purposes.TECHNICAL FIELD

[0002] The present disclosure generally relates to trispecific T cell engaging antibody constructs that can comprise a first binding domain capable of binding a first antigen on a cytotoxic effector cell, a second binding domain capable of binding a second antigen on a cytotoxic effector cell, and a third binding domain capable of binding a tumor-associated antigen (TAA), such as Delta-like Ligand 3 (also known and referred to herein as DLL3, or delta-like canonical Notch ligand 3), on a tumor cell. The disclosure also relates to pharmaceutical compositions comprising such constructs, as well as methods of using such constructs for the treatment of a disease such as cancer.BACKGROUND

[0003] Cancer continues to pose a major unmet medical need, despite the considerable progress that has been made in its treatment over the past decades. While the current standard of care, as well as more recently developed anti-cancer therapies, have shown some clinical progress, various indications such as those with low T cell tumor infiltration still present a major clinical challenge.

[0004] Small cell lung cancer (SCLC), for instance, is an aggressive neuroendocrine cancer with a poor prognosis and high unmet medical need. Current standards of care, including chemotherapy, targeted therapy and immunotherapy, are associated with limited duration of response. Furthermore, SCLC is characterized by an immunosuppressive microenvironment and poor T cell infiltration which present additional treatment challenges.

[0005] DLL3 is an inhibitory Notch ligand that is aberrantly overexpressed in SCLC and other neuroendocrine tumors, but minimally expressed in normal tissues, and thus can be potentially exploited as a therapeutic target.SUMMARY

[0006] In various embodiments, the present disclosure describes trispecific and trivalent antibody constructs that are capable of engaging two different antigens on one or more immune cell(s) (e.g., T cell(s)), as well as a TAA on a tumor cell. In certain embodiments of the present disclosure, described are trivalent and trispecific antibody constructs that comprise three binding domains, wherein a first binding domain is capable of binding a first antigen on a cytotoxic effector cell, a second binding domain is capable of binding a second antigen on a cytotoxic effector cell, and a third binding domain is capable of binding a TAA on a tumor cell.

[0007] In some embodiments, described herein is an antibody construct, comprising: (i) a first fragment antigen binding (Fab) domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first single-chain variable fragment (scFv) domain capable of binding a second antigen on a cytotoxic effector cell; (iii) a third binding domain capable of binding DLL3 on a tumor cell, wherein the third binding domain is a second Fab domain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to (1) the C-terminus of the light chain of either the first Fab domain or the second Fab domain, if the third binding domain is the second Fab domain, or (2) the C-terminus of the light chain of the first Fab domain if the third binding domain is the second scFv domain.

[0008] In some embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding a TAA on a tumor cell; (iii) an scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of a Fab light chain.

[0009] In some embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domaincapable of binding DLL3 on a tumor cell; (iii) an scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of a Fab light chain.

[0010] In some embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding cluster of differentiation 3 (CD3, e.g., CD3s) on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) an scFv domain capable of binding cluster of differentiation 28 (CD28) on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the Fab light chain of the first Fab domain.

[0011] In some embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a third binding domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain. In some embodiments, the third binding domain is a second Fab domain. In other embodiments, the third binding domain is a second scFv domain.

[0012] In some embodiments, described herein is an antibody construct, comprising (from N- to C-terminus): (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 Fab heavy chain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer H1-H2; (iii) a first light chain (LI), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acidsequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein LI and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain; and (iv) a second light chain (L2), comprising an anti-DLL3 Fab light chain, wherein L2 and the anti-DLL3 Fab heavy chain of H2 associate to form an anti-DLL3 Fab domain.

[0013] In some embodiments, described herein is an antibody construct, comprising: (i) a Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; 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 C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide.

[0014] In some embodiments, described herein is an antibody construct, comprising: (i) a Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; 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 C-terminus of the Fab light chain N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide.

[0015] In some embodiments, described herein is an antibody construct, comprising: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 scFv domain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer H1-H2; and (iii) a light chain (L), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein L and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain.

[0016] In some embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a third binding domain capable of binding a TAA on a tumor cell, wherein the third binding domain comprises or consists of a second Fab domain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0017] Further described herein are pharmaceutical compositions comprising one or more of the trivalent and trispecific DLL3 -targeting antibody constructs of the present disclosure, nucleic acid molecule(s) encoding the one or more (e.g., two, three, four, or more) polypeptide chains of the antibody constructs described herein, as well as methods of producing and using such antibody constructs, e.g., for the treatment of DLL3 -positive cancers.

[0018] In some embodiments, the disclosure relates to a trivalent and trispecific antibody construct of the present disclosure for use in the treatment of cancer.

[0019] In some embodiments, the disclosure relates to a trivalent and trispecific antibody construct of the present disclosure for use in the manufacture of a medicament for the treatment of cancer.

[0020] In various embodiments, the cancer in the treatment of which a trivalent and trispecific antibody construct of the present disclosure can be used is a DLL3-positive cancer.BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The implementations disclosed herein are illustrated by way of example, and not by way of limitation, in the accompanying drawings. The description and drawings are only for the purpose of illustration and as an aid to understanding and are not intended as a definition of the limits of the antibody constructs, pharmaceutical compositions, and methods of the present disclosure.

[0022] FIGS. 1A-1D show schematic representations of the format and geometry of trivalent and trispecific antibody constructs that comprise binding domains capable of engaging CD3, CD28, and DLL3, according to embodiments of the present disclosure. FIGS. 1E-1F show schematic representations of bispecific control constructs used as described in this disclosure. “A” and “B” refer to two different heavy chains (can include homodimeric and heterodimeric Fc domains) andthe dark and light shadings represent heavy and light chains, respectively, according to certain embodiments of the present disclosure.

[0023] FIGS. 1G-1H show schematic representations of the format and geometry of trivalent and trispecific antibody constructs that comprise binding domains capable of engaging CD3 (e.g., via a Fab domain) and CD28 (e.g., via an scFv domain) on a cytotoxic effector cell, and a TAA on a tumor cell (e.g., via a Fab or an scFv domain), according to certain embodiments of the present disclosure.

[0024] FIGS. 2A-2C show dose-response curves of primary human T cells in co-culture with DLL3+ NCI-H82-RFP cells as well as primary human Pan-T cells in co-culture with DLL3+ NCI- H82-RFP cells following treatment with the respective anti-(DLL3xCD3xCD28) trivalent and trispecific antibody constructs and bispecific controls, respectively.

[0025] FIGS. 3A-3B show binding curves for certain trivalent and trispecific antibody constructs and bispecific control constructs to (i) wildtype Jurkat cells (e.g., T cells expressing both CD3 and CD28), (ii) CD28 knock-out (KO) Jurkat cells, and (iii) CD3 knock-out (KO) Jurkat cells, respectively.

[0026] FIG. 4A shows concentration response curves and in vitro cytotoxicities for certain tested constructs in various DLL3-positive cell lines in co-culture with human T cells. For this study, T cells were thawed at 37 °C and mixed with DLL3+ NCI-H82-RFP (top left), SHP-77-RFP (top right) or COR-L279-RFP (bottom) cells such that the ratio of T cells to tumor cells was adjusted to an effector to target cell (E:T) ratio of 1 :5, 1 :2 and 1 :2 for the H82, SHP-77 and COR-L279 cell lines, respectively. FIG. 4B shows concentration response curves and in vitro cytotoxicities for certain tested constructs in various DLL3-positive cell lines in co-culture with human T cells in a follow-on study in which human T cells were thawed at 37 °C and mixed with DLL3+ NCI-H82- RFP (top left), SHP-77-RFP (top middle), COR-L279-RFP (top right), NCI-H69-RFP (bottom left), NCI-H524-RFP (bottom middle) and NCI-H526-RFP (bottom right) cells such that the ratio of T cells to tumor cells was adjusted to an effector to target cell (E:T) ratio of 1 :5 for NCI-H82 and 1 :2 for the remaining cell lines.

[0027] FIGS. 5A-5B show T cell proliferation in the presence of certain antibody constructs. FIG. 5A shows T cell proliferation in the presence of the tested antibody constructs and DLL3+ NCI- H82 cells, and FIG. 5B shows T cell proliferation in the presence of the tested antibody constructs only (i.e., no DLL3+ cells present).

[0028] FIGS. 6A-6B show upregulation of Bcl-XL expression in activated T-cells using certain trivalent and trispecific antibody constructs of the present disclosure as well as bispecific and reference constructs.

[0029] FIG. 7 shows tumor growth delay observed in an in vivo study using donor A-engrafted mice treated with 2.5 mg / kg of the trivalent and trispecific antibody construct v38967 when compared to a clinical benchmark (arrows indicate time points of test article administration). FIG. 8 shows that percent survival in this in vivo study was significantly higher in mice receiving 2.5 mg / kg of v38967 compared to mice receiving equivalent doses of reference construct (Long-rank p = 0.0047). FIG. 9 shows that body weights of mice treated with v38967 over the course of this in vivo study remained stable.

[0030] FIG. 10 shows tumor regression observed in donor A-engrafted mice treated with the trivalent and trispecific antibody construct v39917 compared to the reference construct (v38589) and irrelevant mAb (i.e., control antibody against an irrelevant target, v22277; arrows indicate time points of test article administration). FIG. 11 and FIG. 12 show the percent survival and change in body weight, respectively, over the course of this in vivo study.

[0031] FIG. 13A depicts curves showing the percent of T cells proliferated with the respective titrated antibody constructs tested and in either the presence or absence of target cells. Corresponding histograms of proliferation at 200 pM are displayed in FIG. 13B.

[0032] FIG. 14 shows tumor regression observed in donor B-engrafted mice treated with v41086, CD28-KO (v41753), and the reference construct (v38589), compared to the irrelevant mAb (v40552). Treatment with 6.6 nmol / kg and 3.3 nmol / kg of v41086 resulted in full tumor regression in 7 out of 7 mice (arrows indicate time points of test article administration). Body weights remained stable for all mice as shown in FIG. 15.

[0033] FIG. 16 shows tumor growth inhibition as observed in donor-engrafted mice treated with 2.85 nmol / kg of the trivalent and trispecific antibody construct v41086 when compared to treatment with the reference construct (v38985) (arrows indicate time points of test article administration).

[0034] FIG. 17 shows serum cytokine (IFNy (or IFNg), IL-2, TNFa (or TNFa), and IL-6; in pg / mL) concentrations measured 72 hours post-treatment of humanized mice with the respective construct.

[0035] FIG. 18 shows serum concentrations of the measured cytokines IL-6, IL-10, MCP-1 and IL-2 in cynomolgus monkeys that were given two intravenous (i.v.) doses of 10 mg / kg of the respective tested constructs.DETAILED DESCRIPTION

[0036] In various embodiments, the present disclosure describes trivalent and trispecific antibody constructs capable of binding two different antigens on a cytotoxic effector cell (e.g., a T cell), and a TAA on a tumor cell.

[0037] In certain embodiments, the two different effector cell antigens to which a trivalent and trispecific antibody construct of the present disclosure binds are located on the same cytotoxic effector cell (e.g., a T cell).

[0038] In other embodiments, the two different effector cell antigens to which a trivalent and trispecific antibody construct of the present disclosure binds are located on different cytotoxic effector cells.

[0039] In embodiments in which the two different effector cell antigens to which a trivalent and trispecific antibody construct binds are located on the same cytotoxic effector cell (e.g., a T cell), such antibody construct comprises an anti-CD28 scFv domain linked to the C-terminus of an anti- CD3 Fab light chain (e.g., as depicted in FIGs. IB, ID, 1G and 1H) of an anti-CD3 Fab domain and can cross-link the two effector cell antigens (e.g., CD3 epsilon (CD3e) and CD28) on the same cell, thereby stimulating a specific effector cell response, e.g., an anti-tumor response, in the presence of tumor cells expressing a TAA (e.g., DLL3). Such anti-tumor response may be differentiated from effector cell responses elicited by antibody constructs that merely target one effector cell antigen (e.g., CD3 or CD28, but not both, such as a bispecific anti-(CD3xTAA) or anti-(CD28xTAA) antibody construct), and / or it can be differentiated from a response elicited by trivalent and trispecific antibody constructs in which the anti-CD3 and anti-CD28 antigen binding domains are in a different configuration and proximity to each other, e.g., constructs that have formats as depicted in FIG. 1A or FIG. 1C in which the anti-CD28 scFv domain is not linked to the C-terminus of the anti-CD3 Fab light chain.

[0040] In various embodiments, the trivalent and trispecific (e.g., Het-Fab and scFv2) antibody constructs described herein bind CD3 and CD28 on the same cytotoxic effector cell (e.g., T cell) and not - or at least to a significantly lower degree - on two different effector cells, thereby eliciting no (or reduced) cross-linking of two cytotoxic effector cells compared to trivalent andtrispecific antibody construct with a different format and / or geometry, e.g., constructs in which the anti-CD28 scFv and anti-CD3 Fab domains are in a different configuration to each other, in which the anti-CD28 scFv domain is attached to either the N-terminus of the anti-CD3 Fab domain (e.g., as depicted in FIG. 1A) or a C-terminus of the Fc domain (e.g., as depicted in FIG. 1C).

[0041] In various embodiments, the TAA that is targeted by a trivalent and trispecific antibody construct of this disclosure is DLL3, and thus, in various embodiments, the tumor cells that are targeted by the constructs described herein comprise DLL3 -expressing tumor cells which can have the same or different tissue(s) of origin.

[0042] A trivalent and trispecific antibody construct of the present disclosure can comprise a first binding domain capable of binding a first antigen on a cytotoxic effector cell, a second binding domain capable of binding a second antigen on a cytotoxic effector cell, and a third binding domain capable of binding DLL3 on a tumor cell. In certain embodiments, described herein are trivalent and trispecific T cell engaging antibody constructs capable of co-stimulating an effector cell (e.g., a T cell) by engaging two effector cell antigens (e.g., CD3 and CD28) on the same effector cell while being capable of binding DLL3 on a tumor cell.

[0043] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises (i) a first Fab domain (i.e., the first binding domain) that is capable of binding a first antigen on a cytotoxic effector cell, (ii) a first scFv domain (i.e., the second binding domain) that is capable of binding a second antigen on a cytotoxic effector cell, and (iii) a third binding domain capable of binding DLL3, wherein the third binding domain can comprise or consist of a second Fab domain or a second scFv domain.

[0044] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises (i) a first Fab domain (i.e., the first binding domain) that is capable of binding a first antigen on a cytotoxic effector cell, (ii) a second Fab domain (i.e., the third binding domain) capable of binding DLL3 on a tumor cell, and (iii) a first scFv domain (i.e., the second binding domain) that is capable of binding a second antigen on a cytotoxic effector cell.

[0045] In other embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises (i) a first Fab domain (i.e., the first binding domain) that is capable of binding a first antigen on a cytotoxic effector cell, (ii) a first scFv domain (i.e., the second binding domain) that is capable of binding a second antigen on a cytotoxic effector cell; and (iii) a second scFv domain (i.e., the third binding domain) that is capable binding DLL3 on a tumor cell.

[0046] In certain embodiments, the first and second antigens on the cytotoxic effector cell are independently selected from CD3 and CD28. In certain embodiments, the first antigen on the cytotoxic effector cell (e.g., engaged via a Fab domain) is CD3 and the second antigen on the cytotoxic effector cell (e.g., engaged via an scFv domain) is CD28.

[0047] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises (i) a first Fab domain (i.e., the first binding domain) that is capable of binding CD3 on a cytotoxic effector cell, (ii) a first scFv domain (i.e., the second binding domain) capable of binding CD28 on a cytotoxic effector cell, and (iii) a third binding domain capable of binding DLL3, wherein the third binding domain can comprise or consist of a second Fab domain or a second scFv domain, and wherein the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0048] Any of the trivalent and trispecific antibody constructs described herein can further comprise an Fc domain (e.g., a homodimeric or a heterodimeric Fc domain) comprising a first Fc polypeptide and a second Fc polypeptide and to which the various binding domains are coupled, either directly (e.g., without a linker) or indirectly, e.g., via a linker and / or via another binding domain (e.g., in instances in which a first binding domain is coupled indirectly to an Fc polypeptide via a second binding domain, e.g., a C-terminus of the first binding domain is coupled to an N- terminus of the second binding domain, and such second binding domain is in turn coupled to the N-terminus of the Fc polypeptide, thereby indirectly coupling the first binding domain to the Fc polypeptide).

[0049] As further described herein, FIGS. 1A-1D and FIGS. 1G-1H depict certain trivalent and trispecific antibody construct formats and geometries with different relative orientations of the three binding domains, according to certain embodiments of this disclosure.

[0050] Further described herein are pharmaceutical compositions comprising one or more of the trivalent and trispecific antibody constructs disclosed herein, and a pharmaceutically acceptable carrier or excipient.

[0051] Other embodiments of this disclosure relate to a nucleic acid molecule, or a set of nucleic acid molecules, that encode the one or more (e.g., 2, 3, 4 or more) polypeptide chains (e.g., Hl, H2, LI, L2, etc.) of an antibody construct described herein. Certain other embodiments of this disclosure can relate to a vector, or a set of vectors, that comprise a nucleic acid molecule or a set of nucleic acid molecules encoding an antibody construct of the present disclosure. Host cellscomprising such nucleic acid molecule, or set of nucleic acid molecules, and / or such vector, or set of vectors, for expressing a trivalent and trispecific antibody construct are also described herein.

[0052] Other embodiments of this disclosure describe methods of producing and using the trivalent and trispecific antibody constructs described herein, e.g., for the treatment of a cancer in a subject (such as a rodent or a human).I. DEFINITIONS

[0053] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.

[0054] The term “about,” as used herein in the context of a numerical value or range, generally refers to ±10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, or ±1% of the numerical value or range recited or claimed, unless otherwise specified. In various embodiments, the term “about” refers to an approximately ±10% variation from a given value or range. In other embodiments, the term “about” refers to an approximately ±5% variation from a given value or range. In yet other embodiments, the term “about” refers to an approximately ±1% variation from a given value or range. Unless otherwise recited, it is to be understood that ±10% variation is always included in any given value provided herein, whether it is specifically referred to or not.

[0055] The use of the word “a” or “an,” when used herein in conjunction with the term “comprising,” can mean “one,” but it is also consistent with the meaning of “one or more,” “at least one” and “one or more than one.”

[0056] As used herein, the terms “comprising,” “having,” “including” and “containing,” and grammatical variations thereof, are inclusive or open-ended and do not exclude additional, unrecited elements and / or method steps. The term “consisting essentially of’ when used herein in connection with a construct, composition, use or method, denotes that additional features, elements and / or method steps can be present, but that these additions do not materially affect the manner in which the recited construct, composition, method or use functions. The term “consisting of,” when used herein in connection with a construct, composition, use or method, excludes the presence of additional elements and / or method steps. An antibody construct, composition, use, or method described herein as comprising certain elements and / or steps can also, in certain embodiments, consist essentially of those elements and / or steps, and in other embodiments, consist of those elements and / or steps, whether or not these embodiments are specifically referred to.

[0057] The terms “subject” and “patient” can be used interchangeably herein and generally refer to an animal in need of treatment. An animal in need of treatment can be a human or a non-human animal, such as a mammal, bird, or fish. In certain embodiments, the subject is a mammal. In some embodiments, the subject is a human. In other embodiments, the subject is a rodent or a non-human primate.

[0058] An “effective amount” of a trivalent and trispecific antibody construct described herein, or a pharmaceutical composition comprising such antibody construct, in respect of a particular result to be achieved is an amount sufficient to achieve the desired result. For example, an “effective amount” of an antibody construct or pharmaceutical composition when referred to in respect of the killing of cancer cells, refers to an amount of antibody construct or composition comprising the antibody construct sufficient to produce a killing effect.

[0059] Unless specified otherwise, the terms “Fc region,” “Fc” and “Fc domain” are used interchangeably herein and refer to a C-terminal region of an immunoglobulin (Ig) heavy chain that contains at least a portion of a constant region. In various embodiments, an Fc domain herein can be dimeric. Such dimeric Fc domain can comprise a first Fc polypeptide and a second Fc polypeptide, wherein each Fc polypeptide can comprise a CH2 domain and a CH3 domain. Such dimeric Fc can either be homodimeric, i.e., comprising first and second Fc polypeptides that have identical amino acid sequences, or heterodimeric, i.e., comprising first and second Fc polypeptides that have different amino acid sequences, e.g., sequences that share about 95%, 96%, 97%, 98%, or about 99% sequence identity. In some embodiments, an antibody construct of the present disclosure comprises a homodimeric Fc domain. In yet other embodiments, and as further described herein, an antibody construct comprises a heterodimeric Fc domain in which at least one of the CH2 and / or CH3 domains of the first and second Fc polypeptides have amino acid sequences that share about 99%, 98%, or about 97% sequence identity.

[0060] The term “multispecific,” as used herein in the context of an antibody construct, refers to a biologically functional protein (e.g., an antibody construct as described herein) which is “at least bispecific,” i.e., it comprises at least a first binding domain and a second binding domain, wherein such first and second binding domain can bind specifically two distinct epitopes, e.g., a first epitope and a second epitope. Such first and second epitopes can be located on the same antigen or on different antigens, e.g., a first epitope on cluster of differentiation 3 (CD3) or cluster of differentiation 28 (CD28) and a second epitope on a TAA. Accordingly, in some embodiments,antibody constructs according to the present disclosure can comprise specificities for at least two different antigens or for at least three different antigens or targets. Hence, the term “multispecific” in the context of an antibody construct herein encompasses antibody constructs that are at least bispecific (i.e., comprising two binding domains with specificities for two different antigens or targets), or at least trispecific (i.e., comprising three binding domains with specificities for three different antigens or targets, e.g., CD3, CD28, and a TAA such as DLL3).

[0061] In various embodiments, an antibody construct of the present disclosure that is capable of binding CD3 refers to a construct that is able to bind an epitope on the epsilon chain of CD3, also referred to herein as “CD3s,” “CD3e” or “CD3 epsilon.”

[0062] The term “trispecific,” as used herein in the context of an antibody construct, refers to a biologically functional protein (e.g., an antibody construct as described herein) which is “at least trispecific,” i.e., it comprises at least a first binding domain, a second binding domain and a third binding domain, wherein such first, second and third binding domains can bind specifically three distinct epitopes, e.g., a first epitope, a second epitope and a third epitope. Such first, second and third epitopes can be located on the same antigen or on different antigens, e.g., a first epitope on CD3, a second epitope on CD28, and a third epitope on a TAA (e.g., DLL3). Accordingly, in some embodiments, trispecific antibody constructs according to the present disclosure can comprise specificities for at least three different antigens or targets. Hence, the specificity in the context of an antibody construct herein describes the total number of different epitopes and / or antigens an antibody construct can specifically bind to, e.g., a monospecific antibody construct comprises one or more binding domain(s) with a specificity for one epitope or antigen, a trispecific antibody construct comprises three or more binding domains with specificities for three different epitopes and / or antigens, and so forth.

[0063] The term “trivalent,” as used herein in the context of an antibody construct, refers to a biologically functional protein (e.g., an antibody construct as described herein) which is “at least trivalent,” i.e., it comprises at least 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, second, and third binding domains is capable of specifically binding an epitope and / or antigen, e.g., CD3, CD28, and / or a TAA. The three binding domains can either have specificities for three different epitopes or antigens, or two or more of the three binding domains have a specificity for the same epitope or antigen. Hence, the valency in the context of an antibody construct herein, e.g., being mono-, bi-,or trivalent, describes the total number of antigen binding domains of an antibody construct. Accordingly, the valency of an antibody construct is generally at least equal to its specificity, i.e., a trispecific antibody construct has to be at least trivalent. In embodiments in which the antibody construct is trivalent and trispecific, each of the three binding domains of the construct is capable of binding a different epitope or antigen, and thus the construct engages each of the three epitopes and / or antigens monovalently.

[0064] As used herein, the term “binding domain” in the context of an antibody construct described herein can generally be used interchangeably with the term “antigen binding domain,” unless the binding domain has been specified to bind to a moiety other than an antigen on the surface of a target cell.

[0065] As used herein, the term “format” in the context of an antibody construct described herein generally describes attributes of the antibody construct including its antigen valency (e.g., a construct being mono- or bivalent for a given antigen), the type(s) of binding domain(s) (e.g., possessing one or more scFv domain(s), one or more Fab domain(s), etc.) present in an antibody construct, as well as the presence, absence, and / or type of an Fc domain (e.g., homodimeric, heterodimeric, containing one or more constant heavy domains, CH2, CH3, etc.). As an example, in some embodiments, an antibody construct of the present disclosure can be trivalent and trispecific in a 1+1+1 format illustrating that the construct contains three binding domains, wherein each binding domain has an affinity for a different antigen (e.g., CD3, CD28 and a TAA), i.e., the trivalent construct is monovalent (as indicated by “1”) for each of the three antigens.

[0066] As used herein, the term “geometry” in the context of an antibody construct described herein generally describes the overall (e.g., 3 -dimensional) structure of an antibody construct, including the relative spatial localization and arrangement and / or connectivity of the various domains of an antibody construct, e.g., the relative arrangement and connectivity of binding domains and Fc domains, as further described herein and as illustrated in, e.g., FIGS. 1A-1D and FIGS. 1G-1H, according to certain embodiments of the present disclosure.

[0067] Generally, and unless specified otherwise, an amino acid sequence of a polypeptide described herein is described and defined in the direction from N- to C-terminus. As an example, a polypeptide described as comprising an scFv domain coupled to an Fc polypeptide is defined herein as a polypeptide in which the C-terminus of the scFv domain is coupled, either with orwithout a linker, to the N-terminus of the Fc polypeptide and the domain structure can be described as: scFv-Fc, or with the inclusion of a linker as: scFv-LinkerscFv'Fc-Fc.

[0068] As used herein, abbreviations such as “Hl” and “H2,” or “A” and “B,” are generally used as generic heavy chain identifiers and broadly refer to a first heavy chain and a second heavy chain of an antibody construct, respectively, and thus are not intended to be limited to any specific heavy chain amino acid (or polynucleotide) sequence(s).

[0069] The term “amino acid modification,” as used herein in the context of an amino acid sequence of a polypeptide, generally refers to an amino acid sequence of a polypeptide in which one or more amino acid substitution(s), one or more amino acid insertion(s), and / or one or more amino acid deletion(s) have been introduced relative to a corresponding unmodified (e.g., WT or reference) amino acid sequence of the polypeptide.

[0070] Descriptions of antibody constructs such as “anti-(DLL3xCD28xCD3)” and “anti- DLL3 / anti-CD28 / anti-CD3” can be used interchangeably herein and generally refer to an antibody construct that is at least trispecific and thus contains at least three binding domains that are capable of binding an epitope on DLL3, an epitope on CD28 and an epitope on CD3, respectively. In various embodiments, such description refers to a trivalent and trispecific antibody construct having three binding domains, one capable of binding an epitope on DLL3, one an epitope on CD28 and one an epitope on CD3, respectively, e.g., as shown in FIGS. 1A-1D and FIGS. 1G-1H

[0071] The term “dissociation constant (KD or Ka)” as used herein, is intended to refer to the equilibrium dissociation constant of a particular ligand-protein interaction. As used herein, ligandprotein interactions refer to, but are not limited to protein-protein interactions or antibody-antigen interactions. The KD measures the propensity of two proteins complexed together (e.g. AB) to dissociate reversibly into constituent components (A+B), and is defined as the ratio of the rate of dissociation, also called the “off-rate (koff)”, to the association rate, or “on-rate (kon)”. Thus, KD equals koff / konand is expressed as a molar concentration (M). It follows that the smaller the KD, the stronger the affinity of binding, and thus a decrease in KD indicates an increase in affinity. Therefore, a KD of 1 mM indicates weak binding affinity compared to a KD of 1 nM. Affinity is sometimes measured in terms of a KA or Ka, which is the reciprocal of the KD or Ka. KD values for antigen-binding constructs can be determined using methods well established in the art. One method for determining the KD of an antigen-binding construct is by using surface plasmonresonance (SPR), typically using a biosensor system such as a Biacore® system. Isothermal titration calorimetry (ITC) is another method that can be used to measure KD.

[0072] Generally, it is to be understood that the positive recitation of a feature in one embodiment serves as a basis for excluding the feature in an alternative embodiment. In particular, where a list of options is presented for a given embodiment or claim, it is to be understood that one or more option can be deleted from the list and the shortened list can form an alternative embodiment, whether or not such an alternative embodiment is specifically referred to.

[0073] It is further contemplated that any embodiment discussed herein can be implemented with respect to any antibody construct, method, use, or composition disclosed herein, and vice versa. Furthermore, modifications of the specific embodiments described herein that would be apparent to those skilled in the art are intended to be included within the scope of the claims recited herein.II. ANTIBODY CONSTRUCTS

[0074] The present disclosure describes trivalent and trispecific T cell engaging antibody constructs comprising three binding domains capable of engaging two different antigens on a cytotoxic effector cell and a TAA on a tumor cell, and which can have varying formats and structures, e.g., as shown in FIGS. 1G-1H, according to certain embodiments of this disclosure.

[0075] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises: (i) a first binding domain capable of binding a first antigen on a cytotoxic effector cell, (ii) a second binding domain capable of binding a second antigen on a cytotoxic effector cell, (iii) a third binding domain capable of binding DLL3 on a tumor cell, and (iv) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0076] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell, (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell, (iii) a third binding domain capable of binding DLL3 on a tumor cell, and (iv) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, and wherein the first scFv domain is coupled to the C- terminus of the light chain of the first Fab domain, e.g., as shown in FIGS. IB, ID, 1G and 1H, according to certain embodiments of this disclosure.A. Structure of Antibody Constructs and Their Biological Targets

[0077] In various embodiments, the antibody constructs described herein can have various formats and geometries, e.g., the number and type of binding domains may vary, and the binding domainsand the Fc domain can be interconnected (e.g., covalently or non-covalently) in various configurations, as further described herein.

[0078] In certain embodiments, the antibody constructs described in the present disclosure are trivalent and trispecific and thus comprise at least three antigen binding domains, with each of the three binding domains being capable of binding a different antigen. In various embodiments, the trivalent and trispecific antibody constructs of the present disclosure can have a format described as “1+1+1”, indicating that each of the three binding domains binds a different antigen, and hence such constructs bind each antigen in a monovalent (“1”) manner. FIGS. 1A-1D depict the formats and structures of certain trivalent and trispecific antibody constructs according to embodiments of the present disclosure.

[0079] Described herein are trivalent and trispecific antibody constructs comprising one or more Fab domain(s) and one or more scFv domain(s) as binding domains, as well as an Fc domain, and, optionally, one or more linkers that couple (e.g., covalently couple) at least some of these domains to each other.

[0080] In various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a first Fab domain capable of binding a first antigen on a cytotoxic effector cell, a first scFv domain capable of binding a second antigen on a cytotoxic effector cell, a third binding domain capable of binding a TAA on a tumor cell, wherein the third binding domain can comprise or consist of a second Fab domain or a second scFv domain, and an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0081] In some embodiments, described herein are trivalent and tri specific antibody constructs that comprise two Fab domains and one scFv domain as binding domains. Such antibody constructs can also be referred to as trivalent and trispecific heterodimeric Fab (or “Het-Fab”) antibody constructs because each Fab domain binds a different antigen and as such contains different (i.e., antigen specific) variable domain sequences; optionally, such Het-Fab antibody construct can further contain one or more amino acid modifications in the CHI and / or CL sequences of each Fab domain to facilitate correct heavy and light chain pairing, as further described herein.

[0082] In some embodiments, described herein is a trivalent and trispecific Het-Fab antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding a TAA on a tumor cell; (iii) a first scFvdomain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of a Fab light chain. FIG. ID shows a schematic of such an antibody construct.

[0083] In various embodiments, the TAA is DLL3.

[0084] In some embodiments, described herein is a trivalent and trispecific Het-Fab antibody construct comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of a Fab light chain.

[0085] In some embodiments, the first antigen is CD3 or CD28. In some embodiments, the second antigen is CD3 or CD28. In some embodiments, the first antigen is CD3 or CD28 and the second antigen is CD3 or CD28, and wherein the first antigen and the second antigen are different antigens. In certain embodiments, the first antigen is CD3 (e.g., engaged via the first Fab domain) and the second antigen is CD28 (e.g., engaged via the first scFv domain).

[0086] In various embodiments, described herein is a trivalent and trispecific Het-Fab antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of a Fab light chain. In some embodiments, the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain, e.g., as depicted in FIG. ID.

[0087] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises or consists of one or more polypeptide chains. In some embodiments, a trivalent and trispecific Het-Fab antibody construct of the present disclosure comprises or consistsof four polypeptide chains. Such four polypeptide chains can comprise or consist of two immunoglobulin heavy chains (e.g., Hl and H2) and two immunoglobulin light chains (e.g., LI and L2).

[0088] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises or consists of a first immunoglobulin heavy chain (Hl) comprising at least a first Fab heavy chain, a second immunoglobulin heavy chain (H2) comprising at least a second Fab heavy chain, a first immunoglobulin light chain (LI) comprising at least a first Fab light chain and a second immunoglobulin light chain (L2) comprising at least a second Fab light chain.

[0089] In some embodiments of a trivalent and trispecific Het-Fab antibody construct that is capable of engaging CD3, CD28 and DLL3 and comprises four polypeptide chains Hl, H2, LI and L2, Hl comprises, from N- to C-terminus, a first Fab heavy chain comprising a first VH sequence and a first CHI sequence, coupled to a first Fc polypeptide comprising a first CH2 sequence and a first CH3 sequence, and wherein the first Fab heavy chain of Hl and a first Fab light chain of LI associate to form the first Fab domain that is capable of binding CD3. In some embodiments, Hl further comprises a first linkerFab'Fcthat couples the C-terminus of the first Fab heavy chain to the N-terminus of the first Fc polypeptide. In some embodiments, such trivalent and trispecific Het-Fab antibody construct comprises a H2 comprising, from N- to C-terminus, a second Fab heavy chain comprising a second VH sequence and a second CHI sequence, coupled to a second Fc polypeptide comprising a second CH2 sequence and a second CH3 sequence, wherein the second Fab heavy chain of H2 and a second Fab light chain of L2 associate to form the second Fab domain that is capable of binding DLL3. In some embodiments, H2 further comprises a second linkerFab'Fcthat couples the C-terminus of the second Fab heavy chain to the N-terminus of the second Fc polypeptide.

[0090] In some embodiments, the first linkerFab'Fcand the second linkerFab'Fcof a trivalent and trispecific Het-Fab antibody construct are polypeptide linkers, each independently comprising from about 5 to about 50, from about 5 to about 25 or from about 5 to about 15 consecutive amino acid residues. In some embodiments, the first linkerFab'Fcand the second linkerl ab-l care each derived independently from an IgGl, IgG2 or IgG4 hinge region.

[0091] In various embodiments of a Het-Fab antibody construct of the present disclosure, the first scFv domain can be coupled to the C-terminus of a Fab light chain, e.g., the C-terminus of LI or L2. In various embodiments, the first scFv domain is coupled to the C-terminus of the first Fablight chain, LI, such that the two binding domains that engage CD3 and CD28 on a cytotoxic effector cell are configurated as depicted in, e.g., FIGS. IB, ID, 1G and 1H.

[0092] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, LI comprises, from N- to C-terminus, the first Fab light chain comprising a first VL sequence and a first CL sequence, coupled to the first scFv domain comprising, from N- to C-terminus, either (i) a third VH sequence coupled to a third VL sequence or (ii) a third VL sequence coupled to a third VH sequence. In some embodiments, the first scFv domain comprises, from N- to C-terminus, the third VL sequence coupled to the third VH sequence. In some embodiments, LI further comprises a linkerFab'scFvthat couples the C-terminus of the first Fab light chain to the N-terminus of the first scFv domain (e.g., the N-terminus of the third VL sequence). In some embodiments, the linkerFab'scFvcomprises the amino acid sequence (G4S)X, wherein x is 1, 2, 3, 4 or 5 (SEQ ID NO: 119). In some of these embodiments, the linkerFab'scFvcomprises the amino acid sequence G4SG (SEQ ID NO: 58).

[0093] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, L2 comprises, from N- to C-terminus, the second Fab light chain comprising a second VL sequence and a second CL sequence.

[0094] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises: (i) a first Fab domain, wherein the first Fab domain is capable of binding CD3 on a cytotoxic effector cell; (ii) a second Fab domain, wherein the second Fab domain is capable of binding DLL3 on the tumor cell; (iii) a first scFv domain, wherein the first scFv domain is capable of binding CD28 on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising the first Fc polypeptide and the second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N- terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0095] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain is capable of binding CD3 and comprises a VH domain comprising a HCDR1 sequence comprising the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 sequence comprising the sequence ITSSGGRI (SEQ ID NO: 50) or ITRSGGRI (SEQ ID NO: 51), and a HCDR3 sequence comprising the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 sequence comprising the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 sequencecomprising the sequence RND, and a LCDR3 sequence comprising the sequence QSYSSGFI (SEQ ID NO: 55).

[0096] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain is capable of binding CD3 and comprises a VH domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 5.

[0097] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising a HCDR1 sequence comprising the sequence SYGVH (SEQ ID NO: 28), a HCDR2 sequence comprising the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 sequence comprising the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL sequence comprising a LCDR1 sequence comprising the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), aLCDR2 sequencecomprising the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 sequence comprising the sequence QQSRKVPFT (SEQ ID NO: 48).

[0098] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti- CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 16.

[0099] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fc polypeptide and the second Fc polypeptide each comprise a CH2 domain that can independently comprise an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 8. In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, 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 IgGl CH2 domain sequence, wherein the one or more amino acidsubstitutions reduce or eliminate binding of the antibody construct to an Fcy-receptor (FcyR). In some embodiments, the CH2 domain sequences of the first and second Fc polypeptides comprise the amino acid substitutions L234A L235A D265S, and wherein the numbering of amino acid residues in the first and second Fc polypeptides is according to the EU numbering system.

[0100] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, the dimeric Fc domain is a heterodimeric Fc domain in which one Fc polypeptide has an amino acid sequence that differs in at least one residue from the amino acid sequence of the other Fc polypeptide. In some embodiments, one of the Fc polypeptides comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 9, and the other Fc polypeptide comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 10.

[0101] In certain embodiments, a trivalent and trispecific Het-Fab antibody construct of the present disclosure comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 Fab heavy chain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein the first Fc polypeptide of Hl and the second Fc polypeptide of H2 associate to form a heterodimer H1-H2; (iii) a first light chain (LI), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13 or 107, and wherein the anti-CD3 Fab light chain of LI and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain; and (iv) a second light chain (L2), comprising an anti-DLL3 Fab light chain, and wherein the anti-DLL3 Fab light chain of L2 and the anti-DLL3 Fab heavy chain of H2 associate to form an anti-DLL3 Fab domain.

[0102] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, one or more of the Fab domains comprise one or more amino acid modifications (e.g., relative to unmodified reference Fab domain sequences) that promote correct heavy and light chain pairing to produce the correctly paired one or more Fab domains (e.g., association of the Fab portions ofHl with LI and H2 with L2), respectively, and relative to the formation of incorrectly paired Fab domains (e.g., association of the Fab portions of Hl with L2 and H2 with LI). Thus, in some embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain, the second Fab domain, or both Fab domains, comprise one or more amino acid modifications that promote correct heavy and light chain pairing to produce the first and second Fab domains, respectively, and relative to the formation of incorrectly paired Fab domains. In some embodiments, the one or more amino acid modifications comprise or consist of one or more amino acid substitutions.

[0103] Amino acid modifications that can promote heavy and light chain pairing are known in the art and are described in, e.g., International Patent Publication Nos. WO 2014 / 082179, WO 2015 / 181805, and WO 2017 / 059551, which are incorporated herein by reference in their entireties.

[0104] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain and the second Fab domain each comprise a kappa light chain. In other embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain and the second Fab domain each comprise a lambda light chain. In yet other embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain comprises a kappa light chain and the second Fab domain comprises a lambda light chain. In yet other embodiments of a trivalent and trispecific Het-Fab antibody construct, the first Fab domain comprises a lambda light chain and the second Fab domain comprises a kappa light chain.

[0105] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct in which one Fab domain comprises a kappa light chain and the other Fab domain comprises a lambda light chain, the Fab domain comprising the kappa light chain comprises one or more, or all of the amino acid substitutions 143E, 145T, 179E and 228D in the CHI sequence of the heavy chain and one or more, or all of the amino acid substitutions 121K, 124R and 178R in the CL sequence of the kappa light chain, and the other Fab domain comprising the lambda light chain comprises one or more, or all of the amino acid substitutions 125R and 188K in the CHI sequence of the heavy chain and one or more, or all of the amino acid substitutions 122D, 129T, 176E and 178E in the CL sequence of the lambda light chain, wherein the amino acid residues are identified according to the Kabat numbering system.

[0106] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct in which one Fab domain comprises a kappa light chain and the other Fab domain comprises a lambda lightchain, the Fab domain comprising the kappa light chain comprises one or more, or all of the amino acid substitutions 124R and 186K in the CHI sequence of the heavy chain and one or more, or all of the amino acid substitutions 133G, 135W, 176D and 180E in the CL sequence of the kappa light chain, and the other Fab domain comprising the lambda light chain comprises one or more, or all of the amino acid substitutions 139W, 143D and 145T in the CHI sequence of the heavy chain and one or more, or all of the amino acid substitutions 124Q and 131R in the CL sequence of the lambda light chain, wherein the amino acid residues are identified according to the Kabat numbering system.

[0107] In certain embodiments of a trivalent and trispecific Het-Fab antibody construct, the second Fab domain that is capable of binding DLL3 has a KD value for binding DLL3 from about 0.01 nM to about 100 nM or from about 0.1 nM to about 50 nM.

[0108] In various other embodiments, described herein are trivalent and trispecific antibody constructs that comprise one Fab domain and two scFv domains as binding domains. Such antibody constructs can also be referred to as trivalent and trispecific “scFv2” antibody constructs indicating that such constructs comprise two scFv binding domains.

[0109] In certain embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the first Fab light chain, and (c) the second scFv domain is coupled to the N- terminus of the second Fc polypeptide. FIG. IB shows a schematic of such an antibody construct according to certain embodiments of this disclosure.

[0110] In some embodiments of a trivalent and trispecific scFv2antibody construct, the first antigen on a cytotoxic effector cell is CD3 or CD28. In some embodiments, the second antigen on a cytotoxic effector cell is CD3 or CD28. In some embodiments, the first antigen is CD3 or CD28 and the second antigen is CD3 or CD28, wherein the first antigen and the second antigen are different. In certain embodiments, the first antigen is CD3 and the second antigen is CD28.[oni] In certain embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell;(ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the first Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide.

[0112] In certain embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure can comprise or consist of one or more polypeptide chains. In some embodiments, a trivalent and trispecific scFv2antibody construct herein comprises or consists of two immunoglobulin heavy chains (e.g., Hl and H2) and one immunoglobulin light chain (e.g., L). In some embodiments, a trivalent and trispecific scFv2antibody construct comprises or consists of a first immunoglobulin heavy chain (Hl) comprising at least a first Fab heavy chain, a second immunoglobulin heavy chain (H2) comprising at least a second Fab heavy chain and an immunoglobulin light chain (L) comprising at least a first Fab light chain.

[0113] In some of these embodiments in which a trivalent and tri specific scFv2antibody construct comprises three polypeptide chains Hl, H2 and L, Hl comprises, from N- to C-terminus, the first Fab heavy chain comprising a first VH sequence and a CHI sequence, coupled to the first Fc polypeptide comprising a first CH2 sequence and a first CH3 sequence, and wherein the first Fab heavy chain of Hl and the first Fab light chain of L associate to form the first Fab domain. In some embodiments, Hl further comprises a linkerFab'Fcthat couples the C-terminus of the first Fab heavy chain to the N-terminus of the first Fc polypeptide.

[0114] In certain embodiments of a trivalent and trispecific scFv2antibody construct, H2 comprises, from N- to C-terminus, the first scFv domain, comprising, from N- to C-terminus, either (i) an anti-DLL3 VH sequence coupled to an anti-DLL3 VL sequence or (ii) an anti-DLL3 VL sequence coupled to an anti-DLL3 VH sequence, coupled to the second Fc polypeptide comprising a second CH2 sequence and a second CH3 sequence, and wherein the first Fc polypeptide of Hl and the second Fc polypeptide of H2 associate to form a heterodimer H1-H2. In some of these embodiments, the first scFv domain comprises, from N- to C-terminus, the anti-DLL3 VL sequence coupled to the anti-DLL3 VH sequence. In some embodiments, H2 further comprises a linkerscFv'Fcthat couples the C-terminus of the first scFv domain to the N-terminus of the second Fc polypeptide.

[0115] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the linkerl ab"Fcand the linkerscFv'Fcare polypeptide linkers, each independently comprising from about 5 to about 50 consecutive amino acid residues. In some embodiments, the linkerl ab-l cand the linkerscFv'Fcare each independently derived from an IgGl, IgG2, or IgG4 hinge region.

[0116] In certain embodiments of a trivalent and trispecific scFv2antibody construct, L comprises, from N- to C-terminus, a first Fab light chain comprising a first VL sequence and a CL sequence, coupled to the second scFv domain comprising, from N- to C-terminus, either (i) a second VH sequence coupled to a second VL sequence or (ii) a second VL sequence coupled to a second VH sequence. In some embodiments, the second scFv domain comprises, from N- to C-terminus, the second VL sequence coupled to the second VH sequence. In some embodiments, L further comprises a linkerFab'scFvthat couples the C-terminus of the first Fab light chain to the N-terminus of the second scFv domain. In some embodiments, the linkerFab'scFvcomprises the amino acid sequence (G4S)X, wherein x is 1, 2, 3, 4 or 5.

[0117] In certain embodiments, a trivalent and trispecific scFv2antibody construct comprises: (i) a first Fab domain, wherein the first Fab domain is capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain, wherein the first scFv domain is capable of binding CD28 on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the first Fab light chain N-terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide.

[0118] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the first Fab domain is capable of binding CD3 and comprises a VH domain comprising a HCDR1 sequence comprising the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 sequence comprising the sequence ITSSGGRI (SEQ ID NO: 50) or ITRSGGRI (SEQ ID NO: 51), and a HCDR3 sequence comprising the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 sequence comprising the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 sequence comprising the sequence RND (SEQ ID NO: 54), and a LCDR3 sequence comprising the sequence QSYSSGFI (SEQ ID NO: 55).

[0119] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the first Fab domain is capable of binding CD3 and comprises a VH domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 5. In some embodiments, the anti-CD3 Fab domain comprises a VH domain comprising the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2 and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 5.

[0120] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising a HCDR1 sequence comprising the sequence SYGVH (SEQ ID NO: 28), a HCDR2 sequence comprising the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 sequence comprising the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL sequence comprising a LCDR1 sequence comprising the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 sequence comprising the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 sequence comprising the sequence QQSRKVPFT (SEQ ID NO: 48).

[0121] In some embodiments, the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 95% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 97% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 99% sequence identity to the sequence set forth in SEQ ID NO: 16. In some embodiments, the anti-CD28 scFv domain comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 16.

[0122] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the first Fc polypeptide and the second Fc polypeptide each comprise a CH2 domain that can independently comprise an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 8. In some embodiments of a trivalent and trispecific scFv2antibody construct, 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 IgGl CH2 domain sequence, wherein the one or more amino acid substitutions reduce or eliminate binding of the antibody construct to an FcyR. In some embodiments, the CH2 domain sequences of the first and second Fc polypeptides comprise theamino acid substitutions L234A L235A D265S, and wherein the numbering of amino acid residues in the first and second Fc polypeptides is according to the EU numbering system.

[0123] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the dimeric Fc domain is a heterodimeric Fc domain in which one Fc polypeptide has an amino acid sequence that differs in at least one residue from the amino acid sequence of the other Fc polypeptide. In some embodiments, one of the Fc polypeptides comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 9, and the other Fc polypeptide comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 10.

[0124] In certain embodiments, a trivalent and trispecific scFv2antibody construct comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 scFv domain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein the first Fc polypeptide of Hl and the second Fc polypeptide of H2 associate to form a heterodimer H1-H2; and (iii) a light chain (L), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein the anti-CD3 Fab light chain of L and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain.

[0125] In certain embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; (iii) a third binding domain capable of binding DLL3 on a tumor cell, wherein the third binding domain is a second Fab domain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to: the C-terminus of the light chain of eitherthe first Fab domain or the second Fab domain if the third binding domain is the second Fab domain, or the C-terminus of the light chain of the first Fab domain if the third binding domain is the second scFv domain. In some embodiments, the third binding domain is a second Fab domain. In other embodiments, the third binding domain is a second scFv domain. In various embodiments, the first scFv domain is coupled to the C-terminus of the Fab light chain of the first Fab domain, independent of structure or format (e.g., scFv or Fab) of the third binding domain.

[0126] In certain embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a third binding domain capable of binding a TAA on a tumor cell, wherein the third binding domain is a second Fab domain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain. In some embodiments, the third binding domain is a second Fab domain. In other embodiments, the third binding domain is a second scFv domain.

[0127] In various embodiments, the TAA is DLL3.

[0128] In certain embodiments of a trivalent and trispecific scFv2antibody construct, the first scFv domain has a KD value for binding DLL3 from about 0.01 nM to about 100 nM or from about 0.1 nM to about 50 nM.

[0129] In certain embodiments of a trivalent and trispecific antibody construct (e.g., a Het-Fab or scFv2antibody construct), the first Fc polypeptide and the second Fc polypeptide comprise IgGl or IgG4 CH3 sequences comprising one or more amino acid modifications compared to wildtype CH3 sequences that promote preferential pairing of the first and second Fc polypeptide to form a heterodimeric Fc domain compared to the formation of a corresponding homodimeric Fc domain. In some embodiments, the CH3 sequence of one of the Fc polypeptides 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 CH3 sequence of 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 thenumbering of amino acid residues in the Fc polypeptides is according to the EU numbering system. In some embodiments of a trivalent and trispecific Het-Fab antibody construct (e.g., a Het-Fab or scFv2antibody construct), the first or second Fc polypeptide comprises the amino acid substitutions T350V_L351Y_F405A_Y407V, and the other Fc polypeptide comprises the amino acid substitutions T350V_T366L_K392L_T394W, and wherein the numbering of amino acid residues in the first and second Fc polypeptides is according to the EU numbering system.

[0130] In certain embodiments of a trivalent and trispecific antibody construct (e.g., a Het-Fab or scFv2antibody construct), the first and second effector cell antigens that the antibody construct binds to can either be located on the same cytotoxic effector cell or on different cytotoxic effector cells. In some embodiments, the one or more cytotoxic effector cell(s) comprise one or more T cell(s). In various embodiments, the first and second effector cell antigens that the antibody construct binds to are located on the same cytotoxic effector cell. In other embodiments, the first and second effector cell antigens that the antibody construct binds to are located on different cytotoxic effector cells.B. Domains of an Antibody Construct

[0131] As further described herein, a trivalent and trispecific antibody construct of the present disclosure can comprise one or more antibody (or immunoglobulin (“Ig”)) domains. In various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a plurality of (i.e., two or more) antibody domains. Such plurality of antibody domains can comprise (i) one or more immunoglobulin Fc domain(s), wherein an Fc domain can comprise a first Fc polypeptide and a second Fc polypeptide, and can be either homodimeric or heterodimeric, (ii) one or more Fab domain(s), wherein a Fab domain can comprise a Fab heavy chain polypeptide comprising a heavy variable domain (VH) sequence and a heavy constant domain (CHI) sequence and a Fab light chain polypeptide comprising a light variable domain (VL) sequence and light constant domain (CL) sequence, and / or (iii) one or more scFv domains, wherein an scFv domain can comprise an scFv VH sequence coupled to an scFv VL sequence (from either N- to C-terminus or C- to N-terminus). The various domains an antibody construct can comprise are further described herein.

[0132] An Ig structural unit as known in the art is typically composed of two pairs of polypeptide chains, each pair having one “light” chain (about 25 kilodalton (kD)) and one “heavy” chain (about 50-70 kD). Light chains can be classified as either kappa or lambda. The “class” of an Ig refers tothe 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 can be further divided into subclasses (isotypes), for example, IgGl, IgG2, IgG3, IgG4, IgAl and IgA2. The heavy chain constant domains that correspond to the different classes of immunoglobulins (Igs) are called alpha (a), delta (8), epsilon (a), gamma (y) and mu (p), respectively.

[0133] In various embodiments, a trivalent and trispecific antibody construct described herein is based on an IgG class immunoglobulin, for example, an IgGl, IgG2, IgG3 or IgG4 immunoglobulin. In some embodiments, an antibody construct described herein is based on an IgGl, IgG2 or IgG4 immunoglobulin. In certain embodiments, an antibody construct described herein is based on an IgGl immunoglobulin. In the context of the present disclosure, when an antibody construct is based on a specified Ig isotype, it refers to an antibody construct that comprises either all or a portion of the constant region (i.e., a portion of an Fc domain) of the specified Ig isotype. It is to be understood that an antibody construct can also comprise hybrids of isotypes and / or subclasses, according to certain embodiments of this disclosure.

[0134] Generally, in antibodies, the N-terminal domain of each polypeptide chain usually defines a variable region (e.g., Vu or VL) of about 100 to 110 or more amino acids in length that is primarily responsible for antigen recognition. The terms variable light chain (VL) and variable heavy chain (VH) refer to these domains in the light and heavy chain, respectively. As described herein, in various embodiments, a trivalent and trispecific antibody construct of the present disclosure can comprise two or more variable domain sequences as part of its antigen binding domains. In various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises two variable domain sequence per binding domain (e.g., per Fab and / or scFv), and thus by being trivalent (i.e., containing three binding domains) and trispecific (i.e., each of the three binding domains targets a different antigen) can comprise a total of six variable domain sequences, e.g., three VH domain sequences and three VL domain sequences.

[0135] In some embodiments, two or more of such variable domains are coupled to one another in tandem and in a single polypeptide chain format, e.g., as described for scFv-type binding domains which contain (from either N- or C-terminus) a VH domain coupled to a VL domain. The two or more variable domains can be coupled to each other in tandem either directly or via a linker.

[0136] Accordingly, an antibody construct of the present disclosure that is derived from an Ig molecule (i.e., an antibody construct that comprises at least one Ig domain, wherein the one Igdomain can comprise a wildtype amino acid sequence and / or an amino acid sequence that contains one or more amino acid modifications relative to wildtype) can comprise different Ig domains within its heavy and light chain(s). Heavy chain domains can include the Fc domain (or Fc region), e.g., comprising a CH2 domain and CH3 domain, a hinge domain (or hinge region), and a heavy chain Fab domain comprising a variable heavy domain (VH) and a constant heavy domain (CHI), and light chain domains can include the variable light domain (VL) and the light constant domain (CL). In some embodiments, and as further described herein, a light chain of an antibody construct can further comprise an scFv domain coupled to, e.g., C-terminus of the CL constant domain of the light chain. In some embodiments, and according to certain nomenclatures, the “Fc domain” or “Fc region” can include the CH2 and CH3 domains as well as a hinge domain (or hinge region).

[0137] In some embodiments, an antibody construct described herein can further comprise another domain or moiety that may not be derived from an Ig molecule. Such a non-Ig domain can be referred to as a moiety. Such moiety can be a detectable label (e.g., a radiolabel or fluorescent label), a low molecular weight (e.g., <750 Da) drug molecule, another peptide (e.g., signal peptide(s)) or polypeptide molecule, or combinations thereof.B.l Complementarity Determining Regions (CDRs) and Binding Domains

[0138] In each of the VH and VL domains of an antibody construct herein are three loops which are hypervariable in sequence and form an antigen-binding site. Each of these loops is referred to as a “hypervariable region” or “HVR,” or “complementarity determining region” or “CDR.” The terms hypervariable region (HVR) and complementarity determining region (CDR) are used herein interchangeably in reference to the portions of the variable domain (e.g., VH or VL) that form the antigen-binding site. With the exception of CDR1 in VH, CDRS generally comprise the amino acid residues that form the hypervariable loops. The VH and VL domains consist of relatively invariant stretches called framework regions (FRs) of between about 15 to 30 amino acids in length separated by the shorter CDRs, which are each typically between about 5 and 15 amino acids in length, although can occasionally be longer or shorter. The three CDRs and four FRs that make up each VH and VL domain are generally arranged from N- to C-terminus as follows: FR1-CDR1- FR2-CDR2-FR3-CDR3-FR4.

[0139] Several different definitions and numbering conventions of the CDR regions in Ig molecules are in common use, including those described by Kabat et al. (1983, Sequences of Proteins of Immunological Interest, NIH Publication No. 369-847, Bethesda, MD), by Chothia etal. (1987, J Mol Biol, 196:901-917), as well as the IMGT, AbM and Contact definitions. These different definitions include overlapping or subsets of amino acid residues when compared against each other. By way of example, CDR definitions according to Kabat, Chothia, IMGT, AbM and Contact are provided in TABLE 1 below.

[0140] Accordingly, as can be readily apparent to one skilled in the art, the exact numbering and placement of CDRs can differ based on the numbering system employed. However, it is to be understood that the disclosure herein of a variable heavy domain (VH) includes the disclosure of the associated (inherent) heavy chain CDRs (HCDRs or HCDR1 -3) as defined by any of the known numbering systems. Similarly, disclosure herein of a variable light domain (VL) includes the disclosure of the associated (inherent) light chain CDRs (LCDRs or LCDR1-3) as defined by any of the known numbering systems. In case a set of CDR sequences (e.g., the 3 HCDRs and the 3 LCDRs) for a binding domain against a certain target is given or claimed using a specific numbering system (e.g., Kabat or IMGT), it is to be understood that such claim also encompasses the corresponding sets of CDR sequences as determined using any other of the known numbering systems.

[0141] One skilled in the art can appreciate that a limited number of amino acid substitutions can be introduced into the CDR sequences or to the VH or VL sequences of known antibodies without the antibody losing its ability to bind its target, e.g., a reduction in binding affinity of at least about 10-fold, 100-fold, or 1000-fold or more. Candidate amino acid substitutions can be identified by computer modeling or by techniques such as alanine scanning, with the resulting variants being tested for binding activity (e.g., expressed as binding affinity, e.g., given as the measured KD value) by standard techniques. As an example, in certain embodiments, the CD3 binding domain(s) of antibody constructs described herein can comprise a set of CDRs (i.e., heavy chain CDR1, CDR2 and CDR3 (HCDR1-3), and light chain CDR1, CDR2 and CDR3 (LCDR1-3) that have 90% or greater, 95% or greater, 98% or greater, 99% or greater, or 100% sequence identity to the amino acid sequences set forth in SEQ ID NOs: 49-55, respectively, wherein the binding domain retains or substantially retains the ability to bind CD3. In this context, the term “substantially” refers to a change in binding affinity of less than about 5%, less than about 4%, less than about 3%, less than about 2%, or less than about 1%. However, in certain other instances, one or skill in the art can also appreciate that introduction of one or more amino acid modifications into a set of six CDRs for a particular binding domain can result in substantial changes in one or more of the followingproperty parameters: antigen binding affinity, stability of the binding domain, and expression or production yields.TABLE 1: Common CDR Definitions11Either the Kabat or Chothia numbering system can be used for HCDR2, HCDR3 and the light chain CDRs for all definitions except Contact, which uses Chothia numbering.2Using Kabat numbering. The position in the Kabat numbering scheme that demarcates the end of the Chothia and IMGT CDR-H1 loop varies depending on the length of the loop due to the placement of insertions outside of those CDR definitions at positions 35 A and 35B in Kabat. The IMGT and Chothia CDR-H1 loop can be unambiguously 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.

[0142] In some embodiments, the antibody constructs described herein comprise at least one Ig domain from a mammalian Ig, such as a bovine Ig, a human Ig, a camelid Ig, a rat Ig, or a mouse Ig. In some embodiments, an antibody construct herein can be a chimeric construct comprising two or more Ig domains, in which at least one domain is from a first mammalian Ig, for example a human Ig, and at least a second domain is from a second mammalian Ig, for example, a mouse or rat Ig. In other embodiments, an antibody construct can be derived from Igs that are from different species, for example, an antibody construct can be chimeric or humanized. A “chimeric antibody construct” refers to an antibody that typically comprises at least one variable domain from a rodent antibody (usually a murine antibody) and at least one constant domain from a human antibody. A “humanized antibody construct” is a type of chimeric antibody that contains minimal sequence derived from a non-human antibody. In some embodiments, an antibody construct herein can comprise at least one Ig constant domain from a human Ig. In various embodiments, all domains of an antibody construct described herein can be (or be derived from) from a human Ig.

[0143] In some embodiments, and as further described herein, modifications (e.g., to the amino acid sequence) to one or more domains of an antibody construct can be made to further refine the properties and performance of the antibody construct (e.g., antigen affinity, stability, and / orpharmacokinetics, tumor cell killing, etc.). For example, framework region (FR) residues of a human Ig can be replaced by corresponding non-human residues, or the humanized antibodies can comprise residues that are not found in either the recipient antibody or the donor antibody. In general, a variable domain in a humanized antibody or a humanized antibody domain comprises all or substantially all of the hypervariable regions from a non-human Ig and all or substantially all of the FRs from a human Ig sequence. As further described herein, modifications in the Fc domain can enable preferential pairing of the Fc polypeptides to form a heterodimeric Fc domain rather than a homodimeric Fc domain, and / or alter interaction of the Fc domain with one or more FcyRs.

[0144] In some embodiments, the present disclosure relates to antibody constructs that can have certain valencies, e.g., can be trivalent. Hence, in various embodiments, an antibody construct herein comprises three antigen binding domains, i.e., is at least trivalent. In various embodiments of this disclosure, an antibody construct can be trispecific and trivalent, and thus such antibody construct can comprise three binding domains, wherein each of the three binding domains has a unique binding specificity for a different epitope (located on either the same antigen and / or on different antigens). In some of these embodiments, a tri specific and trivalent antibody construct of the present disclosure comprises three binding domains, e.g., one or more Fab domain(s) and / or one or more scFv domain(s), each capable of binding an epitope on one of three different antigens (e.g., CD3, CD28, and a TAA such as DLL3).

[0145] In some embodiments, an antibody construct of the present disclosure can comprise (i) one or more Fab domain(s), (ii) one or more scFv domain(s), and (iii) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0146] In various embodiments, a trivalent and trispecific antibody construct as described herein comprises (i) a first Fab domain capable of binding a first antigen, (ii) a first scFv domain capable of binding a second antigen, (iii) a third binding domain capable of binding a third antigen, wherein the third binding domain comprises or consists of a second Fab domain or a second scFv domain, and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0147] In some embodiments, a trivalent and trispecific antibody construct as described herein comprises (i) a first Fab domain capable of binding a first antigen, (ii) a second Fab domain capable of binding a second antigen, (iii) a first scFv domain capable of binding a third antigen, and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide. In otherembodiments, a trivalent and trispecific antibody construct as described herein comprises (i) a Fab domain capable of binding a first antigen, (ii) a first scFv domain capable of binding a second antigen, (iii) a second scFv domain capable of binding a third antigen, and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide.

[0148] Generally, a “Fab domain,” as used herein (also commonly known as an abbreviation of “fragment, antigen binding”), comprises a constant region comprising the constant domain (CL) of the light chain and the first constant domain (CHI) of the heavy chain, and a variable region comprising the variable domains VL and VH on the light and heavy chains, respectively, which comprise the CDRs as described herein. In some embodiments, a Fab domain can be a single chain Fab. A single chain Fab can be a Fab molecule in which 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 formats are also encompassed herein. In various embodiments herein, however, a Fab domain of an antibody construct is formed by two separately expressed polypeptide chains, i.e., a Fab light chain and a Fab heavy chain. The heavy chain and light chain parts of the Fab domain can, however, be interconnected by covalent bonds, such as disulfide bonds.

[0149] An “scFv domain,” as used herein (also commonly known as an abbreviation of “singlechain fragment, variable”), generally comprises a heavy chain variable domain (VH) and a light chain variable domain (VL) in a single polypeptide chain format. The scFv can optionally comprise a peptide linker between the VH and VL domains which can assist the scFv in forming a functional structure for antigen binding. Hence, in various embodiments, an scFv domain herein can include a VL domain that is coupled via its C-terminus to the N-terminus of a VH domain by a linkerscFv, i.e., an scFv domain can have the domain structure (from N- to C-terminus): VL-linkerscFv-VH, or alternatively, an scFv can comprise a VH connected by its C-terminus to the N-terminus of a VL by a linkerscFv, i.e., having the domain structure (from N- to C-terminus): VH-linkerscFv-VL.B.2 Binding Domains against Antigens on Cytotoxic Effector Cells

[0150] As further described herein, in various embodiments, an antibody construct of the present disclosure can comprise at least two binding domains capable of binding to one or more molecule(s), e.g., a polypeptide(s), on the surface of one or more cytotoxic effector cell(s). Such one or more cytotoxic effector cell(s) can be one or more immune cell(s). Such one or moreimmune cell(s) can comprise a T cell, a macrophage, a dendritic cell, a neutrophil, a B-cell, an NK cell, or a combination thereof.

[0151] In various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises at least one binding domain capable of binding a first antigen on a cytotoxic effector cell, and at least one binding domain capable of binding a second antigen on a cytotoxic effector cell.

[0152] In various embodiments, the cytotoxic effector cell is one cell, i.e., the first and the second antigens that an antibody construct of this disclosure can bind to are located on the surface of the same cell. The binding of an antibody construct to two different antigens located on one cytotoxic effector cell can also be referred to as “cis”-binding. In other embodiments, the cytotoxic effector cell is more than one cell, i.e., the first and the second antigens that an antibody constructs of this disclosure can bind to are located on the surfaces of different cells. The binding of an antibody construct to two different antigens located on two different cytotoxic effector cells can also be referred to as “trans”-binding or cross-linking of effector cells (e.g., T cells).

[0153] In some embodiments of a trivalent and trispecific DLL3-engaging antibody construct herein, the first antigen on a cytotoxic effector cell is CD3, and the second antigen on a cytotoxic effector cell is CD28. In some of these embodiments, both CD3 and CD28 can be engaged by a trivalent and trispecific antibody construct when located on the surface of the same cytotoxic effector cell (i.e., the antibody construct is capable of cis-binding CD3 and CD28 on the same effector cell). Trivalent and trispecific antibody constructs of the present disclosure that are capable of CD3 and CD28 cis-binding comprise a light chain comprising, from N- to C-terminus, an anti-CD3 VL sequence, a CL sequence, an anti-CD28 VL sequence, and an anti-CD28 VH sequence. Such light chain can further comprise one or more linkers, as described herein. Antibody constructs comprising such light chains are depicted in, e.g., FIGS. IB, ID, 1G and 1H. In some embodiments, such antibody constructs nearly exclusively engage CD3 and CD28 in cis (i.e., on the same effector cell). Hence, the type of anti-CD3 (Fab) and anti-CD28 (scFv) binding domains as well as their relative orientation and configuration can drive cis-binding of the antibody construct.

[0154] In other embodiments, both antigens CD3 and CD28 can be engaged by a trivalent and trispecific antibody construct when located on the surfaces of different cytotoxic effector cells (i.e., the antibody construct is capable of trans-binding CD3 and CD28 on two effector cells). Antibodyconstructs capable of engaging CD3 and CD28 in trans can have a format and structure as depicted, e.g., FIGS. 1A and 1C.

[0155] In some embodiments, the present disclosure relates to trivalent and trispecific antibody constructs comprising a first binding domain capable of binding CD3, a second binding domain capable of binding CD28, and a third binding domain capable of binding a DLL3. Hence, the antibody constructs described herein can also be referred to as “T cell engagers,” “TCEs” or “T cell engager molecules,” describing the ability of the construct to bind both, antigens on one or more effector cell(s) (e.g., T cell(s)) as well as DLL3 on a tumor cell. In some embodiments, the engagement of a trivalent and trispecific antibody construct with two different effector cell antigens on an effector cell and DLL3 on a tumor cell can be - at least temporarily - simultaneous, and thereby establish a TCR-independent immune synapse, and direct T cell-mediated cytotoxic activity to a tumor environment which contains tumor cells expressing DLL3. In various embodiments, and as further described herein, a trivalent and trispecific antibody construct may cause a significantly reduced immune cell (e.g., T cell) activation in the absence of DLL3, e.g., when the immune synapse cannot be fully formed due to an absence of DLL3, and relative to immune cell activation in the presence of DLL3.

[0156] In various embodiments, a first antigen binding domain capable of binding CD3 on a cytotoxic effector cell can be a Fab domain or an scFv domain. In these embodiments, a second antigen binding domain capable of binding CD28 on a cytotoxic effector cell can also be a Fab domain or an scFv domain.

[0157] In some embodiments, a first binding domain capable of binding CD3 on a cytotoxic effector cell is a Fab domain, and a second binding domain capable of binding CD28 on a cytotoxic effector cell is an scFv domain.

[0158] In various embodiments of a trivalent and trispecific antibody construct herein, the Fab domain that is capable of binding CD3 can comprise a heavy chain constant domain (CHI) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, such CHI domain sequence comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, such CHI domain sequence comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acidsequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, such CHI domain sequence comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, such CHI domain sequence comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, such CHI domain sequence comprises or consists of the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4.

[0159] In various embodiments of a trivalent and trispecific antibody construct, the Fab domain that is capable of binding CD3 can comprise a light chain constant domain (CL) comprising or consisting of an amino acid sequence having at least about 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, such CL domain sequence comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, such CL domain sequence comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, such CL domain sequence comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, such CL domain sequence comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, such CL domain sequence comprises or consists of the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7.B.2.1 Binding Domains Against CD 3

[0160] As described herein, in some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a binding domain capable of binding CD3 (e.g., CD3s) on a cytotoxic effector cell (e.g., T cell).

[0161] In various embodiments, such anti-CD3 binding domain (e.g., an scFv domain or a Fab domain) of an antibody construct herein can have an affinity for CD3 (given as a KD value for binding CD3) that is not more than about 1 nM, 5 nM, 10 nM, 20 nM, or not more than about 30 nM. In various embodiments, an anti-CD3 binding domain herein has a KD value for binding CD3that is from about 20 nM to about 80 nM, from about 30 nM to about 60 nM, or from about 40 nM to about 50 nM. In some embodiments, an anti-CD3 binding domain herein has a KD value for binding CD3 that is not more than about 30 nM, 40 nM, 50 nM, or about 60 nM. In various embodiments, an antibody construct herein comprises an anti-CD3 binding domain that has a KD value for binding CD3 from about 20 nM to about 40 nM, e.g., of about 30 nM.

[0162] In some embodiments, the anti-CD3 binding domain is capable of binding CD3 with a KD value from about 20 nM to about 200 nM, from about 30 nM to about 150 nM, from about 40 nM to about 100 nM, or from 50 nM to about 80 nM.

[0163] In various embodiments, an antibody construct herein comprises a binding domain capable of binding CD3 on a T cell, wherein such binding domain has a KD value for binding CD3 from about 20 nM to about 40 nM, e.g., of about 30 nM, and comprises the CDR sequences of the VH sequence as set forth in SEQ ID NOS: 49, 50 and 52 or SEQ ID NOS: 49, 51 and 52, and the CDR sequences of the VL sequence as set forth in SEQ ID NO: 53-55.

[0164] In various embodiments, an antibody construct herein comprises a binding domain capable of binding CD3 on a T cell, wherein such binding domain has a KD value for binding CD3 from about 20 nM to about 40 nM, e.g., of about 30 nM, and comprises a VH domain comprising a HCDR1 sequence comprising the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 sequence comprising the sequence ITSSGGRI (SEQ ID NO: 50) or ITRSGGRI (SEQ ID NO: 51), and a HCDR3 sequence comprising the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 sequence comprising the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 sequence comprising the sequence RND (SEQ ID NO: 54), and a LCDR3 sequence comprising the sequence QSYSSGFI (SEQ ID NO: 55).

[0165] In some embodiments, the anti-CD3 binding domain comprises a VH domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, and a VL domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 5. In certain embodiments, the anti-CD3 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5.

[0166] In some embodiments, the CDRs of an anti-CD3 paratope used in an antibody construct of the present disclosure comprise one or more amino acid modifications in one or more of the CDR sequences set forth in SEQ ID NOs: 49, 50 and 52-55 or SEQ ID NOS: 49 and 51-55, wherein at least about 80%, 90%, or 95% binding affinity to CD3 is retained compared to the paratope without such amino acid modifications.

[0167] In various embodiments of a trivalent and trispecific antibody construct, the anti-CD3 binding domain is a Fab domain as described herein.

[0168] The VH domain of an anti-CD3 Fab domain of an antibody construct herein can comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2. In some embodiments, the VH domain of such Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2. In some embodiments, the VH domain of such Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2. In some embodiments, the VH domain of such Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2. In some embodiments, the VH domain of such Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2. In yet other embodiments, the VH domain of such Fab domain comprises or consists of the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2.

[0169] The VL domain of an anti-CD3 Fab domain of an antibody construct herein, which can be part of a light chain that pairs with an anti-CD3 Fab domain sequence of a heavy chain to form the Fab domain, can comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 5. In some embodiments, the VL domain of such scFv or Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 5. In some embodiments, the VL domain of such scFv or Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 5. In some embodiments, the VL domain of such scFv or Fab domain comprises or consists of an amino acid sequence having at least about 97%sequence identity to the amino acid sequence set forth in SEQ ID NO: 5. In some embodiments, the VL domain of such scFv or Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 5. In various embodiments, the VL domain of such scFv or Fab domain comprises or consists of the amino acid sequence set forth in SEQ ID NO: 5.

[0170] In embodiments in which the anti-CD3 domain is a Fab domain, which further comprises a CHI domain and a CL domain in its heavy and light chains, respectively, the CHI domain of an anti-CD3 Fab domain of an antibody construct herein can comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, the CHI domain of such Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, the CHI domain of such Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, the CHI domain of such Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, the CHI domain of such Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4. In some embodiments, the CHI domain of such Fab domain comprises or consists of the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4.

[0171] Moreover, the CL domain of an anti-CD3 Fab domain of an antibody construct herein can comprise or consist of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, the CL domain of such Fab domain comprises or consists of an amino acid sequence having at least about 90% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, the CL domain of such Fab domain comprises or consists of an amino acid sequence having at least about 95% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In some embodiments, the CL domain of such Fab domain comprises or consists of an amino acid sequence having at least about 97% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ IDNO: 7. In some embodiments, the CL domain of such Fab domain comprises or consists of an amino acid sequence having at least about 99% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7. In yet other embodiments, the CL domain of such Fab domain comprises or consists of the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7.

[0172] In some embodiments, an anti-CD3 binding domain of an antibody construct is a Fab domain. In some embodiments, such an anti-CD3 Fab domain comprises a heavy chain, or a portion thereof (e.g., in cases in which the heavy chain further comprises an Fc portion, etc.), and a light chain. The heavy chain of the anti-CD3 Fab domain can comprise or consist of a VH domain coupled to a CHI domain, from N- to C-terminus. In various embodiments, the anti-CD3 Fab domain comprises (i) a VH domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, (ii) a VL domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 5, (iii) a CHI domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4, and (iv) CL domain comprising or consisting of an amino acid sequence having at least about 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 6 or SEQ ID NO: 7.

[0173] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 3, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 6.

[0174] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 4, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 7.

[0175] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1, a CHI domain comprising orconsisting of the amino acid sequence set forth in SEQ ID NO: 4, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 7.

[0176] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 3, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 7.

[0177] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 3, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 6.

[0178] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 2, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 4, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 7.

[0179] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 2, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 3, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 7.

[0180] In certain embodiments, the anti-CD3 Fab domain comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 2, a CHI domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 3, a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 5, and a CL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 6.

[0181] As further described herein, in some embodiments, the binding affinity of a trivalent and trispecific antibody construct of the present disclosure for one or more specific targets (e.g., CD3) can - at least in part - be dependent on its format and / or geometry. As an example, the relativeorientation and vicinity of an anti-CD3 binding domain in an antibody construct can affect the binding domain’s ability to interact with the target epitope it is binding, e.g., through steric hindrance, conformational changes occurring when the construct interacts with one or more of its targets (e.g., conditional degrees of freedom for a binding domain such as its steric flexibility), etc.

[0182] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a first Fab domain capable of binding CD3 and comprising the HCDR1-3 sequences of the VH sequence set forth in SEQ ID NO: 2, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 5, wherein the CDR sequences are determined using any of the numbering system known in the art.

[0183] In further embodiments, the CD3 binding affinity of an anti-CD3 binding domain of an antibody construct herein can be engineered and altered (e.g., increased / decreased relative to unmodified domains), e.g., by using one or more amino acid modifications relative to one or more reference sequence(s). In some embodiments, a trispecific and trivalent antibody construct of the present disclosure can comprise a variant anti-CD3 binding domain that comprises one or more amino acid modifications in its VH and / or VL domain(s) compared to the anti-CD3 binding domains described herein which comprise a VH sequence set forth in, e.g., SEQ ID NOs: 1 or 2 and a VL sequence set forth in, e.g., SEQ ID NO: 5. Such one or more amino acid modifications can reduce or increase the binding affinity of the variant anti-CD3 binding domain to CD3 when compared to the binding affinity of a corresponding anti-CD3 binding domain that does not comprise such one or more amino acid modifications.

[0184] In some embodiments, the one or more amino acid modifications used to alter the binding affinity of an anti-CD3 binding domain can include one or more amino acid substitution(s), one or more amino acid addition(s), and / or one or more amino acid deletion(s). In certain embodiments, the one or more amino acid modifications used to alter the binding affinity of an anti-CD3 binding domain comprise or consist of one or more amino acid substitution(s) relative to an unmodified binding domain sequence (e.g., an anti-CD3 VH or VL sequence).

[0185] In some embodiments, the anti-CD3 affinity of an affinity-altered CD3 binding domain can be about ±2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold higher or lower than a corresponding parental anti-CD3 binding domain.

[0186] In some embodiments, other anti-CD3 binding domains, such as those known in the art (e.g., SP34, OKT3, etc.), as well as variants thereof, can also be used in the trivalent and trispecific antibody constructs described herein.

[0187] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises an anti-CD3 binding domain comprising a VH sequence comprising the VHCDRsl-3 as set forth in SEQ ID NOs: 84-86, and a VL sequence comprising the VLCDRsl-3 as set forth in SEQ ID NOs: 87-89.

[0188] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises an anti-CD3 binding domain comprising a VH sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 82, and a VL sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 83. In some embodiments, such anti-CD3 binding domain comprises a VH sequence comprising the sequence set forth in SEQ ID NO: 82, and a VL sequence comprising the sequence set forth in SEQ ID NO: 83.B.2.2 Binding Domains Against CD28

[0189] As further described herein, in various embodiments, a trivalent and trispecific antibody construct of the present disclosure (e.g., a Het-Fab or scFv2antibody construct) comprises a binding domain capable of binding CD28.

[0190] In various embodiments, such anti-CD28 binding domain (e.g., an scFv domain or a Fab domain) of an antibody construct herein can have an affinity for CD28 (given as a KD value for binding CD28) that is from about 10 nM to about 500 nM or from about 20 nM to about 250 nM.

[0191] In some embodiments, an antibody construct herein comprises a binding domain (e.g., an scFv domain or a Fab domain) capable of binding CD28 on a T cell, wherein such binding domain has an affinity for CD28 that is from about 15 nM to about 35 nM or from about 5 nM to about 50 nM and comprises an anti-CD28 VH sequence comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and an anti-CD28 VL sequence comprising a LCDR1 having the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48).

[0192] In such embodiments, an antibody construct herein comprises a binding domain (e.g., an scFv domain or a Fab domain) capable of binding CD28 on a T cell, wherein such binding domain comprises a VH domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:15, and a VL domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:16. In some of these embodiments, the anti-CD28 binding domain comprises a VH domain comprising or consisting of the sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the sequence set forth in SEQ ID NO: 16.

[0193] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a first scFv domain capable of binding CD28 and comprising the HCDR1-3 sequences of the VH sequence set forth in SEQ ID NO: 15, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 16, wherein the CDR sequences are determined using any of the numbering system known in the art.

[0194] In other embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises an anti-CD28 binding domain comprising (i) a VH domain that comprises a sequence having one or more amino acid substitution(s) compared to the sequence set forth in SEQ ID NO: 15, and / or (ii) a VL domain that comprises a sequence having one or more amino acid substitution(s) compared to the sequence set forth in SEQ ID NO: 16, wherein the position of such amino acid substitution can be provided herein according to the IMGT numbering system.

[0195] In some embodiments, such one or more amino acid substitutions in either the anti-CD28 VH domain and / or the anti-CD28 VL domain can reduce the binding affinity of the corresponding anti-CD28 binding domain to CD28 by about 1.5-fold, 1.6-fold, 1.7-fold, 1.8-fold, 2.0-fold, 3.0- fold, 3.3-fold, 3.5-fold, 3.7-fold, 3.9-fold, 5.0-fold, 5.2-fold, 5.5-fold, 6.0-fold, 7.0-fold, 8.0-fold, 8.5-fold, 9.0-fold, 10-fold, 20-fold, or 25-fold, or from about 1.5-fold to about 25 -fold, from about 2.0-fold to about 20-fold, from about 3.0-fold to about 20-fold, or from about 5.0-fold to about 10- fold, when compared to the binding affinity of an anti-CD28 binding domain that does not contain a VH and / or VL sequence with such amino acid substitutions (e.g., a binding domain comprising the VH and VL sequence set forth in SEQ ID NOs: 15 and 16, respectively).

[0196] Hence, in some embodiments, an anti-CD28 binding domain (e.g., an scFv domain or a Fab domain) of an antibody construct herein can have an affinity for CD28 (given as a KD valuefor binding CD28) that is from about 10 nM to about 500 nM, from about 20 nM to about 600 nM, from about 20 nM to about 250 nM, from about 20 nM to about 150 nM, from about 20 nM to about 100 nM, or from about 20 nM to about 50 nM.

[0197] In various embodiments, the anti-CD28 binding domain of an antibody construct herein comprises or consists of an scFv domain.

[0198] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution P1058A relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 21.

[0199] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution G1064S relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 24.

[0200] In some embodiments, an anti-CD28 binding domain herein comprises a VL domain comprising the amino acid substitution V1035G relative to the amino acid sequence set forth in SEQ ID NO: 16. In some of these embodiments, the anti-CD28 VL domain comprises the amino acid sequence set forth in SEQ ID NO: 17.

[0201] In some embodiments, an anti-CD28 binding domain herein comprises a VL domain comprising the amino acid substitution D1068E relative to the amino acid sequence set forth in SEQ ID NO: 16. In some of these embodiments, the anti-CD28 VL domain comprises the amino acid sequence set forth in SEQ ID NO: 26.

[0202] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution E1080K relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 20.

[0203] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution Y1110S relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 18.

[0204] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution N111 laA relative to the amino acid sequence set forth inSEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 25.

[0205] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution Y1112S relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 76.

[0206] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution Li l 12aN relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 23.

[0207] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution Y1113S relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 27.

[0208] In some embodiments, an anti-CD28 binding domain herein comprises a VH domain comprising the amino acid substitution Y1037A relative to the amino acid sequence set forth in SEQ ID NO: 15. In some of these embodiments, the anti-CD28 VH domain comprises the amino acid sequence set forth in SEQ ID NO: 22.

[0209] In some embodiments, an anti-CD28 binding domain herein comprises a VL domain comprising the amino acid substitution Y1031A relative to the amino acid sequence set forth in SEQ ID NO: 16. In some of these embodiments, the anti-CD28 VL domain comprises the amino acid sequence set forth in SEQ ID NO: 77.

[0210] In some embodiments, an anti-CD28 binding domain herein comprises a VL domain comprising the amino acid substitution N1066A relative to the amino acid sequence set forth in SEQ ID NO: 16. In some of these embodiments, the anti-CD28 VL domain comprises the amino acid sequence set forth in SEQ ID NO: 19.

[0211] In some embodiments, an antibody construct comprising an anti-CD28 binding domain comprising one or more of the VH and / or VL domain substitutions described herein can exhibit a reduced non-specific anti-tumor activity in vitro and / or in vivo. In one such embodiment, an antibody construct comprising a mutated (e.g., relative to the huTN228 wildtype sequence) anti- CD28 binding domain can induce less non-specific (e.g., in the absence of the TAA) immune cellactivity, e.g., having reduced non-specific cytokine production by the immune cells. In certain embodiments, an antibody construct comprising an anti-CD28 binding domain that carries the N1066A substitution in the VL domain relative to huTN228 wildtype can induce a reduced nonspecific T cell activity, e.g., cytokine production, e.g., in the absence of the TAA. In another embodiment, an antibody construct comprising an anti-CD28 binding domain that carries the Y1031A substitution in the VL domain relative to huTN228 wildtype can induce a reduced nonspecific T cell activity, e.g., cytokine production. In some embodiments, non-specific T cell activity, e.g., cytokine production such as production of TNFa, IL-2, etc., can be reduced by about 10-fold, 20-fold, 30-fold, 50-fold, 60-fold, 70-fold, or about 100-fold.

[0212] In various embodiments, an antibody construct herein comprises a binding domain capable of binding CD28 on a T cell, wherein such binding domain has a KD value for binding CD28 from about 20 nM to about 600 nM and comprises a VH domain comprising a HCDR1 having the sequence SXiGVH (SEQ ID NO: 30), a HCDR2 having the sequence VIWX2GGX3TNFNSALMS (SEQ ID NO: 34), and a HCDR3 having the sequence DRAX4GX5YX6X7AMDY (SEQ ID NO: 40) and a VL sequence comprising a LCDR1 having the sequence RASESVEYYXsTSLMQ (SEQ ID NO: 43), a LCDR2 having the sequence AASX9VX10S (SEQ ID NO: 47), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), and wherein Xi = Y, A; X2 = P, A; X3 = G, S; X4 = S, Y; X5= N, A; X6= L, N; X7= S, Y; X8= G, V; X9= N, A; and X10 = E, D.

[0213] In some embodiments, an anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:15, and a VL domain comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO:16. In some of these embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0214] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 21, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0215] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 24, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0216] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 17.

[0217] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 26.

[0218] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 20, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0219] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 18, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0220] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 25, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0221] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 76, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0222] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ IDNO: 23, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0223] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 27, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0224] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 22, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 16.

[0225] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 77.

[0226] In some embodiments, the anti-CD28 binding domain of an antibody construct herein comprises a VH domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 29.

[0227] In various embodiments of a trivalent and trispecific antibody construct, the anti-CD28 binding domain comprises or consists of an scFv domain. As further described herein, such anti- CD28 binding domain can have the domain structure, from N- to C-terminus, of: VL-LinkerscFv- VH or VH-LinkerscFv-VL.

[0228] In embodiments in which the anti-CD28 binding domain of an antibody construct is an scFv domain, such anti-CD28 scFv domain can comprise, from either N- to C-terminus of C- to N-terminus, a VH domain comprising the amino acid sequence set forth in any one of SEQ ID NOs: 15, 18, 20, 21, 22, 23, 24, 25, 27, or 76, coupled to a VL domain comprising the amino acid sequence set forth in any one of SEQ ID NOs: 16, 17, 19, 26, or 77, via a linkerscFvwhich sequence is (GnS)m, wherein n, m can independently be 1, 2, 3, 4 or 5, and as set forth in SEQ ID NO: 69. In some of these instances, the linkerscFvcomprises or consists of the amino acid sequence set forth in SEQ ID NO: 59. And in some embodiments, the anti-CD28 scFv domain has a domain structure,from N- to C-terminus, of: VH-linkerscFv-VL. In other embodiments, the anti-CD28 scFv domain has a domain structure, from N- to C-terminus, of: VL-linkerscFv-VH.

[0229] In certain embodiments, described herein is an antibody construct comprising a binding domain capable of binding CD28, wherein the binding domain comprises a VH sequence comprising a HCDR1 having the sequence SXiGVH (SEQ ID NO: 30), a HCDR2 having the sequence VIWX2GGX3TNFNSALMS (SEQ ID NO: 34), and a HCDR3 having the sequence DRAX4GX5YX6X7AMDY (SEQ ID NO: 40), and a VL sequence comprising a LCDR1 having the sequence RASESVEYYX8TSLMQ (SEQ ID NO: 43), a LCDR2 having the sequence AASX9VX10S (SEQ ID NO: 47), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), and comprises one or more of the following amino acid substitutions at the positions as identified in the CDR sequences: Xi: Y to A, X2: P to A, X3: G to S, X4: S to Y, X5: N to A, Xe: L to N, X7: S to Y, X8: G to V, X9: N to A, or X10: E to D.

[0230] In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution Xi: Y to A. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X2: P to A. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X3: G to S. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X4: S to Y. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X5: N to A. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution Xe: L to N. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X7: S to Y. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X8: G to V. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X$>: N to A. In some embodiments, the antibody construct comprises an anti-CD28 binding domain comprising the substitution X10: E to D.

[0231] In some embodiments, such antibody construct can have a binding affinity for CD28 that is reduced by about 1.5-fold to about 25-fold, by about 2.0-fold to about 20-fold, by about 3.0-fold to about 20-fold, or by about 5.0-fold to about 10-fold, when compared to the binding affinity of an antibody construct comprising an anti-CD28 binding domain that does not contain the one or more amino acid substitutions in one or more of the CDR sequences.

[0232] In some embodiments, such antibody construct can comprise an anti-CD28 binding domain that comprises the CDRs of a VH domain as set forth in SEQ ID NOS: 28, 31, and 35, and the CDRs of the VL sequence set forth in SEQ ID NOS: 41, 44, and 48.

[0233] In some embodiments, such antibody construct can comprise an anti-CD28 binding domain that comprises an anti-CD28 VH sequence comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and an anti-CD28 VL sequence comprising a LCDR1 having the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48).

[0234] As further described herein, in various embodiments, an anti-CD28 binding domain of a trivalent and trispecific antibody construct is an scFv domain. Such anti-CD28 scFv domain can be coupled to a Fab light chain sequence, e.g., to the C-terminus of a Fab light chain sequence.

[0235] In some embodiments, other anti-CD28 binding domains, such as those known in the art (e.g., CD28.3, TGN1412, etc.), as well as variants thereof, can also be used in the trivalent and trispecific antibody constructs described herein.B.3 Binding Domains Against DLL3

[0236] As described herein, in various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a binding domain capable of binding DLL3. In various embodiments, an antibody construct herein can be trivalent and trispecific and comprises a first binding domain capable of binding a first antigen on a cytotoxic effector cell, a second binding domain capable of binding a second antigen on a cytotoxic effector cell, and a third binding domain, wherein such third binding domain is capable of binding DLL3.

[0237] Generally, the anti-DLL3 binding domain of an antibody construct herein can comprise or consist of an scFv domain or a Fab domain. In various embodiments of a trivalent and trispecific Het-Fab antibody construct, the anti-DLL3 binding domain comprises a Fab domain. In other embodiments of a trivalent and trispecific Het-Fab antibody construct, the anti-DLL3 binding domain comprises an scFv domain.

[0238] The anti-DLL3 binding domains that can be part of a trivalent and trispecific antibody construct described herein can encompass binding domains capable of recognizing and binding any one of the epitopes, portions, domains or motifs located on, e.g., mouse, cynomolgus and / orhuman DLL3. In various embodiments, an anti-DLL3 binding domain of a trivalent and trispecific antibody construct described herein can bind to an epitope of human DLL3, which can include an N-terminal ECD domain, EGF1-6, and / or the membrane proximal peptide.

[0239] Anti-DLL3 binding domains against one or more of these human DLL3 epitopes are known in the art, and any such binding domain which is capable of binding DLL3 can generally be used in the trivalent and trispecific antibody constructs described herein.

[0240] In some embodiments, an anti-DLL3 binding domain, e.g., an anti-DLL3 VH and / or an anti-DLL3 VL sequence, of a trivalent and trispecific antibody construct of this disclosure can be one that is known in the art, or a variant thereof, e.g., an anti-DLL3 binding domain that is described in any one of International Patent Publication Nos. WO 2013 / 126746, WO 2016 / 138038, WO 2017 / 021349, WO 2021 / 200898, WO 2019 / 234220, WO 2021 / 173307, WO 2021 / 226204 and WO 2021 / 155380, which are incorporated herein by reference in their entireties.

[0241] In various embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that has a KD value for binding DLL3 from about 0.01 nM to about 100 nM or from about 0.1 nM to about 50 nM. In some embodiments, a trivalent and trispecific antibody construct has a binding affinity (e.g., provided as a KD value) for DLL3 of at least about 40 nM, 30 nM, 20 nM, 10 nM, 5 nM, 3 nM, 2 nM, 1 nM, 0.5 nM, or at least about 0.2 nM. In some embodiments, a trivalent and trispecific antibody construct herein has a binding affinity (e.g., provided as a KD value) for DLL3 of from about 40 nM to about 0.1 nM, from about 20 nM to about 0.2 nM, from about 10 nM to about 0.2 nM, or from about 5 nM to about 0.2 nM, e.g., using SPR or other methods known in the art.

[0242] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VH sequence comprising the HCDR1 sequence as set forth in SEQ ID NO: 90, the HCDR2 sequence as set forth in SEQ ID NO: 91 and the HCDR3 sequence as set forth in SEQ ID NO: 92, and a VL sequence comprising the LCDR1 sequence as set forth in SEQ ID NO: 93, the LCDR2 sequence as set forth in SEQ ID NO: 94 and the LCDR3 sequence as set forth in SEQ ID NO: 95.

[0243] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti-DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 99, and a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 100.

[0244] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises a third binding domain (e.g., a Fab domain or an scFv domain) capable of binding DLL3 on a tumor cell comprising the HCDR1-3 sequences of the VH sequence set forth in SEQ ID NO: 99, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 100, wherein the CDR sequences are determined using any of the numbering system known in the art.

[0245] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VH sequence comprising the HCDR1 sequence as set forth in SEQ ID NO: 90, the HCDR2 sequence as set forth in SEQ ID NO: 96 and the HCDR3 sequence as set forth in SEQ ID NO: 92, and a VL sequence comprising the LCDR1 sequence as set forth in SEQ ID NO: 93, the LCDR2 sequence as set forth in SEQ ID NO: 94 and the LCDR3 sequence as set forth in SEQ ID NO: 95.

[0246] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 101, and a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 102.

[0247] In certain embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VH sequence comprising the HCDR1 sequence as set forth in SEQ ID NO: 90, the HCDR2 sequence as set forth in SEQ ID NO: 96 and the HCDR3 sequence as set forth in SEQ ID NO: 92, and a VL sequence comprising the LCDR1 sequence as set forth in SEQ ID NO: 97, the LCDR2 sequence as set forth in SEQ ID NO: 98 and the LCDR3 sequence as set forth in SEQ ID NO: 95.

[0248] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure, e.g., a trivalent and trispecific Het-Fab or scFv2antibody construct, comprises an anti- DLL3 binding domain (e.g., a Fab domain or an scFv domain, respectively) that comprises a VHsequence comprising the amino acid sequence set forth in SEQ ID NO: 103, and a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 104.C. Fc Domains

[0249] As described herein, a trivalent and trispecific antibody construct of this disclosure can comprise an Fc domain (or Fc region, or Fc, also commonly known as an abbreviation of “fragment, crystallizable”) 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 at least about 90%, 95%, 97%, or at least about 99% amino acid sequence identity. In some embodiments, each Fc polypeptide comprises one or more asymmetric amino acid substitutions that can promote preferential pairing of the Fc polypeptides to form the heterodimeric Fc domain (e.g., H1-H2) compared to formation of a respective homodimeric Fc domain (e.g., Hl-Hl or H2-H2).

[0250] The term “Fc domain,” as used herein, includes native (or wildtype) sequence Fc domains as well as variant Fc domains comprising one or more amino acid modifications relative to a corresponding native or wildtype Fc domain. Unless otherwise specified herein, numbering of amino acid residues in the Fc domain or constant region is according to the EU numbering system, also called the EU index, as described, e.g., in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. 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 and second Fc polypeptide) forming the dimeric (e.g., heterodimeric) Fc domain. In some embodiments, Fc polypeptides can comprise a C-terminal constant region of an Ig heavy chain that is capable of stable self-association. In various embodiments, and as further described herein, an Fc polypeptide (e.g., a first or a second Fc polypeptide) comprises at least one of a CH2 domain and / or a CH3 domain. In certain embodiments, an Fc polypeptide of an antibody construct described herein comprises a CH2 domain and a CH3 domain.

[0251] As disclosed herein, a trivalent and trispecific antibody construct can comprise an Fc domain, wherein such Fc domain can be a heterodimeric Fc domain. The Fc domain, e.g., heterodimeric Fc domain, of an antibody construct, unless otherwise specified, comprises a first Fc polypeptide and a second Fc polypeptide. Generally, each Fc polypeptide of a (e.g.,heterodimeric) Fc domain can comprise a CH2 domain, a CH3 domain, or, as described in various embodiments herein, both a CH2 domain and a CH3 domain.

[0252] In certain embodiments, an antibody construct herein comprises an Fc domain based on a human IgG Fc domain. In some embodiments, an antibody construct comprises an Fc domain based on a human IgGl or IgG4 Fc domain. In various embodiments, an antibody construct comprises a heterodimeric IgG Fc domain comprising two different Fc polypeptides, e.g., a first Fc polypeptide and a second polypeptide, wherein the first and second Fc polypeptides have different amino acid sequences, e.g., amino acid sequences that have about 90%, 95%, 97%, or 99% sequence identity when compared and aligned to one another, e.g., as further described herein. In some embodiments, the differences in the amino acid sequences of a first and second Fc polypeptide can be due to asymmetric amino acid substitutions that can be introduced into each Fc polypeptide chain to promote preferential paring of the heavy chains to form the heterodimeric Fc domain, compared to a corresponding homodimeric Fc domain.

[0253] In various embodiments, a trivalent and trispecific antibody construct herein comprises an Fc domain that is a modified IgG Fc domain, and in which at least the CH3 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to a respective wildtype CH3 domain. In some embodiments, an antibody construct herein comprises an Fc domain that is a modified IgG Fc domain in which at least the CH2 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to a respective wildtype CH2 domain. In some embodiments, an antibody construct comprises an Fc domain that is a modified IgG Fc domain in which both the CH3 domain and the CH2 domain of at least one Fc polypeptide comprises one or more amino acid modifications compared to respective wildtype CH3 and CH2 domains. In various embodiments, both Fc polypeptides of a heterodimeric Fc domain can comprise one or more amino acid modifications in their CH3 domains. In some embodiments, both Fc polypeptides of a heterodimeric Fc domain can comprise one or more amino acid modifications in their CH2 domains. In yet other embodiments, both Fc polypeptides of a heterodimeric Fc domain can comprise one or more amino acid modifications in their CH2 domains and CH3 domains.C.l Modified Fc Domains

[0254] In some embodiments, the present disclosure relates to trivalent and trispecific antibody constructs that can comprise a heterodimeric Ig (e.g., IgG) Fc domain comprising a modifiedheterodimeric CH3 domain, wherein the modified heterodimeric CH3 domain comprises one or more asymmetric amino acid modifications, i.e., one or both the first and the second Fc polypeptides each comprise one or more amino acid modifications in their CH3 domain sequences compared to respective wildtype sequences. As used herein, the term “asymmetric amino acid modification” generally refers to a modification in which an amino acid at a specific position on the first Fc polypeptide is different to the amino acid at the corresponding position on the second Fc polypeptide. These asymmetric amino acid modifications can comprise modifications of only one of the two amino acids at the corresponding position on each Fc polypeptide, or they can comprise modifications of both amino acids at the corresponding positions on each of the first and second Fc polypeptides. In various embodiments, an “asymmetric amino acid modification” is an asymmetric amino acid substitution.

[0255] In some embodiments, an antibody construct herein comprises a heterodimeric Fc domain comprising a modified CH3 domain (i.e., a heterodimeric CH3 domain consisting of the two CH3 domains of the first and second Fc polypeptides), wherein the modified CH3 domain comprises one or more asymmetric amino acid modifications that promote formation of the heterodimeric Fc domain (e.g., pairing of a first Fc polypeptide with a second Fc polypeptide) over formation of a corresponding homodimeric Fc domain (e.g., pairing of a first Fc polypeptide with another first Fc polypeptide). Amino acid modifications that can be made to the CH3 domain of an Fc domain in order to promote formation of a heterodimeric Fc domain are known in the art and include, for example, those described in International Publication No. WO 96 / 027011 (“knobs into holes”), Gunasekaran etal., 2010, J Biol Chem, 285, 19637-46 (“electrostatic steering”), Davis etal., 2010, Prot Eng Des Sei, 23(4): 195-202 (strand exchange engineered domain (SEED) technology) and Labrijn et al., 2013, Proc Natl Acad Sci USA, 110(13):5145-50 (Fab-arm exchange). Other examples include approaches combining positive and negative design strategies to produce stable asymmetrically modified Fc regions as described in International Patent Publication Nos. WO 2012 / 058768 and WO 2013 / 063702.

[0256] In certain embodiments, an antibody construct described herein comprises a heterodimeric Fc domain comprising a modified heterodimeric CH3 domain in which 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.

[0257] In some embodiments, an antibody construct described herein comprises a heterodimeric human IgGl Fc domain having a modified CH3 domain. TABLE 2 herein provides the amino acid sequence of a human IgGl Fc domain sequence (e.g., a sequence that a first and / or a second Fc polypeptide can be derived from), corresponding to amino acids 231 to 447 of a full-length human IgGl heavy chain (e.g., one that comprises VH, CHI, Hinge, CH2 and CH3 domains), and identified by SEQ ID NO: 78. The CH2 domain is typically defined as comprising amino acids 231-340 of the full-length human IgGl heavy chain and the CH3 domain is typically defined as comprising amino acids 341-447 of the full-length human IgGl heavy chain.

[0258] As described herein, an antibody construct can comprise a heterodimeric Fc domain having a modified CH3 domain comprising one or more asymmetric amino acid modifications that promote formation of the heterodimeric Fc domain over formation of a homodimeric Fc domain, and in which the modified CH3 domain comprises a first Fc polypeptide including amino acid modifications at positions F405 and Y407, relative to SEQ ID NO: 78, and a second Fc polypeptide including amino acid modifications at positions T366 and T394, relative to SEQ ID NO: 78. In various embodiments, the one or more amino acid modifications comprise one or more amino acid substitutions. Hence, 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 a first Fc polypeptide further includes an amino acid modification at position L351, relative to SEQ ID NO: 78. In some embodiments, the amino acid modification at position L351 in 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: 78. In some embodiments, the amino acid modification at position K392 in the second Fc polypeptide of the modified CH3 domain is K392F, K392L or K392M. In some embodiments, one or both of the first and second Fc polypeptides of the modified CH3 domain further comprises the amino acid modification T350V.

[0259] In certain embodiments, an antibody construct herein comprises a heterodimeric Fc domain having a modified CH3 domain comprising one or more asymmetric amino acid modifications that promote formation of the heterodimeric Fc domain over formation of a homodimeric Fc domain, and in which the modified CH3 domain comprises a first Fc polypeptide including the amino acid modification F405A, F405I, F405M, F405S, F405T or F405V together with the amino acid modification Y407I or Y407V, and relative to SEQ ID NO: 78, and a second Fc polypeptide including the amino acid modification T366I, T366L or T366M, together with the amino acid modification T394W, and relative to SEQ ID NO: 78. In some embodiments, the first Fc polypeptide of the modified CH3 domain further includes the amino acid modification L351Y. In some embodiments, the second Fc polypeptide of the modified CH3 domain further includes the amino acid modification K392F, K392L or K392M. In some embodiments, one or both of the first and second Fc polypeptides having a modified CH3 domain further comprises the amino acid modification T350V.TABLE 2: Exemplary Human IgGl Fc Domain Sequences and Variants Thereof* “A” corresponds to a first Fc polypeptide chain and “B” to corresponds to a second Fc polypeptide chain.

[0260] In certain embodiments, an antibody construct herein comprises a heterodimeric Fc domain comprising a modified CH3 domain having a first Fc polypeptide that comprises amino acid modifications at positions F405 and Y407, and optionally further comprises an amino acid modification at position L351, and a second Fc polypeptide that comprises amino acidmodifications at positions T366 and T394, and optionally further comprises 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 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, relative to SEQ ID NO: 78.

[0261] In some embodiments, an antibody construct comprises a heterodimeric Fc domain comprising a modified CH3 domain comprising the modifications of any one of Variant 1, Variant 2, Variant 3, Variant 4 or Variant 5, as shown in TABLE 2.

[0262] In various embodiments, an antibody construct of the present disclosure can 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 the amino acid substitutions in their CH3 domains according to variant #1 as shown in TABLE 2. In other embodiments, an antibody construct of the present disclosure can 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 the amino acid substitutions in their CH3 domains according to variant #2 as shown in TABLE 2. In some embodiments, an antibody construct of the present disclosure can 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 the amino acid substitutions in their CH3 domains according to variant #3 as shown in TABLE 2. In some embodiments, an antibody construct of the present disclosure can 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 the amino acid substitutions in their CH3 domains according to variant #4 as shown in TABLE 2. In yet other embodiments, an antibody construct of the present disclosure can 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 the amino acid substitutions in their CH3 domains according to variant #5 as shown in TABLE 2

[0263] In certain embodiments, the CH3 domain of a first Fc polypeptide of an 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 set forth in SEQ ID NO: 9. In certain embodiments, the CH3 domain of a second Fc polypeptide of an 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 set forth SEQ ID NO: 10. In some embodiments, the CH3 domain of a first Fc polypeptide of an antibody construct herein has the amino acid sequence set forth in SEQ ID NO: 9 and a second Fc polypeptide of the antibody construct herein has the amino acid sequence set forth SEQ ID NO: 10.

[0264] In certain embodiments, an antibody construct herein comprises a heterodimeric Fc domain based on an IgG Fc domain having a modified CH2 domain (i.e., a heterodimeric CH2 domain consisting of the two CH2 domain sequences of the respective first and second Fc polypeptides). In some embodiments, an antibody construct comprises a heterodimeric Fc domain based on an IgG Fc domain having a modified CH2 domain, wherein the modification(s) of the CH2 domain result(s) in altered (e.g., reduced or abated) binding to one or more Fc receptors (FcRs) such as receptors of the FcyRI, FcyRII and FcyRIII subclasses.

[0265] Several amino acid modifications to the CH2 domain of the first and / or second Fc polypeptide(s) of an Fc domain that selectively alter the affinity of such Fc domain for different FcyRs are known in the art. Amino acid modifications that result in increased binding and amino acid modifications that result in decreased binding can both be useful in certain indications. For example, increasing binding affinity of an Fc for FcyRIIIa (an activating receptor) can result in increased antibody dependent cell-mediated cytotoxicity (ADCC), which in turn can result in increased lysis of the target cell. Decreased binding to FcyRIIb (an inhibitory receptor) likewise can be beneficial in some circumstances. In certain indications, a decrease in, or elimination of, ADCC and complement-mediated cytotoxicity (CDC) can be desirable. In such embodiments, modified CH2 domains comprising amino acid modifications that result in increased binding to FcyRIIb or amino acid modifications that can decrease or eliminate binding of the Fc region to all of the FcyRs (“knock-out” variants) can be useful.

[0266] Non-limiting examples of amino acid modifications to the CH2 domain that alter binding of the Fc domain by FcyRs include, but are not limited to, the following: S298A / E333A / K334A and S298A / E333A / K334A / K326A (increased affinity for FcyRIIIa) (Lu, et al., 2011, J ImmunolMethods, 365(1-2): 132-41); F243L / R292P / Y300L / V305I / P396L (increased affinity for FcyRIIIa) (Stavenhagen, et al., 2007, Cancer Res, 67(18):8882-90); F243L / R292P / Y300L / L235V / P396L (increased affinity for FcyRIIIa) (Nordstrom JL, et al., 2011, Breast Cancer Res, 13(6):R123); F243L (increased affinity for FcyRIIIa) (Stewart, et al., 2011, Protein Eng Des SeL, 24(9):671-8); S298A / E333A / K334A (increased affinity for FcyRIIIa) (Shields, et al., 2001, J Biol Chem, 276(9):6591-604); S239D / I332E / A330L and S239D / I332E (increased affinity for FcyRIIIa) (Lazar, et al., 2006, Proc Natl Acad Sci USA, 103(11):4005-10), and S239D / S267E and S267E / L328F (increased affinity for FcyRIIb) (Chu, et al., 2008, Mol Immunol, 45(15):3926-33). Additional modifications that affect Fc domain binding to FcyRs are described in Therapeutic Antibody Engineering (Strohl & Strohl, Woodhead Publishing series in Biomedicine No 11, ISBN 1 907568 37 9, Oct 2012, page 283).

[0267] In various embodiments, an antibody construct of the present disclosure comprises a heterodimeric Fc domain based on an IgG Fc domain having a modified CH2 domain, in which one or both of the CH2 sequences (i.e., of the first / second Fc polypeptide) of the modified dimeric CH2 domain comprise one or more amino acid modifications that can result in decreased or eliminated binding of the Fc domain to one or more, or all of the FcyRs (i.e., a “knock-out” or “KO” variant).

[0268] Various publications describe strategies that have been used to engineer antibodies to produce “knock-out” Fc variants (see, for example, 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 reduction of effector function through modification of glycosylation, use of IgG2 / IgG4 scaffolds, or the introduction of mutations in the hinge or CH2 domain of the Fc (see also, U.S. Patent Publication No. 2011 / 0212087, International Patent Publication No. WO 2006 / 105338, U.S. Patent Publication No. 2012 / 0225058, U.S. Patent Publication No. 2012 / 0251531, and Strop et a!., 2012, J. Mol. Biol., 420: 204-219).

[0269] In some embodiments, an Fc domain of an antibody construct can comprise one or more of known amino acid modifications to reduce FcyR and / or complement binding of the Fc domain. In some embodiments, such modifications can include those identified in TABLE 3.TABLE 3: Modifications to Reduce Fey Receptor or Complement Binding

[0270] Additional examples herein include Fc domains engineered to include the amino acid modifications L235A / L236A / D265S, e.g., based on the sequence set forth in SEQ ID NO: 78. In addition, asymmetric amino acid modifications in the CH2 domain that decrease binding of the Fc to all FcyRs are described in International Patent Publication No. WO 2014 / 190441.

[0271] In certain embodiments, the CH2 domain of a first and a 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 set forth in SEQ ID NO: 8. In some embodiments, the CH2 domain of a first and / or a second Fc polypeptide herein comprises or consists of the sequence set forth in SEQ ID NO: 8.

[0272] In certain embodiments, an antibody construct herein comprises a heterodimeric Fc domain in which native glycosylation has been modified. As is known in the art, glycosylation of an 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) results in an aglycoslated Fc that lacks all effector function (Bolt et a!.. 1993, Eur. J. Immunol., 23:403-411; Tao & Morrison, 1989, J. Immunol., 143:2595-2601). Conversely, removal of fucose from heavy chain N297-linked oligosaccharides has been shown to enhance ADCC, based on improved binding to FcyRIIIa (see, for example, 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 fucosyltransf erase (FUT8) (Yamane-Ohnuki et al. , 2004, BiotechnoL Bioeng., 87:614-622), in the variant CHO cell line, Lee 13, that has a reduced ability to attach fucose to N297-linked carbohydrates (International Publication No. WO 03 / 035835), or in other cells that generateafucosylated antibodies (see, for example, 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). In addition, International Patent Publication No. WO 2009 / 135181 describes the addition of fucose analogs to culture medium during antibody production to inhibit incorporation of fucose into the carbohydrate on the antibody.D. Linkers

[0273] In various embodiments of this disclosure, a trivalent and trispecific antibody construct described herein can comprise one or more linkers. In some embodiments, such one or more linkers are peptide (also referred to herein as “peptitic”) linkers comprising or consisting of an amino acid sequence of about 1, 2, 3, 5, 10, 15, 20, 25, 30, 40, or about 50 consecutive amino acid residues in length. The one or more peptide linker of an antibody construct can comprise or consist of an amino acid sequence from 1 to about 50, from 2 to about 40, from 3 to about 30, or from 5 to about 25 consecutive amino acid residues in length.

[0274] Such peptide linkers can couple, or link, two or more polypeptide sequences and / or domains of an antibody construct to each other. In various embodiments, a linker herein can couple a first polypeptide chain, e.g., a heavy chain constant domain (CHI), to an Fc polypeptide. Thus, a linker can be used to couple one domain of an antibody construct to another domain, from N- to C-terminus, e.g., a Fab domain to an Fc domain, e.g., a linkerFab'Fc, a VH domain to a VL domain, e.g., linkerscFv, and so forth. In embodiments in which an antibody construct comprises two scFv domains (e.g., an scFv2antibody construct), both such scFv domains of an antibody construct can contain a linkerscFvwith identical amino acid sequence, and such antibody construct can be described as comprising a linkerscFv, instead of specifying that it contains a linkerscFvland a linkerscFv2. In other embodiments, however, in which the linkerscFvof both scFv domains have different amino acid sequences, such antibody construct can be described as comprising a linkerscFvland a linkerscFv2.

[0275] In embodiments in which a linker couples, e.g., a heavy chain variable domain (VH) to, e.g., a light chain variable domain (VL), the linker can be of sufficient length to allow both domains to elicit their biological function. In addition to providing a spacing function, a linker herein (e.g., a peptide linker) can provide flexibility or rigidity suitable for properly orienting the one or more domains of an antibody construct, both within the antibody construct itself and between the antibody construct and its target(s).

[0276] Further, a linker herein (e.g., a peptide linker) can support (i) expression of a full-length fusion protein, e.g., a full-length polypeptide chain Hl, LI, H2, and / or L2 etc. of an antibody construct, and (ii) provide increased stability of the purified protein both in vitro and in vivo, e.g., following administration to a subject in need thereof, such as a human. The one or more linkers used in antibody constructs herein are generally non-immunogenic or poorly immunogenic in mammalian subjects that a construct may be administered to. In certain embodiments, one or more of the linkers used in an antibody construct herein can comprise part or all of a human Ig hinge region, a stalk region of C-type lectins, a family of type II membrane proteins, or combinations thereof. In certain embodiments, one or more of the linker used in an antibody construct herein can comprise part or all of a human Ig hinge region, such as an IgGl hinge region, such as a linkerFab'Fcor a linkerscFv'Fc.

[0277] In certain embodiments, each linker used in an 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 linker used in an antibody construct herein can comprise or consist of an amino acid sequence having a length from about 3 to about 40 amino acids, from about 10 to about 50 amino acids, from about 2 to about 40 amino acids, from about 5 to about 30 amino acids, from about 5 to about 25 amino acids, from about 4 to about 30 amino acids, from about 10 to about 30 amino acids, or from about 15 to about 25 amino acids. In some embodiments, the one or more linkers of an antibody construct can 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.

[0278] In certain embodiments, a linker (e.g., a linkerscFv, linkerFab'scFv, linkerFab'Fc, etc.) of an antibody construct herein comprises of consists of the amino acid sequence (EAAAK)nwherein n is an integer from 1 to 5 (SEQ ID NO: 60). In some embodiments, a linker comprises or consists of the sequence EAAAK (SEQ ID NO: 61). In some embodiments, a linker comprises or consists of the sequence EAAAKEAAAK (SEQ ID NO: 62). In some embodiments, a linker comprises a polyproline linker, e.g., having an amino acid sequence of PPP (SEQ ID NO: 63) or PPPP (SEQ ID NO: 64). In certain embodiments, a linker is a glycine (G)-proline (P) polypeptide linker, e.g., comprising or consisting of one or more of GPPPG (SEQ ID NO: 65), GGPPPGG (SEQ ID NO: 66), GPPPPG (SEQ ID NO: 67), or GGPPPPGG (SEQ ID NO: 68). In some embodiments, a linkerherein is a (GnS)mlinker, wherein n and m are independently integers from 1 to 5 (SEQ ID NO: 69). In certain embodiments, a linker comprises or consists of an amino acid sequence of (G3S)n(G4S)i (SEQ ID NO: 70), (G3S)i(G4S)n(SEQ ID NO: 71), (G3S)n(G4S)n(SEQ ID NO: 72), wherein each n is an integer from 1 to 5, or (G4S)X(SEQ ID NO: 119) wherein each x is an integer from 1 to 5. In certain embodiments, a linker herein is suitable for connecting two different domains of an antibody construct and comprises a sequence comprising glycine-serine linkers, for example, but not limited to, (GmS)n-GG (SEQ ID NO: 73), (SGn)m (SEQ ID NO: 74), or (SEGn)m (SEQ ID NO: 75), wherein m and n are independently integers from 1 to 20.

[0279] In some embodiments, an antibody construct described herein comprises any one or more of the linkers described herein. In some embodiments, an antibody construct comprises a plurality of linkers, e.g., 1, 2, 3, 4, or 5 linkers, which can include one or more linkerl ab-l c, one or more linkerFab'scFv, and / or one or more linkerscFv.

[0280] In certain embodiments, a linkerscFvof an antibody construct herein comprises or consists of the amino acid sequence (GnS)mlinker, wherein n and m are independently integers from 1 to 5 (SEQ ID NO: 69). In such embodiments, n and m can both be 4, and thus the one or more linkerscFvof an antibody construct can comprise or consist of the sequence (G4S)4(SEQ ID NO: 59). In some embodiments, the one or more linkerscFvof an antibody construct (e.g., a linkerscFvland a linkerscFv2) can comprise or consist of an amino acid sequence having about 80%, 90%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 59.

[0281] In some embodiments, an antibody construct of the present disclosure can comprise a linkerFab'scFvthat couples the C-terminus of a Fab light chain to the N-terminus of an scFv domain. In some embodiments, such linkerFab'scFvcan comprise or consist of the amino acid sequence set forth in SEQ ID NO: 58 (GGGGSG), or a sequence having about 40%, 60%, or 80% sequence identity thereto.

[0282] In certain embodiments, one or more of the linkers that an antibody construct comprises can be an amino acid sequence obtained, derived, or designed from an antibody hinge region sequence. In some embodiments, such linker can have at least one cysteine capable of participating in at least one disulfide bond under physiological conditions or other standard peptide conditions (e.g., peptide purification conditions, conditions for peptide storage, etc.). In certain embodiments, a linker corresponding to, or similar to, an Ig hinge region peptide retains a cysteine that corresponds to the hinge cysteine disposed toward the amino (or N-) terminus of that hinge region.In further embodiments, a linker is derived from an IgGl hinge region and can be modified to remove any one or more of the cysteine residues. In some embodiments, a linker comprises an IgGl hinge region that has one cysteine or two cysteines that correspond to wildtype hinge cysteine residues.

[0283] In certain embodiments, a linker of an antibody construct described herein can comprise an “altered wildtype Ig hinge region” or an “altered Ig hinge region”. Such altered hinge regions can refer to (a) a wild type Ig hinge region with up to 30 percent amino acid changes (or up to 25 percent, 20 percent, 15 percent, 10 percent, or 5 percent amino acid substitutions, insertions or deletions), (b) a portion of a wild type Ig hinge region that is at least 10 amino acids (e.g., at least 12, 13, 14 or 15 amino acids) in length with up to 30 percent amino acid changes (or up to 25 percent, 20 percent, 15 percent, 10 percent, or 5 percent amino acid substitutions or deletions), (c) a portion of a wild type Ig hinge region that comprises the core hinge region, which portion can be 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15, or at least 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 amino acids in length, or (d) a combination of any of (a)-(c). In certain embodiments, one or more cysteine residues in a wildtype Ig hinge region, such as an IgGl hinge region comprising the upper and core regions, can be substituted by one or more other amino acid residues (e.g., one or more serine residues). An altered Ig hinge region can alternatively or additionally have a proline residue of a wildtype Ig hinge region, such as an IgGl hinge region, substituted by another amino acid residue (e.g., a serine residue).

[0284] In some embodiments, an antibody construct of this disclosure comprises a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 56. In some embodiments, an antibody construct comprises a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 57. In some embodiments, an antibody construct comprises a linker comprising or consisting of an amino acid sequence having at least about 80%, 90%, 95%, 97%, 99%, or 100% sequence identity to the sequence set forth in SEQ ID NO: 58.

[0285] In certain embodiments, a trivalent and trispecific Het-Fab antibody construct of this disclosure can comprise a first linkerl ab-l cthat couples one Fab heavy chain to one Fc polypeptide and a second linkerFab'Fcthat couples the other Fab heavy chain to the other Fc polypeptide. In some embodiments, the first linkerFab'Fccan comprise or consist of the amino acid sequence setforth in SEQ ID NO: 56, and the second linkerFab'Fccan comprise or consist of the amino acid sequence set forth in SEQ ID NO: 120.

[0286] In various embodiments, an antibody construct of this disclosure comprises a linker that couples a Fab domain to a first Fc polypeptide, linkerl ab-l c, wherein the linkerFab'Fccomprises or consists of the amino acid sequence set forth in SEQ ID NO: 56, and another linker coupling a first or second scFv domain to a first or second Fc polypeptide, linkerscFv'Fc, wherein the linkerscFv'Fccomprises or consists of the amino acid sequence set forth in SEQ ID NO: 57.E. Certain Embodiments of Antibody Constructs

[0287] The present disclosure describes trivalent and trispecific antibody constructs that can have different formats and / or geometries, e.g., spatial molecular structures. The trivalent and trispecific antibody constructs described herein have the ability to engage three different antigens, two different antigens on one or more cytotoxic effector cell(s) (e.g., T cell(s)) and DLL3 on a tumor cell, wherein each of these antigens is engaged in a monovalent manner.

[0288] In various embodiments, the present disclosure relates to trivalent and trispecific Het-Fab antibody constructs. Such Het-Fab antibody constructs can comprise three antigen binding domains, two of which are Fab domains. Since both Fab domains have binding specificities for different antigens, the two Fab domains comprise different heavy and light chain amino acid sequences (e.g., different sequences in the corresponding VH / VL pairs) they can be described as heterodimeric-Fabs or Het-Fabs (e.g., compared to Fabs that share heavy and light chains with identical amino acid sequences and target the same antigen, e.g., homodimeric-Fabs).

[0289] In certain embodiments, the trivalent and trispecific Het-Fab antibody constructs described herein are capable of binding DLL3 on a target cell (e.g., a tumor cell).

[0290] In various embodiments, described herein is a trivalent and trispecific Het-Fab antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) an scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of a Fab light chain. In some embodiments, the scFv domain is coupled to the C- terminus of the light chain of the first Fab domain.

[0291] In some embodiments herein, a trivalent and trispecific Het-Fab antibody construct is capable of monovalently engaging the following three antigens: CD3, CD28 and DLL3.

[0292] In some of these embodiments, a trivalent and trispecific Het-Fab antibody construct comprises: (i) the first Fab domain, wherein the first Fab domain is capable of binding CD3 on a cytotoxic effector cell; (ii) the second Fab domain, wherein the second Fab domain is capable of binding DLL3 on the tumor cell; (iii) the scFv domain, wherein the scFv domain is capable of binding CD28 on a cytotoxic effector cell; and (iv) the dimeric Fc domain comprising the first Fc polypeptide and the second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0293] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure that is capable of binding CD3, CD28 and DLL3 comprises or consists of four polypeptide chains, two immunoglobulin heavy chains (e.g., Hl and H2) and two immunoglobulin light chains (e.g., LI and L2). In some embodiments, the four polypeptide chains associate, either covalently and / or non-covalently, to form the trivalent and trispecific Het-Fab antibody construct.

[0294] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a heterodimeric Fc domain. In such embodiments, Hl and / or H2 can comprise one or more amino acid modifications that promote formation of the heterodimer Hl -H2 compared to formation of respective homodimers Hl -Hl or H2-H2. In some of these embodiments, and as further described herein, Hl and H2 each comprise a CH3 sequence, wherein each CH3 sequence comprises one or more amino acid substitutions, relative to a respective wildtype CH3 sequence, that promote heterodimer H1-H2 formation compared to homodimer formation.

[0295] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, the CH3 sequence of one of the Fc polypeptides 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 CH3 sequence of 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 polypeptides is according to the EU numbering system.

[0296] In some embodiments of a trivalent and trispecific Het-Fab antibody construct, Hl, H2, LI and / or L2 each comprise one or more amino acid modifications that promote correct heavy and light chain pairing to form the desired Hl -LI and H2-L2 heterodimer pairs, when compared to the formation of incorrectly paired heterodimers, e.g., H1L2 and H2L1. In some embodiments, Hl, H2, LI and / or L2 each comprise one or more amino acid substitutions in their CHI and CL sequences, respectively, that promote correct heavy and light chain pairing, and relative to unmodified parental and / or wildtype CHI and CL sequences.

[0297] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises one Fab domain comprising one or more of the amino acid substitutions 143E, 145T, 179E and 228D in the CHI sequence of the heavy chain and one or more of the amino acid substitutions 121K, 124R and 178R in the CL sequence of the light chain, and another Fab domain comprising one or more of the amino acid substitutions 125R and 188K in the CHI sequence of the heavy chain and one or more of the amino acid substitutions 122D, 129T, 176E and 178E in the CL sequence of the light chain, wherein these amino acid substitutions are according to the Kabat numbering scheme and relative to unmodified parental and / or wildtype CHI and CL sequences and promote correct heavy and light chain pairing for the formation of the desired Hl -LI and H2-L2 heterodimers, compared to formation of the incorrectly paired heterodimers H1L2 and H2L1.

[0298] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 Fab heavy chain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer; (iii) a first light chain (LI), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein LI and the anti-CD3 Fab heavy chain of Hl associate toform an anti-CD3 Fab domain; and (iv) a second light chain (L2), comprising an anti-DLL3 Fab light chain, and wherein L2 and the anti-DLL3 Fab heavy chain of H2 associate to form an anti- DLL3 Fab domain.

[0299] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 Fab heavy chain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer; (iii) a first light chain (LI), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having the sequence set forth in SEQ ID NO: 13, and wherein LI and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain; and (iv) a second light chain (L2), comprising an anti-DLL3 Fab light chain, and wherein L2 and the anti-DLL3 Fab heavy chain of H2 associate to form an anti- DLL3 Fab domain.

[0300] In certain embodiments, a trivalent and trispecific Het-Fab antibody construct of the present disclosure comprises: (i) a first Fab domain that binds CD3 comprising a first Fab heavy chain comprising a VH sequence comprising a HCDR1 sequence as set forth in SEQ ID NO: 49, a HCDR2 sequence as set forth in SEQ ID NO: 51, and a HCDR3 sequence as set forth in SEQ ID NO: 52, and a first Fab light chain comprising a VL domain comprising a LCDR1 sequence as set forth in SEQ ID NO: 53, aLCDR2 sequence as set forth in SEQ ID NO: 54, and aLCDR3 sequence as set forth in SEQ ID NO: 55, as identified according to IMGT numbering; (ii) an scFv domain that binds CD28 comprising a VL sequence comprising a LCDR1 sequence as set forth in SEQ ID NO: 41, a LCDR2 sequence as set forth in SEQ ID NO: 44, and a LCDR3 sequence as set forth in SEQ ID NO: 48, and a VH sequence comprising a HCDR1 sequence as set forth in SEQ ID NO: 28, a HCDR2 sequence as set forth in SEQ ID NO: 31, and a HCDR3 sequence as set forth in SEQ ID NO: 35, as identified according to Kabat numbering; (iii) a second Fab domain that binds DLL3 comprising a second Fab heavy chain comprising a VH sequence comprising a HCDR1 sequence as set forth in SEQ ID NO: 90, a HCDR2 sequence as set forth in SEQ ID NO: 91, and a HCDR3 sequence as set forth in SEQ ID NO: 92, and a second Fab light chain comprising a VL domain comprising a LCDR1 sequence as set forth in SEQ ID NO: 93, a LCDR2 sequence as set forth inSEQ ID NO: 94, and a LCDR3 sequence as set forth in SEQ ID NO: 95, as identified according to Kabat numbering; and (iv) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (1) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide via the C-terminus of the first Fab heavy chain, (2) the scFv domain is coupled to the C-terminus of the first Fab light chain, and (3) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide via the C-terminus of the second Fab heavy chain.

[0301] In certain embodiments, a trivalent and trispecific Het-Fab antibody construct of the present disclosure comprises: (i) a first Fab domain that binds CD3 comprising a first Fab heavy chain comprising a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 2, coupled to a first CHI sequence, and a first Fab light chain comprising a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 5, coupled to a first CL sequence; (ii) an scFv domain that binds CD28 comprising a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 16, coupled to a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 15; (iii) a second Fab domain that binds DLL3 comprising a second Fab heavy chain comprising a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 99, coupled to a second CHI sequence, and a second Fab light chain comprising a VL sequence comprising the amino acid sequence set forth in SEQ ID NO: 100, coupled to a first CL sequence; and (iv) an Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein (1) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide via the C-terminus of the first Fab heavy chain, (2) the scFv domain is coupled to the C-terminus of the first Fab light chain, and (3) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide via the C-terminus of the second Fab heavy chain.

[0302] In various embodiments, the present disclosure relates to trivalent and trispecific scFv2antibody constructs. Such scFv2antibody constructs can comprise three antigen binding domains, two of which are scFv domains, thus the descriptor “scFv2”.

[0303] In certain embodiments, the trivalent and trispecific scFv2antibody constructs described herein are capable of binding DLL3 on a target cell (e.g., a tumor cell).

[0304] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) a Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domaincomprising 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 C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide.

[0305] In some embodiments herein, a trivalent and trispecific scFv2antibody construct is capable of monovalently engaging the following three antigens: CD3, CD28 and DLL3.

[0306] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) the Fab domain, wherein the Fab domain is capable of binding CD3 on a cytotoxic effector cell; (ii) the first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) the second scFv domain, wherein the second scFv domain is capable of binding DLL3 on a tumor cell; and (iv) the dimeric Fc domain comprising the first Fc polypeptide and the 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 C-terminus of the Fab light chain N- terminus of the second Fc polypeptide, and (c) the second scFv domain is coupled to the N- terminus of the second Fc polypeptide.

[0307] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure that is capable of binding CD3, CD28 and DLL3 comprises or consists of three polypeptide chains, two immunoglobulin heavy chains (e.g., Hl and H2) and one immunoglobulin light chain (e.g., L). In some embodiments, the three polypeptide chains associate, either covalently and / or non-covalently, to form the trivalent and trispecific scFv2antibody construct.

[0308] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises a heterodimeric Fc domain. In such embodiments, Hl and / or H2 can comprise one or more amino acid modifications that promote formation of the heterodimer Hl -H2 compared to formation of respective homodimers Hl -Hl or H2-H2. In some of these embodiments, and as further described herein, Hl and H2 each comprise a CH3 sequence, wherein each CH3 sequence comprises one or more amino acid substitutions, relative to a respective wildtype CH3 sequence, that promote heterodimer H1-H2 formation compared to homodimer formation.

[0309] In some embodiments of a trivalent and trispecific scFv2antibody construct, the CH3 sequence of one of the Fc polypeptides comprises a set of amino acid substitutions selected from the group consisting of: L351Y_F405A_Y407V, T350V_L351Y_F405A_Y407V andT350V_L351Y_S400E_F405A_Y407V, and the CH3 sequence of 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 polypeptides is according to the EU numbering system.

[0310] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 scFv domain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer H1-H2; and (iii) a light chain (L), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein L and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain.

[0311] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 scFv domain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having the sequence set forth in SEQ ID NO: 12, and wherein Hl and H2 associate to form a heterodimer H1-H2; and (iii) a light chain (L), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having the sequence set forth in SEQ ID NO: 13, and wherein L and the anti-CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain.

[0312] In certain embodiments, described herein is an antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; (iii) a third binding domain capable of binding a TAA on a tumor cell, wherein the third binding domain comprises or consists of a second Fabdomain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain. In some embodiments, the third binding domain comprises or consists of a second Fab domain. In other embodiments, the third binding domain comprises or consists of a second scFv domain. The dimeric Fc domain can be a heterodimeric Fc domain, as further described herein. FIGS. 1G (e.g., third binding domain is a second scFv domain) and 1H (e.g., third binding domain is a second Fab domain) show the format and geometry of such trivalent and trispecific antibody constructs, according to certain embodiments of this disclosure.

[0313] In some embodiments, such trivalent and trispecific antibody construct comprises a light chain having the following domain structure, from N- to C-terminus, of VL-CL-scFv, wherein the scFv domain can comprise, from N- to C-terminus, a VH coupled to a VL sequence, or a VL sequence coupled to a VH sequence. Thus, in certain embodiments, such Fab light chain has the domain structure, from N- to C-terminus, of (VL-CL)Fab-(VL-VH)SCFv. In other embodiments, the light chain has the domain structure, from N- to C-terminus, of (VL-CL)Fab-(VH-VL)SCFv.

[0314] In certain embodiments, described herein is a trivalent and trispecific antibody construct comprising a light chain having the following domain structure, from N- to C-terminus, of (VL- CL)Fab-(VL-VH)scFv. In some embodiments, such light chain can further comprise one or more linkers, as further described herein. In some embodiments, the light chain comprises a peptide linkerFab'scFvbetween the Fab portion and the scFv portion which couples the Fab portion to the scFv portion to yield the light chain with the domain structure (VL-CL)Fab-LinkerFab'scFv-(VL- VH)SCFV. The linkerFab'scFvcan comprise or consist of the amino acid sequence set forth in SEQ ID NO: 58. As further described herein, the scFv domain of the light chain can also comprise a linkerscFvthat coupled the VL sequence to the VH sequence. Thus, in certain embodiments, the trivalent and trispecific antibody construct comprises a light chain with the domain structure (VL- CL)Fab-LinkerFab'scFv-(VL-LinkerscFv-VH)scFv. The linkerscFvcan comprise or consist of the amino acid sequence set forth in SEQ ID NO: 59.

[0315] In some embodiments, described herein is a trivalent and trispecific antibody construct comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell and comprising a VH domain comprising a HCDR1 having the sequence GVTFNYYG (SEQ ID NO:49), a HCDR2 having the sequence ITRSGGRI (SEQ ID NO: 51), and a HCDR3 having the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 having the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 having the sequence RND, and a LCDR3 having the sequence QSYSSGFI (SEQ ID NO: 55), according to the IMGT numbering system.; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell and comprising a VH domain comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL domain comprising a LCDR1 having the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), according to the Kabat numbering system.; (iii) a third binding domain capable of binding DLL3 on a tumor cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain. In some embodiments, the third binding domain is a second Fab domain. In other embodiments, the third binding domain is a second scFv domain.

[0316] In some embodiments, described herein is a trivalent and trispecific antibody construct comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell and comprising a VH domain comprising a HCDR1 having the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 having the sequence ITRSGGRI (SEQ ID NO: 51), and a HCDR3 having the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 having the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 having the sequence RND, and a LCDR3 having the sequence QSYSSGFI (SEQ ID NO: 55), according to the IMGT numbering system.; (ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell and comprising a VH domain comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL domain comprising a LCDR1 having the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), according to the Kabat numbering system.; (iii) a third binding domain capable of binding DLL3 on a tumorcell and comprising a VH domain comprising a HCDR1 having the sequence HYYVH (SEQ ID NO: 90), a HCDR2 having the sequence IIDPGGGTTSYAQKFLG (SEQ ID NO: 91) and a HCDR3 having the sequence GERVTGNYFYYGMDV (SEQ ID NO: 92), and a VL domain comprising a LCDR1 having the sequence VANQGISNYLV (SEQ ID NO: 93), a LCDR2 having the sequence SVNSLYS (SEQ ID NO: 94) and a LCDR3 having the sequence LQHDSYPYT (SEQ ID NO: 95), according to the Kabat numbering system; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C- terminus of the light chain of the first Fab domain. In some embodiments, the third binding domain is a second Fab domain. In other embodiments, the third binding domain is a second scFv domain.

[0317] In certain embodiments, described herein is a trivalent and trispecific antibody construct comprising a light chain comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 13. In some embodiments, the light chain of a trivalent and trispecific antibody construct can comprise an amino acid sequence having the amino acid sequence set forth in SEQ ID NO: 13.

[0318] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first immunoglobulin (Ig) heavy chain (Hl) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105, a second Ig heavy chain (H2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106, a first light chain (LI) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 107, and a second light chain (L2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 108.

[0319] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 105, a second Ig heavy chain (H2) comprising the amino acid sequence set forth in SEQ ID NO: 106, a first light chain (LI) comprising the amino acid sequence set forth in SEQ IDNO: 107, and a second light chain (L2) comprising the amino acid sequence set forth in SEQ ID NO: 108.

[0320] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first immunoglobulin (Ig) heavy chain (Hl) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 105, a second Ig heavy chain (H2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 115, a first light chain (LI) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 107, and a second light chain (L2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 116.

[0321] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 105, a second Ig heavy chain (H2) comprising the amino acid sequence set forth in SEQ ID NO: 115, a first light chain (LI) comprising the amino acid sequence set forth in SEQ ID NO: 107, and a second light chain (L2) comprising the amino acid sequence set forth in SEQ ID NO: 116.

[0322] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first immunoglobulin (Ig) heavy chain (Hl) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 117, a second Ig heavy chain (H2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 106, a first light chain (LI) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 107, and a second light chain (L2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 118.

[0323] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 117, a second Ig heavy chain (H2) comprising the amino acid sequence set forth inSEQ ID NO: 106, a first light chain (LI) comprising the amino acid sequence set forth in SEQ ID NO: 107, and a second light chain (L2) comprising the amino acid sequence set forth in SEQ ID NO: 118.

[0324] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 109, a second Ig heavy chain (H2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 110, and a light chain (L) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 111.

[0325] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 109, a second Ig heavy chain (H2) comprising the amino acid sequence set forth in SEQ ID NO: 110, and a light chain (L) comprising the amino acid sequence set forth in SEQ ID NO: 111.

[0326] In some embodiments, a trivalent and trispecific scFv2antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 112, a second Ig heavy chain (H2) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 113, and a light chain (L) comprising an amino acid sequence having at least about 80%, 85%, 90%, 95%, 97%, 99%, or 100% sequence identity to the amino acid sequence set forth in SEQ ID NO: 114.

[0327] In some embodiments, a trivalent and tri specific Het-Fab antibody construct of the present disclosure comprises a first Ig heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 112, a second Ig heavy chain (H2) comprising the amino acid sequence set forth in SEQ ID NO: 113, and a light chain (L) comprising the amino acid sequence set forth in SEQ ID NO: 114.

[0328] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure can further comprise a C-terminal lysine (K) residue in the sequence of one or both of its heavy chain(s).F. Certain Properties of Costimulatory DLL3-Tar eting Antibody Constructs

[0329] The trivalent and trispecific antibody constructs of the present disclosure (e.g., Het-Fab as well as scFv2antibody constructs as described herein) can have several specific properties due to their specific combination of format, geometry and antigen affinities. In some embodiments, a trivalent and trispecific antibody construct herein comprises a light chain comprising a Fab portion targeting CD3 and an scFv portion targeting CD28 and thus can possess certain properties that can be unique to a construct with that specified light chain format and geometry.

[0330] In some embodiments, the engagement of a herein disclosed trivalent and trispecific antibody construct with two different antigens, CD3 and CD28, on one or more T cell(s) and with DLL3 on a tumor cell, e.g., in a tumor (micro)environment, can be - at least temporarily - simultaneous, thereby establishing a TCR-independent immune synapse, and directing T cell- mediated cytotoxic activity to a tumor environment which contains tumor cells expressing DLL3. In various embodiments, and as further described herein, a trivalent and trispecific antibody construct may cause a significantly reduced immune cell (e.g., T cell) activation in the absence of DLL3, e.g., when the immune synapse cannot be fully formed due to an absence of DLL3. Such property can be advantageous over those of conventional constructs as it allows, for example, activation of a subject’s immune system in a more TAA (e.g., DLL3)-dependent manner, and thus may cause less off-target effects in the subject compared to conventional constructs or antibodies that act in a less DLL3 -dependent manner.

[0331] In some embodiments, the antibody constructs described herein can exhibit an enhanced anti-tumor activity in tumors that have a relatively low T cell infiltration when compared to conventional constructs that target only one immune cell antigen (e.g., CD3 or CD28, but not both), at least in part, due to their co-stimulatory activity by being able to engage both CD3 and CD28 on the same immune cell (e.g., T cell).

[0332] In various embodiments, binding of a trivalent and trispecific antibody construct of the present disclosure to CD3 comprises or consists of binding of such construct to CD3e, i.e., the epsilon subunit of the T cell receptor complex.

[0333] Generally, and as it may be appreciated by a person of ordinary skill in the art, natural T cell activation can require both TCR (e.g., involving the CD3 complex) and CD28 co-stimulation. The antibody constructs of the present disclosure have been specifically designed, e.g., through their format, geometry and antigen affinities, to provide both CD3 and CD28 co-stimulation. Furthermore, and according to various embodiments of the present disclosure, the anti-CD3 and anti-CD28 binding affinities of the anti-CD3 and anti-CD28 binding domains (e.g., scFv(s), Fab(s), etc.) of antibody constructs described herein, in combination with their relative positioning within the construct, have been specifically selected and engineered to generate signals for immune cell (e.g., T cell) activation with appropriate strength to reduce both T cell anergy on the one side and T cell overreaction and dysfunction on the other side of the spectrum, and to provide an improved ratio of anti-tumor, on-target to healthy tissue, off-target activity. In various embodiments, such improved on-target-to-off-target activities can be achieved by optimizing the format and geometry of the antibody constructs in a way such that engagement of all three antigens, namely CD3, CD28 and DLL3, at the same time allows for the most potent anti -turn or activity, compared to instances in which, e.g., only one of the immune cell antigens CD3 and / or CD28 are engaged and bound by the construct.

[0334] As further described herein, an antibody construct of the present disclosure can be trivalent and trispecific and bind each antigen, e.g., CD3, CD28, DLL3, monovalently via one of its three antigen binding domains.

[0335] In some embodiments, a trivalent and trispecific antibody construct has a melting temperature at Tml, Tm2, and / or Tm3 that is within 10 °C, 5 °C, 2 °C, or within 1 °C degree of that of a bivalent and monospecific (e.g., monospecific for either CD3, CD28 or DLL3) IgGl monoclonal antibody, as determined using, e.g., differential scanning calorimetry (DSC) or differential scanning fluorometry (DSF). In some embodiments, such bivalent and monospecific IgGl monoclonal antibody can be any conventional IgGl antibody capable of binding a specific antigen. In certain embodiments, such monoclonal, monospecific and bivalent IgGl reference antibody is one that comprises two Fab domains as described herein in the context of trispecific antibody constructs. In some embodiments, the bivalent and monospecific IgGl monoclonal antibody comprises two anti-CD3 Fab domains as described herein, two anti-CD28 Fab domains as described herein, or two anti-DLL3 Fab domains as described herein.

[0336] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure binds a cytotoxic effector cell (e.g., a T cell) that expresses CD3 and CD28 with an affinity (i.e., a KD value) from about 5 nM to about 100 pM, from about 1 nM to about 100 pM, from about 1 nM to about 250 pM, from about 1 nM to about 500 pM, or from about 1 nM to about 750 pM. In certain embodiments, such antibody construct binds the cytotoxic effector cell that expresses CD3 and CD28 with an affinity that is about 2-fold, 5-fold, 10-fold, 20-fold, 50-fold, 100-fold, or about 200-fold higher, and / or from about 2-fold to about 200-fold, from about 2-fold to about 150-fold, from about 2-fold to about 100-fold, or from about 20-fold to about 200-fold higher than that of a corresponding bispecific anti-CD3xDLL3 and / or anti-CD28xDLL3 antibody construct. As described herein, KD values can be determined using methods known in the art, e.g., surface plasmon resonance (SPR).

[0337] In some embodiments, using the engineered anti-CD28 binding domains with reduced CD28 affinity, e.g., compared to the parental huTN228 paratope, can result in trivalent and trispecific antibody constructs that may induce less CD28-mediated toxicities.

[0338] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure exhibits an IC50 value from about 50 pM to about 0.01 pM, from about 25 pM to about 0.01 pM, from about from about 10 pM to about 0.05 pM, from about 10 pM to about 0.1 pM, from about 10 pM to about 1 pM, from about 5 pM to about 1 pM for killing TAA-expressing tumor cells that express at least about 100,000 DLL3 / cell by TDCC in the presence of the cytotoxic effector cell and using an E:T ratio of 2: 1 and an incubation period of 72 hours. In some of such embodiments, the antibody construct achieves a maximum killing of DLL3 -expressing tumor cells of at least about 60%, 65%, 70%, 75%, or 80%, 85%, or 90%, 100%, or from about 60% to about 100%, from about 70% to about 90%, or from about 75% to about 85%.

[0339] Furthermore, in various embodiments, the trivalent and trispecific antibody constructs of the present disclosure may provide a (e.g., strictly) target cell (e.g., tumor cell) dependent cytotoxicity profile, e.g., as shown herein when cytokine release is significantly reduced in the presence of only isolated T cells compared to conditions in which the T cells are in co-culture with DLL3 -expressing tumor cells.

[0340] In some embodiments, the trivalent and trispecific antibody constructs disclosed herein can have a thermal stability when measured at 40 °C and over a time period of about 2, 3, 5, 7, 10, or 14 days of at least about 90%, 95%, 97%, 98% or 99%, i.e., 90%, 95%, 97%, 98% or 99% of intactconstruct is when measured using, e.g., size-exclusion chromatography, or other methods known in the art. In certain embodiments, such constructs can comprise one Fab domain capable of binding CD3, and two scFv domains, wherein one such scFv domain is capable of binding either CD28, and one scFv domain is capable of binding DLL3. In other embodiments, such constructs can comprise two Fab domains, one of which is capable of binding CD3 and the other is capable of binding DLL3, and one scFv domain capable of binding CD28. In some embodiments, the stability of a trispecific and trivalent antibody construct over a period of 14 days at 40 °C is at least about 97% or 98% percent (i.e., at least about 97% or 98% of construct is intact as measured, e.g., using size-exclusion chromatography). In some embodiments, the concentration of the construct in such stability experiment is about 1 mg / mL.

[0341] In various embodiments, and in order to address certain treatment challenges with known and conventional therapeutic modalities, e.g., durability of response and sustained T cell activation, the herein disclosed trivalent and trispecific co-stimulatory T cell engager antibody constructs have been rationally designed to, e.g., optimally engage CD3 and CD28 and redirect and enhance cytotoxic T cell responses to DLL3 -expressing tumor cells while maintaining an improved safety profile when compared to conventional therapeutic modalities.

[0342] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure can mediate a sustained T cell-mediated cytotoxicity toward tumor cells over multiple cycles of T cell stimulation. In some embodiments, such trispecific antibody construct can be used to stimulate a T cell population about 4, 5 or 6 times over a 14-day period (e.g., first day of stimulation is on day 0) and achieve anti-tumor cell cytotoxicity of at least about 70%, 75%, 80%, 85%, or 90% and T cell viability of at least about 60%, 65%, 70%, 75%, or 80% as measured at certain time points over the 14-day period.

[0343] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure can exhibit conditional binding to CD28, i.e., binds CD28 with a significantly lower dissociation constant (i.e., higher affinity) when co-engaged with CD3, and compared to binding of CD28 alone, e.g., in the absence of CD3 (e.g., when using a CD3 knock-out cell line). In some embodiments, a trivalent and trispecific antibody construct of the present disclosure binds CD28 with an at least about 2-fold, 3-fold, 5-fold, 7-fold, 8-fold, or at least about 10-fold increased affinity (e.g., expressed in change in KD) when co-engaged with CD3, and when compared to the binding of CD28 in the absence of CD3.

[0344] In some embodiments, a trivalent and trispecific antibody construct of the present disclosure has the ability to bind CD3 and CD28 on the same T cell, which can also be referred to herein as “cis” (on the same cell) T cell binding. Hence, in various embodiments, a trivalent and trispecific antibody construct that comprises a light chain comprising a Fab portion and an scFv portion does not bridge two T cells by, e.g., binding CD3 on one T cell and CD28 on another T cell, which can also be referred to herein as “trans” (on different cells) T cell binding.III. SEQUENCE IDENTITY OF AMINO ACID AND NUCLEIC ACID SEQUENCES

[0345] As described in other parts of this disclosure, certain embodiments herein relate to an isolated polypeptide or a set of isolated polypeptides (e.g., polypeptide chains Hl, H2, LI, L2 or Hl, H2 and L etc., or portions, e.g., domains, thereof) of a trivalent and trispecific antibody construct, as well as to a polynucleotide or a set of polynucleotides encoding the one or more polypeptide chains of an antibody construct described herein. A polynucleotide in this context can encode all or part of an antibody construct, such as one or more polypeptide chains (e.g., Hl, H2, LI, L2 etc.) of an antibody construct.

[0346] In some embodiments, described herein is a nucleic acid molecule or a set of nucleic acid molecules encoding one or more, two or more, three or more, or four or more polypeptide chains that form any of the trivalent and trispecific antibody constructs disclosed herein.

[0347] The terms “nucleic acid,” “nucleic acid molecule” and “polynucleotide” are used interchangeably herein and refer to a polymeric form of nucleotides of any length, either deoxyribonucleotides or ribonucleotides, or analogues thereof. Non-limiting examples of polynucleotides include a gene, a gene fragment, messenger RNA (mRNA), cDNA, recombinant polynucleotides, plasmids, vectors, isolated DNA of any sequence, isolated RNA of any sequence, nucleic acid probes, and primers.

[0348] In some embodiments, described herein is a vector or a set of vectors comprising the nucleic acid molecule or the set of nucleic acid molecules that encode the one or more polypeptide chains (e.g., one or more of Hl, H2, LI, L2 etc.) of an antibody construct disclosed herein.

[0349] A polynucleotide that “encodes” a given polypeptide is a polynucleotide that is transcribed (in the case of DNA) and translated (in the case of mRNA) into a 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' (carboxy) terminus. A transcription termination sequence can be located 3' to the coding sequence.

[0350] In certain embodiments, the present disclosure relates to polynucleotide and / or polypeptide sequences that are identical or substantially identical to another polynucleotide and / or polypeptide sequence. The term “identical,” in the context of two or more polynucleotide or polypeptide sequences, refers to two or more sequences or subsequences that are the same, i.e., have the identical sequence of nucleotide or amino acid monomers (i.e., 100% sequence identity), respectively. Polypeptide or polynucleotide sequences herein share “sequence identity” if they have a percentage or a certain number of amino acid residues or nucleotides, respectively, that are at least about 80%, about 85%, about 90%, about 95%, about 97%, or at least about 99% identity over a specified region when compared and aligned for maximum correspondence over a comparison window or over a designated region as measured using one of the commonly used sequence comparison algorithms as known to persons of ordinary skill in the art or by manual alignment and visual inspection. This definition also refers to the complement of a test polynucleotide sequence. The identity can exist over a region that is at least about 50 amino acids or nucleotides in length, or over a region that is from about 75 to about 100 amino acids or nucleotides in length, or, where not specified, across the entire sequence of a polypeptide or polynucleotide. For sequence comparison, typically test sequences are compared to a designated reference sequence. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters can be used, or alternative parameters can be designated. The sequence comparison algorithm then calculates the percent (%) sequence identities for the test sequences relative to the reference sequence, based on the program parameters.

[0351] The term “comparison window,” as used herein, refers to a segment of a sequence comprising contiguous amino acid or nucleotide positions which can be from about 20 to about 1000 contiguous 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 contiguous amino acid or nucleotide positions over which a test sequence can be compared to a reference sequence of the same number of contiguous positions after the two sequences are optimally aligned. Longer segments up to and including the full-length sequence may also be used as a comparison window in certain embodiments. Methods of alignment of sequences for comparison are known to those of ordinary skill in the art. Optimal alignment of sequences for comparison can be conducted, for example, bythe local homology algorithm of Smith & Waterman, 1970, Adv. Appl. Math., 2:482c; by the homology alignment algorithm of Needleman & Wunsch, 1970, J. Mol. Biol., 48:443; by the search for similarity method of Pearson & Lipman, 1988, Proc. Natl. Acad. Sci. USA, 85:2444, or by computerized implementations of these algorithms (for example, GAP, BESTFIT, FASTA or TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, Madison, WI), or by manual alignment and visual inspection (see, for example, Ausubel et al., Current Protocols in Molecular Biology, (1995 supplement), Cold Spring Harbor Laboratory Press). Examples of available algorithms suitable for determining percent sequence identity are the BLAST and BLAST 2.0 algorithms, which are described in Altschul et al., 1997, Nuc. Acids Res., 25:3389- 3402, and Altschul et al., 1990, J. Mol. BioL, 215:403-410, respectively. Software for performing BLAST analyses is publicly available through the website for the National Center for Biotechnology Information (NCBI).

[0352] Certain embodiments described herein relate to variant sequences (e.g., variant VH domains, variant Fc polypeptides, etc.) that comprise one or more amino acid modification, e.g., one or more amino acid insertions, one or more amino acid deletions, and / or one or more amino acid substitutions, when compared to, e.g., a reference such as a wildtype sequence (e.g., a sequence that is disclosed in a known reference). In certain embodiments, the one or more amino acid modification of a variant sequence comprises one or more amino acid substitutions when compared to a reference such as a wildtype sequence. In some embodiments, the one or more amino acid substitutions are one or more non-conservative substitutions. In other embodiments, the one or more amino acid substitutions are one or more conservative substitutions.

[0353] In general, a “conservative substitution,” as used herein, is considered to be a substitution of one amino acid with another amino acid having similar physical, chemical and / or structural properties. Common conservative substitutions are listed under Column 1 of TABLE 4.TABLE 4: Conservative Amino Acid Substitutions

[0354] One skilled in the art will appreciate that the main factors in determining what constitutes a conservative substitution are usually the size of the amino acid side chain and its physical / chemi cal properties, but that certain environments allow for substitution of a given amino acid with a broader range of amino acids than those listed in Column 1 of TABLE 4. These additional amino acids tend to either have similar properties to the amino acid being substituted but to vary more widely in size or be of similar size but vary more widely in phy si cal / chemi cal properties. This broader range of conservative substitutions is listed under Column 2 of TABLE 4. The skilled person can readily ascertain the most appropriate group of substituents to select from in view of the particular protein environment in which the amino acid substitution is being made.IV. PHARMACEUTICAL COMPOSITIONS

[0355] In certain embodiments, the present disclosure relates to pharmaceutical compositions that can comprise one or more of the trivalent and trispecific antibody constructs described herein. Invarious embodiments, a pharmaceutical composition herein can further comprise a pharmaceutically acceptable excipient, carrier, buffer, stabiliser, or other materials well known to those skilled in the art. Such materials are generally non-toxic and do not interfere with the efficacy of the active ingredient (i.e., antibody construct). The precise nature of a carrier or other material can depend on the route of administration. Hence, a pharmaceutical composition herein can be formulated for various used and administration routes, e.g., for oral, intravenous, cutaneous, subcutaneous, nasal, intramuscular, or intraperitoneal administration routes.

[0356] A pharmaceutical composition for oral administration can be in tablet, capsule, powder, or liquid form. A tablet can include a solid carrier such as gelatin or an adjuvant. Liquid pharmaceutical compositions generally include a liquid carrier such as water, petroleum, animal or vegetable oils, mineral oil, or synthetic oil. Physiological saline solution, dextrose or other saccharide solution or glycols such as ethylene glycol, propylene glycol or polyethylene glycol can be included.

[0357] For intravenous, cutaneous, or subcutaneous injection, or injection at the site of affliction (e.g., at a tumor site), the active ingredient (i.e., antibody construct) can be in the form of a parenterally acceptable aqueous solution which is pyrogen-free and has suitable pH, isotonicity and stability. Those of relevant skill 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, stabilisers, buffers, antioxidants and / or other additives can be included, as required.

[0358] For antibody constructs according to the present disclosure that are administered to a subject, administration is preferably in a “therapeutically effective amount” that is sufficient to show benefit to the individual, as further described herein. The actual amount administered, and rate and time-course of administration, can depend on the nature and severity of the disease (e.g., cancer) being treated. Prescription of treatment, e.g., decisions on dosage etc., is within the responsibility of general practitioners and other medical doctors, and typically takes account of the disorder to be treated, the condition of the individual patient, the site of delivery, the method of administration and other factors known to practitioners. Examples of the techniques and protocols mentioned above can be found in Remington's Pharmaceutical Sciences, 16th edition, Osol, A. (ed), 1980.

[0359] In some embodiments, a pharmaceutical composition can comprise a second active ingredient (e.g., another protein or small molecule) in addition to an antibody construct described herein.

[0360] Hence, also described herein is a pharmaceutical composition comprising any one or more of the trivalent and trispecific antibody construct(s) disclosed herein, and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.V. KITS

[0361] The present disclosure also describes kits comprising one or more of the trivalent and trispecific antibody constructs described herein, or a pharmaceutical composition as described herein and that comprises such antibody construct(s), as well as instructions for use. Thus, in certain embodiments, described herein are kits comprising vectors for expressing an antibody construct described herein and instructions for use. In certain embodiments, described herein are kits comprising host cells comprising a vector for expressing an antibody construct and instructions for use. In some embodiments, the present disclosure relates to kits comprising a purified antibody construct and instructions for use. The purified antibody construct can be lyophilized or provided in a dry form, such as a powder or granules, and the kit can additionally contain a suitable solvent for reconstitution of the lyophilized or dried component(s).

[0362] A kit can further comprise a container and a label and / or package insert on or associated with the container. The label or package insert contains instructions customarily included in commercial packages of therapeutic products, providing information or instructions about the indications, usage, dosage, administration, contraindications and / or warnings concerning the use of such therapeutic products (e.g., an antibody construct described herein). The label or package insert can further include a notice in the form prescribed by a governmental agency regulating the manufacture, use or sale of pharmaceuticals or biological products, which notice reflects approval by the agency of manufacture, for use or sale for human or animal administration. The container can hold a composition comprising an antibody construct of this disclosure. In some embodiments, the container can have a sterile access port. For example, the container can be an intravenous solution bag or a vial having a stopper that can be pierced by a hypodermic injection needle.

[0363] In addition to the container containing a composition comprising an antibody construct, the kit can further comprise 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 can be included in such kit.

[0364] Suitable containers can include, for example, bottles, vials, syringes, intravenous solution bags, and the like. The containers can be formed from a variety of materials such as glass or plastic. If appropriate, one or more components (e.g., an antibody construct) of the kit can be lyophilized or provided in a dry form, such as a powder or granules, and the kit can additionally contain a suitable solvent for reconstitution of the lyophilized or dried component(s).

[0365] A kit herein can further include other materials desirable from a commercial or user standpoint, such as filters, needles, and syringes.VI. METHODS

[0366] Further described herein are methods of producing and using the trivalent and trispecific antibody constructs of the present disclosure.A. Methods of Producing an Antibody Construct

[0367] In some embodiments, the present disclosure relates to methods for preparing the trivalent and trispecific antibody constructs described herein. In various embodiments, an antibody construct of the present disclosure can be produced using standard recombinant methods known in the art (see, for example, U.S. Patent No. 4,816,567 and “Antibodies: A Laboratory Manual ” 2ndEdition, Ed. Greenfield, Cold Spring Harbor Laboratory Press, New York, 2014).

[0368] For recombinant production of an antibody construct described herein, a polynucleotide or set of polynucleotides encoding the antibody construct can be generated and inserted into one or more vectors for further cloning and / or expression in a host cell. Polynucleotide(s) encoding the antibody construct can be produced by standard methods known in the art (see, for example, Ausubel el al., Current Protocols in Molecular Biology, John Wiley & Sons, New York, 1994 & update, and “Antibodies: A Laboratory Manual,” 2ndEdition, Ed. Greenfield, Cold Spring Harbor Laboratory Press, New York, 2014). As would be appreciated by one of skill in the art, the number of polynucleotides required for expression of the antibody construct may be dependent on the format and / or geometry of the antibody construct, including, for example, the number of polypeptide chains that the antibody construct is comprised of. For example, when an antibody construct comprises three polypeptide chains (e.g., Hl, H2 and L), three polynucleotides each encoding one polypeptide chain can be used, and when an antibody construct comprises four polypeptide chains (e.g., Hl, H2, LI and L2), four polynucleotides each encoding one polypeptidechain can be used. In embodiments in which two or more polynucleotides are used, such two or more polynucleotides can be incorporated into one vector or into more than one vector (e.g., two or three separate vectors).

[0369] Generally, for expression, the polynucleotide or set of polynucleotides encoding an antibody construct herein can be incorporated into an expression vector together with one or more regulatory elements, such as transcriptional elements, which can be used for efficient transcription of the polynucleotide(s). Examples of such regulatory elements include, but are not limited to, promoters, enhancers, terminators, and polyadenylation signals. One skilled in the art will appreciate that the choice of regulatory elements can be dependent on the host cell selected for expression of the polypeptides of the antibody construct and that such regulatory elements can be derived from a variety of sources, including bacterial, fungal, viral, mammalian or insect genes. The expression vector can optionally further contain heterologous nucleic acid sequences that facilitate expression or purification of the expressed protein. Examples include, but are not limited to, signal peptides and affinity tags such as metal-affinity tags, histidine tags, avi din / streptavidin encoding sequences, glutathione-S-transferase (GST) encoding sequences and biotin encoding sequences. The expression vector can be an extrachromosomal vector or an integrating vector. Hence, in some embodiments, the amino acid sequences of the polypeptide chains of an expressed antibody construct described herein, e.g., chains Hl, H2, LI and / or L2 etc., can comprise a signal peptide sequence. Such signal peptide sequences may vary depending on the expression system and conditions used for producing an antibody construct. Exemplary signal peptide sequences can comprise the amino acid sequence METDTLLLWVLLLWVPGSTG (SEQ ID NO: 79) or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 80), e g., for Hl, H2, etc., or MRPTWAWWLFLVLLLALWAPARG (SEQ ID NO: 80) or MGWSCIILFLVATATGVHS (SEQ ID NO: 81), e.g., for LI, L2, etc. In certain embodiments, one or more heavy chains (e.g., Hl, H2, etc.) of an antibody construct described herein can comprise a C-terminal lysine residue following expression of the polypeptide chains inside the cell. In various embodiments, such C- terminal lysine residue may be enzymatically cleaved from the polypeptide chains prior to further processing (e.g., purification, formulation, etc.) and prior to use of the corresponding antibody construct, e.g., prior to administration of the construct to a subject in need thereof.

[0370] Certain embodiments for producing an antibody construct of the present disclosure relate to vectors (such as expression vectors) comprising one or more polynucleotides encoding at leasta portion of an antibody construct described herein. The polynucleotide(s) can be comprised by a single vector or by more than one vector. In some embodiments, the polynucleotides are comprised by a multi-cistronic vector. Expression vectors that can be used to express polynucleotides include but are not limited to pTT5 and pUC15 cells comprising vectors encoding an antibody construct.

[0371] Suitable host cells for cloning or expression of the 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, E. coli, A. salmonicida or B. subtilis cells. In certain embodiments, an antibody construct can be produced in bacteria, in particular when glycosylation and Fc effector function are not needed or desired for the indented purpose of the antibody construct, as described for example in U.S. Patent Nos. 5,648,237, 5,789,199, and 5,840,523, and in Charlton, Methods in Molecular Biology, Vol. 248, pp. 245-254, B.K.C. Lo, ed., Humana Press, Totowa, N.J., 2003. Eukaryotic microbes such as filamentous fungi or yeast are suitable expression host cells in certain embodiments, in particular fungi and yeast strains whose glycosylation pathways have been “humanized” resulting in the production of an antibody with a partially or fully human glycosylation pattern (see, for example, Gerngross, 2004, Nat. Biotech. 22: 1409-1414, and Li et al., 2006, Nat. Biotech. 24:210-215).

[0372] Suitable host cells for the expression of glycosylated antibody constructs are, in various embodiments, eukaryotic cells. For example, U.S. Patent Nos. 5,959,177, 6,040,498, 6,420,548, 7,125,978 and 6,417,429 describe PLANTIBODIES™ technology for producing antibodies and portions thereof (e.g., scFv(s), Fab(s), etc.) in transgenic plants. Mammalian cell lines adapted to grow in suspension are particularly useful for the expression of 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, for example, Graham et al., 1977, J. Gen Virol., 36:59), baby hamster kidney cells (BHK), mouse sertoli TM4 cells (see, for example, 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 (BRL 3 A), human lung cells (W138), human liver cells (Hep G2), mouse mammary tumour (MMT 060562), TRI cells (see, for example, Mather et al., 1982, Annals N. Y. 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 celllines (such as YO, NSO and Sp2 / 0). Exemplary mammalian host cell lines suitable for production of antibodies are reviewed in Yazaki & Wu, Methods in Molecular Biology, Vol. 248, pp. 255-268 (B.K.C. Lo, ed., Humana Press, Totowa, N.J., 2003).

[0373] In certain embodiments, the host cell used to produce a trivalent and trispecific antibody construct 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, NSO or COS cell. In some embodiments, the host cell is a stable cell line that allows for mature glycosylation of the antibody construct.

[0374] The host cells comprising the expression vector(s) encoding the antibody construct can be cultured using routine methods to produce the antibody construct. Alternatively, in some embodiments, host cells comprising the expression vector(s) encoding the antibody construct can be used therapeutically or prophylactically to deliver the antibody construct to a subject, or polynucleotides or expression vectors can be administered to a cell from a subject ex vivo and the cell then returned to the body of the subject.

[0375] In some embodiments, a host cell comprises (for example, has been transformed with) a vector comprising a polynucleotide that encodes a VL and a VH of a binding domain of an antibody construct described herein. In some embodiments, a host cell comprises (for example, has been transformed with) a vector comprising a polynucleotide that encodes a full-length polypeptide chain of an antibody construct described herein, e.g., Hl, H2, LI and / or L2 as described herein. In another example, a host cell comprises a first vector comprising a polynucleotide that encodes the VL of a binding domain and a second vector comprising a polynucleotide that encodes the corresponding VH of the binding domain. In various embodiments, the host cell is eukaryotic, for example, a Chinese Hamster Ovary (CHO) cell, a human embryonic kidney (HEK) cell or a lymphoid cell (e.g., Y0, NSO, Sp20 cell). In certain embodiments, the host cell is Expi293™ (Thermo Fisher, Waltham, MA). In certain embodiments, the host cell used herein is CHO-S cells (National Research Council Canada) or HEK293 cells.

[0376] Certain embodiments of the present disclosure relate to a method of making an antibody construct comprising culturing a host cell into which one or more polynucleotides encoding the antibody construct, or one or more expression vectors encoding the antibody construct, have been introduced, under conditions suitable for expression of the antibody construct. Such method can further comprise recovering the antibody construct from the host cell (or from host cell culturemedium). In some embodiments, such method can further comprise purifying the antibody construct.

[0377] Cell culture media that can be used include, but are not limited to, DMEM (Thermo Fisher, Waltham, MA), Opti-MEM™ (Thermo Fisher, Waltham, MA), Opti-MEM™ I Reduced Serum Medium (Thermo Fisher, Waltham, MA), RPMI-1640 medium, Expi293™ Expression Medium (Thermo Fisher, Waltham, MA), and FreeStyle CHO expression medium (Thermo Fisher Scientific, Waltham, MA). The cell culture medium 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 antimycotics, e.g., amphotericin, or any other supplements routinely used in the to support cell culture.

[0378] In various embodiments, an antibody construct of the present disclosure is purified after expression. Proteins, such as an antibody construct of the present disclosure, can be isolated or purified in a variety of ways known to those skilled in the art (see, for example, Protein Purification: Principles and Practice, 3rdEd., Scopes, Springer-Verlag, NY, 1994). Standard purification methods that can be used for the antibody constructs disclosed herein include chromatographic techniques, including ion exchange, hydrophobic interaction, affinity, sizing or gel filtration, and reverse-phase, carried out at atmospheric pressure or at high pressure using systems such as FPLC and HPLC. Additional purification methods include electrophoretic, immunological, precipitation, dialysis and chromatofocusing techniques. Ultrafiltration and diafiltration techniques, in conjunction with protein concentration, can also be used. As is well known in the art, a variety of natural proteins bind Fc domains and other structural elements of an antibody construct, and, in some embodiments, these proteins can be used for purification of an antibody construct. For example, the bacterial proteins A and G can bind to the Fc domain of some antibody constructs. Likewise, the bacterial protein L can bind to the Fab domain of some antibody constructs. Purification can also be enabled by a particular fusion partner. For example, antibody constructs can be purified using glutathione resin if a GST fusion is employed, Ni+2affinity chromatography if a His-tag is employed or immobilized anti-flag antibody if a flag-tag is used. The degree of purification necessary may vary depending on the use of the antibody constructs. Hence, in some embodiments, no purification may be necessary.

[0379] In certain embodiments, an antibody construct of this disclosure is substantially pure. The term “substantially pure” (or “substantially purified”) when used in reference to an antibodyconstruct described herein, refers to an antibody construct as substantially or essentially free of components that normally accompany or interact with the protein as found in its naturally occurring environment, such as a native cell, or a host cell in the case of a recombinantly produced antibody construct. In certain embodiments, an antibody construct that is substantially pure is an antibody construct 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) of other contaminating protein species.

[0380] 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-high performance liquid chromatography-size exclusion chromatography (UPLC-SEC), High Performance Liquid Chromatography (HPLC), mass spectrometry, multi angle light scattering (MALS), and dynamic light scattering (DLS).

[0381] In certain embodiments, an antibody construct described herein can comprise one or more post-translational modifications. Such post-translational modifications can occur in vivo, or they be conducted in vitro after isolation of the antibody construct from the host cell.

[0382] Post-translational modifications can include various modifications as are known in the art (see, for example, Proteins - Structure and Molecular Properties, 2nd Ed., T. E. Creighton, W. H. Freeman and Company, New York, 1993; Post-Translational Covalent Modification of Proteins, B. C. Johnson, Ed., Academic Press, New York, pgs. 1-12, 1983; Seifter et al., 1990, Meth. Enzymol., 182:626-646, and Rattan et al., 1992, Ann. N.Y. Acad. Sci., 663:48-62). In those embodiments in which an antibody construct comprises one or more post-translational modifications, the antibody construct can comprise the same type of modification at one or several sites (e.g., amino acid residues), or it can comprise different modifications at different sites.

[0383] Examples of post-translational modifications can 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.

[0384] Other examples of post-translational modifications can include, for example, addition or removal of N-linked or O-linked carbohydrate chains, chemical modifications of N-linked or O- linked carbohydrate chains, processing of N-terminal or C-terminal ends, attachment of chemical moieties to the amino acid backbone, and addition or deletion of an N-terminal methionine residueresulting from prokaryotic host cell expression. Post-translational modifications can also include modification with a detectable label, such as an enzymatic, fluorescent, isotopic or affinity label to allow for detection and isolation of the protein. 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, dichlorotriazinylamine fluorescein, dansyl chloride and phycoerythrin. An example of a luminescent material is luminol, examples of bioluminescent materials include luciferase, luciferin and aequorin, and examples of suitable radioactive materials include iodine, carbon, sulfur, tritium, indium, technetium, thallium, gallium, palladium, molybdenum, xenon, and fluorine.

[0385] Additional examples of post-translational modifications can include acylation, ADP- ribosylation, amidation, covalent attachment of flavin, covalent attachment of a heme moiety, covalent attachment of a nucleotide or nucleotide derivative, covalent attachment of a lipid or lipid derivative, covalent attachment of phosphotidylinositol, 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.

[0386] In some embodiments, described herein is a method of producing a trivalent and trispecific antibody construct of the present disclosure, 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 four or more polypeptide chains that form the antibody construct; and (b) recovering the antibody construct from the host cell culture. In some embodiments, such method can further comprise, subsequent to step (b), purifying the antibody construct.B. Methods of Using an Antibody Construct of the Present Disclosure

[0387] In certain embodiments, the present disclosure relates to methods of using a trivalent and trispecific antibody construct of the present disclosure. In some embodiments, described hereinare methods of using an antibody construct described herein for the treatment of a disease or condition in a subject in need thereof.

[0388] Such method can comprise administering a trivalent and trispecific antibody construct, or a pharmaceutical composition comprising such antibody construct, to a subject in need thereof. In certain embodiments, the subject is a mammal. In some embodiments, the subject is a human.

[0389] In some embodiments, the present disclosure relates to a method of treating a cancer in a subject in need thereof, the method comprising administering to the subject a trivalent and trispecific antibody construct of the present disclosure, or a pharmaceutical composition comprising such antibody construct. Cancers that can be treated using the methods and antibody constructs disclosed herein can include, but are not limited to, hematologic neoplasms (including leukemias, myelomas and lymphomas), carcinomas (including adenocarcinomas and squamous cell carcinomas), melanomas and sarcomas. Carcinomas and sarcomas are also frequently referred to as “solid tumors”. In certain embodiments, the cancer is a solid tumor. In certain embodiments, the cancer is leukemia. In certain embodiments, the cancer is lymphoma. In certain embodiments, the cancer is a DLL3 -positive (or DLL3 -expressing) cancer.

[0390] When used in a method described herein, an antibody construct of this disclosure can exert either a cytotoxic or cytostatic effect that can result in one or more of a reduction in the size of a tumor, the slowing or prevention of an increase in the size of a tumor, an increase in the disease- free survival time between the disappearance or removal of a tumor and its reappearance, prevention of an initial or subsequent occurrence of a tumor (e.g., metastasis), an increase in the time to progression, reduction of one or more adverse symptom(s) associated with a tumor, an increase in the overall survival time of a subject having a tumor, or a combination of the above.

[0391] The methods described herein can comprise administering a trivalent and trispecific antibody construct to a subject in need thereof. An antibody construct can be administered to a subject by any appropriate route of administration. As will be appreciated by the person of skill in the art, the route and / or mode of administration can vary depending upon the desired therapeutic results. In various embodiments, antibody constructs of this disclosure can be administered by systemic administration or local administration. Local administration can be at the site of a tumor or into a tumor draining lymph node. Generally, the antibody constructs can be administered by parenteral administration, for example, by intravenous, intramuscular, intradermal, intraperitoneal, subcutaneous, or spinal administration, such as by injection or infusion.

[0392] A treatment (e.g., of a cancer in a subject) can be achieved by administration of a therapeutically effective amount of a trivalent and trispecific antibody construct to a subject in need thereof. A “therapeutically effective amount,” as used herein, generally refers to an amount of an antibody construct described herein that is effective, at dosages and for periods of time necessary, to achieve a desired therapeutic result. A therapeutically effective amount can vary according to factors such as the disease state, age, sex, and weight of the subject. A therapeutically effective amount is also one in which any potential toxic or detrimental effects of the antibody constructs are outweighed by the therapeutically beneficial effects. “Sufficient amount” generally refers to an amount sufficient to produce a desired effect, e.g., an amount sufficient to generate an anti-tumor immune response to a target (e.g., tumor) cell or tissue, e.g., by engaging an immune cell (e.g., T cell) using a trivalent and trispecific antibody construct described herein.

[0393] A suitable dosage of a trivalent and trispecific antibody construct described herein can be determined by a skilled medical practitioner. The selected dosage level may depend upon a variety of pharmacokinetic factors including the activity (e.g., antigen affinity(ies)) of the particular antibody construct employed, the route of administration, the time of administration, the rate of excretion of the construct, the duration of the treatment, other drugs, compounds and / or materials used in combination with the antibody construct, e.g., anti-cancer agents, the age, sex, weight, condition, general health and prior medical history of the subject being treated, and like factors well known in the medical arts.

[0394] In some embodiments, a method of treating a disease (e.g., a cancer) in a subject comprises administering a second active ingredient (e.g., another protein or small molecule) in addition to an antibody construct described herein. Such second active ingredient can be administered simultaneously or sequentially with an antibody construct dependent upon the condition to be treated.

[0395] In some embodiments, the present disclosure relates to a method of eliciting an anti-tumor immune response in a cell population comprising immune cells expressing CD3 and CD28 and tumor cells expressing DLL3, the method comprising contacting the cell population with an effective amount of a trivalent and trispecific antibody construct of the present disclosure. In some embodiments, such trivalent and trispecific antibody construct binds CD3 and CD28 on one or more cytotoxic effector cell, such as an immune cell, and DLL3 on a tumor cell, thereby forming a TCR-independent immune synapse, and the antibody construct is either an scFv2antibodyconstruct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a first scFv domain capable of binding CD28 on an immune cell, (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide, or the antibody construct is a Het-Fab antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on an immune cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0396] In some embodiments, the present disclosure relates to a method of inhibiting the proliferation of tumor cells expressing DLL3 in a cell population comprising the tumor cells and immune cells expressing CD3 and CD28, the method comprising contacting the cell population with an effective amount of a trivalent and trispecific antibody construct of the present disclosure. In some embodiments, such trivalent and trispecific antibody construct binds CD3 and CD28 on one or more immune cell(s) and DLL3 on a tumor cell, thereby forming a TCR-independent immune synapse, and the antibody construct is either an scFv2antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a first scFv domain capable of binding CD28 on an immune cell, (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide, or the antibody construct is a Het-Fab antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on an immune cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptideand a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0397] In some embodiments, the present disclosure relates to a method of killing tumor cells expressing DLL3, such method comprising contacting a cell population comprising the tumor cells and immune cells expressing CD3 and CD28 with an effective amount of a trivalent and trispecific antibody construct of the present disclosure. In various embodiments, such trivalent and trispecific antibody construct binds CD3 and CD28 on one or more immune cell(s) and DLL3 on a tumor cell, thereby forming a TCR-independent immune synapse, and the antibody construct is either an scFv2antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a first scFv domain capable of binding CD28 on an immune cell, (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N- terminus of the second Fc polypeptide, or the antibody construct is a Het-Fab antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on an immune cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0398] In any of the methods described herein, the immune cell is a T cell, and an immune cell population can comprise or consist of one or more type(s) of T cells.

[0399] As described herein, in various embodiments, an antibody construct herein can be administered to a subject in need thereof, for example, a subject having cancer, in order to modulate an immune response in the subject. The immune response that can be modulated using an antibody construct of this disclosure can be an anti-tumor immune response in the subject, e.g., in various embodiments, such modulated immune response can occur locally at a tumor site. Thus,in certain embodiments, an antibody construct described herein can initiate and / or upregulate a local immune response, e.g., an anti-tumor response of a subject’s immune system in order to elicit a localized cytotoxic effect against the tumor at the tumor site.

[0400] In various embodiments, an antibody construct described herein, e.g., a trivalent and trispecific antibody construct capable of monovalent binding of CD3 (e.g., via one Fab domain), monovalent binding of CD28 (e.g., via one scFv domain), and monovalent binding of DLL3 (e.g., via an scFv or Fab domain), can have a broader therapeutic window, compared to comparable conventional molecules, and can allow administration of higher doses of the herein described constructs, leading potentially to increased anti-tumor effects without inducing, or inducing lower grades of, side effects and / or off-target effects. Such broader therapeutic window can be due to certain properties of the antibody constructs described herein, including higher ratios of anti-tumor activity compared to cytokine induction, i.e., higher tumor cell killing activities can be achieved at lower cytokine induction levels.

[0401] In some embodiments, the present disclosure relates to a method of inhibiting the growth of a DLL3 -expressing tumor and / or reducing the volume of the tumor in a subject in need thereof, the method comprising administering to the subject an effective amount of a trivalent and trispecific antibody construct of the present disclosure. In various embodiments, such in vivo antitumor effect is elicited by the trivalent and trispecific antibody construct that binds (e.g., simultaneously) CD3 and CD28 on one or more immune cell(s) and DLL3 on a tumor cell, thereby forming a TCR-independent immune synapse, and the antibody construct is either an scFv2antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a first scFv domain capable of binding CD28 on an immune cell, (iii) a second scFv domain capable of binding DLL3 on a tumor cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the first scFv domain is coupled to the C-terminus of the Fab light chain, and (c) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide, or the antibody construct is a Het-Fab antibody construct as described herein, comprising: (i) a first Fab domain capable of binding CD3 on an immune cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on an immune cell; and (iv) a heterodimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of thefirst Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0402] In various embodiments, and as further described herein, inhibition of tumor growth and / or a reduction in tumor volume in a subject can be elicited by simultaneous binding of the antibody construct to CD3 and CD28 on one or more immune cell(s) and to DLL3 on a tumor cell and formation of a TCR-independent artificial immune synapse within a tumor environment in the subject, thereby eliciting an anti -turn or cytotoxic effect mediated by the activated immune cell and directed against the tumor cell. The immune cell can be a T cell.

[0403] In various embodiments of the methods described herein, the trivalent and trispecific antibody constructs binds CD3 and CD28 on the same immune cell (e.g., T cell). In other embodiments, the trivalent and trispecific antibody constructs binds CD3 on a first immune cell (e.g., T cell) and CD28 on a second immune cell (e.g., T cell), wherein the first and the second immune cells are different cells.

[0404] In various embodiments of the methods described herein, the subject is a rodent, a nonhuman primate, or a human.

[0405] In further embodiments, and in relation to a method described herein, administration of a sufficient amount of a trivalent and trispecific antibody construct to a subject in need thereof can provide one or more of the following to activate or upregulate an immune response in the subject: (i) modulation of T-cell receptor signaling, (ii) modulation of T-cell activation, (iii) modulation of pro-inflammatory cytokines, (iv) modulation of interferon-y (IFNy) production by T cells, (v) modulation of T-cell suppression, (vi) modulation of M2 -type tumor associated macrophages (TAM) or myeloid-derived suppressor cell (MDSC) survival and / or differentiation, and / or (vii) modulation of cytotoxic or cytostatic effects on cells, e.g., cancer cells.

[0406] In some embodiments, the present disclosure relates to methods of modulating an immune response in a cell population or in a subject using one or more of the trivalent and trispecific antibody construct(s) of the present disclosure, wherein such modulation can comprise one or more of (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 of the above. As described herein, in certain embodiments, an antibody construct ofthe present disclosure activates T effector cells. In some embodiments, and as demonstrated herein, an antibody construct increases production of one or more cytokines and / or signalling molecules, e.g., compared to baseline or untreated conditions, such as GM-CSF, TNFa, a MIP-1, IFN-y, IL- 2, IL-6, IL-10, IL-12, IL-17, IL-21 and / or C-X-C motif ligand 13 (CXCL13) by T effector cells.

[0407] As described herein, in various embodiments, a trivalent and trispecific antibody construct of the present disclosure comprises an Fc domain comprising a first and a second Fc polypeptide, wherein one or more of the Fc polypeptides can comprise a modified CH2 domain (e.g., compared to a WT domain) that comprises one or more amino acid modifications that can result in a decrease or elimination of binding of the Fc domain to one or more, or to all of the FcyRs (also referred to herein as an Fc “knock-out” or “KO” variant).C. Experimental Methods

[0408] In some embodiments, the present disclosure relates to experimental methods for analyzing and / or detecting a trivalent and trispecific antibody construct of the present disclosure. Such methods can be used, for example, to assess in vitro and / or in vivo properties of such antibody construct such as its pharmacokinetic (PK) and pharmacodynamic (PD) properties. Other properties and characteristics of an antibody construct can be evaluated such as its stability under certain conditions (e.g., temperature, pH, etc.), its solubility, or its behaviour in the presence of certain other chemical components such as other proteins or cells.

[0409] Specific binding of a trivalent and trispecific antibody construct described herein to an antigen (e.g., CD3, CD28, TAA) can be measured, for example, through an enzyme-linked immunosorbent assay (ELISA), a surface plasmon resonance (SPR) technique (employing, 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 the extent of binding to an unrelated protein being less than about 10% of the binding to the target antigen (e.g., CD3, CD28, TAA, etc.) as measured by SPR, for example.

[0410] Additional experimental methods are described in, e.g., EXAMPLES 1-28 herein.VII. CERTAIN EMBODIMENTS OF THE DISCLOSURE

[0411] Certain embodiments of the present disclosure relate to embodiments 1-146 below, as well as to any combination of one or more of embodiments 1-146.

[0412] Embodiment 1. An antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a first scFv domain capable of binding asecond antigen on a cytotoxic effector cell; (iii) a third binding domain capable of binding DLL3 on a tumor cell, wherein the third binding domain is a second Fab domain or a second scFv domain; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to: the C-terminus of the light chain of either the first Fab domain or the second Fab domain, if the third binding domain is the second Fab domain, or the C-terminus of the light chain of the first Fab domain, if the third binding domain is the second scFv domain.

[0413] Embodiment 2. The antibody construct of embodiment 1, wherein the third binding domain is the second Fab domain.

[0414] Embodiment 3. An antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding a tumor-associated antigen (TAA) on a tumor cell; (iii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain or the second Fab domain.

[0415] Embodiment 4. The antibody construct of embodiment 3, wherein the TAA is DLL3.

[0416] Embodiment 5. An antibody construct, comprising: (i) a first Fab domain capable of binding a first antigen on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding a second antigen on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain or the second Fab domain.

[0417] Embodiment 6. The antibody construct of any one of embodiments 1-5, wherein the first antigen is CD3 or CD28.

[0418] Embodiment 7. The antibody construct of any one of embodiments 1-6, wherein the second antigen is CD3 or CD28.

[0419] Embodiment 8. The antibody construct of any one of embodiments 1-7, wherein the first antigen is CD3 or CD28 and the second antigen is CD3 or CD28, and wherein the first antigen and the second antigen are different.

[0420] Embodiment 9. The antibody construct of any one of embodiments 1-8, wherein the first antigen is CD3 and the second antigen is CD28.

[0421] Embodiment 10. An antibody construct, comprising: (i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell; (ii) a second Fab domain capable of binding DLL3 on a tumor cell; (iii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell; and (iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain or the second Fab domain.

[0422] Embodiment 11. The antibody construct of any one of embodiments 1-10, wherein the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0423] Embodiment 12. The antibody construct of any one of embodiments 1-11, wherein the antibody construct comprises a first immunoglobulin heavy chain (Hl) comprising at least a first Fab heavy chain, a second immunoglobulin heavy chain (H2) comprising at least a second Fab heavy chain, a first immunoglobulin light chain (LI) comprising at least a first Fab light chain, and a second immunoglobulin light chain (L2) comprising at least a second Fab light chain.

[0424] Embodiment 13. The antibody construct of embodiment 12, wherein Hl comprises, from N- to C-terminus, the first Fab heavy chain comprising a first VH sequence and a first CHI sequence, coupled to the first Fc polypeptide, and wherein the first Fab heavy chain of Hl and the first Fab light chain of LI associate to form the first Fab domain.

[0425] Embodiment 14. The antibody construct of embodiment 13, wherein Hl further comprises a first linkerFab'Fcthat couples the C-terminus of the first Fab heavy chain to the N-terminus of the first Fc polypeptide.

[0426] Embodiment 15. The antibody construct of any one of embodiment 12-14, wherein H2 comprises, from N- to C-terminus, the second Fab heavy chain comprising a second VH sequence and a second CHI sequence, coupled to the second Fc polypeptide, and wherein the second Fab heavy chain of H2 and the second Fab light chain of L2 associate to form the second Fab domain.

[0427] Embodiment 16. The antibody construct of embodiment 15, wherein H2 further comprises a second linkerl ab-l cthat couples the C-terminus of the second Fab heavy chain to the N-terminus of the second Fc polypeptide.

[0428] Embodiment 17. The antibody construct of any one of embodiments 14 or 16, wherein the first linkerFab'Fcand the second linkerl ab-l care polypeptide linkers, each independently comprising from about 5 to about 50 consecutive amino acid residues.

[0429] Embodiment 18. The antibody construct of embodiment 17, wherein the first linkerFab'Fcand the second linkerl ab-l care each derived independently from an IgGl, IgG2 or IgG4 hinge region.

[0430] Embodiment 19. The antibody construct of any one of embodiments 12-18, wherein LI further comprises the first scFv domain, and wherein the N-terminus of the first scFv domain is coupled to the C-terminus of a first constant domain (CL) sequence of LI.

[0431] Embodiment 20. The antibody construct of embodiment 19, wherein LI comprises, from N- to C-terminus, the first Fab light chain comprising a first VL sequence and the first CL sequence, and the first scFv domain comprising, from N- to C-terminus, either (i) a third VH sequence coupled to a third VL sequence or (ii) a third VL sequence coupled to a third VH sequence.

[0432] Embodiment 21. The antibody construct of embodiment 20, wherein the first scFv domain comprises, from N- to C-terminus, the third VL sequence coupled to the third VH sequence.

[0433] Embodiment 22. The antibody construct of any one of embodiments 19-21, wherein LI further comprises a linkerFab'scFvthat couples the C-terminus of the first Fab light chain to the N- terminus of the first scFv domain.

[0434] Embodiment 23. The antibody construct of embodiment 22, wherein the linkerFab'scFvcomprises the amino acid sequence (G4S)X, wherein x is 1, 2, 3, 4 or 5 (SEQ ID NO: 119).

[0435] Embodiment 24. The antibody construct of any one of embodiments 12-23, wherein L2 comprises the second Fab light chain comprising a second VL sequence and a second CL sequence.

[0436] Embodiment 25. The antibody construct of any one of embodiments 1-24, comprising: (i) the first Fab domain, wherein the first Fab domain is capable of binding CD3 on a cytotoxic effector cell; (ii) the second Fab domain, wherein the second Fab domain is capable of binding DLL3 on the tumor cell; (iii) the first scFv domain, wherein the first scFv domain is capable of binding CD28 on a cytotoxic effector cell; and (iv) the dimeric Fc domain comprising the first Fcpolypeptide and the second Fc polypeptide, wherein: (a) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide, (b) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide, and (c) the scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

[0437] Embodiment 26. The antibody construct of any one of embodiments 1-25, wherein the first antigen and the second antigen are located on the same cell or on different cells.

[0438] Embodiment 27. The antibody construct of any one of embodiments 1-26, wherein the first Fab domain is capable of binding CD3 and comprises a VH domain comprising a HCDR1 sequence comprising the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 sequence comprising the sequence ITSSGGRI (SEQ ID NO: 50) or ITRSGGRI (SEQ ID NO: 51), and a HCDR3 sequence comprising the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 sequence comprising the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 sequence comprising the sequence RND (SEQ ID NO: 54), and a LCDR3 sequence comprising the sequence QSYSSGFI (SEQ ID NO: 55).

[0439] Embodiment 28. The antibody construct of any one of embodiments 1-27, wherein the first Fab domain is capable of binding CD3 and comprises a VH domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 5.

[0440] Embodiment 29. The antibody construct of any one of embodiments 1-28, wherein the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising a HCDR1 sequence comprising the sequence SYGVH (SEQ ID NO: 28), a HCDR2 sequence comprising the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 sequence comprising the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL sequence comprising a LCDR1 sequence comprising the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), aLCDR2 sequence comprising the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 sequence comprising the sequence QQSRKVPFT (SEQ ID NO: 48).

[0441] Embodiment 30. The antibody construct of any one of embodiments 1-29, wherein the first scFv domain is capable of binding CD28 and comprises a VH sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to thesequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 16.

[0442] Embodiment 31. The antibody construct of any one of embodiments 1-30, wherein the first Fc polypeptide and the second Fc polypeptide each comprise a CH2 domain that can independently comprise an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 8.

[0443] Embodiment 32.The antibody construct of any one of embodiments 1-31, wherein the dimeric Fc domain is a heterodimeric Fc domain in which one Fc polypeptide has an amino acid sequence that differs in at least one residue from the amino acid sequence of the other Fc polypeptide.

[0444] Embodiment 33. The antibody construct of any one of embodiments 1-32, wherein one of the Fc polypeptides comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 9, and the other Fc polypeptide comprises a CH3 domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 10.

[0445] Embodiment 34. The antibody construct of any one of embodiments 1-33, wherein the first Fab domain, the second Fab domain, or both Fab domains, comprise one or more amino acid modifications that promote correct heavy and light chain pairing to produce the first and second Fab domains, respectively, relative to the formation of incorrectly paired Fab domains.

[0446] Embodiment 35. The antibody construct of any one of embodiments 1-34, wherein (i) both Fab domains each comprise a kappa light chain; (ii) both Fab domains each comprise a lambda light chain; or (iii) one Fab domain comprises a kappa light chain and the other Fab domain comprises a lambda light chain.

[0447] Embodiment 36. The antibody construct of embodiment 34 or embodiment 35, wherein one Fab domain comprises a kappa light chain and comprises one or more of the amino acid substitutions 143E, 145T, 179E and 228D in the CHI sequence and one or more of the amino acid substitutions 121K, 124R and 178R in the CL sequence of the kappa light chain, and the other Fab domain comprises a lambda light chain and comprises one or more of the amino acid substitutions 125R and 188K in the CHI sequence and one or more of the amino acid substitutions 122D, 129T,176E and 178E in the CL sequence of the lambda light chain, and wherein the amino acid residues are identified according to the Kabat numbering system.

[0448] Embodiment 37. The antibody construct of embodiment 34 or embodiment 35, wherein one Fab domain comprises a kappa light chain and comprises one or more of the amino acid substitutions 124R and 186K in the CHI sequence and one or more of the amino acid substitutions 133G, 135W, 176D and 180E in the CL sequence of the kappa light chain, and the other Fab domain comprises a lambda light chain and comprises one or more of the amino acid substitutions 139W, 143D and 145T in the CHI sequence and one or more of the amino acid substitutions 124Q and 131R in the CL sequence of the lambda light chain, and wherein the amino acid residues are identified according to the Kabat numbering system.

[0449] Embodiment 38. An antibody construct, comprising: (i) a first heavy chain (Hl), comprising an anti-CD3 Fab heavy chain coupled to a first Fc polypeptide, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 11; (ii) a second heavy chain (H2), comprising an anti-DLL3 Fab heavy chain coupled to a second Fc polypeptide, wherein the second Fc polypeptide comprises an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 12, and wherein the first Fc polypeptide of Hl and the second Fc polypeptide of H2 associate to form a heterodimer; (iii) a first light chain (LI), comprising an anti-CD3 Fab light chain coupled to an anti-CD28 scFv domain, and comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 13, and wherein the anti-CD3 Fab light chain of LI and the anti- CD3 Fab heavy chain of Hl associate to form an anti-CD3 Fab domain; and (iv) a second light chain (L2), comprising an anti-DLL3 Fab light chain, and wherein L2 and the anti-DLL3 Fab heavy chain of H2 associate to form an anti-DLL3 Fab domain.

[0450] Embodiment 39. The antibody construct of any one of embodiments 1-38, wherein the second Fab domain has a KD value for binding DLL3 from about 0.01 nM to about 1...

Claims

CLAIMSWHAT IS CLAIMED IS:

1. An antibody construct, comprising:(i) a first Fab domain capable of binding CD3 on a cytotoxic effector cell;(ii) a first scFv domain capable of binding CD28 on a cytotoxic effector cell;(iii) a third binding domain capable of binding DLL3 on a tumor cell; and(iv) a dimeric Fc domain comprising a first Fc polypeptide and a second Fc polypeptide, wherein:(a) the first Fab domain is coupled to the N-terminus of the first Fc polypeptide,(b) the third binding domain is coupled to the N-terminus of the second Fc polypeptide, and(c) the first scFv domain is coupled to the C-terminus of the light chain of the first Fab domain.

2. The antibody construct of claim 1, wherein the first Fab domain comprises a VH domain comprising a HCDR1 having the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 having the sequence ITSSGGRI (SEQ ID NO: 50) or ITRSGGRI (SEQ ID NO: 51), and a HCDR3 having the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 having the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 having the sequence RND, and a LCDR3 having the sequence QSYSSGFI (SEQ ID NO: 55), according to the IMGT numbering system.

3. The antibody construct of claim 1 or claim 2, wherein the first Fab domain comprises a VH domain comprising a HCDR1 having the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 having the sequence ITRSGGRI (SEQ ID NO: 51), and a HCDR3 having the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 having the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 having the sequence RND, and a LCDR3 having the sequence QSYSSGFI (SEQ ID NO: 55), according to the IMGT numbering system.

4. The antibody construct of any one of claims 1-3, wherein the first Fab domain comprises a VH domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 2, and a VLdomain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 5.

5. The antibody construct of any one of claims 1-4, wherein the first Fab domain comprises a VH domain comprising the amino acid sequence set forth in SEQ ID NO: 2, and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 5.

6. The antibody construct of any one of claims 1-5, wherein the first scFv domain comprises a VH domain comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL domain comprising a LCDR1 having the sequence RASESVEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), according to the Kabat numbering system.

7. The antibody construct of any one of claims 1-6, wherein the first scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL domain comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 16.

8. The antibody construct of any one of claims 1-7, wherein the first scFv domain comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 15, and a VL domain comprising the amino acid sequence set forth in SEQ ID NO: 16.

9. The antibody construct of any one of claims 1-8, wherein the third binding domain is a second Fab domain or a second scFv domain.

10. The antibody construct of any one of claims 1-9, wherein the third binding domain comprises a VH domain comprising a HCDR1 having the sequence HYYVH (SEQ ID NO: 90), a HCDR2 having the sequence IIDPGGGTTSYAQKFLG (SEQ ID NO: 91) and a HCDR3 having the sequence GERVTGNYFYYGMDV (SEQ ID NO: 92), and a VL domain comprising a LCDR1 having the sequence VANQGISNYLV (SEQ ID NO: 93), a LCDR2 having the sequence SVNSLYS (SEQ ID NO: 94) and a LCDR3 having the sequence LQHDSYPYT (SEQ ID NO: 95), according to the Kabat numbering system.

11. The antibody construct of any one of claims 1-10, wherein the third binding domain comprises a VH sequence comprising the amino acid sequence set forth in SEQ ID NO: 99, and a VL domain comprising the amino acid sequence the sequence set forth in SEQ ID NO: 100.

12. The antibody construct of claim 1, wherein (i) the first Fab domain comprises a VH domain comprising a HCDR1 having the sequence GVTFNYYG (SEQ ID NO: 49), a HCDR2 having the sequence ITRSGGRI (SEQ ID NO: 51), and a HCDR3 having the sequence TLDGRDGWVAY (SEQ ID NO: 52), and a VL domain comprising a LCDR1 having the sequence TGNIGSNY (SEQ ID NO: 53), a LCDR2 having the sequence RND, and a LCDR3 having the sequence QSYSSGFI (SEQ ID NO: 55), according to the IMGT numbering system, (ii) the first scFv domain comprises a VH domain comprising a HCDR1 having the sequence SYGVH (SEQ ID NO: 28), a HCDR2 having the sequence VIWPGGGTNFNSALMS (SEQ ID NO: 31), and a HCDR3 having the sequence DRAYGNYLYAMDY (SEQ ID NO: 35), and a VL domain comprising a LCDR1 having the sequence RASES VEYYVTSLMQ (SEQ ID NO: 41), a LCDR2 having the sequence AASNVDS (SEQ ID NO: 44), and a LCDR3 having the sequence QQSRKVPFT (SEQ ID NO: 48), according to the Kabat numbering system, and (iii) the third binding domain comprises a VH domain comprising a HCDR1 having the sequence HYYVH (SEQ ID NO: 90), a HCDR2 having the sequence IIDPGGGTTSYAQKFLG (SEQ ID NO: 91) and a HCDR3 having the sequence GERVTGNYFYYGMDV (SEQ ID NO: 92), and a VL domain comprising a LCDR1 having the sequence VANQGISNYLV (SEQ ID NO: 93), a LCDR2 having the sequence SVNSLYS (SEQ ID NO: 94) and a LCDR3 having the sequence LQHDSYPYT (SEQ ID NO: 95), according to the Kabat numbering system.

13. The antibody construct of claim 1, wherein (i) the first Fab domain comprises the HCDR1-3 sequences of the VH sequence set forth in SEQ ID NO: 2, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 5, (ii) the first scFv domain comprises the HCDR1-3 sequences of the VH sequence set forth in SEQ ID NO: 15, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 16, and (iii) the third binding domain comprises the HCDR1-3 sequences of the VH sequence set forth inSEQ ID NO: 99, and the LCDR1-3 sequences of the VL sequence set forth in SEQ ID NO: 100.

14. The antibody construct of claim 1, wherein (i) the first Fab domain comprises a VH sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 2, and a VL sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 5, (ii) the first scFv domain comprises a VH sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 15, and a VL sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 16, and (iii) the third binding domain comprises a VH sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 99, and a VL sequence comprising an amino acid sequence having at least about 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 100.

15. The antibody construct of any one of claims 1-14, wherein the third binding domain is a second Fab domain.

16. The antibody construct of claim 15, wherein (i) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide via a first linkerFab'Fcand (ii) the second Fab domain is coupled to the N-terminus of the second Fc polypeptide via a second linkerFab'Fc.

17. The antibody construct of claim 16, wherein the first linkerFab'Fcand the second linkerl ab-Fcindependently comprise or consist of the amino acid sequence set forth in SEQ ID NOs: 56 or 120.

18. The antibody construct of any one of claims 15-17, wherein one Fab domain comprises a kappa light chain and the other Fab domain comprises a lambda light chain.

19. The antibody construct of claim 18, wherein the Fab domain comprising the kappa light chain comprises one or more of the amino acid substitutions 143E, 145T, 179E and 228D in the CHI sequence and one or more of the amino acid substitutions 121K, 124R and 178R in the CL sequence of the kappa light chain, and the Fab domain comprising the lambda light chain comprises one or more of the amino acid substitutions 125R and 188K in the CHI sequence and one or more of the amino acid substitutions 122D, 129T, 176Eand 178E in the CL sequence of the lambda light chain, and wherein the amino acid residues are identified according to the Kabat numbering system.

20. The antibody construct of claim 19, wherein the Fab domain comprising the kappa light chain comprises the amino acid substitutions 143E, 145T, 179E and 228D in the CHI sequence and the amino acid substitutions 121K, 124R and 178R in the CL sequence of the kappa light chain, and the Fab domain comprising the lambda light chain comprises the amino acid substitutions 125R and 188K in the CHI sequence the amino acid substitutions 122D, 129T, 176E and 178E in the CL sequence of the lambda light chain.

21. The antibody construct of any one of claims 1-14, wherein the third binding domain is a second scFv domain.

22. The antibody construct of claim 21, wherein (i) the first Fab domain is coupled to the N- terminus of the first Fc polypeptide via a first linkerFab'Fcand (ii) the second scFv domain is coupled to the N-terminus of the second Fc polypeptide via a first linkerscFv'Fc.

23. The antibody construct of claim 22, wherein the first linkerFab'Fcand the first linkerscFv'Fcindependently comprise or consist of the amino acid sequence set forth in SEQ ID NOs: 56, 120 or 57.

24. The antibody construct of any one of claims 1-23, wherein the first Fc polypeptide and the second Fc polypeptide each comprise a CH2 domain that independently comprise an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 8.

25. The antibody construct of claim 24, wherein the first Fc polypeptide and the second Fc polypeptide each comprise the amino acid substitutions L234A L235A D265S that reduce binding of the Fc domain to one or more FcyRs, and wherein the numbering of amino acid residues in the first and second Fc polypeptides is according to the EU numbering system.

26. The antibody construct of any one of claims 1-25, wherein the first Fc polypeptide and the second Fc polypeptide each comprise an IgGl CH3 domain, and wherein each CH3 domain comprises one or more amino acid substitutions compared to a corresponding wild-type IgGl CH3 domain sequence, wherein the one or more amino acid substitutions promote preferential pairing of the first and second Fc polypeptides to form a heterodimeric Fc domain.

27. The antibody construct of claim 26, wherein the first or the second Fc polypeptide comprises the amino acid substitutions T350V_L351Y_F405A_Y407V, and the other Fc polypeptide comprises the amino acid substitutions T350V T366L K392L T394W, and the numbering of amino acid residues in the first and second Fc polypeptides is according to the EU numbering system.

28. The antibody construct of any one of claims 1-27, wherein the C-terminus of the light chain of the first Fab domain is coupled to the N-terminus of the first scFv domain via a linkerFab'scFvcomprising the amino acid sequence set forth in SEQ ID NO: 58.

29. The antibody construct of any one of claims 1-28, wherein the first scFv domain has the domain structure, from N- to C-terminus, of: VL-LinkerscFv-VH, and wherein the LinkerscFvcomprises or consists of the amino acid sequence set forth in SEQ ID NO: 59.

30. The antibody construct of any one of claims 1-20, comprising: (i) a first heavy chain (Hl) comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 105; (ii) a second heavy chain (H2) comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 106; (iii) a first light chain (LI) comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 107; and (iv) a second light chain (L2) comprising an amino acid sequence having at least about 90%, 95%, 97%, 99% or 100% sequence identity to the sequence set forth in SEQ ID NO: 108.

31. The antibody construct of claim 30, comprising: (i) a first heavy chain (Hl) comprising the amino acid sequence set forth in SEQ ID NO: 105; (ii) a second heavy chain (H2) comprising the amino acid sequence set forth in SEQ ID NO: 106; (iii) a first light chain (LI) comprising the amino acid sequence set forth in SEQ ID NO: 107; and (iv) a second light chain (L2) comprising the amino acid sequence set forth in SEQ ID NO: 108.

32. The antibody construct of any one of claims 1-31, wherein the cytotoxic effector cell is a T cell, and wherein the antibody construct binds CD3 and CD28 on the same T cell as compared to binding CD3 and CD28 on different T cells, as measured in a cell-cell bridging assay with recorded double positive events of less than about 5, less than about 3, or less than about 2.

33. A pharmaceutical composition comprising the antibody construct of any one of claims 1- 32, and a pharmaceutically acceptable carrier, excipient, diluent, or combination thereof.

34. A nucleic acid molecule or a set of nucleic acid molecules encoding one or more, two or more, three or more, or four or more polypeptide chains that form the antibody construct of any one of claims 1-32.

35. A vector or a set of vectors comprising the nucleic acid molecule or the set of nucleic acid molecules of claim 34.

36. A cell comprising the nucleic acid molecule or the set of nucleic acid molecules of claim 34, or the vector or set of vectors of claim 35.

37. A method of producing an antibody construct of any one of claims 1-32, 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 four or more polypeptide chains that form the antibody construct; and(b) recovering the antibody construct from the host cell culture.

38. A method of eliciting an anti-tumor immune response in a cell population comprising T cells and tumor cells, the method comprising contacting the cell population with an effective amount of the antibody construct of any one of claims 1-32, wherein the T cells express CD3 and CD28 and the tumor cells express DLL3.

39. A method of treating a cancer in a subject in need thereof, the method comprising administering to the subject an antibody construct of any one of claims 1-32.

40. An antibody construct of any one of claims 1-32 for use in the treatment of cancer.

41. Use of an antibody construct of any one of claims 1-32 in the manufacture of a medicament for the treatment of cancer.

Images