Methods of producing heterodimeric antibodies

Abstract

This invention relates to a method for generating heterodimeric antibodies.

Description

[0001] Cross Reference to Related Applications

[0002] This application claims the benefit of U.S. Provisional Application Serial No. 62 / 781,180, filed December 18, 2018, the entire contents of which are incorporated herein by reference. TECHNICAL FIELD

[0003] The present invention relates to methods of producing heterodimeric antibodies.

[0004] SEQUENCE LISTING

[0005] The instant application contains a Sequence Listing which has been submitted via EFS-Web and is hereby incorporated by reference in its entirety. The ASCII text file was created on December 10, 2019, is named JBI6029Wopct1_st25.txt, and is 21 kilobytes in size. BACKGROUND

[0007] Monoclonal antibodies have proven successful as therapeutic molecules, particularly in the treatment of cancer. Bispecific antibodies are expected to increase the potency and efficacy of monoclonal antibody therapy, as they can be used to direct drugs or toxic compounds to target cells, redirect effector mechanisms to disease-associated locations, or increase specificity for tumor cells, for example, by binding to one or more target molecules expressed on tumor cells. In addition, by combining the specificities of two monoclonal antibodies into one, bispecific antibodies can potentially engage in a large array of mechanisms of action.

[0008] Different formats of bispecific antibodies and methods of producing them have been described, but there are challenges in producing bispecific antibodies at large scale by optimizing manufacturing processes.

[0009] SUMMARY The present invention provides methods of producing heterodimeric antibodies, comprising:

[0010] providing a first homodimeric antibody comprising a first Fc region of an immunoglobulin comprising a first CH3 region and a second homodimeric antibody comprising a second Fc region of an immunoglobulin comprising a second CH3 region, wherein the amino acid sequences of the first CH3 region and the second CH3 region are different and such that the heterodimeric interaction between the first CH3 region and the second CH3 region is stronger than the homodimeric interaction between the first CH3 regions or the homodimeric interaction between the second CH3 regions;

[0011] combining the first homodimeric antibody and the second homodimeric antibody into a mixture; and

[0012]

[0013] incubating the mixture in the presence of a reducing agent and ambient dissolved oxygen (D02) to produce the heterodimeric antibody, wherein the method lacks one or more of the following steps during production of the heterodimeric antibody: measuring the percent (%) D02 in the mixture, controlling the % D02 in the mixture, or adding oxygen to the mixture.

[0014] The present invention also provides a method of producing a heterodimeric antibody, comprising:

[0015] providing a first homomultimeric antibody comprising a first Fc region of an immunoglobulin comprising a first CH3 region and a second homomultimeric antibody comprising a second Fc region of an immunoglobulin comprising a second CH3 region, wherein the amino acid sequences of the first CH3 region and the second CH3 region are different and such that a heterodimeric interaction between the first CH3 region and the second CH3 region is stronger than a homodimeric interaction between the first CH3 regions or a homodimeric interaction between the second CH3 regions;

[0016] combining the first homodimeric antibody and the second homodimeric antibody into a mixture;

[0017] incubating the mixture in the presence of a reducing agent; and

[0018] removing the reducing agent to produce the heterodimeric antibody, wherein the percent (%) dissolved oxygen (D02) is controlled to be about 30% or less during the step of incubating the mixture in the presence of a reducing agent, during the step of removing the reducing agent to produce the heterodimeric antibody, or during the steps of incubating the mixture in the presence of a reducing agent and removing the reducing agent to produce the heterodimeric antibody. SUMMARY

[0020] Figure 1 A summary showing Fab-arm exchange is shown.

[0021] Figure 2 A summary of process steps during reduction and ultrafiltration / diafiltration (UF / DF) during the manufacture of bispecific antibodies using Fab-arm exchange is shown.

[0022] Figure 3 Molecular details of the thiol-disulfide exchange reaction are shown.

[0023] Figure 4 A summary of reactions that can occur during Fab-arm exchange in the presence of 2-MEA is shown.

[0024] Figure 5 A sketch of the UF / DF setup and DO2 and pH sensor locations are shown, where RmV = relative millivolts.

[0025] Figure 6Percent (%) DO2 measured during UF / DF is shown for retentate, permeate, and inlet line during the manufacture of bispecific antibody A under low DO2 conditions during UF / DF. a.s.: air saturated.

[0026] Figure 7 Percent (%) DO2 measured during UF / DF is shown for retentate, permeate, and inlet line during the manufacture of bispecific antibody B under low DO2 conditions during UF / DF. a.s.: air saturated.

[0027] Figure 8 Percent (%) DO2 over time during reduction under low DO2 conditions (low DO reduction) or ambient DO2 (target reduction) is shown. a.s.: air saturated. DETAILED DESCRIPTION

[0029] DEFINITIONS

[0030] All publications, including but not limited to patents and patent applications, cited in this specification are herein incorporated by reference as if each individual publication were specifically and individually indicated to be incorporated by reference herein.

[0031] It should be understood that the terms used herein are for the purpose of describing particular embodiments and are not intended to be limiting. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0032] Although any methods and materials similar or equivalent to those described herein can be used in the practice of the present application, exemplary materials and methods are described herein. In describing and claiming the present application, the following terminology will be used.

[0033] As used in the specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a cell" includes mixtures of two or more cells, and the like.

[0034] The transitional terms “comprising,” “consisting essentially of,” and “consisting of” are intended to have their plain, ordinary meanings in accordance with patent language; that is, (i) “comprising” is synonymous with “including,” “containing,” or “characterized by,” which is inclusive or open-ended, and does not exclude additional, unrecited elements or method steps; (ii) “consisting of” excludes any element, step, or ingredient not specified in the claim; (iii) “consisting essentially of’ limits the scope of a claim to the specified materials or steps “plus those that do not materially affect the basic and novel characteristic(s)” of the claimed application. Embodiments described according to the phrase “comprising” (or its equivalents) also provide for the independent description of embodiments according to “consisting of’ and “consisting essentially of’ as embodiments.

[0035] An “antibody” refers to an immunoglobulin molecule having two heavy chains (HC) and two light chains (LC) interconnected by disulfide bonds. Each heavy chain is comprised of a heavy chain variable region (VH) and a heavy chain constant region, which is comprised of a CH1 region, a hinge region, a CH2 region, and a CH3 region. Each light chain is comprised of a light chain variable region (VL) and a light chain constant region (CL). The VHand VLcan be further subdivided into regions of hypervariability, termed complementarity determining regions (CDRs), interspersed with framework regions (FRs). Each VHand VLis composed of three CDRs and four FR segments, arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4. Antibodies include monoclonal antibodies, including murine, human, humanized, and chimeric antibodies, bispecific or multispecific antibodies.

[0036] "Complementarity determining region (CDR)" is an antibody region that binds to an antigen. There are three CDRs in VH (HCDR1, HCDR2, HCDR3) and three CDRs in VL (LCDR1, LCDR2, LCDR3). CDRs can be delineated using various depictions, such as Kabat (Wu et al. (1970) J Exp Med 132:211-50) (Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md., 1991), Chothia (Chothia et al. (1987) J Mol Biol 196:901-17), IMGT (Lefranc et al. (2003) Dev Comp Immunol 27:55-77), and AbM (Martin and Thornton (1996) J Bmol Biol 263:800-15). Correspondences between the various delineations and variable region numbering are described (see, e.g., Lefranc et al. (2003) Dev Comp Immunol 27:55-77; Honegger and Pluckthun, J Mol Biol (2001) 309:657-70; the International ImMunoGeneTics (IMGT) database; World Wide Web resource, http: / / www_imgt_org). Programs available, such as abYsis by UCL Business PLC, can be used to delineate CDRs. Unless otherwise specified in the specification, the terms "CDR", "HCDR1", "HCDR2", "HCDR3", "LCDR1", "LCDR2", and "LCDR3" include CDRs defined by any of the methods described above, Kabat, Chothia, IMGT, or AbM.

[0037] Depending on the amino acid sequence of the constant region, immunoglobulins can be assigned to five major classes: IgA, IgD, IgE, IgG, and IgM. IgA and IgG are further divided into subclasses, e.g., IgGl, IgG2, IgG3, and IgG4, IgAl, and IgA2. The antibody light chains of any vertebrate species can be assigned to one of two clearly distinct types, namely kappa (K) and lambda (l), based on the amino acid sequences of their constant domains.

[0038] "Antigen binding fragment" refers to a portion of an immunoglobulin molecule that retains the antigen binding properties of the parent full-length antibody. Exemplary antigen binding fragments are heavy chain complementarity determining regions (HCDR) 1, 2, and / or 3, light chain complementarity determining regions (LCDR) 1, 2, and / or 3, VH, VL, VH and VL, Fab, F(Ab')2, Fd and Fv fragments and a domain antibody (DAB) consisting of either a VH domain or a VL domain. The VH and VL domains can be joined together by a synthetic linker to form a variety of types of single-chain antibody designs in which the VH / VL domains pair within the molecule, or pair between molecules in the case of those expressed by separate chains, to form a monovalent antigen binding site, such as a single-chain Fv (scFV) or diabody; as described, for example, in International Patent Publication No. WO 1998 / 44001, International Patent Publication No. WO 1988 / 01649; International Patent Publication No. WO 1994 / 13804; International Patent Publication No. WO 1992 / 01047.

[0039] "Monoclonal antibody" refers to an antibody obtained from a population of substantially homogeneous antibody molecules, i.e., the individual antibodies comprising the population are identical except for possibly well-known alterations such as the removal of C-terminal lysine from the antibody heavy chain or post-translational modifications such as amino acid isomerization or deamidation, methionine oxidation or asparagine or glutamine deamidation. Monoclonal antibodies typically bind one antigenic epitope. Bispecific monoclonal antibodies bind two different antigenic epitopes. Monoclonal antibodies can have heterogeneous glycosylation within the antibody population. Monoclonal antibodies can be monospecific or multispecific, e.g., bispecific, monovalent, bivalent or multivalent.

[0040] "Fab-arm" refers to one heavy chain-light chain pair of an antibody.

[0041] "Homodimerization" refers to the interaction of two heavy chains having the same CH3 amino acid sequence.

[0042] "Homodimer" refers to an antibody having two heavy chains having the same CH3 region amino acid sequence.

[0043] "Heterodimerization" refers to the interaction of two heavy chains having different CH3 amino acid sequences.

[0044] "Heterodimer" refers to an antibody having two heavy chains having amino acid sequences that differ by one or more amino acids in the CH3 region.

[0045] An "Fc region" or "Fc domain" refers to the region of an antibody comprising at least a portion of the hinge region, a CH2 region and a CH3 region. The Fc region can be generated by digestion of an antibody with papain, wherein the Fc region is the fragment obtained that includes one or both CH2-CH3 regions and a portion of the hinge region of an immunoglobulin. The constant domains of the heavy chain of an antibody define the antibody's isotype, such as IgGl, IgG2, IgG3, IgG4, IgAl, IgA2, IgE. The Fc region mediates the effector functions of antibodies, such as binding to cellular Fc receptors and proteins of the complement system.

[0046] A "CH1 region" or "CH1 domain" refers to the CH1 region of an immunoglobulin. The CH1 region of a human IgGl antibody corresponds to amino acid residues 118-215. However, the CH1 region can also be of any other antibody isotype as described herein.

[0047] A "CH2 region" or "CH2 domain" refers to the CH2 region of an immunoglobulin. The CH2 region of a human IgGl antibody corresponds to amino acid residues 231-340. However, the CH2 region can also be of any other antibody isotype as described herein.

[0048] A "CH3 region" or "CH3 domain" refers to the CH3 region of an immunoglobulin. The CH3 region of a human IgGl antibody corresponds to amino acid residues 341-446. However, the CH3 region can also be of any other antibody isotype as described herein.

[0049] A "hinge" or "hinge region" refers to the hinge region of an immunoglobulin. The hinge region of a human IgGl antibody generally corresponds to amino acids 216-230 according to the EU numbering system. It is also recognized that "hinge" includes additional residues referred to as upper and lower hinge regions, e.g., from amino acid residues 216 to 239.

[0050] A "mixture" refers to an aqueous solution of two or more antibodies.

[0051] A "reducing agent" refers to an agent capable of reducing interchain disulfide bonds in the hinge region of an antibody.

[0052] "Compliance with GMP" refers to manufacturing according to the regulations of the Good Manufacturing Practice (CGMP) enforced by the FDA. CGMP provides for systems that assure proper design, monitoring, and control of manufacturing processes and facilities. Adherence to the CGMP regulations is designed to ensure the safety, strength, quality, and purity of drug products. This includes establishing strong quality management systems, obtaining appropriate quality raw materials, establishing strong operating procedures, testing and investigating product quality deviations, and maintaining reliable testing laboratories. This formal control system of a pharmaceutical company, when fully implemented, helps to prevent instances of contamination, mix-ups, deviations, failures, and errors. This ensures that drug products meet their quality standards.

[0053] “Drug substance” or “DS” means any substance or mixture of substances intended to be used in the manufacture of a drug (medicinal) product and that, when used in the production of a drug, becomes an active ingredient of the drug product. These substances are intended to furnish pharmacological activity or other direct effect in the diagnosis, cure, mitigation, treatment, or prevention of disease, or to affect the structure or function of the body.

[0054] “Drug product” or “DP” means a finished dosage form, for example, a tablet, capsule, or solution containing a active pharmaceutical ingredient (e.g., drug substance), usually but not necessarily in combination with inactive ingredients.

[0055] “Reference product” means an approved biological product against which a biosimilar product is compared. The reference product is approved based on, inter alia, a full complement of safety and efficacy data, and is approved in at least one of the United States, Europe, or Japan.

[0056] A“biological generic” of an (approved reference product / biologic drug) refers to a biological product that, despite minor differences in clinically inactive components, is highly similar to the reference product, based on data derived from: (a) analytical studies that have shown that, despite minor differences in clinically inactive components, the biological product is highly similar to the reference product; (b) animal studies, including toxicity assessments; and / or (c) one or more clinical studies (including assessments of immunogenicity and pharmacokinetics or pharmacodynamics) adequate to demonstrate safety, purity, and potency in one or more appropriate conditions of use for which the reference product is licensed and for which the biological generic seeks licensure. A biological generic can be an interchangeable product, which can be substituted for the reference product without the intervention of an intervening medical professional at the pharmacy. To meet the additional criteria of“interchangeability,” the biological generic is expected to produce the same clinical result as the reference product in any given patient, if the biological generic is administered to an individual more than once, the risk of an adverse effect or diminished efficacy with respect to the interchange or switch between use of the biological generic and the reference product is no greater than the risk if the reference product were used without such interchange or switch. The biological generic utilizes the same mechanism of action for the proposed conditions of use as is known for the reference product. The one or more conditions of use specified, recommended, or suggested in the proposed label for the biological generic must have been previously approved for the reference product. The route of administration, dosage form, and / or strength of the biological generic must be the same as those of the reference product, and the biological generic must be manufactured, processed, packaged, or held in accordance with the standards that have been designed to assure that the biological generic will remain safe, pure, and potent. The biological generic can include minor modifications in the amino acid sequence, such as N- or C-terminal truncations, that are not expected to alter the performance of the biological generic when compared to the reference product. The reference product can be approved in at least one of the United States, Europe, or Japan.

[0057] “Isolated” refers to a homogeneous group of molecules (e.g., a synthetic polynucleotide or protein such as an antibody) that have been substantially separated and / or purified from other components of the system from which the molecule is produced (e.g., a recombinant cell), as well as a protein that has undergone at least one purification or isolation step. An“isolated antibody” refers to an antibody that is substantially free of other cellular material and / or chemicals, and includes an antibody that is isolated to a high degree of purity, such as 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% purity.

[0058] A "humanized antibody" refers to an antibody in which at least one CDR is derived from a non-human species and at least one framework is derived from a human immunoglobulin sequence. A humanized antibody can include substitutions in the framework, so the framework can not be an exact copy of an expressed human immunoglobulin or human immunoglobulin germline gene sequence.

[0059] A "human antibody" refers to an antibody that is optimized to have minimal immune response when administered to a human subject. The variable regions of a human antibody are derived from human immunoglobulin sequences. If a human antibody contains a constant region or a portion of a constant region, the constant region is also derived from human immunoglobulin sequences. A human antibody comprises heavy and light chain variable regions "derived from" human-derived sequences if the variable regions are obtained from a system that utilizes human germline immunoglobulin or rearranged immunoglobulin genes. An exemplary system is a human immunoglobulin gene library displayed on phage and transgenic non-human animals, such as mice or rats, that carry human immunoglobulin loci. Due to differences between the systems used to obtain human antibodies and human immunoglobulin loci, "human antibodies" often contain amino acid differences when compared to immunoglobulins expressed in humans, either due to introduction of somatic mutations or intentional introduction of substitutions in the framework or CDRs or both. Typically, the amino acid sequence of a "human antibody" is at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identical to the amino acid sequence encoded by a human germline immunoglobulin or rearranged immunoglobulin gene. In some cases, a "human antibody" can contain a consensus framework sequence derived from human framework sequence analysis, e.g., as described in Knappik et al., (2000) J Mol Biol 296:57-86, or synthetic HCDR3s incorporated into a human immunoglobulin gene library displayed on phage, e.g., as described in Shi et al., (2010) J Mol Biol 397:385-96 and WO2009 / 085462. Antibodies in which at least one CDR is derived from a non-human species are not included in the definition of "human antibodies."

[0060] "Recombinant" refers to DNA, antibodies, and other proteins that are prepared, expressed, created, or isolated by recombinant means, such as by ligating segments of DNA from different sources into a recombinant DNA, antibody, or protein. "Recombinant antibody" includes all antibodies that are prepared, expressed, created, or isolated by recombinant means, such as antibodies isolated from animals (e.g., mice) that are transgenic or transchromosomal for human immunoglobulin genes or a hybridoma prepared therefrom (described further below), antibodies isolated from a host cell transformed to express the antibody, antibodies isolated from a recombinant, combinatorial antibody library, and antibodies prepared, expressed, created, or isolated by any other means that involve splicing of human immunoglobulin gene sequences to other DNA sequences, or antibodies generated in vitro using Fab-arm exchange, such as bispecific antibodies.

[0061] "Bispecific" refers to an antibody that specifically binds to two different antigens or two different epitopes within the same antigen. A bispecific antibody can have cross-reactivity to other related antigens, or can bind to an epitope shared between two or more different antigens.

[0062] "Multispecific" refers to an antibody that specifically binds to at least two different antigens or at least two different epitopes within the same antigen. A multispecific antibody can bind, for example, two, three, four, or five different antigens or different epitopes within the same antigen.

[0063] "About" means within an acceptable error range for the particular value as determined by one of ordinary skill in the art, which will depend in part on how the value is measured or determined, i.e., the limitations of the measurement system. Unless otherwise stated in the embodiments or elsewhere in the specification, "about" as used herein will mean ± one standard deviation, or ± 5% of the specified value, whichever is greater, according to the practice in the art.

[0064] "Variant" refers to a polypeptide or polynucleotide that differs from a reference polypeptide or reference polynucleotide by one or more modifications, such as substitutions, insertions, or deletions.

[0065] "Mutation" refers to an engineered or naturally occurring alteration in a polypeptide or polynucleotide sequence when compared to a reference sequence. The alteration can be a substitution, insertion, or deletion of one or more amino acids or polynucleotides.

[0066] Unless explicitly stated otherwise, the numbering of amino acid residues in the constant region of an antibody is according to the EU index as set forth in Kabat et al. (Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD. (1991)). Antibody constant chain numbering can be found in, for example, the IMGT Scientific Charts resource on the ImMunoGeneTics website.

[0067] Mutations in the CH3 region described herein are denoted as modified position in the first CH3 domain of the first heavy chain / modified position in the second CH3 domain of the second heavy chain. For example, F405L / K409R refers to a F405L mutation in the first CH3 region and a K09R mutation in the second CH3 region. L351Y_F405A_Y407V / T394W refers to L351Y, F40FA, and Y407V mutations in the first CH3 region and a T394W mutation in the second CH3 region. D399FHKRQ / K409AGRH refers to a mutation where D399 can be substituted with F, H, K, R, or Q and K409 can be substituted with A, G, R, or H.

[0068] As shown in Table 1, conventional one-letter and three-letter amino acid codes are used herein.

[0069] Table 1

[0070] Amino acid Three letter code One letter code Amino acid Three letter code One letter code Alanine Ala A Leucine Leu L Arginine Arg R Lysine Lys K Asparagine Asn N Methionine Met M Aspartic acid Asp D Phenylalanine Phe F Cysteine Cys C Proline Pro P Glutamic acid Gln E Serine Ser S Glutamine Glu Q Threonine Thr T Glycine Gly G Tryptophan Trp W Histidine His H Tyrosine Tyr Y Isoleucine Ile I Valine Val V

[0071] Methods of the application

[0072] The present application provides methods for producing hetero-dimeric antibodies using Fab-arm exchange. The present application is based at least in part on the identification that, contrary to what is disclosed in the literature (see, e.g., US 2014 / 0303356), oxygen is not required for disulfide bonds to reform after their reduction during Fab-arm exchange to form stable hetero-dimeric antibodies, thus providing an alternative to process control during large-scale production of hetero-dimeric antibodies (e.g., bispecific antibodies).

[0073] The present invention provides a method of producing a heterodimeric antibody, comprising: providing a first homodimeric antibody comprising a first Fc region of an immunoglobulin comprising a first CH3 region and a second homodimeric antibody comprising a second Fc region of an immunoglobulin comprising a second CH3 region, wherein the amino acid sequences of the first CH3 region and the second CH3 region are different and such that the heterodimeric interaction between the first CH3 region and the second CH3 region is stronger than the homodimeric interaction between the first CH3 regions or the homodimeric interaction between the second CH3 regions;

[0074] combining the first homodimeric antibody and the second homodimeric antibody into a mixture; and

[0075] incubating the mixture in the presence of a reducing agent and ambient dissolved oxygen (DO2) to produce the heterodimeric antibody, wherein the method lacks one or more of the following steps during production of the heterodimeric antibody: measuring the percent (%) DO2 in the mixture, controlling the % DO2 in the mixture, or adding oxygen to the mixture.

[0076] In some embodiments, the % DO2 in the mixture during production of the heterodimeric antibody is about 10% to about 90%.

[0077] In some embodiments, the % DO2 in the mixture prior to adding the denaturant is about 30% or more.

[0078] The present invention also provides a method of producing a heterodimeric antibody, comprising:

[0079] providing a first homodimeric antibody comprising a first Fc region of an immunoglobulin comprising a first CH3 region and a second homodimeric antibody comprising a second Fc region of an immunoglobulin comprising a second CH3 region, wherein the amino acid sequences of the first CH3 region and the second CH3 region are different and such that the heterodimeric interaction between the first CH3 region and the second CH3 region is stronger than the homodimeric interaction between the first CH3 regions or the homodimeric interaction between the second CH3 regions;

[0080] combining the first homodimeric antibody and the second homodimeric antibody into a mixture;

[0081] incubating the mixture in the presence of a reducing agent; and

[0082] removing the reducing agent to produce the heterodimeric antibody, wherein

[0083] controlling the percent (%) dissolved oxygen (DO2) to be about 30% or less in step c), step d), or both step c) and step d).

[0084] In some embodiments, the mixture has a %DO2of about 25% or less, 20% or less, about 15% or less, about 10% or less, about 9% or less, about 8% or less, about 7% or less, about 6% or less, about 5% or less, about 4% or less, about 3% or less, about 2% or less, or about 1% or less.

[0085] In some embodiments, the %DO2is controlled by displacing oxygen from the mixture with an inert gas, such as nitrogen. The dissolved oxygen concentration can be monitored by known methods, for example, using a dissolved oxygen probe.

[0086] “Stronger” in terms of heterodimeric interaction of the first CH3 region and the second CH3 region can be more than twice as strong, for example, more than three times as strong, more than four times as strong, or more than five times as strong, than the strongest of the homodimeric interaction between the first CH3 regions or the homodimeric interaction between the second CH3 regions. The strength of CH3 domain interaction can be measured using mass spectrometry. In an exemplary assay, a construct comprising the first CH3 region and the second CH3 region, or alternatively both comprising a CH2 region, is prepared using standard molecular biology techniques. Samples comprising the first CH3 domain, the second CH3 domain, or both the first CH3 domain and the second CH3 domain are prepared using a 10 kDa MWCO spin filtration column and the buffer is exchanged to 100 mM ammonium acetate pH 7. Aliquots (about 1 μΐ) of serially diluted samples (20 μΜ - 25 nM; monomer equivalents) are loaded into gold-coated borosilicate capillaries and analyzed on a LCT mass spectrometer (Waters). The monomer signal Ms is defined as the monomer peak area as a fraction of the total peak area in the spectrum (Ms / (Ms+Ds), where Ds = dimer signal). The monomer concentration at equilibrium [M]eq is defined as Ms·[M]0, where [M]0 is the total protein concentration in terms of monomer. The dimer concentration at equilibrium [D]eq is defined as ([M]0-[M]eq) / 2. Then, the K 2 D .

[0087] In some embodiments, the first homodimeric antibody and the second homodimeric antibody are combined in the mixture at a molar ratio of about 1 : 1.

[0088] In some embodiments, the first homodimeric antibody and the second homodimeric antibody are combined in the mixture at a molar ratio of about 1.05: 1.

[0089] ​In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.03 to about 1 :2. In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.05 to 1 : 1.5. In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.1 to 1 : 1.5. In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.1 to 1 : 1.4. In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.15 to 1 : 1.35. In some embodiments, the first and second homodimeric antibodies are combined in a mixture in a molar ratio of about 1 : 1.2 to 1 : 1.3.

[0090] In some embodiments, the total concentration of immunoglobulin in the mixture is about 1 g / L to 70 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 8 g / L to about 50 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 8 g / L to about 13 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 9.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 9.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 10 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 10.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 11.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 11.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 12.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 12.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 13.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 13.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 14.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 14.5 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 15.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 20.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 25.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 30.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 35.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 40.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 45.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 50.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 55.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 60.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 65.0 g / L. In some embodiments, the total concentration of immunoglobulin in the mixture is about 70.0 g / L.

[0091] In some embodiments, the mixture is incubated in the presence of the reducing agent for about 10 minutes or more. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 10 minutes to about 30 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 10 minutes to about 24 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 15 minutes. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 20 minutes. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 30 minutes. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 40 minutes. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 50 minutes. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 1 hour. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 2 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 3 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 4 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 4 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 5 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 6 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 7 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 8 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 9 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 10 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 11 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 12 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 13 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 14 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 15 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 16 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 17 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 18 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 19 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 20 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 21 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 22 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 23 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 24 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 25 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 26 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 27 hours.In some embodiments, the mixture is incubated in the presence of the reducing agent for about 28 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 29 hours. In some embodiments, the mixture is incubated in the presence of the reducing agent for about 30 hours.

[0092] In some embodiments, the reducing agent is 2-mercaptoethylamine (2-MEA). In some embodiments, the reducing agent is a chemical derivative of 2-MEA. In some embodiments, the reducing agent is L-cysteine. In some embodiments, the reducing agent is D-cysteine. In some embodiments, the reducing agent is glutathione. In some embodiments, the reducing agent is tris(2-carboxyethyl)phosphine.

[0093] In some embodiments, the concentration of the reducing agent in the mixture is about 0.1 mM to about 1 M. In some embodiments, the concentration of the reducing agent in the mixture is about 1.0 mM to about 500 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 5.0 mM to about 100 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 10 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 15 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 20 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 25 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 30 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 35 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 40 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 50 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 60 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 70 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 80 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 90 mM. In some embodiments, the concentration of the reducing agent in the mixture is about 100 mM.

[0094] In some embodiments, the concentration of 2-MEA in the mixture is about 10 mM to about 100 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 90 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 80 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 70 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 60 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 50 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM to about 40 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 20 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 25 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 30 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 35 mM. In some embodiments, the concentration of 2-MEA in the mixture is about 40 mM.

[0095] In some embodiments, the mass ratio of total immunoglobulin to total reducing agent in the mixture is about 1.0 to about 5.0. In some embodiments, the mass ratio of total immunoglobulin to total reducing agent in the mixture is about 1.4 to about 3.8. In some embodiments, the mass ratio of total immunoglobulin to total reducing agent in the mixture is about 1.4 to about 3.5. In some embodiments, the mass ratio is about 1.8 to about 3.8. In some embodiments, the mass ratio is about 2.3 to about 3.0. In some embodiments, the mass ratio is about 1.6 to about 2.1. In some embodiments, the mass ratio is about 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.0, 2.8..3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 4.0. 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7, 4.8, 4.9, or 5.0. By "mass ratio" is meant the total immunoglobulin in grams per liter to the total reducing agent in grams per liter. By "total immunoglobulin" is meant the amount of the two parent antibodies in the mixture at the beginning of the reduction step.

[0096] In some embodiments, the mixture comprises a buffer. In some embodiments, the buffer comprises a sodium acetate buffer. In some embodiments, the buffer further comprises NaCl. In some embodiments, the buffer comprises about 100 mM sodium acetate and about 30 mM NaCl. In some embodiments, the buffer has a pH of about 7.3. Other buffers can be used, such as IX Dulbecco’s Phosphate Buffered Saline (DPBS), a sodium phosphate buffer, a potassium phosphate buffer, a Tris buffer, a histidine buffer, or a citrate buffer.

[0097] In some embodiments, the method further comprises the step of removing the reducing agent from the mixture.

[0098] In some embodiments, the reducing agent is removed by filtration.

[0099] In some embodiments, the filtration is diafiltration.

[0100] In some embodiments, the first and second homodimeric antibodies are of IgGl, IgG2, or IgG4 isotype.

[0101] In some embodiments, the first and second homodimeric antibodies are of IgGl isotype. In some embodiments, the first and second homodimeric antibodies are of IgG2 isotype. In some embodiments, the first and second homodimeric antibodies are of IgG4 isotype.

[0102] In some embodiments, the first CH3 domain and the second CH3 domain comprise the following mutations when compared to the wild-type IgGl of SEQ ID NO: 1 : F405L / K409R, T350I_K370T_F405L / K409R, K370W / K409R, D399AFGHILMNRSTVWY / K409R, T366ADEFGHILMQVY / K409R, L368ADEGHNRSTVQ / K409AGRH, D399FHKRQ / K409AGRH, F405IKLSTVW / K409AGRH, Y407LWQ / K409AGRH, T366Y / F405A, T366W / F405W, F405W / Y407A, T394W / Y407T, T394S / Y407A, T366W / T394S, F405W / T394S, T366W / T366S_L368A_Y407V, L351Y_F405A_Y407V / T394W, T366I_K392M_T394W / F405A_Y407V, T366L_K392M_T394W / F405A_Y407V, L351Y_Y407A / T366A_K409F, L351Y_Y407A / T366V_K409F, Y407A / T366A_K409F, T350V_L351Y_F405A_Y407V / T350V_T366L_K392L_T394W, K409D / D399K, K409E / D399R, K409D_K360D / D399K_E356K, K409D_K360D / D399E_E356K, K409D_K370D / D399K_E357K, K409D_K370D / D399E_E357K, K409D_K392D / D399K_E356K_E357K, or K409D_K392D / D399E_E356K_E357K. Mutations can also be introduced into the wild-type IgG2 of SEQ ID NO: 2 or the wild-type IgG4 of SEQ ID NO: 3, in which case it is well known that due to the sequence differences between wild-type IgGl, IgG2, and IgG4, the wild-type residue at a particular mutation site can not correspond to the residue of IgGl. For example, the IgG4 mutation wild-type / F405L_R409K corresponds to the IgGl mutation F405L / K409R, while the IgG4 mutation R409D / D399K corresponds to the IgGl mutation K409D / D399K. Mutations are compared to the reference wild-type IgGl of SEQ ID NO: 1, IgG2 of SEQ ID NO: 2, and IgG4 of SEQ ID NO: 3.

[0103] In some embodiments, the first Fc region and / or the second Fc region comprises one or more mutations that modulate binding of the first Fc region and / or the second Fc region to an Fc gamma receptor (FcγR), FcRn, or Protein A, when compared to a wild-type IgG1 of SEQ ID NO: 1, a wild-type IgG2 of SEQ ID NO: 2, a wild-type IgG4 of SEQ ID NO: 3.

[0104] Wild-type IgG1 (SEQ ID NO: 1)

[0105]

[0106] Wild-type IgG2 (SEQ ID NO: 2)

[0107]

[0108] Wild-type IgG4 (SEQ ID NO: 3)

[0109]

[0110] In some embodiments, the FcγR is FcγRI, FcγRIIa, FcγRIIb, or FcγRIII.

[0111] In some embodiments, the one or more substitutions that modulate binding of the first Fc region and / or the second Fc region to an Fcγ is L234A_L235A, F234A_L235A, S228P_F234A_L235A, S228P_L234A_L235A, V234A_G237A_P238S_H268A_V309L_A330S_P331S, V234A_G237A, H268Q_V309L_A330S_P331S, S267E_L328F, L234F_L235E_D265A, L234A_L235A_G237A_P238S_H268A_A330S_P331S, or S228P_F234A_L235A_G237A_P238S.

[0112] In some embodiments, the one or more substitutions that modulate binding of the first Fc region and / or the second Fc region to FcRn is M428L_N434S, M252Y_S254T_T256E, T250Q_M428L, N434A and T307A_E380A_N434A, H435A, P257I_N434H, D376V_N434H, M252Y_S254T_T256E_H433K_N434F, T308P_N434A, or H435R.

[0113] In some embodiments, the one or more substitutions that modulate the Protein A binding of the first Fc region and / or the second Fc region are Q311R, Q311K, T307P_L309Q, T307P_V309Q, T307P_L309Q_Q311R, T307P_V309Q_Q311R, H435R, or H435R_Y436F.

[0114] In some embodiments, the step of producing the heterodimeric antibody is performed under GMP-compliant conditions.

[0115] In some embodiments, the step of producing the heterodimeric antibody is performed during the manufacture of a bulk drug substance comprising the heterodimeric antibody.

[0116] In some embodiments, the heterodimeric antibody is a bispecific antibody.

[0117] In some embodiments, the bispecific antibody binds CD3, BCMA, or CD123.

[0118] In some embodiments, the step of producing the heterodimeric antibody is performed during the manufacture of an innovative drug product.

[0119] In some embodiments, the step of producing the heterodimeric antibody is performed during the manufacture of a generic drug product.

[0120] Fab-arm exchange and Fc region mutations that promote heterodimerization

[0121] Heterodimeric antibodies can be produced using Fab-arm exchange, in which one heavy chain of a parent homodimeric antibody and its attached light chain (half-arm) is exchanged with one heavy chain of another parent homodimeric antibody and its attached light chain, to form a heterodimeric antibody composed of two heavy chains and two attached light chains (van der Neut Kolfschoten et al., (2007) Science 317: 1554-1557).

[0122] Two homodimeric parent antibodies are engineered to have asymmetric mutations in their CH3 regions that favor Fab-arm exchange and heterodimeric antibody formation upon reduction and reformation of the disulfide bridges in the hinge region of the antibodies. A schematic of the Fab-arm exchange reaction is shown in Figure 1 Upon introduction of a reducing agent to a mixture of the two parent homodimeric antibodies, the half-arms dissociate and the asymmetric CH3 mutations favor reformation of the heterodimeric antibody. Subsequent reformation of the disulfide bridges between the half-arms stabilizes the formed heterodimeric antibody (Gramer et al. (2013) MAbs 5: 962-973).

[0123] In the methods of the application, any CH3 region mutation that promotes CH3 heterodimer formation can be used, such as those described herein. Several methods are known to make modifications in the CH3 region to promote heterodimerization. Generally, in all of these methods, the first CH3 region and the second CH3 region are engineered in a complementary fashion, such that each CH3 region (or heavy chain comprising it) can no longer homodimerize with itself, but is forced to heterodimerize with the other CH3 region engineered in a complementary fashion (such that the first and second CH3 regions heterodimerize and no homodimers are formed between two first or two second CH3 regions).

[0124] CH3 mutations that favor Fab-arm exchange include Mutations (Genmab), Knob-in-Hole mutations (Genentech), Electrostatically matched mutations (Chugai, Amgen, Novo Nordisk, Oncomed), SEEDbody (EMD Serono) and other asymmetric mutations (e.g., Zymeworks).

[0125] Mutations (Genmab) are disclosed, e.g., in US 9,150,663 and US 2014 / 0303356, including mutations F405L / K409R, wild type / F405L_R409K, T350I_K370T_F405L / K409R, K370W / K409R, D399AFGHILMNRSTVWY / K409R, T366ADEFGHILMQVY / K409R, L368ADEGHNRSTVQ / K409AGRH, D399FHKRQ / K409AGRH, F405IKLSTVW / K409AGRH, and Y407LWQ / K409AGRH.

[0126] Knob-in-Hole is disclosed, e.g., in WO 1996 / 027011, including mutations at the CH3 region interface, where an amino acid with a small side chain (hole) is introduced into the first CH3 region and an amino acid with a large side chain (knob) is introduced into the second CH3 region, resulting in a preferential interaction between the first CH3 region and the second CH3 region. Exemplary CH3 region mutations that form a knob and a hole are T366Y / F405A, T366W / F405W, F405W / Y407A, T394W / Y407T, T394S / Y407A, T366W / T394S, F405W / T394S, and T366W / T366S_L368A_Y407V.

[0127] Heterodimerization of the heavy chains can be promoted by substituting positively charged residues on the first CH3 domain for negatively charged residues on the second CH3 domain to promote heavy chain heterodimer formation by making use of electrostatic interactions, as described in US2010 / 0015133, US2009 / 0182127, US2010 / 028637 or US2011 / 012353.

[0128] Other asymmetric mutations that can be used to promote heavy chain heterodimerization are L351Y_F405A_Y407V / T394W, T366I_K392M_T394W / F405A_Y407V, T366L_K392M_T394W / F405A_Y407V, L351Y_Y407A / T366A_K409F, L351Y_Y407A / T366V_K409F, Y407A / T366A_K409F, or T350V_L351Y_F405A_Y407V / T350V_T366L_K392L_T394W, as described in US2012 / 0149876 or US2013 / 0195849.

[0129] SEEDbody mutations involve the substitution of IgA residues for selected IgG residues to promote heavy chain heterodimerization, as described in US20070287170.

[0130] Other exemplary mutations that can be used are R409D_K370E / D399K_E357K, S354C_T366W / Y349C_T366S_L368A_Y407V, Y349C_T366W / S354C_T366S_L368A_Y407V, T366K / L351D, L351K / Y349E, L351K / Y349D, L351K / L368E, L351Y_Y407A / T366A_K409F, L351Y_Y407A / T366V_K409F, K392D / D399K, K392D / E356K, K253E_D282K_K322D / D239K_E240K_K292D, K392D_K409D / D356K_D399K, as described in WO2007 / 147901, WO 2011 / 143545, WO2013157954, WO2013096291 and US2018 / 0118849.

[0131] These different methods for Fab-arm exchange can be combined with various bispecific antibody formats, for example those involving VH / VL engineering, such as VH / VL domain swapping, CH1 / CL domain swapping, use of a common light chain as described in WO98050431, or use of tethered light chains, including inside-out tethered light chains as described in US9062120.

[0132] Further engineering of heterodimeric antibodies

[0133] Fc engineering

[0134] In addition to CH3 region mutations that promote heterodimerization, antibodies used in the methods of the application can comprise mutations in the Fc region that modulate antibody effector function or half-life. Furthermore, antibodies used in the methods of the application can comprise Fc mutations that modulate antibody binding to protein A, thus facilitating purification of the antibody.

[0135] Fc positions that can be mutated to modulate antibody half-life (e.g., binding to FcRn) include positions 250, 252, 253, 254, 256, 257, 307, 376, 380, 428, 434, and 435. Exemplary mutations that can be made alone or in combination are mutations T250Q, M252Y, I253A, S254T, T256E, P257I, T307A, D376V, E380A, M428L, H433K, N434S, N434A, N434H, N434F, H435A, and H435R. Exemplary single or combination mutations that can be made to increase antibody half-life are mutations M428L / N434S, M252Y / S254T / T256E, T250Q / M428L, N434A, and T307A / E380A / N434A. Exemplary single or combination mutations that can be made to decrease antibody half-life are H435A, P257I / N434H, D376V / N434H, M252Y / S254T / T256E / H433K / N434F, T308P / N434A, and H435R.

[0136] Mutations can be introduced into the Fc region that reduce antibody binding to activating Fcy receptors (FcyR) and decrease Fc effector functions, such as Clq binding, complement-dependent cytotoxicity (CDC), antibody-dependent cell-mediated cytotoxicity (ADCC), or phagocytosis (ADCP).

[0137] Fc positions that can be mutated to reduce antibody binding to activating FcyRs and subsequently reduce effector functions include positions 214, 233, 234, 235, 236, 237, 238, 265, 267, 268, 270, 295, 297, 309, 327, 328, 329, 330, 331, and 365. Exemplary mutations that can be made alone or in combination are mutations K214T, E233P, L234V, L234A, G236 deletion, V234A, F234A, L235A, G237A, P238A, P238S, D265A, S267E, H268A, H268Q, Q268A, N297A, A327Q, P329A, D270A, Q295A, V309L, A327S, L328F, A330S, and P331S in IgGl, IgG2, IgG3, or IgG4. Exemplary combination mutations that result in antibodies with reduced ADCC are mutations L234A / L235A on IgGl, V234A / G237A / P238S / H268A / V309L / A330S / P331S on IgG2, F234A / L235A on IgG4, S228P / F234A / L235A on IgG4, N297A on all Ig isotypes, V234A / G237A on IgG2, K214T / E233P / L234V / L235A / G236 deletion / A327G / P331A / D365E / L358M on IgGl, H268Q / V309L / A330S / P331S on IgG2, S267E / L328F on IgGl, L234F / L235E / D265A on IgGl, L234A / L235A / G237A / P238S / H268A / A330S / P331S on IgGl, S228P / F234A / L235A / G237A / P238S on IgG4, and S228P / F234A / L235A / G236 deletion / G237A / P238S on IgG4. A hybrid IgG2 / 4 Fc domain can also be used, for example, an Fc with residues 117-260 from IgG2 and residues 261-447 from IgG4.

[0138] Exemplary mutations that result in antibodies with reduced CDC are the K322A mutation.

[0139] The well-known S228P mutation can be made in IgG4 antibodies to enhance stability of IgG4.

[0140] Mutations can be introduced into the Fc region that enhance binding of the antibody to Fcy receptors (FcyRs) and enhance Fc effector functions, such as Clq binding, complement dependent cytotoxicity (CDC), antibody dependent cell-mediated cytotoxicity (ADCC), and / or phagocytosis (ADCP).

[0141] Mutations can be made to increase the Fc positions of the antibody that bind to activating FcyRs and enhance antibody effector functions, including positions 236, 239, 243, 256, 290, 292, 298, 300, 305, 312, 326, 330, 332, 333, 334, 345, 360, 339, 378, 396, or 430 (residue numbering according to the EU index). Exemplary mutations that can be made alone or in combination are G236A, S239D, F243L, T256A, K290A, R292P, S298A, Y300L, V305L, K326A, A330K, I332E, E333A, K334A, A339T, and P396L. Exemplary combinations of mutations that result in antibodies with increased ADCC or ADCP are S239D / I332E, S298A / E333A / K334A, F243L / R292P / Y300L, F243L / R292P / Y300L / P396L, F243L / R292P / Y300L / V305I / P396L, and G236A / S239D / I332E on IgGl.

[0142] Fc positions that can be mutated to enhance CDC of the antibody include positions 267, 268, 324, 326, 333, 345, and 430. Exemplary mutations that can be made alone or in combination are S267E, F1268F, S324T, K326A, K326W, E333A, E345K, E345Q, E345R, E345Y, E430S, E430F, and E430T. Exemplary combinations of mutations that result in increased CDC of the antibody are K326A / E333A, K326W / E333A, H268F / S324T, S267E / H268F, S267E / S324T, and S267E / H268F / S324T on IgGl.

[0143] "Antibody-dependent cellular cytotoxicity," "antibody-dependent cell-mediated cytotoxicity," or "ADCC" is a mechanism of inducing cell death that relies on the interaction of antibody-coated target cells with effector cells, such as natural killer cells (NK), monocytes, macrophages, and neutrophils, that have lytic activity via Fcy receptors (FcyRs) expressed on the effector cells. For example, NK cells express FcyRIIIa, while monocytes express FcyRI, FcyRII, and FcyRIIIa. The ADCC activity of an antibody can be assessed using an in vitro assay with cells expressing the antibody-bound protein as target cells and NK cells as effector cells. Cell lysis can be detected by release of a marker from the lysed cells, such as a radioactive substrate, a fluorescent dye, or a native intracellular protein. In an exemplary assay, target cells are used at a ratio of 1 target cell to 4 effector cells. The target cells are pre-labeled with BATDA and combined with effector cells and test antibody. The samples are incubated for 2 hours and cell lysis is measured by measuring the BATDA released into the supernatant. The data are normalized to the maximum cytotoxicity obtained using 0.67% Triton X-100 (Sigma Aldrich) and the minimum control determined by spontaneous release of BATDA from the target cells in the absence of any antibody.

[0144] "Antibody-dependent cellular phagocytosis" ("ADCP") refers to a mechanism of eliminating antibody-coated target cells through internalization by phagocytic cells, such as macrophages or dendritic cells. ADCP can be assessed using monocyte-derived macrophages as effector cells and cells expressing the antibody-bound protein as target cells, which are also engineered to express GFP or other marker molecule. In an exemplary assay, the effector: target cell ratio can be, for example, 4: 1. Effector cells can be incubated with target cells for 4 hours with or without an antibody of the application. After incubation, cells can be isolated using accutase. Macrophages can be identified with anti-CDl lb and anti-CD14 antibodies coupled to a fluorescent marker and the percentage phagocytosis can be determined based on CDl lb + CD14 + GFP fluorescence in macrophages.

[0145] "Complement-dependent cytotoxicity" or "CDC" refers to a mechanism of inducing cell death in which the Fc effector domain of a target-bound antibody binds and activates complement component Clq, which in turn activates the complement cascade, leading to target cell death. Activation of complement can also result in deposition of complement components on the surface of the target cell, which promote CDC by binding to complement receptors (e.g., CR3) on leukocytes. CDC of cells can be measured, for example, by plating Daudi cells at 1 x 105cells / well in a 96-well plate and incubating with test antibody for 30 minutes at 37°C. The cells are washed and incubated with rabbit complement for 2 hours at 37°C. The cells are washed and incubated with a fluorescent dye, such as CellTracker Blue (ThermoFisher), for 30 minutes at 37°C. The cells are washed and the fluorescence is measured. The data are normalized to the maximum cytotoxicity obtained using 0.67% Triton X-100 (Sigma Aldrich) and the minimum control determined by spontaneous release of CellTracker Blue from the target cells in the absence of any antibody. 5Cells / well (50 μL / well) are seeded in RPMI-B (RPMI supplemented with 1% BSA) and 50 μL of test antibody is added to the wells at a final concentration of 0-100 μg / mL, the reaction is incubated at room temperature for 15 minutes, 11 μL of pooled human serum is added to the wells and the reaction is incubated at 37°C for 45 minutes. The percentage (%) lysis of cells can be detected as % propidium iodide staining of cells in a FACS assay using standard methods.

[0146] Binding of antibodies to FcyR or FcRn can be assessed using flow cytometry on cells engineered to express each receptor. In an exemplary binding assay, 2 x 105cells / well are seeded in a 96-well plate and blocked in BSA Staining Buffer (BD Biosciences, San Jose, USA) for 30 minutes at 4°C. Cells are incubated with test antibody for 1.5 hours at 4°C on ice. After two washes with BSA Staining Buffer, cells are incubated with R-PE labeled anti-human IgG secondary antibody (Jackson Immunoresearch Laboratories) for 45 minutes at 4°C. Cells are washed twice in Staining Buffer and then resuspended in 150 μL of Staining Buffer containing 1:200 dilution of DRAQ7 Live / Dead Stain (Cell Signaling Technology, Danvers, USA). PE and DRAQ7 signals of stained cells are detected by a Miltenyi MACSQuant flow cytometer using B2 and B4 channels (Miltenyi Biotec, Auburn, USA), respectively. Live cells are gated on DRAQ7 exclusion and the geometric mean fluorescence signal of at least 10,000 live events collected. FlowJo software (TreeStar) is used for analysis. Data are plotted as log of antibody concentration versus mean fluorescence signal. Non-linear regression analysis is performed. 5

[0147] "Enhanced" or "enhancement" refers to an enhancement in effector function (e.g., ADCC, CDC, and / or ADCP) or binding to Fcy receptor (FcyR) or FcRn of an antibody of the application having at least one mutation in the Fc region as compared to the parent antibody without the mutation. "Enhanced" can be an enhancement of about 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, or more, or a statistically significant enhancement.

[0148] ​"Decreased" or "decrease" refers to a decrease in effector function (e.g., ADCC, CDC, and / or ADCP) or binding to Fcy receptor (FcyR) or FcRn of an antibody of the application having at least one mutation in the Fc region as compared to the parent antibody without the mutation. A "decrease" can be a decrease of about 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100% or more, or a statistically significant decrease.

[0149] "Modulated" refers to an increase or decrease in effector function (e.g., ADCC, CDC, and / or ADCP) or binding to Fcy receptor (FcyR) or FcRn of an antibody of the application having at least one mutation in the Fc region as compared to the parent antibody without the mutation.

[0150] Mutations that modulate antibody binding to protein A can be introduced into antibodies used in the methods of the application. Production and purification of full-length bispecific therapeutic antibodies requires efficient separation of the bispecific antibody from excess parent and / or intermediate molecules. Thus, bispecific antibodies having Fc mutations that modulate protein A binding in an asymmetric fashion (e.g., only in one heavy chain) can be purified from parent antibodies based on their differential elution profile from a protein A affinity column. Exemplary mutations that can be introduced to modulate protein A binding of antibodies used in the methods of the application are Q311R, Q311K, T307P_L309Q, T307P_V309Q, T307P_L309Q_Q311R, T307P_V309Q_Q311R, H435R, or H435R_Y436F.

[0151] Glycoengineering

[0152] The ability of the antibodies used in the methods of the application to induce ADCC can be enhanced by engineering their oligosaccharide component. Human IgGl is N-glycosylated at Asn297, with the majority of the glycans being in the well-known biantennary G0, G0F, G1, G1F, G2 or G2F form. Antibodies produced from non-engineered CHO cells typically have a fucose content of about at least 85% of their oligosaccharides. Removal of core fucose from biantennary complex-type oligosaccharides attached to the Fc region can enhance ADCC of antibodies through improved FcyRIIIa binding, without altering antigen binding or CDC activity. Such mAbs can be obtained using different approaches reported to result in successful expression of relatively higher defucosylated antibodies carrying biantennary complex-type Fc oligosaccharides, such as controlling the osmolality of the culture, applying the variant CHO line Lec13 as host cell line, applying the variant CHO line EB66 as host cell line, applying the rat hybridoma cell line YB2 / 0 as host cell line, introducing small interfering RNA specific for the a1,6-fucosyltransferase (FUT8) gene, or co- expressing β-1,4-N-acetylglucosaminyltransferase III and Golgi a-mannosidase II or the potent a-mannosidase I inhibitor kifunensine.

[0153] The heterodimeric antibodies used in the methods of the application can have a biantennary glycan structure with a fucose content of about 0% to about 15%, e.g., 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or 0%.

[0154] The heterodimeric antibodies used in the methods of the application can have a biantennary glycan structure with a fucose content of about 50%, 40%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 14%, 13%, 12%, 11% 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or 0%.

[0155] "Fucose content" refers to the amount of fucose monosaccharide at Asn297 within the sugar chain. The relative amount of fucose is the percentage of fucose-containing structures relative to all sugar structures. These can be characterized and quantified by a variety of methods, for example: 1) MALDI-TOF using samples treated with N-glycosidase F (e.g., complex, hybrid, and oligo and high mannose structures); 2) by enzymatic release of Asn297 glycans, followed by derivatization and detection / quantification by HPLC (UPLC) and / or HPLC-MS (UPLC-MS) with fluorescence detection; 3) intact protein analysis of native or reduced mAbs, with or without treatment of Asn297 glycans with Endo S or other enzymes that cleave between the first and second GlcNAc monosaccharides, leaving fucose attached to the first GlcNAc; 4) digestion of mAbs into constituent peptides by enzymatic digestion (e.g., trypsin or endopeptidase Lys-C), followed by separation, detection, and quantification by HPLC-MS (UPLC-MS); or 5) isolation of mAb oligosaccharides from mAb proteins by specific enzymatic deglycosylation at Asn297 with PNGase F. The released oligosaccharides can be labeled with a fluorophore, separated, and identified by various complementary techniques that allow fine characterization of glycan structures by matrix-assisted laser desorption ionization (MALDI) mass spectrometry by comparing experimental mass to theoretical mass, determination of sialylation by ion exchange HPLC (GlycoSep C), separation and quantification of oligosaccharide forms by normal phase HPLC (GlycoSep N) according to hydrophilicity standards, and separation and quantification of oligosaccharides by high performance capillary electrophoresis-laser induced fluorescence (HPCE-LIF).

[0156] "Low fucose" or "low fucose content" refers to an antibody having a fucose content of about 0% to 15%.

[0157] "Normal fucose" or "normal fucose content" refers to an antibody having a fucose content of about more than 50%, typically about more than 60%, 70%, 80%, or more than 85%.

[0158] Heterodimeric antibodies used in the methods of the application can be post-translationally modified by processes such as glycosylation, isomerization, deglycosylation, or non-naturally occurring covalent modifications such as the addition of polyethylene glycol moieties (PEGylation) and lipidation. Such modifications can occur in vivo or in vitro. For example, heterodimeric antibodies used in the methods of the application can be conjugated with polyethylene glycol (PEGylated) to improve their pharmacokinetic profile. Conjugation can be performed by techniques known to those of skill in the art. Conjugation of therapeutic antibodies with PEG has been shown to enhance pharmacodynamics without interfering with function.

[0159] C-terminal lysine

[0160] The C-terminal lysine (CTL) of the heterodimeric antibody used in the methods of the application can be partially removed during manufacturing. During manufacturing, the CTL can be removed to a level below the maximum by controlling the concentration of extracellular Zn 2+ , EDTA or EDTA - Fe 3+ as described in U.S. Patent Publication No. US20140273092. The CTL content in an antibody can be measured using known methods.

[0161] The heterodimeric antibody used in the methods of the application can have a C-terminal lysine content of about 10% to about 90%, about 20% to about 80%, about 40% to about 70%, about 55% to about 70%, or about 60%.

[0162] The heterodimeric antibody used in the methods of the application can have a C-terminal lysine content of about 0%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100%.

[0163] Antibody allotype

[0164] The heterodimeric antibody used in the methods of the application can be of any allotype. The allotype is not expected to have an effect on Fab-arm exchange. The immunogenicity of a therapeutic antibody is associated with an increased risk of infusion reactions and a shortened duration of therapeutic response (Baert et al., (2003) N Engl J Med 348:602-08). The extent to which a therapeutic antibody induces an immune response in a host can be determined in part by the allotype of the antibody (Stickler et al., (2011) Genes and Immunity 12:213-21). Antibody allotypes are associated with amino acid sequence variations at specific positions in the constant region sequence of the antibody. Table 2 shows selected IgGl, IgG2, and IgG4 allotypes.

[0165] Table 2

[0166]

[0167] Target antigen

[0168] The methods of the application can be used to generate hetero-dimeric antibodies with any specificity using Fab-arm exchange, as the reduction and reformation of disulfide bonds during manufacture is not expected to be affected by the antibody VH / VL regions, which is also demonstrated herein. Depending on the specificity of each VH / VL pair, the hetero-dimeric antibody can bind the same target antigen at two different epitopes, or can bind two or more target antigens. When the hetero-dimeric antibody binds two target antigens, the target antigens can be on the same cell or on two different cells. The target antigens can be tumor-associated antigens, antigens that play a role in inflammation, modulating T or B cell activity, or generally any target for which it is desirable to modulate its biological activity or for which it is desirable to reduce the number of cells expressing the target.

[0169] Exemplary antigens that can be bound by the heterodimeric antibodies used in the methods of the application are one or more of the following: ABCF1, ACVR1, ACVR1B, ACVR2, ACVR2B, ACVRL1, ADORA2A, Aggrecan, AGR2, AICDA, AIF1, AIG1, AKAP1, AKAP2, albumin, AMH, AMHR2, ANGPT1, ANGPT2, ANGPTL3, ANGPTL4, ANPEP, APC, APOC1, APOE, AR, AZGP1 (zinc-a-glycoprotein), B7.1, B7.2, BAD, BAFF, BAG1, BAI1, BCL2, BCL6, BCMA, BDNF, BLNK, BLR1 (MDR15), BlyS, BMP1, BMP2, BMP3B (GDF10), BMP4, BMP6, BMP8, BMPR1A, BMPR1B, BMPR2, BPAG1 (netrin), BRCA1, BTLA, C19orf10 (IL27w), C3, C4A, C5, C5R1, CANT1, CASP1, CASP4, CAV1, CCBP2 (D6 / JAB61), CCL1 (1-309), CCL11 (eotaxin), CCL13 (MCP-4), CCL15 (MIP-1d), CCL16 (HCC-4), CCL17 (TARC), CCL18 (PARC), CCL19 (MIP-3b), CCL2 (MCP-1), MCAF, CCL20 (MIP-3a), CCL21 (MIP-2), SLC, exodus-2, CCL22 (MDC / STC-1), CCL23 (MPIF-1), CCL24 (MPIF-2 / eotaxin-2), CCL25 (TECK), CCL26 (eotaxin-3), CCL27 (CTACK / ILC), CCL28, CCL3 (MIP-1a), CCL4 (MIP-1b), CCL5 (RANTES), CCL7 (MCP-3), CCL8 (mcp-2), CCNA1, CCNA2, CCND1, CCNE1, CCNE2, CCR1 (CKR1 / HM145), CCR2 (mcp-1RB / RA), CCR3 (CKR3 / CMKBR3), CCR4, CCR5 (CMKBR5 / ChemR13), CCR6 (CMKBR6 / CKR-L3 / STRL22 / DRY6), CCR7 (CKR7 / EBI1), CCR8 (CMKBR8 / TER1 / CKR-L1), CCR9 (GPR-9-6), CCRL1 (VSHK1),CCRL2 (L-CCR), CD123, CD137, CD164, CD16a, CD16b, CD19, CD1C, CD20, CD200, CD-22, CD24, CD28, CD3, CD3 epsilon, CD30, CD32a, CD32b, CD33, CD37, CD38, CD39, CD4, CD40, CD40L, CD44, CD45RB, CD47, CD52, CD69, CD72, CD73, CD74, CD79A, CD79B, CD8, CD80, CD81, CD83, CD86, CD89, CD96, CDH1 (E-cadherin), CDH10, CDH12, CDH13, CDH18, CDH19, CDH20, CDH5, CDH7, CDH8, CDH9, CDK2, CDK3, CDK4, CDK5, CDK6, CDK7, CDK9, CDKN1A (p21Wap1 / Cip1), CDKN1B (p27Kip1), CDKN1C, CDKN2A (p16INK4a), CDKN2B, CDKN2C, CDKN3, CEBPB, CER1, CHGA, CHGB, chitinase, CHST10, CKLFSF2, CKLFSF3, CKLFSF4, CKLFSF5, CKLFSF6, CKLFSF7, CKLFSF8, CLDN3, CLDN7 (claudin-7), CLN3, CLU (clusterin), CMKLR1, CMKOR1 (RDC1), CNR1, COL18A1, COL1A1, COL4A3, COL6A1, CR2, CRP, CSF1 (M-CSF), CSF2 (GM-CSF), CSF3 (GCSF), CTLA4, CTNNB1 (b-catenin), CTSB (cathepsin B), CX3CL1 (SCYD1), CX3CR1 (V28), CXCL1 (GRO1), CXCL10 (IP-10), CXCL11 (I-TAC / IP-9), CXCL12 (SDF1), CXCL13, CXCL14, CXCL16, CXCL2 (GRO2), CXCL3 (GRO3), CXCL5 (ENA-78 / LIX), CXCL6 (GCP-2), CXCL9 (MIG), CXCR3 (GPR9 / CKR-L2), CXCR4, CXCR6 (TYMSTR / STRL33 / Bonzo), CYB5, CYC1, CYSLTR1, DAB2IP, DES, DKFZp451J0118, DNAM-1, DNCL1, DPP4, E2F1, ECGF1, EDG1, EFNA1, EFNA3, EFNB2, EGF, EGFR,ELAC2, ENG, ENOl, EN02, EN03, EPHB4, EPO, ERBB2 (Her-2), EREG, ERK8, ESRl, ESR2, F3 (TF), FADD, FasL, FASN, FCER1A, FCER2, FCGR3A, FGF, FGF1 (aFGF), FGF10, FGF11, FGF12, FGF12B, FGF13, FGF14, FGF16, FGF17, FGF18, FGF19, FGF2 (bFGF), FGF20, FGF21, FGF22, FGF23, FGF3 (int-2), FGF4 (HST), FGF5, FGF6 (HST-2), FGF7 (KGF), FGF8, FGF9, FGFR, FGFR3, FIGF (VEGFD), FIL1 (EPSILON), FIL1 (ZETA), FLJ12584, FLJ25530, FLRT1 (Fibrillin), FLT1, FOS, FOSL1 (FRA-1), FY (DARC), GABRP (GABAa), GAGEB1, GAGEC1, GALNAC4S-6ST, GATA3, GDF5, GFI1, GGT1, GITR, GITRL, GM-CSF, GNAS1, GN RH1, GPR2 (CCR10), GPR31, GPR44, GPR81 (FKSG80), GRCC10 (C10), GRP, GSN (Gelsolin), GSTP1, HAVCR2, HDAC4, HDAC5, HDAC7A, HDAC9, HGF, HIF1A, HIP1, Histamine and Histamine Receptors, HLA, HLA-A, HLA-DRA, HM74, HMOX1, HUMCYT2A, HVEM, ICEBERG, ICOS, ICOSL, IDO, ID2, IFN-a, IFNA1, IFNA2, IFNA4, IFNA5, IFNA6, IFNA7, IFNB1, IFNγ, IFNW1, IGBP1, IGF1, IGF1R, IGF2, IGFBP2, IGFBP3, IGFBP6, IL-1, IL10, IL10RA, IL10RB, IL11, IL11RA, IL-12, IL12A, IL12B, IL12RB1, IL12RB2, IL13, IL13RA1, IL13RA2, IL14, IL15, IL15RA, IL16, IL17, IL17B, IL17C, IL17R, IL18, IL18BP, IL18R1, IL18RAP, IL19, IL1A, IL1B, IL1F10, IL1F5, IL1F6, IL1F7, IL1F8, IL1F9,IL1HY1, IL1R1, IL1R2, IL1RAP, IL1RAPL1, IL1RAPL2, IL1RL1, IL1RL2, IL1RN, IL2, IL20, IL20RA, IL21R, IL22, IL22R, IL22RA2, IL23, IL24, IL25, IL26, IL27, IL28A, IL28B, IL29, IL2RA, IL2RB, IL2RG, IL3, IL30, IL3RA, IL4, IL4R, IL5, IL5RA, IL6, IL6R, IL6ST (Glycoprotein 130), IL7, IL7R, IL8, IL8RA, IL8RB, IL8RB, IL9, IL9R, ILK, INHA, INHBA, INSL3, INSL4, Insulin, Insulin Receptor, IRAK1, IRAK2, ITGA1, ITGA2, ITGA3, ITGA6 (a6 Integrin), ITGAV, ITGB3, ITGB4 (b4 Integrin), JAG1, JAK1, JAK3, JUN, K6HF, KAI1, KDR, KITLG, KIR, KLF5 (GC Box BP), KLF6, KLK10, KLK12, KLK13, KLK14, KLK15, KLK3, KLK4, KLK5, KLK6, KLK9, KRT1, KRT19 (Keratin 19), KRT2A, KRTHB6 (Hair-specific Keratin Type II), LAG-3, LAMA5, LDL, LEP (Leptin), LFA, Lingo-p75, Lingo-Troy, LPS, LTA (TNF-b), LTB, LTB4R (GPR16), LTB4R2, LTBR, MACMARCKS, MAG or Omgp, MAP2K7 (c-Jun), MDK, Mesothelin, c-Met, MIB1, Midkine, MIF, MIP-2, MKI67 (Ki-67), MMP2, MMP9, MS4A1, MSMB, MT3 (Metallothionein-III), MTSS1, MUC1 (Mucin), c-MYC, MYD88, NCK2, Neuronan, NFKB1, NFKB2, NGFB (NGF), NGFR, NgR-Lingo, NgR-Nogo66 (Nogo), NgR-p75, NgR-Troy, NKG2D, NKp46, NME1 (NM23A), NOX5, NPPB, NR0B1, NR0B2, NR1D1, NR1D2, NR1H2, NR1H3, NR1H4, NRII2, NRII3, NR2C1, NR2C2, NR2E1, NR2E3, NR2F1, NR2F2, NR2F6, NR3C1, NR3C2, NR4A1,NR4A2, NR4A3, NR5A1, NR5A2, NR6A1, NRP1, NRP2, NT5E, NTN4, ODZ1, OPRD1, OX-40, OX-40L, P2RX7, PAP, PART1, PATE, PAWR, PCA3, PCNA, PD-1, PD-L1, PDGFA, PDGFB, PECAM1, PF4 (CXCL4), PGF, PGR, phosphoglycoprotein, PIAS2, PIK3CG, PLAU (uPA), PLG, PLXDC1, PPBP (CXCL7), PPID, PR1, PRKCQ, PRKD1, PRL, PROC, PROK2, PSAP, PSCA, PSMA, PTAFR, PTEN, PTGS2 (COX-2), PTN, RAC2 (p21 Rac2), RARB, RGS1, RGS13, RGS3, RNF110 (ZNF144), ROBO2, ROR1, SI00A2, SCGB1D2 (lipophilin B), SCGB2A1 (mammaglobin 2), SCGB2A2 (mammaglobin 1), SCYE1 (endothelial monocyte-activating cytokine), SDF2, SERPINA1, SERPINA3, SERPINB5 (maspin), SERPINE1 (PAI-1), SERPINF1, SHBG, SLA2, SLC2A2, SLC33A1, SLC43A1, SLIT2, SPP1, SPRR1B (Spr1), ST6GAL1, STAB1, STAT6, STEAP, STEAP2, TB4R2, TBX21, TCP10, TDGF1, TEK, TF (transferrin receptor), TGFA, TGFB1, TGFB111, TGFB2, TGFB3, TGFBI, TGFBR1, TGFBR2, TGFBR3, TH1L, THBS1 (thrombospondin-1), THBS2, THBS4, THPO, TIE (Tie-1), TIGIT, TIM-3, TIMP3, tissue factor, TLR10, TLR2, TLR3, TLR4, TLR5, TLR6, TLR7, TLR8, TLR9, TNF, TNF-a, TNFAIP2 (B94), TNFAIP3, TNFRSF11A, TNFRSF1A, TNFRSF1B, TNFRSF21, TNFRSF5, TNFRSF6 (Fas), TNFRSF7, TNFRSF8, TNFRSF9, TNFSF10 (TRAIL), TNFSF11 (TRANCE), TNFSF12 (APO3L), TNFSF13 (April), TNFSF13B,TNFSF14 (HVEM-L), TNFSF15 (VEGI), TNFSF18, TNFSF4 (OX40 ligand), TNFSF5 (CD40 ligand), TNFSF6 (FasL), TNFSF7 (CD27 ligand), TNFSF8 (CD30 ligand), TNFSF9 (4-1BB ligand), TOLLIP, Toll-like receptor, TOP2A (topoisomerase Iia), TP53, TPM1, TPM2, TRADD, TRAF1, TRAF2, TRAF3, TRAF4, TRAF5, TRAF6, TREM1, TREM2, TRPC6, TSLP, TWEAK, VEGF, VEGFB, VEGFC, versican, VHLC5, VISTA, VLA-4, XCL1 (lymphotactin), XCL2 (SCM-1b), XCR1 (GPR5 / CCXCR1), YY1, and ZFPM2.

[0170] In some embodiments, the heterodimeric antibody used in the methods of the application binds CD3.

[0171] In some embodiments, the heterodimeric antibody used in the methods of the application binds CD3 and a tumor associated antigen (TAA).

[0172] In some embodiments, the heterodimeric antibody used in the methods of the application binds both CD123 and CD3 and both BCMA and CD3.

[0173] In some embodiments, the heterodimeric antibody used in the methods of the application binds EGFR and c-Met.

[0174] In some embodiments, the heterodimeric antibody that binds EGFR and c-Met comprises a first heavy chain (HC1) of SEQ ID NO: 4, a first light chain (LC1) of SEQ ID NO: 5, a second heavy chain (HC2) of SEQ ID NO: 6, and a second light chain (LC2) of SEQ ID NO: 7.

[0175] The application also provides a bispecific antibody produced by the methods of the application.

[0176] The application also provides a bispecific antibody that binds EGFR and c-Met produced by the methods of the application.

[0177] The present invention also provides a bispecific antibody that binds EGFR and c-Met, generated by the method of the present invention, comprising a first heavy chain (HC1) of SEQ ID NO:4, a first light chain (LC1) of SEQ ID NO:5, a second heavy chain (HC2) of SEQ ID NO:6, and a second light chain (LC2) of SEQ ID NO:7.

[0178] Production of homodimer and heterodimer antibodies

[0179] The first and second homodimer antibodies can be produced together or separately in a culture vessel such as a bioreactor. The first and second homodimer antibodies can be produced by co-expression in host cells or by generating individual homodimer antibodies using separate host cells. In the latter case, the host cells can be of the same or different origins.

[0180] "Host cell" refers to a cell into which a vector expressing at least one antibody heavy chain and one antibody light chain has been introduced. "Host cell" refers not only to the specific test cell but also to its offspring and stable cell lines derived from the specific test cell. Because certain modifications may occur in subsequent generations due to mutations or environmental influences, such offspring may differ from the parent cell but are still included within the scope of the term "host cell" as used herein.

[0181] Such host cells can be eukaryotic cells, prokaryotic cells, plant cells, or archaea cells. Examples of prokaryotic host cells include *Escherichia coli*, bacilli such as *Bacillus subtilis*, and other Enterobacteriaceae species such as *Salmonella*, *Serratia*, and various *Pseudomonas* species. Other microorganisms such as yeast can also be used for expression. *Yeast* genera (e.g., *Saccharomyces cerevisiae*) and *Pichia pastoris* are examples of suitable yeast host cells. Exemplary eukaryotic cells can be of mammalian, insect, bird, or other animal origin. Mammalian eukaryotic cells include immortalized cell lines such as hybridoma or myeloma cell lines such as SP2 / 0 (American Center for Type Culture Collection (ATCC), Manassas, VA, CRL-1581), NSO (European Center for Animal Cell Collection (ECACC), Salisbury, Wiltshire, UK, ECACC No. 85110503), FO (ATCC CRL-1646), and Ag653 (ATCC CRL-1580) mouse cell lines. An exemplary human myeloma cell line is U266 (ATTC CRL-TIB-196). Other useful cell lines include those derived from Chinese hamster ovary (CHO) cells, such as CHOK1SV (Lonza Biologics, Walkersville, MD). CHOK2SV (Lonza), CHO-K1 (ATCC CRL-61), or DG44.

[0182] Exemplary vectors that can be used to express one or more antibody heavy and light chains include pWLneo, pSV2cat, pOG44, PXR1, pSG (Stratagene), pSVK3, pBPV, pMSG, and pSVL (Pharmacia), pEE6.4 (Lonza), and pEE12.4 (Lonza), and vectors used in the Lonza Xceed system.

[0183] The first and second homodimeric antibodies can be produced by expressing them in separate host cells. In this case, the first and second homodimeric antibodies can be purified prior to mixing them in the presence of a reducing agent to initiate Fab-arm exchange. Purification can be accomplished using known methods such as protein A chromatography. Alternatively, the media containing the first and second homodimeric antibodies can be combined, where a reducing agent can be added to initiate Fab-arm exchange. If the first and second homodimeric antibodies are produced by co-expression, protein A chromatography can also be used to purify the antibodies.

[0184] Described herein is the production of heterodimeric antibodies by Fab-arm exchange. Process conditions such as the concentration of the first and second homodimeric antibodies, the reducing agent, the concentration of the reducing agent, the reduction time, the dissolved oxygen concentration, the temperature at which the reaction is performed, and the buffer used are optimized to produce heterodimeric antibodies in high yield and purity, as described herein.

[0185] Following Fab-arm exchange, the resulting heterodimeric antibodies can be further purified. Purification methods include protein A, protein G chromatography, other affinity chromatography modalities, such as affinity chromatography based on antigen binding or binding to anti-idiotypic antibodies, thioaffinity, ion exchange, hydrophobic interaction, hydroxyapatite chromatography, and other mixed mode resins. Other methods can use precipitation with, for example, salt or polyethylene glycol to obtain purified heterodimeric antibodies.

[0186] Apparatus suitable for use in the process of the application is known. Expression of the homodimeric antibody by the host cell can be carried out, for example, in a reaction vessel such as a bioreactor. The reduction and oxidation steps can be carried out in the same bioreactor as the expression of the first and / or second homodimeric antibody, or it can be carried out in a separate reactor vessel. The reaction vessels and supporting process tubing can be disposable or reusable, and made of standard materials (plastic, glass, stainless steel, etc.). The reaction vessels can be equipped with mixing, sparging, headspace gassing, temperature control, and / or monitoring with probes for temperature, weight / volume, pH, dissolved oxygen (DO), and oxidation-reduction potential measurements. All of these techniques are common in standard unit operations in manufacturing plants and are well known to those skilled in the art.

[0187] While the application has been generally described, embodiments of the application will be further disclosed in the following examples, which should not be construed as limiting the scope of the claims.

[0188] Example 1. General Methods

[0189] cSDS

[0190] Capillary sodium dodecyl sulfate-polyacrylamide gel electrophoresis separates proteins based on molecular weight. The analysis employs a commercial capillary electrophoresis system and a fused silica capillary in a temperature controlled cartridge. The test article is mixed with internal standards and an alkylating reagent, heated for a defined time and temperature to denature the proteins. The sample is then injected electrokinetically by voltage, followed by analysis for a predetermined duration by applying a larger electric field. Detection is accomplished by absorbance at 220 nm, and purity is determined by calculating the corrected peak area of the main peak relative to all other peaks in the post electropherogram to the internal standards.

[0191] IEX-HPLC

[0192] Percent bispecific antibody relative to residual parental homodimeric antibody is determined by ion exchange high performance liquid chromatography (IEX-HPLC). Ion exchange chromatography achieves separation by exploiting differences in surface charge on the molecules. IEX-HPLC is performed using a Thermo Scientific Propac WCX-10 (4 mm ID x 150 mm, packed with 10 μιη particles). Salt and / or pH gradients are used to resolve the separation of the material. As the gradient concentration increases, the overall surface interaction of each antibody to the column is neutralized and the eluate is detected at 280 nm. The relative amount of each IgG is measured by comparing the peak area counts.

[0193] HIC-HPLC

[0194] Hydrophobic interaction chromatography separates proteins based on their hydrophobicity. Proteins are retained on the column under high salt conditions and eluted by decreasing the salt concentration. High salt buffer conditions expose hydrophobic regions due to decreased protein solvation and promote binding to the HIC stationary phase. Samples were injected into a Tosoh TSKgel Butyl-NPR (4.6 mm x 10 cm, 2.5 um non-porous resin based beads) column and eluted with an ammonium sulfate gradient. Absorbance was monitored at 280 nm and peaks were identified by comparison to reference standards.

[0195] RP-HPLC (Detection of Cystamine)

[0196] Reversed phase high performance chromatography separates solute molecules in the mobile phase based on hydrophobic interactions with the stationary phase media. Utilizing the hydrophobicity difference between 2-MEA and the oxidized dimeric form of cystamine, resolution and quantification was performed by RP-HPLC by comparing measured values to a standard curve. A Phenomenex Luna 18 (4.6 mm x 150 mm, 3 um) column was employed using a hexanesulfonate / acetonitrile mobile phase system to separate the two species by a step gradient of increasing organic solvent. Species were detected by monitoring absorbance at 220 nm (2-MEA) and 255 nm (cystamine) and peak areas were compared to their respective standard curves.

[0197] General antibody production protocol for monospecifics and bispecifics (performed at 200 ml scale) and scaling

[0198] Each monospecific antibody was thawed from individual mammalian cell banks and amplified and expressed separately in bioreactors. Production bioreactors were supplemented with rich media and harvested prior to significant cell death. The bioreactor was harvested by centrifugation and / or filtration and the monospecific antibodies were captured separately by affinity protein A chromatography. The resulting captured monospecific antibodies were then stored frozen until combined in the FAE stage.

[0199] Example 2: Process improvements during the manufacture of bispecific antibodies using Fab-arm exchange

[0200] Bispecific antibodies are produced using Fab-arm exchange by introducing asymmetric mutations in the CH3 domains of two homodimeric parental antibodies, followed by reduction and reformation of intermolecular disulfide bonds. The asymmetric CH3 domain mutations drive preferential formation of the heterodimeric bispecific antibody over the homodimeric parental antibodies.

[0201] To ensure process robustness during manufacturing, experiments were designed to understand the possible thresholds and ranges of various parameters during manufacturing. It has been shown that antibody disulfide bond reformation is dependent on the presence of oxygen and free metals, as EDTA and anaerobic conditions inhibit the re-oxidation of cysteines to form disulfide bonds in antibodies (US 2014 / 0303356). Therefore, studies were initiated to investigate the necessity of controlling dissolved oxygen (D02) and metal amounts during the production of bispecific antibodies by Fab-arm exchange.

[0202] Figure 2 Manufacturing steps during the disulfide bond reduction and reformation process during the production of bispecific antibodies by Fab-arm exchange are shown.

[0203] Theoretically, antibody disulfide bond reformation after reduction with a reducing agent such as 2-mercaptoethylamine (2-MEA) can occur by multiple pathways. For example, thiol-disulfide exchange can lead to disulfide bond reduction and reformation. In order to perform thiol-disulfide exchange, the reducing agent such as 2-MEA needs to be in its thiol (RS-) form to perform thiol-disulfide exchange. Figure 3 Molecular details of the thiol-disulfide exchange reaction are shown. In order to reduce the disulfide bonds in the parent antibody, the thiolate anion (circled in Figure 3 ) attacks the sulfur atom of the disulfide bond, displacing one of the sulfur atoms and forming a new disulfide bond with the 2-MEA thiolate.

[0204] Using the Henderson-Hasselbalch equation shown below, the ratio of 2-MEA thiol (SH) and thiolate anion (S - ) is highly dependent on the pH and the 2-MEA thiol pKa.

[0205]

[0206] Therefore, the reduction of disulfide bonds and the overall Fab-arm exchange are inhibited at low pH, where the protonated thiol form of 2-MEA is favored over its deprotonated thiolate anion form. At higher pH, the equilibrium shifts towards the thiolate, enabling the reduction of disulfide bonds and thiol-disulfide exchange.

[0207] Figure 4 A summary of reactions that can occur during Fab-arm exchange in the presence of 2-MEA is shown. Removal of the reducing agent (e.g., 2-MEA) by buffer exchange (UF / DF) during Fab-arm exchange can enable disulfide bond reformation in the heterodimeric bispecific antibody, but multiple reaction pathways allow the disulfide reformation process to occur. In the presence of oxygen, oxygen can react with the protonated thiolate, facilitating the reformation of disulfide bonds in the heterodimeric bispecific antibody, as shown in Figure 4as shown in reaction 2 in Scheme 1. The presence of oxygen can also reduce the residual 2-MEA thiol in solution, forming cystamine dimer and water, as shown in reaction 4 in Scheme 1. Figure 4 Thus, a sufficiently high concentration of 2-MEA should be used during the Fab-arm exchange to improve the efficiency and yield of the bispecific antibody.

[0208] However, in the absence of oxygen, thiolate disulfide exchange and disulfide bond reformation can still be possible. Under oxygen-deprived conditions, due to the reduction in the concentration of 2-MEA during the buffer exchange, the antibody thiol groups and 2-aminoethyl disulfide Fab intermediates can drive the reaction in the opposite direction Figure 4 reaction 1 in Scheme 1, re-forming the disulfide bond in the bispecific heterodimeric antibody, with the production of 2-MEA thiolate. This reaction will continue to drive in the opposite direction as more 2-MEA is removed during the buffer exchange process.

[0209] Example 3: Manufacture of bispecific antibody A under low DO2 conditions during UF / DF

[0210] Experimental Methods

[0211] In this experiment, ambient DO2 levels were present during the reduction period and for up to 23 hours after the addition of 2-MEA, after which a nitrogen blanket was included to minimize the %DO2 during the UF / DF. This method reflects conditions where oxygen is present during the reduction phase and absent during the disulfide reformation.

[0212] Bispecific antibody A binds to BCMA and CD3 and is of the IgG4 isotype with S228P, F234A, L235A substitutions (“PAA substitutions”) in both heavy chains, F at position 405 and R at position 409 in one heavy chain and L at position 405 and K at position 409 in the second heavy chain to drive heterodimer formation. The parental antibodies are designated p1A-IgG4PAAF405R409 and p2A-IgG4PAAL405K409.

[0213] The parental antibodies p1A-IgG4PAAF405R409 and p2A-IgG4PAAL405K409 were harvested from cell culture bioreactors and purified by protein A affinity chromatography.

[0214] A solution was prepared using p1A-IgG4P AAF405R409 and p2A-IgG4P AAL405K409 at a molar ratio or 1 : 1.06 to intentionally limit one of the parental antibodies. The mixture was then adjusted to pH 7.3 and diluted to a total IgG concentration of 10.5 g / L using 101 mM sodium acetate, 105 mM Tris base, pH 7.3. A 50 mM sodium acetate, 800 mM 2-MEA pH 5.0 stock was added to the parental mAb mixture. The final reduction buffer composition prior to UF / DF was about 100 mM sodium acetate, about 35 mM 2-MEA, about 30 mM NaCl pH 7.3. The reduced parental mAb solution was incubated at 24 °C for 23 hours, followed by transfer to the UF / DF retentate vessel. Prior to the start of ultrafiltration (UF) and diafiltration (DF), a nitrogen blanket was delivered to the retentate vessel via the headspace of the vessel to abate DO2during the period of disulfide bridge reforming and to maintain throughout the UF and DF processes. To prevent the addition of DO2by the diafiltration buffer, the diafiltration buffer (100 mM Tris-acetate, 30 mM NaCl pH 7.5) was sparged with nitrogen gas throughout the process. The use of an online DO2sensor verified that the retentate DO2had reached a level of less than 5% DO2and the diafiltration buffer had reached a level of less than 1% DO2prior to the start of the UF and DF steps.

[0215] Throughout the UF and DF processes, DO2levels were measured in the feed inlet, retentate, and permeate lines during UF / DF and were maintained at <4% DO2in all. The diafiltration buffer was maintained at a level of <1% DO2. The UF / DF system was maintained at a crossflow flow rate of 54 mL / min, targeting a transmembrane pressure of 14.5 psi for 11 diafiltration volumes (DV). All processes following incubation were performed at room temperature (18-22 °C). Upon completion of the UF / DF, the recovered retentate was quenched with 1.0 M acetic acid and adjusted to pH 5.0 to minimize further disulfide bond formation, followed by storage at 2-8 °C. No additional buffer recovery flush was performed upon completion of the UF / DF to prevent any possible oxidation of the material from the introduction of additional diafiltration buffer.

[0216] The UF / DF setup was equipped to measure DO2at 3 locations during the process. The % DO2was monitored online in the feed inlet, retentate, and permeate lines as shown in Figure 5 Upon reduction of DO2, the levels in the permeate were maintained below 5% and the levels in the inlet / retentate were maintained below 3%. Table 3 shows a summary of the UF / DF process parameters and performance.

[0217] Table 3

[0218]

[0219]

[0220] Table 4 shows a summary of off-line %DO2, pH, and conductivity measurements before and after the addition of 2-MEA. Figure 6 The %DO2 during UF / DF measured in the retentate, permeate, and inlet line is shown. After nitrogen overlay of the retentate in the recirculation loop, the %DO2 stabilized at 3.2% in the permeate, 1.5% in the retentate, and 1.4% in the feed over time.

[0221] Non-reducing cSDS analysis was used to measure the disulfide bond integrity of the heterodimer bispecific antibody formed (% purity of the heterodimer antibody on non-reducing cSDS). The purity of antibody A was determined to be 97.59%.

[0222] This study shows that bispecific antibody A was formed during reduction at ambient %DO2 (ranging from 89-13%) and at low %DO2 environments (<4% DO2 during UF / DF) with high levels of disulfide bond formation (reflected as % purity on non-reducing cSDS). It was previously thought that DO2 was required to oxidize the thiols to disulfides to stabilize the bispecific antibody.

[0223] Table 4

[0224] Step Dissolved oxygen (%) pH Conductivity (mS / cm) Parent IgG mixture 88.1 4.99 3.37 pH adjusted parent pool 87.5 7.35 4.43 Diluted pH adjusted parent pool 88.9 7.33 4.39 Post 2-MEA addition (T=0) 36.9 6.99 7.73 Post 2-MEA addition (T=2 hr) 18.7 Post 2-MEA addition (T=23 hr) 13.2

[0225] The oxygen levels present during the study were calculated to be insufficient to contribute to the oxidation of thiols during UF / DF. The following calculations show that using the reaction represented by Reaction 2 in Figure 4 requires a DO2 concentration in solution higher than 30% for the oxidation of thiols to occur.

[0226] For this study, the following assumptions were made to calculate the theoretical % DO2 saturated solution to oxidize all thiols present in the amount of antibody:

[0227] • IgG solution: 10 g / L

[0228] • moles S-S (disulfides to be reduced and oxidized) / moles mAb

[0229] • ½ moles O2 / moles S-S. If oxygen is required to generate disulfides from thiols, see Reaction 2 in Figure 4

[0230] • Oxygen solubility in air-saturated water = 7 ppm (g O2 / 106 g solution = 100% saturation

[0231] • mAb MW: 150 kDa

[0232] ​To calculate the % dissolved DO2 concentration of an oxidized 10 mg / mL antibody solution when 4 disulfides / mole of mAb are oxidized and the O2 required to oxidize all thiols:

[0233] 1. (10 mg mAb / mL solution) * (moles mAb / 150,000 g mAb) * (2 moles S-S / mole mAb) * (1 / 2 moles O2 / mole S-S) * (1000 mL / 1 L) (1 g / 1000 mg) = 6.67 x 10 -4 moles O2 / L solution

[0234] 2. (6.67 x 10 -5 moles O2 / L solution) * (31.998 g O 2 / 2.13 x 10 -6 g O2 / g solution

[0235] 3. (2.13 x 10 -6 g O2 / g solution) * (10 6 g solution) = 2.13 g O2 / 10 6 g solution = 2.13 ppm

[0236] 4. Air saturated water is about 7 g O2 / 10 6 g solution = 7 ppm

[0237] 5. 2.13 ppm / 7 ppm = 30% O2 saturated solution is required

[0238] 2% DO2 at operating temperature is equivalent to 0.192 mg / mL or 6.00 μΜ, 3% DO2 is equivalent to 0.288 mg / L or 8.99 μΜ, 4% DO2 is equivalent to 0.384 mg / L or 11.29 μΜ, 5% DO2 is equivalent to 0.480 mg / mL or 14.99 μΜ. 30% DO2 is equivalent to 2.88 mg / mL or 90 μΜ.

[0239] This indicates that there is a chemical reaction pathway that is independent of oxygen that allows for reformation of disulfide bonds in the absence of oxygen during the UF / DF step. When 2-MEA is removed from the solution, Figure 4 the reverse reaction of reaction 1 of

[0240] Conclusion: Maintaining DO2 levels <4% during UF / DF does not significantly affect disulfide bond formation. Purity of the bispecific antibody is >97% at completion of the UF / DF.

[0241] Example 4: Manufacture of bispecific antibody B at low DO2 during UF / DF

[0242] In this experiment, ambient DO2levels were present during reduction and for 23 hours after addition of 2-MEA, after which a nitrogen blanket was included to minimize %DO2during UF / DF. This approach reflects conditions where oxygen is present during the reduction phase and absent during disulfide bridge reformation.

[0243] Bispecific antibody B binds to CD123 and CD3 and is of IgG4 isotype with S228P, F234A, L235A substitutions (“PAA substitutions”) in both heavy chains, F at position 405 and R at position 409 in one heavy chain, and L at position 405 and K at position 409 in the second heavy chain to drive heterodimer formation. Parental antibodies are designated p1B-IgG4PAAF405R409 and p2B-IgG4PAAL405K409.

[0244] Parental antibodies p1B-IgG4PAAF405R409 and p2B-IgG4PAAL405K409 were harvested from cell culture bioreactors and purified by protein A affinity chromatography.

[0245] Solutions of p1B-IgG4PAAF405R409 and p2B-IgG4PAAL405K409 were prepared and adjusted to pH 7.3 and diluted to a total IgG concentration of 10.5 g / L using 101 mM sodium acetate, 105 mM Tris base, pH 7.3. A 50 mM sodium acetate, 800 mM 2-MEA pH 5.0 stock was added to the parental mixture. The final reduction buffer composition prior to UF / DF was about 100 mM sodium acetate, about 35 mM 2-MEA, about 30 mM NaCl, pH 7.3. The reduced parental solutions were incubated at 24 °C for 23 hours, after which they were transferred to the UF / DF retentate vessel. Prior to the start of ultrafiltration (UF) and diafiltration (DF), a nitrogen blanket was delivered to the retentate vessel via the headspace of the vessel to abate DO2during disulfide bridge reformation and to maintain throughout the UF and DF processes. To prevent addition of DO2from the diafiltration buffer, the diafiltration buffer (100 mM Tris-acetate 30 mM NaCl pH 7.5) was sparged with nitrogen gas throughout the process. An online DO2sensor was used to verify that the retentate had reached a level of less than 5% DO2and the diafiltration buffer had reached a level of less than 1% prior to the start of the UF and DF steps.

[0246] During the entire UF and DF process, DO2levels in the feed inlet, retentate, and permeate lines during UF / DF were measured and were maintained at <4% DO2. The diafiltration buffer was maintained at a level of <1% DO2. The UF / DF system was maintained at a crossflow rate of 42 mL / min, targeting a transmembrane pressure of 15.0 psi for 11 diafiltration volumes (DV). All processes following incubation were performed at room temperature (18-22°C). Upon completion of the UF / DF, a 3 mL sample was taken, left unadjusted at pH 7.5, and immediately stored at -70°C to prevent further oxidation. The remaining volume of recovered retentate was quenched with 1.0 M acetic acid and adjusted to pH 5.0 to minimize further disulfide bond formation. The bulk material was then stored at -70°C. No additional buffer recovery flush was performed upon completion of the UF / DF to prevent any possible oxidation of the material from the introduction of additional diafiltration buffer, but the yield was lower than typically observed.

[0247] The UF / DF setup was equipped to measure DO2at 3 locations during the process. Figure 5 On-line monitoring of % DO2in the feed inlet, retentate, and permeate lines is shown. Upon reduction of DO2, the levels in the permeate remained below 5% and the concentration in the inlet / retentate remained below 3%.

[0248] Table 5 shows a summary of the UF / DF process parameters. Figure 7 % DO2measured during UF / DF in the retentate, permeate, and inlet lines is shown. Table 6 shows a summary of the off-line % DO2, pH, and conductivity measurements.

[0249] Table 5

[0250] Final 2-MEA concentration (mM) 35.0 TMP (psi) 15.0-16.5 Crossflow (mL / min) 42-45 Diafiltration concentration (g / L) 25.0 Total diafiltration volume (DVs) 11 Average permeate flux (L / m2 / hr) 17.7 Loading ratio (g / m2) 320 Yield (%) 85

[0251] Table 6

[0252]

[0253] Non-reducing cSDS analysis was used to measure the disulfide bond integrity in the bispecific antibodies formed in the frozen sample at pH 7.5 and the bulk material at pH 5.0. The purity of the bispecific antibody preparation at pH 7.5 was measured to be 97.16% and the purity of the antibody preparation at pH 5.0 was measured to be 97.27%.

[0254] This study demonstrates that bispecific antibody B is formed with high levels of disulfide bond formation under low % DO2environments during reduction (ranging from 83-24%) and during UF / DF (<4% DO2).

[0255] Both the study described in Example 2 and the study described in this example show that oxygen is not essential for the reformation of disulfide bridges during the Fab-arm exchange process during the UF / DF step.

[0256] Conclusion: Maintaining DO2 levels at <4% during the UF / DF step for the production of bispecific antibody B using Fab-arm exchange does not significantly impact disulfide bond formation. At the completion of the UF / DF, non-reduced cSDS shows % purity >97%.

[0257] Example 5: Production of bispecific antibody A with low DO2 in the presence of EDTA during UF / DF

[0258] During the production of bispecific antibodies there is an opportunity to introduce free metal ions into the manufacturing process. These free metal ions can participate in the redox reactions of the Fab-arm exchange during the preparation of the buffer from raw materials and leaching from the metal components. To investigate the possible impact of trace metals in the Fab-arm exchange, the addition of EDTA during the Fab-arm exchange process was investigated. The addition of EDTA sequesters possible free metal ions in the solution that can catalyze the oxidation reactions to reform disulfide bonds.

[0259] Solutions were prepared using p1A-IgG4P AAF405R409 and p2A-IgG4P AAL405K409 at a molar ratio of 1.05:1.00. The mixture was then adjusted to pH 7.3 and diluted to a total IgG concentration of 10.5 g / L using 101 mM sodium acetate, 105 mM Tris base, pH 7.3. A 50 mM sodium acetate, 800 mM 2-MEA, pH 5.0 stock was added to the parental mAb mixture. The final reduction buffer composition prior to UF / DF was approximately 100 mM sodium acetate, approximately 35 mM 2-MEA, approximately 30 mM NaCl, pH 7.3. The reduced parental solution was incubated at 24 °C for 23.5 hours. At the completion of the incubation, a 500 mM EDTA, pH 8.0 stock was added to the reduced parental mixture to a target of 2 mM EDTA to sequester free metal ions that would be present in the solution prior to the start of the UF / DF. EDTA was also added to the diafiltration buffer to a target of 100 mM Tris-acetate, 30 mM NaCl, 2 mM EDTA, pH 7.5. This was done to ensure that no additional free metal ions could be introduced during the UF / DF buffer exchange. No EDTA was added during the reduction step to allow for any possible disulfide oxidation and cystamine formation that can occur spontaneously during this step.

[0260] The reduced parent solution containing EDTA was then transferred to the UF / DF retentate vessel, and a nitrogen blanket was delivered to the retentate vessel via the headspace of the vessel prior to the start of UF to abate DO2from the retentate and maintain throughout the UF and DF processes. To prevent the addition of DO2through the diafiltration buffer, the diafiltration buffer was sparged with nitrogen gas throughout the process. In-line DO2sensors confirmed that the retentate DO2levels reached a level of less than 2% DO2and the diafiltration buffer reached a DO2level of less than 1% DO2prior to the start of the UF and DF steps.

[0261] Throughout the UF and DF processes, the DO2levels in the feed inlet, retentate, and permeate lines during UF / DF were measured and were all maintained at <4% DO2. The diafiltration buffer was maintained at a level of <1% DO2. The UF / DF system was maintained at a cross-flow rate of 58 mL / min, targeting a transmembrane pressure of 14.5 psi for 11 diafiltration volumes (DV). All processes following incubation were performed at room temperature (18-22 °C). Upon completion of the UF / DF, a 9 mL sample was taken, left unadjusted at pH 7.5, and immediately stored at -70 °C to prevent further oxidation. The remaining volume of recovered retentate was quenched with 1.0 M acetic acid and adjusted to pH 5.0 to minimize further disulfide bond formation. The bulk bispecific antibody sample at pH 5.0 was then stored at -70 °C. No buffer recovery flush was performed upon completion of the UF / DF to prevent possible oxidation of the material from the introduction of additional diafiltration buffer. Table 7 summarizes the UF / DF process parameters and performance.

[0262] Table 7

[0263] Final 2-MEA concentration (mM) 35.0 TMP (psi) 14.8-15.5 Crossflow (mL / min) 58 Diafiltration concentration (g / L) 25.0 Total diafiltration volume (DVs) 11 Average permeate flux (L / m2 / hr) 30.9 Loading ratio (g / m2) 318 Yield (%) N / A

[0264] Non-reducing cSDS analysis was used to measure the disulfide bond integrity in the formed bispecific antibody of the frozen sample at pH 7.5 and the bulk material at pH 5.0. The purity of the bispecific antibody preparation at pH 7.5 was 97.44% and the purity of the bispecific antibody preparation at pH 5.0 was measured to be 97.39%.

[0265] This study demonstrates that bispecific antibody A was formed during UF / DF under low % DO2environments and minimal free metal ions available, with high levels of disulfide bond formation. This study demonstrates that Fab-arm exchange can occur during UF / DF without the need for free metal ions to catalyze the oxidation reaction.

[0266] Conclusion: The experimental results demonstrate that the addition of EDTA to the reduced parent mixture prior to UF / DF in Fab-arm exchange did not affect disulfide bond formation. Non-reducing cSDS resulted in >97% purity at the completion of UF / DF.

[0267] Example 6: Comparison of Fab-arm exchange under environmental and low DO2 conditions during reduction and UF / DF

[0268] This study was designed to assess the hypothesis that certain DO2-mediated oxidation can occur during 2-MEA reduction.

[0269] For this study, Fab-arm exchange was performed using parental antibodies p1B-IgG4PAAF405R409 and p2B-IgG4PAAL405K409. A mixture of p1B-IgG4PAAF405R409 and p2B-IgG4PAAL405K409 was prepared in solution and adjusted to pH 7.3, then diluted to a total IgG concentration of 10.5 g / L using 101 mM sodium acetate, 105 mM Tris base, and pH 7.3. The mixture containing the parental antibodies was aliquoted into two separate containers. In one container, a nitrogen-coated layer was delivered via headspace to reduce DO2 from the parental mixture until %DO2 < 5%. The other parental mixture container was kept under ambient air conditions as a reduction control. A stock solution of 50 mM sodium acetate and 800 mM 2-MEA at pH 5.0 was added to each parental mixture. The final reduction buffer composition prior to UF / DF was approximately 100 mM sodium acetate, approximately 35 mM 2-MEA, and approximately 30 mM NaCl, pH 7.3. Both parental reduction solutions were incubated at 24°C for 24 hours, and the %DO2 in both containers was measured. The results showed... Figure 8 Throughout the incubation process, the low DO2 reduction vessel was kept under a nitrogen cover and maintained at <2% DO2 for 24 hours. In the vessel kept under ambient air, %DO2 initially decreased to approximately 40% DO2 after 2-MEA doping, and thereafter gradually increased linearly back to >90% DO2. A summary of the UF / DF process parameters and performance is shown in Table 8.

[0270] Table 8

[0271] Final 2-MEA concentration (mM) 35.0 TMP (psi) 14.5-16.0 Crossflow (mL / min) 55 Diafiltration concentration (g / L) 25.0 Total diafiltration volume (DVs) 12 Average permeate flux (L / m2 / hr) 19.5 Loading ratio (g / m2) 251 Yield (%) N / A

[0272] During reduction, samples were obtained from both containers through sampling ports on the containers. Before each sample extraction, the ports were flushed, and the samples were immediately quenched by adding 10% v / v cystamine to the RP-HPLC mobile phase, 20 mM hexanesulfonate, and pH 2.0 buffer, verifying that each sample had been reduced to pH < 5 before analysis. Disulfide bond integrity at all time points was determined by non-reducing cSDS, and residual 2-MEA / cystamine concentration was determined by RP-HPLC.

[0273] The residual 2-MEA / cystamine assay requires additional sample preparation, which involves removing the antibody through a 5 kDa molecular weight retention filter. The results for non-reduced cSDS% purity and residual 2-MEA / cystamine are reported in Table 9.

[0274] Upon completion of the 24-hour reduction incubation, the low-DO2 container was transferred to the UF / DF system. A nitrogen overlay continued to be delivered to the leucovorin container via the top space of the container to continue reducing DO2 from the leucovorin and was maintained throughout the UF and DF processes. The percolation buffer was also nitrogen-protected throughout the treatment to prevent the introduction of DO2 through the buffer. Verification using an online DO2 sensor confirmed that the leucovorin had reached a level below 5% DO2 and the percolation buffer had reached a level below 1% DO2 prior to the start of the UF and DF steps.

[0275] Throughout the UF and DF processes, DO2 levels in the feed inlet, permeate, and permeate lines were measured during UF / DF and maintained at <4% DO2. The permeate buffer was maintained at <1% DO2. The UF / DF system was maintained at a crossflow rate of 55 mL / min, targeting a transmembrane pressure of 14.5 psi for 12 permeate volumes (DV), using a 100 mM Tris-acetate, 30 mM NaCl, pH 7.5 buffer. All processes after incubation were performed at room temperature (18–22 °C). After UF / DF, the recovered permeate was quenched with 1.0 M acetic acid and adjusted to pH 5.0 to minimize further disulfide bond formation. The entire material was then stored at -70 °C. No additional buffer recovery rinse was performed after UF / DF to prevent any potential oxidation of the material due to the introduction of additional permeate buffer.

[0276] For containers with low DO2% and control containers with ambient DO2, disulfide integrity was measured by non-reducing cSDS and residual 2-MEA / cysteine ​​was measured by RP-HPLC at different time points throughout reduction and UF / DF.

[0277] Table 9 presents a summary of the data. In this experiment, Fab-arm exchange reactions under oxygen deprivation conditions were continued only via UF / DF. For both Fab-arm exchange conditions, residual 2-MEA and cystamine were measured only by reduction; however, for oxygen deprivation conditions during UF / DF, samples were obtained to measure possible disulfide bond reformation by non-reducing cSDS.

[0278] Table 9

[0279]

[0280]

[0281] *Initial t=0 samples were collected immediately after the addition of 2-MEA to the parent mixture. The sampling time points were used to determine the rates of disulfide bond reduction and reformation during reduction and UF / DF.

[0282] Conclusion: This data supports the hypothesis that disulfide bond formation is mediated by an oxygen-independent pathway in the absence of oxygen during reduction and UF / DF.

[0283] During Fab-arm exchange under deprivation conditions, minimal cystamine formation and disulfide bond reformation were observed during reduction. Only 1.6 mM cystamine and 14.74% intact antibody were observed after 24 hours of incubation in the presence of 2-MEA. However, once the antibody sample was treated with UF / DF, a progressively increasing level of disulfide bond reformation was evident with the removal of 2-MEA. At the end of DV (DV12), the bispecific antibody formulation formed at pH 5.0 was measured to have a purity of 90.59%.

[0284] In ambient DO2, disulfide bond reduction occurs rapidly, with approximately 86% of the parental antibody reduced within 30 minutes of the start of reduction. Disulfide bond reformation is observed during the remainder of the reduction, even with the continued presence of residual 2-MEA in solution. At the completion of the reduction phase, 71% of the antibody retains intact disulfide bonds. The formation of cystamine dimers and the increasing levels of bispecific antibody formation during the reduction incubation indicate that oxygen was consumed by reactions 4 and 2, respectively. Figure 4 As shown in the figure. This is supported by RP-HPLC data, which showed a significant reduction in 2-MEA and its conversion to the dimer compound cystamine ( Figure 4 Reaction 4 in the middle.

[0285] This study indicates that DO2 catalyzes the formation of interchain disulfide bonds in IgG HC-HC and LC-HC during the reduction phase, but it is not essential for the formation of intact bispecific antibodies because the disulfide bonds reform even under oxygen deprivation conditions. Even in the absence of oxygen, the removal of 2-MEA during percolation leads to the formation of... Figure 4 The reverse reaction of reaction 1 in the reaction drives the reformation of disulfide bonds.

[0286] Example 7: Generation of bispecific EGFR / c-Met antibodies using high concentrations of parental antibodies via Fab-arm exchange in environmental DO2.

[0287] This study was designed to assess the hypothesis that higher homodimer mixture protein concentrations and different reducing agent-to-protein mass ratios produce oxidized heterodimer antibodies.

[0288] This study used a bispecific EGFR / cMet antibody. The EGFR / cMet antibody contains the first heavy chain (HC1) of SEQ ID NO:4, the first light chain (LC) of SEQ ID NO:5, the second HC (HC2) of SEQ ID NO:6, and the second LC (LC2) of SEQ ID NO:7.

[0289] For this study, Fab-arm exchange was performed using parental EGFR and c-Met antibodies (both IgG1 isotypes). Neutralized parental antibody protein A eluent mixtures were prepared by filtering the solution and adjusting the pH to 7.9. These mixtures were then diluted to a total IgG concentration of 10–35 g / L using 209.5 mM sodium acetate and 300 mM NaCl at pH 7.9. 50 mM sodium acetate and 800 mM 2-MEA stock solution at pH 5.0 were added to each parental mixture to target reducing agent concentrations of 35, 50, or 100 mM. Both reduced parental solutions were incubated for 3 hours under controlled conditions of 21.5 °C, 24 °C, or uncontrolled conditions at ambient room temperature. After 3 hours, the incubated mixtures were ultrafiltered and percolated against 10 mM Tris, 7.8 mM acetate, and pH 7.5. Table 10 summarizes the operating conditions and analytical results. This study demonstrates the oxidation of heterodimers under the following conditions: a temperature range of 21.5–24 °C, a reducing agent concentration of 35–100 mM, and a homodimer mixture protein concentration of 10–35 g / L. The integrity and purity of the bispecific antibody were assessed by NR-cSDS and IHC-HPLC using the methods described in Example 1.

[0290] Table 10

[0291]

[0292]

[0293] SEQ ID NO:4(HC1)

[0294]

[0295] SEQ ID NO:5(LC1)

[0296]

[0297] SEQ ID NO:6(HC2)

[0298]

[0299] SEQ ID NO:7(LC2)

[0300] sequence list <110> Janssen Biotech, Inc. Alfonso Martin, Pedro Jose Capaldi, Michael Cohen, Jeffrey Detzel, Andrew Sakyiama, Joseph <120> Methods for generating heterodimeric antibodies <130> JBI6029WOPCT1 <140> Pending allocation <141> 2019-12-xx <160> 7 <170> PatentIn version 3.5 <210> 1 <211> 330 <212> PRT <213> Homo sapiens <400> 1 Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro Ser Ser Lys 1 5 10 15 Ser Thr Ser Gly Gly Thr Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr 20 25 30 Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala Leu Thr Ser 35 40 45 Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser 50 55 60 Leu Ser Ser Val Val Thr Val Pro Ser Ser Ser Leu Gly Thr Gln Thr 65 70 75 80 Tyr Ile Cys Asn Val Asn His Lys Pro Ser Asn Thr Lys Val Asp Lys 85 90 95 Lys Val Glu Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys 100 105 110 Pro Ala Pro Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro 115 120 125 Lys Pro Lys Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys 130 135 140 Val Val Val Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp 145 150 155 160 Tyr Val Asp Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu 165 170 175 Glu Gln Tyr Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu 180 185 190 His Gln Asp Trp Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn 195 200 205 Lys Ala Leu Pro Ala Pro Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly 210 215 220 Gln Pro Arg Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Asp Glu 225 230 235 240 Leu Thr Lys Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr 245 250 255 Pro Ser Asp Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn 260 265 270 Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe 275 280 285 Leu Tyr Ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn 290 295 300 Val Phe Ser Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr 305 310 315 320 Gln Lys Ser Leu Ser Leu Ser Pro Gly Lys 325 330 <210> 2 <211> 326 <212> PRT <213> Homo sapiens <400> 2 Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro Cys Ser Arg 1 5 10 15 Ser Thr Ser Glu Ser Thr Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr 20 25 30 Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala Leu Thr Ser 35 40 45 Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser 50 55 60 Leu Ser Ser Val Val Thr Val Pro Ser Ser Asn Phe Gly Thr Gln Thr 65 70 75 80 Tyr Thr Cys Asn Val Asp His Lys Pro Ser Asn Thr Lys Val Asp Lys 85 90 95 Thr Val Glu Arg Lys Cys Cys Val Glu Cys Pro Pro Cys Pro Ala Pro 100 105 110 Pro Val Ala Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro Lys Asp 115 120 125 Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val Val Asp 130 135 140 Val Ser His Glu Asp Pro Glu Val Gln Phe Asn Trp Tyr Val Asp Gly 145 150 155 160 Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Phe Asn 165 170 175 Ser Thr Phe Arg Val Val Ser Val Leu Thr Val Val His Gln Asp Trp 180 185 190 Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys Gly Leu Pro 195 200 205 Ala Pro Ile Glu Lys Thr Ile Ser Lys Thr Lys Gly Gln Pro Arg Glu 210 215 220 Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Glu Glu Met Thr Lys Asn 225 230 235 240 Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp Ile 245 250 255 Ser Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys Thr 260 265 270 Thr Pro Pro Met Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Lys 275 280 285 Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn Val Phe Ser Cys 290 295 300 Ser Val Met His Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser Leu 305 310 315 320 Ser Leu Ser Pro Gly Lys 325 <210> 3 <211> 327 <212> PRT <213> Homo sapiens <400> 3 Ala Ser Thr Lys Gly Pro Ser Val Phe Pro Leu Ala Pro Cys Ser Arg 1 5 10 15 Ser Thr Ser Glu Ser Thr Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr 20 25 30 Phe Pro Glu Pro Val Thr Val Ser Trp Asn Ser Gly Ala Leu Thr Ser 35 40 45 Gly Val His Thr Phe Pro Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser 50 55 60 Leu Ser Ser Val Val Thr Val Pro Ser Ser Ser Leu Gly Thr Lys Thr 65 70 75 80 Tyr Thr Cys Asn Val Asp His Lys Pro Ser Asn Thr Lys Val Asp Lys 85 90 95 Arg Val Glu Ser Lys Tyr Gly Pro Pro Cys Pro Ser Cys Pro Ala Pro 100 105 110 Glu Phe Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro Lys 115 120 125 Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val Val 130 135 140 Asp Val Ser Gln Glu Asp Pro Glu Val Gln Phe Asn Trp Tyr Val Asp 145 150 155 160 Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Phe 165 170 175 Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu His Gln Asp 180 185 190 Trp Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys Gly Leu 195 200 205 Pro Ser Ser Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg 210 215 220 Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Gln Glu Glu Met Thr Lys 225 230 235 240 Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp 245 250 255 Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys 260 265 270 Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser 275 280 285 Arg Leu Thr Val Asp Lys Ser Arg Trp Gln Glu Gly Asn Val Phe Ser 290 295 300 Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser 305 310 315 320 Leu Ser Leu Ser Leu Gly Lys 325 <210> 4 <211> 455 <212> PRT <213> artificial sequence <220> <223> EGFR / cMet antibody HC1 <400> 4 Gln Val Gln Leu Val Glu Ser Gly Gly Gly Val Val Gln Pro Gly Arg 1 5 10 15 Ser Leu Arg Leu Ser Cys Ala Ala Ser Gly Phe Thr Phe Ser Thr Tyr 20 25 30 Gly Met His Trp Val Arg Gln Ala Pro Gly Lys Gly Leu Glu Trp Val 35 40 45 Ala Val Ile Trp Asp Asp Gly Ser Tyr Lys Tyr Tyr Gly Asp Ser Val 50 55 60 Lys Gly Arg Phe Thr Ile Ser Arg Asp Asn Ser Lys Asn Thr Leu Tyr 65 70 75 80 Leu Gln Met Asn Ser Leu Arg Ala Glu Asp Thr Ala Val Tyr Tyr Cys 85 90 95 Ala Arg Asp Gly Ile Thr Met Val Arg Gly Val Met Lys Asp Tyr Phe 100 105 110 Asp Tyr Trp Gly Gln Gly Thr Leu Val Thr Val Ser Ser Ala Ser Thr 115 120 125 Lys Gly Pro Ser Val Phe Pro Leu Ala Pro Ser Ser Lys Ser Thr Ser 130 135 140 Gly Gly Thr Ala Ala Leu Gly Cys Leu Val Lys Asp Tyr Phe Pro Glu 145 150 155 160 Pro Val Thr Val Ser Trp Asn Ser Gly Ala Leu Thr Ser Gly Val His 165 170 175 Thr Phe Pro Ala Val Leu Gln Ser Ser Gly Leu Tyr Ser Leu Ser Ser 180 185 190 Val Val Thr Val Pro Ser Ser Ser Leu Gly Thr Gln Thr Tyr Ile Cys 195 200 205 Asn Val Asn His Lys Pro Ser Asn Thr Lys Val Asp Lys Arg Val Glu 210 215 220 Pro Lys Ser Cys Asp Lys Thr His Thr Cys Pro Pro Cys Pro Ala Pro 225 230 235 240 Glu Leu Leu Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Lys Pro Lys 245 250 255 Asp Thr Leu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val Val 260 265 270 Asp Val Ser His Glu Asp Pro Glu Val Lys Phe Asn Trp Tyr Val Asp 275 280 285 Gly Val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu Glu Gln Tyr 290 295 300 Asn Ser Thr Tyr Arg Val Val Ser Val Leu Thr Val Leu His Gln Asp 305 310 315 320 Trp Leu Asn Gly Lys Glu Tyr Lys Cys Lys Val Ser Asn Lys Ala Leu 325 330 335 Pro Ala Pro Ile Glu Lys Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg 340 345 350 Glu Pro Gln Val Tyr Thr Leu Pro Pro Ser Arg Glu Glu Met Thr Lys 355 360 365 Asn Gln Val Ser Leu Thr Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp 370 375 380 Ile Ala Val Glu Trp Glu Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys 385 390 395 400 Thr Thr Pro Pro Val Leu Asp Ser Asp Gly Ser Phe Leu Leu Tyr Ser 405 410 415 Lys Leu Thr Val Asp Lys Ser Arg Trp Gln Gln Gly Asn Val Phe Ser 420 425 430 Cys Ser Val Met His Glu Ala Leu His Asn His Tyr Thr Gln Lys Ser 435 440 445 Leu Ser Leu Ser Pro Gly Lys 450 455 <210> 5 <211> 214 <212> PRT <213> artificial sequence <220> <223> EGFR / cMet antibody LC1 <400> 5 Ala Ile Gln Leu Thr Gln Ser Pro Ser Ser Leu Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Thr Cys Arg Ala Ser Gln Asp Ile Ser Ser Ala 20 25 30 Leu Val Trp Tyr Gln Gln Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile 35 40 45 Tyr Asp Ala Ser Ser Leu Glu Ser Gly Val Pro Ser Arg Phe Ser Gly 50 55 60 Ser Glu Ser Gly Thr Asp Phe Thr Leu Thr Ile Ser Ser Leu Gln Pro 65 70 75 80 Glu Asp Phe Ala Thr Tyr Tyr Cys Gln Gln Phe Asn Ser Tyr Pro Leu 85 90 95 Thr Phe Gly Gly Gly Thr Lys Val Glu Ile Lys Arg Thr Val Ala Ala 100 105 110 Pro Ser Val Phe Ile Phe Pro Pro Ser Asp Glu Gln Leu Lys Ser Gly 115 120 125 Thr Ala Ser Val Val Cys Leu Leu Asn Asn Phe Tyr Pro Arg Glu Ala 130 135 140 Lys Val Gln Trp Lys Val Asp Asn Ala Leu Gln Ser Gly Asn Ser Gln 145 150 155 160 Glu Ser Val Thr Glu Gln Asp Ser Lys Asp Ser Thr Tyr Ser Leu Ser 165 170 175 Ser Thr Leu Thr Leu Ser Lys Ala Asp Tyr Glu Lys His Lys Val Tyr 180 185 190 Ala Cys Glu Val Thr His Gln Gly Leu Ser Ser Pro Val Thr Lys Ser 195 200 205 Phe Asn Arg Gly Glu Cys 210 <210> 6 <211> 449 <212> PRT <213> Artificial sequence <220> <223> HC2 of EGFR / cMet antibody <400> 6 Gln Val Gln Leu Val Gln Ser Gly Ala Glu Val Lys Lys Pro Gly Ala 1 5 10 15 Ser Val Lys Val Ser Cys Glu Thr Ser Gly Tyr Thr Phe Thr Ser Tyr 20 25 30 Gly Ile Ser Trp Val Arg Gln Ala Pro Gly His Gly Leu Glu Trp Met 35 40 45 Gly Trp Ile Ser Ala Tyr Asn Gly Tyr Thr Asn Tyr Ala Gln Lys Leu 50 55 60 Gln Gly Arg Val Thr Met Thr Thr Asp Thr Ser Thr Ser Thr Ala Tyr 65 70 75 80 Met Glu Leu Arg Ser Leu Arg Ser Asp Asp Thr Ala Val Tyr Tyr Cys 85 90 95 Ala Arg Asp Leu Arg Gly Thr Asn Tyr Phe Asp Tyr Trp Gly Gln Gly 100 105 110 Thr Leu Val Thr Val Ser Ser Ala Ser Thr Lys Gly Pro Ser Val Phe 115 120 125 Pro Leu Ala Pro Ser Ser Lys Ser Thr Ser Gly Gly Thr Ala Ala Leu 130 135 140 Gly Cys Leu Val Lys Asp Tyr Phe Pro Glu Pro Val Thr Val Ser Trp 145 150 155 160 Asn Ser Gly Ala Leu Thr Ser Gly Val His Thr Phe Pro Ala Val Leu 165 170 175 Gln Ser Ser Gly Leu Tyr Ser Leu Ser Ser Val Val Thr Val Pro Ser 180 185 190 Ser Ser Leu Gly Thr Gln Thr Tyr Ile Cys Asn Val Asn His Lys Pro 195 200 205 Ser Asn Thr Lys Val Asp Lys Arg Val Glu Pro Lys Ser Cys Asp Lys 210 215 220 Thr His Thr Cys Pro Pro Cys Pro Ala Pro Glu Leu Leu Gly Gly Pro 225 230 235 240 Ser Val Phe Leu Phe Pro Pro Lys Pro Lys Asp Thr Leu Met Ile Ser 245 250 255 Arg Thr Pro Glu Val Thr Cys Val Val Val Asp Val Ser His Glu Asp 260 265 270 Pro Glu Val Lys Phe Asn Trp Tyr Val Asp Gly Val Glu Val His Asn 275 280 285 Ala Lys Thr Lys Pro Arg Glu Glu Gln Tyr Asn Ser Thr Tyr Arg Val 290 295 300 Val Ser Val Leu Thr Val Leu His Gln Asp Trp Leu Asn Gly Lys Glu 305 310 315 320 Tyr Lys Cys Lys Val Ser Asn Lys Ala Leu Pro Ala Pro Ile Glu Lys 325 330 335 Thr Ile Ser Lys Ala Lys Gly Gln Pro Arg Glu Pro Gln Val Tyr Thr 340 345 350 Leu Pro Pro Ser Arg Glu Glu Met Thr Lys Asn Gln Val Ser Leu Thr 355 360 365 Cys Leu Val Lys Gly Phe Tyr Pro Ser Asp Ile Ala Val Glu Trp Glu 370 375 380 Ser Asn Gly Gln Pro Glu Asn Asn Tyr Lys Thr Thr Pro Pro Val Leu 385 390 395 400 Asp Ser Asp Gly Ser Phe Phe Leu Tyr Ser Arg Leu Thr Val Asp Lys 405 410 415 Ser Arg Trp Gln Gln Gly Asn Val Phe Ser Cys Ser Val Met His Glu 420 425 430 Ala Leu His Asn His Tyr Thr Gln Lys Ser Leu Ser Leu Ser Pro Gly 435 440 445 Lys <210> 7 <211> 214 <212> PRT <213> Artificial sequence <220> <223> LC2 of EGFR / cMet antibody <400> 7 Asp Ile Gln Met Thr Gln Ser Pro Ser Ser Val Ser Ala Ser Val Gly 1 5 10 15 Asp Arg Val Thr Ile Thr Cys Arg Ala Ser Gln Gly Ile Ser Asn Trp 20 25 30 Leu Ala Trp Phe Gln His Lys Pro Gly Lys Ala Pro Lys Leu Leu Ile 35 40 45 Tyr Ala Ala Ser Ser Leu Leu Ser Gly Val Pro Ser Arg Phe Ser Gly 50 55 60 Ser Gly Ser Gly Thr Asp Phe Thr Leu Thr Ile Ser Ser Leu Gln Pro 65 70 75 80 Glu Asp Phe Ala Thr Tyr Tyr Cys Gln Gln Ala Asn Ser Phe Pro Ile 85 90 95 Thr Phe Gly Gln Gly Thr Arg Leu Glu Ile Lys Arg Thr Val Ala Ala 100 105 110 Pro Ser Val Phe Ile Phe Pro Pro Ser Asp Glu Gln Leu Lys Ser Gly 115 120 125 Thr Ala Ser Val Val Cys Leu Leu Asn Asn Phe Tyr Pro Arg Glu Ala 130 135 140 Lys Val Gln Trp Lys Val Asp Asn Ala Leu Gln Ser Gly Asn Ser Gln 145 150 155 160 Glu Ser Val Thr Glu Gln Asp Ser Lys Asp Ser Thr Tyr Ser Leu Ser 165 170 175 Ser Thr Leu Thr Leu Ser Lys Ala Asp Tyr Glu Lys His Lys Val Tyr 180 185 190 Ala Cys Glu Val Thr His Gln Gly Leu Ser Ser Pro Val Thr Lys Ser 195 200 205 Phe Asn Arg Gly Glu Cys 210

Claims

1. A method of producing a heterodimeric antibody, comprising: a) providing a first homodimeric antibody comprising a first Fc region of an immunoglobulin comprising a first CH3 region and a second homodimeric antibody comprising a second Fc region of an immunoglobulin comprising a second CH3 region, wherein the amino acid sequence of the first CH3 region is different from the second CH3 region and such that the heterodimeric interaction between the first CH3 region and the second CH3 region is stronger than the homodimeric interaction between the first CH3 regions or the homodimeric interaction between the second CH3 regions; b) combining the first homodimeric antibody and the second homodimeric antibody into a mixture; wherein the mixture comprises a sodium acetate buffer; c) incubating the mixture in the presence of a reducing agent; and d) removing the reducing agent, thereby producing the heterodimeric antibody, wherein the percentage of dissolved oxygen is controlled to be 15% or less in step d) or in both steps c) and d).

2. The method of claim 1, wherein the percentage of dissolved oxygen is controlled to be 10% or less in step d).

3. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 9% or less in step d).

4. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 8% or less in step d).

5. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 7% or less in step d).

6. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 6% or less in step d).

7. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 5% or less in step d).

8. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 4% or less in step d).

9. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 3% or less in step d).

10. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 2% or less in step d).

11. The method of claim 1, wherein the percentage of dissolved oxygen of the mixture is 1% or less in step d).

12. The method of claim 1, wherein the percentage of dissolved oxygen is controlled by an overlay of nitrogen.

13. The method of claim 1, wherein the first homodimeric antibody and the second homodimeric antibody are combined into the mixture in a molar ratio of 1 : 1 to 1 :2 in step b).

14. The method of claim 1, wherein the first homodimeric antibody and the second homodimeric antibody are combined into the mixture in a molar ratio of 1.05: 1 in step b).

15. The method of claim 1, wherein the total concentration of immunoglobulin in the mixture is 8 g / L to 13 g / L.

16. The method of claim 1, wherein the total concentration of immunoglobulin in the mixture is 10.5 g / L.

17. The method of claim 1, wherein the mixture is incubated in the presence of the reducing agent for 10 minutes or more.

18. The method of claim 1, wherein the mixture is incubated in the presence of the reducing agent for 10 minutes to 24 hours.

19. The method of claim 1, wherein the reducing agent is 2-mercaptoethylamine (2-MEA), a chemical derivative of 2-MEA, L-cysteine, or D-cysteine.

20. The method of claim 19, wherein the concentration of 2-MEA in the mixture is 20 mM to 40 mM.

21. The method of claim 20, wherein the concentration of 2-MEA in the mixture is 35 mM.

22. The method of claim 1, wherein the buffer further comprises NaCl.

23. The method of claim 22, wherein the buffer comprises 100 mM sodium acetate and 30 mM NaCl.

24. The method of claim 23, wherein the pH of the buffer is 7.

3.

25. The method of claim 1, comprising the step of removing the reducing agent from the mixture.

26. The method of claim 25, wherein the reducing agent is removed by filtration.

27. The method of claim 26, wherein the filtration is diafiltration.

28. The method of claim 1, wherein the first homodimeric antibody and the second homodimeric antibody are of IgGl, IgG2, or IgG4 isotype.

29. The method of claim 28, wherein the first CH3 domain and the second CH3 domain comprise the following mutations when compared to a wild-type IgGl of SEQ ID NO: 1, a wild-type IgG2 of SEQ ID NO: 2, or a wild-type IgG4 of SEQ ID NO: 3: F405L / K409R, wild-type / F405L_R409K, T350I_K370T_F405L / K409R, K370W / K409R, D399AFGHILMNRSTVWY / K409R, T366ADEFGHILMQVY / K409R, L368ADEGHNRSTVQ / K409AGRH, D399FHKRQ / K409AGRH, F405IKLSTVW / K409AGRH, Y407LWQ / K409AGRH, T366Y / F405A, T366W / F405W, F405W / Y407A, T394W / Y407T, T394S / Y407A, T366W / T394S, F405W / T394S, T366W / T366S_L368A_Y407V, L351Y_F405A_Y407V / T394W, T366I_K392M_T394W / F405A_Y407V, T366L_K392M_T394W / F405A_Y407V, L351Y_Y407A / T366A_K409F, L351Y_Y407A / T366V_K409F, Y407A / T366A_K409F, T350V_L351Y_F405A_Y407V / T350V_T366L_K392L_T394W, K409D / D399K, K409E / D399R, K409D_K360D / D399K_E356K, K409D_K360D / D399E_E356K, K409D_K370D / D399K_E357K, K409D_K370D / D399E_E357K, K409D_K392D / D399K_E356K_E357K, or K409D_K392D / D399E_E356K_E357K.

30. The method of claim 29, wherein the first Fc region and / or the second Fc region comprises one or more mutations that modulate binding of the first Fc region and / or the second Fc region to an Fc gamma receptor (FcyR), FcRn, or Protein A when compared to a wild-type IgGl of SEQ ID NO: 1, a wild-type IgG2 of SEQ ID NO: 2, or a wild-type IgG4 of SEQ ID NO:

3.

31. The method of claim 30, wherein the FcyR is FcyRI, FcyRIIa, FcyRIIb or FcyRIII.

32. The method of claim 30, wherein the one or more substitutions that modulate the binding of the first Fc region and / or the second Fc region to Fcy is L234A_L235A, F234A_L235A, S228P_F234A_L235A, S228P_L234A_L235A, V234A_G237A_P238S_H268A_V309L_A330S_P331S, V234A_G237A, H268Q_V309L_A330S_P331S, S267E_L328F, L234F_L235E_D265A, L234A_L235A_G237A_P238S_H268A_A330S_P331S, or S228P_F234A_L235A_G237A_P238S.

33. The method of claim 30, wherein the one or more substitutions that modulate the binding of the first Fc region and / or the second Fc region to FcRn is M428L_N434S, M252Y_S254T_T256E, T250Q_M428L, N434A and T307A_E380A_N434A, H435A, P257I_N434H, D376V_N434H, M252Y_S254T_T256E_H433K_N434F, T308P_N434A, or H435R.

34. The method of claim 30, wherein the one or more substitutions that modulate the binding of the first Fc region and / or the second Fc region to Protein A is Q311R, Q311K, T307P_L309Q, T307P_V309Q, T307P_L309Q_Q311R, T307P_V309Q_Q311R, H435R or H435R_Y436F.

35. The method of claim 1, wherein the step of producing the heterodimeric antibody is performed under GMP-compliant conditions.

36. The method of claim 1, wherein the step of producing the heterodimeric antibody is performed during the manufacture of a medicament comprising the heterodimeric antibody.

37. The method of claim 1, wherein the heterodimeric antibody is a bispecific antibody.

38. The method of claim 37, wherein the bispecific antibody binds CD3, CD123, BCMA, EGFR, or c-Met, or any combination thereof.

39. The method of claim 1, wherein the mass ratio of total immunoglobulin in the mixture to total reducing agent in the mixture is 1.0 to 5.

0.

40. The method of claim 39, wherein the mass ratio is 1.4 to 3.

8.

41. The method of claim 39, wherein the mass ratio is 3.3 to 4.

4. ​ 42. The method of claim 37, wherein the bispecific antibody binds EGFR and c-Met.

43. The method of claim 37, wherein the bispecific antibody comprises a first heavy chain (HC1) of SEQ ID NO: 4, a first light chain (LC1) of SEQ ID NO: 5, a second heavy chain (HC2) of SEQ ID NO: 6, and a second light chain (LC2) of SEQ ID NO:

7.

44. The method of claim 37, wherein the bispecific antibody binds CD3.

Images