Fc region-based heterodimer molecule and use thereof

Abstract

Provided are an immunoglobulin Fc region-based heterodimeric Fc stent, a heterologous multimeric protein containing the heterodimeric Fc stent, in particular a single or multispecific binding protein, and a composition comprising same and a use.

Description

Fc region-based heterodimeric molecules and uses thereof TECHNICAL FIELD

[0001] The present invention relates to Fc scaffolds based on immunoglobulin Fc regions, heteromultimeric proteins, in particular mono- or multispecific binding proteins, comprising said Fc scaffolds, and compositions and uses comprising the same. BACKGROUND

[0002] Bispecific antibodies are a rapidly growing class of novel antibody drugs in recent years. Such antibodies can bind to two different antigens or two different epitopes on the same antigen at the same time, thus promising additional clinical benefits for complex disease treatment.

[0003] At present, there are many different structural forms of bispecific antibodies, including non-IgG-like forms based on scFv, Fab, VH, VL, etc. antibody fragment fusion and IgG-like forms based on Fc domain. Compared with non-IgG-like forms, bispecific antibodies similar to conventional IgG structural forms can induce better physicochemical properties, stability, pharmacokinetic characteristics, and have wide application prospects.

[0004] It is known that the Fc region of an antibody has a dimerization function, which plays a key role in maintaining the in vivo function of the antibody. At present, various heterodimeric Fc technologies have been reported to assist bispecific antibody production (Moore GL, Methods 2019, 154:38-50).

[0005] For example, Carter et al. invented the knob-into-hole model, which successfully realized the preparation of bispecific antibodies (patent US7951917B1). In this method, a specific small side chain amino acid located at the interface of the CH3 region of the first heavy chain is mutated to a large side chain amino acid (e.g. T366Y), and a specific amino acid located at the interface of the CH3 region of the second heavy chain is mutated to a small side chain amino acid (e.g. Y407T), which causes the interface mutation residues to be complementary in spatial structure, thereby promoting the formation of Fc heterodimers. In addition to the above example mutations, the mutation combinations more widely used in this method include knob T366W and hole T366S, L368A and Y407V mutation combinations, and the proportion of Fc heterodimers can reach 80-90%. However, the ability of the knob-hole model to inhibit the formation of homodimers is still not enough.

[0006] Electrostatic steering is also widely used to promote Fc heterodimer formation. This method mutates a specific amino acid in one heavy chain CH3 to a positively charged amino acid (Lys or Arg) and another specific amino acid in the other heavy chain CH3 to a negatively charged amino acid (Asp or Glu), promoting Fc heterodimer formation through electrostatic attraction between the mutated residues. For example, Kannam et al. generated a high proportion of heterodimers by introducing mutations D399K and E356K in the CH3 region of the first heavy chain and mutations K409D and K392D in the CH3 region of the second heavy chain (Patent US8592562B2). Igawa et al. also promoted Fc heterodimer formation by introducing mutations E356K and D399K in the CH3 region of the first heavy chain and mutations K439E and K409D in the CH3 region of the second heavy chain (Patent US10011858B2).

[0007] Genmab invented DuoBody technology. This technology involves introducing F405L mutation in the CH3 of one antibody and CH3 K409R mutation in the CH3 of the other antibody. By expressing and purifying these two antibodies separately, and then mixing them in the presence of a reducing agent (such as 2-MEA), controlled Fab arm exchange between the two antibodies is achieved. This technology generates Fc heterodimers at a proportion of about 90-95% (Patent US9150663B2).

[0008] SEED (strand exchange engineered domain) technology is also used to promote Fc heterodimer formation (Muda et al. Proteins Eng. Des. Sel. 2011, 24:447-454). This technology takes advantage of the sequence differences between IgA and IgG in the CH3 domain to generate Fc heterodimers by pairing complementary CH3 domains, avoiding homodimer formation.

[0009] Zymeworks reported a scaffold structure to improve the stability of Fc heterodimers (Patent US20120149876A1). For example, a scaffold 1a design includes T366I, K392M, T394W mutations in the CH3 region of one heavy chain and F405A, Y407A mutations in the CH3 region of the other heavy chain, and the Tm value of this scaffold reaches 74°C. In addition, a scaffold 2a design includes L351Y, Y047A mutations in the CH3 region of one heavy chain and T366V, K409F mutations in the CH3 region of the other heavy chain, and the Tm value of this scaffold reaches 75.5°C. The proportion of Fc heterodimers based on such scaffold structures reaches more than 90%.

[0010] Although various strategies for Fc heterodimer formation have been proposed, these strategies have shown certain defects in application. For example, most Fc-containing bispecific antibodies have the problem of homologous mispairing and the problem of reduced thermal stability, thereby affecting the drugability of antibody drugs.

[0011] For example, it has been found that most Fc heterodimers formed by the prior art have lower thermal stability than wild-type IgG1 mAbs. The Tm value of the CH3 domain of wild-type IgG1 mAbs is generally greater than 80°C (Ionescu RM, J Pharm Sci. 2008, 97: 1414-1426). However, the CH3 amino acid mutations introduced by Fc heterodimerization technology can cause the melting temperature (Tm) of the CH3 domain to decrease significantly. For example, the Tm value of the CH3 heterodimer produced by the knob-hole model (T366W / T366S, L368A, Y407V) is 69.4°C (Atwell S, J. Mol. Biol. 1997, 270: 26-35). The Tm value of the CH3 heterodimer produced by the electrostatic steering strategy (D399K, E356K / K409D, K392D) is 68.8°C (Gunasekaran K, J. Biol. Chem. 2010, 285: 19637-19646). For complex antibody drugs, a lower Tm value will affect the manufacture and storage of the drug. Although the introduction of non-native disulfide bonds (such as Y349C and S354C) in the knob-hole model can improve the efficiency of hetero-Fc dimer formation and thermal stability, the introduction of additional disulfide bonds will have an adverse effect on product quality for some drugs, such as antibody conjugate drugs (ADC).

[0012] In addition, the proportion of Fc heterodimers produced by the prior art is generally between 80-95%, and other impurities include Fc monomers, Fc homodimers, and high molecular weight aggregates. These impurities need to be further removed by ion exchange chromatography and other methods, and the removal of homologous heavy chain mispairing impurities in particular poses a challenge to downstream antibody purification processes. Moreover, the higher the impurity content, the lower the recovery rate of the product after purification, which also increases production costs.

[0013] In view of the wide application prospects of IgG-like bispecific antibodies, there is still an urgent need in the art to develop Fc heterodimerization technology to solve the problems of chain mispairing and thermal stability of bispecific antibodies in order to improve the drugability of bispecific antibodies.

[0014] SUMMARY

[0015] The present inventors have found that in a scaffold structure based on a dimerized Fc region, on the basis of introducing a knob-into-hole (KIH) mutation, further introducing specific CH3 domain interface mutations can make the antibody comprising the Fc scaffold superior to the corresponding antibody comprising only the KIH mutation in terms of thermal stability; while maintaining the advantages brought by the KIH mutation (especially, the advantage of promoting the pairing of heterologous dimeric heavy chains of interest), and maintaining biological effects comparable to natural antibodies, including antigen-specific binding and FcRn receptor and FcγR receptor binding. Based on this finding, the present inventors propose a combination of Fc heterodimerization mutations capable of reducing the proportion of chain mispairing and increasing the stability of the antibody product of interest, and application of the mutation combination of the present application can effectively improve the drugability and stability of bispecific antibodies.

[0016] In a first aspect, therefore, the present application provides a CH3 heterodimer and a heterodimeric Fc scaffold comprising the mutation combination of the present application.

[0017] In a second aspect, the present application provides the use of the heterodimeric Fc scaffold and the CH3 heterodimer of the present application as components for constructing heteromultimeric proteins, especially single / multispecific binding proteins, and the corresponding heteromultimeric proteins comprising the Fc scaffold or CH3 heterodimer.

[0018] In a third aspect, the present application provides, inter alia, a binding protein comprising the heterodimeric Fc scaffold according to the present application.

[0019] In a fourth aspect, the present application provides an isolated polynucleotide encoding the heterodimeric Fc scaffold, CH3 heterodimer, heteromultimeric protein or binding protein according to the present application; a vector (especially an expression vector) comprising the isolated polynucleotide of the present application and a host cell comprising the isolated polynucleotide or vector (especially an expression vector) of the present application. The present application also provides the use and method of using the polynucleotide, vector or host cell of the present application to produce the heterodimeric Fc scaffold, CH3 heterodimer, heteromultimeric protein and binding molecule of the present application.

[0020] In a fifth aspect, the present application provides a composition comprising the heteromultimeric protein or binding protein according to the present application and a conjugate comprising a therapeutic agent conjugated to the heteromultimeric protein or binding protein according to the present application. In some embodiments, the composition is a pharmaceutical composition comprising the heteromultimeric protein of the present application (especially the multispecific binding protein of the present application) and a pharmaceutically acceptable carrier.

[0021] In a sixth aspect, the present application also provides methods of using and uses of the heteromultimeric proteins, binding proteins, compositions and conjugates according to the present application. In some embodiments, the present application provides uses of the heteromultimeric proteins, binding proteins, compositions or conjugates according to the present application as a medicament. In some embodiments, the present application provides uses of the heteromultimeric proteins, binding proteins, compositions or conjugates according to the present application for treating a disease in a subject in need thereof. In specific embodiments, the disease is cancer. The present application also provides uses of the heteromultimeric proteins, binding proteins, compositions or conjugates according to the present application in the manufacture of a medicament for treating a disease in a subject in need thereof; and methods of treating a disease in a subject using the heteromultimeric proteins, binding proteins, compositions or conjugates according to the present application, wherein the method comprises administering to the subject a therapeutically effective amount of the heteromultimeric proteins, binding proteins, compositions or conjugates of the present application. In some embodiments, the heteromultimeric proteins, binding proteins, compositions or conjugates are formulated in a pharmaceutically acceptable form. In some embodiments, the disease is cancer. In some embodiments, the subject is preferably a mammal, particularly a human.

[0022] BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 shows a partial crystal structure representation of the CH3-CH3 region of an Fc dimer.

[0024] Figure 1A is a representation of the hydrophobic interactions at the CH3-CH3 interface of wild-type IgGl (PDB: 1HZH).

[0025] Figure IB is a representation of the hydrophobic interactions at the CH3-CH3 interface of the knob-hole model (PDB: 4NQS). The left side of the figure is the knob chain and the right side is the hole chain.

[0026] Figure 1C is a representation of the 3D structure of the CH3-CH3 region based on computer simulation of the M2 combination of mutations. The hole L351Y mutation forms new hydrogen bonds (represented by dashed lines) with the knob T366W and E357 amino acids, respectively.

[0027] Figure ID is a representation of the 3D structure of the CH3-CH3 region based on computer simulation of the M3 combination of mutations. The knob K409D and hole D399R mutations form a new hydrogen bond network (represented by dashed lines) with multiple amino acids at the CH3-CH3 interface.

[0028] Figure 2 shows a non-reduced SDS-PAGE electropherogram of VHH-Fc / Fc mutant combinations.

[0029] The VHH-Fc and Fc plasmid transfection ratio in Figure 2A is 1:1. The VHH-Fc and Fc plasmid transfection ratio in Figure 2B is 0.6:1. For each mutant combination shown, the heterodimer (AB) and other impurities (including homodimer and free monomer) contained in the protein A purified product present different migration distances due to the difference in molecular weight.

[0030] Figure 3 shows the Tm values of VHH-Fc / Fc mutant combinations determined by DSC method.

[0031] The Tm values of the samples were analyzed by using a Microcal PEAQ-DSC. Each mutant combination has one to three Tm values according to the Cp value change, and the highest Tm value reflects the stability of the CH3 domain.

[0032] Figure 4 shows the non-reduced SDS-PAGE electropherogram of VHH-Fc / Fab-Fc mutant combinations.

[0033] For each mutant combination shown, the heterodimer (ABC) and other impurities (homodimer, half antibody, free monomer) contained in the protein A purified product present different migration distances due to the difference in molecular weight.

[0034] Figure 5 shows the non-reduced SDS-PAGE electropherogram of Fab-Fc mutants.

[0035] For each mutant shown, the heavy chain and light chain plasmid transfection ratio is 1:1.5. For each mutant shown, the half antibody monomer, disulfide-linked homodimer and other impurities (free light chain and free heavy chain) contained in the protein A purified product present different migration distances due to the difference in molecular weight.

[0036] Figure 6 shows the in vitro assembly effect of Fab-Fc mutant combinations detected by HIC-HPLC.

[0037] The protein A affinity purified half antibody presents two main peaks, which are monomers or homodimers, respectively. The bispecific antibody generated by in vitro assembly presents one heterodimer main peak, and other small peaks are half antibody monomers or homodimer impurities.

[0038] Figure 7 shows the binding activity of bispecific antibodies to antigens determined by ELISA.

[0039] Figure 8 shows the binding affinity constant of bispecific antibodies to FcRn and FcyRI determined by Fortebio.

[0040] DETAILED DESCRIPTION

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In addition, the materials, methods, and examples described herein are illustrative only and are not intended to be limiting. Other features, objects, and advantages of the application will be apparent from the description and drawings, and from the claims.

[0042] I. Definitions

[0043] The term "about" when used in connection with a numerical value means encompassing numerical values within a range having a lower limit that is 5% less than the specified numerical value and an upper limit that is 5% greater than the specified numerical value.

[0044] In this document, the term "comprising" or "including" means including the recited elements, integers or steps, but not excluding any other elements, integers or steps. For example, referring to an Fc region comprising a CH3 domain, this encompasses an Fc region consisting of or consisting essentially of a CH3 domain, but also an Fc region further comprising other constant domains such as CH2.

[0045] In this document, the term "Fc scaffold" refers to a scaffold protein dimer consisting of or consisting essentially of paired and dimerized immunoglobulin Fc regions. This dimeric structure can serve as a scaffold for attachment or conjugation of other functional molecules at its N- or C-terminus, including, but not limited to, binding domains, biologically active polypeptides or fragments thereof, toxins. In some aspects, the scaffold can be used, inter alia, as a building block for constructing multispecific (e.g. bispecific) antibodies or immunoadhesins.

[0046] In this document, the term "heteromultimeric protein" refers to a molecule comprising at least a first polypeptide and a second polypeptide, wherein the second polypeptide differs from the first polypeptide in at least one amino acid residue in the amino acid sequence. Preferably, the heteromultimeric protein is a monospecific or multispecific binding protein comprising at least one target binding domain. In some aspects, the first polypeptide and the second polypeptide of the heteromultimeric protein comprise a multimerization domain (e.g. the first and second CH3 domains of a CH3 heterodimer according to the application; or the first and second Fc regions of a heterodimeric Fc scaffold according to the application), wherein said multimerization domain promotes the interaction of the first and second polypeptides at an interface and the formation of a "heterodimer". In some aspects, the heteromultimeric protein can comprise a "heterodimer" formed by the first and second polypeptides, or can comprise additional polypeptides in addition to the first and second polypeptides to form a higher order tertiary structure. Examples of heteromultimeric proteins include, but are not limited to, monospecific or multispecific binding proteins such as bispecific antibodies, bispecific immunoadhesins, or antibody / immunoadhesin chimeras.

[0047] As used herein, the terms "binding protein" and "binding molecule" are used interchangeably and in their broadest sense, and refer to a proteinaceous molecule that specifically binds to at least one target. One example of a binding protein is an antibody and an immunoadhesin.

[0048] In the present context, the term "Fc scaffold-based binding protein" refers to a binding protein comprising, or consisting of, or consisting essentially of, an Fc scaffold and at least one target binding domain linked thereto. In some embodiments, the binding protein comprises one (and only one) Fc scaffold. In other embodiments, the binding protein can comprise more than one Fc scaffold.

[0049] In the present context, the term "monospecific" refers to a binding molecule capable of specifically binding to one and only one specific target (e.g. one epitope). The term "multispecific" refers to a binding molecule capable of specifically binding to at least two different targets (e.g. at least two different epitopes). Typically, a multispecific binding molecule comprises at least two target binding domains each having specificity for a different target. In some embodiments, the different targets can be polypeptides or proteins or portions thereof, e.g. receptors, ligands, or antigenic epitopes, located on the same cell, different cells, the same antigen or different antigens. In some embodiments, a multispecific binding molecule is bispecific, capable of simultaneously binding two different targets, e.g. two different epitopes, e.g. two different epitopes expressed on two different cells, or two different epitopes on two different antigens, or two different epitopes on one and the same antigen.

[0050] In the present context, the term "valency" denotes the presence of a specific number of target binding domains in a binding molecule. Thus, the term "monovalent" in relation to a binding molecule denotes that one (and no more than one) target specific binding domain is present in said binding molecule. Correspondingly, the term "multivalent" denotes that a plurality of target specific binding domains are present in a binding molecule. For example, a "tetravalent" binding molecule denotes that a total of 4 target specific binding domains are present in the binding molecule, regardless of whether the targets to which said binding domains are directed are the same or not.

[0051] In the present context, the term "target binding domain" refers to a portion of a polypeptide or protein that provides for interaction with a target. Exemplary binding domains include the antigen binding domain of an antibody, a ligand, the receptor binding domain of a ligand, and the ligand binding domain of a receptor. In some aspects of the present application, the preferred target binding domain is the antigen binding domain of an antibody. The antigen binding domain of an antibody typically comprises amino acid residues from the complementarity determining regions (CDRs) of the antibody. A native immunoglobulin molecule typically has two antigen binding domains, a Fab molecule typically has a single antigen binding domain.

[0052] In the present context, the term "antigen binding domain" refers to an antibody or fragment thereof that specifically binds to an antigenic epitope. In some embodiments, the antigen binding domain comprises an antibody heavy chain variable region (VH) and an antibody light chain variable region (VL), or in the case of an antibody being a heavy chain antibody, a single heavy chain variable region (VHH). In some embodiments, the antigen binding domain can further comprise an antibody constant region. Useful heavy chain constant regions include any of the following five types: alpha, delta, epsilon, gamma, or mu. Useful light chain constant regions include any of the following two types: kappa and lambda. Thus, in some aspects, a particular antigen binding domain can be selected from, including but not limited to: an antibody fragment comprising an antibody heavy chain variable region (VH) and an antibody light chain variable region (VL), such as Fv, scFv, Fab, scFab, crossFab, and an antibody fragment comprising a heavy chain variable region of a heavy chain antibody, such as VHH.

[0053] In the present context, the term "antigenic determinant" is used interchangeably with "epitope" and refers to a site on a macromolecule, such as a polypeptide and a protein, with which an antigen binding domain interacts and forms a complex. In some aspects, an epitope can be, for example, a linear epitope consisting of a contiguous stretch of amino acids, or a conformational configuration composed of non-contiguous amino acid residues located in different regions. Useful antigenic determinants according to the present application can be, for example, on the surface of a tumor cell, a virus-infected cell, other diseased cell, an immune cell, or present in a free macromolecule of a body fluid and / or extracellular matrix (ECM).

[0054] In the present context, the term "antigen" refers to a macromolecule that can elicit an immune response in a mammal, thereby producing specific antibodies against it. In some aspects, an antigen useful according to the present application is a polypeptide or protein, including but not limited to, a naturally occurring protein from a vertebrate source and variants or fragments or derivatives thereof. The vertebrate includes a mammal, such as a primate, e.g., a human.

[0055] In the present context, the term "immunoglobulin" refers to a protein having the structure of a naturally occurring antibody. For example, an IgG class immunoglobulin is a heterotetrameric glycoprotein of about 150,000 daltons, composed of two light chains and two heavy chains that are disulfide-bonded. From N- to C-terminus, each immunoglobulin heavy chain has one variable region (VH), also called heavy chain variable domain, followed by three constant regions (CHI, CH2, and CH3), also called heavy chain constant regions. Similarly, from N- to C-terminus, each immunoglobulin light chain has one variable region (VL), also called light chain variable domain, followed by one constant region (CL), also called light chain constant region. In an IgG antibody molecule, typically the Vh-CHI of the heavy chain pairs with the VL-CL of the light chain to form a Fab fragment that specifically binds to an antigen. Thus, an IgG immunoglobulin is essentially composed of two Fab molecules and two dimerized Fc regions connected by an immunoglobulin hinge region. The heavy chains of an immunoglobulin can be assigned to one of five types, called alpha (IgA), delta (IgD), epsilon (IgE), gamma (IgG), or mu (IgM), based on the amino acid sequences of their constant regions. Certain classes can be further divided into subtypes, e.g., gamma 1 (IgG1), gamma 2 (IgG2), gamma 3 (IgG3), gamma 4 (IgG4), alpha 1 (IgA1), and alpha 2 (IgA2). The light chains of an immunoglobulin can also be assigned to one of two types, called kappa and lambda, based on the amino acid sequences of their constant domains.

[0056] In the present context, the term "antibody" refers to a polypeptide comprising one or more domains that bind to an epitope on an antigen of interest, wherein the binding domain(s) has / have sequence(s) derived from or having sequence identity with a variable region of an immunoglobulin. The term encompasses various antibody structures including, but not limited to, single chain antibodies, mono- / multi-specific antibodies, chimeric antibodies, humanized antibodies, human sequence antibodies, full-length antibodies, and antibody fragments, as well as antigen-binding proteins assembled from any of the same as components, provided that they exhibit the desired antigen-binding activity. Typically, an antibody forms an antigen-binding domain at the surface of the VH-VL dimer through 3 complementarity determining regions (HCDR1-3) in the heavy chain variable region (VH) and 3 complementarity determining regions (LCDR1-3) in the light chain variable region (VL). The 6 CDRs confer the specific binding of an antibody to an antigen. However, in the case of heavy chain antibodies, e.g., from camelids, an antibody can confer the specific binding of an antibody to an antigen through 3 complementarity determining regions (CDR1-3) in its single VHH domain (also referred to as VHH domain herein). The VHH domain, like the heavy and light chain variable regions of conventional IgG antibodies, comprises four conserved framework regions (FRs) and three complementarity determining regions (CDRs), and is arranged in the order FR1-CDR1-FR2-CDR2-FR3-CD3-FR4.

[0057] In the present context, the term "complementarity determining region" or "CDR region" or "CDR" or "hypervariable region" is a region in an antibody variable region that is highly variable in sequence and forms structurally defined loops ("hypervariable loops") and / or contains antigen contact residues ("antigen contacts"). CDRs are primarily responsible for binding to an epitope. In the VHH domains and the VH / VL domains of the antibodies of the present application, CDRs are sequentially numbered from the N-terminus, and are generally referred to as CDR1, CDR2 and CDR3. The CDR sequences in a given VHH domain and a given VH / VL domain can be determined using art-known schemes. The skilled person can readily determine the CDR sequence range of any given antibody variable region amino acid sequence at http: / / www.abysis.org / abysis / , including the CDR ranges defined by the Kabat, AbM, Chothia, Contact and IMGT schemes, and combinations thereof. Unless otherwise indicated, in the present application, the term "CDR" or "CDR sequence" encompasses CDR sequences determined in any of the above ways, and combinations thereof, but preferably CDR sequences defined according to the Kabat scheme.

[0058] In the present context, the term "antibody fragment" refers to a non- intact portion of an antibody that comprises an antigen binding domain. Examples of antibody fragments include, but are not limited to, Fv, Fab, Fab', Fab'-SH, F(ab')2, diabodies, linear antibodies, single-chain antibodies (e.g. scFv, scFab) and single-domain antibodies (sdAb). The amino acid sequence of a VH or VL in an antibody fragment can be modified by substitution, deletion, addition and / or insertion, as long as antigen binding ability is maintained. Furthermore, the variable region of an antibody fragment can be chimerized and humanized.

[0059] In the present context, the term "Fv domain" refers to the smallest antibody fragment that includes a complete antigen recognition and binding site. "Fv" is a dimer (VH-VL dimer) formed from a VH and a VL by strong association through non-covalent bonds.

[0060] In the present context, the term "scFv domain" refers to a single chain polypeptide in which two variable regions, typically a VH and a VL, are connected via a linker to form an antigen binding domain necessary for antigen binding.

[0061] In the present context, the term "Fab domain" refers to an antigen binding domain similar to the one formed in conventional four-chain IgG antibodies by pairing of a heavy chain variable region VH and a heavy chain constant region CH1 (VH-CH1) with a complementary light chain variable region VL and a light chain constant region CL (VL-CL). In the present context, the term "Fab domain" refers to an antigen binding domain similar to the one formed in conventional four-chain IgG antibodies by pairing of a heavy chain variable region VH and a heavy chain constant region CH1 (VH-CH1) with a complementary light chain variable region VL and a light chain constant region CL (VL-CL).

[0062] In the present context, the term "crossFab" or "crossFab domain" refers to a Fab domain, wherein CH1 and CL are exchanged, i.e. an antigen binding domain formed by pairing of VH-CL with VL-CH1.

[0063] In the present context, the term "scFab" refers to a Fab domain linked by an artificial linker into a single polypeptide chain.

[0064] In the present context, the term "VHH" or "VHH domain" is used to refer to a heavy chain variable domain from a heavy chain antibody that lacks a light chain. Thus, a VHH differs from a conventional VH of a four-chain immunoglobulin in that it does not need to pair with a light chain variable domain to form an antigen binding domain. Such VHH molecules can be derived from antibodies produced in camelids (e.g. camels, llamas, dromedaries, alpacas and guanacos). Other species than camelids can also produce heavy chain antibodies that naturally lack a light chain, and such VHHs are also within the scope of the present application. In some cases, for therapeutic applications of VHHs, it is desirable to reduce their immunogenicity. Thus, preferably, a VHH domain used in a binding protein of the present application comprises a humanized sequence.

[0065] In the present context, an "immunoglobulin constant domain" refers to a constant domain from or obtained or derived from an immunoglobulin heavy chain (e.g. a human IgGl heavy chain) or light chain, including the heavy chain constant domains CH1, CH2, CH3, and optionally CH4; and the light chain constant domain CL. The term includes native sequence and variant sequence constant domains.

[0066] In the present context, the "type" or "isotype" of an immunoglobulin constant domain refers to the type or isotype as determined based on its amino acid sequence. Heavy chain constant domains can be assigned to one of five distinct types, IgA, IgD, IgE, IgG, and IgM, or further assigned to a subtype (isotype), e.g. IgGl, IgG2, IgG3, IgG4, IgAl and IgA2, according to their sequence. Thus, in the present context, reference to an IgGl Fc region or CH3 domain refers to an Fc region or CH3 domain which can be classified as IgG type and further classified as IgGl subtype based on its amino acid sequence. Similarly, light chain constant domains can be assigned to kappa and lambda light chain CL domains according to their sequence. A person of skill in the art can readily determine the type or isotype to which a constant domain belongs by comparing its amino acid sequence to the sequences of the corresponding constant domains of naturally occurring immunoglobulins of different types or isotypes. In the present context, when a heterodimeric Fc scaffold or a heteromultimeric protein molecule of the application comprises multiple immunoglobulin heavy chain constant domains, it is understood that they can be selected independently of each other according to the intended function or use of the molecule. As an example, for a heteromultimer of the application comprising a CHI domain and CH2 and CH3 domains, the three constant domains can all be of IgGl subtype, e.g. of human IgGl subtype, or only the CH3 domain or the CH2 and CH3 domains are of IgGl subtype, e.g. of human IgGl subtype.

[0067] In the present context, an "IgG format" refers to an antibody which has the same Y-shaped structural features as an IgG immunoglobulin, essentially consisting of two Fab domains linked via an immunoglobulin hinge region (or, where appropriate, via a flexible linker peptide) to the N-terminus of a dimerized Fc domain. Thus, in a typical case, an antibody of IgG format consists of two heavy chains and two light chains, wherein each heavy chain has from N- to C-terminus VH, CHI, CH2 and CH3 domains; and each light chain has from N- to C-terminus VL and CL domains.

[0068] In the present context, an "IgG-like format" refers to an antibody which still retains the same Y-shaped structural features as an IgG immunoglobulin, but wherein one Fab domain is missing and / or at least one Fab domain is replaced by a binding domain of a different format (e.g. a scFv, a VHH, a ligand, or a ligand binding domain of a receptor). Such IgG-like format antibodies can be described by indicating the type of binding domain linked to the Fc dimeric scaffold. For example, VHH-Fc / Fc is a monovalent IgG-like format antibody consisting of a VHH-Fc polypeptide and a Fc polypeptide; a VHH-Fc / Fab-Fc antibody is a bivalent IgG-like format antibody consisting of a VHH-Fc polypeptide and a Fab-Fc polypeptide.

[0069] In the present context, the term "immunoadhesin" refers to an antibody-like protein molecule formed by fusion of a non-immunoglobulin binding domain having a desired binding specificity (e.g. a binding domain from a cell surface receptor or ligand, or a ligand itself) to an immunoglobulin constant domain (e.g. an Fc region or CH3 domain). The immunoglobulin constant domain can be obtained from any immunoglobulin, e.g. IgG, especially IgGl, IgG2, IgG3 or IgG4.

[0070] In the present context, the term "Fc domain" or "Fc region" is used to define the part of an immunoglobulin heavy chain constant region comprising a CH3 domain or fragments thereof. In some cases, the immunoglobulin part can also comprise one or more additional immunoglobulin constant domains or fragments thereof, including hinge regions, CHI or CH2 domains. In the present context, reference to the EU numbering system as described in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD, 1991 (also referred to as EU index) is used when referring to the amino acid of the Fc region and constant domains. Reference to the EU numbering of amino acid residues can also be made to the IMGT Scientific chart (https: / / www.imgt.org / IMGTScientificChart / Numbering / Hu_IGHGnber.html), which is hereby incorporated by reference, when referring to the human IgGl Fc region. According to this numbering, in the human IgGl immunoglobulin heavy chain, amino acids 118-215 are the CHI domain, amino acids 216-230 are the hinge region, amino acids 231-340 are the CH2 domain, and amino acids 341-447 are the CH3 domain. In some cases, the C-terminal lysine (Lys447) of the CH3 domain can be absent. One of skill in the art can readily determine the constant domains, their type / subtype and species origin comprised in the heterodimeric Fc regions and binding proteins according to the present application by sequence alignment of published natural immunoglobulin constant regions. In some embodiments, the first Fc region and the second Fc region of the heterodimeric Fc scaffold according to the present application comprise or consist of a CH3 domain. In other embodiments, the Fc region further comprises a CH2 domain. In yet other embodiments, the Fc region further comprises a hinge region or a partial hinge region, e.g. a lower hinge region portion comprising the sequence "CPPCP". In the present context, as understood by the skilled person, a CH3 heterodimer according to the present application is a special form of the heterodimeric Fc scaffold according to the present application, in which the first and second Fc region of the Fc scaffold consist or essentially consist of a CH3 domain. Thus, the description of the heterodimeric Fc scaffold according to the present application also equally applies to the CH3 heterodimer according to the present application, unless explicitly stated to the contrary.

[0071] In the context of the present application, the term "Fc domain" or "Fc region" encompasses native sequence Fc regions and variant sequence Fc regions. In the context of the present application, the term "native sequence Fc region" encompasses the naturally occurring Fc region sequences found in various immunoglobulin subtypes, as well as allotypes thereof (Gestur Vidarsson et al., IgG subclasses and allotypes: from structure to effector functions, 20 October 2014, doi: 10.3389 / fimmu.2014.00520.). In the context of the present application, the term "variant sequence Fc region" refers to an Fc region polypeptide comprising a modification relative to a native sequence Fc region polypeptide. The modification can be an addition, deletion, or substitution of an amino acid residue. Substitutions can include naturally occurring amino acids and non-naturally occurring amino acids. The purpose of the modification can be to alter the physicochemical properties of the Fc region, such as thermal stability and heterodimerization propensity, and / or the binding of the Fc region to its receptor and its effector functions.

[0072] In the context of the present application, an "Fc scaffold" is also referred to as an "Fc scaffold dimer" or "dimeric Fc scaffold". In the context of the present application, the term "homo- dimeric" Fc scaffold refers to a Fc scaffold in which the first and second Fc regions making up the Fc scaffold are identical in sequence. Accordingly, a "hetero-dimeric" Fc scaffold refers to a Fc scaffold in which the first and second Fc regions making up the Fc scaffold differ in sequence by at least one amino acid residue.

[0073] In the context of the present application, the term "CH3 dimer" refers to a pair of CH3 domains that are paired and dimerized. The term "CH3 homo-dimer" refers to a dimer in which the first and second CD3 domains making up the dimer are identical in sequence. Accordingly, a "CH3 hetero-dimer" refers to a dimer in which the first and second CD3 domains making up the dimer differ in sequence by at least one amino acid residue.

[0074] In the context of heteromultimeric proteins based on an Fc scaffold (e.g., monospecific / multispecific binding proteins), when such a multimeric protein comprises only one Fc scaffold, such a multimeric protein can be considered as a dimer composed of two protein units (i.e., monomers) depending on the two Fc regions of the scaffold, wherein one Fc member of the Fc scaffold and the polypeptides (e.g., target binding domains, which can be 0, 1 or more) linked thereto constitute one protein unit (i.e., monomer) of the dimer, and the other Fc member of the Fc scaffold and the polypeptides (e.g., target binding domains, which can be 0, 1 or more) linked thereto constitute the other protein unit (i.e., monomer) of the dimer. In this case, it should be understood that each monomer constituting the dimer can be a single or multiple polypeptide chain. When the two monomers are identical in sequence structure, i.e., are copies of each other, such a dimer is referred to herein as a homodimer. For example, a native IgG immunoglobulin is a typical example of such a homodimer. When the two monomers are not identical, e.g., differ in Fc sequence and / or binding domain sequence linked thereto, such a dimer is referred to herein as a heterodimer. For example, a bispecific IgG antibody or IgG-like antibody is a typical example of such a heterodimer. Monospecific / multispecific antibodies based on a heterodimeric Fc scaffold are another typical example of such a heterodimer.

[0075] In the context of heteromultimeric proteins according to the present application (e.g., binding proteins according to the present application), the term "purity" refers to the proportion of the heteromultimeric protein in the total protein in the purified product after expression and purification of the heteromultimeric protein from the host cell. The purity of the product can be determined by SEC-HPLC to determine the proportion of the desired heteromultimeric protein product to various chain mispairing products in the purified product. Preferably, the purity is greater than 90%.

[0076] In the context of heteromultimeric proteins according to the present application (e.g., binding proteins according to the present application), the term "thermal stability" refers to the fact that the protein exhibits a CH3 Tm value greater than 70°C as determined by DSC; or, when the protein is an antibody in the IgG configuration, exhibits a CH3 Tm value similar to that of a native human IgG1 immunoglobulin (e.g., ± 2°C, preferably ± 1°C). For heteromultimeric proteins according to the present application based on a heterodimeric Fc scaffold, preferably, the Tm value is determined without introducing a non-native disulfide bond in the heterodimeric Fc scaffold.

[0077] In the present text, when referring to a single mutation, the amino acid residue position at which the mutation occurs and the amino acid residues before and after the mutation are described, denoted as [original amino acid residue] mutation residue position [amino acid residue after mutation]. For example, a threonine to tryptophan substitution at position 366 of the Fc region is denoted as T366W; a tyrosine to valine substitution at position 409 of the Fc region is denoted as Y407V. When referring to a combination of multiple mutations, the combination of mutations occurring on the same polypeptide chain are connected using the symbol "-", for example, a combination of L351Y and D399R mutations occurring on the hole chain can be denoted as "L351Y-D399R"; for a combination of mutations occurring on different polypeptide chains, the mutations are separated using the symbol " / ", for example, a combination of L351Y and D399R mutations occurring on the hole chain and a K409D mutation occurring on the knob chain can be denoted as "hole L351Y-D399R / knob K409D".

[0078] In the present text, the term "flexible linker peptide" or "peptide linker" is used interchangeably to refer to a short amino acid sequence composed of amino acids, for example, glycine (G) and / or serine (S) and / or threonine residues (T), used individually or in combination, or a hinge region from an immunoglobulin or a modified version thereof.

[0079] In the present text, the term "conjugate" refers to a binding protein modifier or derivative formed by covalently linking or conjugating other molecules (e.g., therapeutic agent molecules or diagnostic agent molecules) to a protein of interest (e.g., a heteromultimeric protein or binding protein according to the present application). Examples of said other molecules that can be mentioned include, but are not limited to, proteins / polypeptides / peptides, labels, drugs, and cytotoxic agents, for example: radioisotopes; chemotherapeutic agents; growth inhibitory agents; enzymes and fragments thereof; fluorescent reporter proteins; antibiotics; toxins such as small molecule toxins or enzymatically active toxins of bacterial, fungal, plant or animal origin, including fragments and / or variants thereof; and known anti-tumour or anti-cancer agents of various kinds.

[0080] In the present text, the "percent (%) identity" of an amino acid sequence refers to the number of positions in the candidate sequence at which the amino acid residue is identical with the specified amino acid sequence shown herein in the comparison window and, if necessary to achieve the maximum percent sequence identity, gaps are introduced in the candidate sequence for optimal alignment with the specified amino acid sequence and the comparison window is compared, and the number of positions in the candidate sequence that do not match the specified amino acid sequence are counted, and the percent of total residues in the comparison window that are not matched is calculated. In the absence of specific indication, the comparison window is the full length of the specified amino acid sequence.

[0081] In the context of reference to constant domains CH2, CH3, Fc domain, the term "wild type" refers to a constant domain having a sequence from a native immunoglobulin constant domain, or having no more than 1-5 amino acid residue changes (preferably no more than 1, 2, 3, 4, or 5 amino acid residue changes, more preferably conservative amino acid substitutions) compared to that native sequence.

[0082] For polypeptide sequences, "conservatively modified" includes substitutions, deletions or additions to a polypeptide sequence which result in the substitution of an amino acid with a chemically similar amino acid. Conservative substitution tables providing functionally similar amino acids are well known in the art. The following eight groups contain amino acids that are conservative substitutions for one another: 1) Alanine (A), Glycine (G); 2) Aspartic acid (D), Glutamic acid (E); 3) Asparagine (N), Glutamine (Q); 4) Arginine (R), Lysine (K); 5) Isoleucine (I), Leucine (L), Methionine (M), Valine (V); 6) Phenylalanine (F), Tyrosine (Y), Tryptophan (W); 7) Serine (S), Threonine (T); and 8) Cysteine (C), Methionine (M) (see, e.g., Creighton, Proteins (1984)).

[0083] In the context of the present application, the term "host cell" refers to a cell into which an exogenous polynucleotide has been introduced, including the progeny of the cell. Host cells include "transformants" and "transformed cells," which include the primary transformed cell and progeny derived therefrom. Host cells are any type of cellular system that can be used to produce a polypeptide or protein molecule of the present application, including eukaryotic cells, e.g., mammalian cells, insect cells, yeast cells; and prokaryotic cells, e.g., E. coli cells. Host cells include cells in culture as well as cells within a transgenic animal, transgenic plant, or plant tissue or animal tissue in culture.

[0084] In the context of the present application, the term "expression vector" refers to a vector comprising a recombinant polynucleotide comprising an expression control sequence operably linked to a nucleotide sequence to be expressed. Expression vectors comprise sufficient cis-acting elements for expression; additional elements for expression can be provided by the host cell or in an in vitro expression system. Expression vectors include all those known in the art, including but not limited to, cosmids, plasmids (e.g., naked or contained within liposomes) and viruses (e.g., lentivirus, retrovirus, adenovirus, and adeno-associated virus).

[0085] In the context of the present application, the term "individual" refers to a mammal. Mammals include, but are not limited to, domesticated animals (e.g., cows, sheep, cats, dogs, and horses), primates (e.g., humans and non-human primates such as monkeys), rabbits, and rodents (e.g., mice and rats). In particular, the individual is a human.

[0086] As used herein, the term "treatment" refers to clinical intervention designed to alter the natural course of the disease in an individual receiving treatment. Desired effects of treatment include, but are not limited to, preventing occurrence or reoccurrence of disease, alleviating symptoms, reducing any pathological consequences of the disease, either directly or indirectly, preventing metastasis, decreasing rate of disease progression, ameliorating or palliating disease state, and improving or alleviating prognosis.

[0087] As used herein, the terms "cancer" and "tumor" are used interchangeably to refer to or describe a physiological condition in a mammal typically characterized by unregulated cell growth. Examples of cancer include, but are not limited to, carcinoma, solid tumor, and liquid tumor.

[0088] II. Heterodimeric Fc Scaffold

[0089] As shown in the examples, by introducing the specific Fc mutation combination of the present application, the Fc heterodimer ratio of the heterodimeric Fc scaffold according to the present application can reach more than 90% after expression in cells and purification by affinity chromatography column, higher than the control Fc protein containing only the corresponding knob-into-hole mutation; and can reach more than 98% purity after further purification by ion exchange column. Moreover, compared with the control Fc protein, the Fc heterodimer after purification has better thermal stability, in which the CH3 domain Tm value determined by DSC method can reach more than 74°C.

[0090] In the first aspect, therefore, the present application provides a heterodimeric Fc scaffold that can be used to enhance the formation of a heteromultimer of interest. By using the heterodimeric scaffold according to the present application, the yield of the heteromultimer of interest relative to the undesired heteromultimer and homomultimer impurities can be greatly improved, improving the production efficiency and cost of the heteromultimer of interest; at the same time, the thermal stability of the heteromultimer of interest product is improved.

[0091] The heterodimeric Fc scaffold according to the present application comprises two immunoglobulin Fc regions paired and heterodimerized, wherein the two Fc regions respectively comprise a CH3 domain. The heterodimeric Fc scaffold according to the present application is characterized in that, in the CH3 domain, not only a knob-into-hole (KIH) mutation is contained, but also the following mutations are contained:

[0092] (a) the Fc region containing the hole mutation further contains a L351Y mutation; or

[0093] (b) the Fc region containing the hole mutation further contains L351Y and D399R mutations, and the Fc region containing the knob mutation further contains K409D mutation.

[0094] In the present context, in some cases, the above-mentioned combination of mutations is also referred to as "CH3 interface characteristic mutations of the present application" for brevity. Likewise, for brevity, Fc regions having hole mutations and polypeptides comprising the same are also referred to as hole chains; and Fc regions having corresponding knob mutations and polypeptides comprising the same are referred to as knob chains. Accordingly, when referring to other mutations introduced on a hole chain, the expression "hole" or "hole" is added in front of the mutation to make clear that the mutation is introduced on a hole chain; and likewise, when referring to other mutations introduced on a knob chain, the expression "knob" or "knob" is added in front of the mutation to make clear that the mutation is introduced on a knob chain. For example, hole L351Y refers to the introduction of a L to Y residue substitution at position L351 of the Fc region of a hole chain; knob K409D refers to the introduction of a K to D residue substitution at position K409 of the Fc region of a knob chain.

[0095] Knob-into-hole (KIH) mutation technology is known in the art. See, e.g., US 7951917 B1 ; US 5,731,168; US 7,695,936; Ridgway et al., Prot Eng 9, 617-621 (1996) and Carter, J Immunol Meth 248, 7-15 (2001 ). Generally, this mutation technology involves introducing a protuberance ("knob" or knob mutation) in the Fc region CH3 domain interface of a first polypeptide and a corresponding cavity ("hole" or hole mutation) in the Fc region CH3 domain interface of a second polypeptide, such that the protuberance can be positioned in the cavity in order to promote "heterodimer" formation and impede homodimer formation of the first and second polypeptides. The protuberance can be constructed by replacing a small amino acid side chain from the interface of the first polypeptide with a larger side chain (e.g., tyrosine or tryptophan). By replacing a large amino acid side chain with a smaller amino acid (e.g., alanine or threonine), a complementary cavity of the same or similar size as the protuberance can be created in the interface of the second polypeptide. Such a combination of a spatially positioned and sized complementary protuberance and cavity mutation is referred to in the art as a KIH mutation, wherein the protuberance mutation is referred to as the knob mutation of the KIH mutation and the cavity mutation is referred to as the hole mutation of the KIH mutation.

[0096] In one embodiment, the Fc scaffold according to the application comprises a KIH mutation, wherein the CH3 domain of one of the first and second Fc regions of said Fc scaffold comprises a knob mutation; the CH3 domain of the other of the first and second Fc regions of said Fc scaffold comprises a complementary, corresponding hole mutation. In a more specific embodiment, in the CH3 domain of one of the first and second Fc regions of the Fc scaffold according to the application (e.g. the first Fc region), an amino acid residue substitution is introduced such that one or more amino acid residues are replaced by an amino acid residue having a larger side chain volume, thereby creating a protuberance in the CH3 domain of this Fc region; and simultaneously in the CH3 domain of the other of the first and second Fc regions of the Fc scaffold according to the application (e.g. the second Fc region), an amino acid residue substitution is introduced such that one or more amino acid residues are replaced by an amino acid residue having a smaller side chain volume, creating a cavity in the CH3 domain of this Fc region which is spatially complementary to the aforementioned protuberance, whereby the protuberance present on the CH3 domain of one Fc region can be positioned in the cavity present on the CH3 domain of the other Fc region. The protuberance and cavity can be made by altering the nucleic acid encoding the polypeptide, e.g. by site-specific mutagenesis, or by peptide synthesis. In some more specific embodiments, the Fc scaffold according to the application comprises a KIH mutation, wherein the introduced knob mutation is T366W, and the introduced hole mutation is Y407V. In other more specific embodiments, the Fc scaffold according to the application comprises a KIH mutation, wherein the introduced knob mutation is T366W, and the introduced hole mutation is T366S-L368A-Y407V.

[0097] In some embodiments, the Fc region of the heterodimeric Fc scaffold according to the application consists of or consists essentially of a CH3 domain. In this embodiment, the application thus provides a CH3 heterodimer according to the application.

[0098] In some embodiments, the heterodimeric Fc scaffold according to the application or the heteromultimeric protein according to the application comprising said scaffold comprises, in addition to the CH3 domain, further immunoglobulin constant domains, e.g. comprising CH2 and / or hinge regions in the Fc region, and / or comprising CH1 and / or CL domains linked to VH or VL in the target binding domain. In such embodiments, these immunoglobulin constant domains can have the same or different species origin (but preferably are both of human origin), and / or have the same or different immunoglobulin class or subtype (but preferably have the same IgG class or IgG subtype). In one embodiment, the Fc scaffold according to the application comprises CH2 and CH3 domains from an IgG, in particular human IgG, immunoglobulin. In one embodiment, the Fc scaffold according to the application further comprises a hinge region and / or a CH1 domain from an IgG, in particular human IgG, immunoglobulin. As the skilled person understands, when referring to the species origin and class of an immunoglobulin constant domain, it is meant that said constant domain comprises the natural amino acid sequence from an immunoglobulin of said species and class or a variant sequence thereof, wherein the amino acid changes in said variant sequence do not affect the species origin and class or subtype determination of said constant domain. Typically, for the purposes of the present application, said variant sequence has no more than 10 amino acid residue changes relative to the corresponding natural sequence. As an example, the structural features of the various IgG subtypes can be found in the review by Gestur Vidasson et al. (IgG Subclasses and Allotypes: From Structure to Effector Functions, Front Immunol. 2014; 5:520, doi: 10.3389 / fimmu.2014.00520), which is hereby incorporated by reference in its entirety. In some embodiments, the immunoglobulin constant domains comprised in the heterodimeric Fc scaffold according to the application or the heteromultimeric protein according to the application are constant domains of IgG immunoglobulins. In some embodiments, said constant domains can be independently of each other selected from different IgG1 subtypes, e.g. IgG1, IgG2, IgG3 and IgG4 subtypes, but preferably are both of IgG1 subtype or of IgG4 subtype. In some embodiments, said constant domains are human-derived IgG immunoglobulin constant domains.

[0099] In some specific embodiments, the heterodimeric Fc scaffold according to the application comprises an IgG Fc region. In some more specific embodiments, said Fc region is an IgG1 Fc region. In other specific embodiments, said Fc region is an IgG4 Fc region.

[0100] In some embodiments, the Fc region of the heterodimeric Fc scaffold according to the application has effector function. In some embodiments, the effector function is Fcy receptor binding and / or FcRn receptor binding. In some embodiments, the Fcy receptor is human FcyRIIa, FcyRI and / or FcyRIIIa, especially FcyRI. In some embodiments, the heterodimeric Fc scaffold according to the application (or binding molecule comprising said heterodimeric Fc scaffold) has not less than 60%, for example not less than 50%, 40%, 30%, 20% or 10% of the Fcy receptor binding affinity compared to a native IgGl dimeric Fc domain (or binding molecule comprising a native IgGl dimeric Fc domain). In some embodiments, the Fcy receptor is FcyRI. In some embodiments, the heterodimeric Fc scaffold according to the application (or binding molecule comprising said heterodimeric Fc scaffold) shows substantially similar FcRn binding affinity with respect to binding to the FcRn receptor as a native IgGl dimeric Fc domain (or binding molecule comprising a native IgGl dimeric Fc domain), i.e. more than about 70%, in particular more than about 80%, more particularly more than about 90% of the FcRn binding affinity of a native IgGl dimeric Fc domain.

[0101] Fc domains can confer advantageous pharmacokinetic properties to heteromultimeric proteins comprising them, e.g., bispecific antibodies, including prolonged serum half-life, facilitating good accumulation in target tissues and favorable tissue-blood partitioning ratios. However, Fc domains can at the same time cause heteromultimeric proteins comprising them, e.g., bispecific antibodies, to be targeted more to Fc gamma receptor-expressing cells than to desired cells and / or tissues bearing the target of interest. In addition, activation of the Fc gamma receptor signaling pathway can lead to cytokine release, causing heteromultimeric proteins comprising Fc domains to cause severe side effects upon systemic administration. In some embodiments, therefore, in addition to the CH3 interface characterizing mutations of the present application, the heterodimeric Fc scaffolds of the present application can comprise or not comprise additional mutations, as desired, to preserve or affect their effector function(s). For example, it is known that the upper part of the CH2 domain and the hinge region of an antibody are involved in various effector functions, e.g., Fc gamma receptor binding and effector functions such as ADCC; and that some CH2 and CH3 interface residues of an antibody are involved in FcRn receptor binding. In some cases, the Fc region used in the heterodimeric scaffolds of the present application can be altered (reduced or enhanced) in one or more desired effector functions, e.g., binding affinity of the Fc region to Fc gamma receptors and / or FcRn, by comprising additional mutations introduced in the CH2 domain and / or the hinge region, if present. Such amino acid modifications are known in the art, including, but not limited to, amino acid substitutions at one or more positions selected from the group consisting of E233, L234, L235, N297, P331, and P329, and more particularly, one or more amino acid substitutions selected from the group consisting of E233P, L234A, L235A, L235E, N297A, N297D, and P331S. In some cases where the Fc scaffold has a human IgGl Fc region, the binding of the Fc scaffold to Fc gamma receptors can be substantially reduced by comprising L234A-L235A mutations, L234A-L235A-P329G mutations, and / or N297A mutations in the Fc region.

[0102] In the present application, some exemplary sequences of the Fc scaffolds according to the present application are provided. In some embodiments, the heterodimeric Fc scaffold of the present application comprises a first polypeptide and a second polypeptide, wherein the first polypeptide comprises a first Fc region, and the second polypeptide comprises a second Fc region, wherein:

[0103] (a) the CH3 domains of the first and second Fc regions comprise, in addition to the above-mentioned characteristic CH3 interface mutations of the application, 0-3 amino acid residue alterations relative to a wild-type CH3 domain, preferably the wild-type CH3 domain is a native IgG (in particular IgG1 or IgG4, preferably human IgG1) immunoglobulin CH3 domain, more preferably comprises the amino acid sequence of SEQ ID NO: 19 or 20;

[0104] (b) the first and second Fc regions comprise a native IgG (in particular IgG1 or IgG4, preferably human IgG1) immunoglobulin CH2 domain, or a CH2 domain comprising the amino acid sequence set forth in SEQ ID NO: 21 or having 1-5 amino acid residue alterations thereto;

[0105] (c) the Fc region comprising the knob mutation comprises the amino acid sequence of SEQ ID NO: 1 or 2 or an amino acid sequence having at least 95%, 96%, 97%, 98% or 99% identity thereto;

[0106] (d) the Fc region comprising the hole mutation comprises the amino acid sequence of SEQ ID NO: 7 or 8 or an amino acid sequence having at least 95%, 96%, 97%, 98% or 99% identity thereto, and / or

[0107] (e) the Fc scaffold does not form a non-native disulfide bond between the first Fc region and the second Fc region.

[0108] In some preferred embodiments, the Fc region comprising the knob mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 1 or 2, or an amino acid sequence having at least 95% identity thereto; and the Fc region comprising the hole mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 7 or 8, or an amino acid sequence having at least 95% identity thereto. In some embodiments, the Fc region comprising the knob mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 1 or 2, and the Fc region comprising the hole mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 7 or 8. In some embodiments, the Fc region comprising the knob mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 1, and the Fc region comprising the hole mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 7. In some embodiments, the Fc region comprising the knob mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 2, and the Fc region comprising the hole mutation comprises a CH3 domain of the amino acid sequence set forth in SEQ ID NO: 8. In any of the above-mentioned embodiments, the Fc region comprising the knob mutation can be the first Fc region and the Fc region comprising the hole mutation is the second Fc region, or vice versa.

[0109] In some embodiments, the heterodimeric Fc scaffold according to the present application does not comprise a non-native disulfide bond introduced into the interface of the CH3 domains of the first and second Fc regions, and still has good thermal stability, e.g., the Tm value of the CH3 domains is 70 °C or higher according to DSC assay.

[0110] In some embodiments, the heterodimeric Fc scaffold according to the present application has better thermal stability compared to a corresponding control Fc scaffold having the same KIH mutations. In this context, the “corresponding control Fc scaffold” refers to a heterodimeric Fc scaffold that is identical to the heterodimeric Fc scaffold according to the present application in all sequence structures including the KIH mutations, except that it has unmutated native residues at positions L351, D366 and K409 relative to the heterodimeric Fc scaffold according to the present application with which it is compared.

[0111] In some embodiments, the heterodimeric Fc scaffold according to the present application has comparable purity compared to a corresponding control Fc scaffold having the same KIH mutations. In some embodiments, after expression and assembly in host cells and purification under the same conditions, the heterodimeric Fc scaffold according to the present application has a purity that is at least 90%, 95%, 100%, 110% or more of the purity of the control Fc scaffold, and preferably has a purity greater than 90% according to SEC-HPLC assay.

[0112] III. Heteromultimeric Proteins

[0113] In a second aspect, the present application provides a heteromultimeric protein comprising a CH3 heterodimer according to the present application or a heterodimeric Fc scaffold according to the present application.

[0114] In some embodiments, the heteromultimeric protein according to the present application comprises at least a first and a second polypeptide, wherein the first and second polypeptide comprise a first and a second CH3 domain, respectively, wherein the first and second CH3 domains in the first and second polypeptide meet at the interface and form a CH3 heterodimer according to the present application.

[0115] In some embodiments, the heteromultimeric protein according to the present application comprises at least a first and a second polypeptide, wherein the first and second polypeptide comprise a first and a second Fc region, respectively, wherein the first and second Fc regions in the first and second polypeptide meet at the interface and form a Fc scaffold according to the present application.

[0116] In some embodiments, the heteromultimeric protein is a mono- or multispecific binding protein comprising at least one target binding domain, wherein the at least one target binding domain is linked to a CH3 heterodimer according to the present application or a heterodimeric Fc scaffold according to the present application. In some embodiments, the heteromultimeric protein according to the present application comprises more than one of said Fc scaffold. In some embodiments, the heteromultimeric protein according to the present application comprises only one of said Fc scaffold. Such mono- or multispecific binding proteins comprising the Fc scaffold of the present application are also referred to herein as "Fc scaffold based binding proteins".

[0117] Fc scaffold based binding proteins

[0118] In some aspects, the present application thus provides binding proteins based on the heterodimeric Fc scaffold of the present application. In some embodiments, the binding protein according to the present application comprises at least one target binding domain linked to the Fc scaffold.

[0119] Types of target binding domains

[0120] The type of target binding domain that can be used in the binding proteins of the application is not particularly limited, so long as it has the ability to bind a target of interest. Such binding domains, as are known in the art, can be from binding partners, including non-immunoglobulin-based binding pairs, such as receptor / ligand binding pairs, enzyme / substrate pairs, or immunoglobulin-based binding pairs, such as antigen / antibody binding pairs, antibody / anti-idiotypic antibody. Where the target binding domain is from an immunoglobulin-based binding pair, the binding protein comprising it can be an antibody, e.g., a mono / multispecific antibody. Where the target binding domain is from a non-immunoglobulin-based binding pair, the binding protein comprising it can be an immunoadhesin. In addition, the application contemplates binding proteins comprising binding domains from both of these two different classes of binding pairs, such antibody / immunoadhesin chimera proteins, which are sometimes referred to herein as antibodies.

[0121] In some embodiments, the binding proteins according to the application comprise at least one antigen binding domain. The antigen that can be bound by the antigen binding domain according to the application can be selected from, for example, but not limited to, a tumor associated antigen, an immune checkpoint molecule, an angiogenic factor, or a combination thereof. The tumor associated antigen can include, but is not limited to, MUC1-1, BCMA, CLDN18.2, HER2, BRAF, EGFR, CD20, CD38, FolR1, and CD52. The immune checkpoint molecule can include, but is not limited to, PD-L1, PD-1, PD-L2, CTLA-4, B7-H3, TIM3, LAG-3, VISTA, ICOS, 4-1BB, OX40, GITR, and CD40. The angiogenic factor can include, but is not limited to, basic FGF, HGF, Stie-2, Svegfr-1, Svegfr-2, EGF, IL-6, IL-8, PLGF, VEGF, PDGF-bb, ANG1, ANG2, SDF-1a, MDC, Galectin, TSP-1, Endocan, Enos, HIF-1a. In some embodiments, the antigen is selected from a tumor associated antigen, e.g., EGFR and HER2.

[0122] In some embodiments, the binding proteins according to the application comprise at least one non-immunoglobulin binding domain.

[0123] In some embodiments, examples of such non-immunoglobulin binding domains that can be used include, but are not limited to, ligand binding domains from the following receptor proteins: 4-1BB; adrenocorticotropic hormone receptor; activin receptor; BLTR (blotin leukotriene B4 receptor); BMP receptor; C3a receptor; C5a receptor; chemokine receptor; cytokine receptor; growth hormone receptor; BTLA; interferon-alpha receptor; interferon-beta receptor; interferon-gamma receptor; IL-1 type I receptor; IL-1 type II receptor; IL-10 receptor; IL-11 receptor; IL-12 receptor; BCMA; TACI; BAFF receptor; immunomodulatory signaling receptor CD72; Kaposi sarcoma-associated herpesvirus GPCR; lipoxin A4 receptor; lymphotoxin beta receptor; RON receptor; SCF receptor; somatostatin receptor; T1 / ST2; TGF-beta receptor; tumor necrosis factor receptor; TNFRSF19; erythropoietin receptor; leukemia inhibitory factor receptor; and C-kit receptor.

[0124] In some embodiments, examples of such non-immunoglobulin binding domains that can be used include, but are not limited to, receptor binding domains from the following ligands or the ligands themselves: alpha-MSH; 9E3 / cCAF; adrenocorticotropic hormone; activin; AK155; angiogenesis inhibitor; Apo2L / TRAIL; BLR1 ligand / BCA-1 / BLC / CXCL13; calcitonin gene-related peptide; CD27 ligand; CD30 ligand; CD40 ligand; endorphin; endostatin; erythropoietin; Fas ligand; Flt-3 ligand; G-CSF; GCP-2 / CXCL6; GM-CSF; various cytokines, e.g., IFN alpha, IFN beta; interferon gamma; IL-1 alpha; IL-1 beta; IL-10; IL-11; IL-12; IL-13; IL-15; IL-16; IL-2; IL-27; IL-3; IL-4; IL-5; IL-6; IL-7; IL-8 / CXCL8; IL-9; somatostatin; stem cell factor; substance P; TARC / CCL17; TCA3 / mouse CCL1; TECK / CCL25; TGF beta; thrombopoietin; TNF alpha.

[0125] In some embodiments, the non-immunoglobulin binding domain is from a protein selected from the group consisting of: VEGF receptor; TNF receptor; IL-1; lymphocyte function-associated antigen 3 (LFA-3 / CD58); lymphotoxin beta receptor (LTBR); CTLA-4; IL-12; activin receptor; TACI; BR3; T cell receptor; CD4; L-selectin; homing receptor; CD44; NP receptor; interferon gamma receptor; 4-1BB; and IgE receptor. In other embodiments, the non-immunoglobulin binding domain is from a receptor binding domain of a ligand selected from the group consisting of: IL-1, IL-12, or the ligand itself. Examples of other useful non-immunoglobulin binding domains can also be found, for example, in US7951917B1 and US20220275048A1.

[0126] Manner of linking target binding domains

[0127] In the binding proteins according to the application, the position of linking the target binding domains to the Fc scaffold is not particularly limited. In some embodiments, the at least one target binding domain is independently linked to the N- or C-terminus of the first and / or second polypeptide of the Fc scaffold. In some embodiments, at least two or more or all of the at least one target binding domain are linked to each other, optionally via a peptide linker, and in turn to the heterodimeric Fc scaffold according to the application. In other embodiments, at least two or more or all of the at least one target binding domain are separately linked to different termini of the Fc scaffold, optionally via a peptide linker. For example, when the binding protein comprises 2 binding domains, one binding domain can be linked to the N-terminus of the first polypeptide of the Fc scaffold, and the other binding domain is linked to a different terminus of the Fc scaffold, e.g., the C-terminus of the first polypeptide, the N-terminus of the second polypeptide, or the C-terminus of the second polypeptide. Thus, in the binding proteins according to the application, at least one, two, three, or all of the four termini of the heterodimeric Fc scaffold (i.e., the N- and C-termini of the first and second polypeptides) can each be linked to at least one (preferably 1 or 2) target binding domain.

[0128] The at least one target binding domain comprised in the binding protein can be the same or different from each other. In one embodiment, at least two, more or all of the at least one target binding domain are the same. In other embodiments, at least two, more or all of the at least one target binding domain are different from each other.

[0129] In some embodiments, the binding protein is monospecific, and has a single or multiple binding domains for this specificity. In some embodiments, the binding protein is multispecific, and preferably, wherein for each specificity, there is a single or multiple binding domains.

[0130] In some embodiments, the at least one target binding domain is, independently of each other, an antigen binding domain selected from the group consisting of Fab, VHH, scFv, scFab and crossFab. In case the binding domain is a Fab or crossFab, the Fab or crossFab can be linked to another binding domain or Fc scaffold of the binding protein, e.g. to the N-terminus of the first or second polypeptide of a heterodimeric Fc scaffold according to the application, either by the C-terminus of the chain comprising the VH domain or the C-terminus of the chain comprising the VL domain, optionally via a peptide linker.

[0131] In some embodiments, at least one of the at least one target binding domain is a VHH domain. In other embodiments, at least one of the at least one target binding domain is a Fab domain. In other embodiments, at least one of the at least one target binding domain is a VHH domain, and at least one of the at least one target binding domain is a Fab domain. In other embodiments, at least two, multiple or all of the at least one target binding domain are Fab domains, and the Fab domains have the same or different VLCL light chain.

[0132] In some embodiments, the binding protein is a monospecific monovalent protein comprising a VHH domain, and preferably the VHH domain is linked at the N-terminus of the first or second polypeptide of the Fc scaffold. In some embodiments, the binding protein is a bispecific bivalent binding protein comprising a VHH and / or Fab binding domain. In some embodiments, the binding protein is a bispecific antibody of IgG configuration comprising a first and a second Fab domain and an Fc scaffold, preferably the first and second Fab domains are linked at the N-terminus of the first and second polypeptide of the Fc scaffold, respectively, by the C-terminus of the chain comprising the VH domain thereof. In some embodiments, the binding protein is a bispecific antibody of IgG-like configuration comprising a VHH and a Fab, wherein the antibody consists of one heavy chain comprising a VHH-Fc domain and one heavy chain comprising a VH-Fc domain and one light chain comprising a VL-CL domain.

[0133] In the context of the present application, the term "linking" as used in connection with different components of a scaffold or binding protein means that the components are fused or conjugated, either directly or via a suitable linker. Peptide linkers typically used for this purpose are short amino acid polypeptides of some flexibility, generally between 1 and 50 amino acid residues in length, but not limited thereto. In some cases, any chain of amino acid residues located between two components to be linked can be considered a peptide linker, as long as it does not interfere with the intended function of the two components to be linked. Thus, in some aspects, a heterodimeric Fc scaffold according to the present application, or a CH3 heterodimer according to the present application, is linked to a polypeptide of interest, such as a target binding domain according to the present application, or other functional polypeptide or protein domain, by a peptide linker, are within the contemplation of the present application.

[0134] In the heteromultimeric protein according to the present application, the skilled person can readily determine the available linker sequences depending on the components to be linked and the position of linkage. In some embodiments, a linker is used to link the Fc scaffold according to the present application and the antigen binding domain; and / or to link two antigen binding domains. In some embodiments, the available linker is a flexible linking peptide of 5-50 amino acids, preferably comprising glycine (G) and / or serine (S) and / or threonine residues (T). In some embodiments, the linker has a length of 5-30 amino acids, e.g. 8, 10, 15, 20, 25 or 30 amino acids, or a length of amino acids falling in between any two integers. In some embodiments, the linker comprises the amino acid sequence (G4S) nwherein n is an integer equal to or greater than 1, for example, n is an integer of 2, 3, 4, 5, 6 or 7. In a preferred embodiment, the linker consists of the amino acid sequence (G4S)2. In other embodiments, the linker is a hinge region from an immunoglobulin or a derivative thereof. Examples of other useful linkers include, for example but not limited to, the following amino acid sequences: (Gly3Ser)2(SEQ ID NO: 22), (Gly4Ser)2(SEQ ID NO: 12), (Gly3Ser)3(SEQ ID NO: 24), (Gly4Ser)3(SEQ ID NO: 25), (Gly3Ser)4(SEQ ID NO: 26), (Gly4Ser)4(SEQ ID NO: 27), (Gly3Ser)5(SEQ ID NO: 28), (Gly4Ser)5(SEQ ID NO: 29), (Gly3Ser)6(SEQ ID NO: 30), (Gly4Ser)6(SEQ ID NO: 31), GGG (SEQ ID NO: 32), DGGGS (SEQ ID NO: 33), TGEKP (SEQ ID NO: 34), GGRR (SEQ ID NO: 35), EGKSSGSGSESKVD (SEQ ID NO: 36), KESGSVSSEQLAQFRSLD (SEQ ID NO: 37), GGRRGGGS (SEQ ID NO: 38), LRQRDGERP (SEQ ID NO: 39), LRQKDGGGSERP (SEQ ID NO: 40) and GSTSGSGKPGSGEGSTKG (SEQ ID NO: 23). Alternatively, if necessary, suitable flexible linking peptides can be rationally designed using computer programs to model the three-dimensional structure of proteins and peptides, or by phage display methods.

[0135] In a preferred embodiment, a peptide linker having the sequence shown in SEQ ID NO: 12 is used to link the Fc scaffold and the antigen binding domain, e.g. VHH or Fab, at its N-terminus.

[0136] In some cases, chemical synthetic linkers can also be used to link binding domains to Fc scaffolds. In this case, the linkage is also referred to as "conjugation." Examples of chemical synthetic linkers are: N-hydroxysuccinimide (NHS), disuccinimidyl suberate (DSS), bis(sulfosuccinimidyl) suberate (BS3), dithiobis(succinimidyl propionate) (DSP), dithiobis(sulfosuccinimidyl propionate) (DTSSP), ethylene glycol bis(succinimidyl succinate) (EGS), ethylene glycol bis(sulfosuccinimidyl succinate) (Sulfo-EGS), tartaric acid disuccinimidyl ester (DST), tartaric acid disulfosuccinimidyl ester (Sulfo-DST), bis[2-(succinimidooxycarbonyloxy)ethyl]sulfone (BSOCOES), bis[2-(sulfosuccinimidooxycarbonyloxy)ethyl]sulfone (Sulfo-BSOCOES), and the like.

[0137] When multiple linkers are used in the heteromultimeric proteins of the application, it is understood that the linkers can be the same as or different from each other.

[0138] In some cases, chemical synthetic linkers can also be used to conjugate chemical synthetic therapeutic agents (e.g., small molecule toxins) or diagnostic agents to the heteromultimeric proteins of the application, e.g., bispecific antibodies, e.g., to the sugar chains of the CH2 domains. Such conjugates are also contemplated by the application.

[0139] Exemplary heteromultimeric proteins

[0140] In some embodiments, the heteromultimeric protein of the application is a multispecific binding protein, in particular a bispecific antibody, comprising binding to different epitopes of the same antigen. Bispecific antibodies with binding specificity for at least two different antigens or antigenic epitopes have a wide range of potential clinical applications, as targeted drugs for in vitro and in vivo immunodiagnostic and therapeutic uses, and for diagnostic immunoassay tests. See, e.g., WO9850431A2. Bispecific antibodies can be used in in vitro assays for probing the functional properties of cell surface molecules and for determining the cytotoxicity mediated by different Fc receptors (Fanger et al., Crit. Rev. Immunol. 12:101-124 (1992)); for enzyme-linked immunoassays (Nolan et al., Biochem. Biophys. Acta. 1040:1-11 (1990); Hammerling et al., J. Exp. Med. 128:1461-1473 (1968)); and for immunodiagnosis of various diseases, e.g., cancer, in vitro or in vivo (Songsivilai et al., Clin. Exp. Immunol. 79:315 (1990)). To facilitate diagnostic use of BsAb, one arm of the BsAb can bind to an antigen on the surface of diseased tissue or cells (e.g., tumor-associated antigens), while the other arm can bind to a detectable label, e.g., a chelator that tightly binds a radionuclide. In therapeutic applications, bispecific antibodies can be used to direct a patient's cellular immune defense mechanisms specifically against diseased tissue / cells, e.g., tumor cells, or pathogens. To this end, one arm of the BsAb can bind to an antigen on the surface of an immune cell (e.g., T cell, NK cell), while the other arm can bind to an antigen on the surface of the diseased tissue / cells.

[0141] In some embodiments, the multispecific antibody according to the application is a heterodimer consisting of or mainly consisting of one heterodimeric Fc scaffold according to the application and at least one antigen binding domain attached thereto, having a configuration selected from the group consisting of:

[0142] -(antigen binding domain)n-Fc / (antigen binding domain)n-Fc,

[0143] -(antigen binding domain)n-Fc-(antigen binding domain)m / (antigen binding domain)n-Fc,

[0144] -(antigen binding domain)n-Fc-(antigen binding domain)m / (antigen binding domain)n-Fc-(antigen binding domain)m,

[0145] wherein each of n and m above is independently of the other selected from the group consisting of 0, 1 and 2;

[0146] each of the above antigen binding domains is independently selected from the group consisting of: Fv, scFv, Fab, scFab, crossFab, VHH, a ligand binding domain of a ligand or a receptor;

[0147] the symbol "-Fc" denotes a linkage at the N-terminus of the Fc;

[0148] the symbol "Fc-" denotes a linkage at the C-terminus of the Fc;

[0149] wherein the two monomers of the heterodimer are separated by the symbol " / ";

[0150] wherein either of the two Fc is a knob chain, while the other is a hole chain.

[0151] In some preferred embodiments, n is 1 and m is 0. In some preferred embodiments, n is 0 and m is 1. In some embodiments, n is 1 and m is 1. In some preferred embodiments, each of the above antigen binding domains is independently selected from the group consisting of: scFv, VHH, and Fab.

[0152] In some specific embodiments, the multispecific heterodimeric protein according to the application is bispecific, and has an IgG configuration or an IgG-like configuration.

[0153] In some specific embodiments, the multispecific heterodimeric protein according to the application is a bispecific antibody, and has a configuration selected from the group consisting of:

[0154] - scFv-Fc / scFv-Fc, wherein the scFv domains for two different specificities are respectively linked at the N-terminus of the Fc scaffold according to the application;

[0155] - Fab-Fv / Fab-Fc, wherein the Fab domains for two different specificities are respectively linked at the N-terminus of the Fc scaffold according to the application;

[0156] - VHH-Fc / VHH-Fv, wherein the VHH domains for two different specificities are respectively linked at the N-terminus of the Fc scaffold according to the application;

[0157] - scFv-Fc / Fab-Fc, wherein the scFv domains and Fab domains for two different specificities are respectively linked at the N-terminus of the Fc scaffold according to the application;

[0158] - scFv / VHH-Fc, wherein the scFv domains and VHH domains for two different specificities are respectively linked at the N-terminus of the Fc scaffold according to the application;

[0159] - VHH / Fab-Fc, wherein a VHH domain and a Fab domain for two different specificities are respectively connected at the N-terminus of the Fc scaffold according to the application.

[0160] In embodiments where the heteromultimeric protein according to the application comprises two or more Fab domains for different specificities, for example in the case of a bispecific antibody of Fab-Fc / Fab-Fc configuration, light chain mispairing can be prevented by using a common light chain. See for example US 7951917B. Alternatively, the first Fab-Fc half antibody and the second Fab-Fc half antibody can be produced separately, followed by in vitro assembly into the desired heterodimer by addition of an appropriate reducing agent in their mixture.

[0161] In some embodiments, the heteromultimeric protein of the application is a bispecific antibody comprising a VHH domain and a Fab domain that bind different antigens, respectively. In some embodiments, the bispecific antibody binds HSA and B7-H3. In some embodiments, the VHH domain comprises the CDR1, CDR2 and CDR3 sequences in the VHH amino acid sequence set forth in SEQ ID NO: 11; or comprises the amino acid sequence set forth in SEQ ID NO: 11, or an amino acid sequence at least 95% identical thereto. In some embodiments, the Fab domain comprises a VH and a VL, wherein the VH comprises the CDR1, CDR2 and CDR3 sequences in the VH amino acid sequence set forth in SEQ ID NO: 13; or comprises the amino acid sequence set forth in SEQ ID NO: 13, or an amino acid sequence at least 95% identical thereto; and wherein the VL comprises the CDR1, CDR2 and CDR3 sequences in the VL amino acid sequence set forth in SEQ ID NO: 14; or comprises the amino acid sequence set forth in SEQ ID NO: 14, or an amino acid sequence at least 95% identical thereto. In some embodiments, the VHH domain comprises the amino acid sequence of SEQ ID NO: 11; and the Fab comprises a VH of the sequence set forth in SEQ ID NO: 13 and a VL of the sequence set forth in SEQ ID NO: 14. In some embodiments, the bispecific antibody is a bivalent antibody. In some embodiments, the VHH and Fab domains are respectively connected at the N-terminus of the first and second polypeptides of the heterodimeric Fc scaffold according to the application.

[0162] In some embodiments, the binding protein of the application is a bispecific antibody comprising first and second Fab domains that bind different antigens, respectively. In some embodiments, the bispecific antibody binds EGFR / HER2. In some embodiments, the first Fab domain comprises a VH and a VL, wherein the VH comprises CDR1, CDR2, and CDR3 sequences in the VH amino acid sequence set forth in SEQ ID NO: 15; or comprises the amino acid sequence set forth in SEQ ID NO: 15, or an amino acid sequence at least 95% identical thereto; and wherein the VL comprises CDR1, CDR2, and CDR3 sequences in the VL amino acid sequence set forth in SEQ ID NO: 16; or comprises the amino acid sequence set forth in SEQ ID NO: 16, or an amino acid sequence at least 95% identical thereto. In some embodiments, the second Fab domain comprises a VH and a VL, wherein the VH comprises CDR1, CDR2, and CDR3 sequences in the VH amino acid sequence set forth in SEQ ID NO: 17; or comprises the amino acid sequence set forth in SEQ ID NO: 17, or an amino acid sequence at least 95% identical thereto; and wherein the VL comprises CDR1, CDR2, and CDR3 sequences in the VL amino acid sequence set forth in SEQ ID NO: 18; or comprises the amino acid sequence set forth in SEQ ID NO: 18, or an amino acid sequence at least 95% identical thereto. In some embodiments, the first Fab domain comprises a VH of the sequence set forth in SEQ ID NO: 15 and a VL of the sequence set forth in SEQ ID NO: 16; and the second Fab comprises a VH of the sequence set forth in SEQ ID NO: 17 and a VL of the sequence set forth in SEQ ID NO: 18. In some embodiments, the bispecific antibody is a bivalent antibody. In some embodiments, the first and second Fab domains are linked at the N-terminus of the first and second polypeptides, respectively, of a heterodimeric Fc scaffold according to the application.

[0163] IV. Polynucleotides, vectors, host cells

[0164] In some aspects, the present application provides one or more polynucleotides encoding a CH3 heterodimer, a heterodimeric Fc scaffold, a heteromultimeric protein or a binding protein according to the present application; one or more vectors comprising the one or more polynucleotides, wherein the one or more polynucleotides can be present in a single vector or in separate multiple vectors. In other aspects, the present application also provides host cells comprising the one or more nucleic acids or the one or more vectors. Depending on the case, the single or multiple vectors comprising the one or more polynucleotides can be introduced into the same host cell for expression in one host cell to produce the desired product; or the single or multiple vectors comprising the one or more polynucleotides can be introduced into different host cells separately for expression in different host cells to produce intermediates comprising different chains or different combinations of chains (e.g., different monomeric proteins) of the heteromultimer according to the present application, respectively, and by mixing the intermediates under conditions suitable for assembly of the heteromultimer according to the present application, the heteromultimer according to the present application is produced.

[0165] In the present application, the expression vector that can be used is not particularly limited, and includes, but is not limited to, a virus, a plasmid, a cosmid, a lambda phage, or a yeast artificial chromosome (YAC).

[0166] In the present application, the host cell that can be used is not particularly limited. Suitable host cells include prokaryotic microorganisms such as E. coli, eukaryotic microorganisms such as filamentous fungi or yeast, or various eukaryotic cells such as mammalian host cells, insect cells, and the like. Examples of useful mammalian host cell lines include the SV40-transformed monkey kidney CV1 line (COS-7), human embryonic kidney line (HEK293 or 293F cells), baby hamster kidney cells (BHK), monkey kidney cells (CV1), African green monkey kidney cells (VERO-76), human cervical carcinoma cells (HELA), canine kidney cells (MDCK), buffalo rat liver cells (BRL 3A), human lung cells (W138), human liver cells (Hep G2), CHO cells, NSO cells, myeloma cell lines such as YO, NS0, P3X63, and Sp2 / 0, and the like. In some embodiments, a mammalian cell line suitable for suspension culture can be used. In some preferred embodiments, the host cell is a CHO or HEK293 cell.

[0167] V. Production and purification of the heteromultimeric protein of the present application

[0168] In a further aspect, the present application provides a method for producing a heteromultimeric protein of the present application. For producing a heteromultimeric protein of the present application, polypeptide chains of the heteromultimeric protein of the present application can be obtained, e.g., by solid state peptide synthesis (e.g., Merrifield solid phase synthesis) or recombinant production, and assembled under suitable conditions. As previously described, in the case of recombinant production, the assembly can occur in the host cell used for expressing the heteromultimeric protein; or, if necessary, the assembly of the heteromultimeric protein can be performed in vitro after the expressed intermediate polypeptide chains have been harvested from the host cell.

[0169] In one embodiment, the present application thus provides a method for producing a heteromultimeric protein of the present application, the method comprising: culturing a host cell comprising a polynucleotide encoding a polypeptide chain of the heteromultimeric protein under conditions suitable for expression of the polypeptide chain; and assembling the polypeptide chain under conditions suitable for assembly of the polypeptide chain into the heteromultimeric protein to produce the heteromultimeric protein.

[0170] In some embodiments, the method comprises: culturing a host cell comprising a polynucleotide encoding a polypeptide chain of the heteromultimeric protein under conditions suitable for expression of the polypeptide chain; and recovering the heteromultimeric protein produced by the host cell from the cell culture. Preferably, the recovery product has a purity of greater than 80%, preferably greater than 85%, more preferably greater than 90% of the heteromultimeric protein of the present application as determined by SEC-HPLC after protein A affinity chromatography.

[0171] In other embodiments, the method comprises: culturing a host cell comprising a polynucleotide encoding a polypeptide chain of the heteromultimeric protein under conditions suitable for expression of the polypeptide chain; and recovering an intermediate of the heteromultimeric protein produced by the host cell from the cell culture; mixing the intermediates under conditions suitable for assembly to produce the heteromultimeric protein, and recovering the produced heteromultimeric protein. In the case of a heterodimeric protein, the monomer comprising the knob chain and the monomer comprising the hole chain are separately expressed in different host cells, and the heteromultimeric protein is produced and recovered from the mixture of the monomers under conditions suitable for assembly. Preferably, the recovery product has a purity of greater than 80%, preferably greater than 85%, more preferably greater than 90% of the heteromultimeric protein of the present application as determined by SEC-HPLC after ion exchange chromatography.

[0172] The antibodies prepared by the methods described herein can be purified by known art such as high performance liquid chromatography, ion exchange chromatography, gel electrophoresis, affinity chromatography, size exclusion chromatography, and the like. After purification, the purity of the heteromultimeric protein of the application can be determined by any of a number of well-known analytical methods including size exclusion chromatography, gel electrophoresis, high performance liquid chromatography, and the like. The physical / chemical properties and / or biological activities of the heteromultimeric proteins provided herein can be identified, screened for, or characterized by a variety of assays known in the art.

[0173] VI. Compositions and Uses

[0174] The present application also relates to a composition comprising a heteromultimeric protein of the application (e.g., a binding protein according to the application). In embodiments where the composition is a pharmaceutical composition, the composition further comprises a pharmaceutically acceptable carrier. As used herein, "pharmaceutically acceptable carrier" includes any and all solvents, dispersion media, isotonic agents, and absorption delaying agents, and the like that are physiologically compatible.

[0175] In the present application, a pharmaceutical composition generally refers to a drug for the treatment or prevention of a disease or for examination, diagnosis. In some preferred embodiments relating to the pharmaceutical composition of the present application, the heteromultimeric protein of the present application is a bispecific antibody, e.g., a bispecific antibody in which at least one specificity targets a disease-associated antigen. In some embodiments, the heteromultimeric protein of the present application is the only active ingredient in the pharmaceutical composition. In other embodiments, the pharmaceutical composition can comprise a heteromultimeric protein described herein in combination with one or more other therapeutic agents.

[0176] The pharmaceutical composition of the present application can be formulated into preparations according to a method well known to those skilled in the art. In addition, it can be considered that a pharmaceutically acceptable carrier or medium, e.g., sterilized water or physiological saline, vegetable oil, emulsifier, suspending agent, surfactant, stabilizer, flavoring agent, excipient, solvent, preservative, binding agent, etc., is appropriately combined with the heteromultimeric protein of the present application to formulate a unit dose form suitable for the needs of drug administration. For example, a sterile solution or suspension can be prepared using a pharmaceutically acceptable carrier to be administered parenterally in the form of an injection. The amount of active ingredient in the preparation will be set to produce a predetermined range of effective amount after administration to an individual.

[0177] The pharmaceutical compositions of the present application are suitable for various routes of administration, including but not limited to, intravenous, intramuscular, subcutaneous, parenteral, rectal, spinal or epidermal administration (e.g., by injection or infusion). Accordingly, the pharmaceutical compositions of the present application can be formulated into dosage forms for administration by the intended route of administration, e.g., injection dosage forms, nasal administration dosage forms, pulmonary administration dosage forms, transdermal administration dosage forms. For injection dosage forms, the route of administration can include, but is not limited to, intravenous injection, intramuscular injection, intraperitoneal injection, subcutaneous injection, intratumoral injection, and the like for systemic or local administration. In addition, the specific method of administration of the pharmaceutical compositions of the present application can be appropriately selected depending on the age, symptoms of the patient.

[0178] In addition, detection or diagnostic kits comprising the heteromultimeric proteins described herein are also within the scope of the present application. The kits can comprise one or more other elements, including, for example: instructions for use; other reagents, such as labels or reagents for conjugation; pharmaceutically acceptable carriers; and devices or other materials for administration to a subject.

[0179] The following examples are set forth to assist in understanding the application. No limitation to the scope of the application is intended or should be inferred as a result of the inclusion of the examples. Examples

[0180] Example 1 Designing Fc Mutation Candidate Combination Sequences

[0181] The Fc heterodimer crystal structures of knob-into-hole (KIH) models (e.g., 4NQS and 5ID8) were searched from the Protein Data Bank (www.pdb.org) and compared with the wild-type IgG1 Fc crystal structure (e.g., 1HZH) to analyze the reasons for the reduced thermal stability of the KIH models. It was found that the KIH mutations reduced the number and strength of the inter-contacting amino acids of the two CH3 domains.

[0182] In wild-type IgGl Fc, some key amino acids at the CH3-CH3 interface form a hydrophobic core, including L351, L368, Y407, etc., covering a large surface area (see FIG. 1A). However, in the KIH model, because L368A and Y407V are mutated into small side chain amino acids, the surface area of the hydrophobic core is significantly reduced (see FIG. IB). In addition, the contact between some amino acids near the hydrophobic core is also affected. For example, in wild-type IgGl Fc, the K409 side chain of one CH3 domain directly contacts L368, K370, D399, F405, Y407 amino acids of the other CH3 domain; in particular, K409 forms a complex hydrogen bond network with D399, S364, T411 and a water molecule, while preventing electrostatic clash between K409 and K370. But in the KIH structure, the K409 side chain in the knob chain is affected by the T366W mutation, losing direct contact with L368A, Y407V, and also losing the stabilizing water molecule and the hydrogen bond network mediated by it.

[0183] To enhance the CH3 interaction strength and improve thermal stability, we selected amino acids near the Fc contact interface of the KIH model as mutation objects as much as possible, and finally designed a series of Fc heterodimer mutation combinations. The specific amino acid mutation information is shown in Table 1. The KIH combination is the knob-hole mutation combination disclosed in patent US7951917B1. M1-M7 are newly designed candidate mutation combinations.

[0184] Table 1. Fc heterodimer mutation combination design, where A chain is the chain containing knob mutation, and B chain is the chain containing corresponding hole mutation

[0185] Example 2 VHH-Fc / Fc heterodimer expression and purification

[0186] In this example, in order to verify the effect of the Fc mutant combinations in Table 1 on promoting heterodimer formation, we first designed A chain VHH-Fc fusion molecules and B chain Fc molecules. On this basis, we synthesized VHH-Fc and Fc genes, constructed expression plasmids, collected culture supernatant after transfection of 293F cells, purified by Protein A affinity, and analyzed the content of heterodimer and other impurities in the purified product using SDS-PAGE and SEC-HPLC methods to determine the effect of different mutation combinations on the expression yield and purity of VHH-Fc / Fc heterodimer.

[0187] Expression vector construction

[0188] According to the constant region amino acid sequence of human immunoglobulin IgG1 (P01857) on the protein database Uniprot, the human IgG1-Fc amino acid sequence was obtained, including the hinge region, CH2 domain and CH3 domain. Then 10 Fc mutant amino acid sequences were designed according to Table 1 (see SEQ ID NO: 1 to SEQ ID NO: 10). A VHH fragment (from Ozoralizumab antibody, see SEQ ID NO: 11) and a flexible linker polypeptide (see SEQ ID NO: 12) were introduced at the N-terminus of the A chain Fc fragment to construct the A chain VHH-Fc fusion molecule; at the same time, the B chain Fc fragment remained unchanged. The molecular weight difference between VHH-Fc and Fc is beneficial to distinguish the proportion of AB heterodimer, AA / BB homodimer, A / B monomer or high molecular weight aggregate (HMWS) formed during recombinant expression.

[0189] A chain VHH-Fc and B chain Fc DNA coding genes were designed using codon optimization for mammalian / human expression, cloned into the expression vector pcDNA3.4 (Invitrogen) to obtain plasmids for expressing VHH-Fc and Fc. Gene synthesis and plasmid construction were completed by Suzhou Genwiz Company (Genwiz).

[0190] 293F cell transfection

[0191] The VHH-Fc and Fc plasmids were mixed in a molar ratio of 1:1 or 0.6:1, and then 1 / 10 transfection volume of OPM-CD Trans293 medium (Opmi, P82019) was added. FectoPRO transfection reagent (PolyPlus, PT-116-010) was added to the medium at a ratio of 1 μl / 1.2 ml cells. The medium containing the plasmid and transfection reagent was mixed and allowed to stand for 10 min. The mixture was added to a shake flask containing Expi293F cells (Invitrogen, A14635) and incubated at 37°C, 8% CO2. After 16 h of transfection, 10% transfection volume of feed (Opmi, F081918-001) was added and incubated for another 6 days.

[0192] Protein A purification

[0193] 1 ml Protein A Diamond column (BioRad, AA0273) was equilibrated with 5 column volumes of equilibration buffer (20 mM sodium phosphate, pH 7.4). The clarified cell culture supernatant was loaded onto the column at a flow rate of 0.5 column volumes / min. The column was then washed sequentially with 5 column volumes of equilibration buffer (20 mM sodium phosphate, pH 7.4) and pre-elution buffer (20 mM acetic acid-sodium acetate, pH 5.2). Finally, the column was eluted with 10 column volumes of elution buffer (20 mM acetic acid-sodium acetate, pH 3.2) and the collected eluate was adjusted to pH 6.0 with 1.5 M Tris base.

[0194] SDS-PAGE analysis

[0195] 5 μg of the Protein A column purified product was mixed with non-reducing loading buffer (without DTT) and run on a 10% SurePAGE gel (Genscript, M00666). After electrophoresis, the gel was stained with Coomassie Brilliant Blue and destained before taking a picture with a gel imager.

[0196] The results are shown in Figure 2. The co-expression products of VHH-Fc plasmid and Fc plasmid mainly contained five different molecular weight proteins: AB heterodimer, AA homodimer, BB homodimer, A monomer and B monomer. At the plasmid ratio of 1:1, the heterodimer content of KIH control, M1, M2, M5, M6 combinations was high, greater than 90%, but the heterodimer content of M3, M4, M7 combinations was low, between 80-90%. At the plasmid ratio of 0.6:1, the heterodimer content of M1, M3, M5, M6, M7 combinations was high, greater than 90%, but the heterodimer content of KIH, M2, M4 combinations was low, between 80-90%. The results showed that M1 and M6 combinations had higher ability to promote AB heterodimer formation and inhibit AA / BB homodimer formation compared to KIH. M2 combination had similar ability to promote AB heterodimer and AA / BB homodimer formation compared to KIH control. M3 combination had similar ability to promote AB heterodimer formation compared to KIH combination, but M3 combination induced more AA homodimer and less BB homodimer formation.

[0197] SEC-HPLC analysis

[0198] SEC test was performed on 1 mg / ml Protein A column purified product using Theromo Vanquish Core high performance liquid chromatograph, the chromatographic column was Waters BioResolve SEC mAb (specification: 2.5 μm, 7.8 x 300 mm), isocratic elution was performed with PBS as the mobile phase, the elution time was 30 minutes, the flow rate was 0.5 ml / min, the detection wavelength was 280 nm, the column temperature was 20°C, the injection volume was 10 μl, and the temperature control plate temperature was 5°C.

[0199] The results are shown in Table 2. Overall, the ratio of dimers produced by each mutant combination calculated by SEC-HPLC analysis was similar to the results of SDS-PAGE analysis. Under the condition of plasmid ratio 1:1, the content of heterodimers of KIH control, M1, M2, M5, M6 combinations was high, more than 90%, but the content of heterodimers of M3, M4, M7 combinations was low, between 80-90%. Under the condition of plasmid ratio 0.6:1, the content of heterodimers of M1, M3, M5, M6, M7 combinations was high, more than 90%, but the content of heterodimers of KIH, M2, M4 combinations was low, between 80-90%.

[0200] Table 2 SEC-HPLC analysis of VHH-Fc / Fc mutant combinations

[0201] Based on the above results, it can be seen that the introduction of hole L351Y mutation on the basis of KIH basically does not change the formation ratio of heterodimers and homodimers. The introduction of K409D / D399R double mutation on the basis of KIH-L351Y basically does not change the heterodimer ratio, but induces more AA homodimers and less BB homodimers.

[0202] Example 3 Thermal stability analysis of VHH-Fc / Fc heterodimers

[0203] Heterodimer purification

[0204] 1 ml Mono S ion exchange column (Cytiva, 17516801) was washed with 5 column volumes of equilibration buffer (20 mM sodium phosphate, pH 6.0) or to baseline. The pH of the Protein A affinity purified sample was adjusted to 6.0 with 1.5 M Tris base or 50 mM sodium acetate (pH 3.2), and the sample was loaded onto the Mono S column at a flow rate of 0.5 column volumes / min. Then wash with 10 column volumes of equilibration buffer. Finally, elution was performed with elution buffer (20 mM sodium phosphate + 1 M sodium chloride, pH 6.0) according to a 0-30% gradient, and the eluate was collected in separate tubes. SEC-HPLC was used to detect the heterodimer purity of each tube sample, and samples with a purity greater than 98% were combined.

[0205] Thermal stability assay

[0206] Thermal stability of the sample was analyzed by using Microcal PEAQ-DSC. The method was as follows: the protein sample was diluted to 1 mg / mL in PBS, and the temperature range was set from 20 to 100 °C with a heating rate of 90 °C / h. PBS was added in the cell as a blank buffer, and temperature scanning was performed. Then the solution in the sample cell was removed, and the sample to be tested was added, and temperature scanning was performed. The experimental data were processed using the analysis software provided by the instrument, and the Tm value was calculated.

[0207] The Tm determination results of the VHH-Fc / Fc mutant combinations are shown in Table 3 and Figure 3. The KIH control has at least two Tm values, wherein the first Tm value (Tm1) 65.68 °C reflects the thermal stability of the VHH fragment and the CH2 domain, and the second Tm value (Tm2) 71.48 °C reflects the thermal stability of the CH3 domain. The Tm values of the M1 and M6 mutant combinations are only 67.56 °C and 66.31 °C, which are significantly lower than the KIH control. The results show that the introduction of electrostatic steering design on the basis of KIH structure will cause the unexpected decrease of the thermal stability of the CH3 domain. Interestingly, the CH3 domain Tm value of the M2 mutant combination reaches about 74.03 °C, which is about 2.5 °C higher than the KIH control; and the CH3 domain Tm value of the M3 mutant combination is even higher, reaching about 75.84 °C. The results show that the single amino acid mutation of B chain Fc L351Y can improve the stability of the CH3 domain of KIH form; and further, the combination of this mutation with A chain Fc K409D and B chain Fc D399R amino acid mutations has a synergistic enhancement effect, further improving the stability of the CH3 domain. Other mutant combinations, including M4, M5, M7, have poor thermal stability effect, and the Tm values are all less than 70 °C. These results show that the introduction of A chain L351Y or B chain L351Y-D399K on the basis of KIH cannot improve the thermal stability of the CH3 domain.

[0208] Table 3 Tm values of VHH-Fc / Fc mutant combinations

[0209] Example 4 Expression and purification of VHH-Fc / Fab-Fc heterodimer

[0210] To verify the versatility of the mutation combination of Example 2 on different antibody sequences and molecular formats, a VHH-Fc / Fab-Fc heterodimer molecule was additionally constructed. A VHH antibody fragment (amino acid sequence see SEQ ID NO: 11) and a flexible linker polypeptide (see SEQ ID NO: 12) were introduced at the N-terminus of the first Fc fragment, forming a VHH-Fc fusion protein of A chain. A Fab antibody fragment was introduced at the N-terminus of the second Fc fragment, constituting a Fab-Fc half-antibody. The Fab amino acid sequence was derived from a B7-H3 mouse monoclonal antibody obtained by hybridoma technology, containing a light chain domain VL-CL (amino acid sequence see SEQ ID NO: 14) and a heavy chain domain VH-CH1 (amino acid sequence see SEQ ID NO: 13). The heavy chain of the Fab-Fc antibody was designated as B chain, and the light chain as C chain. The molecular weight difference between the VHH-Fc fusion protein and the Fab-Fc half-antibody facilitated the differentiation of various forms of protein components formed during recombinant expression, including ABC heterodimer, AA homodimer, BB homodimer, BBCC homodimer, A monomer, B monomer and C monomer. Among them, the target bispecific antibody was the ABC heterodimer, and the others were impurities.

[0211] Similar to that described in Example 2, the A chain, B chain and C chain genes were synthesized, plasmids expressing A chain, B chain and C chain respectively were constructed, 293F cells were transfected, culture medium supernatant was collected, Protein A affinity purification was performed, and the purified product was analyzed using SDS-PAGE and SEC-HPLC methods to analyze the content of ABC heterodimer and other impurities.

[0212] The SDS-PAGE analysis results are shown in Figure 4. Two different plasmid transfection ratios (A chain plasmid: B chain plasmid: C chain plasmid) conditions were used, which were 1:1:1.5 or 0.6:1:1.5, respectively. The KIH control had a purity of less than 90% ABC heterodimer under both conditions. Other mutant combinations, M1, M2, M3, M5, M6 and M7, had a purity of more than 90% ABC heterodimer under at least one condition. However, the M4 mutant combination was ineffective, and the purity of the ABC heterodimer was less than that of the KIH control under both conditions.

[0213] The SEC-HPLC analysis results are shown in Table 4. Under both conditions of ABC plasmid ratio of 1 : 1 : 1.5 and 0.6: 1 : 1.5, the expression of M1-M7 mutant combination molecules were all greater than 200 mg / L, similar to or higher than the KIH control. The proportion of ABC heterodimers produced by each mutant combination calculated by SEC-HPLC was generally similar to the results of SDS-PAGE analysis. The purity of ABC heterodimers of the KIH control was less than 90% under both conditions. Under the condition of ABC plasmid ratio of 1 : 1 : 1.5, the proportion of ABC heterodimers of M1-M7 mutant combinations were all higher than the KIH control, among which the proportion of M1, M2, M3, M5, M7 heterodimers was higher than 90%. Under the condition of ABC plasmid ratio of 0.6: 1 : 1.5, the proportion of M1, M6 and M7 combination heterodimers was higher than the KIH control, and was greater than 90%.

[0214] Table 4 SEC-HPLC analysis of VHH-Fc / Fab-Fc mutant combinations

[0215] Example 5 Fab-Fc half antibody expression and purification

[0216] To verify the universality of the mutant combinations of Example 2 on different antibody sequences and molecular forms, two different Fab-Fc half antibody mutants were also constructed by connecting two Fab fragments binding to different antigens in the Fc region. The A chain Fc designed in Table 1 of Example 1 was connected to the N-terminal of the first Fab to obtain A half antibody, wherein the Fab is from anti-EGFR monoclonal antibody Zalutumumab, containing heavy chain domain VH-CH1 (amino acid sequence see SEQ ID NO: 15) and light chain domain VL-CL (amino acid sequence see SEQ ID NO: 16). The B chain Fc designed in Table 1 of Example 1 was connected to the N-terminal of the second Fab to obtain B half antibody, wherein the Fab fragment is from anti-HER2 monoclonal antibody Pertuzumab, containing heavy chain domain VH-CH1 (amino acid sequence see SEQ ID NO: 17) and light chain domain VL-CL (amino acid sequence see SEQ ID NO: 18).

[0217] Similar to that described in Example 2, the light chain and heavy chain genes of A and B half antibodies were synthesized, plasmids expressing the genes respectively were constructed, 293F cells were transfected to express A half antibody and B half antibody respectively, the culture supernatant was collected, Protein A affinity purification was performed, and the purified products were analyzed by SDS-PAGE and SEC-HPLC methods.

[0218] The SDS-PAGE analysis results are shown in Figure 5. For all the tested half antibody mutants, the 293 cell expression products mainly formed half antibody monomers with a molecular weight of 75 KDa and homodimers with a molecular weight of 150 KDa, as well as a small amount of light chains with a molecular weight of 25 KDa and heavy chains with a molecular weight of 50 KDa. Among them, the monomer proportion of Fab-Fc mutants K0, K2, K7, H0, H2 and H7 was greater than 80%; the monomer proportion of mutants K9 and H8 was less than 50%.

[0219] The SEC-HPLC analysis results are shown in Table 5. The expression amount of all the half antibody mutants in 293 cells was greater than 100 mg / L. The half antibody monomer, homodimer and multimer proportions in the 293 cell expression products were calculated by using the analysis software of SEC-HPLC. The sum of the monomer and dimer contents produced by all the Fab-Fc mutants was greater than 95%, and the multimer content was less than 5%. K0 produced 46.27% dimers, and K2 produced 56.17% dimers, which should include non-covalent dimers and disulfide bond linked covalent dimers. This result indicates that the K409D mutation has no effect on the homodimerization of CH3 domain. K9 produced about 86.2% dimers. K9 produced significantly more dimers than K0, indicating that the L351Y mutation promotes the homodimerization of CH3 domain. K7 produced 26.51% dimers and 73.24% monomers. H0 produced 30.4% dimers, which should include non-covalent dimers and disulfide bond linked covalent dimers. H2 produced 14.38% dimers and H7 produced 15.05% dimers. H2 and H7 produced significantly less homodimers than H0, indicating that the D399R mutation prevents the homodimerization of CH3 domain by electrostatic repulsion. H8 produced 64.71% dimers. H8 produced significantly more dimers than H0, indicating that the L351Y mutation promotes the homodimerization of CH3 domain.

[0220] Table 5 SEC-HPLC analysis results of Fab-Fc half antibodies

[0221] Example 6 Preparation and purification of Fab-Fc / Fab-Fc heterodimers

[0222] In vitro assembly

[0223] The half antibody solution obtained by affinity chromatography in Example 5 was mixed with the corresponding half antibody combination in equimolar ratio (1:1) according to Table 6, and then neutralized to pH 8-8.2 with 1.5M Tris base, with a total protein amount of 2mg in the reaction system. After standing at room temperature for 30 minutes, 200x molar ratio of GSH solution (Sigma, S0073) was added. The reaction sample was incubated at 37°C for 4 hours, and samples were taken at 0.5, 1, 2, and 4 hours, respectively, and the in vitro assembly efficiency was detected immediately. The reaction sample was transferred to a 4°C refrigerator and stood overnight.

[0224] HIC-HPLC (hydrophobic interaction high performance liquid chromatography) analysis of purity

[0225] HIC test was performed using Theromo Vanquish Core high performance liquid chromatograph, the chromatographic column was TOSOH TSKgel Bulty-NPR (specification: 2.5μm, 4.6x100mm), 1.5M ammonium sulfate and 20mM phosphate mixed solution (pH 7.0) was used as mobile phase A, 20mM phosphate solution (pH 7.0) was used as mobile phase B for gradient elution (0-15min 0% B-100% B, 15.5-22min 0% B-0% B), the flow rate was 0.6ml / min, the detection wavelength was 280nm, the column temperature was 25°C, the injection volume was 10μl, and the temperature control disc temperature was 5°C.

[0226] The HIC-HPLC chromatograms of half antibodies and double antibodies reacted for 4 hours are shown in Figure 6. This method can sensitively detect the assembly efficiency of double antibody heterodimer molecules. Because the monomers or homodimers formed by half antibodies have different hydrophobicities, they have different peak times in the HIC chromatogram. The peak times of the double antibody molecules generated by the in vitro assembly of two Fab-Fc are also different from those of the half antibodies.

[0227] The HIC-HPLC analysis results are shown in Table 6. Under the condition of incubation at 37°C for 0.5 hours, the proportion of AB heterodimers generated by the KIH, M1, M2 and M3 mutant combinations was already greater than 90%. Under the condition of incubation at 37°C for 2 hours and 4 hours, the proportion of AB heterodimers generated by the KIH, M1, M2 and M3 mutant combinations increased to about 95%.

[0228] Table 6 HIC-HPLC analysis of double antibody assembly efficiency

[0229] Heterodimer purification

[0230] 1 ml Capto S ImpAct ion exchange column (Cytiva, 17371751) was rinsed with 5 column volumes of equilibration buffer (20 mM sodium phosphate, pH 6.0) or to baseline. In vitro assembled samples were adjusted to pH 5.8-6.2 with 1 M acetic acid, then diluted 1-2 fold with ultrapure water, and filtered through a 0.22 um PES filter. Samples were loaded onto the Capto S column at a flow rate of 0.5 column volumes / minute. The column was then rinsed with 10 CV of equilibration buffer. Finally, the column was eluted with an elution buffer (20 mM sodium phosphate + 1 M sodium chloride, pH 6.0) according to a 0-30% gradient, and the eluate was collected in fractions. SEC-HPLC was used to analyze the purity of the dimer peak for each fraction, and fractions with purity greater than 98% were combined.

[0231] The results of SEC-HPLC analysis are shown in Table 7. The combination of KIH, M1, M2, and M3 mutants produced a dimer ratio greater than 96% under the condition of 4 hours incubation at 37 °C, with the main component being AB heterodimer and a small amount of AA or BB homodimer. Other impurities included A or B monomer, or high molecular weight aggregates, all with a ratio less than 2%.

[0232] Table 7 SEC-HPLC analysis of bispecific purity produced by 4 hours incubation at 37 °C

[0233] Example 7 Bispecific binding activity to antigen

[0234] The purified bispecific antibody of Example 6 was tested for binding activity to antigens. Recombinant human EGFR protein (ACRO Biosystems, EGR-H5222) and HER2 protein (ACRO Biosystems, HE2-H5225) were diluted to 1 ug / ml with coating buffer (Solarbio, C1055) and 100 ul was added to each well and incubated at 4°C for 16 hours. Blocking buffer was prepared by adding 3% BSA (Sangon, 9048-46-8) and 0.05% Tween 20 (Sangon, A600560-0500) to PBS solution. After washing the plate with 0.05% PBST for 3 times, 300 ul of blocking buffer was added to each well and incubated at room temperature for 2 hours. The bispecific antibody was diluted with 0.5% BSA-0.05% PBST blocking buffer, with a maximum concentration of 15 ug / ml and a 3-fold serial gradient. After washing the plate with 0.05% PBST for 3 times, 100 ul of bispecific antibody solution was added to each well and incubated at room temperature for 1 hour. After washing the plate with 0.05% PBST for 3 times, 100 ul of HRP-labeled secondary antibody (Bethyl, A80-304P, diluted 1:5000) was added to each well and incubated at room temperature for 1 hour. After washing the plate with 0.05% PBST for 3 times, 100 ul of color developing solution (Solarbio, PR1210) was added to each well and incubated at room temperature in the dark for 10 minutes. 100 ul of stop solution (Solarbio, C1058) was added to each well and the OD450 wavelength absorbance value was read using a multifunctional enzyme label instrument (Tecan, Spark).

[0235] The ELISA results are shown in Figure 7. The bispecific antibodies produced by the KIH, M1, M2 and M3 mutant combinations have very similar antigen binding curves. The bispecific antibodies bind to EGFR with an IC50 value of between 17.0-24.9 ng / ml and to HER2 with an IC50 value of between 12.8-14.6 ng / ml. The results show that mutations in the Fc region CH3 domain do not affect the antigen binding ability of the bispecific antibodies.

[0236] Example 8 Binding activity of bispecific antibodies to Fc receptors

[0237] The affinity of the purified bispecific antibody of Example 6 for binding to human FcyRI receptor and human FcRn was determined using the Bio-Layer Interferometry (BLI) technique. The Fortebio Octet Red 96 instrument was used for detection and calculation of the equilibrium dissociation constant (KD).

[0238] A column of HIS1 K (Fortebio, Cat 18-5120) sensors were soaked in regular buffer (1 x PBS, pH 7.4 with 0.1% BSA, 0.02% Tween-20). After the sensors were equilibrated in regular buffer, 5 ug / ml of recombinant human FcyRI protein (ACRO Biosystems, Cat FCA-H52H1) was immobilized, followed by 100 nM antibody binding and dissociation in regular buffer. Instrument run steps: Baseline1 (60 s), Loading (~15 s, 0.5 nm), Baseline2 (60 s), Association (60 s) and Dissociation (120 s) at 1000 rpm and 30 °C.

[0239] A column of SA (Fortebio, Cat 18-5019) sensors were soaked in pH 6.0 buffer (1 x PBS, pH 6.0, with 0.1% BSA, 0.02% Tween-20). After the sensors were equilibrated in pH 6.0 buffer, 200 ul of pH 6.0 buffer, 100 nM of antibody and human FcRn were added to a 96-well black polystyrene microplate (Greiner, 655209) respectively. Fortebio Octet Red96 was used for detection, according to the sample position layout, the sensor position was selected. After the sensors were equilibrated in pH 6.0 buffer, 2 ug / ml of recombinant human FcRn protein (ACROBiosystems Cat FCM-H8286) was immobilized, followed by 100 nM antibody binding and dissociation in pH 6.0 buffer. Instrument run steps: Baseline1 (60 s), Loading (~100 s, 3.0 nm), Baseline2 (60 s), Association (60 s) and Dissociation (60 s) at 1000 rpm and 30 °C.

[0240] The results of Fortebio test are shown in Figure 8. The affinity of control antibody IgG1 to human FcyRI was 6.33 x 10 -9 The affinity of KIH, M1, M2 and M3 mutant combinations to human FcyRI was between 1.01 x 10 -8 and 1.03 x 10 -8 The affinity of control antibody IgG1 to human FcRn was 1.41 x 10 -8 The affinity of KIH, M1, M2 and M3 mutant combinations to FcRn was slightly lower, between 1.44 x 10 -8 and 1.85 x 10 -8The affinity difference is within the detection error of the instrument. Therefore, the Fc mutant combinations obtained in this study substantially maintained the binding to human FcRn and human FcyRI. Moreover, the affinity values of the Ml, M2 and M3 mutant combinations were very similar to the KIH control.

[0241] Example 9 Accelerated stability experiment of bispecific antibody

[0242] The purified bispecific antibody of Example 6 was buffer exchanged to 20 mM histidine and diluted to 1 mg / ml, and then filtered through a 0.22 um PES filter. The sample was placed in an oven at 42°C, and samples were taken after 7 days and 14 days, and analyzed for purity by SEC-HPLC and SDS-PAGE.

[0243] The SEC-HPLC analysis results are shown in Table 9. After 7 days and 14 days of storage at 42°C, the purity of the heterodimer of each mutant combination only slightly decreased compared to 0 days of storage. Among them, the purity of the KIH, Ml, M2 and M3 mutant combinations decreased by 0.21%, 0.73%, 0.47% and 0.57% respectively after 7 days, and decreased by 1.43%, 1.50%, 1.46% and 0.84% respectively after 14 days. The above results show that the newly designed Fc mutant combinations have good stability under high temperature storage conditions.

[0244] The SDS-PAGE analysis results also show that after 14 days of storage at 42°C, the proportions of protein bands of different molecular weights of the KIH, Ml, M2 and M3 mutant combinations did not change substantially compared to 0 days of storage.

[0245] Table 9 SEC-HPLC analysis of storage stability of bispecific antibody in accelerated experiment

[0246] Discussion

[0247] By analyzing the reasons for the reduction in thermal stability of the knob-hole model, we selected amino acids near the contact interface of the Fc fragment CH3 as the mutation objects to investigate the effects of different mutant combinations on the production performance, physicochemical properties and biological functions of Fc-based IgG-like single / multispecific antibodies.

[0248] Without wishing to be bound by theory, it is believed that in wild-type IgGl Fc, the L351 amino acid in one chain makes direct contact with L351, P352, P353, S354, T366 in the opposite chain; while in KIH structure, L351 in the hole chain makes direct contact with L351, P352, P353, S354, T366W in the knob chain, but the contact distance is increased, and the interaction force is expected to be weakened. Mutating the hole L351, such as to tyrosine (L351Y) with a larger side chain, is speculated to make the hole L351 residue more closely contact with the knob chain amino acids P352, P353, S354, T366W, thereby generating stronger hydrophobic interactions (see FIG. 1C), and creating new contacts with other knob chain amino acids, such as the hydroxyl group (OH) in L351Y forming a new hydrogen bond network with T366W and E357.

[0249] Based on the mutation of the hole chain L351, further introduce the combination of knob chain K409D and hole chain D399R mutations. Without wishing to be bound by theory, it is speculated that this combination of mutations can lead to a similar pair of salt bridges as in IgGl Fc; moreover, the hole chain D399R will make new contacts with the knob chain K392, T411, T366W amino acids, while the knob chain K409D mutation will retain direct contact with the hole chain F405 (see FIG. 1D).

[0250] The experimental results of the above examples show that combining the above mutations in the KIH structure, i.e., hole chain L351Y or hole chain L351Y + knob chain K409D-D399R mutations, can confer good promoting effects on the production and physicochemical properties of heterodimer Fc scaffolds and single / multispecific proteins (especially bispecific antibodies) containing the same. The Fc heterodimer scaffold protein containing the combination of mutations can achieve a ratio of Fc heterodimer of more than 90% after expression in cells and affinity chromatography column purification, higher than the control Fc protein containing only the corresponding knob-into-hole mutations; and can achieve a purity of more than 98% after further purification by ion exchange column. Moreover, compared with the control Fc protein, the purified Fc heterodimer scaffold protein has better thermal stability, with a CH3 domain Tm value of more than 74°C determined by DSC method.

[0251] Further, the experimental results of the above examples also show that combining the above mutations in the KIH structure, i.e., hole chain L351Y or hole chain L351Y + knob chain K409D-D399R mutations, has essentially no effect on the antigen binding function and FcRn and FcyR receptor binding function of heterodimer Fc scaffolds and single / multispecific proteins (especially bispecific antibodies) containing the same.

[0252] Taken together, these results suggest that the introduction of the hole L351Y mutation or the combination of the hole L351Y-D399R with the knob K409D mutation on the KIH background will favor the production and stability of the heterodimeric antibody of interest.

[0253] SEQUENCE LISTING

Claims

1. A heterodimeric Fc scaffold comprising a first polypeptide and a second polypeptide, wherein the first polypeptide comprises a first Fc region and the second polypeptide comprises a second Fc region, wherein the first Fc region comprises a first CH3 domain and the second Fc region comprises a second CH3 domain, wherein wherein the first and second Fc regions comprise a knob mutation and a hole mutation, respectively, of a KIH mutation in their CH3 domains relative to a wild-type CH3 domain, and wherein, (a) the Fc region comprising the hole mutation further comprises a L351Y mutation; or (b) the Fc region comprising the hole mutation further comprises L351Y and D399R mutations, and the Fc region comprising the knob mutation further comprises a K409D mutation, wherein the first and second polypeptides pair and heterodimerize via the first and second Fc regions to form the Fc scaffold, wherein the amino acid residues are numbered according to the EU index in Kabat.

2. The heterodimeric Fc scaffold of claim 1, wherein, The first and second Fc regions comprise CH2 and CH3 domains, or comprise a hinge region, CH2 and CH3 domains.

3. The heterodimeric Fc scaffold of any one of claims 1-2, wherein, The first and second Fc regions are of an IgG type, such as an IgG1, IgG2, IgG3 or IgG4 subtype, preferably the first and second Fc regions are of the human IgG1 subtype.

4. The heterodimeric Fc scaffold of any one of claims 1-3, wherein the knob mutation is T366W; and the hole mutation is Y407V or is T366S-L368A-Y407V, preferably the first Fc region comprises T366W and the second Fc region comprises T366S-L368A-Y407V-L351Y; or the first Fc region comprises T366W-K409D and the second Fc region comprises T366S-L368A-Y407V-L351Y-D399R.

5. The heterodimeric Fc scaffold of any one of claims 1-4, wherein, (a) the first and second CH3 domains further comprise 0-3 amino acid residue alterations relative to a wild-type CH3 domain, preferably the wild-type CH3 domain is a native human IgG1 immunoglobulin CH3 domain, more preferably comprises the amino acid sequence of SEQ ID NO: 19 or 20; (b) the first and second Fc regions comprise a native human IgG1 immunoglobulin CH2 domain, or a CH2 domain comprising the amino acid sequence set forth in SEQ ID NO: 21 or having 1-5 amino acid residue alterations thereto; (c) the Fc region comprising the knob mutation comprises the amino acid sequence of SEQ ID NO: 1 or 2 or an amino acid sequence having at least 95%, 96%, 97%, 98% or 99% identity thereto; (d) the Fc region comprising the hole mutation comprises the amino acid sequence of SEQ ID NO: 7 or 8 or an amino acid sequence having at least 95%, 96%, 97%, 98% or 99% identity thereto, and / or (e) the Fc scaffold does not form a non-native disulfide bond between the first Fc region and the second Fc region.

6. The heterodimeric Fc scaffold of any one of claims 1-5, wherein the Fc scaffold has increased thermal stability compared to a corresponding control Fc scaffold comprising only the KIH mutations, and preferably, the CH3 domains of the Fc scaffold have a Tm value of about 70°C or higher as determined by DSC.

7. [Rule 91 correction 09.02.2026] A CH3 heterodimer comprising a first immunoglobulin CH3 domain and a second immunoglobulin CH3 domain that are heterodimerized, wherein: the first CH3 domain and the second CH3 domain comprise a knob mutation and a hole mutation, respectively, in the KIH mutations, and wherein, (a) the CH3 domain comprising the hole mutation further comprises a L351Y mutation; or (b) the CH3 domain comprising the hole mutation further comprises L351Y and D399R mutations, and the CH3 domain comprising the knob mutation 8. A heteromultimeric protein comprising the heterodimeric Fc scaffold of any one of claims 1-6 or the CH3 heterodimer of claim 7.

9. A binding protein comprising the heterodimeric Fc scaffold of any one of claims 1-6 and at least one target binding domain attached thereto.

10. The binding protein of claim 9, wherein the at least one target binding domain is independently selected from the group consisting of: a ligand, a receptor binding domain of a ligand, a ligand binding domain of a receptor, and an antigen binding domain of an antibody, respectively.

11. The binding protein of any one of claims 9-10, wherein the binding protein is monospecific or bispecific.

12. The binding protein of any one of claims 9-11, wherein the binding protein comprises at least one target binding domain selected from the group consisting of Fv, scFv, Fab, scFab, crossFab, VHH and ligand, independently of each other, preferably 1-6, more preferably 1, 2 or 3 of said target binding domains.

13. The binding protein of any one of claims 9-12, wherein the binding protein comprises only one of said heterodimeric Fc scaffolds.

14. The binding protein of any one of claims 9-13, wherein the binding protein has an IgG configuration or an IgG-like configuration, preferably, the binding protein has a configuration selected from the group consisting of: scFv-Fc / scFv-Fc, Fab-Fc / Fab-Fc, VHH-Fc / VHH-Fc, VHH-Fc / Fab-Fc, VHH-Fc / scFv-Fc, VHH-Fc / Fc, scFv-Fc / Fc, Fab-Fc / Fc, ligand-Fc / ligand-Fc, ligand-Fc / scFv-Fc, ligand-Fc / VHH-Fc, ligand-Fc / Fab-Fc, ligand-Fc / Fc.

15. The binding protein of any one of claims 9-14, wherein the binding protein is bispecific and comprises: (a) two different Fab domains attached N-terminal to the first and second Fc regions, respectively, of the heterodimeric Fc scaffold, wherein the Fab domains have the same or different VLCL light chains; or (a) two different Fab domains attached N-terminal to the first and second Fc regions, respectively, of the heterodimeric Fc scaffold, wherein the Fab domains have the same or different VLCL light chains; or (b) one VHH domain and one Fab domain linked respectively at the N-terminus of the first and second Fc region of said heterodimeric Fc scaffold.

16. The binding protein of any one of the preceding claims 9-15, having one or more properties selected from the group consisting of: - thermal stability, wherein, - a CH3 domain having a Tm value greater than or equal to 74°C as determined by DSC; - a purity of 90% or more; - a binding affinity to the FcRn receptor not less than 80%, preferably not less than 85%, 90%, or 95% of the binding affinity of a native IgGl monoclonal antibody to the FcRn receptor; - a binding affinity to the FcyRl receptor not less than 50% of the binding affinity of a native IgGl monoclonal antibody to the FcyRl receptor; - a storage stability, wherein the purity changes less than 2% as determined by SEC-HPLC after 2 weeks storage at 42°C.

17. A polynucleotide encoding the heterodimeric Fc scaffold of any one of claims 1-6, the CH3 heterodimer of claim 7, the heteromultimeric protein of claim 8, or the binding protein of any one of claims 9-16.

18. A vector or host cell comprising the polynucleotide of claim 17.

19. A composition comprising the heteromultimeric protein of claim 8 or the binding protein of any one of claims 9-16.

20. A conjugate comprising a therapeutic or diagnostic agent conjugated to the heteromultimeric protein of claim 8 or the binding protein of any one of claims 9-16.

21. Use of the composition of claim 19 or the conjugate of claim 20 as a medicament or in the manufacture of a medicament, preferably the medicament is for cancer treatment.

Images