Engineered immunoglobulin heavy chain-light chain pairs and uses thereof

Abstract

The invention provides heterodimer pairs that may include a first heterodimer and a second heterodimer, wherein each heterodimer includes an immunoglobulin heavy chain or a fragment thereof and an immunoglobulin light chain or a fragment thereof. At least one of the heterodimers may include one or more amino acid modifications in the CH and / or CL domains, and one or more amino acid modifications or combinations thereof in the VH and / or VL domains. The modified amino acid may be part of the interface between the light chain and the heavy chain, and is usually modified to produce a preferential pairing between each heavy chain and the desired light chain, so that when two heavy chains and two light chains of the heterodimer are co-expressed in a cell, the heavy chain of the first heterodimer preferentially pairs with one chain of the light chain rather than another chain. Similarly, the heavy chain of the second heterodimer may preferentially pair with the second light chain rather than the first light chain.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Application No. 61 / 730,906, filed November 28, 2012, U.S. Provisional Application No. 61 / 761,641, filed February 6, 2013, U.S. Provisional Application No. 61 / 818,874, filed May 2, 2013, and U.S. Provisional Application No. 61 / 869,200, filed August 23, 2013, each of which is hereby incorporated by reference in its entirety for all purposes.

[0003] Sequence Listing

[0004] This application contains a sequence listing that has been submitted through EFS-net and is hereby incorporated by reference in its entirety. The ASCII copy (created on X / X / 2013) is named XXX_sequencelisting.txt and is XXX bytes in size.

[0005] background

[0006] Bispecific antibodies are capable of binding to two different epitopes. The epitope may be on the same antigen, or each epitope may be on a different antigen. This characteristic of bispecific antibodies makes them attractive tools for various therapeutic applications, where there is a therapeutic benefit of targeting or recruiting more than one molecule in the treatment of a disease. One of the methods for forming bispecific antibodies may include the accompanying expression of two unique antibody heavy chains and two unique antibody light chains. Correctly forming bispecific antibodies in a form similar to the wild type remains a challenge, because the antibody heavy chain has evolved to bind the antibody light chain in a relatively promiscuous manner. As a result of this promiscuous pairing, the accompanying expression of two antibody heavy chains and two antibody light chains naturally results in a scrambled heavy chain-light chain pairing. This mismatching remains a major challenge for the generation of bispecific therapeutic agents, where uniform pairing is a basic requirement for good manufacturability and bioefficacy.

[0007] Several methods for preparing bispecific antibodies in which a specific antibody light chain or fragment is paired with a specific antibody heavy chain or fragment have been described. A review of various approaches to this problem can be found in Klein et al., (2012) mAbs 4:6, 1-11. International Patent Application No. PCT / EP2011 / 056388 (WO 2011 / 131746) describes an in vitro method for generating heterodimeric proteins in which asymmetric mutations are introduced into the CH3 regions of two monospecific starting proteins to drive directed "Fab-arm" or "half-molecule" exchange between two monospecific IgG4- or IgG4-like antibodies after incubation under reducing conditions.

[0008] Schaefer et al. (Roche Diagnostics GmbH) described a method for assembling two heavy chains and two light chains from two existing antibodies into a human bivalent bispecific IgG antibody without the use of an artificial linker (PNAS (2011) 108 (27): 11187-11192). The method involves exchanging the heavy and light chain domains within one half of the antigen-binding fragment (Fab) of the bispecific antibody.

[0009] Strop et al. (Rinat-Pfizer Inc.) describe a method for producing stable bispecific antibodies by separately expressing and purifying two antibodies of interest and then mixing them under specified redox conditions (J. Mol. Biol. (2012) 420:204-19).

[0010] Zhu et al. (Genentech) have engineered mutations in the VL / VH interface of a diabody construct composed of variant domain antibody fragments that is completely devoid of constant domains, and produced heterodimeric diabodies (Protein Science (1997) 6: 781-788). Similarly, Igawa et al. (Chugai) have also engineered mutations in the VL / VH interface of single-chain diabodies to promote selective expression of diabodies and inhibit their isomerization (Protein Engineering, Design & Selection (2010) 23: 667-677).

[0011] U.S. Patent Publication No. 2009 / 0182127 (Novo Nordisk, Inc.) describes the generation of bispecific antibodies by modifying amino acid residues at the Fc interface and the CH1:CL interface of light chain-heavy chain pairs that reduce the ability of the light chain of one pair to interact with the heavy chain of the other pair.

[0012] Overview

[0013] Described herein is an isolated antigen-binding polypeptide construct comprising at least a first heterodimer and a second heterodimer, wherein the first heterodimer comprises a first immunoglobulin heavy chain polypeptide sequence (H1) and a first immunoglobulin light chain polypeptide sequence (L1); and the second heterodimer comprises a second immunoglobulin heavy chain polypeptide sequence (H2) and a second immunoglobulin light chain polypeptide sequence (L2), wherein at least one of the H1 or L1 sequences of the first heterodimer is different from the corresponding H2 or L2 sequence of the second heterodimer, and wherein H1 and H2 each comprise at least one heavy chain variable domain (V H domain) and the heavy chain constant domain (C H1L1 and L2 each contain at least a light chain variable domain (V L domain) and the light chain constant domain (C L domain); and at least one of H1, H2, L1 and L2 comprises at least one amino acid modification in at least one constant domain and / or at least one variable domain, wherein H1 preferentially pairs with L1 over L2, and H2 preferentially pairs with L2 over L1.

[0014] In some aspects, the construct further comprises a heterodimeric Fc comprising at least 2 C H3 Sequence, wherein the Fc is coupled to the first heterodimer and the second heterodimer through or without one or more linkers, wherein the dimerized C H3 The sequence has a melting temperature (Tm) of about 68°C or higher as measured by differential scanning calorimetry (DSC), and wherein the construct is bispecific.

[0015] In some aspects, the at least one amino acid modification is selected from at least one amino acid modification shown in the Tables or Examples.

[0016] In some aspects, H1 preferentially pairs with L1 over L2, and H2 preferentially pairs with L2 over L1 when H1, H2, L1, and L2 are co-expressed in a cell or mammalian cell, or when H1, H2, L1, and L2 are co-expressed in a cell-free expression system, or when H1, H2, L1, and L2 are co-expressed, or when H1, H2, L1, and L2 are co-produced by a redox production system.

[0017] In some aspects, at least one of H1, H2, L1, and L2 comprises V H and / or V L At least one amino acid modification of the C H1 and / or C L At least one amino acid of the domain is modified so that H1 preferentially pairs with L1 over L2, and / or H2 preferentially pairs with L2 over L1.

[0018] In some respects, if H1 is in C H1 The structural domain contains at least one amino acid modification, then at least one of L1 and L2 is in C L domain contains at least one amino acid modification; and / or if H1 is in V H The structural domain contains at least one amino acid modification, then at least one of L1 and L2 is in V L The domain contains at least one amino acid modification.

[0019] In some aspects, H1, L1, H2 and / or L2 comprise at least 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acid mutations. In some aspects, at least one of H1, H2, L1 and L2 comprises at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acid modifications of at least one constant domain and / or at least one variable domain.

[0020] In some aspects, when L1 and L2 are co-expressed with at least one of H1 and H2, the relative pairing of at least one of the H1-L1 and H2-L2 heterodimer pairs to at least one of the respective corresponding H1-L2 or H2-L1 heterodimer pairs is greater than 50%, 51%, 52%, 53%, 54%, 55%, 56%, 57%, 58%, 59%, 60%, 61%, 62%, 63%, 64%, 65%, 66%, 67%, 68%, 69%, 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 100%, 101%, 102%, 103%, 104%, 105%, 106%, 107%, 108%, 109%, 110%, 111%, 112%, 113%, 114%, 115%, 116%, 117%, 118%, 119%, 120%, 121%, 122%, 123%, 124%, 125 , 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%, and wherein the relative pairing of the modified H1-L1 or H2-L2 heterodimer pair is greater than the respective relative pairing observed in the corresponding H1-L1 or H2-L2 heterodimer pair without at least one amino acid modification.

[0021] In some aspects, the thermal stability measured by melting temperature (Tm) (as measured by DSC of at least one of the first and second heterodimers) is within about 0°C, 1°C, 2°C, 3°C, 4°C, 5°C, 6°C, 7°C, 8°C, 9°C or 10°C of the Tm of the corresponding heterodimer without at least one amino acid modification. In some aspects, the thermal stability measured by melting temperature (Tm) (as measured by DSC of each heterodimer comprising at least one amino acid modification) is within about 0°C, 1°C, 2°C, 3°C, 4°C, 5°C, 6°C, 7°C, 8°C, 9°C or 10°C of the Tm of the corresponding heterodimer without at least one amino acid modification.

[0022] In some aspects, the affinity of each heterodimer for the antigen to which it binds is within about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 25, 30, 35, 40, 45, or 50 times the affinity of the respective unmodified heterodimer for the same antigen as measured by surface plasmon resonance (SPR) or FACS.

[0023] In some aspects, at least one of H1 and L1 comprises at least one domain comprising at least one amino acid modification, which results in greater spatial complementarity of amino acids when H1 is paired with L1 compared to L2. In some aspects, at least one of H2 and L2 comprises at least one domain comprising at least one amino acid modification, which results in greater spatial complementarity of amino acids when H2 is paired with L2 compared to L1. In some aspects, at least one of H1 and L1 comprises at least one domain comprising at least one amino acid modification, which results in greater electrostatic complementarity between charged amino acids when H1 is paired with L1 compared to L2. In some aspects, at least one of H2 and L2 comprises at least one domain comprising at least one amino acid modification, which results in greater electrostatic complementarity between charged amino acids when H2 is paired with L2 compared to L1.

[0024] In some aspects, the at least one amino acid modification is a mutation shown in at least one of a set of tables or examples. In some aspects, the at least one modification is not H1-Q39E, L1-Q38K, H2-Q39K and L2-Q38E. In some aspects, the at least one modification is not H1-Q39E, L1-Q38E, H2-Q39K and L2-Q38K.

[0025] In some aspects, the construct further comprises a H3 The Fc of the sequence, wherein the Fc is coupled to the first heterodimer and the second heterodimer through or without one or more linkers.

[0026] In some aspects, the Fc is human Fc, human IgG1 Fc, human IgA Fc, human IgG Fc, human IgD Fc, human IgEFc, human IgM Fc, human IgG2 Fc, human IgG3 Fc or human IgG4 Fc. In some aspects, the Fc is a heterodimeric Fc. In some aspects, the Fc is in C H3 In some aspects, at least one of the dimerization sequences comprises one or more modifications. H3The sequence has a melting temperature of about 68°C, 69°C, 70°C, 71°C, 72°C, 73°C, 74°C, 75°C, 76°C, 77°C, 77.5°C, 78°C, 79°C, 80°C, 81°C, 82°C, 83°C, 84°C or 85°C or more (as measured by DSC). In some aspects, the Fc is a heterodimer formed with a purity of greater than about 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% when produced; or wherein The Fc is a heterodimer formed when expressed or when expressed by a single cell at a purity of greater than about 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%. H3 At least one of the sequences comprises one or more modifications that promote the formation of heterodimeric Fcs having stability compared to wild-type homodimeric Fcs. In some aspects, the Fc further comprises at least one C H2 In some aspects, the C of the Fc H2 The sequence comprises one or more modifications. In some aspects, the Fc comprises one or more modifications to promote selective binding of Fc-gamma receptors.

[0027] In some aspects, the Fc is coupled to the heterodimer via one or more joints, or wherein the Fc is coupled to H1 and H2 via one or more joints. In some aspects, the one or more joints are one or more polypeptide joints. In some aspects, the one or more joints comprise one or more antibody hinge regions. In some aspects, the one or more joints comprise one or more IgG1 hinge regions. In some aspects, the one or more joints comprise one or more modifications. In some aspects, one or more modifications to the one or more joints promote selective binding of Fc-γ receptors.

[0028] In some aspects, the at least one amino acid modification is at least one amino acid mutation or wherein the at least one amino acid modification is at least one amino acid substitution.

[0029] In some aspects, the sequence of each of H1, H2, L1, and L2 is derived from a human sequence.

[0030] In some aspects, the construct is multispecific or bispecific. In some aspects, the construct is multivalent or bivalent.

[0031] Also described herein is an isolated polynucleotide or a set of isolated polynucleotides comprising at least one sequence encoding a construct described herein. In some aspects, the polynucleotide or a set of polynucleotides is a cDNA.

[0032] Also described herein is a vector or a group of vectors, the vector comprising one or more of the polynucleotides described herein or a group of polynucleotides. In some aspects, the vector or the group of vectors is selected from a plasmid, a polycistronic vector, a viral vector, a non-additional mammalian vector, an expression vector, and a recombinant expression vector.

[0033] Also described herein is an isolated cell comprising a polynucleotide or set of polynucleotides described herein or a vector or set of vectors described herein.In some aspects, the cell is a hybridoma, a Chinese hamster ovary (CHO) cell, or a HEK293 cell.

[0034] Also described herein is a pharmaceutical composition comprising a construct described herein and a pharmaceutically acceptable carrier. In some aspects, the composition further comprises one or more substances selected from a buffer, an antioxidant, a low molecular weight molecule, a drug, a protein, an amino acid, a carbohydrate, a lipid, a chelating agent, a stabilizer, and an excipient.

[0035] Also described herein are uses of a construct described herein or a pharmaceutical composition described herein for treating a disease or disorder or cancer or vascular disease in a subject or for the manufacture of a medicament.

[0036] Also described herein is a method of treating a subject having a disease or disorder or cancer or vascular disease comprising administering to the subject a construct described herein or a composition described herein.

[0037] Also described herein is a method of inhibiting, attenuating or blocking a signal in or to a cell comprising contacting the cell with a construct described herein or a composition described herein.

[0038] Also described herein is a method for obtaining a construct described herein from a host cell culture, the method comprising the steps of: (a) obtaining a host cell culture comprising at least one host cell containing one or more nucleic acid sequences encoding the construct; and (b) recovering the construct from the host cell culture.

[0039] Also described herein is a method of obtaining the construct described herein, comprising the steps of: (a) obtaining H1, L1, H2, and L2; (b) allowing H1 to preferentially pair with L1 over L2, and allowing H2 to preferentially pair with L2 over L1; and (c) obtaining the construct.

[0040] Also described herein is a method for preparing a construct described herein, comprising: obtaining a polynucleotide or a group of polynucleotides encoding at least one construct; determining an optimal ratio of each polynucleotide or group of polynucleotides for introduction into at least one host cell, wherein the optimal ratio is determined by estimating the amount of H1-L1 and H2-L2 heterodimer pairs formed when H1, L1, H2 and L2 are expressed compared to the amount of mismatched H1-L2 and H2-L1 heterodimer pairs formed when H1, L1, H2 and L2 are expressed; selecting a preferred optimal ratio, wherein transfection of at least one host cell with a polynucleotide or a group of polynucleotides at a preferred optimal ratio results in expression of the construct; transfecting at least one host cell with a polynucleotide or a group of polynucleotides at an optimal ratio; and culturing at least one host cell to express the construct.

[0041] In some aspects, the optimal ratio is selected by estimating the transfection in a transient transfection system. In some aspects, transfection of at least one host cell with a polynucleotide or a group of polynucleotides at a preferred optimal ratio results in optimal expression of the construct. In some aspects, the construct comprises a C H3 In some aspects, the Fc is a heterodimer, optionally comprising one or more amino acid modifications.

[0042] Also described herein is a computer-readable storage medium storing a data set comprising data representing complementary mutations in a first heterodimer comprising a first immunoglobulin heavy chain polypeptide sequence (H1) and a first immunoglobulin light chain polypeptide sequence (L1); and complementary mutations in a second heterodimer comprising a second immunoglobulin heavy chain polypeptide sequence (H2) and a second immunoglobulin light chain polypeptide sequence (L2), wherein H1 and H2 each comprise at least one heavy chain variable domain (V H domain) and the heavy chain constant domain (C H1 wherein L1 and L2 each comprise at least a light chain variable domain (V L domain) and the light chain constant domain (C L domain), and wherein the dataset of complementary mutations comprises data representing those mutations or a subset of those mutations listed in the Tables or Examples; and computer executable code for determining the likelihood that H1 will prefer to pair with L1 over L2 and / or that H2 will prefer to pair with L2 over L1.

[0043] Also described herein is a computer-implemented method for determining preferential pairings, comprising: obtaining a data set comprising data representing complementary mutations in a first heterodimer comprising a first immunoglobulin heavy chain polypeptide sequence (H1) and a first immunoglobulin light chain polypeptide sequence (L1); and complementary mutations in a second heterodimer comprising a second immunoglobulin heavy chain polypeptide sequence (H2) and a second immunoglobulin light chain polypeptide sequence (L2), wherein H1 and H2 each comprise at least a heavy chain variable domain (V H domain) and the heavy chain constant domain (C H1 wherein L1 and L2 each comprise at least a light chain variable domain (V L domain) and the light chain constant domain (C L domain), and wherein the data set of complementary mutations comprises data representing those mutations listed in the table or example or a subset of those mutations; and determining by a computer processor the likelihood that H1 will preferentially pair with L1 over L2 and / or that H2 will preferentially pair with L2 over L1. In some aspects, the method further comprises generating a construct described herein.

[0044] Also described herein is a method for producing a bispecific antigen-binding polypeptide construct, the bispecific construct comprising a first heterodimer comprising a first immunoglobulin heavy chain polypeptide sequence (H1) and a first immunoglobulin light chain polypeptide sequence (L1) from a first monospecific antigen-binding polypeptide; and a second heterodimer comprising a second immunoglobulin heavy chain polypeptide sequence (H2) and a second immunoglobulin light chain polypeptide sequence (L2) from a second monospecific antigen-binding polypeptide, wherein H1 and H2 each comprise at least a heavy chain variable domain (V H domain) and the heavy chain constant domain (C H1 wherein L1 and L2 each comprise at least a light chain variable domain (V L domain) and the light chain constant domain (C L domain), the method comprising: obtaining a data set comprising data representing a set of amino acid modifications within H1, H2, L1 and L2, so that when a subset of modifications are introduced into H1, H2, L1 and / or L2, in the system H1 preferentially pairs with L1 over L2 and H2 preferentially pairs with L2 over L1; introducing a subset of one or more modifications from the data set into the first heterodimer and / or the second heterodimer; and co-expressing the first heterodimer and the second heterodimer in at least one host cell to produce an expression product comprising the bispecific construct.

[0045] In some aspects, the method further comprises determining the amount of bispecific construct in the expression product relative to other polypeptide products. In some aspects, the bispecific construct is produced at a purity greater than 70% compared to other polypeptide products. In some aspects, the data set is a data set described herein. In some aspects, the method further comprises the step of adding additional amino acid modifications to at least one of H1, H2, L1 or L2 to increase the purity of the bispecific construct compared to other polypeptide products. In some aspects, the construct comprises at least 2 C H3 In some aspects, the antigen-binding polypeptide is an antibody, a Fab or a scFv. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A table showing the preferred pairing of heterodimers from the antigen binding constructs described herein in the LCCA design set is provided, wherein V H or V L The domains were amino acid modified.

[0048] Figure 2 A table showing the preferred pairings of heterodimers in the LCCA design set is provided, where C H1 or C L The domains were amino acid modified.

[0049] Figure 3 Expressions showing the thermal stability of selected preferentially paired or mismatched heterodimers from the antigen-binding constructs described herein and the affinity of the antigen therefor are provided.

[0050] Figure 4 Thermal unfolding curves of selected heterodimers are provided.

[0051] Figure 5 Size exclusion chromatography profiles of selected heterodimers are provided.

[0052] Fig. 6 depicts the D3H44 heavy chain and light chain amino acid sequences aligned for the variable region, constant region and J-region segments of homologous human germline sequences.(symbols in the figure: * sequence identity, #interface hot spots (specific drivers), + residues mutated tested in design). Fig. 6A Human VH germline subgroups are depicted (one representative sequence is shown for each family). Sequence identities are based on alignment of D3H44 to VH3 and IGHJ3*02. Figure 6BHuman kappa VL germline subgroups are depicted (one representative sequence is shown for each family). Sequence identities are based on alignment of D3H44 to VKI and IGKJ1*01. Figure 6C Human lambda VL germline subgroups are depicted (one representative sequence is shown from each family). Sequence identities are based on alignment of D3H44 to VL1 and IGLJ1*01. Fig.6D Depicts the human CH1 allele sequence. Fig. 6E Human kappa and lambda allele sequences are depicted.

[0053] Figure 7 Depicted is a flow chart outlining the strategy for designing bispecific antibodies.

[0054] Figure 8 A high-level schematic overview of the engineering requirements for forming bispecific Mabs (monoclonal antibodies) and the assay requirements required for quantitative heavy chain light chain pairs is illustrated. The design goal of engineering bispecific Mabs with high purity (i.e., few or no mismatched HL associations) can be achieved by rationally engineering (through the introduction of specific amino acid mutations) the preferential pairing of two unique heavy chains for their unique cognate light chains. The method is schematically shown; here H1 has been engineered to preferentially pair with L1 rather than L2. Similarly, H2 has been engineered to preferentially pair with L2 rather than L1. Experimental screening of bispecific Mab designs requires assays that can simultaneously quantify H1-L1:H1-L2 and H2-L2:H2-L1. These assay requirements can be simplified by assuming that each bispecific Fab arm can be engineered independently. In this case, the assay may only need to quantify H1-L1:H1-L2 or H2-L2:H2-L1, rather than both simultaneously.

[0055] Fig. 9 A schematic depiction of how to mark heavy and light chains and determine preferential pairing is provided. In this schematic diagram, circles represent cells in which 3 constructs are transfected. The expression product is secreted from the cells, and the supernatant (SPNT) is flowed over the detection device (SPR chip in this case). Based on the detection levels of two different tags fused to the light chains of two competing heavy chain pairings, the quantitative assessment of the preferential pairing of the heavy chain to the two light chains can be determined.

[0056] Fig.10 Depicted are the expected heavy chain related products when each of the two full length heavy chains is co-expressed with two different light chains. Preferential pairing was assessed using MCA.

[0057] Fig.11Depicted is an exemplary set of H1, L1, H2, L2 chains designed so that H1 pairs with L1 in preference to L2 and H2 pairs with L2 in preference to L1. A cartoon representation of the 3D crystal structure of the variable region heavy and light chain interface is provided. The mutations introduced on the interface achieve two sets of electrostatic and steric complementarity in the variable region interface that are each used to preferentially form a forced pair. On the other hand, there are unfavorable steric and electrostatic mismatches in incorrect pairs, which can result in reduced pairing propensity for mispairing and reduced stability.

[0058] Fig.12 Describes the Fig.11 An exemplary set of H1, L1, H2, L2 chains shown in , LC / MS spectra generated when H1, L1 and L2 are co-expressed (left frame) and when H2, L1 and L2 are co-expressed (right frame).

[0059] Fig.13 Describes the Fig.11 Evaluation of the biophysical properties of the designed H1-L1 and H2-L2 pairs are shown. Fig.13 A shows non-reducing SDS-PAGE analysis of H1-L1 and H2-L2 pairs before and after protein A purification, and the yields of the products are shown at the bottom of the gel; Fig.13 B shows the DSC thermogram of the products of H1-L1 and H2-L2 pairing; and Fig.13 C shows the UPLC-SEC (H2-L2) characteristic spectrum of the H2-L2 pair.

[0060] Fig.14 Depicted are the expected heavy chain related products when two different light chains are co-expressed with two different heavy chains in cells. Preferential pairing was assessed using SMCA (monoclonal antibody competition assay).

[0061] Fig.15 Describes the Fig.11 LC / MS spectra of the designed bispecific antigen-binding construct (H1-L1_H2-L2) shown in .

[0062] Fig.16 AThe left frame depicts the Fig.11 Evaluation of the purity of the bispecific antigen-binding constructs of the design set depicted in . The figure shows a Coomassie-stained non-reducing SDS-PAGE of the SMCA variants and a control variant consisting of a Her2-binding Mab and a heterodimeric Fc. The purity of the SMCA variants after protein A (ProtA) purification was high and comparable in nature to the control. The estimated post-ProtA yield of the SMCA variants was 20 mg / L and was comparable to the control yield of 40 mg / ml. The right frame ( Fig.16B) Depicts the SEC profiles of the above bispecific constructs. The major peak observed (>90% of the total peak area) runs at a molecular weight of ~150KDa and is composed of Mab monomers. The minor peak observed runs at ~75KDa and is composed of half antibodies. No significantly higher molecular weight peaks (potentially indicating aggregate species) were observed.

[0063] Fig.17 A depiction shows based on Fig.11 SPR data for monospecific binding of the bispecific constructs of the design set depicted in FIG. 1 for either TF or Her2 antigens (Note: This SMCA design replacement sheet (Article 26) utilizes the same Fig.11 Same H1-L1 and H2-L2 MCA designs as depicted in ).

[0064] Fig.17 B depicts SPR data showing simultaneous bispecific binding of the bispecific antigen binding construct to TF and Her2 antigens.

[0065] Fig.18 A flow chart is depicted for the identification of critical interface residues and for computer modeling regarding the design of preferential heavy chain-light chain pairings.

[0066] Fig.19 A schematic diagram is provided depicting a method for preparing bispecific antibodies using a forced mutation pair library provided in the present invention.

[0067] Details

[0068] Provided herein is an antigen-binding polypeptide construct (also referred to as a heterodimer pair) that may include a first heterodimer and a second heterodimer, wherein each heterodimer includes an immunoglobulin heavy chain or a fragment thereof and an immunoglobulin light chain. At least one of the heterodimers may include one or more amino acid modifications in an immunoglobulin heavy chain constant domain 1 (CH1) and one or more amino acid modifications in an immunoglobulin light chain constant domain (CL); one or more amino acid modifications in an immunoglobulin heavy chain variable domain (VH) and one or more amino acid modifications in an immunoglobulin light chain variable domain (VL); or a combination of the aforementioned amino acid modifications for the constant and variable domains of heavy and light chains. The modified amino acid is typically part of the interface between the light chain and the heavy chain, and is modified to produce a preferential pairing between each heavy chain and the desired light chain, so that the heavy chain of the first heterodimer is preferentially paired with one of the light chains rather than the other. Similarly, the heavy chain of the second heterodimer may preferentially pair with the second light chain rather than the first light chain.

[0069] As noted above, the specific combination of amino acid modifications described herein promotes preferential pairing of heavy chains with specific light chains, thereby enabling bispecific monoclonal antibody (Mab) expression to occur with negligible or limited mispairing, and minimizing the need to purify the desired heterodimer from unwanted or mispaired products. The heterodimer can exhibit thermal stability comparable to heterodimers that do not include the amino acid modifications, and can also exhibit binding affinity for antigens comparable to heterodimers that do not include the amino acid modifications.

[0070] The design of the first and second heterodimers can be used to generate bispecific antibodies that target two different therapeutic targets or that target two different epitopes (overlapping or non-overlapping) within the same antigen.

[0071] The present invention also provides a method for preparing a heterodimer pair according to the present invention.

[0072] definition

[0073] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the claimed subject matter belongs. Where multiple definitions exist for a term herein, those in this section shall prevail. When a URL or other such identifier or address is referred to, it is understood that such identifiers may change, and that particular information on the Internet may vary, but equivalent information may be found by searching the Internet. Reference to such information demonstrates their availability and public dissemination.

[0074] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of any subject matter claimed.In this application, the use of the singular includes the plural unless expressly stated otherwise.

[0075] In this specification, unless otherwise specified, any concentration range, percentage range, ratio range or integer range should be understood to include any integer value within the listed range, and its fraction (such as one tenth and one hundredth of an integer) when appropriate. As used herein, unless otherwise specified, "about" means ± 10% of the indicated range, value, sequence or structure. It should be understood that, unless otherwise specified or specified by its context, as used herein, the terms "one (a)" and "an" refer to "one or more" in the listed components. The use of substitutions (e.g., "or") should be understood to mean any one, both or any combination thereof in the substitutions. As used herein, the terms "include" and "comprising" are used synonymously. In addition, it should be understood that the separate single-chain polypeptides or immunoglobulin constructs derived from the different combinations of structures and substituents described herein are disclosed by the present application, and the degree of disclosure is just like each single-chain polypeptide or heterodimer is separately described. Therefore, selecting specific components to form separate single-chain polypeptides or heterodimers is within the scope of the present disclosure.

[0076] The section headings used herein are for organizational purposes only and should not be construed as limiting the subject matter described. All documents or portions of documents cited in this application, including but not limited to patents, patent applications, articles, books, manuals, and treatises, are hereby expressly incorporated by reference in their entirety for any purpose.

[0077] It should be understood that the methods and compositions described herein are not limited to the specific methods, protocols, cell lines, constructs and reagents described herein, and these may vary. It should also be understood that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the scope of the methods and compositions described herein, which will only be limited by the appended claims.

[0078] All publications and patents mentioned herein are incorporated herein by reference in their entirety for the purpose of describing and disclosing the construct and method described in the publication, which can be used in conjunction with the methods, compositions and compounds described herein. The publications discussed herein are provided only because of their disclosures prior to the filing date of the present application. Any content herein should not be construed as approving that the inventor described herein has no right to precede the disclosure due to prior inventions or for any other reason.

[0079] In the present application, amino acid names and atom names (e.g., N, O, C, etc.) are used as defined by the Protein Data Bank (PDB) (www.pdb.org), which is based on the IUPAC nomenclature (IUPAC Nomenclature and Symbolism for Amino Acids and Peptides (residue names, atom names, etc.), Eur. J. Biochem., 138, 9-37 (1984) together with its amendments in Eur. J. Biochem., 152, 1 (1985). The term "amino acid residue" is mainly intended to refer to an amino acid residue contained in the group consisting of 20 naturally occurring amino acids, namely alanine (Ala or A), cysteine ​​(Cys or C), aspartic acid (Asp or D), glutamic acid (Glu or E), phenylalanine (Pyr or Pyr), thiazolyl acetate ... Alanine (Phe or F), glycine (Gly or G), histidine (His or H), isoleucine (Ile or I), lysine (Lys or K), leucine (Leu or L), methionine (Met or M), asparagine (Asn or N), proline (Pro or P), glutamine (Gln or Q), arginine (Arg or R), serine (Ser or S), threonine (Thr or T), valine (Val or V), tryptophan (Trp or W) and tyrosine (Tyr or Y) residues.

[0080] The terms "polypeptide," "peptide," and "protein" are used interchangeably herein to refer to a polymer of amino acid residues. That is, a description of a polypeptide applies equally to a description of a peptide and a description of a protein, and vice versa. The terms apply to naturally occurring amino acid polymers, as well as amino acid polymers in which one or more amino acid residues is a non-naturally encoded amino acid. As used herein, the terms encompass amino acid chains of any length, including full-length proteins, in which the amino acid residues are linked by covalent peptide bonds.

[0081] The term "nucleotide sequence" or "nucleic acid sequence" is intended to mean a contiguous stretch of two or more nucleotide molecules. The nucleotide sequence may be of genomic, cDNA, RNA, semisynthetic or synthetic origin, or any combination thereof.

[0082] "Cell", "host cell", "cell line" and "cell culture" are used interchangeably herein, and all such terms should be understood to include the progeny produced by the growth or cultivation of the cell. "Transformation" and "transfection" are used interchangeably to refer to the process of introducing a nucleic acid sequence into a cell.

[0083] The term "amino acid" refers to naturally occurring and non-naturally occurring amino acids, as well as amino acid analogs and amino acid mimetics that function in a manner similar to naturally occurring amino acids. Naturally encoded amino acids are the 20 common amino acids (alanine, arginine, asparagine, aspartic acid, cysteine, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, and valine) as well as pyrrolysine and selenocysteine. Amino acid analogs refer to compounds having the same basic chemical structure (i.e., alpha carbon bound to hydrogen, carboxyl group, amino group, and R group) as naturally occurring amino acids, such as homoserine, norleucine, methionine sulfoxide, methionine methylsulfonium. Such analogs have modified R groups (such as norleucine) or modified peptide backbones, but retain the same basic chemical structure as naturally occurring amino acids. Reference to amino acids includes, for example, naturally occurring proteinaceous L-amino acids; D-amino acids; chemically modified amino acids, such as amino acid variants and derivatives; naturally occurring non-proteinaceous amino acids, such as β-alanine, ornithine, etc.; and chemically synthesized compounds having properties known in the art as characteristic of amino acids. Examples of non-naturally occurring amino acids include, but are not limited to, □-methyl amino acids (e.g., methylalanine), D-amino acids, histidine-like amino acids (e.g., 2-amino-histidine, hydroxy-histidine, homohistidine), amino acids with additional methylene groups in the side chain ("homo" amino acids), and amino acids in which the carboxylic acid functional group in the side chain is replaced by a sulfonic acid group (e.g., cysteine). Incorporating non-natural amino acids, including synthetic non-natural amino acids, substituted amino acids, or one or more D-amino acids into the proteins of the present invention can be advantageous in a variety of different ways. Peptides and the like containing D-amino acids exhibit increased stability in vitro or in vivo compared to their L-amino acid counterparts. Therefore, when greater intracellular stability is needed or required, the construction of peptides and the like incorporating D-amino acids can be particularly advantageous. More specifically, D-peptides and the like are resistant to endogenous peptidases and proteases, thereby providing improved bioavailability of the molecule and extended in vivo life when such properties are desired. In addition, D-peptides and the like cannot be efficiently processed for class II major histocompatibility complex-restricted presentation to T helper cells and are therefore less likely to induce a humoral immune response in an intact organism.

[0084] Amino acids may be referred to herein by either their commonly known three letter symbols or by the one-letter symbols recommended by the IUPAC-IUB Biochemical Nomenclature Commission. Likewise, nucleotides may be referred to by their commonly recognized single-letter codes.

[0085] "Conservatively modified variants" apply to amino acids and nucleic acid sequences. With respect to specific nucleic acid sequences, "conservatively modified variants" refer to those nucleic acids that encode the same or substantially the same amino acid sequence, or when the nucleic acid does not encode an amino acid sequence, it refers to substantially the same sequence. A large number of functionally identical nucleic acids encode any given protein due to the degeneracy of the genetic code. For example, the codons GCA, GCC, GCG, and GCU all encode the amino acid alanine. Therefore, at each position where alanine is specified by a codon, the codon can be changed to any of the corresponding codons described without changing the encoded polypeptide. Such nucleic acid variations are "silent variations", which are a type of variation with conservative modifications. Each nucleic acid sequence encoding a polypeptide herein also describes each possible silent variation of the nucleic acid. One of ordinary skill in the art will recognize that each codon in a nucleic acid (except AUG, which is usually the only codon for methionine, and TGG, which is usually the only codon for tryptophan) can be modified to produce functionally identical molecules. Therefore, each silent variation of a nucleic acid encoding a polypeptide is implicit in each described sequence.

[0086] With respect to amino acid sequences, one of ordinary skill in the art will recognize that individual substitutions, deletions or additions to a nucleic acid, peptide, polypeptide or protein sequence that alter, add or delete a single amino acid or a small percentage of amino acids in the coded sequence are "conservatively modified variants," wherein the changes produce a deletion of an amino acid, an addition of an amino acid, or an amino acid replacement by a chemically similar amino acid. Conservative substitution tables providing functionally similar amino acids are known to one of ordinary skill in the art. Such conservatively modified variants are in addition to and do not exclude polymorphic variants, interspecies homologs and alleles of the present invention.

[0087] Conservative substitution tables providing functionally similar amino acids are known to those of ordinary skill in the art. The following eight groups each contain amino acids that are conservative substitutions of each other:

[0088] 1) Alanine (A), glycine (G);

[0089] 2) Aspartic acid (D), Glutamic acid (E);

[0090] 3) Asparagine (N), glutamine (Q);

[0091] 4) Arginine (R), Lysine (K);

[0092] 5) Isoleucine (I), leucine (L), methionine (M), valine (V);

[0093] 6) Phenylalanine (F), tyrosine (Y), tryptophan (W);

[0094] 7) serine (S), threonine (T); and

[0095] 8) Cysteine ​​(C), Methionine (M)

[0096] (See, e.g., Creighton, Proteins: Structures and Molecular Properties (W H Freeman & Co.; 2nd ed. (December 1993)).

[0097] The term "identical" or percentage "identity" refers to two or more sequences or subsequences being identical in the context of two or more nucleic acids or polypeptide sequences. When comparing and comparing for maximum correspondence on a comparison window or a designated area, as measured by using one of the following sequence comparison algorithms (or other algorithms available for use by those of ordinary skill in the art) or by manual comparison and visual inspection, if the sequence has a certain percentage of identical amino acid residues or nucleotides (i.e. about 50% identity, about 55% identity, 60% identity, about 65%, about 70%, about 75%, about 80%, about 85%, about 90% about 95% or about 99% identity in a designated area), then the sequence is "substantially identical". This definition also refers to the complementary sequence of a test sequence. Identity can be present in a region of at least about 50 amino acids or nucleotides in length, or in a region of 75 to 100 amino acids or amino acids or nucleotides in length, or, in the case of unspecified, in the entire sequence of a polynucleotide or polypeptide. The polynucleotides encoding the polypeptides of the present invention (including homologs from species other than humans) can be obtained by a method comprising the following steps: screening a library using a labeled probe having a polynucleotide sequence of the present invention or a fragment thereof under stringent hybridization conditions, and isolating full-length cDNA and genomic clones containing the polynucleotide sequence. Such hybridization techniques are well known to those of ordinary skill in the art.

[0098] If the amino acid sequence of a derivative or variant of a polypeptide is at least 50% identical to a sequence of 100 amino acids from the original peptide, the derivative or variant is considered to share "homology" or "homologous" with the peptide. In certain embodiments, the derivative or variant is at least 75% identical to the peptide or a derivative or variant of a fragment of the peptide having the same number of amino acid residues as the derivative. In certain embodiments, the derivative or variant is at least 85% identical to the peptide or a derivative or variant of a fragment of the peptide having the same number of amino acid residues as the derivative as a derivative. In certain embodiments, the amino acid sequence of the derivative is at least 90% identical to the peptide or a fragment of the peptide having the same number of amino acid residues as the derivative. In some embodiments, the amino acid sequence of the derivative is at least 95% identical to the derivative or variant of the fragment of the peptide having the same number of amino acid residues as the derivative. In certain embodiments, the derivative or variant is at least 99% identical to the derivative or variant of the fragment of the peptide having the same number of amino acid residues as the derivative.

[0099] As used herein, an "isolated" polypeptide or construct means a construct or polypeptide that has been identified and separated and / or recovered from a component of its natural cell culture environment. Contaminating components of its natural environment are materials that may normally interfere with the diagnostic or therapeutic use of the heteromultimer and may include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes.

[0100] In certain embodiments, as used herein, an "isolated" antigen-binding construct described herein comprises a heterodimer pair or an "isolated" heterodimer comprising a heterodimer or a heterodimer pair that has been identified and separated from and / or recovered from components of its natural cell culture environment. Contaminating components of the natural environment are materials that may interfere with the diagnosis or therapeutic use of the heterodimer or antigen-binding construct, and may include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes.

[0101] Heterodimers, as well as antigen-binding constructs and heterodimer pairs, are typically purified to substantially homogeneity. The phrases "substantially homogeneous," "substantially homogeneous form," and "substantially homogeneous" are used to indicate that the product is substantially free of byproducts derived from unwanted polypeptide combinations (e.g., homodimers). Expressed in terms of purity, substantially homogeneous means that the amount of byproducts does not exceed 10%, preferably less than 5%, more preferably less than 1%, and most preferably less than 0.5%, wherein the percentages are calculated by weight.

[0102] The phrase "selectively (or specifically) hybridizes to" refers to the binding, duplexing or hybridization of a molecule only to a particular nucleotide sequence under stringent hybridization conditions when the sequence is present in a complex mixture (including but not limited to, total cellular or library DNA or RNA).

[0103] Unless explicitly defined differently herein, terms understood by those skilled in the art of antibody technology are each given the meaning acquired in the art. Antibodies are known to have variable regions, hinge regions, and constant domains. Immunoglobulin structure and function are reviewed in, for example, Harlow et al., eds., Antibodies: A Laboratory Manual, Chapter 14 (Cold Spring Harbor Laboratory, Cold Spring Harbor, 1988).

[0104] As used herein, the terms "antibody" and "immunoglobulin" or "antigen-binding polypeptide" are used interchangeably. "Antigen-binding polypeptide" refers to a polypeptide that specifically binds to an analyte (antigen) and is generally encoded by an immunoglobulin gene or one or more fragments thereof. Recognized immunoglobulin genes include κ, λ, α, γ, δ, ε, and μ constant region genes, as well as countless immunoglobulin variable region genes. Light chains are classified as κ or λ. Heavy chains are classified as γ, μ, α, δ, or ε, which in turn define immunoglobulin subtypes IgG, IgM, IgA, IgD, and IgE, respectively. In addition, antibodies can belong to one isotype among many subtypes, for example, IgG can belong to IgG1, IgG2, IgG3, or IgG4 subclasses.

[0105] An exemplary immunoglobulin (antibody) structural unit consists of two pairs of polypeptide chains, each pair having one "light chain" (about 25 kD) and one "heavy chain" (about 50-70 kD). The term "light chain" includes full-length light chains and fragments thereof having sufficient variable region sequences to confer binding specificity. Full-length light chains include variable region domains VL and constant region domains CL. The variable region domains of the light chain are at the amino terminus of the polypeptide. Light chains include kappa chains and lambda chains. The term "heavy chain" includes full-length heavy chains and fragments thereof having sufficient variable region sequences to confer binding specificity. Full-length heavy chains include variable region domains VH and three constant region domains CH1, CH2, and CH3. The VH domain is at the amino terminus of the polypeptide, and the CH domain is at the carboxyl terminus, with CH3 closest to the carboxyl terminus of the polypeptide. The heavy chain may be of any isotype, including IgG (including IgG1, IgG2, IgG3 and IgG4 subclasses), IgA (including IgA1 and IgA2 subclasses), IgM and IgE. The term "variable region" or "variable domain" refers to the portion of the light chain and / or heavy chain of an antibody that is generally responsible for antigen recognition, generally including approximately 120 to 130 amino acids at the amino terminal end of the heavy chain (VH) and approximately 100 to 110 amino terminal amino acids in the light chain (VL). "Complementarity determining region" or "CDR" is an amino acid sequence that contributes to antigen binding specificity and affinity. The "framework" region (FR) can help maintain the correct conformation of the CDR to promote the binding between the antigen binding region and the antigen. Structurally, the framework region can be located between the CDRs in an antibody. The variable region generally exhibits the same general structure of a relatively conservative framework region (FR) connected by three hypervariable region CDRs. The CDRs from the two chains of each pair are generally aligned by the framework region, which can enable binding to a specific epitope. From N-terminal to C-terminal, light chain and heavy chain variable region usually comprise domain FR1, CDR1, FR2, CDR2, FR3, CDR3 and FR4.Unless otherwise noted, the distribution of amino acid to each domain is usually based on the definition of the Kabat sequence of the protein of immunological concern (National Institutes of Health, Bethesda, Md. (1987 and 1991)).In certain embodiments, immunoglobulin construct comprises at least one immunoglobulin domain from IgG, IgM, IgA, IgD or IgE connected to therapeutic polypeptide.In some embodiments, the immunoglobulin domain included in the immunoglobulin construct provided herein is from a construct based on immunoglobulin, such as a double antibody or nano antibody.In certain embodiments, immunoglobulin construct described herein comprises at least one immunoglobulin domain from a heavy chain antibody such as a camelid antibody.In certain embodiments, the immunoglobulin constructs provided herein comprise at least one immunoglobulin domain from a mammalian antibody, such as a bovine antibody, a human antibody, a camelid antibody, a mouse antibody, or any chimeric antibody.

[0106] "Bispecific", "dual specific" or "bifunctional" antigen binding proteins or antibodies are hybrid antigens with two different binding sites. Bispecific antigen binding proteins and antibodies are types of multispecific antigen binding protein antibodies. The two binding sites of a bispecific antigen binding protein or antibody will bind to two different epitopes, which may be located in the same or different molecular targets. "Multispecific antigen binding proteins" or "multispecific antibodies" are antigen binding proteins or antibodies that target more than one antigen or epitope. "Bivalent antigen binding proteins" or "bivalent antibodies" contain two antigen binding sites. In some cases, the two binding sites have the same antigen specificity. Bivalent antigen binding proteins and bivalent antibodies can be bispecific, see, below. Bivalent antibodies, except for "multispecific" or "multifunctional" antibodies, are generally understood to have each of their binding sites identical in certain embodiments.

[0107] The term "preferential pairing" is used herein to describe a first polypeptide and a second polypeptide, such as an immunoglobulin heavy chain and an immunoglobulin light chain pairing pattern in the antigen-binding constructs and heterodimer pairs described herein. Thus, "preferential pairing" refers to the preferred pairing of the first polypeptide and the second polypeptide when pairing occurs between the first and second polypeptides when one or more additional different polypeptides are present simultaneously. Typically, preferential pairing occurs as a result of modification (e.g., amino acid modification) of one or both of the first and second polypeptides. Typically, preferential pairing results in the first and second polypeptides being paired being the most abundant dimers present after pairing occurs. It is known in the art that an immunoglobulin heavy chain (H1), if co-expressed with two different immunoglobulin light chains (L1 and L2), will be statistically paired equally with the two light chains, which results in a mixture of approximately 50:50 H1 paired with L1 and H1 paired with L2. In this context, "preferential pairing" can occur, for example, between H1 and L1 if the amount of H1-L1 heavy chain-light chain heterodimers is greater than the amount of H1-L2 heterodimers when H1 is co-expressed with L1 and L2. Thus, in this case, H1 preferentially pairs with L1 relative to L2.

[0108] The antibody heavy chain is paired with the antibody light chain and meets or contacts each other at an "interface". The "interface" includes one or more "contact" amino acid residues in the first polypeptide that interact with one or more "contact" amino acid residues of the second polypeptide. In one context, the term interface can be used to describe the interface of the dimerized CH3 domain of Fc, wherein the Fc is preferably derived from an IgG antibody, such as IgG1, most preferably a human IgG1 antibody.

[0109] The antibody heavy chains associated with the antibody light chains usually meet or contact each other at an "interface". The immunoglobulin light chain is operably associated with the immunoglobulin heavy chain through the "interface". The "interface" comprises one or more "contact" amino acid residues in the immunoglobulin heavy chain that interact with one or more "contact" amino acid residues in the interface of the immunoglobulin light chain. As used herein, the interface may comprise the VH and CH1 domains of the immunoglobulin heavy chain and the VL and CL domains of the immunoglobulin light chain. The "interface" may be derived from an IgG antibody, most preferably a human IgG1 antibody.

[0110] As used herein, the term "amino acid modification" includes, but is not limited to, amino acid mutation, insertion, deletion, substitution, chemical modification, physical modification, and rearrangement.

[0111] Antigen binding constructs and heterodimer pairs

[0112] The antigen binding constructs described herein may comprise a first heterodimer and a second heterodimer; each heterodimer is obtained by pairing an immunoglobulin heavy chain with an immunoglobulin light chain. The structure and organization of the constant and variable domains of the immunoglobulin heavy and light chains are well known in the art. An immunoglobulin heavy chain typically comprises a variable (VH) domain and three constant domains CH1, CH2, and CH3. An immunoglobulin light chain typically comprises a variable (VL) domain and a constant (CL) domain. Various modifications of these typical forms can be produced.

[0113] Antigen-binding constructs and heterodimer pairs described herein may include a first heterodimer and a second heterodimer, each heterodimer comprising an immunoglobulin / antibody heavy chain or a fragment thereof having at least a VH and CH1 domain, and an immunoglobulin / antibody light chain having a VL domain and a CL domain. In one embodiment, the heterodimer and antigen-binding construct of the heterodimer pair comprise a full-length immunoglobulin heavy chain. In another embodiment, the heterodimer or antigen-binding construct of the heterodimer pair comprises a fragment of an immunoglobulin heavy chain including at least a VH and CH1 domain. In one embodiment, both heterodimers of the heterodimer pair comprise an amino-terminal fragment of an immunoglobulin heavy chain comprising at least a VH and CH1 domain. In another embodiment, both heterodimers of the heterodimer pair comprise a carboxyl-terminal fragment of an immunoglobulin heavy chain comprising at least a VH and CH1 domain.

[0114] Each heterodimer of the heterodimer pair can specifically bind to an antigen or epitope. In one embodiment, the immunoglobulin heavy chain and immunoglobulin light chain of each heterodimer are derived from a known therapeutic antibody or are engineered from it. Therapeutic antibodies are antibodies that are effective in treating the disease or disorder of a mammal having a disease or disorder or being susceptible to the disease or disorder. The suitable therapeutic antibody from which each heterodimer is derived can be derived from (including but not limited to) abavoumab, adalimumab, alemtuzumab, aurograb, bapinux monoclonal antibody, basiliximab, belimumab, bevacizumab, briakinumab, canakinumab, catumaxomab, certolizumab monoclonal antibody, cetuximab, zenipax, denosumab, efalizumab, galiximab, gemtuzumab monoclonal antibody, golimumab, ibritumomab tiuxetan, infliximab, ipilimumab, ruximab, mepolizumab, movizumab monoclonal antibody. Anti-cancer drugs include leukemia, leukemia, nab, leukemia, selegiline, avizumab, avizumab, rituximab, teplizumab, tocilizumab / atlizumab, tositumomab, trastuzumab, ProxiniumTM, RencarexTM, ustekinumab, and zalutumumab.

[0115] In one embodiment, the immunoglobulin heavy chain and immunoglobulin light chain of each dimer are derived from or engineered from an antibody, the antibody binding molecule including but not limited to the following proteins and subunits, domains, motifs and epitopes of proteins belonging to the following list: renin; growth hormone, including human growth hormone and bovine growth hormone; growth hormone releasing factor; parathyroid hormone; thyroid stimulating hormone; lipoprotein; alpha-1-antitrypsin; insulin A-chain; insulin B-chain; proinsulin; follicle stimulating hormone; calcitonin; luteinizing hormone; glucagon; coagulation factors such as factor VII, factor VIIIC, factor IX, tissue factor (TF) and von Willebrand factor; anticoagulant factor activators such as protein C; atrial natriuretic factor; pulmonary surfactant; plasminogen activators such as urokinase or human urine or tissue-type plasminogen activator (t-PA); bombesin; thrombin; hematopoietic growth factor; tumor necrosis factor alpha and beta; enkephalinase; RANTES (regulation of activation normal T cell expressed and secreted); human macrophage inflammatory protein (MIP-1-alpha); serum albumin such as human serum albumin; Mullerian inhibiting substance; relaxin A chain; relaxin B chain; prorelaxin; mouse gonadotropin-related peptide; microbial proteins such as beta-lactamase; DNA enzymes; IgE; cytotoxic T lymphocyte-associated antigen (CTLA) such as CTLA-4; inhibins; activins; vascular endothelial growth factor-1; VEGF; receptors for hormones or growth factors such as, for example, EGFR, VEGFR; interferons such as alpha interferon (α-IFN), beta interferon (β-IFN) and gamma interferon (γ-IFN); protein A or D; rheumatoid factor; neurotrophic factors such as bone-derived neurotrophic factor (BDNF), neurotrophin-3, -4, -5 or -6 (NT-3, NT-4, NT-5 or NT-6), or nerve growth factor; platelet-derived growth factor (PDGF); fibroblast growth factors such as AFGF and PFGF; epidermal growth factor (EGF); transforming growth factors (TGF) such as TGF-α and TGF-β, including TGF-1, TGF-2, TGF-3, TGF-4 or TGF-5; insulin-like growth factor-I and -II (IGF-I and IGF-II); de (1-3)-IGF-I (brain IGF-I), insulin-like growth factor binding protein; CD proteins such as CD2, CD3, CD4, CD8, CD11a, CD14, CD18, CD19, CD20, CD22, CD23, CD25, CD33, CD34, CD40, CD40L, CD52, CD63, CD64, CD80 and CD147; erythropoietin; osteoinductive factors; immunotoxins; bone morphogenic proteins (BMPs); interferons such as interferon alpha, beta and gamma;Colony stimulating factors (CSFs), such as M-CSF, GM-CSF, and G-CSF; interleukins (ILs), e.g., IL-1 to IL-13; TNF-α, superoxide dismutase; T cell receptors; surface membrane proteins; decay accelerating factors; viral antigens such as, e.g., a portion of the AIDS envelope, e.g., gp120; transporters; homing receptors; addressins; regulatory proteins; cell adhesion molecules such as LFA-1, Mac 1, p150.95, VLA-4, ICAM-1, ICAM-3 and VCAM, α4 / β7 integrin and αv / β3 integrin, including any one or their subunits, integrin α subunits such as CD49a, CD49b, CD49c, CD49d, CD49e, CD49f, α7, α8, α9, αD, CD11a, CD11b, CD51, CD11c, CD41, αIIb, αIELb; integrin β subunits such as CD29, CD 18, CD61, CD104, β5, β6, β7 and β8; integrin subunit combinations, including but not limited to αVβ3, αVβ5 and α4β7; members of the apoptosis pathway; IgE; blood group antigens; flk2 / flt3 receptors; obesity (OB) receptor; mp1 receptor; CTLA-4; protein C; Eph receptors such as EphA2, EphA4, EphB2, etc.; human leukocyte antigens (HLA) such as HLA-DR; complement proteins such as complement receptors CR1, C1Rq and other complement factors such as C3 and C5; glycoprotein receptors such as GpIb.α., GPIIb / IIIa and CD200; and fragments of any of the polypeptides listed above. ;

[0116] In embodiments, the immunoglobulin heavy and light chains of each heterodimer are derived from or engineered from antibodies that specifically bind to cancer antigens, including but not limited to ALK receptor (pleiotrophin receptor), pleiotrophin, KS1 / 4 pan-cancer antigen; ovarian cancer antigen (CA125); prostatic acid phosphate; prostate specific antigen (PSA); melanoma-associated antigen p97; melanoma antigen gp75; high molecular weight melanoma antigen (HMW-MAA); prostate specific membrane antigen; carcinoembryonic antigen (CEA); polymorphic epithelial mucin antigen; human milk fat globule antigen; colorectal tumor-associated antigens such as: CEA, TAG-72, CO17-1A, GICA 19-9, CTA-1, LEA; Burkitt's lymphoma antigen 38.13; CD19; human B lymphoma antigen CD20; CD33; melanoma-specific antigens such as ganglioside GD2, ganglioside GD3, ganglioside GM2 and ganglioside GM3; tumor-specific transplantable cell surface antigen (TSTA); virus-induced tumor antigens, including T-antigen, envelope antigens of DNA tumor viruses and RNA tumor viruses; carcinoembryonic antigen alpha-fetoprotein such as CEA of colon cancer, 514 carcinoembryonic antigen Trophoblastic glycoprotein and bladder tumor carcinoembryonic antigen, differentiation antigens such as human lung cancer antigens L6 and L20, antigens of fibrosarcoma; human leukemia T cell antigen Gp37; neoglycoprotein, sphingolipids; breast cancer antigens such as EGFR (epidermal growth factor receptor); NY-BR-16; NY-BR-16 and HER2 antigen (p185HER2); polymorphic epithelial mucin (PEM); malignant human lymphocyte antigen APO-1; differentiation antigens such as I antigen found in fetal erythrocytes; found in adult erythrocytes Primary endoderm I antigen; preimplantation embryo; I (Ma) found in gastric adenocarcinoma; M18, M39 found in mammary epithelial cells; SSEA-1 found in bone marrow cells; VEP8; VEP9; Myl; Va4-D5; D156-22 found in colorectal cancer; TRA-1-85 (blood group H); SCP-1 found in testicular and ovarian cancers; C14 found in colon adenocarcinoma; F3 found in lung adenocarcinoma; AH6 found in gastric cancer; Y hapten; found in embryonal carcinoma cells Ley found in tumours; TL5 (blood type A); EGF receptor found in A431 cells; E1 series (blood type B) found in pancreatic cancer; FC10.2 found in embryonic carcinoma cells; gastric adenocarcinoma antigen; CO-514 (blood type Lea) found in adenocarcinoma; NS-10 found in adenocarcinoma; CO-43 (blood type Leb); G49 found in the EGF receptor of A431 cells; MH2 (blood type ALeb / Ley) found in colon adenocarcinoma, 19.9; gastric cancer mucin; T5A7 found in bone marrow cells; R24 found in melanoma; 4.2, GD3, D1.1, OFA-1, GM2, OFA-2, GD2, and M1:22:25:8 found in embryonic carcinoma cells and SSEA-3 and SSEA-4 found in 4- to 8-cell stage embryos; cutaneous T-cell lymphoma antigen, MART-1 antigen; Sialy Tn (STn) antigen; colon cancer antigen NY-CO-45, lung cancer antigen NY-LU-12 variant A; adenocarcinoma antigen ART1; paraneoplastic brain-testis cancer antigen (cancer neuron antigen MA2; paraneoplastic neural antigen); neuro-oncology abdominal antigen 2 (NOVA2); hepatocellular carcinoma antigen gene 520; tumor-associated antigen CO-029; tumor-associated antigen MAGE-C1 (cancer / testis antigen CT7), MAGE-B1 (MAGE-XP antigen), MAGE-B2 (DAM6), MAGE-2, MAGE-4-a, MAGE-4-b and MAGE-X2; cancer-testis antigen (NY-EOS-1) and fragments of any of the polypeptides listed above. .

[0117] Human antibodies can be classified into isotypes, including IgG, IgA, IgE, IgM and IgD. In one embodiment, Fc is derived from the IgG isotype. In another embodiment, Fc is derived from the IgA isotype. In another embodiment, Fc is derived from the IgE isotype. In another embodiment, Fc is derived from the IgM isotype. In another embodiment, Fc is derived from the IgD isotype.

[0118] Human IgG antibodies can also be subdivided into subclasses IgG1, IgG2, IgG3 and IgG4. Thus, in some embodiments, it is contemplated that the Fc may be derived from the IgG1, IgG2, IgG3 or IgG4 subclass of an antibody.

[0119] Each heterodimer of a heterodimer pair can specifically bind to an epitope or antigen. In one embodiment, each heterodimer of a heterodimer pair binds to the same epitope. In another embodiment, the first heterodimer of a heterodimer pair specifically binds to an epitope on an antigen and the second heterodimer of a heterodimer pair specifically binds to a different epitope on the same antigen. In another embodiment, the first heterodimer of a heterodimer pair specifically binds to an epitope on a first antigen, and the second heterodimer of a heterodimer pair specifically binds to an epitope on a second antigen different from the first antigen. For example, in one embodiment, the first heterodimer specifically binds to tissue factor, and the second heterodimer specifically binds to the antigen Her2 (ErbB2). In another embodiment, the first heterodimer specifically binds to the above-mentioned molecule or cancer antigen. In another embodiment, the second heterodimer specifically binds to the above-mentioned molecule or cancer antigen. In another embodiment, the first heterodimer specifically binds to the antigen CD3, and the second heterodimer specifically binds to the antigen CD19.

[0120] As indicated above, in some embodiments, the immunoglobulin heavy chain and immunoglobulin light chain of each heterodimer can be derived from or engineered from a known therapeutic antibody, or from an antibody that binds to a different target molecule or cancer antigen. The amino acid and nucleotide sequences of many such molecules are readily available (see, e.g., GenBank: AJ308087.1 (humanized anti-human tissue factor antibody D3H44 light chain variable region and CL domain); GenBank: AJ308086.1 (humanized anti-human tissue factor antibody D3H44 heavy chain variable region and CH1 domain); GenBank: HC359025.1 (pertuzumab Fab light chain gene module); GenBank: HC359024.1 (pertuzumab Fab heavy chain gene module); GenBank: GM685465.1 (antibody trastuzumab Fab heavy chain gene module); GenBank: GM685465.1 (antibody trastuzumab Fab light chain gene module); GenBank: GM685466.1 (antibody trastuzumab Fab heavy chain gene module); GenBank: GM685467.1 (antibody trastuzumab Fab heavy chain gene module); GenBank: GM685468.1 (antibody trastuzumab Fab light chain gene module); GenBank: GM685469.1 (antibody trastuzumab Fab heavy ... The sequences of each of the GenBank numbers described herein are available from the NCBI website as of November 28, 2012, and are each incorporated by reference in their entirety for all purposes.

[0121] In some aspects, the separated antigen-binding construct comprises an amino acid sequence having at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to the amino acid sequence shown in the table or accession number disclosed herein, or a fragment thereof. In some aspects, the separated antigen-binding construct comprises an amino acid sequence encoded by a polynucleotide having at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to the nucleotide sequence shown in the table or accession number disclosed herein, or a fragment thereof.

[0122] Amino acid modifications to immunoglobulin heavy and light chains

[0123] At least one of the heterodimers of the heterodimer pair can include one or more amino acid modifications to their heavy chains of immunoglobulins and / or light chains of immunoglobulins, so that the heavy chain of the first heterodimer preferentially matches with one of the light chains rather than another. Similarly, the heavy chain of the second heterodimer preferentially matches with the second light chain rather than the first light chain. This preferential pairing of a heavy chain with one of two light chains can be based on a design group comprising a heavy chain of immunoglobulin and two light chains of immunoglobulins, wherein when the heavy chain of immunoglobulin and two light chains of immunoglobulins are co-expressed, the heavy chain of immunoglobulin preferentially matches with a light chain of two light chains of immunoglobulins relative to another light chain of two light chains of immunoglobulins. Therefore, the LCCA design group can include a heavy chain of immunoglobulin, the first light chain of immunoglobulin and the second light chain of immunoglobulin.

[0124] In one embodiment, the one or more amino acid modifications comprise one or more amino acid substitutions.

[0125] In one embodiment, the preferential pairing shown in the LCCA design group is established by modifying one or more amino acids as the part of the interface between the light chain and the heavy chain.In one embodiment, the preferential pairing shown in the LCCA design group is established by modifying one or more amino acids in at least one of the CH1 structure of the immunoglobulin heavy chain, the CL domain of the first immunoglobulin light chain and the CL domain of the second immunoglobulin light chain.

[0126] In one embodiment, one or more amino acid modifications are limited to the conserved framework residues of the variable (VH, VL) and constant (CH1, CL) domains as indicated by the Kabat numbering of the residues. For example, Almagro [Frontiers In Bioscience (2008) 13: 1619-1633] provides definitions of framework residues based on Kabat, Chotia and IMGT numbering schemes.

[0127] In one embodiment, at least one of the heterodimers comprises one or more mutations introduced into an immunoglobulin heavy chain and an immunoglobulin light chain that are complementary to each other. Complementarity at the interface of the heavy chain and the light chain can be achieved based on steric and hydrophobic contacts, electrostatic / charge interactions, or a combination of multiple interactions. Complementarity between protein surfaces is widely described in the literature by lock and key fits, button holes, protrusions and cavities, donors and acceptors, etc. (all implying the properties of structural and chemical matching between two interacting surfaces). In one embodiment, at least one of the heterodimers comprises one or more mutations, wherein the mutations introduced into the immunoglobulin heavy chain and the immunoglobulin light chain introduce new hydrogen bonds across the light chain and the heavy chain at the interface. In one embodiment, at least one of the heterodimers comprises one or more mutations, wherein the mutations introduced into the immunoglobulin heavy chain and the immunoglobulin light chain introduce new salt bridges across the light chain and the heavy chain at the interface.

[0128] The non-limiting examples of suitable LCCA design groups are shown in Table 1, showing amino acid modifications in one immunoglobulin heavy chain CH1 domain (H1) and two immunoglobulin light chain CL domains (L1 and L2) of a heterodimer, wherein when H1, L1 and L2 are co-expressed, H1 preferentially pairs with L1. The amino acid modifications shown in these LCCA design groups are based on the amino acid sequences of anti-tissue factor antibody D3H44 immunoglobulin heavy and light chains.

[0129] Table 1: LCCA design set with options for constant domain modifications of one immunoglobulin heavy chain (H1) and two immunoglobulin light chains L1 and L2, with H1 preferentially pairing with L1

[0130]

[0131] *Kabat number; # WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0132] Additional non-limiting examples of suitable LCCA design groups are shown in Table 2, showing amino acid modifications in one immunoglobulin heavy chain CH1 domain (H2) and two immunoglobulin light chain CL domains (L1 and L2) of a heterodimer, wherein when H2, L1 and L2 are co-expressed, H2 preferably pairs with L2:

[0133] Table 2: LCCA design set with selections for constant domain modifications of one immunoglobulin heavy chain (H2) and two immunoglobulin light chains L1 and L2, where H2 preferentially pairs with L2

[0134] Group No. H2_Mutation* L1_Mutation L2_Mutation C501 A139W_V190S WT# F116A C502 A139W_V190A WT F116A C504 F100W WT F98L C506 A139W F116S F118W_V133S C508 A139V F116A F118W_V133S C512 A139I WT F118W_V133S C516 A139W L135W_N137A F116A_L135A C518 A139W L135W F116A_L135A C520 A139W L135W F116A_L135V C522 WT WT S176V_T178L C526 F174V_P175S_S188G V133S S176L C528 F174V_S188L S176L WT C529 S188L S176L WT C531 K145T_Q179D_S188L Q124E_Q160E_T178D Q160K_T178R C533 K145T_Q179D_S188F Q124E_V133W_Q160E_T180E V133A_Q160K_T178R

[0135] *Kabat number; # WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0136] Additional non-limiting examples of suitable LCCA design sets are described in the Examples, Tables, and Figures.

[0137] In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least one amino acid modification in the CH1 domain, a first immunoglobulin light chain with at least one amino acid modification in the CL domain, and a second immunoglobulin light chain without any amino acid modification in the CL domain. In another embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least one amino acid modification in the CH1 domain, a first immunoglobulin light chain with at least one amino acid modification in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In another embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least one amino acid modification in the CH1 domain, a first immunoglobulin light chain with at least two amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the CL domain. In another embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least one amino acid modification in the CH1 domain, a first immunoglobulin light chain with at least two amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain.

[0138] In one embodiment, the LCCA design group includes an immunoglobulin heavy chain without amino acid modifications in the CH1 domain, a first immunoglobulin light chain without amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain without amino acid modifications in the CH1 domain, a first immunoglobulin light chain without amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the CL domain.

[0139] In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least two amino acid modifications in the CH1 domain, a first immunoglobulin light chain without amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least two amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least one amino acid modification in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least two amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least one amino acid modification in the CL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least two amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least two amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the CL domain. In one embodiment, the LCCA design set comprises an immunoglobulin heavy chain having at least two amino acid modifications in the CH1 domain, a first immunoglobulin light chain having at least three amino acid modifications in the CL domain, and a second immunoglobulin light chain having at least two amino acid modifications in the CL domain.

[0140] In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least three amino acid modifications in the CH1 domain, a first immunoglobulin light chain without amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least three amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least one amino acid modification in the CL domain, and a second immunoglobulin light chain with at least one amino acid modification in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least three amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least three amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the CL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain with at least three amino acid modifications in the CH1 domain, a first immunoglobulin light chain with at least four amino acid modifications in the CL domain, and a second immunoglobulin light chain with at least three amino acid modifications in the CL domain.

[0141] In one embodiment, the preferential pairing shown in the LCCA design group is set up by modifying one or more amino acids in at least one of the VH domain of a heavy chain immunoglobulin, the VL domain of a first light chain immunoglobulin, and the VL domain of a second light chain immunoglobulin. The limiting examples of suitable LCCA design groups are shown in Table 3, showing amino acid modifications in a heavy chain immunoglobulin VH domain (H1) and two light chain immunoglobulin VL domains (L1 and L2) of a heterodimer, wherein when H1, L1 and L2 are co-expressed, H1 preferentially pairs with L1:

[0142] Table 3: LCCA design set with selections for variable domain modifications of 1 immunoglobulin heavy chain and 2 immunoglobulin light chains, where H1 is preferably paired with L1

[0143] Group No. H1_Mutation* L1_Mutation L2_Mutation

[0144] V001 V37W_W103H F98L F98W V004 V37W_W103H F98L P44W V005 V37A_W103H P44W F98L V006 V37W_W103F F98L F98W V007 V37W F98A F98W V009 V37W F98A WT# V011 V37I WT F98L V013 V37A_W103V P44W F98A V015 V37A_W103H P44W F98A V016 V37A_W103H P44W F98W V020 L45W Y87 P44W V022 WT F98W F98A V023 WT WT F98A V024 Q39R Q38E F98A V026 Q39R Q38E WT V028 Q39R Q38E Q38R V030 Q39R Q38D Q38R V032 Q39M Q38M Q38E V034 Q39K Q38N_T85E Q38N_T85K V037 Q39E Q38R F98A V039 Q39D Q38R Q38D V040 V37E L89R_F98T WT V042 V37E_F100D L89R_F98W WT V044 V37E_F100D L89R_F98W F98

[0145] *Kabat numbering; #WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0146] Additional non-limiting examples of suitable LCCA design groups are described in Table 4, showing amino acid modifications in one immunoglobulin heavy chain VH domain (H2) and two immunoglobulin light chain VL domains (L1 and L2) of a heterodimer, wherein when H2, L1 and L2 are co-expressed, H2 preferentially pairs with L2:

[0147] Table 4: LCCA design set with selections for variable domain modifications of 1 immunoglobulin heavy chain and 2 immunoglobulin light chains, where H2 is preferably paired with L2

[0148] Group No. H2_Mutation* L1_Mutation L2_Mutation V002 V37I F98L F98W V003 WT# F98L F98W V005 V37A_W103H F98L P44W V007 V37W F98W F98A V008 V37I F98A F98W V009 V37W WT F98A V010 V37I F98A WT V012 F100W WT F98L

[0149] V014 V37W P44W F98A V017 V37A_W103V P44W F98W V018 V37A_W103V P44W F98L V019 V37I_F100W P44W F98L V021 V37A_W103H Y87 P44W V025 V37W Q38E F98A V027 WT Q38E WT V029 Q39E Q38E Q38R V030 Q39R Q38R Q38D V031 Q39E Q38D Q38R V033 Q39R Q38M Q38E V035 Q39D Q38N_T85E Q38N_T85K V036 Q39E Q38N_T85E Q38N_T85K V038 V37W Q38R F98A V041 WT L89R_F98T WT V043 WT L89R_F98W WT V045 V37S_A97K L89R_F98W F98

[0150] *Kabat numbering; #WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0151] In one embodiment, the LCCA design group includes an immunoglobulin heavy chain without amino acid modifications in the VH domain, a first immunoglobulin light chain without amino acid modifications in the VL domain, and a second immunoglobulin light chain with at least one amino acid modification in the VL domain. In one embodiment, the LCCA design group includes an immunoglobulin heavy chain without amino acid modifications in the VH domain, a first immunoglobulin light chain without amino acid modifications in the VL domain, and a second immunoglobulin light chain with at least two amino acid modifications in the VL domain.

[0152] In one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least one amino acid modification in the VH domain, the first immunoglobulin light chain without amino acid modification in the VL domain, and the second immunoglobulin light chain with at least one amino acid modification in the VL domain. In one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least one amino acid modification in the VH domain, the first immunoglobulin light chain with at least one amino acid modification in the VL domain, and the second immunoglobulin light chain with at least one amino acid modification in the VL domain. In one embodiment, in one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least one amino acid modification in the VH domain, the first immunoglobulin light chain with at least two amino acid modifications in the VL domain, and the second immunoglobulin light chain with at least two amino acid modifications in the VL domain.

[0153] In one embodiment, in one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least two amino acids modified in the VH domain, the first immunoglobulin light chain without amino acids modified in the VL domain and the second immunoglobulin light chain with at least one amino acid modified in the VL domain. In one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least two amino acids modified in the VH domain, the first immunoglobulin light chain with at least two amino acids modified in the VL domain and the second immunoglobulin light chain with at least one amino acid modified in the VL domain. In one embodiment, the LCCA design group is included in the immunoglobulin heavy chain with at least two amino acids modified in the VH domain, the first immunoglobulin light chain with at least one amino acids modified in the VL domain and the second immunoglobulin light chain with at least one amino acid modified in the VL domain.

[0154] In one embodiment, the LCCA design groups shown in Tables 1 to 4 are combined to provide a combination comprising two different immunoglobulin heavy chains (H1 and H2) and two different immunoglobulin light chains (L1 and L2), wherein when H1, H2, L1 and L2 are co-expressed, H1 preferentially pairs with L1 and H2 preferentially pairs with L2. In one embodiment, the LCCA design group comprising the CH1 domain of the heavy chain and / or the CL domain of the light chain from Table 1 is combined with a modified LCCA design group further comprising the CH1 domain of the heavy chain and / or the CL domain of the light chain from Table 2.

[0155] Non-limiting examples of design groups derived from combinations of LCCA design groups are shown in Table 5:

[0156] Table 5: Design groups containing constant domain modifications

[0157]

[0158]

[0159] *Kabat numbering; #WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0160] In one embodiment, a V comprising a heavy chain from Table 3 is H domain and / or light chain V L The LCCA design set of domains and the V from Table 4 also containing the heavy chain H domain and / or light chain V L The LCCA design group combination of the modification of the domain. The non-limiting example of the design group of such combination derived from the LCCA design group is shown in Table 6:

[0161] Table 6: Design groups containing variable domain modifications

[0162] Group No. Group No. H1_Mutation* L1_Mutation H2_Mutation L2_Mutation V001 V002 V37W_W103H F98L V37I F98W V001 V003 V37W_W103H F98L WT# F98W V004 V005 V37W_W103H F98L V37A_W103H P44W V006 V002 V37W_W103F F98L V37I F98W V006 V003 V37W_W103F F98L WT F98W V007 V008 V37W F98A V37I F98W V009 V010 V37W F98A V37I WT V011 V012 V37I WT F100W F98L V013 V014 V37A_W103V P44W V37W F98A V015 V014 V37A_W103H P44W V37W F98A V016 V017 V37A_W103H P44W V37A_W103V F98W V005 V018 V37A_W103H P44W V37A_W103V F98L V005 V019 V37A_W103H P44W V37I_F100W F98L V020 V021 L45W Y87 V37A_W103H P44W V022 V007 WT F98W V37W F98A V023 V009 WT WT V37W F98A V024 V025 Q39R Q38E V37W F98A V026 V027 Q39R Q38E WT WT

[0163] V028 V029 Q39R Q38E Q39E Q38R V030 V031 Q39R Q38D Q39E Q38R V032 V033 Q39M Q38M Q39R Q38E V034 V035 Q39K Q38N_T85E Q39D Q38N_T85K V034 V036 Q39K Q38N_T85E Q39E Q38N_T85K V037 V038 Q39E Q38R V37W F98A V039 V030 Q39D Q38R Q39R Q38D V040 V041 V37E L89R_F98T WT WT V042 V043 V37E_F100D L89R_F98W WT WT V044 V045 V37E_F100D L89R_F98W V37S_A97K F98

[0164] *Kabat numbering; #WT refers to the wild-type immunoglobulin chain without amino acid mutations

[0165] Portability of specific amino acid modifications identified herein to other antibodies:

[0166] Although the specific amino acid modifications to the immunoglobulin heavy and light chains identified above have been described in terms of the D3H44 anti-tissue factor extracellular domain antibody immunoglobulin heavy and light chains, it is contemplated and demonstrated herein (see the Examples, Figures and Tables) that these amino acid modifications can be transferred to other immunoglobulin heavy and light chains, thereby resulting in a similar pattern of preferential pairing of one immunoglobulin heavy chain with one of two immunoglobulin light chains based on the following.

[0167] The VH:VL and CH1:CL interface residues in the interface between the immunoglobulin heavy and light chains are relatively well conserved (Padlan et al., 1986, Mol. Immunol. 23(9):951-960). This sequence conservation, a result of evolutionary constraints, increases the likelihood that a functionally active antibody binding domain will form during the combinatorial pairing of the light and heavy chains. As a result of this sequence conservation, it can be concluded that the sequence modifications in the specific example noted above for the D3H44 sequence, which drive preferential pairing, can be transferred to other heavy and light chain pairs of heterodimers, obtaining roughly the same results for preferential pairing, because this region shows high sequence conservation between antibodies; in addition, when sequence differences exist, these modifications are generally located at the distal end of the CH1:CL interface. This is particularly true for the CH1 and CL domains. However, there is some sequence variability in the antigen binding site for the CDR (complementarity determining region) loop residues (and lengths), particularly for CDR-H3. Therefore, in one embodiment, when the amino acid sequence of the antigen binding site is significantly different from the amino acid sequence of the antigen binding site of D3H44 antibody, the heterodimer according to the present invention is to comprising a heterodimer, wherein at least one heterodimer comprises one or more amino acid modifications at the near end or distal end of the CDR loop in VH and / or VL domains. In another embodiment, when the amino acid sequence of the antigen binding site is substantially the same as the amino acid sequence of the antigen binding site of D3H44 antibody, the heterodimer according to the present invention is to comprising a heterodimer, wherein at least one heterodimer comprises one or more amino acid modifications at the near end or distal end of the CDR loop in VH and / or VL domains.

[0168] In one embodiment, the amino acid modifications described herein can be transferred to the immunoglobulin heavy and light chains of antibodies based on human or humanized IgG1 / κ. Non-limiting examples of such IgG1 / κ chains include orvacumab (for humans) or trastuzumab, pertuzumab or bevacizumab (for humanized).

[0169] In another embodiment, the amino acid modifications described herein can be transferred to the immunoglobulin heavy and light chains of antibodies utilizing the commonly used VH and VL subgroups. Non-limiting examples of such antibodies include Pertuzumab.

[0170] In one embodiment, the amino acid modifications described herein can be transferred to the immunoglobulin heavy and light chains of antibodies having a framework close to the germline. Examples of such antibodies include atezolizumab.

[0171] In one embodiment, the amino acid modifications described herein can be transferred to immunoglobulin heavy and light chains of antibodies having VH: VL interdomain angles close to the average values ​​observed for heavy and light chain pairs. Examples of antibodies of this type include, but are not limited to, pertuzumab. In another embodiment, the amino acid modifications described herein can be transferred to immunoglobulin heavy and light chains of antibodies having classical CL and CH1 domains. Suitable examples of such antibodies include, but are not limited to, trastuzumab.

[0172] In some embodiments, certain subsets of the amino acid modifications described herein are used in the variable domains in the antigen-binding constructs provided above.

[0173] The Examples, Figures and Tables show that amino acid modifications (e.g., within one or more Fab fragments comprising the variable and constant regions) can be transferred to other immunoglobulin heavy and light chains, resulting in a similar pattern of preferential pairing of one immunoglobulin heavy chain with one of two immunoglobulin light chains.

[0174] Priority Matching

[0175] As described above, at least one heterodimer of the antigen-binding construct / heterodimer pair according to the present invention may contain one or more amino acid modifications to their immunoglobulin heavy chains and / or immunoglobulin light chains, so that the heavy chain of one heterodimer, such as H1, preferentially pairs with one of the light chains, such as L1, rather than the other light chain L2, and the heavy chain H2 of the other heterodimer preferentially pairs with the light chain L2 rather than the light chain L1. In other words, the desired preferential pairing is considered to be between H1 and L1 and between H2 and L2. If the yield of H1-L1 heterodimers is greater than the yield of mismatched H1-L2 heterodimers when H1 is combined with L1 and L2, then the preferential pairing between, for example, H1 and L1 is considered to occur relative to the respective pairing of the corresponding H1 / L1 pair to the H2 / L2 pair without the one or more amino acid modifications. Similarly, if when H2 is combined with L1 and L2, the yield of H2-L2 heterodimers is greater than the yield of mispaired H2-L1 heterodimers, then the preferential pairing between H2 and L2 is considered to occur relative to the respective pairing of the corresponding H1 / L1 pair to the H2 / L2 pair without one or more amino acid modifications. In this context, heterodimers comprising H1 and L1 (H1-L1) or H2 and L2 (H2-L2) are referred to herein as preferentially paired, correctly paired, forcedly paired or required heterodimers, and heterodimers comprising H1 and L2 (H1-L2) or H2 and L1 (H2-L1) are referred to herein as mispaired heterodimers. The group of mutations corresponding to two heavy chains and two light chains intended to achieve the selective pairing of H1-L1 and H2-L2 is referred to as a design group.

[0176] Therefore, in one embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 55%. In another embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 60%. In another embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 70%. In another embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 80%. In another embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 90%. In another embodiment, when one heavy chain immunoglobulin of heterodimer is co-expressed with two light chains immunoglobulin, the relative yield of the desired heterodimer is greater than 95%.

[0177] In the above example, if the amount of H1-L1 heterodimer required is greater than the amount of mismatched H1-L2 heterodimer when H1 is co-expressed with L1 and L2, preferential pairing between H1-L1 is considered to occur. Similarly, if the amount of H2-L2 heterodimer required is greater than the amount of mismatched H2-L2 heterodimer when H2 is co-expressed with L1 and L2, preferential pairing between H2-L2 is considered to occur. Therefore, in one embodiment, when one immunoglobulin heavy chain of the heterodimer is co-expressed with two immunoglobulin light chains, the ratio of the desired heterodimer to the mismatched heterodimer is greater than 1.25:1. In one embodiment, when one immunoglobulin heavy chain of the heterodimer is co-expressed with two immunoglobulin light chains, the ratio of the desired heterodimer to the mismatched heterodimer is greater than 1.5:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimers to mismatched heterodimers is greater than 2:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimers to mismatched heterodimers is greater than 3:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimers to mismatched heterodimers is greater than 5:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimers to mismatched heterodimers is greater than 10:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimers to mismatched heterodimers is greater than 25:1. In another embodiment, when one immunoglobulin heavy chain of a heterodimer is co-expressed with two immunoglobulin light chains, the desired ratio of heterodimer to mismatched heterodimers is greater than 50:1.

[0178] Thermal stability of heterodimers

[0179] In one embodiment, each heterodimer of a heterodimer pair according to the present invention has a thermal stability comparable to the thermal stability of a heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to CH1, CL, VH or VL domains described herein. In one embodiment, thermal stability is determined by the measurement of melting temperature or or Tm. Therefore, in one embodiment, the thermal stability of a heterodimer according to the present invention is within about 10 ° C of the thermal stability of a heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to CH1, CL, VH or VL domains described herein. Therefore, in one embodiment, the thermal stability of a heterodimer according to the present invention is within about 5 ° C of the thermal stability of a heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 3°C ​​of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 2°C of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 1.5°C of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 1°C of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 0.5°C of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein. In another embodiment, the thermal stability of the heterodimer according to the present invention is within about 0.25°C of the thermal stability of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modifications to the CH1, CL, VH or VL domains described herein.

[0180] Affinity of heterodimer for antigen

[0181] In one embodiment, each heterodimer of a heterodimer pair has an affinity for its respective antigen that is the same or comparable to the affinity for its respective antigen of a heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In one embodiment, the heterodimer of a heterodimer pair has an affinity for its respective antigen within about 50 times or one order of magnitude of the affinity for its respective antigen of a heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In one embodiment, the heterodimer of a heterodimer pair has an affinity for its respective antigen within about 25 times or one order of magnitude of the affinity for its respective antigen of a heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In one embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen within about 10 times or one order of magnitude of the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In another embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen within about 5 times or one order of magnitude of the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In another embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen within about 2.5 times or one order of magnitude of the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy chain and light chain but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In another embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen within about 2 times or one order of magnitude of the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In another embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen within about 1.5 times or one order of magnitude of the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein. In another embodiment, the heterodimer of the heterodimer pair has an affinity for its respective antigen that is approximately the same as the affinity for its respective antigen of the heterodimer comprising the same immunoglobulin heavy and light chains but without the amino acid modified heterodimer to CH1, CL, VH or VL domains described herein.

[0182] Additional optional modifications

[0183] In one embodiment, the immunoglobulin heavy and light chains of the heterodimer pairs according to the invention may be further modified (i.e., by covalent attachment of different types of molecules) so that the covalent attachment does not interfere with the preferential pairing between the heavy and light chains or affect the ability of the heterodimer to bind its antigen, or affect its stability. Such modifications include, but are not limited to, restriction, glycosylation, acetylation, PEGylation, phosphorylation, amidation, derivatization with known protecting / blocking groups, proteolytic cleavage, attachment to cellular ligands or other proteins, and the like. Any of a number of chemical modifications may be performed by known techniques, including, but not limited to, specific chemical cleavage, acetylation, formylation, metabolic synthesis of tunicamycin, and the like.

[0184] In another embodiment, the immunoglobulin heavy and light chains of the heterodimer pair according to the invention can be conjugated to a therapeutic agent or drug moiety that modifies (directly or indirectly) a given biological response. The therapeutic agent or drug moiety will not be construed as being limited to classical chemotherapeutic agents. For example, the drug moiety can be a protein or polypeptide having a desired biological activity. Such proteins may include, for example, toxins such as abrin, ricin A, onconase (or another cytotoxic RNase), Pseudomonas exotoxin, cholera toxin, or diphtheria toxin; proteins such as tumor necrosis factor, interferon alpha, interferon beta, nerve growth factor, platelet-derived growth factor, tissue plasminogen activator, apoptotic agents, e.g., TNF-α, TNF-β, AIM I (see International Publication No. WO 97 / 33899), AIM II (see International Publication No. WO 97 / 34911), and Fas ligand (Takahashi et al., 1994, J. Immunol., 6:1567), and VEGI (see International Publication No. WO 97 / 34911). 99 / 23105), thrombotic or anti-angiogenic agents, such as angiostatin or endostatin; or biological response modifiers, such as, for example, lymphokines (e.g., interleukin-1 ("IL-1"), interleukin-2 ("IL-2"), interleukin-6 ("IL-6"), granulocyte macrophage colony stimulating factor ("GM-CSF") and granulocyte colony stimulating factor ("G-CSF")) or growth factors (e.g., growth hormone ("GH")).

[0185] In addition, in alternative embodiments, the antibody can be conjugated to a therapeutic moiety such as a radioactive material or a macrocyclic chelator for conjugating radioactive metal ions (see above for examples of radioactive materials). In certain embodiments, the macrocyclic chelator is 1,4,7,10-tetraaza-N,N′,N″,N″-tetraacetic acid (DOTA) that can be attached to the antibody via a linker molecule. Such linker molecules are well known in the art and are described in Denardo et al., 1998, Clin Cancer Res. 4:2483; Peterson et al., 1999, Bioconjug. Chem. 10:553 and Zimmerman et al., 1999, Nucl. Med. Biol. 26:943.

[0186] In some embodiments, the immunoglobulin heavy chain and light chain of the heterodimer are expressed as a fusion protein containing a tag to facilitate purification and / or testing, etc. As mentioned herein, a "tag" is any added series of amino acids provided in a protein at the C-terminus, N-terminus, or internally, which facilitates the identification or purification of the protein. Suitable tags include, but are not limited to, tags known to those of ordinary skill in the art for purification and / or testing. Such as albumin binding domain (ABD), His tag, FLAG tag, glutathione S-transferase, hemagglutinin (HA), and maltose binding protein. Such labeled proteins can also be engineered to include a cleavage site, such as a thrombin, enterokinase, or factor X cleavage site, to facilitate removal of the tag before, during, or after purification.

[0187] In some embodiments, one or more cysteine ​​residues forming interchain disulfide bonds on the bottom of the Fab domain of the light chain (position 214, Kabat numbering) and the heavy chain (position 233, Kabat numbering) can be modified to serine or alanine or non-cysteine ​​or a different amino acid.

[0188] It is expected that additional amino acid modifications may be made to the immunoglobulin heavy chain to increase the level of preferential pairing and / or enhance the thermal stability of the heterodimer pair. For example, additional amino acid modifications may be made to the immunoglobulin heavy chain Fc domain to drive preferential pairing between heterodimer pairs relative to homodimer pairs. Such amino acid modifications are known in the art, including, for example, those described in U.S. Patent Publication No. 2012 / 0149876. Alternatively, alternative strategies to drive preferential pairing between heterodimer pairs relative to homodimer pairs may also be applied, such as "buttonholes", charged residues with ionic interactions, and chain exchange engineered domain (SEED) technology. The latter strategy has been described in the art and reviewed in Klein et al. (supra). Further discussion of the Fc domain is provided below.

[0189] Fc domain

[0190] The constructs described herein may also include an Fc. In some aspects, the Fc comprises at least one or two C H3 In some aspects, the Fc is coupled to the first heterodimer and / or the second heterodimer via one or more linkers. In some aspects, the Fc is a human Fc. In some aspects, the Fc is a human IgG or IgG1 Fc. In some aspects, the Fc is a heterodimer Fc. In some aspects, the Fc comprises at least one or two C H2 Domain sequence.

[0191] In some aspects, the Fc is in C H3 In some aspects, the Fc comprises one or more modifications in at least one of the C H2 In some aspects, the Fc is a single polypeptide. In some aspects, the Fc is a plurality of peptides, such as two polypeptides.

[0192] In some aspects, the Fc is in C H3 In some aspects, the Fc comprises one or more modifications in at least one of the C H2 In some aspects, Fc is a single polypeptide. In some aspects, Fc is a plurality of peptides, such as two polypeptides.

[0193] In some aspects, the Fc is an Fc described in patent application PCT / CA2011 / 001238 filed on November 4, 2011 or PCT / CA2012 / 050780 filed on November 2, 2012 (the entire disclosure of each of which is hereby incorporated by reference in its entirety for all purposes).

[0194] In some aspects, the constructs described herein comprise a heterodimeric Fc comprising a modified CH3 domain that has been asymmetrically modified. The heterodimeric Fc may comprise two heavy chain constant domain polypeptides: a first heavy chain polypeptide and a second heavy chain polypeptide, which may be used interchangeably as long as the Fc comprises a first heavy chain polypeptide and a second heavy chain polypeptide. Typically, the first heavy chain polypeptide comprises a first CH3 sequence, and the second chain polypeptide comprises a second CH3 sequence.

[0195] When two CH3 sequences containing one or more amino acid modifications introduced in an asymmetric manner dimerize, the two CH3 sequences generally result in a heterodimer Fc rather than a homodimer. As used herein, "asymmetric amino acid modification" refers to any modification in which the amino acid at a specific position on the first CH3 sequence is different from the amino acid at the same position on the second CH3 sequence, and the first and second CH3 sequences preferentially pair to form a heterodimer rather than a homodimer. The heterodimerization can be the result of modification of only one of the two amino acids at the same respective amino acid position on each sequence; or the result of modification of two amino acids on each sequence at the same respective position on each of the first and second CH3 sequences. The first and second CH3 sequences of the heterodimer Fc may contain one or more asymmetric amino acid modifications.

[0196] Table X provides the amino acid sequence of the human IgGl Fc sequence corresponding to amino acids 231 to 447 of a full-length human IgGl heavy chain. The CH3 sequence comprises amino acids 341-447 of a full-length human IgGl heavy chain.

[0197] Typically, Fc may include two continuous heavy chain sequences (A and B) capable of dimerization. In some aspects, one or two sequences of Fc include one or more mutations or modifications at the following positions: L351, F405, Y407, T366, K392, T394, T350, S400 and / or N390 (using EU numbering). In some aspects, Fc includes the mutant sequences shown in Table X. In some aspects, Fc includes the mutation of variant 1A-B. In some aspects, Fc includes the mutation of variant 2A-B. In some aspects, Fc includes the mutation of variant 3A-B. In some aspects, include the mutation of variant 4A-B. In some aspects, include the mutation of variant 5A-B.

[0198]

[0199] The first and second CH3 sequences may comprise amino acid mutations as described herein, with reference to amino acids 231 to 447 of a full-length human IgG1 heavy chain. In one embodiment, the heterodimer Fc comprises a first CH3 sequence having amino acid modifications at positions F405 and Y407 and a second CH3 sequence having amino acid modifications at position T394. In one embodiment, the heterodimer Fc comprises a first CH3 sequence having one or more amino acid modifications selected from L351Y, F405A, and Y407V and a second CH3 sequence having one or more amino acid modifications selected from T366L, T366I, K392L, K392M, and T394W. The modified CH3 domain.

[0200] In one embodiment, the heterodimer Fc comprises a modified CH3 domain, the modified CH3 domain having a first CH3 sequence with amino acid modifications at positions L351, F405 and Y407 and a second CH3 sequence with amino acid modifications at positions T366, K392 and T394, and one of the first or second CH3 sequences also comprises an amino acid modification at position Q347, and the other CH3 sequence also comprises an amino acid modification at position K360. In another embodiment, the heterodimer Fc comprises a modified CH3 domain, the modified CH3 domain having a first CH3 sequence with amino acid modifications at positions L351, F405 and Y407, and a second CH3 sequence with amino acid modifications at positions T366, K392 and T394, one of the first or second CH3 sequences also comprises an amino acid modification at position Q347, and the other CH3 sequence also comprises an amino acid modification at position K360, and one or both of the CH3 sequences also comprise an amino acid modification T350V.

[0201] In one embodiment, the heterodimeric Fc comprises a modified CH3 domain having a first CH3 sequence having amino acid modifications at positions L351, F405, and Y407 and a second CH3 sequence having amino acid modifications at positions T366, K392, and T394, and one of the first and second CH3 sequences further comprises an amino acid modification of D399R or D399K, and the other CH3 sequence comprises one or more of T411E, T411D, K409E, K409D, K392E, and K392D. In another embodiment, the heterodimeric Fc comprises a modified CH3 domain having a first CH3 sequence having amino acid modifications at positions L351, F405, and Y407 and a second CH3 sequence having amino acid modifications at positions T366, K392, and T394, and one of the first and second CH3 sequences further comprises an amino acid modification of D399R or D399K, the other CH3 sequence comprises one or more of T411E, T411D, K409E, K409D, K392E, and K392D, and one or both of the CH3 sequences further comprises amino acid modification T350V.

[0202] In one embodiment, the heterodimeric Fc comprises a modified CH3 domain having a first CH3 sequence having amino acid modifications at positions L351, F405 and Y407 and a second CH3 sequence having amino acid modifications at positions T366, K392 and T394, wherein one or both of the CH3 sequences further comprises an amino acid modification of T350V.

[0203] In one embodiment, the heterodimeric Fc comprises a modified CH3 domain, wherein the modified CH3 domain comprises the following amino acid modifications, wherein "A" represents an amino acid modification to the first CH3 sequence, and "B" represents an amino acid modification to the second CH3 sequence: A: L351Y_F405A_Y407V, B: T366L_K392M_T394W, A: L351Y_F405A_Y407V, B: T366L_K392L_T394W 4W, A: T350V_L351Y_F405A_Y407V, B: T350V_T366L_K392L_T394W, A: T350V_L351Y_F405A_Y407V, B: T350 V_T366L_K392M_T394W, A: T350V_L351Y_S400E_F405A_Y407V, and / or B: T350V_T366L_N390R_K392M_T394W.

[0204] One or more asymmetric amino acid modifications may promote the formation of heterodimeric Fc, wherein the heterodimeric CH3 domain has a stability comparable to that of a wild-type homodimeric CH3 domain. In an embodiment, one or more asymmetric amino acid modifications promote the formation of heterodimeric Fc domains, wherein the heterodimeric Fc domain has a stability comparable to that of a wild-type homodimeric Fc domain. In an embodiment, one or more asymmetric amino acid modifications promote the formation of heterodimeric Fc domains, wherein the heterodimeric Fc domain has a stability observed by a melting temperature (Tm) in a differential scanning calorimetry study, and wherein the melting temperature is within 4°C of the melting temperature observed for the corresponding symmetric wild-type homodimeric Fc domain. In some aspects, the Fc is at C H3 At least one of the sequences comprises one or more modifications that promote the formation of a heterodimeric Fc having a stability comparable to that of a wild-type homodimeric Fc.

[0205] In one embodiment, the stability of the CH3 domain can be assessed by, for example, measuring the melting temperature of the CH3 domain using differential scanning calorimetry (DSC). Thus, in other embodiments, the CH3 domain has a melting temperature of about 68°C or higher. In another embodiment, the CH3 domain has a melting temperature of about 70°C or higher. In another embodiment, the CH3 domain has a melting temperature of about 72°C or higher. In another embodiment, the CH3 domain has a melting temperature of about 73°C or higher. In another embodiment, the CH3 domain has a melting temperature of about 75°C or higher. In another embodiment, the CH3 domain has a melting temperature of about 78°C or higher. In some aspects, the dimerized CH3 domain has a melting temperature of about 75°C or higher.H3 The sequence has a melting temperature (Tm) of about 68°C, 69°C, 70°C, 71°C, 72°C, 73°C, 74°C, 75°C, 76°C, 77°C, 77.5°C, 78°C, 79°C, 80°C, 81°C, 82°C, 83°C, 84°C or 85°C or more.

[0206] In some embodiments, the heterodimer Fc comprising a modified CH3 sequence can be formed in the expressed product at a purity of at least about 75% compared to the homodimer Fc. In another embodiment, the heterodimer Fc is formed with a purity greater than about 80%. In another embodiment, the heterodimer Fc is formed with a purity greater than about 85%. In another embodiment, the heterodimer Fc is formed with a purity greater than about 90%. In another embodiment, the heterodimer Fc is formed with a purity greater than about 95%. In another embodiment, the heterodimer Fc is formed with a purity greater than about 97%. In some aspects, Fc is a heterodimer formed with a purity greater than about 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% when expressed. In some aspects, Fc is a heterodimer formed with a purity greater than about 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% when expressed by a single cell.

[0207] Additional methods for modifying monomeric Fc polypeptides to promote heterodimeric Fc formation are described in International Patent Publication No. WO 96 / 027011 (Buttonhole), Gunasekaran et al. (Gunasekaran K. et al. (2010) J Biol Chem. 285, 19637-46, electrostatic design to achieve selective heterodimerization), Davis et al. (Davis, JH. et al. (2010) Prot Eng Des Sel; 23(4): 195-202, strand exchange engineered domain (SEED) technology) and Labrijn et al. [Efficient generation of stable bispecific IgG1 by controlled Fab-arm exchange. Labrijn AF, Meesters JI, de Goeij BE, van den Bremer ET, Neijssen J, van Kampen MD, Strumane K, Verploegen S, Kundu A, Gramer MJ, van Berkel PH, van deWinkel JG, Schuurman J, Parren PW. Proc Natl Acad Sci US A. 2013 Mar 26;110(13):5145-50.

[0208] In some embodiments, the isolated constructs described herein include antigen-binding constructs that bind to antigens; and dimeric Fc polypeptide constructs that have superior biophysical properties such as stability and ease of manufacture relative to antigen-binding constructs that do not include the same Fc polypeptide. Many mutations in the heavy chain sequence of Fc are known in the art to selectively change the affinity of antibody Fc for different Fcγ receptors. In some aspects, the Fc comprises one or more modifications to promote selective binding of Fc-γ receptors.

[0209] The CH2 domain is amino acids 231-340 of the sequence shown in Table X. Exemplary mutations are listed below:

[0210] S298A / E333A / K334A, S298A / E333A / K334A / K326A (Lu Y, Vernes JM, Chiang N, et al. J Immunol Methods. 2011 Feb 28;365(1-2):132-41);

[0211] F243L / R292P / Y300L / V305I / P396L, F243L / R292P / Y300L / L235V / P396L (StaVenhagen JB, Gorlatov S, Tuaillon N, et al. Cancer Res. 2007 Sep 15;67(18):8882-90; Nordstrom JL, Gorlatov S, Zhang W, et al. Breast Cancer Res. 2011 Nov 30;13(6):R123);

[0212] F243L (Stewart R, Thom G, Levens M, et al. Protein Eng Des Sel. 2011 Sep;24(9):671-8.), S298A / E333A / K334A (Shields RL, Namenuk AK, Hong K, et al. J Biol Chem. 2001 Mar 2;276(9):6591-604);

[0213] S239D / I332E / A330L, S239D / I332E (Lazar GA, Dang W, Karki S, et al. Proc Natl Acad Sci US A. 2006 Mar 14; 103(11):4005-10);

[0214] S239D / S267E, S267E / L328F (Chu SY, Vostiar I, Karki S, et al. Mol Immunol. 2008 Sep; 45(15): 3926-33);

[0215] S239D / D265S / S298A / I332E, S239E / S298A / K326A / A327H, G237F / S298A / A330L / I332E, S239D / I332E / S298A, S239D / K326E / A330L / I332E / S298A, G236A / S239D / D270L / I332E, S239E / S267E / H268D, L234F / S267E / N325L, G237F / V266L / S267D and other mutations listed in WO2011 / 120134 and WO2011 / 120135 (incorporated herein by reference). Therapeutic Antibody Engineering (edited by William R. Strohl and Lila M. Strohl, Woodhead Publishing series in Biomedicine No 11, ISBN 1 907568 379, October 2012) lists mutations on page 283.

[0216] In some embodiments, the CH2 domain comprises one or more asymmetric amino acid modifications. In some embodiments, the CH2 domain comprises one or more asymmetric amino acid modifications to promote selective binding of FcγRs. In some embodiments, the CH2 domain allows for the isolation and purification of the isolated constructs described herein.

[0217] FcRn binding and PK parameters

[0218] As is known in the art, binding to FcRn recycles endocytosed antibodies from endosomes back into the blood (Raghavan et al., 1996, Annu Rev Cell Dev Biol 12: 181-220; Ghetie et al., 2000, Annu Rev Immunol 18: 739-766). This process, coupled with exclusion of kidney filtration due to the large size of the full-length molecule, results in a favorable antibody serum half-life ranging from 1 to 3 weeks. Fc binding to FcRn also plays a crucial role in antibody trafficking. Thus, in one embodiment, the constructs of the invention are capable of binding to FcRn.

[0219] Additional modifications to enhance effector function

[0220] In some embodiments, the constructs described herein may be modified to enhance their effector function. Such modifications are known in the art and include afucosylation or engineering of the affinity of the Fc portion of the antibody for activating receptors, primarily FCGR3a (for ADCC), and for C1q (for CDC). Table Y below summarizes various designs reported herein for engineering effector function.

[0221]

[0222]

[0223] Thus, in one embodiment, the constructs described herein may include a dimeric Fc comprising one or more of the amino acid modifications indicated in the table above that confer enhanced effector function. In another embodiment, the constructs may be afucosylated to enhance effector function.

[0224] Connectors

[0225] Construct described herein can include one or more heterodimers described herein that are operably coupled to Fc described herein. In some aspects, Fc is coupled to one or more heterodimers by or without one or more joints. In some aspects, Fc is directly coupled to one or more heterodimers. In some aspects, Fc is coupled to one or more heterodimers by one or more joints. In some aspects, Fc is coupled to the heavy chain of each heterodimer by a joint.

[0226] In some aspects, the one or more linkers are one or more polypeptide linkers. In some aspects, the one or more linkers are one or more antibody hinge regions. In some aspects, the one or more linkers comprise one or more IgG1 hinge regions.

[0227] Methods for preparing heterodimer pairs

[0228] As described above, the heterodimer pair according to the present invention can include the first heterodimer and the second heterodimer, each heterodimer including an immunoglobulin heavy chain or a fragment thereof having at least a VH and a CH1 domain, and an immunoglobulin light chain having a VL domain and a CL domain. The immunoglobulin heavy chain and the immunoglobulin light chain of the heterodimer can be easily prepared using recombinant DNA techniques known in the art. Standard techniques such as, for example, Sambrook and Russell, Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 3rd edition, 2001); Sambrook et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2nd edition, 1989); Short Protocols in Molecular Biology (Ausubel et al., John Wiley and Sons, New York, 4th edition, 1999); and Glick and Pasternak, Molecular Biotechnology: Principles and Applications of Recombinant DNA (ASM Press, Washington, DC, 2nd edition, 1998) can be used for recombinant nucleic acid methods, nucleic acid synthesis, cell culture, transgene incorporation, and recombinant protein expression. Alternatively, heterodimers and heterodimer pairs according to the invention can be chemically synthesized.

[0229] The nucleic acid and amino acid sequences of the immunoglobulin heavy and light chains of the antibodies from which the heterodimers are derived are known in the art or can be easily determined using nucleic acid or protein sequencing methods. Methods for fusing the tag genes described herein to immunoglobulin heavy and / or light chains are known in the art, and some methods are described below and in the Examples.

[0230] For example, methods for expressing and co-expressing immunoglobulin heavy and light chains in host cells are well known in the art. In addition, methods for labeling heavy and / or light chains using recombinant DNA techniques are also well known in the art. Expression vectors and host cells suitable for expression of heavy and light chains are also well known in the art, as described below.

[0231] Bispecific antibody production methods that do not rely on the use of only a single clone or transient cell line expressing all four chains are known in the art (Gramer, et al. (2013) mAbs 5, 962; Strop et al. (2012) J Mol Biol 420, 204.). These methods rely on the generation of two pairs of light chains and heavy chains involved in the formation of bispecific antibodies (redox generation) under redox conditions and then arm exchange. In this method, H1: L1 and H2: L2 pairs are expressed in two different cell lines to independently produce two Fab arms. Subsequently, the two Fab arms are mixed under selected redox conditions to achieve the reassociation of two unique heavy chains H1 and H2 to form a bispecific antibody comprising L1: H1: H2: L2 chains. It is conceivable that the designed library / data set described herein is used in the generation of bispecific antibodies using redox generation methods or improved forms of the method.

[0232] In certain embodiments, a cell-free protein expression system is used to co-express polypeptides (e.g., heavy and light chain polypeptides) without the use of living cells. Instead, all components required for DNA transcription into RNA and RNA translation into protein (e.g., ribosomes, tRNA, enzymes, cofactors, amino acids) are provided in solution for use in vitro. In certain embodiments, in vitro expression requires (1) a genetic template (mRNA or DNA) encoding heavy and light chain polypeptides and (2) a reaction solution containing the necessary transcription and translation molecular machinery. In certain embodiments, a cell extract generally provides the components of the reaction solution, e.g., RNA polymers for mRNA transcription, ribosomes, tRNA, amino acids, enzyme cofactors, energy sources, and cellular components necessary for correct protein folding for polypeptide translation. Lysates derived from bacterial cells, yeast cells, insect cells, plant cells, mammalian cells, human cells, or combinations thereof can be used to prepare a cell-free protein expression system. Such cell lysates can provide the correct composition and proportion of enzymes and building blocks required for translation. In some embodiments, the cell membrane is removed to leave only the cytoplasm and organelle components of the cell.

[0233] Several cell-free protein expression systems are known in the art, as reviewed in Carlson et al. (2012) Biotechnol. Adv. 30: 1185-1194. For example, cell-free protein expression systems are obtainable from prokaryotic or eukaryotic cells. Examples of prokaryotic cell-free expression systems include those from Escherichia coli (E. coli.). Eukaryotic cell-free protein expression systems are obtainable based on extracts from, for example, rabbit reticulocytes, wheat germ, and insect cells. Such prokaryotic and eukaryotic cell-free protein expression systems are commercially available from companies such as Roche, Invitrogen, Qiagen, and Novagen. One of ordinary skill in the art can easily select a suitable cell-free protein expression system that can produce polypeptides (e.g., heavy chain and light chain polypeptides) that can pair with each other. In addition, the cell-free protein expression system can also be supplemented with chaperone molecules (e.g., BiP) and isomerases (e.g., disulfide isomerases) to improve the efficiency of IgG folding.

[0234] In some embodiments, a cell-free expression system is used to co-express heavy and light chain polypeptides from a DNA template (transcription and translation) or an mRNA template (translation only).

[0235] Vectors and host cells

[0236] The recombinant expression of heavy and light chains requires the construction of an expression vector containing a polynucleotide encoding a heavy chain or light chain (e.g., an antibody or a fusion protein). After obtaining a polynucleotide encoding a heavy chain or a light chain, a vector for producing a heavy chain or a light chain can be produced by recombinant DNA technology using techniques known in the art. Therefore, the method for preparing a protein by expressing a polynucleotide containing a heavy chain or a light chain encoding nucleotide sequence is described herein. Methods known to those of ordinary skill in the art can be used to construct an expression vector containing a heavy chain or a light chain encoding sequence and appropriate transcription and translation control signals. These methods include, for example, in vitro recombinant DNA technology, synthetic techniques, and in vivo genetic recombination. Therefore, the invention provides a reproducible vector comprising a nucleotide sequence encoding a heavy chain or a light chain operably connected to a promoter.

[0237] The expression vector is transferred to a host cell by conventional techniques, and the transfected cells are subsequently cultured by conventional techniques to produce a modified heavy or light chain for use in the methods of the invention. In a specific embodiment, the heavy and light chains used in the methods are co-expressed in a host cell for expressing a complete immunoglobulin molecule, as described in detail below.

[0238] A variety of host-expression vector systems can be used to express modified heavy and light chains. Such host-expression systems represent vehicles by which the target coding sequences can be produced and subsequently purified, and also represent cells that can express the modified heavy and light chains in situ when transformed or transfected with appropriate nucleotide coding sequences. These cells include, but are not limited to, microorganisms such as bacteria (e.g., E. coli and Bacillus subtilis) transformed with recombinant bacteriophage DNA, plasmid DNA, or cosmid DNA expression vectors containing modified heavy and light chain coding sequences; yeast (e.g., Saccharomyces Pichia) transformed with recombinant yeast expression vectors containing modified heavy and light chain coding sequences; insect cell systems infected with recombinant viral expression vectors (e.g., baculovirus) containing modified heavy and light chain coding sequences; or cells infected with recombinant Viral expression vectors (e.g., cauliflower mosaic virus, CaMV; tobacco mosaic virus, TMV) infected or transformed with recombinant plasmid expression vectors (e.g., Ti plasmid) containing modified heavy and light chain coding sequences; or mammalian cell systems (e.g., COS, CHO, BHK, HEK-293, NSO and 3T3 cells) with recombinant expression constructs containing promoters derived from the genome of mammalian cells (e.g., metallothionein promoter) or from mammalian viruses (e.g., adenovirus late promoter; vaccinia virus 7.5K promoter). In certain embodiments, bacterial cells such as Escherichia coli, or eukaryotic cells are used to express modified heavy and light chains, which are recombinant antibodies or fusion protein molecules. For example, mammalian cells such as Chinese hamster ovary cells (CHO), combined with vectors such as the major intermediate early gene promoter element from human cytomegalovirus, are effective expression systems for antibodies (Foecking et al., 1986, Gene 45:101; and Cockett et al., 1990, Bio / Technology 8:2). In specific embodiments, the expression of the nucleotide sequences encoding the immunoglobulin heavy and light chains of each heterodimer is regulated by a constitutive promoter, an inducible promoter, or a tissue-specific promoter.

[0239] In mammalian host cells, many virus-based expression systems can be used. In the case where adenovirus is used as an expression vector, the modified target heavy and light chain coding sequences can be connected to adenovirus transcription / translation control complexes, such as late promoters and tripartite leader sequences. The chimeric gene can then be inserted into the adenovirus genome by in vitro or in vivo recombination. Insertion of the viral genome (e.g., region E1 or E3) in a non-essential region will result in a recombinant virus that can survive in an infected host and can express the modified heavy and light chains (e.g., see Logan & Shenk, 1984, Proc. Natl. Acad. Sci. USA 81: 355-359). Specific initiation signals may also be required for efficient translation of the inserted antibody coding sequence. These signals include the ATG initiation codon and adjacent sequences. In addition, the initiation codon must be in phase with the reading frame of the desired coding sequence to ensure translation of the complete insert. These exogenous translation control signals and initiation codons can have a variety of sources (natural and synthetic). Expression efficiency can be enhanced by including appropriate transcription enhancer elements, transcription terminators, etc. (see, e.g., Bittner et al., 1987, Methods in Enzymol. 153: 516-544).

[0240] The expression of the immunoglobulin heavy and light chains of the heterodimers can be controlled by any promoter or enhancer element known in the art. Promoters that can be used to control the expression of genes encoding modified heavy and light chains (e.g., antibodies or fusion proteins) include, but are not limited to, the SV40 early promoter region (Bernoist and Chambon, 1981, Nature 290: 304-310), the promoter contained in the 3′ long terminal repeat of Rous sarcoma virus (Yamamoto et al., 1980, Cell 22: 787-797), the herpes thymidine kinase promoter (Wagner et al., 1981, Proc. Natl. Acad. Sci. USA 78: 1441-1445), the regulatory sequence of the metallothionein gene (Brinster et al., 1982, Nature 296: 39-42), the tetracycline (Tet) promoter (Gossen et al., 1995, Proc. Nat. Acad. Sci. USA 89:5547-5551); prokaryotic expression vectors such as the β-lactamase promoter (Villa-Kamaroff et al., 1978, Proc. Natl. Acad. Sci. USA 75:3727-3731) or the tac promoter (DeBoer et al., 1983, Proc. Natl. Acad. Sci. USA 80:21-25; see also "Useful proteins from recombinant bacteria" in Scientific American, 1980, 242:74-94); containing the nopaline synthase promoter region (Herrera-Estrella et al., Nature 303:209-213) or the cauliflower mosaic virus 35S RNA promoter (Gardner et al., 1981, Nucl. Acids Res.9:2871) and the promoter of the photosynthetic enzyme ribulose bisphosphate carboxylase (Herrera-Estrella et al., 1984, Nature 310:115-120); promoter elements from yeast or other fungi such as the Gal 4 promoter, ADC (alcohol dehydrogenase) promoter, PGK (phosphoglycerol kinase) promoter, alkaline phosphatase promoter, and the following animal transcription control regions, which show tissue specificity and have been used in transgenic animals: the control region of the elastase I gene active in pancreatic cancer cells (Swift et al., 1984, Cell 38:639-646; Ornitz et al., 1986, Cold Spring Harbor Symp. Quant. Biol.50:399-409; MacDonald, 1987, Hepatology 7:425-515); the insulin gene control region active in pancreatic β cells (Hanahan, 1985, Nature 315:115-122), the immunoglobulin gene control region active in lymphoid cells (Grosschedl et al., 1984, Cell 38:647-658; Adames et al., 1985, Nature 318:533-538; Alexander et al., 1987, Mol. Cell. Biol. 7:1436-1444), the mouse mammary tumor virus control region active in testicular, mammary, lymphoid and mast cells (Leder et al., 1986, Cell 45:485-495), the albumin gene control region active in liver (Pinkert et al., 1987, Genes and Cells 45:485-495), the 1:268-276), the alpha-fetoprotein gene control region active in liver (Krumlauf et al., 1985, Mol. Cell. Biol. 5:1639-1648; Hammer et al., 1987, Science 235:53-58; the alpha1-antitrypsin gene control region active in liver (Kelsey et al., 1987, Genes and Devel. 1:161-171), the beta-globin gene control region active in bone marrow cells (Mogram et al., 1985, Nature 315:338-340; Kollias et al., 1986, Cell 46:89-94; the myelin basic protein gene control region active in oligodendrocytes of the brain (Readhead et al., 1987, Cell 48:703-712); myosin light chain-2 gene control region active in skeletal muscle (Sani, 1985, Nature 314:283-286); neuron-specific enolase (NSE) active in neurons (Morelli et al., 1999, Gen. Virol. 80:571-83); brain-derived neurotrophic factor (BDNF) gene control region active in neurons (Tabuchi et al., 1998, Biochem. Biophysic. Res. Com. 253:818-823); glial fibrillary acidic protein (GFAP) promoter active in astrocytes (Gomes et al., 1999, Braz J Med Biol Res 32(5):619-631; Morelli et al., 1999, Gen. Virol.80:571-83) and a gonadotropin-releasing hormone gene control region active in the hypothalamus (Mason et al., 1986, Science 234:1372-1378).

[0241] In addition, host cell strains that regulate the expression of inserted sequences or modify or process gene products in a desired specific manner can be selected. Expression from certain promoters can be increased in the presence of certain inducers; therefore, the expression of genetically engineered fusion proteins can be controlled. In addition, different host cells have characteristic and specific mechanisms for transcriptional and post-translational processing and modification (e.g., glycosylation, phosphorylation of proteins). Appropriate cell lines or host systems can be selected to ensure the desired modification and processing of the expressed exogenous proteins. For example, expression in bacterial systems will produce non-glycosylated products, and expression in yeast will produce glycosylated products. Eukaryotic host cells with cellular mechanisms for the correct processing (e.g., glycosylation and phosphorylation) of the primary transcripts of gene products can be used.Such mammalian host cells include, but are not limited to, CHO, VERY, BHK, Hela, COS, MDCK, HEK-293, 3T3, WI38, NSO, in particular neuronal cell lines such as, for example, SK-N-AS, SK-N-FI, SK-N-DZ human neuroblastoma (Sugimoto et al., 1984, J. Natl. Cancer Inst. 73:51-57), SK-N-SH human neuroblastoma (Biochim. Biophys. Acta, 1982, 704:450-460), Daoy human brain cell medulloblastoma (He et al., 1992, Cancer Res. 52:1144-1148), DBTRG-05MG glioblastoma cells (Kruse et al., 1992, In Vitro Cell. Dev. Biol. 28A:609-614), IMR-32 human neuroblastoma (Cancer Res., 1970, 30: 2110-2118), 1321N1 human astrocytoma (Proc. Natl. Acad. Sci. USA, 1977, 74: 4816), MOG-G-CCM human astrocytoma (Br. J. Cancer, 1984, 49: 269), U87MG human glioblastoma-astrocytoma (Acta Pathol. Microbiol. Scand., 1968, 74: 465-486), A172 human glioblastoma (Olopade et al., 1992, Cancer Res.52:2523-2529), C6 rat glioma cells (Benda et al., 1968, Science161:370-371), Neuro-2a mouse neuroblastoma (Proc.Natl.Acad.Sci.USA, 1970,65:129-136), NB41A3 mouse neuroblastoma (Proc.Natl.Acad.Sci.USA, 1962,48:1184-1190), SCP sheep choroid plexus papilloma (Bolin et al., 1994, J.Virol.Methods 48:211-221), G355-5, PG-4 cat normal astrocytes (Haapala et al., 1985, J.Virol.53:827-833), Mpf ferret brain (Trowbridge et al., 1982, In Vitro 18:952-960) and normal cell lines such as, for example, CTX TNA2 rat normal cortical brain (Radany et al., 1992, Proc. Natl. Acad. Sci. USA 89:6467-6471) such as, for example, CRL7030 and Hs578Bst.Furthermore, different vector / host expression systems can affect the processing response to varying degrees.

[0242] For the long-term high yield production of recombinant proteins, stable expression is usually preferred. For example, the cell line of modified heavy chain and light chain (for example, antibody or fusion protein) of the present invention can be stably expressed by engineering. The host cell can be transformed by DNA controlled by appropriate expression control elements (for example, promoter, enhancer, sequence, transcription terminator, polyadenylation site, etc.) and selectable markers rather than by using expression vectors containing viral replication start points. After introducing foreign DNA, the engineered cells can be grown in enriched medium for 1-2 days, and then switched to selection medium. The selectable marker in the recombinant plasmid confers resistance to selection and allows cells to stably integrate plasmids into their chromosomes, and grow to form colonies (foci), which can then clone the colonies and expand them into cell lines.

[0243] A number of selection systems can be used, including but not limited to the herpes simplex virus thymidine kinase (Wigler et al., 1977, Cell 11:223), hypoxanthine-guanine phosphoribosyltransferase (Szybalska & Szybalski, 1962, Proc. Natl. Acad. Sci. USA 48:2026), and adenine phosphoribosyltransferase (Lowy et al., 1980, Cell 22:817) genes for tk-, hgprt-, or aprt- cells, respectively. Similarly, antimetabolite resistance can also be used as the basis for selection for the following genes: dhfr (which confers resistance to methotrexate) (Wigler et al., 1980, Natl. Acad. Sci. USA 77:3567; O'Hare et al., 1981, Proc. Natl. Acad. Sci. USA 78:1527); gpt (which confers resistance to mycophenolic acid) (Mulligan & Berg, 1981, Proc. Natl. Acad. Sci. USA 78:2072); neo (which confers resistance to the aminoglucosidase G-418) (Colberre-Garapin et al., 1981, J. Mol. Biol. 150:1); and hygro (which confers resistance to hygromycin) (Santerre et al., 1984, Gene 30:147) genes.

[0244] Co-expression of heavy and light chains

[0245] According to the right heavy chain immunoglobulin and light chain of heterodimer of the present invention can be co-expressed in mammalian cells, as pointed out above.In one embodiment, a heavy chain and two different light chains are co-expressed in above-mentioned LCCA design group, wherein heavy chain preferentially matches with one of two light chains.In another embodiment, two heavy chains and two light chains are co-expressed, wherein each heavy chain preferentially matches with a light chain in the light chain.

[0246] Testing of heterodimer pairs

[0247] As described above, at least one heterodimer according to the right heterodimer of the present invention can comprise one or more amino acid modifications to their heavy chains of immunoglobulins and / or light chains of immunoglobulins, so that when two heavy chains and two light chains of the right heterodimer are co-expressed in mammalian cells, the heavy chain of the first heterodimer is preferentially paired with one of the light chains rather than another. Similarly, the heavy chain of the second heterodimer is preferentially paired with the second light chain rather than the first light chain. The degree of preferential pairing can be assessed, for example, by using the method described above. Each heterodimer of the right heterodimer can be tested as described below for its affinity for each antigen. The thermal stability of each heterodimer of the right heterodimer can also be tested as described below.

[0248] Methods for measuring preferred pairing

[0249] LCCA

[0250] In one embodiment, the preferential pairing between an immunoglobulin heavy chain and a light chain is determined by performing a light chain competitive assay (LCCA). Commonly owned patent application PCT / US2013 / 063306, filed October 3, 2013, describes various embodiments of LCCA and is incorporated herein by reference in its entirety for all purposes. The method allows for quantitative analysis of the pairing of heavy chains with specific light chains in a co-expressed protein mixture and can be used to determine whether a specific immunoglobulin heavy chain selectively associates with either of two immunoglobulin light chains when the heavy and light chains are co-expressed. The method is briefly described as follows: at least one heavy chain and two different light chains are co-expressed in a cell in such a ratio that the heavy chain is the limiting pairing reactant; optionally separating the secreted protein from the cells; separating the immunoglobulin light chain polypeptides that are bound to the heavy chain from the rest of the secreted protein to produce an isolated heavy chain pairing fraction; detecting the amount of each different light chain in the isolated heavy chain fraction; and analyzing the relative amount of each different light chain in the isolated heavy chain fraction to determine the ability of at least one heavy chain to selectively pair with one of the light chains.

[0251] The method provides a reasonable throughput and is robust (i.e., insensitive to slight changes in operation such as the user or flow rate) and accurate. The method provides a sensitive assay for the effect of small variations in measurable protein sequences. Mixed protein-protein, domain-domain, chain-chain interactions on large surface areas typically require multiple mutations (exchanges) to introduce selectivity. The protein product does not have to be separated and purified, which enables more efficient screening. Further details of the embodiments of the method are described in the examples.

[0252] Alternative methods for determining preferential pairing

[0253] Alternative methods for detecting preferential pairing include the use of LC-MS (liquid chromatography-mass spectrometry) to quantify the relative heterodimer population (including each light chain) by using differences in their molecular weight to identify each different species. Antigenic activity assays can also be used to quantify the relative heterodimer population containing each light chain, whereby the measured degree of binding (relative to a control) can be used to assess each respective heterodimer population.

[0254] Additional methods such as SMCA are described in the Examples, Figures and Tables.

[0255] Thermal stability

[0256] The thermal stability of the heterodimers can be determined according to methods known in the art. The melting temperature of each heterodimer is indicative of its stability. The melting point of the heterodimer can be measured using techniques such as differential scanning calorimetry (Chen et al. (2003) Pharm Res 20: 1952-60; Ghirlando et al. (1999) Immunol Lett 68: 47-52). Alternatively, circular dichroism (Murray et al. (2002) J. Chromatogr Sci 40: 343-9) can be used to measure the thermal stability of the heterodimers.

[0257] Affinity for antibodies

[0258] The binding affinity of heterodimers for their respective antigens and the dissociation rate of the interaction can be determined by competitive binding assays according to methods known in the art. An example of a competitive binding assay is a radioimmunoassay, which includes incubating labeled antigen (e.g., for target molecules (e.g., heterodimers of the present invention), 3H or 125I) in the presence of increasing amounts of unlabeled antigen, and detecting molecules bound to the labeled ligand. The affinity of the heterodimers of the present invention for antigens and the binding dissociation rate can be determined from saturation data by Scatchard analysis.

[0259] The kinetic parameters of the heterodimers according to the present invention can also be determined using surface plasmon resonance (SPR)-based assays known in the art (e.g., BIAcore kinetic analysis). For a review of SPR-based techniques, see Mullet et al., 2000, Methods 22: 77-91; Dong et al., 2002, Review in Mol. Biotech., 82: 303-23; Fivash et al., 1998, Current Opinion in Biotechnology 9: 97-101; Rich et al., 2000, Current Opinion in Biotechnology 11: 54-61. In addition, any SPR device and SPR-based method described in U.S. Pat. Nos. 6,373,577, 6,289,286, 5,322,798, 5,341,215, 6,268,125 for measuring protein-protein interactions can be expected to be used in the methods of the present invention. FACS can also be used to measure affinity, as is known in the art.

[0260] Pharmaceutical composition

[0261] The present invention also provides pharmaceutical compositions comprising heterodimers or heterodimer pairs described herein. Such compositions comprise a therapeutically effective amount of heterodimers or heterodimer pairs and a pharmaceutically acceptable carrier. In a specific embodiment, the term "pharmaceutically acceptable" means approved by the regulatory mechanisms of the federal or state governments or listed in the U.S. Pharmacopoeia or other generally recognized pharmacopeias for animals, more particularly for humans. The term "carrier" refers to a diluent, adjuvant excipient or vehicle used together with the therapeutic agent. Such pharmaceutical carriers can be sterile liquids, such as water and oils, including those of petroleum, animal, vegetable or synthetic origin, such as peanut oil, soybean oil, mineral oil, sesame oil, etc. When the pharmaceutical composition is administered intravenously, water is a preferred carrier. Saline solutions and aqueous glucose and glycerol solutions can also be used as liquid carriers, particularly for injectable solutions. Suitable pharmaceutical excipients include starch, glucose, lactose, sucrose, gelatin, malt, rice, flour, chalk, silica gel, sodium stearate, glycerol monostearate, talc, sodium chloride, skim milk powder, glycerol, propylene, ethylene glycol, water, ethanol, etc. If necessary, the composition may also contain a small amount of a wetting or emulsifying agent, or a pH buffer. These compositions may take the form of solutions, suspensions, emulsions, tablets, pills, capsules, powders, sustained-release preparations, etc. The composition may be formulated into a suppository with conventional binders and carriers such as triglycerides. Oral preparations may include standard carriers such as pharmaceutical grade mannitol, lactose, starch, magnesium stearate, sodium saccharin, cellulose, magnesium carbonate, etc. Examples of suitable pharmaceutical carriers are described in "Remington's Pharmaceutical Sciences" compiled by EWMartin. Such compositions will contain (preferably in purified form) a therapeutically effective amount of the compound, and an appropriate amount of a carrier to provide a form for correct administration to the patient. The formulation should be suitable for the mode of administration.

[0262] In certain embodiments, the composition comprising heterodimer or heterodimer is formulated as a pharmaceutical composition suitable for intravenous administration to humans according to conventional methods. Typically, the composition for intravenous administration is a solution in a sterile isotonic aqueous buffer. If necessary, the composition may also include a solubilizer and a local anesthetic such as lidocaine to relieve pain at the injection site. Generally, the composition is provided separately or mixed together in a sealed container such as an ampoule or a pouch indicating the amount of the active agent in a unit dose form (e.g., as a dry lyophilized powder or anhydrous concentrate). When the composition is to be administered by infusion, it can be distributed with an infusion bottle containing sterile pharmaceutical grade water or saline. When the composition is administered by injection, an ampoule of sterile water for injection or saline may be provided so that the composition can be mixed before administration.

[0263] In certain embodiments, the compositions described herein are formulated in neutral or salt form. Pharmaceutically acceptable salts include those formed with anions, such as those derived from hydrochloric acid, phosphoric acid, acetic acid, oxalic acid, tartaric acid, and the like, and those formed with cations, such as those derived from sodium hydroxide, potassium hydroxide, ammonium hydroxide, calcium hydroxide, ferric hydroxide, isopropylamine, triethylamine, 2-ethylaminoethanol, histidine, procaine, and the like.

[0264] The amount of the compositions described herein that is effective in the treatment, inhibition and prevention of diseases or conditions associated with abnormal expression and / or activity of therapeutic proteins can be determined by standard clinical techniques. In addition, in vitro assays can optionally be used to help identify the optimal dose range. The precise dosage to be used for the formulation will also depend on the route of administration and the severity of the disease or condition, and should be determined according to the doctor's judgment and the situation of each patient. Extrapolate the effective dose from the dose-response curve derived from an in vitro or animal model test system.

[0265] Uses of heterodimer pairs

[0266] As described above, the heterodimer pair according to the present invention may comprise a first heterodimer and a second heterodimer, wherein the immunoglobulin heavy chain and immunoglobulin light chain of each heterodimer are derived from known therapeutic antibodies or engineered from them, or from known antibodies of binding molecules. Therefore, it is expected that the heterodimers derived from these antibodies or engineered from them can be used to treat or prevent the same diseases, disorders or infections for which known therapeutic antibodies or known antibodies can be used.

[0267] Therefore, in one embodiment, a heterodimer pair according to the present invention comprising a heterodimer having a heavy chain and a light chain derived from a therapeutic antibody useful for treating and / or preventing cancer or related disorders can also be used to treat or prevent cancer and related disorders.

[0268] In another embodiment, a heterodimer pair according to the present invention comprising a heterodimer having a heavy chain and a light chain derived from a therapeutic antibody that can be used to prevent, treat or manage the symptoms of an inflammatory disorder in a subject can also be used to prevent, treat or manage the symptoms of an inflammatory disorder in a subject.

[0269] In another embodiment, a heterodimer pair according to the present invention comprising a heterodimer having a heavy chain and a light chain derived from a therapeutic antibody that can be used to treat or prevent an autoimmune disorder or an inflammatory disease in a subject can also be used to treat or prevent an autoimmune disorder or an inflammatory disease in a subject.

[0270] In another embodiment, a heterodimer pair according to the present invention comprising a heterodimer having a heavy chain and a light chain derived from a therapeutic antibody useful for treating or preventing an infectious disease in a subject can also be used to treat or prevent an infectious disease in a subject.

[0271] In another embodiment, a heterodimer pair according to the present invention comprising a heterodimer having a heavy chain and a light chain derived from a therapeutic antibody useful for treating a vascular disease in a subject can also be used to treat a vascular disease in a subject.

[0272] In another embodiment, the heterodimer according to the present invention may also be advantageously used in combination with other therapeutic agents known in the art for treating or preventing cancer, autoimmune diseases, inflammatory conditions or infectious diseases. In a specific embodiment, the heterodimer according to the present invention may be used in combination with, for example, monoclonal or chimeric antibodies, lymphokines or or hematopoietic growth factors (such as, for example, IL-2, IL-3 and IL-7) for increasing the number or activity of effector cells, the effector cells interact with the molecules to enhance immune response. The heterodimer according to the present invention may also be advantageously used in combination with one or more drugs (such as, for example, anticancer agents, anti-inflammatory agents or antiviral agents) for treating diseases, conditions or infections.

[0273] Reagent test kit

[0274] The present invention additionally provides a kit comprising one or more heterodimer pairs. The individual components of the kit may be packaged in separate containers and, in conjunction with such containers, a notice in the form prescribed by a governmental regulatory agency governing the manufacture, use or sale of a drug or biological product, the notice reflecting the regulatory approval of the manufacture, use or sale. The kit may optionally contain instructions or instructions outlining the use and administration regimen of the heterodimer.

[0275] When one or more components of the kit are provided in the form of a solution, e.g., an aqueous solution or a sterile aqueous solution, the container means may itself be an inhaler, syringe, pipette, eye dropper or other such similar device from which the solution may be administered to a subject or used in or mixed with additional components of the kit.

[0276] The components of the kit may also be provided in dry or lyophilized form, and the kit may additionally contain a suitable solvent for reconstitution of the lyophilized components. Regardless of the number or type of containers, the kit of the present invention may also include a device to assist in administering the composition to a patient. Such a device may be an inhaler, a nasal spray device, a syringe, a pipette, tweezers, a measuring spoon, an eye dropper, or a similar medically approved delivery vehicle.

[0277] Computer Implementation

[0278] In one embodiment, the computer comprises at least one processor coupled to a chipset. Also coupled to the chipset are memory, storage devices, keyboards, and network adapters. The display is coupled to a graphics adapter. In one embodiment, the function of the chipset is provided by a memory controller hub and an I / O controller hub. In another embodiment, the memory is directly coupled to the processor rather than the chipset.

[0279] A storage device is any device capable of holding data, such as a hard drive, compact disk read-only memory (CD-ROM), DVD, or solid-state storage device. The memory holds instructions and data used by the processor. A pointing device may be a mouse, trackball, or other type of pointing device, and is used in combination with a keyboard to input data into a computer system. A graphics adapter displays images and other information on a display. A network adapter couples a computer system to a local or wide area network.

[0280] As known in the art, a computer may have components different from the previously described components and / or other components. In addition, a computer may lack certain components. In addition, the storage device may be from a local and / or remote computer (such as embodied in a storage area network (SAN)).

[0281] As known in the art, a computer is suitable for executing computer program modules for providing the functionality described herein. As used herein, the term "module" refers to computer program logic for providing a specific function. Therefore, a module can be implemented in hardware, firmware, and / or software. In one embodiment, a program module is stored on a storage device, loaded into a memory, and executed by a processor.

[0282] It should be understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in view thereof will be suggested to those skilled in the art and are to be included within the spirit and purview of the present application and the scope of the appended claims. Example

[0283] Below is the example for carrying out the specific embodiment of the present invention.Described embodiment is provided for illustration purpose only, and is not intended to limit the scope of the present invention in any way.Efforts have been made to ensure the accuracy about the numbers used (e.g., amount, temperature, etc.), but of course certain experimental errors and deviations should be allowed.

[0284] Unless otherwise indicated, the practice of the present invention will employ conventional methods of protein chemistry, biochemistry, recombinant DNA technology, and pharmacology within the capabilities of those of ordinary skill in the art. Such techniques are fully explained in the literature. See, for example, TECreighton, Proteins: Structures and Molecular Properties (WH Freeman and Company, 1993); A Lehninger, Biochemistry (Worth Publishers, Inc., current edition); Sambrook, et al., Molecular Cloning: A Laboratory Manual (2nd edition, 1989); Methods In Enzymology (S. Colowick and N. Kaplan, ed., Academic Press, Inc.); Remington's Pharmaceutical Sciences, 18th edition (Easton, Pennsylvania: Mack Publishing Company, 1990); Carey and Sundberg Advanced Organic Chemistry 3rd edition (Plenum Press) Volumes A and B (1992).

[0285] Example 1: Preparation of constructs encoding D3H44 IgG heavy chain and D3H44 IgG light chain.

[0286] The wild-type heavy chain and light chain of the anti-tissue factor antibody D3H44 of the coexpression group described herein are prepared as follows. D3H44 Fab light chain (AJ308087.1) and heavy chain (AJ308086.1) sequences are obtained from GenBank (Table A, A1 and A2), for mammalian expression synthetic genes and optimized codons. The light chain vector insert composed of 5'-EcoRI cleavage site-HLA-A signal peptide-HA or FLAG tag-light chain Ig clone-'TGA termination'-BamH1 cleavage site-3' is connected to the pTT5 vector (Durocher Y et al., Nucl.Acids Res.2002; 30, No.2e9). The resulting vector+insert is sequenced to confirm the correct reading frame and the sequence of the coding DNA. Similarly, the heavy chain vector insert consisting of 5'-EcoR1 cleavage site-HLA-A signal peptide-heavy chain clone (terminated at T238; see Table A1)-ABD2-His6 tag-TGA termination-BamH1 cleavage site-3' was ligated into the pTT5 vector (ABD; albumin binding domain). The resulting vector + insert was also sequenced to confirm the correct reading frame and sequence of the coding DNA. Different D3H44 constructs were generated by gene synthesis or by site-directed mutagenesis (Braman J, Papworth C & Greener A., ​​Methods Mol. Biol. (1996) 57: 31-44).

[0287] The heavy and light chains were tagged at the C and N termini, respectively, to facilitate evaluation of preferential pairing by competitive assay-SPR screening. The ABD2-His6 heavy chain tag specifically allowed HC-LC complexes to be captured on the anti-his tag SPR chip surface, while the FLAG and HA light chain tags allowed relative LC1 and LC2 populations to be quantified.

[0288] Example 2: Evaluation of preferential pairing of heterodimers in co-expression panels containing variable domain modifications in D3H44 IgG light and / or heavy chains

[0289] The heterodimers were determined to preferentially bind to the modified V in a co-expression panel containing the D3H44 heavy and light chains. L and / or V H The ability of domain pairing is shown in Figure 1 middle. Figure 1 The results provided in are preliminary and a more complete set of results is provided below. Figure 1-3 The amino acid modifications shown in are identified. See Tables A, A1 and A2.

[0290] One D3H44 heavy chain construct was co-expressed with two unique D3H44 light chain constructs and screened according to the competition assay-SPR ( Figure 1 Relative light chain pairing specificity (e.g., H1-L1:H1-L2) was determined by light chain competition assay validation, where the L1:L2 DNA ratio was varied during transfection to 40:60, 50:50, and 60:40 ( Figure 1 ). For competition assay screening and validation, the heavy chain was kept at limiting amounts (i.e., HC < L1 + L2). A schematic diagram representing the design of the assay is shown in Figure 8 middle.

[0291] The method is performed as follows: The light chain competition assay (LCCA) quantifies the relative pairing of one heavy chain for at least two unique light chains. The assay and the aforementioned steps can be summarized as follows: 1. Concomitant expression of heavy and light chains, with the heavy chain present in limited amounts (e.g. HC:LC1:LC2=1:1:1), 2. Separation of HC-LC complexes - achieved by binding the heavy chain to an SPR chip via his-tag pull-down, and 3) quantification of relative HC-LC populations (i.e. H1-L1:H1-L2). In the SPR format, antibodies specific for the unique light chain labeled populations are used for quantification. Note: This assay can be performed with or without HL disulfide. A schematic representing the method is shown in Fig. 9 middle.

[0292] Transfection method

[0293] The co-expression set containing one heavy chain and two light chain constructs prepared as described in Example 1 was transfected into CHO-3E7 cells as follows. CHO-3E7 cells were cultured at 1.7-2×10 6 The cells were cultured at a density of 10 cells / ml at 37°C in FreeStyle 4x10 medium supplemented with 4 mM glutamine and 0.1% Pluronic F-68 (Invitrogen cat#24040-032). TM F17 medium (Invitrogen cat#A-1383501). A total volume of 2 ml was transfected with a total of 2 ug DNA at a DNA:PEI ratio of 1:2.5 using PEI-pro (Polyplus cat#115-010). 24 hours after adding the DNA-PEI mixture, the cells were transferred to 32°C. On the 7th day, the expression of the supernatant was tested by non-reducing SDS-PAGE analysis followed by Coomassie blue staining to visualize the bands. The HC:LC ratios are as specified in Table 7.

[0294] Table 7:

[0295]

[0296] ^Filler DNA (Filler DNA): No DNA insert P TT5 vector.

[0297] Competitive assay SPR method

[0298] The extent of preferential D3H44 light chain pairing to the D3H44 heavy chain in the co-expression group was assessed using an SPR-based readout of a unique epitope tag located on the N-terminus of each light chain.

[0299] Surface plasmon resonance (SPR) equipment. GLM sensor chip, Biorad ProteOn amine coupling kit (1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysulfosuccinimide (sNHS) and ethanolamine) and 10 mM sodium acetate buffer were purchased from Bio-Rad Laboratories (Canada) Ltd. (Mississauga, ON). Recombinant Her-2 protein was purchased from eBioscience (San Diego, CA). 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, ethylenediaminetetraacetic acid (EDTA) and NaCl were purchased from Sigma-Aldrich (Oakville, ON). 10% Tween 20 solution was purchased from Teknova (Hollister, CA).

[0300] SPR mutual sensor determination. All surface plasmon resonance measurements were performed at a temperature of 25° C. using a BioRad ProteOn XPR36 instrument (Bio-Rad Laboratories (Canada) Ltd. (Mississauga, ON)) with PBST electrophoresis buffer (PBS containing 0.05% Tween 20 Teknova Inc). The anti-penta His capture surface was generated using a GLM sensor chip activated by injecting a 1:5 dilution of a standard BioRad sNHS / EDC solution at 100 μL / min (for 140 seconds) in the analyte (horizontal) direction. Immediately after activation, a solution of 25 μg / mL of anti-penta His antibody (Qiagen Inc.) in 10 mM NaOAc pH 4.5 was injected at a flow rate of 25 μL / min in the analyte (vertical) direction until approximately 3000 resonance units (RU) were immobilized. Remaining reactive groups were quenched by injecting 1 M ethanolamine at 100 μL / min for 140 sec in the analyte direction, which also ensured that an interspot of analog activation was generated for blank reference.

[0301] The screening of the binding of anti-FLAG (Sigma Inc.) and anti-HA (Roche Inc.) monoclonal antibodies of heterodimers is carried out in the following two steps: by indirectly capturing heterodimers on the anti-penta His surface in the ligand direction, and then injecting anti-FLAG and anti-HA injections in the analyte. First, a buffer injection (lasting 30 seconds) carried out in the ligand direction at 100uL / minute is used to stabilize the base baseline. For each heterodimer capture, the unpurified heterodimer in the cell culture medium is diluted to 4% in PBST. In a single ligand channel, 1 to 5 heterodimers or controls (i.e., controls containing 100% HA-light chain or 100% FLAG-light chain) are injected simultaneously at a flow rate of 25 μL / minute (lasting 240 seconds). This results in about 300 to 400RU of saturated heterodimer capture on the anti-penta His surface. If necessary, the first ligand channel is left empty to be used as a blank control. After the heterodimer capture step, two buffer injections were immediately performed in the analyte direction to stabilize the baseline, followed by two injections of 5nM anti-FLAG and 5nM anti-HA at 50 μL / min each for 120 seconds, with a 180 second dissociation period, thereby generating a set of binding sensorgrams with buffer reference for each captured heterodimer. If possible, the antigen bound by the heterodimer can also be injected above the last remaining analyte channel as an activity control. The heterodimer was regenerated by an 18 second pulse of 0.85% phosphoric acid (18 seconds at 100 μL / min) to prepare the anti-penta His surface for the next injection cycle. The sensorgrams were compared and double referenced using buffer blank injections and midpoints, and the resulting sensorgrams were analyzed using ProteOn Manager software v3.0.

[0302] The total percentage of L1 and L2 should theoretically add up to 100%. In practice, it was observed for some variants that the total amount of L1 and L2 added up to significantly less than 100%. This difference in the total light chain percentage is believed to be partially due to the occurrence of variable nonspecific binding during the initial heterodimer capture process on the SPR chip.

[0303] Example 3: Heterodimers containing constant (C L or C H1 Evaluation of preferential pairings in co-expression groups of ) domain modifications

[0304] As described in Example 2 for heterodimers with variable domain modifications, it was determined that heterodimers preferentially associate with the modified C in a co-expression set comprising the D3H44 heavy and light chains. L and / or C H1 The ability of domain pairing is shown in Figure 2 One D3H44 heavy chain construct was co-expressed with two unique D3H44 light chain constructs and screened according to the competitive assay-SPR ( Figure 2 Relative light chain pairing specificity (e.g., H1-L1:H1-L2) was determined by the column titled "Competition Assay Results" in the Figure 1A . Selected heterodimer hits were validated by a modified competition assay in which the L1:L2 DNA ratio was varied during the transfection process through 40:60, 50:50, and 60:40 ( Figure 2 ). For competition assay screening and validation, the heavy chains were kept at limiting amounts (ie, HC < L1 + L2) as described in Example 2. The evaluation of preferential pairing was performed as described in Example 2.

[0305] Example 4: Scaling up for biophysical characterization

[0306] Selected heterodimers (paired and mismatched) are scaled up (usually up to 50 ml) and purified to test thermal stability and antigen binding. Figure 3 The heterodimers HD100-HD115 shown in . The heavy and light chains of each heterodimer were expressed in 50 ml cultures of CHO-3E7 cells under the above culture conditions. The cells were centrifuged and the heterodimers were purified by loading the supernatant onto a Fractogel column packed with nickel as described below.

[0307] Purification on a Fractogel column packed with nickel (His)

[0308] The column was loaded with nickel: 5 column volumes (CV) of 0.5M NaCl (no pH adjustment), followed by 4CV 200mM NiCl2 (nickel) and 2CV 0.5M NaCl pH 5.0 for consecutive washes. Sample loading and elution: The column was equilibrated with 10CV PBS. The sample was loaded and washed with 10CV of wash buffer #1 (50mM sodium phosphate pH 7.0, 300mM NaCl), followed by 10CV of wash buffer #2 (50mM sodium phosphate pH 7.0, 300mM NaCl, 25mM imidazole) to remove impurities bound to the column. The heterodimers were eluted in fractions with wash buffer #1 + 300mM imidazole. The protein content of each fraction was tested by Bradford protein assay. The fractions containing protein were mixed. The antigen binding and thermal stability of the purified heterodimers were then determined as described in Example 5.

[0309] Example 5: Thermal stability and antigen affinity measurements of heterodimers

[0310] The thermal stability and antigen affinity of the selected heterodimer pairs were measured to compare these properties with those of the wild-type unmodified heavy chain-light chain pairs. The correctly paired and mismatched heterodimers from the co-expression group were individually scaled up, purified (i.e., His-tag affinity purification) as described below, and evaluated for thermal stability and antigen binding. The results are shown in Figure 3 middle.

[0311] Measurement of thermal stability

[0312] The thermal stability of selected heterodimer pairs was measured using differential scanning calorimetry (DSC) as follows.

[0313] Each heterodimer was purified as described in Example 3, diluted to 0.2 mg / mL in PBS, and a total of 400 μL was used for DSC analysis using a VP-Capillary DSC (GE Healthcare). At the beginning of each DSC run, 5 buffer blank injections were performed to stabilize the baseline, and a buffer injection was set up before each heterodimer injection for reference. Each sample was scanned from 20°C to 100°C at a rate of 60°C / hour using low feedback, 8 seconds of filtration, 5 minutes of preTstat, and 70 psi of nitrogen pressure. The resulting thermogram was referenced and analyzed using Origin 7 software.

[0314] The thermal unfolding curves of the tested heterodimers are shown in Figure 4 The results show that correctly paired heterodimers (from a design point of view) are generally significantly more stable than intentionally mismatched heterodimers (e.g., HD107 vs. HD108). In addition, many correctly paired heterodimers exhibit thermal stability close to that of wild-type Fab (e.g., HD114).

[0315] Measurement of antigen affinity

[0316] The affinity of the heterodimer pairs for the antigen (tissue factor extracellular domain) was measured using a surface plasmon resonance (SPR) assay. All surface plasmon resonance assays were performed at 25°C using a BioRad ProteOn XPR36 instrument (Bio-Rad Laboratories (Canada) Ltd. (Mississauga, ON) using PBST electrophoresis buffer (PBS containing 0.05% Tween 20 Teknova Inc). The purified tissue factor (TF) surface was generated using a GLM sensor chip activated by injecting a 1:10 dilution of a standard BioRad sNHS / EDC solution at 100 μL / min (for 140 seconds) in the analyte (perpendicular) direction. Immediately after activation, a solution of 25 μg / mL of TF in 10 mM NaOAc pH 4.5 was injected in the ligand direction at a flow rate of 25 μL / min until approximately 1000 resonance units (RU) were immobilized (or when 60 nM was flowing). FAB was sufficient to produce a maximum response of 100RU). Remaining active groups were quenched by injecting 1M ethanolamine at 100 μL / min for 140 seconds in the direction of the analyte. For each injection series, two buffer blank injections were performed as horizontal injections before the purified heterodimer. A 3-fold dilution series (60nM, 20nM, 6.7nM, 2.2nM) and a blank buffer control were injected simultaneously at 50 μL / min (for 120 seconds) for each heterodimer, and a set of binding sensorgrams with buffer reference were generated for each heterodimer. The heterodimer:TF complex on the SPR surface was regenerated by an 18-second pulse of 0.85% phosphoric acid (at 100 μL / min for 18 seconds) to prepare the TF surface for the next injection cycle. The sensorgrams were aligned and double-referenced using buffer blank injections and midpoints, and the resulting sensorgrams were analyzed using a 1:1 binding model in ProteOn Manager software v3.0.

[0317] The results indicate that correctly paired heterodimers (from a design standpoint) display a range of affinities for the antigen, with some designs showing wild-type-like binding affinities for the antigen (eg, HD107 and HD114).

[0318] Example 6: Size Exclusion Chromatography (SEC) Profiles of Representative Samples of Wild-Type Labeled D3H44 Heterodimers and Individual Preferentially Paired Heterodimers

[0319] Wild-type D3H44 heterodimers (one heavy chain and one light chain) with a C-terminal ABD2-His6 tag on the heavy chain and an N-terminal FLAG tag on the light chain were expressed and purified according to methods known in the art and similar to those described in Examples 1 and 4. Preferred or correctly paired heterodimers from the co-expression group were individually scaled up ( Figure 3 The heterodimers HD100, HD105 and HD107 shown in ) were purified by His-tag affinity chromatography and SEC as described in Example 4.

[0320] SEC was performed as follows. The heterodimer samples were separated on a Superdex 200HR 10 / 30 Pharmacia (GE Healthcare) column on the Purifier system. The heterodimer samples (0.3-0.5 ml) in PBS were manually loaded into a 0.5 ml loop filled with PBS. The samples were then automatically injected onto the column and separated at 0.5 ml / min using a 1 CV elution volume. At OD 280 Protein elution was monitored and collected in 1 ml fractions.

[0321] like Figure 5 As shown in , the correctly paired heterodimer exhibits a SEC profile close to that observed for the wild-type heterodimer without amino acid modification (main peak [*]: heterodimer). Equivalent results were obtained when the light chain of the wild-type heterodimer had an N-terminal HA tag.

[0322] Example 7: Additional data related to the stability of the heterodimer

[0323] The designs shown in Table 8 are highlighted as a combination of design drivers with enhanced HC-LC selectivity.

[0324] Table 8:

[0325]

[0326]

[0327] The residue numbering follows the Kabat nomenclature (Kabat EA et al., (1983) Sequence of Proteins of Immunological Interest National Institutes of Health, Bethesda).

[0328] Most designs maintained wild-type-like thermal stability (Tm) and TF binding affinity shown in Table 9.

[0329] Table 9:

[0330]

[0331] Design 12 contained the wild-type H2-L2 pairing (see Table 8). As a result, the H2-L2 Fab retained the wild-type binding affinity. Wild-type anti-TF D3H44 Fab Tm = ~76°C (data not shown).

[0332] Example 8: Additional heterodimers and testing of said heterodimers

[0333] Additional heterodimer pairs were prepared and tested as described in Table 10. These heterodimers were designed to increase Fab hotspot coverage.

[0334] Table 10:

[0335]

[0336]

[0337] *The residue numbering in Table 10 follows the convention used for the residues in the crystal structure of D3H44 Fab (PDB ID=1JPT [Faelber K et al., J. Mol. Biol. (2001) 313:83-97]; www.rcsb.org / pdb / explore / explore.do?structureld=1JPT).

[0338] The stability, ability to bind to target, and ability to selectively pair of these heterodimers were determined as described in Example 5 and are shown in Table 11.

[0339] Table 11:

[0340]

[0341] Wild-type anti-TF D3H44 Fab KD = 0.052 nM; Wild-type anti-TF Fab Tm = ~76°C

[0342] Example 9: Additional Heterodimers

[0343] The following heterodimer pairs were also prepared and tested for their ability to selectively pair.

[0344] Table 12:

[0345]

[0346] The residue numbering in Table 12 follows the crystal structure for D3H44 Fab (PDB ID = 1JPT [Faelber K et al., J. Mol. Biol. (2001) 313: 83-97]; www.rcsb.or g / pdb / explore / explore.do? structureId=1JPT ) in the residues.

[0347] Table 13:

[0348]

[0349]

[0350] The residue numbering in Table 13 follows the crystal structure for D3H44 Fab (PDB ID = 1JPT [Faelber K et al., J. Mol. Biol. (2001) 313: 83-97]; www.rcsb.o rg / pdb / explore / explore.do? structureId=1JPT ) in the residues.

[0351] Example 10: Evaluation of preferential pairing of heterodimers in co-expression panels containing constant domain and / or variable domain modifications in D3H44 heavy chain and light chain Fab formats

[0352] Design Figure 1 and 2 The co-expression groups other than those co-expression groups are shown in Tables 14 and 15. As described in Example 1, constructs encoding D3H44 IgG heavy and light chains in Fab form were prepared, containing amino acid modifications according to the design of the co-expression group. As described in Examples 2 and 3, the ability of D3H44 heavy and light chain Fabs to preferentially pair was evaluated. As described in Example 5, the stability and binding affinity of the designs were determined. The results shown in Tables 14 and 15 are cumulative and include all constructs other than the new designs. Figure 1 and 2 The results of the designs shown in the table. The amino acid modifications shown in these tables are identified with reference to the amino acid sequences of the D3H44 heavy chain and the D3H44 light chain. See Tables A, A1 and A2.

[0353] It is noted that "design" or "design set" in this application refers to a set of mutations on the H1, L1, H2 and L2 chains. "LCCA design" refers to a set of mutations in H1, L1 and L2.

[0354] Each unique group of H1, L1 and L2 mutations (LCCA format) is assigned a unique number, or so-called 'unique identifier'. When data is provided in H1 L1 H2 L2 format (Fab pair format or SMCA), such a design group is represented by a 'unique identifier group' consisting of two unique identifiers that make up the LCCA (e.g., 1-2). First, the design characterized in the D3H44 LCCA data set is specified. Designs that are not present in this group but are present in different homogeneous or mixed system data or / and in different formats (MCA) are additionally represented by *. In this exercise of specifying unique identifiers, certain redundancy has been generated through the automated processing of many tables. The cases in which different WT amino acids occupy the same position in the system except D3H44 are: 309*=319*=47, 316*=101, 317*=183, 318*=182, 310*=48, 311*=102, 323*=180, 324*=179. The cases where additional mutations were incorporated for LC / MS were: 442*=326* and 443*=23.

[0355] Note that most LCCA experiments were performed on constructs without the interchain Fab disulfide bonds located in the constant domains (H / C233S-L / C214S).

[0356] In Table 14, designs showing correct pairing specificity of 55%:45% (H1-L1:H1-L2 and H2-L2:H2-L1) or greater are provided. To highlight the success of a particular design with respect to preferential pairing, two complementary LCCA groups (H1, L1, L2 and H2, L2, L1) are represented in the form of Fab pairs.

[0357] The presence of tags (L: HA and FLAG and H: ABD2) did not affect the -50%:50% neutral pairing expected for D3H44 WT (this was further supported by the actual designed pairing results; therefore tag information is not included in the table).

[0358] In the table, the measured amounts of the relevant Fab species (measured amounts of H1-L1, H1-L2 and H2-L2, H2-L1) are included in the form of ratios (H1-L1:H1-L2 and H2-L2:H2-L1). In most cases, several LCCA experimental replicates (screening and validation) were performed. A summary column in the format of normalized ratios (i.e., the sum of H1-L1 and H1-L2 to reach 100%) of the intermediate H1-L1:H1-L2 and H2-L2:H2-L1 is also provided.

[0359] The data were clustered according to the categories of thermal stability (Tm) and antigen affinity (in the table buckets, 'TF' notation is used for tissue factor affinity).

[0360] Tm1 (x=>71°C); Tm2 (71°C>x=>66°C); Tm3 (66°C>x). The Tm3 category also includes ND (not tested) cases.

[0361] For reference, the Tm of D3H44WT Fab (without disulfide bonds) is -76°C.

[0362] TF1 (x = <5X median KD of WT); TF2 (5 < x = <20X median KD of WT); and a miscellaneous category including cases with x > 20X median KD of WT, where NB cases are marked separately (no binding: for KD greater than 500 nM, it is ~10000X median KD of WT). This last type also includes ND cases (experiment not performed).

[0363] For example, the median KD of D3H44 WT Fab was 0.06 nM (range 0.1).

[0364] Note: In the columns referring to antigen affinity and thermal stability, if the number of experiments performed was greater than 1 (n>1), the range (minimum - maximum readings) is indicated.

[0365] Within each bucket, designs were ranked in descending order of pair specificity of H1-L1:H1-L2, followed by pair specificity of H2-L2:H2-L1.

[0366] Examples of the reading bucket categories in the table:

[0367] Tm1 only (H1-L1 and H2-L2Tm both belong to Tm1 category)

[0368] Tm1 / Tm2 (H1-L1 or H2-L2Tm belongs to Tm1 and the other belongs to Tm2 category)

[0369] The same logic is used for the 'TF' category.

[0370] An additional set of LCCA results (Table 15) (also in the form of Fab pairs, obtained after additional design cycles) are included in a separate table. This set of data is arranged in order of decreasing pair specificity and contains a few limited data on thermal stability.

[0371] The results in Tables 14 and 15 show that our chip design method results in the realization of the preferential pairing of H1-L1 across different groups of designs and their variants relative to H1-L2 and H2-L2 relative to the preferential pairing of H2-L1. These designs are usually divided into two main categories: electrostatic (based on the specificity driver utilizing hydrogen bonding or charge-to-charge interaction) and spatial complementation. For two LCCA designs, such pairing specificity varies from medium to ~100% correct pairing. As shown in Table, some designs in these designs do not affect thermal stability (Tm) or antigen binding affinity, but some display different degrees of influence on these two properties. In addition, successful designs are present in constant and variable domains, and are present in domain design combination forms.

[0372] Example 11: Evaluation of preferential pairing of heterodimers in co-expression panels containing constant domain and / or variable domain modifications in mixed Ab or pure Ab heavy chain and light chain Fab formats

[0373] In the system of heterodimer to deriving from different Ab (referring to D3H44) or two different antibodies, test some design described in the aforementioned embodiment, whether to cause preferential pairing in the system of these types with the design of assessment coexpression group.Test many different systems.In one example, a kind of heterodimer is to deriving from the D3H44 heavy chain and light chain existing with Fab form and the second heterodimer is to deriving from the pertuzumab heavy chain and light chain existing with Fab form.In another example, a kind of heterodimer is to deriving from the D3H44 heavy chain and light chain existing with Fab form, and the second heterodimer is to deriving from the trastuzumab heavy chain and light chain existing with Fab form.

[0374] Prepare construct as described in Example 1, described construct encoding comprises D3H44IgG, pertuzumab and trastuzumab heavy chain and light chain that exist with Fab form according to the amino acid modified design of coexpression group.The basic DNA sequence of the basic DNA sequence of pertuzumab heavy chain, the basic DNA sequence of the light chain of pertuzumab, the basic DNA sequence of trastuzumab heavy chain and the basic DNA sequence of the light chain of trastuzumab are shown in Table A, A1 and A2.Through site-directed mutagenesis, amino acid modified is introduced into these sequences, or synthesizes described DNA sequence dna, comprises the amino acid modified from the base sequence described in Example 1.

[0375] The ability of the heterodimer designs to preferentially pair was evaluated as described in Examples 2 and 3, except for the fact that when mixed systems were tested, the chains used were optionally from different Fabs.

[0376] The results are shown in Table 16. The amino acid modifications shown in Table 16 were identified with reference to the amino acid sequences of the heavy chain of D3H44 and the light chain of D3H44, the heavy chain of pertuzumab and the light chain of pertuzumab, and the heavy chain of trastuzumab and the light chain of trastuzumab. See Tables A, A1 and A2.

[0377] The designs of representative and diverse subsets that exhibited successful preferential pairing in the D3H44 system were tested in different systems (trastuzumab (TRAS) and pertuzumab (PERT)) as well as in mixed systems (D3H44 / TRAS, D3H44 / PERT, and TRAS / PERT) (as with D3H44LCCA, the constructs lacked Fab disulfides).

[0378] The data are provided in LCCA (Table 16) and Fab pair format (Table 17). The LCCA data reflect the minimum 'competition unit' (i.e. H1-L1:H1-L2) and is the best format for interpreting whether the LCCA design can be successfully transferred between Fabs. Analysis of the Fab pair format (including the second Fab pair (i.e. H2-L2:H2-L1)) further illustrates the extent of translation into the complete design (i.e. H1-L1 and H2-L2) and its effectiveness in these different Fab systems. In addition to the ratio of H1-L1:H1-L2, we also provide the relative tendency of correct pairing relative to incorrect pairing as a scalar, where scalar=ln(H1-L1:H1-L2).

[0379] In some mixed systems, inherent cross-system pairing preferences were observed (e.g., H_D3H44 pairs were preferentially paired with L_PERT compared to L_D3H44) (Table 17). In some instances, the preference of H_D3H44 to pair with L_PERT was also supported by thermal stability measurements (Tm), which showed that the H_D3H44-L_PERT species was more stable than H_D3H44-L_D3H44. Therefore, together with the amount of the actual species measured by the report, standardized data for the behavior of each WTLCCA system (REF) are provided in the form of ΔScalar (VAR-REF_WT), where ΔScalar = ln (H1-L1: H1-L2 / H2-L2: H2-L1). This metric is an indicator of the actual effectiveness of the LCCA design. The data included in the table include designs that produce equivalents with a pairing specificity of 55%: 45% or greater (i.e., ΔScalar (VAR-REF_WT) > 0.2).

[0380] Unlike in the D3H44 system, some species ratios appear to be sometimes affected by the light chain tag in the WT PERT system and in the WTTRAS (to a lesser extent) system (Table 18). This appears to be due to a random event of HA-tag cutting (LC / MS evidence is available when the system is tested in Mab form), rather than the tag interfering with pairing. Therefore, the tag is considered when providing results in the relevant tables.

[0381] An overview of the results expressed in the form of LCCA and Fab pairs across different systems is reported in Tables 19 and 20. The results show the design transferability across different tested Fab systems. These results do not necessarily reflect that some designs are better than other designs; they do not reflect a wider range of portability assessments. The successful LCCA designs (e.g., median ΔScalar (VAR-REF_WT) > 0.2) in two or more systems shown in Table 19 constitute the LCCA designs of about 30% of the tests in these different systems. This represents transferability, depending on the specific system and design. Therefore, a collection with a unique design allows for the generation of bispecific pairs for many systems and highlights the utility of the library of the design group that can be evaluated in the context of any target antibody (or antibody pair). The present embodiment shows that mutation / design groups can be used to realize the preferential pairing of heterodimers in the co-expression group containing constant domains and / or variable domain modifications in the form of mixed Ab or pure Ab heavy chains and light chain Fabs.

[0382] Example 12: Evaluation of preferential pairing of heterodimers in D3H44 heavy and light chains in co-expression panels containing constant domain and / or variable domain modifications in full-length heavy chain (Mab) format

[0383] Heterodimer co-expression panel designs were evaluated to determine if they also allowed preferential pairing in a format where the heavy chain was a full-length heavy chain rather than a Fab portion (Mab format).

[0384] Preparation of constructs:

[0385] The following preparation coding comprises the amino acid modified D3H44IgG heavy chain and the construct of light chain according to the design of coexpression group.As described in Example 1, prepare D3H44 Fab light chain construct.Except producing full-length D3H44 heavy chain by attaching the IgG1*01 DNA sequence encoding hinge-CH2-CH3 domain to the C-terminal of D3H44 Fab heavy chain, as described in Example 1, prepare D3H44 heavy chain sequence.Notably, remove the classic C-terminal heavy chain lysine residue to prevent the LC-MS signal heterogeneity caused by the C-terminal lysine clipping (Lawrence W.Dick Jr. et al., Biotechnol.Bioeng.(2008)100:1132-43).

[0386] Mab assay formats

[0387] The ability of the D3H44 heavy and light chains to preferentially pair was assessed as follows: one full-length D3H44 heavy chain construct was co-expressed with two unique D3H44 light chain constructs, resulting in three possible antibody species: H1-L1:H1-L1, H1-L2:H1-L2, and H1-L1:H1-L2 (see Fig.10 ). The relative light chain pairing specificity (based on the preferred H1-L1:H1-L1 species relative to the amount of the other species) was determined using LC-MS after protein A (pA) purification. If possible, the chains were not labeled as long as the 3 possible Mab species generated by co-transfection of the 3 chains differed from each other by at least 50 Da. When mass differences ruled out this possibility, at least one of the light chains was constructed with an N-terminal HA or FLAG tag fusion to provide sufficient mass distinction between the species. As described in Example 2, the heavy chains were kept in limited amounts (i.e., HC < L1 + L2).

[0388] Transfection Methods for Mab Assay Formats

[0389] The co-expression set containing one heavy chain and two light chain constructs prepared as described in Example 12 was transfected into CHO-3E7 cells as follows. CHO-3E7 cells were cultured at 1.7-2×10 6 Cells were cultured at a density of 10 cells / ml at 37°C in FreeStyle(TM) F17 medium (Invitrogen cat# A-1383501) supplemented with 4 mM glutamine and 0.1% Pluronic F-68 (Invitrogen cat# 24040-032). A total volume of 50 ml was transfected with a total of 50 ug DNA at a DNA:PEI ratio of 1:2.5 using PEI-pro (Polyplus cat# 115-010). 24 hours after the addition of the DNA-PEI mixture, the cells were transferred to 32°C. On day 7, the supernatant was tested for expression by non-reducing SDS-PAGE analysis followed by Coomassie blue staining (to visualize the bands). The HC:L1:L2 ratio used was 1:1:1.

[0390] Mass spectrometry for Mab assay formats

[0391] After protein A purification and deglycosylation, mass spectrometry was used to assess the degree to which the D3H44 light chain preferentially paired with the D3H44 heavy chain in the co-expression group. The purified samples were deglycosylated using PNGaseF as follows: 0.2U PNGaseF / μg of antibodies in 50mMTris-HCl pH 8.0, incubated overnight at 37°C, with a final protein concentration of 0.45mg / mL. Protein samples were analyzed by LC-MS using an Agilent 1100 HPLC system coupled to a LTQ-Orbitrap XL hybrid mass spectrometer (ThermoFisher Scientific) via a high-flow electrospray interface. Samples (2.5 μg) were injected onto a 2.1×10 mm Poros R2 column (Applied Biosystems) and eluted using the following gradient conditions: 0-3 minutes: 20% solvent B; 3-6 minutes: 20-90% solvent B. Solvent A was 0.1% formic acid in water, and solvent B was 65% ACN, 25% THF, 9.9% ddH2O, 0.1% formic acid. The flow rate was 1 mL / min. The flow was split after the column and injected into the electrospray interface at 100 μL / min. The column and solvent were heated to 80°C to improve protein peak shape. The LTQ-Orbitrap XL was calibrated using ThermoFisher Scientific's LTQ Positive Ion ESI calibration fluid (caffeine, MRFA and Ultramark 1621) and tuned using a 10 mg / mL CsI solution. The cone voltage (source fragmentation setting) was 40 V, the FT resolution was 7,500 and the scan range was m / z 400-4,000.

[0392] Protein spectra were deconvoluted using the MaxEnt module of the MassLynx instrument control and data analysis algorithm (Waters). Briefly, the raw protein LC-MS data were first opened in QualBrower (spectral viewing module of Xcalibur (Thermo Scientific)) and converted to be compatible with MassLynx using Databridge (a file conversion program provided by Waters). The converted protein spectra were viewed in the spectral module of MassLynx and deconvoluted using MaxEnt. The abundance of different antibody species in each sample was determined directly from the resulting molecular weight profiles.

[0393] The results are shown in Table 21. The amino acid modifications shown in Table 21 were identified with reference to the amino acid sequences of the D3H44 heavy chain and the D3H44 light chain. See Tables A, A1 and A2.

[0394] A subset of designs that showed successful preferential pairing in the D3H44 system, which also represented the diverse design set, was selected for evaluation of the transferability of the format, ie, from LCCA utilizing Fab structures to Mab competition assays (MCA) based on Mab structures.

[0395] As shown in Table 21, the D3H44 WT for Mab competitive assays exhibits a deviation from the expected theoretical species distribution of 50% H1-L1_H1-L2 and 25% of H1-L1_H1-L1 and H1-L2_H1-L2, respectively, where the light chain differs in the presence or absence of a tag (HA or FLAG). This deviation may be a result of tag-dependent expression levels rather than tag-dependent pairing (e.g., based on experiments performed at different WT H1:L1:L2 ratios, and indirect observations based on designed behavior). The median values ​​of all measured species between variants including experimental repetitions and / or differences in tag identity are also included.

[0396] Examples of LC / MS spectra of two successful Mab assay variants comprising a given design ( Fig.11 ) Fig.12 In the case of these variants, most of the measured species matched the desired H1-L1_H1-L1 species. Fig.13 The expression profiles reported in, UPLC-SEC (shown only for the H2-L2_H2-L2 species (in the figure, denoted as H2-L2)) and DSC thermograms were generally used for the characterization of variants performed for certain hits. In this particular case, these data showed a well-expressed, homogeneous species with relatively little impact on Fab stability.

[0397] The result display in table 21 is for being remarkably successful in LCCA design, and the transferability in Mab form is also achieved.There are 11 demonstrations equal to or greater than 25% theoretical correct pairing species in 12 LCCA designs of test.These data show that the LCCA form used as the initial screening of design library is suitable for evaluating and / or predicting the success of Mab form.

[0398] Example 13: Evaluation of preferential pairing of heterodimers in co-expression panels containing constant domain and / or variable domain modifications in a bispecific antibody format

[0399] Heterodimer co-expression panel designs were evaluated to determine if they also allow preferential pairing of bispecific Mab antibody formats. In this example, the Fc region of the full-length heavy chain of each heterodimer was asymmetrically modified to promote heterodimerization (vs. homodimerization) of the unique heavy chains.

[0400] Preparation of constructs:

[0401] The heterodimer coexpression set design was tested in the context of the following bispecific antibodies: a) D3H44 / pertuzumab, b) D3H44 / trastuzumab, c) D3H44 / ramucirumab and d) trastuzumab / ramucirumab. In addition to introducing complementary Fc heterodimerization mutations into each of the two heavy chains of the coexpression design set, D3H44, pertuzumab and trastuzumab heavy and light chains containing amino acid modifications in the constant and / or variable domains were prepared as described in Example 12. Ramucirumab heavy and light chains were prepared based on the basic DNA sequence of the ramucirumab heavy and light chains. See Table A. The CH3 sequence of the heavy chain includes the following amino acid modifications:

[0402] a) D3H44 / Pertuzumab: Chain A: T371V_T389L_K420L_T422W, Chain B: T371V_L372Y_F436A_Y438V

[0403] b) D3H44 / trastuzumab: Chain A: T371V_T389L_K420L_T422W, Chain B: T371V_L372Y_F436A_Y438V

[0404] c) D3H44 / ramucirumab: Chain A: T371V_T389L_K420L_T422W, Chain B: T371V_L372Y_F436A_Y438V

[0405] d) Trastuzumab / ramucirumab: Chain A: T371V_T389L_K420L_T422W, Chain B: T371V_L372Y_F436A_Y438V

[0406] Assay format (SMCA)

[0407] As described below, the heterodimer coexpression group design is evaluated to preferentially pair with the ability to form bispecific antibodies. The assay is based on coexpression of 4 chains (2 from A1, two from Ab2) and the presence of the correctly formed bispecific antibodies using mass spectrometry (LC-MS). The assay is performed as follows. Fig.14Provide a schematic diagram depicting 4 starting polypeptide chains and potential products produced by the coexpression of these starting polypeptide chains in the absence of preferential pairing between the heavy chain and light chain of the heterodimer pair. Two full-length heavy chain constructs are coexpressed with two unique light chain constructs to produce possible antibody species: H1-L1: H1-L1, H1-L2: H1-L2, H1-L1: H1-L2, H2-L1: H2-L1, H2-L2: H2-L2, H2-L1: H2-L2, H1-L1: H2-L1, H1-L2: H2-L2, H1-L2: H2-L1 and H1-L1: H2-L2. The latter species is a correctly paired heterodimer (see figure below). Relative light chain pairing specificity (according to the preferred species H1-L1: H2-L2 species relative to the amount of other species) is determined using LC-MS after protein A purification. If possible, chains were not labeled as long as all Mab and half-Ab species differed from each other by at least 50Da. When mass differences ruled out this possibility, at least one of the light chains was constructed using an N-terminal HA tag fusion to provide sufficient mass distinction between species. We refer to this assay, which includes the expression and screening steps of bispecific antibodies, as SMCA.

[0408] Transfection method

[0409] The co-expression set containing two heavy chain and two light chain constructs prepared as described in Example 1 was transfected into CHO-3E7 cells as follows. CHO-3E7 cells were cultured at 1.7-2×10 6 Cells were cultured at a density of 10 cells / ml at 37°C in FreeStyle(TM) F17 medium (Invitrogen cat#A-1383501) supplemented with 4 mM glutamine and 0.1% Pluronic F-68 (Invitrogen cat#24040-032). A total volume of 20 ml was transfected with a total of 20 ug DNA at a DNA:PEI ratio of 1:2.5 using PEI-pro (Polyplus cat#115-010). 24 hours after the addition of the DNA-PEI mixture, the cells were transferred to 32°C. On day 7, the supernatant was tested for expression by non-reducing SDS-PAGE analysis followed by Coomassie blue staining (to visualize the bands). Initially, the H1:H2:L1:L2 ratio used was kept neutral (15:15:35:35) to assess expression efficiency. A set of H1:H2:L1:L2 DNA ratios were then tested in CHO expression to assess which conditions produced a mixture that reflected the theoretical distribution of the different species when all chains were wild type. These ratios compensated for natural differences in expression levels of the two different antibodies and / or inherent pairing preferences between heavy and light chains.

[0410] SPR biosensor assay

[0411] EDC: 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride; sNHS: N-hydroxysulfosuccinimide; SPR: surface plasmon resonance; EDTA: ethylenediaminetetraacetic acid; HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; TF: tissue factor.

[0412] SPR equipment. GLC sensor, Biorad ProteOn amine coupling kit (EDC, sNHS and ethanolamine) and 10 mM sodium acetate buffer were purchased from Bio-Rad Laboratories (Canada) Ltd. (Mississauga, ON). Recombinant Her-2 protein was purchased from eBioscience (San Diego, CA). PBS electrophoresis buffer containing 0.05% Tween20 (PBST) was purchased from Teknoca Inc. (Hollister, CA). Goat polyclonal anti-human Fc antibody was purchased from Jackson Immuno Research Laboratories Inc. (West Grove, PA).

[0413] All surface plasmon resonance measurements were performed using a BioRad ProteOn XPR36 instrument (Bio-Rad Laboratories (Canada) Ltd. (Mississauga, ON)) with PBST electrophoresis buffer at a temperature of 25°C. The anti-human Fc capture surface was generated using a GLM sensor chip activated by injecting a 1:5 dilution of a standard BioRad sNHS / EDC solution at 100 μL / min (for 140 seconds) in the analyte (horizontal) direction. Immediately after activation, a 10 μg / mL solution of anti-human Fc antibody in 10 mM NaOAc pH 4.5 was injected at a flow rate of 25 μL / min in the ligand (vertical) direction until approximately 3000 resonance units (RU) were immobilized. Remaining active groups were quenched by injecting 1 M ethanolamine at 100 μL / min for 140 seconds in the analyte direction, which also ensured that an intermediate point of the mimic activation was generated for blank reference. The screening of antibody variants for binding to Her2 or TF antigen targets is performed in the following two steps: the antibody variants are indirectly captured on the anti-human Fc surface in the ligand direction, followed by simultaneous injection of 5 concentrations of purified antigen and a buffer blank for double reference in the analyte. First, a buffer injection (lasting 30 seconds) at 100uL / min in the ligand direction is used to stabilize the base line. For each antibody variant capture, the unpurified variant in the cell culture medium is diluted to 4% in PBST. 1 to 5 variants or controls are injected simultaneously (lasting 240 seconds) at a flow rate of 25μL / min in a single ligand channel. This results in approximately 400 to 600RU of capture on the anti-human Fc surface. If necessary, leave the first ligand channel empty to serve as a blank control. After the capture step, two buffer injections were performed in the analyte direction to stabilize the baseline, followed by simultaneous injections (for 120 seconds) of 60nM, 20nM, 6.7nM, 2.2nM and 0.74nM antigen (TF or Her2) at 50μL / min together with a buffer blank for a 300 second dissociation period. The capture antibody surface was regenerated by an 18 second pulse of 0.85% phosphoric acid (18 seconds at 100μL / min) to prepare for the next injection cycle. The sensorgrams were aligned and double referenced using buffer blank injections and midpoints, and the resulting sensorgrams were analyzed using ProteOn Manager software v3.1. The double referenced sensorgrams were fitted to a 1:1 binding model. The Rmax values ​​for each antigen were normalized for the antigen capture level of each variant and compared to the 100% control.

[0414] The stability of the bispecific antibodies was tested as described in Example 5. LCMS analysis of the bispecific antibodies was performed using the method described in Example 12.

[0415] The D3H44LCCA designs of many selections were tested in SMCA form. In addition, some designs were only directly evaluated in SMCA form. Most designs were tested in the D3H44 / pertuzumab system. The design of a subgroup selected in the system (comprising multiple designs) was further tested in 3 other bispecific Mab systems: D3H44 / trastuzumab, D3H44 / ramucirumab (RAMU) and trastuzumab / ramucirumab systems.

[0416] The inherent cross-system light chain / heavy chain preference observed in the LCCA experiment for D3H44 / pertuzumab and D3H44 / trastuzumab (Example 11) is also reproducible in the full-length antibody format. Such, similar behavior is also observed for the other two systems, although there are different cross-system pairing preferences.

[0417] It is usually not possible to experimentally distinguish the desired bispecific species H1-L1_H2-L2 from the mismatch type: H1-L2_H2-L1 based on LC / MS. Thus, when the bispecific content is reported in the table, it cannot be completely excluded that it does not contain mismatch species of this type. However, the extremely low contents observed for species such as H1-L2_H1-L2 and H2-L1_H2-L1 and H1-L2 and H2-L1 half antibodies indicate that if contamination of bispecific species occurs, it is only in small amounts. All other species present in a particular sample measured by LC / MS are included in the table. In most cases, the MS peak is accompanied by side peaks; however, these cannot be annotated as bispecific species alone. These adduct species are not considered when normalizing the MS peak intensity. The numbers representing the side peaks in the table are obtained by comparing the intensity with the intensity of the main bispecific peak. Some preliminary analyses indicate that the side peaks are associated with the presence of the light chain tag (involving the formation of adducts or heterogeneity of the cleavage of the leader peptide) and that they may represent the main peak species.

[0418] Finally, all paired species, except the mismatch species, are summed up to obtain the ratio reported in the paired: mismatch column. This is further used to calculate ΔScalar(VAR-REF_WT) according to the mathematical method described in Example 11 to show the strength of a particular design. Within the system, the scalar index values ​​of the designs are sorted in descending order.

[0419] Due to the cross-system natural preference for pairing of light and heavy chains described in Example 12, the DNA ratio of H1:L1:H2:L2 needs to be changed (e.g., overexpression of H1 (D3H44) relative to H2 (PERT)) to produce the highest content of the desired bispecific species (H1-L1_H2-L2). The optimal ratio can therefore also vary to some extent with the specific design. The DNA ratio mainly affects the actual ratio of bispecific species: paired half antibody species (usually one type: H1-L1 or H2-L2). Examples are included in Table 23, where DNA titration varies with this ratio, but the overall ratio of paired: mismatched species is relatively constant.

[0420] In the case of the D3H44 / pertuzumab system, a limited set of DNA titrations were performed for most designs. Data on the ratio of the species content that resulted in the highest pairing: mispairing are shown in Tables 22 and 24. For the other 3 systems, transfection was performed only at one ratio, as reported in the table. If the experiment was repeated at the reported ratio, the mean value is included in the table. Information about tag identity is not included because no tag influence is observed for WT (Table 26). The WT reference provided was selected in this way to represent the most common DNA ratio between the reported design data.

[0421] LC / MS analysis was performed on samples stored at pH 4 as well as pH 7. Experiments on samples stored at pH 7 were tested with the D3H44 / trastuzumab, D3H44 / ramucirumab (RAMU) and trastuzumab / ramucirumab systems. The data are therefore presented in two tables (Tables 22 (pH 4) and 24 (pH 7)). A satisfactory correlation was observed for the paired: mismatched species ratios between the two experiments, suggesting that pH is unlikely to play a significant role in the LC / MS experiments, i.e., in the characterization of the type of species present.

[0422] Fig.15 Provides a Fig.11 and Fig.12 Results of LC / MS analysis of the designed bispecific constructs targeting TF and Her2 in MCA format are shown in. Nearly 92% of the preferred bispecific antibodies with the correct pairing of the mandatory heavy and light chains were observed. Fig.16 The expression profiles of the protein product in the supernatant (after Protein A purification in SDS PAGE) and the SEC profiles of the Protein A purified compounds are provided. Fig.17 The bispecific target binding properties of the bispecific molecules were provided, first against two targets (TF and Her2 individually) and subsequently in a sandwich (bridging) format.

[0423] In some cases, the mutations at H / S 115 and H / S 156 characterized in a limited set of variants were not part of the actual design, but were instead added for practical reasons to obtain the necessary mass differences (for the purposes of LC / MS experiments). These amino acids are located on the surface of the constant domains, sufficiently far from the actual design set of mutations and are not expected to affect the behavior of the antibody.

[0424] The thermal stability and antigen affinity of many designs were evaluated in the D3H4 / pertuzumab system and are shown in Table 25. Tm values ​​(in italics) were annotated manually. The homodimeric Mab control (WT) showed the following stability ranges of thermal melting (Tm) for products expressed in transfection replicates (at pH 7): PERT (72.03-77.72) and D3H44 (77.97-78.88). m ). The wider range observed for pertuzumab may be due to its intrinsic properties. Affinity measurements were performed only after pA purification. The observed KD ranges for homodimeric Mabs were: TF: 0.04-0.076nM, HER2: 1.84-6.3nM. In most cases, the two melting transitions can be expected to be due to the different ranges of the two Fabs indicated above. In cases where only one Tm value is reported, it is potentially increased for one of two reasons: changes in pertuzumab stability and / or destabilization of the design of D3H44 Fab (which may be consistent with the destabilization of pertuzumab Fab). The results show that the choice of SMCA as a body can vary from variants that minimally affect thermal stability and antigen binding to variants that exhibit the described behavior to varying degrees.

[0425] Some designs were tested using two possible Fc mutation layouts. These can be identified by having the same unique identifier and are marked with *D3H44 (or using # before the unique identifier setting in Table 25). As can be seen from Table 22, with limited exceptions, the layout of the Fc mutations does not seem to affect the type of mismatch results of the pairing.

[0426] The SMCA result shows that the design of the representative diversity group of a considerable number can overcome the natural cross-category pairing tendency of Mab form.Among these designs, the design of the test close to 1 / 4 falls into high pairing: mismatch ratio (>=80: 20) category (number is provided for the system of more exhaustive test, D3H44 / pertuzumab).The ratio of the design effectiveness across different systems shows different degrees of transferability.In addition, the design listed in about 60% of the table 27 (transferred to SMCA form from D3H44LCCA) causes high priority pairing (>75%: 25%) in at least one of the SMCA systems of 4 tests.

[0427] For designs tested in systems other than D3H44 / pertuzumab, the design layout was also reversed for the binding domain, such as the H1-L1 design on the D3H44 binding domain, the H2-L2 design on TRAS, and the H1-L1 design on TRAS and the H2-L2 design on D3H44. The results show that the efficacy of the design can be affected by such a reversed layout in most cases.

[0428] The results discussed above indicate that the transferability of a design can be affected by the combination of antigen binding domains coupled to the driving potential of the design components (e.g. H1L1L2 can drive better than H2L2L1). The results indicate that a large amount of basic core design work covers a range of Mab pairs to form bispecific Abs with greater than 75% selectivity pairing.

[0429] Example 14: Molecular Modeling and Computer-Guided Engineering of Fab Interfaces

[0430] We used a method guided by structure and computer molecular modeling to generate a library of heavy and light chain mutation designs that can be screened in the context of other antibodies or fragments thereof to identify mutations that display the desired specificity in the target antibody. The design strategy for engineering priority HC-LC pairings involves first identifying a representative Ab that works together (i.e., D3H44). The key criteria for this Ab are shown in Table 28.

[0431] Table 28:

[0432]

[0433]

[0434] As indicated in Table 28, the key criteria provided by this antibody was that it was a human / humanized Ab, with commonly used VH and VL subgroups and minimal framework region mutations. In addition, the structural consideration was that the VH:VL interdomain angle should be close to the average observed for the Ab. After the Fab was selected, chip analysis of the Fab interface was performed with the goal of identifying and understanding important residues. A two-pronged approach was adopted. First, a global analysis of sequence conservation across the Fab variable and constant interfaces was performed by sequence and structure alignment of publicly available Abs. The alignment of constant and variable domain sequences from different antibody subgroups is shown in Figure 6. In parallel, the constant and variable domain sequences were aligned using Fig.18 Many molecular modeling tools listed in (e.g. ResidueContacts TM ) analyzed the crystal structure interface D3H44. These analyses resulted in the identification of a list of hotspot positions for engineering priority HC-LC pairings. The hotspot positions determined from this analysis are listed in Table 29.

[0435] Table 29: hot spots amino acid positions on the interface of heavy chain (H) and light chain (L) in typical Fabs derived from people's VH and VLκ chains. These positions and amino acids are also the most conservative in VLλ chains. Except for a few being located in the DCR3 ring, these amino acids are mainly framework residues. The amino acids in the parental D3H44 sequence (using Kabat numbering) are provided in Tables A1-A2.

[0436] H(Kabat) L(Kabat) V37 Y36 Q39 Q38 L45 P44 W47 L89 F100 F98 W103 F116 L124 F118 A139 V133 F174 L135

[0437] Subsequently, potential mutations and designs at simulated hotspot locations and positions adjacent to the target hotspots in the 3D crystal structure were performed by in silico mutagenesis and packaging / modeling (using Zymepack TM ) were identified and scored based on a number of factors including steric and electrostatic complementarity. Fig.11 Provided are the hotspot positions on the heavy chain and light chain interfaces in a limited number of variable domains, and how mutations can be introduced at these interface positions to help the selective pairing of the forced chains, while being unfavorable for the formation of incorrect chain pairing. Spatial complementarity is modeled, and the complementarity is also calculated based on energy factors such as van der Waals stacking, cavitation effect, and close contact of hydrophobic groups. Similarly, electrostatic interaction energy is modeled, and evaluation is performed based on Coulomb interactions, hydrogen bonds, and desolvation effects between charges. Simulate the preferred heavy chain and light chain pairing models such as H1: L1 (or H2: L2) and incorrect pairings such as H1: L2 (and H2: L1) obtained by introducing target mutations, and calculate relative spatial and electrostatic scores. This allows us to determine whether a specific mutation group leads to favorable energy, i.e., preferred (forced) heavy chain-light chain pairs relative to incorrect (non-forced) pairs of greater spatial or electrostatic complementarity. The calculated spatial and electrostatic energies are components of the free energy associated with light chain and heavy chain pairing. Therefore, greater spatial and electrostatic complementarity indicates a greater free energy change associated with the pairing of forced pairs relative to the pairing of non-forced pairs. Larger space or electrostatic complementarity leads to the preferential (selective) pairing of mandatory heavy chain and light chain relative to non-mandatory pairing, and can be detected according to the percentage of the two products (mandatory pairing to non-mandatory pairing heavy chain and light chain) after co-expression. Larger space or electrostatic complementarity in mandatory pairing can also be observed according to better / larger thermal stability relative to non-mandatory pairing. List the candidate list of candidate designs and rate them. Use LCCA system in vitro test design at the beginning. Further improve the design of moderate performance of non-optimal biophysical characteristics such as poor HC-LC pairing specificity, low thermal stability or reduced antigen binding affinity through additional rounds of chip design and in vitro screening. The best design is then tested in bispecific Mab format; in vitro screening is mainly carried out by SMCA format.

[0438] Example 15: Generation of bispecific antibodies derived from Mab1 and Mab2 using a library of bispecific antibody mutation design panels

[0439] In one embodiment, a bispecific antibody mutation design group is provided herein that aims to selectively form a bispecific antibody derived from two classical antibodies Mab1 and Mab2 (composed of antigen-binding fragments Fab1 and Fab2, respectively). The design group consists of homologous mutations corresponding to Fab1, Fab2 and Fc, respectively. Mutations are introduced at the interface of the light chain and the heavy chain of Fab1 to achieve selective pairing between two mandatory chains in the presence of the competitive light chain and heavy chain of Fab2. By introducing favorable complementary mutations in two mandatory light chains and heavy chains based on the space, hydrophobicity or electrostatic complementarity between certain hot spot framework residues on the interface to achieve selective pairing, while the residues containing these mutations are not conducive to the interface interaction of non-mandatory chain pairs. In each design group, selective pairing mutations can also be introduced at the interface of the light chain of Fab2 and the antibody to achieve selective pairing between the two mandatory chains in the presence of the competitive light chain and heavy chain of Fab1. The purpose of the mutation is to reduce the mismatching of the light chain from Fab1 with the heavy chain of Fab2, and vice versa. Mutations are introduced on the Fc interface to achieve alternative pairing of heavy chains to form asymmetric antibody molecules containing two different heavy chains.

[0440] Engineering can usually lead to deleterious effects on certain interface residue positions of the light chain and heavy chain of the antibody, such as the loss of the antigen binding affinity, stability, solubility, and aggregation tendency of the antibody. Many related properties can be affected, such as kon and koff rates, melting temperature (Tm), stability to stress conditions such as acid, alkali, oxidation, freezing / thawing, stirring, pressure, etc. This is usually affected by the complementary determining region (CDR) of the target antibody. In view of the fact that the CDR of the antibody is usually different, the impact of the mutation design group can be different between all antibodies. In another embodiment, many different bispecific mutation design groups constituting the library of the bispecific antibody mutation design group are defined, and the mutation design group comprises mutations at different hot spots on the interface. The library of such a bispecific antibody mutation group is shown in Table 30. Provided herein is a method for producing a bispecific antibody with significant purity relative to an antibody-like structure containing incorrect pairing (derived from any two available antibodies Mab1 and Mab2). After introducing homologous mutations in each mutation design group, the light and heavy chains of Mab1 and Mab2 are co-expressed, and the expressed antibody products are analyzed and screened to evaluate the purity of the preferred bispecific antibody relative to other expressed Mab-like species in the protein product. In some embodiments, the analytical screening method may be based on LC-MS technology. In some embodiments, the analytical screening method may be based on charge-based separation such as capillary isoelectric focusing (cIEF) technology. An example of a screening technique is shown in Example 13 based on the SMCA method. In some embodiments, a significant purity of the bispecific antibody is defined as greater than 70% of all obtained Mab-like species in the expressed protein product. In some embodiments, a significant purity of the bispecific antibody is defined as greater than 90% of all obtained Mab-like species in the expressed protein product. The method for preparing and selecting a bispecific Mab design group derived from Mab1 and Mab2 is schematically shown in Fig.19 middle.

[0441] All publications, patents, and patent applications mentioned in this specification are incorporated by reference into the specification to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference herein.

[0442] While the invention has been particularly shown and described with reference to a preferred embodiment and various alternative embodiments, it will be understood by those skilled in the relevant art that various changes in form and detail may be made without departing from the spirit and scope of the invention.

[0443] All references, issued patents, and patent applications cited in the body of this specification are hereby incorporated by reference in their entirety for all purposes.

[0444]

[0445]

[0446]

[0447]

[0448]

[0449]

[0450]

[0451]

[0452]

[0453]

[0454]

[0455]

[0456]

[0457]

[0458]

[0459]

[0460]

[0461]

[0462]

[0463]

[0464]

[0465]

[0466]

[0467]

[0468]

[0469]

[0470]

[0471]

[0472]

[0473]

[0474]

[0475]

[0476]

[0477]

[0478]

[0479]

[0480]

[0481]

[0482]

[0483]

[0484]

[0485]

[0486]

[0487]

[0488]

[0489]

[0490]

[0491]

[0492]

[0493]

[0494]

[0495]

[0496]

[0497]

[0498]

[0499]

[0500]

[0501]

[0502]

[0503]

[0504]

[0505]

[0506]

[0507]

[0508]

[0509]

[0510]

[0511]

[0512]

[0513]

[0514]

[0515]

[0516]

[0517]

[0518]

[0519]

[0520]

[0521]

[0522]

[0523]

[0524]

[0525]

[0526]

[0527]

[0528]

[0529]

[0530]

[0531]

[0532]

[0533]

[0534]

[0535]

[0536]

[0537]

[0538]

[0539]

[0540]

[0541]

[0542]

[0543]

[0544]

[0545]

[0546]

[0547]

[0548]

[0549]

[0550]

[0551]

[0552]

[0553]

[0554]

[0555]

[0556]

[0557]

[0558]

[0559]

[0560]

[0561]

[0562]

[0563]

[0564]

[0565]

[0566]

[0567]

[0568]

[0569]

[0570]

[0571]

[0572]

[0573]

[0574]

[0575]

[0576]

[0577]

[0578]

[0579]

[0580]

[0581]

[0582]

[0583]

[0584]

[0585]

[0586]

[0587]

[0588]

[0589]

[0590] Table 30

[0591]

[0592]

[0593]

[0594]

[0595]

[0596]

[0597]

[0598]

[0599]

[0600]

[0601]

[0602]

[0603]

[0604] Table 31

[0605]

[0606]

Claims

1. An isolated antigen-binding polypeptide construct comprising at least a first heterodimer and a second heterodimer, The first heterodimer comprises a first human or humanized immunoglobulin G heavy chain polypeptide sequence H1 and a first human or humanized immunoglobulin light chain polypeptide sequence L1, and binds to a first epitope; and the second heterodimer comprises a second human or humanized immunoglobulin G heavy chain polypeptide sequence H2 and a second human or humanized immunoglobulin light chain polypeptide sequence L2, and binds to a second epitope, wherein at least one of the H1 or L1 sequences of the first heterodimer is different from the corresponding H2 or L2 sequence of the second heterodimer, H1 and H2 each comprise at least a heavy chain variable domain and a heavy chain constant domain 1; and L1 and L2 each comprise at least a light chain variable domain and a light chain constant domain; and wherein a) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitutions 143K and 146G; and L2 comprises amino acid substitutions 124E and 133D; b) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 179K; and L2 comprises amino acid substitutions 124E, 160E, and 180E; c) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 179R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; d) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 186R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; e) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124R, 160K, and 178R; H2 comprises amino acid substitutions 146G and 179R; and L2 comprises amino acid substitutions 124E, 160E, and 178D; f) H1 ​​comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124R, 160K, and 178R; H2 comprises amino acid substitutions 146G and 186R; and L2 comprises amino acid substitutions 124E, 160E, and 178D; g) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124R, 160K, and 178R; H2 comprises amino acid substitutions 146G and 179K; and L2 comprises amino acid substitutions 124E, 160E, and 178D; h) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 186R; and L2 comprises amino acid substitutions 124E, 160E, and 178D; i) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 179R; and L2 comprises amino acid substitutions 124E, 160E, and 178D; j) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 160K and 178R; H2 comprises amino acid substitutions 146G and 179K; and L2 comprises amino acid substitutions 124E, 160E, and 178D; k) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124R, 160K, and 178R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; 1) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124K and 178R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; m) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124K and 178R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitution 124E; n) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143R; and L2 comprises amino acid substitution 124E; o) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 does not comprise an amino acid substitution that promotes preferential pairing; and L2 comprises amino acid substitution 124E; p) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitution 124E; q) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitution 124E; r) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitution 178E; s) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; t) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124K; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitution 124E; u) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124K; H2 comprises amino acid substitution 143R; and L2 comprises amino acid substitution 124E; v) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143R; and L2 comprises amino acid substitution 124E; w) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 does not comprise an amino acid substitution that promotes preferential pairing; and L2 comprises amino acid substitution 124E; x) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitution 124E; y) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitution 124E; z) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124K; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitution 124E; aa) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 160E, and 178D; bb) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitution 133E; cc) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143R; and L2 comprises amino acid substitution 133E; dd) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143R; and L2 comprises amino acid substitutions 124E and 133E; ee) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitutions 124E and 133E; ff) H1 ​​comprises amino acid substitutions 143E, 145T, and 179E; L1 comprises amino acid substitutions 124K and 178R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 178E, and 180E; gg) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 178E and 180E; hh) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 160E and 178E; ii) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 143K; and L2 comprises amino acid substitutions 124E and 178E; jj) H1 comprises amino acid substitutions 143E, 145T, and 188L; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E and 178E; kk) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 does not comprise an amino acid substitution that promotes preferential pairing; and L2 does not comprise an amino acid substitution that promotes preferential pairing; 11) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitution 124R; H2 comprises amino acid substitutions 146G and 179K; and L2 comprises amino acid substitutions 124E, 160E, and 180E; mm) H1 comprises amino acid substitutions 143E and 145T; L1 comprises amino acid substitutions 124R, 160K, and 178R; H2 comprises amino acid substitutions 146G and 179K; and L2 comprises amino acid substitutions 124E, 160E, and 180E; or nn) H1 comprises amino acid substitutions 143E, 145T, and 179E; L1 comprises amino acid substitutions 124K and 178R; H2 comprises amino acid substitution 186R; and L2 comprises amino acid substitutions 124E, 160E, and 180E; and The numbering of amino acid residues is according to Kabat, and the amino acid substitutions promote preferential pairing of H1 with L1 over L2, and preferential pairing of H2 with L2 over L1.

2. The construct of claim 1, wherein when H1, H2, L1 and L2 are co-expressed in a cell or mammalian cell, or when H1, H2, L1 and L2 are co-expressed in a cell-free expression system, or when H1, H2, L1 and L2 are co-produced, or when H1, H2, L1 and L2 are co-produced by a redox production system, H1 preferentially pairs with L1 over L2, and H2 preferentially pairs with L2 over L1.

3. The construct of any one of claims 1 to 2, wherein human or humanized L1 and / or human or humanized L2 is a kappa light chain polypeptide.

4. A construct according to claim 3, wherein when L1 and L2 are co-expressed with at least one of H1 and H2, the pairing of the modified H1-L1 or H2-L2 heterodimer pair is greater than the respective pairing observed in the corresponding H1-L1 or H2-L2 heterodimer pair without the amino acid substitution.

5. The construct according to claim 3, wherein the construct further comprises a H3 The invention relates to an Fc having a sequence, wherein the Fc is coupled to the first heterodimer and the second heterodimer through or without one or more linkers.

6. The construct of claim 5, wherein the Fc is human Fc, human IgG1 Fc, human IgAFc, human IgG Fc, human IgD Fc, human IgE Fc, human IgM Fc, human IgG2 Fc, human IgG3 Fc or human IgG4 Fc.

7. The construct of claim 5, wherein the Fc is a heterodimeric Fc.

8. The construct according to claim 7, wherein the Fc is in C H3 At least one of the sequences comprises one or more modifications that promote the formation of a heterodimeric Fc having a stability comparable to that of a wild-type homodimeric Fc.

9. The construct of claim 7, wherein the heterodimeric Fc is human IgG1, IgG2 or IgG4 Fc, and: a) C H3 One of the sequences comprises amino acid substitutions L351Y, F405A, Y407V; and the other comprises amino acid substitutions T366L, K392M, T394W; b) C H3 One of the sequences comprises amino acid substitutions L351Y, F405A, Y407V; and the other comprises amino acid substitutions T366L, K392L, T394W; c) C H3 One of the sequences comprises amino acid substitutions T350V, L351Y, F405A, Y407V; and the other comprises amino acid substitutions T350V, T366L, K392M, T394W; d) C H3 One of the sequences comprises the amino acid substitutions T350V, L351Y, F405A, Y407V; and the other comprises the amino acid substitutions T350V, T366L, K392L, T394W; or e) C H3 One of the sequences comprises the amino acid substitutions T350V, L351Y, S400E, F405A, Y407V; and the other comprises the amino acid substitutions T350V, T366L, N390R, K392M, T394W, Wherein C H3 The numbering of the amino acid residues of the sequences is according to the EU numbering system.

10. The construct of claim 5, wherein the Fc further comprises at least one C H2 sequence.

11. The construct of claim 10, wherein the Fc comprises one or more modifications to promote selective binding of Fc-γ receptors.

12. The construct of claim 5, wherein the Fc is coupled to the heterodimer via one or more linkers, or wherein the Fc is coupled to H1 and H2 via one or more linkers.

13. The construct of claim 12, wherein the one or more linkers are one or more polypeptide linkers.

14. The construct of claim 13, wherein the one or more linkers comprise one or more antibody hinge regions.

15. The construct of claim 13, wherein the one or more linkers comprise one or more IgG1 hinge regions.

16. The construct of claim 3, wherein the construct is multispecific or bispecific.

17. The construct according to claim 16, wherein the construct is multivalent or bivalent.

18. The construct of claim 17, wherein the construct is conjugated to a therapeutic agent or drug.

19. An isolated polynucleotide or a group of isolated polynucleotides comprising at least one sequence encoding a construct according to any one of claims 1 to 17.

20. A vector or a set of vectors comprising one or more of the polynucleotide or the set of polynucleotides according to claim 19.

21. An isolated cell comprising the polynucleotide or group of polynucleotides according to claim 19, or the vector or group of vectors according to claim 20.

22. A pharmaceutical composition comprising the construct according to any one of claims 1 to 18 and a pharmaceutically acceptable carrier.

23. A method for obtaining a construct according to any one of claims 1 to 17 from a host cell culture, the method comprising the steps of: (a) obtaining a host cell culture comprising at least one host cell containing one or more nucleic acid sequences encoding said construct; and (b) recovering the construct from the host cell culture.

Images