HK1169132B - Bispecific death receptor agonistic antibodies - Google Patents

Description

Bispecific death receptor-activating antibodies

The present invention relates to bispecific antibodies comprising a first antigen-binding site specific for a death receptor (detahrecter) and a second antigen-binding site specific for a second antigen, methods for their production, pharmaceutical compositions comprising said antibodies, and uses thereof.

Due to the selective targeting of antigens that are differentially expressed on cancer cells, monoclonal antibodies have proven to be effective therapeutic agents in the treatment of cancer. Most of the therapeutic strategies currently developed for monoclonal antibodies include targeting tumor-associated antigens to modify tumor cell biology, inhibit growth factor receptors, inhibit angiogenesis, induce apoptosis and cytotoxicity through complement fixation or antibody-dependent cytotoxicity. Some antibodies target growth factor receptors that are critically important for cancer cell survival, e.g., trastuzumabAnd cetuximabTargeting TRAIL death receptors on cancer cells with agonist monoclonal antibodies represents a new generation of monoclonal antibody therapy because of its ability to directly induce apoptosis of the targeted cells. It may be advantageous to use an agonistic monoclonal antibody directed against the death receptor (rather than TRAIL): TRAIL targets multiple receptors including death receptors and trap receptors (decoyrecteptors), thus affecting selectivity. In addition, TRAIL has a much shorter blood half-life than monoclonal antibodies, a factor that affects dosage and schedule parameters. The very short blood half-life of TRAIL compared to monoclonal antibodies would require large and frequent doses. In addition, the production of recombinant TRAIL is very difficult and time-consuming.

Michaelson J.S. et al (mAbs, Vol.1, No.2, p: 128-141; 3/4 months 2009) describe engineered IgG-like bispecific antibodies that target 2 TNF family member receptors, TRAIL-R2 (TNF-related apoptosis-inducing ligand receptor-2) and LT β R (lymphotoxin β receptor).

HerrmannT et al (cancer Res 2008; 68 (4); p: 1221-1227) describe the expression of a protein against CD95/Fas/Apo-1 cell surface receptors and 3 target antigens on glioblastoma cells: bispecific monovalent chemically conjugated Fab molecules of NG2, EGFR and CD 40.

The present invention relates to antibodies that combine an antigen-binding site that targets a death receptor and a second antigen-binding site that targets a second antigen. Whereby the death receptor becomes cross-linked and induces apoptosis of the target cell. These bispecific death receptor-activated antibodies have the advantage over traditional antibodies targeting death receptors of inducing the specificity of apoptosis only at the site where the second antigen is expressed.

In a first aspect (object), the invention relates to a bispecific antibody comprising a first antigen-binding site specific for a death receptor antigen and a second antigen-binding site specific for a second antigen.

In a preferred embodiment of the bispecific antibody, the death receptor is selected from the group consisting of a death receptor 4 polypeptide (DR4), a death receptor 5 polypeptide (DR5) or a FAS polypeptide, preferably a human DR4 polypeptide (seq. id.no.1), a human DR5 polypeptide (seq. id.no.2) or a human FAS polypeptide (seq. id.no. 3).

In a further preferred embodiment of the bispecific antibody, the second antigen is associated with a tumor disease or rheumatoid arthritis.

In a further preferred embodiment of the bispecific antibody, the second antigen is selected from the group consisting of a carcinoembryonic antigen (CEA) polypeptide, a CRIPTO protein, a magic detour (magic fundabout) homolog 4(ROBO4) polypeptide, a melanoma-associated chondroitin sulfate proteoglycan (MCSP) polypeptide, a tenascin C polypeptide and a Fibroblast Activation Protein (FAP) polypeptide, preferably a human CEA polypeptide (seq.id.no.4), a human CRIPTO polypeptide (seq.id.no.5), a human ROBO4 polypeptide (seq.id.no.6), a human MCSP polypeptide (seq.id.no.7), a human tenascin C polypeptide (seq.id.no.8) and a human FAP polypeptide (seq.id.id.no. 9).

In a further preferred embodiment of the bispecific antibody, the bispecific antibody is a dimeric molecule comprising a first antibody and a second antibody, said first antibody comprising a first antigen-binding site and said second antibody comprising a second antigen-binding site.

In a preferred embodiment of the dimeric bispecific antibody of the invention, the first and second antibodies comprise the Fc part of the heavy chain of the antibody, wherein the Fc part of the first antibody comprises a first dimerization module and the Fc part of the second antibody comprises a second dimerization module, thereby allowing heterodimerization of the 2 antibodies.

In a further preferred embodiment of the dimeric bispecific antibody, the first dimerization module comprises a knob (knob) and the second dimerization module comprises a knob (hole) according to the knob-hole structure (knobs) strategy (see Carter P.; RidgwayJ. B.; PrestaL. G.: Immunotechnology, Vol.2, No.1, 2 months 1996, pp.73-73 (1)).

In a further preferred embodiment of the dimeric bispecific antibody, the first antibody is an immunoglobulin (Ig) molecule comprising a light chain and a heavy chain, and the second antibody is selected from scFv, scFab, Fab or Fv.

In a further preferred embodiment, the bispecific antibody comprises a modified Fc part having reduced binding affinity for fey receptors compared to the wild-type Fc part, e.g. a LALA modification.

In a further preferred embodiment of the dimeric bispecific antibody, the Ig molecule comprises a first antigen-binding site specific for a death receptor and the second antibody comprises a second antigen-binding site specific for a second antigen.

In a further preferred embodiment of the bispecific antibody, the Ig molecule comprises a second antigen-binding site specific for a second antigen, and the second antibody comprises an antigen-binding site specific for a death receptor.

In a further preferred embodiment of the dimeric bispecific antibody, the second antibody is fused to the N-or C-terminus of the heavy chain of the Ig molecule.

In a further preferred embodiment of the dimeric bispecific antibody, the second antibody is fused to the N-or C-terminus of the light chain of the Ig molecule.

In a further preferred embodiment of the dimeric bispecific antibody, the Ig molecule is an IgG. In a further preferred embodiment of the dimeric bispecific antibody, the second molecule is fused to the Ig molecule by a peptide linker, preferably a peptide linker having a length of about 10-30 amino acids.

In a further preferred embodiment of the dimeric bispecific antibody, the second antibody comprises an additional cysteine residue to form a disulfide bond.

Bispecific antibodies according to the invention are at least bivalent and may be trivalent or multivalent, e.g. tetravalent or hexavalent.

In a second aspect, the invention relates to a pharmaceutical composition comprising a bispecific antibody of the invention.

In a third aspect, the invention relates to a bispecific antibody of the invention for use in the treatment of cancer or rheumatoid arthritis.

In further aspects, the invention relates to a nucleic acid sequence comprising a sequence encoding a heavy chain of a bispecific antibody of the invention, a nucleic acid sequence comprising a sequence encoding a light chain of a bispecific antibody of the invention, an expression vector comprising a nucleic acid sequence of the invention, and to a prokaryotic or eukaryotic host cell comprising a vector of the invention.

Detailed Description

The term "polypeptide" as used herein refers to polypeptides of the invention, i.e. native amino acid sequences and sequence variants of DR4, DR5, FAS, CEA, CRIPTO, ROBO4, MCSP, tenascin C and FAP from any animal, e.g. mammalian species (including humans).

"native polypeptide" refers to a polypeptide having the same amino acid sequence as a naturally occurring polypeptide, regardless of its mode of preparation. The term "native polypeptide" specifically encompasses naturally occurring truncated or secreted forms, naturally occurring variant forms (e.g., additional spliced forms), and naturally occurring allelic variants of the polypeptides of the invention. The amino acid sequence in the sequence listing (seq. Id. No.1-9) refers to the natural human sequence of the protein of the invention.

The term "polypeptide variant" refers to an amino acid sequence variant of a native sequence that comprises one or more amino acid substitutions and/or deletions and/or insertions in the native sequence. Amino acid sequence variants typically have at least about 75%, preferably at least about 80%, more preferably at least about 85%, even more preferably at least about 90%, and most preferably at least about 95% sequence identity with the amino acid sequence of the native sequence of the polypeptide of the invention.

The term "antibody" encompasses various forms of antibody structures, including, but not limited to, whole antibodies and antibody fragments. The antibody according to the invention is preferably a fully human antibody, a humanized antibody, a chimeric antibody, or other genetically engineered antibody, as long as it retains the characteristic properties according to the invention.

An "antibody fragment" comprises a portion of a full-length antibody, preferably its variable domain, or at least its antigen-binding site. Examples of antibody fragments include diabodies, single chain antibody molecules, and multispecific antibodies formed from antibody fragments. scFv antibodies are described, for example, in Houston, J.S., MethodsinEnzymol.203(1991) 46-96). In addition, antibody fragments comprise single chain polypeptides having the characteristics of a VH domain (i.e., capable of assembly with a VL domain), or having the characteristics of a VL domain (i.e., capable of assembly with a VH domain), to form a functional antigen binding site and thereby provide the antigen binding properties of a full-length antibody.

As used herein, the term "monoclonal antibody" or "monoclonal antibody composition" refers to a preparation of antibody molecules of a single amino acid composition.

The term "chimeric antibody" refers to an antibody comprising a variable or binding region from one source or species, and at least a portion of a constant region from a different source or species, typically prepared by recombinant DNA techniques. Chimeric antibodies comprising a murine variable region and a human constant region are preferred. Other preferred forms of "chimeric antibodies" encompassed by the present invention are those in which the constant regions have been modified or altered from the constant regions of the original antibody to produce a characteristic according to the invention, particularly with respect to C1q binding and/or Fc receptor (FcR) binding. Such chimeric antibodies are also referred to as "class switch antibodies". Chimeric antibodies are the expression product of an immunoglobulin gene comprising a DNA segment encoding an immunoglobulin variable region and a DNA segment encoding an immunoglobulin constant region. Methods for producing chimeric antibodies involve conventional recombinant DNA and gene transfection methods well known in the art. See, e.g., Morrison, S.L., et al, Proc.Natl.Acadsi.USA 81(1984) 6851-6855; U.S. patent nos. 5,202,238 and 5,204,244.

The term "humanized antibody" refers to an antibody in which the framework or "complementarity determining regions" (CDRs) have been modified to comprise the CDRs of an immunoglobulin with different specificity compared to the parent immunoglobulin. In a preferred embodiment, murine CDRs are grafted to the framework regions of a human antibody to make a "humanized antibody". See, e.g., Riechmann, L., et al, Nature332(1988) 323-327; and Neuberger, M.S. et al, Nature314(1985)268- > 270. Particularly preferred CDRs correspond to those providing sequences that recognize the antigen of the chimeric antibody described above. Other forms of "humanized antibodies" encompassed by the present invention are those in which the constant regions have been modified or altered from those of the original antibody to produce a humanized antibody having properties according to the present invention, particularly with respect to C1q binding and/or Fc receptor (FcR) binding.

As used herein, the term "human antibody" is intended to include antibodies having variable and constant regions derived from human germline immunoglobulin sequences. Human antibodies are well known in the art (vanDijk, m.a. and vandeWinkel, j.g., curr. opin. chem. biol.5(2001) 368-374). Human antibodies can also be produced in transgenic animals (e.g., mice) that are immunized to produce all or selected human antibodies without endogenous immunoglobulin production. Transfer of a human germline immunoglobulin gene array into such germline mutant mice results in the production of human antibodies upon antigen stimulation (see, e.g., Jakobovits, A. et al, Proc. Natl. Acad. Sci. USA90(1993) 2551-2555; Jakobovits, A. et al, Nature362(1993) 255-258; Bruggemann, M. et al, Yeast Immunol.7(1993) 33-40). Human antibodies can also be produced in phage display libraries (Hoogenboom, H.R. and Winter, G., J.Mol.biol.227(1992) 381-59388; Marks, J.D. et al, J.Mol.biol.222(1991) 581-597). Human monoclonal antibodies can also be prepared using the techniques of Cole et al and Boerner et al (Cole et al, monoclonal antibodies and cells therapy, Anan R. Liss, p.77 (1985); and Boerner, P.et al, J.Immunol.147(1991) 86-95). As already mentioned the chimeric and humanized antibodies according to the invention, the term "human antibody" as used herein also comprises antibodies in which the constant region is modified, for example by "class switching" (i.e. altering or mutating the Fc part, e.g. a mutation from IgG1 to IgG4 and/or IgG1/IgG 4) to produce such antibodies according to the invention, in particular with regard to C1q binding and/or Fc receptor (FcR) binding.

As used herein, the term "recombinant human antibody" is intended to include all human antibodies prepared, expressed, created or isolated by recombinant means, e.g., antibodies isolated from host cells such as NS0 or CHO cells or from animals (e.g., mice) transgenic for human immunoglobulin genes or antibodies expressed using recombinant expression vectors transfected into host cells. Such recombinant human antibodies have variable and constant regions in rearranged form. Recombinant human antibodies according to the invention can be highly mutated in vivo by somatic cells. Thus, the amino acid sequences of the VH and VL regions of the recombinant antibody are sequences that do not naturally occur within the human antibody germline repertoire in vivo, despite their origin and association with human germline VH and VL sequences.

As used herein, "variable domain" (light chain variable domain (VL), heavy chain variable domain (VH)) means each pair of light and heavy chain domains that are directly involved in binding of an antibody to an antigen. The light and heavy chain variable domains have the same general structure, and each region comprises four Framework (FR) regions widely conserved in sequence connected by three "hypervariable regions" (or complementarity determining regions, CDRs). The framework regions adopt a beta sheet conformation and the CDRs may form loops connecting the beta sheet structures. The CDRs in each chain are fixed in their three-dimensional structure by framework regions and form together with the CDRs from the other chains the antigen binding site. The antibody heavy and light chain CDR3 regions play a particularly important role in the binding specificity/affinity of the antibodies according to the invention and therefore provide further objects of the invention.

As used herein, the term "antigen-binding portion of an antibody" refers to the amino acid residues of an antibody that are responsible for antigen binding. The antigen-binding portion of an antibody comprises amino acid residues from a "complementarity determining region" or "CDR". The "framework" or "FR" regions are those variable region regions that are not hypervariable region residues as defined herein. Thus, the light and heavy chain variable domains of an antibody comprise, from N-terminus to C-terminus, the regions FR1, CDR1, FR2, CDR2, FR3, CDR3 and FR 4. In particular, the CDR3 of the heavy chain is the region that is most conducive to antigen binding and defines antibody performance. CDRs and FRs are determined according to the standard definition of Kabat et al, sequence of proteins of immunological interest, 5 th edition, public health service, national institutes of health, Bethesda, Md. (1991) and/or residues from "hypervariable loops".

Antibody specificity refers to the selective recognition of a particular epitope of an antigen by an antibody. For example, natural antibodies are monospecific. According to the invention, a "bispecific antibody" is an antibody having two different antigen binding specificities. The antibodies of the invention are specific for two different antigens, namely a death receptor antigen as the first antigen and a second antigen.

As used herein, the term "monospecific" antibody refers to an antibody having one or more binding sites, wherein each binding site binds to the same epitope of the same antigen.

The term "bispecific" antibody, as used herein, refers to an antibody having at least two binding sites, wherein each binding site binds to the same antigen or a different epitope of a different antigen.

As used herein, the term "valency" refers to the specific number of binding sites present in an antibody molecule. Thus, the terms "bivalent", "tetravalent", and "hexavalent" refer to the presence of two binding sites, four binding sites, and six binding sites, respectively, in an antibody molecule. Bispecific antibodies according to the present invention are at least "bivalent" and may be "trivalent" or "multivalent" (e.g., "tetravalent" or "hexavalent").

The antibodies of the invention have two or more binding sites and are bispecific. That is, an antibody can be bispecific even in the presence of more than two binding sites (i.e., the antibody is trivalent or multivalent). Bispecific antibodies of the invention include, for example, multivalent single chain antibodies, diabodies and triabodies, as well as antibodies having the constant domain structure of a full-length antibody to which additional antigen binding sites (e.g., single chain Fv, VH and/or VL domains, Fab or (Fab)2) can be attached by one or more peptide linkers. The antibody may be a full length antibody from a single species, or chimeric or humanized.

A "single chain Fab fragment" is a polypeptide consisting of an antibody heavy chain variable domain (VH), an antibody constant domain 1(CH1), an antibody light chain variable domain (VL), an antibody light chain constant domain (CL) and a linker, wherein the antibody domain and the linker have one of the following sequences in the N-terminal to C-terminal direction:

a) VH-CH 1-linker-VL-CL, b) VL-CL-linker-VH-CH 1, c) VH-CL-linker-VL-CH 1 or d) VL-CH 1-linker-VH-CL; wherein the linker is a polypeptide of at least 30 amino acids, preferably between 32 and 50 amino acids. The single chain Fab fragment a) VH-CH 1-linker-VL-CL, b) VL-CL-linker-VH-CH 1, c) VH-CL-linker-VL-CH 1 and d) VL-CH 1-linker-VH-CL are stabilized by a natural disulfide bond between the CL domain and the CH1 domain. In addition, these single chain Fab molecules may be further stabilized by the formation of interchain disulfide bonds between the positions 44 and 100 of the heavy chain variable region and the light chain variable region, respectively, by insertion of cysteine residues (e.g., between numbering according to Kabat). The term "N-terminal" refers to the last amino acid of the N-terminus. The term "C-terminal" refers to the last amino acid at the C-terminus.

As used herein, the term "nucleic acid" or "nucleic acid molecule" is intended to include DNA molecules and RNA molecules. The nucleic acid molecule may be single-stranded or double-stranded, but is preferably double-stranded DNA.

The term "amino acid" as used in the present application refers to the group of naturally occurring carboxy α -amino acids, which includes alanine (three letter code: ala, one letter code: A), arginine (arg, R), asparagine (asn, N), aspartic acid (asp, D), cysteine (cys, C), glutamine (gln, Q), glutamic acid (glu, E), glycine (gly, G), histidine (his, H), isoleucine (ile, I), leucine (leu, L), lysine (lys, K), methionine (met, M), phenylalanine (phe, F), proline (pro, P), serine (ser, S), threonine (thr, T), tryptophan (trp, W), tyrosine (tyr, Y), and valine (val, V).

A nucleic acid is "operably linked" when it is in a functional association with another nucleic acid sequence. For example, a DNA of a presequence or secretory leader is operably linked to a DNA of a polypeptide if they are expressed as a preprotein that participates in the secretion of the polypeptide; a promoter or enhancer is operably linked if it affects the transcription of the coding sequence; or they may be operably linked if the ribosome binding site is aligned with the coding sequence to facilitate translation. Generally, "operably linked" means that the DNA sequences being linked are contiguous and, in the case of a secretory leader, contiguous and in reading frame. However, enhancers need not be contiguous. The connection is made by a connection at a convenient restriction site. If such sites are not present, synthetic oligonucleotide aptamers (adaptors) or linkers are used in accordance with conventional procedures.

As used herein, the terms "cell," "cell line," and "cell culture" are used interchangeably and all names include progeny thereof. Thus, the words "transfectants" and "transfected cells" include primary subject cells and cultures derived therefrom, regardless of the number of passages. It is also understood that all progeny are not precisely identical in DNA content, due to deliberate or inadvertent mutations. Variant progeny that have the same function or biological activity as the function or biological activity screened for by the originally transformed cell are included.

As used herein, the term "binding" or "specific binding" refers to the binding of an antibody to an epitope in an in vitro assay, preferably in a surface plasmon resonance (SPR, BIAcore, GE-healthcare uppsala, Sweden) assay. Binding affinity is defined by the terms ka (the association rate constant of an antibody in an antibody/antigen complex), kD (dissociation constant) and kD (kD/ka). Binding or specific binding means 10^-8mol/l or less, preferably 10^-9M to 10^-13Binding affinity (KD) in mol/l.

Binding of the antibody to the death receptor can be studied by BIAcore assay (GE-healthcare uppsala, Sweden). Binding affinity is defined by the terms ka (the association rate constant of an antibody in an antibody/antigen complex), kD (dissociation constant) and kD (kD/ka).

The term "epitope" includes any polypeptide determinant capable of specific binding to an antibody. In certain embodiments, epitope determinants include chemically active surface groups of a molecule, such as amino acids, sugar side chains, phosphoryl groups, or sulfonyl groups, which, in certain embodiments, may have specific three-dimensional structural characteristics and/or specific charge characteristics. An epitope is a region of an antigen-binding antibody.

The "Fc portion" of an antibody is not directly involved in binding of the antibody to an antigen, but exhibits a variety of effector functions. The "Fc portion of an antibody" is a term well known to those skilled in the art and is defined based on the papain cleavage of antibodies. Depending on the amino acid sequence of the constant region of the heavy chain of an antibody, antibodies or immunoglobulins are classified as: IgA, IgD, IgE, IgG and IgM types, and several of these can be further divided into subtypes (isotypes), such as IgG1, IgG2, IgG3 and IgG4, IgA1 and IgA 2. The different classes of immunoglobulins are called α, γ, and μ, respectively, based on the heavy chain constant region. The Fc portion of the antibody is directly involved in ADCC (antibody-dependent cell-mediated cytotoxicity) and CDC (complement-dependent cytotoxicity) based on complement activation, C1q binding and Fc receptor binding. Complement activation (CDC) is initiated by the binding of complement factor C1q to the Fc portion of most IgG antibody subtypes. Although the effect of the antibody on the complement system depends on the specific conditions, binding to C1q is caused by a defined binding site in the Fc portion. Such binding sites are known in the art and are described, for example, by Boakle et al, Nature282(1975)742-743, Lukas et al, J.Immunol.127(1981)2555-2560, BrunhouseandCebra, mol.Immunol.16(1979)907-917, Burton et al, Nature288(1980)338-344, Thommesen et al, mol.Immunol.37(2000)995-1004, Idusogene et al, J.Immunol.164(2000)4178-4184, Hezareh et al, J.Virology75 (1212001) 2001-61-12168, Morgan et al, Immunogloy 86(1995) 1995-324, EP 03074319. Such binding sites are, for example, L234, L235, D270, N297, E318, K320, K322, P331 and P329 (numbering according to EU index of Kabat, see below). Antibodies of the IgG1, IgG2, and IgG3 subclasses generally show complement activation and binding to C1q and C3, whereas IgG4 does not activate the complement system and does not bind to C1q and C3.

The antibodies according to the invention are produced by recombinant means. Thus, one aspect of the invention is a nucleic acid encoding an antibody according to the invention, and a further aspect is a cell comprising said nucleic acid encoding an antibody according to the invention. Methods for recombinant production are well known in the art and include expression of proteins in prokaryotic and eukaryotic cells, followed by isolation of the antibody polypeptide, and usually purification to a pharmaceutically acceptable purity. To express the antibodies as previously described in a host cell, the nucleic acids encoding the respective modified light and heavy chains are inserted into an expression vector by standard methods. Expression is carried out in suitable prokaryotic or eukaryotic host cells such as CHO cells, NS0 cells, SP2/0 cells, HEK293 (including HEK293EBNA) cells, COS cells, PER. C6 cells, yeast or E.coli cells, and the antibody is recovered from the cells (supernatant or cells after lysis). General methods for recombinant production of antibodies are well known in the art and are described, for example, in Makrides, S.C., ProteinExpr. Purif.17(1999) 183-202; geisse, S.et al, protein Expr. Purif.8(1996) 271-282; kaufman, R.J., mol.Biotechnol.16(2000) 151-161; werner, R.G., drug Res.48(1998)870-880 for review.

The antibody according to the present invention can be suitably isolated from the culture medium by conventional immunoglobulin purification methods, such as, for example, protein a-sepharose, hydroxylapatite chromatography, gel electrophoresis, dialysis or affinity chromatography. DNA and RNA encoding monoclonal antibodies are readily isolated and sequenced using conventional methods. Hybridoma cells can be used as a source of such DNA and RNA. Once the DNA is isolated, the DNA can be inserted into an expression vector and subsequently transfected into a host cell that would not otherwise produce immunoglobulin, such as a HEK293 cell, CHO cell, or myeloma cell, to effect synthesis of the recombinant monoclonal antibody in the host cell.

Amino acid sequence variants (or mutants) of the antibodies according to the invention are prepared by introducing appropriate nucleotide changes into the antibody DNA, or by nucleotide synthesis. Such modifications may be made, however, only to a very limited extent, e.g., as described above. For example, the modifications do not alter the antibody characteristics described above, such as IgG isotype and antigen binding, but may improve yield of recombinant product, protein stability, or aid in purification.

As used herein, the term "host cell" refers to any type of cellular system that can be engineered to produce an antibody according to the invention. In one embodiment, HEK293 cells and CHO cells are used as host cells. As used herein, the terms "cell," "cell line," and "cell culture" are used interchangeably and all names include progeny thereof. Thus, the words "transfectants" and "transfected cells" include primary subject cells and cultures derived therefrom, regardless of the number of passages. It is also understood that all progeny are not precisely identical in DNA content, due to deliberate or inadvertent mutations. Variant progeny that have the same function or biological activity as used to screen the originally transformed cell are included.

Expression in NS0 cells is determined by, for example, Barnes, L.M., et al, Cytotechnology32(2000) 109-123; barnes, L.M., et al, Biotech.Bioeng.73(2001) 261-. Transient expression is described, for example, by Durocher, Y. et al, Nucl. acids. Res.30(2002) E9. Cloning of the variable region was performed by Orlandi, R.et al, Proc.Natl.Acad.Sci.USA86(1989) 3833-3837; carter, P.et al, Proc.Natl.Acad.Sci.USA89(1992) 4285-; and Norderhaug, l, et al, j.immunol.methods204(1997) 77-87. A preferred transient expression system (HEK293) is described by Schlaeger, E.J., and AndChristensen, K., in Cytotechnology30(1999)71-83 and by Schlaeger, E.J., in J.Immunol.Methods194(1996) 191-199.

Regulatory element sequences suitable for prokaryotes, for example, include promoters, alternative operator sequences and ribosome binding sites. Eukaryotic cells are known to utilize promoters, enhancers and polyadenylation signals.

Purification of the antibody to remove other cellular components or other impurities, such as other cellular nucleic acids or proteins, is performed by standard techniques, including alkali/SDS treatment, cesium chloride banding (CsClbanding), column chromatography, agarose gel electrophoresis, and the like, as are well known in the art (see Ausubel, f. et al, editors, currentprotocol analytical biology, greene publishing and wiley inter science, new york (1987)). Different methods have been established and widely used for protein purification, such as affinity chromatography using microbial proteins (e.g. protein a or protein G affinity chromatography), ion exchange chromatography (e.g. cation exchange (carboxymethyl resin), anion exchange (aminoethyl resin) and mixed mode exchange), thiophilic adsorption (e.g. with β -mercaptoethanol and other SH ligands), hydrophobic interaction or aromatic adsorption chromatography (e.g. with phenyl sepharose, aza-arylophilic resin or m-aminophenylboronic acid (m-aminophenylboronic acid)), metal chelate affinity chromatography (e.g. with ni (ii) and cu (ii) affinity materials), size exclusion chromatography and electrophoretic methods (e.g. gel electrophoresis, capillary electrophoresis) (Vijayalakshmi, m.a.application.biochem.biotech.75 (1998) 93-102).

As used herein, "pharmaceutical carrier" includes any and all solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic and absorption delaying agents, and the like, which are physiologically compatible. Preferably, the carrier is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal or epidermal administration (e.g., by injection or infusion).

The compositions of the present invention may be administered by a variety of methods well known in the art. As the skilled person will appreciate, the route and/or manner of administration will vary with the desired result. For administration of the compounds of the present invention by certain routes of administration, it may be desirable to coat them with materials that prevent inactivation of the compounds, or to co-administer the compounds therewith. For example, the compound may be administered to the subject in a suitable carrier, such as a liposome or diluent. Pharmaceutically acceptable diluents include saline solutions and aqueous buffered solutions. Pharmaceutically acceptable carriers include sterile aqueous solutions or dispersions and sterile powders for the extemporaneous preparation of sterile injectable solutions or dispersions. The use of such media and agents for pharmaceutically active substances is well known in the art.

The phrases "parenteral administration" and "administered parenterally" as used herein, refer to modes of administration other than enteral and topical administration, typically by injection, including, but not limited to: intravenous, intramuscular, intraarterial, intrathecal, intracapsular, intraorbital, intracardiac, intradermal, intraperitoneal, transtracheal, subcutaneous, subcuticular, intraarticular, subcapsular, subarachnoid, intraspinal, epidural, and intrasternal injection and infusion.

As used herein, the term cancer refers to a proliferative disease, such as lymphoma, lymphocytic leukemia, lung cancer, non-small cell lung (NSCL) cancer, bronchiolar cell lung cancer, bone cancer, pancreatic cancer, skin cancer, cancer of the head or neck, cutaneous or intraocular melanoma, uterine cancer, ovarian cancer, rectal cancer, cancer of the anal region, gastric cancer (stomachcancer), gastric cancer (gastromiccancer), colon cancer, breast cancer, uterine cancer, cancer of the fallopian tubes, cancer of the endometrium, cancer of the cervix, cancer of the vagina, cancer of the vulva, Hodgkin's disease, cancer of the esophagus, cancer of the small intestine, cancer of the endocrine system, cancer of the thyroid gland, cancer of the parathyroid gland, cancer of the adrenal gland, sarcoma of soft tissue, cancer of the urethra, cancer of the penis, prostate cancer, cancer of the bladder, cancer of the kidney or ureter, renal cell cancer, cancer of the renal pelvis, mesothelioma, hepatocellular carcinoma, biliary tract, neoplasms of the Central Nervous System (CNS), Spinal axis tumors (spinoaxostomators), brain stem gliomas, glioblastoma multiforme, astrocytomas, stanwoma, ependymomas, medulloblastomas, meningiomas, squamous cell carcinomas, pituitary adenomas, and ewing's sarcoma, including refractory types of any of the foregoing cancers or combinations of one or more of the foregoing cancers.

These compositions may also contain adjuvants such as preserving, wetting, emulsifying, and dispersing agents. Prevention of the presence of microorganisms can be ensured by the aforementioned sterilization procedures, as well as by the inclusion of various antibacterial and antifungal agents, such as parabens, chlorobutanol, phenol, sorbic acid, and the like. It may also be desirable to include isotonic agents, for example, sugars, sodium chloride, and the like in the compositions. Additionally, delayed absorption of injectable pharmaceutical forms can be brought about by the incorporation of agents which delay absorption such as aluminum monostearate and gelatin.

Regardless of the route of administration chosen, the compounds of the present invention and/or the pharmaceutical compositions of the present invention, which may be used in the form of suitable hydrates, are formulated into pharmaceutically acceptable dosage forms by conventional methods well known in the art.

Actual dosage levels of the active ingredient in the pharmaceutical compositions according to the invention can be varied to obtain effective amounts of the active ingredient to achieve the desired therapeutic response for a particular patient, composition and mode of administration without toxicity to the patient. The selected dosage level will depend upon a variety of pharmacokinetic factors including the activity of the particular composition of the invention employed, the route of administration, the time of administration, the rate of excretion of the particular compound employed, other drugs, compounds and/or materials used in combination with the particular composition employed, the age, sex, body weight, condition, general health and past medical history of the patient being treated, and like factors well known in the medical arts.

The composition must be sterile and flowable so that the composition can be delivered by syringe. In addition to water, the carrier may also be an isotonic buffered saline solution.

Proper fluidity can be maintained, for example, by the use of a coating such as lecithin, by the maintenance of the required particle size in the case of dispersants, and by the use of surfactants. In many cases, it will be preferable to include isotonic agents, for example, sugars, polyalcohols such as mannitol or sorbitol, and sodium chloride in the composition.

As used herein, the term "transformation" refers to a process of transferring a vector/nucleic acid into a host cell. If cells without a robust cell wall barrier are used as host cells, transfection is carried out, for example, by calcium phosphate precipitation methods as described by Graham, F.L., and van der Eb, A.J., Virology52(1973) 456-467. However, other methods of introducing DNA into cells may also be used, such as by nuclear injection or by protoplast fusion. If prokaryotic cells or cells containing a robust cell wall structure are used, for example, one method of transfection is calcium treatment with calcium chloride as described by Cohen, S.N. et al, PNAS.69(1972) 2110-2114.

As used herein, "expression" refers to the process of transcription of a nucleic acid into mRNA and/or the subsequent translation of the transcribed mRNA (also referred to as transcript) into a peptide, polypeptide or protein. Both the transcript and the encoded polypeptide are referred to as gene products. If the polynucleotide is derived from genomic DNA, expression in eukaryotic cells includes splicing of the mRNA.

A "vector" is a nucleic acid molecule that is inserted into and/or between host cells, particularly a self-replicating nucleic acid molecule. The term includes vectors that are used primarily for the insertion of DNA or RNA into a cell (e.g., chromosomal integration), replication of vectors used primarily for the replication of DNA or RNA, and expression vectors used for transcription and/or translation of DNA or RNA. Also included are vectors that provide more than one of the functions described above.

An "expression vector" is a polynucleotide that is capable of being transcribed and translated into a polypeptide when introduced into an appropriate host cell. An "expression system" generally refers to a suitable host cell containing an expression vector that can be used to produce a desired expression product.

The following examples, sequence listing and figures are provided to aid the understanding of the present invention, the true scope of which is set forth in the appended claims. It is to be understood that changes may be made to the operations shown without departing from the spirit of the invention.

Brief Description of Drawings

FIG. 1: FACS binding analysis of CEA, DR5 and FAS expression levels in different human cell lines (Lovo, OVCAR-3, AsPC-1, BxPC3, LS174T and MKN-45) were detected using unlabeled, commercially available murine IgG1 antibody (CEA: Abcam # 11330; DR 5: R & D # MAB 631; FAS: BD #555671) and the commonly used goat anti-mouse FITC labeled IgG (SerotecStar 105F). Samples containing cells only or cells and secondary antibody only were used as controls. All cell lines tested expressed large amounts of DR5 and FAS on the surface, except Lovo cells. In contrast, CEA expression is very low. When the same cells were detected with other antibodies against 3 antigens, the Lovo cells were also DR5 in FACS analysis, FAS and CEA expression positive (data not shown).

FIG. 2: apoptosis induction (DNA fragmentation assay) of different cell lines was analyzed after 4 hours incubation with a commercially available non-crosslinked antibody capable of inducing apoptosis in solution (DR 5: R & D # MAB 631; FAS: Millipore/Upstate: CH 11). To detect apoptosis, an elispaplus kit was used for cell death detection for analysis of histone-associated DNA fragmentation. In BxPC-3, Lovo and LS174T cells, apoptosis was clearly induced by DR5 and FAS, whereas ASPC-1 cells did not undergo apoptosis at all. MKN-45 cells were more resistant to DR5 than other cell lines.

FIG. 3: apoptosis induction of LS174T cells after 4 hours incubation with ApomAb (white bars), ApomAb cross-linked to anti-human Fc antibody (hatched grey bars), ApomAb _ sm3e _ a (black bars) and ApomAb _ sm3e _ a1 (dotted grey bars) (DNA fragmentation assay). The induction of highly cross-linked CEA binding dependent apoptosis targeted by bispecific antibodies can be detected. This effect was in the same range as apoptosis induced by cross-linking of ApomAb and could be completely abolished by pre-incubation with excess sm3 eIgG. No apoptosis was observed in the control (cells only or sm3eIgG) and ApomAb alone did not induce apoptosis at the concentration used (1 μ g/ml).

FIG. 4: comparison of the apoptosis-inducing activity of different ApomAb _ sm3e bispecific molecules in DNA fragmentation assays using LS174T cells incubated for 4 hours with apoptosis-inducing agents compared to ApomAb alone (white bars) or ApomAb cross-linked to anti-human Fc antibody (hatched gray bars). In general, the molecule in which sm3escFv was fused to the C-terminus of the ApomAb heavy chain (form a, black bars) appeared to be more active than the construct in which sm3escFv was fused to the C-terminus of the ApomAb light chain (form B, grey bars). Furthermore, bispecific antibodies comprising disulfide stabilized scfvs (format a1, dotted grey bars and B1, small grid bars) appeared to be slightly less active than molecules with wild-type scFv.

FIG. 5: apoptosis induction assay (DNA fragmentation assay) of LS174T cells after 4 hours incubation with ApomAb (white bars), ApomAb cross-linked to anti-human Fc antibody (hatched grey bars), or ApomAb _ PR1a3_ a bispecific construct (black bars). In each case, apoptosis induction was dependent on the antibody concentration used. ApomAb alone also induced low levels of apoptosis at high concentrations, but the level of apoptosis was significantly increased by cross-linking. The bispecific ApomAb _ PR1a3_ a molecule is even more active than the cross-linked ApomAb without the second cross-linker.

FIG. 6: apoptosis induction analysis (DNA fragmentation assay) of Lovo cells after 4 hours incubation with ApomAb (white bars), ApomAb cross-linked with anti-human Fc antibody (grey bars) or ApomAb _ PR1a3_ a bispecific construct (black bars). In each case, apoptosis induction was dependent on the antibody concentration used. ApomAb alone also induced low levels of apoptosis at high concentrations (as described above), but levels of apoptosis were significantly increased by cross-linking. The bispecific ApomAb _ PR1a3_ a molecule is itself as active as the cross-linked ApomAb.

FIG. 7: comparison of DNA fragmentation in LS174T cells after 4 hours of incubation with different apoptosis-inducing bispecific antibodies. The molecule used is an ApomAb _ PR1A3 bispecific molecule in which PR1A3scFv (wt ═ a/B or disulfide stabilized ═ a1/B1) is fused to the C-terminus of the heavy chain (a, hatched gray bars) or the light chain (B, dotted bars). Although the fusion position of the scFv does not appear to differ in this case with respect to apoptosis induction, the type of fusion scFv used is important: the induction of apoptosis was almost completely eliminated using disulfide stabilized scFv (grey and black bars, respectively) compared to wtsc fv comprising fusion with ApomAb. Due to the lower affinity of PR1A3 compared to sm3e, the overall induction of apoptosis with bispecific molecules comprising PR1A3 was also lower.

FIG. 8: FACS binding analysis of ApomAb-CEA (PR1A3) bispecific constructs on MKN-45 cells. ApomAb _ PR1A3 bispecific constructs with wild type (a) or disulfide stabilized scFv (a1) were compared. Both bispecific constructs bound to the target cells in a concentration-dependent manner, but the molecule comprising PR1A3 in the form of the wild-type scFv bound antigen with much higher affinity than the molecule comprising the disulfide-stabilized PR1A3 scFv.

FIG. 9: CRIPTO, FAS and DR5 surface expression analysis was performed by FACS binding experiments on NCCIT and recombinant CRIPTO expressing HEK293 cells. NCCIT cells do not express FAS, express only a small amount of CRIPTO but a similar amount of DR5 compared to recombinant HEK293-CRIPTO cells. The latter cells showed low levels of FAS, significant levels of DR5 and fairly high levels of CRIPTO expression.

FIG. 10: comparison of apoptosis induction (DNA fragmentation in HEK293-CRIPTO cells) using FAS (HFE7AIgG), FAS (HFE7AIgG) crosslinked by anti-human Fc antibodies and FAS-CRIPTO bispecific molecules (HFE7A _ LC020H3L2D1, with wt (a) or disulfide stabilized (a1) criptospcfv fused to the C-terminus of HFE7A heavy chain). Fagg alone, crptiogg alone and FAS-MCSP bispecific molecules did not induce apoptosis, while cross-linked FAS and HFE7A-CRIPTO bispecific molecules showed DNA fragmentation after 4 hours of incubation, which was partially eliminated by pre-incubation with excess anti-crptiogg.

FIG. 11: induction of apoptosis by HFE7A-CRIPTO bispecific molecules in recombinant HEK293-CRIPTO cells (black bars) compared to recombinant HEK293-FAP (fibroblast activation protein) cells (white bars) (DNA fragmentation assay). In both cell lines, apoptosis could be induced using a commercially available antibody that induces apoptosis (CH11) and using HFE7AIgG crosslinked by a second Fc-specific antibody, while HFE7A alone did not induce apoptosis under the conditions used. Apoptosis induction using the bispecific FAS-CRIPTO molecule was higher than with cross-linked HFE7AIgG, but could not be completely inhibited by pre-incubation with excess anti-CRIPTOIgG. Some low background apoptosis was observed in HEK293-FAP cells, which also could not be completely competed by excess crptotigg (out-completed). Even the negative control molecule, in which the disulfide-stabilized MCSP-specific scFv was fused to the C-terminus of the HFE7A heavy chain, showed a low degree of apoptosis in HEK293-FAP cells.

FIG. 12: FACS binding analysis of MCSP surface expression levels was determined on different cell lines (MCF7, SkBr3, A431, A549, HCT-116 and U87-MG) using 2 different antibodies. The same level of MCSP expression was detectable using all 2 antibodies, indicating that U87-MG showed the highest MCSP expression, HCT-116 had low MCSP expression, while all other cell lines detected were MCSP negative (in the context of negative controls such as unstained cells).

FIG. 13: the apoptotic capacity of U87-mg (a) and HCT-116(B) cells was assessed using soluble and cross-linked ApomAb (black bars) and HFE7A (grey bars) and related control molecules (anti FAS _ CH11, anti DR5_ R2 and anti Fc-IgG alone). Although apoptosis could only be induced by DR5 receptor after 4 hours in HCT-116 cells and not by FAS, it was different in U87-MG cells. Here, significant apoptosis was observed after 24 hours. In contrast to HCT-116 cells, the U87-MG apoptosis induction efficiency by cross-linked HFE7A was 2-fold higher than with cross-linked ApomAb. Control antibodies that have been rendered apoptotic in solution are even more effective.

FIG. 14: apoptosis induction assay in U87-MG glioma cells after 24 hours incubation with bispecific HFE7A-MCSP antibody (mab9.2.27) in which wild type (form a) or disulfide stabilized MCSPscFv (form a1) is fused to the C-terminus of ApomAb heavy chain. In this case, constructs comprising disulfide-stabilized scfvs demonstrated significantly higher apoptosis than molecules comprising wild-type scfvs (although the amount of apoptosis measured by DNA fragmentation was relatively low). However, in both cases, induction of apoptosis could be completely abolished by preincubation of cells with excess competing mcspiggs.

FIG. 15: FACS binding analysis of human Fibroblast Activation Protein (FAP) expression levels of 2 different cell lines (SW872 and GM05389) (a). The fluorescence intensity measured using different concentrations of anti-FAP antibody is shown over a range of 3 orders of magnitude (black, grey and shaded lines). Negative control reactions for only secondary antibody and cells are shown as dotted and white bars, respectively. Although GM05389 cells exhibited FAP expression above background at all antibody concentrations tested, FAP expression was detectable in SW872 cells only at the highest antibody concentration used (10 μ g/ml), indicating that these cells were not suitable for FAP-based binding/apoptosis induction experiments. In addition, this cell line was shown to undergo little ApomAb mediated apoptosis (B). ApomAb alone or another commercially available anti-DR 5 antibody did not induce relevant DNA fragmentation. Only when ApomAb was cross-linked to anti-human Fc antibody, a detectable low level of apoptosis induction was observed.

FIG. 16: apoptosis induction assay of GM05389 (white bar) and MDA-MB-231 (grey bar) alone, compared to co-culturing both cell lines (black bar). In all cell lines, ApomAb alone had only a weak effect, whereas cross-linking of ApomAb resulted in significant induction of apoptosis in MDA-MB-231 cells. DNA fragmentation induced with the death receptor agonist bispecific construct (ApomAb-FAP) only occurred at high levels when both cell lines were co-cultured. Here, ApomAb cross-linking alone did not increase apoptosis in the same range, suggesting that two cell lines are necessary for optimal induction of apoptosis: one expressing the death receptor and a second expressing the FAP antigen.

FIG. 17: results of an apoptosis induction assay (24 hours) on MKN-45 cells using a tetravalent bispecific ApomAb _ PR1a3_ scFab molecule in which scFab is fused to the C-terminus of the ApomAb (a form) heavy chain. Apoptosis induction was compared to ApomAb (+/-cross-linked with 10-fold excess anti-human Fc antibody) and negative controls. All constructs were used at concentrations of 0.1 and 1.0. mu.g/ml. Under the assay conditions used, the bispecific ApomAb _ PR1a3_ scFab construct (black bars) clearly showed a concentration-dependent induction of apoptosis, which was in the same range as observed for the highly cross-linked ApomAb (grey bars) and which was significantly higher than the apoptosis induced using the ApomAb alone (shaded bars).

FIG. 18: apoptosis induction of LS174T cells by ApomAb (alone, shaded bars or highly cross-linked, grey bars) was analyzed compared to bispecific trivalent construct (ApomAb _ sm3e _ scFab; 2x1 valency, black bars) and negative control. The construct was used at concentrations of 0.1 and 1.0. mu.g/ml for 4 hours. The bispecific ApomAb _ sm3e _ scFab construct was able to induce apoptosis in a concentration-dependent manner, in the same range as that induced by highly cross-linked ApomAb.

FIG. 19: ApomAb and bispecific DR5 agonist antibodies ApomAb _ sm3e _ a1 were analyzed for in vivo efficacy in an intra-splenic transfer model using human colon cancer cell line LS174T, compared to vehicle controls. A random group of 10 mice per group was treated with PBS (black line), ApomAb (black circle) or ApomAb-sm3e _ a1 bispecific antibody (black square). Percent survival was plotted against time course of the experiment.

Examples

Example 1: design of bispecific antibody recognizing human death receptor 5 and human CEA

Described below are tetravalent bispecific antibodies comprising a combination of a full-length antibody binding to a first antigen (human death receptor, DR5) and 2 single-chain Fv fragments bound to a second antigen (human carcinoembryonic antigen, CEA), said single-chain Fv fragments being fused to the C-terminus of the heavy or light chain of said full-length antibody by a peptide linker. The antibody domains and linkers in the single chain Fv have the following orientation: VH-linker-VL.

The sequences of ApomAb antibodies described by Adams in US2007/0031414a1 were used as the light and heavy chain variable regions of DR5 recognition antibodies.

scFv for binding CEA antigen uses the sequences of the light and heavy chain variable regions of PR1A3(Bodmer et al 1999; US5965710) and sm3e (Begent et al 2003; US7232888B 2).

The VH and VL of the corresponding CEA antibody are linked by a glycine-serine (G4S)4 linker through gene synthesis and recombinant DNA techniques to produce a single chain Fv fused to the C-terminus of the heavy or light chain of ApomAbIgG1 through a (G4S) n linker (where n is 2 or 4).

In addition to the "wild-type" scFv, variants were prepared comprising cysteine residues at Kabat position 44 of the heavy chain variable region and Kabat position 100 of the light chain variable region to create an interchain disulfide bond between VH and VL. This has the purpose of stabilizing the scFv molecule to minimize potential aggregation propensity.

To avoid non-specific cross-linking of the bispecific molecule, e.g. via Fc receptor acting as human Fc γ RIIIa, 2 amino acids in the Fc region of the IgG part of the bispecific molecule were changed. The 2 th leucine residues at positions 234 and 235 in the Fc region were replaced with alanine residues by site-directed mutagenesis. This so-called LALA mutation is described to eliminate Fc-FcR interaction (Hessell et al, Nature449(2007), 101 ff).

All of these molecules were recombinantly expressed, prepared and purified using standard antibody purification techniques, including size exclusion chromatography followed by protein a affinity chromatography. The molecules were characterized in terms of expression yield, stability and biological activity.

A summary of the different bispecific death receptor agonist antibody molecules consisting of ApomAb-CEA combinations is provided in table 1. The design description of the different molecules can be deduced from the molecular names where the first part characterizes IgG targeting death receptors (e.g. ApomAb), the second name describes the source of scFv targeting CEA (e.g. PR1a3 or sm3e), and the letter and number combination describes the fusion site and disulfide-stabilizing properties of the scFv.

Table 1: description of different bispecific death receptor-activating antibodies targeting human DR5 and human CEA and their related features.

Example 2: expression and purification of bispecific death receptor agonist antibodies

Separate expression vectors for the light and heavy chains of each bispecific antibody were constructed. These vectors contain a prokaryotic selectable marker, regulatory elements for gene expression in mammalian cells, and an origin of replication, oriP, from Ebstein-Barr virus for autonomous replication of the plasmid in EBNA-containing HEK293 cells. The plasmid was propagated in e.coli, purified and co-transfected into HEK293EBNA cells using calcium phosphate mediated precipitation for transient transfection. Cell culture supernatants were harvested 7 days later and the antibodies were purified by protein a and size exclusion chromatography. Purified molecules were analyzed for homogeneity, stability and integrity by analytical size exclusion chromatography (before and after one freeze-thaw step) and SDS-PAGE analysis (under non-reducing and reducing conditions).

Table 2: summary of purification yields and monomer content of different death receptor agonist bispecific antibodies

All molecules in sufficient quantities and with appropriate masses can be prepared and purified for further characterization and detection. The yield after purification was in the range of about 5mg/L, with some molecules having some deviations. For example, the yield of ApomAb-sm3e-B1 was significantly lower (2.19mg/L), while the corresponding construct ApomAb-PR1A3_ B1 could even be purified to over 11 mg/L.

The measurement of aggregate formation after freeze/thaw and the increasing concentration of antibody indicate that depending on the molecule, stabilization by interchain disulfide bonds may have a favorable effect on the propensity to form aggregates. In general, disulfide stabilization results in higher monomer content of the molecule, at least at higher concentrations (table 3).

Table 3: aggregate formation of bispecific death receptor-activated antibodies correlated with protein concentration

The tendency to form aggregates depends not only on the disulfide bond stabilization of the scFv, but also on the antigen-binding scFv used. As is evident from table 3, bispecific ApomAb molecules comprising PR1A3scFv undergo significant aggregation upon increasing protein concentration. Only 80% of the material at concentrations above 3mg/ml appeared monomeric, whereas these molecules did not form aggregates at the concentrations used after the introduction of 2 additional cysteine residues (VH 44/VL100 according to Kabat numbering).

The extent of aggregate formation of bispecific ApomAb molecules comprising sm3escFv was not as significant, since here the monomer content was: it is still about 94% when there is no disulfide stabilization and 97% when there is disulfide stabilization.

Example 3: induction of apoptosis by death receptor bispecific DR5-CEA antibody molecules

Human DR5 death receptor-activating antibody ApomAb induces apoptosis in tumor cells expressing DR5, such as colon cancer cell line LS180 or Colo-205. In vitro, ApomAb itself mediates significant apoptosis that can be significantly enhanced by cross-linking of DR5 bound to ApomAb with antibodies that bind to the human Fc region of ApomAb. This induction of apoptosis can also be transferred in vivo, where different tumor models can show that ApomAb exhibits significant efficacy (Jin et al, 2008; Adams et al, 2008), most likely through a crosslinking event via the human Fc receptor. To assess the potential of DR5-CEA bispecific antibodies in targeting DR5 cross-linking at tumor sites and subsequent induction of apoptosis, the activity of ApomAb-CEA bispecific molecules in apoptosis mediation was analyzed in vitro.

To determine whether the DR5-CEA bispecific antibody molecule is capable of inducing tumor antigen binding-dependent apoptosis of target cells, DNA fragmentation in tumor cells following incubation with death receptor agonist bispecific antibodies was analyzed as a measure of apoptosis using a cell death detection ELISA assay.

To find cell lines suitable for measuring the antigen binding dependent DR5 cross-linking leading to apoptosis induction, several different tumor cell lines were analyzed for surface expression of DR5, FAS and CEA.

All target cell lines used were analyzed for relative expression levels of tumor associated antigen and FAS or DR5 death receptor prior to performing the following apoptosis assay.

Cell number and viability were determined. For this purpose, adherent growing cells were detached using cell isolation buffer (Gibco-Invitrogen # 13151-014). Cells were harvested by centrifugation (4 min, 400Xg), washed with FACS buffer (PBS/0.1% BSA) and the cell number adjusted to 1.111X10 in FACS buffer⁶Cells/ml. Using 180. mu.l of this cell suspension per well of a 96-well round bottom plate, 2X10 per well was obtained⁵A cell. Cells were incubated with primary antibody at 4 ℃ for 30 minutes at appropriate dilutions. Then go toCells were harvested by centrifugation (4 min, 400Xg), the supernatant was removed thoroughly and the cells were washed once with 150. mu.l FACS buffer. The cells were resuspended in 150. mu.l FACS buffer and incubated with the secondary antibody (in the case of the use of unlabelled primary antibody) in the dark for 30 minutes at 4 ℃. After 2 steps of washing with FACS buffer, cells were resuspended in 200 μ Ι FACS buffer and analyzed in htsfacscanntoii (BD, software FACSDiva). Alternatively, cells can be fixed with 200 μ l of 2% PFA (paraformaldehyde) in FACS buffer for 20 minutes at 4 ℃ and analyzed later. All assays were performed in triplicate.

The results of FACS binding analysis of different tumor cell lines using 3 antibodies specifically recognizing CEA, DR5 or FAS are shown in figure 1. All other cell lines tested, except the Lovo cells, expressed the antigen tested at different levels. CEA expression was highest in MKN-45 cells and was more or less similar in OVCAR-3, AsPC-1, BxPC-3 and LS 174T. In terms of DR5 expression, AsPC-1 and bxpc.3 cells expressed the most receptors compared to the other cell lines, followed by OVCAR-3 and MKN-45, while LS174T had the lowest DR5 expression level. With respect to FAS expression, cell lines were differentially expressed but all showed significant FAS expression. When Lovo cells (which were negative in this assay) were later analyzed with different antibodies against CEA, DR5 and FAS, they also showed significant expression of the detection antigen (data not shown).

To determine the induced apoptosis, the elispaplus kit was used for cell death detection from Roche. Briefly, each well of a 96-well plate was inoculated with 10 in 200. mu.l of the appropriate medium⁴Cells (cell number and survival after isolation) and in 5% CO₂The culture was carried out overnight at 37 ℃ in the atmosphere. The next day the medium was replaced with fresh medium containing appropriate concentrations of apoptosis-inducing antibodies, control antibodies and other controls:

bispecific antibody is used at a final concentration of 0.01-10 μ g/ml; control antibody was used at 0.5. mu.g/ml and cross-linking antibody was used at 100. mu.g/ml. A 100-fold excess of competing antibody was used.

At 37 ℃ 5% CO₂Cells were cultured for 4-24 hours to allow induction of apoptosis. Cells were harvested by centrifugation (10 min, 200Xg) and incubated in 200. mu.l lysis buffer (provided in the kit) for 1 hour at room temperature. Intact cells were pelleted by centrifugation (10 min, 200Xg) and 20. mu.l of supernatant was analyzed according to the manufacturer's recommendations to detect induction of apoptosis.

One set of cell lines was also analyzed for the ability to undergo apoptosis by incubation with a commercially available antibody directed against DR5 or FAS, which is known to crosslink death receptors already in solution (fig. 2).

Here, significant differences in apoptosis induction between the cell lines as shown in figure 2 were observed. Although the induction of apoptosis by DR5 and FAS was similar in MKN-45 and BxPC-3 (although the value of DNA fragmentation in MKN-45 only reached 50% in BxPC-3), in LS174T and Lovo cells, apoptosis induced by DR5 cross-linking antibodies was much better than that induced by antibodies binding to FAS. In LS174T cells, induction of apoptosis by DR5 cross-linking was approximately 2-fold more efficient than induction by FAS cross-linking. In Lovo cells, this difference in apoptosis induction was even 4-fold. ASPC-1 cells are very resistant to apoptosis induction by death receptor cross-linking. Based on these results, cell lines Lovo and LS174T were selected for analysis of apoptosis induced by tumor antigen-targeted DR5 cross-linking.

The results of apoptosis induction in LS174T cells after treatment with bispecific DR5-CEA molecules (ApomAb-sm3e) compared to the effect of ApomAb or cross-linked ApomAb are illustrated in fig. 3. Under the assay conditions used (incubation for 4 hours at a concentration of 1 μ g/ml), ApomAb alone or sm3e in the form of IgG1 showed no detectable DNA fragmentation (normalized to the value of "cell only"), whereas the bispecific ApomAb-sm3e molecule (wild-type (form a) or disulfide-stabilized (form a1) scFv) showed significant apoptosis induction comparable to the theoretical maximum value of highly cross-linked ApomAb. The 2 bispecific molecules showed very similar activity, demonstrating that the stabilization of the molecules by insertion of interchain disulfides did not affect the biological activity. When cells were preincubated with excess sm3eIgG (100-fold higher concentration compared to bispecific constructs), no longer induced apoptosis, indicating that sm3eIgG blocked all CEA antigen on the cell surface and prevented additional binding of bispecific death receptor agonist molecules. This demonstrates that the specificity of apoptosis induced is dependent on the cross-linking of DR5 death receptors by tumor antigens.

The results of comparing the induction of apoptosis in LS174T cells by different molecular forms of the bispecific ApomAb-sm3e construct are summarized in fig. 4. Induction of apoptosis was carried out at a concentration of 1. mu.g/ml for 4 hours. Again, the bispecific ApomAb-sm3e molecule, in which sm3escFv was fused to the C-terminus of the ApomAb heavy chain (a and a1 forms), demonstrated significant induction of apoptosis, in this case even more potent than the highly cross-linked ApomAb. ApomAb alone does not induce detectable DNA fragmentation under the conditions of use. 2 additional bispecific constructs (sm3escFv fused to the C-terminus of ApomAb light chain, wild-type or disulfide-stabilized form B1) also demonstrated high levels of apoptosis induction, with at least form B displaying a similar range to that of cross-linked ApomAb, indicating that 2 forms are essentially functional. Fusion of scFv to the C-terminus of the heavy chain of ApomAb appears to be slightly more advantageous than fusion to the light chain. Compared to the results shown in fig. 4, the disulfide-stabilized molecules also appeared to exhibit slightly reduced activity compared to the molecules with the wild-type scFv.

As shown in fig. 3 and 4, the ApomAb-sm3e construct described above plays a very good role in antigen-dependent induction of specific apoptosis. This CEA antibody sm3e showed very high affinity for its antigen (low picomolar range). To assess whether the apoptosis-inducing effect of the bispecific DR5-CEA construct could also be mediated by molecules with lower binding affinity for tumor antigens, additional constructs similar to the former were generated. The sequences of CEA antibody PR1a3, which have a rather low affinity in the micromolar range for CEA, were used to engineer scFv targeting CEA. To evaluate this antibody, bispecific constructs were generated in which PR1A3scFv (wild-type or disulfide-stabilized) was fused to the C-terminus of the ApomAbIgG heavy or light chain. The resulting molecule is named similarly to what has been described: ApomAb _ PR1A3_ A/A1/B/B1, where A and A1 describe a fusion to the C-terminus of the heavy chain and B1 show a fusion to the C-terminus of the light chain. A and B comprise wild-type scFv, whereas a1 and B1 indicate disulfide-stabilized scFv.

In fig. 5, the induction of apoptosis of LS174T cells by ApomAb, cross-linked ApomAb and ApomAb _ PR1A3 bispecific antibody (wild-type PR1A3scFv fused to the C-terminus of the ApomAb heavy chain) at concentrations ranging from 0.01 to 10.0 μ g/ml is shown. ApomAb itself exhibits a degree of concentration-dependent induction of apoptosis, which can be significantly enhanced by cross-linking of ApomAb with anti-human Fc antibodies. The bispecific ApomAb-PR1a3 molecule also demonstrated a concentration-dependent induction of apoptosis, even higher at a concentration of 10.0 μ g/ml than the same concentration of cross-linked ApomAb, indicating that in order to obtain good in vitro efficacy in apoptosis induction, it is not absolutely necessary to use the highest affinity tumor antigen binding agent in this bispecific death receptor-agonistic antibody format.

To investigate whether the induction of apoptosis observed after incubation with DR5-CEA bispecific molecule could be applied to other cell lines, a similar experiment as shown in figure 6 was performed using another colon cancer cell line, Lovo cells.

The results of inducing apoptosis in Lovo cells using the death receptor agonist bispecific molecule ApomAb _ R1a3_ a (DR5-CEA) compared to induction of apoptosis by ApomAb and cross-linked ApomAb are shown in fig. 6. Concentration-dependent induction of apoptosis was observed for all constructs. Here, ApomAb alone achieved 20% activity of the cross-linked ApomAb when used at a concentration of 10. mu.g/ml. Below this concentration, apoptosis was induced much lower than cross-linked ApomAb. The ApomAb _ PR1a3 bispecific antibody showed the same degree of induction of DNA fragmentation as the highly cross-linked ApomAb antibody in the absence of any cross-linking molecule, demonstrating that the apoptosis-inducing effect using death receptor agonist antibodies is a universal phenomenon applicable to all apoptosis-capable cell lines.

In FIG. 7, the results of a comparison between the different ApomAb-PR1A3 and ApomAb-sm3e constructs are shown. Here the induction of apoptosis in LS174T cells after 4 hours incubation with a concentration of 1 μ g/ml is summarized. It is evident from the results that the affinity for CEA antigen may indeed play a role in mediating apoptosis through death receptor cross-linking. Constructs comprising high affinity CEA binding agents have a significant difference in apoptosis induction compared to low affinity binding agents. Compared to ApomAb-sm3e, ApomAb-PR1A3 showed only about 1/3 induction of apoptosis in LS174T cells. In addition there appears to be inherent differences in different molecules, which are also reflected in apoptosis-inducing ability. In the case of fusion of PR1A3scFv to ApomAb, there was no difference in activity between scFv fused to the C-terminus of the heavy or light chain. All 2 molecules showed the same induction of apoptosis. In contrast, the constructs comprising sm3escFv performed differently. Here, fusion of scFv to the C-terminus of the heavy chain is more efficient than fusion to the C-terminus of the light chain.

Another difference between the 2 series of constructs is the fact that there is a different effect of disulfide stabilization of the scFv. While the disulfide-stabilized sm3 escFv-containing construct was not affected in apoptosis induction, the opposite was true for PR1A3 scFv. No significant induction of apoptosis is shown anymore if a disulfide-stabilized form is used.

Example 4: generation of bispecific death receptor agonist antibodies targeting FAS (CD95) and CRIPTO as tumor antigens and in vitro evaluation of these molecules:

CRIPTO is a GPI-anchored growth factor that is reported to be overexpressed in cancer cells, but underexpressed or absent in normal cells. CRIPTO is found to be upregulated in colon tumors and liver metastases. As a member of the EGF family, it is considered to be an autocrine growth factor that plays a role in the proliferation, metastasis and/or survival of tumor cells. This growth factor activates several signal transduction pathways through several potential receptors or co-receptors.

To figure out whether CRIPTO is a suitable target for the death receptor agonist bispecific antibody approach, tetravalent bispecific antibodies were generated that target FAS as the death receptor and CRIPTO as the tumor antigen. These molecules consist of a full-length IgG1 antibody (recognizing FAS) with a CRIPTO-targeted scFv fused to the C-terminus of the heavy chain of the antibody.

The sequences of the HFE7A antibody (Haruyama et al, 2002) were used for the heavy and light chains of the IgG part of the molecule targeting FAS, and the HFE7A antibody was a human/mouse cross-reactive antibody against CD 95. Criptostfv was generated from the sequence of a humanized anti-CRIPTO antibody produced by immunization (LC020_ H3L2D 1). The scFv was generated using standard recombinant DNA techniques and fused to the C-terminus of the fagg 1 heavy chain via a short peptide linker. The order of the individual domains in the scFv is VH- (G4S)4 linker VL.

Unfortunately, there are not many suitable existing cell lines available for CRIPTO targeting. The potential of 2 cell lines to serve as target cell lines for induction of apoptosis mediated by FAS cross-linking of bispecific FAS/CRIPTO antibodies was therefore evaluated. Results of the evaluation of the surface expression of FAS, DR5 and CRIPTO in NCCIT and recombinant human CRIPTO expressing HEK cells (hereinafter referred to as HEK-CRIPTO) are shown in fig. 9. In contrast to HEK-CRIPTO cells, NCCIT expressed little FAS on the surface, expressed only very low levels of CRIPTO, whereas DR5 expression appeared to be normal. In contrast, HEK-CRIPTO cells expressed high levels of CRIPTO, significant levels of DR5 and appropriate levels of FAS, and these cells were therefore selected for in vitro analysis of FAS-CRIPTO bispecific antibody-induced apoptosis.

Figure 10 summarizes the results of in vitro experiments using the induction of apoptosis on HEK-CRIPTO cells using HFE7A, cross-linked HFE7A, or HFE7A-CRIPTO bispecific constructs. No significant induction of apoptosis was achieved using HFE7A or CRIPTO (LC020) alone. Crosslinking of HFE7A with anti-human Fc antibodies resulted in high levels of DNA fragmentation as caused by the bispecific HFE7A-CRIPTO molecule. In this case, the bispecific molecule comprises a wild-type criptostfv (HFE7A _ LC020_ a) or a disulfide-stabilized scFv (HFE7A _ LC020_ a1) fused to the C-terminus of the HFE7A heavy chain. Little difference in apoptosis induction between these two molecules was observed.

In both cases, pre-incubation with excess crptotigg significantly reduced apoptosis induction, but this reduction was incomplete. The reason for this is unclear and remains to be evaluated. A similar construct (HFE7A _ LC007_ a1) comprising MCSP-targeting scFv fused to the C-terminus of the HFE7A heavy chain did not induce any apoptosis of HEK-CRIPTO cells, indicating that the observed apoptosis using the bispecific HFE7A-CRIPTO molecule is tumor antigen specific.

The results of comparison of apoptosis induction between HEK-CRIPTO and recombinant human FAP (fibroblast activation protein) -expressing HEK cells (HEK-FAP) after treatment with HFE7A-CRIPTO bispecific antibody are shown in fig. 11. Both 2 cell lines undergo apoptosis if incubated with a positive control antibody that has conferred apoptosis in solution or when treated with cross-linked HFE 7A. The anti-FAS antibody HFE7A itself did not mediate apoptosis in these cell lines. The bispecific HFE7A-CRIPTO molecule induced apoptosis only in HEK-CRIPTO cells but not in control HEK-FAP cells. There appears to be a low level of DNA fragmentation also in HEK-FAP cells, but this is a non-specific basal activity as it can also be observed with the unrelated HFE7A-MCSP control molecule and even with anti-CRIPTO and anti-Fc antibodies alone. As observed in the experiment depicted in figure 10, inhibition of apoptosis by pre-incubation with excess crispoigg was incomplete in this case.

Example 5: generation of FAS-MCSP bispecific death receptor agonist antibodies and evaluation of apoptosis-inducing potential thereof

Among the antigens that are directly expressed and displayed on the surface of tumor cells, other antigens are also contemplated for targeted cross-linking of death receptors to induce apoptosis. In particular antigens from stroma or neovasculature. An example of the latter is melanoma-associated chondroitin sulfate proteoglycan (MCSP). MCSP is expressed on most melanoma cells as well as glioma cells and neovasculature. Several monoclonal antibodies targeting human MCSP have been described, but none are suitable for cancer therapy due to lack of potency (e.g., lack of ADCC). Thus the MSCP antibodies would be exploited if used in a bispecific format capable of mediating apoptosis targeting to tumor sites.

To assess simultaneous tumor/neovasculature targeting with respect to apoptosis induction, bispecific death receptor agonist antibodies were generated in which an MCSP-specific scFv (wild-type or disulfide-stabilized) was fused to the C-terminus of the anti-FAS antibody HFE 7A. These scfvs are fused to HFE7A via a short peptide linker. The sequences used to generate the light and heavy chain variable regions of MCSP-targeted scFv were derived from MCSP antibody 9.2.27(Beavers et al, 1996; US 5580774).

To define a cell line suitable for analysis of apoptosis induction in vitro, MCSP expression was examined for several stem cell lines by FACS binding analysis (figure 12). In the cell lines tested, only HCT-116 and U-87MG displayed significant MCSP expression as detected with 2 anti-MCSP antibodies (9.2.27 and LC 007). All other cell lines tested showed only very low or no expression of MCSP. Both cell lines were therefore analyzed for whether they underwent apoptosis when treated with cross-linked agonist death receptor antibodies or with control antibodies that had conferred apoptosis in solution. In U-87MG cells, both anti-FAS and anti-DR 5 antibodies induced apoptosis (FIG. 13A), whereas HCT-116 cells were different. Here only the anti-DR 5 antibody induced apoptosis (fig. 13B). U-87MG cells were therefore selected as target cells for subsequent apoptosis induction experiments.

FIG. 14 shows the results of an apoptosis induction experiment using the glioma cell line U-87MG after treatment with a FAS agonistic bispecific antibody consisting of FAS-targeting HFE7AIgG combined with MCSP-binding scFv (9.2.27) at a concentration of 1. mu.g/ml. Both wild type (form a) and H44/L100 disulfide stabilized scFv (form a1) were compared to HFE7A alone or HFE7A crosslinked by a second anti-human Fc antibody. Although in general the induction of apoptosis of these U-87MG cells was rather low (even after 24 hours incubation), significant DNA fragmentation was observed when using the bispecific FAS agonistic antibody. In this case, constructs comprising disulfide-stabilized scFv appeared to be more efficient than constructs comprising wild-type scFv, and both showed much higher apoptosis-inducing ability than cross-linked HFE7AIgG molecules. Preincubation of cells with 100-fold excess MCSP (9.2.27) IgG completely inhibited apoptosis induction of bispecific constructs, indicating that DNA fragmentation/apoptosis observed in the absence of competing antibodies is specific and dependent on cross-linking of FAS by MCSP antigen.

Example 6: DR5-FAP death receptor agonist bispecific antibodies are capable of mediating apoptosis in one cell line by cross-linking of a second cell line

Another approach to apoptosis induction through cross-linking of the death receptor DR5 (in addition to cross-linking through tumor cell-expressed antigens) is to target the stroma around the tumor. In this case, the targeted antigen is not displayed directly on the tumor cell, but on a second, different cell type. An example of such an antigen is FAP (fibroblast activation protein). This protein is expressed on activated fibroblasts as it is found in the tumor stroma.

To investigate the possibility of inducing tumor-targeted apoptosis using bispecific death receptor agonist antibodies targeting human DR5 and derived from tumor stromal antigens, bispecific molecules consisting of an IgG1 moiety that recognizes DR5 and an FAP-binding scFv fused to the C-terminus of the antibody heavy chain were generated. The sequence of the IgG targeting DR5 was derived from ApomAb sequence as described in US2007/0031414a 1. The sequences of the heavy and light chain variable regions of the scFv that bind FAP are from Fab anti-FAP molecules isolated by phage display as shown in sequences #1 and 2. Fascfv was fused to the C-terminus of the anti-DR 5IgG heavy chain via a (G4S)2 linker.

In such a setup 2 different cell lines have to be used for in vitro activity assays: one cell line (the target cell line) should express human DR5, must be capable of apoptosis but need not express FAP. The second cell line (effector cell line) must be apoptosis negative (either by apoptosis resistance or by not expressing DR5) but requires expression of FAP on the surface.

One possible effector cell line meeting the desired criteria is the human fibroblast cell line GM 05389. As shown in figure 15A, this cell line expressed significant levels of FAP compared to cell line SW872 which only showed FAP expression at the highest antibody concentration detected (10 μ g/ml), but which did not undergo apoptosis induced by uncrosslinked ApomAb as shown in figure 15B. This cell line thus appears to be a potential effector cell line in an apoptosis assay in which DNA fragmentation of a target cell line is induced by cross-linking of antigens expressed on a second cell line.

Human breast cancer cell line MDA-MB-231 expressing low levels of DR5 and sensitive to DR 5-mediated induction of apoptosis was used as the target cell line. In fig. 16, the induction results of DNA fragmentation by tumor-targeted cross-linked GM05389 cells and MDA-MB-231 cells of DR5 of FAP compared to the two cell line combinations are summarized. Significant induction of apoptosis after incubation with death receptor agonist antibody was observed only when 2 cell lines were co-cultured (black bars), whereas only a lower degree of apoptosis induced by cross-linking of DR5 with anti-human Fc-targeted ApomAb was detectable in 2 cell lines alone (white and grey bars, respectively). We interpret this result by the fact that the DR5 receptor on MDA-MB-231 cells is cross-linked by binding to the FAP antigen expressed by the fibroblast line GM 05389.

Example 7: CEA single chain Fab molecule (scFab) fused to ApomAb to generate DR5-CEA bispecific agonistic antibody

In addition to stabilizing bispecific antibodies against aggregate formation by insertion of internal cysteine residues as defined in the heavy and light chains of scFv, the use of single chain fab (scfab) is another possible strategy to stabilize whole bispecific antibodies against non-specific cross-linking.

To assess whether this format (scFab fused to DR5 agonist antibody) displays apoptosis-inducing activity similar to that of the corresponding scFv-containing molecule, different bispecific antibodies were generated by standard recombinant DNA techniques, in which CEAscFab was fused to the C-terminus of the heavy or light chain of ApomAb.

The orientation of the different domains of scFab is as follows: VL-CL-VH-CH 1. The C-terminus of the light chain constant region (CL) is linked to the N-terminus of the heavy chain variable region (VH) by a 34 amino acid peptide linker. Fusion of scFab was performed via the G4S linker (2 amino acids or 4 amino acids).

2 bispecific antibodies comprising a single chain Fab were generated which differed in basic form: in one format, 2 scfabs are fused to the C-terminus of the heavy or light chain of an ApomAb (bispecific, tetravalent homodimeric molecules). On the other hand, bispecific molecules (bispecific, trivalent heterodimeric molecules) were constructed in which only 1 scFab was fused to only one ApomAb heavy chain C-terminus. This heterodimerization is obtained by using the so-called knob and hole technique, wherein Fc mutations are used which only allow the formation of heterologous IgG molecules.

In fig. 17, the results of an apoptosis induction experiment are shown, in which ApomAb _ PR1a3_ scFab is compared to ApomAb or highly cross-linked ApomAb. The gastric cancer cell line MKN-45 was used in this assay and apoptosis was measured after 24 hours using a DNA fragmentation assay. Clearly, the bispecific constructs displayed the same range of apoptosis-inducing activity as observed with ApomAb cross-linked by anti-Fc antibodies, and the activity was significantly higher than that observed with ApomAb alone. However, ApomAb-self-induced apoptosis was rather high, most likely due to the extended incubation time of 24 hours, which is the time necessary to show maximal apoptosis induction on the MKN-45 cell line used (in contrast to, for example, LS174T cells, assays using LS174T cells were only performed for 4 hours).

To assess whether bispecific trivalent DR5 agonist antibodies (monovalent for tumor target CEA, and bivalent for DR5) were also able to induce apoptosis in targeted tumors, molecules were generated in which CEAscFab (sm3e specific) was fused to the C-terminus of the ApomAb heavy chain (containing the knob mutation). This heavy chain was co-expressed with the corresponding ApomAb heavy chain and ApomAb light chain comprising a "hole" mutation. The results of the 4 hour apoptosis induction assay are summarized in fig. 18, where bispecific trivalent molecules (concentrations of 0.1 and 1.0 μ g/ml) were analyzed on LS174T cells. From these results it is clear that the described trivalent form is also capable of inducing targeted apoptosis in the same range as the apoptosis induced by highly cross-linked ApomAb. At lower concentrations, the bispecific format appears to be even more active than the cross-linked ApomAb.

Example 8: the DR5-CEA bispecific agonist antibodies have higher in vivo potency than ApomAb

To evaluate that the apoptotic activity of the death receptor agonist antibody, which has been demonstrated in vitro, is sufficient to also have high in vivo efficacy, an in vivo experiment using human colon cancer cell line LS174T as a model was established.

Briefly, on day 1 of the experiment, 3X10 was used⁶Intrasplenic injection of tumor cells female SCID beige mice were treated. On day 7, tumor transplantation of the animals was examined as a standard for starting the treatment with the antibody one day later. Treatment consisted of a series of 3 injections (10 mg/kg each, i.v. 7 days apart). Animals were analyzed daily to show termination criteria.

Figure 19 summarizes the results obtained in this in vivo experiment. Here the survival of 3 groups of mice (each group consisting of the initial 10 animals and treated with different molecules) was compared. Although the control group (PBS, black line) was completely terminated 37 days after tumor injection, the group treated with ApomAb (filled circles) showed prolonged survival (maximum 44 days). The group treated with bispecific antibody (ApomAb _ sm3e _ a1, black squares) showed even longer survival (52 days) than the group obtained with ApomAb alone. Mathematical analysis of the data obtained demonstrated that these results were statistically significant (p-value less than 0.05), indicating that ApomAb showed in vivo efficacy compared to PBS control, and that bispecific ApomAb _ sm3e _ a1 demonstrated even higher in vivo efficacy than ApomAb.

Materials and methods:

transfection of HEK293EBNA cells

All (bispecific) antibodies used herein were transiently produced in HEK293EBMA cells using calcium phosphate dependent co-transfection procedures for heavy and light chain vectors as described below.

At 37 ℃ with 5% CO₂Cells were cultured in standard DMEM medium (Invitrogen) containing 10% FCS (Gibco, #16000) in an atmospheric humidified incubator. 48 hours before transfection, 3X10⁷Cells were seeded in roller bottles (Falcon #353069, 1400 cm)²) 200ml DMEM/10% FCS in roller bottle incubator (0.3rpm) at 37 ℃. For transfection, with H₂O.mu.g of total DNA (440. mu.g of each of the heavy and light chain vectors) +4.4ml of CaCl₂Fill to a total volume of 8.8 ml. The solution was mixed briefly. After mixing, 8.8ml of 1.5mM phosphate buffer (50mM hepes, 280mM NaCl, 1.5mM NaH) was added₂PO₄(ii) a pH7.05) was used for DNA precipitation. After mixing for an additional 10 seconds and a short incubation at room temperature (20 seconds), 200ml of DMEM/2% FCS was added to the DNA solution. media/DNA solutions were used to replace the original media in the roller bottles to transfect the cells. After 48 hours of incubation at 37 ℃, the transfection medium was replaced with 200ml dmem/10% FCS and antibodies were produced for 7 days.

The supernatant was harvested after production and the antibody-containing supernatant was filtered through a 0.22 μm sterile filter and stored at 4 ℃ until purification.

Purification of

Proteins were produced by transient expression in HEK293EBNA cells. Using standard procedures, e.g. protein A affinity purification: (Explorer) and size exclusion chromatography all bispecific molecules described herein were purified in two steps.

The supernatant was adjusted to ph8.0 (using 2mtris ph8.0) and applied to a magelect sure resin (GEHealthcare) contained in a tricorn 5/50 column (GEHealthcare, column volume (cv) ═ 1ml) equilibrated with buffer a (50mM sodium phosphate, ph7.0, 250mM naci). After washing with 10 column volumes (cv) of buffer A, 20cv of buffer B (50mM sodium phosphate, pH7.0, 1M NaCl) and once more 10cv of buffer A, a transition to buffer B (50mM phosphate) of more than 20cv was usedSodium, 50mM sodium citrate pH3.0, 250mM nacl) to elute the protein. The protein containing fractions were pooled and the solution pH was slowly adjusted to pH6.0 (using 2 MTRISpH8.0). The sample was concentrated to 2ml using an ultracentrifuge (Vivaspin15R30.000MWCOHY, Sartorius) and subsequently applied to HiLoadTM16/60Superdex TM200 preparative grade (GEHealthcare) equilibrated with 20mM histidine, pH6.0, 150mM NaCl. The eluted fractions were analyzed for aggregate content by analytical size exclusion chromatography. Therefore 50. mu.l of each fraction were loaded onto a column using 2mM MAPS, pH7.4, 150mM NaCl, 0.02% w/vNaN₃Equilibrated Superdex (TM) 20010/300GL column (GEHealthcare). Fractions of less than 2% oligomers were combined and concentrated to a final concentration of 1-1.5mg/ml using an ultracentrifuge (Vivaspin15R30.000MWCOHY, Sartorius). The purified protein was frozen in liquid nitrogen and stored at-80 ℃.

FACS binding analysis

Prior to the apoptosis assay, all target cell lines used were analyzed for relative expression levels of tumor associated antigen and FAS or DR5 death receptor.

Cell number and viability were determined. For this purpose, adherent growing cells were detached using cell isolation buffer (Gibco-Invitrogen # 13151-014). Cells were harvested by centrifugation (4 min, 400Xg), washed with FACS buffer (PBS/0.1% BSA) and the cell number adjusted to 1.111X10 in FACS buffer⁶Cells/ml. Using 180. mu.l of this cell suspension per well of a 96-well round bottom plate, 2X10 per well was obtained⁵A cell. Cells were incubated with primary antibody at 4 ℃ for 30 minutes at appropriate dilutions. The cells were then harvested by centrifugation (4 min, 400Xg), the supernatant was removed thoroughly and the cells were washed once with 150. mu.l FACS buffer. The cells were resuspended in 150. mu.l FACS buffer and incubated with the secondary antibody (in the case of the use of unlabelled primary antibody) in the dark for 30 minutes at 4 ℃. After 2 steps of washing with FACS buffer, cells were resuspended in 200 μ Ι FACS buffer and analyzed in htsfacscanntoii (BD, software FACSDiva). Alternatively, cells can be fixed with 200 μ l of 2% PFA (paraformaldehyde) in FACS buffer for 20 minutes at 4 ℃ and analyzed later. All assays were performed in triplicate.

Antibodies and concentrations used:

biacore analysis (surface plasmon resonance, SPR)

SPR experiments were performed on BiacoreT100 using HBS-EP (0.01MHEPESpH7.4, 0.15MNaCl, 3mMEDTA, 0.005% surfactant P20, GEHealthcare) as the running buffer. Direct ligation of the biotinylated antigens in 1220, 740 and 300 Resonance Units (RU) was performed on streptavidin chips using standard methods (GEHealthcare), respectively. Different concentrations of bispecific death receptor-activating antibody were passed through the flow cell at 278K at a flow rate of 40 μ l/min for 90 seconds to record the binding phase. The 300 second dissociation phase was monitored and triggered by switching the sample solution to HBS-EP. The overall refractive index difference was corrected by subtracting the response obtained from an empty streptavidin surface. Kinetic constants were obtained using BiacoreT100 evaluation software (vAA, Biacore, Freiburg/Germany) to fit a rate equation for 1: 1 Langmuir binding by numerical integration. Since the antigen is immobilized, the use of Langmuir binding by numerical integration of 1: 1 gives only apparent KD values or avidity.

Induction of apoptosis

To determine the induced apoptosis, the elispaplus kit was used for cell death detection from Roche. Briefly, each well of a 96-well plate was inoculated with 10 in 200. mu.l of the appropriate medium⁴Cells (cell number and survival after isolation) and in 5% CO₂The culture was carried out overnight at 37 ℃ in the atmosphere. The next day the medium was replaced with fresh medium containing appropriate concentrations of apoptosis-inducing antibodies, control antibodies and other controls:

bispecific antibody is used at a final concentration of 0.01-10 μ g/ml; control antibody was used at 0.5. mu.g/ml and cross-linking antibody was used at 100. mu.g/ml. A 100-fold excess of competing antibody was used.

At 37 ℃ 5% CO₂Cells were cultured for 4-24 hours to allow induction of apoptosis. Cells were harvested by centrifugation (10 min, 200Xg) and incubated in 200. mu.l lysis buffer (provided in the kit) for 1 hour at room temperature. Intact cells were pelleted by centrifugation (10 min, 200Xg) and 20. mu.l of supernatant was analyzed according to the manufacturer's recommendations to detect induction of apoptosis.

While the preferred embodiments of the invention have been illustrated and described, it is to be clearly understood that this invention is not limited to the preferred embodiments and may be otherwise variously embodied and practiced within the scope of the following claims.

Claims (19)

1. A dimeric bispecific antibody comprising a first antigen-binding site specific for a death receptor 5 polypeptide (DR5) and a second antibody comprising a second antigen-binding site specific for Fibroblast Activation Protein (FAP), wherein the dimeric bispecific antibody is bivalent for DR 5.

2. The bispecific antibody of claim 1, wherein the first and second antibodies comprise an Fc portion of an antibody heavy chain, wherein the Fc portion of the first antibody comprises a first dimerization module and the Fc portion of the second antibody comprises a second dimerization module that allows heterodimerization of the two antibodies.

3. The bispecific antibody of claim 2, wherein the first dimerization module comprises a knob and the second dimerization module comprises a hole according to a knob-and-hole structural strategy.

4. The bispecific antibody of claim 1, wherein the first antibody is an immunoglobulin Ig molecule comprising a light chain and a heavy chain and the second antibody is selected from scFv, scFab, Fab or Fv.

5. The bispecific antibody of claim 4, wherein the second antibody is fused to the N-or C-terminus of the heavy chain of an Ig molecule.

6. The bispecific antibody of claim 4, wherein the second antibody is fused to the N-or C-terminus of the light chain of an Ig molecule.

7. The bispecific antibody of any one of claims 4-6, wherein the second antibody is fused to the Ig molecule by a peptide linker.

8. The bispecific antibody of claim 7, wherein the peptide linker is a peptide linker having a length of 10-30 amino acids.

9. The bispecific antibody of any one of claims 4-6 and 8, wherein the second antibody comprises an additional cysteine residue to form a disulfide bond.

10. The bispecific antibody of claim 7, wherein the second antibody comprises an additional cysteine residue to form a disulfide bond.

11. The bispecific antibody of any one of claims 4-6, 8 and 10, wherein the Ig molecule comprises an Fc variant having a reduced affinity for an Fcyreceptor compared to a wild-type Fc region.

12. The bispecific antibody of claim 7, wherein the Ig molecule comprises an Fc variant having reduced affinity for an Fcyreceptor compared to a wild-type Fc region.

13. The bispecific antibody of claim 9, wherein the Ig molecule comprises an Fc variant having reduced affinity for an Fcyreceptor compared to a wild-type Fc region.

14. The bispecific antibody of any one of claims 4-6, 8, 10, 12 and 13, wherein the Ig molecule is an IgG.

15. The bispecific antibody of claim 7, wherein the Ig molecule is an IgG.

16. The bispecific antibody of claim 9, wherein the Ig molecule is an IgG.

17. The bispecific antibody of claim 11, wherein the Ig molecule is an IgG.

18. A pharmaceutical composition comprising the bispecific antibody of any one of claims 1-17.

19. Use of a bispecific antibody of any one of claims 1-17 for the manufacture of a medicament for the treatment of cancer.