Bispecific antigen binding construct

Abstract

To provide multispecific, e.g., bispecific, antigen binding constructs, comprising a conventional antibody and a VHH antibody for improved binding to protein complexes.SOLUTION: A bispecific antigen binding construct comprises a single domain antibody (VHH) binding region which binds a first target antigen and is operatively linked to a first IgG Fc domain polypeptide.SELECTED DRAWING: None

Description

Technical Field

[0001] (Related Applications) This application claims the benefit of U.S. Provisional Patent Application No. 62 / 609,523, filed Dec. 22, 2017, the entire contents of which are incorporated herein by reference.

[0002] (Sequence Listing) This application includes a sequence listing submitted electronically in ASCII format, the entire contents of which are incorporated herein by reference. The name of the ASCII copy created on Dec. 21, 2018 is 60 7772_AGX5-031_ST25.txt, and the size is 59,420 bytes.

[0003] (Field of the Invention) The present invention relates to multi-specific, e.g., bispecific, antigen-binding constructs that include conventional antibodies and VHH antibodies with improved binding to protein complexes. The antigen-binding constructs can target at least two different antigens in a protein complex. The present invention further relates to multi-specific antigen-binding constructs that include conventional antibodies and VHH antibodies and have improved purification, e.g., by size exclusion chromatography (SEC).

Background Art

[0004] (Background of the Invention) Natural antibodies, including bivalent antibodies, exhibit immunoreactivity against specific epitopes on a particular target antigen. Multi-specific antibodies are antibodies that are engineered to potentially recognize and bind to multiple epitopes on different target antigens, as the name implies.

[0005] ​​​​​​Conventional natural antibodies typically contain a combination of immunoglobulin heavy and light chains, and here they divide The antigen-binding properties of this child are determined by the variable regions of the heavy and light chains, i.e., the VH and VL domains, respectively. Determined by the variant domain. More specifically, the antigen-binding site of conventional antibodies is typically The residues that receive contributions from three complementarity-determining regions (CDRs) within each of the VH and VL domains include.

[0006] Multispecific antibodies typically recognize and bind to multiple epitopes. In that it incorporates pairings of multiple VH-VL domains, it differs from natural antibodies in that it does so. These antibodies are commercially very important because they can bind to multiple target antigens. However, mispairs between different heavy and light chains incorporated into the same antibody molecule can occur. The rings present significant difficulties in the production and isolation of multispecific antibodies. Sparing results in monospecific antibodies or antibodies with non-functional or non-productive antigen-binding sites. There is a possibility that antibodies may be produced unintentionally, which could reduce the yield of the target multispecific antibody. It decreases.

[0007] Figure 1 shows the production of bispecific antibodies that exhibit immunoreactivity against two different epitopes. The difficulties that may arise are illustrated. The bispecific antibody shown in (A) has two different heavy chains and two different It contains light chains. However, only the proper pairing of these four immunoglobulin chains is necessary. However, it produces an antibody with the desired binding profile, i.e., an antibody with specificity for both target antigens. To cause this. In fact, nine other origins can be formed from the mixture of the four heavy and light chains shown. There are possible combinations, resulting in bivalent monospecific antibodies (E and H) and monovalent monospecific antibodies. Sexual antibodies (B, C, G, and J) and unbound antibodies (D, F, and I) are produced. This problem The more complex the multispecific antibody molecule, the more complex it is intended to bind to the antibody. The more epitopes or antigens there are, the more serious the condition becomes.

[0008] Improving the production of multispecific antibodies by addressing the problem of improper chain pairing. Various attempts are being made to achieve this. Some approaches involve manipulating antibodies to obtain VH-VL The focus is on facilitating correct pairing between mains. For example, US patent applications. Issue 2010 / 0254989 A1 involves genetically modifying the VH and VL of individual antibodies via a GlySer linker. The construction of a fused, bispecific cMet-ErbB1 antibody is described. Another approach is To produce bispecific antibodies, rat-mouse quadromas were used, and here Rat and rat antibodies primarily form the original VH-VL pairing, while bispecific antibodies are rat and It consists of mouse Fc (Lindhofer et al., J Immunol (1995) 155:1246-1252).

[0009] Furthermore, regarding bispecific antibodies containing an Fc domain, researchers have found that the constant region of the heavy chain undergoes a sudden change. The focus is on introducing heterodimerization to promote proper heterodimerization of the Fc portion. Such techniques have been reviewed in the literature by Klein et al. (mAbs (2012) 4:6, 1-11), The contents of this document are incorporated herein by reference in their entirety. These technologies include antibodies "Knob-in" involves introducing a bulky residue into one of the CH3 domains of the heavy chain. One approach is the "Two-Hole" (KiH) approach. This bulky residue is paired with the heavy chain. fits into complementary "holes" in the other CH3 domain, promoting proper pairing of the heavy chains to promote.

[0010] In addition, the researchers are also trying to solve the problem of properly associating pairs of heavy and light chains One approach is to use the CrossMab principle (reviewed in Klein et al.) which involves domain exchange between heavy and light chains to promote the formation of proper pairings (reviewed in Klein et al.). Other researchers have sought to enhance the affinity between a heavy chain and its cognate light chain by manipulating the contact surface between paired VH-VL domains or paired CH1-CL domains of the heavy and light chains (Lewis et al., Nature Biotechnology (2014)32:191-198 ). Techniques such as those described above that require extensive antibody engineering have met with some success; however, the production of antibodies with specific mutations is labor-intensive and can result in antibodies that are highly immunogenic in humans and / or have problems with loss of effector function . ). Another approach to the production of multispecific antibody preparations with appropriate antigen specificity is the development of methods to enrich antibodies with proper heavy-chain / light-chain pairing. For example, Spiess et al . describe a method for producing a MET-EGFR bispecific antibody from co-cultures of bacteria expressing two different half-antibodies (Nature Biotechnology (2013)31:753-758 ).

[0011] In addition, the constant region of at least one heavy chain of a bispecific antibody is mutated to increase the affinity ​​​​​​​​​​A method for altering the binding affinity of a substance, such as protein A, is described. This allows for a purification technique that utilizes differential bonding between two heavy chain affinity substances. Based on this, it becomes possible to isolate appropriately paired heavy chain heterodimers (US See National Patent Application Publication No. 2010 / 0331527 and WO 2013 / 136186. Based on differential combination Therefore, the constraints on the method for selecting an appropriate heavy chain heterodimerization are that these methods are appropriate for heavy chains - Antibodies that have light chain pairing are not selected, and these techniques typically involve shared or common light chains. This is applied to multispecific antibodies that possess the following properties.

[0013] International patent application PCT / EP2012 / 071866 (WO 2013-064701) concerns an anti-idiotype binding compound. Methods for isolating multispecific antibodies based on quality, particularly the use of anti-idiotype antibodies, are used. This addresses the problem of improper chain pairing. Anti-idiotype binding substances are used in two processes. It is used in a selective assay method. In this method, a first substance is used to produce a VH-VL dopamine specific to the first antigen. The antibody having pairing is captured, and then a second substance is used to target a second antigen specific to that antibody. It captures antibodies that also possess a second VH-VL domain pairing.

[0014] A drawback of this method is that the anti-idiotype binding substance used to isolate the antibody may not conform to its antigen. It must be specific to each multispecific antibody produced according to its binding profile. Therefore, the principle of the method described in PCT / EP2012 / 071866 is generally applicable to any multiplexing. Applicable to the isolation of specific antibodies, but the reagent, i.e., the anti-idiotype binding substance, The multispecific antibodies to be isolated must be produced according to specific VH-VL domain pairings. I don't know. [Overview of the Initiative]

[0015] (Summary of the invention) This invention relates to at least one conventional Fab-binding region and one single-domain antibody (VHH) binding. By providing a multispecific antigen-binding construct having a region, we can leverage cutting-edge technology. To improve it. This format, when combined with a heterodimerization method, forces one This process generates a bispecific antibody composition. The heterodimerization method used is forced Fab The heavy chain region and the complete VHH chain of the heavy chain are bound together. Since the VHH chain does not associate with the light chain, Fab A portion of the light chain region associates only with its corresponding heavy chain.

[0016] This invention allows for the purification of desired multispecific antigen binding from undesirable antibodies based on size. By providing constructs, such as bispecific antibodies, we can further improve cutting-edge technologies. The desired multispecific antigen-binding construct has a size of approximately 112 kDa, while the desired The size of non-existent antibodies is approximately 150 kDa and / or 75 kDa.

[0017] In a first embodiment, the present invention is a bispecific antigen-binding construct: (a) A first IgG Fc domain polypeptide that binds to the first target antigen and is operablely linked to it. A single-domain antibody (VHH) binding region; and (b) A second IgG Fc domain polypeptide that binds to a second target antigen and is operably linked to it. The first and second IgG Fc domain polypeptides contain the Fab portion of a conventional IgG antibody, and the two IgG Fc domain polypeptides are The bispecific antigen-binding construct is merged to form the bispecific antigen-binding construct. Provides an instruct.

[0018] In one embodiment, dimerization of the Fc domain is performed through knob-into-hole interaction. Fab arm exchange (FAE), electrostatic steering interaction, or hydrophobic phase It occurs through interaction.

[0019] In one embodiment, the first IgG Fc domain polypeptide comprises a knob substitution, and the second IgG Fc domain polypeptides include hole substitutions.

[0020] In one embodiment, the knob substitution is arginine (R), phenylalanine (F), tyrosine Selected from the group consisting of phenyl (Y), tryptophan (W), and any combination thereof. .

[0021] In one embodiment, the hole substitution is alanine (A), asparagine (N), asparagine Acids (D), glycine (G), serine (S), threonine (T), valine (V), and their It is selected from a group consisting of any combination.

[0022] In one embodiment, the bispecific antigen-binding construct is approximately 100 kDa to approximately 120 kDa It has a molecular weight within a range.

[0023] In one embodiment, the desired bispecific antigen-binding construct is approximately 112 kDa. .

[0024] In another embodiment, the present invention relates to a method for purifying a bispecific antigen-binding construct. hand: (a) A mixed antigen-binding construct composition containing antigen-binding constructs of different sizes. The process of providing goods; and (b) A step of separating mixed antigen-binding construct compositions based on size, The desired bispecific antigen-binding construct is (i) A first IgG Fc domain polypeptide that binds to the first target antigen and is operablely linked to it. A single-domain antibody (VHH) binding region; and (ii) A second IgG Fc domain polypeptide that binds to a second target antigen and is operablely linked to it. The Fab portion of a conventional IgG antibody is bonded, and the first and second IgG Fc domain polypeptides are The preceding step includes dimerizing to form the bispecific antigen-binding construct. A notation method is provided.

[0025] In one embodiment, size-based separation includes size exclusion chromatography.

[0026] In one embodiment, the mixed antigen-binding construct composition is first prepared with protein A-, p Purification is performed by rhotein G-, protein L-, or CH1-selective chromatography.

[0027] In another embodiment, the present invention provides a desired dual specificity in a mixture of other antigen-binding constructs A method for determining the amount of a sex antigen-binding construct: (a) A mixed antigen-binding construct composition containing antigen-binding constructs of different sizes. The process of providing goods; and (b) A step of separating mixed antigen-binding construct compositions based on size, The desired antigen-binding construct is (i) A first IgG Fc domain polypeptide that binds to the first target antigen and is operablely linked to it. A single-domain antibody (VHH) binding region; and (ii) A second IgG Fc domain polypeptide that binds to a second target antigen and is operablely linked to it. The Fab portion of a conventional IgG antibody is bonded, and the first and second IgG Fc domain polypeptides are The preceding step includes dimerizing to form the bispecific antigen-binding construct. A notation method is provided.

[0028] In one embodiment, size-based isolation includes gel electrophoresis.

[0029] In one embodiment, the desired bispecific antigen-binding construct is approximately 100 to approximately 120 kDa That is the case.

[0030] In one embodiment, the desired bispecific antigen-binding construct is approximately 112 kDa. .

[0031] In one embodiment, the other antigen-binding construct is approximately 75 kDa or approximately 150 kDa. [Brief explanation of the drawing]

[0032] (Brief explanation of the drawing) [Figure 1] This shows 10 different pairings of heavy and light chain combinations ((A)-(J)) that can result from the production of bispecific antibodies having two different heavy chains and two different light chains. [Figure 2] Figures 2A-2C show schematic diagrams of antigen-binding constructs. Figure 2A shows a schematic diagram of a conventional antibody. Figure 2B shows a schematic diagram of the VHH domain. Figure 2C shows a schematic diagram of the Fab domain / VHH bispecific antibody of the present invention. [Figure 3]A schematic diagram of the antigen-binding construct of a bispecific antibody having a knob-into-hole Fc mutation and an Fc-dead aglycosylation N297A mutation in both Fc portions is shown. The conventional Fab portion contains a hole mutation, and the VHH portion contains a knob T366W mutation. [Figure 4] The following shows the amino acid sequence alignments of representative IgG2 Fc regions. The sequences are as follows: mIgG2aHole, SEQ ID NO: 32; mIgG2aKnob, SEQ ID NO: 33; mFcFusionIgG2aHole, SEQ ID NO: 34; mFcFusionIgG2aKnob, SEQ ID NO: 35; and pUPEX36-mIgG2.P9, SEQ ID NO: 36. [Figure 5] Figures 5A-5B show SDS-PAGEs of antigen-binding constructs with Coomassie brilliant blue staining. Stained gels are shown for both non-reducible and reduced samples. Figure 5A shows SDS-PAGEs of antigen-binding constructs for targets A and B. Figure 5B shows SDS-PAGEs of antigen-binding constructs for targets C and D. [Figure 6] Figures 6A-6C show size exclusion chromatography (SEC) data for the antigen-binding constructs VHH2H3-mFc_Hole+4R36B7mIgG_Knob, VHH3H2-mFc_Hole+4R36B7mIgG_Knob, and VHH3H2-mFc_Hole alone, respectively. [Figure 7] Figures 7A-7D show the binding affinity of the antigen-binding construct to (Figure 7A) immobilized target antigen B, (Figure 7B) immobilized target A, (Figure 7C) PBS, and (Figure 7D) isotype control, as measured by Biacore (SPR). [Figure 8] Figures 8A-8C show the binding dynamics of the antigen-binding construct to (Figure 8A) target A, (Figure 8B) target B, and (Figure 8C) the target A-target B complex. [Modes for carrying out the invention]

[0033] (Detailed description of the invention) (A. Definition) “Antibody” or “Immunoglobulin” — The term “immunoglobulin” as used herein It has two heavy chains and two heavy chains, regardless of whether it has any associated specific immunoreactivity. Contains polypeptides having a combination of light chains. "Antibody" refers to the target antigen (for example, This refers to aggregates that exhibit significant known specific immune response activity against human antigens. The "specificity" of a substance for a human antigen does not exclude cross-reactivity with species homologs of that antigen. Antibodies and immunoglobulins contain light chains and heavy chains, and there are covalent interchain crosslinks between them. The absence of immunoglobulins is not a concern. The basic immunoglobulin structure of vertebrate systems is relatively well understood. .

[0034] The general term "immunoglobulin" refers to a group of five different classes of antibodies that are biochemically distinguishable. Yes. All five classes of antibodies are within the scope of this invention, and in the following discussion, immunoglobulins are generally used. This study focuses on the IgG class of immunoglobulin molecules. Regarding IgG, immunoglobulins have a molecular weight of approximately Two identical light chain polypeptides with a molecular weight of 23,000 daltons and two identical light chain polypeptides with molecular weights of 53,000 to 70,000. It contains heavy chains. The four chains are linked together in a Y-shape by disulfide bonds. In this configuration, the light chain starts from a Y-shaped opening and supports the heavy chain that follows, passing through a variable region (b (racket).

[0035] The light chain of an antibody is classified as kappa or lambda (κ, λ). The heavy chain of each class is kappa or It can bind to either of the λ light chains. Generally, the light and heavy chains are covalently bonded to each other, and immunoglobulins Robulin is transmitted by hybridomas, B cells, or genetically modified host cells. When produced, the "tail" portions of the two heavy chains are either covalent disulfide bonds or non-covalent bonds. They are bonded to each other by bonding. In heavy chains, the amino acid sequence is at the N-terminus of the branched end of a Y-shaped configuration. It runs to the C-terminus at the bottom of each chain. As those skilled in the art will know, the heavy chain consists of gamma, muon, and alpha. They are classified into γ, μ, α, δ, ε, and several between them. It will be recognized that there are subclasses of antibodies (e.g., γ1-γ4). This is the property of this chain that determines whether it is IgG, IgM, IgA IgG, or IgE, respectively. The subclasses (isotypes) of brin (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, etc.) are These classes are known to be well-characterized and to be given functional specialization. Modified forms of the isotype can be easily recognized by those skilled in the art in light of the brief disclosure, and therefore This falls within the scope of the present invention.

[0036] As described above, the variable region of the antibody allows the antibody to selectively recognize epitopes on the antigen. Furthermore, it becomes possible to bind specifically to the VL domain and VH domain of the antibody. The ions are combined to form a variable region that defines the three-dimensional antigen-binding site. This quadruple antibody The structure forms antigen-binding sites located at the ends of each Y-shaped arm. More specifically, anti The primordial binding site is defined by three complementarity-determining regions (CDRs) on each VH and VL chain.

[0037] "Isolated antibody" - As used herein, "isolated antibody" refers to an antibody that has been isolated from its natural environment. These are antibodies that have been and / or recovered. The impurities of the natural environment are the diagnostic or Substances that interfere with therapeutic use, including enzymes, hormones, and other proteinaceous or non-proteinaceous substances. It may contain protein components. This is because at least one component of the antibody's natural environment is absent. The isolated antibody contains antibodies that are present in their original location (in situ) within the recombinant cell. However, Typically, isolated antibodies are prepared in at least one purification step.

[0038] "Affinity variant"—The term "affinity variant" as used herein refers to the following: A variant antibody refers to an antibody that shows one or more changes in the amino acid sequence compared to the original antibody. Ant shows a change in affinity for the protein compared to the reference antibody. Typically, affinity Sex variants show improved protein affinity compared to the reference antibody. This improvement is due to... Decreased KD, increased off-rate, or altered cross-reaction pattern with non-human homologs of the protein. It can be modified. Affinity variants typically differ in the amino acid sequence of the CDR compared to the reference antibody. It shows one or more changes. Such substitutions result in the original ami present at a given position in the CDR. The amino acid may be a different amino acid, either a naturally occurring amino acid residue or a non-naturally occurring amino acid residue. Substances can be substituted. Amino acid substitutions can be conservative or non-conservative.

[0039] "Binding site" - As used herein, the term "binding site" refers to the target antigen ( For example, it includes a polypeptide region responsible for selective binding to human antigens. The binding domain is It contains at least one binding site. An example of a binding domain is an antibody variable domain. The antibody molecule of the present invention may contain multiple (for example, two, three, or four) binding sites.

[0040] "Derived from camelids" - In one preferred embodiment, the antigen-binding construct of the present invention The ct is derived from conventional camelid antibodies produced by active immunization of camelids. This includes a framework amino acid sequence and / or a CDR amino acid sequence. However, The antibody of the present invention, which contains an amino acid sequence derived from a camelid, contains a human amino acid sequence (i.e., Framework and / or constant region derived from human antibodies or other non-camelid mammal species. The sequence may be manipulated to include, for example, human or non-human primate framework regions, heavy chains. The portion and / or hinge portion may be included in the antibody. In one embodiment, 1 The non-camelid amino acids listed above may be present in the framework region of "camelid-derived" antibodies. For example, the amino acid sequence of a camelid animal framework corresponds to that of a human or non-human primate. It may contain one or more amino acid mutations that result in the presence of amino acid residues. Furthermore, it may originate from camelid animals. The original VH and VL domains, or their humanized variants, are conjugated to the constant domain of a human antibody. This allows for the creation of chimeric molecules.

[0041] "Conservative amino acid substitution" - A "conservative amino acid substitution" is the substitution of amino acid residues with similar side chains. This involves substituting amino acid residues that have a similar side chain. Milly is defined in the art as a family that includes basic side chains (e.g., lysine). (arginine, histidine), acidic side chain (e.g., aspartic acid, glutamic acid), non Charged side chains (e.g., glycine, asparagine, glutamine, serine, threonine, cyanoacrylate) (Leucine, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine) (Proline, phenylalanine, methionine, tryptophan), β-branched side chain (for example) (e.g., threonine, valine, isoleucine), and aromatic side chains (e.g., tyrosine, phenyl (Alanine, tryptophan, histidine) are present. Therefore, immunoglobulin polypeptides The non-essential amino acid residues within can be substituted with other amino acid residues from the same side-chain family. can.

[0042] "Epitope" - The term "epitope" refers to a peptide to which an antibody or antibody fragment binds. Alternatively, it refers to a specific sequence of amino acids located on a protein. An epitope is an amino acid or Often composed of chemically active surface molecules such as sugar side chains, and possessing specific three-dimensional structural properties. It has specific charge properties. The epitopes are linear, i.e., a single sequence of amino acids and Involved in or conforming to the binding of antigens, i.e., not necessarily contiguous. It can be involved in binding to two or more amino acid sequences in various regions.

[0043] "Heavy chain portion"—as used herein, the term "heavy chain portion" refers to the immunoglobulin heavy chain. Contains amino acid sequences derived from the constant domain. Polypeptides containing the heavy chain portion are:CH1 domain , hinge (e.g., upper, middle, and / or lower hinge regions) domain, CH2 domain, CH 3 domains, or at least one of their variants or fragments. In one embodiment, In this invention, the binding molecule is the Fc portion of the immunoglobulin heavy chain (for example, the hinge portion, CH2 The domain may include the CH3 domain. In another embodiment, the binding molecule of the present invention It lacks at least a portion of its constant domain (for example, all or part of the CH2 domain). In one embodiment, at least one, preferably all, of the constant domains are human immune It originates from the globulin heavy chain. For example, in one preferred embodiment, the heavy chain portion is completely hydroxylated. It contains a tohinge domain. In another preferred embodiment, the heavy chain portion is a complete human Fc portion ( For example, it includes hinge, CH2, and CH3 domain sequences derived from human immunoglobulin.

[0044] In one embodiment, the constituent constant domains of the heavy chain portion are different immunoglobulin molecules. It originates from. For example, the heavy chain portion of a polypeptide is derived from the CH2 domain and IgG1 molecule. It may include a hinge region derived from 3 or IgG4 molecules. In other embodiments, a constant domain It is a chimeric domain containing parts of different immunoglobulin molecules. For example, the hinge is IgG1 It may include a first part derived from a molecule and a second part derived from an IgG3 or IgG4 molecule. As described above, Those skilled in the art will know that the constant domain of the heavy chain portion is derived from a natural (wild-type) immunoglobulin molecule. It should be understood that the amino acid sequence can be modified to change. That is, The polypeptides of the present invention disclosed herein have one or more heavy chain constant domains (CH1, hin This may include changes or modifications to the (CH2, or CH3) and / or the light chain constant domain (CL). Exemplary modifications include the addition, deletion, or substitution of one or more amino acids in one or more domains. ru.

[0045] "Variable region" or "variable domain" - The term "variable" refers to the variable domains VH and VL. A specific part of the sequence differs widely among antibodies, and each specific antibody against its target antigen. This refers to the fact that it is used for binding and specificity. However, variability is the variable nature of the antibody. It is not uniformly distributed throughout the main body. The variability forms the antigen-binding site. In each of the VL and VH domains, there are three sequences called "hypervariable loops" The focus is on the ligment. The first, second, and third hypervariable loops of the Vλ light chain domain are described in this specification. In this text, these are referred to as L1(λ), ​​L2(λ), and L3(λ), and the residues in the VL domain are 24-33(L1( λ), consisting of 9, 10, or 11 amino acid residues, 49-53 (L2(λ), consisting of 3 residues), and It can be defined as containing 90-96 (L3(λ), consisting of 5 residues) (Morea et al., Method 20:267-279(2000)). The first, second, and third hypervariable loops of the Vκ light chain domain are... In the specification, these are referred to as L1(κ), L2(κ), and L3(κ), and residues 25-33(L) in the VL domain. (consisting of 1(κ), 6, 7, 8, 11, 12, or 13 residues), 49-53(consisting of L2(κ), 3 residues) It can be defined as including , and 90-97 (L3(κ), consisting of 6 residues) (Morea et al. (Reference, Method 20:267-279(2000)). The first, second, and third hypervariable loops of the VH domain are revealed. In detail, these are referred to as H1, H2, and H3, and consist of residues 25-33 (H1, 7, 8, or 9) in the VH domain. ), 52-56 (H2, consisting of 3 or 4 residues), and 91-105 (H3, highly variable in length) It can be defined as including (Morea et al., Method 20:267-279(2000)).

[0046] Unless otherwise indicated, the terms L1, L2, and L3 refer to the first, second, and third of the VL domains, respectively. This refers to a hypervariable loop of type 3, encompassing hypervariable loops obtained from both Vκ and Vλ isotypes. The terms H1, H2, and H3 refer to the first, second, and third hypervariable loops of the VH domain, respectively. Ultra-transferable cells obtained from any known heavy-chain isotype containing γ, ε, δ, α, or μ It includes variable loops.

[0047] The highly variable loops L1, L2, L3, H1, H2, and H3 each have a "complementarity determination region" as defined below. In other words, it may include the "CDR" portion. The terms are not strictly synonymous. The hypervariable loop (HV) is defined based on its structure. In contrast, the complementarity determination region (CDR) is defined based on sequence variability (Kabat et al.). Literature, Sequences of Proteins of Immunology Interest), 5th edition, Public Health Service, National Institutes of Health, Bethes da, MD., 1983), and the constraints of HV and CDR may differ in some VH and VL domains. That is the reason.

[0048] The CDRs of the VL and VH domains are typically defined as containing the following amino acids. Possible: Residues 24-34 (CDRL1), 50-56 (CDRL2), and 89-97 (CDRL3) in the light chain variable domain. ), as well as residues 31-35 or 31-35b (CDRH1), 50-65 (CDRH2) in the heavy chain variable domain, and 95-102(CDRH3)(Kabat et al., "Sequences of Proteins Targeted in Immunology", 5th edition, P Public Health Service, National Institutes of Health, Bethesda, MD. (1991)). subordinate HV is included in the corresponding CDR, and unless otherwise indicated, VH and VL are used herein. The main reference to the "super variable loop" also includes the corresponding CDR, and vice versa. This is how it should be interpreted.

[0049] The more highly preserved portion of the mutable domain is defined as the framework area (FR) below. ) are called. The native heavy chain and light chain variable domains each consist of four FRs (FR1, FR2, etc.). It employs a β-sheet configuration mainly connected by three highly variable loops, including FR3 and FR4. The hypervariable loops of each chain are held together in close proximity by FR, and the hypervariable loops of the other chains are separated. Along with the progenitor, it contributes to the formation of the antigen-binding site of the antibody. Complementarity is determined by the structural analysis of the antibody. The relationship between the arrangement and shape of the binding sites formed by the region was clarified (Chothia References by et al., J. Mol Biol. 227:799-817 (1992); References by Tramontano et al., J. Mol. Biol, 21 5:175-182 (1990))). Despite the high variability of that sequence, 5 out of 6 loops This refers to the main chain conformation, which has a repertoire of only a few structures called "canonical structures". These conformations are adopted, firstly by the length of the loop. Secondly, it is determined that its packing, hydrogen bonding, or abnormal back chain conformation is assumed. A specific position in a loop and framework domain that determines conformation through capability. It is determined by the presence of key residues in that position.

[0050] "CDR"—as used herein, the terms "CDR" or "complementarity-determining region" refer to heavy chains and This refers to discontinuous antigen-binding sites found in both variable regions of the light chain polypeptide. The specific areas mentioned are Kabat et al.'s literature, J. Biol. Chem. 252, 6609-6616 (1977), and Kabat et al. The literature, "Sequences of Proteins Targeted in Immunology" (1991), and the literature by Chothia et al., J. Mo l. Biol. 196:901-917 (1987) and the literature by MacCallum et al., J. Mol. Biol. 262:732-745 (19 It is described in 96). Comparing these documents with each other, there are overlapping amino acid residues in the definition. It contains a group or a subset of amino acid residues. It encompasses the CDR as defined in each of the above references. The amino acid residues contained are listed for comparison. Preferably, the term "CDR" refers to the sequence. This is a CDR defined by Kabat based on comparison.

[0051] Table 1: Definition of CDR [Table 1] 1 The residue numbering follows the nomenclature used in the literature by Kabat et al., as described above. 2 The residue numbering follows the nomenclature used in the literature by Chothia et al., as described above. 3 The residue numbering follows the nomenclature used in the literature by MacCallum et al., as shown above.

[0052] The amino acid residues of the VHH domain are the general VH domain as given by Kabat et al. Numbering will be done according to the standard numbering system. VH domains and VHH domains are well known in the art. As indicated, the total number of amino acid residues in each CDR may vary, depending on the Kabat numbering system. The total number of amino acid residues indicated may not match (i.e., 1 according to Kabat numbering). The above positions may not be occupied in the actual array, or the actual array may contain Kabat (This may mean that the numbering may contain more amino acid residues than the number permitted by the numbering system.) Please note that, generally speaking, Kabat numbering is based on the amino acid residues in the actual sequence. This means that it may or may not correspond to the original actual numbering. VHH For details on domain numbering, see the United States, incorporated by reference herein. It is listed in issue 8,703,131.

[0053] "Framework domain" - as used herein, "framework domain" or "FR domain" The term refers to a portion of the variable region that includes amino acid residues that are not part of the CDR (for example) (If using the Kabat definition for CDR). Therefore, the length of the variable region framework is approximately 100-120. It contains amino acids, but only those located outside the CDR. Specific examples of heavy chain variable regions and Kab Regarding the CDR as defined in the literature by at et al., framework region 1 encompasses amino acids 1-30. The variable region corresponds to the domain of the variable region; framework region 2 is a variable region encompassing amino acids 36-49. Corresponding to the domain of the region; framework region 3 is a variable region encompassing amino acids 66-94. It corresponds to the main region, and framework region 4 is variable from amino acid 103 to the end of the variable region. It corresponds to the variable domain. The framework domain of the light chain is determined by each light chain variable domain CDR. They are similarly classified. Similarly, the definition of CDR according to Chothia et al. or McCallum et al. Using this, the boundaries of the framework area are defined by the end of each CDR as described above. It is classified. In a preferred embodiment, the CDR is defined by Kabat.

[0054] In natural antibodies, the six CDRs present on each monomeric antibody allow the antibody to maintain its three-dimensional structure in an aqueous environment. When exhibiting its function, it takes on a specific configuration to form an antigen-binding site, consisting of a short, discontinuous amino acid sequence. Yes. The remaining parts of the heavily variable domain and lightly variable domain are shown in the amino acid sequence. The variability between child elements is smaller, and these are designated as framework domains. Generally, it adopts a β-sheet conformation, and the CDR forms a loop, β-sheet structure It connects to and, in some cases, forms part of it. Therefore, these frameworks The crust region provides the proper orientation of the six CDRs through non-covalent interactions between the chains. It acts to form a scaffold. Antigens are formed by the placed CDRs. The binding site defines a surface complementary to the epitope on the immunoreactive antigen. The surface promotes non-covalent binding of antibodies to immunoreactive antigen epitopes. The location can be easily identified by someone skilled in the art.

[0055] "Hinge region" - As used herein, the term "hinge region" refers to a CH1 domain being connected to CH2 It contains the heavy chain molecule portion that binds to the domain. This hinge region contains approximately 25 residues, and Due to its flexibility, the two N-terminal antigen-binding regions can move independently. (Hinge region) The region consists of three distinct domains: the upper, middle, and lower hinge domains (Roux et al., J. It can be subdivided into Immunol 161:4083 (1998). Anti-"complete human" hinge region including The body may include one of the hinge region arrangements shown in Table 2 below.

[0056] Table 2: Human hinge arrangement [Table 2]

[0057] "Fragment" - The term "fragment" refers to a smaller portion of an intact or complete antibody or antibody chain. This refers to an antibody or part or portion of an antibody chain that contains an amino acid residue that is not present. It is also called an "antigen-binding fragment." The term refers to binding to an antigen, or for antigen binding (i.e., specific binding to a human antigen). Immunoglobulins compete with intact antibodies (i.e., with the intact antibodies from which they originate). This refers to a polypeptide fragment of a robulin or antibody. The “fragment” of an antibody molecule as used herein and The term refers to the antigen-binding fragment of an antibody, such as the antibody light chain (VL), antibody heavy chain (VH), and single-chain antibody. ScFv), F(ab')2 fragment, Fab fragment, Fd fragment, Fv fragment, and single-domain antibody fragment (DAb) This includes the fragments, for example, the chemical or complete antibodies or antibody chains. It can be obtained by enzymatic treatment or recombinant means.

[0058] "Specificity" - The term "specificity" refers to the ability to identify a given target antigen, such as a human target antigen, specifically. This refers to the ability to bind to an antibody or its antigenic binding (for example, immunoreactivity with the target antigen). The composite fragment is single-specific and contains one or more binding sites that specifically bind to the target, or The antibody or its antigen-binding fragment is multispecific and specific to the same or different targets. It may contain two or more binding sites that bind to the target. In one embodiment, the antigen-binding site of the present invention The instruct is specific to multiple target antigens. In one embodiment, the antigenic ligation of the present invention The compound construct is specific to two target antigens. In one embodiment, the anti- The protoconjugate construct is specific to multiple human target antigens. In one embodiment, The antigen-binding construct of the invention is specific to two human target antigens. For example, one embodiment In this case, the bispecific binding molecule of the present invention binds target antigen A (Target A) and target antigen B (Target B). It binds to the target antibody. As another example, in one embodiment, the bispecific binding molecule of the present invention is a target antibody It binds to proto-C (target C) and target antigen D (target D).

[0059] "Synthesis"—As used herein, the term "synthesis" in relation to polypeptides is derived from naturally occurring It includes polypeptides containing amino acid sequences that do not appear in natural polypeptides. For example, the form of a natural polypeptide. Modified (including mutations such as addition, substitution, or deletion), or amino acids In the linear sequence, a second amino acid sequence that does not naturally bind due to its properties (in the case of nature) The first amino acid sequence that binds to (even in the natural case, or not) Polypeptides that do not occur naturally, including (though they may not be present).

[0060] "Manipulated"—as used herein, the term "engineered" refers to a composite means (for example) If, by recombinant technology, in vitro peptide synthesis, enzymatic or chemical synthesis of peptides nucleic acids or polymers (by plucking, or by a combination of some of these techniques) This involves the manipulation of cydoplasmic molecules. Preferably, the antigen-binding construct of the present invention is, for example, humanized. Antibodies and / or chimeric antibodies, as well as antigen-binding properties, stability / half-life, or effector function. The antibody is modified to include one or more of the following properties:

[0061] "Fc region" - As used herein, the term "Fc region" refers to the Fc domain of its two heavy chains. This refers to the portion of the native immunoglobulin formed by the nucleotide. The native Fc region is It is a homodimer. The term “variant Fc region” as used herein refers to the native region. This refers to an Fc region that has one or more modifications to the Fc region. The Fc region is an amino acid substitution, addition, and / or modified by deletion, addition, and / or modification of the native glycan. To obtain. This term refers to a heterodimer Fc, where each constituent Fc domain contains a different Fc region. Examples of the domain include, but are not limited to, the "knob" described in U.S. Patent No. 8,216,805. There is an Fc region fabricated using the "into-hole" technique. Furthermore, this term constitutes... For example, the Fc domain is found in US Patent Application Publication No. 2009 / 0252729 A1 and US 2011 / 0081345 A1. It includes a single-chain Fc region that is bound together by the linker components described below. (See separate reference) Unless otherwise specified, the numbering of all antibody constant regions follows the literature of Edelman et al., Proc. Natl. Acad. Sci. This refers to the EU numbering scheme described in USA 63(1):78-85 (1969).

[0062] "Humanized substitution"—as used herein, the term "humanized substitution" refers to the VH or VL domain. Refer to amino acid residues located at specific positions within an antibody (e.g., antibodies derived from camelids). This refers to amino acid substitutions that involve substitution with an amino acid residue appearing at an equivalent position in the VH or VL domain. The reference human VH or VL domain is a VH or VL domain encoded by the human germline. It may be an in. Humanization substitution is the framework region and / or C of the antibody as defined herein. It can be added during DR.

[0063] "Humanized antibody or variant" - as used herein, "humanized antibody" or "humanized variant" The term "ant" refers to a variant antibody that contains one or more "humanization substitutions" compared to the reference antibody. Refers to a body, and a portion of the reference antibody (e.g., VH domain and / or VL domain or at least 1 including two CDRs, a part thereof) has amino acids derived from non-human species, and "humanized replacement" appears in the amino acid sequence derived from non-human species.

[0064] "Heavy-chain only antibody" or "VHH antibody" - As used herein, the term "heavy-chain only antibody" or "V HH antibody" refers to a second type of antibody produced only by camelid species including camels, llamas, alpacas. A heavy-chain only antibody is composed of two heavy chains and lacks a light chain. Each heavy chain has a variable domain at the N-terminus, and these variable domains are referred to as "VHH" domains to distinguish them from the heavy-chain variable domains of conventional heterotetrameric antibodies, i.e., the VH domains described above.

[0065] "Modified antibody" - As used herein, the term "modified antibody" refers to a synthetic form of an antibody modified in a way that does not occur naturally, e.g., an antibody that contains at least two heavy-chain portions but does not contain two complete heavy-chain portions (e.g., domain-deleted antibody or minibody); a multispecific form of an antibody modified to bind to two or more different antigens or different epitopes on a single antigen (e.g., bispecific, trispecific, etc.); a heavy-chain molecule bound to a ScFv molecule, etc. ScFv molecules are known in the art and are described, for example, in U.S. Patent No. 5,892,019. Furthermore, the term "modified antibody" includes multivalent forms of antibodies (e.g., trivalent, tetravalent, etc., antibodies that bind to three or more copies of the same antigen). In another embodiment, the modified antibody of the present invention is a polypeptide that does not have a CH2 domain and contains a binding portion of one member of a receptor-ligand pair. It is a fusion protein containing at least one heavy chain moiety, including a ptide-binding domain.

[0066] Furthermore, the term "modified antibody" used herein refers to the amino acid sequence of an antibody. It may also refer to the antibody. A person skilled in the art can modify the antibody and compare it to the antibody from which it originated. It will be understood that this allows for the production of variant antibodies in which the amino acid sequence is altered. For example, a nucleotide or an amino acid residue that causes a conservative substitution or modification at a "non-essential" amino acid residue. Amino acid substitutions can be made (e.g., CDRs and / or framework residues). Amino acid substitutions may include substitution of one or more amino acids with natural or unnatural amino acids. .

[0067] "Target A"—as used herein, the terms "Target A" or "Target Antigen A" refer to the present invention. This refers to the target antigen to which the conventional Fab portion of an antibody specifically binds.

[0068] "Target B"—as used herein, the terms "Target B" or "Target Antigen B" refer to the present invention. This refers to the target antigen to which the VHH-containing portion of the antibody specifically binds. Target A and Target B are combined. It can form a body. In one embodiment, the present invention has specificity for target A and target B. The bispecific antibody can also bind to the complex of target A and target B.

[0069] "Target C"—as used herein, the terms "Target C" or "Target Antigen C" refer to the present invention. This refers to the target antigen to which the conventional Fab portion of an antibody specifically binds.

[0070] "Target D"—as used herein, the terms "Target D" or "Target Antigen D" refer to the present invention. This refers to the target antigen to which the VHH-containing portion of the antibody specifically binds.

[0071] "Antigen-binding construct" - The term "antigen-binding construct" as used herein refers to The term refers to a single-domain antibody (VHH) binding region fused to the IgG Fc domain; conventional IgG It includes the heavy chain Fc domain portion of the antibody and the light chain portion of a conventional IgG antibody. The VHH binding region is single It is an antibody fragment consisting of a monomeric variable antibody domain. The VHH binding region is a specific first antigen and It binds selectively. The heavy chain Fc domain portion of conventional IgG antibodies is variable and fixed. Includes the constant domain. The light chain portion of conventional IgG antibodies contains the variable and constant domains of the conventional IgG light chain. This includes the heavy and light chain portions of conventional IgG antibodies, which selectively bind to a specific second antigen.

[0072] (B. Multispecific antigen-binding construct) One component of the multispecific antigen-binding construct of the present invention is the heavy chain of a conventional antibody. and light chains, or antigen-binding fragments thereof, wherein this specification, "conventional antibodies" and Using the term, immunoglobulin heavy chains and arranged according to the "Y-shaped" configuration shown in Figure 1 This describes heterotetrameric antibodies containing light chains. Conventional antibodies of this type are not limited to these. However, llamas, mice, rats, rabbits, goats, hamsters, chickens, monkeys, and These may be derived from any suitable species, including human-derived antibodies.

[0073] In one exemplary embodiment, a conventional antibody is described herein by reference in its entirety. As described in incorporated U.S. Patent No. 8,524,231 and as stated in its claims This is a SIMPLE antibody. SIMPLE antibodies have a heavy chain variable domain (VH) and a light chain variable domain. It may include (VL), where VH and / or VL domains or one or more complementary thereof. The sex-determining region (CDR) originates from camelid animals; that is, camelid animals (e.g., It is derived from conventional antibodies produced by immunization of llamas. It shows high homology with humans. or having a CDR sequence, VH and / or VL domain derived from at least one camelid animal. The antibody or antigen-binding fragment is derived from the VH or VL domain of a conventional antibody from a camelid animal. It may be a humanized or germline variant. Here, "humanized" and "germline variant" The term "..." is defined elsewhere in this specification.

[0074] In one embodiment, the conventional antibody-antigen binding region is called "Fab" (antigen-binding fragment). This is possible. Fab extracts one constant domain and one variable domain from each heavy and light chain. Includes. The variable heavy and light chains contain CDRs responsible for antigen binding. The Fab portion of a conventional antibody is shown in Figure 2A. It can also be found in the schematic diagram of 2C.

[0075] Another component of the multispecific antigen-binding construct of the present invention is a VHH domain or a VHH anti It contains heavy chains of the body or nanobody. VHH antibodies, which are heavy chain antibodies derived from camelid animals, are two It consists of heavy chains and lacks light chains (Hamers-Casterman et al., Nature. 1993; 363; 446). -8). Each heavy chain of the VHH antibody has a variable domain at the N-terminus, and these variable domains are conventional In order to distinguish it from the variable domain of the antibody heavy chain, i.e., the VH domain, in this technical field, This is called the "VHH" domain. Similar to conventional antibodies, the VHH domain of the molecule contributes to antigen binding specificity. To this end, the isolated VHH domain or other VHH antibody or fragment is used as the antigen-binding conjugate of the present disclosure. It is suitable as a component of a tract.

[0076] The multispecific antigen-binding constructs of this disclosure each have their own selective target antigens. Specific conventional heterotetrameric antibodies or VHH antibodies have a polypeptide containing their antigen. Antibodies can be produced or obtained by active immunization of a host species. The host species are: mice, rats, rabbits, goats, hamsters, chickens, monkeys, Alternatively, one of the camelid species may be selected. For the production of VHH antibodies, the llama species may be selected. Immunizing any species from the camelid family with polypeptides containing their respective antigens It is possible.

[0077] An exemplary antigen-binding construct of the present invention is shown below.

[0078] Table 3: Antigen-binding construct sequences for target A / target B antibodies. HCDR1-H sequences assigned by Kabat. CDR3 and LCDR1-LCDR3 are shown in bold. [Table 3] TIFF2023139025000004.tif244170TIFF2023139025000005.tif244170TIFF2023139025000006.tif119170

[0079] Table 4: Antigen-binding construct sequences for target C / target D antibodies. [Table 4]

[0080] (Dimerization of C. VHH and conventional antibodies) This invention addresses the problem of chain mispairing present in conventional bispecific antibody formats. This overcomes the limitations. For example, with conventional bispecific antibody chain pairing, up to 10 different antibodies can be paired. Body types can be generated (Figure 1). For the 10 different possible options, the desired bispecific antibody Body yields are generally low, making it difficult to isolate desired bispecific antibodies from populations of other antibodies. This becomes difficult. In some cases, heterodimerization methods are used to force a specific desired configuration. This can partially overcome this problem. For example, the following Hete The rodimerization method forces the interaction of specific Fc domains, resulting in one of four possible configurations. Take one (Figure 1(A)~(D)). Unfortunately, however, these heterodimerization methods depend However, random pairing of light chains is possible, and it is a bispecific antibody with a single configuration. The uniform generation of is hindered.

[0081] To overcome the shortcomings in this field, the multispecific antigen-binding compound of the present invention The struct contains a single VHH-binding region that binds to the first target antigen. The VHH binding region is a conventional I-binding domain that binds to the first IgG Fc domain polypeptide and the second target antigen. The Fab portion of the IgG antibody is operably linked. The Fab portion is a second IgG Fc domain polypept The IgG Fc domain polypeptides are operably linked to the cytoplasm and dimerize. It forms a bispecific antigen-binding construct (Figure 2C). The light chain portion of Fab is the heavy chain portion of Fab. It can only be paired with the chain portion. Therefore, this allows for the desired bispecific configuration. It is guaranteed that you will gain something.

[0082] In one aspect of the present invention, the two Fc domains of the antigen-binding construct are knob-int Heterodimerization occurs through u-hole pairing. In this dimerization technique, CH3 domain The protrusions (knobs) and recesses (holes) provided in the contact surface of the unit are utilized. Suitable positions and The knob or hole used for measurement is located on the contact surface of the first or second CH3 domain, and is required. This simply involves providing holes or knobs corresponding to the adjacent contact surfaces. The pairing of the Fc domain is promoted and enhanced at the contact surface of the CH3 / CH3 domain. Fusion to VHH. The IgG Fc domain of the modified antibody is provided with a knob, while the IgG Fc domain of the conventional antibody is provided with a knob. A hole is provided that is designed to accommodate, and vice versa. The "knob" is the first Fc door At least one amino acid side chain protruding from the contact surface of the main CH3 moiety, typically relatively It refers to a large side chain. The projection is complementary to and receives the "hole" in the CH3 portion of the second Fc domain. It creates a "knob" that is accepted. The "hole" is drawn from the contact surface of the CH3 portion of the second Fc domain. It contains at least one amino acid side chain, typically a relatively small side chain. The technology is cited, for example, in U.S. Patent No. 5,821,333, the literature by Ridgway et al., and Protein Engineering 9:617- This is described in 621 (1996), and in the literature of Carter P, J. Immunol Methods 248:7-15 (2001). ru.

[0083] Examples of amino acid residues that can act as knobs include arginine(R) and phenylalanine. There are ammonium compounds (F), tyrosine (Y), and tryptophan (W). Existing amino acids in the CH3 domain The no-acid residue can be exchanged or substituted with an amino acid residue of the knob. The amino acids that are most beneficial include alanine (A), asparagine (N), aspartic acid (D), and glycan. These have small side chains, such as syn (G), serine (S), threonine (T), or valine (V). It may contain any amino acid.

[0084] Examples of amino acid residues that can function as holes include alanine (A), serine (S), and thresholds. There are onin (T) and valine (V). The amino acid residues present in the CH3 domain are H It can be replaced or substituted with the amino acid residue of . Preferred amino acid to be substituted This includes any amino acid having a large or bulky side chain, such as arginine(R), fe It may contain nyalanin (F), tyrosine (Y), or tryptophan (W).

[0085] The CH3 domain is preferably derived from a human IgG1 antibody. An example of an analogy for the CH3 domain is... Mino acid substitutions include T366Y, T366W, F405A, F405W, Y407T, Y407A, Y407V, T394S, and others. These are some of the combinations. A preferred exemplary combination is a knob mutation on the first CH3 domain. Regarding the difference, T366Y or T366W, and regarding the hole mutation on the second CH3 domain, Y407T or It is Y407V.

[0086] In one embodiment of the present invention, the two Fc domains of the antigen-binding construct are Fab arms Heterodimerization occurs through exchange (FAE). Human IgG1 with the P228S hinge mutation is F It may contain 405L or K409R CH3 domain mutations. When the two antibodies are mixed with a reducing agent, FAE is This technology is based on U.S. Patent No. 9,212,230 and the literature of Labrijn AF, Proc Natl Aca This is described in d Sci USA 110(13):5145-5150 (2013).

[0087] In one embodiment of the present invention, the two Fc domains of the antigen-binding construct are electrostatic Heterodimation occurs through electrostatic induction. This dimerization technique utilizes electrostatic induction. This promotes and enhances Fc domain pairing at the CH3 / CH3 domain contact surface. By altering the charge complementarity between domains, heterodimerization (same-charge pairing) is preferred over homodimerization (same-charge pairing). This promotes dimerization (opposite charge pairing). In this method, electrostatic repulsion leads to homodimerization. This prevents the substitution of amino acid residues, including K409D, K392D, and in the first CH3 domain. / or K370D, D399K, E356K, and / or E357K in the second CH3 domain may be included. This technology This refers to U.S. Patent Application Publication No. 2014 / 0154254 A1 and the literature of Gunasekaran K., J Biol Chem 285 ( 25) It is described in 19637-19646 (2010).

[0088] In one embodiment of the present invention, the two Fc domains of the antigen-binding construct are hydrophobic. Heterodimerization occurs through sex interaction effects. In this dimerization technique, electrostatic interactions Alternatively, hydrophobic interactions are used for Fc domain pairing at the CH3 / CH3 domain contact surface. To promote and enhance. An example of amino acid residue substitution is K409 in the first CH3 domain. W, K360E, Q347E, Y349S, and / or S354C, D399V, F405T, Q347R in the second CH3 domain This may include E357W and / or Y349C. Examples of amino acid residual substitution pairs include K409W:D399V, K409W:F405T, K360E:Q347R, and Y349 S:E357W and S354C:Y349C are included. This technology is described in U.S. Patent Application Publication No. 2015 / 0307628 A1. It is described there.

[0089] In one embodiment of the present invention, heterodimerization is achieved through the use of a leucine zipper fusion. Obtained. The leucine zipper domain fused to the C-terminus of each CH3 domain in the antibody chain is forcibly This technique induces heterodimerization. This technique is described in the literature by Wranik B., J Biol Chem 287(52): 43331- It is described in 43339 (2012).

[0090] In an embodiment of the present invention, heterodimerization is performed by the strand exchange operation domain (SEED) body This can be done through the use of IgG and IgA. The CH3 domain derived from IgG and IgA formats is forcibly heated. This technique involves ronidimerization. This technique is described in the literature by M. Muda, Protein Eng. Des. Sel. 24(5):447-45. It is described in 4 (2011).

[0091] Unless otherwise stated, all antibody constant region numbering used herein is by Edelma. n et al.'s paper, Proc. Natl. Acad. Sci. USA 63(1):78-85 (1969), describes the EU numbering system. Compatible with Keym.

[0092] Further methods for heterodimerization of heavy and / or light chains and for the production and purification of asymmetric antibodies are described below. , which are publicly known in the art. For example, each of them is incorporated herein by reference. See also Klein C., mABs 4(6) 653-663 (2012), and U.S. Patent No. 9,499,634. .

[0093] (D. Modifications not based on heterodimerization of the Fc region) The antibody molecule of the present invention has one or more amino acids in the constant regions of the heavy chain and / or light chain, particularly in the Fc region. It may have acid substitutions, insertions, or deletions. Amino acid substitutions involve different substituted amino acids from natural ones. Replace with amino acids or unnatural or modified amino acids. Also, for example, glycosylated patterns Changes in the molecule (for example, due to the addition or deletion of N or O bonded glycosylation sites), Other structural modifications are also permitted.

[0094] The bispecific antigen-binding construct of the present invention has binding affinity to the neonatal Fc receptor (FcRn). To increase acidity, modifications can be made to the Fc region. Acidic pH (for example, approximately The increased binding affinity can be measured at pH 5.5 to approximately pH 6.0. Affinity can be measured at a neutral pH (for example, approximately pH 6.9 to approximately pH 7.4). "Increased binding affinity" means increased binding affinity to FcRn for unmodified Fc regions. Typically, the unmodified Fc region is the wild-type amino acid sequence of human IgG1, IgG2, IgG3, or IgG4. In such embodiments, the increased FcRn nodes of the antibody molecule with modified Fc region are present. Binding affinity is measured against the binding affinity of FcRn to wild-type IgG1, IgG2, IgG3, or IgG4. ru.

[0095] Several Fc substitutions have been reported that increase binding to FcRn, thereby improving antibody pharmacokinetics. Such substitutions are, for example, incorporated herein by reference in their entirety. This is included in the literature by Zalevsky et al. (2010) Nat. Biotechnol. 28(2):157-9; Hinton et al. References (2006) J Immunol.176:346-356; Yeung et al. (2009) J Immunol.182:7663-7671; P References of resta LG. (2008) Curr.Opin.Immunol.20:460-470; and references of Vaccaro et al. (2005) N This is reported in at. Biotechnol. 23(10):1283-88.

[0096] In one embodiment, one or more bispecific antigen-binding constructs described herein The antibody molecule contains amino acid substitutions H433K and N434F, or modifications comprising these substitutions. Includes human IgG Fc domains. Here, the Fc domain numbering follows EU numbering. Further implementation In this embodiment, one or more antibody molecules of the combination described herein are amino acid substitutions M252Y, S2 Modified human IgG Fc domain containing or comprising 54T, T256E, H433K, N434F It includes the 'n'. Fc domain numbering follows EU numbering.

[0097] Furthermore, bispecific antigen-binding constructs can be used with cytotoxic substances, such as chemotherapeutic agents and toxins. (For example, enzyme-active toxins of bacterial, fungal, plant or animal origin, or fragments thereof), and It contains antibodies conjugated with radioactive isotopes (i.e., radioactive conjugates). It can be modified to form an immunoconjugate. Furthermore, the Fc region is as shown by reference. The Chan and Carter reference (2010) included in the detailed text: Nature Reviews: Immunology As described in 10:301-316, it can be manipulated to extend the half-life.

[0098] In one embodiment, the Fc region is manipulated to create a device that does not have an effector function, for example If it has the ability to induce antibody-dependent cytotoxicity (ADCC) or complement-dependent cytotoxicity (CDC), It can be made so that it does not exist. The Fc region that does not have such manipulated effects is In this specification, this is referred to as "Fc dead". In one embodiment, the bispecific antigen binding of the present invention The construct is a natural IgG isotype with reduced effector function, such as IgG4. It may have an Fc region derived from it. The Fc region derived from IgG4 is, for example, between IgG4 molecules in vivo. By introducing modifications that minimize arm replacement, further enhancing therapeutic utility Modification is possible. The Fc region derived from IgG4 can be modified to include an S228P substitution.

[0099] In one embodiment, the bispecific antigen-binding construct of the present invention is glycosylated It will be corrected in relation to. For example, an aglycosylated antibody (i.e., one that has glycosylation) It is possible to produce antibodies (that do not exist) by changing glycosylation, for example, against a target antigen. Such glycosylation can increase the affinity of antibodies. This can be achieved by changing one or more glycosylation sites. For example, one or more Performing the above amino acid substitution eliminates one or more variable region framework glycosylation sites. This eliminates glycosylation at that site. Sialylation can increase the affinity of an antibody for an antigen. In certain embodiments it is possible to engineer the Fc region to reduce or eliminate glycosylation. For example, mutating the N297 site in the CH2 domain can reduce or eliminate glycosylation (the numbering of the Fc domain follows the EU numbering). In certain embodiments, the bispecific antigen-binding construct of the invention contains one or more N297A mutations. The N297A mutation has been shown to reduce antibody-dependent cellular cytotoxicity (ADCC). See, for example, the publication by Borrok et al., (2012) ACS Chem Biol. 7(9):1596-1602.

[0100] (E. Polynucleotide encoding a multispecific antigen-binding construct) In addition, the invention provides a polynucleotide molecule encoding the multispecific antigen-binding construct of the invention, operably linked to regulatory sequences that enable the expression of the multispecific antigen-binding construct polypeptide in a host cell system or cell-free expression system, as well as an expression vector comprising the nucleotide sequence encoding the multispecific antigen-binding construct of the invention. The invention also provides a host

[0101] cell or cell-free expression system comprising these expression vectors. The polynucleotide molecules encoding the multispecific antigen-binding construct of this invention include, for example, recombinant DNA molecules. As used herein, the terms "nucleic acid", "polynucleotide", When discussing, the sequence or structure of a specific nucleic acid molecule provides the sequence in the 5' to 3' direction It may be described according to common convention. In some embodiments of the invention, the nucleic acid or polynucleotide is "isolated". When applied to a nucleic acid molecule, this term refers to a nucleic acid molecule separated from other components. This includes the sequences immediately adjacent in the natural genome of the organism from which it is derived. For example, an "isolated nucleic acid" can include a DNA molecule inserted into a vector such as a plasmid or viral vector, or integrated into the genomic DNA of a prokaryotic cell, eukaryotic cell, or non-human host organism. When applied to RNA, the term "isolated polynucleotide" mainly refers to an RNA molecule encoded by an isolated DNA molecule as previously defined. Alternatively, this term may refer to an RNA molecule purified / separated from other nucleic acids that are associated in its natural state (i.e., in a cell or tissue). An isolated polynucleotide (DNA or RNA) can further be a molecule produced directly by biological or synthetic means and separated from other components present during its production. including the sequences immediately adjacent in the natural genome of the organism from which it is derived. For example, an "isolated nucleic acid" can include a DNA molecule inserted into a vector such as a plasmid or viral vector, or integrated into the genomic DNA of a prokaryotic cell, eukaryotic cell, or non-human host organism. When applied to RNA, the term "isolated polynucleotide" mainly refers to an RNA molecule encoded by an isolated DNA molecule as previously defined. Alternatively, this term may refer to an RNA molecule purified / separated from other nucleic acids that are associated in its natural state (i.e., in a cell or tissue). An isolated polynucleotide (DNA or RNA) can further be a molecule produced directly by biological or synthetic means and separated from other components present during its production. For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture ー such as a plasmid or viral vector, or integrated into the genomic DNA of a prokaryotic cell, eukaryotic cell, or non-human host organism. When applied to RNA, the term "isolated polynucleotide" mainly refers to an RNA molecule encoded by an isolated DNA molecule as previously defined. Alternatively, this term may refer to an RNA molecule purified / separated from other nucleic acids that are associated in its natural state (i.e., in a cell or tissue). An isolated polynucleotide (DNA or RNA) can further be a molecule produced directly by biological or synthetic means and separated from other components present during its production. For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture

[0102] For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture For recombinant production of the multispecific antigen-binding constructs according to the invention, recombinant polynucleotides encoding various construct components are prepared (using standard molecular biology techniques) and inserted into a replicable vector for expression in a selected host cell or cell-free expression system. Suitable host cells can be prokaryotes, yeast, or higher eukaryotic cells, specifically mammalian cells. Examples of useful mammalian host cell lines are the monkey kidney CV1 line transformed by SV40 (COS-7, ATCC CRL 1651); the human embryonic kidney line (in suspension culture Subcloned 293 or 293 cells for proliferation, Graham et al., J. Gen. Viro l. 36:59 (1977)); Baby hamster kidney cells (BHK, ATCC CCL 10); Chinese hamster Ovarian cells / -DHFR (CHO, Urlaub et al., Proc. Natl. Acad. Sci. USA 77:4216 (198) 0)); Mouse Sertoli cells (TM4, mother, Biol. Reprod. 23:243-251 (1980)); Mouse myeloma Cell SP2 / 0-AG14 (ATCC CRL 1581; ATCC CRL 8287) or NS0 (HPA culture cell line preservation institution number No. 85110503); Monkey kidney cells (CV1 ATCC CCL 70); African green monkey kidney cells (VERO-76, ATC) C CRL-1587); human cervical cancer cells (HELA, ATCC CCL 2); canine kidney cells (MDCK, ATCC CCL 34); Buffalo rat hepatocytes (BRL 3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL 75); Human hepatocytes (Hep G2, HB 8065); Mouse mammary carcinoma (MMT 060562, ATCC CCL51); TRI cells (Mather et al., Annals NYAcad.Sci.383:44-68 (1982)); MRC 5 cells; FS4 cells; and human liver These are cell tumor lines (Hep G2) and DSM's PERC-6 cell line. In each of these host cells... Expression vectors suitable for this use are also generally known in the art.

[0103] It should be noted that the term "host cell" generally refers to a cultured cell line. Any human being into whom an expression vector encoding a primordial polypeptide has been introduced is a "host cell Explicitly exclude it from the definition of "".

[0104] (F. Therapeutic utility of multispecific antigen-binding constructs) As used herein, the terms “to treat,” “treating,” or “treatment” refer to the disease. The intention is to delay, interfere with, suppress, manage, halt, or reduce the severity of a condition, disability, state, or disease. However, this does not necessarily mean that all disease-related symptoms, conditions, or disorders can be completely eliminated. Not included.

[0105] For therapeutic use in humans, the multispecific antigen-binding construct described herein It can be administered to human controls that require treatment in an "effective dose". The term refers to a sufficient amount or dose of a drug to achieve the desired result. In some embodiments, The term "effective dose" refers to the amount administered to a subject in a single or multiple dose for the treatment of a disease. This refers to the amount of drug that produces therapeutic effect. In one embodiment, the "effective dose" is defined as The term refers to the therapeutic effect in treating a disease when a single or multiple dose is administered to a human patient. It refers to the amount of a drug that produces an effect. For example, the term "effective dose" refers to the amount of a drug that has a disease. When administered as a single or multiple dose to human patients, it produces therapeutic efficacy in the treatment of the disease. This may refer to the amount of the multispecific antigen-binding construct of the present invention.

[0106] The effective therapeutic dose of the multispecific antigen-binding construct is approximately 0.1 mg / kg body weight per single dose. Contains an amount in the range of approximately 20 mg / kg body weight. The therapeutically effective dose of tract is approximately 1 mg / kg body weight to approximately 10 mg / kg body weight per single dose. This may include: The effective therapeutic dose for any individual patient depends on the patient's age and overall condition, and It can be determined according to factors such as the nature and severity of the disease to be treated. For any individual The therapeutically effective amount for a separate patient can be determined by a medical professional, for example, by monitoring the effect of the antigen-binding construct on a biomarker, such as the cell surface expression of the target antigen in tumor tissue, or symptoms such as tumor regression. The amount of antigen-binding construct administered at any given time point can be adjusted so that the optimal amount of antigen-binding construct is administered during the course of treatment, whether used alone or in combination with any other therapeutic agent.

[0107] In addition, the antigen-binding constructs described herein or pharmaceutical compositions containing such constructs are intended to be administered in combination with any other treatment as a combination therapy.

[0108] (G. Pharmaceutical Composition) The scope of the present invention includes pharmaceutical compositions formulated with the antigen-binding constructs of the present invention together with one or more pharmaceutically acceptable monomers or excipients. Techniques for formulating antigen-binding constructs such as monoclonal antibodies for therapeutic use in humans are well known in the art and are reviewed, for example, in the literature by Wang et al., J Pharm Sci 96: 1-26 (2007).

[0109] The pharmaceutical composition can be formulated for administration by any suitable route of administration, including, without limitation, intravenous (i.v.), intraperitoneal (i.p.), intramuscular (i.m.), intratumoral, oral, other enteral, subcutaneous (s.c.), and pulmonary.

[0110] (H. Purification of Multispecific Antigen-Binding Constructs)​​​​​​​​ The scope of the present invention includes methods for purifying the multispecific antigen-binding construct of the present invention. Purification methods may include methods based on physicochemical fractionation. Such methods may include size Exclusion chromatography (SEC), ammonium sulfate precipitation, ion exchange chromatography - (IEC), immobilized metal chelate chromatography (IMAC) is one of the purification methods. This may include purity purification. Such a method may involve the conservation region of the antigen-binding construct. Protein A, Protein G, and / or Protein L as affinity ligands. It depends on the affinity ligand. The affinity ligand is conjugated into the resin to facilitate purification. Further affinity purification methods can improve the specificity of antibodies to the IgG CH1 domain. This can depend on specific bonding. This is a CH1-selective chromatography method.

[0111] The multispecific binding constructs disclosed herein can be purified by SEC. A VHH binding region operably linked to the Fab portion of a conventional IgG antibody (e.g., SIMPLE antibody) The size of the multispecific binding construct of the present disclosure, including the body, is expected to be approximately 112 kDa. It is formed by the dimerization of two Fab moieties, each operably linked to an Fc domain. Conventional IgG antibodies are predicted to be approximately 150 kDa. Two functionally linked Fc domains are present. The VHH antibody formed by the dimerization of the VHH portion is expected to have a magnitude of approximately 75 kDa. Isomerized construct (~112 kDa) and two possible undesirable alternative binding constructs Considering the size difference between tracts (~150 kDa and / or 75 kDa), a person skilled in the art would know The desired multispecific binding construct is removed from the undesirable binding construct by SEC. It will be readily apparent that it can be easily purified.

[0112] Whether the resulting multispecific antigen-binding construct was stabilized by crosslinking. Regardless, the method of the present invention, in some embodiments, involves multispecific antigen binding The process may include a step to further purify the construct. It may also include a multispecific antigen-binding construct. The mixture can be analyzed using standard chromatography techniques, for example (but not limited to) standard Quasi-protein A chromatography, protein G, protein L, cation / anion exchange Conversion chromatography, size exclusion chromatography, hydrophobic interaction chromatography By using thiophyllite chromatography, or by setting up to capture IgG molecules Purification using the measured ligand (protein A mimetic, llama VHH ligand, etc.) It is possible. Alternatively, the multispecific antigen-binding construct can undergo salting-in. Add ed precipitation (ammonium sulfate), organic solvent (DMSO, ethanol), pH change Using standard techniques such as precipitation or nonionic polymers (polyethylene glycol), It can be made to work. In another setting, a multispecific antigen-binding construct can be used with a membrane. It can be applied to filtration techniques to concentrate multispecific antigen-binding constructs. The substance still contains the parent's monospecific IgG molecule and the multispecific antigen-binding construct. Therefore, combining all these technologies is possible for a multispecific antigen-binding construct. This may be necessary to purify and homogenize the substance. Subsequently, multispecific antigen-binding substances Additional purification steps may be required to separate the tract from the parent's single-specific IgG molecule. This involves using an affinity column for first binding specificity and for binding and elution. For subsequent binding and elution using an affinity column for second binding specificity. This can be achieved through further purification.

[0113] The quantity, quality, and purity of the (purified) multispecific antigen-binding construct are determined by absorbance measurement. Routine biochemical techniques such as HP-SEC, SDS-PAGE, native PAGE, and RP-HPLC It can be analyzed using the following technique: multispecific antigen-binding constructs from parent IgG molecules. Additional technologies that can identify these include, but are not limited to, IEF, cIEF, and CIEX. There are also mass spectrometry (ESI, MALDI), which analyze molecules based on their charge and / or mass. High-precision separation and detection become possible. The two bonds of the double-bond specific antigen-binding construct Specificity can be measured using methods such as ELISA, RIA, and surface plasma resonance (SPR). Various different methods using biolayer interferometry, DELFIA, FRET, ECL, Gyros, and AlfaScreen The evaluation can be performed using any of the following binding assay formats. The purification method It is described in U.S. Patent No. 9,212,230.

[0114] (Integrated by reference) Various patents, published patent applications, and publications have precedented the descriptions and the entirety of the following embodiments. It is cited throughout, and each of them is incorporated herein in its entirety by reference. ru. [Examples]

[0115] (Examples) The present invention can be further understood by referring to the following non-limiting embodiments.

[0116] (Example 1 - Preparation and purification of anti-target A / anti-target B antigen-binding constructs) Anti-target A is a conventional antibody that has binding specificity to target A. Anti-target A is mouse C It can neutralize the activation of target A in D4+ spleen cells. Anti-target B can neutralize target B. This is a VHH with binding specificity. For binding to target A, target B, or the target A-target B complex. The bispecific antigen-binding constructs of anti-target A and anti-target B are knob-into-hole ( Generated using KIH technology. Anti-target B VHH was fused to mIgG2α_hole. Anti-target A-VH was fused to mIg The G2a_knob was cloned, and the anti-target A-VL was cloned as CLVλ.

[0117] To produce the anti-target A / anti-target B antigen binding construct, the following was added to 10 × 100 ml of 293E cells: Constructs: mFcFusionIgG2aKnob_Anti-Target B-VHH, mIgG2aHole_Anti-Target A _VH and CLVλ_Anti-Target A were transfected at 20 μg each per 100 ml. gG2aHole, mIgG2aKnob, mFcFusion IgG2aHole, and mFcFusionIgG2aKnob construct It was created by cloning into pUPE (pCDNA3.1-like vector) using KpnI-NotI. Ant i-Target A_VH was cloned into the pUPE-mIgG2aHole vector using BsmBI. Anti-Ta rget B-VHH was cloned into the pUPE-mFcFusionIgG2aKnob vector using BsmBI. The sample was purified using protein A beads. The bound antigen-binding construct was then analyzed using SDS-PAGE. Elution and analysis were performed using the following method. Figure 5A shows the SDS-PAGE gel stained with Coomassie blue. The diagram shows that the purified antigen-binding construct is of the appropriate size of 112 kDa. Note that the molecular marker moved faster than the sample. Antigen-binding structure The IgG had a lower band compared to complete IgG. In a normal scenario, mispearl The compound produced a 150 kDa complete IgG, a 75 kDa VHH-Fc fusion, and a bispecificity folic acid of approximately 112 kDa. - A mat is formed. However, in Figure 5A, only a 112 kDa product was produced. This indicates that the desired bi-singularity format was preferentially produced. Under reducing conditions, two bands were observed for the two heavy chains. mFcFusionIgG2aKnob_A nti-Target B-VHH does not contain a CH1 domain, lacks a light chain, and yields a product of approximately 37 kDa. Please note the following. The heavy chain of mIgG2aHole_Anti-Target A-VH is expected to be approximately 50 kDa. The light chain of target A is approximately 25 kDa.

[0118] Therefore, the bispecificity format of the present invention yields only the appropriate bispecificity antibodies. This guarantees that, in particular, only the mIgG2aHole_Anti-Target A_VH construct is compatible with mFcFusionIgG2. aKnob_Anti-Target pairs with the B-VHH construct. The light chain pairs with the VHH chain. Therefore, only the remaining light chain CLVλ_Anti-Target A construct is used with the Anti-Target A_VH chain. Pairing is possible. To those skilled in the art, the present invention is not limited to the above-mentioned KIH heterodimerization method. It will soon become clear that this is not determined. Any suitable heterogeneity described elsewhere in this specification The dimerization method can be used with the above-described bispecific antibody format.

[0119] (Example 2 - Characterization of the Anti-Target A / Anti-Target B Antigen-Binding Construct) The antigen-binding construct of Example 1 was examined for its binding properties to target A and target B, and B was analyzed by Bia. The analysis was performed using core(SPR). This construct was able to capture both target A and target B. (Compare Figures 7A and 7B with the PBS and isotype controls in Figures 7C and 7D.)

[0120] The affinity of the antigen-binding construct for the target A-target B complex is such that it is effective against either target A or target B. The affinity is much higher compared to other factors, indicating that it is a divalent bond (Figures 8A-8C and Table 5). Since the detection limit of Biacore is + / -1E-12, the affinity shown in Figure 8 is an overestimation. Please be mindful.

[0121] Table 5: Antigen binding conjugates of target A-target B complex, target A, and target B by Biacore Affinity measurement of ct. [Table 5]

[0122] The divalent interaction of antigen-binding constructs, as demonstrated in Figures 8A-8C and Table 5, is Instead of the tract binding to target B and target A within the same complex molecule, it binds to the target of one of the complexes. This may be the result of binding to target B and target A of the adjacent complex. Antigen-binding construct To determine whether they bind in the same complex molecule, the Biacore CM5 chip is used for antigen binding. The construct was coated. The target A-target B complex, target A, and target B were placed on the chip. The antigen-binding construct was conjugated (Table 6). The affinity with target B was determined by whether target B was immobilized and the antigen-binding construct was conjugated. The affinity was similar to that shown when the molecules were bound. Since target A is a dimer, target A is fixed. It bound to the modified antigen-binding construct with relatively high affinity. This allowed target A to bind to the antigen. Divalent interactions with the chip coated with the original bonding construct become possible. The affinity of the body to the immobilized antigen-binding construct is the immobilized antigen of target B. The fact that it is clearly higher than the affinity for the binding construct suggests that the antigen-binding construct This suggests that lacto binds within a single complex molecule.

[0123] Table 6: Affinity measurement of the target A-target B complex, target A, and target B using Biacore. [Table 6]

[0124] (Example 3 - Preparation and purification of target C / target D antigen-binding constructs) VHH2H3 and VHH3H2 are different VHH antibodies with binding specificity to target D. 4R36B7 This is a conventional antibody that has binding specificity to target C. It binds to both target C and target D. For the bispecific antigen-binding constructs VHH2H3 or VHH3H2 and 4R36B7, knob-in It was fabricated using the KIH (Knowledge-Inducing Hole) technology. VHH2H3 and VHH3H2 were fused with mIgG2a_hole. 4R36B7-VH was cloned as mIgG2a_knob, and 4R36B7-VL was cloned as CLVλ. I did it.

[0125] Target C / target D antigen-binding constructs were prepared using 293E cells. The sample was protein A beads. The antigen-binding construct was purified using [method / method]. The bound antigen-binding construct was eluted and analyzed using SDS-PAGE. Figure 5B shows an SDS-PAGE gel stained with Coomassie blue, and the antigen-binding construct The diagram shows that the purified ct is of an appropriate size of approximately 120 kDa. The molecular weight marker is Note that it migrated faster compared to the sample. The antigen-binding construct is complete Ig It has a lower band compared to G. In a normal scenario, due to mispairing, 150 kDa A complete IgG, a 75 kDa VHH-Fc fusion, and a bispecific format of approximately 120 kDa are formed. However, Figure 5B shows that only a 120 kDa product was produced, and this suggests that This indicates that the desired bispecific format was preferentially produced.

[0126] Therefore, the bispecificity format of the present invention yields only the appropriate bispecificity antibodies. This guarantees that, in particular, only the 4R36B7 mIgG2a Fc dead_Knob construct is VHH2H3-mFc dea Pairs with the d_Hole or VHH3H2-mFc dead_Hole construct. The light chain pairs with the VHH chain. Since ringing is not possible, only the remaining light chain CLVλ_4R36B7 construct remains as 4R36B7 mIgG2a Fc dea It can be paired with a d-heavy chain. For those skilled in the art, the present invention is the above KIH heterodimerization method. It will soon become clear that this is not limited to law. Any appropriate provision described elsewhere in this specification may apply. The heterodimerization method can be used with the bispecific antibody format described above.

[0127] (Example 4 - Size exclusion chromatography (SEC) of antigen-binding constructs) The chromatography system used was the Agilent 1260 Infinity II, with a four-component pump. Automatic injector, cooled autosampler (6°C), online degassing device, and DAD detector. It is equipped with an output device. The column is Waters XBridge BEH 200 Å SEC (3.5 μm, 7.8 × 300 mm, Wa Waters XBridge BEH 200 Å SEC Preguard Column (3) (catalog number 176003596) The sample was bonded to a 0.5 μm, 7.8 × 30 mm Waters column (catalog number 176003591). This is first equilibrated with a 10-column volume (CV) mobile phase, which is a thermostat-controlled compartment It was not retained in the ment. The operating flow rate was set to 1 mL / min for 30 minutes, during which time PBS was used as the mobile phase. It was used as follows (Sigma, catalog number D8537). The detector was set to wavelengths of 280 and 214 nm simultaneously. (Reference wavelength 360 nm, cutoff 100 nm). Subsequently, it was found that it had 3.5% aggregation. Aggregation monitoring was performed at a 214 nm channel in the presence of a highly agglomerated control sample (HAC). The samples that had been monitored several times were continued to be used as QC samples (taken along). Data was acquired using Chemstation OpenLab CDC software (Agilent).

[0128] Figures 6A-6C and Table 7 below show the results of SEC analysis for bispecific antibodies targeting C / targeting D. vinegar.

[0129] Table 7: SEC results for target C / target D antibodies. [Table 7]

[0130] (Example 5 - Characterization of Target C / Target D Antigen-Binding Construct) The affinity of the target C / target D binding construct was measured using Biacore 3000. Using HS Chemistry, we manufactured CM5 chips coated with target C or target D of approximately 500 RU. The antibody concentration range was added to the chip (9 concentrations, starting from 10 μg / ml). The molecular weights were determined to be as follows: mIgG ~ 150 kDa; mIgG-VHH-Fc ~ 120 kDa; VHH-Fc dimer The body was 80 kDa. Affinity was determined using BiaEvaluation software, 1:1 langum The parameters were determined by fitting them to a coupled model. The results are as follows: This is shown in Table 8. We found that both antibody arms were functional.

[0131] Table 8: Affinity measurement of target C and target D by Biacore. "NB" means no binding. [Table 8]

Claims

[Claim 1] The novel products, methods and processes substantially as herein described.

Images