Methods for producing bispecific proteins

Abstract

The present invention provides a method for producing a heterodimeric protein comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface. The present invention also provides an isolated heterodimeric protein obtainable by the method. The present invention also provides an isolated heterodimeric antibody comprising a first IgG CH3 domain and a second IgG CH3 domain, wherein the first CH3 domain and the second CH3 domain are capable of forming a CH3-CH3 interface, as well as a pharmaceutical composition comprising the isolated heterodimeric protein.

Description

[Technical Field]

[0001] Provided herein is a method for producing a heterodimeric protein comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface. Also provided herein is an isolated heterodimeric protein obtained by the method. Also provided herein is an isolated heterodimeric antibody comprising a first IgG CH3 domain and a second IgG CH3 domain, wherein the first CH3 domain and the second CH3 domain are capable of forming a CH3-CH3 interface capable of generating a multispecific binding domain, as well as a pharmaceutical composition comprising the isolated heterodimeric protein having multiple binding specificities. [Background technology]

[0002] Monospecific antibodies play an important role as therapeutic molecules for the treatment of many diseases, particularly cancer. Monospecific antibodies bind to a single specific region or epitope of an antigen and are often selected for their desirable functional properties for therapeutic use (e.g., killing tumor cells, blocking receptor-ligand interactions, neutralizing viruses, etc.). Monospecific antibodies have many beneficial properties, including the ability to be mass-produced, their biophysical and biochemical properties can be extensively analyzed, and batch-to-batch consistency can be ensured, facilitating regulatory acceptance.

[0003] Despite these favorable properties, monospecific antibodies have several drawbacks related to their specificity. For this reason, bispecific and multispecific antibodies have recently begun to play an even more important role, as they may overcome some of the limitations of monospecific antibody therapy. For example, they can be used as mediators to target drugs or toxic compounds to target cells, to retarget effector mechanisms to disease-related sites, or to enhance specificity for tumor cells, for example, by binding to a combination of target molecules found only on tumor cells.

[0004] While desirable, the production and testing of multispecific antibodies still present challenges. For example, bispecific antibodies in the IgG format, consisting of two heavy chains and two light chains, have been produced using a variety of methods. For example, bispecific antibodies can be generated by fusing two antibody-secreting cell lines to create a new cell line or by expressing two antibodies in a single cell using recombinant DNA technology. These approaches result in multiple antibody species because the heavy chains of each antibody can form monospecific dimers (also known as homodimers or homodimeric antibodies) containing two identical heavy chains with the same specificity, or bispecific dimers (also known as heterodimers or heterodimeric antibodies) containing two different heavy chains with different specificities. Furthermore, the light and heavy chains of each antibody can pair randomly, forming inappropriate and non-functional combinations. This problem is known as heavy-light chain mispairing. This issue can be resolved by selecting antibodies with a common light chain for expression as bispecific antibodies. When using a common light chain, the desired bispecific antibody must be purified from the resulting antibody mixture, since expression of two heavy chains and one common light chain in a single cell can potentially result in three different antibody species (i.e., two monospecific "parent" antibodies and the bispecific antibody).

[0005] Although it is possible to produce essentially a single antibody species in a single cell through the use of heterodimerization techniques to pair bispecific antibody constant regions, obtaining this product requires the use of techniques that limit the ability to combine different binding domains (e.g., a common light chain binding domain with a non-common light chain antibody). Thus, there remains a need for a multispecific format that can generate essentially a single product that allows for combinations of different binding domains (common light chains and non-common light chains) that can be reliably manufactured for preclinical testing and clinical and commercial development. Summary of the Invention

[0006] The present disclosure is based on the inventors' development of a novel method for generating heterodimeric proteins, in particular heterodimeric antibodies, which employs a novel Fab-arm exchange method.

[0007] The method described herein is based on the inventors' development of a novel Fab-arm exchange method. In this method, mutations are introduced into the CH3 domain, allowing Fab-arm exchange to occur between all IgG subtypes without destabilizing the core hinge region. As discussed in more detail in the Examples section below, the inventors have shown that Fab-arm exchange in IgG1 molecules can occur when one CH3 domain contains amino acids 351D and 368E and the other CH3 domain contains amino acids 366K and 351K. Although the Examples employ these specific amino acids at positions 351, 366, and 368, this method would also work if different positively charged amino acid residues were present at positions 351 and 366 in one CH3 domain and negatively charged amino acid residues corresponding to positions 351 and 368 were present in the other CH3 domain.

[0008] One advantage of the present disclosure is that it allows for the exchange of a Fab domain with any previously disclosed antibody, given its amino acid sequence, by introducing positively charged amino acids at positions 351 and 366 in its CH3 domain and exchanging the Fab domain with a second antibody whose CH3 domain contains negatively charged amino acids at positions 351 and 368. Currently, there are over 800 known antibodies approved or in development from the International Nonproprietary Names (INN) list maintained by the World Health Organization (WHO), all of which can be combined with another antibody as long as both antibodies contain two different IgG CH3 domains capable of forming a CH3-CH3 interface, one antibody having a CH3 domain with positively charged amino acids at positions 351 and 366 and the other antibody having a CH3 domain with negatively charged amino acids at positions 351 and 368.

[0009] Producing multispecific antibodies containing binding domains that do not share the same light chain and expressing such products from a single cell can result in light chain mispairing, resulting in numerous different antibody species with lost or reduced affinity and specificity, necessitating laborious, time-consuming, and inefficient separation procedures to select the desired antibody species. The present disclosure allows for combining antibodies with a common light chain binding domain with antibodies with non-common light chain binding domains to produce multispecific antibody species with relatively high purity, yield, and efficiency. The present disclosure also now allows for combining antibodies with non-common light chain binding domains with other antibodies with non-common light chain binding domains. The light chains can be any member of both the kappa and lambda families. The Fab domains can be from any source, including from the general light chain repertoire.

[0010] In addition to enabling Fab-arm exchange for all IgG formats, the present disclosure also offers certain unexpected advantages. For example, it has been found that when half antibodies containing the amino acid combinations 366K and 351K, or 351D and 368E, are subjected to reducing and reoxidizing conditions, they remain primarily as half antibodies rather than binding to other half antibodies containing the same residues. As will be apparent to those skilled in the art, this is advantageous when generating bispecific antibodies, as the tendency of half antibodies to form homodimers can hinder efficient production and purification of bispecific antibodies. Furthermore, the presence of homodimers can hinder the screening of large repertoires of multispecific antibodies for functional activity. As shown in more detail in the Examples section of this application, compared to art-known Fab-arm exchange methods that rely on the use of the 405L / 409R mutations, the novel method described herein may be significantly superior in reducing the formation of potentially undesirable homodimers. Specifically, as described in Example 7 of the present disclosure, the 405L / 409R method yielded 5-13% homodimers, whereas the method of the present invention yielded only about 1% homodimers. In the same sense, as shown in Example 7, the method of the present invention can produce a larger amount of heterodimers.

[0011] Thus, in a first aspect, there is provided herein a method for producing a heterodimeric protein comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first protein comprising a first CH3 domain comprising positively charged amino acid residues at positions 351 and 366; and (b) a second protein comprising a second CH3 domain comprising negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering, providing; incubating the proteins of (a) and (b) together under reducing conditions to provide a reduced first protein and a reduced second protein; and Reoxidizing the reduced protein to obtain the heterodimeric protein.

[0012] In certain embodiments, the first protein and / or the second protein is selected from the group comprising or consisting of a monomeric protein, a homodimeric protein, and a heterodimeric protein.

[0013] In certain embodiments, the IgG CH3 domain is an IgG1, IgG2, IgG3, or IgG4 CH3 domain.

[0014] In one embodiment, the IgG CH3 domain is an IgG1 CH3 domain.

[0015] In some embodiments, the IgG CH3 domains of the first protein and the second protein are IgG1. In some embodiments, the IgG CH3 domains of the first protein and the second protein are IgG2. In some embodiments, the IgG CH3 domains of the first protein and the second protein are IgG3. In some embodiments, the IgG CH3 domains of the first protein and the second protein are IgG4.

[0016] In some embodiments, the IgG CH3 domain is a human IgG CH3.

[0017] In some embodiments, the first protein and the second protein comprise the same hinge region. In some embodiments, the first protein and the second protein comprise an IgG1 hinge region. In some embodiments, the first protein and the second protein comprise an IgG2 hinge region. In some embodiments, the first protein and the second protein comprise an IgG3 hinge region. In some embodiments, the first protein and the second protein comprise an IgG4 hinge region.

[0018] In certain embodiments, the first protein and / or the second protein is selected from the group comprising or consisting of an antibody and a half-antibody, or a fragment thereof.

[0019] In certain aspects, the antibody, half antibody, or fragment thereof is a human antibody, half antibody, or fragment thereof.

[0020] In certain embodiments, the fragment is a monomeric Fc region or a dimeric Fc region.

[0021] In certain embodiments, the antibody, half antibody, or fragment thereof is an IgG1, IgG2, IgG3, or IgG4 antibody, half antibody, or fragment thereof. In certain embodiments, the first protein can be an antibody that comprises a first binding specificity and the second protein can be an antibody that comprises a second, different binding specificity.

[0022] In one embodiment, the first protein can be an antibody that comprises a binding specificity and the second protein can be an antibody that comprises another, different binding specificity.

[0023] In some embodiments, the first protein is a multispecific antibody (e.g., a bispecific or trispecific antibody). In some embodiments, the second protein is a multispecific antibody (e.g., a bispecific or trispecific antibody). In some embodiments, both the first and second proteins are multispecific antibodies (e.g., a bispecific or trispecific antibody).

[0024] In certain embodiments, the first protein and / or the second protein is a homodimeric antibody.

[0025] In certain embodiments, the first protein and / or the second protein are heterodimeric antibodies that comprise a common light chain.

[0026] In one embodiment, the resulting heterodimeric protein is a heterodimeric antibody.

[0027] In certain embodiments, the heterodimeric antibody is multivalent, and optionally the multivalent antibody comprises more than one valency, including a bivalent, trivalent, or tetravalent antibody.

[0028] The present disclosure is suitable for generating multivalent multimers known in the art, which multivalent multimers include and do not include the use of a common light chain. WO2019 / 190327 is incorporated by reference, and in particular Figures 1a-1u thereof.

[0029] In certain embodiments, the heterodimeric antibody is a multispecific antibody, and optionally the multispecific antibody comprises more than one valency and is a bispecific antibody, a trispecific antibody, a tetraspecific antibody, or has up to six valencies.

[0030] In certain embodiments, a heterodimeric antibody may comprise two non-identical light chains.

[0031] In certain embodiments, the first CH3 domain can comprise 351K and 366R, 351R and 366K, 351K and 366K, or 351R and 366R.

[0032] In certain embodiments, the second CH3 domain can comprise 351D and 368E, 351E and 368D, 351D and 368D, or 351E and 368E.

[0033] In certain embodiments, the second CH3 domain may comprise 351D and 368E, and the first CH3 domain may comprise 366K and 351K.

[0034] In some embodiments, the second protein is obtained independently of the first protein.

[0035] In some embodiments, the reducing conditions can include: (a) incubating the protein in the presence of any suitable reducing agent known in the art, optionally wherein the reducing agent is selected from the group comprising or consisting of 2-mercaptoethylamine (2-MEA), dithiothreitol (DTT), dithioerythritol (DTE), glutathione, tris(2-carboxyethyl)phosphine (TCEP), L-cysteine, beta-mercaptoethanol, thioglycolate, cysteamine, homocysteine, penicillamine, and sodium borohydride; and / or (b) incubating the protein at pH 6.0 and 12.0, optionally with a pH between 7.0 and 11.0; and / or (c) incubating the protein at a redox potential of −150 to −600 mV, optionally, the redox potential is −250 to −400 mV;

[0036] In certain embodiments, the method may further comprise the step of enriching and / or isolating the heterodimeric protein obtained after reoxidation.

[0037] In one embodiment, the heterodimeric protein obtained after reoxidation is concentrated and / or isolated using a method selected from the group including or consisting of precipitation, centrifugation, filtration, size exclusion chromatography, affinity chromatography, cation and / or anion exchange chromatography, and hydrophobic interaction chromatography.

[0038] The methods described herein provide a method for producing a heterodimeric antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first IgG antibody or half-body, wherein each CH3 domain comprises positively charged amino acid residues at positions 351 and 366; and (b) a second IgG antibody or half-body, wherein each CH3 domain comprises negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering, providing; Incubating the antibodies and / or half-bodies of (a) and (b) under conditions sufficient to cause disulfide bond isomerization of cysteines in the core hinge region of the antibodies or half-bodies to yield heterodimeric proteins.

[0039] In a further aspect, provided herein is an isolated heterodimeric protein obtainable by the methods of the present disclosure.

[0040] In one aspect, the heterodimeric protein obtained by the methods of the present disclosure is an IgG antibody.

[0041] In certain aspects, the heterodimeric protein obtained by the methods of the present disclosure is a multispecific IgG antibody.

[0042] In certain embodiments, an IgG antibody comprises two light chains with non-identical sequences. In certain embodiments, an IgG antibody comprises a binding domain comprising non-identical light chain sequences.

[0043] 1. An isolated heterodimeric antibody comprising a first IgG CH3 domain and a second IgG CH3 domain, wherein the first CH3 domain and the second CH3 domain are capable of forming a CH3-CH3 interface, and wherein the first CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R, and the second CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D, and wherein said heterodimeric antibody further comprises two light chains having non-identical sequences.

[0044] In one embodiment, the second CH3 domain comprises 351D and 368E, and the first CH3 domain comprises 366K and 351K.

[0045] In a further aspect, provided herein is a pharmaceutical composition comprising an isolated heterodimeric protein of the present disclosure and a pharmaceutically acceptable carrier.

[0046] In some aspects, the heterodimeric protein is obtained by the methods of the present disclosure.

[0047] Throughout this specification and the claims, the terms "comprise" and "contain" and variations thereof mean "including, but not limited to," and are not intended to (and do not) exclude other moieties, additives, ingredients, integers, or steps.

[0048] Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification should be understood as contemplating the plural as well as the singular, unless the context otherwise requires.

[0049] It should be understood that any feature, integer, property, compound, chemical moiety, or group described in connection with one aspect or embodiment of the present disclosure is applicable to any other aspect or embodiment described herein, unless inconsistent therewith.

[0050] Various aspects of the disclosure are described in further detail below. [Brief explanation of the drawings]

[0051] Aspects of the present disclosure will now be further described with reference to the accompanying drawings. [Figure 1] Top: SDS / PAGE labChip results for reactions #1-#12 from Table 2 without FAE. Bottom: Reactions #1-#12 from Table 2 after FAE. [Figure 2a] HP-CIEX results for reactions #1-#12 in Table 2 without FAE and after FAE. Figure 2a-2f: The top two graphs show the reaction results without FAE, and the bottom two graphs show the reaction results with FAE. Figure 2a: The left two graphs show the results for #1, and the right two graphs show the results for #2. Figure 2b: The left and right two graphs show the results for #3 and #4, respectively. Figure 2c: The left and right two graphs show the results for #5 and #6, respectively. Figure 2d: The left and right two graphs show the results for #7 and #8, respectively. Figure 2e: The left and right two graphs show the results for #9 and #10, respectively. Figure 2f: The left and right two graphs show the results for #11 and #12, respectively. [Figure 2b] Same as above [Figure 2c] Same as above [Figure 2d] Same as above [Figure 2e] Same as above [Figure 2f] Same as above [Figure 3a]HP-SEC results for reactions #1-#12 in Table 2 without FAE and after FAE. Figures 3a-3d: The top three graphs show the reaction results without FAE, and the bottom three graphs show the reaction results with FAE. Figure 3a: From left to right: Results for #1, #5, and #6. Figure 3b: From left to right: Results for #3, #9, and #10. Figure 3c: From left to right: Results for #2, #7, and #8. Figure 3d: From left to right: Results for #4, #11, and #12. [Figure 3b] Same as above [Figure 3c] Same as above [Figure 3d] Same as above [Figure 4-1] Gel filtration purification and LabChip analysis results for reactions #1–8 in Table 3 after FAE. [Figure 4-2] Same as above [Figure 4-3] Same as above [Figure 4-4] Same as above [Figure 5] LabChip analysis of gel filtration purified FAE products from reactions #1–8 in Table 3 under non-reducing (top panel) and reducing (bottom panel) conditions. [Figure 6a] CIEX results for samples obtained before FAE and after FAE and gel filtration. Figure 6a: Results for reactions #1 and #2 in Table 3 show the formation of IgG heterodimers before (top graph) and after (bottom graph) FAE. Figure 6b: Results for reactions #3 and #4 in Table 3 show the formation of IgG heterodimers before (top graph) and after (bottom graph) FAE. Figure 6c: Results for reactions #5 and #6 in Table 3 show the formation of IgG heterodimers before (top graph) and after (bottom graph) FAE. Figure 6d: Results for reactions #7 and #8 in Table 3 show the formation of IgG heterodimers before (top graph) and after (bottom graph) FAE. [Figure 6b] Same as above [Figure 6c] Same as above [Figure 6d] Same as above DETAILED DESCRIPTION OF THE INVENTION

[0052] Human immunoglobulin G (IgG) antibodies exist in four subclasses with distinct structural and functional properties. IgG consists of two pairs of heavy and light chains (half molecules) that are bound together via inter-heavy chain disulfide bonds located in the hinge region.

[0053] Provided herein are methods for producing heterodimeric proteins comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface. In some embodiments, the methods are in vitro methods.

[0054] As used herein, the term "heterodimeric protein" refers to a protein comprising two monomers with non-identical polypeptides linked covalently or non-covalently. One or each of the monomers may be paired with a light chain. When used as the first and second proteins provided as methods (a) and (b) of the present disclosure, the heterodimeric protein comprises a CH3 domain comprising positively charged amino acid residues at positions 351 and 366 or negatively charged amino acid residues at positions 351 and 368. The heterodimeric protein obtained by the disclosed method comprises positively charged amino acid residues at positions 351 and 366 and negatively charged amino acid residues at positions 351 and 368.

[0055] Heterodimeric protein products As will be apparent to those skilled in the art, the characteristics of the heterodimeric protein produced by the disclosed method are determined by the starting materials, i.e., the first and second proteins provided as steps (a) and (b) of the disclosed method. For example only, if the first and second proteins are fragments of an IgG antibody (e.g., comprising or consisting of an IgG CH3 domain), the heterodimeric protein obtained by the disclosed method will be a heterodimeric fragment of an IgG antibody (e.g., the heterodimeric fragment comprises or consists of two IgG CH3 domains). Similarly, if the first and second proteins are IgG antibodies or IgG half-bodies, the heterodimeric protein produced will be an IgG antibody. In some embodiments, the first and / or second proteins comprise a CH3 domain, a hinge region, and a Fab arm. In principle, the present technology allows the production of heterodimers in which the first and / or second proteins may comprise any multimerization domain, including, but not limited to, a variable heavy domain, a CH1 domain, a CH2 domain, a variable light chain, etc. In certain embodiments, the first protein may comprise an antibody binding domain and the second protein may comprise a cytokine, ligand, scFv, or other domain that provides therapeutic potential (e.g., a bifunctional or multifunctional fusion protein).

[0056] Examples of the first and second proteins are described below.

[0057] With respect to heterodimeric proteins produced by the methods of the present disclosure, the term "heterodimeric protein" refers to a protein comprising two monomers having non-identical polypeptides linked either covalently or non-covalently, wherein the two monomers are two differentThe IgG CH3 domains comprise or consist of an IgG CH3 domain. IgG CH3 domains can be distinguished by differences in their polypeptide sequences. Specifically, two different IgG CH3 domains differ in amino acids at least at positions 351, 366, and / or 368. In certain embodiments, one of the two CH3 domains comprises positively charged amino acid residues at positions 351 and 366, and the second of the two CH3 domains comprises negatively charged amino acid residues at positions 351 and 368. Amino acid residue numbers are according to EU numbering (available at https: / / www.imgt.org / IMGTScientificChart / Numbering / Hu_IGHGnber.html, last updated 2020-01-20 21:00:03 CET). A CH3 domain comprising positively charged amino acid residues at positions 351 and 366 is referred to herein as a "351 / 366 positive CH3 domain" or "first CH3 domain." In the same sense, a CH3 domain containing negatively charged amino acid residues at positions 351 and 368 is referred to herein as a "351 / 368 negative CH3 domain" or "second CH3 domain."

[0058] In this disclosure, when referring to a 351 / 366 positive CH3 domain or a 351 / 368 negative CH3 domain, this refers specifically to the side chain charges of residues 351, 366, and 368, and not necessarily all charges throughout the CH3 domain. It will be understood that the heterodimeric proteins produced by the methods of the present disclosure are heterodimers because they have at least two different CH3 domains.

[0059] As used herein, the term "CH3 domain" refers to the CH3 domain of immunoglobulins, particularly IgG immunoglobulins. CH3 domains and their sequences are well known in the art. "IgG immunoglobulin" (also referred to herein as "IgG," "IgG molecule," or "IgG antibody") refers to a polypeptide belonging to the class of antibodies substantially encoded by art-recognized immunoglobulin gamma genes. In humans, the IgG immunoglobulin class includes the subclasses IgG1, IgG2, IgG3, and IgG4. Conventional IgG immunoglobulins are typically heterotetramers with two heavy chains and two light chains linked together by disulfide bonds (-SS-) at the hinge region. However, IgG immunoglobulins are often referred to in the art as dimers (e.g., homodimers or heterodimers). A dimer is formed by two monomers, each of which contains a heavy chain and a light chain. The heavy and light chains are linked together by disulfide bonds (-SS-). Such monomers are referred to as "half antibodies" or "half bodies."

[0060] The methods of the present disclosure generate heterodimeric proteins as a result of the ability of different (i.e., 351 / 366 positive and 351 / 368 negative) IgG CH3 domains to form a CH3-CH3 interface. In certain embodiments, two different IgG CH3 domains preferentially bind, i.e., have a greater tendency to bind to each other than to another identically charged IgG CH3 domain. In other words, a 351 / 368 negative IgG CH3 domain described herein may have a greater tendency to bind to a 351 / 366 positive IgG domain described herein than to another 351 / 368 negative IgG CH3 domain. Similarly, a 351 / 366 positive IgG CH3 domain described herein may have a greater tendency to bind to a 351 / 368 negative IgG CH3 domain described herein than to another 351 / 366 positive IgG CH3 domain.

[0061] As used herein, the term "CH3-CH3 interface" refers to the bond between two different CH3 domains resulting from interacting amino acid residues, i.e., at least one interaction between an amino acid in a first CH3 domain and an amino acid in a second CH3 domain. Such interactions may occur, for example, through van der Waals forces, hydrogen bonds, water-mediated hydrogen bonds, salt bridges or other electrostatic forces, attractive interactions between aromatic side chains, disulfide bond formation, or other forces known to those skilled in the art. When two different CH3 domains form a CH3-CH3 interface, they will be understood to form a heterodimeric protein (a heterodimer due to the different sequences of at least the CH3 domains).

[0062] The interaction between two CH3 domains (such as the two CH3 domains of two individual heavy chains) is known to play an important role in driving heavy chain dimerization. Thus, the CH3 domain directs the binding of antibody heavy chains, and the interface between the CH3 domains is known to contain more than 20 contact residues from each chain that play a role in CH3-CH3 interactions (Deisenhofer J., Biochemistry 1981(20)2361-2370; Miller S., J. Mol. Biol. 1990(216)965-973; Padlan, Advances in Protein Chemistry 1996(49)57-133). The CH3 variants of the present disclosure (positively charged amino acid residues at positions 351 and 366 and negatively charged amino acid residues at positions 351 and 368, specific examples of which are discussed in more detail elsewhere in this disclosure) can therefore be used in conjunction with other antibody domains to generate full-length antibodies that are either bispecific or monospecific. Antibody specificity, defined by the VH / VL combination, does not normally affect the dimerization behavior of the heavy chain, which is driven by the CH3 domain.

[0063] In certain embodiments, the IgG CH3 domain of the heterodimeric protein produced by the methods of the present disclosure does not comprise an arginine at position 409 and / or does not comprise a leucine at position 405. In some examples, the IgG CH3 domain of the heterodimeric protein produced by the methods of the present disclosure comprises a lysine at position 409 and / or a phenylalanine at position 405 (EU numbering).

[0064] The heterodimeric protein produced by the methods described herein is a heterodimeric IgG antibody or a heterodimeric fragment of an IgG antibody. In this regard, a heterodimeric fragment of an IgG antibody refers to a molecule containing at least two different IgG CH3 domains. In certain embodiments, the fragment may further contain one or more domains (such as CH2, CH1, VH, CL, and / or VL) and / or specific binding portions typically present in an IgG antibody.

[0065] In certain embodiments, a heterodimeric IgG antibody or heterodimeric fragment of an IgG antibody may comprise two IgG CH3 domains selected from the group comprising or consisting of IgG1, IgG2, IgG3, and IgG4 CH3 domains. In certain embodiments, each IgG CH3 domain in a heterodimeric IgG antibody or heterodimeric fragment thereof is an IgG1 CH3 domain. It will be understood that each of the monomers forming a heterodimeric IgG antibody or heterodimeric fragment of an IgG antibody need not comprise an IgG CH3 domain of the same subclass. In certain embodiments, each of the CH3 domains of a heterodimeric IgG antibody or heterodimeric fragment of an IgG antibody is of a different subclass. By way of example only, one of the monomers may comprise an IgG1 CH3 domain, and the other may comprise an IgG2, IgG3, or IgG4 CH3 domain.

[0066] In certain embodiments, the IgG CH3 domain in the heterodimeric protein produced by the methods described herein is a human IgG CH3 domain (e.g., a human IgG1, IgG2, IgG3, or IgG4 CH3 domain). In certain embodiments, the IgG CH3 domain in the heterodimeric IgG antibody or heterodimeric fragment of an IgG antibody produced by the methods described herein is a human IgG CH3 domain (e.g., a human IgG1, IgG2, IgG3, or IgG4 CH3 domain).

[0067] In a specific example, the heterodimeric protein produced by the methods described herein is a heterodimeric IgG antibody. In this specific example, the antibody is an IgG1, IgG2, IgG3, or IgG4 antibody. In some embodiments, it is a human IgG1. In some embodiments, it is a human IgG2. In some embodiments, it is a human IgG3. In some embodiments, it is a human IgG4 antibody.

[0068] In another specific example, the heterodimeric protein produced by the methods described herein is a heterodimeric fragment of an IgG antibody. In this specific example, the heterodimeric fragment is a heterodimeric fragment of an IgG1, IgG2, IgG3, or IgG4 antibody. In certain embodiments, it is a heterodimeric fragment of a human IgG1, IgG2, IgG3, or IgG4 antibody.

[0069] In certain embodiments, when the heterodimeric protein produced by the methods described herein is a heterodimeric fragment of an IgG antibody, the fragment may comprise two IgG CH3 domains and two IgG CH2 domains (e.g., each monomer in the heterodimeric protein may comprise an IgG CH3 domain and an IgG CH2 domain).

[0070] In a specific example, the heterodimeric protein produced by the methods described herein comprises or consists of an IgG Fc region. In this specific example, the IgG Fc region is an IgG Fc region of an IgG1, IgG2, IgG3, or IgG4 antibody. In certain embodiments, it is an IgG Fc region of a human IgG1, IgG2, IgG3, or IgG4 antibody.

[0071] As used herein, the term "IgG Fc region" refers to a fragment of the crystallizable C-terminal region of an immunoglobulin heavy chain. The human IgG heavy chain Fc region is generally defined as comprising amino acid residues from P230 to the carboxyl terminus of an IgG antibody. The numbering of residues in the Fc region is according to the EU index. In certain embodiments, the Fc region may include a hinge region. The hinge region (e.g., residues 216-230 in the EU numbering system for IgG1) may extend from the N-terminus of the Fc region. Two monomeric IgG Fc domains are linked together by disulfide bonds (-SS-) in the hinge region, which aids in the formation and / or maintenance of heterodimeric proteins. The number of disulfide bonds in the hinge varies depending on the immunoglobulin subclass (Papadea and Check 1989). In vivo, Fc region dimers or heterodimers interact with the complement system and specific receptors on various cell surfaces. In certain embodiments, an IgG Fc region can comprise an IgG1 or IgG2 core hinge region, CPPC. Alternatively, an IgG Fc region can comprise an IgG3 core hinge region, CPRC. Alternatively, an IgG Fc region can comprise an IgG4 core hinge region, CPSC. As used herein, the term "core hinge region" refers to the four amino acids corresponding to positions 226-229 (EU numbering) of a human IgG1 antibody.

[0072] In some embodiments, the first protein and the second protein comprise the same hinge region. In some embodiments, the first protein and the second protein comprise an IgG1 hinge region. In some embodiments, the first protein and the second protein comprise an IgG2 hinge region. In some embodiments, the first protein and the second protein comprise an IgG3 hinge region. In some embodiments, the first protein and the second protein comprise an IgG4 hinge region.

[0073] In some embodiments, the heterodimeric protein produced by the methods described herein is a heterodimeric antibody. The antibodies produced by the methods described herein can have sequences of any origin, including mouse and human sequences. The antibodies may consist of sequences of only one origin, such as fully human antibodies, or may have sequences of two or more origins, such as chimeric antibodies and humanized antibodies. It is desirable for therapeutic antibodies to resemble as closely as possible the natural antibodies of the subject to be treated (e.g., human antibodies for human subjects).

[0074] Antibody binding can be expressed in terms of specificity and affinity. Specificity determines which antigen or epitope is bound by the binding domain. Affinity is a measure of the strength of binding to a particular antigen or epitope.

[0075] As used herein, the term "antigen" refers to a substance or molecule that, when introduced into the body, elicits antibody production by the immune system. Antigens can be derived from, among other things, pathogenic organisms, tumor cells or other abnormal cells, haptens, or even self-structures. At the molecular level, antigens are characterized by their ability to bind to the antigen-binding site of an antibody. Mixtures of antigens may also be considered "antigens"; that is, while tumor cell lysates or virus particles may be referred to as "antigens," those skilled in the art will understand that such tumor cell lysates or virus particle preparations contain many antigenic determinants. An antigen contains at least one, and often more, epitopes. As used herein, the term "epitope" refers to the portion of an antigen that is recognized by the immune system, particularly antibodies, B cells, or T cells. While epitopes are typically thought of as derived from non-self proteins, host-derived sequences that can be recognized are also classified as epitopes.

[0076] It will be understood that the heterodimeric proteins (e.g., heterodimeric IgG antibodies or heterodimeric fragments thereof) produced by the methods described herein are composed of two monomers, each of which comprises or consists of a different IgG CH3 domain (i.e., a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative IgG CH3 domain). Additionally, there may be additional differences (i.e., sequence differences) in one or both of the monomers (e.g., half-bodies or fragments thereof).

[0077] In certain embodiments, these differences can result in a multivalent and / or multispecific heterodimeric protein (such as an IgG antibody or heterodimeric fragment thereof). The term "multivalent" (e.g., a multivalent antibody or heterodimeric fragment thereof) refers to a molecule having multiple valencies, where "valency" is described as the number of antigen-binding moieties present per molecule (e.g., an antibody or heterodimeric fragment thereof). In this manner, a single binding molecule can bind to multiple binding sites on a target antigen. Examples of multivalent antibodies include, but are not limited to, bivalent antibodies, trivalent antibodies, tetravalent antibodies, pentavalent antibodies, etc., including antibodies with at least hexavalent antibodies.

[0078] As used herein, the term "multispecific" (e.g., a multispecific antibody or heterodimeric fragment thereof) refers to a single molecule that binds to two or more different epitopes on at least two or more different antigens. The term "multispecific antibody" includes, but is not limited to, bispecific antibodies, trispecific antibodies, tetraspecific antibodies, and the like. In some embodiments, the term "multispecific antibody" refers to a bispecific antibody. In some embodiments, the term "multispecific antibody" refers to a trispecific antibody. In some embodiments, the term "multispecific antibody" refers to an antibody with a valency of tetravalent, pentavalent, or hexavalent. In some embodiments, the term "multispecific antibody" refers to an antibody with a valency greater than 6.

[0079] In certain embodiments, the heterodimeric protein produced by the methods described herein is a heterodimeric IgG antibody, wherein the antibody is multivalent and / or multispecific.

[0080] In certain embodiments, a heterodimeric protein (such as an IgG antibody or heterodimeric fragment thereof) can comprise two or more variable regions. In certain embodiments, each of the variable regions can specifically bind to a different epitope. In certain embodiments, the different epitopes are located on different antigens. In certain embodiments, the different antigens are expressed on the same cell or on different cells.

[0081] In certain embodiments, a heterodimeric antibody may comprise two non-identical light chains. Alternatively, a heterodimeric antibody may comprise two identical light chains.

[0082] In embodiments in which the heterodimeric protein produced by the methods of the present disclosure is an IgG antibody, the antibody may comprise two monomers of the same IgG subclass, for example, two IgG1 monomers, two IgG2 monomers, two IgG3 monomers, or two IgG4 monomers.

[0083] In certain aspects, heterodimeric proteins (such as IgG antibodies or heterodimeric IgG fragments thereof) produced according to the present disclosure may comprise two identical light chains. Alternatively, heterodimeric proteins (such as IgG antibodies or heterodimeric IgG fragments thereof) produced according to the present disclosure may comprise two non-identical light chains.

[0084] In one aspect, provided herein is a method for generating a DEKK heterodimeric protein comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first protein comprising a first CH3 domain comprising positively charged amino acid residues at positions 351 and 366, wherein the positively charged amino acid residues at positions 351 and 366 are K and K, respectively; (b) a second protein comprising a second CH3 domain comprising negatively charged amino acid residues at positions 351 and 368, wherein the negatively charged amino acid residues at positions 351 and 368 are D and E, respectively; where the numbering follows EU numbering, providing; incubating the proteins of (a) and (b) together under reducing conditions to provide a reduced first protein and a reduced second protein; and Reoxidizing the reduced first and second proteins to obtain the DEKK heterodimer protein.

[0085] In one aspect, provided herein is a method for producing a heterodimeric antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first antibody or half-body, wherein each CH3 domain comprises positively charged amino acid residues at positions 351 and 366; and (b) a second antibody or half-body, wherein each CH3 domain comprises negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering, providing; incubating the antibodies and / or half-bodies of (a) and (b) together under reducing conditions to provide a reduced first antibody or half-body and a reduced second antibody or half-body; and Re-oxidizing the reduced first antibody and / or second half-body to obtain the heterodimeric antibody.

[0086] In one aspect, provided herein is a method for producing an IgG heterodimeric antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first IgG antibody or IgG half-body, wherein each CH3 domain comprises positively charged amino acid residues at positions 351 and 366; and (b) a second IgG antibody or IgG half-body, wherein each CH3 domain comprises negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering, providing; incubating the IgG antibodies and / or IgG half-bodies of (a) and (b) together under reducing conditions to provide a reduced first IgG antibody or IgG half-body and a reduced second IgG antibody or IgG half-body; and Re-oxidizing the reduced IgG antibody and / or IgG half-bodies to obtain the heterodimeric IgG antibody.

[0087] In one aspect, provided herein is a method for producing a DEKK heterodimeric antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first antibody or half-body, wherein each CH3 domain comprises positively charged amino acid residues at positions 351 and 366, and the positively charged amino acid residues at positions 351 and 366 are K and K, respectively; and (b) a second antibody or half-body, wherein each CH3 domain comprises negatively charged amino acid residues at positions 351 and 368, and the negatively charged amino acid residues at positions 351 and 368 are D and E, respectively; where the numbering follows EU numbering, providing; incubating the antibodies and / or half-bodies of (a) and (b) together under reducing conditions to provide a reduced first antibody or half-body and a reduced second antibody or half-body; and Re-oxidizing the reduced antibody and / or half-bodies to obtain the DEKK heterodimeric antibody.

[0088] In one aspect, provided herein is a method for producing a DEKK heterodimeric IgG antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, said method comprising the steps of: (a) a first IgG antibody or IgG half-body, wherein each CH3 domain comprises positively charged amino acid residues at positions 351 and 366, and the positively charged amino acid residues at positions 351 and 366 are K and K, respectively; and (b) a second IgG antibody or IgG half-body, wherein each CH3 domain comprises negatively charged amino acid residues at positions 351 and 368, and the negatively charged amino acid residues at positions 351 and 368 are D and E, respectively; where the numbering follows EU numbering, providing; incubating the IgG antibodies and / or IgG half-bodies of (a) and (b) together under reducing conditions to provide a reduced first IgG antibody or IgG half-body and a reduced second IgG antibody or IgG half-body; and Re-oxidizing the reduced IgG antibody and / or IgG half-bodies to obtain the DEKK heterodimeric IgG antibody.

[0089] Method for producing heterodimeric proteins The disclosed methods include providing (a) a first protein comprising a CH3 domain comprising positively charged amino acid residues at positions 351 and 366 (also referred to herein as a 351 / 366 positive CH3 domain), and (b) a second protein comprising a CH3 domain comprising negatively charged amino acid residues at positions 351 and 368 (also referred to herein as a 351 / 368 negative CH3 domain), where numbering is according to EU numbering.

[0090] In certain embodiments, the first and / or second protein may be selected from the group comprising or consisting of a monomeric protein, a homodimeric protein, and a heterodimeric protein.

[0091] Those skilled in the art will understand that the terms "monomeric protein" and "monomer," used interchangeably herein, generally refer to a single, non-aggregated protein or polypeptide molecule. However, as noted above, in the field of antibodies, the term "monomer" can also refer to a half-body. For example, an IgG monomer is an IgG half-body, i.e., a molecule comprising or consisting of one IgG heavy chain linked to one IgG light chain. In an IgG half-body, one light chain and one heavy chain are linked by a disulfide bond. Thus, for the purposes of this disclosure, the term "monomer" can refer to a single, non-aggregated protein or polypeptide molecule (e.g., a single IgG CH3 domain or a monomeric IgG Fc region) or a half-body (e.g., an IgG half-body), depending on the context.

[0092] Similarly, in the art, the term "homodimeric protein" or "homodimer" generally refers to a dimer formed from two identical polypeptides (e.g., two 351 / 368 negative CH3 domains or two 351 / 366 positive CH3 domains) linked covalently or non-covalently. However, in the antibody field, the term homodimer can also refer to an antibody (such as an IgG antibody) having two identical half-bodies. Therefore, for the present disclosure, the term homodimer refers to a dimer formed from two identical polypeptides (e.g., two identical IgG CH3 domains) linked either covalently or non-covalently, or a homodimeric antibody (e.g., a homodimeric IgG antibody), depending on the context.

[0093] Thus, when the first protein is a homodimer, the disclosed method includes (a) providing a first protein (e.g., a homodimeric IgG antibody or fragment thereof) comprising two (identical) CH3 domains, each comprising positively charged amino acid residues at positions 351 and 366 (also referred to herein as a 351 / 366 positive CH3 domain). Similarly, when the second protein is a homodimer, the disclosed method includes (b) providing a second protein (e.g., a homodimeric IgG antibody or fragment thereof) comprising two (identical) CH3 domains, each comprising negatively charged amino acid residues at positions 351 and 368 (also referred to herein as a 351 / 368 negative CH3 domain).

[0094] As described elsewhere herein, the term "heterodimeric protein" or "heterodimer" refers to a protein comprising two monomers having covalently or non-covalently linked non-identical polypeptides. With respect to the first and second proteins described herein, the term refers to a protein comprising two monomers containing IgG CH3 domains that are different (i.e., have non-identical polypeptide sequences) but have the same charge (i.e., the CH3 domains in the first protein heterodimer are both 351 / 366 positive IgG CH3 domains, and the CH3 domains in the second protein heterodimer are both 351 / 368 negative IgG CH3 domains). Thus, when the first protein is a heterodimer, the disclosed method includes (a) providing a first protein (e.g., a heterodimeric IgG antibody or fragment thereof) comprising two CH3 domains, each containing positively charged amino acid residues at positions 351 and 366 (also referred to herein as a 351 / 366 positive CH3 domain). Similarly, when the second protein is a heterodimer, the disclosed method includes (b) providing a second protein (e.g., a heterodimeric IgG antibody or fragment thereof) that includes two CH3 domains, each of which includes a negatively charged amino acid residue at positions 351 and 368 (also referred to herein as a 351 / 368 negative CH3 domain).

[0095] In one embodiment, the two IgG CH3 domains of the two monomers forming the heterodimer of the first protein have the same amino acids at positions 351 and 366, and similarly, the two IgG CH3 domains of the two monomers forming the heterodimer of the second protein have the same amino acids at positions 351 and 368. In one embodiment, the IgG CH3 domains of the two monomers forming the heterodimer are identical (i.e., the polypeptide sequences of the IgG CH3 domains are identical over their entire length).

[0096] In the field of antibodies, the term heterodimer can also refer to an antibody (such as an IgG antibody) that comprises or consists of non-identical half-bodies. Therefore, in the context of the present disclosure, the term heterodimeric protein can also refer to a heterodimeric antibody. Such heterodimeric antibodies are made from two non-identical half-bodies, and the two half-bodies contain IgG CH3 domains with the same charge (i.e., 351 / 366 positive or 351 / 368 negative IgG CH3 domains). In some embodiments, the IgG CH3 domains of the two half-bodies forming a heterodimeric antibody have the same amino acids at positions 351 and 366 or 351 and 368 (depending on whether the IgG CH3 domain is 351 / 366 positive or 351 / 368 negative). In some embodiments, the IgG CH3 domains of the two half-bodies forming a heterodimeric antibody are identical (i.e., the polypeptide sequences of the IgG CH3 domains are identical throughout their entire length).

[0097] In some embodiments, the first and second proteins provided in (a) and (b) of the disclosed method are monomers (e.g., the first and second proteins are first and second half-bodies, respectively). In some embodiments, the first protein (e.g., the first half-body) can comprise a 351 / 366 positive IgG CH3 domain, and the second protein (e.g., the second half-body) can comprise a 351 / 368 negative IgG CH3 domain. Upon completion of the reduction and reoxidation steps of the disclosed method, the resulting heterodimeric protein (e.g., a heterodimeric antibody) comprises the first and second proteins (e.g., a front half-body and a back half-body).

[0098] In another example, the first protein provided in (a) of the disclosed method is a monomer (e.g., the first protein is a half-body), and the second protein provided in (b) is a homodimeric protein (e.g., the second protein is a homodimeric antibody). The monomeric protein (e.g., half-body) may contain a 351 / 366-positive IgG CH3 domain, and the homodimeric protein (e.g., homodimeric antibody) may contain two 351 / 368-negative IgG CH3 domains. Upon completion of steps ii) and iii) of the disclosed method, the generated heterodimeric protein (e.g., heterodimeric antibody) will contain the monomeric protein (which is the half-body) and one monomer of the homodimeric protein (e.g., one half-body of the homodimeric antibody). Thus, the generated heterodimeric protein (e.g., heterodimeric antibody) will contain a 351 / 366-positive IgG CH3 domain and a 351 / 368-negative IgG CH3 domain.

[0099] In another example, the first protein provided in (a) of the disclosed method is a homodimeric protein (e.g., the first protein is a homodimeric antibody), and the second protein provided in (b) is a monomer (e.g., the second protein is a half-body). The homodimeric protein (e.g., a homodimeric antibody) may contain two 351 / 366-positive IgG CH3 domains, and the monomeric protein (e.g., a half-body) may contain a 351 / 368-negative IgG CH3 domain. Upon completion of steps ii) and iii) of the disclosed method, the generated heterodimeric protein (e.g., a heterodimeric antibody) will contain one monomer of the homodimeric protein (e.g., one half-body of the homodimeric antibody) and the monomeric protein (the half-body). Thus, the generated heterodimeric protein (e.g., a heterodimeric antibody) contains a 351 / 366-positive IgG CH3 domain and a 351 / 368-negative IgG CH3 domain.

[0100] In another example, the first protein provided in (a) of the disclosed method can be a homodimeric protein (e.g., a first homodimeric antibody), and the second protein provided in (b) can also be a homodimeric protein (e.g., a second homodimeric antibody). The first homodimeric protein (e.g., the first homodimeric antibody) can contain two 351 / 366 positive IgG CH3 domains, and the second protein (e.g., the second homodimeric antibody) can contain two 351 / 368 negative IgG CH3 domains. Upon completion of the reduction and reoxidation steps of the disclosed method, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains one monomer of the first homodimeric protein (e.g., one half-body of the first homodimeric antibody) and one monomer of the second homodimeric protein (e.g., one half-body of the second homodimeric antibody). The resulting heterodimeric protein (eg, heterodimeric antibody) therefore contains a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative IgG CH3 domain.

[0101] In another example, the first protein provided in (a) of the method of the present disclosure is a monomer (e.g., the first protein is a half-body), and the second protein provided in (b) is a heterodimeric protein (e.g., the second protein is a heterodimeric antibody). The monomeric protein (e.g., half-body) may contain a 351 / 366 positive IgG CH3 domain, and the heterodimeric protein (e.g., heterodimeric antibody) may contain a 351 / 368 negative IgG CH3 domain. Upon completion of the reduction and reoxidation steps of the method of the present disclosure, the resulting heterodimeric protein (e.g., heterodimeric antibody) contains the monomeric protein (half-body) and one monomer of the heterodimeric protein (e.g., one half-body of a homodimeric antibody). Thus, the resulting heterodimeric protein (e.g., heterodimeric antibody) contains a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative IgG CH3 domain.

[0102] In another example, the first protein provided in (a) of the disclosed method is a heterodimeric protein (e.g., the first protein is a heterodimeric antibody), and the second protein provided in (b) is a monomer (e.g., the second protein is a half-body protein). The heterodimeric protein (e.g., a heterodimeric antibody) may contain two 351 / 366-positive IgG CH3 domains, and the monomeric protein (e.g., a half-body) may contain one 351 / 368-negative IgG CH3 domain. Upon completion of the reduction and reoxidation steps of the disclosed method, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains one monomer of the heterodimeric protein (e.g., one half-body of the heterodimeric antibody) and the monomeric protein (half-body). Thus, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains a 351 / 366-positive IgG CH3 domain and a 351 / 368-negative IgG CH3 domain.

[0103] In another example, the first protein provided in (a) of the method of the present disclosure can be a heterodimeric protein (e.g., a first heterodimeric antibody), and the second protein provided in (b) can also be a homodimeric protein (e.g., a second homodimeric antibody). The first heterodimeric protein (e.g., a first heterodimeric antibody) can contain two 351 / 366 positive IgG CH3 domains, and the second protein (e.g., a second homodimeric antibody) can contain two 351 / 368 negative IgG CH3 domains. Upon completion of the reduction and reoxidation steps of the method of the present disclosure, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains one monomer of the first heterodimeric protein (e.g., one half-body of the first heterodimeric antibody) and one monomer of the second homodimeric protein (e.g., one half-body of the second homodimeric antibody). The heterodimeric protein (eg, heterodimeric antibody) thus produced contains a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative CH3 domain.

[0104] In another example, the first protein provided in (a) of the method of the present disclosure may be a homodimeric protein (e.g., a first homodimeric antibody), and the second protein provided in (b) may also be a heterodimeric protein (e.g., a second heterodimeric antibody). The first homodimeric protein (e.g., a first homodimeric antibody) may contain two 351 / 366 positive IgG CH3 domains, and the second protein (e.g., a second heterodimeric antibody) may contain two 351 / 368 negative IgG CH3 domains. Upon completion of the reduction and reoxidation steps of the method of the present disclosure, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains one monomer of the first homodimeric protein (e.g., one half-body of the first homodimeric antibody) and one monomer of the second heterodimeric protein (e.g., one half-body of the second heterodimeric antibody). The heterodimeric protein (eg, heterodimeric antibody) thus produced contains a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative CH3 domain.

[0105] In another example, the first protein provided in (a) of the method of the present disclosure can be a heterodimeric protein (e.g., a first heterodimeric antibody), and the second protein provided in (b) can also be a heterodimeric protein (e.g., a second heterodimeric antibody). The first heterodimeric protein (e.g., the first heterodimeric antibody) can contain two 351 / 366 positive IgG CH3 domains, and the second protein (e.g., the second heterodimeric antibody) can contain two 351 / 368 negative IgG CH3 domains. Upon completion of the reduction and reoxidation steps of the method of the present disclosure, the resulting heterodimeric protein (e.g., a heterodimeric antibody) contains one monomer of the first heterodimeric protein (e.g., one half-body of the first heterodimeric antibody) and one monomer of the second heterodimeric protein (e.g., one half-body of the second heterodimeric antibody). The heterodimeric protein (eg, heterodimeric antibody) thus produced contains a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative CH3 domain.

[0106] In certain embodiments, in which the first and / or second protein provided in (a) of the method is a monomeric protein, it is a monomeric Fc region.

[0107] In certain embodiments, in which the first and / or second protein provided in (a) of the method is a dimer (heterodimer or homodimer), it is a dimeric Fc region.

[0108] In some embodiments, the first protein and / or the second protein is selected from the group comprising or consisting of antibodies and half-antibodies, or fragments thereof. Antibodies, half-bodies, and fragments thereof are described elsewhere herein. In some embodiments, the fragment is a monomeric or dimeric Fc region.

[0109] In certain embodiments, the first and / or second proteins can have sequences of any origin, e.g., mouse and human sequences. The first and / or second proteins can consist of sequences from only one origin, such as a fully human antibody, or can have sequences from multiple origins, e.g., chimeric or humanized antibodies. It may be desirable for a therapeutic first and / or second protein (e.g., a first and / or second antibody) to be as close as possible to the natural antibody of the subject being treated (e.g., a human antibody for a human subject). Thus, in certain embodiments, the antibody, half antibody, or fragment thereof is a human antibody, human half antibody, or fragment thereof. It will be understood that the fragment comprises at least an IgG CH3 domain, a hinge region, and a Fab arm. In certain embodiments, the Fab arm comprises a variable heavy chain domain. In certain embodiments, the Fab arm comprises a variable heavy chain domain and a variable light chain. In certain embodiments, the Fab arm comprises a variable heavy chain domain but does not comprise (i.e., lacks) a variable light chain.

[0110] In certain embodiments, when the first and / or second protein is an antibody, it is a homodimeric antibody or a heterodimeric antibody.

[0111] In certain embodiments, the first protein is an antibody having a first binding specificity and the second protein is an antibody having a second, different binding specificity.

[0112] In one embodiment, the first protein can be an antibody that comprises a binding specificity and the second protein can be an antibody that comprises another, different binding specificity.

[0113] As previously described, generation of heterodimeric proteins by the methods described herein is facilitated by the preferential binding of the 351 / 366 positive CH3 domain (containing positively charged amino acid residues at positions 351 and 366) to the 351 / 368 negative CH3 domain (containing negatively charged amino acid residues at positions 351 and 368).

[0114] In certain embodiments, the 351 / 366-correct CH3 domain may contain an amino acid K or R at position 351 and an amino acid K or R at position 366. For example, the 351 / 366-correct CH3 domain may contain an amino acid K at position 351 and an amino acid R at position 366, an amino acid R at position 351 and an amino acid K at position 366, an amino acid K at position 351 and an amino acid K at position 366, or an amino acid R at position 351 and an amino acid R at position 366. In certain embodiments, the 351 / 366-correct CH3 domain may contain an amino acid K at position 351 (i.e., 351K) and an amino acid K at position 366 (i.e., 366K). A first protein comprising two of the latter 351 / 366-correct CH3 domains is referred to herein as a KKKK first protein. In certain embodiments, the KKKK first protein is a homodimer or a heterodimer. In some embodiments, the KKKK first protein is a homodimeric IgG antibody or fragment thereof, or a heterodimeric IgG antibody or fragment thereof. In some embodiments, the KKKK IgG antibody or fragment thereof is an IgG1, IgG2, IgG3, or IgG4 antibody or fragment thereof.

[0115] In certain embodiments, the 351 / 368 negative CH3 domain can comprise an amino acid D or E at position 351 and an amino acid D or E at position 368. For example, the 351 / 368 negative CH3 domain can comprise an amino acid D at position 351 and an amino acid E at position 368, or an amino acid E at position 351 and an amino acid D at position 368, or an amino acid E at position 351 and an amino acid E at position 368, or an amino acid D at position 351 and an amino acid D at position 368. In certain embodiments, the 351 / 368 negative CH3 domain can comprise an amino acid D at position 351 (i.e., can comprise 351D) and an amino acid E at position 368 (i.e., can comprise 368E). A second protein comprising two of the latter 351 / 368 negative CH3 domains is referred to herein as a DEDE second protein. In certain embodiments, the DEDE second protein is a homodimer or a heterodimer. In certain embodiments, the DEDE second protein is a homodimeric IgG antibody or fragment thereof, or a heterodimeric IgG antibody or fragment thereof. In certain embodiments, the DEDE IgG antibody or fragment thereof is an IgG1, IgG2, IgG3, or IgG4 antibody or fragment thereof.

[0116] In some embodiments, the 351 / 366-positive CH3 domain may comprise an amino acid K at position 351 (i.e., may comprise 351K) and an amino acid K at position 366 (i.e., may comprise 366K), and the 351 / 368-negative CH3 domain may comprise an amino acid D at position 351 (i.e., may comprise 351D) and an amino acid E at position 368 (i.e., may comprise 368E). A heterodimer having one 351 / 366-positive CH3 domain and one 351 / 368-negative CH3 domain, as described above, is referred to herein as a DEKK heterodimer. In some embodiments, the DEKK heterodimer is a heterodimeric IgG antibody or fragment thereof. In some embodiments, the DEKK IgG antibody or fragment thereof is an IgG1, IgG2, IgG3, or IgG4 antibody or fragment thereof.

[0117] It will be understood that the above amino acid positions are based on the human IgG CH3 domain, however, in embodiments in which a non-human IgG CH3 domain is utilized, the same amino acid substitutions are introduced at the corresponding amino acid residues.

[0118] In one embodiment, the first and second proteins are provided in a ratio that is favorable for producing a heterodimeric protein by the methods described herein. By "favorable," we mean that the ratio of heterodimeric protein produced is increased compared to when equal amounts of the first and second proteins are provided (the heterodimeric protein has a 351 / 366 positive CH3 domain and a 351 / 368 negative CH3 domain). In one embodiment, the ratio of the first protein (containing a 351 / 366 positive IgG CH3 domain) to the second protein (containing a 351 / 368 negative IgG CH3 domain) is 20:1 to 1:20 (w / w). Ratios exceeding these amounts can be used, but may actually result in less efficient protein use. In one embodiment, the ratio is 10:1 to 1:10. In another embodiment, the ratio is 5:1 to 1:5. In one embodiment, the ratio is at least 1:1 (w / w), such as at least 1.2:1 (w / w), such as at least 1.5:1 (w / w) or at least 2:1 (w / w). In one embodiment, the ratio of the first protein (comprising the 351 / 366 positive IgG CH3 domain) to the second protein (comprising the 351 / 368 negative IgG CH3 domain) is between 1:1 (w / w) and 2:1 (w / w).

[0119] In one embodiment, the ratio of the KKKK first protein (comprising a 351 / 366 positive IgG CH3 domain) to the DEDE second protein (comprising a 351 / 368 negative IgG CH3 domain) is at least 1:1 (w / w), e.g., at least 1.2:1 (w / w), e.g., at least 1.5:1 (w / w), or at least 2:1 (w / w). In one embodiment, the ratio of the KKKK first protein (comprising a 351 / 366 positive IgG CH3 domain) to the DEDE second protein (comprising a 351 / 368 negative IgG CH3 domain) is between 1:1 (w / w) and 2:1 (w / w).

[0120] In one embodiment, the ratio of the KK first protein (comprising a 351 / 366 positive IgG CH3 domain) to the DEDE second protein (comprising a 351 / 368 negative IgG CH3 domain) is 20:1 to 1:20 (w / w). In one embodiment, the ratio is 10:1 to 1:10. In one embodiment, the ratio is 5:1 to 1:5. In one embodiment, the ratio is at least 1:1 (w / w), e.g., at least 1.2:1 (w / w), e.g., at least 1.5:1 (w / w), or at least 2:1 (w / w). In one embodiment, the ratio of the KK first protein (comprising a 351 / 366 positive IgG CH3 domain) to the DEDE second protein (comprising a 351 / 368 negative IgG CH3 domain) is 1:1 (w / w) to 2:1 (w / w).

[0121] When incubating the first and second proteins under reducing conditions and subsequently reoxidizing them, it will be understood that any suitable amount of the first protein and the second protein can be used. Appropriate concentrations can be determined by one of skill in the art, for example, using the methodology described in the Examples section below. As a non-limiting example, each of the first and second proteins is used under the reducing conditions at a concentration of at least 50 μg / ml, e.g., at least 0.1 mg / ml, at least 1.0 mg / ml, at least 10 mg / ml, or at least 25 mg / ml, but not more than 100 mg / ml. In one embodiment, when each of the first and second proteins is an antibody, each of the first and second proteins is used under the reducing conditions at a concentration of at least 50 μg / ml, e.g., at least 0.1 mg / ml, or at least 1.0 mg / ml. Typically, each can be used in a concentration range of 50 μg / ml to 2 mg / ml, e.g., 50 μg / ml to 10 mg / ml. By way of example only, each can be used at a concentration of about 1.1 mg / ml.

[0122] It will also be understood that additional reagents may be present under the reducing conditions to facilitate the process of obtaining a reduced first protein and a reduced second protein. For example, in one embodiment, cystamine is added under reducing conditions at an appropriate concentration, e.g., in the range of 2 to 70 mM, to further reduce any homodimers present in the reaction. For example, if the first and second proteins are antibodies (e.g., homodimeric antibodies), cystamine is added at a concentration of about 2 to about 70 mM under reducing conditions. In one embodiment, cystamine is added at a concentration of 2 to 32 mM under reducing conditions to further reduce any homodimers present in the reaction. For example, if the first and second proteins are antibodies (e.g., homodimeric antibodies), cystamine is added at a concentration of about 2 to about 32 mM under reducing conditions.

[0123] In some embodiments, the first and second proteins are obtained independently of each other. By way of example only, the first and second proteins are produced by different host cells. Exemplary methods for obtaining the first and second proteins are described in the Examples section below. However, other methods are known to those skilled in the art.

[0124] A "host cell" can be any host cell capable of expressing a recombinant DNA molecule and expressing a binding moiety known in the art.

[0125] In one embodiment, the first protein (such as a homodimeric antibody) and / or the second protein (such as a homodimeric antibody) are obtained under serum-free conditions (e.g., by culturing host cells in FreeStyle 293 medium or FreeStyle Cho medium, Invitrogen).

[0126] In one embodiment, between providing the first and second proteins in (a) and (b) and incubating under reducing conditions in the method, the first and / or second proteins are purified using methods known in the art. Such methods may include precipitation, centrifugation, filtration, size exclusion chromatography, affinity chromatography, cation and / or anion exchange chromatography, and hydrophobic interaction chromatography. For a mixture of antibodies containing IgG molecules, Protein A or Protein G affinity chromatography can be used (see, e.g., U.S. Patent Nos. 4,801,687 and 5,151,504).

[0127] The incubation under reducing and reoxidizing conditions of this method promotes recombination of the first and second proteins (e.g., antibodies, or dimeric fragments thereof), or binding of the first and second proteins (e.g., half-bodies or monomeric fragments thereof) to generate a heterodimeric protein. Recombination of the first and second proteins occurs when the first and second proteins exchange monomers (e.g., half-bodies or dimeric fragments thereof) to generate a heterodimeric protein containing two different IgG CH3 domains that can form a CH3-CH3 interface.

[0128] As used herein, the term "incubate" or "incubating" refers to holding, storing, or maintaining a first and second protein together under relevant conditions (i.e., reducing conditions). The first and second proteins are incubated together in a composition or preparation comprising the first and second proteins.

[0129] Said incubation under reducing conditions of this method includes incubating the proteins (a) and (b) together under reducing conditions to provide a reduced first protein and / or a reduced second protein.

[0130] The term "reducing conditions" refers to an environment in which the first and / or second protein are more likely to be reduced than oxidized. In certain embodiments, reducing conditions can result in "disulfide bond reduction" (i.e., the process of cleaving a disulfide bond, thereby generating two thiol groups (-SH groups)). One of skill in the art will understand that incubating an antibody (homodimeric or heterodimeric antibody) under reducing conditions can reduce disulfide bonds (such as those in the core hinge region) and separate the antibody into two half-bodies.

[0131] The step of incubating the first and second proteins under reducing conditions can include incubating the first and second proteins in the presence of a reducing agent. The term "reducing agent" refers to a compound that reduces surrounding molecules, i.e., a compound that changes surrounding molecules to become more reduced. A reducing agent can act by donating electrons, thereby reducing a substrate (i.e., the first and / or second protein) and then becoming oxidized itself.

[0132] Examples of reducing agents include 2-mercaptoethylamine (2-MEA), dithiothreitol (DTT), dithioerythritol (DTE), glutathione, tris(2-carboxyethyl)phosphine (TCEP), L-cysteine, beta-mercaptoethanol, thioglycolate, cysteamine, homocysteine, penicillamine, and / or sodium borohydride.

[0133] In one embodiment, when the reducing agent is 2-MEA, its concentration is about 25 mM to about 100 mM, for example, about 50 mM to about 75 mM. In one embodiment, the concentration of 2-MEA is about 75 mM.

[0134] It will be appreciated that the time of incubation with the reducing agent may depend on the concentration and / or temperature at which the incubation is carried out, for example, a higher concentration of the reducing agent may allow for a shorter incubation time and / or a lower incubation temperature.

[0135] In one embodiment, when the concentration of 2-MEA is about 75 mM, the incubation continues for at least 300 minutes at about 31° C. In one embodiment, when the concentration of 2-MEA is about 75 mM, the incubation continues for about 300 minutes at about 31° C.

[0136] In one embodiment, the reducing agent does not include an enzyme.

[0137] Additionally or alternatively, incubating the first and second proteins under reducing conditions can include incubating the first and second proteins at pH 6.0 or higher, e.g., pH 7.0 or higher, pH 8.0 or higher, pH 9.0 or higher, pH 10.0 or higher, pH 11.0 or higher, or pH 12.0. For example, the first and second proteins are incubated at pH 6.0-11.0 or pH 6.0-10.0, optionally at a pH of 7.0-8.0 (e.g., pH 7.3-7.5). In one embodiment, the pH is 7.4. Additionally or alternatively, incubating the first and second proteins under reducing conditions can include incubating the proteins at a redox potential of -150 mV to -600 mV, optionally at a redox potential of -250 mV to -400 mV. Examples of suitable reducing conditions are known in the art. Some examples are described in Labrijn AF., Nature Protocols 2014, Vol. 9, No. 10, pp 2450-2463.

[0138] The method further comprises reoxidizing the reduced protein to obtain the heterodimeric protein.

[0139] As used, the term "reoxidation" or "oxidation" means to undergo or cause a reaction in which electrons are lost to another species. In the context of the present disclosure, the reoxidation step allows the resulting reduced first protein and reduced second protein to combine together to form a heterodimeric protein comprising two different IgG CH3 domains that can form a CH3-CH3 interface.

[0140] A first protein with a 351 / 366 positive IgG CH3 domain and a second protein with a 351 / 368 negative IgG CH3 domain preferentially bind to each other, resulting in a higher proportion of heterodimeric proteins containing two different IgG CH3 domains than dimeric proteins (homodimers or heterodimers) with identically charged IgG CH3 domains. Therefore, the present disclosure provides a method for the efficient and controlled production of well-defined mixtures of Ig antibodies or heterodimeric fragments thereof, including a high proportion of bispecific antibodies in the mixture. In systems where bispecific antibodies are desired, the proportion of bispecific antibodies can be at least 80%, at least 85%, at least 90%, at least 95%, or at least 97% or more. This means that only 5% or less, or even 3% or less, of monospecific bivalent by-products can be obtained. Notably, the monomeric by-products, i.e., half-molecules, produced by the methods described herein are more stable than the monomeric by-products produced by at least some other methods known in the art (such as those described in WO2011131746), which is advantageous because these half-molecules can be subjected to reducing conditions (described elsewhere in this disclosure) to produce the desired heterodimeric protein without the need for repeated oxidation steps.

[0141] In some embodiments, reoxidation of the reduced protein is achieved by removing the reducing agent from a mixture of the reduced first protein and the reduced second protein. By way of example only, the reducing agent is removed by diafiltration (e.g., as described in Labrijn AF., Nature Protocols 2014, Vol. 9, No. 10, pp. 2450-2463). It will be appreciated that complete removal of the reducing agent is not necessary to reoxidize the reduced first and second proteins. For example, with respect to 2-MEA as the reducing agent, lowering the concentration to less than 50 μM may be sufficient to reoxidize the reduced first and second proteins. Another example of a method for reoxidizing the reduced protein is buffer exchange. In some embodiments, buffer exchange is performed with PBS. In some cases, PBS has a pH of 7.4. Methods for performing buffer exchange are known to those skilled in the art. In some embodiments, buffer exchange is performed using a Zeba plate, as exemplified in the Examples section of this disclosure. As another example, buffer exchange into PBS can be performed using an AKA Pure 25 system equipped with a Spark ALIAS autosampler and a desalting column (such as a HiPrep 26 / 10). In such an example, buffer exchange can occur at a flow rate of 6 mL / min and 20°C.

[0142] Once buffer exchange is complete, the sample can be maintained at a temperature of about 4° C. for about 24, 48, 64 hours or longer to allow for complete reoxidation.

[0143] Alternatively or additionally, the reoxidation step comprises incubating the reduced first protein and the reduced second protein with an oxidizing agent.

[0144] In certain embodiments, the method further comprises a step of concentrating and / or isolating the resulting heterodimeric protein. The heterodimeric protein can be concentrated and / or isolated from any contaminants that may result from the methods described herein by routine methods, such as routine purification methods. Such contaminants may include homodimeric proteins (proteins containing two monomers, each having the same IgG CH3 domain) and / or monomeric proteins (e.g., half-bodies). Methods for purifying the resulting heterodimeric protein may include precipitation, centrifugation, filtration, size exclusion chromatography, affinity chromatography, cation and / or anion exchange chromatography, hydrophobic interaction chromatography, etc. For antibody mixtures containing IgG molecules, Protein A or Protein G affinity chromatography can be used (see, e.g., U.S. Patent Nos. 4,801,687 and 5,151,504).

[0145] In certain embodiments, the incubation under reducing conditions and reoxidizing the reduced proteins together can also be described as incubating the first and second proteins under conditions sufficient for cysteines in the CH3 regions to undergo disulfide bond isomerization to form a heterodimeric protein. In certain embodiments, when the first and second proteins are antibodies or half-antibodies, the incubation under reducing conditions and reoxidizing the reduced proteins together can also be described as incubating the first and second proteins under conditions sufficient for cysteines in the core hinge regions of the first and second proteins to undergo disulfide bond isomerization to form a heterodimeric protein.

[0146] In certain embodiments, the first protein and / or the second protein comprise a fusion protein. The fusion protein may comprise an antibody binding domain, an scFv, a ligand, a protein receptor, or a cytokine. In certain embodiments, the first and / or second protein comprise an antibody binding domain, an scFv, a ligand, a protein receptor, or a cytokine. In certain embodiments, the heterodimeric protein is a bifunctional or multifunctional fusion protein.

[0147] In a further aspect, the description provides an isolated heterodimeric protein obtainable by the methods of the disclosure.

[0148] In certain embodiments, the heterodimeric protein obtained by the methods of the disclosure is an IgG antibody, such as an IgG1, IgG2, IgG3, or IgG4.

[0149] In certain embodiments, an IgG antibody may comprise two light chains with non-identical sequences.

[0150] In a further aspect, the present disclosure provides an isolated heterodimeric antibody comprising a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative IgG CH3 domain, wherein the 351 / 366 positive CH3 domain and the 351 / 368 negative CH3 domain are capable of forming a CH3-CH3 interface, wherein the 351 / 368 negative CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D, and said 351 / 366 positive CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R, and said heterodimeric antibody further comprises two light chains having non-identical sequences.

[0151] In one embodiment, the 351 / 368 negative CH3 domain comprises 351D and 368E, and the 351 / 366 positive IgG CH3 domain comprises 366K and 351K.

[0152] It will be understood that aspects relating to heterodimeric proteins mentioned herein in the context of the methods of the present disclosure equally apply to heterodimeric proteins obtainable by the methods described herein, and to isolated heterodimeric antibodies comprising a 351 / 366 positive IgG CH3 domain and a 351 / 368 negative IgG CH3 domain, wherein the 351 / 366 positive CH3 domain and the 351 / 368 negative CH3 domain are capable of forming a CH3-CH3 interface, wherein the 351 / 368 negative CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D, and wherein said 351 / 366 positive CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R, and wherein said heterodimeric antibody further comprises two light chains with non-identical sequences.

[0153] Mutations at positions 351 / 366 and / or 351 / 368 can be combined with any of the modifications described in WO2020 / 226502A2, which is incorporated herein by reference in its entirety. Similarly, mutations at positions 351 / 366 and / or 351 / 368 can be combined with any of the modifications described in WO2021 / 235936A1, which is incorporated herein by reference in its entirety.

[0154] In a further aspect, the present specification provides a pharmaceutical composition comprising the disclosed isolated heterodimeric protein and a pharmaceutically acceptable carrier. The term "pharmaceutical composition" refers to a formulation in a form in which the biological activity of the active ingredient (e.g., a heterodimeric protein of the present disclosure) is effective and the formulation does not contain additional ingredients that are unacceptably toxic to the subject to which it is administered. The term "pharmaceutically acceptable carrier" refers to any carrier useful for solubilizing and delivering an agent (e.g., a heterodimeric protein of the present disclosure) to a subject. Many pharmaceutically acceptable carriers are known in the art. By way of example only, these include saline, phosphate buffer, or phosphate-buffered saline. The composition may further typically contain pharmaceutically acceptable concentrations of salts, buffers, preservatives, compatible carriers, supplemental immune-enhancing agents such as adjuvants and cytokines, and any other therapeutic agents. The composition may also contain antioxidants and / or preservatives. Antioxidants can include thiol derivatives (e.g., thioglycerol, cysteine, acetylcysteine, cystine, dithioerythreitol, dithiothreitol, glutathione), tocopherol, butylhydroxyanisole, butylhydroxytoluene, sulfites (e.g., sodium sulfate, sodium bisulfite, sodium acetone bisulfite, sodium metabisulfite, sodium sulfite, sodium formaldehyde sulfoxylate, sodium thiosulfate), and nordihydroguaiaretic acid. Suitable preservatives can be, for example, phenol, chlorobutanol, benzyl alcohol, methylparaben, propylparaben, benzalkonium chloride, and cetylpyridinium chloride.

[0155] term [1] 1. A method for producing a heterodimeric protein comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, comprising: (a) a first protein comprising a first CH3 domain comprising positively charged amino acid residues at positions 351 and 366; and (b) a second protein comprising a second CH3 domain comprising negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering, providing; incubating the proteins of (a) and (b) together under reducing conditions to provide a reduced first protein and a reduced second protein; and reoxidizing the reduced first and second proteins to obtain a heterodimeric protein; A method comprising: [2] The method according to [1] above, wherein the first protein and / or the second protein comprises or is selected from the group consisting of a monomeric protein, a homodimeric protein, and a heterodimeric protein. [3] The method of any preceding clause, wherein the IgG CH3 domain is an IgG1, IgG2, IgG3, or IgG4 CH3 domain, and optionally the CH3 domain is a human CH3 domain. [4] 5. The method of any preceding clause, wherein the first protein and / or second protein comprises or is selected from the group consisting of an antibody and a half antibody, or fragment thereof; and optionally, the antibody, half antibody, or fragment thereof is a human antibody, half antibody, or fragment thereof. [5] The method according to [4] above, wherein the fragment is a hinge comprising a monomeric Fc region or a hinge comprising a dimeric Fc region. [6] The method according to [4] or [5] above, wherein the antibody, half antibody, or fragment thereof is an IgG1, IgG2, IgG3, or IgG4 antibody, half antibody, or fragment thereof. [7] The method of any preceding clause, wherein the first protein is an antibody having a first binding specificity and the second protein is an antibody having a second, different binding specificity. [8] The method of any preceding clause, wherein the first protein and / or the second protein is a homodimeric antibody. [9] The method of any preceding clause, wherein the first protein and / or the second protein is a heterodimeric antibody.

[10] Item 11. The method of any of the preceding clauses, wherein the heterodimeric protein obtained after reoxidation of the reduced protein is a heterodimeric antibody.

[11] The method of any one of claims 9 to 10, wherein the heterodimeric antibody is multivalent, and optionally the multivalent antibody is a bivalent, trivalent, tetravalent, or up to hexavalent antibody.

[12] The method according to any one of [9] to

[11] above, wherein the heterodimeric antibody is a multispecific antibody, and optionally the multispecific antibody is a bispecific antibody, a trispecific antibody, or a tetraspecific antibody.

[13] The method according to any one of [9] to

[12] above, wherein the heterodimeric antibody comprises two non-identical light chains.

[14] The method according to any one of [9] to

[12] above, wherein the heterodimeric antibody comprises a single light chain.

[15] The method according to any one of [9] to

[12] above, wherein the heterodimeric antibody comprises three light chains.

[16] The method according to any one of [9] to

[12] above, wherein the heterodimeric antibody comprises three non-identical light chains.

[17] The method according to any one of [9] to

[12] above, wherein the heterodimeric antibody comprises two identical light chains and further comprises a light chain that is non-identical to the identical light chains.

[18] The method of any of the preceding clauses, wherein the light chain is a member from either the kappa or lambda family.

[19] The method of any preceding clause, wherein the first protein and the second protein comprise identical hinge regions.

[20] In certain embodiments, the method of any preceding clause, wherein the first protein and the second protein comprise an IgG1 hinge region. [twenty one] The method of any preceding clause, wherein the first protein and the second protein comprise an IgG2 hinge region. [twenty two] The method of any preceding clause, wherein the first protein and the second protein comprise an IgG3 hinge region. [twenty three] The method of any preceding clause, wherein the first protein and the second protein comprise an IgG4 hinge region.

[24] The method of any preceding clause, wherein the first CH3 domain comprises 351K and 366R, 351R and 366K, 351K and 366K, or 351R and 366R. [twenty five] The method of any preceding clause, wherein the second CH3 domain comprises 351D and 368E, 351E and 368D, 351D and 368D, or 351E and 368E.

[26] The first CH3 domain comprises 351K and 366K; and The method of any preceding clause, wherein the second CH3 domain comprises 351D and 368E.

[27] The method of any of the preceding clauses, wherein the first and / or second proteins of a) and b) comprise one or more Fc modifications.

[28] The method of any preceding clause, wherein the first and / or second proteins of a) and b) are Fc engineered.

[29] The method of any of the preceding clauses, wherein the first and / or second proteins of a) and b) are Fc-silenced or Fc-enhanced.

[30] The method of any of the preceding clauses, wherein the first and / or second proteins of a) and b) comprise a CH2 domain with a mutation at positions 235 and / or 236 that affects ADCC.

[31] 10. The method of any preceding claim, wherein the second protein is obtained independently from the first protein.

[32] 10. The method of any preceding claim, wherein the reducing conditions are: (a) incubating the protein in the presence of a reducing agent, optionally comprising or selected from the group consisting of 2-mercaptoethylamine (2-MEA), dithiothreitol (DTT), dithioerythritol (DTE), glutathione, tris(2-carboxyethyl)phosphine (TCEP), L-cysteine, beta-mercaptoethanol, thioglycolate, cysteamine, homocysteine, penicillamine, and sodium borohydride; and / or (b) incubating the protein at a pH of 6.0 to 12.0, optionally a pH of 7.0 to 8.0; and / or (c) incubating the protein at a redox potential of -150 to -600 mV, optionally the redox potential is -250 to -400 mV; A method comprising:

[33] 10. The method of any of the preceding claims, further comprising the step of enriching and / or isolating the heterodimeric protein obtained after reoxidizing the reduced protein.

[34] 33. The method of claim 33, wherein the heterodimeric protein is concentrated and / or isolated using a method selected from the group consisting of precipitation, centrifugation, filtration, size exclusion chromatography, affinity chromatography, cation exchange chromatography and / or anion exchange chromatography, and hydrophobic interaction chromatography.

[35] An isolated heterodimeric protein obtainable by the method according to any of the preceding paragraphs.

[36] 1. An isolated heterodimeric antibody comprising a first IgG CH3 domain and a second IgG CH3 domain, the first CH3 domain and the second CH3 domain are capable of forming a CH3-CH3 interface; the first CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R; and the second CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D; An isolated heterodimeric antibody, wherein said heterodimeric antibody further comprises two or more light chains having non-identical sequences.

[37] The isolated heterodimeric antibody according to

[36] , wherein the first CH3 domain comprises 351K and 366K, and the second CH3 domain comprises 351D and 368E.

[38] A pharmaceutical composition comprising the isolated heterodimeric protein according to

[35] or the isolated heterodimeric antibody according to

[36] or

[37] , and a pharmaceutically acceptable carrier.

[39] The method according to any one of [1] to

[34] above, wherein the first protein and / or the second protein comprises an antibody binding domain, an scFv, a ligand, a protein receptor, or a cytokine.

[40] The method according to

[28] , wherein the heterodimeric protein is a di- or multifunctional fusion protein. [Example]

[0156] Example 1: Transfection, expression, and purification of IgG antibodies used in Fab arm exchange (FAE) The aim of the following experiments was to directly compare the results of FAE using the various mutations listed in Table 1 (i.e., the DEKK mutation and the mutation 405L-409R).

[0157] Expi293F (商標) Cells (ThermoFisher Scientific) were cultured in 100 mL Expi293F (商標) The cells were cultured in expression medium (ThermoFisher Scientific, cat.# A14351010) and treated with ExpiFectamine. (商標) Various expression vectors (see Table 1 ) encoding heavy chains (HC) and light chains (LC) were transiently transfected using a 293 transfection kit (ThermoFisher Scientific, cat.# A14635) and OptiMEM I reduced serum medium (Gibco, cat.# 31985062). [Table 1] The expression vector contains DNA constructs encoding the CH3 domain with the indicated modifications, a model heavy chain (HC), and a model light chain (LC).

[0158] Briefly, DNA-Opti-MEM and Expifectamine-Opti-MEM mixtures were prepared by adding 50 μg of plasmid DNA (0.1 mL from a 0.5 mg / mL DNA stock) to 3 mL of Opti-MEM. (商標) Dilute with medium I, and 0.16 mL of ExpiFectamine (商標) The 293 reagent was diluted with 2.8 mL of Opti-MEM by swirling or inverting, followed by a 5-minute incubation. The DNA-Opti-MEM mixture was then added to the Expifectamine-Opti-MEM mixture, and the tube was inverted 4-5 times and incubated at room temperature for 15 minutes. (商標) The total volume of the 293 / plasmid DNA complex (approximately 6.06 mL) was added dropwise to the cells, with gentle mixing of the flask during the addition, followed by incubation (37°C incubator, ≥80% relative humidity, 8% CO2, 155 rpm on an orbital shaker).

[0159] 18-22 hours after transfection, ExpiFectamine (商標) Enhancer 1 and Enhancer 2 from the 293 transfection kit were added to the transfection flask with gentle mixing. Six days after transfection, the medium was collected, and the cells were centrifuged at 500 g for 10 minutes at room temperature. The supernatant was collected and transferred to a new 50 ml tube, and the cells were centrifuged at 3000 g for 20 minutes. The supernatant was filtered using a 0.45 μm bottle-top filter, and the IgG concentration was measured using a ForteBIO Octet-QK system based on biolayer interferometry (BLI). This allows for real-time quantification and kinetic characterization of biomolecular interactions. The supernatant was used for AKTA purification.

[0160] Culture supernatants containing 9-12 mg of IgG were purified using an AKTA pure system (Cytiva, EN490, serial number: 2031829) with a Protein A column (GE Healthcare / cat#11-0034-95, according to GE Healthcare's instructions) and eluted with 0.1 M citrate buffer (pH 3.0). The eluate was immediately neutralized with an equal volume of 1.0 M Tris-HCl pH 8.0 or directly rebuffered to PBS using a desalting column (Cytiva #17-1408-01). The purified IgG molecules were used in FAE as described in Example 2.

[0161] The concentration of all samples was determined by measuring the absorbance of the protein solution at 280 nm according to the Beer-Lambert law, correcting for PBS as a blank, and using a typical molar extinction coefficient of 1.45 mol -1 dm 3 cm -1 The amino acid composition was adjusted using HCl. The antibody concentrations were all 1.2 to 2.1 mg / mL in PBS pH 7.4.

[0162] Example 2: FAE Protocol Briefly, IgG molecules were incubated without shaking in the presence of 75 mM MEA (β-mercaptoethylamine hydrochloride) at 31°C, pH 7.4. After 5 hours, samples were buffer exchanged (Bex) into PBS pH 7.4 at room temperature using a Zeba plate. Samples were stored at least overnight at 4°C for reoxidation.

[0163] FAE reactions were performed in a 96-well format on the IgG molecules produced in Example 1, as shown in Table 2. [Table 2] FAE reactions #1-4 contain a mixture of two IgG molecules, whereas FAE reactions #5-12 contain individual IgGs as controls.

[0164] Fresh MEA stock solution (750 mM) was prepared by dissolving 852 mg of cysteamine hydrochloride (Sigma cat# 30078) in 5-6 mL of PBS (pH 7.4). The pH was adjusted to 7.3-7.5 by adding NaOH (5 M) to the solution at room temperature, and the solution was filled up with PBS pH 7.4 to a final volume of 10 mL to obtain a 750 mM stock solution. The solution was filtered through a 0.2 μm filter before use.

[0165] Briefly, the IgG samples listed in Table 2 were prepared in a deep-well plate (Plate 1, final volume 175 μL, adjusted with PBS pH 7.4) using IgG at a concentration of 1.1 mg / mL. 11 μL of 750 mM MEA reducing agent was pipetted into 12 wells of another deep-well plate (Plate 2). 100 μL of each prepared sample from Plate 1 was gently mixed with MEA from Plate 2, covered with an aluminum foil seal, and stored at 31 °C for 5 hours without shaking. The remaining material (75 μL) on Plate 1 served as an unreacted control, i.e., not exposed to MEA or Bex, and was stored at 4 °C in the dark until further use. After incubation, 100 μL of each sample was transferred to Zeba®. (商標)A Spin Desalting Plate, 96 well (Thermo Fisher Scientific, cat. # 89807) was used for buffer exchange (bex).

[0166] Buffer exchange: Samples were buffer exchanged into PBS pH 7.4 (1x, Gibco cat. # 10010-015) using a Zeba™ Spin Desalting kit (Thermo Fisher Scientific, cat. # 89807) to remove the reducing agent and allow reformation of disulfide bonds.

[0167] In simple terms, Zeba (商標) The spin desalting plate was equilibrated to room temperature and assembled on top of the wash plate. The plate assembly was centrifuged to remove the storage solution, and the wash plate was blotted dry on paper towels. Three wash steps were performed by adding wash buffer (1x, Gibco cat. # 10010-015), after which the plate was centrifuged, the flow-through discarded, and blotted dry on paper towels.

[0168] Next, Zeba (商標) The desalting plate was stacked on top of the collection plate, and the IgG samples were loaded into the wells and centrifuged. The flow-through containing the IgG was collected in another Zeba (商標) The flow-through was transferred to a plate and centrifuged. To achieve a more thorough buffer exchange, the flow-through was transferred to another Zeba (商標) The plates were re-transferred and centrifuged. All centrifugations were performed at 1000xg for 2 minutes. (商標)The third flow-through, containing the sample that had passed through the plate three times, was retained for IgG concentration measurement. The buffer-exchanged samples were stored overnight (without further pipetting, shaking, or mixing) in the capped collector plate at 4°C to allow for complete reoxidation of the material. The buffer-exchanged samples and the unreacted control stored at 4°C on Plate 1 were measured on the Lunatic system as described in Example 1. The resulting IgG molecules were subjected to SDS-PAGE using a Labchip under reducing and non-reducing conditions, using CIEX analysis and HP-SEC, as appropriate.

[0169] Example 3: LabChip analysis Approximately 1 μg of purified reaction product IgG, including the unreacted control, was analyzed on a LabChip (LabChip GXII Touch HT; Perkin Elmer) under nonreducing conditions using the Protein Clear HR Reagent kit (staining solution, sample buffer, protein gel matrix, protein ladder, lower limit marker, wash buffer; Perkin Elmer CLS960014) and Protein Express Assay LabChip for use with the GXII Touch HT (Elmer 760499) according to the manufacturer's instructions.

[0170] The IgG, half-body, and IgG-1 LC bands were quantified using Labchip RX Reviewer software from samples run under non-reducing conditions. Only samples in which the IgG-1 LC band accounted for less than 15% of the signal were analyzed. Care was also taken to ensure that only samples containing >95% intact molecules (IgG dimers or half-bodies) were included in the starting material. The results are shown in Figure 1.

[0171] Samples taken before the FAE reaction behaved as follows: species containing the CH3-DE mutation formed primarily DEDE homodimers, while samples containing the CH3 KK mutation formed primarily half-bodies. Single-arm products with CH3 mutations 405L or 409R formed primarily homodimers. Samples with a mixture of DE / KK or 405L / 409R showed bands similar to those shown in Figure 1. Samples from reactions containing the DE / KK CH3 mutation showed a mixture of homodimers and half-bodies, while samples from proteins containing the CH3 mutation 405L / 409R showed primarily homodimers.

[0172] The single products from samples obtained after the FAE reaction with the DE or KK CH3 mutations mainly showed half-bodies, whereas samples with the 405L or 409R CH3 mutations formed homodimers.

[0173] The sample containing the mixture of CH3 species including DE / KK obtained after the FAE reaction showed an IgG peak containing half-body contamination, whereas the sample containing the 405L / 409R species showed an IgG peak containing unresolved homodimer contamination.

[0174] Example 4: HP-CIEX analysis To confirm that the FAE reaction actually led to the production of bispecific antibodies, all samples (reaction products and unreacted controls) produced in Example 2 were analyzed by CIEX (Agilent 1260 series). The column (TSK gel SP-STAT 7 μm, 4.6 mm ID x 10 cm L; Tosoh cat# 21964) was equilibrated using low ionic strength (25 mM sodium phosphate pH 6.0 ± 0.05, sodium dihydrogen phosphate dihydrate (NaH2PO4, 2 H2O; Sigma, ref. 71500) and disodium hydrogen phosphate dihydrate (Na2HPO4, 2 H2O; Sigma, ref. 71643) dissolved in Milli-Q water and filtered through a 0.45 μm membrane filter. The column was then equilibrated using a phosphate buffer containing salt (buffer B) (25 mM sodium phosphate, 1 M NaCl pH 6.0, sodium dihydrogen phosphate dihydrate (NaH2PO4, 2 H2O; Sigma, ref. 71500) and disodium hydrogen phosphate dihydrate (Na2HPO4, 2 H2O; Sigma, ref. 71643) dissolved in Milli-Q water. The antibody was removed from the column by increasing the Na+ concentration through a gradient of phosphate buffer containing increasing percentages of NaCl (Sigma, ref. S3014) and NaCl (Sigma, ref. S3014) dissolved in Milli-Q water and filtered through a 0.45 μm membrane filter. The injected sample mass for all test samples and controls was 10 μg with injection volumes ranging from 10 to 100 μL. The chromatograms were analyzed for peak pattern and retention time. The peak area of ​​the major peak was determined based on the results at 220 nm.

[0175] The percentage of bispecific antibody in samples that underwent FAE was used to calculate efficiency as a percentage of the FAE reaction. The difference between samples that underwent FAE and unreacted control samples was used to distinguish between bispecific antibody and contaminants after FAE. The results are shown in Figure 2a-f.

[0176] FAE with the DEKK mutant produced approximately 1% homodimers in reaction #1 and no detectable homodimers in reaction #3, whereas FAE with the 405L / 409R mutation produced approximately 3.5% homodimers in reaction #2 and approximately 4.6% homodimers in reaction #4 (see Figure 2a / b).

[0177] Additionally, in samples taken from reactions 1 and 3 after FAE, a peak corresponding to the DEKK heterodimer appeared (approximately 21 min and 16 min, respectively), which was absent in the individual samples but accounted for approximately 88% or 96% of the material eluting from CIEX. In samples from FAE reactions 5 and 9, an early-eluting peak (approximately 8.5 min and 1-6 min, respectively) appeared, indicating the formation of DE half-bodies. In samples taken from reactions 2 and 4 after FAE, a peak representing the heterodimer (approximately 16 min) accounted for approximately 83% or 81%, respectively, of the material eluting from CIEX, which was absent in the individual samples.

[0178] Example 5: HP-SEC analysis To detect the content of aggregates and IgG dimers and half antibodies in the samples before and after FAE under native conditions, the samples (i.e., reaction products and unreacted controls) produced according to Example 2 were analyzed using HP-SEC (Agilent 1260 series) using a TSK-gel G3000SWxl (Tosoh Bioscience - 808541), a TSK guard column SWXL (Tosoh Bioscience - 808543), and a 0.45 μm membrane filtered HP-SEC buffer (200 mM sodium phosphate, 50 mM NaCl, pH 7.0; sodium dihydrogen phosphate dihydrate (NaH2PO4, 2H2O; Sigma, ref. 71500), disodium dihydrogen phosphate hydrate (Na2HPO4, 2H2O; Sigma, ref. 71643), and NaCl (Sigma, ref. S3014) dissolved in Milli-Q water) was used.

[0179] The same amount of sample was injected (20 μg of each sample in an injection volume of 10–100 μL). The data obtained was used to estimate the purity (% of intact IgG dimer or half-body) of the sample. To ensure sufficient sample quality, the starting material must contain more than 95% intact molecules. If not, additional purification or repeated purification was considered. The chromatograms were analyzed for the retention time and relative peak area of ​​the peaks observed at 280 nm. The results are shown in Figure 3. When analyzed individually, no aggregates were present within the same sample, so the main peak represents the IgG-IgG dimer.

[0180] After FAE, a main peak of size corresponding to IgG was detected in reactions containing the DE and KK CH3 mutants (i.e., #1 and #3) with a small amount of half-bodies and no aggregates. After FAE, a main peak of size corresponding to IgG was detected in reactions containing the 405L and 409R CH3 mutants (i.e., #2 and #4) with the 405L and 409R CH3 mutants (i.e., #2 and #4) with a small amount of half-bodies and no aggregates. After FAE of individual single products from samples obtained after FAE reactions with the DE (#5 and #9) or KK (#6 and #10) CH3 mutants, they mainly showed half-bodies, whereas samples containing the 405L (#8 and #12) or 409R (#7 and #11) CH3 mutants formed homodimers.

[0181] Example 6: Expanded FAE Protocol Briefly, IgG molecules were incubated in the presence of β-mercaptoethylamine hydrochloride without shaking. The samples were then buffer exchanged (Bex) into PBS using an AKTA Pure 25 system equipped with a Spark ALIAS autosampler and a single HiPrep 26 / 10 desalting column, Cytiva. For reoxidation, the samples were stored at 4°C for a time sufficient for complete reoxidation to occur. FAE reactions were performed on the IgG molecules produced in Example 1, as shown in Table 3. [Table 3]

[0182] Fresh MEA stock solution (750 mM) was prepared by dissolving 852 mg of cysteamine hydrochloride (Sigma cat# 30078) in 5-6 mL of PBS (pH 7.4). The pH was adjusted to 7.3-7.5 by adding NaOH (5 M) to the solution at room temperature, and the solution was filled up with PBS pH 7.4 to a final volume of 10 mL to obtain a 750 mM stock solution. The solution was then filtered through a 0.2 µm filter before use.

[0183] The IgG mixture shown in Table 3 was prepared to a final concentration of 1.1 mg / mL in a final volume of 3.15 mL by adding PBS (10x stock Gibco Cat# 70011-051 pH 7.4 diluted with Fresenius Versylene sterile endotoxin-free water, Cat# B230531). 100 uL of this mixture was removed and stored at 4°C as an unreacted control, i.e., without exposure to MEA or Bex, and kept at 4°C in the dark until further use. To initiate the FAE, 350 uL of 75 mM MEA was added to the remaining 3.05 mL, and the solution was gently mixed and kept at 31°C for 5 hours without shaking.

[0184] Buffer exchange: After 5 hours, the samples were buffer exchanged (Bex) into PBS (diluted from 10x stock Gibco Cat# 70011-051 pH 7.4 using Fresenius Versylene sterile endotoxin-free water, Cat# B230531) at a flow rate of 6 mL / min and 20°C using an AKTA Pure 25 system equipped with a Spark ALIAS autosampler and one HiPrep 26 / 10 desalting column, Cytiva Cat# 17-5087-01. For reoxidation, samples were kept at 4°C for 64 hours (without pipetting, shaking, or mixing) to ensure complete reoxidation of the material. For quality control purposes, 60-100 μg of each sample was retained before gel filtration purification (i.e., pre-GF control).

[0185] Gel filtration The buffer-exchanged sample was concentrated to a volume of 2.4 + / - 0.4 mL (approximately 1.5 + / - 0.25 mg / mL protein concentration) using an Amicon 15 Ultra (30 kDa molecular weight cutoff) device (Merck / Millipore Cat# UFC903096). The material was then loaded onto an AKTA Pure 25 system equipped with a Spark ALIAS autosampler. Gel filtration purification was performed to separate the sample based on size using a Superdex 200 increase 16 / 40 column (Cytiva Cat# 29321905) with PBS (prepared as described above) as the mobile phase at a flow rate of 1 mL / min (fraction size 0.5 mL, autosampler injection loop 10 mL, temperature 20 °C). Sample detection was performed by ultraviolet light according to the manufacturer's instructions. The results for reactions #1–8 in Table 3 are shown in the top panels of Figure 4a–4h.

[0186] LabChip analysis Fractions from the gel filtration column were analyzed on a LabChip GXII Touch instrument under non-reducing conditions to visualize the possible presence of IgG dimers and half-bodies in the resulting gel filtration fractions. LabChip analysis was performed using HT Protein Express chips (Perkin-Elmer Cat# 760499) and the Protein Clear HR Reagent Kit (Perkin-Elmer CL8960014). LabChip analysis conditions: Input samples and fractions with concentrations greater than 1 mg / mL were diluted to 1 mg / mL using PBS. In a PCR plate, 1 μL of sample was mixed with 7 μL of non-reducing sample buffer (containing 9 mM N-ethylmaleimide) and incubated at 70°C for 10 minutes.

[0187] Afterwards, 14 μl of HO was added, and the plate was centrifuged at 2800 x g for 2 minutes. Sample analysis was performed using a LabChip GXII Touch instrument, using the standard "HT Protein Express" script according to the manufacturer's instructions. The results for reactions 1–8 in Table 3 are shown in the lower panels of Figure 4A–4h. Samples collected after FAE primarily contain IgG dimers. Very little aggregates are observed. Asterisks in samples 1, 2, 5, and 6 indicate the IgG half-body peaks separated from the IgG dimers by gel filtration and LabChip analysis.

[0188] Fractions containing IgG dimers but no half-bodies were pooled, and the protein concentration of the pooled sample was measured using a Little Lunatic UV / Vis spectrophotometer according to the manufacturer's instructions (Unchained Labs). Absorbance was measured at 280 nm using the total protein program, with water as the blank and an extinction coefficient set to 1.45 mL / mg / cm.

[0189] Pooled samples from reactions #1–8 in Table 3 were analyzed again using the LabChip under non-reducing conditions as described above, as well as under reducing conditions using sample buffer containing 35 mM DL-dithiothreitol (Sigma cat# 43819). The results are shown in Figure 5 (non-reducing: upper panel, reducing: lower panel). The results from non-reducing conditions indicate that all IgG antibodies were at least 98% pure, with only trace amounts of IgG with a single LC. Therefore, FAE did not cause significant loss of light chains using these samples. The results from reducing conditions indicate that the bispecific antibodies from reactions #1–4 and #5–8 share the same mixture of light chains, indicating the presence of two distinct Fab arms in the IgG samples.

[0190] Additionally, "pre-GF control" samples from reactions 1-8, taken before and after FAE, were analyzed by reducing and non-reducing LabChip analysis according to the procedure described above. For all reactions, samples were verified to contain half-bodies before FAE, but samples after FAE showed fewer half-bodies (data not included).

[0191] Run 7: HP-CIEX analysis To confirm that the FAE reaction actually led to the production of bispecific antibodies, all samples (reaction products and unreacted controls) generated in Example 2 were analyzed by CIEX (Agilent 1260 series). The column (TSK gel SP-STAT 7 μm, 4.6 mm ID x 10 cm L; Tosoh cat# 21964) was equilibrated with a low ionic strength phosphate buffer (Buffer A) (25 mM sodium phosphate pH 6.0 ± 0.05, composed of sodium dihydrogen phosphate dihydrate (NaH2PO4, 2 H2O; Sigma, ref. 71500) and disodium hydrogen phosphate dihydrate (Na2HPO4, 2 H2O; Sigma, ref. 71643), dissolved in Milli-Q water and filtered through a 0.45 μm membrane filter). Next, the antibody was removed from the column by applying a gradient of increasing salt concentrations of phosphate buffer (Buffer B) (25 mM sodium phosphate, 1 M NaCl pH 6.0, sodium dihydrogen phosphate dihydrate (NaH2PO4, 2 H2O; Sigma, ref. 71500), disodium hydrogen phosphate dihydrate (Na2HPO4, 2 H2O; Sigma, ref. 71643), and NaCl (Sigma, ref. S3014), dissolved in Milli-Q water and filtered through a 0.45 μm membrane filter). The injected sample mass for all test samples and controls was 10 μg, and the injection volume ranged from 10 to 100 μL. Chromatograms were analyzed for peak pattern, retention time, and peak area of ​​the major peak observed at 220 nm.

[0192] The results in each bottom panel of Figure 6 show that FAE using the DE / KK mutations produced only approximately 1% homodimers (see reactions #1 and #5) or undetectable amounts of homodimers (see reactions #2 and #6). FAE using the 405L / 409R mutations produced approximately 5–13% homodimers (see reactions #3, #4, #7, and #8).

[0193] The results shown in the bottom panels of Figure 6a and Figure 6c show the appearance of peaks that were not present before FAE, accounting for approximately 87–89% or 95–96% of the material eluting from CIEX corresponding to the DEKK heterodimer (i.e., the peaks at approximately 21 min for reactions #1 and #2 and approximately 17 min for reactions #5 and #6).

[0194] The results, shown in the bottom panels of Figure 6b and Figure 6d, indicate the appearance of peaks not present prior to FAE, which account for approximately 77–84% or 83–88% of the material corresponding to the 405L / 409R heterodimer eluting from CIEX (i.e., the peaks at approximately 21 min for reactions #3 and #4, and approximately 16 min for #7 and #8). The asterisks in Figure 6a–d indicate peaks detected in each sample from #1 to #8, indicating that they arise independently of the DE / KK or 405L / 409R Fc format used. Based on their position relative to the main heterodimer peak, these are expected to be easily removed using standard separation techniques.

[0195] Taken together, these results demonstrate that using either the DE / KK or 405L / 409R substitutions, heterodimers can be produced on a larger scale than shown in Examples 1-5, and even with unequal ratios of starting material. Gel filtration-purified samples carrying the 405L / 409R substitution contained 5-13% unwanted homodimers, whereas samples carrying the DE / KK substitution more effectively removed contaminants, with a maximum reported level of homodimer of 1%. Higher purity was obtained using the DEKK-based substitution compared to the 405L / 409R substitution (87-96% vs. 77-88%, respectively).

[0196] We also report that the estimated total recovery of bispecific antibodies in milligrams of protein after the FAE reaction is comparable between the two substitution systems, but that samples subjected to FAE using the DE / KK substitution are more pure.

[0197] Notably, purified bispecific antibodies generated by FAE were confirmed to still bind to their cognate antigens by ELISA. Binding to other targets was not observed. All antibodies prepared by FAE exhibited specific binding to their targets, leading to the conclusion that antibody specificity is maintained after FAE despite reduction and reoxidation of the Fab arms. Furthermore, bispecific antibodies were found to bind similarly to each antigen, regardless of the DE:KK ratio selected before FAE.

[0198] array JPEG2026500710000004.jpg245159 JPEG2026500710000005.jpg242158 JPEG2026500710000006.jpg21162

Claims

1. 1. A method for producing a heterodimeric antibody comprising two different IgG CH3 domains capable of forming a CH3-CH3 interface, comprising: (a) a first antibody or half antibody comprising a first CH3 domain comprising positively charged amino acid residues at positions 351 and 366; and (b) a second antibody or half antibody comprising a second CH3 domain comprising negatively charged amino acid residues at positions 351 and 368; where the numbering follows EU numbering. providing incubating the antibodies and / or half antibodies of (a) and (b) together under reducing conditions to provide a reduced first antibody or half antibody and a reduced second antibody or half antibody; and re-oxidizing the reduced antibody and / or half antibody to obtain the heterodimeric antibody; wherein the IgG CH3 domain of the heterodimeric antibody does not comprise an arginine at position 409, and the heterodimeric antibody comprises an IgG Fc region comprising an IgG1 or IgG2 core hinge region CPPC.

2. 2. The method of claim 1, wherein the IgG CH3 domain is an IgG1 or IgG2 CH3 domain, and optionally the CH3 domain is a human CH3 domain.

3. The method of claim 1 , wherein the first antibody or half antibody and / or the second antibody or half antibody is a human antibody or half antibody.

4. The method of claim 3, wherein the antibody or half antibody is an IgG1 or IgG2 antibody or half antibody.

5. 2. The method of claim 1, wherein the first antibody or half antibody is an antibody having a first binding specificity and the second antibody or half antibody is an antibody having a second, different binding specificity.

6. The method of claim 1 , wherein the first antibody and / or the second antibody is a homodimeric antibody.

7. The method of claim 1 , wherein the first antibody and / or the second antibody is a heterodimeric antibody.

8. 8. The method of claim 7, wherein the heterodimeric antibody is multivalent, and optionally the multivalent antibody is a bivalent, trivalent, tetravalent, or up to hexavalent antibody.

9. 8. The method of claim 7, wherein the heterodimeric antibody is a multispecific antibody, optionally wherein the multispecific antibody is a bispecific antibody, a trispecific antibody, or a tetraspecific antibody.

10. The method of claim 7 , wherein the heterodimeric antibody comprises two non-identical light chains.

11. 2. The method of claim 1, wherein the first CH3 domain comprises 351K and 366R, 351R and 366K, 351K and 366K, or 351R and 366R.

12. 2. The method of claim 1, wherein the second CH3 domain comprises 351D and 368E, 351E and 368D, 351D and 368D, or 351E and 368E.

13. The first CH3 domain comprises 351K and 366K; and 2. The method of claim 1, wherein the second CH3 domain comprises 351D and 368E.

14. The method of claim 1, wherein the first and / or second antibodies or half antibodies of a) and b) are Fc engineered.

15. The method of claim 1, wherein the first and / or second antibodies or half antibodies of a) and b) are Fc-silenced or Fc-enhanced.

16. The method of claim 1, wherein the first and / or second antibody or half antibody of a) and b) comprises a CH2 domain having a mutation at positions 235 and / or 236 that affects ADCC.

17. The method of claim 1 , wherein the second antibody or half antibody is obtained independently from the first antibody or half antibody.

18. 10. The method of claim 1, wherein the reducing conditions are: (a) incubating the antibody or half-antibody in the presence of a reducing agent, optionally comprising or selected from the group consisting of 2-mercaptoethylamine (2-MEA), dithiothreitol (DTT), dithioerythritol (DTE), glutathione, tris(2-carboxyethyl)phosphine (TCEP), L-cysteine, beta-mercaptoethanol, thioglycolate, cysteamine, homocysteine, penicillamine, and sodium borohydride; and / or (b) incubating the antibody or half-antibody at a pH of 6.0 to 12.0; and / or (c) incubating the antibody or half-antibody at a redox potential of −150 to −600 mV, optionally the redox potential is −250 to −400 mV; A method comprising:

19. The method of claim 1, further comprising a step of concentrating and / or isolating the heterodimeric antibody obtained after reoxidation.

20. 20. The method of claim 19, wherein the heterodimeric antibody is concentrated and / or isolated using a method selected from the group comprising or consisting of precipitation, centrifugation, filtration, size exclusion chromatography, affinity chromatography, cation exchange chromatography and / or anion exchange chromatography, and hydrophobic interaction chromatography.

21. 1. An isolated heterodimeric antibody comprising a first IgG CH3 domain and a second IgG CH3 domain, the first CH3 domain and the second CH3 domain are capable of forming a CH3-CH3 interface; the first CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R; and the second CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D; the heterodimeric antibody further comprises two light chains having non-identical sequences; An isolated heterodimeric antibody, wherein the IgG CH3 domain of the heterodimeric antibody does not comprise an arginine at position 409, and the heterodimeric antibody comprises an IgG Fc region comprising an IgG1 or IgG2 core hinge region CPPC.

22. 22. The isolated heterodimeric antibody of claim 21, wherein the first CH3 domain comprises 351K and 366K, and the second CH3 domain comprises 351D and 368E.

23. A mixture of heterodimeric antibodies, the heterodimeric antibody comprises a first IgG CH3 domain and a second IgG CH3 domain; the first CH3 domain comprises one or more of the amino acid variants 366K, 366R, 351K, or 351R; and the second CH3 domain comprises one or more of the amino acid variants 351E, 351D, 368E, or 368D; the heterodimeric antibody further comprises two light chains having non-identical sequences; A mixture, wherein the IgG CH3 domain of the heterodimeric antibody does not contain an arginine at position 409, and the heterodimeric antibody comprises an IgG Fc region comprising an IgG1 or IgG2 core-hinge region CPPC.

24. 24. The mixture of claim 23, wherein the first CH3 domain comprises 351K and 366K and the second CH3 domain comprises 351D and 368E.

25. A pharmaceutical composition comprising the isolated heterodimeric antibody of any one of claims 21 to 22 and a pharmaceutically acceptable carrier.

26. The method of any one of claims 1 to 20, wherein the first antibody or half antibody and / or the second antibody or half antibody comprises an antibody binding domain, an scFv, a ligand, a protein receptor, or a cytokine.

27. The method of claim 26, wherein the heterodimeric antibody is a bifunctional or multifunctional fusion protein.

28. Use of the isolated heterodimeric antibody of any of claims 21 to 22 for the manufacture of a medicament.

Images