Multispecific antigen-binding molecules with improved internalization properties

Abstract

The present invention relates to a multispecific antigen-binding molecule comprising a first antigen-binding domain and a second antigen-binding domain, wherein the first domain specifically binds to a target molecule (T), and the second domain specifically binds to an internalizing effector protein (E), and the second antigen-binding domain binds to E with a dissociation constant (K D )10 -9 ~10 -8 M. The multispecific antigen-binding molecule is useful in methods for treating and / or preventing cancer.

Description

[Technical field]

[0001] FIELD OF THEINVENTION The present invention relates to multispecific antigen-binding molecules, compositions comprising said multispecific antigen-binding molecules, and the use of said multispecific antigen-binding molecules in the treatment of disease. [Background technology]

[0002] 2. Background of the Invention Since the development of the first monoclonal antibodies, research has been directed towards further optimization of antibodies for human therapy. The first monoclonal antibodies were of mouse and rat origin. Developments in antibody technology have led to the availability of humanized and human antibodies with reduced immunogenicity risk profiles. Genetic and chemical modifications have led to the development of more potent antibodies with increased therapeutic potential. This includes antibodies with optimized Fc-mediated effector functions, or with optimized binding properties, or with optimized antitumor activity through conjugation with toxic molecules (antibody drug conjugates).

[0003] Antibody drug conjugates (ADCs) have emerged as powerful therapeutic agents for the treatment of cancer, since they combine antibody-mediated tumor targeting with the cytotoxic activity of toxins. ADCs contain antibodies (e.g., monoclonal antibodies, single-chain variable fragments [scFv], or bispecific antibodies) linked to cytotoxic payloads or drugs. The advantage of directing cytotoxic drugs to tumors by antibodies against tumor-associated antigens is that the therapeutic window can be improved compared to unconjugated cytotoxic drugs. This allows the application of cytotoxic payloads with increased efficacy.

[0004] Currently, two ADCs have already been approved for therapeutic use: brentuximab vedotin (Adcetris) for the treatment of recurrent Hodgkin's lymphoma and recurrent sALCL, and trastuzumab emtansine (Kadcyla) for the treatment of HER2-positive metastatic breast cancer patients who have previously received trastuzumab and taxanes, either separately or in combination. In addition, more than 50 different ADCs are currently under clinical evaluation. In many cases, ADCs rely on the internalization of toxin-conjugated antibody molecules into target cells to release their payload and induce subsequent cytotoxicity. Most ADCs in clinical development are designed to be stable in the circulation and release their cytotoxic payload after internalization and lysosomal processing of the antigen / ADC complex.

[0005] However, the requirement for antigen- and antibody-mediated internalization limits the number of suitable ADC targets. Many tumor-associated antigens do not internalize well or traffic to lysosomes well, and therefore are not promising candidate targets for ADC-based therapeutics. Also, in many cases, intracellular processing of ADCs is inefficient. After internalization, receptors such as transferrin, HER2, cell adhesion molecule L1, and integrins are continuously recycled from endosomal compartments back to the cell membrane. Thus, to achieve maximum killing activity by ADCs, high antigen turnover, efficient trafficking to lysosomes, and highly toxic payloads are required.

[0006] Methods for enhancing ADC internalization, lysosomal targeting, and intratumoral and intracellular processing can be used to enhance the tumor cell killing activity of ADC. One approach to optimize ADC activity is by selecting specific epitopes on tumor-associated targets, since the specific epitopes recognized can affect internalization and lysosomal trafficking. For example, it has been previously shown that the efficacy of HER2-ADC can be improved by selecting HER2-ADCs that allow enhanced internalization by loading HER2 onto other ErbB molecules via heterodimerization. This provides an attractive strategy to increase the delivery and tumor cell killing ability of ADCs to both high and low HER2 expressing tumor cells.

[0007] In general, efficient internalization of ADC and subsequent translocation to lysosomes where proteolysis can occur is preferred.However, for many cell surface proteins and carbohydrate structures on tumor cells, the magnitude of these processes is insufficient to allow sufficiently potent cell killing by ADC.

[0008] International Patent Application WO2013 / 138400 describes a multispecific antibody having a first domain that specifically binds to a target antigen, such as IL-4R or SOST, and a second domain that specifically binds to an internalizing effector protein. When the target antigen is a tumor-associated antigen, binding of the tumor-associated antigen and the internalizing effector protein by the multispecific antibody facilitates targeted killing of tumor cells.

[0009] It is an object of the present invention to provide bispecific or multispecific antigen-binding molecules that have increased internalization capabilities compared to monospecific antigen-binding molecules against tumor-associated targets.

[0010] It is another object of the present invention to provide bispecific or multispecific antibody drug conjugate molecules that have increased internalization capabilities compared to monospecific antibody drug conjugate molecules against tumor-associated targets.

[0011] It is another object of the present invention to provide ADCs that have enhanced internalization, lysosomal targeting, and / or intracellular processing in tumor cells.

[0012] It is another object of the present invention to provide ADCs that have increased cytotoxicity against tumor cells and / or fewer side effects.

[0013] It is another object of the present invention to provide ADCs with an increased therapeutic window.

[0014] It is another object of the present invention to enable the generation of effective ADCs against tumor-associated antigens that are poorly internalized or poorly trafficked to lysosomes. [Prior art documents] [Patent documents]

[0015] [Patent Document 1] WO2013 / 138400 Summary of the Invention

[0016] In a first aspect, the present invention relates to a multispecific antigen-binding molecule comprising a first antigen-binding domain and a second antigen-binding domain, wherein the first domain specifically binds to a target molecule (T) and the second domain specifically binds to an internalizing effector protein (E), and the second antigen-binding domain has a dissociation constant K D 10 -9 ~10 -8 Has M.

[0017] The present inventors have found that the specific affinity range of the second antigen-binding domain confers surprisingly beneficial properties to the multispecific antigen-binding molecules of the present invention. The second antigen-binding domain is found to readily induce internalization of the multispecific molecule and induce cytotoxicity of the drug-conjugated multispecific molecule within the specific binding affinity range. This is particularly true when the first domain specifically binds to a tumor-associated antigen, such as HER2. At the same time, the second antigen-binding domain exerts only limited internalization and cytotoxicity (when used in the context of ADC) within the specific binding affinity range in the absence of binding of the first binding domain, i.e., in the absence of tumor-associated target (T), it shows surprisingly low cytotoxicity in cells expressing an internalizing effector protein (E). Thus, the multispecific antigen-binding molecules of the present invention show binding, internalization, lysosomal translocation, and toxin release in tumor cells, with minimal internalization, lysosomal translocation, and toxin release in non-tumor cells.

[0018] It has also been found by the inventors that the multispecific antigen binding molecules of the invention allow the accessibility of tumor antigens that are generally not or poorly internalized, thereby greatly enhancing the pool of potential ADC targets.

[0019] In another aspect, the present invention relates to a multispecific antigen-binding molecule, which is conjugated to a cytotoxic moiety, a radioisotope, a drug, a cytokine, or an RNA silencing vehicle. In another aspect, the present invention relates to the use of a multispecific antigen-binding molecule in a method for treating and / or preventing cancer, and to the use of a multispecific antigen-binding molecule in a method for targeting a tumor in a subject, the method for targeting a tumor comprising administering a multispecific antibody or ADC to the subject.

[0020] In other aspects, the present invention relates to pharmaceutical compositions comprising the multispecific antigen-binding molecules as active ingredients, nucleic acids encoding the multispecific antigen-binding molecules, expression vectors containing the nucleic acids and capable of expressing the nucleic acids in a single or multiple prokaryotic or eukaryotic host cell lines, as appropriate, and prokaryotic or eukaryotic host cell lines comprising the vectors.

[0021] definition Binding, affinity, and K D The term "binding" as used herein means, for example, about 10 -8 M or less K D "Avidity" refers to the binding of an antigen-binding molecule, such as an antibody, to a predetermined antigen or target with a binding affinity corresponding to a given value.

[0022] Those skilled in the art will recognize the affinity and equilibrium dissociation constants, K D You are probably familiar with the concept of the dissociation constant, K D can be measured by the biolayer.

[0023] K D Values ​​can be determined by biolayer interferometry (BLI) on an Octet HTX machine using an antigen-binding molecule, e.g., an antibody, as the immobilized ligand and the antigen as the analyte.

[0024] K D (M) refers to the dissociation equilibrium constant of a particular interaction between a multispecific antigen-binding molecule and an antigen, preferably an interaction between a single binding arm of an antibody molecule and an antigen, k d k a can be obtained by dividing by "k d ”(seconds -1 The term k ) as used herein refers to the dissociation rate constant of a particular interaction between a multispecific antigen-binding molecule and an antigen. This value is k dis Value, k off Also called the k value or off rate. a " " -1 × seconds-1 The term k ) as used herein refers to the association rate constant of a particular interaction between a multispecific antigen-binding molecule and an antigen. This value is k on Also called the on-speed or on-value.

[0025] antibody As used herein, the term "antibody" (Ab) in the context of the present invention refers to an immunoglobulin molecule, a fragment of an immunoglobulin molecule, or any derivative thereof, that has the ability to specifically bind to an antigen under typical physiological conditions for a suitable, functionally defined period of time, and preferably dually binds to two different antigens, as in the case of bispecific antibodies, in order to induce, promote, enhance and / or modulate physiological responses associated with antibody binding to the antigen.

[0026] The variable regions of an immunoglobulin molecule (either heavy and / or light chains or only heavy chains) contain the binding domains that interact with antigens. The constant region of an antibody (Ab) can mediate the binding of the immunoglobulin to FcRn. Antibodies include, but are not limited to, DuoBody molecules, tandem scFv, tandem scFv-Fc, knob-into-hole IgG, scFv-Fc knob-into-hole, scFv-Fc-scFv, F(ab')2, Fab-scFv, (Fab'scFv)2, diabody, sc diabody, sc diabody-Fc, or sc diabody-CH3, Triomab, kih The antibodies may be bispecific or multispecific, such as IgG common LC, CrossMab, DVD-Ig, 2-in-1-IgG, IgG-scFv, bi-Nanobody, BiTE, TandAb, DART, DART-Fc, scFv-HSA-scFv, orthoFab-IgG, tetravalent Tv-IgG, dock-and-lock (DNL) formats such as DNL-Fab3, and Azymetric scaffolds, or similar molecules.

[0027] As stated above, the term antibody, as used herein, unless otherwise stated or clearly contradicted by the context, includes fragments of antibodies that are antigen-binding fragments, i.e. that retain the ability to specifically bind to an antigen. It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of antigen-binding fragments encompassed by the term "antibody" include (i) a Fab' fragment or a monovalent fragment consisting of the VL, VH, CL and CH1 domains, or a monovalent antibody (Genmab) as described in WO2007059782; (ii) an F(ab')2 fragment, which is a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) an Fd fragment, which consists essentially of the VH and CH1 domains; (iv) an Fv fragment, which consists essentially of the VL and VH domains of a single domain of an antibody, (v) a dAb fragment, which consists essentially of the VH domain and is also called a domain antibody; (vi) a camelid or nanobody, and (vii) an isolated complementarity determining region (CDR). Furthermore, the two domains VL and VH of the Fv fragment are encoded by separate genes, but can be joined using recombinant methods by a synthetic linker that allows the VL and VH regions to be made as a single protein chain paired to form a monovalent molecule (also known as a single chain antibody or single chain Fv (scFv)). Such single chain antibodies are encompassed by the term antibody, unless stated otherwise or clearly indicated by the context. These and other useful antibody fragments in the context of the present invention, as well as bispecific formats of such fragments, are further described herein. Unless otherwise indicated, it should also be understood that the term antibody includes polyclonal antibodies, monoclonal antibodies (mAbs), antibody-like polypeptides, chimeric antibodies, humanized antibodies, fully human antibodies, and antibody fragments that retain the ability to specifically bind to an antigen (antigen-binding fragments) provided by known techniques such as enzymatic cleavage, peptide synthesis, and recombinant techniques. The antibodies generated may possess any isotype.

[0028] As used herein, "isotype" refers to the immunoglobulin (sub)class (e.g., IgG1, IgG2, IgG3, IgG4, IgD, IgA, IgE, or IgM) that is encoded by heavy chain constant region genes.

[0029] The term "monovalent antibody" means in the context of the present invention that an antibody molecule is capable of binding only one antigen molecule.

[0030] The term "bivalent antibody" means in the context of the present invention that an antibody molecule contains two binding domains for a specific antigen and is therefore capable of binding to one or two molecules of that antigen.

[0031] Generation of multispecific antibodies Multispecific antibodies, in particular bispecific antibodies, of the present invention can be generated through controlled Fab-arm exchange (FAE) as described in Labrijn et al., Efficient generation of stable bispecific IgG1 by controlled Fab-arm exchange, PNAS, vol.110, no.13, pp.5145-5150, March 2013, WO 2011 / 131746 A2, or Labrijn et al. Controlled Fab-arm exchange for the generation of stable bispecific IgG1, Nat Protoc, 2014 Oct;9(10):2450-63, also known as DuoBody® technology. Briefly, in this in vitro method, two different antibodies are provided, both comprising an immunoglobulin Fc region together with a CH3 region, where the sequence of each CH3 region is different and contains corresponding mutations. As a result, the heterodimeric interaction between the first and second CH3 regions is stronger than the homodimeric interaction of each of the first and second CH3 regions. The antibodies are incubated under reducing conditions to allow the cysteines in the hinge regions to undergo disulfide bond isomerization and obtain the bispecific antibody by regulated Fab arm exchange. The reducing agent is then removed from the mixture (containing the bispecific antibody) to allow oxidation of the disulfide bonds. The sequences of the CH3 regions contain corresponding mutations, i.e., mutations at different positions of the two CH3 regions, preferably a mutation at position 405 of the CH3 region of one IgG1 molecule and a mutation at position 409 of the CH3 region of the other IgG1 molecule.Multispecific antibodies may be generated using other technologies and formats such as, but not limited to, tandem scFv, tandem scFv-Fc, knob-into-hole IgG, scFv-Fc knob-into-hole, scFv-Fc-scFv, F(ab')2, Fab-scFv, (Fab'scFv)2, diabody, sc diabody, sc diabody-Fc, or sc diabody-CH3, triomab, kih IgG common LC, CrossMab, DVD-Ig, 2-in-1-IgG, IgG-scFv, bi-nanobody, BiTE, TandAb, DART, DART-Fc, scFv-HSA-scFv, orthoFab-IgG, tetravalent Tv-IgG, dock-and-lock (DNL) formats such as DNL-Fab3, and azimetric scaffolds, or similar molecules.

[0032] Induction of internalization of multispecific antigen-binding molecules Binding of both T and E by a multispecific antigen-binding molecule preferably induces internalization of the multispecific antigen-binding molecule of the invention to a greater extent than binding to the target T alone, and also induces cytotoxicity of the drug-conjugated multispecific antibody of the invention. For example, internalization induced by binding, preferably simultaneous binding, of T and E by a multispecific antigen-binding molecule can be greater than 10%, e.g., greater than 20%, greater than 30%, greater than 40%, greater than 50%, greater than 60%, greater than 70%, greater than 80%, greater than 90%, greater than 100%, greater than 110%, greater than 150%, greater than 200%, greater than 300%, greater than 400%, greater than 500%, greater than 750%, greater than 1000%, greater than 2000%, or greater than 5000% higher than the level of internalization measured in the presence of a control construct that contains only binding to T and not to E.

[0033] Non-limiting examples of determining whether the multispecific antigen-binding molecule of the present invention enhances internalization of the multispecific antigen-binding molecule to a greater extent than binding of the target molecule (T) by the first domain alone are shown in the colocalization assays of Examples 5 and 8 or the HER2 downmodulation assay of Example 9 described below. In these examples, T is HER2 and E is CD63. With regard to the concept of enhanced internalization, in Examples 13 and 14, enhanced internalization of a multispecific antibody is demonstrated using integrin β1 (CD29) as the target (T) and CD63 as the internalizing effector protein (E) that induces enhanced internalization.

[0034] The inventors have found that the specific affinity range of the second antigen-binding domain confers surprisingly beneficial properties to the bispecific antigen-binding molecules of the invention, particularly when E is CD63. In particular, when the first domain specifically binds to a tumor-associated antigen such as HER2, the monovalent binding of the antigen-binding domain to E has a dissociation constant K D 10 -9 ~10 -8 Within the specific affinity range of M, the second antigen-binding domain readily induces internalization of the multispecific antigen-binding molecule and cytotoxicity of the drug-conjugated multispecific antibody. Furthermore, when only E is present and no tumor-associated target antigen (T) is present, the second antigen-binding domain exerts only a limited contribution to the internalization of the multispecific antigen-binding molecule and cytotoxicity of the drug-conjugated multispecific antibody within this specific affinity range, i.e., in the absence of the tumor target (T), it exhibits surprisingly low or no cytotoxicity in cells expressing the internalizing effector protein (E). Thus, the multispecific antigen-binding molecule of the present invention exhibits binding, internalization, and lysosomal accumulation in tumor cells with minimal internalization in non-tumor cells.

[0035] CD63 Cluster of differentiation 63 (CD63, Uniprot ID P08962) molecule is also known as lysosomal associated membrane glycoprotein 3 (LAMP-3). CD63 is also known as platelet glycoprotein 40 (Pltgp40), melanoma antigen ME491 or MLA1, ocular melanoma associated antigen (OMA81H), tetraspanin 30 (TSPAN30), granulophysin, or lysosomal integral membrane protein 1 (LIMP-1). CD63 is a member of the tetraspanin superfamily and is widely expressed. The CD63 gene is located on human chromosome 12q13 and was the first tetraspanin to be characterized. CD63 was originally discovered as a protein present on the cell surface of activated platelets, known as Pltgp40, and was also found in early human melanoma cells, where it was known as ME491.

[0036] CD63 is expressed in many cell types. In particular, CD63 is expressed intracellularly in the lysosomes, endosomes, and granules of resting platelets and basophils. Cell surface expression of CD63 can be detected in activated basophils and platelets, monocytes, macrophages, and granulocytes. CD63 is also expressed in endothelial cells, fibroblasts, osteoblasts, nerve tissue, melanoma cells, smooth muscle cells, and mast cells. CD63 has been described to shuttle between the cell membrane and intracellular compartments.

[0037] The major pool of CD63 resides in intracellular compartments such as endosomes and lysosomes, although some expression can also be found on the cell surface. CD63 has been described to control the trafficking of other proteins, typically through endocytosis. In addition, CD63 has been described to control the surface expression of membrane type 1 matrix metalloproteinase by targeting the enzyme for lysosomal degradation, and silencing CD63 in endothelial cells prevents the internalization of vascular endothelial growth factor receptor 2 (VEGFR2) in response to the ligand VEGF. Also, in different tumor types, CD63 has been shown to shuttle continuously between the cell membrane and lysosomes, which was dependent on the presence of AP2 and clathrin. Thus, CD63 appears to be an attractive antigen for facilitating internalization and lysosomal delivery, a feature suitable for enhancing the efficacy of certain antibody drug conjugates (ADCs) that target tumor antigens that do not sufficiently internalize and / or shuttle to lysosomes alone.

[0038] CD63 was first discovered as a surface antigen abundantly expressed on early melanoma cells, but CD63 expression does not show tumor-specific expression. CD63 cell surface expression is decreased during the progression of malignant melanoma, indicating an inverse correlation between CD63 cell surface expression and tumor aggressiveness.

[0039] Tumor-associated antigens Tumor-associated antigens are antigens that are expressed on the surface of (certain types of) tumor cells and that are found to a lesser extent on normal cells. As used herein, the term "tumor-associated antigen" refers not only to proteins or polypeptides that are preferentially expressed on the (external) surface of tumor cells, but also to carbohydrates, glycoproteins, lipids, lipoproteins, lipopolysaccharides, or other non-protein polymers. The term "preferentially expressed", as used in this context, means that the antigen is expressed in tumor cells at a level that is at least 10%, such as at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 100%, at least 110%, at least 150%, at least 200%, at least 300%, at least 400%, at least 500%, at least 750%, at least 1000%, at least 2000%, or at least 5000% greater than the expression level of the antigen in non-tumor cells. Tumor-associated antigens can be derived from any protein, glycoprotein, or other macromolecule synthesized by tumor cells.

[0040] Tumor-associated antigens can be divided into different antigen classes: (1) class I HLA-restricted cancer-testis antigens that are normally expressed in the testis or some tumors but not in normal tissues, including antigens from the MAGE, BAGE, GAGE, NY-ESO, and BORIS families; (2) class I, including melanocyte differentiation antigens such as MART-1, gp100, PSA, tyrosinase, TRP-1, and TRP-2; HLA-restricted differentiation antigens; (3) broadly expressed antigens, which are antigens or altered translation products expressed in both normal and tumor tissues at different levels, including CEA, HER2 / neu, hTERT, MUC1, MUC2, and WT1; (4) tumor-specific antigens, which are unique antigens arising from mutations in normal genes, including β-catenin, α-fetoprotein, MUM, RAGE, SART, etc.; (5) viral antigens, such as HPV, EBV; and (6) fusion proteins, which are proteins created by chromosomal rearrangements, such as deletions, translocations, inversions, or duplications, resulting in new proteins expressed exclusively by tumor cells, such as Bcr-Abl.

[0041] Preferred examples of tumor associated antigens include 5T4, A33, activin receptor, adrenomedullin receptor, AFP, AGS-5, ALK, annexin, AXL, B7-H3, B7-H4, BAGE protein, BCMA, bombesin, C33 antigen, C4.4a, C-type lectin-like (receptor), CA19.9, CA-125, CADM1, CAIX, CanAg, CAR, carbonic anhydrase, caveolin 1, CCK2R, CD4, CD10, CD19, CD20, CD21, CD22, CD25, CD27, CD30, CD33, CD37, CD38, CD44, CD51, CD57, CD70, CD73, CD74, CD79a, CD79b, CD80, CEA, CEACAM, c-kit, claudins, chemokine receptors (i.e., CXCR4, CXCR5), c-Met, Cripto-1, DEC-205, Derlin 1, desmoglein 3, Dlk-1, DLL3, DS6, E-cadherin, E-selectin, EAG-1, ED-B, EpCAM, EGFR, EGFRvIII, emmprin, endothelin receptor, ErbB2 / Her2, ErbB3, ErbB4, ETV6-AML, ephrin type A receptor, epiregulin, ETA, FAPα, FcyR, FGFR, FOLR1, Frizzled, Fyn3, galectin, ganglioside, GCC, GD2, GD3, GloboH, lypican3, GLUT3, GPNMB, G protein-coupled receptor (i.e., GPR49), gp100, Hsp, HLA / B-raf, HLA-DR, HLA / k-ras, HLA MAG E-A3, HMW-MAA, hTERT, ICAM-3, IGF-R, IL-13-R, L1CAM, laminin receptor, LIV1, LMP2, LRP5, LRP6, MAGE proteins, MART-1, melanotransferrin, mesothelin, metalloproteinases, ML-IAP, mucin, Mud, Mud 6 (CA-125), MU M1, N-cadherin, NA17, NCAM-1, nectin 4, notch, NP-55, NRP1, NY-BR1, NY-BR62, NY-BR85, NY-ES01, PLAC1, PRLR, PRAME, prominin 1, PSMA (FOLH 1), RON, SLC44A4, SLITRK6, Steap-1, Steap-2, surviving, syndecan, TAG-72, TF, TGFβ, TMPRSS2, TMEFF2, TNFR, Tn, TROP2, TRP-1, TRP-2, TWEAKR, tyrosinase, uroplakin 3, and VEGFR.

[0042] Preferred examples of tumor-associated antigens include those that are highly overexpressed but lack sufficient lysosomal trafficking, such as glycosylphosphatidylinositol (GPI)-anchored proteins (i.e., glypican family, uPAR, folate-binding receptor, prostasin, FcgRIIIb [CD16b], alkaline phosphatase, acetylcholinesterase, 5'-nucleotidase [p36], Cripto, LFA-3 [CD58], DAF [CD55], Thy-1 [CD90], Qa-2, Ly-6A, and MIRL [CD59]), adhesion molecules (i.e., selectins, L1CAM, N-CAM, LRP1, TAG1, cadherins), which are often recycled back to the cell membrane after endocytosis, with only a small fraction targeted for lysosomal degradation.

[0043] Preferred examples of tumor-associated antigens include tumor-specific antigens known to interact with CD63, which at low copy number can enhance bivalent binding of bsADCs. For example, CD63 has been described to interact with other tetraspanins (i.e., CD81, CD82, CD9, and CD151), integrins, MHCII, CXCR4, TM4SF5, syntenin 1, TIMP-1, H,K-ATPase, L6 antigen, and MT1-MMP.

[0044] Examples of tumor-associated antigens include Lewis Y (CD174), Lewis X (CD15), SLe X , SLe A These include the selection of a glycotarget such as sTn, Fucosyl GM1, Globo H, SSEA-3, GM2, GD2, GD3, polysialic acid, or a glycoprotein (i.e., mucin). [Brief description of the drawings]

[0045] [Figure 1] 1 shows dose response curves for binding of anti-CD63 antibodies to recombinant human CD63 as measured by ELISA. [Diagram 2]1 shows affinity measurements of affinity variants of anti-CD63 antibodies measured by label-free biolayer interferometry. [Diagram 3] 1 shows the results of a viability assay to test the cytotoxicity of monovalent bsCD63N74H×b12-Duo3 and bsHER2×CD63N74H-Duo3 ADCs in Colo205 cells. [Figure 4] 1 shows the results of a viability assay to test the cytotoxicity of monovalent bsCD63N74H×b12-Duo3 and bsHER2×CD63N74H-Duo3 ADCs in SK-OV-3 cells. [Diagram 5] 1 shows the results of a viability assay to test the cytotoxicity of monovalent bsCD63N74H×b12-Duo3 and bsHER2×CD63N74H-Duo3 ADCs in HCC1954 cells. [Figure 6] FIG. 13 shows lysosomal co-localization of monovalent bsCD63N74H×b12 bispecific molecules measured in SK-OV-3 cells by confocal microscopy. [Figure 7] Shows binding of bsHER2xCD63N74H to SK-OV-3 cells as determined by flow cytometry. [Figure 8] 1 shows the intracellular accumulation of FITC-conjugated CD63 antibody and CD63 affinity variant antibody in granulocytes and spindle cells. [Figure 9] Shows lysosomal colocalization of bsHER2×CD63N74H in SK-OV-3 cells followed over time. [Figure 10] Shown is the total amount of HER2 protein in tumor cell lines with different expression levels of HER2 quantified by ELISA after 3 days of incubation with bsHER2xCD63N74H compared to untreated cells. [Figure 11] 1 shows the results of a viability assay to test the cytotoxicity of Duostatin 3-conjugated bispecific ADCs in vitro. [Figure 12]1 shows the mean tumor size and tumor-free survival in mice subcutaneously inoculated with SK-OV-3 tumor xenografts and subsequently treated with Duostatin 3-conjugated multispecific ADCs. [Figure 13] Shows binding of bsβ1×CD63N74H to SK-OV-3 cells as assessed by flow cytometry. [Figure 14] Shows lysosomal co-localization of bsβ1×CD63N74H in SK-OV-3 cells. [Figure 15] Shows lysosomal co-localization of bsβ1×CD63N74H in SK-OV-3 cells followed over time. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0046] Detailed Description of the Invention In one aspect, the present invention relates to a multispecific antigen-binding molecule comprising a first antigen-binding domain and a second antigen-binding domain, wherein the first antigen-binding domain specifically binds to a target molecule (T) and the second domain specifically binds to an internalizing effector protein (E), and the second antigen-binding domain has a dissociation constant K D Value 10 -9 ~10 -8 Has M.

[0047] To provide tumor specificity for the multispecific antigen-binding molecules of the present invention, in the absence of the first antigen-binding domain, the second domain that specifically binds to the internalizing effector protein (E) advantageously does not bind or binds only with low affinity and then does not internalize or at least internalizes to a significantly lesser extent. D Value 10 -9 ~10 -8 It has been found by the present inventors that a multispecific antigen-binding molecule having M satisfies these criteria.

[0048] The target molecule (T), preferably a tumor-associated target molecule, can be a protein, a polypeptide, a lipid, or other macromolecule. In a preferred embodiment, the target molecule is a protein. In another embodiment, the target molecule, preferably a tumor-associated target molecule, is a polypeptide. In some embodiments, T is a target protein or target polypeptide expressed on the cell surface. In other embodiments, T is a soluble target protein or target polypeptide, preferably one that interacts with a cell surface receptor. Target binding by the multispecific antigen-binding molecule may occur extracellularly or on the cell surface.

[0049] In one embodiment, the target molecule is a receptor expressed on the cell surface. In a preferred embodiment, the target molecule is a tyrosine kinase receptor, preferably a transmembrane tyrosine kinase receptor. In another embodiment, the target molecule is a membrane-bound ligand.

[0050] In other embodiments, the multispecific antigen-binding molecule, preferably the bispecific antigen-binding molecule, binds to E on the cell surface or intracellularly.

[0051] It is particularly preferred that the target molecule is a tumor-associated antigen, such as a tumor-associated protein or polypeptide. Advantageously, the tumor-associated antigen is an antigen that is not normally internalized or is poorly internalized. Preferably, the tumor-associated antigen is an antigen that shows inefficient translocation to the lysosomal compartment.

[0052] The internalizing effector protein (E) may be tumor-associated or tumor-specific, hi other embodiments, the internalizing effector protein (E) may be expressed in both tumor and non-tumor cells.

[0053] An internalizing effector protein (E) is a protein that can be internalized into a cell or that is otherwise involved in or contributes to internalization. In some embodiments, the internalizing effector protein is a protein that undergoes transcytosis; that is, the protein is internalized on one side of the cell and transported to the other side of the cell. Preferably, the internalizing effector protein is a membrane protein or a soluble extracellular protein that binds to a membrane-bound receptor. In a preferred embodiment, the internalizing effector protein is a protein that exhibits efficient translocation to the lysosomal compartment of the cell.

[0054] The binding of the second domain to an internalizing effector protein advantageously results in the internalization of the multispecific antigen-binding molecule and the associated target molecule into the cell. In a preferred embodiment, the internalizing effector protein is a membrane-associated protein having at least one extracellular domain or region, which is internalized and preferably processed via intracellular degradation and / or recycling pathways. Specific examples of internalizing effector proteins that are directly internalized into cells include, for example, CD63, MHC-I (e.g., HLA-B27), Kremen 1, Kremen 2, LRP5, LRP6, LRP8, transferrin receptor, LDL receptor, LDL-related protein 1 receptor, ASGR1, ASGR2, amyloid precursor protein-like protein 2 (APLP2), apelin receptor (APLNR), MAL (Myelin And Lymphocyte Protein, VIP17), IGF2R, vacuolar H+ATPase, diphtheria toxin receptor, folate receptor, glutamate receptor, glutathione receptor, leptin receptor, scavenger receptor (e.g., SCARA1-5, SCARB1-3, CD36). In a preferred embodiment, the internalizing effector protein E is a cell surface internalizing receptor. In a preferred embodiment, the internalizing effector protein E is CD63.

[0055] In a preferred embodiment, the multispecific antigen-binding molecule comprises (i) a first binding arm comprising a first antigen-binding domain and (ii) a second binding arm comprising a second antigen-binding domain.

[0056] It is particularly preferred that the multispecific antigen-binding molecule is a bispecific antigen-binding molecule.

[0057] In one embodiment of the invention, the second antigen-binding domain comprises 10 -9 10 higher than M -8 Dissociation constant K with E is lower than M D According to another embodiment, the second antigen-binding domain has a dissociation constant K D 2.0×10 -9 ~9.0×10 -9 According to another embodiment, the second antigen-binding domain has a dissociation constant K D 2.0×10 -9 ~7.3×10 -9 Has M.

[0058] According to another embodiment, E is a cell surface expressed molecule that is internalized, preferably directly internalized, into the cell. Preferably, the multispecific antigen-binding molecule is internalized into the cell by binding to E only in the presence of a target molecule (T). It is also preferred that the multispecific antigen-binding molecule is internalized into the cell by binding to E only when the first domain specifically binds to the target molecule (T).

[0059] In one embodiment, the multispecific antigen binding molecule, upon binding to E, is internalized more efficiently into cells expressing T compared to cells that do not express T.

[0060] In another embodiment, the multispecific antigen binding molecule, upon binding to T, is internalized more efficiently into cells that express E compared to cells that do not express E.

[0061] In another embodiment, the multispecific antigen-binding molecule, upon binding to E, is transported to the lysosomal compartment in cells expressing T.

[0062] In another embodiment, the multispecific antigen-binding molecule, upon binding to E, is transported to the lysosomal compartment more efficiently in cells expressing T compared to cells that do not express T.

[0063] In another embodiment, the multispecific antigen-binding molecule, upon binding to T, is transported to the lysosomal compartment more efficiently in cells expressing E compared to cells that do not express E.

[0064] According to another embodiment, E is selected from the group consisting of CD63, MHC-I, Kremen 1, Kremen 2, LRP5, LRP6, transferrin receptor, LDLr, MAL, V-ATPase, and ASGR. In a preferred embodiment, E is CD63.

[0065] According to another embodiment, E is a soluble ligand that is internalized into cells via interaction between E and an internalizing receptor molecule expressed on the cell surface.

[0066] According to another embodiment, T is a target molecule expressed on the cell surface. In a particularly preferred embodiment, T is a tumor-associated antigen. Thus, the strategy of enhancing internalization of the present invention can include combining a tumor-associated target antigen with the internalization ability of an antigen, such as CD63. In particular, a bispecific antibody that binds both CD63 and a tumor-associated target can be useful in therapeutic settings where specific targeting and enhanced internalization of antibody-drug conjugates are desired. According to one embodiment, T is HER2.

[0067] According to another embodiment, the first and / or second antigen binding domain comprises at least one antibody variable region, preferably at least two antibody variable regions.

[0068] According to some embodiments, the multispecific antigen-binding molecule is a multispecific antibody, preferably a bispecific antibody, or a multispecific, preferably bispecific, antibody fragment or recombinant modified part thereof.In a particularly preferred embodiment, the multispecific antigen-binding molecule is a bispecific antibody.A bispecific antibody can be used to use the internalization-enhancing properties of one antigen by binding with one arm of the bispecific antibody, and to bind to a target molecule, such as a tumor-associated target molecule, with the other arm.Then, such a bispecific antibody can be loaded with a cytotoxic conjugate to induce cell death upon internalization of the ADC.

[0069] In one embodiment, the antibody is a bispecific antibody comprising (i) a first antibody comprising a first antigen-binding domain that specifically binds to a target molecule (T) as defined herein, and (ii) a second antibody comprising a second antigen-binding domain that specifically binds to an internalizing effector protein (E) as defined herein.

[0070] In a preferred embodiment, the multispecific antigen-binding molecule is a bispecific antibody comprising a first binding arm comprising a first antigen-binding domain and a second binding arm comprising a second antigen-binding domain. Advantageously, the first antigen-binding domain comprises a first heavy chain variable sequence (VH) and a first light chain variable sequence (VL), and the second antigen-binding domain comprises a second heavy chain variable sequence (VH) and a second light chain variable sequence (VL), and each of the variable sequences comprises three CDR sequences, CDR1, CDR2, and CDR3.

[0071] In a preferred embodiment, (i) the first binding arm comprises a first heavy chain comprising a first heavy chain variable sequence (VH) and a first heavy chain constant sequence (CH), and a first light chain comprising a first light chain variable sequence (VL) and a first light chain constant sequence (CL), and (ii) the second binding arm comprises a second heavy chain comprising a second heavy chain variable sequence (VH) and a second heavy chain constant sequence (CH), and a second light chain comprising a second light chain variable sequence (VL) and a second light chain constant sequence (CL).

[0072] According to another embodiment, the first binding arm is derived from a chimeric antibody or a humanized antibody or a human antibody. According to another embodiment, the second binding arm is derived from a chimeric antibody or a humanized antibody or a human antibody. Thus, in one embodiment, the first binding arm is derived from a human antibody and the second binding arm is derived from a humanized or chimeric antibody.

[0073] It is preferred that the multispecific antigen-binding molecules of the present invention are bispecific antibodies, which are full-length antibodies, preferably IgG1 antibodies.

[0074] In a preferred embodiment, the multispecific antigen-binding molecule of the present invention is isolated. As used herein, the term "isolated multispecific antigen-binding molecule" refers to a multispecific antigen-binding molecule, such as a bispecific antibody, that is substantially free of other antigen-binding molecules or antibodies with different antigen specificities. Furthermore, an isolated multispecific antigen-binding molecule is substantially free of other cellular material and / or chemicals.

[0075] According to another embodiment, the second antigen binding domain has one or more mutations that modulate the affinity of the second antigen binding domain with E. In another embodiment, the second antigen binding domain is derived from an antibody having one or more mutations in VH and / or VL that modulate the affinity of the second antigen binding domain with E. The affinity is determined by the dissociation constant K D Value 10 -9 ~10 -8 It is preferred that the β-terminal amino acid sequence is modulated to have M.

[0076] According to another embodiment, when E is CD63, the antibody has one or more mutations in the anti-CD63 Fab region that modulate the affinity of the second antigen-binding domain with E. According to another embodiment, the mutation is a single amino acid substitution, preferably a single amino acid histidine substitution.

[0077] According to another embodiment, the one or more mutations in the VH and / or VL are amino acid substitutions, preferably histidine substitutions, at position 54 of the VL according to SEQ ID NO:5 in Table 1 below, or at positions 71, 72 and / or 74 of the VH according to SEQ ID NO:1 in Table 1 below. For example, anti-CD63-N74H has an asparagine to histidine mutation at position 74 of the heavy chain according to SEQ ID NO:1 and anti-CD63-LN54H has an asparagine to histidine mutation at position 54 of the light chain according to SEQ ID NO:5. According to another embodiment, the second antigen-binding domain has a K within the preferred affinity range. D is selected to bind to target E at

[0078] In a preferred embodiment, the amino acid substitution, preferably a histidine substitution, is at position 74 of VH according to SEQ ID NO:1. In a preferred embodiment, the mutation is N74H of VH according to SEQ ID NO:1.

[0079] In a preferred embodiment, the second domain, preferably as part of the second binding arm, comprises: (a) VH CDRs 1, 2, and 3 as provided in SEQ ID NOs:2, 3, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs:9, 7, and 8, respectively; or (b) VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 10, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively; or (c) VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 11, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively; or (d) VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 12, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively. Includes.

[0080] In a particularly preferred embodiment, the second domain preferably comprises, as part of the second binding domain, VH CDRs 1, 2 and 3 as provided in SEQ ID NOs:2, 12 and 4, respectively, and VL CDRs 1, 2 and 3 as provided in SEQ ID NOs:6, 7 and 8, respectively.

[0081] According to another embodiment, the multispecific antigen-binding molecule comprises a variant Fab region of a CD63-specific monoclonal antibody. According to another embodiment, the multispecific antigen-binding molecule comprises a variant Fab region of CD63-specific monoclonal Ab 2192. According to another embodiment, the multispecific antigen-binding molecule comprises an antigen-binding region specific for CD63 selected from a hybridoma or phage display library.

[0082] According to another embodiment, the first and second antigen binding domains are each a pair of an antibody heavy chain variable domain and an antibody light chain variable domain.

[0083] The bispecific antigen-binding molecule may preferably further comprise an antibody constant region.

[0084] According to another embodiment, the antigen binding molecule is a tumor-associated target (T) x CD63 bispecific antibody. In a preferred embodiment, the (T) x CD63 bispecific antibody is conjugated to a cytotoxic drug. In one embodiment, the (T) x CD63 bispecific antibody is conjugated to Duostatin 3. In one embodiment, the (T) x CD63 bispecific antibody is conjugated to Duostatin 3, and the second binding domain, the anti-CD63 binding domain, comprises VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 12, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively. In particular, the tumor-associated target (T) x CD63 bispecific antibody drug conjugates (ADCs) are useful in therapeutic settings where specific targeting and enhanced internalization of the antibody drug conjugate is desired. The (T)×CD63 bispecific ADCs of the invention have been found by the inventors to be more efficient in killing cells expressing the target T compared to targeting only tumor-associated antigens using monospecific ADCs. Such ADCs are found to be more potent in eradicating tumor cells in vitro and in animal models than prior art bispecific ADCs.

[0085] The (T)xCD63 bispecific ADCs of the present invention are found to be advantageous by their ability to bind both tumor-associated targets and the strong internalization properties of the CD63 antigen, thereby enhancing the internalization of the ADC and thereby inducing more potent killing of the cells through more efficient payload delivery within the target cells. To address the need to increase the efficacy of antibodies targeting various tumor antigens, a narrow range of surprisingly efficient CD63 affinity variants (spanning a range of CD63 affinities) have been found to enhance the efficacy of bispecific ADCs.

[0086] According to another embodiment, the antigen binding molecule is a HER2xCD63 bispecific antibody.

[0087] According to another embodiment, the antigen binding molecule has an EC of less than 5.0 μg / ml, such as less than 4.0 μg / ml, for example less than 3.0 μg / ml, such as less than 2.0 μg / ml, for example less than 1.0 μg / ml, such as less than 0.9 μg / ml, for example less than 0.8 μg / ml, such as less than 0.7 μg / ml, for example less than 0.6 μg / ml, for example less than 0.5 μg / ml, such as less than 0.4 μg / ml, for example less than 0.3 μg / ml, for example less than 0.2 μg / ml, such as less than 0.1 μg / ml, for example less than 0.05 μg / ml, for example less than 0.01 μg / ml, for binding to cells expressing a tumor-associated target (T), such as cells expressing HER2, as determined by flow cytometry. 50 It has a value.

[0088] In another preferred embodiment, binding of T and E by the multispecific antigen-binding molecule induces internalization of the multispecific antigen-binding molecule to a greater extent than binding of T by the first domain alone. Similarly, binding of T and E by a drug-conjugated multispecific antibody of the invention preferably induces cytotoxicity to a greater extent than binding of T by the first domain alone.

[0089] In another preferred embodiment, binding of T and E by a multispecific, preferably bispecific, antigen-binding molecule induces internalization of the multispecific antigen-binding molecule to a greater extent than a corresponding bivalent, monospecific antibody that binds T. Similarly, binding of T and E by a drug-conjugated, multispecific, preferably bispecific, antibody of the invention preferably induces cytotoxicity to a greater extent than a corresponding bivalent, monospecific ADC that binds T.

[0090] According to another aspect, K D The value is determined by biolayer interferometry at 30° C. D is determined by biolayer interferometry at a pH of 7.2 to 7.5, e.g., 7.3 to 7.4, e.g., pH 7.4, at 30° C. Dis determined by biolayer interferometry at 1000 RPM shaker speed and 30° C. D is determined by biolayer interferometry using an Octet system such as the Octet HTX (ForteBio).

[0091] In a preferred embodiment, the antigen-binding molecule comprises (i) a first binding arm comprising a first heavy chain comprising a first heavy chain constant sequence (CH), the first CH comprising a first CH3 region; (ii) a second binding arm comprising a second heavy chain comprising a second heavy chain constant sequence (CH), the second CH comprising a second CH3 region; and a bispecific antibody comprising The sequences of the first CH3 region and the second CH3 region are different and are such that the heterodimer interaction between the first binding arm and the second binding arm is stronger than the homodimer interaction of each of the first binding arm and the second binding arm.

[0092] Preferably, in the first heavy chain CH3 region at least one of the amino acids at a position corresponding to positions T366, L368, K370, D399, F405, Y407 or K409 of a human IgG1 heavy chain is substituted and in the second heavy chain CH3 region at least one of the amino acids at a position corresponding to positions T366, L368, K370, D399, F405, Y407 or K409 of a human IgG1 heavy chain is substituted, the first and second heavy chains not being substituted at identical positions, and the amino acid positions are numbered according to the EU index.

[0093] In another embodiment, (i) the first CH3 region has a F405L substitution and the second CH3 region has a K409R substitution, or (ii) the first CH3 region has a K409R substitution and the second CH3 region has a F405L substitution.

[0094] According to another embodiment, the multispecific antigen-binding molecule is conjugated to a cytotoxic moiety, a radioisotope, a drug, a cytokine, or an RNA silencing vehicle.Preferably, the multispecific antigen-binding molecule is conjugated to a cytotoxic moiety, a radioisotope, or a drug.Preferably, the multispecific antigen-binding molecule is conjugated to a cytotoxic moiety.

[0095] Cytotoxic moieties include duostatin 3, duostatin 5, pyrrolobenzodiazepines or analogues or derivatives, IGN-based toxins or analogues or derivatives, alpha-amanitin or analogues or derivatives, dolastatin or analogues or derivatives, taxol; cytochalasin B; gramicidin D; ethidium bromide; emetine; mitomycin; etoposide; tenoposide; vincristine; vinblastine; colchicine; doxorubicin; daunorubicin; Dihydroxyanthracin diones; tubulin inhibitors such as maytansine or its analogues or derivatives; mitoxantrone; mithramycin; actinomycin D; 1-dehydrotestosterone; glucocorticoids; procaine; tetracaine; lidocaine; propranolol; puromycin; calicheamicin or its analogues or derivatives, methotrexate, 6-mercaptopurine, 6-thioguanine, cytarabine, fludarabine, 5-fluorouracil, decal antimetabolites such as decarbazine, hydroxyurea, asparaginase, gemcitabine, or cladribine; mechlorethamine, thioepa, chlorambucil, melphalan, carmustine (BSNU), lomustine (CCNU), cyclophosphamide, busulfan, dibromomannitol, streptozotocin, dacarbazine (DTIC), procarbazine, mitomycin C, cisplatin, carboplatin, duocarmycin A, duocarmycin SA, laccase, alkylating agents such as rachelmycin (CC-1065) or an analogue or derivative thereof; antibiotics such as dactinomycin, bleomycin, daunorubicin, doxorubicin, idarubicin, mithramycin, mitomycin, mitoxantrone, plicamycin, anthramycin (AMC); mitotic inhibitors such as auristatin or an analogue or derivative thereof, monomethylauristatin E or F, or an analogue or derivative thereof;Diphtheria toxin and related molecules such as diphtheria A chain and active fragments and hybrid molecules thereof, ricin toxins such as ricin A or deglycosylated ricin A chain toxins, cholera toxin, shiga-like toxins such as SLT I, SLT II, ​​SLT IIV, LT toxin, C3 toxin, shiga toxin, pertussis toxin, tetanus toxin, soybean Bowman-Birk protease inhibitors, Pseudomonas exotoxins, alorin, saporin, modeccin, geranin, abrin A chain, modeccin A chain, alpha-sarcin, Aleurites fordii proteins, dianthin proteins, Phytolacca americana proteins such as PAPI, PAPII, and PAP S, momordica charantia inhibitors, curcin, crotin, sapaonaria officinalis inhibitor, gelonin, mitogellin, restrictocin, phenomycin, and enomycin toxins; ribonuclease (RNase); DNase I, Staphylococcus aureus enterotoxin A; pokeweed antiviral protein; diphtherin toxin, Pseudomonas endotoxin, and RNAi (i.e., siRNA, shRNA conjugated to an antibody or delivered on a nanoparticle);

[0096] According to another embodiment, the cytotoxic moiety is selected from the group consisting of maytansine, calicheamicin, duocarmycin, duostatin, duostatin 3, duostatin 5, rachelmycin (CC-1065), auristatin, monomethylauristatin E, monomethylauristatin F, doxorubicin, dolastatin, pyrrolobenzodiazepines, IGN-based toxins, alpha-amanitin, or an analog, derivative, or prodrug of any of them.

[0097] In one embodiment, the cytotoxic moiety, drug, or radioisotope is linked to the antibody or fragment thereof by a cleavable linker, such as N-succinimidyl 4-(2-pyridyldithio)-pentanoate (SSP), maleimidocaproyl-valine-citrulline-p-aminobenzyloxycarbonyl (mc-vc-PAB), or AV-1 K-locked valine-citrulline.

[0098] The term "cleavable linker", as used herein, refers to a subset of linkers that are catalyzed by specific proteases in the target cell or tumor microenvironment, resulting in the release of the cytotoxic agent. Examples of cleavable linkers are linkers based on chemical motifs including disulfides, hydrazones, or peptides. Another subset of cleavable linkers adds an additional linker motif between the cytotoxic agent and the primary linker, i.e., a site for attaching the linker-drug combination to the antibody. In some embodiments, the additional linker motif is cleavable by a cleavable agent present in the intracellular environment (e.g., inside a lysosome or endosome or caveolae). The linker can be, for example, a peptidyl linker that is cleaved by intracellular peptidases or protease enzymes, including but not limited to lysosomal or endosomal proteases. In some embodiments, the peptidyl linker is at least 2 amino acids long or at least 3 amino acids long. Cleavage agents can include cathepsin B and D and plasmin, all of which are known to hydrolyze dipeptide drug derivatives in target cells, resulting in the release of active drugs (see, for example, Dubowchik and Walker, 1999, Pharm. Therapeutics 83:67-123). In a specific embodiment, the peptidyl linker that can be cleaved by intracellular proteases is a Val-Cit (valine-citrulline) linker or a Phe-Lys (phenylalanine-lysine) linker (see, for example, US6214345, which describes the synthesis of doxorubicin with a Val-Cit linker). The advantage of using intracellular proteolytic release of therapeutic substances is that the substances are typically attenuated when conjugated, and the serum stability of the conjugate is typically high.

[0099] In another embodiment, the cytotoxic agent, drug, or radioisotope is linked to the antibody or fragment thereof by a non-cleavable linker, such as succinimidyl-4(N-maleimidomethyl)cyclohexane-1-carboxylate (MCC) or maleimidocaproyl (MC).

[0100] The term "non-cleavable linker", as used herein, refers to a subset of linkers that, in contrast to cleavable linkers, do not contain motifs that are specifically and predictably recognized by intracellular or extracellular proteases.Thus, ADCs based on non-cleavable linkers are not released or cleaved from the antibody until the complete antibody-linker-drug complex is degraded in the lysosomal compartment.An example of a non-cleavable linker is a thioether.In yet another embodiment, the linker unit is not cleavable and the drug is released by antibody degradation.

[0101] In particularly preferred embodiments, binding of T and E by the multispecific antigen-binding molecule induces internalization of the multispecific antigen-binding molecule to a greater extent than binding to the target T alone.

[0102] In another aspect, the present invention relates to multispecific antibodies generated by using technologies or formats such as, but not limited to, DuoBody, CrossMab, triomab, kih IgG common LC, DVD-Ig, 2-in-1-IgG, IgG-scFv, bi-nanobody, BiTE, TandAb, DART, DART-Fc, scFv-HSA-scFv, orthoFab-IgG, tetravalent Tv-IgG, dock-and-lock (DNL) formats such as DNL-Fab3, or fragments such as tandem scFv, tandem scFv-Fc, knob-into-hole IgG, scFv-Fc knob-into-hole, scFv-Fc-scFv, F(ab')2, Fab-scFv, (Fab'scFv)2, diabody, sc diabody, sc diabody-Fc, sc diabody-CH3, or azimetric scaffolds.

[0103] In another aspect, the present invention provides a tandem scFv, a tandem scFv-Fc, a scFv-Fc knob-into-hole, a scFv-Fc-scFv, a F(ab')2, a Fab-scFv, a (Fab'scFv)2, a diabody, a sc diabody, a sc diabody-Fc, a sc diabody-C H 3. The bispecific antibody fragment of the multispecific antigen-binding molecule.

[0104] In another aspect, the present invention relates to a multispecific antigen-binding molecule or bispecific antibody fragment for use in a method for treating and / or preventing cancer.The subject treated in such a method is preferably a human individual who needs such treatment, such as a cancer patient.In one embodiment, the cancer is breast cancer, including primary, metastatic and refractory breast cancer.

[0105] In some embodiments, the cancer is endometrial / cervical cancer, lung cancer, malignant melanoma, ovarian cancer, pancreatic cancer, prostate cancer, testicular cancer, soft tissue tumors such as synovial sarcoma, breast cancer, brain cancer, leukemia, lymphoma, mast cell tumor, kidney cancer, cervical cancer, bladder cancer, esophageal cancer, gastric cancer, or colorectal cancer.

[0106] Effective dosages and dosing regimens for the multispecific antigen-binding molecules depend on the cancer being treated. An exemplary non-limiting range for a therapeutically effective amount of a bispecific antibody of the present invention is about 0.1-100 mg / kg, such as about 0.1-50 mg / kg, such as about 0.1-20 mg / kg, such as about 0.1-10 mg / kg, such as about 0.5 mg / kg, such as about 0.3 mg / kg, about 1 mg / kg, about 3 mg / kg, about 5 mg / kg, or about 8 mg / kg.

[0107] In some embodiments, the multispecific antigen binding molecule may be administered prophylactically to reduce the risk of developing cancer, to delay the onset of occurrence of events during cancer progression, or in an adjuvant setting and / or to reduce the risk of recurrence when the cancer is in remission.

[0108] In one embodiment, a method for treating or preventing cancer comprises administering to a subject in need thereof a therapeutically effective amount of a multispecific antigen-binding molecule of the present invention and at least one additional therapeutic agent. In some embodiments, such additional therapeutic agents may be selected from antimetabolites such as methotrexate, 6-mercaptopurine, 6-thioguanine, cytarabine, fludarabine, 5-fluorouracil, decarbazine, hydroxyurea, asparaginase, gemcitabine, or cladribine. In other embodiments, such additional therapeutic agents may be selected from alkylating agents such as mechlorethamine, thioepa, chlorambucil, melphalan, carmustine (BSNU), lomustine (CCNU), cyclophosphamide, busulfan, dibromomannitol, streptozotocin, dacarbazine (DTIC), procarbazine, mitomycin C, cisplatin, and other platinum derivatives such as carboplatin. In other embodiments, such additional therapeutic agents may be selected from mitotic inhibitors such as taxanes, e.g., docetaxel and paclitaxel, and vinca alkaloids, e.g., vindesine, vincristine, vinblastine, and vinorelbine, hi other embodiments, such additional therapeutic agents may be selected from topoisomerase inhibitors, such as topotecan or irinotecan, or cytostatics, such as etoposide and teniposide. In other embodiments, such additional therapeutic agents may be selected from a growth factor inhibitor such as an inhibitor of ErbB1 (EGFR) (such as an EGFR antibody, e.g., zalutumumab, cetuximab, panitumumab, nimotuzumab, gefitinib, or other EGFR inhibitors such as erlotinib), another inhibitor of ErbB2 (HER2 / neu) (such as a HER2 antibody, e.g., trastuzumab, trastuzumab-DM1, or pertuzumab), or an inhibitor of both EGFR and HER2, such as lapatinib. In other embodiments, such additional therapeutic agents may be selected from a tyrosine kinase inhibitor, such as imatinib or lapatinib.

[0109] In another aspect, the present invention relates to a multispecific antigen-binding molecule for use in a method for targeting a tumor in a subject, the method comprising administering the multispecific antigen-binding molecule to the subject.Tumors that can be targeted by the present invention include malignant and non-malignant tumors.Malignant (including primary and metastatic) tumors that can be treated include, but are not limited to, those that occur in the adrenal gland; bladder; bone; breast; cervix; endocrine glands (including thyroid, pituitary, and pancreas); colon; rectum; heart; hematopoietic tissue; kidney; liver; lung; muscle; nervous system; brain; eye; oral cavity; pharynx; larynx; ovary; penis; prostate; skin (including melanoma); testis; thymus; and uterus.Examples of such tumors include apudomas, choriocarcinomas, branchiomas, malignant carcinoid syndrome, carcinoid heart disease, cell tumors (e.g., Walker carcinoma, basal cell carcinoma, basosquamous cell carcinoma, Brown-Pierce carcinoma, ductal carcinoma, Ehrlich carcinoma, carcinoma in situ, Krebs 2 carcinoma, Merkel cell carcinoma, mucinous carcinoma, non-small cell lung carcinoma, oat cell carcinoma, papillary carcinoma, scirrhous carcinoma, bronchogenic carcinoma, squamous cell carcinoma, and transitional cell carcinoma). ), plasmacytoma, melanoma, chondroblastoma, chondroma, chondrosarcoma, fibroma, fibrosarcoma, giant cell tumor, histiocytoma, lipoma, liposarcoma, mesothelioma, myxoma, myxosarcoma, osteoma, osteosarcoma, Ewing's sarcoma, synovium, adenofibroma, adenolymphoma, carcinosarcoma, chordoma, mesenchymoma, mesonephroma, sarcoma, ameloblastoma, cementoma, odontoma, teratoma, thymoma, trophoblastoma, adenocarcinoma, adenoma, chorionic hemangioma, cholesteatoma, cylindroma, cystadenocarcinoma, cystadenoma, granulosa Cell tumor, Intersex positive ovarian tumor, Liver cancer, Sweat adenoma, Islet cell tumor, Leydig cell tumor, Papilloma, Sertoli cell tumor, Theca cell tumor, Leiomyoma, Leiomyosarcoma, Myoblastoma, Myoma, Myosarcoma, Rhabdomyoma, Rhabdomyosarcoma, Ependymoma, Ganglioneuroma, Glioma, Medulloblastoma, Meningioma, Schwannoma, Neuroblastoma, Neuroepithelioma, Neurofibroma, Neuroma, Paraganglioma, Nonchromaffin paraganglioma, Angiokeratoma, Eosinophilic angiolymphocytosis, Sclerosing hemangioma, Hematoma, Included are angiomatosis, glomus hemangiomas, hemangioendotheliomas, hemangiomas, hemangiopericytomas, angiosarcomas, lymphangiomas, lymphangioleiomas, lymphangiosarcomas, pinealomas, carcinosarcomas, chondrosarcomas, phyllodes cystsarcomas, fibrosarcomas, angiosarcomas, leiomyosarcoma, leukemia sarcomas, liposarcomas, lymphangiosarcomas, myxosarcomas, ovarian carcinomas, rhabdomyosarcomas, sarcomas (e.g., Ewing experimental sarcoma, Kaposi's sarcoma, and mast cell sarcoma), neoplasms, and other tumors of such cells. Administration can include intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, intranasal, epidural, and oral routes.

[0110] Similarly, the present invention relates to a method for killing tumor cells expressing a tumor-associated target molecule, such as HER2, comprising the step of administering to a subject in need thereof an effective amount of a multispecific antigen-binding molecule, such as a bispecific antibody, such as an antibody drug conjugate (ADC), of the present invention.

[0111] In one embodiment, the tumor cells are involved in a type of cancer selected from the group consisting of breast cancer, prostate cancer, non-small cell lung cancer, bladder cancer, ovarian cancer, gastric cancer, colorectal cancer, esophageal cancer, and squamous cell carcinoma of the head and neck, cervical cancer, pancreatic cancer, testicular cancer, malignant melanoma, and soft tissue cancer (e.g., synovial sarcoma).

[0112] In yet another aspect, the present invention relates to a pharmaceutical composition comprising the multispecific antigen-binding molecule as an active ingredient. Advantageously, such pharmaceutical composition is formulated with suitable excipients such as antioxidants, antibacterial agents, chelating agents, buffers, colorants, flavoring agents, diluents, emulsifiers, and / or suspending agents. The pharmaceutical composition may be administered by infusion, bolus injection, or by absorption through epithelial or mucocutaneous linings. In some embodiments, the pharmaceutical composition of the present invention may contain one or more additional pharmacologic active ingredients, such as cytotoxic agents or anticancer drugs.

[0113] In another aspect, the present invention relates to a method for treating a disease comprising the step of administering to a subject in need thereof a multispecific antigen-binding molecule of the present invention or a pharmaceutical composition of the present invention.

[0114] In another aspect, the present invention relates to a nucleic acid, such as a DNA molecule, encoding a multispecific antigen-binding molecule according to the invention. The nucleic acid may encode the heavy and light chains of a bispecific antigen-binding molecule, such as an antibody, of the present invention.

[0115] In another aspect, the present invention relates to an expression vector or a set of expression vectors that contain the nucleic acid and can express the nucleic acid in a prokaryotic or eukaryotic host cell line. The heavy and light chains of the antibody may be encoded by the same vector or by different vectors, depending on the bispecific antibody technology used. Such expression vectors can be used for the recombinant production of the antibody of the present invention.

[0116] The expression vector in the context of the present invention can be any suitable vector, including chromosomal vectors, non-chromosomal vectors, and synthetic nucleic acid vectors (nucleic acid sequences that contain an appropriate set of expression control elements). Examples of such vectors include derivatives of SV40, bacterial plasmids, phage DNA, baculovirus, yeast plasmids, vectors derived from a combination of plasmids and phage DNA, and viral nucleic acid (RNA or DNA) vectors. In one embodiment, the nucleic acid encoding the antibody is contained in, for example, a naked DNA vector or RNA vector that contains linear expression elements, a small nucleic acid vector, a plasmid vector such as pBR322, pUC19 / 18, or pUC118 / 119, a "midge" minimum size nucleic acid vector, or as a precipitated nucleic acid vector construct such as CaP04 precipitated construct.

[0117] In another aspect, the present invention relates to a prokaryotic or eukaryotic host cell line comprising the vector. The host cell is a cell into which an expression vector, i.e. an expression vector encoding a homodimeric monospecific precursor molecule, is introduced when Duobody technology is used to generate the bispecific antigen-binding molecule of the present invention, or a single host cell comprising a nucleic acid encoding the bispecific molecule of the present invention. Recombinant host cells include, for example, transfectomas such as CHO cells, HEK293 cells, NS / 0 cells, and lymphoid cells.

[0118] In another embodiment, the present invention relates to anti-idiotypic antibodies raised against the multispecific antigen-binding molecules of the present invention as defined above. Anti-idiotypic (Id) antibodies are antibodies that recognize unique determinants generally associated with the antigen-binding site of an antibody. Id antibodies can be prepared by immunizing an animal with a multispecific antigen-binding molecule, such as the bispecific antibody described above, from which the anti-Id is prepared. The immunized animal typically recognizes the idiotypic determinants of the immunizing bispecific antibody and is able to respond thereto by producing antibodies against these idiotypic determinants (anti-Id antibodies).

[0119] In one embodiment, anti-idiotypic antibodies are used to detect the level of said defined multispecific antigen-binding molecules in a sample. EXAMPLES

[0120] Example 1: Antibody Generation, Site-Directed Mutagenesis, and Duostatin 3 Conjugation The cloning and production of human HER2 antibody IgG1-153 has been described elsewhere; de ​​Goeij BECGMAbs, 2014.6(2):p.392-402. The heavy and light chain variable domain regions of mouse monoclonal CD63 antibody 2192 (see Table 1 below, SEQ ID NO:1 and 5) were obtained from hybridoma 2.19 (Metzelaar MJVirchows Arch B Cell Pathol Incl Mol Pathol, 1991.61(4):p.269-77) by 5'RACE and sequencing of the variable regions from hybridoma-derived RNA. The variable regions were cloned into the mammalian expression vector pcDNA3.3 (Invitrogen) containing the appropriate human light chain constant domain (codon-optimized, Invitrogen) with the appropriate human heavy chain constant domain mutation (K409R or F405L). The human-mouse chimeric CD63 antibody was called wild type IgG1-CD63. Mutations of the IgG1-CD63 antibody were introduced into the variable domain by either site-directed mutagenesis or direct gene synthesis with the aim of generating a panel of IgG1-CD63 affinity variants. The amino acid mutations are indicated in the antibody name (i.e., anti-CD63-N74H has an asparagine to histidine mutation at amino acid position 74 of the heavy chain (SEQ ID NO:1, Table 1) and anti-CD63-LN54H has an asparagine to histidine mutation at position 54 of the light chain as numbered in SEQ ID NO:5, Table 1). The antibodies were generated by co-transfection and transient expression of heavy and light chain vectors in HEK-293 freestyle cells (Invitrogen) as described by Vink T. Methods, 2014.65(1):p.5-10. Duobodies were generated by regulatory Fab arm exchange as described by Labrijn AFNat Protoc, 2014.9(10):p.2450-63. Human antibody IgG1-b12 specific for HIV gp120 was included as an isotype control (see Parren PWHIAIDS, 1995.9(6):p.F1-6).

[0121] Duostatin 3-conjugated antibodies were generated by covalent conjugation of valine-citrulline-Duostatin 3 (Duo3) at the antibody lysine groups of IgG1-HER2-F405L and IgG1-b12-F405L as described by de Goeij BECG Mol Cancer Ther. 2015.14(5):1130-40. The bispecific ADCs, bsHER2×CD63-Duo3 and bsHER2×b12-Duo3, were generated by Fab arm exchange of Duo3-conjugated antibody IgG1-HER2-F405L-Duo3 with unconjugated IgG1-CD63-K409R or IgG1-b12-K409R. The bispecific ADC, bsCD63xb12-Duo3, was generated by Fab arm exchange of IgG1-b12-F405L-Duo3 with IgG1-CD63-K409R. All bispecific ADCs had a DAR of 1. To generate control ADCs with a DAR of 1, IgG1-HER2-F405L-Duo3 and IgG1-b12-F405L-Duo3 were Fab arm exchanged with IgG1-HER2-K409R and IgG1-b12-K409R to generate IgG1-HER2-Duo3 and IgG1 b12-Duo3, respectively. The DAR of the ADCs was determined by hydrophobic interaction chromatography (HIC).

[0122] Table 1. Heavy chain variable region (VH), light chain variable region (VL), and CDR sequences of anti-CD63 antibody 2192 TIFF0007676100000001.tif212170

[0123] Example 2: CD63 Binding ELISA To improve the efficacy of ADCs, bispecific ADCs were generated in which one Fab arm specifically binds to a target protein (T) and a second Fab arm binds to an effector protein (E) that facilitates internalization and lysosomal delivery of the cytotoxic payload. The resulting bsADC should induce cytotoxicity in cells expressing both T and E. Some cytotoxicity may also be induced in cells expressing T but not E, but in cells expressing E but not T, the bsADC should not induce cytotoxicity. CD63 is used as the effector protein (E). To ensure tumor specificity of the bsAb, the anti-CD63 arm (E) of the bsAb should preferably not bind and internalize in the absence of the tumor-specific arm (T), or do so only to a very limited extent. A panel of IgG1-CD63 variants with mutations in the variable region was generated as described above. The IgG1-CD63 antibody variants were screened for binding to soluble CD63 by ELISA. Briefly, ELISA plates (Greiner) were coated with 0.8 μg / mL goat anti-human IgG (Jackson) overnight at 4 °C. Plates were blocked with 2% chicken serum and incubated with 1 μg / mL histidine mutant variant of anti-CD63 mAb 2192. Serially diluted (1-0.0005 μg / mL) recombinant human CD63 (Creative Biomart) was added, followed by 1 μg / mL mouse anti-polyhistidine biotin (R&D). Reactions were visualized using ABTS and stopped with oxalic acid. Fluorescence at 405 nm was measured and illustrated using GraphPad Prism 6 software.

[0124] Diverse CD63 binding curves were found for the different histidine mutant anti-CD63 antibodies, as can be seen in Figure 1. For some of the affinity variants, binding to CD63 was comparable to wild-type (wt) IgG1-CD63 (referred to as wt IgG1-CD63), whereas others showed partial or complete loss of binding.

[0125] Example 3: CD63 affinity measurements The binding kinetics of anti-CD63 antibodies (Creative BioMart) to recombinant human CD63 second extracellular domain (Ala 103-Val 203) fused to a polyhistidine tag at the C-terminus and a signal peptide at the N-terminus was assessed using label-free biolayer interferometry on an Octet HTX (ForteBio). wt IgG1-CD63 or its affinity variants were immobilized at 1 μg / mL to Anti-Human IgG Fc Capture Biosensors (ForteBio) for 1000 s. The association and dissociation kinetics of human His-tagged CD63 (100 nM, 50 nM, 25 nM, and 12.5 nM, concentrations calculated using a predicted molecular weight of 13 kDa) were determined in Sample Diluent (ForteBio) at 30 °C using an association time of 1000 s, a dissociation time of 2000 s, and a shaker speed of 1000 rpm. Data traces were corrected using a reference sensor exposed only to Sample Diluent during the association and dissociation steps, the Y-axis was baseline aligned, and inter-step correction and Savitzky-Golay filtering were applied. The association rate constant, K on (1 / Ms), dissociation rate constant K dis (1 / s), and the equilibrium dissociation constant K D (M) was determined by ForteBio Data Analysis Software v8.1 using a 1:1 model and a global full fit. K was determined using a dissociation time of 1000 s, except for affinity variants T71, P72, N74, Y121, LV49, and LY51, where a dissociation time of 200 s was used. D was used to calculate.

[0126] 3.6 × 10 for different anti-CD63 antibody variants -10 ~2.7×10 -8A wide range of Ab affinities was measured, spanning M (see FIG. 2 and Table 2). Thus, it was possible to reduce the affinity of wt IgG1-CD63 by introducing a single amino acid histidine substitution.

[0127] (Table 2) TIFF0007676100000002.tif107128

[0128] Example 4: Cytotoxicity of affinity variants of bsCD63xHER2-Duo3 ADC and affinity variants of monovalent bsCD63xb12-Duo3 ADC using HCC1954, SK-OV-3, and Colo205 cells Binding of bsADC to tumor cells expressing both target protein (T) and effector protein (E) should preferentially result in cytotoxicity. However, in the absence of tumor-associated target protein (T), bsADC should preferably not induce cytotoxicity. Using different anti-CD63 affinity variants, multiple bsADCs targeting CD63 (E) and HER2 (T) were generated. The same anti-CD63 affinity variants were also used to generate bsADCs targeting CD63 and HIV gp120. HIV gp120 is a viral protein that is not expressed in the tumor cells tested. Thus, bsADCs targeting CD63 and gp120 (i.e., bsADCs containing binding domains derived from CD63 antibody and gp120-specific antibody IgG1-b12) can only bind to CD63, which reflects the activity of ADCs against normal tissues that lack expression of T.

[0129] The cytotoxicity of bsADC was tested using HCC1954, SK-OV-3, and Colo205 cells. Cells were seeded in 96-well tissue culture plates (5,000 cells / well) and incubated at 37°C for 6 hours. Serially diluted ADC (10-0.0005 μg / mL) was added and cells were incubated at 37°C for 4 days. Cell viability was assessed using CellTiter-GLO (Promega) according to the manufacturer's guidelines. The percentage of viable cells was expressed as a percentage compared to untreated cells (0% cell death) and cells treated with staurosporine (100% cell death). Percentage of viable cells = (RFU of cells treated with ADC - RFU of cells treated with staurosporine) x 100 / (RFU of cells treated with untreated - RFU of cells treated with staurosporine). RFU = relative fluorescence units.

[0130] Figures 3-5 show cell viability after 4 days of treatment with serially diluted ADCs as a percentage compared to untreated cells. Data shown are the mean ± standard deviation of at least two different experiments. IC for cytotoxicity 50 Values ​​were determined using GraphPad Prism 6 software and are illustrated in Table 3.

[0131] (Table 3) IC50 values TIFF0007676100000003.tif131166

[0132] The bsADCs in Figures 3A, 4A, and 5A showed the highest affinity for CD63 (3.6 × 10 -10 ~7.8×10 -10 M range; Table 2) and had low IC when tested as bsCD63×HER2-ADCs. 50 Although the bsADCs in Figures 3B, 4B, and 5B induced cytotoxicity at 0.1×10 β-ADC levels, they also showed some cytotoxicity when tested as bsCD63×b12-ADCs. -9 ~7.3×10 -9M range) and had low IC when tested as bsCD63×HER2-ADCs 50 values ​​and showed limited cytotoxicity when tested as bsCD63xb12-ADC. The antibodies in Figures 3C, 4C, and 5C induced cytotoxicity at low affinity (1.7x10 -8 ~2.7×10 -8 M) and when tested as a bsCD63×HER2-ADC, insufficient IC 50 These antibodies induced cytotoxicity at low antibody levels and showed little cytotoxicity when combined when tested as bsCD63xb12-ADC.

[0133] In conclusion, 2.0 x 10 -9 ~7.3×10 -9 The CD63 antibodies shown in Figures 3B, 4B, and 5B, with affinities spanning M, demonstrated the most favorable properties for enhanced delivery of ADCs. These antibodies were able to induce cytotoxicity with low IC50 values ​​as bsCD63xHER2-ADCs, but induced limited cytotoxicity as bsCD63xb12-ADCs.

[0134] Example 5: Lysosomal co-localization of affinity variants of monovalent bsCD63xb12 ADC To confirm that CD63xb12 bispecific antibodies with reduced affinity for CD63 show less internalization and lysosomal trafficking, confocal microscopy experiments were performed. SK-OV-3 cells were cultured on glass coverslips (Thermo Fisher Scientific) for 16 hours at 37°C. Antibodies (2μg / mL and 10μg / mL) were added and cells were incubated for 16 hours at 37°C. Cells were fixed, permeabilized, and incubated for 45 minutes with goat anti-human IgG1-FITC (Jackson) to stain for human IgG and mouse anti-human CD107a-APC (BD) to stain for lysosomes. Coverslips were mounted on microscope slides (Calbiochem) and imaged with a Leica SPE-II confocal microscope (Leica Microsystems) equipped with LAS-AF software. 12-bit grayscale TIFF images were analyzed for colocalization using MetaMorph® software (Molecular Devices). Colocalization was illustrated as arbitrary units [AU] representing the total pixel intensity of the antibody overlapping with the lysosomal marker LAMP1. To correct for differences in cell density between different images, this value was divided by the total pixel intensity of LAMP1.

[0135] As can be seen in Figure 6, wild type IgG1-CD63 showed the highest colocalization values, significantly higher than its monovalent counterpart (bsCD63 WT This was followed by bsCD63-Y79H×b12) and bsCD63-Y79H×b12. The lysosomal colocalization values ​​for bsP72H×b12, bsY121H×b12, bsLN54H×b12, and bsN74H×b12 were approximately 10-fold lower compared to bsY79H×b12 and wild-type bsCD63×b12. The values ​​for bsV52H×b12, bsG76H×b12, bsLV49H×b12, and bsLY51H×b12 were even lower, indicating that lysosomal trafficking of monovalent CD63 antibodies was almost nonexistent. Asterisks ( * ) were not imaged. Data shown are the mean ± standard deviation of triplicate images.

[0136] Example 6: bsHER2xCD63 with SK-OV-3 cells as determined by flow cytometry N74H Combination of Clone anti-CD63-N74H was selected for further analysis based on its ability to induce limited cytotoxicity as a bsCD63xb12-ADC, but with low IC50 values ​​as a bsCD63xHER2-ADC. N74H The binding of was tested using flow cytometry (FACS Canto II, BD Biosciences). Serially diluted antibodies were incubated with SK-OV-3 cells for 30 min at 4° C. Antibody binding was detected using a phycoerythrin-conjugated goat anti-human IgG antibody (Jackson) and samples were analyzed on a flow cytometer. IgG1-b12 was used as an isotype control antibody. The resulting data shown in FIG. 7 are the average of two experiments.

[0137] As can be seen in Figure 7, bsHER2×CD63 N74H and the monovalent HER2 antibody bsHER2×b12 were identical. N74H This indicates that tumor cell binding of IgG1-CD63 occurs through monovalent binding to HER2. N74H and bsCD63 N74H ×b12 showed no binding to SK-OV-3 cells, consistent with the low expression of CD63 on the cell membrane.

[0138] Example 7: bsHER2xCD63 in whole blood cells N74H and bsCD63 N74H mAb-FITC accumulation assay with ×b12 In healthy tissues that do not express the model tumor antigen HER2, bsHER2×CD63 N74H To demonstrate that bsHER2×CD63 does not bind and accumulate, N74Hand monovalent and bivalent control antibodies were conjugated with FITC. Accumulation was investigated in granulocytes and platelets of healthy donors that do not express HER2. Whole blood samples from healthy donors were collected in heparin tubes. Whole blood was diluted 1:2 with RPMI-1640 supplemented with 10% heat-inactivated cosmic calf serum. Anti-CD63 antibodies were conjugated with FITC (Thermo Scientific) according to the manufacturer's instructions and added to whole blood cells at a final concentration of 10 μg / mL.

[0139] After 1 h incubation at 4°C or 3 and 16 h incubation at 37°C, red blood cells were lysed by incubation with red blood cell lysis buffer (155 mM NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA, pH 7.4) for 15 min at 4°C. FITC fluorescence intensity was measured with a flow cytometer (BD). Granulocytes were gated using mouse anti-human CD66b-PerCP-Cy5.5 (BD) and platelets were gated using mouse anti-human CD62-APC (BD).

[0140] Figure 8 shows that there is little or very low levels of binding of the CD63 antibody to granulocytes or platelets after 1 hour. However, the FITC fluorescence of IgG1-CD63 in granulocytes clearly increases after 16 hours of incubation, indicating accumulation of IgG1-CD63 on granulocytes. In contrast, bsCD63 N74H ×b12 and bsHER2×CD63 N74H FITC fluorescence of bsHER2xb12 showed little increase after 16 hours (see FIG. 8). Similarly, IgG1-HER2 and bsHER2xb12 showed no binding or intracellular accumulation in granulocytes or platelets, consistent with the lack of HER2 expression in these cell types. Thus, by using a low affinity CD63-specific Fab arm, it was possible to minimize the binding and intracellular accumulation of monovalent CD63 Ab in normal cells.

[0141] Example 8: Confocal microscopy, bsHER2xCD63 tracked over time N74H Lysosomal colocalization of bsHER2×CD63 N74H The internalization and lysosomal colocalization of was followed over time. SK-OV-3 cells (20.000) were grown on coverslips (Thermo Fisher Scientific) for 16 h at 37°C. One hour before antibody treatment, cells were preincubated with 50 μg / mL leupeptin (Sigma) to block lysosomal activity. Antibodies (5 μg / mL or 1 μg / mL) were added and cells were incubated for 1, 3, or 16 h at 37°C. Cells were fixed, permeabilized, and incubated for 45 min with goat anti-human IgG1-FITC (Jackson) to stain for human IgG and mouse anti-human CD107a-APC (BD) to stain for lysosomes. Hoechst (Molecular Probes, 1:10.000) was added (5 min at RT) to stain nuclei. Coverslips were mounted on microscope slides (Calbiochem) and imaged with a Leica SPE-II confocal microscope (Leica Microsystems) equipped with LAS-AF software. 12-bit grayscale TIFF images were analyzed for colocalization using MetaMorph® software (Molecular Devices). Colocalization was illustrated as arbitrary units [AU] representing the total pixel intensity of the antibody overlapping with the lysosomal marker LAMP1. To correct for differences in cell density between different images, this value was divided by the total pixel intensity of LAMP1. Total IgG staining was illustrated as the total pixel intensity of FITC divided by the total pixel intensity of LAMP1.

[0142] The grey bars in Figure 9 represent total IgG staining (illustrated as arbitrary units). The black bars in Figure 9 represent lysosomal colocalization (illustrated as arbitrary units). Both IgG1-HER2 and bsHER2xb12 showed similar staining of SK-OV-3 cells after 1, 3, and 16 hours (grey bars). A small fraction of IgG1-HER2 and bsHER2xb12 showed lysosomal colocalization after 16 hours of antibody exposure (black bars). Antibody targeting CD63 IgG1-CD63 N74H showed no staining (gray bars) or lysosomal colocalization (black bars) after 1 and 3 h, whereas 16 h of antibody exposure resulted in Ab staining of cells, all of which colocalized with the lysosomal marker LAMP1. Monovalent CD63 antibody bsCD63 N74H bsHER2×CD63 did not show mAb staining or lysosomal colocalization at any of the time points measured. N74H Lysosomal colocalization of increased gradually over time (black bars). Data shown are the mean ± standard deviation of triplicate images.

[0143] Example 9: bsHER2×CD63 N74H induces HER2 downmodulation Using HER2 downmodulation ELISA, bsHER2×CD63 N74HWe investigated whether the strong lysosomal targeting observed by leads to increased down-modulation of the target antigen. AU565, SK-OV-3, and Colo205 cells were seeded (1 million cells / flask) in T25 flasks (Greiner) and incubated overnight at 37°C to obtain confluent monolayers. Antibodies were added (10 μg / mL) and cells were cultured for an additional 3 days at 37°C, washed, and lysed. Total protein levels were quantified using the bicinchoninic acid (BCA) protein assay reagent (Pierce) according to the manufacturer's instructions. ELISA plates (Greiner) were then coated with 1 μg / mL rabbit anti-human HER2 (Cell Signalling Technology), blocked with 2% chicken serum (Hyclone), and incubated with 50 μL cell lysate. To detect HER2, goat anti-human HER2-biotin (R&D, 50 ng / mL) was added, followed by streptavidin-poly-HRP (Sanquin, 100 ng / mL). The reaction was visualized using ABTS and stopped with oxalic acid. Fluorescence was measured at 405 nm, and the amount of HER2 was expressed as a percentage of untreated cells.

[0144] The total amount of HER2 protein in tumor cell lines with different expression levels of HER2; AU565 (500,000 HER2 / cell, FIG. 10A), SK-OV-3 (200,000 HER2 / cell, FIG. 10B), and Colo205 (50,000 HER2 / cell, FIG. 10C) was quantified after 3 days of incubation with HER2 antibodies and compared to untreated cells (see FIG. 10). IgG1-HER2 induced approximately 40% down-modulation of total HER2 in AU565 cells expressing high levels of HER2. Despite the fact that monovalent bsHER2×b12 antibody showed dose-dependent binding to HER2-positive SK-OV-3 cells (FACS binding example), no down-modulation of HER2 by bsHER2×b12 was observed. This highlights that bivalent antibody binding was important to increase HER2 degradation. bsHER2×CD63 N74H was able to restore HER2 downmodulation in AU565 cells. Moreover, bsHER2×CD63 was also able to restore HER2 downmodulation in cell lines with lower HER2 expression, such as SK-OV-3 and Colo205. N74H induced down-modulation of HER2, whereas IgG1-HER2 had no effect on HER2 protein levels.

[0145] Example 10: Cytotoxicity Induced by ADC Conjugated with Duostatin 3 Cells were seeded in 96-well tissue culture plates (5,000 cells / well) and allowed to attach for 6 h at 37°C. Serially diluted ADCs (10-0.0005 μg / mL) were added and cells were incubated for an additional 3 days at 37°C. Cell viability was assessed using CellTiter-GLO (Promega) according to the manufacturer's guidelines. The percentage of viable cells was plotted as a percentage of untreated cells.

[0146] As can be seen in Figure 11, IgG1-HER2-Duo3 was able to kill approximately 80% of HCC1954 cells that exhibit high HER2 expression (500.000 HER2 / cell). In SK-OV-3 cells expressing 200.000 HER2 / cell, IgG1-HER2-Duo3 killed only approximately 30% of the cells, whereas the viability of low HER2 expressing Colo205 cells (50.000 HER2 / cell) was not affected.

[0147] Monovalent bsHER2×b12 killed a similar percentage of cells compared to IgG1-HER2-Duo3, but with approximately 10-fold reduced IC 50 The bsHER2×CD63 expression in HCC1954 cells was N74H The cytotoxicity induced by bsHER2×CD63 was equivalent to that of IgG1-HER2-Duo3. However, in cells with lower copy numbers of HER2 (SK-OV-3, and to a lesser extent Colo205), N74H -Duo3 induced much greater cytotoxicity compared to ADCs targeting HER2 alone (see Figure 11). Data shown are the mean ± standard deviation of at least two separate experiments.

[0148] Example 11: bsHER2xCD63 on SK-OV-3 tumor xenografts N74H -Antitumor effect of ADC bsHER2×CD63 N74H The antitumor efficacy of the -ADC was investigated in SK-OV-3 tumor xenografts. Female SCID mice (CB-17 / IcrPrkdc-scid / CRL) aged 6-11 weeks were purchased from Charles River. Mice were inoculated with 5 × 10 6 Subcutaneous tumors were induced by inoculating SK-OV-3 cells. Tumor volume was calculated as 0.52 × length × width. 2 (mm 3 ) was calculated from digital caliper measurements. 3When tumor size reached 100%, mice were sorted into groups of 7 mice with equal tumor size distribution and intraperitoneally injected with mAb (8 mg / kg). During the study, blood samples were collected in heparin-containing tubes to confirm the presence of human IgG in the plasma. IgG levels were quantified using a nephelometer (Siemens Healthcare). Mice that did not show human IgG in the plasma were excluded from the analysis.

[0149] As shown in Figure 12, bsHER2×CD63 N74H -ADC induced significant inhibition of tumor growth, whereas monovalent bsHER2×b12-ADC or bsCD63 N74H bsHER2×CD63×b12-duo3 had no effect on tumor growth, demonstrating that low affinity CD63-specific Fab arms can be used to induce lysosomal delivery and toxin release of poorly internalizing ADCs in tumors in vivo. Mantel-Cox analysis of Kaplan-Meier plots showed that bsHER2×CD63×b12-duo3 had no effect on tumor growth, demonstrating that low affinity CD63-specific Fab arms can be used to induce lysosomal delivery and toxin release of poorly internalizing ADCs in tumors in vivo. N74H - ADC showed significant inhibition of tumor growth, P value < 0.0001.

[0150] Example 12: bsβ1×CD63 with SK-OV-3 cells detected by flow cytometry N74H Combination of We investigated whether the low affinity binding domain for E could be used to enhance the internalization and lysosomal targeting of other tumor antigens as well. Integrins have been described to rely on clustering for internalization. Therefore, monovalent integrin antibodies are expected to show minimal internalization and lysosomal targeting, and therefore may represent a suitable model system to test whether internalization can be enhanced in a bispecific format that targets T and E. For this purpose, we selected the antibody huK20, which targets integrin β1. The sequence of antibody huK20 was obtained from WO1996 / 008564 and cloned and produced as described in Example 1 therein.

[0151] IgG1-β1 antibody, monovalent control bsβ1×b12, and bispecific antibody bsβ1×CD63 with SK-OV-3 N74H The binding of was investigated using flow cytometry (FACS Canto II, BD Biosciences). Serially diluted antibodies were incubated with SK-OV-3 cells for 30 min at 4°C. Antibody binding was then detected using a phycoerythrin-conjugated goat anti-human IgG antibody (Jackson) and samples were analyzed on a flow cytometer. IgG1-b12 was used as an isotype control antibody.

[0152] As can be seen in Figure 13, bsβ1×CD63 N74H The binding curves of bsβ1×CD63 and the monovalent integrin β1 antibody bsβ1×b12 were very similar. N74H These results indicate that tumor cell binding of IgG1-CD63 occurs through monovalent binding to integrin β1. N74H and bsCD63 N74H ×b12 showed no binding to SK-OV-3 cells, consistent with the low expression of CD63 on the cell membrane.

[0153] Example 13: bsβ1×CD63 measured by confocal microscopy N74H Lysosomal colocalization of Dual targeting of integrin β1 and CD63 is bsβ1×CD63 N74HTo investigate whether integrin β1 leads to increased lysosomal colocalization of β1, confocal microscopy experiments were performed with tumor cell lines carrying different copy numbers of integrin β1 on the cell membrane. 20.000 SK-OV-3, NCI-H1975, and MDA-MB-468 cells were grown on glass coverslips (Thermo Fisher Scientific) for 4 h at 37°C. One hour before antibody treatment, cells were preincubated with 50 μg / mL leupeptin (Sigma) to block lysosomal activity. Antibodies (2 μg / mL, 0.4 μg / mL, and 0.08 μg / mL) were added and cells were incubated for 16 h at 37°C. Cells were fixed, permeabilized, and incubated for 45 min with goat anti-human IgG1-FITC (Jackson) to stain for human IgG and mouse anti-human CD107a-APC (BD) to stain for lysosomes. Hoechst (Molecular Probes, 1:10.000) was added to stain the nuclei (5 min at RT). Coverslips were mounted on microscope slides (Calbiochem) and imaged by a Leica SPE-II confocal microscope (Leica Microsystems) equipped with LAS-AF software. 12-bit grayscale TIFF images were analyzed for colocalization using MetaMorph® software (Molecular Devices). Colocalization was illustrated as arbitrary units [AU] representing the total pixel intensity of the antibody overlapping with the lysosomal marker LAMP1. To correct for differences in cell density between different images, this value was divided by the total pixel intensity of LAMP1.

[0154] As can be seen in Figure 14, bsβ1×CD63 N74H showed the strongest amount of lysosomal colocalization across all tested cell lines and mAb concentrations (only shown for SK-OV-3). IgG1-β1 and bsAb-β1×b12 showed moderate lysosomal colocalization that was not affected by mAb concentration. IgG1-CD63 N74Hshowed substantial lysosomal colocalization in NCI-H1975 cells only at 2 and 0.4 μg / mL. The monovalent control bsCD63 N74H ×b12 showed lysosomal colocalization only at 2 μg / mL in NCI-H1975 cells, which was consistent with CD63 N74H This correlated with the decreased affinity of bsβ1×CD63 N74H Increased lysosomal colocalization of was most evident in cells expressing high integrin β1 copy numbers (SK-OV-3 > NCI-H1975 > MDA-MB-468). For some clones, lysosomal colocalization was not measured, so AUs are not shown. Data shown are the mean ± standard deviation of triplicate images.

[0155] Example 14: Confocal microscopy, bsβ1×CD63 tracked over time N74H Internalization and lysosomal colocalization of bsβ1×CD63 N74H To better understand the kinetics of internalization and lysosomal colocalization of bsβ1×CD63 N74HThe internalization and lysosomal colocalization of was followed over time. SK-OV-3 cells (20.000 cells) were grown on glass coverslips (Thermo Fisher Scientific) for 16 h at 37 °C. One hour before antibody treatment, cells were preincubated with 50 μg / mL leupeptin (Sigma) to block lysosomal activity. Antibodies (2 μg / mL) were added and cells were incubated for 1, 3, or 16 h at 37 °C. Cells were fixed, permeabilized, and incubated for 45 min with goat anti-human IgG1-FITC (Jackson) to stain for human IgG and mouse anti-human CD107a-APC (BD) to stain for lysosomes. Hoechst (Molecular Probes, 1:10.000) was added to stain the nuclei (5 min at RT). Coverslips were mounted on microscope slides (Calbiochem) and imaged with a Leica SPE-II confocal microscope (Leica Microsystems) equipped with LAS-AF software. 12-bit grayscale TIFF images were analyzed for colocalization using MetaMorph® software (Molecular Devices). Colocalization was illustrated as arbitrary units [AU] representing the total pixel intensity of the antibody overlapping with the lysosomal marker LAMP1, divided by the total pixel intensity of LAMP1. Total IgG staining was illustrated as the total pixel intensity of FITC, divided by the total pixel intensity of LAMP1.

[0156] The grey bars in Figure 15 represent total IgG staining (illustrated as arbitrary units). The black bars in Figure 15 represent lysosomal colocalization (illustrated as arbitrary units). Both IgG1-β1 and bsβ1×b12 showed similar staining of SK-OV-3 cells after 1 hour, 3 hours, and 16 hours (grey bars), but little lysosomal colocalization was measured (black bars). Thus, IgG1-β1 and bsβ1×b12 were able to bind to SK-OV-3 cells but were not transported to lysosomes. Antibody targeting CD63, IgG1-CD63 N74H and bsCD63 N74H×b12 showed no staining (gray bars) or lysosomal colocalization (black bars) after 1 hour. However, extended incubation resulted in Ab staining of cells, all of which colocalized with the lysosomal marker LAMP1, indicating that both antibodies were immediately transported to lysosomes. This effect was not observed with IgG1-CD63 N74H This is most evident for bsCD63 N74H The results were relatively unremarkable for β1×CD63. N74H showed equal staining of SK-OV-3 cells after 1, 3, and 16 hours (gray bars). Lysosomal colocalization gradually increased over time (black bars), indicating that β1 × CD63 N74H was shown to initially bind to tumor cells through integrin β-1 and then be transported to lysosomes. Data shown are the mean ± standard deviation of at least three images.

Claims

1. A multispecific antibody comprising a first antigen-binding domain and a second antigen-binding domain, wherein the first antigen-binding domain specifically binds to a target molecule (T) that is a cell surface-expressed target protein or polypeptide, and the second antigen-binding domain specifically binds to an internalizing effector protein (E), where E is CD63 and the second antigen-binding domain has a dissociation constant K Ｄ Value 2.0 x 10 -9 ~7.3 x 10 -9 M, K Ｄ A multispecific antibody, the value of which is measured by biolayer interferometry.

2. 2. The multispecific antibody of claim 1, comprising (i) a first binding arm comprising the first antigen-binding domain, and (ii) a second binding arm comprising the second antigen-binding domain.

3. 3. The multispecific antibody of claim 1 or 2, which is a bispecific antibody.

4. The multispecific antibody according to any one of claims 1 to 3, wherein E is a molecule expressed on the cell surface, which molecule is internalized into the cell.

5. The multispecific antibody of claim 4, which is internalized into the cell by binding to E only in the presence of the target molecule (T).

6. 6. The multispecific antibody of claim 4 or 5, wherein the multispecific antibody is internalized into the cell by binding to E only if the first antigen-binding domain specifically binds to the target molecule (T).

7. The multispecific antibody of any one of claims 1 to 6, which upon binding to E is internalized more efficiently into cells expressing T compared to cells that do not express T.

8. The multispecific antibody of any one of claims 1 to 7, which upon binding to T is internalized more efficiently into cells expressing E compared to cells not expressing E.

9. The multispecific antibody of any one of claims 1 to 8, which upon binding to E is transported to the lysosomal compartment in cells expressing T.

10. The multispecific antibody of any one of claims 1 to 9, which, upon binding to E, is transported to the lysosomal compartment in cells expressing T more efficiently compared to cells not expressing T.

11. The multispecific antibody of any one of claims 1 to 10, which, upon binding to T, is transported to the lysosomal compartment in cells expressing E more efficiently compared to cells not expressing E.

12. The multispecific antibody according to any one of claims 1 to 11, wherein T is a target molecule expressed on a cell surface.

13. The multispecific antibody according to any one of claims 1 to 12, wherein T is a tumor associated antigen.

14. The multispecific antibody of any one of claims 1 to 13, wherein T is HER2.

15. The multispecific antibody of any one of claims 1 to 14, wherein the first antigen-binding domain and / or the second antigen-binding domain comprises at least one antibody variable region.

16. 16. The multispecific antibody according to any one of claims 1 to 15, which is a multispecific, preferably a bispecific, antibody fragment or a recombinantly modified part thereof, which is a multispecific, preferably bispecific, antibody fragment or recombinantly modified part thereof.

17. 17. The multispecific antibody of any one of claims 1 to 16, which is a bispecific antibody comprising a first binding arm comprising the first antigen-binding domain and a second binding arm comprising the second antigen-binding domain.

18. 18. The multispecific antibody of claim 17, wherein the first antigen-binding domain comprises a first heavy chain variable sequence (VH) and a first light chain variable sequence (VL), and the second antigen-binding domain comprises a second heavy chain variable sequence (VH) and a second light chain variable sequence (VL), and the variable sequences each comprise three CDR sequences CDR1, CDR2, and CDR3.

19. 19. The multispecific antibody of claim 17 or 18, wherein (i) the first binding arm comprises a first heavy chain comprising a first heavy chain variable sequence (VH) and a first heavy chain constant sequence (CH), and a first light chain comprising a first light chain variable sequence (VL) and a first light chain constant sequence (CL), and (ii) the second binding arm comprises a second heavy chain comprising a second heavy chain variable sequence (VH) and a second heavy chain constant sequence (CH), and a second light chain comprising a second light chain variable sequence (VL) and a second light chain constant sequence (CL).

20. The multispecific antibody of any one of claims 17 to 19, wherein the first binding arm is derived from a chimeric antibody, a humanized antibody or a human antibody.

21. The multispecific antibody according to any one of claims 17 to 20, wherein the second binding arm is derived from a chimeric antibody or from a humanized antibody or from a human antibody.

22. The multispecific antibody according to any one of claims 1 to 21, which is a bispecific antibody, wherein said bispecific antibody is a full length antibody, preferably an IgG1 antibody.

23. The multispecific antibody of any one of claims 1 to 22, wherein the second antigen-binding domain comprises one or more mutations that modulate the affinity of the second antigen-binding domain to E.

24. The multispecific antibody of claim 23, wherein the second antigen-binding domain is derived from an antibody having one or more mutations in VH and / or VL that modulate the affinity of the second antigen-binding domain to E.

25. 25. The multispecific antibody according to claim 24, wherein the mutation is a single amino acid substitution, preferably a single histidine amino acid substitution.

26. the second antigen-binding domain comprising: a. VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 3, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 9, 7, and 8, respectively; or b. VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 10, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively; or c. VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 11, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively; or d. VH CDRs 1, 2, and 3 as provided in SEQ ID NOs: 2, 12, and 4, respectively, and VL CDRs 1, 2, and 3 as provided in SEQ ID NOs: 6, 7, and 8, respectively. The multispecific antibody according to any one of claims 1 to 25, comprising:

27. 27. The multispecific antibody of claim 26, wherein the second antigen-binding domain comprises VH CDRs 1, 2 and 3 as provided in SEQ ID NOs: 2, 12 and 4, respectively, and VL CDRs 1, 2 and 3 as provided in SEQ ID NOs: 6, 7 and 8, respectively.

28. 28. The multispecific antibody of any one of claims 1 to 27, wherein the first antigen-binding domain and the second antigen-binding domain are each a pair of an antibody heavy chain variable domain and an antibody light chain variable domain.

29. The multispecific antibody according to any one of claims 1 to 28, which is a tumor-associated target (T) x CD63 bispecific antibody.

30. The multispecific antibody of any one of claims 1 to 29, which is a HER2xCD63 bispecific antibody.

31. 31. The multispecific antibody according to any one of claims 1 to 30, having an EC50 value, determined by flow cytometry, for binding to cells expressing a tumor-associated target (T), such as cells expressing HER2, of less than 5.0 μg / ml, such as less than 0.5 μg / ml.

32. K Ｄ The multispecific antibody according to any one of claims 1 to 31, wherein is determined by biolayer interferometry at 30°C.

33. (i) a first binding arm comprising a first heavy chain comprising a first heavy chain constant sequence (CH), the first CH comprising a first CH3 region; (ii) a second binding arm comprising a second heavy chain comprising a second heavy chain constant sequence (CH), the second CH comprising a second CH3 region; and a bispecific antibody comprising 33. The multispecific antibody according to any one of claims 1 to 32, wherein the sequences of the first CH3 region and the second CH3 region are different and are such that a heterodimeric interaction between the first binding arm and the second binding arm is stronger than a homodimeric interaction of each of the first binding arm and the second binding arm.

34. 34. The multispecific antibody of claim 33, wherein in the first heavy chain CH3 region at least one of the amino acids at a position corresponding to positions T366, L368, K370, D399, F405, Y407 or K409 of a human IgG1 heavy chain is substituted and in the second heavy chain CH3 region at least one of the amino acids at a position corresponding to positions T366, L368, K370, D399, F405, Y407 or K409 of a human IgG1 heavy chain is substituted, and wherein the first heavy chain and the second heavy chain are not substituted in the same positions, and the amino acid positions are numbered according to the EU index.

35. 35. The multispecific antibody of claim 34, wherein (i) the first CH3 region has a F405L substitution and the second CH3 region has a K409R substitution, or (ii) the first CH3 region has a K409R substitution and the second CH3 region has a F405L substitution.

36. The multispecific antibody according to any one of claims 1 to 35, which is conjugated to a cytotoxic moiety, to a radioisotope or to a drug.

37. 37. The multispecific antibody of claim 36, wherein the cytotoxic moiety is selected from the group consisting of maytansine, calicheamicin, duocarmycin, duostatin, duostatin 3, duostatin 5, rachelmycin (CC-1065), auristatin, monomethyl auristatin E, monomethyl auristatin F, doxorubicin, dolastatin, pyrrolobenzodiazepines, IGN-based toxins, alpha-amanitin, or an analog, derivative, or prodrug of any of them.

38. The multispecific antibody of any one of claims 1 to 37, wherein binding of T and E by the multispecific antibody induces internalization of the multispecific antibody to a greater extent than binding of T alone.

39. 39. The multispecific antibody of any one of claims 1 to 38, wherein the antibody is a bispecific antibody selected from tandem scFv, tandem scFv-Fc, scFv-Fc knob-into-hole, scFv-Fc-scFv, F(ab')2, Fab-scFv, (Fab'scFv)2, diabody, sc diabody, sc diabody-Fc, sc diabody-CH3, or an azymetric scaffold.

40. A multispecific antibody according to any one of claims 1 to 39 for use in a method for treating and / or preventing cancer.

41. 41. The multispecific antibody for use according to claim 40, wherein the cancer is endometrial / cervical cancer, lung cancer, malignant melanoma, ovarian cancer, pancreatic cancer, prostate cancer, testicular cancer, soft tissue tumour such as synovial sarcoma, breast cancer, brain cancer, leukemia, lymphoma, mast cell tumor, kidney cancer, cervical cancer, bladder cancer, esophageal cancer, gastric cancer or colorectal cancer.

42. 40. The multispecific antibody of any one of claims 1 to 39 for use in a method of targeting a tumor in a subject comprising administering said multispecific antibody to said subject.

43. A pharmaceutical composition comprising as an active ingredient a multispecific antibody according to any one of claims 1 to 38.

44. A nucleic acid encoding a multispecific antibody according to any one of claims 1 to 38.

45. 45. An expression vector containing the nucleic acid of claim 44, capable of expressing said nucleic acid in a prokaryotic or eukaryotic host cell line.

46. 46. ​​A prokaryotic or eukaryotic host cell line comprising the vector of claim 45.

Images