CA2638794A1 - Functional antibodies - Google Patents

Abstract

The invention is directed to novel antibodies, which comprise single domain binding sites. The antibodies can be bivalent or multivalent, and can be bispecific. The invention is further directed to monospecific and bispecific antibodies that bind to mPDGFR.alpha.. The antibodies can be administered alone or in combination with anti- angiogenic or anti-neoplastic drugs.

Description

FUNCTIONAL ANTIBODIES
CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Application No.60/773,994, filed February 15, 2006, which is incorporated herein by reference in its entirety.

FILED OF THE INVENTION

[0002] The present invention is directed to novel antibodies, which comprise single domain binding sites. The antibodies can be bivalent or multivalent, and can be bispecific or multispecific. Also disclosed are bispecific and multi specific antibodies that bind to cell surface receptor proteins. When administered to a subject alone or in combination with anti-angiogenic or anti-neoplastic drugs, the antibodies can be used to inhibit tumor growth as well as to inhibit inhibit hyperproliferative diseases.

BACKGROUND

[0003] Bispecific antibodies (BsAb) are immunoglobulin (Ig)-based molecules that bind to two different epitopes on either the same or distinct antigens. Both laboratory and early clinical studies have demonstrated that BsAb may have significant applications in cancer therapy, for example by targeting tumor cells with cytotoxic agents such as effector cells, radionuclides, drugs and toxins (Weiner et al., 1997, Cancer Immunol.
hnmunother.
45:190-2.; van Spriel et al., 2000, Immuol. Today 21:391-7; Segal et al., 2000, J. Immunol.
Methods 248:1-6.), or by simultaneously targeting two different tumor targets (or epitopes) in order to enhance the biological activities of individual antibody therapeutics (Lu et al., 1999, J. Immunol.Methods 230:159-71; Lu et al., 2001, Cancer Res. 61:7002-8.; Lu et al., 2002, J.
Inununol. Methods 267:213-26). A major obstacle in the development of BsAb-based therapeutics has been the difficulty in producing the materials in sufficient quantity and quality for clinical studies via traditional methods, including the hybrid hybridoma and chemical conjugation (Carter et al., 1995, J. Hematotherapy 4:463-70).
Coexpression of two different sets of IgG light and heavy chains can result in a variety of light-and heavy-chain pairs, only one of which is the desired functional bispecific heterodimer (Suresh et al., 1986, Methods Enzymol. 121:210-28). On the other hand, chemical cross-linking of two IgGs or their fragments is often inefficient arid can lead to the loss of antibody activity (Zhu et al., 1994, Cancer Lett. 86:127-34). In both methods, purification of the BsAb from the non-functional species, such as homodimers and mispaired heterodimers of non-cognate Ig light and heavy chains produced by the hybrid hybridoma, and multimeric aggregates resulting from chemical conjugation, is often difficult and the yield is usually low (Cao et al., 1998, Bioconj. Chem. 9:635-44).

[0004] To improve efficiency, a variety of recombinant methods have been developed for production of BsAb, both as antibody fragments (Carter et a1.,1995;
Pluckthun et al., 1997, Immunotechology 3:83-105; Todorovska et al., 2001, J. Immunol. Methods 248:47-66) and full length IgG formats (Carter, 2001, J. Immunol. Methods 248:7-15). For example, production of homogeneous full-length IgG-like BsAb has been achieved by the so-called "knobs-into-holes" engineering for efficient Ig CH3 domain heterodimerization (Ridgway et al., 1996, Protein Eng. 9:617-21; Merchant et al., 1998, Nat. Biotech. 16:677-8 1) and by fusing two single chain Fv (scFv) of different specificities onto either the N-or the C-terminus of a full-length IgG molecule (Zhuang et al., 2000, Protein Eng.
13:361-7; Coloma and Morrison, 1997, Nat. Biotechnol. 15:159-63). BsAbs have also been constructed by genetically fusing two single chain Fv (scFv) or Fab fragments with or without the use of flexible linkers (Mallender et al., 1994, J. Biol. Chem. 269:199-206; Mack et al., 1995, Proc.
Natl. Acad. Sci. USA. 92:7021-5; Zapata et al., 1995, Protein Eng. 8:1057-62), via a dimerization device such as leucine zipper (Kostelny et al., 1992, J. Immunol.
148:1547-53;
de Kruif et al., 1996, J. Biol. Chem. 271:7630-4), and Ig CL/CH1 domains (Muller et al., 1998, FEBS Left. 422:259-64); by diabody (Holliger et al., 1993, Proc. Nat.
Acad. Sci. USA.
90:6444-8; Zhu et al., 1996, Bio/Technology (NY) 14:192-6); Fab-scFv fusion (Lu et al., 2002; Schoonjans et al., 2000, J. Immunol. 165:7050-7); and miniantibody formats (Pack et al., 1992, Biochemistry 31:1579-84; Pack et al., 1993, Bio/Technology 11:1271-7). In majority of the cases, these recombinant approaches result in the production of divalent bispecific antibody molecules that are monovalent to each of their target antigens. Further, examples IgG-like bispecific antibodies comprising functional Fc domains are limited.
SUMMARY OF THE INVENTION

[0005] The invention provides novel bispecific antibodies which comprise single variable domain (sVD) antigen binding sites. The antibodies of the invention can also include antigen binding sites that coinprise antibody Fvs in addition to sVD
antigen binding sites.

[0006] Antibodies of the invention can be specific for any antigen. In one embodiment, an antibody of the invention binds to a cell surface antigen, which can be a receptor tyrosine kinase (including, but not limited to PDGFR, VEGFRI, VEGFR2, EGFR).
In another embodiment, an antibody of the invention binds to a ligand of a cell surface receptor. In a preferred embodiment, an antibody of the invention has receptor neutralizing activity. The invention further provides pharmaceutical compositions of the antibodies.

[0007] The present invention provides methods of inhibiting activation of one or more receptor tyrosine kinases. Tumor growth in a mammal can be treated or inhibited by administering to the mammal an effective amount of a present antibody. The present antibodies can be coadministered with antibodies that bind to other cell surface antigen (including, e.g., RTKs) or cytokines (including, e.g., RTK ligands). In certain embodiments, the methods also comprise administering to the mammal an anti-neoplastic agent or treatment, including, for example, a chemotherapeutic agent and/or radiation.
In another embodiment, the invention provides a method of treating a non-cancer hyperproliferative disease, e.g., psoriasis, in a mammal.

BRIEF DESCRIPTION OF THE FIGURES

[0008] Figure 1 is a schematic diagram depicting examples of single domain antibody-based bispecific antigen binding proteins of the invention. The single domain (designatetd VH2 in the drawing) can be incorporated by fusion to the N-ten.~ninus or C-terminus of other antibody domains.

[0009] Figure 2 depicts expression and purification of mPDGFRa-specific Fabs.
The Fabs were expressed in E. coli host HB2151 cells, purified by affinity chromatography, and analyzed by SDS-PAGE. A) "Standard" mPDGFRat specific Fabs having heavy (VH-CHI) and light (VL-CL) chains of similar molecular weight. B) IF2 and 1F9 Fab fragments lacking VL domains and a standard Fab control in the presence or absence of DTT.
Molecular weight markers are in kilodaltons.

[0010] Figure 3 depicts binding and blocking assays of purified Fabs. A) quantitative binding ELISA of purified Fabs to murine PDGFRa- Various amount of Fabs were first incubated in a 96-well plate precoated with mPDGFRcx Fc fusion protein followed by incubation with a rabbit anti-anti-human-rc antibody HRP conjugate. Plate-bound antibody-HRP was quantified by addition of peroxidase substrate, and absorbance was read at A450 nm. B) inhibition of binding of mPDGFRa to immobilized PDGF-AA by purified Fabs.
Various amounts of Fabs were incubated with a fixed amount of mPDGFRa/Fc in solution.
The mixtures were incubated in 96-well plates coated with PDGF-AA followed by anti-human-Fc antibody-HRP conjugate. Bound mPDGFRa was then quantified by addition of peroxidase substrate, and absorbance was read at A450 nm. Data shown represent the mean =LSD of duplicate samples.

[0011] Figure 4 depicts structures of monovalent and bivalent Fabs of the invention.

[0012] Figure 5 depicts expression and purification of bivalent mPDGFRa specific Fabs. The Fabs were expressed in E. coli host HB2151 cells, purified by affinity chromatography, and analyzed by SDS-PAGE. Bivalent 1F2-2H Fab is compared with a standard Fab (5C5) under non-reducing conditions. Molecular weight markers are in kilodaltons.

[0013] Figure 6 depicts binding and blocking assays of purified antibodies. A) quantitative binding assay of purified anti-n-iPDGFRa monovalent and bivalent 1 F2 Fab antibodies. Various amounts of antibodies were incubated in a 96-well plate precoated with mPDGFRa/Fc, followed by incubation with an anti-human--c antibody HRP
conjugate. The plate-bound antibody-HRP was then quantified by the addition of peroxidase substrate, and absorbance was read at A450 nm. B) inhibition of the binding of mPDGFRa to immobilized PDGF-AA by the purified Fab antibodies. Various amounts of antibodies were incubated with a fixed amount of mPDGFRadFc and the mixtures were transferred to plates precoated with PDGF-AA. After wash, the plates were then incubated with an anti-human-Fe antibody-HRP conjugate. Bound mPDGFR[x was quantified by addition of peroxidase substrate, and absorbance was read at A450 nn-i. Data shown represent the mean SD of duplicate samples.
2B4 is an antibodies directed against mouse VEGFR2 (Flkl).

[0014] Figure 7 provides a schematic representation of the 1F2-based full length IgGs. The 1 F2 VH is expressed as a fusion protein to CL and/or CH, providing tetravalent (panel A; MW -150,000) or bivalent (panels B and C:11IW-125,000) antibodies.

[0015] Figure 8 depicts expression and purification of anti-mPDGFRa antibodies. In the right panel, antibodies were treated with DTT prior the electrophoresis.

[0016] Figure 9 depicts binding and ligand blocking assays for purified antibodies. A) quantitative binding purified anti-mPDGFRa antibodies. Various amounts of antibodies were incubated in a 96-well plate precoated with mPDGFRtx/Fc, followed by incubation with an anti-human-K antibody HRP conjugate. Plate-bound antibody-HRP was quantified by the addition of peroxidase substrate, and absorbance was read at A450 nm. B) inhibition of binding of mPDGFRce to immobilized PDGF-AA by the purified antibodies. Various amounts of antibodies were incubated with a fixed amount of mPDGFRa/Fc and transferred to plates precoated with PDGF-AA. Plates were washed and incubated with an anti-human-Fc antibody-HRP conjugate. Bound mPDGFRa was then quantified by addition of peroxidase substrate, and absorbance was read at A450 nm. Data represent the mean =LSD of duplicate samples. 2B4 is an antibody directed against mouse VEGFR2 (Flkl).

[0017] Figure 10 provides a schematic representation of selected 1F2-based tetravalent bispecific antibodies that bind to mPDGFRa and flk-1. A) IgG-like antibody designated 1F2/scFv2B4. 1F2 VH is expressed as a fusion to CL, and 2B4 is expressed as a scFv fusion to CH. B) The IgG-like antibody designated 1F2-2B4. 1F2 Vx is expressed as a fusion at the amino terminus of 2B4 VL.

[0018] Figure 11 depicts expression and purification of the 1F2-based bispecific antibodies. The antibodies were expressed in mammalian cells, purified by affinity chromatography, and analyzed by SDS-PAGE. The antibodies were treated with/without (+/-) DTT prior the electrophoresis. Molecular weight markers are in kilodaltons.

[0019] Figure 12 depicts quantitative binding assays of purified anti-mPDGFRa x anti-Flk-1 bispecific antibodies. A) Various amounts of antibodies were first incubated in a 96-well plate precoated with mPDGFRa/Fc, followed by incubation with an anti-human-K
antibody HlZP conjugate. B). Various amounts of antibodies were first incubated in a 96-well plate precoated with Flk-1/Fc, followed by'incubation with an anti-human-K
antibody HRP
conjugate. Data shown represent the mean tSD of duplicate samples.

[0020] Figure 13 depicts blocking assays of purified bispecific anti-mPDGFRCY
x anti-Flk-1 bispecific antibodies. A) inhibition of the binding of mPDGFRa to immobilized PDGF-AA by the purified bispecific antibodies. B) inhibition of the binding ofF1k-1 to immobilized VEGF165 by the purified bispecific antibodies. In both assays, various amounts of antibodies were incubated with a fixed amount of receptor (mPDGFRa/Fc or Flk-1) in solution and transferred to plates precoated with ligand (PDGF-AA or VEGF165).
The plates were then incubated with an anti-human-Fc antibody-HRP conjugate. Bound mPDGFRa was then quantified by addition of peroxidase substrate, and absorbance was read at A450 nm.
Data shown represent the mean SD of duplicate samples.

[0021] Figure 14 depicts inhibition ofPDGF and VEGF-stimulated receptor phosphorylation by the bispecific 1F2-2B4IgG. eEnd.1 cells were first incubated with various antibodies at 37 C for 30 min, followed by stimulation with VEGF or PDGF at 37 C
for 15 min. After cell lysis, the receptors were immunoprecipitated from the cell lysate supernatant by incubation with an anti-mPDGFRa or an anti-mVEGFR2 antibody, followed by ProA/G-sepharose beads. The precipitated receptor proteins were resolved on a 4-12%
Nupage Bis-Tris gel and transferred to a polyvinylidene difluoride membrane.
Phospho-mVEGFR2 and phospho-mPDGFRa were detected on the blot using an anti-phospho-tyrosine antibody-HRP conjugate. Total receptor proteins loaded on the gel were assayed with antibodies to mPDGFRa or mVEGFR2.

[0022] Figure 15 depicts the amino acid sequences of VH and VL domains of several identified Fab proteins. Fabs 1E10, 1A11, 3B2, 1C10, 3G7, 1F9, and 1F2 were identified in a screen for Fab-phage that bind to murine PDGFRu- Fab 2B4 was identified in a screen for Fab-phage that bind to murine VEGFR2. ^: frame-shift mutation; * stop codon resulting from frame-shift mutation; @...@: in-frame deletion.

DETAILED DESCRIPTION OF THE INVENTION

[0023] The invention provides novel antibodies that comprise single variable domain (sVD) antigen binding sites. The invention also provides antigen binding proteins which comprise both single domain antigen binding sites and two-domain, Fv-like antigen binding sites.

[0024] Single domain antigen binding sites are similar to immunoglobulin variable domains (e.g., VH or VL) but are capable of specific binding to antigen in the absence of a second antigen binding domain (e.g., a counterpart VL or VH). Such single domains may in fact be incapable of stable association with counterpart binding domains.
However, single domain antibodies can bind antigen with affinities and avidities similar to those of antibodies that include both VL and VH domains. Moreover, like natural antibodies, single domain antibodies can block receptor-ligand interactions. Accordingly, antibodies of the invention that comprise single domain binding sites are used for the same 'purposes as antibodies obtained from, for example, hybridomas, transgenic mice expressing human antibodies, and phage display libraries of Fab or scFv binding domains.

[0025] Antibodies of the invention are immunoglobulin-like in several respects. 1) antibodies of the invention that are IgG-like are heterotetramers, consisting of two of each of two dissimilar polypeptide chains. 2) The dissimilar polypeptide chains associate and may be covalently linked, for example by disulfide bonds in same manner of heavy and light chain constant regions of standard, naturally occurring immunoglobulins. 3) One of the polypeptide chains is capable of stable self association in the manner of naturally occurring immunoglobulin heavy chains. For example, one of the polypeptide chains can include one or more domains corresponding to CH2, CH3, or CH4 of a naturally occurring immunoglobulin. In a preferred embodiment, an antibody of the invention comprises the constant domain structure of a naturally occurring IgG.

[0026] One enormous advantage of an immunoglobulin-like structure is that antibody chains pair naturally upon expression. Unlike many tetrameric antibodies of the prior art, no further manipulations are necessary to obtain a preparation of homogeneous antibodies.

[0027] Antigen-binding proteins of the invention comprise antigen binding sites that are single variable domains (sVDs). In certain embodiments of the invention, an sVD
binding site substitutes for one variable domain of an IgG-like antibody, and a scFv substitutes for the other variable domain. Examples of such embodiments are depicted in Fig.
1 A. In other embodiments of the invention, an sVD binding site is grafted to an IgG antibody at the N terminus or C terminus of the IgG heavy or light chain. Examples of such embodiments are depicted in Fig. 1B. The sVDs can also be incorporated into an immunoglobulin-like antibody at other positions. For example, the sVDs can be attached at the C terminus of CH and/or CL constant domains. Moreover, by combining substitutions of sVD binding sites as in Fig. 1 with additions of sVD binding sites to the C
terminus of CH
and/or CL constant domains, IgG-like antibodies that are multispecific and/or multivalent for a particular antigen can be produced.

[0028] As known in the art, an Fv consists of two domains (e.g., a VH and a VL
domain). In a standard antibody, the VH and VL domains of the Fv, though associated to form an antigen binding site, are not directly linked, but are each joined to linked antibody constant regions. The VH and VL domains of the Fv can also be joined with a synthetic linker to form a single chain Fv (scFv).

[0029] Antibody specificity refers to selective recognition of the antibody for a particular epitope of an antigen. Natural antibodies, for example, are monospecific.
Bispecific antibodies (BsAbs) are antibodies which have two different antigen-binding specificities or sites. Where an antigen-binding protein has more than one specificity, the recognized epitopes may be associated with a single antigen or with more than one antigen.

[0030] A natural antibody molecule is composed of two identical heavy chains and two identical light chains. Each light chain is covalently linked to a heavy chain by an interchain disulfide bond. The two heavy chains are further linked to one another by multiple disulfide bonds at the hinge region. The individual chains fold into domains having similar sizes (about 110-125 amino acids) and structures, but different functions. The light chain comprises one variable domain (VL) and one constant domain (CL). The heavy chain comprises one variable domain (VH) and, depending on the class or isotype of antibody, three or four constant domains (CH1, CH2, CH3 and CH4). In mice and humans, the isotypes are IgA, IgD, IgE, IgG, and IgM, with IgA and IgG further subdivided into subclasses or subtypes. The portion of an antibody consisting of VL and VH domains is designated "Fv"
and constitutes the antigen-binding site. A single chain Fv (scFv) is an engineered protein containing a VL domain and a VH domain on one polypeptide chain, wherein the N
terminus of one domain and the C terminus of the other domain are joined by a flexible linker. "Fab"
refers to the portion of the antibody consisting of VL, VH, CL and Cxl domains.

[0031] The variable domains show considerable amino acid sequence variablity from one antibody to the next, particularly at the location of the antigen binding site. Three regions, called "hypervariable" or "complementarity-determining regions"
(CDRs) are found in each of VL and VH.

[0032] "Fc" ? is the designation for the portion of an antibody which comprises paired heavy chain constant domains. In an IgG, antibody, for example, the Fc comprises CH2 and CH3 domains. The Fc of an IgA or an IgM antibody further comprises a CH4 domain. The Fc is associated with Fc receptor binding, activation of complement-mediated cytotoxicity and antibody-dependent cellular-cytotoxicity. For natural antibodies such as IgA
and IgM, which are complexes of multiple IgG like proteins, complex formation requires Fe constant domains.

[0033] Finally, the "hinge" region separates the Fab and Fc portions of the antibody, providing for mobility of Fabs relative to each other and relative to Fc, as well as including multiple disulfide bonds for covalent linkage of the two heavy chains.

[0034] Antibodies of the invention have a combination of desirable features.
First, they are essentially homogeneous. By design, mispairing of antibody heavy and light chains is greatly reduced or eliminated. For example, some bispecific antibodies make use of two different heavy chains to provide two specificities. Four combinations are possible when such heavy chains are arranged into an IgG type molecule. Two of those consist of mispaired heavy chains such that the product is monospecific. In antibodies of the invention, mispairing is substantially eliminated.

[0035] A second feature of the antibodies of the invention is that they are bivalent for each binding specificity. Many bispecific antibodies are monovalent for each of the antibody binding sites that are comprised. This is significant for antibody function because bivalency allows for cooperativity of binding and a significant increase in binding avidity with resepect to a molecule comprising a single antigen-specific binding site.

[0036] A third advantage of the antibodies of the invention is that one or more heavy ~
chain constant domains which constitute the Fc region (e.g., CH2 and/or CH3 for an IgG, molecule) of a natural antibody and which provide other antibody functions are present.
Furthermore, the multiple binding domains are separated from the constant domains such that functions provided by the constant domains are not impaired. Constant domain functions include binding to certain accessory molecules (e.g., binding to cell surface and soluble Fe receptors, J chain association for IgA and IgM, S protein for IgA), activation of the complement pathway (complement dependent cytoxicity, CDC), recognition of antibody bound to target cells by several different leukocyte populations (antibody-dependent cell-mediated cytoxicity, ADCC) and opsonization (enhancement of phagocytosis).
Also, the Fc heavy chain constant domain(s) can confer increased serum half-life.

[0037] A fourth advantage of proteins of the invention is that there is no requirement for processing in vitro to obtain the complete product. Though rearranged in an artificial manner, each of the domains has a natural character which allows expression in a biological system. For example, bispecific antibodies can be expressed in prokaryotic and eukaryotic expression systems. The proteins that are produced are substantially bispecific.
[0038] sVD antigen binding sites and Fv region-containing binding sites for use in an antibody of the invention can be obtained by a variety of methods. The amino acid sequences of the VH and/or VL portions of a selected binding domain can be obtained from a naturally-occurring antibody or are chosen or be modified, screened, or selected for desired binding characteristics. For example, VH and/or VL domains can be obtained directly from a monoclonal antibody which has the desired binding characteristics.
Alternatively, VH and/or VL domains can be from libraries of V gene sequences from a mammal of choice.
Elements of such libraries express random combinations of VH and/or VL domains and are screened with any desired antigen to identify those elements which have desired binding characteristics. Particularly preferred is a human V gene library. Methods for such screening are known in the art. Alternatively, VH and/or VL domains from a selected non-human source may be incorporated into chimeric antibody which comprises human constant domains. For example, for administration to a human, it may be desired to use an antibody with one or more functional human constant domains wherein the VH and VL domains have been selected from a non-human source. To maximize constant domain associated function or to reduce immunogenicity of the antibody, human constant domains are preferred.

[0039] AIternatively, a V domain can be made that is "humanized." Humanized variable domains are constructed in which amino acid sequences which comprise one or more complementarity determining regions (CDRs) of non-human origin are grafted to human framework regions (FRs). For examples, see: Jones, P. T. et al., 1996, Nature 321, 522-25;
Riechman, L. et al., 1988, Nature 332, 323-27; and U.S. Patent No. 5,530,101 to Queen et al.
A humanized construct is particularly valuable for elimination of adverse immunogenic characteristics, for example, where an antigen binding domain from a non-human source is desired to be used for treatment in a human. Variable domains have a high degree of structural homology, allowing easy identification of amino acid residues within variable domains which corresponding to CDRs and FRs. See, e.g., Kabat, E.A., et al., 1991, Sequences of Proteins of Immunological Interest. 5th ed. National Center for Biotechnology Information, National Institutes of Health, Bethesda, MD. Thus, amino acids which participate in antigen binding are easily identified. In addition, methods have been developed to preserve or to enhance affinity for antigen of humanized binding domains comprising grafted CDRs. One way is to include in the recipient variable domain the foreign framework residues which influence the conformation of the CDR regions. A secoirid way is to graft the foreign CDRs onto human variable domains with the closest homology to the foreign variable region. Queen, C. et al., 1989, Proc. Natl. Acad. Sci. USA 86, 10029-33. CDRs are most easily grafted onto different FRs by first amplifying individual FR sequences using overlapping primers which include desired CDR sequences, and joining the resulting gene segments in subsequent amplification reactions. Grafting of a CDR onto a different variable domain can further involve the substitution of amino acid residues which are adjacent to the CDR in the amino acid sequence or packed against the CDR in the folded variable domain structure which affect the conformation of the CDR. Humanized variable domains of the invention therefore include human domains which comprise one or more non-human CDRs as well as such domains in which additional substitutions or replacements have been made to preserve or enhance binding characteristics.

[0040] An antibody of the invention may also employ variable domains which have been made less immunogenic by replacing surface-exposed residues so as to make the antibody appear as self to the immune system (Padlan, E.A., 1991, Mol.
Irnrnunol. 28, 489-98). Antibodies have been modified by this process with no loss of affinity (Roguska et al., 1994, Proc. Natl. Acad. Sci. USA 91, 969-973). Because the internal packing of amino acid residues in the vicinity of the antigen binding site remains unchanged, affinity is preserved.
Substitution of surface-exposed residues according to the invention for the purpose of reduced immunogenicity does not mean substitution of CDR residues or adjacent residues which influence binding characteristics.

[0041] It is often preferable to employ variable domains that are essentially human.
Human binding domains can be obtained from phage display libraries wherein combinations of human heavy and light chain variable domains are displayed on the surface of filamentous phage (See, e.g., McCafferty et al., 1990, Nature 348, 552-54; Aujame et aI., 1997, Human Antibodies 8, 155-68). Combinations of variable domains are typically displayed on filamentous phage in the form of Fabs or scFvs. The library is screened for phage bearing combinations of variable domains having desired antigen binding characteristics. Preferred single domain and variable domain combinations display high affinity for a selected antigen and little cross-reactivity to other related antigens. By screening very large repertoires of antibody fragments, (see e.g., Griffiths et al., 1994, EMBO J. 13, 3245-60) a good diversity of high affinity binding domains are isolated, with many expected to have sub-nanomolar affinities for the desired antigen.

[0042] Alternatively, human binding domains can be obtained from transgenic animals into which unrearranged human Ig gene segments have been introduced and in which the endogenous mouse Ig genes have been inactivated (reviewed in Bruggemann and Taussig, 1997, Curr. Opin. Biotechnot. 8, 455-58). Preferred transgenic animals contain very large contiguous Ig gene fragments that are over 1 Mb in size (Mendez et al., 1997, Nature Genet. 15, 146-56) but human Mabs of moderate affinity can be raised from transgenic animals containing smaller gene loci (See, e.g., Wagner et al., 1994, Eur. J.
Immunol. 42, 2672-81; Green et al., 1994, Nature Genet. 7, 13-21).

[0043] The sVD binding sites can be obtained from antigen specific Fv regions (which comprise both VH and VL domains). Often, it can be shown that the binding affinity and specificity of an Fv region is contributed primarily by one of the variable domains.
Alternatively, the scFv can be obtained directly. Direct sources of sVDs include mammals (e.g., camelids) that naturally express antibodies containing only VH domain and phage display libraries constructed to express only a single variable domain. For example, a human domain antibody phage display library is commercially available from Domantis (Cambridge, UK). As exemplified herein, sVD binding sites specific for PDGFRa were obtained from an Fab library containing variants in which the VL domain had been spontaneously deleted.

[0044] In a physiological immune response, mutation and selection of expressed antibody genes leads to the production of antibodies having high affinity for their target antigen. The VH and VL domains incorporated into antibodies of the invention can similarly be subject to in vitro mutation and screening procedures to obtain high affinity variants.
Thus, binding domains of the invention include those for which binding characteristics have been improved by direct mutation or by methods of affinity maturation.
Affinity and specificity may be modified or improved by mutating CDRs and screening for antigen binding sites having the desired characteristics (See, e.g., Yang et al., 1995, J. Mol. Bio. 254, 392-403). It is understood that amino acid residues that are primary determinants of binding of single domain antibodies can be within Kabat defined CDRs, but may include other residues as well, such as, for example, residues that would otherwise be buried in the VH-VL
interface of a VH-VL heterodimer. CDRs or other residues that determine binding are mutated in a variety of ways. One way is to randomize individual residues or combinations of residues so that in a population of otherwise identical antigen binding sites, all twenty amino acids, or a subset thereof, are found at particular positions.
Alternatively, mutations are induced over a range of CDR residues by error prone PCR methods (See, e.g., Hawkins et al., 1992, J. Mol. Bio. 226, 889-96). Phage display vectors containing genes encoding binding domains (e.g., heavy and/or light chain variable region genes) can be propagated in mutator strains of E. coli (See, e.g., Low et al., 1996, J Mol. Bio. 250, 359-68). These methods of mutagenesis are illustrative of the many methods known to one of skill in the art.

[0045] Each variable domain of the antibodies of the present invention may be a complete immunoglobulin heavy or light chain variable domain, or it may be a functional equivalent or a mutant or derivative of a naturally occurring domain, or a synthetic domain constructed, for example, in vitro using a technique such as one described in (Medical Research Council / Griffiths et al.). For instance, it is possible to incorporate domains corresponding to antibody variable domains which are missing one or more amino acids. The important characterizing feature is the ability of each variable domain to associate with a complementary variable domain to form an antigen binding site.

[0046] Antigen-binding proteins of the invention have binding sites for any epitope, antigenic site or protein. Of particular interest are antibodies that are useful for treatment of disease. Preferred antibodies neutralize receptor proteins, such as receptors which are involved in angiogenesis and/or oncogenesis. Neutralizing a receptor means inactivating the intrinsic kinase activity of the receptor to transduce a signal. A reliable assay for receptor neutralization is the inhibition of receptor phosphorylation. The present invention is not limited by any particular mechanism of receptor neutralization. Some possible mechanisms include preventing binding of the ligand to the extracellular binding domain of the receptor, and preventing dimerization or oligomerization of receptor. Other mechanisms cannot, however, be ruled out.

[0047] Neutralization of activation of a receptor in a sample of endothelial or non-endothelial cells, such as tumor cells, may be performed in vitro or in vivo.
Neutralizing activation of a receptor in a sample of receptor expressing cells comprises contacting the cells with an antibody of the invention. In vitro, the cells are contacted with the antibody before, simultaneously with, or after, adding VEGF to the cell sample. In vivo, an antibody of the invention is contacted with a receptor by administration to a mammal. Methods of administration to a mammal include, for example, oral, intravenous, intraperitoneal, subcutaneous, or intramuscular administration.

[0048] Examples of such receptors include, but are not limited to platelet derived growth factor receptor (PDGF-R), VEGF receptors (e.g., VEGFR-2/KDR/Flk-1, VEGFRI/Flt-1, VEGFR3lFlt-4), epidennal growth factor receptor (EGFR), insulin-like growth factor receptor (IGFR) and the like. Additional non-limiting examples of receptor tyrosine kinases include Flt-4, HER2/neu, Tek and Tie2.

[0049] Other factors implicated as possible regulators of angiogenesis and/or growth of tumors in vivo include fibroblast growth factor (FGF), and nerve growth factor (NGF).
The corresponding receptors are fibroblast growth factor (FGF-R), and nerve growth factor receptor (NGFR). Another receptor implicated in cell migration, morphology changes, and invasiveness is macrophage-stimulating protein receptor ("MSP-R" or "RON").
Receptors of interest include human proteins and homologues from other mammals.

[0050] Antibodies of the invention can incorporate Ig antigen binding domains from any source. For example, antibodies are known for the above listed receptors and are sources of VH and VL domains for use in antibodies of the present invention. Examples of scFv variable region binding domains specific for KDR include, for example, VH and VL domains of IMC-1C11 (nucleotide and amino acids sequences of VH: SEQ ID NOS:1 and 2;
nucleotide and amino acid sequences of VL: SEQ ID NOS:3 and 4) (see, WO
00/44777), IMC-2C6 (nucleotide and amino acids sequences of VH: SEQ ID NOS:5 and 6;
nucleotide and amino acid sequences of VL: SEQ ID NOS:7 and 8) (see, WO 03/075840), and IMC-1121 (nucleotide and amino acids sequences of VH: SEQ ID NOS:5 and 6;
nucleotide and amino acid sequences of VL: SEQ ID NOS:9 and 10) (see, WO 03/075840). Examples of binding domains specific for Flt-1 include 6.12 (nucleotide and amino acids sequences of VH:
SEQ ID NOS: 11 and 12; nucleotide and amino acid sequences of VL: SEQ ID
NOS:13 and 14) and IMC-18F1 (nucleotide and amino acids sequences of VH: SEQ ID NOS:27 and 28;
nucleotide and amino acid sequences of VL: SEQ ID NOS:29 and 30).

[0051] Binding domains specific for EGFR include, for example, ERBITUX
(Cetuximab; IMC-C225) (nucleotide and amino acids sequences of VH: SEQ ID
NOS:15 and 16; nucleotide and amino acid sequences of VL: SEQ ID NOS:17 and 18) as disclosed in WO
96/40210 and IMC11F8 (nucleotide and amino acids sequences of VH: SEQ ID
NOS:19 and 20; nucleotide and amino acid sequences of VL: SEQ ID NOS:21 and 22). An example of a binding domain specific for IGFR is IMC-A12 (nucleotide and amino acids sequences of VH:
SEQ ID NOS: 23 and 24; nucleotide and amino acid sequences of VL: SEQ ID NOS:
25 and 26). Antibodies that bind to FGF receptors (see, WO 20051037235) include, for example, FRl-H7 (nucleotide and amino acids sequences of VH: SEQ ID NOS: 31 and 32;
nucleotide and amino acid sequences of VL: SEQ ID NOS: 33 and 34), FRl-A1 (nucleotide and amino acids sequences of VH: SEQ ID NOS: 35 and 36; nucleotide and amino acid sequences of VL:
SEQ ID NOS: 37 and 38), and FR1-4H (nucleotide and amino acids sequences of VH: SEQ
ID NOS: 39 and 40; nucleotide and amino acid sequences of VL: SEQ ID NOS: 41 and 42).
Antibodies that bind to RON or MSP-R (see, WO 2005/120557) include IMC-41A10 (nucleotide and amino acids sequences of VH: SEQ ID NOS: 43 and 44; nucleotide and amino acid sequences of VL: SEQ ID NOS: 45 and 46) and IMC-41B12 (nucleotide and amino acids sequences of VH: SEQ ID NOS: 47 and 48; nucleotide and amino acid sequences of VL: SEQ ID NOS: 49 and 50). Antibodies that bind to PDGFRa include, for example, 3G3 and 7G11 (Loizos et al., 2005, Mol. Cancer Ther. 4:369).

[0052] Further, portions of the above listed binding domains, such as the CDR
regions, may be incorporated into binding domains used to make the binding proteins described herein.

[0053] Certain preferred antibodies bind to two of the above listed receptors.
In an embodiment of the invention, a bispecific antigen-binding protein binds to and blocks activation of two different receptor tyrosine kinases involved in angiogenesis. In one such embodiment, the antibody binds to PDGFR and a VEGF receptor such as, for example, VEGFR2/Flk-1/KDR. In another such embodiment the antibody binds to KDR and FLT-1.

[0054] In another embodiment, an antibody of the invention binds to HER2 and EGFR. In yet another preferred embodiment, an antibody of the invention binds to EGFR
and IGFR.

[0055] In another embodiment, an antigen-binding protein of the invention binds to EGFR and a VEGFR. In a preferred embodiment, the VEGFR is VEGFR2. Such an antibody is useful for blocking stimulation of vascular epithelial cells, by blocking signal transduction through both EGFR and VEGFR. This is particularly useful where angiogenesis occurs in response to EGFR ligands, particularly TGFc4 secreted by tumor cells.

[0056] Antibodies of the invention can be used to cross-link antigens on target cells with antigens on immune system effector cells. This can be useful, for example, for promoting immune responses directed against cells which have a particular antigens of interest on the cell surface. According to the invention, immune system effector cells include antigen specific cells such as T cells which activate cellular immune responses and nonspecific cells such as macrophages, neutrophils and natural killer (NK) cells which mediate cellular immune responses.

[0057] Antibodies of the invention can have a binding site for any cell surface antigen of an immune system effector cell. Such cell surface antigens include, for example, cytokine and lymphokine receptors, Fc receptors, CD3, CD16, CD28, CD32 and CD64. In additions to antigen binding sites provided by sVDs, the bispecific antibodies can include binding sites provided by Fvs. Such Fvs may be obtained from antibodies to the aforementioned antigens, from combinatorial libraries, as well as by other methods known in the art.
The bispecific antibodies which are specific for cytokine and lymphokine receptors can also include binding sites comprising sequences of amino acids that correspond to all or part of the natural ligand for the receptor. For example, where the cell-surface antigen is an IL-2 receptor, an bispecific antibody of the invention can have an antigen-binding site which comprises a sequence of amino acids corresponding or IL-2. Other cytokines and lymphokines include, for example, interleukins such as interleukin-4 (IL-4) and interleukin-5 (IL-5), and colony-stimulating factors (CSFs) such as granulocyte-macrophage CSF (GM-CSF), and granulocyte CSF (G-CSF).

[0058] Antibodies of the invention are made by expressing two polypeptide chains, which taken together, comprise at least one single domain antigen binding site. The two polypeptide chains each comprise at least one heavy chain constant domain that is capable of dimerization (e.g., CH2 and/or CH3). Antibodies are conveniently produced in E. coli using DNA constructs which comprise bacterial secretion signal sequences at the start of each polypeptide chain. A variety of bacterial signal sequences are known in the art. A perferred signal sequence is from the pelB gene of Erwinia. carotovora. The DNA
fragments coding for the antibodies can be cloned, e.g., into vectors employing human cytomegalovirus (HCMV) promoters and enhancers for high level expression in mammalian cells, such as, for example, CHO, NSO, COS-7, and PER.C6 cells, and cell lines of lymphoid origin such as lymphoma, myeloma, or hybridoma cells. (See, e.g., Bendig, et al., U.S. Patent 5,840,299;
Maeda, et al. (1991) Hum. Antibod. Hybridomas 2, 124-34).

[0059] A selectable marker is a gene which encodes a protein necessary for the survival or growth of transformed host cells grown in a selective culture medium. Typical selectable markers encode proteins that (a) confer resistance to antibiotics or other toxins, e.g.
ampicillin, neomycin, methotrexate, or tetracycline, (b) complement auxotrophic deficiencies, or (c) supply critical nutrients not available from complex media, e.g. the gene encoding D-alanine racemase for Bacilli. A particularly useful selectable marker confers resistance to methotrexate. For example, cells transformed with the DHFR
selection gene are first identified by culturing all of the transformants in a culture medium that contains methotrexate (Mtx), a competitive antagonist of DHFR. An appropriate host cell when wild-type DHFR is employed is the Chinese hamster ovary (CHO) cell line deficient in DHFR activity, prepared and propagated as described by Urlaub and Chasin (1980) Proc.
Natl. Acad. Sci. USA 77, 4216. The transformed cells are then exposed to increased levels of methotrexate. This leads to the synthesis of multiple copies of the DHFR gene, and, concomitantly, multiple copies of other DNA comprising the expression vectors, such as the DNA encoding the antibody or antibody fragment.

[0060] Where it is desired to express a gene construct in yeast, an example of a suitable selection gene for use in yeast is the trpl gene present in the yeast plasmid YRp7.
Stinchcomb et al. (1979) Nature, 282, 39; Kingsman et al. (1979) Gene 7, 141.
The trpl gene provides a selection marker for a mutant strain of yeast lacking the ability to grow in tryptophan, for example, ATCC No. 44076 or PEP4-1. Jones (1977) Genetics 85, 12. The presence of the trpl lesion in the yeast host cell genome then provides an effective environment for detecting transformation by growth in the absence of tryptophan. Similarly, Leu2-deficient yeast strains (ATCC 20,622 or 38,626) are complemented by known plasmids bearing the Leu2 gene.

[0061] Transformed host cells are cultured by methods known in the art in a liquid medium containing assimilable sources of carbon, e.g. carbohydrates such as glucose or lactose, nitrogen, e.g. amino acids, peptides, proteins or their degradation products such as peptones, ammonium salts or the like, and inorganic salts, e.g. sulfates, phosphates and/or carbonates of sodium, potassium, magnesium and calcium. The medium furthermore contains, for example, growth-promoting substances, such as trace elements, for example iron, zinc, manganese and the like.

[0062] Antibodies that bind to growth factor receptors are preferably capable of blocking activation of receptor tyrosine kinase (RTK) activity. Tyrosine kinase inhibition can be determined using well-known methods, for example, by measuring the autophosphorylation level of recombinant kinase receptor, and/or phosphorylation of natural or synthetic substrates. Thus, phosphorylation assays are useful in determining RTK
antagonists of the present invention. Phosphorylation can be detected, for example, using an antibody specific for phosphotyrosine in an ELISA assay or on a western blot.
Some assays for tyrosine kinase activity are described in Panek et al., J. Pharmacol. Exp.
Thera. (1997) 283:1433-44 and Batley et al., Life Sci. (1998) 62:143-50.

[0063] In addition, methods for detection of protein expression can be utilized to determine RTK antagonists, wherein expression of the proteins being measured is mediated by the RTK. These methods include immunohistochemistry (IHC) for detection of protein expression, fluorescence in situ hybridization (FISH) for detection of gene amplification, competitive radioligand binding assays, solid matrix blotting techniques, such as Northem and Southem blots, reverse transcriptase polymerase chain reaction (RT-PCR) and ELISA.
See, e.g., Grandis et al., Cancer, (1996) 78:1284-92; Shimizu et al., Japan J.
Cancer Res., (1994) 85:567-71; Sauter et al., Am. J. Path., (1996) 148:1047-53; Collins, Glia, (1995) 15:289-96; Radinsky et al., Clin. Cancer Res., (1995) 1:19-31; Petrides et al., Cancer Res., (1990) 50:3934-39; Hoffmann et al., Anticancer Res., (1997) 17:4419-26;
Wikstrand et al., Cancer Res., (1995) 55:3140-48.

[0064] The ability of an antibody to block ligand binding can be measured, for example, by an in vitro competitive assay. In such an assay, a ligand of the RTK (e.g., EGF
for EGFR) is immobilized, and a binding assay is carried out to determine the effectiveness of the antibody to competitively inhibit binding of the RTK to the immobilized ligand.

[0065] In vivo assays can also be utilized to determine RTK antagonists. For example, receptor tyrosine kinase inhibition can be observed by mitogenic assays using cell lines stimulated with receptor ligand in the presence and absence of inhibitor. For example, A431 cells (American Type Culture Collection (ATCC), Rockville, MD) stimulated with EGF
can be used to assay EGFR inhibition. Another method involves testing for inhibition of growth of EGFR-expressing tumor cells, using for example, human tumor cells injected into a mouse. See U.S. Patent No. 6,365,157 (Rockwell et al.).

[0066] Preferred antibodies of the instant invention have dual specificity and are capable of binding to two different antigens simultaneously. The different antigens can be located on different cells or on the same cell. Cross-linking of antigen can be shown in vitro, for example by providing a solid surface to which a first antigen has been bound, adding a bispecific antibodies specific for the first antigen and a second antigen for which the binding protein is also specific and detecting the presence of bound second antigen.

[0067] Preferred antibodies of the invention are capable of blocking the interaction between two receptors and their respective ligands. For example, an antibody specific for KDR and Flt-1 inhibits VEGF induced cell migration as well as PIGF induced cell migration.
Combination of two receptor binding specificities in a bispecific antibodies can be more efficacious in inhibiting cell migration than the individual parent antibodies (see, e.g., Zhu, Z., WO 2004/003211).

[0068] Compared to antibodies that are monospecific, bispecific antibodies can be more potent inhibitors of cellular function. For example, VEGF-stimulated cellular functions such as, for example, proliferation of endothelial cells and VEGF- and P1GF-induced migration of human leukemia cells can be more efficiently inhibited by bispecific antibodies, even where affinity for one or both of the two target antigens is reduced. An antibody specific (monovalent) for both KDR and Flt-1 can be more effective to inhibit VEGF or P1GF
induced cell migration than a monospecific scFv directed at either of the target antigens (WO
2004/003211).

[0069] In another example, an antibody having dual specificity for both EGFR
(or Her2/neu) and IGFR that is capable of binding to both receptors and blocking interaction with their specific ligands is used to neutralizing both EFG- and IGF-stimulated receptor activation and downstream signal transduction. Stimulation of either EGFR or IGFR is observed to result in activation (e.g:, phosphorylation) of common downstream signal transduction molecules, including Akt and p44/42, although to different extents. In certain tumor cells, inhibition of EGFR function can be compensated by upregulation of other growth factor receptor signaling pathways, and particularly by IGFR
stimulation. In contrast to treatment with an antibody that binds to one receptor, and does not completely block phosphorylation of either Akt or p44/42, incubation of tumor cells with an antibody that binds to both EGFR and IGFR blocks phosphorylation of both Akt and p44/42.
Accordingly, inhibition of IGFR signaling results in inhibition of tumor growth and increased sensitivity of tumor cells to certain therapeutic agents.

[0070] Inhibition of phosphorylation of such common signal transduction cascade components is also observed with antibodies that bind to other RTKs, such as, for example, RON. Accordingly, the antigen-binding proteins are generally useful for treating neoplastic diseases characterized by cell growth or transformation resulting from activation of multiple signal transduction pathways.

[0071] The antibodies of the invention are useful for treatment of a variety of proliferative disorders. For example, the present invention provides for treatment of tumors that express and are stimulated through more than one receptor tyrosine kinase. Stimulation through more that one receptor can result in uncontrolled growth that is insensitive to blockage of each receptor alone. Alternatively, stimulation of a second receptor can add to the activation observed in response to stimulation through a first receptor.
Alternatively, the contributions from the individual receptors can be multiplicative. In each of the above instances, significantly improved inhibition of tumor growth is observed in the presence of an antigen-binding protein that.blocks both of the receptors.

[0072] The antibodies of the invention are useful for treating diseases in which receptor stimulation is through an EGFR paracrine and/or autocrine loop. For example, EGFR expressing tumors are characteristically sensitive to EGF present in their environment, and can further be stimulated by tumor produced EGF or TGF-a. While not iritending to be bound to any particular mechanism, the diseases and conditions that may be treated or prevented by the present methods include, for example, those in which tumor growth is stimulated. The method is therefore effective for treating a solid tumor that is not vascularized, or is not yet substantially vascularized.

[00731 Certain antibodies of the invention are useful for inhibiting angiogenesis associated with a hyperproliferative disease. For example, by blocking tumor associated angiogenesis, tumor growth may be inhibited. In one embodiment, the antibody binds to a tumor associated RTK and inhibits production of angiogenic ligands (i.e., VEGF) by the tumor, and also binds to a VEGF receptor associated with cells of the vasculature to inhibit proliferation of such cells. In a different embodiment, the antibody binds to multiple VEGF
receptors, such that VEGF or other ligand of VEGFR (e.g., P1GF) ligand is blocked from binding to more than one type of VEGF receptor.
[0074J Tumors that may be treated include primary tumors and metastatic tumors, as well as refractory tumors. Refractory tumors include tumors that fail to respond or are resistant to treatment with chemotherapeutic agents alone, antibodies alone, radiation alone or combinations thereof. Refractory tumors also encompass tumors that appear to be inhibited by treatment with such agents, but recur up to five years, sometimes up to ten years or longer after treatment is discontinued. The tumors may express EGFR or other RTK at normal levels or they may overexpress the RTK at levels, for example, that are at least 10, 100, or 1000 times normal levels.
[0075] Examples of tumors that express EGFR and are stimulated by a ligand of EGFR include carcinomas, gliomas, sarcomas, adenocarcinomas, adenosarcomas, and adenomas. Such tumors can occur in virtually all parts of the body, including, for example, breast, heart, lung, small intestine, colon, spleen, kidney, bladder, head and neck, ovary, prostate, brain, pancreas, skin, bone, bone marrow, blood, thymus, uterus, testicles, cervix or liver. Some tumors observed to overexpress EGFR that may be treated according to the present invention include, but are not limited to, colorectal and head and neck tumors, especially squamous cell carcinoma of the head and neck, brain tumors such as glioblastomas, and tumors of the lung, breast, pancreas, esophagus, bladder, kidney, ovary, cervix, and prostate. Non-limiting examples of tumors observed to have constitutively active (i.e., unregulated) receptor tyrosine kinase activity include gliomas, non-small-cell lung carcinomas, ovarian carcinomas and prostate carcinomas. Other examples of tumors include Kaposi's sarcoma, CNS neoplasms, neuroblastomas, capillary hemangioblastomas, meningiomas and cerebral metastases, melanoma, gastrointestinal and renal carcinomas and sarcomas, rhabdomyosarcoma, glioblastoma, preferably glioblastoma multiforme, and leiomyosarcoma. Overexpression of other RTKs can produce similar growth defects. For example, most metastatic bone cancers arise from primary tumors of prostate, breast, or lung.
Prostate tumors initially may be hormone dependent, but loss of such dependence coincides with IGFR mediated stimulation of cells that migrate to bone.
[0076] The antibodies are also useful for treating hyperproliferative disesases other than tumors comprising administering to the mammal an effective amount of the antibody of the present invention. As disclosed herein, "hyperproliferative disease" is defined as a condition caused by excessive growth of non-cancer cells that express a member of the EGFR
family or other tyrosine kinase receptors. The excess cells generated by a hyperproliferative disease express the RTK at normal levels or they may overexpress the RTK.

[0077] Examples of hyperproliferative disease include psoriasis, actinic keratoses, and seborrheic keratoses, warts, keloid scars, and eczema. Also included are hyperproliferative diseases caused by virus infections, such as papilloma virus infection. For example, psoriasis comes in many different variations and degrees of severity.
Different types of psoriasis display characteristics such as pus-like blisters (pustular psoriasis), severe sloughing of the skin (erythrodermic psoriasis), drop-like dots (guttae psoriasis) and smooth inflained lesions (inverse psoriasis). The treatment of all types of psoriasis (e. g., psoriasis vulgaris, psoriasis pustulosa, psoriasis erythrodermica, psoriasis arthropathica, parapsoriasis, palmoplantar pustulosis) is contemplated by the invention.
[0078] According to the invention, antibodies can be chemically or biosynthetically conjugated to other agents such as antineoplastic or anti-angiogenic agents for treatment of disease. Anti-tumor agents linked to an antibody include any agents which destroy or .
damage a tumor to which the antibody has bound or in the environment of the cell to which the antibody has bound. For example, an anti-tumor agent is a toxic agent such as a chemotherapeutic agent or a radioisotope. The chemotherapeutic agents are conjugated to the antibody using conventional methods (See, e.g., Hermentin and Seiler (1988) Behring Inst.
Mitt. 82, 197-215), including by peptide and non-peptide linkers.
[0079] Antibodies of the invention can also be linked to detectable signal-producing agents useful in vivo and in vitro for diagnostic purposes. The signal producing agent produces a measurable signal which is detectible by external means, usually the measurement of electromagnetic radiation. For the most part, the signal producing agent is an enzyme or chromophore, or emits light by fluorescence, phosphorescence or chemiluminescence.
Chromophores include dyes which absorb light in the ultraviolet or visible region, and can be substrates or degradation products of enzyme catalyzed reactions.
[0080] The invention further contemplates the use of antibodies with treatment or diagnostic agents incorporated into secondary reagents. For example, one member of a binding pair is linked to the antibody of the invention. Anti-neoplastic agents, for example, are conjugated to second members of such pairs and are thereby directed to the site where the antibody is bound. In a preferred embodiment, biotin is conjugated to an antibody of the invention, and thereby provides a target for an anti-neoplastic agent or other moiety which is conjugated to avidin or streptavidin. Alternatively, biotin or another such moiety is linked to an antibody of the invention and used as a reporter, for example in a diagnostic system where a detectable signal-producing agent is conjugated to avidin or streptavidin.
[0081] Antibodies can be administered in combination with one or more suitable adjuvants, such as, for example, cytokines (IL-10 and IL-13, for example) or other inunune stimulators, such as, but not limited to, chemokine, tumor-associated antigens, and peptides.
It should be appreciated, however, that administration of an antibody alone is sufficient to prevent, inhibit, or reduce the progression of the tumor in a therapeutically effective manner.
[0082] In certain embodiments, it can be desirable to administer an antibody of the invention that binds to an RTK and blocks ligand binding in combination with another antigen-binding protein that binds to ligand. Ligand binding antibodies are well known in the art, and include, e.g., anti-VEGF (Avastin ; bevacizumab).
[0083] The antibodies of the invention are also to be used in combined treatment methods by administration with an anti-neoplastic agent such as a chemotherapeutic agent or a radioisotope. Suitable chemotherapeutic agents are known to those skilled in the art and include irinotecan (CPT-11), anthracyclines (e.g. daunomycin and doxorubicin), methotrexate, vindesine, neocarzinostatin, cisplatin, chlorambucil, cytosine arabinoside, 5-fluorouridine, melphalan, ricin and calicheamicin. An antibody and an anti-angiogenic or anti-neoplastic agent are admininstered to a patient in amounts effective to inhibit angiogenesis and/or reduce tumor growth. The antibodies are also to be administered in combination with other treatment regimes, for example, with treatments such as radiation therapy. For examples of combination therapies, see, e.g., U.S. Patent No.
6,217,866 (Schlessinger et al.) (Anti-EGFR antibodies in combination with anti-neoplastic agents); WO
99/60023 (Waksal et al.) (Anti-EGFR antibodies in combination with radiation).
[00841 Any suitable anti-neoplastic agent can be used, such as a chemotherapeutic agent, radiation or combinations thereof. The anti-neoplastic agents known in the art or being evaluated can be grouped in to classes based on their target or mode of action. For example, alkylating agents include, but are not limited to, cisplatin, cyclophosphamide, melphalan, and dacarbazine. Examples of anti-metabolites include, but not limited to, doxorubicin, daunorubicin, and paclitaxel, gemcitabine, and topoisomerase inhibitors irinotecan (CPT-11), aminocamptothecin, camptothecin, DX-8951f, and topotecan (topoisomerase I) and etoposide (VP-16) and teniposide (VM-26) (topoisomerase II). For radiation, the source can be either external (external beam radiation therapy - EBRT) or internal (brachytherapy -BT) to the patient being treated. Such classifications can be useful for choosing an antineoplastic agent to use. For example, it has been observed that antibodies that bind IGFR may be particularly effective when administered with a topoisomerase inhibitor.

[0085] The dose of anti-neoplastic agent administered depends on numerous factors, including, for example, the type of agent, the type and severity tumor being treated and the route of administration of the agent. It should be emphasized, however, that the present invention is not limited to any particular dose.
[0086] In a combination therapy, the antibody is administered before, during, or after commencing therapy with another agent, as well as any combination thereof, i.e., before and during, before and after, during and after, or before, during and after commencing the anti-neoplastic agent therapy. For example, for treatment of a tumor or neoplastic disease, the antibody can be administered between 1 and 30 days, preferably 3 and 20 days, more preferably between 5 and 12 days before commencing radiation therapy. In a preferred embodiment of the invention, chemotherapy is administered concurrently with, prior to, or subsequent to antibody therapy.
[0087] In the present invention, any suitable method or route can be used to administer antibodies of the invention, and optionally, to co-administer anti-neoplastic agents, receptor antagonists, or other pharmaceutical composition. For example, anti-neoplastic agent regimens utilized according to the invention include any regimen believed to be optimally suitable for the treatment of a patient's neoplastic condition.
Different malignancies can requiie use of specific anti-tumor antibodies and specific anti-neoplastic agents, which will be determined on a patient to patient basis. Routes of administration include, for example, oral, intravenous, intraperitoneal, subcutaneous, or intramuscular administration. The dose of anti-neoplastic agent administered depends on numerous factors, including, for example, the type of neoplastic agent, the type and severity tumor being treated and the route of administration of the antineoplastic agent. It should be emphasized, however, that the present invention is not limited to any particular method or route of administration.
[0088] It is understood that antibodies of the invention, where used in a mammal for the purpose of prophylaxis or treatment, will be administered in the form of a composition additionally comprising a pharmaceutically acceptable carrier. Suitable pharmaceutically acceptable carriers include, for example, one or more of water, saline, phosphate buffered saline, dextrose, glycerol, ethanol and the like, as well as combinations thereof.
Pharmaceutically acceptable carriers can further comprise minor amounts of auxiliary substances such as wetting or emulsifying agents, preservatives or buffers, which enhance the shelf life or effectiveness of the binding proteins. The compositions of the injection can, as is well known in the art, be formulated so as to provide quick, sustained or delayed release of the active ingredient after administration to the mammal.
[0089] The present invention also includes kits for inhibiting tumor growth and/or angiogenesis, or treating other disesase, comprising a therapeutically effective amount of an antibody of the invention. Human or humanized antibodies are preferred. The kits can further contain any suitable antagonist of, for example, another growth factor receptor involved in tumorigenesis or angiogenesis (e.g., EGFR, VEGFR-1/Flt-1, VEGFR-2/Flk-1/KDR, IGFR, PDGFR, NGFR, FGFR, etc, as described above). Alternatively, or in addition, the kits of the present invention can further comprise an anti-neoplastic agent.
Examples of suitable anti-neoplastic agents in the context of the present invention have been described herein. The kits of the present invention can further comprise an adjuvant;
examples have also been described above.
[0090] Also included within the scope of the present invention is use of the present antibodies in vivo and in vitro for investigative or diagnostic methods, which are well known in the art. The diagnostic methods include kits which contain antibodies of the present invention.
[0091] Accordingly, the present receptor binding antibodies thus can be used in vivo and in vitro for investigative, diagnostic, prophylactic, or treatment methods, which are well known in the art. Of course, it is to be understood and expected that variations in the principles of invention herein disclosed can be made by one skilled in the art and it is intended that such modifications are to be included within the scope of the present invention.
All references mentioned herein are incorporated by reference in their entirety.
EXAMPLES
[0092] The following examples further illustrate the invention, but should not be construed to limit the scope of the invention in any way. Detailed descriptions of conventional methods, such as those employed in the construction of vectors and plasmids, the insertion of genes encoding polypeptides into such vectors and plasmids, the introduction of plasmids into host cells, and the expression and determination thereof of genes and gene products can be obtained from numerous publications, including Sambrook, J et al., (1989) Molecular Cloning: A Laboratory Manual, 2"a ed., Cold Spring Harbor Laboratory Press; and Coligan, J. et al. (1994) Current Protocols in Immunology, Wiley & Sons, Incorporated.
[0093] Selection of human anti-rnPDGFRa antibodies from phage display Fab library. A large naive human Fab phage display library from Dyax (containing 3.7x10'o clones) was used to select antibodies directed against mPDGFRcY Fc protein (R&D Systems (Minneapolis, MN). The library stock (100 l) was grown to log phase in 20 ml of 2YTAG
medium, rescued with M13K07 helper phage, and amplified overnight in 2YTAK
medium (2YT containing 100 iCg/ml of ampicillin and 50 glml of kanamycin) at 30 C.
Phage preparations were precipitated in 4% PEG, 0.5 M NaCl. Phage preparations were resuspended in 1 ml of 3% fat-free milk/PBS containing 240 g unrelated human IgG and incubated at 37 C for 1 h to block nonspecific binding and binding to the Fc-tag of the mPDGFRcx protein.
[0094] Three rounds of selection were performed on immobilized mPDGFR&--Fc with decreasing amounts of protein (50, 10 and 2 jig, respectively) coated on selection tubes. For each round of selection, mPDGFRa-Fc-coated Maxisorp Star tubes (Nunc, Rosklide, DenMark) were first blocked with 3% milk/PBS at 37 C for I h, and then incubated with the phage preparation at room temperature for I h. The tubes were washed 15 times with PBST
(PBS containing 0.1 % Tween 20) followed by 15 washes with PBS. The bound phage was eluted at room temperature for 10 min with lml of a freshly prepared solution of 100 mM
triethylamine (Sigma). The eluted phage were incubated with 10 ml of mid-log phase TGI
cells at 37 C for 30 min stationary and 30 min shaking. The infected TGI cells were pelleted, plated onto three 2YTAG plates, and incubated overnight at 30 C. All of the colonies grown on the plates were scraped into 3-5 ml of 2YTA medium, mixed with glycerol (final concentration: 10%), aliquoted, and stored at -80 C. For subsequent rounds of selection, the phage stock (100 1) from the previous round of selection was amplified and used for selection followed the procedure described above with decreased amount of mPDGFRa/Fc for coating the Maxisorp Star tubes.
[0095] To estimate input phage, 10 t of phage prior of selection was used to infect TG1 cells, and then titrated on 2YTAG plates. To estimate the phage recovery from selection, 10 1 of TG1 infected cell with eluted phage, from the above selection procedure, were titrated on 2YTAG plates. Recovery yields for each round of selection were calculated and increased in successive rounds (first round, 1 x 10"6%; second round 4x 10-6Ofo; third round, 2x 10-4%) despite that coating mPDGFRa/Fc was decreased.

[0096] ELISA screening for antibodies with binding and blocking activities.
After 2a and 3d rounds of selection, 190 clones from each round were randomly picked and tested for binding and blocking activities by both phage ELISA and Fab ELISA.
Briefly, individual TG1 clones recovered after 2 a and 3`d rounds of selection were randomly picked and grown at 37 C in 96-well plates. To produce phage, the cells were rescued with M13K07 helper phage as described above. To produce soluble Fab, the cells were incubated in medium containing 1mM of IPTG. For the binding ELISA, phage preparations and cell culture supematants containing soluble Fab were blocked with 1/6 volume of 18%
milk/PBS
at room temperature for 1 h. The blocked phage preparation or cell culture supematants were then added to 96-well microtiter plates (Nunc) coated with mPDGFRa/Fc (1 g/ml, 50 l, at 4 C overnight), and incubated at room temperature for 1 h. After incubation at room temperature for 1 h, the plates were washed 3 times with PBST.

[0097] For the phage ELISA, the plates were incubated with an anti-M13 phage antibody-HRP conjugate (Amersham Biosciences). For the Fab ELISA the plates were incubated with an anti-human-Fab antibody-HRP conjugate (Jackson TmznunoResearch Laboratory). After three washes, color was developed by addition of TMB
peroxidase substrate(KPL, Gaithersburg, MD), and the absorbance was measured at 450 nM
using a microplate reader (Molecular Device, Sunnyvale, CA). ELISAs for binding to 1C11 IgG
were also performed to eliminate Fc specific antibodies identified as binders in the library screen. More than 77% clones picked after 2"a selection, and 99% of the recovered clones after 3`a selection were positive in mPDGFRa binding assay, suggesting a high efficiency of the selection process.
[0098] For the blocking ELISA, 50 l of phage preparations or Fab culture supernatants were mixed with a fixed amount of mPDGFRcx Fc (0.5 Fcggfml) and incubated at RT for 30 min. The mixture was then transferred to 96-well plates precoated with rhPDGF-AA (0.5 g/ml; R&D Systems, Minneapolis, MN) and incubated at RT for 1 h. To quantify bound mPDGFRcx Fc protein, the plates were then incubated at RT for 1 h with a rabbit anti-human-Fc antibody-HRP conjugate, followed by three washes and addition of the TMS
peroxidase substrate. The bound mPDGFRae-Fc protein was quantified by reading the absorbance at 450 nM. Blocking activity was indicated by decreased ELISA
signal detected by anti-human-Fc antibody-HRP conjugate. About 4.2% of clones that bound xnPDGFRa showed PDGF-AA blocking activity.

[0099] Based on blocking assay results, 14 clones, including a non-blocking binder as a control clone (3F3), were initially selected for further study. Phagemid DNA
were isolated and DNA sequences were determined by dideoxynucleotide sequencing.
Classification and alignments of VH and VL genes were performed using the Bioinformatics web site of Andrew C.R. Martin's Bioinformatics Group (www.bioinf.org.uk) and MagAlign of DNAstar.
[0100] Analysis of selected anti-mPDGFRcx antibodies. Among 14 clones, 9 distinct antibodies were revealed (Table 1). Beside 1C10 and 1F2, no identical VHS or VLS were found. Interestingly, four of the antibodies with stronger blocking activity had incomplete light chains. Clones IF2 and 1F9 had in-frame VL deletions (leader peptide sequence followed by CL sequence). Clone IC10 and 1F2 shared the same VH gene. However, since the VL encoding sequence of 1 C 10 included a stop cordon such that CL was not expressed.
Clone 3G7 included a stop codon appeared at the 5' end of VL gene. Both 1C10 and 3G7 demonstrated strong blocking. Fab expression and binding activity appeared to be very low, but can be accounted for by poor binding of the anti-human-Fab secondary reagent.

Table 1- anti-mPDGFRca antibodies from a Fab phage display library Clone subgroup subgroup CL Binding Blocking Expres ion 1 C 10 I rclI - +I- ++ +/-1F2 I - + +++ +++ ++
1F9 I - + ++ ++ ++
3G7 I - - +/- ++ +/-1All 11 KII + +++ + ++
1E10 II xIII + ++ + ++
3132 III xIII + ++ + ++
3B9 III KIII + ++ + ++
3F3 III KIIT + +++ - +++
[0101] Nucleotide and amino acids sequences for the Fabs are as follows: 1E10 VH
domain: SEQ ID NOS: 51 and 52; 1E10 VL domain: SEQ ID NOS: 53 and 54; IA12 VH
domain: SEQ ID NOS: 55 and 56; 1A12 VL domain: SEQ ID NOS: 57 and 58; 3B2 VH
domain: SEQ ID NOS: 59 and 60; 3B2 VL domain: SEQ ID NOS: 61 and 62; 1C10 VH
domain: SEQ ID NOS: 63 and 64; 1C10 VL domain: SEQ ID NOS: 65 and 66; 3G7 VH
domain: SEQ ID NOS: 67 and 68; 3G7 VL domain nucleotide sequence: SEQ ID NO:
69;
3G7 VL domain: QAW; 1F9 VH domain: SEQ ID NOS: 70 and 71; 1F9 VL domain: SEQ
ID
NOS: 72 and 73; 1F2 VH domain: SEQ ID NOS: 74 and 75; 1F2 VL domain: SEQ ID
NOS:
76 and 77. Fig. 15 shows each of the domains (including the VL truncations or deletions of 3G7, 1F9, and IF2) and the locations of the CDR regions.
[01021 The Fab fragments of six.clones were expressed in E. coli HB2151 host cells, and purified by affinity chromatography Protein G column. Phagemids of the individual selected clones were used to transform a nonsupressor E.coli host HB2151.
Expression of the Fab fragments in HB2151 was induced by culturing the cells'in 2YTA medium containing 1 mM of IPTG at 30 C. A periplasmic extract of the cells was prepared as described by Lu lxl. The soluble Fab protein was purified using a Protein G column followi-ng the manufacturer's protocol (Amersham Biosciences). To examine the purity of the preparation and molecular weight, the purified antibodies were electrophoresed in NuPAGETM
4-12%
Bis-Tris gel (Invitrogen) and visualized by staining with the solution of SimplyBlueTM
SafeStain (Invitrogen).
[0103] The purified binding proteins were analyzed by SDS-PAGE (Fig. 2). Fabs with complete light chains appear in Fig. 2A with molecular weights of about 50,000. In contrast, unreduced fragments of 1F2 and 1F9 yielded small bands (-DTT: MW -37,500) as compared with a standard control Fab (Fig. 2B). Under reducing conditions (+DTT), two bands were evident, the upper band consistent with normal size of VH-CHl fragments, and the lower bands consistent with light chains having CL fragments (MW 12,500) but lackiing variable domains.
[0104] mPDGFRat binding and mPDGFRa/PDGF-AA blocking by the anti-mPDGFRa antibodies. Soluble Fab fragments of six anti-rnPDGFRa clones were compared quantitatively in their antigen binding efficiency and potency in blocking mPDGFRcx/PDGF-AA interaction. In the binding assay, the Fab was first incubated in a 96-well plate precoated with mPDGFRalFc fusion protein (1 ;ug/m1) at room temperature for I h, followed by incubation with a rabbit anti-anti-human-Fab antibody HRP conjugate for an additional hour.
The plate-bound antibody-HRP was then quantified by the addition of peroxidase substrate.
Consistent with the initial screening result (Table 1), 1F2 Fab binds much more efficiently to mPDGFRa than all other Fab proteins (Fig. 3A).
[0105] The binding kinetics of various anti-mPDGFRa antibodies were determined by surface plasmon resonance using a BlAcore 3000 biosensor and evaluated using the program BIA Evaluation 2.0 (Biacore, Inc., Uppsala, Sweden). The affinity constant, Kd, was calculated from the ratio of dissociation rate (koff)/association rate (kon). Reported values represent the mean =L S.E. from at least two determinations for Fabs and three determinations for IgG. (Table 2). Consistent with binding assays, sVD
antibody 1F2 has much higher affinity (Kd) of about 0.5 nM, more than 23-fold higher than the selected normal Fab, l E 10.

[0106] In the Table 2, 1F2-2H Fab is an engineered divalent Fab containing a second identical. VH domain expressed as a fusion to CL. The divalent Fab was expressed in E. coli and purified by Protein G chromatography. SDS-PAGE analysis of the purified 1F2-2H Fab demonstrated one single protein band of approximately 50 kD, similar to that of the standard Fab fragment (Fig. 5). The affinity of the divalent Fab is 79.5 pM, compared to 418 pM for the monovalent 1F2 Fab, indicating a 5.2-fold enhancement (Table 2).

Table 2 - Binding kinetics of anti-mPDGFRa antibodies to immobilized receptor Antibody kon (1/Ms) koff (1/s) Kd (M) 1F2 Fab 9.17=L4.15x105 3.13 0.76x10-4 4.18f2.07x10"1 1F9 Fab 1.75- 0.45x 106 3.20f0.4x 10-3 2.05 0.75x 10-9 IEIO Fab 1.75 0.05x105 2.05z1:0.55x10"3 1.22 0.29x10-8 IF2-2H Fab 9.35 7.65x106 1.40 0.50x10-4 7.95 0.55x10-"
IF2-CL 2.33 0.3Ix106 6.03~1.59x10"5 2.53f0.34x10-"
1F2-CH 1.50+-0.62x106 5.37zh1.65x10-5 3.93f0.81x10-"
IF2-2H 4.23=L0.05x105 1.67zL0.05x10' 3.90-4:0.82x10-10 1E10 1.37 -0.33x105 1.17d:0.62x10' 8.13f3.44x10-10 [0107] In the blocking assay, various amounts of antibodies were incubated with a fixed amount of mPDGFRa/Fc fusion in solution for 30 min, and the mixtures were then transferred to 96-well plates coated with rhPDGF-AA and incubated for 1 h.
Bound mPDGFRa was then quantified by quantifying bound anti-human-Fc-Ab. The IF2 Fab strongly blocked mPDGFRa/PDGF-AA interaction (Fig. 3B), with IC50 value of 12 nM, compared with 57, 140 and 220 nM for 1F9, 1E10 and 1A11, respectively. On anequal molar basis, the divalent Fab showed a moderately enhanced activity in both binding to mPDFGRa and in blocking the receptor/ligand interaction (Fig. 6).
[0108] Cloning, expression, and purication of full IgG. Fabs of selected clones were converted to full length IgGs by cloning the VH and VL domains into the heavy chain expression vector, pDFc, and light chain expression vector, pLCx respectively.
The pDFc has cloning sites Hind III and Nhe I between encoding genes of a leader peptide sequence and CH1. The pLCK has cloning sites Hind III and BsiWI between a leader peptide sequence and CL. Briefly, the genes encoding of VH and VL were amplified by reverse transcription-PCR
from phagmid DNAs and cloned into their respective expression vectors. The heavy chain and light chain expression vectors (VH-pDFc and VL-pLCx) were then co-transfected into COS-7 cells for transient expression as previously described. Cell culture media were collected at 48 and 96 hours after transfection, and pooled. Antibodies were purified from the supernatant of cell cultures by affinity chromatography affinity chromatography using Protein-A columns following the manufacture's protocol (Amersham Biosciences).
After function of IgG was confirmed, expressed pairs of VH and VL genes were cloned into a single expression vectors and transfected into COS-7 cells. Antibody purification was as described above.
[0109] Three IgG like constructs were produced from 1F2 Ig which contains only one VH (see Fig. 7), including tetravalent 1F2-2H IgG with four VHs (A), and divalent 1F2 with 1F2 VH expressed as a fusion either to CH or CL (B and C). The plasmids for expressing the light and heavy chains were used to co-transfect COS-7 cells for transient expression.
Expression levels of antibodies were monitored by ELISA using the supematants of cell cultures. Expression of 1F2-2H was consistently less than both divalent 1F2 antibodies, about one-fourth of divalent IF2 antibodies. The antibodies from the supernatant of cell cultures were purified by affinity chromatography using a Protein A column.

[0110] Clone 1E10, a blocker with normal antibody components, were also converted to a full length IgG. The VH and VL genes of 1E10 were first cloned into pDFc and pLCK, respectively. As above, after DNA sequencing and confirmation of 1E10 IgG
binding activity confirmation, the expressed VH / VL pair was cloned into a single expression vector which was then used to transfect COS-7 cells followed by antibody purification.
[0111] Purified antibodies were analyzed by SDS-PAGE to determine purity of the preparation and molecular weight. Divalent antibodies 1F2-CH/CL and 1F2-CL/CH, with MW -125,000, migrated fast than the normal IgG (1E10 and 2B4; left panel of Fig. 8).
When antibodies were treated with DTT prior the electrophoresis (right penal of Fig. 8) the heavy and light chains components were resolved. Antibody 1F2-CH/CL had a upper band correlated with a heavy chain similar to that of 1E10 and 2B4, and a fast migrating lower band with a mobility close to that anticipated (MW -12,500, CL only). Antibody CL/CH, on the other hand, showed the lower band represents a light chain with mobility similar to that of 1E10 and 2B4, and the upper band correlates with Cx1-Fc polypeptide only (MW-37,500).
[0112] Binding and blocking activities offull length anti-mPDGFRcY antibodies.
All three purified 1F2 variant antibodies, 1F2-2H IgG, 1F2-CHICL and 1F2-CL/CH, and antibody 1 El 0 were compared for their binding affinity to mPDGFRa by ELISA.
Various amount of antibodies were incubated in the ELISA plate coated with mPDGFRalFc, mPDGFRcx bound antibodies were then detected by an anti-human-x antibody HRP
conjugate. The ELISA result indicates that even though 1F2-2H IgG has four binding sites, the binding activity of 1F2-2H was -gx lower than both divalent 1F2 IgGs (1F2-CH/CL and 1F2-CL/CH) (Fig. 9A). On the other hand, both divalent 1F2 (1F2-CH/CL and 1F2-CL/CH) had very similar binding affinity showed by ELISA (Fig.9A) and BlAcore analysis (Table 2).
Affinity of divalent 1F2 IgG increased about one magnitude from the parental 1F2 Fab and 1 E 10 IgG increased 20 times from its Fab (Tables 2). Over all, the binding affinities of both divalent lF2 IgGs to mPDGFRa were about one magnitude higher than 1E10 IgG
(Fig. 9A
and Table 2).
[0113] Quantitative blocking assays were also performed to evaluate the anti-mPDGFRa antibodies. As described above, various amounts of antibodies were incubated with a fixed amount of mPDGFRa/Fc fusion in solution for 30 min, and the mixtures were then transferred to 96-well plates coated with rhPDGF-AA and incubated for I
h. Bound mPDGFRa was then quantified by quantifying bound anti-human-Fc-Ab. Consistent with the binding data, the divalent 1F2-CH/CL and 1F2-CL/CH were better blockers than 1F2-2H
IgG and 1E10 IgG: the IC50 values were 3.2, 2.7, 17, and 9.6 nM, for 1F2-CH/CL,1F2-CL/CH, 1F2-2H IgG and 1E10 IgG, respectively (Fig. 9B).
[0114] Bispecific antibodies. An anti-Flk-1 antibody, 2B4, was also isolated from the Dyax Fab phage library using the same procedure described above except that the Maxisorp Star tubes were coated with mVEGFR2-Fc fusion protein. The Fab of 2B4 was specific to Flk-1 and blocked Flk-1/VEGF165 interaction. VH and VL domains of Fab 2B4 are provided by SEQ ID NOS:79 and 81, respectively. Fig. 15 shows each of the domains and the locations of the CDR regions. The 2B4 Fab was converted to full IgGl as described above.
The binding kinetics of 2B4 Fab and IgG were determined by BIAcore analysis.
The binding affinities of 2B4 Fab and IgG to mVEGFR2 are 6.7 + 3.0 nM and 0.39 - 0.1 nM, respectively. 2B4 IgG blockes VEGFR2/VEGF165 interaction with an IC50 value of approximately 3.5 nM (Fig. 13B).
[0115] Construction of domain-based bispecific antibodies (anti-mPDGFRax anti-Flt-1)-Domain-based bispecific antibodies were constructed in two formats, scFv-format and Fab-format (Fig. 10, A and B). In the scFv-format, PCR fragments encoding the VH and VL genes of 2B4 were first assembled using overlapping PCR. The COOH terminus of 2B4 VL was linked to the NH2 terminus of 2B4 VH via a 5-amino acid linker (Ala-Ser-Thr-Lys-Gly, obtained from the N-terminal of a CL domain)(Fig. l OA). The resulting gene, encoding 2B4 VL-VH, was then cloned into vector pDFc via Hind III/Nhe I ligation for expression of scFv2B4-CH fusion (consisting of 2B4 VL - 2B4 VH - CH1 - CH2 - CH3). The scFv2B4-CH expressing vector was then paired with the 1F2VH-CL expressing vector as described above by co-transfection of COS-7 cells for transient expression.
[0116] In the Fab-format, the VH gene of 1F2 and the VL of 2B4 were first assembled using overlapping PCR. In this fusion, the COOH terminus of 1F2 VH was linked to the amino ternzinus of 2B4 VL via a 5-amino acid linker from 5' end of CH. The 1F2 VL encoding gene was then cloned into vector pLCK via Hind III/BsiWI sites for expression of 1F2 VH-2B4 VL - CL-fusion protein. The 2B4 VH gene was cloned into pDFc vector as described above for VH expression_ The plasmids for expressing light and heavy chains, paired as depicted in Fig. l OB, were used to co-transfect COS-7 cells for transient expression.
[0117] After the clones were tested by ELISA for dual binding activities, the Ig expression constructs were combined into a single expression vector which was then used transfect COS-7 cells for transient expression of designed IgG. Antibodies were purified from the supematant of cell cultures as described above.

[0118] Expression and purifccation of domain-based bispecific antibodies. The 1F2/scFv2B4 and 1F2-2B4 antibodies were produced by transfection and transient expression in COS-7 cells (100-200m1 cell cultures). Antibodies were purified from cell culture supernatants affinity chromatography using Protein A columns.
[0119] Purified bispecific antibodies (1F2/scFv2B4 and 1F2-2B4) and their parental antibodies 1F2 and 2B4 were analyzed by SDS-PAGE (Fig. 8). The lower mobility of untreated bispecific antibodies (-DTT) correlated with their high molecular weight (-175,000), as compared to 2B4 IgG, and divalent 1F2-CH/CL which migrated at the molecular weight position -125,000. The two components of each of the four antibodies were separately resolved after treatment with DTT. Comparing with a standard IgG (2B4), the lower band of bispecific antibody 1F2/scFv2B4 (the VH-CL fusion) is comparable to a standard IgG light chain. The upper band correlates with a scFv-heavy chain fusion protein (MW -62,500). On the other hand, the upper band of 1F2-2B4 corresponds to the heavy chain of a standard IgG (2B4) and the lower band is observed at a position expected for a VH-light chain fusion (MW -37,500).
[0120] Dual specificity of domain-based bispecific antibodies. In a cross-linking assay the bispecific 1F2-2B4 antibody and the monospecific 1F2-CH/CL and 2B4 antibodies were first incubated with mPDGFRa or mVEGFR2 in solution and then transferred to a microtiter plate coated with the second receptor (mVEGFR2 or mPDGFReY, respectively) followed by incubation with a biotin-labeled polyclonal antibody to the first receptor. Only the BsAb, but not the parent monospecific 2B4 IgG and 1F2-CH/CL, was able to cross-link the two target receptors.
[0121] The antigen binding efficiency of the domain-based bispecific antibodies was determined on immobilized mPDGFRa and Flk-1. ELISA showed that the bispecific antibody, IF2/scFv2B4 and 1F2-2B4 bound to mPDGFRa and Flk-1, but not as efficiently as the parental antibody 1F2 and 2B4 (Fig. 12). However, in both cases, the binding of 1F2-2B4 antibody was slightly more efficient compared with 1F2/scFv2B4. As expected, the Fik-1-specific antibody 2B4 did not bind to mPDGFRa (Fig. 12A), nor did the mPDGFRcx specific antibody 1 F2 bind to Flk-1 (Fig. 12B). The binding kinetics of the bispecific antibodies to mPDGFRa and Flk-1 were determined by surface plasmon resonance using a BlAcore instrument (Table 3). Consistent with the observations from ELISA, the 1F2-2B4 antibody has greater affinity to both mPDGFRa and Flk-1 in comparison with 1F2/scFv2B4 antibody.

Table 3- Binding kinetics of sVD-based bispecific antibodies Antibody Target kon (1/Ms) koff (i/s) Kd (M) mVEGFR2 2B4Fab 2.0~ 0.7E5 1.1=E 0_15E-3 6.7 3.0E-9 2B4 IgG 2.6 ~ 0.15E5 1.0 f 0_59E' 3.9 ~ 0.90E-10 1 F2/scFv2B4 1.6 t 0.10E5 5.9 ~ 0.15E' 4.0 ~ 0.45E-9 1 F2-2B4 1.6 ~ 0.40E5 2.4 ~ 0.10E' 1.5 :L 0.00E"9 mPDGFRa 1F2 Fab 9.2 ~ 4.15E5 3.1 ~ 0.76E-4 4.2 2.1E"10 1F2-CHICL 2.0 ~ 0.95E6 6.7 f 0.35E-5 4.5 2.35E-"

1F2/scFv2B4 5.1 + 2.6E5 9.3 .f 1.75E-5 2.2 - 0.75E-10 1F2-2B4 2.4 1.5E5 7.0 ~ 2.0E"5 4.1 + 1. 8E-t [0122] Fig. 13 shows that both bispecific antibodies inhibit mPDGFRa from binding to irnrnobilized PDGF-AA (Fig. 13A) with estimated IC50 of 9.9 nM and 25.3 nM
for 1F2/scFv2B4 and 1F2-2B4, respectively. The antibodies also block Flk-I from binding to immobilized VEGF165 (Fig. 13B) with IC50 value of 9.5 nM and 19.5 nM, respectively. As expected, 2B4 had no effects on mPDGFRadPDGF-AA interaction, whereas 1F2 had no effects on Flk-1/VEGF165 interaction.
[0123] Although the binding affinity of the bispecific antibody is lower in mPDGFa binding and mVEGFR2 binding compared to respective monospecific bivalent molecules, 1F2-CH/CL and 2B4 IgG (Table 3), when examined for binding to cell surface-expressed receptors on eEnd.1 cells (the cells that express both mPDGFRa and mVEGFR2), the BsAb demonstrated higher efficiency than either of the parent monospecific bivalent antibodies.
The mean fluorescence intensity (MFI) on the cells were 15.7, 31 and 36.9, for (mPDGFRa binding), 2B4 IgG (mVEGFR2 binding) and 1F2-2B4IgG (mPDGFRa and mVEGFR2 binding), respectively.

[0124] Inhibition of ligand-induced activation of mWEGFRZ and mPDGFR.
Expression of mVEGFR2 and mPDGFRa was demonstrated in eEnd.I cells via FACS
analysis. The cells were incubated with 2B4 IgG, 1F2-CH/CL or 1F2-2B4IgG (10 g/ml) at 4 C for 1 h, followed by incubation with a PE-labeled goat anti-human Fc antibody (Jackson ImmunoResearch Lab.) for an additional hour, and analysis on a Guava Easycyte System (Guava Technologies, Inc. Hayward, CA).
[0125] To examine receptor phosphorylation, eEnd.l cells were plated onto 6 cm dishes and grown to 70-80% confluence, after which the cells were washed twice in PBS and cultured overnight in serum free medium. The cells were first incubated with various antibodies at 37 C for 30 min, followed by stimulation with VEGF or PDGF-AA at 37 C for 15 min. The cells were lysed in lysis buffer (50 mM Tris-HCI, pH 7.4, 150 rnM
NaCI, 1%
TritonX-100, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 0.5 mM Na3VO4,1 g/ml leupeptin, 1 g/ml pepstatin, and 1 g/ml aprotinin) for 1 h, followed by centrifugation of the lysate at 12, 000 rpm for 10 min at 4 C. Receptors were immunoprecipitated from the cell lysate supernatant using anti-mPDGFRa (eBioscience, San Diego, CA) or anti-mVEGFR2 antibody (Santa Cruz Biotech, Santa Cruz, CA), followed by the addition of 20 l of ProA/G-sepharose beads (Santa Cruz Biotech). The precipitated receptor proteins were resolved on a 4-12% Nupage Bis-Tris gel (Invitrogen) and transferred to a polyvinylidene difluoride membrane. Phospho-mVEGFR2 and phospho-mPDGFR.a were detected on the blot using an anti-phospho-tyrosine antibody-HHRP conjugate (Santa Cruz Biotech). Total receptor proteins loaded on the gel were assayed with antibodies to mPDGFRoc or mVEGFR2 (both from Santa Cruz Biotech).
[0126] As shown in Fig. 14, the BsAb inhibited both PDGF and VEGF-stimulated phosphorylation of mPDGFRa and mVEGFR2 receptors, whereas monospecific parent antibodies only blocked the activation of a single receptor stimulated by its cognate ligand.
As a control, the anti-EGFR antibody, C225, did not have any effect on ligand-stimulated activation of either receptor.

Claims (37)

1. An antigen-binding protein comprising a complex of two first polypeptides and two second polypeptides, each first polypeptide having a first antigen-binding site located to the N
terminus of an immunoglobulin light chain constant domain (C L domain), said C
L
domain capable of stable association with an immunoglobulin heavy chain first constant domain (C H1domain), each second polypeptide having a second antigen-binding site located to the N
terminus of a C H1 domain, said C H1 domain followed by one or more heavy chain constant domains capable of stable self-association;
wherein at least one of said first antigen-binding site and said second antigen-binding site is a single variable domain (sVD).

2. The antigen binding protein of claim 1, wherein both of said first and second antigen binding sites are single variable domains (sVDs).

3. The antigen binding protein of claim 1, wherein one of said first and second antigen binding sites is a single variable domain (sVD) and the other of said first and second antigen binding sites is a single chain Fv (scFv).

4. An antigen-binding protein comprising a complex of two first polypeptides and two second polypeptides;
each first polypeptide comprising an immunoglobulin heavy chain composed of a variable domain and constant domains;
each second polypeptide comprising an immunoglobulin light chain composed of a variable domain and a constant domain;
wherein the two first polypeptides and two second polypeptides stably associate to form an immunoglobulin-like molecule;
wherein the variable domains of the immunoglobulin heavy chains stably associate with the variable domains of the immunoglobulin light chains to form first antigen binding sites; and wherein either or both of said first and second polypeptides further comprise a second single chain variable domain (sVD) antigen binding site at the N
terminus or C
terminus.

5. The antigen-binding protein of any one of claims 1 to 4, wherein said first antigen-binding sites and said second antigen-binding sites have different specificities.

6. The antigen-binding protein of claim 5, wherein said different specificities are for epitopes which reside on different antigens.

7. The antigen-binding protein of claim 5, wherein in said different specificities are for epitopes which reside on the same antigen.

8. The antigen-binding protein of any one of claims 1 and 2, wherein said first antigen-binding sites of said second antigen binding sites have the same specificity.

9. The antigen-binding protein of any one of claims 1 to 4, wherein said constant domains are human constant domains.

10. The antigen-binding protein of any one of claims 1 to 4, wherein said single variable domain is human.

11. The antigen-binding protein of any one of claims 1 to 4, which binds to an Fe receptor.

12. The antigen-binding protein of any one of claims 1 to 4, which effects complement dependent cytotoxicity (CDC).

13. The antigen-binding protein of any one of claims 1 to 4, which effects antibody dependent cell-mediated cytotoxicity (ADCC).

14. The antigen-binding protein of any one of claims 1 to 4, which is linked to an anti-tumor agent.

15. The antigen-binding protein of any one of claims 1 to 4, which is linked to a detectable signal producing agent.

16. The antigen-binding protein of any one of claims 1 to 4, wherein at least one of said first and second antigen-binding sites is specific for a receptor tyrosine kinase (RTK).

17. The antigen-binding protein of claim 16, which blocks binding of ligand to the receptor tyrosine kinase.

18. The antigen-binding protein of claim 16, which neutralizes activation of the receptor tyrosine kinase.

19. The antigen-binding protein of claim 16, which inhibits signal transduction by the receptor tyrosine kinase.

20. The antigen binding protein of claim 16, wherein the receptor tyrosine kinase is human.

21. The antigen binding protein of claim 16, wherein the receptor tyrosine kinase is selected from the group consisting of PDGFR.alpha., VEGFR1, VEGFR2, VEGFR3, EGFR, HER2, IGFR, FGFR, NGFR, RON, Tek, and Tie2.

22. The antigen binding protein of claim 16, wherein one of said first and second antigen binding sites is specific for PDGFR.alpha. and the other of said first and second antigen binding sites is specific for VEGFR2.

23. The antigen binding protein of claim 16, wherein one of said first and second antigen binding sites is specific for VEGFR1 and the other of said first and second antigen binding sites is specific for VEGFR2.

24. The antigen binding protein of claim 16, wherein one of said first and second antigen binding sites is specific for IGFR and the other of said first and second antigen binding sites is specific for EGFR or HER2.

25. A method of neutralizing the activation of a receptor tyrosine kinase, which comprises treating a cell with an antigen-binding protein of any one of claims 1 to 4 specific for said receptor tyrosine kinase in an amount sufficient to neutralize activation of the receptor.

26. A method of inhibiting angiogenesis which comprises treating a mammal with an antigen-binding protein of any one of claims 1 to 4, in an amount sufficient to neutralize activation of the receptor.

27. A method of reducing tumor growth which comprises treating a mammal with an antigen-binding protein of any one of claims 1 to 4, in an amount sufficient to reduce tumor growth.

28. The method of any one of claims 26 and 27, wherein the antigen binding protein binds to PDGFR.alpha. and to VEGFR2 and inhibits ligand induced activation of PDGFR.alpha.
and ligand induced activation of VEGFR2.

29. The method of any one of claims 26 and 27, wherein the antigen binding protein binds to VEGFR1 and to VEGFR2 and inhibits ligand induced activation of VEGFR1 and ligand induced activation of VEGFR2.

30. The method of claim 27, wherein the antigen binding protein binds to EGFR
and to IGFR and inhibits ligand induced activation of EGFR and ligand induced activation of IGFR.

31. The method of claim 27, wherein the antigen binding protein binds to HER2 and to IGFR and inhibits ligand induced activation of HER2 and ligand induced activation of IGFR.

32. The method of claim 27, which further comprises administering an effective amount of an anti-neoplastic agent.

33. A method for making an antigen-binding protein which comprises a) coexpressing in a host cell a recombinant DNA construct encoding a first polypeptide having a first antigen-binding site located to the N terminus of an immunoglobulin light chain constant domain (C L domain), said C L domain capable of stable association with an immunoglobulin heavy chain first constant domain (C H1domain), and a recombinant DNA construct encoding a second polypeptide having a second antigen-binding site located to the N terminus of a CH1 domain, said CH1 domain followed by one or more heavy chain constant domains capable of stable self-association;
wherein at least one of said first antigen-binding site and said second antigen-binding site is a single variable domain (sVD);
for a time and in a manner sufficient to allow expression of the polypeptides and formation of the antibody; and b) recovering the antigen-binding protein.

34. The method of claim 33, wherein the constructs are on the same DNA
expression vector.

35. The method of claim 33, wherein the constructs are on different DNA
expression vectors.

36. The method of claim 33, wherein the host cell is a bacterial cell, a yeast cell or a mammalian cell.

37. The method of claim 33, wherein the antibody is secreted from the host cell.