HK1203079B - Single-chain antibodies and other heteromultimers - Google Patents

Description

Single-chain antibodies and other heteropolymers

This application is related to and claims the benefit of priority from U.S. Provisional Application Serial No. 61/597,486, filed February 10, 2012, the disclosure of which is incorporated herein by reference in its entirety.

Sequence Listing

This application contains a sequence listing in ASCII format submitted via EFS-Web, which is hereby incorporated by reference in its entirety. The ASCII copy was created on January 28, 2013, is named P4733R1WO_PCTSequenceListing.txt, and is 7501 bytes in size.

Field of the Invention

The present invention relates to novel engineered proteins and protein complexes, including heteromultimers with single or multispecificity (e.g., single-chain antibodies, multi-chain antibodies, and immunoadhesin-antibody complexes), and methods for their construction and production. The present invention also relates to novel applications of techniques for obtaining single or multispecific heteromultimers. The heteromultimers produced by the methods provided herein can be used as therapeutic agents for any disease or pathological condition, as well as any other use in which the use of antibodies is advantageous.

Background of the Invention

Developing technologies for producing antibodies or other heteromultimers with diverse binding characteristics (e.g., monospecific or multispecific) that are commercially and therapeutically viable and scalable has been difficult. Many approaches have been attempted, but almost all have significant disadvantages, such as poor solubility or inability to express in mammalian cells, low yields of heterodimer formation, technical difficulties in production or immunogenicity, short half-life in vivo, or instability (e.g., Hollinger et al., (1993) PNAS 90:6444-6448; US 5,932,448; US 6,833,441; US 5,591,828; US 7,129,330; US 7,507,796; Fischer et al., (2007) Pathobiology 74:3-14; Booy (2006) Arch. Immunol. Ther. Exp. 54:85-101; Cao et al., (2003) 55:171-197 and Marvin et al., (2006) Current Opinion in Drug Discovery & Development 9(2):184-193). Therefore, there is a need for improved techniques and processes for producing antibodies or other heteromultimers with different binding characteristics.

SUMMARY OF THE INVENTION

The present invention provides heteromultimers (e.g., new single-chain antibodies (scAbs), multi-chain antibodies (mcAbs), and immunoadhesin-antibody complexes) and methods for creating, producing, and using the heteromultimers. In one aspect, the present invention is characterized by a heteromultimer single-chain antibody comprising a heterodimerization (HD) tether connecting a first heavy chain variable (VH) domain and a second VH domain, wherein the heteromultimer comprises one or more heavy chain constant (CH) domains selected from a first CH2 domain, a first CH3 domain, a second CH2 domain, and a second CH3 domain. In one embodiment, the heteromultimer comprises at least one pair of heavy chain constant domains. In another embodiment, the heteromultimer may comprise a hinge domain between the VH and CH2 domains on one or two heavy chains. In another embodiment, the heteromultimer comprises a first and/or second CH1 domain. The one or two CH1 domains are located at the C-terminus of one or two VH domains and the N-terminus of one or two hinge domains, or, in the absence of a hinge domain, at the N-terminus of one or two CH2 domains. In particular embodiments, the heteromultimers may also comprise one or two light chain variable (VL) domains connected to the N-terminus of the first and/or second VH domains by one or two CLH tethers (cognate LC-HC tethers). In some embodiments, the heteromultimers further comprise one or two light chain constant (CL) domains, each located at the C-terminus of one or two VL domains and immediately adjacent to the N-terminus of one or two CLH tethers.

In another aspect, the invention features a heteromultimeric single-chain antibody comprising a single polypeptide having the following domains positioned relative to each other in the N-terminal to C-terminal direction as follows: VL ₁ -CL ₁ -CLH Tether ₁ -VH ₁ -CH1 ₁ -Hinge ₁ -CH2 ₁ -CH3 ₁ -HD Tether-VL ₂ -CL ₂ -CLH Tether ₂ -VH ₂ -CH1 ₂ -Hinge ₂ -CH2 ₂ -CH3 ₂ .

In another aspect, the invention features a multi-chain antibody heteromultimer comprising three polypeptide chains, wherein the first and second polypeptide chains are identical and each form a light chain (LC), and the third polypeptide chain forms a first heavy chain (HC) and a second HC. The first and second polypeptide chains each comprise a VL and CL domain. The third polypeptide chain comprises two VH domains, an HD tether, one or two hinge domains, and one or more heavy chain constant domains selected from a first CH1 domain, a first CH2 domain, a first CH3 domain, a second CH1 domain, a second CH2 domain, and a second CH3 domain, wherein the components of the second polypeptide chain are arranged relative to each other in the following order from N-terminus to C-terminus: VH ₁ - optional CH1 ₁ - optional hinge ₁ - optional CH2 ₁ - optional CH3 ₁ - HD tether - VH ₂ - optional CH1 ₂ - optional hinge ₂ - optional CH2 ₂ - optional CH3 ₂ .

In another aspect, the present invention features a multi-chain antibody heteromultimer comprising two polypeptide chains, wherein a first polypeptide chain forms a first light chain (LC), and a second polypeptide chain forms a first heavy chain (HC), a second LC, and a second HC. The first polypeptide chain comprises a first VL and a CL domain. The second polypeptide chain comprises two VH domains, an HD tether, a second VL domain, a second CL domain, a CLH tether, one or two hinge domains, and one or more heavy chain constant domains selected from a first CH1 domain, a first CH2 domain, a first CH3 domain, a second CH1 domain, a second CH2 domain, and a second CH3 domain, wherein the components of the second polypeptide chain are arranged relative to each other in the N-terminal to C-terminal direction as follows: VH ₁ - optional CH1 ₁ - optional hinge ₁ - optional CH2 ₁ - optional CH3 ₁ - HD tether - VL ₂ - CL ₂ - CLH tether - VH ₂ - optional CH1 ₂ - optional hinge ₂ - optional CH2 ₂ - optional CH3 ₂ .

In another aspect, the invention features a multi-chain antibody heteromultimer comprising two polypeptide chains, wherein a first polypeptide chain forms a first LC, a first HC, and a second HC, and a second polypeptide chain forms a second LC. The first polypeptide chain comprises two VH domains, an HD tether, a first VL domain, a first CL domain, a CLH tether, one or two hinge domains, and one or more heavy chain constant domains selected from a first CH1 domain, a first CH2 domain, a first CH3 domain, a second CH1 domain, a second CH2 domain, and a second CH3 domain, wherein the components of the second polypeptide chain are positioned relative to each other in the N-terminal to C-terminal direction as follows: VL ₁ -CL ₁ -CLH tether-VH ₁ -optional CH1 ₁ -optional hinge ₁ -optional CH2 ₁ -optional CH3 _{1 -HD} tether-VH 2 -optional CH1 ₂ -optional hinge ₂ -optional CH2 ₂ -optional CH3 _2. The second polypeptide chain comprises a second VL and CL domains.

In another aspect, the invention features a heteromultimer comprising two polypeptide chains, wherein the first polypeptide chain comprises an immunoadhesin comprising an adhesin and one or more heavy chain constant domains (e.g., CH2 ₁ and/or CH3 ₁ ), the second polypeptide chain forms a half-antibody comprising a VH domain and one or more heavy chain constant domains (e.g., CH1, CH2 ₂ and/or CH3 ₂ ), and the first and second polypeptide chains are connected to each other by an HD tether to form a single polypeptide chain. The components of the heteromultimer are positioned relative to each other in the N-terminal to C-terminal direction as follows: adhesin-optional CH2 ₁ -optional CH3 ₁ -HD tether-VH-optional CH1-optional CH2 ₂ -optional CH3 _2. The HD tether facilitates the interaction between the one or more constant domains of the immunoadhesin and the half-antibody. In one embodiment, the CLH tether promotes interaction between the light chain and heavy chain components of the half antibody to obtain a heteromultimer having components positioned relative to each other as follows from the N-terminus to the C-terminus: adhesin-optional CH2 ₁ -optional CH3 ₁ -HD tether-VH-CL-CLH tether-VH-optional CH1-optional CH2 ₂ -optional CH3 _2. In another embodiment, the VL and CL domains of the half antibody light chain are provided by a second polypeptide that associates with the heavy chain of the half antibody of the first polypeptide chain to form a light chain-heavy chain cognate pair. In another embodiment, the immunoadhesin portion of the heteromultimer may include an amino acid spacer between its adhesin and heavy chain constant domain components. In one embodiment, the spacer includes glycine (G) and serine (S) residues, such as GGS repeats. In another embodiment, the spacer is between 10-80 amino acids long, such as between 20-40 residues.

The heteromultimers of the present invention may comprise an HD tether between 15 and 100 amino acids in length. In particular embodiments, the HD tether is between 30 and 39 amino acids in length, such as 30, 31, 32, 33, 34, 35, 36, 37, 38, or 39 amino acids in length. In one embodiment, the tether comprises glycine (G) and serine (S) residues. In another embodiment, the tether comprises GGS repeats. In a preferred embodiment, the tether comprises 8 to 9 GGS repeats (SEQ ID NO: 19).

The heteromultimers of the present invention may also comprise one or more CLH tethers. In one embodiment, the one or more CLH tethers are each between 10-80 amino acids in length. In a particular embodiment, the one or more CLH tethers are each between 20-40 amino acids in length. In one embodiment, the tether comprises glycine (G) and serine (S) residues. In another embodiment, the tether comprises GGS repeats.

In another embodiment, one or more of the HD and CLH tethers of the present invention are cleavable by one or more of the following endopeptidases: furin, urokinase, thrombin, tissue plasminogen activator (tPa), genenase, Lys-C, Arg-C, Asp-N, Glu-C, Factor Xa, tobacco etch virus protease (TEV), enterokinase, human rhinovirus C3 protease (HRV C3), or kininogenase. In a preferred embodiment, at least one tether is cleavable by furin. In another embodiment, at least one of the one or more tethers of the present invention is cleavable at two sites located at or near the N-terminus and C-terminus of the tether. Preferably, for HD tethers, one of the two cleavage sites is a furin cleavage site and the other is a Lys-C cleavage site. Preferably, for CLH tethers, the cleavage sites at both the N-terminus and the C-terminus of one or more CLH tethers are cleavable by furin. In certain embodiments, the furin cleavage site comprises the amino acid sequence RKRKRR (SEQ ID NO:9). In certain other embodiments, the furin cleavage site comprises the amino acid sequence RHRQPR (SEQ ID NO: 10). In one embodiment, endopeptidase cleavage occurs in situ. In certain embodiments, the endopeptidase is recombinantly expressed in a host cell. In another embodiment, endopeptidase cleavage occurs after purification and after addition of the endopeptidase.

The heteropolymers of the present invention may have one or more (e.g., two) CLH tethers, linkers, each comprising one or more cleavage sites for one or more of the following specific endopeptidases: carboxypeptidase A, carboxypeptidase B, plasma carboxypeptidase B (also known as carboxypeptidase U or coagulation-activated fibrinolysis inhibitor (TAFI)), carboxypeptidase D, carboxypeptidase E (also known as enkephalin convertase or carboxypeptidase H), carboxypeptidase M, carboxypeptidase N, or carboxypeptidase Z. In a preferred embodiment, the heteropolymers of the present invention can be cleaved by a carboxypeptidase B exopeptidase. In one embodiment, the exopeptidase cleavage occurs in situ. In certain embodiments, the exopeptidase is recombinantly expressed in a host cell. In another embodiment, the exopeptidase cleavage occurs after purification and after the addition of the exopeptidase. As used herein, carboxypeptidase B may refer to the class of carboxypeptidases or a specific carboxypeptidase. As a class, carboxypeptidase B includes all specific carboxypeptidases except carboxypeptidase A. As a specific carboxypeptidase, carboxypeptidase B is also known as carboxypeptidase U or TAFI. Those skilled in the art can readily distinguish between carboxypeptidase B as a class and carboxypeptidase B as a specific exopeptidase based on the context in which the term is used.

In another embodiment, the heteromultimers of the present invention may comprise one or more hinge domains, including but not limited to a hinge domain comprising Glu216 to Pro230 of human IgG1. In some embodiments, one or both hinge domains comprise a mutation that removes the Lys-C endopeptidase cleavage site. In one example, the mutation that removes the Lys-C endopeptidase cleavage site is a K222A substitution (EU numbering system).

The heteromultimers of the present invention can be monospecific. In one embodiment, the monospecific heteromultimers of the present invention comprise two half antibodies that bind to the same epitope target but can bind with different affinities. In another embodiment, the monospecific heteromultimers of the present invention comprise half antibodies associated with an immunoadhesin, each specific for the same binding partner or epitope.

The heteromultimers of the present invention can be bispecific or multispecific. In one embodiment, the heteromultimers are capable of binding to at least two antigens. In another embodiment, the heteromultimers are capable of binding to at least two epitopes on the same antigen. In another embodiment, the bispecific or multispecific heteromultimers of the present invention comprise half antibodies associated with immunoadhesins, each of which is specific for a different binding partner or epitope.

In another embodiment, the heteromultimers of the invention comprise a constant region conjugated to a cytotoxic agent.

In another embodiment, the heteromultimer may comprise two heavy chain constant domains (e.g., two CH3 domains) having a protrusion or a cavity, wherein the protrusion or cavity of one heavy chain constant domain (e.g., CH3 ₁ domain) can be respectively placed in the cavity or protrusion of the second heavy chain constant domain (e.g., CH3 ₂ domain). Preferably, the two constant domains meet at an interface comprising the protrusion and the cavity. In another embodiment, the heteromultimer may comprise at least one light chain constant domain and a heavy chain constant domain interface (e.g., CL/CH1 interface), wherein the light chain constant domain (e.g., CL domain) and the heavy chain constant domain (e.g., CH1 domain) interact at least in part through protrusion-cavity interactions.

In another embodiment, the heteromultimers of the present invention comprise a CH2 domain mutation located in their CH2 ₁ or CH2 ₂ , wherein the mutation results in an antibody with altered effector function. In a preferred embodiment, the CH2 domain mutation is an N297 mutation. In certain embodiments, the N297 mutation is an N297A mutation. In certain other embodiments, the CH2 domain further comprises a D256A mutation.

In other aspects, the invention features a method for producing heteromultimers. In another aspect, the invention features a polynucleotide encoding a heteromultimer of the invention. In other aspects, the invention features a vector comprising a polynucleotide of the invention and a host cell comprising the vector. In one embodiment, the host cell is a mammalian cell. In a preferred embodiment, the mammalian cell is a CHO cell. In one embodiment, the host cell is a prokaryotic cell. In other embodiments, the prokaryotic cell is an Escherichia coli (E. coli) cell. In other aspects, the invention features a method for producing heteromultimers, the method comprising culturing a host cell comprising a vector having a polynucleotide encoding a heteromultimer in a culture medium. Preferably, the heteromultimer is recovered from the host cell or the culture medium of the host cell.

In other aspects, the present invention provides single-chain antibodies comprising a single polypeptide comprising the following domains positioned relative to each other from the N-terminus to the C-terminus: VL ₁ -CL ₁ -CLH tether ₁ -VH ₁ -CH1 ₁ -hinge ₁ -CH2 ₁ -CH3 ₁ -HD tether -VL ₂ -CL ₂ -CLH tether ₂ -VH ₂ -CH1 ₂ -hinge ₂ -CH2 ₂ -CH3 ₂ , wherein CLH tether 1, CLH tether 2, and HD tether each comprise an amino acid sequence cleavable by a furin endopeptidase. In certain embodiments of the present invention, the furin cleavable sequence comprises the amino acid sequence RKRKRR (SEQ ID NO: 9), while in other embodiments, the furin cleavable sequence comprises the amino acid sequence RHRQPR (SEQ ID NO: 10). In related aspects, the present invention provides polynucleotide molecules encoding the single-chain antibodies of the present invention, vectors comprising the polynucleotides, and host cells comprising the vectors. In certain embodiments, the host cell is a mammalian cell, including but not limited to a CHO cell. In certain other embodiments, the host cell is a prokaryotic cell, including but not limited to an Escherichia coli cell. In other related aspects, the present invention provides a method for producing a single-chain antibody, the method comprising culturing a host cell containing a vector in a culture medium. In certain embodiments, the method further comprises recovering the single-chain antibody from the host cell or the culture medium.

Other features and advantages of the invention will be apparent from the following detailed description, the drawings, and the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram illustrating the structure of an exemplary heteromultimeric single-chain antibody comprising three cleavable tethers. The endopeptidase cleavage site is indicated by a triangle. The optional K222A mutation is also illustrated.

Figure 2 is a schematic diagram illustrating the arrangement of the LC, HC, tethers, and cleavage sites of exemplary heteromultimeric single-chain antibodies. The cleavage sites are exemplified by RKRKRRG(GGS) _6GRSRKRR (SEQ ID NO: 14) and (GGS) _(8-10) RSRKRR (SEQ ID NO: 15-17). The furin recognition site is denoted as RXRXRR (SEQ ID NO: 8).

Figure 3A is an example of a heteromultimeric single-chain antibody after furin cleavage. The residues in brackets (RKRKRR (SEQ ID NO: 9) and RKRKR (SEQ ID NO: 18)) indicate residues that can be removed by endogenous exopeptidases prior to purification on a Protein A column, resulting in a C-terminal defect. The optional K222A mutation is also illustrated.

Figure 3B is an example of a heteromultimeric single-chain antibody after treatment with furin, Lys-C, and an exopeptidase, such as carboxypeptidase B. The optional K222A mutation is also illustrated.

Figure 4 is an exemplary conjugated heteromultimeric single-chain antibody prior to cleavage to remove the tether. Optional conjugated moieties, such as toxins, antibiotics, etc., and the optional K222A mutation are also illustrated.

Figure 5A is a schematic diagram illustrating the structure of an exemplary heteromultimeric multi-chain antibody containing a cleavable tether. The two unlinked LCs can be expressed independently of the polypeptide containing the tether. The unlinked LCs can be expressed in the same or different cells as the linked heavy chain. The unlinked LCs can be expressed in the same or different plasmids. Optional conjugated moieties, such as toxins, antibiotics, etc., and the optional K222A mutation are also illustrated.

Figure 5B is a schematic diagram illustrating the structure of an exemplary heteromultimeric multi-chain antibody containing two cleavable tethers. The HD tether indirectly connects the first HC to the second HC via the attached LC. The unattached LC can be expressed independently of the polypeptide containing the tether. The unattached LC can be expressed in the same or different cells as the attached heavy chain. The unattached LC can be expressed in the same or different plasmids. The optional K222A mutation is also illustrated.

Figure 5C is a schematic diagram illustrating the structure of an exemplary heteromultimeric multi-chain antibody containing two cleavable tethers. The HD tether directly connects the first HC and the second HC. The unlinked LC can be expressed independently of the polypeptide containing the tether. The unlinked LC can be expressed in the same or different cells as the linked heavy chain. The unlinked LC can be expressed in the same or different plasmids. The optional K222A mutation is also illustrated.

Figures 6A-6D are graphs illustrating the results of Octet analysis. (A) Combination of curves in Figures B-D. (B) Simultaneous binding of an exemplary heteromultimeric single-chain antibody to both Antigen 1 and Antigen 2. (C and D) Antibody 1 and Antibody 2 do not cross-react with each other's antigens but bind to their respective antigens. The x-axis is time in seconds. The y-axis is relative absorbance. Reference is made to Example 3.

Detailed Description of the Invention

When generating monospecific or multispecific (e.g., bispecific) antibodies or other heteromultimers with multiple polypeptide chains having different binding properties, undesirable heavy chain homodimerization often occurs. We have found that this common problem can be avoided by generating single-chain monospecific or multispecific heteromultimers whose assembly is guided by one or more tethers. Without being limited to theory, we believe that HD tethers enable different Fc heavy chain components to be bound with high precision and efficiency, resulting in functional heteromultimers comprising two half molecules (e.g., two half antibodies) that bind to the same target or different targets with the same or different affinities. Heteromultimers with linked heavy chain components can additionally comprise different light chain components, resulting in functional single-chain monospecific or multispecific heteromultimers (e.g., antibodies) with complete complementary heavy and light chains. Additional tethers according to the present invention can be used to connect the light and heavy chains of the heteromultimer, thereby facilitating the proper association of each light chain with its cognate heavy chain.

The use of the methods for making heteromultimers described herein allows for the production of substantially homogeneous monospecific or multispecific heteromultimer populations generated from single or multiple polypeptide sequences. The heteromultimers produced using the methods described herein can be used to identify more than one target in a pathogenic pathway, or to co-localize a specific target (e.g., a tumor cell) and an active agent (e.g., a T cell) directed against the target. In addition, the heteromultimers described herein are advantageous because they eliminate the need for combination therapies targeting two antigens and the risks associated with providing two or more therapeutic agents to a subject.

I. Definition

The term "antibody" as used herein is used in the broadest sense to refer to any immunoglobulin (Ig) molecule comprising two heavy chains and two light chains, and any fragment, mutant, variant or derivative thereof as long as it exhibits the desired biological activity (e.g., epitope binding activity). Examples of antibodies include monoclonal antibodies, polyclonal antibodies, bispecific antibodies, multispecific antibodies, and antibody fragments.

When referring to the residues in the variable domain (approximately residues 1-107 in the light chain and residues 1-113 in the heavy chain), the Kabat numbering system is generally used (e.g., Kabat et al., Sequences of Immunological Interest. 5th Edition. Public Health Service, National Institutes of Health, Bethesda, Md. (1991)). When referring to the residues in the immunoglobulin heavy chain constant region, the "EU numbering system" or "EU index" is generally used (e.g., the EU index reported in Kabat et al., supra). The "EU index as in Kabat" refers to the residue numbering of human IgG1 EU antibodies. Unless otherwise indicated herein, reference to the residue numbers in the antibody variable domain refers to the residues numbered by the Kabat numbering system. Unless otherwise indicated herein, reference to the residue numbers in the antibody heavy chain constant domain refers to the residues numbered by the EU numbering system.

The naturally occurring basic four-chain antibody unit is a heterotetrameric glycoprotein composed of two identical light chains (LC) and two identical heavy chains (HC). (IgM antibodies are composed of five basic heterotetrameric units plus an additional polypeptide chain called a J chain and therefore contain 10 antigen-binding sites, while secretory IgA antibodies can polymerize to form multivalent aggregates containing 2-5 basic four-chain units and a J chain.) In the case of IgG, the four-chain unit is typically approximately 150,000 daltons. Each LC is linked to the HC by a single covalent disulfide bond, while the two HCs are linked to each other by one or more disulfide bonds, depending on the HC isotype. Each HC and LC also have regularly spaced intrachain disulfide bridges. Each HC has a variable domain (VH) at its N-terminus, followed by three constant domains (CH1, CH2, CH3) for each of the α and γ chains, and four Cj domains for the μ and ε isotypes. Each LC has a variable domain (VL) at its N-terminus, followed by a constant domain (CL) at its other end. The VL is aligned with the VH, and the CL is aligned with the first constant domain (CH1) of the heavy chain. CH1 can be connected to the second constant domain (CH2) of the heavy chain by a hinge region. It is believed that specific amino acid residues form an interface between the light chain and heavy chain variable domains. The pairing of VH and VL together forms a single antigen binding site. For the structure and properties of different classes of antibodies, see, for example, Basic and Clinical Immunology, 8th edition, Daniel P. Stites, Abba I. Terr and Tristram G. Parslow (eds.), Appleton & Lange, Norwalk, CT, 1994, page 71 and Chapter 6.

The "hinge region" is generally defined as the stretch from Glu216 to Pro230 of human IgG1 (Burton, Molec. Immunol. 22: 161-206 (1985)). Hinge regions of other IgG isotypes can be aligned with the IgG1 sequence by placing the first and last cysteine residues that form the inter-heavy chain SS bond at the same position.

The "lower hinge region" of an Fc region is generally defined as the segment of residues immediately adjacent to the C-terminus of the hinge region, residues 233 to 239 of the Fc region. Prior to the present invention, FcγR binding was generally attributed to amino acid residues within the lower hinge region of the IgG Fc region.

The "CH2 domain" of the human IgG Fc region typically extends from approximately residue 231 to approximately residue 340 of IgG. The CH2 domain is unique in that it is not tightly paired with another domain. Instead, two N-linked branched carbohydrate chains are inserted between the two CH2 domains of an intact native IgG molecule. It is hypothesized that the carbohydrates provide an alternative to domain-domain pairing and help stabilize the CH2 domain. Burton, Molec. Immunol. 22:161-206 (1985).

A "CH3 domain" comprises a segment of the residues C-terminal to the CH2 domain of an Fc region (ie, from about amino acid residue 341 to about amino acid residue 447 of IgG).

Light chains (LC) from any vertebrate species can be assigned to two clearly distinct types, designated kappa and lambda, based on the amino acid sequence of their constant domains. Immunoglobulins can be divided into different classes, or isotypes, based on the amino acid sequence of their heavy chain constant domains (CH). There are five classes of immunoglobulins: IgA, IgD, IgE, IgG, and IgM, which have heavy chains that are α, δ, γ, ε, and μ, respectively. The γ and α classes are further divided into subclasses based on relatively minor differences in CH sequence and function; for example, humans express the following subclasses: IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2.

The term "variable" refers to the fact that the sequences of certain segments of the variable domain vary greatly between antibodies. The V domain mediates antigen binding and defines the specificity of a particular antibody for its specific antigen. However, variability is not evenly distributed across the 110 amino acid span of the variable domain. Instead, the V region consists of relatively constant fragments of 15-30 amino acids called framework regions (FRs), which are separated by shorter, extreme variations called "hypervariable regions" of 9-12 amino acids each. Native heavy and light chains each contain four FRs, primarily in a β-sheet conformation, connected by three hypervariable regions that form loops, connecting, in some cases, parts of the β-sheet structure. The hypervariable regions in each chain are kept in close proximity by the FRs and contribute to the formation of the antibody's antigen-binding site with the hypervariable regions from the other chains (see Kabat et al., Sequences of Proteins of Immunological Interest, 5th ed. Public Health Service, National Institutes of Health, Bethesda, MD, 1991). The constant region is not involved directly in binding an antibody to an antigen, but exhibits various effector functions, such as participation of the antibody in antibody-dependent cellular cytotoxicity (ADCC).

As used herein, the term "hypervariable region" or "HVR" refers to each region of an antibody variable domain whose sequence is hypervariable and/or forms a structurally determined loop ("hypervariable loop"). Typically, a natural four-chain antibody comprises six HVRs; three in VH (H1, H2, H3) and three in VL (L1, L2, L3). HVRs typically comprise amino acid residues from hypervariable loops and/or from "complementarity determining regions" (CDRs), which have the highest sequence variability and/or participate in antigen recognition. As used herein, HVRs comprise any number of residues located at positions 24-36 (for L1), 46-56 (for L2), 89-97 (for L3), 26-35B (for H1), 47-65 (for H2), and 93-102 (for H3). Thus, the HVRs comprise residues in the positions previously described in (A), (B), and (C): (A) 24-34 (L1), 50-52 (L2), 91-96 (L3), 26-32 (H1), 53-55 (H2), and 96-101 (H3) (Chothia and Lesk, J. Mol. Biol. 196:901-917 (1987)); (B) 24-34 of L1, 50-56 of L2, 89-97 of L3, 31-35B of H1, 50-65 of H2, and 95-102 of H3 (Kabat et al., Sequences of Proteins of Immunological Interest, 5th ed. Public Health Service, National Institutes of Health). of Health, Bethesda, MD (1991)); (C) 30-36 (L1), 46-55 (L2), 89-96 (L3), 30-35 (H1), 47-58 (H2), 93-100a-j (H3) (MacCallum et al., J. Mol. Biol. 262:732-745 (1996)). Unless otherwise indicated, HVR residues and other residues in the variable domain (e.g., FR residues) are numbered herein according to Kabat et al., supra.

A variety of HVR descriptions are in use and are included herein. The Kabat complementarity determining region (CDR) is based on sequence variability and is the most commonly used (Kabat et al., Sequences of Proteins of Immunological Interest, 5th Edition. Public Health Service, National Institutes of Health, Bethesda, MD, 1991). Chothia refers to the position of the structural loop (Chothia and Lesk J. Mol. Biol. 196: 901-917 (1987)). AbM HVR represents a compromise between Kabat HVR and Chothia structural loops and is used in Oxford Molecular's AbM antibody modeling software. The "contact" HVR is based on analysis of existing complex crystal structures. The residues from each of the above HVRs are recorded in the table below.

An HVR may comprise an "extended HVR" as follows: 24-36 or 24-34 (L1), 46-56 or 50-56 (L2), and 89-97 or 89-96 (L3) in VL, and 26-35 (H1), 50-65 or 49-65 (H2), and 93-102, 94-102, or 95-102 (H3) in VH. In each of the above definitions, the variable domain residues are numbered according to Kabat et al., supra.

"Framework regions" (FRs) are the variable domain residues that are not CDR residues. Each variable domain typically has four FRs, referred to as FR1, FR2, FR3, and FR4. If the CDRs are defined according to Kabat, the light chain FR residues are located at approximately residues 1-23 (LCFR1), 35-49 (LCFR2), 57-88 (LCFR3), and 98-107 (LCFR4), and the heavy chain FR residues are located at approximately residues 1-30 (HCFR1), 36-49 (HCFR2), 66-94 (HCFR3), and 103-113 (HCFR4) in the heavy chain. If the CDRs comprise amino acid residues from a hypervariable loop, the light chain FR residues are located at approximately residues 1-25 (LCFR1), 33-49 (LCFR2), 53-90 (LCFR3), and 97-107 (LCFR4) in the light chain, and the heavy chain FR residues are located at approximately residues 1-25 (HCFR1), 33-52 (HCFR2), 56-95 (HCFR3), and 102-113 (HCFR4) in the heavy chain. In some cases, when the CDRs comprise amino acids from a CDR and a hypervariable loop as defined by Kabat, the FR residues are adjusted accordingly. For example, when CDRH1 comprises amino acids H26-H35, the heavy chain FR1 residues are located at positions 1-25 and FR2 residues are located at positions 36-49.

A "human consensus framework" is a framework that represents the most common amino acid residues found in selected human immunoglobulin VL or VH framework sequences. Typically, human immunoglobulin VL or VH sequences are selected from a subset of variable domain sequences. Typically, the subset of sequences is a subset as in Kabat. In one embodiment, for VL, the subset is a κI subset as in Kabat. In one embodiment, for VH, the subset is a III subset as in Kabat.

An example of a "complete" antibody is one that comprises an antigen binding site and CL and heavy chain constant domains CH1, CH2 and CH3. The constant domains may be native sequence constant domains (eg, human native sequence constant domains) or amino acid sequence variants thereof.

"Antibody fragments" comprise a portion of an intact antibody, preferably the antigen binding or variable region of an intact antibody. Examples of antibody fragments include Fab, Fab', F(ab')2, and Fv fragments; diabodies (Db); tandem diabodies (taDb), linear antibodies (e.g., Example 2 of U.S. Patent 5,641,870; Zapata et al., Protein Eng. 8(10): 1057-1062 (1995)); one-armed antibodies, single variable domain antibodies, minibodies, single-chain antibody molecules; and multispecific antibodies formed from antibody fragments (e.g., including but not limited to Db-Fc, taDb-Fc, taDb-CH3, and (scFV)4-Fc).

A "Fab" fragment is an antigen-binding fragment produced by papain digestion of an antibody and consists of the entire L and H chain variable domains (VH), and the first constant domain (CH1) of one heavy chain. Papain digestion of an antibody produces two identical Fab fragments. Pepsin treatment of an antibody produces a single large F(ab')2 fragment, which roughly corresponds to two disulfide-linked Fab fragments with divalent antigen-binding activity and is still capable of cross-linking antigen. The Fab' fragment differs from the Fab fragment in having several additional residues at the carboxyl terminus of the CH1 domain, including one or more cysteines from the antibody hinge region. Fab'-SH is herein referred to as Fab' in which the cysteine residues of the constant domains bear a free thiol group. F(ab')2 antibody fragments originally were pairs of Fab' fragments with hinge cysteines between them. Other chemical couplings of antibody fragments are also known.

The "Fc" fragment is the residual antibody fragment produced by papain digestion, a name reflecting its ability to readily crystallize. The Fc fragment comprises the carboxyl-terminal portions of two H chains held together by disulfide bonds. The effector functions of an antibody are determined by sequences within the Fc region; this region is also recognized by Fc receptors (FcRs) present on certain cell types.

The term "Fc region" herein is used to define the C-terminal region of an immunoglobulin heavy chain, including native sequence Fc regions and variant Fc regions. Although the boundaries of the immunoglobulin heavy chain Fc region can vary, the human IgG heavy chain Fc region is generally defined as extending from position Cys226, or from the amino acid residue of Pro230 to its carboxyl terminus. The C-terminal lysine (residue 447 according to the EU numbering system) in the Fc region can be removed, for example, during production or purification of the antibody, or by recombinantly engineering the nucleic acid encoding the heavy chain of the antibody. Thus, the composition of a complete antibody can include antibody groups in which all K447 residues have been removed, antibody groups in which the K447 residue has not been removed, and antibody groups with a mixture of antibodies with and without the K447 residue.

A "functional Fc region" possesses the "effector functions" of a native sequence Fc region. Exemplary "effector functions" include binding to C1q; CDC; binding to Fc receptors; ADCC; phagocytosis; downregulation of cell surface receptors (e.g., B cell receptor, BCR), etc. Such effector functions generally require the Fc region to be combined with a binding domain (e.g., an antibody variable domain) and can be assessed using a variety of assays as disclosed in the definitions section herein.

A "native sequence Fc region" comprises an amino acid sequence identical to the amino acid sequence of a naturally occurring Fc region. Native sequence human Fc regions include native sequence human IgG1 Fc regions (non-A and A allotypes); native sequence human IgG2 Fc regions; native sequence human IgG3 Fc regions; and native sequence human IgG4 Fc regions and naturally occurring variants thereof.

A "variant Fc region" comprises an amino acid sequence that differs from a native sequence Fc region by virtue of at least one amino acid modification, preferably one or more amino acid substitutions. Preferably, the variant Fc region has at least one amino acid substitution compared to a native sequence Fc region or to the Fc region of a parent polypeptide, for example, from about 1 to about 10 amino acid substitutions in a native sequence Fc region or in the Fc region of a parent polypeptide, preferably from about 1 to about 5 amino acid substitutions. The variant Fc region herein preferably has at least about 80% homology to a native sequence Fc region and/or to the Fc region of a parent polypeptide, most preferably at least about 90% homology thereto, and more preferably at least about 95% homology thereto.

As used herein, "Fc complex" refers to two co-interacting CH2 domains and/or two co-interacting CH3 domains of an Fc region, wherein the CH2 domains and/or CH3 domains interact through bonds and/or forces that are not peptide bonds (e.g., van der Waals forces, hydrophobic forces, hydrophilic forces).

As used herein, "Fc component" refers to the hinge region, CH2 domain, or CH3 domain of the Fc region.

As used herein, "Fc CH component" or "FcCH" refers to a polypeptide comprising the CH2 domain, the CH3 domain, or both the CH2 and CH3 domains of an Fc region.

"Fv" consists of a dimer of one heavy chain and one light chain variable region domain in tight, non-covalent association. The folding of these two domains produces six hypervariable loops (3 loops each from the H and L chains) that provide the amino acid residues for antigen binding and confer antigen-binding specificity to the antibody. However, even a single variable domain (or half of an Fv containing only three CDRs specific for an antigen) has the ability to recognize and bind antigen, although generally at a lower affinity than the entire binding site.

As used herein, the term "immunoadhesin" refers to a molecule that combines the binding specificity of a heterologous protein ("adhesin") and the effector function of an immunoglobulin constant domain. Structurally, an immunoadhesin comprises a fusion of an amino acid sequence having the desired binding specificity, different from the antigen recognition and binding site of an antibody (i.e., "heterologous" compared to the constant region of an antibody) and an immunoglobulin constant domain sequence (e.g., the CH2 and/or CH3 sequence of an IgG). The adhesin and immunoglobulin constant domain may optionally be separated by an amino acid spacer. Exemplary adhesin sequences include contiguous amino acid sequences comprising portions of a receptor or ligand that binds to a target protein. An adhesin sequence may also be a sequence that binds to a target protein, but is not a receptor or ligand sequence (e.g., an adhesin sequence in a peptibody). Such polypeptide sequences can be selected or identified using a variety of methods, including phage display technology and high-throughput sorting methods. The immunoglobulin constant domain sequences in the immunoadhesins can be obtained from any immunoglobulin, such as IgG1, IgG2, IgG3, or IgG4 subclasses, IgA (including IgAl and IgA2), IgE, IgD, or IgM.

As used herein, amino acid " spacer " refers to an amino acid sequence of 2 or more amino acids in length, which cannot be cut, for example, by self-cleavage, enzymatic or chemical cleavage. Spacer can be composed of neutral, polar or non-polar amino acids. Amino acid spacer length can be, for example, 2 to 100 amino acids long, for example, 10-80 amino acids long or 20-40 amino acids long, for example, 3, 5, 10, 15, 20, 25, 30, 35 or 40 amino acids long. In some embodiments, amino acid spacer can include glycine (G) and serine (S) residues, such as GGS repeats. In some embodiments, amino acid spacer can include threonine (T) and histidine (H) residues. Exemplary spacers are THT (SEQ ID NO: 1), GGGSTHT (SEQ ID NO: 2) and GGGSGGGSTHT (SEQ ID NO: 3).

"Single-chain Fv," also abbreviated as "sFv" or "scFv," is an antibody fragment comprising VH and VL antibody domains linked into a single polypeptide chain. Preferably, the sFv polypeptide further comprises a polypeptide linker between the VH and VL domains that enables the sFv to form an ideal structure for antigen binding. For a review of sFv, see Pluckthun, The Pharmacology of Monoclonal Antibodies, vol. 113, Rosenburg and Moore, eds., Springer-Verlag, New York, pp. 269-315 (1994); Malmborg et al., J. Immunol. Methods 183: 7-13, 1995.

The term "diabodies" refers to small antibody fragments that are prepared by constructing a short linker (about 5-10 residues) between the VH and VL domains to achieve interchain, rather than intrachain, pairing of the V domains, resulting in a bivalent fragment, i.e., an sFv fragment (see above) with two antigen-binding sites. Bispecific diabodies are heterodimers of two "crossover" sFv fragments in which the VH and VL domains of the two antibodies are located on different polypeptide chains. Diabodies are described more fully in, for example, EP 404,097; WO 93/11161 and Hollinger et al., Proc. Natl. Acad. Sci. USA 90:6444-6448 (1993).

The expressions "single domain antibody" (sdAb) or "single variable domain (SVD) antibody" generally refer to antibodies in which a single variable domain (VH or VL) can confer antigen binding. That is, the single variable domain does not need to interact with another variable domain to recognize the target antigen. Examples of single domain antibodies include those derived from camelids (llamas and camels) and cartilaginous fish (e.g., nurse sharks), as well as those derived from recombinant methods of human and mouse antibodies (Nature (1989) 341:544-546; Dev Comp Immunol (2006) 30:43-56; Trend Biochem Sci (2001) 26:230-235; Trends Biotechnol (2003) 21:484-490; WO 2005/035572; WO 03/035694; Febs Lett (1994) 339:285-290; WO 00/29004; WO 02/051870).

As used herein, the term "half antibody" refers to one arm of an antibody and comprises at least a VH domain and a CH domain. In some embodiments, a half antibody can associate with an immunoadhesin to form a heteromultimer of the present invention. In other embodiments, a first half antibody can associate with a second half antibody having the same or different amino acid sequence (e.g., differing by at least one amino acid residue) to form a symmetric or asymmetric heteromultimer, respectively.

The term "single-chain antibody" is used herein in the broadest sense and specifically covers antibodies with monospecificity or multispecificity (e.g., bispecific) that are initially produced as a single continuous polypeptide chain. Such single-chain antibodies include, but are not limited to, antibodies having two linked HCs that may be different or identical from each other and comprise two different or identical VH domains, a HD tether connecting the two HCs, and at least one heavy chain constant domain selected from two different or identical CH2 domains and two different or identical CH3 domains. Single-chain antibodies may additionally comprise one or two, different or identical CH1 domains. In some embodiments, single-chain antibodies comprise one or two hinge domains connecting one HC domain (e.g., VH or CH1) to a second, continuously positioned domain (e.g., CH2). In other embodiments, single-chain antibodies may comprise one or two linked LCs that may be different or identical from each other and may each comprise two different or identical VL and CL domains, each linked to a specific HC via a CLH tether. As described herein, single-chain antibodies may additionally use knob-into-hole technology to support HC/HC or HC/LC heterodimerization and may comprise a cleavable tether.Single-chain antibodies are heteromultimers of the present invention.

As used herein, the term "multi-chain antibody" refers to an antibody composed of two LCs and two HCs, wherein the two HCs are expressed as a single polypeptide chain and at least one LC is expressed as a separate polypeptide chain. The independently expressed LCs associate with their cognate HCs to form an antibody with two functional arms. Multi-chain antibodies can be monospecific or multispecific. Multi-chain antibodies can additionally utilize knob-into-hole technology to support HC/HC or HC/LC heterodimerization and can include cleavable tethers.

The term "knob-into-hole" or "KnH" technology, as referred to herein, refers to a technique for directing the pairing of two polypeptides in vitro or in vivo by introducing a protrusion (knob) into one polypeptide and a cavity (hole) into another polypeptide at their interacting interface. For example, KnH has been introduced into the Fc:Fc binding interface, CL:CH1 interface, or VH/VL interface of an antibody (e.g., US2007/0178552, WO 96/027011, WO 98/050431, and Zhu et al. (1997) Protein Science 6:781-788). This is particularly useful in driving the pairing of two different heavy chains together during the preparation of multispecific antibodies. For example, a multispecific antibody having a KnH in its Fc region may further comprise a single variable domain linked to each Fc region, or further comprise different heavy chain variable domains paired with similar or different light chain variable domains. KnH technology can also be used to pair together two different receptor extracellular domains, or any other polypeptide sequences containing different target recognition sequences.

The term "multispecific antibody" is used in the broadest sense and specifically covers antibodies with multi-epitope specificity. Such multispecific antibodies include, but are not limited to, antibodies comprising a heavy chain variable domain (VH) and a light chain variable domain (VL), wherein the VH/VL unit has multi-epitope specificity, antibodies with two or more VL and VH domains, each VH/VL unit binding to a different epitope, antibodies with two or more single variable domains, each single variable domain binding to a different epitope, full-length antibodies, antibody fragments such as Fab, Fv, dsFv, scFv, diabodies, bispecific diabodies and trivalent antibodies, antibody fragments that have been covalently or non-covalently linked. "Multi-epitope specificity" refers to the ability to specifically bind to two or more different epitopes on the same or different targets. "Monospecific" refers to the ability to bind only one antigen. In one embodiment, a monospecific heteromultimer binds to two different epitopes on the same target/antigen. According to one embodiment, the multispecific antibody is an IgG antibody that binds each epitope with an affinity of 5 μM to 0.001 pM, 3 μM to 0.001 pM, 1 μM to 0.001 pM, 0.5 μM to 0.001 pM, or 0.1 μM to 0.001 pM.

The antibodies of the present invention may be "chimeric" antibodies, in which a portion of the heavy and/or light chain is identical or homologous to the corresponding sequence in an antibody derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain is identical or homologous to the corresponding sequence in an antibody derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, as long as they exhibit the desired biological activity (U.S. Patent No. 4,816,567; and Morrison et al., Proc. Natl. Acad. Sci. USA 81: 6851-6855 (1984)). The chimeric antibodies of interest herein include primatized antibodies comprising variable domain antigen-binding sequences derived from non-human primates (e.g., Old World monkeys, apes, etc.) and human constant region sequences.

The "humanized" form of a non-human (e.g., rodent) antibody is a chimeric antibody containing the minimum sequence derived from a non-human antibody. In most cases, a humanized antibody is a human immunoglobulin (recipient antibody) in which the residues from the hypervariable region of the receptor are replaced by residues from a hypervariable region of a non-human species (donor antibody), such as mouse, rat, rabbit, or non-human primate, with ideal antibody specificity, affinity, and performance. In some cases, human immunoglobulin framework region (FR) residues are replaced by corresponding non-human residues. In addition, humanized antibodies may include residues that are not present in the receptor antibody or the donor antibody. Such modifications are made to further improve antibody performance. Typically, a humanized antibody will include substantially all or at least one, typically two variable domains, wherein all or substantially all of the hypervariable loops correspond to the hypervariable loops of a non-human immunoglobulin, and all or substantially all of the FRs are FRs of human immunoglobulin sequences. Humanized antibodies also optionally include at least a portion of an immunoglobulin constant region (Fc), typically a constant region of a human immunoglobulin. For more detailed information, see Jones et al., Nature 321:522-525 (1986); Riechmann et al., Nature 332:323-329 (1988); and Presta, Curr. Op. Struct. Biol. 2:593-596 (1992).

As used herein, "complex" or "complexed" refers to an association of two or more molecules that interact with each other through bonds and/or forces other than peptide bonds (e.g., van der Waals forces, hydrophobic forces, hydrophilic forces). In one embodiment, the complex is a heteromultimer. It should be understood that the terms "protein complex" or "polypeptide complex" as used herein include complexes having non-protein substances conjugated to the protein (e.g., including but not limited to chemical molecules such as toxins or detection agents) within the protein complex.

As used herein, the term "heteromultimer" or "heteromultimeric" describes two or more polypeptides having different sequences that interact with each other through non-peptide covalent bonds (e.g., disulfide bonds) and/or non-covalent interactions (e.g., hydrogen bonds, ionic bonds, van der Waals forces, or hydrophobic interactions). Multimeric polypeptides in an initially linked form (e.g., produced as a single continuous polypeptide chain) are also included in this definition. As used herein, heteromultimers include, for example, single-chain antibodies and multi-chain antibodies, as well as multimers having one or more half antibodies associated with one or more immunoadhesins. Heteromultimers include polypeptides and/or polypeptide complexes in which HD tethers are present or absent.

An antibody that "binds" to an antigen of interest of the present invention is one that binds to the antigen with sufficient affinity so that the antibody can be used as a diagnostic and/or therapeutic agent for targeting a protein or a cell or tissue expressing the antigen, and does not significantly cross-react with other proteins. In such embodiments, the extent to which the antibody binds to "non-target" proteins is less than about 10% of the binding of the antibody to its specific target protein, as determined by fluorescence activated cell sorting (FACS) analysis or radioimmunoprecipitation (RIA) or ELISA. With respect to the binding of an antibody to a target molecule, the term "specific binding" or "specifically binds" or "specifically binds to" or "specific for" a specific polypeptide or an epitope on a specific polypeptide target refers to binding that is measurably different from non-specific interactions (e.g., non-specific interactions can be binding to bovine serum albumin or casein). Specific binding can be measured, for example, by comparing the binding of a determined molecule to the binding of a control molecule. For example, specific binding can be determined by competition with a control molecule similar to the target, such as an excess of unlabeled target. In this case, specific binding is indicated if the binding of the labeled target to the probe is competitively inhibited by an excess of unlabeled target. As used herein, the terms "specifically bind" or "specifically binds to" or are "specific for" a particular polypeptide or an epitope on a particular polypeptide target can be displayed by a molecule that has an affinity for the target of about 1 μM to about 1 fM, alternatively about 200 nM to about 1 fM, alternatively about 200 nM to about 1 pM, alternatively about 150 nM to about 1 fM, alternatively about 150 nM to about 1 pM, alternatively about 100 nM to about 1 fM, alternatively about 100 nM to about 1 pM, alternatively about 60 nM to about 1 fM, alternatively about 60 nM to about 1 pM, alternatively about 50 nM to about 1 fM, alternatively about 50 nM to about 1 pM, Alternatively, a Kd of about 30nM to about 1fM, alternatively about 30nM to about 1pM, alternatively about 20nM to about 1fM, alternatively about 20nM to about 1pM, alternatively about 10nM to about 1fM, alternatively about 10nM to about 1pM, alternatively about 8nM to about 1fM, alternatively about 8nM to about 1pM, alternatively about 6nM to about 1fM, alternatively about 6nM to about 1pM, alternatively about 4nM to about 1fM, alternatively about 4nM to about 1pM, alternatively about 2nM to about 1fM, alternatively about 2nM to about 1pM, alternatively about 1nM to about 1fM, alternatively about 1nM to about 1pM. In one embodiment, the term "specific binding" refers to binding wherein a molecule binds to a specific polypeptide or epitope on a specific polypeptide and does not substantially bind to other polypeptides or polypeptide epitopes.

"Binding affinity" generally refers to the strength of the sum of the total non-covalent interactions between a single binding site of a molecule (e.g., an antibody) and its binding partner (e.g., an antigen). Unless otherwise indicated, "binding affinity" as used herein refers to the intrinsic binding affinity that reflects a 1:1 interaction between members of a binding pair (e.g., an antibody and an antigen). The affinity of a molecule X for its partner Y can generally be represented by a dissociation constant (Kd). For example, Kd can be about 200nM, 150nM, 100nM, 60nM, 50nM, 40nM, 30nM, 20nM, 10nM, 8nM, 6nM, 4nM, 2nM, 1nM or stronger. Affinity can be measured using conventional methods well known in the art, including those described herein. Low-affinity antibodies generally bind to antigen slowly and tend to dissociate easily, while high-affinity antibodies generally bind to antigen faster and tend to remain bound longer. A variety of methods for measuring binding affinity are known in the art, any of which can be used for the purposes of the present invention.

In one embodiment, the "Kd" or "Kd value" according to the present invention is measured by surface plasmon resonance assay using a BIAcore ^™ -2000 or BIAcore ^™ -3000 (BIAcore, Inc., Piscataway, NJ) at 25°C with an immobilized antigen CM5 chip at ~10 response units (RU). The antibodies of the invention may have an ATP concentration of about 1 μM to about 1 fM, alternatively about 200 nM to about 1 fM, alternatively about 200 nM to about 1 pM, alternatively about 150 nM to about 1 fM, alternatively about 150 nM to about 1 pM, alternatively about 100 nM to about 1 fM, alternatively about 100 nM to about 1 pM, alternatively about 60 nM to about 1 fM, alternatively about 60 nM to about 1 pM, alternatively about 50 nM to about 1 fM, alternatively about 50 nM to about 1 pM, alternatively about 30 nM to about 1 fM, alternatively about 30 nM to about 1 pM. , alternatively an affinity of a Kd value of about 20 nM to about 1 fM, alternatively about 20 nM to about 1 pM, alternatively about 10 nM to about 1 fM, alternatively about 10 nM to about 1 pM, alternatively about 8 nM to about 1 fM, alternatively about 8 nM to about 1 pM, alternatively about 6 nM to about 1 fM, alternatively about 6 nM to about 1 pM, alternatively about 4 nM to about 1 fM, alternatively about 4 nM to about 1 pM, alternatively about 2 nM to about 1 fM, alternatively about 2 nM to about 1 pM, alternatively about 1 nM to about 1 fM, alternatively about 1 nM to about 1 pM. To measure Kd values using surface plasmon resonance, carboxymethylated dextran biosensor chips (CMS, BIAcore Inc.) were activated with N-ethyl-N'-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) according to the supplier's instructions. Antigen was diluted to 5 μg/ml (~0.2 μM) with 10 mM sodium acetate, pH 4.8, and then injected at a flow rate of 5 μl/min to obtain approximately 10 response units (RU) of coupled protein. Following antigen injection, 1 M ethanolamine was injected to block unreacted groups. For kinetic measurements, two-fold serial dilutions of Fab (e.g., 0.78 nM to 500 nM) in PBS containing 0.05% Tween 20 (PBST) were injected at 25°C at a flow rate of approximately 25 μl/min. The association rate (k _on ) and dissociation rate (k _off ) were calculated using a simple one-to-one Langmuir binding model (BIAcore Evaluation Software Version 3.2) by simultaneously fitting the association and dissociation sensorgrams. The equilibrium dissociation constant (Kd) was calculated as the ratio of k _off /k _on . See, e.g., Chen et al., J. Mol. Biol. 293: 865-881 (1999). If the on-rate measured by the surface plasmon resonance assay described above exceeds 10 ⁶ M ⁻¹ s ⁻¹ , the on-rate can be measured using a fluorescence quenching technique that measures the increase or decrease in fluorescence emission intensity (excitation = 295 nm; emission = 340 nm, 16 nm bandpass) of 20 nM anti-antigen antibody (Fab form) in PBS, pH 7.2 at 25° C. in the presence of increasing concentrations of antigen as measured by a spectrometer such as a spectrophotometer equipped with a stop-flow device (Aviv Instruments) or an 8000 series SLM-Aminco spectrophotometer with a stirring cuvette (ThermoSpectronic).

The "association rate" or "kon" according to the present invention can also be measured using the same surface plasmon resonance technique described above using a BIACore ^™ -2000 or BIAcore ^™ -3000 (BIAcore, Inc., Piscataway, NJ) at 25°C using an immobilized antigen CM5 chip at approximately 10 response units (RU). Briefly, a carboxymethylated dextran biosensor chip (CMS, BIAcore Inc.) was activated with N-ethyl-N'-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) according to the supplier's instructions. Antigen was diluted to 5 μg/ml (~0.2 μM) with 10 mM sodium acetate, pH 4.8, and then injected at a flow rate of 5 μl/min to obtain approximately 10 response units (RU) of coupled protein. Following antigen injection, 1 M ethanolamine was injected to block unreacted groups. For kinetic measurements, two-fold serial dilutions of Fab (e.g., 0.78 nM to 500 nM) in PBS containing 0.05% Tween 20 (PBST) were injected at 25° C. at a flow rate of approximately 25 μl/min. The association rate (k _on ) and dissociation rate (k _off ) were calculated using a simple one-to-one Langmuir binding model (BIAcore Evaluation Software Version 3.2) by simultaneously fitting the association and dissociation sensorgrams. The equilibrium dissociation constant (Kd) was calculated as the ratio of k _off /k _on . See, e.g., Chen et al., J. Mol. Biol. 293: 865-881 (1999). If the on-rate measured by the surface plasmon resonance assay described above exceeds 10 ⁶ M ⁻¹ s ⁻¹ , the on-rate can be measured using a fluorescence quenching technique that measures the increase or decrease in fluorescence emission intensity (excitation = 295 nm; emission = 340 nm, 16 nm bandpass) of 20 nM anti-antigen antibody (Fab form) in PBS, pH 7.2 at 25° C. in the presence of increasing concentrations of antigen as measured by a spectrometer such as a spectrophotometer equipped with a stop-flow device (Aviv Instruments) or an 8000 series SLM-Aminco spectrophotometer with a stirring cuvette (ThermoSpectronic).

Unless otherwise indicated, "biologically active" and "biological characteristics" with respect to a polypeptide of the present invention, such as an antibody, fragment or derivative thereof, refers to the ability to bind a biological molecule.

The heteromultimers of the invention are typically purified to essential homogeneity.The phrases "essentially homogeneous," "essentially homogeneous form," and "essentially homogeneous" are used to indicate that the product is substantially free of by-products derived from undesired combinations of polypeptides.

Expressed in terms of purity, substantially homogeneous means that the amount of by-products does not exceed 10%, 9%, 8%, 7%, 6%, 4%, 3%, 2% or 1% by weight, or is less than 1% by weight. In one embodiment, the by-products are less than 5%.

"Biomolecule" refers to nucleic acids, proteins, carbohydrates, lipids, and combinations thereof. In one embodiment, the biomolecule is naturally occurring.

When used to describe the various heteromultimers disclosed herein, "isolated" means that the heteromultimer has been identified and separated and/or recovered from the cell or cell culture in which it is expressed. Contaminant components of its natural environment are materials that typically interfere with the diagnostic or therapeutic use of the polypeptide and may include enzymes, hormones, and other proteinaceous or non-proteinaceous solutes. In preferred embodiments, the heteromultimer will be purified (1) to a degree sufficient to obtain at least 15 residues of N-terminal or internal amino acid sequence using a spinning cup sequenator, or (2) to homogeneity by SDS-PAGE under non-reducing or reducing conditions using Coomassie blue, preferably silver stain. Isolated heteromultimers include antibodies that are in situ within recombinant cells because at least one component of the polypeptide's natural environment will not be present. However, typically, the isolated polypeptide will be prepared by at least one purification step.

As used herein, "linked" or "linked" refers to a direct peptide bond connection between a first and a second amino acid sequence or a connection involving a peptide bond to a third amino acid sequence between the first and second amino acid sequences. For example, an amino acid linker that is bound to the C-terminus of one amino acid sequence and the N-terminus of another amino acid sequence.

As used herein, a "linker" refers to an amino acid sequence that is two or more amino acids long. A linker can be composed of neutral, polar, or non-polar amino acids. The linker length can be, for example, 2 to 100 amino acids long, for example 2 to 50 amino acids long, for example 3, 5, 10, 15, 20, 25, 30, 35, 40, 45, or 50 amino acids long. A linker can be "cleavable," for example, by self-cleavage or enzymatic or chemical cleavage. Cleavage sites in amino acid sequences and enzymes and chemicals that cut at such sites are well known in the art and are also described herein.

As used herein, an "HD tether" or "heterodimerization tether" refers to an amino acid linker that connects two different heavy chain constant (CH) domain-containing polypeptides. Typically, the two CH domain-containing polypeptides are linked by linking the CH2 or CH3 domain of a first polypeptide to a VL domain, which is itself a component of a second CH domain-containing polypeptide. In some embodiments, the HD tether directly connects the CH3 domain of a first polypeptide to the VH domain of a second CH domain-containing polypeptide. Generally, HD tethers of 15-100 amino acids are effective, as are HD tethers of 20-40 amino acids, 25-40 amino acids, and 30-40 amino acids. In particular embodiments, the HD tether is 30-39 amino acids long (e.g., 30, 31, 32, 33, 34, 35, 36, 37, 38, or 39 amino acids in length). The HD tether can be "cleavable," for example, by autocleavage or by enzymatic or chemical cleavage using methods or reagents standard in the art.

As used herein, "CLH tether" or "cognate LC-HC tether" refers to the amino acid linker that connects a light chain to its cognate heavy chain. A CLH tether generally refers to the amino acids that connect the CL domain of a light chain to the VH domain of a heavy chain. In some embodiments, a CLH tether directly connects the VL domain of a light chain to the VH domain of a heavy chain. Typically, CLH tethers of 10-80 amino acids are effective, as are CLH tethers of 20-40 amino acids, 25-40 amino acids, 30-40 amino acids, and 35-40 amino acids (e.g., 30, 31, 32, 33, 34, or 35 amino acids). The single-chain antibodies of the present invention may have multiple CLH tethers, which may be the same or different in sequence and/or length. In a preferred embodiment, the single-chain antibody has two tethers (CLH tether 1 and CLH tether 2), each connecting a light chain to its cognate heavy chain. The CLH tether can be "cleavable," for example, by self-cleavage, or enzymatic or chemical cleavage using methods or reagents standard in the art.

Enzymatic cleavage of the "linker" or "tether" can involve the use of an endopeptidase, such as, for example, urokinase, Lys-C, Asp-N, Arg-C, V8, Glu-C, chymotrypsin, trypsin, pepsin, papain, thrombin, tissue plasminogen activator (tPa), genenase, factor Xa, TEV (tobacco etch virus protease), enterokinase, HRV C3 (human rhinovirus C3 protease), kininogenase, and a subtilisin-like proprotein convertase (e.g., furin (PC1), PC2, or PC3) or N-arginine dibasic convertase. In a desired embodiment, enzymatic cleavage involves the endopeptidase furin. Chemical cleavage also involves the use of, for example, hydroxylamine, N-chlorosuccinimide, N-bromosuccinimide, or cyanogen bromide.

As used herein, a "furin endopeptidase cleavage site" is an _X1 - _X2 - _X3 -arginine amino acid sequence (SEQ ID NO: 6), wherein _X1 is a basic amino acid residue (natural or non-natural, modified or unmodified), and _X2 and _X3 can each be any amino acid residue (natural or non-natural, modified or unmodified) that can be cleaved by a furin endopeptidase at the C-terminal side. Furin endopeptidase cleaves at the C-terminal side of the arginine residue. In certain embodiments, the furin cleavage site comprises the amino acid sequence RXRXYR, wherein Y is K or R, and X is any amino acid residue (SEQ ID NO: 7), more specifically RXRXRR (SEQ ID NO: 8). In certain embodiments, the furin cleavage site comprises the amino acid sequence RKRKRR (SEQ ID NO: 9). In certain other embodiments, the furin cleavage site comprises the amino acid sequence RHRQPR (SEQ ID NO: 10). In other embodiments, the furin cleavage site comprises the amino acid sequence RSRKRR (SEQ ID NO: 11).

As used herein, a "Lys-C endopeptidase cleavage site" is a lysine residue in an amino acid sequence that can be cleaved at the C-terminal side by a Lys-C endopeptidase. Lys-C endopeptidase cleaves at the C-terminal side of a lysine residue.

Enzymatic cleavage of a "linker" or "tether" may also involve the use of an exopeptidase, such as carboxypeptidase A, carboxypeptidase B, carboxypeptidase D, carboxypeptidase E (also known as carboxypeptidase H), carboxypeptidase M, carboxypeptidase N, or carboxypeptidase Z, to remove residual endopeptidase recognition sequences following endopeptidase cleavage.

Antibody "effector functions" refer to those biological activities attributable to the Fc region (a native sequence Fc region or an amino acid sequence variant Fc region) of an antibody, and may vary with the antibody isotype. Examples of antibody effector functions include: binding to C1q and complement-dependent cytotoxicity; binding to Fc receptors; antibody-dependent cell-mediated cytotoxicity (ADCC); phagocytosis; downregulation of cell surface receptors (e.g., B cell receptors); and B cell activation.

"Antibody-dependent cell-mediated cytotoxicity," or "ADCC," refers to a form of cytotoxicity in which secreted Ig bound to Fc receptors (FcRs) present on certain cytotoxic cells (e.g., natural killer (NK) cells, neutrophils, and macrophages) enables these cytotoxic effector cells to specifically bind to target cells bearing an antigen and then kill the target cells with a cytotoxic agent. Antibodies "arm" cytotoxic cells and are absolutely required for this type of killing. NK cells, the primary cells mediating ADCC, express only FcγRIII, while monocytes express FcγRI, FcγRII, and FcγRIII. The expression of FcRs on hematopoietic cells is summarized in Table 3 on page 464 of Ravetch and Kinet, Annu. Rev. Immunol. 9:457-92 (1991). To assess the ADCC activity of a molecule of interest, an in vitro ADCC assay, such as that described in U.S. Pat. No. 5,500,362 or 5,821,337, can be performed. Useful effector cells for such assays include peripheral blood mononuclear cells (PBMC) and natural killer (NK) cells. Alternatively or additionally, ADCC activity of the target molecule can be assessed in vivo, e.g., in an animal model, such as that disclosed in Clynes et al., Proc. Natl. Acad. Sci. USA 95: 652-656 (1998).

"Fc receptor" or "FcR" describes a receptor that binds to the Fc region of an antibody. A preferred FcR is a native sequence human FcR. In addition, a preferred FcR is an FcR that binds to an IgG antibody (gamma receptor), and includes receptors of the FcγRI, FcγRII, and FcγRIII subclasses, including allelic variants and alternatively spliced forms of such receptors. FcγRII receptors include FcγRIIA ("activating receptor") and FcγRIIB (inhibitory receptor), which have similar amino acid sequences, the main difference being their cytoplasmic domains. The activating receptor FcγRIIA contains an immunoreceptor tyrosine-based activation motif (ITAM) in its cytoplasmic domain. The inhibitory receptor FcγRIIB contains an immunoreceptor tyrosine-based inhibition motif (ITIM) in its cytoplasmic domain (see review M. in Annu. Rev. Immunol. 15: 203-234 (1997)). FcRs are reviewed in Ravetch and Kinet, Annu. Rev. Immunol. 9:457-492 (1991); Capel et al., Immunomethods 4:25-34 (1994) and deHaas et al., J. Lab. Clin. Med. 126:330-41 (1995). The term "FcR" herein encompasses other FcRs, including those to be identified in the future. The term also encompasses the neonatal receptor FcRn, which is responsible for the transport of maternal IgG to the fetus (Guyer et al., J. Immunol. 117:587 (1976) and Kim et al., J. Immunol. 24:249 (1994)).

"Human effector cells" are leukocytes that express one or more FcRs and perform effector functions. Preferably, the cells express at least FcγRIII and perform ADCC effector functions. Examples of human leukocytes that mediate ADCC include peripheral blood mononuclear cells (PBMCs), natural killer (NK) cells, monocytes, cytotoxic T cells, and neutrophils, with PBMCs and NK cells being preferred. Effector cells can be isolated from natural sources, such as blood.

"Complement-dependent cytotoxicity" or "CDC" refers to the lysis of target cells in the presence of complement. Activation of the classical complement pathway is initiated by the binding of the first component of the complement system (C1q) to an antibody (of the appropriate subclass) bound to its cognate antigen. To assess complement activation, a CDC assay as described in Gazzano-Santoro et al., J. Immunol. Methods 202:163 (1996) can be performed.

The term "therapeutically effective amount" refers to the amount of a heteromultimer, antibody, antibody fragment or derivative used to treat a disease or condition in a subject. In the case of a tumor (e.g., a cancer tumor), a therapeutically effective amount of a heteromultimer, antibody or antibody fragment (e.g., a multispecific antibody or antibody fragment) can reduce the number of cancer cells; reduce the size of the primary tumor; inhibit (i.e., slow down to some extent, preferably prevent) the infiltration of cancer cells into peripheral organs; inhibit (i.e., slow down to some extent, preferably prevent) tumor metastasis; inhibit tumor growth to some extent; and/or alleviate one or more symptoms associated with the condition to some extent. To the extent that a heteromultimer, antibody or antibody fragment can prevent growth and/or kill existing cancer cells, it can be cytostatic and/or cytotoxic. For cancer treatment, in vivo efficacy can be measured, for example, by assessing survival time, time to disease progression (TTP), response rate (RR), duration of response and/or quality of life.

"Reduce or inhibit" means the ability to cause an overall decrease of preferably 20% or more, more preferably 50% or more, most preferably 75%, 85%, 90%, 95% or more. Reduce or inhibit can refer to the symptoms of the disorder being treated, the presence or size of metastases, the size of the primary tumor, the size or number of blood vessels in an angiogenic disorder.

The term "cytotoxic agent" as used herein refers to a substance that inhibits or prevents cell function and/or causes cell destruction. The term is intended to include radioisotopes (e.g., At ²¹¹ , I ¹³¹ , I ¹²⁵ , Y ⁹⁰ , Re ¹⁸⁶ , Re ¹⁸⁸ , Sm ¹⁵³ , Bi ²¹² , Ra ²²³ , P ³² and radioisotopes of Lu), chemotherapeutics such as methotrexate, adriamycin, vinca alkaloids (vincristine, vinblastine, etoposide), doxorubicin, melphalan, mitomycin C, chlorambucil, daunomycin or other intercalating agents, enzymes and fragments thereof such as nuclear decomposition enzymes, antibiotics and toxins such as small molecule toxins or enzyme-active toxins of bacterial, fungal, plant or animal origin, including fragments and/or variants thereof, as well as various antitumor, anticancer and chemotherapeutic agents disclosed herein. Other cytotoxic agents are disclosed herein. Tumoricidal agents cause tumor cell destruction.

A "chemotherapeutic agent" is a compound used to treat cancer. Examples of chemotherapeutic agents include alkylating agents such as thiotepa and cyclophosphamide; alkyl sulfonates such as busulfan, improsulfan, and piposulfan; aziridines such as benzodopa, carboquone, meturedopa, and uredopa; ethylenimines and methylamelamines including altretamine, triethylenemelamine, and acetaminophen. , trietylenephosphoramide, triethylenethiophosphoramide, and trimethylolomelamine; acetogenins (especially bullatacin and bullatacinone); delta-9-tetrahydrocannabinol (dronabinol); beta-lapachone; lapachoquinone; colchicine; betulinic acid acid; camptothecin (including the synthetic analogs topotecan, CPT-11 (irinotecan), acetylcamptothecin, scopolectin, and 9-aminocamptothecin); bryostatin; callystatin; CC-1065 (including its synthetic analogs adozelesin, carzelesin, and bizelesin); podophyllotoxin; podophyllinic acid acid; teniposide; cryptophycins (particularly cryptophycin 1 and cryptophycin 8); dolastatin; duocarmycin (including synthetic analogs, KW-2189, and CB1-TM1); eleutherobin; pancratistatin; sarcodictyin; spongistatin; nitrogen mustard mustards such as chlorambucil, chlornaphazine, cholophosphamide, estramustine, ifosfamide, mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembichin, phenesterine, prednimustine, trofosfamide, uracil mustard, mustard; nitrosureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, and ranimnustine; antibiotics such as enediyne antibiotics (e.g., calicheamicin, particularly calicheamicin gamma 1 (see, e.g., Agnew, Chem. Intl. ed. Engl., 33:183-186 (1994); dynemicins, including dynemicin A; esperamicin; and neocarzinostatin chromophores and related pigment proteins (enediyne antibiotic chromophores), aclacinomycins, actinomycin, authramycin, azaserine, bleomycins, cactinomycin, carabicin, carminomycin, carzinophilin, chromomycin, dactinomycin, daunorubicin, bicin, detorubicin, 6-diazo-5-oxo-L-norleucine, doxorubicin (including morpholino-doxorubicin, cyanomorpholino-doxorubicin, 2-pyrrolino-doxorubicin, and deoxydoxorubicin), epirubicin, esorubicin, idarubicin, marcellomycin, mitomycins such as mitomycin C, mycophenolic acid, acid, nogalamycin, olivomycin, peplomycin, potfiromycin, puromycin, quelamycin, rodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-fluorouracil (5-FU); folic acid analogs such as denopterin, methotrexate, pteroyltriglutamate, rin), trimetrexate; purine analogs such as fludarabine, 6-mercaptopurine, thiamiprine, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxifluridine, enocitabine, floxuridine; androgens such as calusterone, dromostanolone propionate, propionate, epitiostanol, mepitiostane, testolactone; antiadrenal drugs such as aminoglutethimide, mitotane, and trilostane; folic acid supplements such as frolinic acid; aceglaone; aldophosphamide glycoside; aminolevulinic acid; eniluracil; amsacrine; bestrabucil; bisantrene; edatraxate; defofamine; demecolcine; diaziquone; elfornithine; elliptinium acetate acetate; epothilone; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidamine; maytansinoids (such as maytansine and ansamitocin); mitoguazone; mitoxantrone; mopidamol; nitracrine; pentostatin; phenamet; pirarubicin; losoxantrone; 2-ethylhydrazide; procarbazine; polysaccharide complex (JHS Natural Products, Eugene, OR; razoxane; rhizoxin; sizofiran; spirogermanium; tenuazonic acid acid; triaziquone; 2,2',2"-trichlorotriethylamine; trichothecenes (especially T-2 toxin, verracurin A, roridin A, and anguidine); urethan; vindesine, dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; gacytosine; arabinoside ("Ara-C"); thiotepa; taxoids, such as paclitaxel (Bristol-Myers Squibb Co., Ltd. Oncology, Princeton, NJ), Cremophor-free, albumin-engineered nanoparticle formulations of paclitaxel (American Pharmaceutical Partners, Schaumberg, Illinois), and doxetaxel (Rhone-Poulenc Rorer, Antony, France); chloranbucil; gemcitabine; 6-thioguanine; mercaptopurine; methotrexate; platinum analogs such as cisplatin and carboplatin; vinblastine; etoposide (VP-16); ifosfamide ( ifosfamide; mitoxantrone; vincristine, oxaliplatin; leucovorin; vinorelbine, novantrone; edatrexate; daunomycin; aminopterin; ibandronate; CPT-11; topoisomerase inhibitors (RFS) 2000; difluoromethylornithine (DMFO); retinoids such as retinoic acid; capecitabine and pharmaceutically acceptable salts, acids or derivatives of any of the foregoing; and combinations of two or more of the foregoing, for example, CHOP, an abbreviation for combination therapy with cyclophosphamide, doxorubicin, vincristine, and prednisolone, and FOLFOX, an abbreviation for treatment with oxaliplatin (ELOXATIN™) in combination with 5-FU and leucovorin.

This definition also includes anti-hormonal agents that regulate, reduce, block or inhibit the effects of hormones that can promote cancer growth, and the anti-hormonal agents are generally in systemic or systemic treatment forms. They can be hormones themselves. Examples include anti-estrogens and selective estrogen receptor modulators (SERMs), including, for example, tamoxifen (tamoxifen) (including tamoxifen), raloxifene (raloxifene), droloxifene (droloxifene), 4-hydroxytamoxifen (4-hydroxytamoxifen), trioxifene (trioxifene), keoxifene, LY117018, onapristone (onapristone) and toremifene (toremifene); anti-progesterone drugs; estrogen receptor down-regulators (ERD); agents that act to inhibit or shut down the ovaries, for example, luteinizing hormone-releasing hormone (LHRH) agonists such as and leuprolide acetate (leuprolide acetate), goserelin acetate (goserelin acetate) androgen inhibitors, such as 4(5)-imidazoles, aminoglutethimide, megestrol acetate, exemestane, formestanie, fadrozole, vorozole, letrozole, and anastrozole. In addition, the definition of chemotherapeutic agents includes bisphosphonates such as clodronate (e.g., clodronate or etidronate), NE-58095, zoledronic acid/zoledronate, alendronate, pamidronate, tiludronate, or risedronate; and troxacitabine (a 1,3-dioxolane nucleoside cytosine analog); antisense oligonucleotides, particularly those that inhibit the expression of genes in signaling pathways involved in abnormal cell proliferation, such as PKC-α, Raf, H-Ras, and epidermal growth factor receptor (EGF-R); vaccines such as vaccines and gene therapy vaccines, for example, vaccines, vaccines, and vaccines; topoisomerase I inhibitors; GnRH antagonists; lapatinib ditosylate; ditosylate) (a small molecule inhibitor of ErbB-2 and EGFR dual tyrosine kinases, also known as GW572016); and pharmaceutically acceptable salts, acids or derivatives of any of the foregoing substances.

As used herein, the term "growth inhibitor" refers to a compound or composition that inhibits cell growth in vitro or in vivo. Therefore, a growth inhibitor can be an agent that significantly reduces the percentage of cells in the S phase. Examples of growth inhibitors include agents that block the advancement of the cell cycle (at a non-S phase position), such as agents that induce G1 arrest and M phase arrest. Classical M phase blockers include vinca alkaloids (such as vincristine and vinblastine), taxanes, and topoisomerase II inhibitors such as doxorubicin, epirubicin, daunorubicin, etoposide, and bleomycin. Agents that arrest G1 also spill over into S phase arrest, for example, DNA alkylating agents such as tamoxifen, prednisone, dacarbazine, mechlorethamine, cisplatin, methotrexate, 5-fluorouracil, and ara-C. For further information, see Molecular Basis of Cancer, Mendelsohn and Israel (eds.), Chapter 1, entitled "Cell cycle regulation, oncogenes, and antineoplastic drugs", Murakami et al. (WB Saunders: Philadelphia, 1995), especially page 13. Taxanes (paclitaxel and docetaxel) are anticancer drugs derived from the yew tree. Docetaxel (Rhone-Poulenc Rorer) is derived from the European yew tree and is a semisynthetic analog of paclitaxel (Bristol-Myers Squibb). Paclitaxel and docetaxel promote the assembly of tubulin dimers into microtubules and stabilize microtubules by preventing depolymerization, leading to the arrest of mitosis in cells.

As used herein, "anti-cancer therapy" refers to a treatment that reduces or inhibits cancer in a subject. Examples of anti-cancer therapies include cytotoxic radiation therapy and administration to a subject of a therapeutically effective amount of a cytotoxic agent, a chemotherapeutic agent, a growth inhibitory agent, a cancer vaccine, an angiogenesis inhibitor, a prodrug, a cytokine, a cytokine antagonist, a corticosteroid, an immunosuppressant, an antiemetic, an antibody or antibody fragment, or an analgesic.

"Target molecule" refers to a molecule that can bind to a protein complex of the invention (preferably with an affinity greater than 1 μM Kd according to Scatchard analysis). Examples of target molecules include, but are not limited to, serum soluble proteins and their receptors, such as cytokines and cytokine receptors, adhesins, growth factors and their receptors, hormones, viral particles (e.g., RSV F protein, CMV, StaphA, influenza virus, hepatitis C virus), microorganisms (e.g., bacterial cell proteins, fungal cells), adhesins, CD proteins and their receptors.

A "subject" is a vertebrate, such as a mammal, such as a human. Mammals include, but are not limited to, livestock (such as cattle), sports animals, pets (such as cats, dogs, and horses), primates, mice, and rats.

Unless otherwise stated, the commercial reagents involved in the examples were used according to the manufacturer's instructions. In the examples below, and throughout the specification, the sources of cells are identified by the ATCC accession number of the American Type Culture Collection, Manassas, VA. Unless otherwise stated, the present invention uses standard experimental protocols of recombinant DNA technology, such as those described above and in the following textbooks: Sambrook et al., supra; Ausubel et al., Current Protocols in Molecular Biology (Green Publishing Associates and Wiley Interscience, NY, 1989); Innis et al., PCR Protocols: A Guide to Methods and Applications (Academic Press, Inc., NY, 1990); Harlow et al., Antibodies: A Laboratory Manual (Cold Spring Harbor Press, Cold Spring Harbor, 1988); Gait, Oligonucleotide Synthesis (IRLPress, Oxford, 1984); Freshney, Animal Cell Culture, 1987; Coligan et al., Current Protocols in Immunology, 1991.

In this specification and claims, the word "comprise", or variations "comprises" or "comprising", will be understood to imply the inclusion of a stated item or group of items but not the exclusion of any other item or group of items.

As used herein, the term "prodrug" refers to a precursor or derivative form of a pharmaceutically active substance that is optionally less cytotoxic to tumor cells than the parent drug and can be enzymatically activated or converted to the more active parent form. See, for example, Wilman, "Prodrugs in Cancer Chemotherapy," Biochemical Society Transactions, 14, pp. 375-382, 615th Meeting Belfast (1986) and Stella et al., "Prodrugs: A Chemical Approach to Targeted Drug Delivery," Directed Drug Delivery, Borchardt et al., (eds.), pp. 247-267, Humana Press (1985). Prodrugs include, but are not limited to, phosphate-containing prodrugs, thiophosphate-containing prodrugs, sulfate-containing prodrugs, peptide-containing prodrugs, D-amino acid-modified prodrugs, glycosylated prodrugs, β-lactam-containing prodrugs, optionally substituted phenoxyacetamide-containing prodrugs or optionally substituted phenylacetamide-containing prodrugs, 5-fluorocytosine and other 5-fluorouridine prodrugs that can be converted into the more active cytotoxic free drug. Examples of cytotoxic drugs that can be derivatized into prodrug forms for use in the present invention include, but are not limited to, the chemotherapeutic agents described above.

The term "cytokine" refers to a general term for proteins released by one cell population that act as intercellular mediators on another cell. Examples of such cytokines are lymphokines, monokines, and traditional polypeptide hormones. Cytokines include growth hormones such as human growth hormone (HGH), N-methionyl human growth hormone, and bovine growth hormone; parathyroid hormone; thyroxine; insulin; proinsulin; relaxin; prorelaxin; glycoprotein hormones such as follicle-stimulating hormone (FSH), thyroid-stimulating hormone (TSH), and luteinizing hormone (LH); epidermal growth factor (EGH); liver growth factor; fibroblast growth factor (FGF); prolactin; placental lactogen; tumor necrosis factor-α and -β; Mullerian inhibitory substance; mouse gonadotropin-related peptide; inhibin; activin; vascular endothelial growth factor; integrin; thrombopoietin (TPO); nerve growth factors such as NGF-α; platelet-derived growth factor; transforming growth factors (TGF) For example, TGF-α and TGF-β; insulin-like growth factor-I and -II; erythropoietin (EPO); osteoinductive factor; interferons such as interferon-α, -β and -γ; colony stimulating factors (CSFs) such as macrophage CSF (M-CSF), granulocyte-macrophage CSF (GM-CSF); and granulocyte CSF (G-CSF); interleukins (ILs), such as IL-1, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-11, IL-12, IL-18; tumor necrosis factors such as TNF-α or TNF-β; and other polypeptide factors, including LIF and kit ligand (KL). The term cytokine as used herein includes proteins from natural sources or recombinant cell culture, as well as biologically active equivalents of native sequence cytokines.

"Cytokine antagonists" refer to molecules that partially or completely block, inhibit, or neutralize the biological activity of at least one cytokine. For example, cytokine antagonists can inhibit cytokine activity by inhibiting cytokine expression and/or secretion, or by binding to cytokines or cytokine receptors. Cytokine antagonists include antibodies, heteromultimers, synthetic or natural sequence peptides, immunoadhesins, and small molecule antagonists that bind to cytokines or cytokine receptors. Cytokine antagonists are optionally conjugated or fused to a cytotoxic agent.

As used herein, the term "immunosuppressant" refers to a substance that acts to suppress or mask the immune system of a subject. This includes substances that inhibit cytokine production, downregulate or suppress autoantibody expression, or mask MHC antigens. Examples of immunosuppressants include 2-amino-6-aryl-5-substituted pyrimidines (see U.S. Patent No. 4,665,077); mycophenolate mofetil (mycophenolate ethyl). mofetil, such as azathioprine/6-mercaptopurine; bromocryptine; danazol; dapsone; glutaraldehyde (which masks MHC antigens, as described in U.S. Pat. No. 4,120,649); anti-idiotypic antibodies to MHC antigens and MHC fragments; cyclosporine A; steroids, such as corticosteroids and glucocorticoids, such as prednisone, prednisolone such as (prednisolone sodium phosphate) or (prednisolone sodium phosphate oral solution), methylprednisolone, and dexamethasone; methotrexate (oral or subcutaneous) (TREXALL ^™ ); hydroxycloroquine/chloroquine; sulfasalazine; leflunomide; cytokine or cytokine receptor antagonists, including anti-interferon-α, -β, or -γ antibodies, anti-tumor necrosis factor-α antibodies (infliximab or adalimumab), anti-TNF-α immunoadhesins (etanercept), anti-tumor necrosis factor-β antibodies, anti-interleukin-2 antibodies, and anti-IL-2 receptor antibodies; anti-LFA-1 antibodies, including anti-CD11a and anti-CD18 antibodies; anti-L3T4 antibodies; heterologous antilymphocyte globulin; polyclonal or pan-T antibodies, or monoclonal anti-CD3 or anti-CD4/CD4a antibodies; soluble peptides containing the LFA-3 binding domain (WO 90/08187); streptokinase; TGF-β; streptodornase; RNA or DNA from the host; FK506; RS-61443; deoxyspergualin; rapamycin; T cell receptor (Cohen et al., U.S. Pat. No. 5,114,721); T cell receptor fragments (Offner et al., Science 2251:430-432 (1991); WO 90/11294; Ianeway, Nature 341:482 (1989); and WO 91/01133); T cell receptor antibodies (EP 340,109) such as T10B9; cyclophosphamide; dapsone; penicillamine; plasma exchange. exchange); or intravenous immunoglobulin (IVIG). The above can be used alone or in combination with each other, particularly a combination of a steroid and another immunosuppressant, or such a combination followed by a maintenance dose of a non-steroidal active agent to reduce the need for steroids.

"Analgesic" refers to a drug that acts to suppress or calm pain in a subject. Exemplary analgesics include nonsteroidal anti-inflammatory drugs (NSAIDs), including ibuprofen, naproxen, acetylsalicylic acid, indomethacin, sulindac, and tolmetin, including their salts and derivatives, as well as a variety of other drugs used to reduce the stinging pain that may occur, including anticonvulsants (gabapentin, phenyloin, carbamazepine) or tricyclic antidepressants. Specific examples include acetaminophen, aspirin, amitriptyline, carbamazepine, phenyltoin, gabapentin, (E)-N-vanillyl-8-methyl-6-nonenamide, or neuroleptics.

"Corticosteroid" refers to any of several synthetic or naturally occurring substances that have the general chemical structure of a steroid and that mimic or enhance the effects of naturally occurring corticosteroids. Examples of synthetic corticosteroids include prednisone, prednisolone (including methylprednisolone), dexamethasone, triamcinolone, and betamethasone.

As used herein, a "cancer vaccine" is a composition that stimulates an immune response against cancer in a subject. Cancer vaccines typically consist of a source of cancer-related material or cells (antigens) that is autologous (from the subject) or allogeneic (from another) to the subject, and other components (e.g., adjuvants) to further stimulate and promote an immune response against the antigens. Cancer vaccines can result in stimulation of the subject's immune system to produce antibodies against one or more specific antigens, and/or to produce killer T cells to attack cancer cells that harbor these antigens.

"Antiemetics" are compounds that reduce or prevent nausea in a subject. Antiemetic compounds include, for example, neurokinin-1 receptor antagonists, 5HT3 receptor antagonists (e.g., ondansetron, granisetron, tropisetron, and zatisetron), GABAB receptor agonists such as baclofen, corticosteroids such as dexamethasone, or antidopaminergic agents, phenothiazines (e.g., prochlorperazine, fluphenazine, thioridazine, and mesoridazine), dronabinol, metoclopramide, domperidone, haloperidol, lorazepam, prochlorperazine, and levomopromazine.

II. Construction of Single- and Multi-Chain Tethered Antibodies and Other Heteromultimers

Heteromultimers, including monospecific and multispecific (eg, bispecific) antibodies described herein, can be constructed through the use of one or more tethers.

The use of tethers allows the construction of relatively pure heteromultimer groups with different heavy chains and/or light chains within a single heteromultimer. Specifically, as described above, antibodies typically comprise two identical heavy chains, each paired with one of two identical light chains. The use of the tether technology of the present invention allows different antibody heavy chains to dimerize with each other in the formation of a single heteromultimer antibody formed by a single polypeptide chain. The tether can connect the first heavy chain and the second heavy chain, typically by connecting the CH3 domain of the first heavy chain and the VL domain of the light chain, and the light chain itself is connected to the second heavy chain through a second tether. In a heteromultimer single-chain antibody, a third tether can directly connect the second light chain and the first heavy chain, thereby producing a heteromultimer comprising two different heavy chains, each heavy chain can each be paired with its cognate light chain. In other embodiments, the tether can directly connect the first heavy chain and the second heavy chain by connecting the C-terminus of the first heavy chain constant domain and the N-terminus of the second heavy chain variable domain. Due to the presence of different heavy and light chain homologous pairs, each pair of heavy and light chains within such heteromultimers can have different binding specificities, and thus heteromultimers can be considered multispecific antibodies. The use of the tethering technology of the present invention also allows the formation of associated heteromultimers (e.g., multi-chain antibodies) comprising three polypeptide chains, wherein one polypeptide chain comprises two HCs directly linked together by an HD tether as described above, and the other two polypeptide chains are identical and form a LC, which associates with the two HCs to form two functional HC/LC homologous pairs. In other embodiments, the use of HD and CLH tethers allows the formation of multispecific antibodies with two polypeptide chains, wherein one polypeptide chain forms a first LC and the other polypeptide chain comprises two HCs and a second LC, the second LC being bound to the first HC via the HD tether and to the second HC via the CLH tether. In other embodiments, the use of HD and CLH tethers allows for the formation of multispecific antibodies with two polypeptide chains, wherein the first polypeptide chain comprises a first LC and two HCs, wherein one HC directly interacts with the first LC via the CLH tether and directly interacts with the second HC via the HD tether, and the second polypeptide forms a second LC that associates with the HC that is not attached to the first LC. In other embodiments, the use of the tether technology of the present invention allows for the formation of heteromultimers in which immunoadhesins are associated with half antibodies via the HD tether. Tether technology can be developed alone or in combination with knob-into-hole ("KnH") technology to engineer the heteromultimers of the present invention. Heteromultimers comprising tethers, with or without KnH technology, and the production of recombinant heteromultimers are described in detail below.

A. Heteropolymeric system chain

The present invention provides heteromultimers constructed with tethers. For example, a heteromultimer may have a tether connecting the C-terminus of an immunoglobulin heavy chain constant domain and the N-terminus of an immunoglobulin heavy chain variable domain. In other embodiments, the heteromultimer further comprises one or more (e.g., two) additional tethers to assist in the proper association of the heavy chain with its cognate light chain (i.e., the association of the heavy chain with the light chain to which it is connected). Such heteromultimers may be constructed with or without additional heteromultimerization domains, such as heteromultimers constructed using KnH technology.

As shown in the schematic diagram in Figure 1, an exemplary heteromultimer is a multispecific single-chain antibody comprising two different heavy chains (HC1 and HC2) and two different light chains (LC1 and LC2), wherein HC1 and LC1 form a first cognate pair and HC2 and LC2 form a second cognate pair. In the exemplary heteromultimer, there are three tethers. The first tether (HD tether) connects HC1 and LC2, and the second tether (CLH tether ₁ ) and the third tether (CLH tether ₂ ) connect LC1 and HC1 and LC2 and HC2, respectively. The tethers help bind the two different heavy chains and their cognate light chains together, thereby generating a heteromultimeric single-chain antibody with multispecificity.

In a specific embodiment, in the assembled heteromultimer ( FIG. 1 ), the HD tether is long enough to span the distance between the C-terminus of HCl and the N-terminus of LC2 (or HC2 in the absence of LC2) to allow proper association between the first and second heavy chains, but short enough to prevent intermolecular homomultimerization of HCl and/or HC2. The distance between the C-terminus of HCl and the N-terminus of LC2 (or HC2 in the absence of LC2) can vary between assembled heteromultimers, and thus the length of the HD tether can also vary between heteromultimers. Generally, HD tethers of 15-100 amino acids are effective, as are HD tethers of 20-40 amino acids, 25-40 amino acids, and 30-40 amino acids. In a specific embodiment, the HD tether is 36-39 amino acids long (e.g., 36, 37, 38, or 39 amino acids long).

In a specific embodiment, in the assembled heteromultimer (Figure 1), each CLH tether is long enough to span the distance between the N-terminus of the heavy chain and the C-terminus of its cognate light chain to allow proper light chain/heavy chain association, but short enough to prevent undesirable interchain association (i.e., association of a light chain with a non-cognate heavy chain to which it is not directly connected). The distance between the N-terminus of a heavy chain and the C-terminus of its cognate light chain can vary between antibodies, and even within an antibody itself (i.e., the distance between the N-terminus of the heavy chain of a first cognate pair and the C-terminus of its cognate light chain is different from that of a second cognate pair). Thus, the length of the HD tether can vary between heteromultimers of the invention, and within the same heteromultimer, the length of CLH tether ₁ can be different from the length of CLH tether ₂ . Typically, CLH tethers of 10-80 amino acids are effective, as are CLH tethers of 20-40 amino acids, 25-40 amino acids, 30-40 amino acids, and 35-40 amino acids (e.g., 30, 31, 32, 33, 34, or 35 amino acids). The heteromultimers of the present invention may have two tethers (CLH tether ₁ and CLH tether ₂ ) that bind the first light chain to the first heavy chain and the second light chain to the second heavy chain.

The HD or CLH tether can remain flexible and not form secondary structure, and for this purpose, tethers containing glycine (G) and serine (S) can be used. The tether can be composed solely of G and S residues, but other residues may also be included as long as the tether remains flexible to allow for the interchain association described above. In specific embodiments, the HD or CLH tether comprises GGS repeats ( FIG. 2 ). In some embodiments, the HD tether comprises at least one GGS repeat. An exemplary HD tether described herein comprises 8-9 GGS repeats (SEQ ID NO: 19) and endopeptidase cleavage sites (furin and Lys-C cleavage sites) at its N- and C-termini ( FIG. 2 ). In another embodiment, a CLH tether comprises at least one GGS repeat. An exemplary CLH tether described herein comprises 6 GGS repeats (SEQ ID NO: 20) and endopeptidase cleavage sites (furin) at its N- and C-termini ( FIG. 2 ).

B. Cleavage of heteropolymer chains

After the heteromultimers of the present invention are assembled, the tether is no longer needed, and in some embodiments, the tether is cleaved from the heteromultimer. A cleavage site that is present within or immediately adjacent to the tether, but not present in the sequence of the non-tethered heteromultimer components, or that cannot be cleaved under the conditions used, can be used to remove the tether.

Figure 2 illustrates the location of exemplary cleavage sites that may be present in the three tethers in an exemplary heteromultimeric single-chain antibody. Typically, the cleavage site within the tether is located at or near the C-terminus or N-terminus of the tether sequence, or within the antibody sequence at or near the site where the antibody and tether are bound. If one or more tethers are cut with a Lys-C endopeptidase (e.g., at the C-terminal lysine residue of the constant heavy chain), the sequence of the heteromultimer may need to be modified to remove the Lys-C endopeptidase cleavage site. An example of such modification is the mutation of lysine in the hinge region to alanine (e.g., K222A, Kabat numbering system in the exemplary heteromultimers described herein; K222A, EU numbering system). When different cleavage agents are selected for use in the present invention, modification of other cleavage sites may be required and performed in a similar manner.

Cutting the amino acid sequence at a specific site is standard in this area and can involve enzymatic cutting, chemical cutting or self-processing. For example, available endopeptidases can be used to cut the tether from the protein. Exemplary endopeptidases include but are not limited to furin, urokinase, Lys-C, Asp-N, Arg-C, Glu-C, thrombin, tissue plasminogen activator (tPa), genenase (variant of subtilisin BPN ' protease), factor Xa, TEV (tobacco etch virus protease), enterokinase, HRV C3 (human rhinovirus C3 protease), kininogenase, chymotrypsin, trypsin, pepsin and papain, and the above enzymes are all commercially available (for example, from Boehringer Mannheim, Thermo Scientific or New England Biolabs). Subtilisin-like proprotein convertases such as furin (PC1), PC2, or PC3 (Steiner (1991) in Peptide Biosynthesis and Processing (Fricker, ed.) pp. 1-16, CRC Press, Boca Raton, FL; Muller et al., JBC 275:39213-39222, (2000)) and N-arginine dibasic convertase (Chow et al., JBC 275:19545-19551 (2000)) cleave at the dibasic site. Lys-C cleaves at the carboxyl side of lysine residues, V8 and Glu-C cleave at the carboxyl side of glutamic acid residues, Arg-C cleaves at the carboxyl side of arginine residues, Asp-N cleaves at the amino side of aspartic acid residues, chymotrypsin cleaves at the carboxyl side of tyrosine, phenylalanine, tryptophan, and leucine residues, and trypsin cleaves at the carboxyl side of arginine and lysine residues. TEV cleaves the amino acid sequence GluAsnLeuTyrPheGlnGly (SEQ ID NO: 4) between the "Gln" and "Gly" residues. The use of such enzymes is standard in the art, and protocols are available from the manufacturer. FIG3 illustrates an exemplary heteromultimeric single-chain antibody after cleavage with furin.

Alternatively, the tether can be cleaved from the protein using chemicals such as hydroxylamine. Hydroxylamine cleaves the asparagine-glycine peptide bond. If hydroxylamine is used to cleave the tether from the protein, multiple glycine or asparagine residues in the protein may need to be mutated to avoid fragmenting the protein.

A variety of other chemicals that can cleave peptide bonds are known in the art. For example, N-chlorosuccinimide cleaves at the C-terminal side of tryptophan residues (Shechter et al., Biochemistry 15:5071-5075 (1976)). N-bromosuccinimide and cyanogen bromide also cleave at the C-terminal side of tryptophan residues. In addition, 2-nitrothiocyanate or organophosphines can be used to cleave proteins at the N-terminal side of cysteine (see, for example, EP 0339217).

Proteins are also known to self-process. For example, the Hedgehog protein is processed by its internal proteolytic activity at the GlyAspTrpAsnAlaArgTrpCysPhe cleavage site (SEQ ID NO: 5). An autologous proteolytic cleavage site may also be included in the linker or tether sequence.

After endopeptidase cleavage, the heteromultimers of the present invention can be further treated with one or more exopeptidases before or after purification. Figure 3B illustrates a heteromultimeric single-chain antibody after purification and treatment with furin, Lys-C, and an exopeptidase (e.g., carboxypeptidase B). After furin treatment, the HD tether still attached to the CH3 ₁ domain of the heteromultimer is removed by Lys-C treatment. Residual furin recognition sequences attached to the CL ₁ and CL ₂ domains are removed by treatment with an exopeptidase (e.g., carboxypeptidase B).

C. Knob-into-hole technology

The heteromultimers of the present invention may additionally comprise heterodimerization domains using knob-into-hole (KnH) technology (see, e.g., U.S. Patent No. 5,731,168, incorporated herein by reference in its entirety), which can engineer the interface between two different antibody heavy chains to promote their association. Typically, this technology involves introducing a "knob" at the interface of the first heavy chain and a corresponding "hole" at the interface of the second heavy chain, such that the knob can be placed in the hole to further promote heterodimerization of the heavy chains.

The first preferred interface comprises at least a portion of the CH3 domain of the first heavy chain constant domain (CH3 ₁ ) and at least a portion of the CH3 domain of the second heavy chain constant domain (CH3 ₂ ). Protrusions can be constructed by replacing small amino acid side chains from the interface of the first domain (e.g., CH3 ₁ ) with larger side chains (e.g., tyrosine or tryptophan). Compensatory cavities of the same or similar size as the protrusions are optionally constructed at the interface of the second domain (e.g., CH3 ₂ ) by replacing large amino acid side chains with smaller side chains (e.g., alanine or threonine). The second preferred interface comprises at least a portion of the CL domain of the light chain and the CH1 domain of the heavy chain, where the protrusion-cavity interaction can be constructed as described above. When appropriately positioned and spaced protrusions or cavities exist at the interface of the first or second domain, it is only necessary to process the corresponding cavities or protrusions at the adjacent interfaces.

D. Monospecific or multispecific heteromultimers with different binding properties

It will be appreciated that the variable domains of such heteromultimers can be derived from a variety of methods. For example, the variable domains of the heteromultimers of the present invention can be identical to existing antibodies known in the art.

As described above, the HD tethers of the present invention can be used to generate heteromultimers with different binding properties. In some embodiments, the present invention can be used to generate single-chain monospecific antibodies comprising two half-antibodies with different binding affinities for the same target epitope. In another embodiment, the present invention can be used to generate single-chain multispecific antibodies (antibodies that bind to at least two antigens or at least two different epitopes on the same antigen). Typically, in naturally occurring IgG antibodies, the variable regions of each pair of heavy and light chains in the antibody are identical. The use of tethers according to the present invention allows the two heavy chains in an antibody to be different, resulting in antibodies with antigen-binding domains with different binding specificities (e.g., epitope targets) or antigen-binding domains with the same binding specificity but different binding affinities. In some embodiments, the HD tether promotes the association of two different heavy chains encoded in a single polypeptide chain to obtain a heavy chain-only single-chain antibody. In other embodiments, the CLH tether promotes the association of a heavy chain with its cognate light chain. In some embodiments, the association of one of the heavy chains with its cognate light chain is established in the absence of the CLH tether to generate the heteromultimers of the present invention. Optionally, one or more tethers comprise an endopeptidase cleavage site (e.g., furin and Lys-C cleavage sites) that can be cleaved to allow removal of one or more tethers from the heteromultimer after assembly. Optionally, the heteromultimers of the invention further comprise one or more protuberance-cavity interfaces generated using KnH technology.

As shown in Figure 1, the exemplary bispecific heteromultimer (in this case, a bispecific antibody) described above contains an HD tether and two CLH tethers, each of which may contain two furin cleavage sites at its N-terminus and C-terminus. Optionally, the exemplary heteromultimer can utilize KnH technology to further promote the association of HCl and HC2.

The heteromultimer intended to cleave the tether should not contain within its sequence a cleavage site for an endopeptidase that cleaves the tether, or if there is a cleavage site within the non-tethered sequence of the heteromultimer, the cleavage site should not be cleaved under the conditions used unless the heteromultimer is also intended to be cleaved. The sequence of the heteromultimer can be scanned to determine if there are any cleavage sites (e.g., furin or Lys-C endopeptidase cleavage sites) in the heavy or light chain sequences of the heteromultimer that should be removed to avoid cleavage of the heteromultimer itself upon removal of the tether.

Single-chain monospecific or multispecific heteromultimers (e.g., antibodies) can be constructed using the methods described herein, wherein the two different heavy chains of the heteromultimer are directly connected to each other via an HD tether. In one embodiment, the single-chain heavy chain-only monospecific or multispecific heteromultimer may additionally lack a CH1 domain (the VH domain is directly connected to at least one CH domain, optionally via a hinge). In another embodiment, the single-chain monospecific or multispecific heteromultimer heavy chain lacks a CH1 domain but is associated with its corresponding light chain lacking a CL domain via a CLH tether. Such heteromultimers can be used to bind two different antigens together, or to associate B cells and T cells. In another embodiment, the tether can be used to associate half antibodies and immunoadhesins.

Furthermore, the present invention encompasses heteromultimers with monospecificity or multispecificity.In one embodiment, a first heavy chain linked to its cognate light chain via a CLH tether associates with a second heavy chain, the cognate light chain of which is associated independently of the tether.

E. Heteromultimers with different effector functions

In some embodiments, the heteromultimers of the present invention (e.g., single-chain or multi-chain antibodies, or antibody-immunoadhesin complexes) constructed using the methods described herein may comprise CH2 domain mutations, allowing for altered effector functions. Typically, the CH2 domain mutation is a mutation at N297, resulting in a heteromultimer with an altered glycosylation state. In certain embodiments, the N297 mutation is an N297A substitution. Altered effector functions include, but are not limited to, C1q binding and complement-dependent cytotoxicity, Fc receptor binding, antibody-dependent cell-mediated cytotoxicity (ADCC), phagocytosis, downregulation of cell surface receptors (e.g., B cell receptors), and B cell activation.

F. Conjugated Protein Complex

As described in detail below, the HD tethers of the present invention can also be used to generate protein complexes such as heteromultimers (e.g., monospecific, bispecific, multispecific single-chain or multi-chain antibodies) in which the constant regions are modified by conjugation to a cytotoxic agent. For example, the HD tether allows the construction of heteromultimers in which one of the heavy chain constant regions (HC1 or HC2) contains a modification that allows conjugation to a cytotoxic agent, while the other heavy chain constant region does not contain such a modification. In one example, HC1 is conjugated to a cytotoxic agent, while HC2 is not conjugated. Figure 4 shows a schematic diagram illustrating an example of a conjugated heteromultimer of the present invention. An exemplary heteromultimeric single-chain antibody comprises two full-length heavy chains and a light chain of the same family, as well as the HD and CLH tethers described above. As shown in Figure 4, one of the heavy chains has been conjugated to a cytotoxic agent (e.g., a toxin or antibiotic). Similarly, in an alternative heteromultimer construct, one of the light chain constant regions can be conjugated to a cytotoxic agent, while the other light chain constant region is not conjugated (e.g., LC1 is conjugated to a cytotoxic agent, and LC2 is not conjugated). As shown in Figure 5A, the conjugated protein complex can be a multi-chain antibody of the present invention. The conjugation of cytotoxic agents is not limited to the specific multi-chain antibody shown in Figure 5A. Multi-chain antibodies, such as those shown in Figures 5B and 5C, and immunoadhesin-antibody heteromultimers can also be conjugated at the constant region.

In one embodiment, the constant region of the heteromultimer can be modified to introduce an electrophilic moiety that can react with a nucleophilic substituent on a linker reagent used to conjugate the cytotoxic agent to the heteromultimer or on the cytotoxic agent itself. The sugars of the glycosylated antibody can be oxidized, for example, with periodate oxidizing agents, to form aldehyde or ketone groups that can react with amine groups of the linker reagent or cytotoxic agent. The resulting imine Schiff base group can form a stable bond or be reduced, for example, with a borohydride reagent, to form a stable amine bond. Nucleophilic groups on the cytotoxic agent include, but are not limited to, amine, sulfhydryl, hydroxyl, hydrazide, oxime, hydrazine, thiosemicarbazone, hydrazine carboxylate, and arohydrazide groups that can react with electrophilic groups on the antibody region and linker reagents to form covalent bonds, including: (i) active esters such as NHS esters, HOBt esters, haloformates, and acid halides; (ii) alkyl and benzyl halides such as haloacetamides; and (iii) aldehyde, keto, carboxyl, and maleimide groups.

G. Other protein characteristics

Heteromultimers according to the present invention may comprise sequences from any source, including human or murine, or a combination thereof. The sequences of certain portions of the protein (e.g., hypervariable regions) may also be artificial sequences, such as sequences identified by screening a library containing random sequences (e.g., a phage display library).

In the case of heteromultimers comprising heteromultimers from different sources, the heteromultimers can be "chimeric" heteromultimers in which a portion of the heavy and/or light chains is identical or homologous to the corresponding sequence in an antibody derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain is identical or homologous to the corresponding sequence in an antibody derived from another species or termed another antibody class or subclass, as well as fragments of such antibodies, as long as they exhibit the desired biological activity (U.S. Patent No. 4,816,567; and Morrison et al., Proc. Natl. Acad. Sci. USA 81: 6851-6855 (1984)). Such chimeric heteromultimers can comprise, for example, a murine variable region (or portion thereof) and a human constant region.

Chimeric single-chain heteromultimers may also optionally be "humanized" single-chain heteromultimers (e.g., single-chain antibodies) that contain minimal sequences derived from non-human antibodies. Humanized heteromultimers are typically derived from human antibodies (receptor antibodies), wherein the residues from the hypervariable region of the receptor are replaced by residues from a hypervariable region of a non-human species (donor antibody) (e.g., mouse, rat, rabbit, or non-human primate) with the desired specificity, affinity, and performance. In some cases, residues in the human immunoglobulin framework region (FR) are replaced by corresponding non-human residues. In addition, humanized heteromultimers may include residues that are not present in the receptor antibody or the donor antibody. Such modifications are performed to further improve heteromultimer performance. Generally, humanized heteromultimers will include at least one, typically two, substantially all variable domains, wherein all or substantially all of the hypervariable loops correspond to the hypervariable loops of non-human immunoglobulins, and all or substantially all of the FRs are FRs of human immunoglobulin sequences. The humanized heteromultimers also optionally contain at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin. For more details, see Jones et al., Nature 321:522-525 (1986); Riechmann et al., Nature 332:323-329 (1988) and Presta, Curr. Op. Struct. Biol. 2:593-596 (1992).

More specifically, humanized heteropolymers may have one or more amino acid residues introduced therein from a non-human source. These non-human amino acid residues are generally referred to as "input" residues, which are generally taken from the "input" variable domain. Humanization can be performed essentially according to the method of Winter and colleagues (Jones et al., Nature 321:522-525 (1986); Riechmann et al., Nature 332:323-327 (1988); Verhoeyen et al., Science 239:1534-1536 (1988)), by replacing the corresponding sequence of a human antibody with one or more CDR sequences of a rodent. Therefore, such "humanized" heteropolymers are chimeric antibodies (U.S. Patent number 4,816,567), in which substantially less than a complete human variable domain has been replaced by a corresponding sequence from a non-human species. In practice, humanized heteromultimers are typically human antibodies in which some CDR residues and possibly some FR residues are substituted by residues from analogous sites in rodent antibodies.

The selection of human light chain and heavy chain variable domains used in the preparation of humanized heteromultimers is very important for reducing antigenicity. According to the so-called "best fit" method, the sequence of the variable domains of rodent antibodies is screened for the library of known human variable domain sequences. The human sequence closest to the rodent sequence is used as the human framework (FR) of the humanized heteromultimer (Sims et al., J.Immunol.151:2296 (1993); Chothia et al., J.Mol.Biol.196:901 (1987)). Another method uses a specific framework of the consensus sequence of all human antibodies of the light chain or heavy chain derived from a specific subgroup. The same framework can be used for multiple different humanized heteromultimers (Carter et al., Proc.Natl.Acad.Sci.USA 89:4285 (1992); Presta et al., J.Immnol.151:2623 (1993)).

More importantly, the heteromultimers retain high affinity and other favorable biological properties for one or more target antigens when humanized. To achieve this goal, according to an exemplary method, humanized heteromultimers were prepared by analyzing the parental sequences of three-dimensional models of the parental and humanized sequences and multiple conceptual humanized products. Three-dimensional immunoglobulin models are generally available and are familiar to those skilled in the art. Computer programs are available that illustrate and display the possible three-dimensional conformational structures of selected candidate immunoglobulin sequences. Inspection of such displays allows analysis of the possible role of the residues in the function of the candidate immunoglobulin sequence, that is, analysis of the residues that affect the ability of the candidate immunoglobulin to bind to its antigen. In this way, FR residues can be selected and combined from the receptor and input sequences to obtain the desired heteromultimer characteristics, such as increased affinity for one or more target antigens. In general, CDR residues are directly and most substantially involved in affecting antigen binding.

III. Vectors, Host Cells, and Recombinant Methods

To recombinantly produce the heteromultimers of the present invention, the nucleic acid encoding the heteromultimer is separated and inserted into a replicable vector for further cloning (amplification of DNA) or for expression. The DNA encoding the heteromultimer is conveniently separated and sequenced using conventional procedures (e.g., by using oligonucleotide probes that specifically bind to genes encoding the heavy and light chains of the heteromultimer). Many vectors are available. The selection of the vector depends in part on the host cell to be used. Typically, preferred host cells are prokaryotic or eukaryotic (usually mammalian, but also including fungi (e.g., yeast), insects, plants, and nucleated cells from other multicellular organisms) sources. It should be understood that any isotype of constant region can be used for this purpose, including IgG, IgM, IgA, IgD, and IgE constant regions, and such constant regions can be obtained from any human or animal species.

A. Production of Heteromultimers Using Prokaryotic Host Cells

i. Vector Construction

The polynucleotide sequences encoding the polypeptide components of the heteromultimers of the present invention can be obtained using standard recombinant techniques. The desired polynucleotide sequences can be isolated and sequenced from cells producing antibodies, such as hybridoma cells. Alternatively, the polynucleotides can be synthesized using a nucleotide synthesizer or PCR technology. Once obtained, the sequence encoding the polypeptide is inserted into a recombinant vector capable of replicating and expressing heterologous polynucleotides in a prokaryotic host. Many vectors available and known in the art can be used for the purposes of the present invention. The selection of an appropriate vector will depend primarily on the size of the nucleic acid to be inserted into the vector and the specific host cell to be transformed with the vector. Each vector comprises multiple components depending on its function (amplification or expression of heterologous polynucleotides, or both) and its compatibility with the specific host cell in which it is located. Vector components typically include, but are not limited to, an origin of replication, a selectable marker gene, a promoter, a ribosome binding site (RBS), a signal sequence, a heterologous nucleic acid insert, and a transcription termination sequence.

In general, plasmid vectors containing replicons and control sequences derived from species compatible with the host cell are used in conjunction with such hosts. Carriers typically carry replication sites and marker sequences that can provide trait selection in transformed cells. For example, Escherichia coli is typically transformed with the plasmid pBR322 derived from Escherichia coli species. pBR322 comprises genes encoding ampicillin (Amp) and tetracycline (Tet) resistance, and therefore provides a simple method for identifying transformed cells. pBR322, its derivatives or other microbial plasmids or phages may also comprise or be modified to comprise a promoter that can be used by microorganisms to express endogenous proteins. Examples of pBR322 derivatives for expressing specific heteromultimeric antibodies are described in detail in Carter et al., U.S. Patent number 5,648,237.

In addition, phage vectors containing replicons and control sequences compatible with the host microorganism can be used as transformation vectors for use with such hosts. For example, bacteriophages such as λGEM-11 ^™ can be used to prepare recombinant vectors that can be used to transform susceptible host cells such as E. coli LE392.

The expression vectors of the present invention may comprise two or more promoter-cistron pairs encoding each polypeptide component. A promoter is a non-translated regulatory sequence located upstream (5') of a cistron that regulates its expression. Prokaryotic promoters are generally divided into two categories, inducible and constitutive. An inducible promoter is a promoter that initiates increased levels of transcription of the cistron under its control in response to changes in culture conditions, such as the presence or absence of nutrients or changes in temperature.

A large number of promoters that are well known to be recognized by a variety of potential host cells. DNA can be removed from the source DNA by restriction enzyme digestion, and the isolated promoter sequence is inserted into the vector of the present invention, and the promoter of selection is effectively connected to the cistron DNA encoding the light chain or heavy chain. Natural promoter sequences and many heterologous promoters can be used to instruct the amplification and/or expression of the target gene. In some embodiments, heterologous promoters are used because they usually allow more transcription and higher yields of the target gene expressed compared with the natural target polypeptide promoter.

Promoters suitable for use with prokaryotic hosts include the PhoA promoter, beta galactamase and lactose promoter systems, tryptophan (trp) promoter systems, and hybrid promoters such as the tac or trc promoters. However, other promoters that are functional in bacteria (e.g., other known bacterial or phage promoters) are also suitable. The nucleotide sequence has been published, allowing one skilled in the art to link it to the cistron encoding the target light and heavy chains (Siebenlist et al., (1980) Cell 20:269), using linkers or adapters to provide any desired restriction sites.

In one aspect of the invention, each cistron in the recombinant vector comprises a secretory signal sequence component that instructs the transmembrane translocation of the expressed polypeptide. Typically, the signal sequence can be a component of the vector, or it can be a part of the target polypeptide DNA inserted into the vector. The signal sequence selected for the purpose of the present invention should be a signal sequence that is recognized and processed (i.e., cut by a signal peptidase) by the host cell. For prokaryotic host cells that do not recognize and process the natural signal sequence for heterologous polypeptides, the signal sequence is replaced with a prokaryotic signal sequence selected from, for example, alkaline phosphatase, penicillinase, Ipp or heat-stable enterotoxin II (STII) guide sequence, LamB, PhoE, PelB, OmpA and MBP. In one embodiment of the invention, the signal sequences used in the two cistrons of the expression system are all STII signal sequences or variants thereof.

On the other hand, the production of heteromultimers according to the present invention can be carried out in the cytoplasm of the host cell, and therefore does not require the presence of a secretion signal sequence in each cistron. In this regard, the heteromultimer light and heavy chains are expressed, folded, and assembled in the cytoplasm to form functional heteromultimers. Certain host strains (e.g., E. coli trxB strains) provide a cytoplasmic environment that is conducive to disulfide bond formation, thereby allowing proper folding and assembly of the expressed protein subunits (Proba and Pluckthun, Gene, 159:203 (1995)).

Prokaryotic host cells suitable for expressing the heteromultimers of the present invention include archaebacteria and eubacteria, such as gram-negative and gram-positive organisms. Examples of useful bacteria include Escherichia (Escherichia) (e.g., Escherichia coli), Bacilli (Bacilli) (e.g., Bacillus subtilis (B.subtilis)), Enterobacteria (Enterobacteria), Pseudomonas (Pseudomonas) (e.g., Pseudomonas aeruginosa (P.aeruginosa)), Salmonella typhimurium (Salmonella typhimurium), Serratia marcescans (Serratia marcescans), Klebsiella (Klebsiella), Proteus (Proteus), Shigella (Shigella), Rhizobium (Rhizobia), Vitreoscilla (Vitreoscilla) or Paracoccus (Paracoccus). In one embodiment, gram-negative bacteria are used. In one embodiment, Escherichia coli is used as a host of the present invention. Examples of E. coli strains include strain W3110 (Bachmann, Cellular and Molecular Biology, Vol. 2 (Washington, D.C.: American Society for Microbiology, 1987), pp. 1190-1219; ATCC Deposit No. 27,325) and derivatives thereof, including strain 33D3 (U.S. Pat. No. 5,639,635) having the genotype W3110ΔfhuA(ΔtonA)ptr3lacIqlacL8ΔompTΔ(nmpc-fepE)degP41kanR. Other strains and their derivatives, such as E. coli 294 (ATCC 31,446), E. coli B, E. coli λ1776 (ATCC 31,537), and E. coli RV308 (ATCC 31,608), are also suitable. The above examples are illustrative and non-limiting. Methods for constructing derivatives of any of the above-mentioned bacteria with defined genotypes are known in the art and are described, for example, in Bass et al., Proteins 8:309-314 (1990). It is generally necessary to consider the replicability of the replicon within the bacterial cell and select an appropriate bacterium. For example, when using well-known plasmids such as pBR322, pBR325, pACYC177, or pKN410 to provide the replicon, Escherichia coli, Serratia, or Salmonella species may be used as hosts as appropriate. Generally, the host cell should secrete minimal amounts of proteolytic enzymes, and ideally, additional protease inhibitors are incorporated into the cell culture.

ii. Heteromultimer production

Host cells are transformed with the above-described expression vectors and cultured in conventional nutrient media modified as appropriate to induce promoters, select transformants, or amplify the genes encoding the desired sequences.

Transformation refers to the introduction of DNA into a prokaryotic host so that the DNA can be replicated as an extrachromosomal element or as a chromosomal integrant. Depending on the host cell used, transformation is carried out using standard techniques suitable for such cells. Calcium treatment using calcium chloride is commonly used for bacterial cells containing substantial cell wall barriers. Another transformation method uses polyethylene glycol/DMSO. Another technology used is electroporation.

Prokaryotic cells used to produce the heteromultimers of the present invention are cultured in a culture medium known in the art and suitable for culturing the selected host cells. Examples of suitable culture media include Luria broth (LB) plus necessary nutritional supplements. In some embodiments, the culture medium also contains a selective agent selected based on the construction of the expression vector to selectively allow the growth of prokaryotic cells containing the expression vector. For example, ampicillin is added to the culture medium to cultivate cells expressing an ampicillin resistance gene.

Any necessary supplements other than the carbon source, nitrogen source, and inorganic phosphate source may also be included at appropriate concentrations, introduced alone or as a mixture with another supplement or culture medium (e.g., a complex nitrogen source). Optionally, the culture medium may contain one or more reducing agents selected from glutathione, cysteine, cystamine, thioglycolate, dithioerythritol, and dithiothreitol.

Prokaryotic host cells are cultured at an appropriate temperature. For example, for the cultivation of E. coli, the preferred temperature range is from about 20°C to about 39°C, more preferably from about 25°C to about 37°C, and even more preferably at about 30°C. The pH of the culture medium can range from about 5 to about 9, depending primarily on the host organism. For E. coli, the pH is preferably from about 6.8 to about 7.4, more preferably about 7.0.

If an inducible promoter is used in the expression vector of the present invention, protein expression is induced under conditions suitable for activating the promoter. In one aspect of the present invention, the PhoA promoter is used to control transcription of the polypeptide. Thus, the transformed host cells are cultured in a phosphate-limited medium for induction. Preferably, the phosphate-limited medium is C.R.A.P medium (see, e.g., Simmons et al., J. Immunol. Methods (2002), 263:133-147). As is known in the art, a variety of other inducers can be used, depending on the vector construct used.

In one embodiment, the expressed heteromultimers of the present invention are secreted into the periplasm of the host cell and recovered. Protein recovery generally involves destroying the microorganism, typically using methods such as osmotic shock, ultrasound or lysis. After the cells are destroyed, cell debris or whole cells can be removed by centrifugation or filtration. The heteromultimers can be further purified, for example, by affinity resin chromatography. Alternatively, the protein can be transported to and separated from the culture medium. The cells can be removed from the culture, and the culture supernatant can be filtered and centrifuged to further purify the produced protein. The expressed polypeptide can be further separated and identified by well-known methods such as polyacrylamide gel electrophoresis (PAGE) and Western blotting.

In one aspect of the invention, heteromultimer production is carried out in large quantities using fermentation processes. A variety of large-scale fed-batch fermentation procedures can be used to produce recombinant proteins. Large-scale fermentation has a capacity of at least 1000 liters, preferably a capacity of about 1000 to 100,000 liters. These fermentors use stirring paddles to distribute oxygen and nutrients, particularly glucose (a preferred carbon source/energy source). Small-scale fermentation generally refers to fermentation in a fermentor tank of no more than about 100 liters of capacity, and can range from about 1 liter to about 100 liters.

In fermentation processes, induction of protein expression is typically initiated after cells have grown to a desired density under appropriate conditions, for example an OD550 of about 180-220, at which stage the cells are in the early stationary phase. As known in the art and as described above, various other inducers may be used, depending on the vector construct used. Cells may be cultured for a shorter period of time prior to induction. Although longer or shorter induction times may be used, cells are typically induced for about 12-50 hours.

To improve the yield and quality of the heteromultimers of the present invention, various fermentation conditions can be modified. For example, to improve the proper assembly and folding of the secreted heteromultimers, the host prokaryotic cells can be co-transformed with additional vectors that overexpress chaperone proteins, such as Dsb proteins (DsbA, DsbB, DsbC, DsbD, and/or DsbG) or FkpA (a peptidylprolyl cis, trans isomerase with chaperone activity). Molecular chaperone proteins have been shown to promote proper folding and solubility of heterologous proteins produced in bacterial host cells (Chen et al., (1999) J. Biol. Chem. 274: 19601-19605; Georgiou et al., U.S. Pat. No. 6,083,715; Georgiou et al., U.S. Pat. No. 6,027,888; Bothmann and Pluckthun (2000) J. Biol. Chem. 275: 17100-17105; Ramm and Pluckthun, (2000) J. Biol. Chem. 275: 17106-17113; Arie et al., (2001) Mol. Microbiol. 39: 199-210).

To minimize proteolysis of expressed heterologous proteins (particularly those that are sensitive to proteolysis), certain proteolytic enzyme-deficient host strains can be used in the present invention. For example, the host cell strain can be modified to produce (one or more) genetic mutations in genes encoding known bacterial proteases such as Protease III, OmpT, DegP, Tsp, Protease I, Protease Mi, Protease V, Protease VI, and combinations thereof. Some E. coli protease-deficient strains are available and are described, for example, in Joly et al., (1998), Proc. Natl. Acad. Sci. USA 95: 2773-2777; Georgiou et al., U.S. Patent No. 5,264,365; Georgiou et al., U.S. Patent No. 5,508,192; Hara et al., Microbial Drug Resistance, 2: 63-72 (1996).

In one embodiment, an E. coli strain deficient in proteolytic enzymes and overexpressing one or more molecular chaperone proteins is used as a host cell for the expression system of the present invention.

In another embodiment, the E. coli cells additionally express an endopeptidase (e.g., furin) to cleave one or more tethers of the heteromultimer of interest prior to purification. In another embodiment, the eukaryotic host cells may express an endopeptidase (e.g., furin) and an exopeptidase (e.g., carboxypeptidase B), wherein the endopeptidase cleaves one or more tethers of the heteromultimer prior to purification and the exopeptidase degrades any residual endopeptidase cleavage sites.

iii. Heteromultimer purification

Standard protein purification methods known in the art can be used. The following procedures are exemplary of suitable purification procedures: fractionation on immunoaffinity or ion exchange columns, ethanol precipitation, reversed-phase HPLC, chromatography on silica or cation exchange resins such as DEAE, chromatofocusing, SDS-PAGE, hydrophobic interaction columns (HIC), ammonium sulfate precipitation, and chromatography using, for example, Sephadex G-75 gel.

In one aspect, protein A immobilized on a solid phase is used for immunoaffinity purification of the full-length heteromultimer product of the present invention. Protein A is a 41 kD cell wall protein from Staphylococcus aureus that binds to the Fc region of antibodies with high affinity. Lindmark et al., (1983) J. Immunol. Meth. 62: 1-13. The solid phase to which protein A is immobilized is preferably a column comprising a glass or silica surface, more preferably a controlled empty glass column or a silicic acid column. In some applications, the column has been coated with a reagent such as glycerol to prevent nonspecific attachment of contaminants.

As the first step of purification, the preparation derived from the cell culture as described above is applied to a protein A-immobilized solid phase to allow specific binding of the target heteromultimer to protein A. The solid phase is then washed to remove contaminants that are non-specifically bound to the solid phase. The target heteromultimer can be recovered from the solid phase by eluting into a solution containing a chaotropic agent or a mild detergent. Exemplary chaotropic agents include, but are not limited to, urea, guanidine-HCl, lithium perchlorate, histidine, and arginine. Exemplary mild detergents include, but are not limited to, Tween (e.g., Tween-20), Triton (e.g., Triton X-100), NP-40 (nonylphenoxypolyethoxyethanol), Nonidet P-40 (octylphenoxypolyethoxyethanol), and sodium dodecyl sulfate (SDS). After elution from a column (e.g., a mAbSure column), the heteromultimer is diluted into a solution containing a chaotropic agent or a mild detergent to maintain the stability of the heteromultimer after elution. In some embodiments, one or more residual endopeptidase cleavage sites of the heteromultimers of the invention that have been treated with endopeptidase can be treated and degraded after purification by adding an exopeptidase (e.g., carboxypeptidase B). In other embodiments, one or more tethers of the heteromultimers of the invention can be cleaved and treated with an endopeptidase (e.g., furin) and an exopeptidase (e.g., carboxypeptidase B) after purification, rather than before purification.

B. Production of Heteromultimers Using Eukaryotic Host Cells

Vector components generally include, but are not limited to, one or more of the following: a signal sequence, an origin of replication, one or more marker genes, an enhancer element, a promoter, and a transcription termination sequence.

i. Signal sequence components

Vectors for eukaryotic host cells may contain a signal sequence or other polypeptide with a specific cleavage site at the N-terminus of the target mature protein or polypeptide. The heterologous signal sequence selected should be recognized and processed by the host cell (i.e., cleaved by a signal peptidase). For mammalian cell expression, mammalian signal sequences and viral secretion leader sequences, such as the herpes simplex gD signal, are available. The DNA of such a precursor region is ligated in frame with the DNA encoding the antibody.

ii. Origin of replication

Typically, mammalian expression vectors do not require an origin of replication component. For example, the SV40 origin is commonly used, but only because it contains the early promoter.

iii. Selecting genomic components

Expression and cloning vectors may contain selection genes, also known as selectable markers. Typical selection genes encode proteins that (a) confer resistance to antibiotics or other toxins, such as ampicillin, neomycin, methotrexate, or tetracycline, (b) compensate for the deficiency of the associated auxotrophic defect, or (c) provide a critical nutrient not found in complex culture media.

One example of a selection scheme utilizes a drug to arrest the growth of host cells. Cells successfully transformed with a heterologous gene produce a protein that confers drug resistance, thereby surviving the selection process. Examples of this type of dominant selection utilize the drugs neomycin, mycophenolic acid, and hygromycin.

Another example of a suitable selectable marker for mammalian cells is one that allows identification of cells that can take up heteromultimeric nucleic acids, such as DHFR, thymidine kinase, metallothionein-I and metallothionein-II, preferably metallothionein genes, adenylate deaminase, ornithine decarboxylase, etc.

For example, cells transformed with the DHFR selection gene are first identified by culturing all transformants in a medium containing methotrexate (Mtx), a competitive antagonist of DHFR. An appropriate host cell when using wild-type DHFR is a Chinese hamster ovary (CHO) cell line deficient in DHFR activity (e.g., ATCC CRL-9096).

Alternatively, host cells (particularly wild-type host cells containing endogenous DHFR) transformed or co-transformed with DNA sequences encoding heteromultimers, wild-type DHFR protein and another selectable marker (e.g., aminoglycoside 3'-phosphotransferase (APH)) can be selected by culturing the cells in a medium containing a selection agent for the selectable marker (e.g., an aminoglycoside antibiotic, such as kanamycin, neomycin, or G418). See, e.g., U.S. Patent No. 4,965,199.

iv. Promoter components

Expression and cloning vectors typically contain a promoter that is recognized by the host organism and is effectively connected to one or more nucleic acid sequences encoding the heteropolymer. Promoters of known eukaryotic organisms. Almost all eukaryotic genes have an AT-rich region located about 25 to 30 bases upstream of the transcription start site. Another sequence located 70 to 80 bases upstream of the transcription start site of many genes is the CNCAAT region (SEQ ID NO: 12), in which N can be any nucleotide. At the 3' end of most eukaryotic genes is the AATAAA sequence (SEQ ID NO: 13), which can be a signal for adding a poly A tail to the 3' end of the coding sequence. The above sequences are all appropriately inserted into eukaryotic expression vectors.

Transcription from vectors encoding heteromultimers in mammalian host cells is controlled by promoters obtained, for example, from the genome of viruses such as, for example, polyoma virus, fowlpox virus, adenovirus (e.g., adenovirus 2), bovine papilloma virus, avian sarcoma virus, cytomegalovirus, retrovirus, hepatitis B virus, and simian virus 40 (SV40), from heterologous mammalian promoters such as the actin promoter or immunoglobulin promoters, or from heat shock promoters, so long as such promoters are compatible with the host cell systems.

The early and late promoters of the SV40 virus are conveniently obtained as SV40 restriction fragments that also contain the SV40 viral origin of replication. The immediate early promoter of the human cytomegalovirus is conveniently obtained as a HindIII E restriction fragment. U.S. Patent No. 4,419,446 discloses a system for expressing DNA in mammalian cells using bovine papilloma virus as a vector. A modification of this system is described in U.S. Patent No. 4,601,978. Alternatively, the Rous sarcoma virus long terminal repeat can be used as a promoter.

v. Enhancer element components

The transcription of DNA encoding one or more heteromultimeric polypeptides by higher eukaryotes can be enhanced by inserting an enhancer sequence in the vector. Currently, many enhancer sequences from mammalian genes (such as globulin, elastase, albumin, α-fetoprotein and insulin genes) are known. Enhancers from eukaryotic cell viruses can also be used. Examples include the SV40 enhancer located at the late side of the replication origin (bp100-270), the cytomegalovirus early promoter enhancer, the polyoma enhancer at the late side of the replication origin, and the adenovirus enhancer. See also Yaniv, Nature 297:17-18 (1982) for an explanation of elements that enhance the activation of eukaryotic promoters. The enhancer can be spliced into the 5' or 3' position of the heteromultimeric polypeptide coding sequence in the vector, as long as enhancement is achieved, but is usually located at the 5' site of the promoter.

vi. Transcription termination components

Expression vectors used in eukaryotic host cells also typically contain sequences required for transcription termination and mRNA stabilization. Such sequences are commonly obtained from the 5', and occasionally 3', untranslated regions of eukaryotic or viral DNA or cDNA. These regions contain nucleotide segments that are transcribed as polyadenylated fragments in the untranslated portion of the mRNA encoding the heteromultimer. One useful transcription termination component is the bovine growth hormone polyadenylation region. See WO 94/11026 and the expression vectors disclosed therein.

vii. Host Cell Selection and Transformation

Suitable host cells for cloning or expressing the DNA in the vectors herein include the higher eukaryotic cells described herein, including vertebrate host cells. Propagation of vertebrate cells in culture (tissue culture) is a routine procedure. Examples of useful mammalian host cells include monkey kidney CV1 cell line transformed with SV40 (COS-7, ATCC CRL1651); human embryonic kidney cell line (293 or 293 cells subcloned for culture in suspension culture, Graham et al., J. Gen. Virol. 36:59 (1977)); baby hamster kidney cells (BHK, ATCC CCL10); Chinese hamster ovary cells/-DHFR (CHO, Urlaub et al., Proc. Natl. Acad. Sci. USA 77:4216 (1980)); mouse Sertoli cells (TM4, Mather, Biol. Reprod. 23:243-251 (1980)); monkey kidney cells (CV1 ATCC CCL 70); African green monkey kidney cells (VERO-76, ATCC CRL-1587); human cervical carcinoma cells (HELA, ATCC CCL 2); canine kidney cells (MDCK, ATCC CCL34); Buffalo rat liver cells (BRL 3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL75); human hepatocytes (Hep G2, HB 8065); mouse mammary tumor (MMT 060562, ATCC CCL51); TRI cells (Mather et al., Annals N.Y. Acad. Sci. 383:44-68 (1982)); MRC5 cells; FS4 cells; and a human hepatoma cell line (Hep G2).

Host cells are transformed with the expression or cloning vectors described above for heteromultimer production and cultured in conventional nutrient media modified as appropriate to induce promoters, select transformants, or amplify the genes encoding the desired sequences.

viii. Cultivating Host Cells

Host cells for producing the heteromultimers of the present invention can be cultured in a variety of culture media. Commercially available culture media such as Ham's F10 (Sigma), Minimal Essential Medium ((MEM) (Sigma), RPMI-1640 (Sigma) and Dulbecco's Modified Eagle's Medium ((DMEM), Sigma) are suitable for culturing host cells. In addition, Ham et al., Meth.Enz.58:44 (1979), Barnes et al., Anal.Biochem.102:255 (1980), U.S. Patent Nos. 4,767,704; 4,657,866; 4,927,762; 4,560,655 or 5,122,469; WO 90/03430; WO Any culture medium described in 87/00195 or U.S. Patent Re. 30,985 can be used as culture medium for the host cells. Any of these media may be supplemented, as necessary, with hormones and/or other growth factors (e.g., insulin, transferrin, or epidermal growth factor), salts (e.g., sodium chloride, calcium, magnesium, and phosphate), buffers (e.g., HEPES), nucleotides (e.g., adenosine and thymidine), antibiotics (e.g., GENTAMYCIN ^™ ), trace elements (defined as inorganic compounds typically present at final concentrations in the micromolar range), and glucose or an equivalent energy source. Any other necessary supplements may also be included at appropriate concentrations known to those skilled in the art. Culture conditions, such as temperature and pH, are those previously used with the host cell selected for expression and will be apparent to those skilled in the art.

ix. Purification of Heteromultimers

When using recombinant technology, heteromultimers can be produced intracellularly or directly secreted into the culture medium. If the heteromultimers are produced intracellularly, as a first step, granular cell debris, host cells or cracked fragments are removed by, for example, centrifugation or ultrafiltration. When the heteromultimers are secreted into the culture medium, the supernatant from such expression system is usually first concentrated using a commercially available protein concentration filter (e.g., Amicon or Millipore Pellicon ultrafiltration unit). Protease inhibitors such as PMSF may be included in any of the above steps to inhibit protein hydrolysis, and antibiotics may be included to prevent the growth of accidental contaminants.

Heteromultimeric compositions prepared from cells can be purified by, for example, hydroxyapatite chromatography, gel electrophoresis, dialysis, and affinity chromatography, with affinity chromatography being a preferred purification technique. The suitability of protein A as an affinity ligand depends on the type and isotype of the immunoglobulin Fc domain present in the heteromultimer. Protein A can be used to purify heteromultimers based on human γ1, γ2, or γ4 heavy chains (Lindmark et al., J. Immunol. Meth. 62: 1-13 (1983)). For all mouse isotypes and human γ3, protein G is recommended (Guss et al., EMBO J. 5: 1567-1575 (1986)). The most commonly used matrix for affinity ligand attachment is agarose, but other matrices are also available. Mechanically stable matrices such as controlled pore glass or poly (styrene divinyl) benzene allow faster flow rates and shorter processing times than can be achieved with agarose. When the heteromultimer contains a CH3 domain, Bakerbond ABX ^™ resin (JT Baker, Phillipsburg, NJ) can be used for purification. Other protein purification techniques, such as fractionation on an ion exchange column, ethanol precipitation, reversed-phase HPLC, chromatography on silica, chromatography on heparin SEPHAROSE ^™ , chromatography on an anion or cation exchange resin (e.g., a polyaspartic acid column), chromatofocusing, SDS-PAG, and ammonium sulfate precipitation, may also be used, depending on the antibody to be recovered.

In one embodiment, the eukaryotic cell additionally expresses an endopeptidase (e.g., furin) to cleave one or more tethers of the heteromultimer of interest prior to purification (e.g., during trans-Golgi transport). In another embodiment, the eukaryotic host cell may express an endopeptidase (e.g., furin or Lys-C) and an exopeptidase (e.g., carboxypeptidase B), wherein the endopeptidase cleaves one or more tethers of the heteromultimer prior to purification and the exopeptidase degrades any residual endopeptidase cleavage sites.

The target heteromultimer can be recovered from the solid phase of the column by eluting into a solution containing a chaotropic agent or a mild detergent. A "chaotropic agent" refers to a water-soluble substance that disrupts the three-dimensional structure of a protein (e.g., an antibody) by interfering with stable intramolecular interactions (e.g., hydrogen bonds, van der Waals forces, or hydrophobic effects). Exemplary chaotropic agents include, but are not limited to, urea, guanidine-HCl, lithium perchlorate, histidine, and arginine. A "mild detergent" refers to a water-soluble substance that disrupts the three-dimensional structure of a protein (e.g., an antibody) by interfering with stable intramolecular interactions (e.g., hydrogen bonds, van der Waals forces, or hydrophobic effects), but does not permanently destroy the protein structure, thereby causing a loss of biological activity (i.e., denaturing the protein). Exemplary mild detergents include, but are not limited to, Tween (e.g., Tween-20), Triton (e.g., Triton X-100), NP-40 (nonylphenoxypolyethoxyethanol), Nonidet P-40 (octylphenoxypolyethoxyethanol), and sodium dodecylsulfonate (SDS).

Following any one or more preliminary separation steps, the mixture comprising the target heteromultimer and contaminants can be subjected to low pH hydrophobic interaction chromatography using an elution buffer having a pH between about 2.5 and 4.5, preferably at a low salt concentration (e.g., from about 0-0.25 M salt).

In some embodiments, one or more residual endopeptidase cleavage sites of the heteromultimers of the invention that have been treated with endopeptidase can be treated and degraded by an exopeptidase (e.g., carboxypeptidase B) after purification. In other embodiments, one or more tethers of the heteromultimers of the invention can be cleaved and treated with an endopeptidase (e.g., furin) and an exopeptidase (e.g., carboxypeptidase B) after purification, but not before purification.

x. Production of Heteromultimers Using Baculovirus

Recombinant baculoviruses can be produced by co-transfecting a plasmid encoding a heteromultimer or heteromultimer fragment with BaculoGold ^™ viral DNA (Pharmingen) into insect cells such as Spodoptera frugiperda cells (e.g., Sf9 cells; ATCC CRL 1711) or Drosophila melanogaster S2 cells using, for example, liposomes (commercially available from GIBCO-BRL). In a specific example, the heteromultimer sequence is fused upstream of an epitope tag contained within a baculovirus expression vector. Such epitope tags include poly-His tags. A variety of plasmids can be used, including plasmids derived from commercially available plasmids such as pVL1393 (Novagen) or pAcGP67B (Pharmingen). In short, the sequence encoding the antibody heteromultimer or its fragment can be amplified by PCR using primers complementary to the 5' and 3' regions. The 5' primer can incorporate flanking (selected) restriction enzyme sites. The product can then be digested with the selected restriction enzyme and subcloned into an expression vector.

After transfection with the expression vector, the host cells (e.g., Sf9 cells) are incubated at 28°C for 4-5 days, and the released virus is then collected for further amplification. Viral infection and protein expression can be performed as described, for example, by O'Reilley et al. (Baculovirus expression vectors: A Laboratory Manual. Oxford: Oxford University Press (1994)).

The expressed poly-His-tagged heteromultimers can then be purified using, for example, Ni ²⁺ chelate affinity chromatography as follows. Extracts can be prepared from recombinant virus-infected Sf9 cells as described by Rupert et al. (Nature 362:175-179 (1993)). Briefly, Sf9 cells were washed, resuspended in ultrasonication buffer (25 mL HEPES pH 7.9; 12.5 mM MgCl ₂ ; 0.1 mM EDTA; 10% glycerol; 0.1% NP-40; 0.4 M KCl) and ultrasonicated on ice twice for 20 seconds each. The ultrasonication product was clarified by centrifugation, and the supernatant was diluted 50-fold in loading buffer (50 mM phosphate; 300 mM NaCl; 10% glycerol pH 7.8) and passed through a 0.45 μm filter. A Ni ²⁺ -NTA agarose column (commercially available from Qiagen) with a bed volume of 5 mL was prepared, washed with 25 mL of water, and balanced with 25 mL of loading buffer. The filtered cell extract was loaded into the column at 0.5 mL/min. The column was washed with loading buffer to a baseline A ₂₈₀ , at which point fraction collection began. The column was then washed with a second elution buffer (50 mM phosphate; 300 mM NaCl; 10% glycerol pH 6.0) to elute non-specifically bound proteins. After reaching the A ₂₈₀ baseline again, the column was developed with a 0 to 500 mM imidazole gradient in the second elution buffer. 1 mL fractions were collected and Western blot analysis was performed using SDS-PAGE and silver staining or with Ni ²⁺ -NTA-conjugated alkaline phosphatase (Qiagen). Fractions containing the eluted His _10- tagged (SEQ ID NO: 2) heteropolymer were pooled and dialyzed with loading buffer.

The purification of heteromultimers can also be performed using known chromatographic techniques, such as protein A or protein G column chromatography. The target heteromultimer can be recovered from the solid phase of the column by eluting into a solution containing a chaotropic agent or a mild detergent. Exemplary chaotropic agents and mild detergents include, but are not limited to, guanidine-HCl, urea, lithium perchlorate, arginine, histidine, SDS (sodium dodecyl sulfate), Tween, Triton, and NP-40, all of which are commercially available. As described above, one or more tethers of the heteromultimer can be cleaved and treated after purification by the addition of endopeptidases (e.g., furin or Lys-C) and exopeptidases (e.g., carboxypeptidase B).

C. Purification Technology

One specific purification method that can be used for HD tethered heteromultimers is as follows: Load the ligated heteromultimers onto a Protein A (e.g., mAbSure) column at 4°C.

↓

Wash the column with KPO ₄ and then with PBS + 0.1% Trition X114

↓

The sample was eluted into Tris pH 8.0 (200 mM)

↓

The sample pH was adjusted to 8.0 and cleaved with Lys-C at 37 °C for 1.5 h.

↓

Repurify the sample using MabSURE columns

↓

The sample pH was adjusted to 8.0 and cleaved with carboxypeptidase B at 37 °C for 4 h.

↓

Adjust sample conductivity and load onto ion exchange column

↓

Fractions were collected, pooled and dialyzed into PBS.

In addition to arginine, other chaotropic agents or mild detergents that can be used in the above purification protocol following the initial Protein A column step include, but are not limited to, guanidine-HCl, urea, lithium perchlorate, histidine, SDS (sodium dodecyl sulfate), Tween, Triton, and NP-40, all of which are commercially available. In some embodiments, one or more residual endopeptidase cleavage sites of the heteromultimers of the present invention that have been treated with endopeptidase can be treated and degraded by an exopeptidase (e.g., carboxypeptidase B) after purification. In other embodiments, one or more tethers of the heteromultimers of the present invention can be cleaved and treated with an endopeptidase (e.g., furin or Lys-C) and an exopeptidase (e.g., carboxypeptidase B) after purification rather than before purification.

IV. Conjugated Proteins

The present invention also provides conjugated proteins, such as conjugated heteromultimers or immunoconjugates (e.g., "antibody-drug conjugates" or "ADCs"), comprising any of the heteromultimers described herein (e.g., single-chain monospecific or multispecific antibodies containing an HD tether, multi-chain monospecific or multispecific heteromultimers containing an HD tether), wherein one of the constant regions of the light or heavy chains is conjugated to a chemical molecule, such as a dye or cytotoxic agent, such as a chemotherapeutic agent, a drug, a growth inhibitory agent, a toxin (e.g., an enzymatically active toxin of bacterial, fungal, plant, or animal origin, or a fragment thereof), or a radioactive isotope (i.e., a radioconjugate). Specifically, as described herein, the use of an HD tether enables the construction of heteromultimers containing two different heavy chains (HCl and HC2) and two different light chains (LC1 and LC2). Immunoconjugates constructed using the methods described herein can include a cytotoxic agent conjugated to the constant region of only one of the heavy chains (HCl or HC2) or only one of the light chains (LC1 or LC2). Likewise, because the immunoconjugate can have the cytotoxic agent attached to only one heavy or light chain, the amount of cytotoxic agent administered to a subject is reduced relative to administration of a heteromultimer having the cytotoxic agent attached to two heavy or light chains. Reducing the amount of cytotoxic agent administered to a subject limits adverse side effects associated with the cytotoxic agent.

The use of antibody-drug conjugates for local delivery of cytotoxic or cytostatic drugs, i.e., drugs that kill or inhibit tumor cells, in the treatment of cancer (Syrigos and Epenetos, Anticancer Research 19:605-614 (1999); Niculescu-Duvaz and Springer, Adv. Drg. Del. Rev. 26:151-172 (1997); U.S. Pat. No. 4,975,278) allows for targeted delivery of the drug moiety to the tumor and intracellular accumulation therein, whereas systemic administration of these unconjugated agents can result in unacceptable levels of toxicity to both normal cells and the tumor cells being sought to be removed (Baldwin et al., Lancet (Mar. 15, 1986):603-605 (1986); Thorpe, (1985) "Antibody Carriers Of Cytotoxic Agents In Cancer Therapy: A Review," in Monoclonal Antibodies '84: Biological And Clinical Applications, A. Pinchera et al. (eds.), pp. 475-506). Thus, maximum efficacy with minimal toxicity is sought. Both polyclonal and monoclonal antibodies have been reported for use in this strategy (Rowland et al., Cancer Immunol. Immunother. 21: 183-187 (1986)). Drugs used in this approach include daunomycin, doxorubicin, methotrexate, and vindesine (Rowland et al., (1986) supra). Toxins used in heteromultimer-toxin conjugates include bacterial toxins such as diphtheria toxin, plant toxins such as ricin, small molecule toxins such as geldanamycin (Mandler et al., Jour. of the Nat. Cancer Inst. 92(19):1573-1581 (2000); Mandler et al., Bioorganic & Med. Chem. Letters 10:1025-1028 (2000); Mandler et al., Bioconjugate Chem. 13:786-791 (2002)), maytansinoids (EP 1391213; Liu et al., Proc. Natl. Acad. Sci. USA 93:8618-8623 (1996)), and calicheamicin (Lode et al., Cancer Res. 58:2928 (1998); Hinman et al., Cancer Res. 53:3336-3342 (1993). Toxins can achieve their cytotoxic and cytostatic effects through mechanisms including tubulin binding, DNA binding, or inhibition of topoisomerases. Some cytotoxic drugs tend to be inactive or less active when conjugated to large heteromultimeric antibodies or protein receptor ligands.

Chemotherapeutic agents useful in generating immunoconjugates are described herein (eg, supra). Enzymatically active toxins and fragments thereof that can be used include diphtheria toxin A chain, nonbinding active fragments of diphtheria toxin, exotoxin A chain (from Pseudomonas aeruginosa), ricin A chain, abrin A chain, modeccin A chain, α-sarcin, Aleurites fordii proteins, dianthin proteins, Phytolaca americana proteins (PAPI, PAPII, and PAP-S), momordica charantia inhibitor, curcin, crotin, sapaonaria officinalis inhibitor, gelonin, mitogillin, restrictocin, phenomycin, enomycin, and the tricothecenes. See, e.g., WO93/21232 published October 28, 1993. A variety of radionuclides are available for the production of radioconjugated antibodies. Examples include ²¹² Bi, ¹³¹ I, ¹³¹ In, ⁹⁰ Y, and ¹⁸⁶ Re. Conjugates of antibodies and cytotoxic agents are prepared using a variety of bifunctional protein coupling agents, such as N-succinimidyl-3-(2-pyridyldithiol)-propionate (SPDP), iminothiolane (IT), bifunctional derivatives of imidoesters (e.g., dimethyl adipimidate HCl), active esters (e.g., disuccinimidyl suberate), and bis-succinimidyl esters. suberate), aldehydes (e.g., glutaraldehyde), bis-azide compounds (e.g., bis(p-azidobenzoyl)hexanediamine), bis-diazonium derivatives (e.g., bis-(p-diazoniumbenzoyl)-ethylenediamine), diisocyanates (e.g., toluene 2,6-diisocyanate), and bis-active fluorine compounds (e.g., 1,5-difluoro-2,4-dinitrobenzene). For example, ricin immunotoxins can be prepared as described in Vitetta et al., Science 238:1098 (1987). Carbon-14 labeled 1-isothiocyanatobenzyl-3-methyldiethylenetriaminepentaacetic acid (MX-DTPA) is an exemplary chelating agent for conjugating radionuclides to antibodies. See, for example, WO94/11026.

Also contemplated herein are conjugates of heteromultimers and one or more small molecule toxins (eg, calicheamicins, maytansinoids, dolastatins, aurostatin, trichothecenes, and CC1065, and toxinically active derivatives of these toxins).

A. Maytansine and maytansine alkaloids

In some embodiments, the immunoconjugate comprises a heteromultimer of the invention conjugated to one or more maytansinoid molecules.

Maytansinoids are mitotic inhibitors that work by inhibiting tubulin polymerization. Maytansine was originally isolated from the East African shrub Maytenus serrata (U.S. Patent No. 3,896,111). Subsequently, it was discovered that certain microorganisms also produce maytansinoids, such as maytansinol and C-3 maytansinol esters (U.S. Patent No. 4,151,042). For example, the following U.S. Patent Nos. 4,137,230; 4,248,870; 4,256,746; 4,260,608; 4,265,814; 4,294,757; 4,307,016; 4,308,268; 4,308,269; 4,309,428; 4,313,946; 4,315,929; 4,317,821; 4,322,348; 4,331,598; 4,361,650; 4,364,866; 4,424,219; 4,450,254; 4,362,663; and 4,371,533.

Maytansinoid drug moieties are attractive drug moieties in heteromultimeric antibody-drug conjugates because they are: (i) relatively easy to prepare by fermentation or chemical modification and derivatization of fermentation products, (ii) can be derivatized with functional groups suitable for conjugation to antibodies via non-disulfide linkers, (iii) are stable in plasma, and (iv) are effective against a variety of tumor cell lines.

Immunoconjugates containing maytansinoids, methods for preparing them, and their therapeutic applications have been disclosed in, for example, U.S. Patent Nos. 5,208,020, 5,416,064, and European Patent EP 0 425 235 B1, the disclosures of which are expressly incorporated herein by reference. Liu et al., Proc. Natl. Acad. Sci. USA 93:8618-8623 (1996) described an immunoconjugate comprising a maytansinoid called DM1 linked to the monoclonal antibody C242 against human colorectal cancer. The conjugate was found to be highly cytotoxic to cultured colon cancer cells and to exhibit antitumor activity in an in vivo tumor growth assay. Chari et al., Cancer Research 52:127-131 (1992) have described such immunoconjugates, wherein maytansinoid is puted together with the mouse antibody A7 of the antigen on the human colon cancer cell line through disulfide bond linker, or puted together with another mouse monoclonal antibody TA.1 of the HER-2/neu proto-oncogene.In vitro, the cytotoxicity of TA.1-maytansinoid conjugate was tested on the human breast cancer cell line SK-BR-3, each cell of which expresses ^3x10 HER-2 surface antigens.Drug conjugates have reached the cytotoxicity of the similar degree to free maytansinoid drug, which can be improved by increasing the number of maytansinoid molecules per antibody molecule.A7-maytansinoid conjugates showed low systemic cytotoxicity in mice.

Can be by heteropolymer and maytansinoid molecule chemical connection, prepare maytansinoid conjugate, without significantly reducing the biological activity of heteropolymer or maytansinoid molecule. See, for example, U.S. Patent No. 5,208,020 (its disclosure is clearly incorporated herein by reference). Each heteropolymer antibody molecule is conjugated to an average of 3-4 maytansinoid molecules, showing the effectiveness of enhancing the cytotoxicity of target cells, without negatively affecting the function or solubility of the antibody, although it is expected that even the toxin/antibody of one molecule will also be more cytotoxic than using naked antibodies. Maytansinoids are well known in the art and can be synthesized or separated from natural sources by known techniques. Suitable maytansinoids are disclosed in, for example, U.S. Patent No. 5,208,020 and other patents and non-patent publications mentioned above. Preferred maytansinoids are maytansinol and maytansinol analogs modified in the aromatic ring or other positions of the maytansinol molecule, such as a variety of maytansinol esters.

There are many linking groups for preparing maytansinoid conjugates known in the art, including, for example, those disclosed in U.S. Patent No. 5,208,020 or European Patent 0 425 235 Bl, Chari et al., Cancer Research 52: 127-131 (1992) and U.S. Patent Application Publication No. 2005/0169933, the disclosure of which is explicitly incorporated herein by reference. Maytansinoid conjugates comprising the linker component SMCC can be prepared as disclosed in U.S. Patent Application Publication No. 2005/0169933. Linking groups include disulfide groups, thioether groups, acid-labile groups, photolabile groups, peptidase-labile groups or esterase-labile groups, as disclosed in the patents described above, preferably disulfide groups and thioether groups. Other linking groups are also described and enumerated herein.

Conjugates of heteropolymers and maytansine alkaloids can be prepared using a variety of bifunctional protein coupling agents, such as N-succinimidyl-3-(2-pyridyldithiol)-propionate (SPDP), succinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC), iminothiolane (IT), bifunctional derivatives of imidoesters (e.g., dimethyl adipimidate HCl), active esters (e.g., disuccinimidyl suberate), aldehydes (e.g., glutaraldehyde), bis-azido compounds (e.g., bis(p-azidobenzoyl)hexanediamine), bis-diazonium derivatives (e.g., bis-(p-diazoniumbenzoyl)-ethylenediamine), diisocyanates (e.g., toluene 2,6-diisocyanate), and bis-active fluorine compounds (e.g., 1,5-difluoro-2,4-dinitrobenzene). Particularly preferred coupling agents include N-succinimidyl-3-(2-pyridyldithio)-propionate (SPDP) (Carlsson et al., Biochem. J. 173:723-737 (1978)) and N-succinimidyl-4-(2-pyridylthio) pentanoate (SPP) for providing disulfide linkages.

In another embodiment, the joint can be attached to a plurality of positions of the maytansinoid molecule.For example, conventional coupling techniques can be used to form an ester bond by reacting with a hydroxyl group. The reaction can occur at the C-3 position with a hydroxyl group, at the C-14 position modified with hydroxyl group, at the C-15 position modified with hydroxyl group, and at the C-20 position with a hydroxyl group. In a preferred embodiment, a key is formed at the C-3 position of maytansinol or a maytansinol analog.

B. Auristatin and Dulostatin

In some embodiments, the immunoconjugate comprises a heteromultimer of the invention conjugated to dolastatin or a dolastatin peptide analog and derivative, auristatin (U.S. Patent Nos. 5,635,483 and 5,780,588). Dolastatin and auristatin have been shown to interfere with microtubule dynamics, GTP hydrolysis, and nuclear and cell division (Woyke et al. (2001) Antimicrob. Agents and Chemother. 45(12): 3580-3584), and have anticancer (U.S. Patent No. 5,663,149) and antifungal activity (Pettit et al., Antimicrob. Agents Chemother. 42: 2961-2965 (1998)). The dolastatin or auristatin drug moiety can be attached to the heteromultimer via the N (amino) or C (carboxyl) terminus of the peptide drug moiety (WO 02/088172).

Exemplary auristatin embodiments include N-terminally linked monomethylauristatin drug moieties DE and DF, which are disclosed in "Monomethylvaline Compounds Capable of Conjugation to Ligands," U.S. Application Publication No. 2005/0238649, the disclosure of which is expressly incorporated herein by reference in its entirety.

Typically, peptide-based drug moieties can be prepared by forming peptide bonds between two or more amino acids and/or peptide fragments. Such peptide bonds can be prepared according to, for example, liquid phase synthesis methods well known in the art of peptide chemistry (see E. and K. Lübke, "The Peptides", Volume 1, pages 76-136, 1965, Academic Press). Auristatin/dolastatin drug moieties can be prepared according to the following methods: U.S. Patent Nos. 5,635,483 and 5,780,588; Pettit et al., J. Nat. Prod. 44: 482-485 (1981); Pettit et al., Anti-Cancer Drug Design 13: 47-66 (1998); Poncet, Curr. Pharm. Des. 5: 139-162 (1999) and Pettit, Fortschr. Chem. Org. Naturst. 70: 1-79 (1997). See also Doronina (2003) Nat Biotechnol 21(7):778-784; and "Monomethylvaline Compounds Capable of Conjugation to Ligands," U.S. Application Publication No. 2005/0238649, incorporated herein by reference in its entirety (disclosing, for example, linkers and methods of preparing monomethylvaline compounds (such as MMAE and MMAF) conjugated to linkers).

C. calicheamicin

In other embodiments, the immunoconjugate comprises a heteropolymer of the present invention conjugated to one or more calicheamicin molecules. The calicheamicin antibiotic family is capable of producing double-stranded DNA breaks at sub-picomolar concentrations. For the preparation of calicheamicin family conjugates, see U.S. Patent Nos. 5,712,374, 5,714,586, 5,739,116, 5,767,285, 5,770,701, 5,770,710, 5,773,001, and 5,877,296 (all belonging to American Cyanamid). Available structural analogs of calicheamicin include, but are not limited to, γ ₁ ^I , α ₂ ^I , α ₃ ^I , N-acetyl-γ ₁ ^I , PSAG, and θ ¹ _I (Hinman et al., Cancer Research 53:3336-3342 (1993), Lode et al., Cancer Research 58:2925-2928 (1998), and the aforementioned U.S. patents to American Cyanamid. Another anti-tumor drug to which antibodies can be conjugated is QFA, an antifolate. Both calicheamicin and QFA have intracellular sites of action and are not easily able to cross the plasma membrane. Therefore, the cellular uptake of such active agents through antibody-mediated internalization greatly enhances their cytotoxic effects.

D. Other cytotoxic agents

Other anti-tumor agents that can be conjugated to the heteromultimers of the present invention or prepared according to the methods described herein include BCNU, streptozoicin, vincristine, 5-fluorouracil, the family of active agents collectively referred to as the LL-E33288 complex described in U.S. Pat. Nos. 5,053,394 and 5,770,710, and esperamicins (U.S. Pat. No. 5,877,296).

Useful enzymatically active agonist toxins and fragments thereof include diphtheria toxin A chain, non-binding active fragments of diphtheria toxin, exotoxin A chain (from Pseudomonas aeruginosa), ricin A chain, abrin A chain, modeccin A chain, α-sarcin, Aleurites fordii proteins, dianthin proteins, Phytolaca americana proteins (PAPI, PAPII, and PAP-S), momordica charantia inhibitor, Curcas curcas toxin, crotonin, phytoncidin inhibitor, gelonin, milincomycin, restrictocin, phenomycin, enomycin, and trichothecenes.

The present invention also contemplates immunoconjugates formed between the heteromultimers and a compound having nucleolytic activity (eg, a ribonuclease or a DNA endonuclease, such as a deoxyribonuclease; DNase).

In order to selectively destroy tumors, heteropolymers may contain highly radioactive atoms. A variety of radioisotopes can be used to produce radioconjugated heteropolymers. Examples include At ²¹¹ , I ¹³¹ , I ¹²⁵ , Y ⁹⁰ , Re ¹⁸⁶ , Re ¹⁸⁸ , Sm ¹⁵³ , Bi ²¹² , P ³² , Pb ²¹² , and radioisotopes of Lu. When the conjugate is used for detection, it may contain radioactive atoms for scintigraphic studies, such as Tc ^99m or I ¹²³ , or a spin label for nuclear magnetic resonance (NMR) imaging (also known as magnetic resonance imaging, mri), such as iodine-123, iodine-131, indium-111, fluorine-19, carbon-13, nitrogen-15, oxygen-17, gadolinium, manganese, or iron.

Radioactive or other labels can be incorporated into the conjugate using known methods. For example, the peptide can be biosynthesized or synthesized by chemical amino acid synthesis using suitable amino acid precursors, involving, for example, fluorine-19 in place of hydrogen. Labels such as tc ^99m or I ¹²³ , Re ¹⁸⁶ , Re ¹⁸⁸ , and In ¹¹¹ can be attached via cysteine residues in the peptide. Yttrium-90 can be attached via lysine residues. The IODOGEN method (Fraker et al., Biochem. Biophys. Res. Commun. 80:49-57 (1978)) can be used to incorporate iodine-123. "Monoclonal Antibodies in Immunoscintigraphy" (Chatal, CRC Press 1989) describes other methods in detail.

Conjugates of heteropolymers and cytotoxic agents can be prepared using a variety of bifunctional protein coupling agents, such as N-succinimidyl-3-(2-pyridyldithiol)-propionate (SPDP), succinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC), iminothiolane (IT), bifunctional derivatives of imidoesters (e.g., dimethyl adipimidate HCl), active esters (e.g., disuccinimidyl suberate), aldehydes (e.g., glutaraldehyde), bis-azido compounds (e.g., bis(p-azidobenzoyl)hexanediamine), bis-diazonium derivatives (e.g., bis-(p-diazoniumbenzoyl)-ethylenediamine), diisocyanates (e.g., toluene 2,6-diisocyanate), and bis-active fluorine compounds (e.g., 1,5-difluoro-2,4-dinitrobenzene). For example, a ricin immunotoxin can be prepared as described in Vitetta et al., Science 238:1098 (1987). Carbon-14 labeled 1-isothiocyanatobenzyl-3-methyldiethylenetriaminepentaacetic acid (MX-DTPA) is an exemplary chelating agent for conjugating radionuclides to heteropolymers. See, for example, WO 94/11026. The linker can be a "cleavable linker" that assists in releasing the cytotoxic drug in the cell. For example, an acid-labile linker, a peptidase-sensitive linker, a photolabile linker, a dimethyl linker, or a disulfide-containing linker can be used (Chari et al., Cancer Research 52:127-131 (1992); U.S. Patent No. 5,208,020).

The compounds of the present invention specifically contemplate, but are not limited to, ADCs prepared with the following cross-linkers: commercially available BMPS, EMCS, GMBS, HBVS, LC-SMCC, MBS, MPBH, SBAP, SIA, SIAB, SMCC, SMPB, SMPH, sulfo-EMCS, sulfo-GMBS, sulfo-KMUS, sulfo-MBS, sulfo-SIAB, sulfo-SMCC, and sulfo-SMPB, and SVSB (succinimidyl-(4-vinylsulfone)benzoate) (available, for example, from Pierce Biotechnology, Inc., Rockford, IL., U.S.A. See 2003-2004 Application Handbook and Catalog, pages 467-498.

E. Preparation of Conjugated Heteromultimers

In the conjugated heteromultimers of the present invention, the heteromultimer is conjugated to one or more moieties (e.g., drug moieties), e.g., about 1 to about 20 moieties per antibody, optionally via a linker. Conjugated heteromultimers can be prepared by several routes using organic chemistry reactions, conditions, and reagents known to those skilled in the art, including: (1) reacting nucleophilic groups of the antibody heteromultimer with a divalent linker reagent via covalent bonds, followed by reaction with the target moiety; and (2) reacting nucleophilic groups of the moiety with a divalent linker reagent via covalent bonds, followed by reaction with nucleophilic groups of the antibody heteromultimer. Other methods for preparing conjugated heteromultimers are described herein.

The linker reagent may be composed of one or more linker components. Exemplary linker components include 6-maleimidoacetyl ("MC"), maleimidopropanoyl ("MP"), valine-citrulline ("val-cit"), alanine-phenylalanine ("ala-phe"), p-aminobenzyloxycarbonyl ("PAB"), N-succinimidyl 4-(2-pyridylmercapto) pentanoate ("SPP"), N-succinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate ("SMCC"), and N-succinimidyl (4-iodo-acetyl) aminobenzoate ("SIAB"). Other linker components are known in the art, some of which are described herein. See also "Monomethylvaline Compounds Capable of Conjugation to Ligands," U.S. Patent Application Publication No. 2005/0238649, the contents of which are incorporated herein by reference in their entirety.

In some embodiments, the linker may comprise an amino acid residue. Exemplary amino acid linker components include dipeptides, tripeptides, tetrapeptides, or pentapeptides. Exemplary dipeptides include: valine-citrulline (vc or val-cit), alanine-phenylalanine (af or ala-phe). Exemplary tripeptides include: glycine-valine-citrulline (gly-val-cit) and glycine-glycine-glycine (gly-gly-gly). The amino acid residues comprising the amino acid linker component include naturally occurring, as well as a few amino acids and non-naturally occurring amino acid analogs, such as citrulline. The selectivity of the amino acid linker component for enzymatic cleavage by a specific enzyme can be designed and optimized, such as tumor-associated proteases, histone proteases-B, C, and D, or plasmin proteases.

The nucleophilic groups of the heteromultimers of the present invention include, but are not limited to: (i) N-terminal amino groups; (ii) side chain amino groups, such as lysine; (iii) side chain sulfhydryl groups, such as cysteine; and (iv) hydroxyl or amino groups of sugars in glycosylated antibodies. Amino groups, sulfhydryl groups, and hydroxyl groups are nucleophilic and can react with electrophilic groups on linker moieties and linker reagents to form covalent bonds, and the electrophilic groups include: (i) active esters such as NHS esters, HOBt esters, haloformates, and acid halides; (ii) alkyl and benzyl halides such as haloacetamides; (iii) aldehyde groups, ketone groups, carboxyl groups, and maleimide groups. Certain heteromultimers have reducible interchain disulfide bonds, i.e., cysteine bridges. The heteromultimers of the present invention can be made reactive for conjugation with linker reagents by treatment with a reducing agent such as DTT (dithiothreitol). Therefore, each cysteine bridge will theoretically form two reactive thiol nucleophiles. Additional nucleophilic groups can be introduced into antibodies by reacting lysine with 2-iminothiolane (Traut's reagent), resulting in conversion of amines to thiols. Reactive thiol groups can be introduced into heteromultimers (or fragments thereof) by introducing one, two, three, four, or more cysteine residues (e.g., preparing mutant antibodies containing one or more non-natural cysteine amino acid residues).

The conjugated heteropolymers of the present invention can also be produced by modifying the heteropolymer to introduce an electrophilic portion that can react with a nucleophilic substituent on a linker reagent or a drug or other moiety. The sugars of the glycosylated heteropolymers of the present invention can be oxidized, for example, with a periodate oxidant, to form an aldehyde or ketone group that can react with the amine group of the linker reagent or the drug or other moiety. The obtained imine Schiff base group can form a stable bond or can be reduced, for example, by a borohydride reagent, to form a stable amine bond. In one embodiment, the reaction of the carbohydrate portion of the glycosylated antibody with galactose oxidase or sodium metaperiodate can generate carbonyl (aldehyde and ketone) groups in the protein, which can react with suitable groups on the drug or other moiety (Hermanson, Bioconjugate Techniques). In another embodiment, proteins containing an N-terminal serine or threonine residue can be reacted with sodium metaperiodate to generate an aldehyde at the first amino acid position (Geoghegan and Stroh, Bioconjugate Chem. 3:138-146 (1992); U.S. Patent No. 5,362,852). Such aldehydes can react with drug moieties or linker nucleophiles.

Similarly, electrophilic groups on moieties (e.g., drug moieties) include, but are not limited to, amine, sulfhydryl, hydroxyl, hydrazide, oxime, hydrazine, thiosemicarbazone, hydrazine carboxylate, and arohydrazide groups that can react to form covalent bonds with electrophilic groups on linker moieties and linker reagents, including: (i) active esters such as NHS esters, HOBt esters, haloformates, and acid halides; (ii) alkyl and benzyl halides such as haloacetamides; and (iii) aldehyde, keto, carboxyl, and maleimide groups.

Alternatively, a fusion protein comprising a heteromultimer and a cytotoxic agent can be prepared, for example, by recombinant techniques or peptide synthesis. The length of DNA can comprise regions encoding the two parts of the conjugate, the regions being adjacent to each other or separated by a region encoding a linker peptide that does not destroy the desired properties of the conjugate. In another embodiment, the heteromultimer can be conjugated to a "receptor" (such as streptavidin) for use in pre-targeting applications of tumors, wherein the heteromultimer-receptor conjugate is administered to an individual, followed by removal of unbound conjugate from the circulation using a scavenger, and then administration of a "ligand" (e.g., avidin) conjugated to a cytotoxic agent (e.g., a radionuclide).

V. Therapeutic Uses

The heteromultimers described herein (e.g., single-chain monospecific or multispecific antibodies containing an HD tether, monospecific or multispecific antibodies containing an HD tether, immunoadhesin-antibody complexes) can be used in therapeutic applications. For example, such heteromultimers can be used to treat any disease for which there is a suitable candidate target for the generation of heteromultimers, including proliferative diseases, cancer, angiogenic disorders, inflammatory disorders, autoimmune diseases, and immune disorders.

In addition to therapeutic uses, the heteromultimers of the invention may be used for other purposes, including diagnostic methods, eg, for any disease or condition for which there is a suitable candidate target for the generation of antibodies, such as the diseases or conditions described above.

VI. Dosage, Formulation, and Duration

The heteromultimers of the present invention are formulated, dosed, and administered in a manner consistent with good medical practice. Factors considered in this context include the specific condition being treated, the specific mammal being treated, the clinical condition of the individual subject, the cause of disease, the site of delivery of the active agent, the method of administration, the dosing schedule, and other factors known to the practitioner. The "therapeutically effective amount" of the protein to be administered is determined by such considerations and is the minimum amount required to prevent, ameliorate, or treat a particular condition (e.g., a proliferative disease, cancer, angiogenic condition, inflammatory condition, autoimmune disease, or immune condition). The protein need not be, but may optionally be, formulated with one or more active agents currently used to prevent or treat the condition. The effective amount of such other active agents depends on the amount of protein in the formulation, the type of condition or treatment, and the other factors mentioned above. They are generally used in the same dosage and route of administration as previously used, or at about 1 to 99% of the dosage used to date.

In one embodiment, the present invention can be used to increase the survival duration of human subjects susceptible to or diagnosed with a specific condition having an appropriate candidate target for producing heteromultimers. Survival duration is defined as the time from the first administration of the drug to death. Survival duration can also be measured by a stratified hazard ratio (HR) of the treatment group versus the control group, which represents the risk of death of the subject during treatment.

In another embodiment, the methods of the present invention significantly increase the response rate in a group of human subjects susceptible to or diagnosed with a condition and treated with one or more therapies for the condition. The response rate is defined as the percentage of treated subjects that respond to the treatment. In one aspect, the combination therapy of the present invention using a protein of the present invention and surgery or other treatment modality (e.g., radiation therapy or chemotherapeutic agents) significantly improves the response rate in the group of treated subjects compared to a group treated with surgery or other treatment modality alone, with a chi- ^square p value of less than 0.005. Additional measures of therapeutic efficacy of cancer treatments are described in U.S. Patent Application Publication No. 20050186208.

Therapeutic formulations are prepared by mixing the active ingredient having the desired degree of purity with optional physiologically acceptable carriers, excipients, or stabilizers using standard methods known in the art (Remington's Pharmaceutical Sciences (20th edition), ed. A. Gennaro, 2000, Lippincott, Williams & Wilkins, Philadelphia, PA). Acceptable carriers include saline or buffers such as phosphate, citric acid, and other organic acids; antioxidants including ascorbic acid; low molecular weight (less than about 10 residues) polypeptides; proteins such as serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, arginine, or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugar alcohols such as mannitol or sorbitol; salt-forming counterions such as sodium; and/or nonionic surfactants such as TWEEN ^™ , PLURONICS ^™ , or PEG.

Optionally, but preferably, the formulation includes a pharmaceutically acceptable salt, preferably sodium chloride, preferably at approximately physiological concentrations. Optionally, the formulations of the present invention may include a pharmaceutically acceptable preservative. In some embodiments, the concentration of the preservative ranges from 0.1-2.0%, typically v/v. Suitable preservatives include those known in the pharmaceutical art. Benzyl alcohol, phenol, m-cresol, methylparaben, and propylparaben are preferred preservatives. Optionally, the formulations of the present invention may include a pharmaceutically acceptable surfactant at a concentration of 0.005-0.02%.

As required for the particular indication being treated, the formulations herein may also contain more than one active compound, preferably compounds with complementary activities that do not adversely affect each other. Such molecules are suitably present in combination in amounts that are effective for the intended purpose.

The active ingredient can also be embedded in microcapsules prepared, for example, by coacervation techniques or by interfacial polymerization (e.g., hydroxymethylcellulose or gelatin microcapsules and poly(methyl methacrylate) microcapsules, respectively), in colloidal drug delivery systems (e.g., liposomes, albumin microspheres, microemulsions, nanoparticles and nanocapsules), or in macroemulsions. Such techniques are disclosed in Remington's Pharmaceutical Sciences, supra.

Sustained-release preparations can be prepared. Suitable examples of sustained-release preparations include semipermeable matrices of solid hydrophobic polymers containing heteropolymers, which are formed products, such as films or microcapsules. Examples of sustained-release matrices include polyesters, hydrogels (e.g., poly(2-hydroxyethyl-methacrylate) or poly(vinyl alcohol)), polylactides (U.S. Patent No. 3,773,919), copolymers of L-glutamic acid and γ-ethyl-L-glutamate, non-degradable ethylene-vinyl acetate, degradable lactic acid-glycolic acid copolymers such as LUPRON DEPOT ^™ (injectable microspheres consisting of lactic acid-glycolic acid copolymer and leuprorelin acetate), and poly-D-(-)-3-hydroxybutyric acid. Although polymers such as ethylene-vinyl acetate and lactic acid-glycolic acid allow the release of molecules for more than 100 days, some hydrogels release proteins for shorter periods of time. When the encapsulated heteropolymers of the present invention are maintained in vivo for a long time, they may denature or aggregate due to exposure to a humid environment at 37°C, resulting in loss of biological activity and possible changes in immunogenicity. Rational strategies can be designed for stabilization based on the relevant mechanisms. For example, if the aggregation mechanism is found to be intermolecular SS bond formation via thiol-disulfide exchange, stabilization can be achieved by modifying sulfhydryl groups, lyophilizing from acidic solutions, controlling humidity, using appropriate additives, and developing specific polymer matrix compositions.

The proteins described herein (e.g., single-chain monospecific or multispecific antibodies containing an HD tether, multi-chain monospecific or multispecific heteromultimers containing an HD tether, immunoadhesin-antibody complexes) are administered to human subjects according to known methods, such as intravenous administration as a bolus or by continuous infusion over a period of time, intramuscularly, intraperitoneally, intracerebrospinal, subcutaneously, intraarticularly, intrasynovially, intrathecally, orally, topically, or by inhalation. Topical administration is particularly desirable if antagonism of the target molecule recognized by the protein is associated with extensive side effects or toxicity. Therapeutic applications may also employ ex vivo strategies. Ex vivo strategies involve transfecting or transducing cells obtained from a subject with a polynucleotide encoding a heteromultimer of the invention. The transfected or transduced cells are then returned to the subject. The cells can be any of a variety of types, including, but not limited to, hematopoietic cells (e.g., bone marrow cells, macrophages, monocytes, dendritic cells, T cells, or B cells), fibroblasts, epithelial cells, endothelial cells, keratinocytes, or muscle cells.

In one embodiment, when the condition or location of the tumor permits, the heteromultimers of the invention are administered locally, for example, by direct injection, and the injections can be repeated periodically. The protein complexes can also be delivered systemically to a subject or directly to tumor cells, for example, to a tumor or to a tumor bed following surgical resection of a tumor, to prevent or reduce local recurrence or metastasis.

VII. Products

Another embodiment of the present invention is an article of manufacture containing one or more heteromultimers described herein, and materials for treating or diagnosing any condition for which there is a suitable candidate target for producing the heteromultimer. The article of manufacture comprises a container and a label or package insert on or associated with the container. Suitable containers include, for example, bottles, vials, syringes, and the like. The container can be made of a variety of materials, such as glass or plastic. The container contains a composition that is effective for treating the condition and may have a sterile access port (for example, the container may be an intravenous solution bag or vial with a stopper that can be punctured by a hypodermic needle). At least one active agent in the composition is a heteromultimer of the present invention. The label or package insert indicates that the composition is used to treat a specific condition. The label or package insert also includes instructions for administering the heteromultimer composition to a subject. Articles of manufacture and kits comprising the combination therapies described herein are also contemplated.

Package insert means instructions customarily included in commercial packages of therapeutic products, that contain information about the indications, usage, dosage, administration, contraindications and/or warnings concerning the use of such therapeutic products.

Additionally, the article of manufacture may comprise a second container comprising a pharmaceutically-acceptable buffer, such as bacteriostatic water for injection (BWFI), phosphate-buffered saline, Ringer's solution, and dextrose solution. It may also include other materials considered appropriate from a commercial and user standpoint, including other buffers, diluents, filters, needles, and syringes.

Also provided are kits for various purposes, such as purification or immunoprecipitation of antigens from cells. For the isolation and purification of antigens, the kit may comprise heteromultimers coupled to beads (e.g., sepharose beads). Kits comprising heteromultimers can be provided for in vitro detection and quantification of antigens, such as in ELISA or Western blots. As with the article, the kit comprises a container and a label or package insert on or associated with the container. The container contains a composition comprising at least one heteromultimer of the present invention. Additional containers may be included, comprising, for example, a diluent and a buffer or a control heteromultimer. The label or package insert may provide a description of the composition and a description of the intended in vitro or diagnostic use.

VIII. Target Molecules

Examples of molecules that can be targeted by the heteromultimers of the invention (e.g., single-chain or multi-chain antibodies, or immunoadhesin-antibody complexes) include, but are not limited to, soluble serum proteins, receptor proteins, membrane-bound proteins (e.g., adhesins), cytokines, chemokines, growth factors, and hormones.

Soluble antigens or fragments thereof, optionally conjugated to other molecules, can be used as immunogens for producing heteromultimers. For transmembrane molecules, such as receptors, fragments thereof (e.g., the extracellular domain of a receptor) can be used as immunogens. Alternatively, cells expressing transmembrane molecules can be used as immunogens. Such cells can be derived from natural sources (e.g., cancer cell lines), or can be cells that have been transformed with recombinant technology to express transmembrane molecules. Other antigens and forms thereof for preparing heteromultimers will be apparent to those skilled in the art.

The above written description is considered sufficient to allow one skilled in the art to practice the present invention. The following examples are provided for illustrative purposes only and are not intended to limit the scope of the present invention in any way. Indeed, various modifications in addition to those shown and described herein will be apparent to those skilled in the art based on the above description and fall within the scope of the appended claims. The illustrations provided in the examples are intended to be considered as part of the present invention as a whole.

Example

Example 1. Construction of a vector for expressing HD-tethered heteromultimers

All heteromultimeric single-chain antibody constructs containing HD tethers were prepared in-house from existing IgG1 monoclonal antibody constructs. LC was prepared using PCR, where the 3' primer for LC extended beyond the LC C-terminus, encoded the N-terminal portion of the CLH tether, and terminated in a BamHI site. The 5' LC primer terminated in a ClaI site. The PCR fragment was then digested and cloned into a similarly digested and dephosphorylated pRK vector (Genentech Inc.; Eaton et al., Biochemistry 25:8343-8347 (1986)). The cognate HC was then prepared using PCR, where the 5' primer extended beyond the HC N-terminus (minus the signal sequence), encoded the C-terminal portion of the CLH tether, and terminated in a BamHI site. The 3' HC primer terminated in an XbaI site. The HC PCR fragment was digested and cloned into a similarly prepared plasmid containing the cognate LC as described above, thereby forming a linked half-antibody. The BamHI site was positioned so that it encodes the amino acid residue "GS" and maintains the integrity of the tethered GGS repeats. The lysine residue at position K222 (Kabat numbering scheme) was then mutated to an alanine residue using Stratagene's Quikchange II XL site-directed mutagenesis kit.

The two CLH-linked half antibodies described above are then joined together by another round of PCR, wherein the half antibody intended to be the N-terminus of the single-chain antibody is amplified using the same 5' LC primer used above and a 3' primer that extends beyond the C-terminus of the HC, encodes the N-terminus of the HD tether, and terminates in a BspEI site. The PCR fragment is digested and cloned into a similarly digested and dephosphorylated pRK vector. The half antibody intended to be the C-terminus of the heteromultimeric single-chain antibody is amplified using the same 3' HC primer used in the construction of the half antibody and a 5' primer that extends beyond the N-terminus of the LC (minus the signal sequence), encodes the C-terminal portion of the HD tether, and terminates in a BspEI site. The PCR product is digested and cloned into a similarly digested and dephosphorylated N-terminal construction vector. Similar to BamHI, BspEI can be positioned so that it maintains the integrity of the GGS repeats of the tether.

The construction of heteromultimeric multi-chain antibodies containing HD tethers is similar, differing only in that the N-terminus of the "half antibody" begins with an HC signal sequence and its cognate LC is expressed separately from a second pRK plasmid.

Example 2. Cleavage and Purification of HD Tethered Heteromultimers

An exemplary protocol that can be used to purify the heteromultimers of the present invention is shown below.

Load the heteromultimer onto a Protein A (e.g. mAbSure) column at 4°C

↓

Wash the column with KPO ₄ and then with PBS + 0.1% Trition X114

↓

Elute the sample with 100 mM acetic acid (pH 2.9)

↓

The pH of the sample was adjusted to pH 8.0 with 1 M Tris and then digested at 37°C for 1.5 hr at a 1:500 (wt:wt) ratio.

Lys-C

↓

The sample was purified on mAbSure resin to remove Lys-C

↓

The samples were eluted with 100 mM acetic acid (pH 2.9) and neutralized to pH 8 with 1 M Tris.

↓

The samples were cleaved with carboxypeptidase B 1:5 (wt:wt) at 37°C for 2.5 hr

↓

Adjust the sample pH and load onto a cation exchange column in 20 mM sodium acetate pH 5.0 with a gradient to 20 mM sodium acetate + 1 M NaCl.

↓

Collect fractions, pool & dialyze into PBS

Specifically, heteromultimers were purified from conditioned medium overnight at 4°C using mAbSure Select resin from GE Healthcare (Sweden). The column was washed with 2 column volumes (CV) of PBS (phosphate buffered saline), followed by 10CV of PBS + 0.1% Triton X114 detergent, and then 10CV of potassium phosphate buffer. The column was eluted with 100mM acetic acid (pH 2.9) and immediately adjusted to pH 8.0 with Tris (final concentration 200mM) pH 8.0. The HD tether was removed from the heteromultimer after treatment with Lys-C endopeptidase () at a ratio of 1:500 (wt:wt) at 37°C for 1-5 hours. The sample was then loaded onto the column to remove the enzyme. The residual furin recognition site was then removed by treatment with porcine carboxypeptidase B (Roche Applied Sciences, Cat# 10103233001) at a ratio of 1:5 (wt:wt). The cleaved sample was loaded onto a cation exchange column to remove the enzyme. Peak fractions were pooled and dialyzed against PBS overnight before mass spectrometry analysis to ensure identity and purity.

Example 3. Identification of engineered heteromultimers

To determine whether exemplary heteromultimers constructed using the HD tethered heteromultimerization technology of the present invention retain the binding properties of the antibody or adhesin from which their sequences were derived, binding assays were performed (Figures 6A-6D). Binding assays were performed using the kinetics wizard program of the ForteBio Octet system. All samples were tested at a concentration of 25 μg/ml, which accounts for saturation of the anti-human IgG probe in replicate experiments and different samples. The probe was loaded with the first sample for 10 minutes, followed by a 30 second wash in PBS. All associations with the second and third samples were performed for 15 minutes, with a 30 second wash in PBS between associations.

Other implementations

All patents, patent applications, patent application publications, and other publications cited or referenced in this specification are hereby incorporated by reference to the same extent as if each independent patent, patent application, patent application publication, or publication was specifically and individually indicated to be incorporated by reference.

Claims (113)

1. A single-chain antibody comprising a single polypeptide, said polypeptide comprising the following domains arranged relative to each other in the N-terminal to C-terminal direction: _VL1 - _CL1 -CLH chain ₁ - _VH1 - _CH11 -hinge ₁ - _CH21 -CH31 _- HD chain- _{VL2 -CL2} -CLH chain _{2- VH2} - _CH12 -hinge _{2 -CH22} - _CH32 , wherein said single-chain antibody is a bispecific or multispecific single-chain antibody, wherein said HD chain is 30-39 amino acids long and contains glycine residues abbreviated as G and serine residues abbreviated as S, and an amino acid sequence cleavable by endopeptidase; said CLH chain ₁ and said CLH chain ₂ are each 20-40 amino acids long and contain G and S residues, and an amino acid sequence cleavable by endopeptidase. 2. The single-chain antibody of claim 1, wherein the HD lineage comprises GGS repeats, and the CLH lineage ₁ and the CLH lineage ₂ comprise GGS repeats. 3. The single-chain antibody of claim 2, wherein the HD lineage comprises 8 to 9 GGS repeats. 4. The single-chain antibody of claim 3, wherein the HD chain comprises the amino acid sequence shown in SEQ ID NO:19. 5. The single-chain antibody of claim 1, wherein the amino acid sequence is cleaved in situ by the endopeptidase. 6. The single-chain antibody of claim 5, wherein the amino acid sequence is cleaved after purification and upon addition of the endopeptidase. 7. The single-chain antibody of claim 5, wherein the HD chain comprises two amino acid sequences that can be cleaved by endopeptidase, the amino acid sequences being located at the N and C ends of the HD chain. 8. The single-chain antibody of claim 1, wherein the endopeptidase is selected from the group consisting of furin protease, urokinase, thrombin, tissue plasminogen activator, genenase, Lys-C, Arg-C, Asp-N, Glu-C, factor Xa, tobacco erosion virus protease, enterokinase, human rhinovirus C3 protease, and kininogenase. 9. The single-chain antibody of claim 8, wherein the endopeptidase is furin protease. 10. The single-chain antibody of claim 9, wherein the HD chain comprises two endopeptidase cleavage sites located at the N and C ends of the HD chain, wherein one of the endopeptidase cleavage sites is a furin cleavage site. 11. The single-chain antibody of claim 9, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RKRKRR shown in SEQ ID NO:9. 12. The single-chain antibody of claim 9, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RHRQPR shown in SEQ ID NO:10. 13. The single-chain antibody of claim 10, wherein one of the endopeptidase cleavage sites is a Lys-C cleavage site. 14. The single-chain antibody of claim 1, wherein the endopeptidase is furin protease. 15. The single-chain antibody of claim 14, wherein the CLH chain ₁ or the CLH chain ₂ comprises two endopeptidase cleavage sites located at the N and C ends of the CLH chain ₁ or the CLH chain ₂ , wherein the endopeptidase cleavage sites are furin cleavage sites. 16. The single-chain antibody of claim 15, wherein each of the CLH chain ₁ and the CLH chain ₂ comprises two endopeptidase cleavage sites located at the N and C ends of the CLH chain ₁ and the CLH chain ₂ , and wherein the endopeptidase cleavage sites are furin protease cleavage sites. 17. The single-chain antibody of claim 14, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RKRKRR shown in SEQ ID NO:9. 18. The single-chain antibody of claim 14, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RHRQPR shown in SEQ ID NO:10. 19. The single-chain antibody of claim 1, wherein the first hinge domain or the second hinge domain comprises Glu216 to Pro230 of human IgG1. 20. The single-chain antibody of claim 19, wherein the first hinge domain or the second hinge domain comprises a Lys-C endopeptidase cleavage site. 21. The single-chain antibody of claim 20, wherein the Lys-C endopeptidase cleavage site comprises an inactivating mutation. 22. The single-chain antibody of claim 21, wherein the inactivating mutation is a K222A substitution in the EU numbering system. 23. The single-chain antibody of claim 1, wherein the single-chain antibody is capable of binding at least two antigens. 24. The single-chain antibody of claim 1, wherein the single-chain antibody is capable of binding to at least two epitopes on the same antigen. 25. The single-chain antibody of claim 1, wherein the single-chain antibody comprises a constant region conjugated to a functional portion on at least one Fc component. 26. The single-chain antibody of claim 1, wherein the single-chain antibody, after cleaving one or more of the HD or CLH lineage chains, contains one or more cleavage sites of an exopeptidase. 27. The single-chain antibody of claim 26, wherein the one or more cleavage sites are cleaved in situ by the exopeptidase, or wherein the one or more cleavage sites are cleaved after purification and subsequent addition of the exopeptidase. 28. The single-chain antibody of claim 26, wherein the exopeptidase is selected from the group consisting of carboxypeptidase A, carboxypeptidase B, carboxypeptidase D, carboxypeptidase E, carboxypeptidase M, carboxypeptidase N, and carboxypeptidase Z. 29. The single-chain antibody of claim 28, wherein the carboxypeptidase B is plasma carboxypeptidase B. 30. The single-chain antibody of claim 28, wherein the exopeptidase cleaves at a basic residue. 31. The single-chain antibody of claim 30, wherein the exopeptidase is carboxypeptidase B. 32. The single-chain antibody of claim 1, wherein the _CH31 and _CH32 domains each comprise a protrusion or a cavity, and wherein the protrusion or cavity in the _CH31 domain can be respectively placed in the cavity or protrusion in the _CH32 domain. 33. The single-chain antibody of claim 32, wherein the _CH31 and _CH32 domains meet at the interface between the protrusion and the cavity. 34. The single-chain antibody of claim 1, wherein the polypeptide further comprises: (a) the protrusions or cavities in the _CH11 domain or the _CH12 domain, or both, and (b) A protrusion or cavity in the CL ₁ domain or the CL ₂ domain, or both; and wherein: (c) The protrusions or cavities in the CH1 ₁ structural domain may be respectively placed in the cavities or protrusions in the CL ₁ structural domain. (d) The protrusions or cavities in the CH1 ₂ structural domain may be respectively placed in the cavities or protrusions in the CL ₂ structural domain; or (e), (c), and (d) are both. 35. The single-chain antibody of claim 34, wherein the _CH11 and _CL1 domains, the _CH12 and _CL2 domains, or all four of the domains meet at the interface between the protrusion and the cavity. 36. The single-chain antibody of claim 1, wherein one or more of the CH2 ₁ or CH2 ₂ domains contain a CH2 domain mutation that affects the function of the antibody effector. 37. The single-chain antibody of claim 36, wherein the CH2 domain mutation affecting the function of the antibody effector is an N297 mutation. 38. A polynucleotide encoding a single-chain antibody of any one of claims 1-37. 39. A vector comprising the polynucleotide of claim 38. 40. A host cell comprising the vector of claim 39. 41. The host cell of claim 40, wherein the host cell is a mammalian cell. 42. The host cell of claim 41, wherein the mammalian cell is a CHO cell. 43. The host cell of claim 40, wherein the host cell is a prokaryotic cell. 44. The host cell of claim 43, wherein the prokaryotic cell is an Escherichia coli cell. 45. A method for producing a single-chain antibody of any one of claims 1-37, the method comprising culturing host cells containing the vector of claim 39 in a culture medium. 46. The method of claim 45, wherein the method further comprises recovering the single-chain antibody from the host cell or the culture medium. 47. Antibodies, comprising the following components: (a) A first polypeptide comprising a first VL (VL ₁ ) domain and a first CL (CL ₁ ) domain; and (b) The second polypeptide, comprising: (i) The first VH(VH ₁ ) and the second VH(VH ₂ ) structural domains, (ii) HD chain. (iii) The second VL (VL ₂ ) and the second CL (CL ₂ ) structural domains, (iv) CLH chain, (v) The first hinge (hinge ₁ ) structural domain or the second hinge (hinge ₂ ) structural domain, or both; and (vi) One or more heavy-chain constant structural domains, The one or more heavy chain constant structural domains are selected from the first CH1 (CH1 ₁ ) structural domain, the first CH2 (CH2 ₁ ) structural domain, the first CH3 (CH3 ₁ ) structural domain, the second CH1 (CH1 ₂ ) structural domain, the second CH2 (CH2 ₂ ) structural domain, and the second CH3 (CH3 ₂ ) structural domain; and The components of the second polypeptide are arranged relative to each other in the N-terminus to C-terminus direction as follows: VH ₁ -CH1 ₁ -hinge ₁ -CH2 ₁ -CH3 ₁ -HD chain -VL ₂ -CL ₂ -CLH chain -VH ₂ -CH1 ₂ -hinge ₂ -CH2 ₂ -CH3 ₂ , wherein the HD chain is 30-39 amino acids long and contains glycine residues abbreviated as G and serine residues abbreviated as S, as well as an amino acid sequence that can be cleaved by endopeptidases; the CLH chain is 20-40 amino acids long and contains G and S residues, as well as an amino acid sequence that can be cleaved by endopeptidases. 48. The antibody of claim 47, wherein at least one of the one or more heavy chain constant domains is paired with another heavy chain constant domain. 49. The antibody of claim 47, wherein the HD lineage comprises GGS repeats. 50. The antibody of claim 49, wherein the HD lineage comprises 8 to 9 GGS repeats. 51. The antibody of claim 50, wherein the HD lineage comprises the amino acid sequence shown in SEQ ID NO:19. 52. The antibody of claim 47, wherein the amino acid sequence is cleaved in situ by the endopeptidase. 53. The antibody of claim 47, wherein the amino acid sequence is cleaved after purification and upon addition of the endopeptidase. 54. The antibody of claim 47, wherein the HD chain comprises two amino acid sequences that can be cleaved by endopeptidase, the amino acid sequences being located at the N and C ends of the HD chain. 55. The antibody of claim 47, wherein the endopeptidase is selected from the group consisting of furin protease, urokinase, thrombin, tissue plasminogen activator (tPa), genenase, Lys-C, Arg-C, Asp-N, Glu-C, factor Xa, tobacco erosion virus protease (TEV), enterokinase, human rhinovirus C3 protease (HRV C3), and kininogenase. 56. The antibody of claim 47, wherein the endopeptidase is furin protease. 57. The antibody of claim 47, wherein the HD lineage includes two endopeptidase cleavage sites located at the N and C ends of the HD lineage, wherein one of the endopeptidase cleavage sites is a furin cleavage site. 58. The antibody of claim 47, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RKRKRR shown in SEQ ID NO:9. 59. The antibody of claim 47, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RHRQPR shown in SEQ ID NO:10. 60. The antibody of claim 47, wherein one of the endopeptidase cleavage sites is a Lys-C cleavage site. 61. The antibody of claim 47, wherein the CLH lineage comprises GGS repeats. 62. The antibody of claim 61, wherein the amino acid sequence is cleaved in situ by the endopeptidase. 63. The antibody of claim 61, wherein the amino acid sequence is cleaved after purification and upon addition of the endopeptidase. 64. The antibody of claim 47, wherein the endopeptidase is selected from the group consisting of furin protease, urokinase, thrombin, tissue plasminogen activator, genenase, Lys-C, Arg-C, Asp-N, Glu-C, factor Xa, tobacco erosion virus protease, enterokinase, human rhinovirus C3 protease, and kininogenase. 65. The antibody of claim 64, wherein the endopeptidase is furin protease. 66. The antibody of claim 65, wherein the CLH lineage includes two endopeptidase cleavage sites located at the N and C ends of the CLH lineage, wherein the endopeptidase cleavage sites are furin cleavage sites. 67. The antibody of claim 65, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RKRKRR shown in SEQ ID NO:9. 68. The antibody of claim 65, wherein the amino acid sequence cleavable by furin protease comprises the amino acid sequence RHRQPR shown in SEQ ID NO:10. 69. The antibody of claim 47, wherein the first hinge domain or the second hinge domain comprises Glu216 to Pro230 of human IgG1. 70. The antibody of claim 69, wherein the first hinge domain or the second hinge domain comprises a Lys-C endopeptidase cleavage site. 71. The antibody of claim 70, wherein the Lys-C endopeptidase cleavage site comprises an inactivating mutation. 72. The antibody of claim 71, wherein the inactivating mutation is a K222A substitution in the EU numbering system. 73. The antibody of claim 47, wherein the antibody is a monospecific antibody. 74. The antibody of claim 47, wherein the antibody is a bispecific or multispecific antibody. 75. The antibody of claim 74, wherein the antibody is capable of binding to at least two antigens. 76. The antibody of claim 75, wherein the antibody is capable of binding to at least two epitopes on the same antigen. 77. The antibody of claim 47, wherein the antibody comprises a constant region conjugated to a cytotoxic agent. 78. The antibody of claim 47, wherein the antibody, after cleaving the HD lineage, contains one or more cleavage sites of an exopeptidase. 79. The antibody of claim 47, wherein the antibody, after cleaving the CLH lineage, contains one or more cleavage sites of an exopeptidase. 80. The antibody of claim 78, wherein the one or more cleavage sites are cleaved in situ by the exopeptidase; or the one or more cleavage sites are cleaved after purification and subsequent addition of the exopeptidase. 81. The antibody of claim 79, wherein the one or more cleavage sites are cleaved in situ by the exopeptidase; or the one or more cleavage sites are cleaved after purification and subsequent addition of the exopeptidase. 82. The antibody of claim 78, wherein the exopeptidase is selected from the group consisting of carboxypeptidase A, carboxypeptidase B, carboxypeptidase D, carboxypeptidase E, carboxypeptidase M, carboxypeptidase N and carboxypeptidase Z. 83. The antibody of claim 79, wherein the exopeptidase is selected from the group consisting of carboxypeptidase A, carboxypeptidase B, carboxypeptidase D, carboxypeptidase E, carboxypeptidase M, carboxypeptidase N and carboxypeptidase Z. 84. The antibody of claim 82 or 83, wherein the carboxypeptidase B is plasma carboxypeptidase B. 85. The antibody of claim 82 or 83, wherein the exopeptidase cleaves at a basic residue. 86. The antibody of claim 85, wherein the exopeptidase is carboxypeptidase B. 87. The antibody of claim 47, wherein each of the _CH31 and _CH32 domains comprises a protrusion or a cavity, and wherein the protrusion or cavity in the _CH31 domain may be respectively placed in the cavity or protrusion in the _CH32 domain. 88. The antibody of claim 87, wherein the _CH31 and _CH32 domains meet at the interface between the protrusion and the cavity. 89. The antibody of any one of claims 47-83, 87-88, wherein the polypeptide further comprises: (a) the protrusions or cavities in the _CH11 domain or the _CH12 domain, or both, and (b) A protrusion or cavity in the CL ₁ domain or the CL ₂ domain, or both; and wherein: (c) The protrusions or cavities in the CH1 ₁ structural domain may be respectively placed in the cavities or protrusions in the CL ₁ structural domain. (d) The protrusions or cavities in the CH1 ₂ structural domain may be respectively placed in the cavities or protrusions in the CL ₂ structural domain; or (e), (c), and (d) are both. 90. The antibody of claim 89, wherein the _CH11 and _CL1 domains, the _CH12 and _CL2 domains, or all four of the domains meet at the interface between the protrusion and the cavity. 91. The antibody of claim 47, wherein one or more of the CH2 ₁ or CH2 ₂ domains contain a CH2 domain mutation that affects the function of the antibody effector. 92. The antibody of claim 91, wherein the CH2 domain mutation affecting the function of the antibody effector is an N297 mutation. 93. A polynucleotide encoding an antibody of any one of claims 47-92. 94. A vector comprising the polynucleotide of claim 93. 95. A host cell comprising the vector of claim 94. 96. The host cell of claim 95, wherein the host cell is a mammalian cell. 97. The host cell of claim 96, wherein the mammalian cell is a CHO cell. 98. The host cell of claim 95, wherein the host cell is a prokaryotic cell. 99. The host cell of claim 98, wherein the prokaryotic cell is Escherichia coli. 100. A method for producing an antibody according to any one of claims 47-92, the method comprising culturing host cells containing the vector of claim 94 in a culture medium. 101. The method of claim 100, wherein the method further comprises recovering the antibody from the host cell or the culture medium. 102. A single-chain antibody comprising a single polypeptide, said polypeptide comprising the following domains arranged relative to each other in the N-terminal to C-terminal direction: _VL1 - _CL1 -CLH chain ₁ - _VH1 - _CH11 -hinge ₁ - _CH21 -CH31 _- HD chain- _{VL2 -CL2} -CLH chain _{2- VH2} - _CH12 -hinge _{2 -CH22} - _CH32 , wherein said single-chain antibody is a bispecific or multispecific single-chain antibody, wherein said HD chain is 30-39 amino acids long and contains glycine residues abbreviated as G and serine residues abbreviated as S, and an amino acid sequence cleavable by furin endopeptidase; said CLH chain ₁ and CLH chain ₂ are each 20-40 amino acids long and contain G and S residues, and an amino acid sequence cleavable by furin endopeptidase. 103. The single-chain antibody of claim 102, wherein the sequence cleavable by furin protease comprises the amino acid sequence RKRKRR shown in SEQ ID NO:9. 104. The single-chain antibody of claim 102, wherein the sequence cleavable by furin protease comprises the amino acid sequence RHRQPR shown in SEQ ID NO:10. 105. A polynucleotide encoding a single-chain antibody of any one of claims 102-104. 106. A vector comprising the polynucleotide of claim 105. 107. A host cell comprising the vector of claim 106. 108. The host cell of claim 107, wherein the host cell is a mammalian cell. 109. The host cell of claim 108, wherein the mammalian cell is a CHO cell. 110. The host cell of claim 107, wherein the host cell is a prokaryotic cell. 111. The host cell of claim 110, wherein the prokaryotic cell is Escherichia coli. 112. A method for producing a single-chain antibody according to any one of claims 102-104, the method comprising culturing host cells containing the vector of claim 106 in a culture medium. 113. The method of claim 112, wherein the method further comprises recovering single-chain antibodies from the host cells or the culture medium.