EA040393B1 - BISPECIFIC IgG ANTIBODIES AS T-CELL RECRUITERS - Google Patents

Description

Technical field

The present invention relates to the field of antibody engineering. In particular, the invention relates to the field of therapeutic (human) antibodies for the treatment of diseases associated with abnormal cells. More specifically, the invention relates to bispecific antibodies for the treatment of tumors.

Prerequisites for the creation of the invention

In laboratories, bispecific antibodies have been widely used to retarget immunocompetent effector cells to tumor cells. In this case, one binding site is directed against a tumor associated antigen (TAA) and the second site is directed against a trigger molecule on effector cells, such as, for example, CD3 on T cells (Kontermann, MABS 2012 (4) 182-197; Chames and Baty, MABS 2009 (1) 539-547 Moore et al Blood 2011 (117) 4542-4551). The first bispecific antibodies targeting CD3 and tumor cell-associated antigen were obtained from rodents and were generated using hybridomas (Liu et al. 1985 PNAS 82: 8648, Staerz et al. 1986 PNAS 83: 1453, Lanzavecchia et al. 1987, Eur.J.Imm.17:105). In the case of these hybrid hybridomas, the recombination of Ig heavy and light chains resulted in the creation of bispecific functional antibody molecules within a much larger pool of monospecific and non-functional bispecific antibodies resulting from heavy and light chain mismatches. Due to their dual specificity, these functional bispecific antibodies were able to bind murine and human cytotoxic T-lymphocytes (CTLs) (CTL) to target cells and include a cytotoxic function leading to the lysis of tumor cells presenting the corresponding antigen on their surface. However, CD3xTAA-mediated bispecific IgG induction of tumor cell lysis by polyclonal human resting T cells could not be achieved unless costimulation was provided by added exogenous IL-2 or an anti-CD28 mAb. This can be illustrated by the hybrid rat/mouse IgG2b CD3xCD19 bispecific molecule that was able to induce lysis of the CD19-positive REH B-ALL tumor cell line by resting human T-lymphocytes only after co-administration of IL-2 (Haagen et al. 1995 Blood 85: 3208). Zeidler et al. demonstrated, using a similar hybrid rat IgG2b/mouse IgG2a bispecific CD3xEpcam molecule, that bispecific IgG-induced lysis of Epsat-positive tumor cells can be achieved in mixed cell cultures including both peripheral blood mononuclear cells (PBMC) and tumor cells. , without the addition of exogenous IL-2 (Zeidler et al. 1999 J. Immunol. 163: 1246). The authors argued that the third arm of the antibody, the Fc region, elicits this effect by interacting with the Fcy receptor-positive helper cells present in the PBMC fraction. In particular, the ability to strongly activate was correlated with a hybrid combination of mouse IgG2a/rat IgG2b subclasses, which, unlike other known combinations (e.g. mouse IgG2a/mouse IgG1 or rat IgG2b/mouse IgG1), not only binds but also activates positive helper cells on the Fcyreceptor. This so-called triomat CD3xEpcam bispecific antibody, also known as catumaxomab, has been clinically tested and registered in Europe for the palliative treatment of abdominal tumors of epithelial origin. Although this bispecific antibody has clearly shown clinical efficacy, its rodent origin elicits immune responses against the product upon repeated dosing and therefore hinders the widespread use of this format.

Alternative CD3xTAA formats have been explored to address both the production-related and immunogenicity problems associated with the rodent hybrid triomat format. Such formats are often immunoglobulin-like molecules that differ from full-length human IgG molecules, and include molecules such as dual affinity retargeting molecules (DART™), which are being developed by MacroGenics, on the worldwide web at macrogenics.com/Platforms-DART.html , bispecific T-cell recruiters (BiTE®), which were developed by Micromet, are now Amgen (Sheridan C, Nat Biotechnol. 2012 (30): 300-1), double variable domain-immunoglobulin (DVD-Ig™) molecules, which are being developed Abbott, and Affimed's TandAb® RECRUIT molecules, on the worldwide web at affimed.com/tandab-recruit. In one of these formats, it was found that successful retargeting of peripheral blood lymphocytes to lyse CD19-positive tumor cells using CD3xCD19 diabody required prior activation of peripheral blood T-lymphocytes, now using an antibody against human CD3 plus IL-2 (Kipriyanov et al., 1998 Int. J. Can 77: 763). Other formats, such as the CD3xTAA format of BiTE®, a bivalent single-stranded Fv molecule (Loffler et al. 2000 Blood 95: 2098), do not require prior activation of resting T cells and are able to induce lysis of antigen-positive tumor cells in a very efficient manner ( Dreier et al 2002 Int J Cane 100: 690). Additional studies using BiTE® targeting various TAAs showed that the greater efficiency of the BiTE® format correlated with the size of the antigen and, in particular, with the distance from the TAA epitope to the tumor cell membrane (Bluemel et al.

- 1 040393

2010 Cancer Immunol. Immunother 59:1197). In the case of BiTE® molecules, the efficient formation of cytolytic synapses with T cells was found to be the structural basis of their efficiency (Offner et al. Molecular Immunology 2006 (43) 763-771), which is also believed to be related to the small size of the BiTE® format. If size matters, this may mean that larger molecules, such as intact IgG, will be too large to form effective cytolytic synapses. CD3xCD19 BiTE®, blinatumomab, has demonstrated remarkable clinical efficacy in patients with resistant non-Hodgkin's lymphoma and acute lymphocytic leukemia (Bargou et al. 2008 Science 321: 974). Although CD3xCD19 BiTE® demonstrates a very efficient lysis of tumor cells at low levels in vitro, administration of this bispecific format to patients is associated with significant problems. Due to their small size, BiTE® are rapidly cleared from the bloodstream and thus require continuous infusion for dosing to patients. Since the total duration of the dosing regimen is more than 2 months, this treatment has a significant impact on the quality of life of patients.

Thus, there remains a need for full-length, bispecific, T-cell-attracting IgG molecules effective in killing abnormal cells that combine a long circulating half-life after intravenous administration without the need for continuous infusion, while not being immunogenic, with only limited side effects. .

Brief description of the drawings

Fig. 1 - CLEC12A and related sequences.

Fig. 2 - T cell activation with various antibodies: anti-CD3 monoclonal bivalent IgG, CD3XCLEC12 bispecific IgG, CD3x bispecific IgG isotype control, anti-CLEC12A monoclonal bivalent IgG, monoclonal bivalent, isotype-matched control IgG.

Fig. 3 - Specific lysis of HL60 cells with CD3XCLEC12A bispecific IgG and control antibodies.

Fig. 4 - Specific lysis of HL60 cells with CD3XCLEC12A bispecific IgG and control antibodies (E:T ratios).

Fig. 5 - Specific lysis of HL60 cells with several CD3XCLEC12A bispecific IgG molecules, consisting of different CLEC12A-specific arms and an unchanged CD3-specific arm, and control antibodies.

Fig. 6 - Specific lysis of HL60 cells with CD3XCLEC12A bispecific IgG coupled with Fc silencing (DM=double mutant; TM=triple mutant; WT=wild type, no Fc silencing).

Fig. 7 - Fc-suppression does not affect binding to FcRn.

Fig. 8 - CD3xCLEC12A target-specific bsAb induction of T cell proliferation.

Fig. 9 - CD8+ T cell fraction of AML patients compared to healthy donors.

Fig. 10 - Specific T cell activation induced by CD3xCLEC12A DM-Fc and lysis of HL60 tumor cells by T cells from AML patients.

Fig. 11 - Specific lysis of AML blast cells with autologous T cells from AML patients.

Fig. 12 - specific lysis of monocytes by T-cells of patients.

Fig. 13 - Fc-suppression significantly eliminates the release of cytokines by observer cells.

Fig. 14A - FACS staining, anti-CD3 antibodies on HPBALL cells.

Fig. 14B - Plate-bound IgG, CFSE labeled T cells, day 5 FACS readout.

Fig. 15 - analysis of cytotoxicity in relation to HL60.

Fig. 16 - FACS staining, anti-CLEC12A antibodies on HL60 cells.

Fig. 17 - analysis of cytotoxicity in relation to HL60.

Fig. 18 - analysis using FACS.

Fig. 19 - analysis of cytotoxicity in relation to HL60.

Fig. 20 - VH sequences of CD3-specific and CLEC12A-specific Fab arms. The VL sequence of the common human light chain. The CDR sequences are in bold and underlined.

Brief summary of the invention

The present invention describes a fully human, bispecific full length IgG antibody for the treatment of AML. One arm of the antibody binds to an epitope of immunocompetent effector cells, preferably CD3, while the target of the other arm is CLEC12A, a myeloid cell-specific target on the surface, which is present in 90-95% of patients.

- 2,040,393 patients with de novo AML and recurrent AML. CLEC12A is present on AML leukemic stem cells but not on normal hematopoietic cells. Unlike CD33, CLEC12A is not present on erythroid progenitors or megakaryocytes, thus the CD3xCLEC12A IgG1 subclass bispecific antibody of the present invention should not cause platelet or erythrocyte depletion. Experiments with colonies of bone marrow cells have shown that depletion of CLEC12A+ cells in normal bone marrow does not affect myeloid lines that give rise to platelets and red blood cells. The CD3xCLEC12A IgG1 bispecific antibody of the present invention in a preferred embodiment comprises a modified Fc region to reduce non-specific activation of the immune system by recruiting T cells and FcyR expressing cells within the PBMC. Based on the data presented for the triomat bispecific antibody in the prior art, it was highly doubtful that the CD3xTAA bispecific IgG of the fully human IgG1 format would be able to induce lytic antitumor activity in resting peripheral blood lymphocytes without the need for prior activation of T cells. In addition, data available for the BiTE® format suggested that a full-length IgG molecule would be too large to create effective cytolytic synapses between tumor cells and effector cells. Surprisingly, the present inventors have shown that a fully human, CD3xCLEC12A bispecific, full-length IgG1 is capable of inducing a very efficient, T-cell mediated lysis of CLEC12A-positive HL60 AML tumor cells in vitro. In fact, efficient lysis was mediated by resting T-lymphocytes purified from PBMC without the need for pre-activation of T cells. In addition, the present inventors have shown that this lytic activity is not necessarily dependent on interactions with FcyR present on HL60 cells, as this lytic activity is not affected by the presence of excess human IgG when assayed in media containing human serum. This is the first time that a bispecific, T-cell-attracting full-length IgG1 antibody induces efficient lysis of tumor cells without the need for pre-activation of T cells or the need for active interactions with FcyR. Efficient lysis is achieved despite the relatively large size of IgG1 compared to BiTE® molecules. Notably, if CH2/lower hinge mutations were introduced into the CD3xCLEC12A bispecific IgG1 molecule to further reduce interactions with the Fc receptor, this still resulted in efficient lysis of tumor cells by immunocompetent effector cells. A bispecific, T-cell-attracting antibody of the human IgG1 subclass has advantages over existing IgG that uses a hybrid combination of mouse IgG2a/rat IgG2b subclasses because human IgG1 will be less immunogenic and can thus be used for retreatment. In addition, a full-length, bispecific, T-cell-attracting antibody of the human IgG1 subclass has advantages over immunoglobulin-like molecules such as DART™, TandAb®, or BiTE®, because full-length human IgG1 is not rapidly cleared from the circulation, and for dose administration to patients, thus, continuous infusion will not be required, which is more beneficial for patients.

Embodiments

The present invention relates to an IgG class bispecific antibody, said IgG class bispecific antibody comprising one arm that specifically recognizes CLEC12A or a functional equivalent thereof, and a second arm that specifically recognizes an antigen of immunocompetent effector cells capable of recruiting such cells to abnormal cells, expressing CLEC12A or the specified functional equivalent.

As used herein, the term specifically recognizes CLEC12A or a functional equivalent means that said arm is capable of specifically recognizing CLEC12A or said functional equivalent when CLEC12A or said functional equivalent is present in the vicinity of said antibody. Similarly, the term specifically recognizes an antigen of immunocompetent effector cells means that said arm has the ability to specifically recognize said antigen when said antigen is present in the vicinity of said antibody. This recognition of an antigen by an antibody is typically mediated by regions of complementarity of the antibody and the specific three-dimensional structure of both the antigen and the antibody arm, allowing the two structures to bind together precisely (a lock and key interaction), as opposed to random, non-specific antibody adhesion. Since an antibody typically recognizes an epitope of an antigen, and since such an epitope may also be present in other compounds, antibodies of the present invention that specifically recognize CLEC12A or a functional equivalent thereof and specifically recognize an immunocompetent effector cell antigen may also recognize other compounds if such other compounds contain the same kind of epitope. Therefore, the terms specifically recognize CLEC12A or its functional equivalent, specifically recognize the antigen of immunocompetent effector cells, and

- 3 040393 specifically recognizes CD3 does not exclude the binding of antibodies to other compounds that contain the same (type) epitope(a). Instead, cross-reactivity is allowed. An antibody of the present invention is generally capable of binding to CLEC12A (or a functional equivalent thereof) and immunocompetent effector cell antigen, preferably CD3, with an affinity of at least 1x10'5 M, as described in more detail below.

As used herein, the term antibody means a protein molecule belonging to the class of immunoglobulin proteins containing one or more domains that bind to an antigen epitope, such domains being derived from or sharing homology with an antibody variable region. Preferably, the antibodies for therapeutic use are as close as possible to the natural antibodies of the subject being treated (eg, human antibodies for human subjects). Antibody binding can be represented in terms of specificity and affinity. Specificity determines which antigen or epitope thereof will be specifically bound by the binding domain. Affinity is a measure of the strength of binding to a particular antigen or epitope. Specific binding, or specifically recognizing, is defined as binding with affinities (KD) of at least 1x10'5 M, more preferably 1x10' ⁷ M, more preferably greater than 1x10' ⁹ M. In general, antibodies for therapeutic applications have affinities components up to 1x10' ¹⁰ M or even more. The antibodies of the present invention are typically bispecific full-length antibodies of the human IgG subclass. Preferably, the antibodies of the present invention are antibodies of the human IgG1 subclass.

The term full-length IgG of the present invention is defined as including essentially a whole IgG, which, however, does not necessarily have all the functions of an intact IgG. For the avoidance of doubt, full length IgG contains two heavy and two light chains. Each chain contains constant (C) and variable (V) regions, which can be divided into domains named CH1, CH2, CH3, VH, and CL, VL. An IgG class antibody binds to an antigen via the variable domains contained in the Fab portion and, once bound, can interact with molecules and cells of the immune system via constant domains, primarily via the Fc portion. The terms variable domain, variable region, variable domain, VH/VL pair, VH/VL, Fab portion, Fab arm, Fab or arm are used interchangeably herein. Full-length antibodies of the present invention include IgG molecules, and mutations may be present that provide the desired characteristics. Such mutations must not be deletions of significant portions of any of the regions. However, IgG molecules in which one or more amino acid residues are deleted without significantly altering the binding characteristics of the resulting IgG molecule are included in the term full-length IgG. For example, such IgG molecules may have one or more deletions of 1-10 amino acid residues, preferably in non-CDR regions, where the deleted amino acids are not important for IgG binding specificity.

Full-length IgG antibodies are preferred because of their appropriate half-life and the need to stay as close as possible to fully autologous (human) molecules for reasons of immunogenicity. The present invention uses bispecific antibodies of the IgG class. In a preferred embodiment, bispecific full-length antibodies of the IgG1 subclass are used. IgG1 is preferred based on its long circulating half-life in humans. To prevent any immunogenicity in humans, it is preferred that the IgG class bispecific antibody of the present invention is human IgG1. The term bispecific (bs) means that one arm of the antibody binds to the first antigen, while the second arm binds to the second antigen, and said first and second antigens are not identical. According to the present invention, said first and second antigens are, in fact, two different molecules that are located on two different types of cells. The term one arm [antibody] preferably means one Fab portion of a full length IgG class antibody. Bispecific antibodies that mediate cytotoxicity by recruiting and activating endogenous immunocompetent cells represent a new class of therapeutic agents based on new generation antibodies. This can be achieved by combining the antigen-binding specificity of target cells (i.e., tumor cells) and effector cells (i.e., T cells, NK cells, and macrophages) in a single molecule (Cui et al. JBC 2012 ( 287) 28206-28214; Kontermann, MABS 2012 (4) 182-197; Chames and Baty, MABS 2009 (1) 539-547; Moore et al. Blood 2011 (117) 4542-4551; Loffler et al. 2000 Blood 95 :2098; Zeidler et al. 1999 J. Immunol. 163: 1246). The present invention provides bispecific antibodies in which one arm binds to the CLEC12A antigen of abnormal (tumor) cells while the second arm binds to the antigen of immunocompetent effector cells.

The invention also relates to antibodies in which the VH is able to specifically recognize the first antigen, and the VL paired with the VH in the variable region of the immunoglobulin is able to specifically recognize the second antigen. The resulting VH/VL pair will bind either antigen 1 or antigen 2. 20, 472-486, October 2011) are also covered by the term bispecific antibody because they also have the ability to bind two different antigens. In one embodiment, a VH is used that specifically recognizes CLEC12A or a functional equivalent thereof, and a VL is used that specifically recognizes an immunocompetent effector cell antigen. Alternatively, an antibody of the present invention contains a VH that specifically recognizes an immunocompetent effector cell antigen and a VL that specifically recognizes CLEC12A or a functional equivalent thereof. In any case, the resulting antibody will typically contain two VH/VL pairs, with each VH/VL pair will bind either CLEC12A (or its functional equivalent) or an immunocompetent effector cell antigen. Two-in-one antibodies will typically either bind two similar antigens (AA or BB; monospecific bivalent) or bind two different antigens (AB; bispecific). Therefore, if two-in-one antibodies are used for the therapeutic applications of the present invention, a portion of these antibodies will not have the desired effect due to their binding to either two CLEC12A molecules (or their functional equivalents) or two immunocompetent effector cell antigens, such as CD3. Since the therapeutic goal can still be achieved with a portion of the administered antibodies, two-in-one antibodies are nonetheless suitable.

As used herein, the term CLEC12A refers to member A of the C-type lectin family of 12 domains, also known as C-type lectin-like molecule-1 (CLL-1), an antigen that is present on leukoblastic cells and on leukemic stem cells in acute myeloid leukemia (AML), including CD34-negative or CD34-low-expressing leukemic stem cells (lateral population) (A.B. Bakker et al. Cancer Res 2004, 64, p8443-50; Van Rhenen et al. 2007 Blood 110:2659 ; Moshaver et al. 2008 Stem Cells 26:3059). Otherwise, CLEC12A expression is restricted to a hematopoietic cell line, in particular myeloid cells in peripheral blood and bone marrow, ie. granulocytes, monocytes, and dendritic cell precursors. More importantly, CLEC12A is absent from hematopoietic stem cells. This expression profile makes CLEC12A a particularly suitable target in AML. Alternate names for CLEC12A include dendritic cell-associated C-type lectin-2 (DCAL-2), a receptor like myeloid cell lineage inhibitory C-type lectin (MICL), and member 1 of the L subfamily of receptors like killer cell lectin 1 ( KLRL1) (Zhang W. et al. GenBankTM accession number: AF247788; A.S. Marshall, et al. J Biol Chem 2004, 279, pl4792-802; GenBankTM accession number: AY498550; Y. Han et al. Blood 2004, 104, p2858-66; H. Floyd, et al. GenBankTM accession number: AY426759; C. H. Chen, et al. Blood 2006, 107, p 1459-67). The combination of these sequences is shown in Fig. 1. The full-length form of CLEC12A contains 275 amino acid residues, including an additional intracellular region of 10 amino acids, which is absent in most other isoforms, and shows a strictly myeloid expression profile (at the level of presentation on the surface and at the level of mRNA). The term CLEC12A, or its functional equivalent, means all variants referred to above and their isoforms, which are characterized by the maintenance of a strictly myeloid expression profile (both surface and mRNA level) as described in Bakker et al. Cancer Res 2004, 64, p8443-50. Therefore, the present invention includes IgG class bispecific antibodies in which one arm specifically recognizes functional equivalents of CLEC12A, including those functional equivalents lacking the aforementioned additional 10 amino acid intracellular region. However, the bispecific IgG antibodies of the present invention are preferred in which one arm specifically recognizes the full length form of CLEC12A.

As used herein, the term abnormal cells includes tumor cells, more specifically, tumor cells of hematological origin, including also pre-leukemia cells such as cells that cause myelodysplastic syndromes (MDS) and leukemic cells such as acute myeloid leukemia (AML) tumor cells or chronic myeloid leukemia (CML) cells.

As used herein, the term immunocompetent effector cell or effector cell refers to a cell in the natural repertoire of cells in the mammalian immune system that can be activated and affect the viability of the target cell.

Immunocompetent effector cells include cells of the lymphoid lineage, such as natural killer (NK), T cells, including cytotoxic T cells, or B cells, but also cells of the myeloid lineage, such as monocytes or macrophages, dendritic cells, and neutrophilic granulocytes may be considered as immunocompetent effector cells. Therefore, said effector cell is preferably an NK cell, a T cell, a B cell, a monocyte, a macrophage, a dendritic cell, or a neutrophil granulocyte. According to the present invention, attracting effector cells to abnormal cells means that immunocompetent effector cells are brought into close proximity to abnormal target cells, so that effector cells can

- 5 040393 directly destroy or indirectly initiate the destruction of abnormal cells to which they are attracted. In order to avoid non-specific interactions, it is preferable that the bispecific antibodies of the present invention specifically recognize antigens of immunocompetent effector cells that are at least overexpressed in these immunocompetent effector cells compared to other cells in the body. Target antigens present on immunocompetent effector cells may include CD3, CD16, CD25, CD28, CD64, CD89, NKG2D and NKp46. Preferably, the antigen of the immunocompetent effector cells is CD3 present on T cells or a functional equivalent (the functional equivalent would be a molecule like CD3 with a similar distribution on T cells and a similar function (by type, not necessarily quantitatively)) . As used herein, the term CD3 also encompasses functional equivalents of CD3. The most preferred immunocompetent effector cell antigen is the CD3ε chain. This antigen has been found to be very effective in attracting T cells to abnormal cells. Therefore, the bispecific IgG antibody of the present invention preferably contains a single arm that specifically recognizes CD3ε.

Thus, the present invention relates to a bispecific full-length IgG antibody, said bispecific antibody comprising one arm that specifically recognizes CLEC12A or a functional equivalent thereof, and a second arm that specifically recognizes an antigen of immunocompetent effector cells capable of recruiting such cells to abnormal a cell expressing CLEC12A or said functional equivalent, said immunocompetent effector cells including T cells. In another preferred embodiment, the present invention provides an IgG class bispecific antibody of the present invention, wherein said antigen of said immunocompetent effector cells is CD3 or a functional equivalent thereof, preferably human CD3ε. In another embodiment, the present invention provides F(ab)' ₂ fragments of such a bispecific CLEC12AxCD3 IgG class antibody.

In one aspect, the present invention provides an IgG bispecific antibody of the present invention wherein both arms comprise a common light chain. The term common light chain of the present invention refers to light chains that may be the same or have some differences in amino acid sequences while retaining antibody binding specificity. For example, it is possible, within the scope of the definition of generic light chains as used herein, to make or find light chains that are not identical but are still functionally equivalent, for example, by introducing and testing for conservative amino acid substitutions, amino acid substitutions at sites , which do not contribute or only partially contribute to binding specificity after pairing with the heavy chain, and the like. All of the terms total light chain, total VL, individual light chain, individual VL, with or without the addition of the term rearranged, are used interchangeably herein. In one aspect, the present invention relates to use as a common light chain of a human light chain that can be combined with various heavy chains to produce antibodies with functional antigen binding domains (WO2004/009618, WO2009/157771, Merchant et al. 1998, Nissim et al ., 1994). Preferably, the overall light chain has a germline sequence. The preferred germline sequence is the light chain variable region, which is frequently used in the human repertoire and has excellent mating ability with many different VH regions and has good thermodynamic stability, yield and solubility. The most preferred germline light chain is O12, preferably a rearranged IgVK1-39*01/IGJK1*01 germline kappa light chain (nomenclature according to the IMGT database, on the web at imgt.org) or a fragment or functional derivative thereof. The terms rearranged germline kappa light chain IgVK1-39*01/IGJK1*01, IGKV1-39/IGKJ1, huVK1-39 light chain, or huVK1-39 for short, are used interchangeably throughout this application. Obviously, a person skilled in the art will understand that the total also refers to functional light chain equivalents, the amino acid sequence of which is not identical. There are many variants of this light chain, in the case of which there are mutations (deletions, substitutions, additions) that do not significantly affect the formation of functional binding regions.

A particularly preferred embodiment relates to an IgG class bispecific antibody of the present invention wherein the arm that specifically recognizes CLEC12A or a functional equivalent thereof comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTSY and a CDR2 sequence a heavy chain consisting of a sequence that is at least 90% identical to IINPSGGS, and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to GTTGDWFD. Preferably, said sequences

- 6 040393

The heavy chain CDRs 1, 2 and 3 consist of a sequence that is at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% identical to the CDR sequences shown. . Typically, 1, 2, or 3 amino acid residues are allowed to deviate from the CDR sequences shown, while maintaining the same type of binding activity (in type, not necessarily in quantitative terms). Therefore, said heavy chain CDR sequences 1, 2 and 3 preferably contain sequences that differ by no more than three, preferably no more than two, more preferably no more than one amino acid from the given CDR sequences. In a particularly preferred embodiment, said heavy chain CDR 1, 2 and 3 sequences are identical to the given CDR sequences. The CDR sequences shown are those of Fab-arm 4327, which, as shown in the examples, has good binding abilities to CLEC12A. The heavy chain sequence of Fab arm 4327, hence the VH from CLEC12A specific antibody 4327, is shown in FIG. 20. In one preferred embodiment, an IgG bispecific antibody of the present invention comprises a heavy chain variable region (VH) sequence that is at least 90% identical to that of the VH of the 4327 antibody. of the invention, wherein the arm that specifically recognizes CLEC12A or a functional equivalent thereof comprises a VH sequence consisting of a sequence that is at least 90%, preferably at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% or even 100% sequence identical

QVQLVQSGAEVKKPGASVKVSCKASGYTFTSYYMHWVRQAPGQGLEWMGIINPSGGSTSYAQK FQGRVTMTRDTSTSTVYMELSSLRSEDTAVYYCAKGTTGDWFDYWGQGTLVTVS.

As shown in the examples, the bispecific antibodies of the present invention containing the above VH sequence, together with the VH sequence of the CD3 recognition Fab arm (and together with the common light chain), have an excellent ability to induce T-cell mediated lysis of CLEC12A-positive tumor cells. AML.

Another preferred embodiment relates to an IgG class bispecific antibody of the present invention, wherein the arm that specifically recognizes CLEC12A or a functional equivalent thereof comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTSY and the sequence A heavy chain CDR2 consisting of a sequence that is at least 90% identical to IINPSGGS and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to GNYGDEFDY. Preferably, said heavy chain CDR 1, 2 and 3 sequences consist of a sequence that is at least 95% identical, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% of the given CDR sequences. As stated above, deviations of 1, 2 or 3 amino acid residues from the given CDR sequences are generally tolerated while maintaining the same type of binding activity (in type, not necessarily in quantitative terms). Therefore, said heavy chain CDR sequences 1, 2 and 3 preferably contain sequences that differ by no more than three, preferably no more than two, more preferably no more than one amino acid from the given CDR sequences. In a particularly preferred embodiment, said heavy chain CDR 1, 2 and 3 sequences are identical to the given CDR sequences.

The CDR sequences shown are those of the VH region of antibody 4331, which, as shown in the examples, has good binding abilities to CLEC12A. The VH sequence of Fab-arm 4331 is also shown in FIG. 20. In one preferred embodiment, an IgG bispecific antibody of the present invention comprises a VH sequence that is at least 90% identical to this Fab-arm 4331 VH. Therefore, an IgG bispecific antibody of the present invention is further provided wherein an arm that specifically recognizes CLEC12A or a functional equivalent thereof comprises a VH sequence consisting of a sequence that is at least 90%, preferably at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% or even 100% sequence identical

EVQLVQSGAEVKKPGASVKVSCKASGYTFTSYYMHWVRQAPGQGLEWMGIINPSGGSTSYAQK FQGRVTMTRDTSTSTVYMELSSLRSEDTAVYYCARGNYGDEFDYWGQGTLVTVSS.

As shown in the examples, the bispecific antibodies of the present invention containing the Fab arm VH sequence 4331, together with the CD3 recognition Fab arm VH sequence (and together with the common light chain), have an excellent ability to induce T-cell mediated lysis of CLEC12A-positive tumor cells. AML.

However, another preferred embodiment relates to a bispecific

- 7 040393 an IgG antibody of the present invention, wherein the arm that specifically recognizes CLEC12A or a functional equivalent thereof comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTGY and a heavy chain CDR2 sequence consisting of from a sequence that is at least 90% identical to WINPNSGG and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to DGYFADAFDY. Preferably, said heavy chain CDR 1, 2 and 3 sequences consist of a sequence that is at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% identical to the given CDR sequences. Again, deviations of 1, 2 or 3 amino acid residues from the given CDR sequences are generally allowed while maintaining the same type of binding activity (in type, not necessarily in quantitative terms). Therefore, said heavy chain CDR sequences 1, 2 and 3 preferably contain sequences that differ by no more than three, preferably no more than two, more preferably no more than one amino acid from the given CDR sequences. In a particularly preferred embodiment, said heavy chain CDR 1, 2 and 3 sequences are identical to the given CDR sequences.

The CDR sequences shown are the CDR sequences in the VH of antibody 3918, which, as shown in the examples, also has good binding abilities to CLEC12A. The VH sequence of antibody 3918 is also shown in FIG. 20. In one preferred embodiment, an IgG bispecific antibody of the present invention comprises a VH sequence that is at least 90% identical to that of the 3918 VH. , which specifically recognizes CLEC12A or a functional equivalent thereof, comprises a VH sequence consisting of a sequence that is at least 90%, preferably at least 95%, more preferably at least 97%, more preferably at least 98 %, more preferably at least 99% or even 100% sequence identical

QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQK FQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARDGYFADAFDYWGQGTLVTVSS.

As shown in the examples, the bispecific antibodies of the present invention containing the 3918 Fab arm VH sequence, together with the CD3 recognition Fab arm VH sequence (and together with the common light chain), have an excellent ability to induce T-cell mediated lysis of CLEC12A-positive tumor cells. AML.

Another preferred embodiment relates to an IgG class bispecific antibody of the present invention wherein the second arm specifically recognizes CD3 and comprises a heavy chain CDR1 sequence consisting of the SYGMH sequence and a heavy chain CDR2 sequence consisting of the IIWYSGSKKNYADSVKG sequence and a heavy chain CDR3 sequence , consisting of the sequence GTGYNWFDP. Preferably, said CD3-specific arm comprises a VH sequence consisting of the sequence

QVQLVESGGGWQPGRSLRLSCAASGFTFRSYGMHWVRQAPGKGLEWVAIIWYSGSKKNYADS VKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARGTGYNWFDPWGQGTLVTVSS.

The CDR sequences and VH sequences shown are those of the 3896 antibody. These sequences are also shown in FIG. 20. A heavy chain comprising these CD3-specific CDR sequences and/or the 3896 Fab arm VH sequence shown is preferred for the IgG bispecific antibody of the present invention because these sequences give the bispecific antibody optimal affinity for CD3-expressing immunocompetent cells, while allowing sufficient binding to CLEC12A-positive AML tumor cells. Without wishing to be bound by any theory, it is believed that the overall effect of a bispecific antibody is determined by the combined effect of the affinity of one arm for antigen 1 and the affinity of the other arm for antigen 2. In the case of an antibody of the present invention having specificity for CLEC12A (or its functional equivalent), and antigen effector immunocompetent cells (preferably CD3), optimized timing of binding to CD3-positive immune cells and CLEC12A-expressing tumor cells is preferred for efficient induction of T-cell mediated lysis of CLEC12A-positive tumor cells. The balance between the affinities of the CLEC12A/CD3 bispecific antibody is believed to be of paramount importance. It is believed that a much higher affinity for CD3 compared to a much lower affinity for CLEC12A (ie, too much affinity for CD3) will lead to a situation where antibodies will mainly bind to CD3-expressing T cells. Such bispecific antibody-laden T cells can either engulf their CD3 or invade tissues, thus leaving the bloodstream before they encounter CLEC12A-positive tumor cells. This will reduce the therapeutic effect of the bispecific antibody.

- 8 040393

In the case of a more appropriate mechanism of action, CLEC12A-positive tumor cells first bind to one or more of the bispecific antibodies of the present invention, with subsequent T cell activation following recruitment of T cells by the free CD3-specific arm of the bispecific antibody. Alternatively, CD3 positive T cells and CLEC12A positive tumor cells bind substantially simultaneously to the bispecific antibody.

Therefore, the affinities for both CLEC12A (or a functional equivalent thereof) and the immunocompetent effector cell antigen (CD3) are preferably chosen or modulated such that the correct balance is achieved, i.e., that the resulting bispecific antibodies also bind to CLEC12A and CD3 substantially simultaneously, or that the bispecific antibodies tend to bind to CLEC12A-positive tumor cells to a sufficient extent, and after T cell activation occurs, the tumor cells are lysed. According to the present invention, this excellent balance between affinities for CD3 and CLEC12A is preferably achieved by combining a VH having CDR sequences (or the entire VH sequence) from Fab arm 3896 (which is specific for CD3) with a VH having CDR sequences (or the entire VH sequence) of one or the other Fab arm 4327 or 4331 or 3918 or 3116 (which are specific for CLEC12A). Such resulting bispecific antibodies exhibit an appropriate balance between CD3 and CLEC12A affinities such that T cells and CLEC12A positive AML tumor cells are efficiently brought together and T cell mediated lysis of CLEC12A positive AML tumor cells is optimally induced.

As previously described herein, an IgG class bispecific antibody of the present invention is preferably provided wherein both arms comprise a common light chain variable domain. A particularly preferred common light chain is the rearranged IgVK1-39*01/IgJK1*01 human kappa light chain, also referred to as 012.

The nucleotide and amino acid sequence of VL 012 is also shown in FIG. 20. CDR sequences are in bold and underlined. A bispecific antibody of the present invention containing a common light chain that at least includes CDR 012 sequences is therefore preferred. Therefore, one aspect of the present invention relates to an IgG class bispecific antibody of the present invention, wherein the first and second arms further comprise a light chain CDR1 sequence consisting of a sequence that is at least 90% identical to RASQSISSYLN and a light chain CDR2 sequence , consisting of a sequence that is at least 90% identical to AASSLQS, and a light chain CDR3 sequence, consisting of a sequence that is at least 90% identical to QQSYSTPPT. Preferably, said light chain CDR 1, 2 and 3 sequences consist of a sequence that is at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% % is identical to the given CDR sequences. Again, deviations of 1, 2 or 3 amino acid residues from the given CDR sequences are generally allowed. Therefore, said light chain CDR sequences 1, 2 and 3 preferably contain sequences that differ by no more than three, preferably no more than two, more preferably no more than one amino acid from the given CDR sequences. In a particularly preferred embodiment, said light chain CDR sequences 1, 2 and 3 are identical to the given CDR sequences. In one preferred embodiment, an IgG bispecific antibody of the present invention comprises a VL sequence that is at least 90% identical to VL chain 012. Therefore, further, an IgG bispecific antibody of the present invention is provided wherein the first and second arms comprise a VL sequence consisting of a sequence that is at least 90%, preferably at least 95%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99%, or even 100% sequence identical

DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFS GSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPTFGQGTKVEIK.

The term % identity with is defined herein as the percentage of residues in a candidate amino acid sequence that match those in a control sequence after matching the two sequences and introducing gaps, if necessary, to achieve the maximum percent identity. Methods and computer programs for alignment are well known in the art. One computer program that can be used or adapted for the purposes of determining whether a candidate sequence falls within this definition is Align 2, by Genentech, Inc., which has been filed with the user documentation with the Copyright Office in the United States. States, Washington, DC 20559, December 10, 1991.

The bispecific full-length IgG antibody of the present invention, by definition, has two distinct antigen-binding sites, but the IgG Fc region also includes a third site.

- 9 040393 binding to the Fc receptor. If a cell has both an Fc receptor and one of the targets of a bispecific antibody, cross-linking of the Fc receptor and said target on the surface of said cell may occur, which may lead to undesirable effects. In a preferred embodiment, the present invention provides an IgG bispecific full-length antibody of the present invention, wherein said IgG bispecific antibody has a lower hinge region and/or CH2 domain mutated such that interaction of said IgG bispecific antibody with Fc-gamma (Fcγ )-receptors is significantly reduced. As used herein, the expression so that the interaction of said IgG bispecific antibody with Fc-gamma receptors is significantly reduced means that the ability of said IgG bispecific antibody to interact with Fc-gamma receptors is reduced if such Fc-gamma receptors are present. near the specified antibody. Thus, according to the present invention, the region of the antibody, preferably the lower hinge region and/or the CH2 domain of the antibody, is mutated (generally by expressing the mutated nucleic acid sequence encoding it), resulting in a decrease in the ability to interact with the Fc receptor. Preferably, if the interaction with the Fc receptor essentially stops. Amino acid residues in human IgG1 that are involved in binding to Fcγ receptors have been mapped previously. In addition to residues that, after alteration, increased binding only to specific receptors, or simultaneously increased binding to one type of receptor and decreased binding to another type, several residues were found that abolished binding to one or more of the receptors (Shields RL et al. JBC 2001 (276) 6591-6604 Armour et al Mol Immunol 2003 (40) 585-593). In a further preferred embodiment, said mutated lower hinge region and/or CH2 domain contain at least one substitution at amino acid positions 235 and/or 236 (numbering according to Kabat). Preferably, the amino acids at both positions 235 and 236 are substituted. The examples show that substitutions at these sites can substantially prevent interaction between the bispecific antibody and the Fc receptor present on tumor cells. In particular, the L235G and/or G236R replacements have been shown to be very suitable for this purpose. Therefore, the present disclosure also provides an IgG bispecific antibody of the present invention wherein said mutated CH2 domain and/or lower hinge region comprise a L235G and/or G236R substitution. Preferably, both L235G and G236R are substituted. Alternatively, one of skill in the art can introduce mutations in the lower hinge and/or CH2 domain that include 234F, 235E and/or 331S substitutions (Oganesyan et al. Biol. Crystall. 2008 (D64) 700). Preferably all three substitutions are included in the case of this alternative.

U.S. Provisional Application 61/635935 filed by the present inventors, followed by U.S. Common Application No. 13/866747 and PCT Application No. PCT/NL2013/050294 (which are incorporated herein by reference), discloses methods and means for producing bispecific antibodies from a single cell, thereby providing means that promote the formation of bispecific antibodies instead of the formation of monospecific antibodies. These methods can also be effectively used in the present invention. Thus, the present invention relates to a method for producing a bispecific IgG full-length antibody according to the invention, starting from a single cell, wherein said IgG bispecific full-length antibody comprises two CH3 domains that are capable of docking, said method comprising delivery to said cell a ) a first nucleic acid molecule encoding a polypeptide chain comprising a first CH3 domain, b) a second nucleic acid molecule encoding a polypeptide chain comprising a second CH3 domain, said nucleic acid molecules being provided with means for preferential pairing of said polypeptides comprising the first and second CH3 domains wherein said method further comprises the step of culturing said host cell and allowing said two nucleic acid molecules to be expressed and said bispecific full-length IgG class antibody to be obtained from the culture. Said first and second nucleic acid molecules may be part of the same vector or gene delivery vehicle and may be integrated into the same site in the genome of the host cell. Alternatively, said first and second nucleic acid molecules are delivered separately to said cell.

A preferred embodiment provides a method for producing a full-length IgG bispecific antibody of the present invention from a single cell, said IgG bispecific antibody comprising two CH3 domains that are dockable, said method comprising providing: - a cell containing a) a first a nucleic acid sequence encoding an IgG heavy chain which specifically recognizes CLEC12A and which contains a first CH3 domain, and b) a second nucleic acid sequence encoding an IgG heavy chain which specifically recognizes an antigen of immunocompetent effector cells, preferably CD3, and which contains a second CH3 domain, and

- 10 040393 said nucleic acid sequences are provided with means for preferential pairing of said first and second CH3 domains, said method further comprising the step of culturing said cell and allowing said two nucleic acid sequences to be expressed and said bispecific IgG class antibody to be obtained from the culture . In a particularly preferred embodiment, said cell also contains a third nucleic acid sequence encoding a common light chain. The preferred common light chain is 012, preferably the rearranged human germline kappa light chain IgVK1-39*01/IGJK1*01 as described above. Preferred mutations to produce substantially only bispecific full-length IgG molecules are the amino acid substitutions L351K and T366K (numbered according to Kabat) in the first CH3 domain and the amino acid substitutions L351D and L368E in the second CH3 domain, or vice versa. Therefore, further provided is a method of the present invention for producing an IgG1 subclass bispecific antibody, wherein said first CH3 domain comprises the amino acid substitutions L351K and T366K (Kabat numbering), and wherein said second CH3 domain comprises the amino acid substitutions L351D and L368E, said method further comprising the step of culturing said cell and allowing expression of said nucleic acid sequences and obtaining said bispecific antibody from the culture. Also provided is a method of the present invention for producing an IgG1 subclass bispecific antibody, wherein said first CH3 domain comprises the amino acid substitutions L351D and L368E (numbered according to Kabat), and wherein said second CH3 domain comprises the amino acid substitutions L351R and T366K, wherein said method, it further comprises the step of culturing said cell and allowing expression of said nucleic acid sequences, and obtaining said bispecific antibody from the culture. Antibodies that can be produced using these methods are also part of the present invention.

The present invention further relates to a pharmaceutical composition comprising an IgG bispecific antibody of the present invention and a pharmaceutically acceptable carrier. As used herein, such a pharmaceutically acceptable carrier includes any and all solvents, salts, dispersion media, coatings, antibacterial and antifungal agents, isotonicizing and absorption delaying agents, and the like, which are physiologically compatible. Depending on the route of administration (eg, intravenous, subcutaneous, intra-articular, etc.), the active compound may be enclosed in a sheath material to protect the compound from acids and other environmental conditions that may inactivate the compound.

The antibodies and pharmaceutical compositions of the present invention find their use in the treatment of various leukemias and pre-leukemic diseases of myeloid origin, as well as B-cell lymphomas. Diseases that may be treated according to the present invention include myeloid leukemias or pre-leukemic diseases such as AML, MDS and CML and Hodgkin's lymphomas and most non-Hodgkin's lymphomas. Thus, the present invention relates to a bispecific full-length IgG antibody of the present invention for use as a pharmaceutical agent for the treatment of myelodysplastic syndrome (MDS), chronic myeloid leukemia (CML), or preferably acute myeloid leukemia (AML). Also provided is the use of an IgG class bispecific antibody of the present invention for the preparation of a medicament for the treatment or prevention of myelodysplastic syndrome (MDS), chronic myeloid leukemia (CML), or preferably acute myeloid leukemia (AML).

The amount of an antibody of the present invention administered to a patient is generally within the therapeutic window, meaning that an amount sufficient to produce a therapeutic effect is used without exceeding a threshold amount resulting in an unacceptable degree of side effects. The smaller the amount of antibody needed to produce the desired therapeutic effect, the larger the therapeutic window will typically be. Therefore, an antibody of the present invention that exhibits sufficient therapeutic effects at a low dose is preferred.

Approximately 30,000 patients are diagnosed with acute myeloid leukemia (AML) each year in Europe and the USA. Most of these patients are 60 years of age or older. Old age is the main negative factor determining the outcome of AML, and long-term (over 5 years) survival of intensively treated elderly patients with AML is observed in approximately 10%. Almost all patients who achieve remission after induction chemotherapy have disease progression within 3 years. Current post-remission treatment has shown limited, if any, utility in elderly patients with AML. Thus, there remains a significant mass of resistant leukemia remaining, and the surviving subpopulation of drug-resistant leukemia cells rapidly relapses. New types of drugs with very different mechanisms of action are needed to target these non-responsive AML tumor cells in an attempt to induce and maintain complete remissions. Although complete remission (CR) can be achieved with a range of intensive chemotherapy combinations in more than 50%

- 11 040393 elderly patients with AML and approximately 80% of younger patients, response success or survival remained the main research issue. In a recently published network meta-analysis of 65 randomized clinical trials (15110 patients) in elderly patients with AML, most of the modified research induction regimens have similar or even worse efficacy profiles compared to the conventional 3+7 induction regimen with daunorubicin and cytarabine. This standard treatment for AML is associated with high rates of disease prevalence and even mortality. Most patients in CR suffer a relapse due to leukemic stem cells remaining after chemotherapy. Further dose escalation is limited due to unacceptable toxicity. Thus, especially in the case of elderly patients with AML, there is an urgent need for new therapies, preferably with less toxicity.

Treatment of non-responders to chemotherapy AML could be achieved by recruiting T-cells of the patient's own immune system and AML tumor cells using a bispecific antibody. Thus, the immune system of patients is strengthened and redirected to attack and destroy AML tumor cells. The present invention provides CD3xCLEC12A IgG class bispecific antibodies that effectively induce lysis of AML tumor cells. CD3xCLEC12A bispecific antibodies thus represent a targeted therapy with fewer side effects that specifically kill leukemic stem cells to improve the prognosis for patients with AML. Because CLEC12A is expressed in leukemic stem cells (LSCs) but not in normal hematopoietic stem cells, therapy directed against this antigen (as shown in vitro) will eradicate LSCs, reserving normal stem cells. This is likely to have the greatest effect in the case of minimal residual disease (MRD). There is a possibility that the recurrence rate will decrease due to elimination of MRD. Thus, the effect on the AML patient of this new treatment would be a less toxic treatment with a lower relapse rate, resulting in an improved outcome associated with a better quality of life. These full-length bispecific IgG antibodies are being tested in clinical trials in patients with recurrent AML. Clinical efficacy is analyzed using the reduction in AML of blast cells in the bone marrow as an objective measure of response. An effective bispecific IgG for AML provides a new therapeutic option for a large subset of patients for whom no treatment is currently available. In addition to providing a means to achieve long-term remissions, this treatment option also has the potential to treat AML when used in remission.

Examples

Example 1 Creation and performance characterization of a CD3xCLEC12 bispecific IgG1 candidate.

To test the concept of targeting an immunocompetent effector cell to an abnormal cell using a bispecific full-length IgG, a CD3XCLEC12A bispecific IgG1 candidate was created in which the CD3- and CLEC12A-specific Fab arms were derived from previously described antibodies. The CD3-specific Fab arm used the VH region from the anti-CD3 antibody 15C3, one of the CD3-specific antibodies described in WO2005/118635, and this VH is referred to as 3056. The CLEC12A-specific Fab arm used the VH region from scFv SC02 -357, one of the CLEC12A-specific antibodies described in WO2005/000894 (hereinafter referred to as the CLEC12A-specific reference [Fab arm or antibody]; alternatively, this VH is referred to as 3116). The nucleotide and amino acid sequences of the VH CD3-specific arm (3056) and the nucleotide and amino acid sequence of the VH CLEC12A-specific arm (3116) of this candidate molecule, which is referred to as candidate 3056x3116, are shown in FIG. 20. The nucleotide and amino acid sequences of the total VL (huVK1-39; 012) are also shown in FIG. 20.

The corresponding VH regions were cloned into expression vectors using methods known in the art for the production of bispecific IgG1 (Gunasekaran et al. JBC 2010 (285) 19637-19646; WO2009/089004) in combination with rearranged human light chain IGKV139/IGKJ1 (huVK1-39). HuVk1-39 has previously been shown to be able to pair with more than one heavy chain, thus resulting in antibodies with different specificities, facilitating the design of bispecific molecules (De Wildt RM et al., J. Mol. Biol. 1999 (285) 895-901; De Kruif et al. J. Mol Biol 2009 (387) 548-58; WO2009/157771).

First, 3056x3116 CD3xCLEC12A bispecific IgG1 candidate binding to CD3ε on HPB-ALL cells was demonstrated by flow cytometry, which was performed according to standard procedures known in the art (Table 1). Binding to cell-presented CD3ε was confirmed using CHO cells transfected with CD3δ/ε or CD3γ/ε. Binding of the 3056x3116 bispecific IgG1 candidate to CLEC12A was determined using CHO cells transfected with the CLEC12A expression construct; a monospecific anti-CD3 antibody (3056x3056) and a monospecific anti-CLEC12A antibody (3116x3116) and an irrelevant IgG1 isotype control mAb were taken as controls.

- 12 040393

Table 1. Binding to cell-presented CD3 and CLEC12A according to flow cytometry

IgG HPB-ALL* cells CHO cells transfected with CLEC12A* 3056x3116 CD3xCLEC12Acandidate 6216 5299 CD3 6899 282 CLEC12A 199 4147 Isotype-matched control 34 289

*results are presented as mean fluorescence intensity.

Affinity determinations of the 3056x3116 bispecific IgG1 candidate for CD3δ/ε and the extracellular domain of CLEC12A are performed using surface plasmon resonance (BIAcore). Briefly, purified recombinant antigens are covalently attached to the surface of the CM5 sensor chip using free amine chemistry: antigens are diluted in kAc buffer to 10 µg/mL and attached to the surface, which is activated with NHS/EDC (according to the manufacturer's recommendations). To determine the affinities of the Fab arms present in the bispecific antibodies, they are serially diluted to 100, 50, 20, 10, 1, and 0.1 nM in Hepes buffered saline (HBS) and surface-bound with high-strength CM5 sensor chip antigen. (30 µl/min) flow rate (to prevent rebinding). Flow cell 1 (FC1) is used as the control surface and the responses (grams on the sensor) resulting from this surface are subtracted from the responses determined in the other flow cells (FC). FC2 and FC3 are used for two different antigens recognized by the bispecific antibody in order to be able to measure the affinities of both Fab arms in one kinetic assay on all three surfaces. Because the concentration of an antibody does not change significantly when flowing over a surface with bound antigen, the association rates (which are concentration dependent) of bispecific antibodies are determined simultaneously for the two different antigens they recognize. In this way, sensorgrams are obtained for the association and dissociation phases of various bispecific proteins. Using BIA software and curve fitting using a 1:1 interaction model (for a univalent interaction), the affinities of the Fab arms are determined. In the event that the binding of the bispecific protein to the antigen-coated surface of the sensor chip is weakened (i.e., when a very small amount of protein binds, resulting in low responses and/or very fast dissociation rates), the experimental setup is reversed: the bispecific antibody is covalently bound with the surface of the sensor chip using free amine chemistry, and the recombinant purified antigen is passed over the surface at a high (30 μl/min) flow rate to measure the affinity of the Fab arm directed against this antigen.

The functionality of the 3056x3116 CD3xCLEC12A bispecific Ig candidate was then tested. First, the ability to stimulate T cells was tested using resting T cells from a healthy donor. Briefly, peripheral blood was obtained from healthy donors after informed consent. T cells were isolated by standard isolation using a density gradient to enrich for peripheral blood mononuclear cells (PBMC), followed by negative selection using magnetic beads (Pan T cell isolation kit, Miltenyi Biotec catalog number 130-091- 155). Using this purification strategy, T cells were so-called intact (ie not stained with antibodies, so-called resting T cells) to limit the possibility of pre-activation. Purified resting T cells were subsequently incubated with cells of the HL60 cell line of leukemic origin in 10% fetal bovine serum (FBS) or 10% normal human serum (HS) at an effector:target cell ratio of 10:1 for two days. The results are presented as the percentage of CD69-positive or CD25-positive cells in the population of CD4-positive or CD8-positive T cells.

Both bivalent anti-CD3 IgG and CD3XCLEC12A bispecific IgG effectively increased CD69 and CD25 T cell activation markers on CD4 positive and CD8 positive T cells (FIG. 2). In the presence of FBS, which did not block Fc receptors present on HL60 cells (Liesveld et al. 1988, J. Immunol. 140 (5), pages 1527-1533), also a control bispecific molecule, CD3X control of the IgG isotype, was found to caused T cell activation. This effect was reduced in the presence of HS, suggesting that the observed activation of T

- 13 040393 cells as a result of univalent binding to CD3 CD3X IgG isotype control depended on crosslinking with Fc. However, T cell activation induced by the 3056x3116 CD3xCLEC12A bispecific IgG candidate was only partially dependent on Fc interactions, as the potential for CD69 and CD25 increase was almost completely preserved in the presence of HS (FIG. 2). This indicated that the intrinsic strength of CD3 univalent binding was sufficient to activate T cells when the binding molecule bound to the CLEC12A antigen on HL60 target cells after binding to the other Fab arm.

To investigate whether the degree of T cell activation by the 3056x3116 CD3XCLEC12A bispecific IgG candidate is sufficient to induce target cell lysis, HL60 cells in this assay were labeled with carboxyfluorescein diacetate succinimide ester (CFSE) and cocultured with T cells at various effector ratios: target cell. After one, two or three days, the number of surviving CFSE-positive HL60 cells was determined using flow cytometry. Results are presented as percent specific lysis referred to PBS.

As expected, CD3 monospecific divalent IgG induced T-cell mediated killing of HL60 cells (FIG. 3). Surprisingly, CD3XCLEC12A bispecific univalent IgG and CD3X IgG isotype control also induced resting T cell-mediated killing of HL60 cells. These effects were most pronounced when the assay was performed in the absence of excess human IgG, ie when Fc receptors on HL60 target cells were not blocked (FBS condition; FIG. 3). Surprisingly, even in the presence of excess human IgG (10% HS usage condition), CD3XCLEC12A bispecific IgG was very effective in killing HL60 cells, meaning that the induction of HL60 lysis is independent of interactions with Fcγ receptors. On day 3, there was also lysis of HL60 induced by CD3X control of the IgG isotype, probably due to incomplete blocking of Fc-gamma receptors after long periods of incubation. Killing of HL60 target cells varied with different ratios of effector:target cell (FIG. 4).

In conclusion, this example demonstrates that the CD3xCLEC12A bispecific molecule is a potent inducer of T-cell mediated tumor cell lysis and supports the hypothesis put forward by the present inventors that recruitment of T cells to effectively kill abnormal cells may be mediated by CD3xCLEC12A bispecific full-length antibody of the IgG1 subclass. Surprisingly, the activity induced by the CD3XCLEC12A bispecific IgG is independent of interactions with Fcγ receptors. To expand the panel of CD3XCLEC12A bispecific full-length IgGs in order to arrive at a final clinical candidate, panels of CD3-specific Fab arms and CLEC12A-specific Fab arms were created. Evaluation of the specificity and functionality of CD3-specific and CLEC12A-specific Fab-arms is carried out without changing the other arm, using the corresponding Fab from 3056x3116 CD3XCLEC12A bispecific IgG candidate, as shown in this example.

Example 2 Generation and characterization of CD3-specific Fab arms for CD3xCLEC12 bsAb.

Example 1 showed that CD3xCLEC12A bispecific molecules can be strong inducers of T-cell mediated tumor cell lysis. For this reason, separate panels of CD3-binding molecules as well as CLEC12A-binding molecules were created to create wider panels of such bispecific molecules.

To create a panel of CD3 binding molecules, CD3ε-specific VH regions are generated by immunizing mice transgenic for the huVK1-39 light chain (WO2009/157771) and human heavy chain (HC) minilocus, CD3ε in various formats: (1) isolated CD3δ /ε or CD3γ/ε, which can be fused or linked to a carrier molecule (such as a human IgG Fc or His tag) as known in the art, with or without the use of an adjuvant, (2) cells expressing CD3δ/ε or CD3γ/ε, or (3) a DNA construct encoding CD3δ/ε or CD3γ/ε, or a combination of these strategies. From immunized mice showing a sufficient titer of antigen-specific antibodies, as determined by ELISA and/or flow cytometry, spleens and/or lymph nodes are obtained, on the basis of which Fab libraries in phages are created. Alternatively, sequences of the VH regions are obtained directly from spleen and lymph node material using deep sequencing (co-pending US Provisional Application No. 61/539116).

Antigen-specific Fab arms are selected from phage libraries derived from immunized mice or from synthetic phage display libraries that contain the huVK1-39 light chain VL region and a set of human VH regions. To generate synthetic libraries, randomized CDR3 primers were used as described in De Kruif et al. 1995, J Mol Biol 248 (1), pages 97105. Bacteriophages from these libraries are selected over several cycles for binding to an isolated CD3δ/ε protein, which can be associated with a carrier molecule (see above) or with cells expressing CD3ε, such as HPB-ALL, or cells transfected to express CD3δ/ε or CD3γ/ε, or a combination of these strategies. Non-binding phages are removed, but

- 14 040393 phages that are elicited are eluted using an acidic buffer or, in order to target the selection of Fabs to the desired specificity, using antibodies against a specific epitope, for example using antibodies that are cross-reactive with cynomolgus monkey CD3ε. These phages are then transfected into competent bacteria which are grown under selection pressure for the phage-containing bacteria. After selection of a number of surviving bacterial colonies, the phages are released and subjected to the next round of selection.

After selection is complete, the remaining phages are screened for binding to the antigen present on the cells by flow cytometry and to the isolated antigen by ELISA. Reference anti-CD3 antibodies such as those known in the art, such as OKT-3, are used as a positive control for binding. Nucleotide material from substantially all phages that have shown specific binding to antigen-expressing cells is subjected to colony PCR to amplify the VH regions and PCR with sequencing to determine the sequence of the VH region. The resulting sequences are grouped based on the uniqueness of their HCDR3. In the case of sequences derived from immunized mice in which (limited) somatic hypermutation may occur, VH sequences are further grouped based on the likelihood of a unique VDJ (i.e., if HCDR3s in different clusters contain a <2 amino acid difference, they are considered part of the same and the same cluster and are grouped together). From each cluster, one or more VH regions per cluster are selected for cloning into vectors for expression as bivalent monospecific IgG in combination with the huVK1-39 light chain. The VH region, for which specific binding to the isolated antigen and the antigen present on the cells is confirmed, is subsequently cloned into vectors for expression in the CD3XCLEC12A bispecific format. Their characteristics are then determined to select a candidate with therapeutic potential (see the following examples).

Example 3 Generation and characterization of CLEC12-specific Fab arms for CD3xCLEC12 bsAb.

Since CD3xCLEC12A bispecific molecules were shown in Example 1 to be capable of inducing T-cell mediated tumor cell lysis, the present inventors then wished to establish a broader panel of such bispecific molecules. In addition to the panel of CD3 binding molecules described in Example 2, the present inventors also created a panel of CLEC12A binders.

Briefly, CLEC12A-specific Fab arms were selected from synthetic Fab phage display libraries that contained a rearranged IGKV1-39/IGKJ1 human VL region and a set of human VH regions (De Kruif et al. Biotechnol Bioeng. 2010 (106) 741-50). Bacteriophages from these banks were selected for two cycles in relation to binding to CLEC12A. This was done by incubation with the extracellular domain of CLEC12A (amino acids 75-275) coupled to a His tag (Sino Biological, cat. no. 11896-H07H) which was layered on the surface. Non-binding phages were removed, binding phages were chemically eluted and used to infect bacteria which were grown under selection pressure against phage-containing bacteria. After selecting a number of surviving bacterial colonies, the phages were released and subjected to the next cycle of selection and propagation.

After selection was complete, the remaining phages were screened for binding to CLEC12A present on the HL60 tumor cell line using flow cytometry. A reference antibody against CLEC12A was used as a positive control for binding. Nucleotide material from virtually all phages that showed specific binding to CLEC12A expressing cells was subjected to colony PCR to amplify the VH regions and PCR with sequencing to determine the sequence of the VH region. The resulting sequences were grouped based on the uniqueness of their HCDR3. VH regions from each unique HCDR3 cassette were cloned into vectors for expression as monospecific or bispecific IgG in combination with rearranged human LC IGKV1-39/IGKJ1.

Three selected CLEC12A binding molecules with a unique HCDR3 sequence showed the desired profile in IgG format, which had the following characteristics (Table 2 and data not shown).

Specific binding to the isolated extracellular domain of CLEC12A.

Specific binding of CLEC12A expressed in a tumor cell line.

Confirmation of a myeloid lineage-specific expression profile in human PBMC.

- 15 040393

Table 2. Characterization of CLEC12A-specific Fab arms

CDR3 length Binding to coated CLEC12A* Linking with CLEC12A- expressing cells* * Epitope competition with reference Fab against CLEC12A*** Reference fab 9 1.422 1467 against CLEC12A 10 1.253 899 Yes 9 1.307 1559 No 9 1.328 1106 Yes

*tested in ELISA, CLEC12A extracellular domain (Sino Biological) layered at concentration=2 μg/ml, results are shown as optical density (isotype-matched control as background signal: 0.127).

** verified by flow cytometry on HL60 cells using optimized IgG concentration, results are shown as mean fluorescence intensity (isotype-matched control as background signal: 116).

*** tested in Fab format ELISA compared to reference IgG at concentration=20 µg/ml.

Example 4 Selection of a functional CLEC12-specific Fab arm for CD3xCLEC12 bsAb.

Selected CLEC12A specific Fab arms as described in Example 3 were subsequently expressed in bispecific IgG format with the new CD3 specific Fab arm as the unchanged arm. This new CD3-specific Fab arm, termed anti-CD3 IgG3896 or 3896 for short, also uses the huVK1-39 light chain. The nucleotide and amino acid sequences of this CD3-specific VH candidate 3896 are shown in FIG. 20. Therefore, different CD3XCLEC12A bispecific molecules were expressed, all of which had the same anti-CD3 arm 3896, but which differed in the CLEC12A-specific arm (containing either an anti-CLEC12A reference arm or any one of the CLEC12A-specific Fab- shoulder candidates 4327, 4331 or 3918). These CD3XCLEC12A bispecific molecules were then functionally tested in a target cell lysis assay as described in Example 1. Results were presented as percent specific lysis relative to an isotype matched control. All candidate CLEC12A-specific Fab arms demonstrated dose-dependent specific lysis of HL60 target cells in a bispecific format, with kinetics that were similar to or better than when the CLEC12A-specific reference Fab arm was used (Fig. 5).

Also CD3x isotype control bsAb demonstrated dose-dependent target cell lysis, although higher log concentrations were required to induce the same degree of specific lysis. Despite the presence of excess human IgG through the addition of HS, the killing activity of this monovalent CD3-specific IgG was still visible, probably as a result of Fc-mediated crosslinking. As will be seen from Example 7, this non-specific target lysis can indeed be completely abolished by inhibition of interaction with the Fc receptor by CH2 engineering.

Example 5 Efficacy of CD3xCLEC12 candidate products using AML T cells and/or AML tumor cells.

Examples 1 and 4 demonstrated the ability of a CD3XCLEC12A bispecific IgG that uses either the CD3-specific Fab arm 3056 or 3896 and that uses the CLEC12A-specific candidate Fab arms 4327, 4331 or 3918 or the CLEC12A-specific reference Fab arm 3116, to the induction of lysis of HL60 target cells, mediated by resting T cells of a healthy donor. This example investigates whether T cells derived from patients with AML, one of the main indications for therapeutic use of a CD3XCLEC12A bispecific drug, can be stimulated to kill target tumors after stimulation of CD3XCLEC12A with a bispecific full-length IgG. It is then determined whether patient-derived T cells can kill autologous AML tumor blast cells after CD3XCLEC12A stimulation with bispecific full-length IgG.

- 16 040393

T cells are isolated from the peripheral blood of AML patients according to the procedures described in Example 1. Purified patient-derived T cells are then incubated with CFSE-labeled HL60 cells and monitored for cell lysis as described in Example 1.

In addition, T-cell mediated target cell lysis assay is performed using AML tumor blast cells isolated from the same patient (Norde et al. Blood 2009 (113) 2312). The isolated blast cells are then labeled with CFSE and co-cultured with autologous patient-derived T cells in the presence of a mixture of cytokines as described in Norde et al. and in the presence of CD3XCLEC12A bispecific IgG or controls. Target cell lysis is controlled as described in Example 1.

Example 6 Release of cytokines by T cells after contact with CD3XCLEC12A bispecific IgG.

Using T cell stimulating biologics, T cell overstimulation is a serious hazard because it can lead to cytokine release syndrome (Suntharalingam et al. 2006, New England J Med 355 (10), pages 1018-1028; Chatenoud et al. 1990, Transplantation 49(4), pages 697-702). To investigate the extent of T cell stimulation induced by CD3XCLEC12A bispecific IgG, T cell cytokine induction was examined in a coculture of T cells and Fc receptor expressing target cells.

Briefly, healthy donor resting T cells were co-cultured with HL60 target cells in the presence of 3056x3116 CD3XCLEC12A bispecific IgG candidate (1 mg/ml) or control IgG as described in Example 1. Two days later, the supernatant was collected and cytokine production levels were determined in Luminex assay, as known in the art, using a 10-Plex human cytokine panel (Invitrogen, catalog number LHC0001). This panel covers ten major Th1 and Th2 cytokines.

As expected, CD3 monospecific divalent IgG induced strong production of IFNy, TNFa and IL-2 (Table 3), which are thought to be the main contributors to the cytokine release syndrome. In addition, the production of IL-4, IL-6, IL-8 and IL-10 was increased when incubated with CD3-specific IgG. In contrast, only CD3XCLEC12A bispecific IgG induced IL-8 production at the same level as CD3-specific IgG; other cytokines were not significantly induced by bispecific IgG. GM-CSF was below the detection limit in all conditions.

Table 3. Antibody-induced release of cytokines by T cells

Cytokine CD3 IgG CD3XCLEC12A IgG SOZHcontrol isotype ΙΕΝγ 484.3+155.0 13.5+19.1 0.0+0.0 TNFa 85.3+23.1 14.5+1.4 4.6+1.1 IL-2 285.6+325.5 3.4+0.8 1.7+0.6 IL-4 23.6+3.7 10.2+0.2 7.3+1.3 IL-6 9.0+1.8 3.7+0.3 2.3+0.0 IL-8 1567.8+5.2 1280.1+118.4 359.6+183.6 IL-10 531.5+224.0 21.1+3.0 3.7+4.0 IL-Ιβ 4.4+0.4 3.3+0.1 2.5+0.1 IL-5 2.1+0.2 0.7+0.0 0.6+0.1

Results are presented as mean cytokine concentration in pg/ml of two donors±SD.

The data presented herein suggest a suitable therapeutic profile for the various CD3XCLEC12A bispecific IgG molecules as they strongly induce target cell lysis (Examples 1 and 4) without triggering T cell secretion of potentially deleterious amounts of pro-inflammatory cytokines, which are noted in the case of CD3-specific IgG.

Example 7 Effect of Fc Suppression on In Vitro Efficiency of CD3XCLEC12A bsAb.

The dose dependent target cell lysis by CD3X bsAb isotype control demonstrated in Example 4 was hypothesized to be due to interaction of the Fc portion of the bsAb with Fc receptors on HL60 target cells. Since such non-specific cell lysis can also occur in vivo, either by interacting with Fc receptors on target cells or on observer cells such as NK cells, CH2/lower hinge engineering was used to cause suppression of Fc-mediated bsAb activity.

For this, two strategies for mutating Fc were considered, using either a double 235G 236R mutation (DM; DM-Fc) or a triple 234F 235E 331S mutation (TM; TM-Fc). CD3XCLEC12A bsAb,

- 17 040393 (3056x3116) with either DM-Fc or TM-Fc were generated and confirmed by flow cytometry to bind to CLEC12A-expressing cells at the same intensity as wild-type Fc bsAb (not shown). data). These bsAbs, and wild-type, DM-Fc and FcTM variants of the CD3x isotype control bsAb were then tested in the HL60 target cell lysis assay (see Examples 1 and 4). Results are presented as percent specific lysis relative to the isotype matched control.

Fc suppression with either DM or TM had little or no effect on the extent of specific lysis of HL60 cells induced by the CD3XCLEC12A bsAb (FIG. 6). However, in the case of CD3x isotype control bsAb, the ability to induce lysis of HL60 cells was significantly reduced with TM and even more with DM.

This shows that Fc-suppression by CH2/lower hinge engineering further contributes to target-specific killing of abnormal cells by generating a CD3xCLEC12A bispecific IgG1 format that effectively and specifically recruits effector cells and reduces potential non-specific immune system activation mediated by helper cells expressing the normal Fcy receptor.

Example 8 Effect of Fc Suppression on FcRn, CD16, CD32, CD64 and C1q Binding.

Binding of 3056x3116 CD3XCLEC12A candidate bsAb to wild-type Fc or DM-repressed Fc or TM-repressed Fc activity to human FcRn was determined by biolayer interferometry (BLI, Octet QK, ForteBio). Briefly, purified CD3XCLEC12A IgG1 with wild-type Fc, IgG1 with DM-Fc, or IgG1 with TM-Fc was captured on protein L biosensors (ForteBio, cat. no. buffer/0.002% Tween 20 containing 1.0 mg/ml BSA pH 6.0 (FcRn binding buffer) at room temperature. Subsequently, soluble human FcRn (Sino Biological Inc., CT009-H08H) was added at a concentration of 1 μg/ml in FcRn binding buffer) at room temperature. Data analysis with Octet QK data analysis software showed, after leading to sensor IgG binding to L protein, that subsequent binding of CD3XCLEC12A bsAb to DM-repressed Fc or TM-repressed Fc was comparable to human FcRn. with CD3XCLEC12A bsAb with a wild-type Fc tail (FIG. 7), and thus Fc suppression did not affect binding to FcRn.

Binding of CD3XCLEC12A bsAb to Fc with repressed activity from CD16, CD32 and CD64 is determined by bilayer interferometry (BLI, Octet QK, ForteBio). Brief Protocol: Purified CD3XCLEC12A IgG1 with wild-type Fc, IgG1 with DM-Fc, or IgG1 with TM-Fc are captured on protein L biosensors (ForteBio, cat. no. 18-5085) at a concentration=50 µg/mL in 1x kinetics buffer (ForteBio 18-5032) at room temperature. Subsequently, recombinant protein CD16 (Sino Biological Inc, 10389-H08H1), CD32 (Sino Biological Inc, 10374-H08H) and CD64 (Sino Biological Inc, 10256-H08H) are added at a concentration=1.0 μg/mL in kinetics buffer (ForteBio 18-5032) at room temperature. Binding of FcR receptors to bsAb was analyzed using Octet QK data analysis software.

Binding of CD3XCLEC12A bsAb to C1q repressed Fc is determined by capture ELISA. To this end, purified CD3XCLEC12A IgG1 with wild-type Fc, IgG1 with DM-Fc, or IgG1 with TM-Fc is layered in PBS on a Nunc-Immuno MaxiSorp F96 plate (Nunc, 439454) overnight at 4°C. Subsequently, human C1q (Quidel, A400) was added at a concentration=2.0 μg/ml in ELISA buffer (2% milk/PBST). The complex is then visualized using ovine anti-human C1q polyclonal IgG (Meridian, K90020C) and rabbit anti-sheep HRP-conjugated polyclonal IgG (Southern Biotech, 6150-05). Finally, using a TMB substrate (BD 51-2606KC/51-2607KC), binding is shown and OD450 is quantified using a microplate reader (Multiskan EX, Thermo Electron Corporation).

Example 9 In vivo evaluation of CD3xCLEC12A bispecific IgG efficacy.

Animal xenograft studies using luciferase-expressing HL60 cells (HL60(-Luc) cells) are performed to confirm and expand in vitro data using CD3xCLEC12A bispecific IgG1. More specifically, these studies are performed to determine steady-state plasma concentrations at effective doses that will be taken into account when setting the starting dose for the phase 1 clinical evaluation. To this end, NOD/SCID mice (or comparable immunocompromised mice) are injected subcutaneously with an amount of viable HL60(-Luc) cells, which results in the generation of subcutaneous HL60 tumors in most animals within two weeks of injection. In parallel with HL60(Luc) inoculation, or after initial tumor acquisition, 5x10E6 or 1x10OE7 human PBMCs are administered. CD3xCLEC12A bispecific IgG or control monospecific or control bispecific IgG is administered intravenously at several dose levels on the first day of PBMC administration and 3, 6 and 9 days later. Tumor size is assessed 1 week after the initial

- 18 040393 inoculations with HL60(Luc). The arithmetic mean of tumor sizes (either indicated as tumor volumes or as total bioluminescence) from each group is plotted against time.

Example 10 Use of a CD3xCLEC12A Bispecific Full-Length Subclass Antibody

IgG1 in a phase Ia/Ib study.

The final master CD3xCLEC12A bispecific full-length IgG1 candidate is used to manufacture material in accordance with GMP (good manufacturing practice) and subjected to clinical trials in AML patients. First, a formal, non-clinical safety study of the candidate product is performed to establish a safe starting dose for first human use studies. An open-label, multicenter, dose escalation phase Ia/b trial is then conducted in patients with relapsed AML and/or resistant AML and in patients not eligible for intensive care to investigate the safety and tolerability of CD3xCLEC12A bispecific IgG following intravenous administration. Secondary endpoints include pharmacokinetic and pharmacodynamic characteristics and preliminary efficacy analysis. Overall response rates are assessed by assessing the reduction in AML of blast cells in the bone marrow. In phase Ia, the maximum tolerated dose (MTD) is determined after single/multiple dose increases. After an interim pharmacokinetic analysis, the phase Ib part of the study provides for a dose escalation group - MTD or entails further dose frequency studies.

Example 11 Ability of CD3xCLEC12A bsAb to induce T cell proliferation.

In patients with AML, T cell counts are low compared to the number of AML blast cells at the time of diagnosis. It is well known that T cells undergo proliferation after activation leading to an increase in the number of T cells. In addition, in Example 1, the present inventors have shown that the CD3xCLEC12A bsAb can activate T cells and has the ability to induce T cell-mediated lysis of tumor cells. The present inventors hypothesized that AML patients treated with CD3xCLEC12A bsAb would benefit from an increase in T cell subsets following CD3xCLEC12A-mediated bispecific molecule T cell activation, as T cell proliferation would lead to an increase in T effector cells. To demonstrate that CD3xCLEC12A bsAb induces T cell proliferation in vitro, resting T cells were purified, labeled with carboxyfluorescein diacetate succinimide ester (CFSE) and co-cultured with autologous CLEC12A+ monocytes in the presence of CD3xCLEC12A bsAbs or control Abs. To study in particular CD3xCLEC12A-induced T cell proliferation without non-specific Fc-gamma activation, the CD3xCLEC12A bsAb with DM-Fc tail described in Examples 7 and 8 was used. , an anti-CD3 monoclonal antibody with wild-type Fc, and an irrelevant, isotype-matched control (IgG with wild-type Fc) were included as controls. Monocytes and T cells from the peripheral blood of healthy donors were isolated by standard isolation using a density gradient to enrich for peripheral blood mononuclear cells (PBMC), followed by selection of CD14-positive monocytes using CD14 microbeads (human CD14 microbeads, Miltenyi Biotec, catalog no. 130-050-201), and negative selection of intact T cells using magnetic beads against other leukocytes (Pan T Cell Isolation Kit, Miltenyi BioTec, cat. no. 130-096-535). The Pan T Cell Isolation Kit allows the isolation of resting (intact) T cells (i.e. not stained with antibodies) to avoid the possibility of pre-activation of T cells.

CFSE-labeled purified resting T cells were then incubated with purified monocytes and bsAbs in 10% normal human serum (HS) media at a ratio of effector:target cells=5:1 for seven days. On day 7, the decrease in CFSE signal as a readout of data regarding T cell proliferation was measured by flow cytometry. The results were presented as CFSE signal for CD3+, CD3+CD4+CD3+ and CD8+ T cells in histograms.

The anti-CD3 Ab positive control with wild type Fc induced T cell proliferation, while the isotype matched control IgG with wild type Fc did not induce T cell proliferation (FIG. 8). As expected, the CD3x isotype control bsAb with wild-type Fc did not induce T cell proliferation, but to a much lesser extent compared to the anti-CD3 bivalent monospecific IgG control. In contrast, the CD3x isotype control bsAb with DMFc did not induce T cell proliferation due to its repressed Fc tail. The CD3xCLEC12A bsAb with DM-Fc also induced the desired T cell proliferation mediated in particular by the association of CD3 with the CLEC12A antigen.

This shows that the CD3xCLEC12A bsAb is not only capable of target-specific induction of T-cell mediated tumor lysis as previously shown, but can also strongly induce target-specific T cell proliferation leading to an increase in T cell numbers. Furthermore, it further demonstrates that Fc silencing by CH2/lower hinge engineering not only contributes to target-specific killing of abnormal cells, but also to target-specific induction of T-cell proliferation by CD3xCLEC12A bsAb with DM -Fc class IgG.

- 19 040393

Example 12 Evaluation of CD3xCLEC12A-induced increase in TEMRA subset in AML patients.

Since activation of T cell proliferation has been demonstrated in the CD3xCLEC12 bsAb with DM-Fc, the present inventors then wanted to investigate whether the CD3xCLEC12A bsAb with DM-Fc was able to induce the proliferation of the CD8+ cytotoxic T cell fraction in AML patients. CD8+ cytotoxic T cells have been recognized as the main effectors mediating tumor regression (Sluijter et al., 2010). CD8+ T cells can be divided into four subsets: naive (CCR7+CD45RA+), central memory cells (T _CM , CCR7+CD45RA-), effector memory cells (T _EM , CCR7-CD45RA-), and CD45RA+ effector memory cells (T _EMRA , CCR7-CD45RA+). Studies have shown that subsets of naive cells and CD8+ memory T cells have different abilities to proliferate and differentiate in response to TCR stimulation (Geginat et al., 2003).

First, the CD8+ fraction in the peripheral blood of AML patients in clinical remission was analyzed in comparison with healthy donors. To this end, PBMCs were isolated from frozen peripheral blood samples from AML patients and healthy donors using standard density gradient isolation. The PBMCs were then stained with antibodies against CCR7, CD3, CD4, CD8, CD45RA and CD45RO for analysis for CD8+ T cell subsets by flow cytometry. The results are presented as a percentage of the subset in the total fraction of CD8+ T cells.

Similar to what was previously described, a subset of untaught CD8+ T cells were found to be reduced in the blood of AML patients compared to a subset of untaught CD8+ T cells from healthy individuals, while the T _EMRA fraction (CCR7-CD45RA+) was increased in AML patients compared to healthy donors (FIG. 9).

Experiments are then carried out to investigate the target tumor-specific proliferation of the T cell fraction in AML patients. More specifically, these experiments were performed to determine whether CD3xCLEC12A bsAb with DM-Fc could enhance T cell proliferation and outgrowth of effector T cell subsets (TEM and T _EMRA ) in AML patients relative to naive CD8+ T cells in AML patients.

For this, resting T cells from AML patients in clinical remission are purified according to Example 11. The composition of CD8+ T cell subsets at day=0 is analyzed by staining PBMC with antibodies against CCR7, CD3, CD4, CD8, CD45RA and CD45RO, followed by analysis using flow cytometry. In addition, resting T cells are either labeled with CFSE or not labeled (CFSE labeling as described in Example 11) and co-cultured with HL60 leukemic cells at an E:T=5:1 ratio with control or test antibodies for 7 days. CFSE labeled T cells are used to quantify T cell proliferation while unlabeled T cells are used to determine the percentage of proliferating T cell subsets. CFSE-labeled and unlabeled T cells are incubated with PBS, isotype-matched control Ab with wild-type Fc, CD3xCLEC12A bsAb with DM-Fc, CD3xisotype control bsAb with DM-Fc, and monoclonal anti-CD3 Ab with wild-type Fc at concentration=1 μg /ml After 7 days, CFSE-labeled T cells are stained with antibodies against CD3, CD4 and CD8 and subjected to analysis using FACS to determine the absolute numbers of T cells and the number of cell divisions, while unlabeled CFSE T cells are stained with antibodies against CCR7, CD3 , CD4, CD8, CD45RA and CD45RO to determine the composition of proliferating CD8+ T cell subsets using flow cytometry. T cell proliferation is represented as a CFSE signal in a subset of T cells in histograms, and the size of the four CD8+ T cell subsets is represented as a percentage of the total CD8+ T cell fraction.

Example 13 Efficacy of CD3xCLEC12A bsAb in inducing T-cell mediated tumor cell lysis in an AML patient.

Example 1 showed that the CD3xCLEC12A bsAb can induce the killing of CLEC12A positive HL60 cells by resting T cells from healthy donors. Next, the present inventors investigated the ability of the CD3xCLEC12A bsAb to induce target-specific activation of AML patient T cells and its ability to induce T cell-mediated killing of HL60 cells from the AML patient.

T cells were isolated from frozen peripheral blood of AML patients (AML-M1/M2, M4 or M5 according to the AML FAB classification) in clinical remission using a Pan T cell isolation kit as described in Example 11. Purified, resting T cells derived from AML patients were subsequently incubated with CSFE-labeled HL-60 cells in medium supplemented with 10% normal HS at a ratio of effector:target cell=5:1 for two days, in the presence of PBS matched by isotype control Ab wt Fc, CD3xCLEC12A DM-Fc, CD3x isotype control DM-Fc and Ab against CD-3 wt Fc as positive control (all antibodies at 1 μg/mL). After coculturing for two days, T cell activation was determined by analysis using flow cytometry for CD3, CD4 and CD25. These results are presented as the percentage of CD25+ cells in CD4+ T cells. In addition, surviving CFSE-20 040393 positive HL60 cells were quantified by flow cytometry. Results are presented as percent specific lysis relative to IgG.

These data show that antigen-specific T cell activation of healthy donor and AML patient mediated by CD3xCLEC12A bsAb with DM-Fc was comparable (FIG. 10A). As expected, CD3x control of the bsAb isotype with DM-Fc did not induce activation of either healthy donor T cells or T cells derived from AML patients. CD3xCLEC12A bsAb with DMFc mediated lysis of HL60 cells by T cells derived from AML patients (68% lysis of HL60 cells) was shown to be comparable to that mediated by healthy donor T cells (69% lysis of HL60 cells, Fig. 10B ). As expected, CD3x control of the bsAb isotype with DM-Fc did not induce killing of HL60 cells, either by T cells from AML patients or T cells from healthy donors. Thus, the CD3xCLEC12A bispecific molecule is a potent inducer of T-cell mediated tumor cell lysis, whether these T cells are derived from AML patients or healthy donors.

Since CD3xCLEC12A bsAb has been shown to have the ability to induce potent lysis of HL60 tumor cells by AML patient T cells, the ability of CD3xCLEC12A bsAb to target specific AML T cell activation was subsequently evaluated. In addition, the ability of CD3xCLEC12A bsAb to induce lysis of primary CLEC12A-positive AML blast cells by autologous T cells derived from AML patients was determined. First, frozen bone marrow samples from AML patients containing >70% primary AML blast cells at diagnosis as determined by flow cytometric analysis were thawed, cultured overnight in IMDM medium supplemented with 10% FCS , 100 ng/ml GM-cSf, 100 ng/ml G-CSF, 50 ng/ml IL-3, 25 ng/ml SCF, and 20 ng/ml Flt3L as previously described (Norde et al., 2009). After overnight culture, primary AML blast cells were phenotyped for surface presentation of CLEC12A, CD3, CD4, CD8, CD14, CD19, CD33, CD34, CD38, CD45, and CD117 by flow cytometry and labeled with CFSE. Resting, autologous, patient-derived T cells harvested at the time the patient achieved clinical remission were isolated from peripheral blood using a Pan T cell isolation kit as described in Example 11. Subsequently, AML blast cells were co-cultured with resting autologous T -cells in a ratio of E:T=5:1 in a medium with 10% HS for two days. Conditions tested included PBS, wild-type Fc isotype matched Ab, CD3xCLEC12A DM-Fc, CD3x isotype control DM-Fc, and Ab against CD3 with wild-type Fc as a positive control (all antibodies at concentration=1 μg/ml). After co-culture for two days, T cell activation was determined by analysis using flow cytometry for CD3, CD4, CD8 and CD25. These results are presented as the percentage of CD25+ cells in CD4+ or CD8+ AML T cells. Lysis of AML blast cells was determined by quantifying the number of surviving double positive CTSE7CD45 ^low levels of AML blast cells by flow cytometry. The results were presented as the percentage of specific lysis of blast cells relative to IgG.

These data show that the CD3xCLEC12A bsAb with DM-Fc has the ability to induce AML-specific target blast AML T cell activation comparable to anti-CD3 monoclonal Ab with wild-type Fc as a positive control (FIG. 11A/B). In addition, these data show that CD3xCLEC12A bsAb-induced potent killing of autologous AML blast cells by AML patient-derived T cells is as potent as that induced by anti-CD3 monoclonal Ab with wild-type Fc as a positive control (Fig. 11C). As expected, no or negligible kill of AML blast cells was induced by CD3xisotype control antibody with DM-Fc, which means that the observed kill of AML blast cells mediated by CD3xCLEC12A bsAb is mainly the result of antigen-specific T-cell activation and specific lysis of CLEC12A+ AML tumor cells. Overall, this study shows that the CD3xCLEC12A bsAb can effectively induce the killing of CLEC12A-positive tumor cells by the T cells of AML patients.

Example 14 Effect of Fc Suppression on Non-Specific Cytokine Release.

In Examples 7 and 8, it was shown that the CD3xCLEC12A bsAb format of the IgG1 subclass with Fc suppression by CH2/lower hinge (DM-Fc) engineering resulted in reduced affinity for Fc-gamma receptors and abolished non-specific, Fc-receptor-mediated cytotoxicity for to the HL60 cell line of leukemic origin. It was then examined whether the IgG1 subclass bsAb format with DM-Fc suppression abolished non-specific Fc receptor mediated cytotoxicity in the presence of Fc receptor positive observer cells such as NK cells. In this study, autologous resting T cells derived from healthy donors were redirected to CLEC12A-positive monocytes in the presence of other Fc-receptor-positive, innate effector bystander cells such as NK cells. For this purpose, PBMCs were isolated from heparinized peripheral blood of healthy donors using density gradient centrifugation and seeded at a density of 1x106 cells/ml. PBMC were cultured for two days in medium with 10% FBS, in the presence of or PBS, isotype matched control Ab, CD3xCLEC12A bsAb with wild-type Fc, CD3xCLEC12A bsAb with DM-Fc, CD3c control

- 21 040393 isotype bsAb with wild-type Fc, CD3x isotype control bsAb with DM-Fc, or monoclonal Ab against

CD3 with wild-type Fc. After cultivation for two days, surviving monocytes were quantified by flow cytometry based on CD14 expression. The results are presented as percent specific lysis referred to IgG.

In the case of the CD3xCLEC12A bispecific antibody, Fc suppression due to the presence of the DM-Fc region was found to have only a minor effect on monocyte lysis (FIG. 12). In contrast, in the case of the CD3x bsAb isotype control, Fc suppression due to the presence of the DMFc region significantly reduced non-specific monocyte lysis. Accordingly, it is concluded that Fc suppression in CD3xCLEC12A bsAb also contributes to target-specific killing: CD3xCLEC12A bsAb with DM-Fc in particular recruits T cells and reduces non-specific immune system activation mediated by normal Fcy receptor-expressing helper cells. .

The question was then asked whether Fc-suppression resulting from a DM mutation in the CD3xCLEC12A bsAb abolishes the Fc receptor-mediated release of cytokines known to be associated with cytokine release syndrome (CRS), a common clinical phenomenon with antibody-based therapies that give rise to helper cells. To this end, the cytokine profile in the supernatants of the monocyte killing assay described in FIG. 13 was analyzed using the 10-Plex human cytokine panel for the Luminex platform (Invitrogen, LHC0001) according to the manufacturer's instructions. The profile of the following human cytokines was measured in the day 2 supernatant: GMCSF, IFN-γ, IL-Q, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10 and TNF-α. The results presented are the concentration of cytokines, determined in pg/ml. Levels of the cytokines GM-CSF, IL-4 and IL-5 were below the detection limit of this assay (data not shown).

The data show that CD3xCLEC12A and CD3c isotype control bsAb, both wild-type Fc-tail, elicited a surge of IL1β, IL-6, TNF-α, IL-10, IL-2 and IFN-γ (FIG. 13). However, no or very low levels of these cytokines were found in the case of CD3xCLEC12A and CD3x control bsAb isotype when they had a DM-Fc tail, with the exception of IL-8. Since monocytes are the main source of IL-8, high levels of IL-8 are expected to be released from lysed monocytes and not as a result of a specific FcR-mediated release. It is concluded that Fc suppression by mutating DM in the bsAb IgG format significantly abolishes the Fc receptor-mediated release of CRS-associated cytokines IL-II, IL-6, TNF-α, IL-2 and IFN-γ. Taken together, these data indicate that Fc-suppression resulting from DM mutation in CH2/lower hinge region contributes to increased efficiency and specific recruitment of effector cells by CD3xCLEC12A DM-bsAb by reducing potential non-specific activation of the immune system mediated by expressing normal Fcγ receptors by helper cells, and the associated release of pro-inflammatory cytokines.

Example 15

The binding of candidate 3896 full-length bivalent anti-CD3 monoclonal IgG to membrane-bound CD3 was compared to candidate 3056 full-length bivalent anti-CD3 monoclonal IgG by FACS analysis using CD3-expressing HPB-ALL cells. Irrelevant human IgG1 served as an IgG isotype matched control. Flow cytometry was performed according to standard procedures known in the art. As shown in FIG. 14A, anti-CD3 IgG 3896 bound CD3 in a dose-dependent manner on HPB-ALL cells, as did anti-CD3 IgG 3056.

The ability of anti-CD3 3896 IgG to induce T cell proliferation was then tested in direct comparison with mouse anti-CD3 OKT3, anti-CD3 IgG 3056 and isotype-matched control IgG. Briefly, the antibodies were serially diluted and immobilized in 96-well plates.

After removing unbound IgG, CFSE-labeled T cells were added and incubated at 37°C. On day 5, the level of induced T cell proliferation was analyzed by flow cytometry. The results were expressed as the percentage of viable T cells showing at least a 2-fold decrease in CFSE expression level and are shown in FIG. 14b. Anti-CD3 IgG 3896 as a bivalent monospecific antibody was shown to be less effective in inducing T cell proliferation than candidate anti-CD3 IgG 3056 and mouse OKT3. These data suggest that the reduction in 3896-induced T cell proliferation compared to 3056 anti-CD3 IgG reflects a reduction in CD3 binding capacity as analyzed by flow cytometry. This difference in binding allows the arm with the desired affinity to be selected, resulting in a bispecific antibody that exhibits an appropriate balance between affinities for CD3 and CLEC12A such that T cells and CLEC12A positive AML tumor cells are effectively brought together and T cell mediated lysis of CLEC12A- positive AML tumor cells are optimally induced.

To test the efficacy of the novel CD3-targeting arm 3896 versus the CD3-targeting arm 3056 in CD3xCLEC12A bispecific antibody format, the 3896xCLEC12A-specific reference bispecific antibody of Example 4 (candidate 3896x3116) and the 3056xCLEC12A-specific reference antibody of Example 1 candidate 3056x3116) were directly compared in an HL60 cytotoxicity assay as described previously. The results are shown in FIG. 15. The 3896xCLEC12A-specific target bsAb was noted to have similar potency to the 3056xCLEC12A-specific target bsAb. Therefore, since both bispecific antibodies differ only in their CD3-specific Fab arm, while the CLEC12A-specific Fab arm is the same, it can be concluded that the functionality of the CD3-specific Fab arm 3896 is similar to that of the CD3-specific Fab. -arm 3056 in CD3xCLEC12A bispecific Ab. It should be noted that at lower concentrations, the 3896x3116 candidate is even better than the 3056x3116 candidate. This is advantageous because it provides a wider therapeutic window, as explained earlier in the present description.

Example 16

In Example 3, a panel of CLEC12A-specific Fab arms was selected from phage display libraries. All CLEC12A binding molecules contained the huVk1-39 light chain. Three CLEC12A binding molecules were selected: Fabs 3918, 4327, and 4331. These Fabs were expressed as full-length human IgG1: anti-CLEC12A IgG 3918, anti-CLEC12A IgG 4327, and anti-CLEC12A IgG 4331. Nucleotide and amino acid sequences of VH in anti-CLEC12A IgG 3912A , VH in anti-CLEC12A IgG 4327, VH in anti-CLEC12A IgG 4331 and total VL (IGKV1-39; O12) are shown in FIG. 20. Full length anti-CLEC12A antibodies were tested for binding to CLEC12A expressed by HL60 cells.

The binding of anti-CLEC12A IgG 3918, anti-CLEC12A IgG 4327 and anti-CLEC12A IgG 4331 to membrane-bound CLEC12A was compared to a reference anti-CLEC12A antibody (3116) by FACS analysis using CLEC12A expressing HL60 cells. Irrelevant human IgG1 served as an IgG isotype matched control. Flow cytometry was performed according to standard procedures known in the art. As shown in FIG. 16, Anti-CLEC12A IgG 4327 bound to CLEC12A in a manner similar to the reference anti-CLEC12A antibody. Two other antibodies, anti-CLEC12A IgG 3918 and anti-CLEC12A IgG 4331, also showed good dose-dependent binding to CLEC12A on HL60 cells. Their binding to CLEC12A seemed to be somewhat less compared to the anti-CLEC12A reference antibody.

In conclusion, Fab 3918, 4327 and 4331 are good CLEC12A binding arms.

Example 17.

Whether the bispecific molecules containing the CD3 specific Fab arm 3896 and the CLEC12A specific Fab arm 3981, 4327 or 4331 were tested were functional. For this, the VH sequence of the CD3-specific Fab arm 3896 and the VH region or reference antibody against CLEC12A, CLEC12A-specific Fab 3918, CLEC12A-specific Fab 4327 or CLEC12A-specific Fab 4331 were cloned into expression vectors using methods known in the art. for the production of bispecific IgG1 (Gunasekaran et al. WO2009/089004) in combination with rearranged huVK1-39 light chain to produce bispecific antibodies; 3896xCLEC12A specific reference, 3896x3918, 3896x4327 and 3896x4331. These bispecific molecules were tested for functionality in the previously described HL60 cytotoxicity assay. Resting T cells from two healthy donors (HD1 and HD2) were co-cultured with CFSE-labeled HL60 cells in the presence of various concentrations of the bispecific antibody in an E:T=5:1 ratio for 48 hours in the presence of 10% HS. The number of surviving CFSE-positive HL60 cells was determined by flow cytometry on day 2. Results in FIG. 17 are presented as percent specific lysis. In the case of two separate experiments with T cells from donor 1 (HD1, Fig. 17 upper panel) and T cells from donor 2 (HD2, Fig. 17 lower panel), it was shown that all bispecific molecules were as effective as as well as 3896xCLEC12A-specific reference bispecific antibody when incubated at high concentration.

It should be noted that, especially at low concentrations of bispecific antibodies, the 3896x4327 and 3896x4331 bispecific antibodies were found to be more effective than the 3896xCLEC12A-specific reference bispecific antibody. Therefore, since these bispecific antibodies differ only in their CLEC12A-specific Fab arm, while the CD3-specific Fab arm is the same, it can be concluded that the functionality of the CLEC12A-specific Fab arms 4327 and 4331 is more potent than with reference CLEC12A-specific Fab-arm. Without wishing to be bound by any theory, the observed differences between 3896x4327 and 3896x4331 versus the 3896xCLEC12A-specific reference bispecific IgG may reflect a difference in the affinity of these new Fab arms against CLEC12A, or they may target a different CLEC12A epitope, making it possible to more efficient cross-linking of tumor cells with CD3-expressing T cells.

Example 18.

In Example 2, anti-CLEC12A Fab fragments 3918 and 4331 were shown to compete for

- 23 040393 binding to the CLEC12A epitope assayed in a Fab ELISA with Fab fragments of a reference anti-CLEC12A antibody. However, the anti-CLEC12A Fab 4327 did not compete with the reference anti-CLEC12A IgG for binding in this assay (Table 2). In this experiment, we tested whether full-length anti-CLEC12A IgG 4327 competes for binding to CLEC12A with a reference anti-CLEC12A antibody. Briefly, HL60 cells were pre-incubated with the first antibody at a concentration=50 μg/ml on ice for 20 minutes. Subsequently, an Oregon Green (OG) labeled (Invitrogen, cat. no. 10476) second antibody was added at a concentration=1 μg/mL to the cells plus the first antibody (first antibody concentration after addition of OG-labeled IgG ~45 μg/mL). After 20 min, the cells were washed and analyzed by FACS.

The results are shown in FIG. 18: It was concluded that anti-CLEC12A IgG 4327 and reference anti-CLEC12A IgG compete for binding to CLEC12A. This suggests that both IgGs either bind to a closely related epitope of the CLEC12A antigen or that they bind to different epitopes that prevent both IgGs from binding simultaneously due to steric hindrance.

Example 19.

In previous examples, the CLEC12A-specific Fab arms 4327, 4331, 3918, and 3116 have been shown to be good CLEC12A binders and potent inducers of T cell-mediated killing in a CD3xCLEC12A bispecific format. Until now, bispecific antibodies have been produced using known methods for promoting heterodimerization of immunoglobulin heavy chains (Gunasekaran et al.). In a concurrent U.S. patent application and PCT application filed by the present inventors (U.S. Common Patent Application No. 13/866747 and PCT/NL2013/050294; which are incorporated herein by reference), the present inventors have disclosed methods and means for producing bispecific antibodies from a single cell, whereby means are provided that promote the formation of bispecific antibodies instead of the formation of monospecific antibodies. These methods can also be used to advantage in the present invention. In particular, preferred mutations for the production of substantially only bispecific full-length IgG molecules are the amino acid substitutions L351K and T366K (numbered according to Kabat) in the first CH3 domain (heavy chain KK variant) and the amino acid substitutions L351D and L368E in the second CH3 domain (DE variant). heavy chain), or vice versa. Previously pending US patent application US 13/866747 and PCT/NL2013/050294 filed by the authors of the present invention, it was shown that the DE variant and the KK variant preferentially pair to form heterodimers (so-called DEKK bispecific molecules) . Homodimerization of the DE heavy chain variant (DEDE homodimers) or the KK heavy chain variant (KKKK homodimers) hardly occurred due to strong repulsion between charged residues at the CH3-CH3 junction between identical heavy chains. To demonstrate that the effect of CD3xCLEC12A bispecific molecules is independent of known mutations for heterodimerization (Gunasekaran) or DEKK mutations, DE variant and KK variant heavy chains were used to promote heterodimerization of different heavy chains to generate CD3xCLEC12A bispecific molecules. In addition, double mutations (L235G and G236R; DM) in the CH2/lower hinge region were introduced into these DE and KK heavy chain variants. The Fc tail of these resulting bispecific molecules is referred to as DM DEKK.

Briefly, the VH regions of CLEC12A-specific Fab arms or 3116, 4327, or 4331 were cloned into expression vectors containing the DE variant+DM heavy chain, while the VH region of the anti-CD3 antibody 3056 was cloned into an expression vector containing KK -variant+DM heavy chain (U.S. Common Patent Application No. 13/866747 and PCT/NL2013/050294), and these expression vectors, together with a nucleic acid molecule encoding a rearranged human light chain IGKV1-39/IGKJ1 (huVK1-39 ) were provided to the host cell such that the host cell expressed and produced the bispecific antibodies. The resulting 3056x3116 DM DEKK, 3056x4327 DM DEKK and 3056x4331 DM DEKK bispecific antibodies were subsequently tested for efficacy in an HL60 cytotoxicity assay as described above. The results are shown in FIG. 19: it was shown that all variants are still capable of effective tumor cell lysis, and, accordingly, it was concluded that DM and DEKK mutations can be introduced into the Fc region of the CD3xCLEC12A bispecific antibody, while maintaining the ability to induce tumor cell lysis.

- 24 040393

Bibliography

Armour et al. Mol. Immunol. 2003 (40) 585-593

Bakker A.B. et al. Cancer Res 2004, 64, p8443-50

Bargou et al. 2008 Science 321:974

Bluemel et al. 2010 Cancer Immunol. Immunother. 59:1197

Chames and Baty, MABS 2009 (1) 539-547

Chatenoud et al. 1990, Transplantation 49(4), pages 697702

Chen C.H. et al. Blood 2006, 107, pl459-67

Cui et al. JBC 2012 (287) 28206-28214

De Kruif et al. 1995, J Mol Biol 248(1), pages 97-105

De Kruif et al. J. Mol. Biol. 2009 (387) 548-58

De Kruif et al. Biotechnol Bioeng. 2010(106)741-50

De Wildt RM et al. J. Mol. Biol. 1999 (285) 895-901;

Dreyer et al. 2002 Int. J. Canc. 100:690

Geginat, J. et al. Blood, 2003. 101(11), p. 4260-6

Gunasekaran et al. JBC 2010 (285) 19637-19646

Haagen et al. 1995 Blood 85:3208

Han Y. et al. Blood 2004, 104, p2858-66

Kipriyanov et al. 1998 Int. J. Can. 77:763

Kontermann, MABS 2012 (4) 182-197

Lanzavecchia et al. 1987, Eur.J.Imm. 17:105

Liu et al. 1985 PNAS 82:8648

Liesveld et al. 1988, J. Immunol. 140(5), pages 1527-1533

Loffler et al. 2000 Blood 95:2098

Marshall A.S. et al. J Biol Chern 2004, 279, p 14792-802

Merchant et al. Nature Biotechnology 1998 Volume 16, pp 677-681

Moore et al. Blood 2011 (117) 4542-4551

Moshaver et al. 2008 Stem Cells 26:3059

Nissim et al. The EMBO Journal vol.13 no. 3 pp.692-698. 1994

Nord WJ. et al. Blood 2009 (113) (10): p. 2312-23

Offner et al. Molecular Immunology 2006 (43) 763-771

Oganesyan et al. Biol. Crystall. 2008(D64)700

Schaefer et al (Cancer Cell 20, 472-486, October 2011

Sheridan C, Nat Biotechnol. 2012(30):300-1

Staerz et al. 1986 PNAS 83:1453

Shields R.L. et al. JBC 2001 (276) 6591-6604

Sluijter, B.J., et al. Clin Immunol, 2010. 137(2), p. 22133

Sunthalingam et al. 2006, New England J Med 355(10), pages 1018-1028

Van Rhenen et al. 2007 Blood 110:2659

Zeidler et al. 1999 J. Immunol. 163:1246

WO2004/009618

WO2005/118635

WO2005/000894

WO2005/000894

WO2008/027236

WO2009/089004

WO2009/157771

WO2010/108127

-

Claims (26)

CLAIM 1. A human IgG bispecific full-length antibody, wherein said IgG bispecific antibody comprises a first arm that specifically recognizes C-type lectin family 12 domain member A (CLEC12A) and a second arm that specifically recognizes an antigen on immunocompetent effector cells capable of attracting such cells to an abnormal cell expressing CLEC12A, where said antigen on said immunocompetent effector cells is CD3. 2. The IgG bispecific antibody of claim 1, wherein said immunocompetent effector cells include T cells. 3. An IgG bispecific antibody according to claim 1, wherein said antibody specifically recognizes CD3ε. 4. An IgG bispecific antibody according to any one of claims 1 to 3, wherein both arms contain a common light chain. 5. The IgG bispecific antibody of claim 4, wherein said common light chain is a germline light chain. 6. An IgG bispecific antibody according to claim 4 or 5, wherein said common light chain is a rearranged human germline kappa light chain IgVK1-39*01/IGJK1*01. 7. An IgG class bispecific antibody according to any one of claims 1 to 6, wherein said bispecific antibody is human IgG1. 8. An IgG bispecific antibody according to any one of claims 1 to 7, wherein the arm that specifically recognizes CLEC12A comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTSY and a heavy chain CDR2 sequence consisting of from a sequence that is at least 90% identical to IINPSGGS, and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to GTTGDWFDY. 9. An IgG bispecific antibody according to claim 8, wherein the arm that specifically recognizes CLEC12A contains a heavy chain variable region sequence consisting of a sequence that is at least 90% identical to QVQLVQSGAEVKKPGASVKVSC ASGYTFTSYYMHWVRQAPGQGLEWMGIINPSGGSTSYAQK FQGRVTMTRDTSTSTVYQMELSSLAVDRSEDT AVGWWGTLVRSEDT AVGWTLCAFDTVSS. 10. An IgG bispecific antibody according to any one of claims 1 to 8, wherein the arm that specifically recognizes CLEC12A comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTSY and a heavy chain CDR2 sequence consisting of from a sequence that is at least 90% identical to IINPSGGS, and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to GNYGDEFDY. 11. The IgG bispecific antibody of claim 10, wherein the arm that specifically recognizes CLEC12A contains a heavy chain variable region sequence consisting of a sequence that is at least 90% identical to EVQLVQSGAEVKKPGASVKVSC ASGYTFTSYYMHWVRQAPGQGLEWMGIINPSGGSTSYAQK FQGRVTMTRDTSTSQTVYMELSTVYSSRSGNY. 12. An IgG bispecific antibody according to any one of claims 1 to 7, wherein the arm that specifically recognizes CLEC12A comprises a heavy chain CDR1 sequence consisting of a sequence that is at least 90% identical to SGYTFTGY and a heavy chain CDR2 sequence consisting of from a sequence that is at least 90% identical to WINPNSGG and a heavy chain CDR3 sequence consisting of a sequence that is at least 90% identical to DGYFADAFDY. 13. The IgG bispecific antibody of claim 12, wherein the arm that specifically recognizes CLEC12A contains a heavy chain variable region sequence consisting of a sequence that is at least 90% identical 14. An IgG bispecific antibody according to any one of claims 1 to 13, wherein the second arm that specifically recognizes CD3 contains - 26 040393 a heavy chain CDR1 sequence consisting of the SYGMH sequence and a heavy chain CDR2 sequence consisting of the IIWYSGSKKNYADSVKG sequence and a heavy chain CDR3 sequence consisting of the GTGYNWFDP sequence. 15. The IgG bispecific antibody of claim 14, wherein the second arm that specifically recognizes CD3 comprises a heavy chain variable region sequence consisting of the sequence QVQLVESGGGWQPGRSLRLSCAASGFTFRSYGMHWVRQAPGKGLEWVAIIWYSGSKKNYA DSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARGTGYNWFDPWGQGTLVSS. 16. An IgG class bispecific antibody according to any one of claims 1 to 15, wherein the first and second arms comprise a light chain CDR1 sequence consisting of a sequence that is at least 90% identical to RASQSISSYLN and a light chain CDR2 sequence consisting of a sequence which is at least 90% identical to AASSLQS, and a light chain CDR3 sequence consisting of a sequence that is at least 90% identical to QQSYSTPPT. 17. The IgG bispecific antibody of claim 16, wherein the first and second arms comprise a light chain variable region sequence consisting of a sequence that is at least 90% identical to DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAAS SLQSGVPSRFSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPTFGQGTKVEIK. 18. An IgG bispecific antibody according to any one of claims 1 to 17, wherein said IgG bispecific antibody has a mutated CH2 domain and/or lower hinge domain such that interaction of said IgG bispecific antibody with Fcγ receptors is significantly reduced. 19. An IgG bispecific antibody according to claim 18, wherein said mutated CH2 domain and/or lower hinge region contain a L235G and/or G236R substitution. 20. An IgG bispecific antibody according to any one of claims 18 to 19, wherein said mutated CH2 domain and/or lower hinge region contains a substitution of L235G and G236R. 21. A method for producing an IgG class bispecific antibody according to any one of claims 1 to 20 from a single cell, wherein said IgG class bispecific antibody contains two CH3 domains, wherein said method comprises: the presence of a cell containing a) a first nucleic acid sequence encoding an IgG heavy chain that specifically recognizes CLEC12A and which contains a first CH3 domain, and b) a second nucleic acid sequence encoding an IgG heavy chain that specifically recognizes CD3 and which contains a second CH3 domain, wherein said nucleic acid sequences provide amino acid mutations for preferential pairing of said first and second CH3 domains, and wherein said cell contains a third nucleic acid sequence encoding a common light chain, wherein said method involves the step of culturing said cell and allowing expression of said two nucleic acid sequences and harvesting said IgG class bispecific antibody from the culture. 22. The method of claim 21 wherein said common light chain is a rearranged human germline kappa light chain IgVK1-39*01/IGJK1*01. 23. The method according to any one of claims 21-22 for producing an IgG class bispecific antibody, wherein said first CH3 domain contains amino acid substitutions L351K and T366K, and where said second CH3 domain contains amino acid substitutions L351D and L368E, wherein said method further comprises the step of cultivating said cell and expressing said nucleic acid sequences; and collecting said bispecific antibody from the culture. 24. An IgG class bispecific antibody comprising a first arm that specifically recognizes C-type lectin family 12 domain member A (CLEC12A) and a second arm that specifically recognizes CD3, which is obtained by the method of any one of claims 21-23. 25. Pharmaceutical composition for the treatment of leukemia or pre-leukemic disease of myeloid origin or B-cell lymphomas, containing an IgG class bispecific antibody according to any one of claims 1-20 or 24 and a pharmaceutically acceptable carrier. 26. The use of an IgG class bispecific antibody according to any one of claims 1 to 20 or 24 as a pharmaceutical agent for the treatment of leukemias or pre-leukemic diseases of myeloid origin or B-cell lymphoma or myelodysplastic syndrome (MDS) or chronic myeloid leukemia (CML) or acute myeloid leukemia (AML). -